These authors contributed equally.

#### **References**

[1] Linden R, Matte U. A snapshot of gene therapy in Latin America. Genet Mol Biol. 2014; 37(1 Suppl): 294–98.


duction of liver in hemophilia by AAV-factor IX and limitations imposed by the host immune response. Nat Genet. 2006; 12: 342–7.

[16] Rogers GL, Herzog RW. Gene therapy for hemophilia. Front Biosci (Landmark Ed). 2015; 20: 556–603.

[2] Chodisetty S, Nelson EJ. Gene therapy in India: A focus. J Biosci. 2014; 39: 537–41.

Pharm Bull. 2013; 3: 249–55. DOI: 10.5681/apb

trophies. Curr Gene Ther. 2012; 12: 307–14. DOI:

10.1038/mtna

198 RNA Interference

270–81. DOI: 10.1016/j.jconrel

DOI: 10.2144/000112792

cure. N C Med J. 2013; 74: 526–9.

10.3389/fphys

[3] Razi Soofiyani S, Baradaran B, Lotfipour F, Kazemi T, Mohammadnejad L. Gene therapy, early promises, subsequent problems, and recent breakthroughs. Adv

[4] Liu J, Harper SQ. RNAi-based gene therapy for dominant limb girdle muscular dys‐

[5] Liu J, Wallace LM, Garwick-Coppens SE, Sloboda DD, Davis CS, Hakim CH, Hauser MA, Brooks SV, Mendell JR, Harper SQ. RNAi-mediated gene silencing of mutant myotilin improves myopathy in LGMD1A mice. Nucleic Acids. 2014; 3: e160. DOI:

[6] Zhou Y, Zhang C, Liang W. Development of RNAi technology for targeted therapy — a track of siRNA based agents to RNAi therapeutics. J Controll Release. 2014; 193:

[7] Castanotto D, Rossi JJ. The promises and pitfalls of RNA-interference-based thera‐

[8] Burnett JC, Rossi JJ. RNA-based therapeutics: current progress and future prospects.

[9] Kim D, Rossi J. RNAi mechanisms and applications. Biotechniques. 2008; 44: 613-6.

[10] Kola VS, Renuka P, Madhav MS, Mangrauthia SK. Key enzyme and protein of crop insects as candidate for RNAi based gene silencing. Front Physiol. 2015; 6: 119. DOI:

[11] Presloid JB, Novella IS. RNA viruses and RNAi: quasispecies implications for viral

[12] Makkonen KE, Airenne K, Ylä-Herttulala S. Baculovirus-mediated gene delivery and

[13] Porada CD, Stem C, Almeida-Porada G. Gene therapy: the promise of a permanent

[14] Kay MA, Manno CS, Ragni MV, Larson PJ, Couto LB, McClelland A, Glader B, Chew AJ, Tai SJ, Herzog RW, Arruda V, Johnson F, Scallan C, Skarsgard E, Flake AW, High KA. Evidence for gene transfer and expression of factor IX in haemophilia B patients

[15] Manno CS, Pierce GF, Arruda VR, Glader B, Ragni M, Rasko JJ, Ozelo MC, Hoots K, Blatt P, Konkle B, Dake M, Kaye R, Razavi M, Zajko A, Zehnder J, Rustagi PK, Nakai H, Chew A, Leonard D, Wright JF, Lessard RR, Sommer JM, Tigges M, Sabatino D, Luk A, Jiang H, Mingozzi F, Couto L, Ertl HC, High KA, Kay MA. Successful trans‐

RNAi applications. Viruses. 2015; 7: 2099–25. DOI: 10.3390/v7042099

peutics. Nature. 2009; 457: 426-33. DOI: 10.1038/nature07758

Chem Biol. 2012; 19: 60-71. DOI: 10.1016/j.chembiol

escape. Viruses. 2015; 7: 3226–40. DOI: 10.3390/v7062768

treated with an AAV vector. Nat Genet. 2000; 24: 257–61.


## **Non-viral siRNA and shRNA Delivery Systems in Cancer Therapy**

Emine Şalva, Ceyda Ekentok, Suna Özbaş Turan and Jülide Akbuğa

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/62826

#### **Abstract**

[28] Sawyers CL. Perspective: combined forces. Nature. 2013; 498:S7. DOI: 10.1038/498S7a [29] Koldehoff M. Targeting bcr-abl transcripts with siRNAs in an imatinib-resistant chronic myeloid leukemia patient: challenges and future directions. Methods Mol Bi‐

[30] Marchwicka A, Cebrat M, Sampath P, Snieżewski L, Marcinkowska E. Perspectives of differentiation therapies of acute myeloid leukemia: the search for the molecular basis of patients' variable responses to 1,25-dihydroxyvitamin D and vitamin D ana‐

[31] Zhou GB, Zhang J, Wang ZY, Chen SJ, Chen Z. Treatment of acute promyelocytic leukaemia with all-trans retinoic acid and arsenic trioxide: a paradigm of synergistic molecular targeting therapy. Philos Trans R Soc Lond B Biol Sci. 2007; 362: 959–71. [32] Pagnano KB, Rego EM, Rohr S, Chauffaille Mde L, Jacomo RH, Bittencourt R, Firma‐ to AB, Fagundes EM, Melo RA, Bernardo W. Guidelines on the diagnosis and treat‐ ment for acute promyelocytic leukemia: Associacao Brasileira de Hematologia, Hemoterapia e Terapia Celular Guidelines Project: Associacao Medica Brasi‐ leira-2013. Rev Bras Hematol Hemoter. 2014; 36: 71–92. DOI:

[33] Guo J, Cahill MR, McKenna SL, O'Driscoll CM. Biomimetic nanoparticles for siRNA delivery in the treatment of leukaemia. Biotechnol Adv. 2014; 32: 1396–409. DOI:

[34] Mutlu P, Kiraz Y, Gündüz U, Baran Y. An update on molecular biology and drug re‐ sistance mechanisms of multiple myeloma. Crit Rev Oncol Hematol. 2015;

[35] Wu SQ, Xu ZZ, Niu WY, Huang HB, Zhan R. ShRNA-mediated Bmi-1 silencing sen‐ sitizes multiple myeloma cells to bortezomib. Int J Mol Med. 2014; 34: 616–23. DOI:

[36] Tiemann K, Rossi JJ. RNAi-based therapeutics–current status, challenges and pros‐

pects. EMBO Mol Med. 2009; 1: 142–51. DOI: 10.1002/emmm.200900023

ol. 2015; 1218: 277–92. DOI: 10.1007/978-1-4939-1538-5\_17

logs. Front Oncol. 2014; 4: 125. DOI: 10.3389/fonc.2014.00125

10.5581/1516-8484.20140018

S1040-8428: 30003-2. DOI: 10.1016/j.critrevonc

10.1016/j.biotechadv

200 RNA Interference

10.3892/ijmm.2014.1798

RNA interference represents a promising therapeutic strategy for the silencing of specific target genes in cancer therapy. Small interfering RNAs and DNA-based vectors encoding short hairpin RNAs provide sequence-specific post-transcriptional gene silencing by binding to its complementary RNA. For the therapeutic use of siRNA in cancer, efficient intracellular delivery is necessary. The efficient cancer therapy with RNAi is not still ac‐ complished because of internalization and intracellular trafficking problems such as low transfection efficiency, enzyme degradation, inappropriate subcellular localization, and endosomal trapping of siRNAs in cells. Cancer is a complex disease including multiple genes and pathways. The most important benefits of siRNA therapy are high target spe‐ cificity and non-toxicity compared with chemotherapy. The uptake of siRNA by cells without a carrier system is possible, but naked siRNA is mostly degraded with nucleases and activates the immune responses. Development of appropriate delivery systems is an important issue in the use of siRNA-based therapeutics. Non-viral delivery systems have great potential for safe and effective delivery of siRNA therapeutics to tumor cells. Nano‐ carriers such as nanoplexes, lipoplexes, nanoparticles, and liposomes have been common‐ ly used for siRNA delivery. This chapter highlights the importance of non-viral delivery systems *in vitro* and *in vivo* cancer therapy.

**Keywords:** Cancer, non-viral vectors, RNAi, siRNA, shRNA

#### **1. Introduction**

RNA interference (RNAi) is a conserved endogenous cellular process for post-transcriptional regulation in sequence-specific gene silencing. The regulatory RNA molecules include small interfering RNAs (siRNAs), and short hairpin RNAs (shRNAs) provide the specific degrada‐ tion of target mRNA in mammalian cells [1]. siRNAs are the products of long double-stranded

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

RNA (dsRNA) molecules in cells, which are expressed transgenically or delivered exogenous‐ ly. Synthetic siRNAs can be transfected into cells that specifically silence the expression of target genes. In the RNAi pathway, dsRNA (over 100 nt) molecules are cleaved into 21- to 23 nucleotide duplexed RNAs, termed as siRNA duplex, by endoribonuclease Dicer or RNAse III-type enzyme. The cleaved siRNA duplexes contain 5′-phosphate and two-base 3′ over‐ hangs. siRNAs are incorporated into the endogenous RNA-induced silencing complex (RISC). One of the two strands of siRNA duplex is guide (antisense) strand, and the other is passenger (sense) strand. siRNA duplex is unwinded by RNA helicase activity. While the guide strand binds to the RISC, the passenger strand is degraded. The activated RISC binds to mRNA with base-pairing for sequence-specific degradation of complementary mRNA. The mRNA fragments cleaved by Argonaute (Ago) proteins are released from RISC and degraded by other endogenous nucleases. After mRNA degradation, the active RISC is rebuilt and can participate in another RNAi pathway [2, 3]. In the event, RNAi process decreases specific mRNA levels, and thus decreases target gene expression.

Among the nucleic acid–based drugs, siRNAs as potential novel drug candidates are offered a highly promising strategy in cancer therapy. The knockdown of abnormal gene overexpres‐ sion, occuring in cancer by using siRNA, has been used in therapeutic applications [2]. Targets of chemical drugs are limited to certain classes of receptors, ion channels, and enzymes. In the treatment with siRNA, known sequence of any gene of interest is sufficient and its target choice is unlimited [4].

The potential advantages of siRNA treatment compared with other treatment methods are that siRNAs [5–7] (i) provide sequence-specific gene therapy; (ii) can specifically target many undruggable genes or downregulate gene products; (iii) are considered the safe therapeutics; (iv) are potent and efficient molecules and possess high gene silencing activity; and (iv) can be easily designed for any disease.

RNAi might be a promising new pharmaceutical area for treatment of incurable and severe diseases such as cancers and infections. RNAi applications have been recently achieved by using synthetic siRNAs and vector-based siRNA expression systems or short hairpin RNAs (shRNAs) synthesized within the cells by vector-mediated production. The expressed shRNAs from plasmid and viral vectors in nucleus are cleaved by Dicer in cytoplasm and siRNAs are formed. There both strategies have advantages and disadvantages. Vector-based siRNA expression systems have several advantages for applying RNAi compared to synthetic siRNAs. Both permanent and transient transfection with vector-based systems can be ach‐ ieved, and thus vector-based system increases the period of siRNA-mediated inhibition of gene expression [8]. In addition, shRNA constructs are more stable than siRNAs [9]. Low amount (nM) of siRNA and less than five copies of shRNA are sufficient for stable transfection and for acheiving gene silencing effect [10]. The synthetic siRNAs can be easily synthesized in large amounts and chemically modified to improve stability, permeability, efficacy, and transfection control; however, the modified siRNAs are highly expensive [11]. siRNAs are not integrated into host genome. The modification of vector-based shRNA systems is difficult, but shRNA expression systems can be regulated or induced by appropriate promoters and termination sequences. Choice of promoter, loop structure of shRNA, length and arrangement of sense and antisense strands, and orientation of restriction enzyme regions are important for shRNA expression cassette preparation. Similar to the various RNAi applications for targeted gene silencing, the chimeric expression cassettes of siRNA and shRNA in the same expression unit might also be made [12].

RNA (dsRNA) molecules in cells, which are expressed transgenically or delivered exogenous‐ ly. Synthetic siRNAs can be transfected into cells that specifically silence the expression of target genes. In the RNAi pathway, dsRNA (over 100 nt) molecules are cleaved into 21- to 23 nucleotide duplexed RNAs, termed as siRNA duplex, by endoribonuclease Dicer or RNAse III-type enzyme. The cleaved siRNA duplexes contain 5′-phosphate and two-base 3′ over‐ hangs. siRNAs are incorporated into the endogenous RNA-induced silencing complex (RISC). One of the two strands of siRNA duplex is guide (antisense) strand, and the other is passenger (sense) strand. siRNA duplex is unwinded by RNA helicase activity. While the guide strand binds to the RISC, the passenger strand is degraded. The activated RISC binds to mRNA with base-pairing for sequence-specific degradation of complementary mRNA. The mRNA fragments cleaved by Argonaute (Ago) proteins are released from RISC and degraded by other endogenous nucleases. After mRNA degradation, the active RISC is rebuilt and can participate in another RNAi pathway [2, 3]. In the event, RNAi process decreases specific mRNA levels,

Among the nucleic acid–based drugs, siRNAs as potential novel drug candidates are offered a highly promising strategy in cancer therapy. The knockdown of abnormal gene overexpres‐ sion, occuring in cancer by using siRNA, has been used in therapeutic applications [2]. Targets of chemical drugs are limited to certain classes of receptors, ion channels, and enzymes. In the treatment with siRNA, known sequence of any gene of interest is sufficient and its target choice

The potential advantages of siRNA treatment compared with other treatment methods are that siRNAs [5–7] (i) provide sequence-specific gene therapy; (ii) can specifically target many undruggable genes or downregulate gene products; (iii) are considered the safe therapeutics; (iv) are potent and efficient molecules and possess high gene silencing activity; and (iv) can be

RNAi might be a promising new pharmaceutical area for treatment of incurable and severe diseases such as cancers and infections. RNAi applications have been recently achieved by using synthetic siRNAs and vector-based siRNA expression systems or short hairpin RNAs (shRNAs) synthesized within the cells by vector-mediated production. The expressed shRNAs from plasmid and viral vectors in nucleus are cleaved by Dicer in cytoplasm and siRNAs are formed. There both strategies have advantages and disadvantages. Vector-based siRNA expression systems have several advantages for applying RNAi compared to synthetic siRNAs. Both permanent and transient transfection with vector-based systems can be ach‐ ieved, and thus vector-based system increases the period of siRNA-mediated inhibition of gene expression [8]. In addition, shRNA constructs are more stable than siRNAs [9]. Low amount (nM) of siRNA and less than five copies of shRNA are sufficient for stable transfection and for acheiving gene silencing effect [10]. The synthetic siRNAs can be easily synthesized in large amounts and chemically modified to improve stability, permeability, efficacy, and transfection control; however, the modified siRNAs are highly expensive [11]. siRNAs are not integrated into host genome. The modification of vector-based shRNA systems is difficult, but shRNA expression systems can be regulated or induced by appropriate promoters and termination sequences. Choice of promoter, loop structure of shRNA, length and arrangement of sense and

and thus decreases target gene expression.

is unlimited [4].

202 RNA Interference

easily designed for any disease.

Design of optimal siRNA sequence plays a key role for successful siRNA therapy. The choice of potent and specific siRNA sequences is important for minimization of immune responses and off-target effects. siRNAs are 19–27 base pairs in length, but mostly preferred to be 21 nt of siRNAs with a structure of 19 nt duplex region and two nucleotide overhangs at the 3′ end, usually TT and UU, which are important for recognition by the RNAi machinery. Increasing the length of the dsRNA may enhance its potency, dsRNAs with 27 nucleotides are up to 100 times more potent than the siRNAs containing 21 nucleotides. The longer (25–30 nt) duplexes act as a substrate for Dicer (Dicer-subtrate siRNAs). This Dicer-substrate siRNAs are more efficiently loaded into the RISC over the siRNAs with 21 bp and newly produced siRNAs from the long dsRNAs directly incorporated into the RISC complex. Thus, gene silencing mecha‐ nism can be facilitated [13, 14]. On the other hand, dsRNAs longer than about 30 bp can lead to interferon response, which is the defense mechanism against viral infection. The activation of interferon pathway causes non-specific mRNA degradation and apoptosis [15, 16]. The long dsRNAs activates innate immune response by interaction with protein kinase receptor (PKR) and Toll-like receptors (TLR7 and 8 are activated by ssRNA; TLR9 activated by unmethylated CpG; and TLR3 activated by dsRNA). The activation of these receptors induces interferon (IFN) and proinflammatory cytokines [10, 17].

In addition to immune responses, another important problem for efficient RNAi therapy is offtarget effects of these molecules. The cause of off-target effects are the suppression of undesired or unpredicted genes other than the desired target genes. The absence of homolog sequences between siRNA and its target mRNA can cause cleavage of non-targeted mRNA regions. Offtarget effects of siRNA can lead to problems in the interpretation of gene silencing studies, serious and unwanted side effects such as potential toxicity and even cell death [18]. There are many factors and mechanisms leading to occurence of off-target effects. These factors are the length of dsRNA or siRNA, the length and position of siRNA-target sequence mismatch, and coding sequences and untranslated regions in genes [19]. The mechanisms leading to off-target effects of siRNAs are (i) the regulation of unwanted transcripts by seed sequence homology to the 3′ UTR of cellular mRNAs, (ii) the saturation of RISC by affecting cellular miRNA activity of siRNAs in large amounts, (iii) the function of miRNAs as siRNAs or shRNAs because of similarities in gene silencing pathway, (iv) non-specific distribution by non-targeting systemic delivery, and (v) immunostimulatory motifs in siRNA sequences [14, 17, 20].

Many preclinical studies with siRNA indicated that it is a hopeful molecule for clinical research of various diseases. Up to date, at least 22 RNAi-based drugs have been evaluated in clinical trials [21]. The first clinical trial with siRNA was made in 2004 by Opko Health. The clinical studies of Bevasiranib, that is, siRNA targeting vascular endothelial growth factor (VEGF) to suppress ocular neovascularization in patients with age-related macular degeneration (AMD), continued to the phase III trial, but the clinical trials was terminated in 2009 because of its poor efficacy and causing vision loss. Allergan company has terminated the phase II clinical trials of siRNA AGN-745, targeting VEGF because of its off-target effects [21]. The clinical translation of RNAi can be possible with development of safe and efficient RNAi delivery systems that lack of off-target effects [7, 21].

Although siRNAs are used as the potential therapeutic molecules in cancer and other diseases, the most important challenge in the development of RNAi-based drugs is efficient and safe delivery of appropriate doses to target cells and tissues. Therefore, the development of siRNAbased viral and non-viral delivery systems are required to have an enhanced efficacy, im‐ proved stability, and minimized non-specific gene silencing such as off-target effects and immune responses. This chapter focuses on recent improvements in the non-viral siRNA delivery systems in cancer therapy.

#### **2. RNAi in cancer therapy**

Cancer is a multistep genetic disease, which develops as a consequence of changes in the control of cell proliferation and differentiation. In the transformation of normal cells to tumor cells, the affected cells undergo mutations such as downregulation of tumor suppressor genes and overexpression of oncogenes. RNAi-based therapeutics have been extensively used for knockdown of cancer-associated genes. The *in vitro* and *in vivo* studies with siRNA and shRNA have shown that silencing of genes related to tumor cell growth, invasion, angiogenesis, metastasis, and chemoresistance in various types of cancer [17]. RNAi technology, as a new approach for cancer therapeutics, offers many advantages over conventional cancer treatment strategies. Advantages of gene silencing by siRNA and shRNA are the high degree of specif‐ icity to target tumor cells and tissues, a capacity to inhibit target gene expression, a simple and rapid design, and synthesis [22]. While non-specific chemotherapy leads to death of cancer cells, it significantly damages the surrounding healthy tissues and organs, causing extensive systemic toxicity. The side-effects of chemotherapeutic drugs can be minimized with siRNA treatment.

Cancer cells have the ability to develop a resistance to chemotherapeutic drugs. RNAi-based therapeutics can simultaneously target multiple genes in cancer signaling pathway. The simultaneous silencing of multiple genes in cancer therapy have importance in terms of minimizing the multiple drug resistance caused by small chemical molecules given in high dose [23, 24]. This therapy inhibits survival signals and pathways that take part in the development of multi-drug resistance in cancer cells.

Oncogenes, mutated tumor supressor genes, survival and apoptotic genes, causing tumor initiation and progression, are major targets for RNAi-based therapy. Simultaneous suppres‐ sion of one target gene or multiple genes has provided a significant advantage in cancer therapy. siRNAs can be designed for effective gene knockdown by targeting any gene or multiple genes in cells [25]. siRNAs are likely to be more effective than other antisense approaches because of many properties such as a highly specific mRNA degradation, cell-tocell spreading of gene silencing effect, long silencing activity, improved stability *in vivo*, and their efficiency in lower concentrations [26, 27]. A single therapeutic strategy is insufficient for the inhibition of cancer growth and progression; RNAi as a new therapeutic strategy may be used as well as with chemotherapy, immunotherapy, anti-hormone therapy, and radiotherapy for achieving synergistic therapeutic effect.

In clinical trials, the most of siRNAs have been given by local administration. When siRNAs are delivered to target tissue locally, lower siRNA doses can be used for pharmacologic effect (e.g., saline-based formulation, or excipiants such as 5% dextrose) and any drug delivery approaches (e.g., liposome, nanoparticle, and complexes) [28]. However, systemic drug administration by intravenous injection is required for cancer diseases [7]. In systemic effect, siRNAs must encounter several extra- and intra-cellular barriers until it reaches the target cell and tissue. siRNAs cannot freely cross physiological and cellular barriers because of their high molecular weight and negative charge. The significant challenges of using siRNA are their poor cellular uptake, degradation by serum nucleases, and rapid elimination. These factors and barriers reduce therapeutic effect of siRNA. Therefore, efficient *in vitro* and *in vivo* delivery of siRNA-based therapeutics in cancer is dependent on the development of appropriate delivery systems. siRNA delivery systems should (a) protect siRNAs against degradation enzymes and serum proteins, (b) prolong the circulation time of siRNA, (c) provide siRNA stability in blood serum, (d) avoid sequestration in the reticuloendothelial system (RES), (e) avoid aggregation in serum, (f) minimize non-specific tissue and cellular uptake, (g) achieve target-specific siRNA delivery, (h) allow for immune evasion, (i) resist rapid renal clearance, (j) enhance vascular permeability to reach cancer tissues, (k) promote trafficking to the cytoplasm and uptake into RISC, and (l) have low or non-toxicity [7, 29, 30].

#### **3. Delivery strategies of RNAi-based therapeutics**

of siRNA AGN-745, targeting VEGF because of its off-target effects [21]. The clinical translation of RNAi can be possible with development of safe and efficient RNAi delivery systems that

Although siRNAs are used as the potential therapeutic molecules in cancer and other diseases, the most important challenge in the development of RNAi-based drugs is efficient and safe delivery of appropriate doses to target cells and tissues. Therefore, the development of siRNAbased viral and non-viral delivery systems are required to have an enhanced efficacy, im‐ proved stability, and minimized non-specific gene silencing such as off-target effects and immune responses. This chapter focuses on recent improvements in the non-viral siRNA

Cancer is a multistep genetic disease, which develops as a consequence of changes in the control of cell proliferation and differentiation. In the transformation of normal cells to tumor cells, the affected cells undergo mutations such as downregulation of tumor suppressor genes and overexpression of oncogenes. RNAi-based therapeutics have been extensively used for knockdown of cancer-associated genes. The *in vitro* and *in vivo* studies with siRNA and shRNA have shown that silencing of genes related to tumor cell growth, invasion, angiogenesis, metastasis, and chemoresistance in various types of cancer [17]. RNAi technology, as a new approach for cancer therapeutics, offers many advantages over conventional cancer treatment strategies. Advantages of gene silencing by siRNA and shRNA are the high degree of specif‐ icity to target tumor cells and tissues, a capacity to inhibit target gene expression, a simple and rapid design, and synthesis [22]. While non-specific chemotherapy leads to death of cancer cells, it significantly damages the surrounding healthy tissues and organs, causing extensive systemic toxicity. The side-effects of chemotherapeutic drugs can be minimized with siRNA

Cancer cells have the ability to develop a resistance to chemotherapeutic drugs. RNAi-based therapeutics can simultaneously target multiple genes in cancer signaling pathway. The simultaneous silencing of multiple genes in cancer therapy have importance in terms of minimizing the multiple drug resistance caused by small chemical molecules given in high dose [23, 24]. This therapy inhibits survival signals and pathways that take part in the

Oncogenes, mutated tumor supressor genes, survival and apoptotic genes, causing tumor initiation and progression, are major targets for RNAi-based therapy. Simultaneous suppres‐ sion of one target gene or multiple genes has provided a significant advantage in cancer therapy. siRNAs can be designed for effective gene knockdown by targeting any gene or multiple genes in cells [25]. siRNAs are likely to be more effective than other antisense approaches because of many properties such as a highly specific mRNA degradation, cell-tocell spreading of gene silencing effect, long silencing activity, improved stability *in vivo*, and their efficiency in lower concentrations [26, 27]. A single therapeutic strategy is insufficient for

development of multi-drug resistance in cancer cells.

lack of off-target effects [7, 21].

204 RNA Interference

delivery systems in cancer therapy.

**2. RNAi in cancer therapy**

treatment.

siRNAs have large molecular weight (~13 kDa) and are polyanionic nature (~40 negative phosphate charge) and are easily degraded by enzymes in cells, tissues, and bloodstream. In addition, siRNAs cannot easily cross the cell membrane [29]. The naked siRNAs are readily degraded by serum endonucleases. The half-life of circulating naked siRNA is less than 10 minutes because of its rapid clearance by the kidneys, so that they cannot reach to target cell efficiently. The gene silencing activity of unmodified or uncomplexed siRNAs is little or absent [31]. To solve this problem, two strategies are used: chemical modifications and conjugation of siRNA molecules or use of gene delivery systems for increasing efficiency of RNAi-based therapeutics.

Chemical modification is the major approach to overcome *in vivo* siRNA delivery problems. Chemical modifications of naked siRNAs have been performed to (i) enhance siRNA stability, (ii) protect siRNA from degradation, (iii) avoid recognition by the innate immune system and minimize immunostimulatory responses, (iv) minimize off-target effects, (v) reduce required dose for gene silencing, (vi) improve pharmacodynamic properties, (vii) increase delivery to target cells, and (viii) allow the delivery by systemic administration. The sugar, backbone and nucleobase modifications of siRNA, can significantly protect siRNA in both serum and cytoplasm. The commonly used chemical siRNA modifications are the incorporation of locked nucleic acids (LNA), phosphorothioate linkages, and 2′-o-methyl, 2′-amine, 2′-fluoro groups [7, 32, 33]. Chemical modifications must increase the stability of siRNA without affecting its gene silencing activity [23]. However, these substitutions may lead to off-target effects, cytotoxicity, reduced RNAi activity, and impaired biological activity [17, 34].

Other chemical strategies for siRNA are cholesterol, folate, and aptamer conjugation and peptide modification. siRNAs can associate with aptamers, ligands, and antibodies by electrostatic interaction or direct conjugation. The conjugation of these functional groups provides cell- or tissue-specific targeting and efficient delivery. As a result, the efficacy of silencing can be increased [1].

Viral and non-viral vectors have been extensively used in the siRNA-based therapy. Viral vectors encoding shRNA have a high gene transduction and gene silencing effects. Adenoassociated viral vectors, lentiviral vectors, and adenoviral vectors have been extensively used in gene knockdown studies [35]. The transfering of shRNA-encoding vectors into the nucleus of cells have obtained high and long-term shRNA expression. In addition, viral vectors can integrate the host genome [14]. Although viral vectors have a high gene transfection efficiency, the challenges such as inflammatory reactions, strong immunogenicity, insertional mutagen‐ esis, and oncogenic transformation of viral vectors can cause important safety concerns. In addition, some viral vectors have low capacity for transgene insertion. To overcome these problems, non-viral vectors have been developed and used in siRNA delivery. Compared to viral vectors, non-viral vectors have several advantages such as lack of immunogenicity, low or no integration into genome, large-scale production, and use of wide variety of nucleic acids size [36]. However, the transfection efficiency of non-viral vectors is not as high as the viral vectors.

#### **3.1. Non-viral vectors**

The non-viral delivery of siRNA and shRNA therapeutics to target tumor cells is a multi‐ step process. To achieve efficient delivery and therapeutic gene silencing, siRNAs should be stable in biological fluids and must have above mentioned properties [37, 38]. The circulat‐ ing siRNAs after systemic administration must be evaded from the reticuloendothelial system (RES). Negatively charged siRNAs gain the positive charge after complexed with cationic charged polymers. This positive charge facilitates cellular internalization of siRNAs; however, the cationic charge increase non-specific interactions by non-target cells, negatively charg‐ ed serum proteins, and extracellular matrix. As a consequence of these non-specific interac‐ tions, clot-like accumulations or aggregations are formed. Complexes are entrapped in the endothelial capillary bed or taken up by RES recognition. While RES organs such as spleen, liver, and bone marrow uptake the major part of injected dose, the minor part of this reaches to tumors [25, 37, 39].

Non-viral delivery vectors prolong the biological half-life and mean residence time of siRNA, and they enhance accumulation of siRNA molecules in tumor tissues. siRNA therapeutics can be accumulated into cancer tissue by enhanced permeability and retention (EPR) effect as a result of discontinuous vasculature (permeation) and poor lymphatic drainage (retention) in the abnormal tumor blood vessels compared to the normal blood vessels. Tumor endothelium allows penetration of macromolecules [37, 38].

The other challanges of RNAi-based therapeutics delivery to the tumor tissues after systemic circulation are crossing of cellular membrane, intracellular traffic into the cells with endoso‐ mal/lysosomal compartments, release of siRNA or shRNA from carriers, and nuclear transport for vector-based siRNA/shRNA therapeutics and entry to cytoplasm for siRNA-based therapeutics. The cell membrane is an important extracellular barrier for siRNA uptake. The average size of a single siRNA molecule is less than 10 nm. Despite their small size, polyanionic nature and hydrophilicity of siRNA make crossing of biological membranes difficult [18]. To overcome this problem, the complexation of negatively charged siRNA with cationic polymers or lipids are performed. The net positive charge of this formulations facilitates binding to negatively charged cell membranes, following internalization by adsorptive pinocytosis. For cell-type specific delivery, targeting ligands, antibodies, and aptamer-binding non-viral vectors pass through the cell membrane by receptor-mediated endocytosis [1]. After crossing from the cell membrane, siRNAs and vector-based siRNAs/shRNAs encounter several intracellular barriers that include the endosomal trafficking, unpackaging of siRNA, and nuclear traffic. The intracellular traffic of endosomal content is important for succesful siRNA delivery. When siRNA released from the carrier reaches cytosol, RNAi mechanism is induced inside the cells. However, for the onset of RNAi effect, transfer of vector-based siRNA/shRNA to the nucleus is required. In the delivery process, early release of siRNA from endosome is required. If siRNA remains inside the endosome for long time, it will be degraded. Therefore, different agents (fusogenic protein) conjugated with polymers disrupt the endosomal mem‐ brane. In addition, polymers possess proton-sponge effect (polyethyleneimine, PEI), which have been used to induce osmotic swelling and subsequent disruption of the endosome [15].

#### *3.1.1. Polymer-based RNAi delivery system in cancer therapy*

Negatively charged siRNAs or shRNAs can readily bind to cationic polymers or load to the nanocarriers by ionic interactions. Nanosized complexes or polyplexes by electrostatic interactions and nanoparticle formulations by encapsulation have been developed for efficient siRNA/shRNA delivery. Thus, siRNAs can be protected from nuclease attack and cellular uptake of siRNAs via endocytic pathway faciliated. Many natural and synthetic polymers are used for gene delivery, such as polyethyleneimine (PEI), poly-l-lysine (PLL), chitosan, protamine, gelatin, atelocollagen, cationic polypeptides, cyclodextran polymers, dendrimers, poly-lactide-co-glycolide (PLGA), and polydimethylaminoethylmethacrylate (PDMAEMA) [34]. In addition, polyethyleneglycol (PEG) is widely used as a linker between polymer and ligand or nucleic acid or for binding of siRNA onto nanocarrier surface [40].

#### *3.1.1.1. Chitosan*

nucleic acids (LNA), phosphorothioate linkages, and 2′-o-methyl, 2′-amine, 2′-fluoro groups [7, 32, 33]. Chemical modifications must increase the stability of siRNA without affecting its gene silencing activity [23]. However, these substitutions may lead to off-target effects,

Other chemical strategies for siRNA are cholesterol, folate, and aptamer conjugation and peptide modification. siRNAs can associate with aptamers, ligands, and antibodies by electrostatic interaction or direct conjugation. The conjugation of these functional groups provides cell- or tissue-specific targeting and efficient delivery. As a result, the efficacy of

Viral and non-viral vectors have been extensively used in the siRNA-based therapy. Viral vectors encoding shRNA have a high gene transduction and gene silencing effects. Adenoassociated viral vectors, lentiviral vectors, and adenoviral vectors have been extensively used in gene knockdown studies [35]. The transfering of shRNA-encoding vectors into the nucleus of cells have obtained high and long-term shRNA expression. In addition, viral vectors can integrate the host genome [14]. Although viral vectors have a high gene transfection efficiency, the challenges such as inflammatory reactions, strong immunogenicity, insertional mutagen‐ esis, and oncogenic transformation of viral vectors can cause important safety concerns. In addition, some viral vectors have low capacity for transgene insertion. To overcome these problems, non-viral vectors have been developed and used in siRNA delivery. Compared to viral vectors, non-viral vectors have several advantages such as lack of immunogenicity, low or no integration into genome, large-scale production, and use of wide variety of nucleic acids size [36]. However, the transfection efficiency of non-viral vectors is not as high as the viral

The non-viral delivery of siRNA and shRNA therapeutics to target tumor cells is a multi‐ step process. To achieve efficient delivery and therapeutic gene silencing, siRNAs should be stable in biological fluids and must have above mentioned properties [37, 38]. The circulat‐ ing siRNAs after systemic administration must be evaded from the reticuloendothelial system (RES). Negatively charged siRNAs gain the positive charge after complexed with cationic charged polymers. This positive charge facilitates cellular internalization of siRNAs; however, the cationic charge increase non-specific interactions by non-target cells, negatively charg‐ ed serum proteins, and extracellular matrix. As a consequence of these non-specific interac‐ tions, clot-like accumulations or aggregations are formed. Complexes are entrapped in the endothelial capillary bed or taken up by RES recognition. While RES organs such as spleen, liver, and bone marrow uptake the major part of injected dose, the minor part of this reaches

Non-viral delivery vectors prolong the biological half-life and mean residence time of siRNA, and they enhance accumulation of siRNA molecules in tumor tissues. siRNA therapeutics can be accumulated into cancer tissue by enhanced permeability and retention (EPR) effect as a result of discontinuous vasculature (permeation) and poor lymphatic drainage (retention) in

cytotoxicity, reduced RNAi activity, and impaired biological activity [17, 34].

silencing can be increased [1].

vectors.

206 RNA Interference

**3.1. Non-viral vectors**

to tumors [25, 37, 39].

Among the non-viral vectors, chitosan or its derivatives are attractive where chitosan has been shown to be biodegradable, biocompatible, non-toxic, mucoadhesive, and non-inflammatory and has low cost of production. Chitosan is a cationic polysaccharide, consisting of *N*-acetyld-glucosamine and d-glucosamine units. In addition, chitosan has been designated as "Gen‐ erally Recognized As Safe (GRAS)" by the FDA [41]. It has been widely used in *in vivo* siRNA and shRNA delivery applications because of positively charged amines, allowing electrostatic interactions with negatively charged nucleic acids to form stable complexes. The protonated amine groups allow transportation to cellular membranes and subsequent endocytosis into cells. Moreover, the high amounts of chitosan in siRNA complexes may lead to increase cellular accumulation of siRNA molecules and facilitate release of siRNA from endosomes to cytosol under high osmotic pressure in the endosomes of cells [42].

Chitosan-based nanocarriers are prepared by three different methods. These include simple complexation, ionic gelation (siRNA entrapment), and adsorption of siRNA onto the surface of chitosan nanoparticles [42]. The molecular weight and degree of deacetylation of chitosan influence its solubility, hydrophobicity, charge density, and thus the interaction ability with nucleic acids. The N/P ratio (ratio between chitosan nitrogen per siRNA phosphate) of chitosan/siRNA nanoplexes is an important factor for optimization of complex properties (size and zeta potential), transfection, and gene silencing efficiency. Increasing the N/P ratio not only helps to obtain a high transfection efficiency but also enhances toxicity. The excess of free chitosan in the formulations can interact with cell membrane and cellular process, and thus, may reduce cell viability [41].

Chitosan has a great potential in siRNA-based cancer therapy studies, because it can be safely and efficiently delivered to cancer cells. It is reported that chitosan or modified chitosan nanoplexes and nanoparticles as delivery system exerted antitumoral effects in different cancers [43–48].

#### **Studies with chitosan formulations in different cancers**

Howard et al. [43] developed chitosan nanoparticles using polyelectrolyte complexation method. The size of nanoparticles was between 40 and 600 nm. The endogenous enhanced green fluorescent protein (EGFP) silencing efficiency with nanoparticles was found to be 77.9 and 89.3% in human lung carcinoma cells (H1299) and murine peritoneal macrophages. The siRNA/chitosan nanoparticles reduced EGFP expression (43%) compared to untreated control in transgenic EGFP mice. They suggested that this chitosan-based system can be used in the treatment of systemic and mucosal diseases.

Salva and Akbuga [44] studied silencing effect of chitosan/*VEGF* shRNA nanoplexes in breast cancer cell lines. A significant *VEGF* gene silencing (60%) was obtained after nanoplexes application in MCF-7 cells. Salva et al. [44, 45] demonstrated the successful application of chitosan/siRNA or shRNA *VEGF* nanoplexes in *in vivo* breast cancer models. After intratu‐ moral and intraperitoneal injection, comparison was made and higher tumor inhibition was obtained with intratumoral injection. qRT-PCR and Western Blot analysis showed that *VEGF* mRNA and protein expression was significantly reduced by chitosan nanoplexes.

Salva et al. [46] also studied the IL-4 encoded plasmid (pIL-4) to improve the therapeutic efficacy of siRNA targeting *VEGF* because of the anti-angiogenic effect of IL-4 molecule. Researchers prepared chitosan nanoparticles containing shRNA *VEGF* and pIL-4, and they have reported that co-delivery of shRNA *VEGF* and pIL-4 into chitosan nanoparticles caused additive effect on breast tumor cell growth in rat model (97% inhibition) [46].

In another study, Salva et al. [47] obtained enhanced silencing effect by using siRNAs targeting to *VEGF* and *HIF-1α* in different breast cancer cell lines such as MCF-7, MDA-MB-231, and MDA-MB-435. Two siRNAs were encapsulated into liposome coated with chitosan, and the co-delivery of siRNA *VEGF* and *HIF-1α* into liposomal form have significantly inhibited *VEGF* (89%) and the *HIF-1α* (62%) [47].

erally Recognized As Safe (GRAS)" by the FDA [41]. It has been widely used in *in vivo* siRNA and shRNA delivery applications because of positively charged amines, allowing electrostatic interactions with negatively charged nucleic acids to form stable complexes. The protonated amine groups allow transportation to cellular membranes and subsequent endocytosis into cells. Moreover, the high amounts of chitosan in siRNA complexes may lead to increase cellular accumulation of siRNA molecules and facilitate release of siRNA from endosomes to cytosol

Chitosan-based nanocarriers are prepared by three different methods. These include simple complexation, ionic gelation (siRNA entrapment), and adsorption of siRNA onto the surface of chitosan nanoparticles [42]. The molecular weight and degree of deacetylation of chitosan influence its solubility, hydrophobicity, charge density, and thus the interaction ability with nucleic acids. The N/P ratio (ratio between chitosan nitrogen per siRNA phosphate) of chitosan/siRNA nanoplexes is an important factor for optimization of complex properties (size and zeta potential), transfection, and gene silencing efficiency. Increasing the N/P ratio not only helps to obtain a high transfection efficiency but also enhances toxicity. The excess of free chitosan in the formulations can interact with cell membrane and cellular process, and thus,

Chitosan has a great potential in siRNA-based cancer therapy studies, because it can be safely and efficiently delivered to cancer cells. It is reported that chitosan or modified chitosan nanoplexes and nanoparticles as delivery system exerted antitumoral effects in different

Howard et al. [43] developed chitosan nanoparticles using polyelectrolyte complexation method. The size of nanoparticles was between 40 and 600 nm. The endogenous enhanced green fluorescent protein (EGFP) silencing efficiency with nanoparticles was found to be 77.9 and 89.3% in human lung carcinoma cells (H1299) and murine peritoneal macrophages. The siRNA/chitosan nanoparticles reduced EGFP expression (43%) compared to untreated control in transgenic EGFP mice. They suggested that this chitosan-based system can be used in the

Salva and Akbuga [44] studied silencing effect of chitosan/*VEGF* shRNA nanoplexes in breast cancer cell lines. A significant *VEGF* gene silencing (60%) was obtained after nanoplexes application in MCF-7 cells. Salva et al. [44, 45] demonstrated the successful application of chitosan/siRNA or shRNA *VEGF* nanoplexes in *in vivo* breast cancer models. After intratu‐ moral and intraperitoneal injection, comparison was made and higher tumor inhibition was obtained with intratumoral injection. qRT-PCR and Western Blot analysis showed that *VEGF*

Salva et al. [46] also studied the IL-4 encoded plasmid (pIL-4) to improve the therapeutic efficacy of siRNA targeting *VEGF* because of the anti-angiogenic effect of IL-4 molecule. Researchers prepared chitosan nanoparticles containing shRNA *VEGF* and pIL-4, and they have reported that co-delivery of shRNA *VEGF* and pIL-4 into chitosan nanoparticles caused

mRNA and protein expression was significantly reduced by chitosan nanoplexes.

additive effect on breast tumor cell growth in rat model (97% inhibition) [46].

under high osmotic pressure in the endosomes of cells [42].

**Studies with chitosan formulations in different cancers**

treatment of systemic and mucosal diseases.

may reduce cell viability [41].

cancers [43–48].

208 RNA Interference

Yang et al. [48] reported that chitosan/siRNA *VEGF* nanoparticles prepared by complex coacervation method showed spherical morphology with a mean diameter of 110–200 nm and positively charged surface (20 mV). Chitosan nanoparticles were effectively transfected to mouse melanoma cells (B16-F10), and they have investigated 40% of the *VEGF* gene silencing efficiency in cells without any cytotoxicity.

Wang et al. [49] prepared the chitosan-TPP (tripolyphosphate) nanoparticles by ionic gelation method for the delivery of shRNA expressing vector to the human rhabdomyosarcoma (RD) cell line and for the inhibition of *TGF-β1* expression. Suppression of *TGF-β1* gene by chitosan nanoparticles containing shRNA has resulted in decrease of RD cell growth *in vitro* and tumorigenicity in nude mice.

Huang et al. [50] studied the effect of chitosan/shRNA *VEGF* nanocomplexes on angiogenesis and tumor growth in hepatocellular carcinoma (HCC). The administration of low molecular weight chitosan/shRNA *VEGF* complexes by intratumoral or intravenous injection demon‐ strated more effective suppression of tumor angiogenesis and tumor growth in the different HCC models. They showed that LMWC could effectively deliver shRNA into tumor tissue. shRNA *VEGF* concentrations in tumor tissue dramatically increased after intravenous administration of chitosan/shRNA *VEGF* complexes.

#### **Studies with chitosan derivatives and conjugation with other polymers and ligands in different cancers**

In order to increase the transfection efficiency of chitosan, different modifications are made on the structure of chitosan. Modified forms of chitosan such as carboxymethyl or trimethyl chitosan, trisaccharide-substituted chitosan oligomers, and succinated or galactosylated chitosan are formed. Chitosan is also conjugated with folic acid or PEG [51].

Jere et al. [52] used chitosan-graft-polyethylenimine (CHI-g-PEI) copolymer for delivery of shRNA *Akt1* expressing plasmid in lung cancer cells. The formed complexes were silenced *Akt1* onco-protein and significantly reduced the survival, proliferation, and growth progres‐ sion of lung cancer cell. *Akt1* silencing induced apoptosis in cancer cells. The suppression of *Akt1* oncoprotein decreased A549 cell malignancy and metastasis. The therapeutic efficiency of CHI-g-PEI-shRNA *Akt* was found higher than PEI25K-shRNA *Akt* compared to carrier.

Noh et al. [53] prepared a copolymer containing additional cationic moieties linked with chitosan to enhance the cationic charge of chitosan. Therefore, chitosan derivation with polyl-arginine (PLR) and polyethyleneglycol (PEG) (PLR-grafted CS) polyplexes were used for *in vitro* and *in vivo* delivery of siRNA *RFP*. PLR alone can be cytotoxic, thus conjugation of PLR to chitosan both decreased cytotoxicity of PLR and enhanced siRNA delivery efficiency. The pegylation of cationic polymers reduces the charge of polymers and limits the interaction with cell membranes. PEG-CS-PLR did not significantly reduce the cellular delivery of siRNA. Three intratumoral injections of 120 μg of PEG-CS-PLR/siRNA *RFP* complexes to B16F10-RFP tumor-bearing mice had decreased *RFP* expression at 90% level in tumor tissues. It is indicated that PEG-CS-PLR can be a useful carrier for delivery of oncogene-specific siRNAs.

Fernandes et al. [54] investigated folate conjugation to improve gene transfection efficiency of chitosan. When chitosan was conjugated with folate, the folate-chitosan-siRNA complexes have increased gene silencing efficiency because of promoted uptake in HeLa and OV-3 cell lines, which are known to have high folate receptor expression. Higher transfection efficiency and lower toxicity of folate-chitosan complexes are reported in folate receptor–positive cells.

Cell penetrating peptide-based systems may improve cellular uptake and gene silencing efficiency of siRNAs without side effects. Protamine is a cationic, non-toxic polypeptide that has membrane translocation and nuclear localization activities because of its arginine-rich amino acid sequences. In addition to its stabilization enhancing properties, protamine is known to exhibit cell penetrating activity and is an important compound for several cancer targeting systems [55].

Salva et al. [46] have developed ternary nanoplexes of chitosan/protamine/siRNA targeting *VEGF* in breast cancer cell lines for efficient siRNA uptake and inhibition effect. Ternary nanoplexes showed the highest cellular uptake than binary nanoplexes.

Erdem-Cakmak et al. [56] reported that addition of protamine to chitosan complexes increased the silencing of *VEGF* genes after using chitosan/shRNA/protamine nanoplexes. In terms of the gene silencing and transfection, when the molecular weight of chitosans were compared at the different cell lines including HEK293, HeLa, and MCF-7, low molecular weight chitosan (70 kDa) proved more efficient than medium molecular weight chitosan. Gene inhibition values in cell lines after transfection of binary and ternary complexes followed the rank HEK293>HeLa>MCF-7. In addition, any cytotoxicity was not found after the complexes.

Song et al. [57] used protamine/antibody fusion protein to deliver siRNAs targeting *c-my*, *MDM2,* and *VEGF* specifically to HIV envelope-expressing B16 melanoma cells and envelopeexpressing subcutaneous B16 tumors. The positively charged protamine served as binding partner for negatively charged siRNA and showed cell internalization and release of the siRNA cargo. The antibody-protamine delivery system can target siRNA specifically to cells.

Choi et al. [58] reported that complexes prepared with low molecular weight protamine (LMWP) inhibited cell growth by suppressing *VEGF* expression in hepatocarcinoma cancer cells. In tumor tissues, the expression of *VEGF* was inhibited through the systemic application of peptide complex, thereby suppressing tumor growth.

#### *3.1.1.2. Polyethylenimine*

Polyethylenimines (PEIs) are water-soluble cationic synthetic polymers. They can be synthe‐ sized in different lengths and different molecular weights such as branched (bPEI) or linear (lPEI) and low molecular weight (<1000 Da) or high molecular weight (>1000 kDa). PEI has a high cationic charge density because of the protonation of its primary, secondary, and tertiary amine groups positioned at every third nitrogen [59]. While in linear PEI all nitrogen atoms are protonable, in the branched form, two-thirds of nitrogens can be charged. PEI can lead to proton accumulation in endosome, which was brought in by endosomal ATPase with an influx of chloride anion. Proton accumulation in endolysosome counteracts pH decrease, inhibits nucleases and unbalances endosome osmolarity depend on CI concentration and results in osmotic swelling of endosome. This effect of PEI is named as "proton sponge effect". PEI may enhance intracellular delivery by facilitating endosomal escape and induce lysosomal dis‐ truption, endosomal release, and DNA/siRNA protection from lysosomal degradation by buffering endosomes [60].

The molecular weight of PEI is important in the development of gene delivery and level of cytotoxicity in cells. The high molecular weight PEI has higher transfection efficiency than low molecular weight PEIs. PEI has a high electrostatic capacity, which can provide strong electrostatic interactions with the siRNA and contribute to cell membrane binding and internalization. Especially, the 25 kDa bPEI is one of the most effective non-viral vectors in gene silencing because of efficient endosomal escape. However, the high positive charge of bPEI leads to severe cytotoxicity and non-specific interactions with serum proteins [61, 62]. The cytotoxicity of PEIs can be decreased with modification of free amine groups or conjuga‐ tion of cell binding and targeting ligands. Therefore, graft copolymers have been usually preferred as a delivery system.

Schiffelers et al. [63] prepared PEGylated PEI nanoplexes with Arg-Gly-Asp (RGD) peptide ligand containing siRNA targeting *VEGFR-2* and investigated the effect of angiogenesis and tumor growth in tumor-bearing mice. This study indicated that nanoplexes containing siRNA *VEGFR-2* reached tumor tissues after systemic administration. This delivery system has sequence-specific inhibition effect and reduced the tumor growth.

Jiang et al. [64] studied anti-*VEGF* siRNA/PEI-HA complex prepared by PEI-hyaluronic acid (PEI-HA). Complexes at the dose of 4.5 μg of siRNA/mouse were applied intratumorally to C57BL/6 mice by daily injection for 3 days. At 24 hours post-injection, the siRNA *VEGF* formulations were distributed mainly in the tumor, spleen, lung, heart, liver, and kidney. This study suggested that siRNA *VEGF*/PEI-HA complexes can be used for the treatment of cancer in the tissues having HA receptors such as the liver and kidney.

Park et al. [61] synthesized siRNA/(PEI-SS)-b-HA complexes and, after characterization, applied to *in vitro* and *in vivo* gene silencing for target-specific tumor treatment. This delivery system demonstrated an excellent *in vitro* gene silencing efficiency (50–80%). siRNA *VEGF*/ (PEI-SS)-b-HA complexes were administrated intratumorally to colorectal tumor bearing mice every 3 days. After the treatment of tumor, *VEGF* gene silencing and reduction in tumor growth were seen.

#### *3.1.2. Other non-viral delivery systems*

cell membranes. PEG-CS-PLR did not significantly reduce the cellular delivery of siRNA. Three intratumoral injections of 120 μg of PEG-CS-PLR/siRNA *RFP* complexes to B16F10-RFP tumor-bearing mice had decreased *RFP* expression at 90% level in tumor tissues. It is indicated

Fernandes et al. [54] investigated folate conjugation to improve gene transfection efficiency of chitosan. When chitosan was conjugated with folate, the folate-chitosan-siRNA complexes have increased gene silencing efficiency because of promoted uptake in HeLa and OV-3 cell lines, which are known to have high folate receptor expression. Higher transfection efficiency and lower toxicity of folate-chitosan complexes are reported in folate receptor–positive cells.

Cell penetrating peptide-based systems may improve cellular uptake and gene silencing efficiency of siRNAs without side effects. Protamine is a cationic, non-toxic polypeptide that has membrane translocation and nuclear localization activities because of its arginine-rich amino acid sequences. In addition to its stabilization enhancing properties, protamine is known to exhibit cell penetrating activity and is an important compound for several cancer

Salva et al. [46] have developed ternary nanoplexes of chitosan/protamine/siRNA targeting *VEGF* in breast cancer cell lines for efficient siRNA uptake and inhibition effect. Ternary

Erdem-Cakmak et al. [56] reported that addition of protamine to chitosan complexes increased the silencing of *VEGF* genes after using chitosan/shRNA/protamine nanoplexes. In terms of the gene silencing and transfection, when the molecular weight of chitosans were compared at the different cell lines including HEK293, HeLa, and MCF-7, low molecular weight chitosan (70 kDa) proved more efficient than medium molecular weight chitosan. Gene inhibition values in cell lines after transfection of binary and ternary complexes followed the rank HEK293>HeLa>MCF-7. In addition, any cytotoxicity was not found after the complexes.

Song et al. [57] used protamine/antibody fusion protein to deliver siRNAs targeting *c-my*, *MDM2,* and *VEGF* specifically to HIV envelope-expressing B16 melanoma cells and envelopeexpressing subcutaneous B16 tumors. The positively charged protamine served as binding partner for negatively charged siRNA and showed cell internalization and release of the siRNA

Choi et al. [58] reported that complexes prepared with low molecular weight protamine (LMWP) inhibited cell growth by suppressing *VEGF* expression in hepatocarcinoma cancer cells. In tumor tissues, the expression of *VEGF* was inhibited through the systemic application

Polyethylenimines (PEIs) are water-soluble cationic synthetic polymers. They can be synthe‐ sized in different lengths and different molecular weights such as branched (bPEI) or linear (lPEI) and low molecular weight (<1000 Da) or high molecular weight (>1000 kDa). PEI has a high cationic charge density because of the protonation of its primary, secondary, and tertiary

cargo. The antibody-protamine delivery system can target siRNA specifically to cells.

of peptide complex, thereby suppressing tumor growth.

nanoplexes showed the highest cellular uptake than binary nanoplexes.

that PEG-CS-PLR can be a useful carrier for delivery of oncogene-specific siRNAs.

targeting systems [55].

210 RNA Interference

*3.1.1.2. Polyethylenimine*

Among cationic polymers, poly (l-Lysine) (PLL) is one of the mostly studied polymers used for nucleic acid delivery. It formed complexes with DNA smaller than 100 nm. Its complexes can target different cells after binding suitable ligands. PLL can be easily produced in large scale and is physiologically stable and biosafe [65]. PLL may protect siRNA from degradation effect of nucleases. However, PLL has some hurdless that impade its clinical application. PLL does not have the proton buffering ability to enhance lysosomal release of transported siRNA. It can be modified also by addition of ligands [66].

The ternary copolymer mPEG-b-PLL-g(ss-IPEI) was used for siRNA delivery to SKOV-3 ovarian cancer treatment. Nanocomplexes were administered to SKOV-3-implated Balb/c mice and tumor growth inhibition was observed [67].

Dendrimers are highly branched spherical and synthetic multifunctional macromolecules. The surface functional groups of dendrimers can be modified to enhance biocompatibility and decrease toxicity. Polycationic dendrimers such as poly(amidoamine) (PAMAM) and poly(propyleneimine) (PPI) dendrimers, because of the high density of positive charges on the surface, are highly attractive for delivery of negatively charged pDNA, antisense oligonucleo‐ tide (AsODN), and siRNAs. PAMAM dendrimers have primary amine groups on their surface and tertiary amine groups inside. Their amine groups are complexed with siRNAs. Thus, compact structure promote cellular uptake of siRNA and tertiary amine groups initiate the proton sponge effect to enhance endosomal release of siRNA [68, 69].

Waite et al. [70] conjugated cationic PAMAM dendrimers with RGD targeting peptides to enhance the delivery efficiency of siRNA to glioma cells. They suggested a promising strategy of RGD-conjugated dendrimers for siRNA delivery to solid tumors.

Liu et al. [71] investigated *in vitro* characterization and anticancer effect of PAMAM dendrimermediated shRNA against human telomerase reverse transcriptase (*hTERT*) in oral cancer. Dendriplexes had 110 nm size and +30 mV zeta potential which were favorable parameters for escape from the vasculature and intracellular delivery. shRNA *hTERT* dendriplexes were applied by intratumoral administration to tumors. Dendrimer-mediated shRNA *TERT* resulted in cell growth inhibition and apoptosis *in vitro* and tumor growth inhibition *in vivo* in the xenograft model. In addition, expression of *HTERT* and *PCNA* proteins was reduced in tumors.

Atelocollagen, which is produced from bovine type I collagen, has biomaterial properties such as high biocompatibility, biodegredability, and low immunogenicity. Atelocollagen forms a helix of three polypeptide chains and has positive charge, which enable its binding to nucleic acid molecules [72]. At low temperature, atelocollagen exists in liquid form (2–10°C), therefore, it can be easily mixed with nucleic acid solutions [72, 73]. Thus, atelocollagen can increase cellular uptake, nuclease resistance, and prolong release of nucleic acids. The size, charge, and sustained release of atelocollagen/siRNA complexes can be altered by ratio of siRNA to atelocollagen [74, 75].

Takei et al. [76] first studied anti-tumoral effect of atelocollagen complexes containing siRNA *VEGF in vitro* and *in vivo*. They showed that siRNA *VEGF* with atelocollagen inhibited tumor angiogenesis and tumor growth in PC-3 cell lines *in vitro* and xenograft tumor *in vivo* model.

Koyanagi et al. [77] reported that siRNA targeting vasohibin-2 (*VASH*-*2*) using atelocollagen complexes siginificantly inhibited ovarian tumor growth and angiogenesis in ovarian cancer xenograft model. The knockdown of *VASH2* with atelocollagen/siRNA *VASH2* complexes exerted a significant antitumor effect and helped in tumor vascularization.

PLGA has been widely used as gene delivery system because of its biodegredability, biocom‐ patibility, and non-toxic properties. FDA has approved PLGA as a pharmaceutical excipient. PLGA nanoparticles enter the cells efficiently by specific and non-specific endocytosis. Nanoparticles can release the encapsulated drug slowly leading to sustained drug effect [25].

Murata et al. [78] investigated anti-tumor effect of long-term sustained release of PLGA microspheres encapsulating anti-*VEGF* siRNA. The release of siRNA from microspheres was sustained for over one month. Intratumoral injection of PLGA microspheres containing siRNA *VEGF* inhibited tumor growth.

Su et al. [79] prepared PEI-coated PLGA nanoparticles loaded with paclitaxel and *Stat3* siRNA. PLGA-PEI nanoparticles more rapidly released *Stat3* siRNA than paclitaxel. Thus, decrease of *Stat3* expression by siRNA in human lung cancer cells (A549) and A549-derived paclitaxelresistant A549/T12 cell lines reduced resistance of cell to paclitaxel. The released paclitaxel from nanoparticles killed the cancer cells that induce microtubule aggregation. In summary, inhibition of *Stat3* expression decreased cell viability, increased apoptosis, and reduced cellular resistance to paclitaxel.

#### *3.1.3. Lipid-based siRNA delivery systems in cancer therapy*

scale and is physiologically stable and biosafe [65]. PLL may protect siRNA from degradation effect of nucleases. However, PLL has some hurdless that impade its clinical application. PLL does not have the proton buffering ability to enhance lysosomal release of transported siRNA.

The ternary copolymer mPEG-b-PLL-g(ss-IPEI) was used for siRNA delivery to SKOV-3 ovarian cancer treatment. Nanocomplexes were administered to SKOV-3-implated Balb/c mice

Dendrimers are highly branched spherical and synthetic multifunctional macromolecules. The surface functional groups of dendrimers can be modified to enhance biocompatibility and decrease toxicity. Polycationic dendrimers such as poly(amidoamine) (PAMAM) and poly(propyleneimine) (PPI) dendrimers, because of the high density of positive charges on the surface, are highly attractive for delivery of negatively charged pDNA, antisense oligonucleo‐ tide (AsODN), and siRNAs. PAMAM dendrimers have primary amine groups on their surface and tertiary amine groups inside. Their amine groups are complexed with siRNAs. Thus, compact structure promote cellular uptake of siRNA and tertiary amine groups initiate the

Waite et al. [70] conjugated cationic PAMAM dendrimers with RGD targeting peptides to enhance the delivery efficiency of siRNA to glioma cells. They suggested a promising strategy

Liu et al. [71] investigated *in vitro* characterization and anticancer effect of PAMAM dendrimermediated shRNA against human telomerase reverse transcriptase (*hTERT*) in oral cancer. Dendriplexes had 110 nm size and +30 mV zeta potential which were favorable parameters for escape from the vasculature and intracellular delivery. shRNA *hTERT* dendriplexes were applied by intratumoral administration to tumors. Dendrimer-mediated shRNA *TERT* resulted in cell growth inhibition and apoptosis *in vitro* and tumor growth inhibition *in vivo* in the xenograft model. In addition, expression of *HTERT* and *PCNA* proteins was reduced in

Atelocollagen, which is produced from bovine type I collagen, has biomaterial properties such as high biocompatibility, biodegredability, and low immunogenicity. Atelocollagen forms a helix of three polypeptide chains and has positive charge, which enable its binding to nucleic acid molecules [72]. At low temperature, atelocollagen exists in liquid form (2–10°C), therefore, it can be easily mixed with nucleic acid solutions [72, 73]. Thus, atelocollagen can increase cellular uptake, nuclease resistance, and prolong release of nucleic acids. The size, charge, and sustained release of atelocollagen/siRNA complexes can be altered by ratio of siRNA to

Takei et al. [76] first studied anti-tumoral effect of atelocollagen complexes containing siRNA *VEGF in vitro* and *in vivo*. They showed that siRNA *VEGF* with atelocollagen inhibited tumor angiogenesis and tumor growth in PC-3 cell lines *in vitro* and xenograft tumor *in vivo* model.

Koyanagi et al. [77] reported that siRNA targeting vasohibin-2 (*VASH*-*2*) using atelocollagen complexes siginificantly inhibited ovarian tumor growth and angiogenesis in ovarian cancer

It can be modified also by addition of ligands [66].

and tumor growth inhibition was observed [67].

tumors.

212 RNA Interference

atelocollagen [74, 75].

proton sponge effect to enhance endosomal release of siRNA [68, 69].

of RGD-conjugated dendrimers for siRNA delivery to solid tumors.

Cationic lipids are used as carrier for siRNA delivery. Liposomes and lipoplexes, as lipid-based delivery systems, have been widely used in local and systemic siRNA or shRNA delivery. Liposomes are microscopic vesicles that consist of single or multiple lipid bilayer, form in a sphere with an aqueous core. Nucleic acids can either be entrapped in the aqueous core of liposomes or attached to the lipid surface for delivery. The advantages of liposomes as delivery system include a high gene transfection efficiency, enhanced stabilization, easy penetration into cell membranes, efficient *in vivo* delivery, and flexible and versatile physicochemical properties. The disadvantages of liposomes are the short half-life in serum, lack of tissue specificity, rapid liver clearence, and cell toxicity [17]. Three different liposomes, such as neutral, anionic, and cationic liposomes, are used in the siRNA delivery studies [22]. Cationic liposomes for siRNA delivery can easily cross the cell membrane, promote escape from the endosomal compartment, and reach the target genes with good biocompatibility. However, cationic lipids can induce an interferon response and cause unwanted interactions with negatively charged serum proteins because of its high cationic charge density [32, 67]. Interferon responses can lead to not only change in gene expression but also show dosedependent cytotoxicity and pulmonary inflammation [80, 81]. The toxicity and transfection efficiency of cationic lipids depend on length and structure of hydrocarbon chains of lipids [82].

Neutral lipids lead to less cellular toxicity and do not induce immune responses without the down-regulation of gene expression. However, neutral liposomes have shown low transfec‐ tion efficiency because of their lack of surface charges [17]. The commonly used cationic lipids for siRNA delivery include 1,2-dioleoyloxy-3-trimethylammonium propane (DOTAP) and 1,2 di-o-octadecenyl-2-trimethylammonium propane (DOTMA) have combined with neutral lipids such as 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE). This combination can enhance transfection efficiency. Because neutral lipids facilitate fusion to the host cell's membrane, cationic lipids can facilitate electrostatically complexation with siRNA to obtain more stable formulation and entry to cells more easily [18]. Liposomes are usually more stable than lipoplex in biological fluids [5].

Lipid-based siRNA delivery strategies have shown as promising in cancer therapy. Tumortargeting approaches have been used to enhance antitumor efficacy of these delivery systems. Specific delivery to target cells can be achieved by conjugation of ligands or molecules such as transferin or PEG on the surface of liposomes [20]. The targeting and prolonged circulation half-life of liposomes allow for the enhanced permeability of tumor vasculature, increased delivery to tumor tissue, and reduced side effects [34, 82]. Cationic liposomes containing siRNA targeting tumor-associated genes have been used to inhibit tumor growth and prolif‐ eration, induce apoptosis, and enhance the radiosensitivity of tumor cells [83–85].

Cationic lipids can interact with negatively charged siRNA by ionic interactions. Thus, selfassembly formed lipoplexes protect siRNA from enzymatic degradation, enhance cellular uptake of siRNA by endocytosis, enhance the release of siRNA from endosomal/lysosomal entrapment, and thus, promote siRNA accumulation in the cytosol [16]. Commercially available cationic lipid formulations such as Lipofectin®, Lipofectamine® (Invitrogen), Dharmafect® (Dharmacon), RNAifect® (Qiagen), and TransIT TKO® (Mirus) have been studied as tranfection reagents for siRNA delivery *in vitro* [86]. The ratio of lipid and siRNA (lipid/siRNA ratio) affects the colloidal properties of the lipoplexes (particle size and zeta potential). Lipid/siRNA or shRNA ratio is important to facilitate the cellular internalization of lipoplexes and to dissociate the nucleic acids in the cytosol. Lipid/siRNA ratio can be optimized in terms of biological activity [16]. Developing a lipid-based delivery system, choice of lipids, and appropriate formulations are essential to decrease cytotoxicity and increase the transfec‐ tion efficiency of formulation.

To overcome the drawbacks of lipoplexes and liposomes, different nanostructures such as neutral lipid-based nanoliposomes, stable nucleic acid lipid particles (SNALP), and solid lipid nanoparticles (SLN) have been developed as siRNA delivery system. SNALPs are composed of cationic, neutral, and fusogenic lipid mixture. SNALPs increase cellular uptake and endosomal release of siRNA [4]. PEG-conjugated SNALPs represent exciting lipid-based systemic RNAi. The PEG-lipid conjugate improves the retention time to >10 hours [87]. Recently, Tekmira Pharmaceuticals [88] has developed siRNA-based drugs that are encapsu‐ lated in the SNALPs for delivery of siRNAs to target tissue by intravenous injection. SNALPencapsulated siRNA targeting *PLK1* initiated phase I trial in December 2010. Alnylam Pharmaceuticals [89] has developed first dual-targeted siRNA drug, SNALP-encapsulated siRNAs targeting *VEGF,* and kinesin spindle protein (*KSP*) for the treatment of hepatocellular carinoma. Phase I trial was initiated in April 2009 [90].

Tekedereli et al. [91] indicated that knockdown of *Bcl-2* by intravenously administered nanoliposomal-siRNA *Bcl2* (150 μg siRNA/kg) twice a week lead to antitumoral activity in breast tumors of orthotopic xenograft models. In addition, nanoliposomal-siRNAs have enhanced the efficacy of chemotherapeutic agents in the breast cancer therapy.

Landen et al. [92] studied neutral nanoliposomes incorporating siRNA targeting *EphA2* in orthotopic mouse model of ovarian cancer. Three weeks of treatment with *EphA2*-targeting siRNA nanoliposomes (150 μg/kg twice weekly) reduced tumor growth. The combination therapy with paclitaxel reduced tumor growth.

Salva et al. [47] investigated the effect of co-delivery of siRNA *HIF1-α* and *VEGF* in liposomal form in the breast cancer cell lines. Chitosan-coated liposomal formulation for co-delivery of siRNA *VEGF* and *HIF1-α* were developed. The co-delivery of siRNA *VEGF* and *HIF1-α* was greatly enhanced *in vitro* gene silencing efficiency in the breast camcer cell lines (95%). In addition, chitosan-coated liposomes showed 96% cell viability. Salva et al. has suggested that siRNA-based therapies with chitosan-coated liposomes may have some promises in cancer therapy [47].

In conclusion, siRNA-based therapeutics are new and potential targets in cancer studies. In cancer, different mechanisms including angiogenesis, and cell growth were studied as target pathways. However, siRNAs have different hurdles in treatment because of their short biological life in blood, instability, and poor cellular internalization. In order to overcome these hurdles two solutions are present: one is modification of siRNA and the other is use of suitable siRNA delivery system. In cancer treatment, viral and non-viral delivery systems are evaluated as siRNA delivery. Although limited information is available related to *in vivo* delivery, more papers are present in literature. Viral delivery systems have serious problems. Therefore, nonviral systems are more attractable than viral systems for siRNAs. Cationic lipids, liposomes, and polymers such as chitosan, PEI, PLL, and PLGA are used as non-viral siRNA delivery system. However, more suitable carriers are needed for siRNA delivery systems.

#### **Author details**

lipids such as 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE). This combination can enhance transfection efficiency. Because neutral lipids facilitate fusion to the host cell's membrane, cationic lipids can facilitate electrostatically complexation with siRNA to obtain more stable formulation and entry to cells more easily [18]. Liposomes are usually more stable

Lipid-based siRNA delivery strategies have shown as promising in cancer therapy. Tumortargeting approaches have been used to enhance antitumor efficacy of these delivery systems. Specific delivery to target cells can be achieved by conjugation of ligands or molecules such as transferin or PEG on the surface of liposomes [20]. The targeting and prolonged circulation half-life of liposomes allow for the enhanced permeability of tumor vasculature, increased delivery to tumor tissue, and reduced side effects [34, 82]. Cationic liposomes containing siRNA targeting tumor-associated genes have been used to inhibit tumor growth and prolif‐

Cationic lipids can interact with negatively charged siRNA by ionic interactions. Thus, selfassembly formed lipoplexes protect siRNA from enzymatic degradation, enhance cellular uptake of siRNA by endocytosis, enhance the release of siRNA from endosomal/lysosomal entrapment, and thus, promote siRNA accumulation in the cytosol [16]. Commercially available cationic lipid formulations such as Lipofectin®, Lipofectamine® (Invitrogen), Dharmafect® (Dharmacon), RNAifect® (Qiagen), and TransIT TKO® (Mirus) have been studied as tranfection reagents for siRNA delivery *in vitro* [86]. The ratio of lipid and siRNA (lipid/siRNA ratio) affects the colloidal properties of the lipoplexes (particle size and zeta potential). Lipid/siRNA or shRNA ratio is important to facilitate the cellular internalization of lipoplexes and to dissociate the nucleic acids in the cytosol. Lipid/siRNA ratio can be optimized in terms of biological activity [16]. Developing a lipid-based delivery system, choice of lipids, and appropriate formulations are essential to decrease cytotoxicity and increase the transfec‐

To overcome the drawbacks of lipoplexes and liposomes, different nanostructures such as neutral lipid-based nanoliposomes, stable nucleic acid lipid particles (SNALP), and solid lipid nanoparticles (SLN) have been developed as siRNA delivery system. SNALPs are composed of cationic, neutral, and fusogenic lipid mixture. SNALPs increase cellular uptake and endosomal release of siRNA [4]. PEG-conjugated SNALPs represent exciting lipid-based systemic RNAi. The PEG-lipid conjugate improves the retention time to >10 hours [87]. Recently, Tekmira Pharmaceuticals [88] has developed siRNA-based drugs that are encapsu‐ lated in the SNALPs for delivery of siRNAs to target tissue by intravenous injection. SNALPencapsulated siRNA targeting *PLK1* initiated phase I trial in December 2010. Alnylam Pharmaceuticals [89] has developed first dual-targeted siRNA drug, SNALP-encapsulated siRNAs targeting *VEGF,* and kinesin spindle protein (*KSP*) for the treatment of hepatocellular

Tekedereli et al. [91] indicated that knockdown of *Bcl-2* by intravenously administered nanoliposomal-siRNA *Bcl2* (150 μg siRNA/kg) twice a week lead to antitumoral activity in breast tumors of orthotopic xenograft models. In addition, nanoliposomal-siRNAs have

enhanced the efficacy of chemotherapeutic agents in the breast cancer therapy.

eration, induce apoptosis, and enhance the radiosensitivity of tumor cells [83–85].

than lipoplex in biological fluids [5].

214 RNA Interference

tion efficiency of formulation.

carinoma. Phase I trial was initiated in April 2009 [90].

Emine Şalva1\*, Ceyda Ekentok2 , Suna Özbaş Turan<sup>2</sup> and Jülide Akbuğa<sup>2</sup>

\*Address all correspondence to: emine\_salva@yahoo.com

1 Faculty of Pharmacy, Department of Pharmaceutical Biotechnology, İnönü University, Turkey

2 Faculty of Pharmacy, Department of Pharmaceutical Biotechnology, Marmara University, Turkey

#### **References**

[1] Dominska M, Dykxhoorn DM. Breaking down the barriers: siRNA delivery and en‐ dosome escape. J Cell Sci. 2010;123:1183-1189. DOI:10.1242/jcs.066399


[17] Deng Y, Wang CC, Choy KW, Du Q, Chen J, Wong Q et al. Therapeutic potentials of gene silencing by RNA interference: Principles, challenges and new strategies. Gene. 2014;538:217-227. DOI:10.1016/j.gene.2013.12.019

[2] Aravin A, Tuschi T. Identification and characterization of small RNA involved in RNA silencing. FEBS Lett. 2005;579:5830-5840. DOI:10.1186/1471-2164-10-443

[3] Liu X, Rocchi P, Peng L. Dendrimers as non-viral vectors for siRNA delivery. New J

[4] Lee JM, Yoon TJ, Cho YS. Recent developments in nanoparticle-based siRNA deliv‐ ery for cancer therapy. Biomed Res Int. 2013;2013:1-10. DOI:10.1155/2013/782041

[5] Chen Y, Huang L. Tumor-targeted delivery of siRNA by non-viral vector: Safe and effective cancer therapy. Expert Opin Drug Del. 2008;5:1301-1311. DOI:

[6] Draz MS, Fang BA, Zhang P, Hu Z, Gu S, Weng KC et al. Nanoparticle-mediated sys‐ temic delivery of siRNA for treatment of cancers and viral infections. Theranostics.

[7] Xu CF, Wang J. Delivery systems for siRNA drug development in cancer therapy.

[8] Jones SW, deSouza PM, Lindsay MA. siRNA for gene silencing: A route to drug tar‐ get discovery. Curr Opin Pharmacol. 2004;4:522-527. DOI:10.1016/j.coph.2004.06.003

[9] Laufer SD, Detzer A, Sczakiel G, Restle T. Selected strategies for the delivery of siR‐ NA in vitro an in vivo. In: Erdmann VA, Barciszewski J, editors. *RNA Technologies*

[10] Rao DD, Vorhies JS, Senzer N, Nemunaitis J. siRNA vs shRNA: Similarities and dif‐ ferences. Adv Drug Deliver Rev. 2009;61:746-759. DOI:10.1016/j.addr.2009.04.004

[12] Amarzguioui M, Rossi JJ, Kim D. Approaches for chemically synthesized siRNA and vector-mediated RNAi. FEBS Lett. 2005;579:5974-5981. DOI:10.1016/j.febslet.

[13] Aigner A. Delivery systems for the direct application of siRNAs to induce RNA inter‐ ference (RNAi) in vivo. J Biomed Biotechnol. 2006;2006:1-15. DOI:10.1155/JBB/

[14] Lam JKW, Chow MYT, Zhang Y, Leung SWS. siRNA versus miRNA as therapeutics for gene silencing. Mol Ther-Nuc Acids. 2005;4:e252. DOI:10.1038/mtna.2015.23

[15] Gary DJ, Puri N, Won YY. Polymer-based siRNA delivery: Perspectives on the fun‐ damental and phenomenological distinctions from polymer-based DNA delivery. J

[16] Wang J, Lu Z, Wientjes MG, Au JLS. Delivery of siRNA therapeutics: Barriers and

Control Release. 2007;121:64-73. DOI:10.1016/j.jconrel.2007.05.021

carriers. AAPS J. 2010;12:492-503. DOI:10.1208/s12248-010-9210-4

Asian J Pharm Sci. 2015;10:1-12. DOI:10.1016/j.ajps.2014.08.011

*and Their Applications*. Berlin: Springer-Verlog; 2010. pp. 29-58.

[11] Latterich M, editor. *RNAi*. New York:Taylor & Francis Group; 2008, 23 p.

Chem. 2012;36:256-263. DOI:10.1039/C1NJ20408D

10.1517/17425240802568505

216 RNA Interference

2005.08.070

2006/71659

2014;4:872-892. DOI:10.7150/thno.9404


pIL-4 into chitosan nanoparticles in the breast tumor model. J Pharm Sci. 2014;103:785-795. DOI:10.1002/jps.23815

[47] Salva E, Turan SO, Eren F, Akbuga J. The enhancement of gene silencing efficiency with chitosan-coated liposome formulations of siRNAs targeting HIF-1α and VEGF. Int J Pharm. 2015;478:147-154. DOI:10.1016/j.ijpharm.2014.10.065

[32] Gao Y, Liu XL, Li XR. Research progress on siRNA delivery with nonviral carriers.

[33] Juliano R, Bauman J, Kang H, Ming X. Biological barriers to therapy with antisense and siRNA oligonucleotides. Mol Pharm. 2009;6:686-695. DOI:10.1021/mp900093r [34] Ozpolat B, Sood AK, Lopez-Berestein G. Nanomedicine based approaches for the de‐ livery of siRNA in cancer. J Intern Med. 2010;267:44-53. DOI:10.1111/j.

[35] Amer MH. Gene therapy for cancer: Present status and future perspective. Mol Cell

[36] Oliveirea S, Storm G, Schiffelers RM. Targeted delivery of siRNA. J Biomed Biotech‐

[37] Zhang Y, Satteriee A, Huang L. In vivo gene delivery by nonviral vectors: Overcom‐

[38] Scomparin A, Polyak D, Krivitsky A, Satchi-Falnaro R. Achieving successful delivery of oligonucleotides – From physico-chemical characterization to in vivo evaluation.

[39] Zhou Y, Zhang C, Liang W. Development of RNAi technology for targeted therapy – A track of siRNA based agents to RNAi therapeutics. J Control Release.

[40] Martimprey H, Vauthier C, Malvy C, Couvreur P. Polymer nanocarriers for the de‐ livery of small fragments of nucleic acids: Oligonucleotides and siRNAs. Eur J Pharm

[41] Ragelle H, Vandermeulen G, Preat V. Chitosan-based siRNA delivery systems. J

[42] Mao S, Sun W, Kissel T. Chitosan-based formulations for delivery of DNA and siR‐ NA. Adv Drug Deliver Rev. 2010;62:12-27. DOI:10.1016/j.addr.2009.08.004

[43] Howard KA, Rahbek UL, Liu X, Damgaard CK, Glud SZ, Andersen MØ, et al. RNA interference in vitro and in vivo using a chitosan/siRNA nanoparticle system. Mol

[44] Salva E, Akbuga J. In vitro silencing effect of chitosan nanoplexes containing siRNA expressing vector targeting VEGF in breast cancer cell lines. Pharmazie.

[45] Salva E, Kabasakal L, Eren F, Ozkan N, Cakalagaoglu F, Akbuga J. Local delivery of chitosan/VEGF siRNA nanoplexes reduces angiogenesis and growth of breast cancer

[46] Salva E, Turan SO, Kabasakal L, Alan S, Ozkan N, Eren F, et al. Investigation of the therapeutic efficacy of codelivery of psiRNA-vascular endothelial growth factor and

in vivo. Nucleic Acid Ther. 2012;22:40-48. DOI:10.1089/nat.2011.0312

Biotechnol Adv. DOI: http://dx.doi.org/10.1016/j.biotechadv.2015.04.008.

ing hurdles? Mol Ther. 2012;20:1298-1304. DOI:10.1038/mt.2012.79

Int J Nanomed. 2011;6:1017-1025. DOI:10.2147/IJN.S17040

1365-2796.2009.02191.x

218 RNA Interference

Ther. 2014;2:27. DOI:10.1186/2052-8426-2-27

nol. 2006;2006:1-9. DOI:10.1155/JBB/2006/63675

2014;193:270-281. DOI:10.1016/j.jconrel.2014.04.044

Biopharm. 2009;71:490-504. DOI:10.1016/j.ejpb.2008.09.024

Ther. 2006;14:476-484. DOI:10.1016/j.ymthe.2006.04.010

2010;65:896-902. DOI:10.1691/ph.2010.0192

Control Release. 2013;172:207-218. DOI:10.1016/j.jconrel.2013.08.005


[71] Liu X, Huang H, Wang J, Wang C, Wang M, Zhang B, et al. Dendrimers-delivered short hairpin RNA targeting HTERT inhibits oral cancer cells growth in vitro and in vivo. Biochem Pharmacol. 2011;82:17-23. DOI:10.1016/j.bcp.2011.03.017

[58] Choi YS, Lee JY, Sun JS, Kwon YM, Lee SJ, Chung JK, et al. The systemic delivery of siRNAs by a cell penetrating peptide, low molecular weight protamine. Biomaterials.

[59] Morille M, Passirani C, Vonarbourg A, Clavreul A, Benoit JP. Progress in developing cationic vectors for non-viral systemic gene therapy against cancer. Biomaterials.

[60] Templeton NS, editor. *Gene and Cell Therapy Therapeutic Mechanisms and Strategies*. 3rd

[61] Park K, Lee MY, Kim KS, Hahn SK. Target specific tumor treatment by VEGF siRNA complexed with reducible polyethylenimine-hyaluronic acid conjugate. Biomaterials.

[62] Yang J, Liu H, Zhang X. Design, preparation and application of nucleic acid delivery carriers. Biotechnol Adv. 2014;32:804-817. DOI:10.1016/j.biotechadv.2013.11.004 [63] Schiffelers RM, Ansari A, Xu J, Zhou Q, Tang Q, Storm G, et al. Cancer siRNA thera‐ py by tumor selective delivery with ligand-targeted sterically stabilized nanoparticle.

[64] Jiang G, Park K, Kim J, Kim KS, Hahn SK. Target specific intracellular delivery of siRNA/PEI-HA complex by receptor mediated endocytosis. Mol Pharm.

[65] Howard KA. Delivery of RNA inerference therapeutics using polycation-based nano‐ particles. Adv Drug Deliver Rev. 2009;61:710-720. DOI:10.1016/j.addr.2009.04.001 [66] Scholz C, Wagner E. Therapeutic plasmid DNA versus siRNA delivery: Common and different tasks for synthetic carriers. J Control Release. 2012;161:554-565. DOI:

[67] Li J, Cheng D, Yin T, Chen W, Lin Y, Chen J, et al. Copolymer of poly(ethylene gly‐ col) and poly-(L-lysine) grafting polyrthylenimine through a reducible disulfide link‐ age for siRNA delivery. Nanoscale. 2014;6:1732-1740. DOI:10.1039/C3NR05024F [68] Zhou J, Shum KT, Burneti JC Rossi JJ. Nanoparticle-based delivery of RNAi thera‐ peutics: Progress and challenges. Pharmaceuticals. 2013;6:85-107. DOI:10.3390/

[69] McCarroll J, Kavallaris M. Nanoparticle delivery of siRNA as a novel therapeutic for

[70] Waite CL, Roth CM. PAMAM-RGD conjugates enhance siRNA delivery through a multicellular spheroşd model of malignant glioma. Bioconjug Chem.

2010;31:1429-1443. DOI:10.1016/j.biomaterials.2009.11.001

2008;29:3477-3496. DOI:10.1016/j.biomaterials.2008.04.036

2010;31:5258-5265. DOI:10.1016/j.biomaterials.2010.03.018

Nucleic Acids Res. 2004;32:e149. DOI:10.1093/nar/gnh140

human disease. Australian Biochemist. 2012;43:9-13.

2009;20:1908-1916. DOI:10.1021/bc900228m

2009;6:727-737. DOI:10.1021/mp800176t

10.1016/j.jconrel.2011.11.014

ph6010085

220 RNA Interference

ed. New York: Taylor & Francis Group; 2008, 330 p.


## **siRNA-Induced RNAi Therapy in Acute Kidney Injury**

### Cheng Yang and Bin Yang

[84] Yang W, Sun T, Cao J, Liu F. Survivin downregulation by siRNA/cationic liposome complex radiosensitises human hepatoma cells in vitro and in vivo. Int J Radiat Biol.

[85] Yao Y, Su Z, Liang Y, Zhang N. pH-Sensitive carboxymethyl chitosan-modified cati‐ onic liposomes for sorafenib and siRNA co-delivery. Int J Nanomedicine.

[86] Zhang S, Zhi D, Huang L. Lipid-based vectors for siRNA delivery. J Drug Target.

[87] Li W, Szoka FC. Lipid-based nanoparticles for nucleic acid delivery. Pharm Res.

[88] Ozcan G, Ozpolat B, Coleman RL, Sood AK, Lopez-Berestein G. Preclinical and clini‐ cal development of siRNA-based therapeutics. Adv Drug Deliv Rev. 2015;87:108-119.

[89] Davidson BL, McCray PB Jr. Current prospects for RNA interference-based thera‐

[90] siRNA clinical trials [Internet]. Available from: https://www.clinicaltrials.gov/ct2/

[91] Tekedereli I, Alpay SN, Akar U, Yuca E, Ayugo-Rodriguez C, Han HD, et al. Thera‐ peutic silencing of Bcl-2 by systemically administered siRNA nanotherapeutics inhib‐ its tumor growth by autophagy and apoptosis and enhances the efficacy of therapy in orthotopic xenograft models of ER(-) and ER(+) breast cancer. Mol Ther-Nucleic

[92] Landen CN, Chavez-Reyes A, Bucana C, Schmandt R, Deavers M, Lopez-Berestein G, et al. Therapeutic EphA2 gene targeting in vivo using neutral liposomal small inter‐ fering RNA delivery. Cancer Res. 2005;65:6910-6918. DOI:

2010;86:445-457. DOI: 10.3109/09553001003668006

2012;20:724-735. DOI: 10.3109/1061186X.2012.719232

pies. Nat Rev Genet. 2011;12:329-340. DOI:10.1038/nrg2968

2007;24:438-449. DOI:10.1007/s11095-006-9180-5

DOI:org/10.1016/j.addr.2015.01.007

show/NCT00882180 [Accessed 2016-01-15]

Acids. 2013;2:e121. DOI:10.1038/mtna.2013.45

10.1158/0008-5472.CAN-05-0530

2015;10:6185–6198. DOI: 10.2147/IJN.S90524

222 RNA Interference

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/61838

#### **Abstract**

siRNA therapy has great potential in humans, and its applications have been significantly improved. The kidney is a comparatively easy target organ of siRNA therapy due to its unique structural and functional characteristics. Here, we reviewed recent achievements in the design, delivery, and utilization of RNAi with a focus on kidney diseases, in partic‐ ular acute kidney injury. In addition, the perspectives and challenges of siRNA therapy such as increasing its serum stability and immune tolerance, targeting single/double/ multiple genes, cell/allele-specific delivery, time-controlled silencing, and siRNA-modi‐ fied stem cell therapy were also discussed. Finally, selecting target genes and therapeutic time windows were addressed.

**Keywords:** Small interfering RNA, kidney diseases, delivery, off-target effects, compen‐ sative responses

#### **1. Introduction**

Acute kidney injury (AKI) is very common and critical in clinical practice. The incidence of hospital-acquired AKI is increasing, and many patients require renal replacement therapy [1]. AKI significantly increases the risk of chronic renal disease, end-stage renal disease (ESRD), and death, presenting a major burden to the patient and the health care system. Because of high metabolic activity in handling and transporting ions, amino acids, and other small molecules, the kidney is highly susceptible to acute injuries from lack of sufficient perfusion, exposure to, and accumulation of nephrotoxic substances. Despite numerous clinical trials, AKI remains a cause of significant morbidity and mortality for which there is no effective intervention [2].

RNA interference (RNAi) is a highly conserved biological phenomenon in all eukaryotes, including renal cells. Although RNAi naturally exists, synthetic artificial siRNA exerts similar

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

effects as natural endogenous microRNA (miRNA). Both sense and antisense strands of siRNA can be synthesized separately and annealed to form double stranded siRNA duplexes *in vitro.* After the siRNA is delivered into the cytoplasm, the artificial siRNA silences the target gene using similar biological processes as endogenous miRNA. Since the introduction of 21 nucleotide artificial siRNAs that triggered gene silencing in mammalian cells [3], synthetic siRNA has generated much interest in biomedical research, in which the kidney is one of important key players. siRNA as a strategic molecule has been highly expected in the field of innovative therapy. Because siRNA is highly efficient at gene silencing, it is possible to develop specific siRNA-based drugs that could target any genes, including those that have unknown pharmacological antagonists or inhibitors. Different types of synthetic siRNA have been tested for their efficacy in various disease models, including cancer [4], autoimmune disorders [5], cardiovascular injuries [6], and organ transplantation [7], including native and transplanted kidney injuries [8].

As siRNA is a posttranscriptional regulator, it must first be absorbed into the target cells. Therefore, the kidney could be an excellent target organ for siRNA therapy because it benefits from rapid and vast blood flow physically, subsequent glomerular filtration, and tubular absorption. In fact, the systemic administration of siRNA leads to rapid uptake by the kidney, yielding a significant decrease of target protein expression [8]. Consequently, RNAi by siRNA has advantages for the treatment of renal diseases due to the unique urological system [9]. In addition, the preservation of donor kidneys before transplantation also provides a suitable time window for the intervention of siRNA.

In this chapter, we highlighted the design and delivery of siRNA and its therapeutic effects with a focus on kidney diseases. We also discussed future challenges of siRNA therapy, targeting single/double/multiple genes, cell/allele specific delivery, time-controlled silence, and siRNA-modified stem cell therapy.

#### **2. Current principle of siRNA design**

The design of potent siRNAs has been greatly improved over the past decade. The basic criteria for choosing siRNAs include the consideration of thermodynamic stability, internal repeats, immunostimulatory motifs, such as GC content, secondary structure, base preference at specific positions in the sense strand, and appropriate length [10].

Chemical modifications significantly enhance the stability and uptake of naked siRNAs. Importantly, siRNAs can be directly modified without crippling the silencing ability. Chemical modifications have been rigorously investigated for virtually every part of siRNA molecules, from the termini and backbone to the sugars and bases, with the goal of engineering siRNA to prolong half-life and increase cellular uptake. The most common chemical modification involves modifying the sugar moiety. For example, the incorporation of 2′-fluoro (2′-F), -O methyl, -halogen, -amine, or -deoxy can significantly increase the stability of siRNA in serum.

Locked nucleic acid (LNA) has been also applied to modify siRNA. The commonly used LNA contains a methylene bridge connecting the 2′-oxygen with the 4′-carbon of the ribose ring. This bridge locks the ribose ring in the 3′-endo conformation characteristic of RNA [11]. Additionally, recent studies, including ours [12], have proven the efficacy of LNA-modified siRNA in terms of prolonged half-life in serum, but without detectable adverse effects, suggesting that the natural RNAi machinery could accommodate a certain degree of alterations in the chemical structure of siRNAs [13].

#### **3. siRNA delivery**

effects as natural endogenous microRNA (miRNA). Both sense and antisense strands of siRNA can be synthesized separately and annealed to form double stranded siRNA duplexes *in vitro.* After the siRNA is delivered into the cytoplasm, the artificial siRNA silences the target gene using similar biological processes as endogenous miRNA. Since the introduction of 21 nucleotide artificial siRNAs that triggered gene silencing in mammalian cells [3], synthetic siRNA has generated much interest in biomedical research, in which the kidney is one of important key players. siRNA as a strategic molecule has been highly expected in the field of innovative therapy. Because siRNA is highly efficient at gene silencing, it is possible to develop specific siRNA-based drugs that could target any genes, including those that have unknown pharmacological antagonists or inhibitors. Different types of synthetic siRNA have been tested for their efficacy in various disease models, including cancer [4], autoimmune disorders [5], cardiovascular injuries [6], and organ transplantation [7], including native and transplanted

As siRNA is a posttranscriptional regulator, it must first be absorbed into the target cells. Therefore, the kidney could be an excellent target organ for siRNA therapy because it benefits from rapid and vast blood flow physically, subsequent glomerular filtration, and tubular absorption. In fact, the systemic administration of siRNA leads to rapid uptake by the kidney, yielding a significant decrease of target protein expression [8]. Consequently, RNAi by siRNA has advantages for the treatment of renal diseases due to the unique urological system [9]. In addition, the preservation of donor kidneys before transplantation also provides a suitable

In this chapter, we highlighted the design and delivery of siRNA and its therapeutic effects with a focus on kidney diseases. We also discussed future challenges of siRNA therapy, targeting single/double/multiple genes, cell/allele specific delivery, time-controlled silence,

The design of potent siRNAs has been greatly improved over the past decade. The basic criteria for choosing siRNAs include the consideration of thermodynamic stability, internal repeats, immunostimulatory motifs, such as GC content, secondary structure, base preference at

Chemical modifications significantly enhance the stability and uptake of naked siRNAs. Importantly, siRNAs can be directly modified without crippling the silencing ability. Chemical modifications have been rigorously investigated for virtually every part of siRNA molecules, from the termini and backbone to the sugars and bases, with the goal of engineering siRNA to prolong half-life and increase cellular uptake. The most common chemical modification involves modifying the sugar moiety. For example, the incorporation of 2′-fluoro (2′-F), -O methyl, -halogen, -amine, or -deoxy can significantly increase the stability of siRNA in serum. Locked nucleic acid (LNA) has been also applied to modify siRNA. The commonly used LNA contains a methylene bridge connecting the 2′-oxygen with the 4′-carbon of the ribose ring.

kidney injuries [8].

224 RNA Interference

time window for the intervention of siRNA.

and siRNA-modified stem cell therapy.

**2. Current principle of siRNA design**

specific positions in the sense strand, and appropriate length [10].

The biggest obstacle faced by siRNA therapies is the *in vivo* delivery of genetic materials. The systemic delivery of synthetic siRNA has the most medical and commercial potential. This type of delivery, however, remains a major challenge for translating siRNA from the research to the clinic. Overcoming the delivery challenge requires effective siRNA delivery vehicles. The virus-based delivery system, while efficient, may be fatally flawed due to raised safety concerns, such as inducing mutations and triggering immunogenic and inflammatory responses [14]. Therefore, extensive research had been performed to develop efficacious nonviral delivery systems, including direct chemical modification of siRNA (as described above) and/or optimization of delivery materials, such as liposome formulation, nanoparticle conjugation and antibodies that target cellular moieties [14].

To date, studies on synthetic siRNA therapy have been performed in a variety of cell culture and rodent models [15] that produced exciting results and were cost effective but failed to faithfully mimic human diseases. Therefore, large animal models, such as porcine models, are indispensable to compensate for the limitations of rodent models due to their greater similarity to human beings. The investigations on siRNA conducted in our laboratory have reflected this trend in the field [7, 12, 16].

#### **3.1. Direct delivery of synthetic siRNA** *in vitro***/***ex vivo*

The siRNA could be easily transduced into various cells for scientific research. For example, we transfected synthetic caspase-3 siRNA in porcine proximal tubular cells (LLC-PK1) using cationic lipid-based transfection reagent. The caspase-3 siRNA inhibited apoptosis and inflammation in LLC-PK1 cells that were subjected to hydrogen peroxide stimulation [17]. In addition to *in vitro* delivery of siRNA, *ex vivo*/*in vivo* siRNA delivery to target organs is an indispensable step before its clinical application. If it was directly delivered into the kidneys, siRNA could obtain higher local concentrations, which would result in improved gene silencing efficacy. During kidney transplantation, *ex vivo* local delivery of siRNA into the donor kidney is feasible because it could be facilitated by the unique structure of the kidney and the characteristics of kidney transplantation. We utilized an *ex vivo* isolated porcine kidney reperfusion system to assess the efficacy of naked caspase-3 siRNA. The caspase-3 siRNA was directly infused into the renal artery (locally) and autologous blood perfusate (mimic systemic delivery) before 24-h cold storage (CS), followed by a further reperfusion for 3 h. The results demonstrated that the caspase-3 siRNA improved ischemia reperfusion (IR) injury with reduced caspase-3 expression and apoptosis, better renal oxygenation, and acid–base homeo‐ stasis [16]. These promising proof-of-principle observations provide valuable guidance for further development before siRNA used in clinical practice.

#### **3.2. Local or systemic delivery of siRNA** *in vivo*

Delivery of siRNA via *ex vivo* route can be applied in donor kidneys, but most renal diseases need *in vivo* delivery. Based on the anatomical and physiological characteristics of the kidney, local delivery can be achieved through several routes: (1) renal artery, first targeting the glomeruli or tubules [18, 19]; (2) renal vein, predominately targeting tubulointerstitium [20]; (3) intraureteral, administered into the renal pelvis and interstitium [21]; and (4) subcapsular administration, achieves intraparenchymal silencing [22]. Due to the rich blood flow through the glomeruli, siRNA injection via the renal artery followed by electroporation could silence specific genes in the glomeruli, such as TGF-β1, which subsequently ameliorates matrix expansion in an experimental glomerulonephritis model [18].

We then used naked caspase-3 siRNA in a porcine kidney autotransplant model for the first time. The left kidney was retrieved from mini pigs and was infused with University of Wisconsin solution, with or without 0.3 mg of naked caspase-3 siRNA, via the renal artery, which was followed by renal artery and renal vein clamping for 24-h cold storage (CS, mimicking donor kidney preservation before transportation in clinic). After right nephrecto‐ my, the left kidney was autotransplanted into the right nephridial pit for 48 h without systemic siRNA treatment (Figure 1). The expression of caspase-3 mRNA and active caspase-3 protein, as well as its precursor, was downregulated by siRNA in the post-CS kidney. In the siRNA preserved posttransplant kidney, however, caspase-3 precursor was further decreased while caspase-3 mRNA, and its activated subunits were upregulated, which resulted in increased apoptosis and inflammation. This study indicated that the naked caspase-3 siRNA was effective for cold preservation but was not effective at protecting posttransplant kidneys, which may be due to systemic compensative responses overcoming local effects. Therefore, to overcome the systemic response and to prolong the therapeutic time window, we subsequently utilized a novel, serum-stable caspase-3 siRNA, both locally as before and systemically via a pretransplantation intravenous injection, and observed the animals for up to 2 weeks post‐ transplantation. The effectiveness of the novel caspase-3 siRNA was confirmed by downre‐ gulated caspase-3 mRNA and protein in the post-CS and/or posttransplant kidneys, as well as reduced apoptosis and inflammation. More importantly, renal function, associated with active caspase-3, HMGB1, apoptosis, inflammation, and tubulointerstitial damage, was improved by this novel, serum-stable caspase-3 siRNA [12].

It has also been revealed that an injection of a single-dose Fas siRNA through the renal vein post ischemia provided a survival advantage in a murine IR model, which was due to the antiapoptosis and antiinflammation effects of the Fas siRNA [20]. Unilateral ureteral obstruc‐ tion (UUO) is a well-established model for tubulointerstitial fibrosis. Xia et al. injected the siRNA of heat shock protein 47 once via the ureter at the time of UUO preparation, leading to significantly reduced fibrosis-related protein expression and a remarkable alleviation of the accompanying interstitial fibrosis [21]. Subcapsular administration is still used in some experiments due to its unique advantages, although it requires an invasive procedure and has

#### siRNA-Induced RNAi Therapy in Acute Kidney Injury http://dx.doi.org/10.5772/61838 227

stasis [16]. These promising proof-of-principle observations provide valuable guidance for

Delivery of siRNA via *ex vivo* route can be applied in donor kidneys, but most renal diseases need *in vivo* delivery. Based on the anatomical and physiological characteristics of the kidney, local delivery can be achieved through several routes: (1) renal artery, first targeting the glomeruli or tubules [18, 19]; (2) renal vein, predominately targeting tubulointerstitium [20]; (3) intraureteral, administered into the renal pelvis and interstitium [21]; and (4) subcapsular administration, achieves intraparenchymal silencing [22]. Due to the rich blood flow through the glomeruli, siRNA injection via the renal artery followed by electroporation could silence specific genes in the glomeruli, such as TGF-β1, which subsequently ameliorates matrix

We then used naked caspase-3 siRNA in a porcine kidney autotransplant model for the first time. The left kidney was retrieved from mini pigs and was infused with University of Wisconsin solution, with or without 0.3 mg of naked caspase-3 siRNA, via the renal artery, which was followed by renal artery and renal vein clamping for 24-h cold storage (CS, mimicking donor kidney preservation before transportation in clinic). After right nephrecto‐ my, the left kidney was autotransplanted into the right nephridial pit for 48 h without systemic siRNA treatment (Figure 1). The expression of caspase-3 mRNA and active caspase-3 protein, as well as its precursor, was downregulated by siRNA in the post-CS kidney. In the siRNA preserved posttransplant kidney, however, caspase-3 precursor was further decreased while caspase-3 mRNA, and its activated subunits were upregulated, which resulted in increased apoptosis and inflammation. This study indicated that the naked caspase-3 siRNA was effective for cold preservation but was not effective at protecting posttransplant kidneys, which may be due to systemic compensative responses overcoming local effects. Therefore, to overcome the systemic response and to prolong the therapeutic time window, we subsequently utilized a novel, serum-stable caspase-3 siRNA, both locally as before and systemically via a pretransplantation intravenous injection, and observed the animals for up to 2 weeks post‐ transplantation. The effectiveness of the novel caspase-3 siRNA was confirmed by downre‐ gulated caspase-3 mRNA and protein in the post-CS and/or posttransplant kidneys, as well as reduced apoptosis and inflammation. More importantly, renal function, associated with active caspase-3, HMGB1, apoptosis, inflammation, and tubulointerstitial damage, was improved by

It has also been revealed that an injection of a single-dose Fas siRNA through the renal vein post ischemia provided a survival advantage in a murine IR model, which was due to the antiapoptosis and antiinflammation effects of the Fas siRNA [20]. Unilateral ureteral obstruc‐ tion (UUO) is a well-established model for tubulointerstitial fibrosis. Xia et al. injected the siRNA of heat shock protein 47 once via the ureter at the time of UUO preparation, leading to significantly reduced fibrosis-related protein expression and a remarkable alleviation of the accompanying interstitial fibrosis [21]. Subcapsular administration is still used in some experiments due to its unique advantages, although it requires an invasive procedure and has

further development before siRNA used in clinical practice.

expansion in an experimental glomerulonephritis model [18].

**3.2. Local or systemic delivery of siRNA** *in vivo*

226 RNA Interference

this novel, serum-stable caspase-3 siRNA [12].

**Figure 1.** Schematic drawing showed a series of our studies using caspase-3 siRNA. The caspase-3 siRNA was first used to protect porcine renal tubular epithelia cells against hydrogen peroxide-induced injury [17]. The renoprotection of naked caspase-3 siRNA with the same sequences was further validated in a porcine *ex vivo* isolated reperfusion model and showed that the siRNA was effective for cold preservation [16], but not in autotransplanted kidneys with‐ out systematic siRNA treatment [7, 12, 16]. Finally, the chemically modified siRNA of caspase-3 via locked nucleic acid stabilized the siRNA in serum and significantly protected autotransplanted kidneys [7, 12, 16].

limitations in clinical practice. Cuevas et al. reported that an infusion of DJ-1 (an antioxidant) specific siRNA into the subcapsule silenced DJ-1 expression in the renal cortex and increased ROS production [22].

Systemic delivery is a common and convenient clinical practice, although current clinical trials using siRNAs are almost directly administered to the target site, such as the nostril, eye, and lung, thereby avoiding the complexity of systemic delivery [23]. The most common method of systemic siRNA delivery is a hydrodynamic intravenous injection with hydraulic pressure to assist siRNA cell entry. However, the pharmacokinetic metabolism of siRNA is more compli‐ cated during systemic delivery because siRNAs can be rapidly degraded by nucleases in the serum and cleared by the kidney and liver. To enhance the *in vivo* efficacy of siRNA treatment, a variety of approaches have been attempted for both siRNA itself and delivery techniques [22–24], as mentioned above.

Due to its anatomical and physiological characteristics, the kidney is the most preferable target organ of systemic siRNA administration. siRNA access to the kidney is thought to be depend‐ ent on the filtration and reabsorption functions of the kidney. Proximal tubule cells (PTCs) are the primary site for rapid and extensive endocytic uptake of siRNA within the kidney following glomerular filtration. In an AKI model, naked synthetic siRNA targeting p53 that was intravenously injected 4 h after renal ischemic injury significantly reduced upregulated p53 expression and protected both the PTCs and kidneys [25]. In another study performed by Zheng et al., siRNA was systematically injected to target complement 3 (C3) and caspase-3 in a murine renal IR injury model. The results showed that the level of serum creatinine and blood urea nitrogen was significantly decreased in the siRNA-treated mice [26]. As most of AKI may be not associated with renal surgery, the systemic siRNA delivery might be a desirable approach. However, for kidney transplantation, in which IR injury is inevitable, local siRNA delivery via any above-mentioned method is feasible and more effective.

#### **3.3. Cell-specific delivery**

As proposed by precision medicine, individual person should receive customized healthcare including diagnosis and intervention. The dysfunctional cells are the true targets for siRNA delivery. For instance, it is known that p53 in PTCs promotes AKI, whereas p53 in other tubular cells does not [27]. It is also expected that apoptosis-inducing siRNA should be directly delivered into tumor cells rather than the surrounding normal cells. Therefore, the cell-specific delivery method is our key point in the next generation of siRNA development.

Recently, antibody conjugation technology has made tumor-targeting drug delivery systems available. The conjugate can be regarded as a "guided molecular missile" that specifically targets unique antigens [28]. Inspired by cancer therapy strategies, siRNAs have also been "packed" to be delivered to target organs, even cells. Recently, a type of asymmetric liposome particle (ALP) has been developed, which highly efficiently encapsulates siRNA without nonspecific cell penetration. The ALPs protected siRNA from ribonuclease degradation. ALPs without any surface modification elicited almost no uptake into cells, while the polyarginine peptide surface-modified ALPs induced nonspecific cell penetration [29]. Leus et al. delivered siRNA targeting vascular cell adhesion molecule-1 (VCAM-1) into inflammation-activated endothelium using anti-VCAM-1-SAINTPEGarg formulated with additional 2 mol% DOPE-PEG2000 *in vivo*. The antibody recognizes VACM-1, which can create specificity for inflamma‐ tion activated endothelial cells. The siRNA homed to VCAM-1 protein expressing vasculature in TNF-α-treated mice without any kidney and liver toxicity [30]. These results represent great progress in siRNA delivery system development. Antibody-mediated specific recognition rather than virus-mediated recognition may be a mainstream in the future.

#### **3.4. Allele-specific RNAi**

RNAi, in addition, discriminates between two sequences only differing by one nucleotide conferring a high specificity of RNAi for its target mRNA. This property was used to develop a particular therapeutic strategy called "allele-specific RNAi" devoted to silence the mutated allele of genes causing dominant inherited diseases without affecting the normal allele. Therapeutic benefit was now demonstrated in cells from patients and animal models, and promising results of the first phase Ib clinical trial using siRNA-based allele-specific therapy were reported in pachyonychia congenita, an inherited skin disorder due to dominant mutations in the *keratin 6* gene [31]. The allele-specific siRNA silencing of the mutant *keratin 12* allele was also applied in corneal limbal epithelial cells grown from patients with Mees‐ mann's epithelial corneal dystrophy [32, 33]. It has also been shown that modified siRNAs conferring allele-specific silencing against disease-causing ALK2 mutants found in fibrodys‐ plasia ossificans progressiva, without affecting normal ALK2 allele [34].

#### **3.5. Delivery of siRNA using a cargo system**

the primary site for rapid and extensive endocytic uptake of siRNA within the kidney following glomerular filtration. In an AKI model, naked synthetic siRNA targeting p53 that was intravenously injected 4 h after renal ischemic injury significantly reduced upregulated p53 expression and protected both the PTCs and kidneys [25]. In another study performed by Zheng et al., siRNA was systematically injected to target complement 3 (C3) and caspase-3 in a murine renal IR injury model. The results showed that the level of serum creatinine and blood urea nitrogen was significantly decreased in the siRNA-treated mice [26]. As most of AKI may be not associated with renal surgery, the systemic siRNA delivery might be a desirable approach. However, for kidney transplantation, in which IR injury is inevitable, local siRNA

As proposed by precision medicine, individual person should receive customized healthcare including diagnosis and intervention. The dysfunctional cells are the true targets for siRNA delivery. For instance, it is known that p53 in PTCs promotes AKI, whereas p53 in other tubular cells does not [27]. It is also expected that apoptosis-inducing siRNA should be directly delivered into tumor cells rather than the surrounding normal cells. Therefore, the cell-specific

Recently, antibody conjugation technology has made tumor-targeting drug delivery systems available. The conjugate can be regarded as a "guided molecular missile" that specifically targets unique antigens [28]. Inspired by cancer therapy strategies, siRNAs have also been "packed" to be delivered to target organs, even cells. Recently, a type of asymmetric liposome particle (ALP) has been developed, which highly efficiently encapsulates siRNA without nonspecific cell penetration. The ALPs protected siRNA from ribonuclease degradation. ALPs without any surface modification elicited almost no uptake into cells, while the polyarginine peptide surface-modified ALPs induced nonspecific cell penetration [29]. Leus et al. delivered siRNA targeting vascular cell adhesion molecule-1 (VCAM-1) into inflammation-activated endothelium using anti-VCAM-1-SAINTPEGarg formulated with additional 2 mol% DOPE-PEG2000 *in vivo*. The antibody recognizes VACM-1, which can create specificity for inflamma‐ tion activated endothelial cells. The siRNA homed to VCAM-1 protein expressing vasculature in TNF-α-treated mice without any kidney and liver toxicity [30]. These results represent great progress in siRNA delivery system development. Antibody-mediated specific recognition

RNAi, in addition, discriminates between two sequences only differing by one nucleotide conferring a high specificity of RNAi for its target mRNA. This property was used to develop a particular therapeutic strategy called "allele-specific RNAi" devoted to silence the mutated allele of genes causing dominant inherited diseases without affecting the normal allele. Therapeutic benefit was now demonstrated in cells from patients and animal models, and promising results of the first phase Ib clinical trial using siRNA-based allele-specific therapy were reported in pachyonychia congenita, an inherited skin disorder due to dominant

delivery via any above-mentioned method is feasible and more effective.

delivery method is our key point in the next generation of siRNA development.

rather than virus-mediated recognition may be a mainstream in the future.

**3.3. Cell-specific delivery**

228 RNA Interference

**3.4. Allele-specific RNAi**

Although lentivirus vectors as vehicles together with liposome reagents are widely applied in the transduction of siRNA, nanoparticle systems have emerged in last few years as an alternative carrier for advanced diagnostic and therapeutic applications. The nanotechnology offers many merits and overcomes the range of challenges/barriers summarized in the previous section, such as the bioavailability and biodistribution of therapeutic agents. Recent reports have demonstrated that the kidney, the glomerulus especially, is a readily accessible site for nanoparticles. Zuckerman at al. intravenously administered nanoparticles containing polycationic cyclodextrin and siRNA/CDP-NPs, most of which deposited in the glomerular mesangial areas. Furthermore, the cultured mouse and human mesangial cells could rapidly internalize siRNA/CDP-NPs. This process could be accelerated by attaching targeting ligand mannose or transferrin to the nanoparticle surface [35].

Complex nanoparticles, especially cationic polyplexes/lipoplexes and liposomes, dominat‐ ed the scene in the early days of RNAi therapeutic development. Their main advantage lies in their endosomal release activity and their ability to concentrate multiple RNAi triggers in one particle [36]. Forbes and Peppas cross-linked polycationic nanoparticle formula‐ tions using ARGET ATRP or UV-initiated polymerization. The advantage of this method is the one-step, one-pot, and surfactant-stabilized monomer-in-water synthesis, which is simpler and faster compared with traditional complicated multistep techniques involving toxic organic solvents [37].

Regardless of how much each mechanism plays in the transport of the drug, cell entry remains a focus for drug design and discovery. An exciting and relatively new approach to transporting pharmaceutical agents into cells is making use of cell-penetrating peptides (CPPs). CPPs are relatively short peptides, typically less than 30 amino acids, and could be vectors for the delivery of genetic and biologic products. CPPs provide a safe, efficient, and noninvasive mode of transport for various cargos into cells [38]. Recently, van Asbeck et al. discovered that CPP/ siRNA complexes with the most negative zeta-potentials in serum were the most resistant to siRNA release over a 20-h incubation period compared to less negatively charged complexes [39]. They also found that the zeta-potential of CPP/siRNA complexes in serum did not correlate with improved cellular association, which might demonstrate the importance of serum proteins or CPP conformation on the ability of CPP/siRNA complexes to associate with the cell membrane. Huang et al. designed a bifunctional peptide named RGD10-10R, by which siRNA was delivered *in vitro* and *in vivo*. Because of their electrostatic interactions with polyarginine (10R), negatively charged siRNAs were readily complexed with RGD10-10R peptides, forming spherical RGD10-10R/siRNA nanoparticles. This is also a novel siRNA delivery tool [40].

Gemini or dimeric lipids (GCLs) are a recent type of amphiphilic molecules that contain two polar headgroups linked by a rigid or flexible spacer that may be hydrophobic or hydrophilic. As each headgroup has a hydrophobic moiety, GCLs may be considered as two conventional monomeric surfactants connected by a spacer group [41]. GCLs have been proved as promising candidates to transfect nucleic acids in gene therapy. The molecular structure of the GCLs offers a high number of alternatives to develop and to improve their capability as transfecting agents.

#### **4. siRNA therapy in AKI**

To date, siRNA therapy has been successfully applied in a variety of acute kidney injuries. IR injury is the primary cause of AKI, particularly during kidney transplantation, in which the kidney is exposed to hypoxia and experiences a series of oxidative, inflammatory and apoptotic responses [42, 43]. Consequently, specific siRNAs targeting critical molecules that are involved in the processes of oxidation, inflammation, and apoptosis have been developed.

Caspase-3, which mediates apoptosis and inflammation, is upregulated by IR injury. Multiple pharmacological interventions against caspase-3, including enzyme inhibitors and genetic modification, have been investigated. In recent years, our group studied the delivery and efficacy of caspase-3 siRNA in *in vitro*, *ex vivo*, and *in vivo* kidney injury models. The synthetic caspase-3 siRNA was initially tested in porcine PTCs (LLC-PK1), with or without hydrogen peroxide (H2O2) stimulation. Apoptotic cells and activated IL-1β protein expression were significantly reduced by the caspase-3 siRNA, with improved cell viability [17]. This outcome led to siRNA application in an isolated organ perfusion system, as described above, and the efficacy of caspase-3 siRNA was further proven, in terms of silenced caspase-3 mRNA and protein expression, attenuated inflammation and apoptosis, and improved renal function and histology [16].

The porcine kidney preserved by caspase-3 siRNA was then autotransplanted in a 2-day model. However, the transplanted kidney was not protected without systemic treatment of the recipient. Moreover, new serum-stabilized caspase-3 siRNAs were applied locally in kidney preservation and intravenously in recipient in a 2-week autotransplant model. The transplanted kidneys were protected without significant off-target effects. These serials of stepby-step studies provided promising evidence to support siRNA treatment to be further applied in clinic.

p53, another pivotal protein in the apoptotic pathway, has been identified as a mediator of transcriptional responses to IR injury [44]. Molitoris et al. revealed that intravenously injected p53 siRNA attenuated ischemic and cisplatin-induced AKI [25]. Fujino et al. also tested the efficacy siRNA targeting p53 via transarterial administration siRNA injected into the left renal artery immediately after ischemia improved tubular injury and downregulated GSK-3β expression [45]. In a diabetic mouse model, p53 inhibition by siRNA also reduced ischemic AKI [46].

Silencing of other important transcription factors or immunity related receptors using siRNAs have also been studied. Renal IR injury and inflammation are related to postsurgical healing and both processes can be influenced by toll-like receptor (TLR) signals. Effective TLR9 silencing by siRNA decreases renal cell apoptosis, mitigates AKI severity, and increases the mice survival [47]. NF-κB, a pro-inflammatory transcription factor induced by TLR and other signals, plays a key role in AKI. NF-κB activation depends on the activation of the inhibitor of κB kinase β (IKKβ). Wan et al. demonstrated that silencing IKKβ using siRNA diminished inflammation and protected the kidneys against IR injury [19]*.* These studies clearly demon‐ strate the therapeutic potential of siRNA-induced silencing of key AKI mediators, which are activated and involved in the pathways of apoptosis, inflammation, immunity, etc.

#### **5. Off-target side effects and toxicities of siRNA**

The siRNA has been likened to a "magic bullet" due to this potency and specificity, but offtarget side effects and toxicities create additional challenges for researchers. The induction of various side effects may be caused by unexpected perturbations between RNAi molecules and cellular components. The off-target effects of siRNA were first reported by Jackson and colleagues in 2003 [48]. Broadly speaking, off-target effects can be siRNA specific or nonspe‐ cific. The former are caused by limited siRNA complementarity to nontargeted mRNAs. The latter, resulting in immune- and toxicity-related responses, are due to the construction of the siRNA sequence, its modification, or the delivery vehicle.

The off-target effects associated with siRNA delivery fall into three broad categories: (1) miRNA-like off-target effects, referring to siRNA-induced sequence-dependent regulation of unintended transcripts through partial sequence complementarity to their 3′UTRs; (2) inflammatory responses through the activation of TLR triggered by siRNAs and/or delivery vehicles (such as cationic lipids and viruses); and (3) widespread effects on miRNA processing and function through the saturation of the endogenous RNAi machinery by exogenous siRNAs [49, 50].

#### **5.1. miRNA-like off-target effects**

Gemini or dimeric lipids (GCLs) are a recent type of amphiphilic molecules that contain two polar headgroups linked by a rigid or flexible spacer that may be hydrophobic or hydrophilic. As each headgroup has a hydrophobic moiety, GCLs may be considered as two conventional monomeric surfactants connected by a spacer group [41]. GCLs have been proved as promising candidates to transfect nucleic acids in gene therapy. The molecular structure of the GCLs offers a high number of alternatives to develop and to improve their capability as transfecting

To date, siRNA therapy has been successfully applied in a variety of acute kidney injuries. IR injury is the primary cause of AKI, particularly during kidney transplantation, in which the kidney is exposed to hypoxia and experiences a series of oxidative, inflammatory and apoptotic responses [42, 43]. Consequently, specific siRNAs targeting critical molecules that are involved

Caspase-3, which mediates apoptosis and inflammation, is upregulated by IR injury. Multiple pharmacological interventions against caspase-3, including enzyme inhibitors and genetic modification, have been investigated. In recent years, our group studied the delivery and efficacy of caspase-3 siRNA in *in vitro*, *ex vivo*, and *in vivo* kidney injury models. The synthetic caspase-3 siRNA was initially tested in porcine PTCs (LLC-PK1), with or without hydrogen peroxide (H2O2) stimulation. Apoptotic cells and activated IL-1β protein expression were significantly reduced by the caspase-3 siRNA, with improved cell viability [17]. This outcome led to siRNA application in an isolated organ perfusion system, as described above, and the efficacy of caspase-3 siRNA was further proven, in terms of silenced caspase-3 mRNA and protein expression, attenuated inflammation and apoptosis, and improved renal function and

The porcine kidney preserved by caspase-3 siRNA was then autotransplanted in a 2-day model. However, the transplanted kidney was not protected without systemic treatment of the recipient. Moreover, new serum-stabilized caspase-3 siRNAs were applied locally in kidney preservation and intravenously in recipient in a 2-week autotransplant model. The transplanted kidneys were protected without significant off-target effects. These serials of stepby-step studies provided promising evidence to support siRNA treatment to be further applied

p53, another pivotal protein in the apoptotic pathway, has been identified as a mediator of transcriptional responses to IR injury [44]. Molitoris et al. revealed that intravenously injected p53 siRNA attenuated ischemic and cisplatin-induced AKI [25]. Fujino et al. also tested the efficacy siRNA targeting p53 via transarterial administration siRNA injected into the left renal artery immediately after ischemia improved tubular injury and downregulated GSK-3β expression [45]. In a diabetic mouse model, p53 inhibition by siRNA also reduced ischemic

in the processes of oxidation, inflammation, and apoptosis have been developed.

agents.

230 RNA Interference

histology [16].

in clinic.

AKI [46].

**4. siRNA therapy in AKI**

The siRNAs and miRNAs share similar machinery downstream of their initial processing. Using several different siRNAs targeting the same gene, microarray profiling showed that each siRNA produced a unique, sequence-dependent signature. Sequence analysis of off-target transcripts revealed that the 3′ UTR regions of these transcripts were complementary to the 5′ end of the transfected siRNA guide strand [48]. It is now understood that for the off-targeting effects to occur, a perfect complementarity between the seed region of the antisense strand such as nucleotide positions 2–7 or 2–8 and the 3′ UTR of the transcript is necessary [49, 51]. Silencing the set of original off-target transcripts could be induced by base mismatches in the 5′ end of siRNA guide strands. However, a new set of off-target transcripts within 3′ UTRs that were complementary to the mismatched guide strand could be generated [49].

RNAi regulation by miRNAs involves partial complementarity between the targeting RNA and miRNA. Because miRNAs cause gene silencing through mRNA degradation and trans‐ lation inhibition, the siRNA-mediated off-target effects may also be acting at two levels. For this reason, there should be greater emphasis on improving siRNA design as well as moni‐ toring gene and protein levels following RNAi therapy to account for any off-target effects.

#### **5.2. Recognition and stimulation of the innate immune system**

The recognition and stimulation of the immune system arenonspecific off-target effects of siRNA therapy. The RNA-sensing pattern recognition receptors (PRRs), localized in endo‐ somes, are the most important components of the innate immune system. The responses of PRRs to siRNAs are either TLR-mediated or non-TLR-mediated. The PRR responses are also associated with siRNA sequence-specific side effects and have recently attracted many attentions from researchers [52]. RNA-sensing TLRs (TLR3 and TLR7) are predominantly located intracellularly and recognize nucleic acids released from invading pathogens. The non-TLR-mediated innate immune responses triggered by siRNA binding are linked to RNAregulated expression of protein kinase (PKR) and retinoic acid inducible gene 1 (RIG1), which further induce caspase-3 and NF-κB expression, respectively. The activation of PRRs generates excessive cytokine release and subsequent inflammation [53].

Based on this second type of off-target RNAi effects, our group further investigated the mechanism of how short-acting caspase-3 siRNA impaired posttransplanted kidneys. The results suggested that the amplified inflammatory responses in caspase-3 siRNA preserved autotransplant kidneys were associated with TLR3, TLR7, and PKR activation, which may be due to systemic compensative responses, although persistent actions initiated by short-acting caspase-3 siRNA cannot be completely excluded [54]. Other studies have also indicated that the horseshoe-like structure of TLR3 facilitates dsRNA recognition [55, 56]. Interactions between TLR3 and dsRNA were originally reported in 2001 when TLR3-deficient mice exhibited reduced immune responses to dsRNA viruses [57].

Several studies have demonstrated that the immune response to siRNAs is cell type-dependent due to the selective expression of TLRs. siRNAs stimulate monocytes and myeloid dendritic cells through TLR8 to produce proinflammatory cytokines, or activate plasmacytoid dendritic cells through TLR7 to produce type I interferons [58–60]. In addition, the volume of hydrody‐ namic naked siRNA delivery influences immune activation. Rácz et al. compared the immune responses induced by 50 μg siRNA dissolved in either low-volume (1 mL/mouse) or highvolume (10% of body weight, 2.5 mL/mouse in average) physiological salt solution delivered *in vivo*. Low-volume hydrodynamic injection induced slight alanine aminotransferase (ALT) elevation and mild hepatocyte injury, whereas high-volume hydrodynamic injection resulted in higher ALT levels and extensive hepatocyte necrosis. High-volume hydrodynamic injection also led to a time-dependent slight increase in IFN-related gene expression [61]. Collectively, these studies suggest that there is a need for improving siRNA design, establishing experi‐ mental controls and carefully interpreting results.

#### **6. From bench to bedside: Clinical trials**

RNAi regulation by miRNAs involves partial complementarity between the targeting RNA and miRNA. Because miRNAs cause gene silencing through mRNA degradation and trans‐ lation inhibition, the siRNA-mediated off-target effects may also be acting at two levels. For this reason, there should be greater emphasis on improving siRNA design as well as moni‐ toring gene and protein levels following RNAi therapy to account for any off-target effects.

The recognition and stimulation of the immune system arenonspecific off-target effects of siRNA therapy. The RNA-sensing pattern recognition receptors (PRRs), localized in endo‐ somes, are the most important components of the innate immune system. The responses of PRRs to siRNAs are either TLR-mediated or non-TLR-mediated. The PRR responses are also associated with siRNA sequence-specific side effects and have recently attracted many attentions from researchers [52]. RNA-sensing TLRs (TLR3 and TLR7) are predominantly located intracellularly and recognize nucleic acids released from invading pathogens. The non-TLR-mediated innate immune responses triggered by siRNA binding are linked to RNAregulated expression of protein kinase (PKR) and retinoic acid inducible gene 1 (RIG1), which further induce caspase-3 and NF-κB expression, respectively. The activation of PRRs generates

Based on this second type of off-target RNAi effects, our group further investigated the mechanism of how short-acting caspase-3 siRNA impaired posttransplanted kidneys. The results suggested that the amplified inflammatory responses in caspase-3 siRNA preserved autotransplant kidneys were associated with TLR3, TLR7, and PKR activation, which may be due to systemic compensative responses, although persistent actions initiated by short-acting caspase-3 siRNA cannot be completely excluded [54]. Other studies have also indicated that the horseshoe-like structure of TLR3 facilitates dsRNA recognition [55, 56]. Interactions between TLR3 and dsRNA were originally reported in 2001 when TLR3-deficient mice

Several studies have demonstrated that the immune response to siRNAs is cell type-dependent due to the selective expression of TLRs. siRNAs stimulate monocytes and myeloid dendritic cells through TLR8 to produce proinflammatory cytokines, or activate plasmacytoid dendritic cells through TLR7 to produce type I interferons [58–60]. In addition, the volume of hydrody‐ namic naked siRNA delivery influences immune activation. Rácz et al. compared the immune responses induced by 50 μg siRNA dissolved in either low-volume (1 mL/mouse) or highvolume (10% of body weight, 2.5 mL/mouse in average) physiological salt solution delivered *in vivo*. Low-volume hydrodynamic injection induced slight alanine aminotransferase (ALT) elevation and mild hepatocyte injury, whereas high-volume hydrodynamic injection resulted in higher ALT levels and extensive hepatocyte necrosis. High-volume hydrodynamic injection also led to a time-dependent slight increase in IFN-related gene expression [61]. Collectively, these studies suggest that there is a need for improving siRNA design, establishing experi‐

**5.2. Recognition and stimulation of the innate immune system**

232 RNA Interference

excessive cytokine release and subsequent inflammation [53].

exhibited reduced immune responses to dsRNA viruses [57].

mental controls and carefully interpreting results.

The numbers of RNAi-based preclinical studies and clinical trials have grown over the past several years. To date, there have been 27 registered clinical trials using siRNA worldwide. These studies include retinal degeneration, dominantly inherited brain and skin diseases, viral infections, respiratory disorders, metabolic diseases, and of particular note, kidney diseases. In 2011, Quark Pharmaceuticals completed a phase I, randomized, double-blind, dose escalation, safety, and pharmacokinetic study (NCT00554359) on QPI-1002, also designated I5NP, which was a synthetic siRNA that temporarily inhibits p53 expression that is in early development for acute kidney failure therapy. I5NP is the first siRNA to be systemically administered in humans. Based on the preclinical data obtained from animal models, the siRNA was intravenously injected within 4 h to bypass surgery patients. Pharmacokinetic data were collected during the first 24 h, and safety and dose-limiting toxicities were monitored until hospital discharge and 6–12 months after surgery. Recently, Quark initiated a subsequent clinical trial to determine whether a single administration of I5NP can prevent delayed graft function in kidney transplant recipients. Data from this study will be used to identify I5NP doses for follow-on efficacy studies (NCT00802347). Another ongoing phase I trial investigat‐ ing solid tumors, including Renal cell carcinoma (RCC), was conducted by Calando Pharma‐ ceuticals. The investigators used CALAA-01, whose active ingredient is a type of siRNA, to inhibit tumor growth and/or reduce tumor size. This siRNA inhibits the expression of the M2 subunit of ribonucleotide reductase and resists nuclease degradation by using a stabilized nanoparticle that targets tumor cells (NCT00689065).Besides, there is an ongoing study, in which patients with melanoma, kidney cancer, pancreatic cancer, or other solid tumors that are metastatic or cannot be removed by surgeryare treated by APN401, siRNA-transfected autologousperipheral blood mononuclear cells.These cells were modified by siRNA targeting factors inhibiting the killing ability of immune cells in vitro and transfused back into the body, in order to kill more cancer cells (NCT02166255, the above clinical trials can be found at ClinicalTrials.gov, Table 1).


**Table 1.** Clinical trials of siRNA therapy in kidney diseases.

#### **7. Perspectives and challenges**

Despite the enormous potential advantages of siRNA therapy, additional research must be performed before its large-scale clinical application.

#### **7.1. Target gene selection**

Genome-wide or pathway-specific siRNA libraries have become available using highthroughput screening approaches. Establishing *in vitro* prescreening leads to signaling pathway prediction and target validation in *in vivo* renal disease. However, choosing one or a set of reasonable target genes is the key for designing specific siRNA treatments. The patho‐ physiological changes during kidney disease, like any other disease, refer to a complex gene and protein regulation network. For example, the network that exists during kidney trans‐ plantation involves the original conditions of the donors and the interactions between the donor kidneys and the recipients, which could direct the progression, as well as the recovery, of the injury. Fortunately, transcriptome measurements of the transplanted kidney may provide a comprehensive understanding of gene regulation and would be beneficial for target gene selection.

Mueller et al. analyzed the transcriptome of postreperfusion implant biopsies in living donors (LD) and deceased donors (DD). Hundreds of mRNAs were identified that predicted delayed graft function [62]. In a recent prospective study using human posttransplant kidney biopsies, 20 mRNAs and two miRNAs were identified as molecular signatures of AKI. Elevated secretory leukocyte peptidase inhibitor in AKI allografts was validated and miR-182-5p was identified as a molecular regulator [63]. These genes could be used as potential targets of siRNA therapy. We recently identified 3 times more differentially expressed genes in renal allograft biopsies between living donors and cadaveric donors at 30 min than 3 months posttransplan‐ tation. The majority of these differentially expressed genes are responsible for acute responses at 30 min, but also involved in inflammation, nephrotoxicity, and proliferation at 3 months. These divergent transcriptome signatures between two types of donors might be linked with not only the initial injury of the donors, but also the immune responses of the recipients.

Another method for selecting target genes is by identifying their translation product proteins. To find a single or a set of crucial proteins involved in kidney allograft rejection, Wu et al. explored potential transcriptional factors and regulation networks in 352 kidney transplant recipients, of which 85 suffered from acute rejection (AR). The results demonstrated that the dominant processes and responses were associated with inflammation and complement activation in AR. A number of transcription factors were identified in AR patients, including NF-κB, signal transducer, and activator of transcription (STAT) 1 and STAT3 [64]. Their recent study further revealed inflammation-derived kidney allograft injury, such as AR, chronic rejection, and impaired renal function without rejection. Wu et al. 12 common proteins and 11 level-specific proteins from the phenotype-related protein–protein interaction networks [65]. These potential biomarkers also provide valuable targets for siRNA design relating to the treatment of transplant-related injury.

#### **7.2. Timely application**

Compared with shRNA, an advantage of siRNA for AKI therapy is time-controlled, transient treatment. Silencing the target gene for a short time or a long time should be assessed before RNAi application. The silenced genes may be multifunctional according to the surrounding milieu. For example, caspase-3, generally considered an executor in cellular apoptosis, should be inhibited in injured tissues. However, it is also a loyal scavenger in malignantly transformed cells, which could be an unavoidable side effect in any caspase-3-targeting siRNA therapy. For AKI, siRNA ineffectiveness is needed after the therapeutic time window. Additionally, siRNA application avoids intracellular traffic. In certain circumstances, shRNA delivery could be harmful to the organ or even fatal.

A study from Grimm et al. investigated the long-term effects of sustained high-level shRNA expression in the livers of adult mice. An evaluation of 49 distinct adeno-associated virus/ shRNA vectors, with unique lengths and sequences that were directed against six targets, showed that 36 vectors resulted in dose-dependent liver injury, with 23 ultimately causing death. The observed morbidity was associated with the downregulation of liver-derived miRNAs, indicating possible competition of the latter with shRNAs (through saturation of the endogenous RNAi machinery by the exogenous siRNAs) for the limited cellular factors required for the processing of various small RNAs [66]. Therefore, controlling intracellular shRNA expression levels will be imperative, but siRNA would not influence the endogenous process of RNA degradation mediated by miRNAs.

#### **7.3. siRNA targeting single, double or multiple genes**

The knockdown of two or more genes simultaneously using siRNA cocktail has been recently reported. Many applications of siRNA cocktail have demonstrated significant benefits compared with siRNA targeted to a single gene, particularly in anticancer and antiviral therapy [67, 68]. A high concentration of individual siRNAs may represent the key off-target effect in terms of competition for endogenous miRNA biogenesis machinery. Therefore, the other advantage of siRNA cocktail is the relatively low concentration of each siRNA, which may also reduce off-target signatures without sacrificing silencing potency [69].

#### **7.4. Cell-specific siRNA targeting**

**7.1. Target gene selection**

234 RNA Interference

gene selection.

treatment of transplant-related injury.

**7.2. Timely application**

Genome-wide or pathway-specific siRNA libraries have become available using highthroughput screening approaches. Establishing *in vitro* prescreening leads to signaling pathway prediction and target validation in *in vivo* renal disease. However, choosing one or a set of reasonable target genes is the key for designing specific siRNA treatments. The patho‐ physiological changes during kidney disease, like any other disease, refer to a complex gene and protein regulation network. For example, the network that exists during kidney trans‐ plantation involves the original conditions of the donors and the interactions between the donor kidneys and the recipients, which could direct the progression, as well as the recovery, of the injury. Fortunately, transcriptome measurements of the transplanted kidney may provide a comprehensive understanding of gene regulation and would be beneficial for target

Mueller et al. analyzed the transcriptome of postreperfusion implant biopsies in living donors (LD) and deceased donors (DD). Hundreds of mRNAs were identified that predicted delayed graft function [62]. In a recent prospective study using human posttransplant kidney biopsies, 20 mRNAs and two miRNAs were identified as molecular signatures of AKI. Elevated secretory leukocyte peptidase inhibitor in AKI allografts was validated and miR-182-5p was identified as a molecular regulator [63]. These genes could be used as potential targets of siRNA therapy. We recently identified 3 times more differentially expressed genes in renal allograft biopsies between living donors and cadaveric donors at 30 min than 3 months posttransplan‐ tation. The majority of these differentially expressed genes are responsible for acute responses at 30 min, but also involved in inflammation, nephrotoxicity, and proliferation at 3 months. These divergent transcriptome signatures between two types of donors might be linked with not only the initial injury of the donors, but also the immune responses of the recipients.

Another method for selecting target genes is by identifying their translation product proteins. To find a single or a set of crucial proteins involved in kidney allograft rejection, Wu et al. explored potential transcriptional factors and regulation networks in 352 kidney transplant recipients, of which 85 suffered from acute rejection (AR). The results demonstrated that the dominant processes and responses were associated with inflammation and complement activation in AR. A number of transcription factors were identified in AR patients, including NF-κB, signal transducer, and activator of transcription (STAT) 1 and STAT3 [64]. Their recent study further revealed inflammation-derived kidney allograft injury, such as AR, chronic rejection, and impaired renal function without rejection. Wu et al. 12 common proteins and 11 level-specific proteins from the phenotype-related protein–protein interaction networks [65]. These potential biomarkers also provide valuable targets for siRNA design relating to the

Compared with shRNA, an advantage of siRNA for AKI therapy is time-controlled, transient treatment. Silencing the target gene for a short time or a long time should be assessed before RNAi application. The silenced genes may be multifunctional according to the surrounding milieu. For example, caspase-3, generally considered an executor in cellular apoptosis, should

We have showed that the apoptosis of different types of cells leads to different outcomes. For instance, the apoptosis of inflammatory cells is associated with inflammation clearance and tissue remodeling, whereas the apoptosis of renal parenchymal cells is link to tubular atrophy and renal fibrosis. Therefore, using the genetic material such as siRNA targeting specific cell at particular time frame is crucial to achieve high efficacy of treatment in AKI and also avoid site-effects [12, 54].

It is still challenging to administrate siRNA cell specifically, but it is feasible as there were a few studies showed delivering siRNA into liver cells and antigen-presenting cells [70–72] using carbon nanotubes and mannose-conjugated liposomes. In addition, surface pegylation and cell-specific targeting ligands incorporation in the carriers may improve the pharmacokinetics, biodistribution, and siRNA selectivity. Choosing appropriate siRNA carriers has to consider the safety, effectiveness, ease of manufacturing, off-target effects [12], and innate immune responses. Of course, the efficacy of siRNA is still a most important factor dominating the selection of its carriers [54].

#### **7.5. siRNA-modified stem cell therapy**

Mesenchymal stem cell (MSC) transplantation has attracted much attention in cell therapy in different organ systems such as myocardial infarction. One of the limitations is the poor survival of grafted cells in the ischemic microenvironment. To tackle this issue, a novel siRNAmediated prolyl hydroxylase domain protein 2 (PHD2) silencing system has been developed based on arginine-terminated generation 4 poly (amidoamine) nanoparticles. This system, for the first time, exhibited effective and biocompatible siRNA delivery and PHD2 silencing in MSCs *in vitro*. After transplant PHD2 siRNA-modified MSC in myocardial infarction models, MSC survival and paracrine function of IGF-1 were enhanced significantly *in* vivo, with decreased cardiomyocyte apoptosis, scar size, and interstitial fibrosis, and increased angio‐ genesis in the diseased myocardium, which ultimately attenuated ventricular remodeling and improved heart function. This study demonstrated that a great potential of siRNA-modified stem cells in therapeutic applications, which, of course, might be used in AKI [73].

#### **8. Conclusions**

The kidney is a comparatively easy target organ for siRNA therapy due to its unique structural and functional characteristics. siRNA intervention is effective, feasible, and has great potential for fighting against kidney diseases. For the next-generation siRNA development, cell-specific precise delivery should be pursued. Although the safety of siRNA therapy has been proven by rapidly emerging clinical studies, off-target and compensative responses still need be overcome via various modification strategies. The time for realizing the therapeutic potential of RNAi has come because optimized siRNA therapy, in conjunction with advanced genetic screening technologies, could facilitate timely and specific treatment of kidney as well as other organ diseases in the near future.

#### **Author details**

Cheng Yang1,2 and Bin Yang3,4\*

\*Address all correspondence to: by5@le.ac.uk

1 Department of Plastic Surgery, Zhongshan Hospital, Fudan University, Shanghai, China

2 Shanghai Key Laboratory of Organ Transplantation, Shanghai, China

3 Medical Research Centre, Medical School of Nantong University; Department of Nephrology, Affiliated Hospital of Nantong University, Nantong, China, United Kingdom

4 Renal Group, Department of Infection, Immunity and Inflammation, University of Leicester,University Hospitals of Leicester, United Kingdom

#### **References**

**7.5. siRNA-modified stem cell therapy**

**8. Conclusions**

236 RNA Interference

**Author details**

organ diseases in the near future.

Cheng Yang1,2 and Bin Yang3,4\*

\*Address all correspondence to: by5@le.ac.uk

Mesenchymal stem cell (MSC) transplantation has attracted much attention in cell therapy in different organ systems such as myocardial infarction. One of the limitations is the poor survival of grafted cells in the ischemic microenvironment. To tackle this issue, a novel siRNAmediated prolyl hydroxylase domain protein 2 (PHD2) silencing system has been developed based on arginine-terminated generation 4 poly (amidoamine) nanoparticles. This system, for the first time, exhibited effective and biocompatible siRNA delivery and PHD2 silencing in MSCs *in vitro*. After transplant PHD2 siRNA-modified MSC in myocardial infarction models, MSC survival and paracrine function of IGF-1 were enhanced significantly *in* vivo, with decreased cardiomyocyte apoptosis, scar size, and interstitial fibrosis, and increased angio‐ genesis in the diseased myocardium, which ultimately attenuated ventricular remodeling and improved heart function. This study demonstrated that a great potential of siRNA-modified

stem cells in therapeutic applications, which, of course, might be used in AKI [73].

The kidney is a comparatively easy target organ for siRNA therapy due to its unique structural and functional characteristics. siRNA intervention is effective, feasible, and has great potential for fighting against kidney diseases. For the next-generation siRNA development, cell-specific precise delivery should be pursued. Although the safety of siRNA therapy has been proven by rapidly emerging clinical studies, off-target and compensative responses still need be overcome via various modification strategies. The time for realizing the therapeutic potential of RNAi has come because optimized siRNA therapy, in conjunction with advanced genetic screening technologies, could facilitate timely and specific treatment of kidney as well as other

1 Department of Plastic Surgery, Zhongshan Hospital, Fudan University, Shanghai, China

Nephrology, Affiliated Hospital of Nantong University, Nantong, China, United Kingdom

3 Medical Research Centre, Medical School of Nantong University; Department of

4 Renal Group, Department of Infection, Immunity and Inflammation, University of

2 Shanghai Key Laboratory of Organ Transplantation, Shanghai, China

Leicester,University Hospitals of Leicester, United Kingdom


chemic and cisplatin-induced acute kidney injury. *J Am Soc Nephrol* 2009; 20:1754– 1764. DOI:10.1681/ASN.2008111204

[26] Zheng X, Zhang X, Sun H, Feng B, Li M, Chen G, Vladau C, Chen D, Suzuki M, Min L, et al: Protection of renal ischemia injury using combination gene silencing of com‐ plement 3 and caspase 3 genes. *Transplantation* 2006; 82:1781–1786. DOI:10.1097/01.tp. 0000250769.86623.a3

[13] Mook OR, Baas F, de Wissel MB, Fluiter K: Evaluation of locked nucleic acid-modi‐ fied small interfering RNA in vitro and in vivo. *Mol Cancer Ther* 2007; 6:833–843. DOI:

[14] Kanasty R, Dorkin JR, Vegas A, Anderson D: Delivery materials for siRNA therapeu‐

[15] Deng Y, Wang CC, Choy KW, Du Q, Chen J, Wang Q, Li L, Chung TK, Tang T: Ther‐ apeutic potentials of gene silencing by RNA interference: principles, challenges, and

[16] Yang B, Hosgood SA, Nicholson ML: Naked small interfering RNA of caspase-3 in preservation solution and autologous blood perfusate protects isolated ischemic por‐ cine kidneys. *Transplantation* 2011; 91:501–507. DOI:10.1097/TP.0b013e318207949f [17] Yang B, Elias JE, Bloxham M, Nicholson ML: Synthetic small interfering RNA downregulates caspase-3 and affects apoptosis, IL-1 beta, and viability of porcine proximal

[18] Takabatake Y, Isaka Y, Mizui M, Kawachi H, Shimizu F, Ito T, Hori M, Imai E: Ex‐ ploring RNA interference as a therapeutic strategy for renal disease. *Gene Ther* 2005;

[19] Wan X, Fan L, Hu B, Yang J, Li X, Chen X, Cao C: Small interfering RNA targeting IKKbeta prevents renal ischemia-reperfusion injury in rats. *Am J Physiol Renal Physiol*

[20] Hamar P, Song E, Kokeny G, Chen A, Ouyang N, Lieberman J: Small interfering RNA targeting Fas protects mice against renal ischemia-reperfusion injury. *Proc Natl*

[21] Xia Z, Abe K, Furusu A, Miyazaki M, Obata Y, Tabata Y, Koji T, Kohno S: Suppres‐ sion of renal tubulointerstitial fibrosis by small interfering RNA targeting heat shock

[22] Cuevas S, Zhang Y, Yang Y, Escano C, Asico L, Jones JE, Armando I, Jose PA: Role of renal DJ-1 in the pathogenesis of hypertension associated with increased reactive oxygen species production. *Hypertension* 2012; 59:446–452. DOI:10.1161/HYPERTEN‐

[23] Gooding M, Browne LP, Quinteiro FM, Selwood DL: siRNA delivery: from lipids to cell-penetrating peptides and their mimics. *Chem Biol Drug Des* 2012; 80:787–809.

[24] Braasch DA, Jensen S, Liu Y, Kaur K, Arar K, White MA, Corey DR: RNA interfer‐ ence in mammalian cells by chemically-modified RNA. *Biochemistry* 2003; 42:7967–

[25] Molitoris BA, Dagher PC, Sandoval RM, Campos SB, Ashush H, Fridman E, Brafman A, Faerman A, Atkinson SJ, Thompson JD, et al: siRNA targeted to p53 attenuates is‐

*Acad Sci U S A* 2004; 101:14883–14888. DOI:10.1073/pnas.0406421101

protein 47. *Am J Nephrol* 2008; 28:34–46. DOI:10.1159/000108759

new strategies. *Gene* 2014; 538:217–227. DOI:10.1016/j.gene.2013.12.019

tubular cells. *J Cell Biochem* 2011; 112:1337–1347. DOI:10.1002/jcb.23050

10.1158/1535-7163.MCT-06-0195

238 RNA Interference

12:965–973. DOI:10.1038/sj.gt.3302480

SIONAHA.111.185744

DOI:10.1111/cbdd.12052

7975. DOI:10.1021/bi0343774

2011; 300:F857–F863. DOI:10.1152/ajprenal.00547.2010

tics. *Nat Mater* 2013; 12:967–977. DOI:10.1038/nmat3765


[48] Jackson AL, Bartz SR, Schelter J, Kobayashi SV, Burchard J, Mao M, Li B, Cavet G, Linsley PS: Expression profiling reveals off-target gene regulation by RNAi. *Nat Bio‐ technol* 2003; 21:635–637. DOI:10.1038/nbt831

[36] Haussecker D: Current issues of RNAi therapeutics delivery and development. *J Con‐*

[37] Forbes DC, Peppas NA: Polycationic nanoparticles for siRNA delivery: comparing ARGET ATRP and UV-initiated formulations. *ACS Nano* 2014; 8:2908–2917. DOI:

[38] Li H, Tsui TY, Ma W: Intracellular delivery of molecular cargo using cell-penetrating peptides and the combination strategies. *Int J Mol Sci* 2015; 16:19518–19536. DOI:

[39] van Asbeck AH, Beyerle A, McNeill H, Bovee-Geurts PH, Lindberg S, Verdurmen WP, Hallbrink M, Langel U, Heidenreich O, Brock R: Molecular parameters of siRNA —cell penetrating peptide nanocomplexes for efficient cellular delivery. *ACS Nano*

[40] Huang Y, Wang X, Huang W, Cheng Q, Zheng S, Guo S, Cao H, Liang XJ, Du Q, Liang Z: Systemic administration of siRNA via cRGD-containing peptide. *Sci Rep*

[41] Junquera E, Aicart E: Recent progress in gene therapy to deliver nucleic acids with multivalent cationic vectors. *Adv Colloid Interface Sci* 2015. DOI:10.1016/j.cis.

[42] Kosieradzki M, Rowinski W: Ischemia/reperfusion injury in kidney transplantation: mechanisms and prevention. *Transplantation Proceedings* 2008; 40:3279–3288. DOI:

[43] Eltzschig HK, Eckle T: Ischemia and reperfusion-from mechanism to translation. *Na‐*

[44] Vaseva AV, Moll UM: The mitochondrial p53 pathway. *Biochim Biophys Acta* 2009;

[45] Fujino T, Muhib S, Sato N, Hasebe N: Silencing of p53 RNA through transarterial de‐ livery ameliorates renal tubular injury and downregulates GSK-3beta expression af‐ ter ischemia-reperfusion injury. *Am J Physiol Renal Physiol* 2013; 305:F1617–F1627.

[46] Peng J, Li X, Zhang D, Chen JK, Su Y, Smith SB, Dong Z: Hyperglycemia, p53, and mitochondrial pathway of apoptosis are involved in the susceptibility of diabetic models to ischemic acute kidney injury. *Kidney Int* 2015; 87:137–150. DOI:10.1038/ki.

[47] Liu L, Li Y, Hu Z, Su J, Huo Y, Tan B, Wang X, Liu Y: Small interfering RNA target‐ ing Toll-like receptor 9 protects mice against polymicrobial septic acute kidney in‐

jury. *Nephron Exp Nephrol* 2012; 122:51–61. DOI:10.1159/000346953

*trol Release* 2014; 195:49–54. DOI:10.1016/j.jconrel.2014.07.056

10.1021/nn500101c

240 RNA Interference

10.3390/ijms160819518

2015.07.003

2014.226

2013; 7:3797–3807. DOI:10.1021/nn305754c

2015; 5:12458. DOI:10.1038/srep12458

10.1016/j.transproceed.2008.10.004

DOI:10.1152/ajprenal.00279.2013

*ture Medicine* 2011; 17:1391–1401. DOI:10.1038/nm.2507

1787:414–420. DOI:10.1016/j.bbabio.2008.10.005


cal siRNA delivery into melanoma. *Biomaterials* 2014; 35:3435–3442. DOI:10.1016/ j.biomaterials.2013.12.079

[72] Chen D, Koropatnick J, Jiang N, Zheng X, Zhang X, Wang H, Yuan K, Siu KS, Shun‐ nar A, Way C, Min WP: Targeted siRNA silencing of indoleamine 2, 3-dioxygenase in antigen-presenting cells using mannose-conjugated liposomes: a novel strategy for treatment of melanoma. *J Immunother* 2014; 37:123–134. DOI:10.1097/CJI. 0000000000000022

[60] Sioud M: Induction of inflammatory cytokines and interferon responses by doublestranded and single-stranded siRNAs is sequence-dependent and requires endoso‐

[62] Mueller TF, Reeve J, Jhangri GS, Mengel M, Jacaj Z, Cairo L, Obeidat M, Todd G, Moore R, Famulski KS, et al: The transcriptome of the implant biopsy identifies do‐ nor kidneys at increased risk of delayed graft function. *Am J Transplant* 2008; 8:78–85.

[63] Wilflingseder J, Sunzenauer J, Toronyi E, Heinzel A, Kainz A, Mayer B, Perco P, Telkes G, Langer RM, Oberbauer R: Molecular pathogenesis of post-transplant acute kidney injury: assessment of whole-genome mRNA and MiRNA profiles. *PLoS One*

[64] Wu D, Zhu D, Xu M, Rong R, Tang Q, Wang X, Zhu T: Analysis of transcriptional factors and regulation networks in patients with acute renal allograft rejection. *J Pro‐*

[65] Wu D, Liu X, Liu C, Liu Z, Xu M, Rong R, Qian M, Chen L, Zhu T: Network analysis reveals roles of inflammatory factors in different phenotypes of kidney transplant pa‐

[66] Grimm D, Streetz KL, Jopling CL, Storm TA, Pandey K, Davis CR, Marion P, Salazar F, Kay MA: Fatality in mice due to oversaturation of cellular microRNA/short hair‐

[67] Liu K, Chen H, You Q, Shi H, Wang Z: The siRNA cocktail targeting VEGF and HER2 inhibition on the proliferation and induced apoptosis of gastric cancer cell. *Mol*

[68] Zhou J, Neff CP, Liu X, Zhang J, Li H, Smith DD, Swiderski P, Aboellail T, Huang Y, Du Q, et al: Systemic administration of combinatorial dsiRNAs via nanoparticles effi‐ ciently suppresses HIV-1 infection in humanized mice. *Mol Ther* 2011; 19:2228–2238.

[69] Ge Q, Xu JJ, Evans DM, Mixson AJ, Yang HY, Lu PY: Leveraging therapeutic poten‐ tial of multi-targeted siRNA inhibitors. *Future Med Chem* 2009; 1:1671–1681. DOI:

[70] Jiang N, Zhang X, Zheng X, Chen D, Siu K, Wang H, Ichim TE, Quan D, McAlister V, Chen G, Min WP: A novel in vivo siRNA delivery system specifically targeting liver cells for protection of ConA-induced fulminant hepatitis. *PLoS One* 2012; 7:e44138.

[71] Siu KS, Chen D, Zheng X, Zhang X, Johnston N, Liu Y, Yuan K, Koropatnick J, Gillies ER, Min WP: Non-covalently functionalized single-walled carbon nanotube for topi‐

pin RNA pathways. *Nature* 2006; 441:537–541. DOI:10.1038/nature04791

mal localization. *J Mol Biol* 2005; 348:1079–1090. DOI:10.1016/j.jmb.2005.03.013 [61] Racz Z, Godo M, Revesz C, Hamar P: Immune activation and target organ damage are consequences of hydrodynamic treatment but not delivery of naked siRNAs in

mice. *Nucleic Acid Ther* 2011; 21:215–224. DOI:10.1089/nat.2010.0248

DOI:10.1111/j.1600-6143.2007.02032.x

242 RNA Interference

DOI:10.1038/mt.2011.207

DOI:10.1371/journal.pone.0044138

10.4155/fmc.09.131

2014; 9:e104164. DOI:10.1371/journal.pone.0104164

*teome Res* 2011; 10:175–181. DOI:10.1021/pr100473w

tients. *J Theor Biol* 2014; 362:62–68. DOI:10.1016/j.jtbi.2014.03.006

*Cell Biochem* 2014; 386:117–124. DOI:10.1007/s11010-013-1850-0

[73] Zhu K, Lai H, Guo C, Li J, Wang Y, Wang L, Wang C: Nanovector-based prolyl hy‐ droxylase domain 2 silencing system enhances the efficiency of stem cell transplanta‐ tion for infarcted myocardium repair. *Int J Nanomedicine* 2014; 9:5203–5215. DOI: 10.2147/IJN.S71586

## **Preclinical Development of RNAi-Inducing Oligonucleotide Therapeutics for Eye Diseases**

Tamara Martínez, Maria Victoria González, Beatriz Vargas, Ana Isabel Jiménez and Covadonga Pañeda

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/61803

#### **Abstract**

RNA interference (RNAi) is a posttranscriptional mechanism of gene regulation present in eukaryotic cells. Inducers of RNAi are small molecules of RNA that act in the cyto‐ plasm where they are able to impair translation of a specific mRNA to protein, hence modifying gene expression. The discovery of this mechanism in mammals led to the de‐ velopment of a new class of therapeutics with the aim of exploiting this endogenous mechanism of action. In the last decade, great efforts have been put into understanding RNAi and translating this accumulated knowledge into the design of modern therapeu‐ tics. With several compounds in phase III clinical development, the field is getting closer to its first market authorization. Here we make a thorough overview of the field of RNAi therapeutics in ophthalmology, one of the fields in which RNAi has been most successful.

**Keywords:** RNAi, eye diseases, ophthalmology, drug development

#### **1. Introduction**

Short interference RNAs (siRNA) are small molecules of double-stranded RNA of around 21 base pair long that specifically downregulate the expression of a target gene [1]. This mecha‐ nism of endogenous gene expression regulation, present in most eukaryotic cells, has been thoroughly used to study gene function [2]. SiRNAs exert their function in the cytoplasm of the cell, where they assemble with a several proteins to yield the RNA-induced silencing complex (RISC), a multimeric RNA–protein complex that recognizes complementary messen‐ ger RNAs (mRNA) and promotes their degradation, thus blocking the synthesis of specific proteins. RNAi may be activated by endogenous siRNAs synthesized in the nucleus of the cell and generated by subsequent processing within the cell cytoplasm to yield siRNAs. Alterna‐

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

tively, siRNAs can be exogenously introduced to mimic the action of endogenous RNAi triggers [3]. Among the benefits of RNAi are the potential of transiently silencing any given gene at any stage of development and to affect gene expression in specific anatomical regions without affecting nontargeted regions. These benefits are being used as a basis to develop a new class of innovative drugs that may reach the market in the next five years; the present report highlights the advances made in RNAi therapeutics on the field of ophthalmology.

#### **2. The special environment of the eye: advantages and disadvantages**

The eye has traditionally been considered a good organ for proof-of-concept studies to assess the efficacy of innovative therapies. It has a very particular anatomy that allows the transfor‐ mation of sensory information into electrical signals that can be thereafter interpreted by the brain. Transformation and partial processing of sensory information takes place in the retina, located at the back of the eye. The correct function of the eye requires light to travel through several anatomical structures in order to reach the photosensitive retina; hence, these struc‐ tures have to be transparent or semi-transparent to allow passage of light. The environment of the eye is extremely specialized to ensure that transparency is maintained, and there are several mechanisms in place to ensure that this specialized environment is preserved. One of the anatomical characteristics of the eye to allow light to travel through its structures is restricting the blood flow to regions where transparency is not strictly required.

In addition, access to the innermost regions of the eye is controlled by several barriers; these barriers isolate the eye from external aggressions and pathogens to but also limit the access of therapeutics. The influence of the particular anatomical features of the eye on drug delivery is further explained in Section 4 of this review. The barriers of the eye do not only isolate this organ from external aggressions or substances but also limit the access of internal elements; as such, the immune system has only limited access to the eye making the eye a partially immune-privileged region. Finally, the aqueous humor, the clear liquid that fills the eye and maintains its shape and pressure, has a very low content in proteins compared to serum. Among the proteins that are significantly reduced compared to other tissues are RNases and elements of the complement cascade. The reduced concentration of RNases increases the halflife of RNAs used as therapeutics, and reduction in the elements of the complement cascade further decreases the likelihood of an unwanted immune reactions to drugs.

In order to preserve its integrity, the eye has efficacious barriers to block the entrance of pathogens and substances that could potentially affect its sensory function. The eye has developed specific features that ensure that light travels through its tissues; this specialization of tissues to preserve visual function is also observed by the immune system [4]. Immune responses change the local environment of tissues and are frequently associated to tissue inflammation. In order to avoid these potentially harmful changes, the eye has a relatively immune privileged status. This immune privilege status is maintained by several mechanisms. Absence of lymphatic and blood vessels in certain areas and abundance of immunosuppressive factors in the aqueous humor are among these mechanisms [5]. On the other hand, the eye needs to be able to respond to situations in which its integrity can be compromised such as viral or bacterial infections. The innate immune response is the first system activated in response to aggressions; it acts like a watchdog mechanism recognizing pathogen-associated molecular patterns (PAMP). Depending on the molecular characteristics and location of the PAMP, different effectors of the innate response are activated; these responses can be mediated by toll-like receptors (TLR) or independent of these receptors. Toll-like receptors discriminate self-motifs from non-self-motifs [6]. There are ten known TLRs, and they differ in their subcellular localization and in the type of non-self-pattern they recognize. TLR1, TLR2, TLR4, TLR5, and TLR6 recognize components of bacterial walls and are located in the cell surface, whereas TLR3, TLR7, TLR8, and TLR9 recognize oligonucleotides and are sequestered in intracellular compartment. TLR3 binds to single- and double-stranded RNA, TLR7 binds to single-stranded RNA, and TLR9 binds to unmethylated CpG motifs, usually found in bacterial DNA. In the eye, the expression pattern of each TLRs varies depending on the anatomical structure; all TLRs are present in the corneal and retinal pigment epithelia; TLR4 is the predominant TLR in the rest of the eye structures where it localizes in resident antigen presenting cells [7, 8]. TLR-independent response mechanisms to cytoplasmic oligonucleotides include dsRNA-binding protein kinase, the RNA helicase (RIG-I), and oligoadenylate synthase enzyme. These proteins are cytoplasmic dsRNA sensors belonging to the antiviral innate immune system, which plays an important role in antiviral defense in response to viral infection and replication [9].

tively, siRNAs can be exogenously introduced to mimic the action of endogenous RNAi triggers [3]. Among the benefits of RNAi are the potential of transiently silencing any given gene at any stage of development and to affect gene expression in specific anatomical regions without affecting nontargeted regions. These benefits are being used as a basis to develop a new class of innovative drugs that may reach the market in the next five years; the present report highlights the advances made in RNAi therapeutics on the field of ophthalmology.

246 RNA Interference

**2. The special environment of the eye: advantages and disadvantages**

restricting the blood flow to regions where transparency is not strictly required.

further decreases the likelihood of an unwanted immune reactions to drugs.

The eye has traditionally been considered a good organ for proof-of-concept studies to assess the efficacy of innovative therapies. It has a very particular anatomy that allows the transfor‐ mation of sensory information into electrical signals that can be thereafter interpreted by the brain. Transformation and partial processing of sensory information takes place in the retina, located at the back of the eye. The correct function of the eye requires light to travel through several anatomical structures in order to reach the photosensitive retina; hence, these struc‐ tures have to be transparent or semi-transparent to allow passage of light. The environment of the eye is extremely specialized to ensure that transparency is maintained, and there are several mechanisms in place to ensure that this specialized environment is preserved. One of the anatomical characteristics of the eye to allow light to travel through its structures is

In addition, access to the innermost regions of the eye is controlled by several barriers; these barriers isolate the eye from external aggressions and pathogens to but also limit the access of therapeutics. The influence of the particular anatomical features of the eye on drug delivery is further explained in Section 4 of this review. The barriers of the eye do not only isolate this organ from external aggressions or substances but also limit the access of internal elements; as such, the immune system has only limited access to the eye making the eye a partially immune-privileged region. Finally, the aqueous humor, the clear liquid that fills the eye and maintains its shape and pressure, has a very low content in proteins compared to serum. Among the proteins that are significantly reduced compared to other tissues are RNases and elements of the complement cascade. The reduced concentration of RNases increases the halflife of RNAs used as therapeutics, and reduction in the elements of the complement cascade

In order to preserve its integrity, the eye has efficacious barriers to block the entrance of pathogens and substances that could potentially affect its sensory function. The eye has developed specific features that ensure that light travels through its tissues; this specialization of tissues to preserve visual function is also observed by the immune system [4]. Immune responses change the local environment of tissues and are frequently associated to tissue inflammation. In order to avoid these potentially harmful changes, the eye has a relatively immune privileged status. This immune privilege status is maintained by several mechanisms. Absence of lymphatic and blood vessels in certain areas and abundance of immunosuppressive factors in the aqueous humor are among these mechanisms [5]. On the other hand, the eye

The first proof-of-concept studies to demonstrate the viability of silencing genes in the eye showed that the injection of siRNAs into the subretinal space or vitreous cavity could indeed downregulate specific genes. In these experiments, the downregulation of genes of the vascular endothelial growth factor (VEGF) family with siRNAs correlated with the inhibition of ocular neovascularization [10, 11]. The first set of these experiments used an adenoviral vector that codified for an siRNA designed against VEGF1. The subretinal injection of this vector 36 h after the induction of coroidal neovascularization (CNV) by laser reduced the areas of neovascula‐ rization compared to areas of mice injected with vector codifying for a scrambled siRNA [10]. In a subsequent study, Campochiaro and coworkers [10] demonstrated that the inhibition of ocular neovascularization could also be achieved by injecting a VEGFR1 siRNA directly, without an expression vector, into the vitreous cavity. The siRNA used in this study had the so-called canonical design, which comprises a 21-nucleotide guide strand and a complemen‐ tary passenger strand annealed to form an siRNA duplex with a 19-bp dsRNA stem and 2 nucleotide 3′ overhangs at both ends [11].

In 2008, Kleinman and coworkers published a study in *Nature* demonstrating that the effect of siRNAs on CNV was mediated by activation of TLR3 rather than an effect on target [12]. The results of these studies showed that the effect of the siRNAs on CNV was dosedependent but not dependent on sequence. In addition, the authors demonstrated that a minimum length of the siRNA was required in order for the molecules to have an effect on CNV; this length was show to be at least 21 nucleotides. The study also showed that the internalization of the siRNAs was not required for the inhibition of CNV as cells of the retinal pigment epithelium (RPE) abundantly express TLR3 on the cytoplasmic surface. The authors used several sequences to point out that the inhibition of CNV was mediated through an off-target effect. Using docking models, the authors showed that TLR3 and siRNAs were indeed able to interact with each other but the interaction was unstable when siRNAs were shorter than 21 nucleotides. A subsequent study by the same group showed that activation of TLR3 by IVT siRNAs led to caspase-3-mediated degeneration of the RPE questioning the safety these compounds as therapeutics for back of the eye diseases [13]. The findings of Kleinman and coworkers boosted research on alternative designs that could successfully block immune recognition; among the most commonly used strategies are incorporation of delivery systems and 2'-ribose modifications.

Finally, exogenous RNAs are quickly degraded by RNases present in tissues and body fluids [14]. These enzymes cut RNA into smaller components that are subsequently incorporated into the route of degradation of nucleotides. RNAses are present at high concentrations both in tissues, such as in spleen, liver, and pancreas, and in biological fluids, such as in serum [15]. In the eye, the tear fluid is rich in nucleases but the presence of these enzymes is considerably lower in most eye tissues, thus allowing for increased half-life of the siRNAs used for thera‐ peutic purposes.

#### **3. Efficacy studies: Animal models to study the eye**

Proof-of-concept studies are required in order to demonstrate that a particular drug has the proposed activity. siRNAs are designed using bioinformatics tools to target specific regions of the human genome. Therefore, assessing the activity of these molecules in animal models requires that the siRNA has biological activity in the species chosen to perform the proof-ofconcept study. With the sequencing of the genome of most animal models used in biomedical research, evaluating the homology of a given sequence between two species is nowadays common practice, but this does not warrant finding a fully homologous siRNA for all targeted genes. In cases where homologous sequences cannot be found, a surrogate compound can be used to perform animal efficacy studies; this entails designing a compound that targets the exact same region as the human version but with the sequence of the gene of the animal species to be used. Several animal models can be used to assess the activity of a compound. Here we highlighted the animal models used in the developmental programs of products included in the ophthalmic siRNA pipeline.

#### **3.1. Ischemic optic neuropathy**

Ischemic optic neuropathy is a sudden loss of vision caused by interruption or decreased blood flow in the optic nerve. There is a disagreement as to its pathogenesis, clinical features, and management because ischemic optic neuropathy is not a unique disease, but a spectrum of different types [16]. Ischemic optic neuropathy can be primarily of two types: anterior (AION) caused by the interruption of blood flow in the optic nerve head and posterior (PION) involving the posterior part the optic nerve. Both types can be further classified into different subtypes. AION comprises arteritic (A-AION) caused by giant cell arteritis and nonarteritic (NA-AION) caused by other than giant cell arteritis. NA-AION is by far the most common type and typically affects individuals between 55 and 67 years of age. The incidence of AION has only been thoroughly studied in the USA where there are 2.3–10.3 cases per 100,000 inhabitants; for the nonarteritic type, the numbers are lower: 0.36 per 100,000 inhabitants. NA-AION is characterized by visual loss, optic disc swelling, sometimes with flame hemorrhages on the swollen disc or nearby neuroretinal layer, and sometimes with nearby cotton wool exudates. Visual loss is usually sudden and may progress over several hours to days and even weeks [17].

Animal modes of this disease are used to assess efficacy of pharmaceuticals in development for these conditions and include the optic nerve crush model and the photoembolic stroke model. The optic nerve crush model is a general model in which surviving of the ganglionar cells can be studied in response to a physical damage to the optic nerve [18]. In the photoem‐ bolic stroke model, a photosensitive dye, such as rose bengal, is injected intravenously, and laser is specifically applied to the optic nerve head to activate the dye. The activation of the dye results in damage to the endothelial cells of the optic nerve vascular system that ultimately leads to thrombosis of vessels and edema of the optic nerve head [19].

#### **3.2. Glaucoma**

through an off-target effect. Using docking models, the authors showed that TLR3 and siRNAs were indeed able to interact with each other but the interaction was unstable when siRNAs were shorter than 21 nucleotides. A subsequent study by the same group showed that activation of TLR3 by IVT siRNAs led to caspase-3-mediated degeneration of the RPE questioning the safety these compounds as therapeutics for back of the eye diseases [13]. The findings of Kleinman and coworkers boosted research on alternative designs that could successfully block immune recognition; among the most commonly used strategies are

Finally, exogenous RNAs are quickly degraded by RNases present in tissues and body fluids [14]. These enzymes cut RNA into smaller components that are subsequently incorporated into the route of degradation of nucleotides. RNAses are present at high concentrations both in tissues, such as in spleen, liver, and pancreas, and in biological fluids, such as in serum [15]. In the eye, the tear fluid is rich in nucleases but the presence of these enzymes is considerably lower in most eye tissues, thus allowing for increased half-life of the siRNAs used for thera‐

Proof-of-concept studies are required in order to demonstrate that a particular drug has the proposed activity. siRNAs are designed using bioinformatics tools to target specific regions of the human genome. Therefore, assessing the activity of these molecules in animal models requires that the siRNA has biological activity in the species chosen to perform the proof-ofconcept study. With the sequencing of the genome of most animal models used in biomedical research, evaluating the homology of a given sequence between two species is nowadays common practice, but this does not warrant finding a fully homologous siRNA for all targeted genes. In cases where homologous sequences cannot be found, a surrogate compound can be used to perform animal efficacy studies; this entails designing a compound that targets the exact same region as the human version but with the sequence of the gene of the animal species to be used. Several animal models can be used to assess the activity of a compound. Here we highlighted the animal models used in the developmental programs of products included in

Ischemic optic neuropathy is a sudden loss of vision caused by interruption or decreased blood flow in the optic nerve. There is a disagreement as to its pathogenesis, clinical features, and management because ischemic optic neuropathy is not a unique disease, but a spectrum of different types [16]. Ischemic optic neuropathy can be primarily of two types: anterior (AION) caused by the interruption of blood flow in the optic nerve head and posterior (PION) involving the posterior part the optic nerve. Both types can be further classified into different subtypes. AION comprises arteritic (A-AION) caused by giant cell arteritis and nonarteritic (NA-AION) caused by other than giant cell arteritis. NA-AION is by far the most common

incorporation of delivery systems and 2'-ribose modifications.

**3. Efficacy studies: Animal models to study the eye**

peutic purposes.

248 RNA Interference

the ophthalmic siRNA pipeline.

**3.1. Ischemic optic neuropathy**

Glaucoma is a group of progressive optic neuropathies characterized by vision impair‐ ment and degeneration of retinal ganglion cells that if left untreated can lead to blind‐ ness. Glaucoma is the second leading cause of global irreversible blindness, and it has been estimated that 60.5 million people were affected by primary open-angle glaucoma (POAG) and primary angle–closure glaucoma (PACG) globally in 2010, a number expected to increase to nearly 80 million by 2020 [20]. The degeneration of the optic nerve is thought to be produced as a result of changes in intraocular pressure (IOP), but specific molecu‐ lar mediators of these changes have yet to be identified. Because glaucoma may be asymptomatic until a relatively late stage where the nerve damage has already occurred, early diagnosis and treatment are crucial for halting the progression of the condition [21]. Reduction of IOP is the only proven strategy to treat the disease; thus, first-line treat‐ ments are aimed toward achieving this goal. There are several compounds currently used to lower IOP, and although most of them are efficacious in lowering IOP, they all come with their own set of side effects and tolerance to the drug is a frequent phenomenon. Tolerance or reduced response of the drug requires change of drug regimen, frequently increasing the dose or combining the prescribed pharmaceutical with another drug [22, 23].

Depending on the specific phase of the disease one wants to model, several animal models can be used [24]. If studying the degeneration of the retina is the main objective, the models mentioned in the previous section can be used. For assessing the IOP lowering efficacy of pharmaceuticals, models with increased IOP are generally used. The increased IOP model induced by oral water overloading in rabbits is a very easy and physiologic model to screen compounds. The basis behind this model is to give the animal an oral overload of water that will result in a transient increase in IOP [25]. Although the specific mechanism behind the increase in IOP following water overloading is uncertain, the model has been extensively used to perform rapid screens of IOP-lowering compounds. The main advantage of this model over other existing models of increased IOP is that the anatomical structure and physiology of eye structures are preserved allowing a normal response to hypotensive drugs. Other models of increased IOP include laser photocoagulation, intracameral injection of latex microspheres, topical application of prednisolone, light-induced reduced outflow facility, subconjunctival injection of betamethasone, or episcleral vein occlusion [24].

#### **3.3. Dry eye disease**

Dry eye disease (DED) is a multifactorial disease of the tear-fluid and ocular surface that results in symptoms of discomfort, visual disturbance, and tear film instability. It is accompanied by increased osmolarity of the tear film and inflammation of the ocular surface [26]. Common symptoms of this condition include blurry vision, tearing, and ocular pain. There are several factors that contribute to the etiology of the disease, among them are insufficient tear secretion, excessive evaporation, and alteration in the composition of the tear film [27]. Temporary changes of the composition of the tear film can cause an acute form of DED; if changes persist, the condition can turn into chronic DED. Damage to the ocular surface is usually more severe in the chronic forms than in the acute types. DED is frequently associated to other conditions such as Sjögren's disease or lachrymal gland dysfunction, but it can also be caused by vitamin deficiency, contact lens wear, and use of several prescription drugs. Acute DED is handled with lubricants and avoiding preservatives in concomitant eye drops. If DED persists, treatment options include procedures that favor tear retention such as punctal occlusion, moisture chamber spectacles, contact lenses, or pharmacologic agents that stimulate tear secretion. More severe forms may require the use of anti-inflammatory therapy [28]. Although some advances have been made toward alleviating some of the symptoms of DED, pain associated to this condition is not usually addressed. Pain in the eye results from stimulation of sensory axons of the trigeminal ganglion neurons innervating the cornea [29]. Animal models to assess the efficacy of ocular analgesics are extremely complex in terms of interpreting efficacy outcomes [30]. One of the commonly used models to study pain is the capsaicininduced ocular pain model developed by Gertrudis and colleagues [31]. This model is based on the evaluation of animal behavior after topical ocular administration of capsaicin, a selective agonist of transient receptor protein vaniloid type 1 (TRPV1). Capsaicin applied locally to the eye activates TRPV1 inducing palpebral closure. Latency to open the eye and time required for complete palpebral opening can be used as measurements of the discomfort caused by capsaicin. Reference products used in this model include analgesics, in particular, capsazepine, the antagonist of TRPV1 channels.

#### **3.4. Ocular allergy**

Ocular allergies constitute a heterogenic group of diseases with a broad spectrum of clinical manifestations and include mild forms such as seasonal allergic conjunctivitis (SAC) and perennial allergic conjunctivitis (PAC), and more severe manifestations such as vernal keratoconjunctivitis (VKC), atopic keratoconjunctivitis (AKC), and giant papillary conjuncti‐ vitis (GPC). The severe forms can be associated to complications such as corneal damage and may cause vision loss. SAC and PAC are commonly IgE-mast cell-mediated hypersensitivity reaction to external allergens, whereas AKC and VKC are characterized by chronic inflamma‐ tion involving several immune cell types. In SAC and PAC, allergens, with the help of antigen presenting cells, trigger a Th2-predominant immune response that induces B cells to release IgE. In SAC and PAC, allergen-induced local release of IgE prompts infiltration and degranu‐ lation of mast cells in Ca2+-dependent mechanism. Mast cells liberate preformed inflammatory mediators such as histamine and leukotriene 4 that subsequently attract eosinophils amplify the allergic response [32]. The prevalence of ocular allergies in the general population is estimated to be around 40% in the United States [33] and up to 35% in Europe and the Middle East [34], but it is probably underestimated in most epidemiologic studies [35]. The primary treatment for ocular allergies includes avoidance of allergens, cold compresses, and lubrica‐ tion. In persisting cases, symptoms can be treated using topical and oral decongestants, antihistamines, mast-cell stabilizers, or anti-inflammatory agents [36]. Allergic conjunctivitis can be modeled in animals by exposing them to allergens in the presence of an adjuvant [37]. The model developed by Magone and coworkers uses Female Balb/C mice that are sensitized with short ragweed and alum and several days later animals receive a topical dose of short ragweed pollen in the eye. A prescreening of mice can be performed in order to select only those animals that respond to allergens.

#### **3.5. Age-related macular degeneration with choroidal neovascularization**

structures are preserved allowing a normal response to hypotensive drugs. Other models of increased IOP include laser photocoagulation, intracameral injection of latex microspheres, topical application of prednisolone, light-induced reduced outflow facility, subconjunctival

Dry eye disease (DED) is a multifactorial disease of the tear-fluid and ocular surface that results in symptoms of discomfort, visual disturbance, and tear film instability. It is accompanied by increased osmolarity of the tear film and inflammation of the ocular surface [26]. Common symptoms of this condition include blurry vision, tearing, and ocular pain. There are several factors that contribute to the etiology of the disease, among them are insufficient tear secretion, excessive evaporation, and alteration in the composition of the tear film [27]. Temporary changes of the composition of the tear film can cause an acute form of DED; if changes persist, the condition can turn into chronic DED. Damage to the ocular surface is usually more severe in the chronic forms than in the acute types. DED is frequently associated to other conditions such as Sjögren's disease or lachrymal gland dysfunction, but it can also be caused by vitamin deficiency, contact lens wear, and use of several prescription drugs. Acute DED is handled with lubricants and avoiding preservatives in concomitant eye drops. If DED persists, treatment options include procedures that favor tear retention such as punctal occlusion, moisture chamber spectacles, contact lenses, or pharmacologic agents that stimulate tear secretion. More severe forms may require the use of anti-inflammatory therapy [28]. Although some advances have been made toward alleviating some of the symptoms of DED, pain associated to this condition is not usually addressed. Pain in the eye results from stimulation of sensory axons of the trigeminal ganglion neurons innervating the cornea [29]. Animal models to assess the efficacy of ocular analgesics are extremely complex in terms of interpreting efficacy outcomes [30]. One of the commonly used models to study pain is the capsaicininduced ocular pain model developed by Gertrudis and colleagues [31]. This model is based on the evaluation of animal behavior after topical ocular administration of capsaicin, a selective agonist of transient receptor protein vaniloid type 1 (TRPV1). Capsaicin applied locally to the eye activates TRPV1 inducing palpebral closure. Latency to open the eye and time required for complete palpebral opening can be used as measurements of the discomfort caused by capsaicin. Reference products used in this model include analgesics, in particular, capsazepine,

Ocular allergies constitute a heterogenic group of diseases with a broad spectrum of clinical manifestations and include mild forms such as seasonal allergic conjunctivitis (SAC) and perennial allergic conjunctivitis (PAC), and more severe manifestations such as vernal keratoconjunctivitis (VKC), atopic keratoconjunctivitis (AKC), and giant papillary conjuncti‐ vitis (GPC). The severe forms can be associated to complications such as corneal damage and may cause vision loss. SAC and PAC are commonly IgE-mast cell-mediated hypersensitivity reaction to external allergens, whereas AKC and VKC are characterized by chronic inflamma‐

injection of betamethasone, or episcleral vein occlusion [24].

**3.3. Dry eye disease**

250 RNA Interference

the antagonist of TRPV1 channels.

**3.4. Ocular allergy**

Age-related macular degeneration (AMD) is the leading cause of severe vision loss in indi‐ viduals over 50 years of age [38, 39]. AMD is caused by a combination of genetic and environ‐ mental factors. Risk factors include hypertension, cardiovascular disease, smoking, and high BMI. Among the genetic factors that confer susceptibility to developing AMD are variants in genes encoding complement pathway proteins [40, 41].

The underlying cause for AMD is accumulation of drusen orresidual material produced by the renewal process of the external part of the photoreceptors of the retina in the retinal pigment epithelium (RPE). The accumulation of this material in the RPE leads to the production of inflammatorymediators that causephotoreceptordegenerationinthemacula andseverevision loss [42]. In the early stages of the disease, accumulated drusen are small; the size and amount of this material increase as the disease progresses and central vision deteriorates.

There are two types of AMD: dry or wet. Dry AMD is characterized by the degeneration of the RPE and photoreceptors along with changes in pigmentation of the RPE. In the wet form, or choroidal neovascularization (CNV), fragile blood vessels of the choriocapillaris grow into the RPE and frequently leak blood and fluid that accumulate between RPE and choriocapillaris. As a result of these abnormal growths, dense scars are formed in the macula, and the RPE can detach. The wet form is more severe than the dry form and sometimes dry AMD can develop into wet AMD [43].

The characteristic invasion of leaky blood to the RPE in wet AMD is mediated by VEGF. The discovery of the relationship between VEGF and changes in vasculature in AMD led to the development of different approaches aimed to decrease the levels of this growth factor. Antibodies targeting VEGF are currently the first-line treatment for wet AMD [44]. The hallmark of wet AMD is CNV; thus, this is the lesion most extensively modeled in animals to assess efficacy of compounds targeting this disease. The laser-induced CNV model is by far the most used animal model. This model, initially developed for nonhuman primates (NHP), was later adapted into rodents. The basis for this model is to induce a break in Bruch's membrane using a high-energy laser. The experimental CNV can be analyzed in vivo using fluorescein angiography or optical coherence tomography or postmortem studying the retina explants. The model has been successfully transferred and validated to rat and mouse, and in both species, the chain of events taking place after lesion induction resembles the events that tale place in humans with the disease. Other models include the injection of subretinal materials such as Matrigel, angiogenic substances, macrophages, lipid peroxides, or polyethy‐ lene glycol. Although these models are promising, they have yet to be appropriately validated in order to be used as a proof of concept tools [45].

#### **3.6. Diabetic retinopathy**

Diabetic retinopathy (DR) is an ocular complication of diabetes mellitus characterized by microaneurysms in the retinal vasculature that eventually lead to ischemia and macular edema. Changes in the retina can cause rapid vision loss, and this complication is the main cause of visual loss in working-age individuals [46, 47].

The initial phase of DR, known as nonproliferative DR, is characterized by the thickening of the capillary basement membrane and apoptosis and migration of pericytes. These microchanges cause microaneurysms and small leakages in the vessels that irrigate the retina. As the disease progresses, interaction between endothelial cells and pericytes weakens and the capillaries become permeable; subsequent accumulation of fluids in the macula leads to edema. The microaneurysms inthe retinal capillaries cause occlusions that compromisebloodflowthrough the retina and cause ischemia. Local hypoxia upregulates angiogenic factors that cause capillaries to grow into the retina, preretinal space, and vitreous cavity; stage known as proliferative DR [48]. Among the upregulated angiogenic factors, one of the most critical is VEGF; the newly formed vessels are structurally deficient and very responsive to this growth factor. As such, antibodies used to treat AMD are also used for the treatment of diabetic retinopathy. DR is usually treated with laser photocoagulation, a procedure that does not cure the disease but mitigates the damage. IVT steroids can also be used to reduce accumulation of fluidswithinthe retina.If accumulationofbloodinthevitreoushumorphysicallyimpedes laser photocoagulation, a vitrectomy has to be performed in order to remove the blood accumulat‐ ed in the vitreous prior to laser photocoagulation.

There are several animal models of diabetic retinopathy, each of them comes with its own set of advantages and disadvantages. One ofthe most extensively used is the streptozotocin (STZ) induced diabetes model. Intravenous or intraperitoneal injection of STZ causes a rapid and selective destruction of β-pancreatic cells leading to hyperglycemia and development of type I diabetes. The model has been used successfully in several animal species including rat, mouse, rabbit, dog, and monkey. Nonproliferative DR develops in this model, but microaneurysms and neovascularization are seldom observed; hence, this model can be complemented with the laser-inducedCNVmodelexplainedintheAMDsection.Largeranimalmodelscanbegenerated by surgically removing the pancreas, but this model is significantly more complicated to generate than the STZ model and has the same drawbacks. Alternatively, animals can be fed a high-galactose diet, but the induction of diabetes is considerably slower [50].

#### **4. Biodistribution studies**

the most used animal model. This model, initially developed for nonhuman primates (NHP), was later adapted into rodents. The basis for this model is to induce a break in Bruch's membrane using a high-energy laser. The experimental CNV can be analyzed in vivo using fluorescein angiography or optical coherence tomography or postmortem studying the retina explants. The model has been successfully transferred and validated to rat and mouse, and in both species, the chain of events taking place after lesion induction resembles the events that tale place in humans with the disease. Other models include the injection of subretinal materials such as Matrigel, angiogenic substances, macrophages, lipid peroxides, or polyethy‐ lene glycol. Although these models are promising, they have yet to be appropriately validated

Diabetic retinopathy (DR) is an ocular complication of diabetes mellitus characterized by microaneurysms in the retinal vasculature that eventually lead to ischemia and macular edema. Changes in the retina can cause rapid vision loss, and this complication is the main

The initial phase of DR, known as nonproliferative DR, is characterized by the thickening of the capillary basement membrane and apoptosis and migration of pericytes. These microchanges cause microaneurysms and small leakages in the vessels that irrigate the retina. As the disease progresses, interaction between endothelial cells and pericytes weakens and the capillaries become permeable; subsequent accumulation of fluids in the macula leads to edema. The microaneurysms inthe retinal capillaries cause occlusions that compromisebloodflowthrough the retina and cause ischemia. Local hypoxia upregulates angiogenic factors that cause capillaries to grow into the retina, preretinal space, and vitreous cavity; stage known as proliferative DR [48]. Among the upregulated angiogenic factors, one of the most critical is VEGF; the newly formed vessels are structurally deficient and very responsive to this growth factor. As such, antibodies used to treat AMD are also used for the treatment of diabetic retinopathy. DR is usually treated with laser photocoagulation, a procedure that does not cure the disease but mitigates the damage. IVT steroids can also be used to reduce accumulation of fluidswithinthe retina.If accumulationofbloodinthevitreoushumorphysicallyimpedes laser photocoagulation, a vitrectomy has to be performed in order to remove the blood accumulat‐

There are several animal models of diabetic retinopathy, each of them comes with its own set of advantages and disadvantages. One ofthe most extensively used is the streptozotocin (STZ) induced diabetes model. Intravenous or intraperitoneal injection of STZ causes a rapid and selective destruction of β-pancreatic cells leading to hyperglycemia and development of type I diabetes. The model has been used successfully in several animal species including rat, mouse, rabbit, dog, and monkey. Nonproliferative DR develops in this model, but microaneurysms and neovascularization are seldom observed; hence, this model can be complemented with the laser-inducedCNVmodelexplainedintheAMDsection.Largeranimalmodelscanbegenerated by surgically removing the pancreas, but this model is significantly more complicated to generate than the STZ model and has the same drawbacks. Alternatively, animals can be fed a

high-galactose diet, but the induction of diabetes is considerably slower [50].

in order to be used as a proof of concept tools [45].

cause of visual loss in working-age individuals [46, 47].

ed in the vitreous prior to laser photocoagulation.

**3.6. Diabetic retinopathy**

252 RNA Interference

Despite the extraordinary potential that RNAi technology displays in the treatment of ocular conditions, the transition of siRNAs programs to the clinical setting still presents challenges. The *in vivo* efficacy of therapies based on siRNAs depends on the ability of a given siRNA to reach the cytoplasm of its target cell in sufficient quantities to achieve its desired biological effect. The intrinsic characteristics of siRNAs such as their sensitivity to degradation by endogenous enzymes, their relative large size, and its negative charge limit their ability to cross biological barriers and reach the cytoplasm. Approaches used to overcome the hurdles associated to the use of siRNAs range from delivery strategies to chemical modifications aimed towards improving the pharmacological properties of the therapeutic siRNAs.

Drug delivery into the eye is challenging due to the presence of static and dynamics barriers that protect the internal tissues. The eye consists of two anatomically differentiated regions: the anterior and posterior segments. The anterior segment includes the cornea, conjunctiva, iris, ciliary body, lens, and anterior and posterior chambers; this segment occupies approxi‐ mately the anterior third of the eyeball. The posterior segment is of greater size and comprises the sclera, choroid, retina, and vitreous cavity. There are significant anatomical, molecular, and immune differences between the two segments; thus, strategies to deliver molecules to the eye will be very different depending on the targeted segment [51]. The anterior region of the eye is protected from exterior aggressions by the cornea and tear film. The former is a specialized tissue composed of five layers that constitutes the main physical barrier to external molecules; the latter is an enzyme-rich fluid that degrades many biological molecules, lubricates the eye surface, and washes away materials from the cornea. In addition, many components of the tear film impede adhesion of molecules to the eye surface further restricting the access of external molecules to the inside of the eye.

Topical ocular administration of drugs is a patient-friendly administration route typically used for the treatment of pathologies affecting the anterior segment of the eye. However, molecules applied as eye drops are quickly cleared from the ocular surface being the bioavailability of a compound administered via this route less than 5% of the initially applied dose. The standard volume of a commercial eye drop is approximately 40 μL whereas the normal volume of the tear fluid in the ocular surface is 7-9 μl. Once an eye drop is instilled in the inferior conjunctival sac, there is a transient increase of volume that activates the blinking reflex and increases the turnover of the tear film. Most of the content of the eye drop is spilled out by the blinking process or drained via the nasolacrimal duct, drastically reducing the amount of compound available to the eye.

The cornea is a specialized tissue covering the anterior part of the eye whose main functions are protecting against harmful agents and provide the eye with a refractive surface that allows the entrance of light. The human cornea is approximately 0.5–0.8 mm thick, and it is comprised of three layers: the outer five cell layer-epithelium, a thick stroma rich in type I collagen fibrils and glycosaminoglycans, and the innermost endothelium consisting in a single layer of cuboidal cells. The corneal epithelium is separated from the stroma by the Bowman´s mem‐ brane, while the stroma and the corneal endothelium are separated by the Descemet´s membrane. There are no blood vessels irrigating the cornea; this provides the required transparency for the transmission and refraction of light. Drugs can take two paths to penetrate the corneal epithelium: the intracellular path crossing through the cells or the paracellular path bypassing between cells. The cells of the corneal epithelium are tightly attached to each other with gap and tight junctions that restrict the diffusion of large molecules between them [52– 54]. The cross-cellular pathway requires molecules to be able to cross cell membranes; thus, lipophilic molecules have an easier access through this route. The stroma is an aqueous matrix composed mainly of hydrated collagen and proteoglycans with few keratinocytes interspersed [55]. The hydophilicity/lipophilicity index determines the diffusion of molecules through this layer [56]. The remaining layers of the cornea do not significantly hamper the diffusion of molecules.

Contrary to the cornea, the conjunctiva is a highly vasculated tissue that covers the sclera and lines the inner surface of the eyelids. Its main functions are producing mucus and tears to lubricate the eye surface and preventing the entrance of pathogens. The human conjunctiva is composed of three layers: the outer epithelium, the substantia propia, containing nerves and blood vessels, and the submucosa layer, which provides a lightweight attachment to the underlyingsclera[57].Thehistologyofthestratifiedouterepitheliumvariesamongthedifferent regions of the conjunctiva, but it is always its apical portion that controls the permeability of the conjunctiva. The conjunctiva offers an attractive route for drug delivery when compared to the cornea as it presents an extended exchange surface as well as a superior rate of permea‐ tion to large hydrophilic molecules. The sclera is structurally continuous with the cornea and extendsposteriorlyfromthe limbus.The compositionofthe sclerais similartothatofthe corneal stroma, mainly collagen and mucopolysaccharides leaving numerous channels through which drugs can freely diffuse [58]. The sclera is poorly vascularized and significantly more permea‐ ble than the cornea but less permeable than conjunctiva. There are contradictory reports on the ability of charged molecules to cross the sclera. Some authors suggest that this layer is more permeable to negatively charged molecules [59, 60], whereas other studies suggest that positively charged molecules cross the sclera more easily [61, 62]. Ranta and colleagues suggestedthatthenegative chargeofmucopolysaccharides inthe sclerapreventedthediffusion ofnegativelychargedmoleculesasaconsequenceofchargerepulsion.Otherstudieshaveshown that negatively charged molecules are indeed able to cross the sclera, pointing out that size is the limiting factor in drug diffusion through this layer [52]. It should be noted, however, that scleral drug binding does not necessarily impair drug delivery to inner structures of the eye; it can also act as a drug-depot if the molecules are subsequently released [63].

Ophthalmic drugs topically administered to de eye can thus be absorbed through two pathways: crossing the cornea to reach the aqueous humor or through the conjunctival-scleral pathway reaching the uvea. The relative quantity that enters through each of the abovementioned routes varies significantly depending on the size and hydrophilic/lipophilic ratio of the molecule. Generally, the conjunctival route is favored for large hydrophilic molecules, whereas small lipophilic drugs are mainly absorbed through the cornea. The ability of siRNAs to penetrate the cornea has been thoroughly demonstrated as well as the ability of these compounds to enter the cytoplasm of cells within the cornea. However, the capacity of siRNAs to cross the cornea is limited, as shown by the limited amount of siRNAs detected in the aqueous humor following eye drop instillation [64].

Increasing the amount of compound in the anterior part may be of interest for treating specific conditions. For this purpose, several strategies can be used in order to improve delivery: (a) increasing the residence time of the compound within the eye surface, (b) directing the molecule to a specific region to increase the concentration locally, and (c) increasing absorption by using physical methods. Increasing the contact time of the molecule with the eye surface can be achieved by the use of formulations or depots. Formulations that increase viscosity and/ or mucoadhesion of ophthalmic solutions are generally believed to increase absorption into the eye. Polymers such as methylcellulose or polyvinyl alcohol can be added to solutions to increase viscosity and consequently increase residence and reduce clearance time. Mucoad‐ hesion may be increased by formulating the oligonucleotides in polymers such as chitosans. These polymers have been used to deliver DNA vectors into the eye [65]. Encapsulation in liposomes and in thermosensitive gels has also been attempted as a means to increase the absorption and residence time of oligonucleotides in the eye [53]. In these studies, a 16-mer was formulated in liposomes, a thermosensitve 27% poloxamer gel, and HEPES; the results of these studies showed that the amount of compound reaching external tissues such as the conjunctiva or the cornea was higher when the compound was prepared in HEPES. By contrast, access to deeper regions of the anterior chamber such as the sclera or the iris benefited from the increased viscosity of the gel formulation [53]. One of the main drawbacks of biodistribu‐ tion studies to assess the fate of a given siRNA in a formulation is that most of these studies focus on the fate of nanocarrier rather than on that of the oligonucleotide and the relative distribution of the molecule among the tissues of the eye. Therefore, thorough biodistribution, studies are required to address the specific characteristics required for improving delivery for specific conditions. Targeting has scarcely been used to deliver oligonucleotides into the eye; there are a few reports using dendrimers with the goal of increasing the intracellular concen‐ tration of therapeutic oligonucleotides in specific regions of the eye, but advances toward this goal are as of today very limited [66]. Physical methods such as iontophoresis have also been studied aiming to increasing the amount of molecule that crosses the cornea and/or the sclera. Although iontophoresis certainly increases the amount of transcorneal and transcleral delivery of oligonucleotides mainly to the anterior chamber but also to some degree to the posterior chamber, the use of this method has not been extensively used most likely because the equipment required to apply the required current would entail in-office administration, which would significantly complicate repeated administrations [67].

membrane. There are no blood vessels irrigating the cornea; this provides the required transparency for the transmission and refraction of light. Drugs can take two paths to penetrate the corneal epithelium: the intracellular path crossing through the cells or the paracellular path bypassing between cells. The cells of the corneal epithelium are tightly attached to each other with gap and tight junctions that restrict the diffusion of large molecules between them [52– 54]. The cross-cellular pathway requires molecules to be able to cross cell membranes; thus, lipophilic molecules have an easier access through this route. The stroma is an aqueous matrix composed mainly of hydrated collagen and proteoglycans with few keratinocytes interspersed [55]. The hydophilicity/lipophilicity index determines the diffusion of molecules through this layer [56]. The remaining layers of the cornea do not significantly hamper the diffusion of

Contrary to the cornea, the conjunctiva is a highly vasculated tissue that covers the sclera and lines the inner surface of the eyelids. Its main functions are producing mucus and tears to lubricate the eye surface and preventing the entrance of pathogens. The human conjunctiva is composed of three layers: the outer epithelium, the substantia propia, containing nerves and blood vessels, and the submucosa layer, which provides a lightweight attachment to the underlyingsclera[57].Thehistologyofthestratifiedouterepitheliumvariesamongthedifferent regions of the conjunctiva, but it is always its apical portion that controls the permeability of the conjunctiva. The conjunctiva offers an attractive route for drug delivery when compared to the cornea as it presents an extended exchange surface as well as a superior rate of permea‐ tion to large hydrophilic molecules. The sclera is structurally continuous with the cornea and extendsposteriorlyfromthe limbus.The compositionofthe sclerais similartothatofthe corneal stroma, mainly collagen and mucopolysaccharides leaving numerous channels through which drugs can freely diffuse [58]. The sclera is poorly vascularized and significantly more permea‐ ble than the cornea but less permeable than conjunctiva. There are contradictory reports on the ability of charged molecules to cross the sclera. Some authors suggest that this layer is more permeable to negatively charged molecules [59, 60], whereas other studies suggest that positively charged molecules cross the sclera more easily [61, 62]. Ranta and colleagues suggestedthatthenegative chargeofmucopolysaccharides inthe sclerapreventedthediffusion ofnegativelychargedmoleculesasaconsequenceofchargerepulsion.Otherstudieshaveshown that negatively charged molecules are indeed able to cross the sclera, pointing out that size is the limiting factor in drug diffusion through this layer [52]. It should be noted, however, that scleral drug binding does not necessarily impair drug delivery to inner structures of the eye; it

can also act as a drug-depot if the molecules are subsequently released [63].

aqueous humor following eye drop instillation [64].

Ophthalmic drugs topically administered to de eye can thus be absorbed through two pathways: crossing the cornea to reach the aqueous humor or through the conjunctival-scleral pathway reaching the uvea. The relative quantity that enters through each of the abovementioned routes varies significantly depending on the size and hydrophilic/lipophilic ratio of the molecule. Generally, the conjunctival route is favored for large hydrophilic molecules, whereas small lipophilic drugs are mainly absorbed through the cornea. The ability of siRNAs to penetrate the cornea has been thoroughly demonstrated as well as the ability of these compounds to enter the cytoplasm of cells within the cornea. However, the capacity of siRNAs to cross the cornea is limited, as shown by the limited amount of siRNAs detected in the

molecules.

254 RNA Interference

Drugs administered systemically enter the eye from the bloodstream crossing the capillaries of the choroid. The choroid is a vascular layer composed of capillaries and supported by Bruch's membrane, a connective membrane of 2–4 μm thickness. Bruch's membrane separates the choroid from the retina, forming the main barrier to permeation across the choroid-Bruch's bilayer [68]. The permeability of Bruch's membrane is relatively high; charge and size do not generally affect drug diffusion through this membrane unless molecules are very big; in this particular case, size can reduce the rate of permeation [52]. The choroid is a thin and permeable membrane that is rich in melanin. Melanin has the ability to bind and retain many drugs hampering their entrance to the retina and inner tissues. Other drug-binding proteins, depending on the kinetic of binding/unbinding retention of drugs by melanin, can completely block the entrance or act as a reservoir for slow release [69]. Studies to assess the binding of oligonucleotides to melanin have yielded different results suggesting that at least some oligonucleotides bind to melanin reducing the rate of entrance to the retina; this is however not the case for all oligonucleotides [52, 70].

The main restriction to free permeation of molecules from systemic circulation to the eye is the blood–retinal barrier (BRB). The BRB is composed of the inner BRB and the outer BRB. The former includes the vessels of the retina, whereas the latter is constituted by the retinal pigment epithelium (RPE). Both barriers possess cells with well-developed intercellular junctions that control the permeation of substances through them. Larger molecules, such as proteins and nucleic acids, are mostly able to permeate through the choroid but have limited ability to cross the inner BRB; thus, drugs need to exit the choroid and penetrate the eye crossing the outer BRB. Crossing through the outer BRB usually requires high systemic doses increasing the likelihood of systemic side effects [71]. Delivery to the posterior segment of systemic or topically applied drugs requires thus crossing several biological barriers. Therefore, invasive administration procedures are frequently used to deliver drugs to the posterior segment. In addition, the outflow mechanisms of the eye rapidly remove drugs from the posterior chamber; thus, reaching clinically meaningful concentrations is challenging. Most of the programs developing siRNAs for eye conditions target the back of the eye; consequently, the route of administration is IVT injection. The concentration of siRNAs administered IVT is highest in the vitreous body, but they are also found in the RPE, choroid, and retina. Depending on the stability of the siRNAs, the compound can also be found in systemic circulation.

There are numerous reports describing strategies that can be of benefit for increasing the concentration of drugs in the back of the eye. Several nonbiodegradable (Retisert™, Illuvien™, and Vitrasert®) and one biodegradable (Ozurdex™) intravitreal ocular inserts are currently used in the clinical practice for delivering small molecules to the back of the eye. It is expected that these advances be soon incorporated into the pipelines of larger molecules such as proteins and oligonucleotides.

#### **5. Toxicology**

siRNAs are chemically synthesized oligonucleotides and are considered New Chemical Entities (NCEs) by the US and European Regulatory Authorities since 2009 when the European Commission excluded siRNAs from the definition of advanced medicinal products [72]. Toxicology assessment of RNAi-based drugs should be carried out following guidelines for NCEs, and the complete toxicology battery is usually performed following the recommenda‐ tions of the ICH M3 (R2) guideline [73]. The guideline recommends the assessment of toxicol‐ ogy in two species, a rodent and a nonrodent, at three dose levels and for a duration that should be similar or superior to the clinical trial to be carried out. This assessment should include acute or maximum tolerated toxicology studies and repeated-dose toxicity studies. Addition‐ ally, pharmacokinetics, safety pharmacology, genotoxicity, carcinogenicity, and specific toxicology studies should be carried out depending of the nature, indication, and route of administration of the product. On the other hand, some aspects of RNAi-based products are closer to new biological entities (NBEs) rather than NCEs; therefore, some of the requirements of the ICH S6 guideline also apply to the design of their developmental programs [74]. As such, a tailored toxicology assessment program should be designed combining the recommenda‐ tions outlined in the above-mentioned guidelines and the accumulated experience of numer‐ ous compounds tested in preclinical and clinical development.

Toxicology of ocular products depends on their biodistribution and on their biological activity. Moreover, the disease process, age, sex, or eye pigmentation are other potential factors affecting the toxicity profile of the ocular drug assessed. Additionally, the bioavailability of the RNAi compound will depend mainly on the route of ocular administration (topical versus injected) and on the physicochemical characteristics of the drug.

Toxicities arising from oligonucleotides, including siRNAs, can be classified into hybridiza‐ tion-dependent toxicities and hybridization-independent toxicities. Hybridization-dependent toxicities can be caused by (a) exaggerated pharmacology: excessive activity on the intended target or by (b) off-target effect: modulating gene expression of an unintended target by an RNAi-mediated mechanism. Hybridization-independent toxicities are often associated to the chemistry of the siRNA. Identified hybridization-independent toxicities include prolongation of activating partial thromboplastin time (aPTT), complement activation, and immunostimu‐ lation [75, 76, 77].

#### **5.1. General toxicology**

oligonucleotides bind to melanin reducing the rate of entrance to the retina; this is however

The main restriction to free permeation of molecules from systemic circulation to the eye is the blood–retinal barrier (BRB). The BRB is composed of the inner BRB and the outer BRB. The former includes the vessels of the retina, whereas the latter is constituted by the retinal pigment epithelium (RPE). Both barriers possess cells with well-developed intercellular junctions that control the permeation of substances through them. Larger molecules, such as proteins and nucleic acids, are mostly able to permeate through the choroid but have limited ability to cross the inner BRB; thus, drugs need to exit the choroid and penetrate the eye crossing the outer BRB. Crossing through the outer BRB usually requires high systemic doses increasing the likelihood of systemic side effects [71]. Delivery to the posterior segment of systemic or topically applied drugs requires thus crossing several biological barriers. Therefore, invasive administration procedures are frequently used to deliver drugs to the posterior segment. In addition, the outflow mechanisms of the eye rapidly remove drugs from the posterior chamber; thus, reaching clinically meaningful concentrations is challenging. Most of the programs developing siRNAs for eye conditions target the back of the eye; consequently, the route of administration is IVT injection. The concentration of siRNAs administered IVT is highest in the vitreous body, but they are also found in the RPE, choroid, and retina. Depending on the

stability of the siRNAs, the compound can also be found in systemic circulation.

There are numerous reports describing strategies that can be of benefit for increasing the concentration of drugs in the back of the eye. Several nonbiodegradable (Retisert™, Illuvien™, and Vitrasert®) and one biodegradable (Ozurdex™) intravitreal ocular inserts are currently used in the clinical practice for delivering small molecules to the back of the eye. It is expected that these advances be soon incorporated into the pipelines of larger molecules such as proteins

siRNAs are chemically synthesized oligonucleotides and are considered New Chemical Entities (NCEs) by the US and European Regulatory Authorities since 2009 when the European Commission excluded siRNAs from the definition of advanced medicinal products [72]. Toxicology assessment of RNAi-based drugs should be carried out following guidelines for NCEs, and the complete toxicology battery is usually performed following the recommenda‐ tions of the ICH M3 (R2) guideline [73]. The guideline recommends the assessment of toxicol‐ ogy in two species, a rodent and a nonrodent, at three dose levels and for a duration that should be similar or superior to the clinical trial to be carried out. This assessment should include acute or maximum tolerated toxicology studies and repeated-dose toxicity studies. Addition‐ ally, pharmacokinetics, safety pharmacology, genotoxicity, carcinogenicity, and specific toxicology studies should be carried out depending of the nature, indication, and route of administration of the product. On the other hand, some aspects of RNAi-based products are closer to new biological entities (NBEs) rather than NCEs; therefore, some of the requirements

not the case for all oligonucleotides [52, 70].

256 RNA Interference

and oligonucleotides.

**5. Toxicology**

Up to date, numerous siRNAs indicated for different eye conditions have entered clinical trials (Table 1); the administration route of four of these compounds is by IVT injection, whereas the remaining two compounds are topically administered in eye drops. The toxicology assessment of these products follows the traditional schedule for NCEs; this schedule entails general toxicology studies in two species, a rodent and a nonrodent species of variable length. Most programs up to date have used NHP as the nonrodent species; this is because siRNAs are species specific, and it is likely that the assessment of toxicology was performed in the only species in which the compound was pharmacologically active. The rabbit is very frequently used to assess toxicology of compounds under development for eye conditions. Many sponsors of programs using NHP or dog as nonrodent species chose to use the rabbit as second species, although this animal is not a rodent per se. Reasons behind this choice include the similarity of the volume of the eye to that of humans and the difficulty of administering controlled doses to smaller animals. This is particularly relevant when the compound is administered by IVT injection. This rationale has also been followed for developing siRNAs for eye conditions; only in one case, PF-04523655, rats were used as the rodent species, and the rest of the programs developing siRNAs for eye indications used the rabbit (New Zealand White rabbits or Dutch Belted rabbit) as second species for toxicology assessment.

Most programs developing siRNAs for eye conditions include acute/maximum tolerated dose and repeat-dose toxicology studies. The length of these studies is determined by the indication, stage of development, and envisioned duration of treatment. In addition, most programs do not only perform toxicology studies using the envisaged route of administration but also include studies using intravenous route to challenge the systemic exposure to the drug and assess potential dose limitations and target organs.

#### **5.2. Genotoxicity**

As mentioned in the previous section, both the ICH M3 (R2) and the ICH S6 guidelines apply to programs developing siRNAs [73]. The ICH S6 states that the range and type of genotoxicity studies routinely conducted for NCEs are usually not applicable to NBEs; pointing out that performance of these studies is only required when there is a cause of concern. The European Medicines Agency (EMA) issued a reflection paper on the assessment of the genotoxic potential of antisense oligodeoxynucleotides in January 2005. This paper recommends addressing at least two issues in regards to oligonucleotides which may indicate a cause of genotoxic concern: (a) analyzing the potential of incorporation of phosphorothioated (PS) oligonucleotides into the DNA and (b) addressing the potential of triplex formation of oligonucleotides with the DNA fiber [78]. Several years of experience with siRNAs indicate that full-length molecules are very unlikely to interact with the DNA. Thus, the potential cause of concern may arise from the genotoxic potential of metabolites or chemical contaminants. The metabolism of nonmo‐ dified oligonucleotides yields naturally occurring nucleotides that are subsequently incorpo‐ rated to the natural degradation pathways of endogenous nucleic acids; thus, toxicities derived from these degradation products are not expected. Modified oligonucleotides, on the other hand, incorporate very frequently backbone modifications to reduce nuclease activity and improve other pharmaceutical properties of the molecule. The most commonly used backbone modification is the replacement of a nonbridging oxygen on the backbone between two ribonucleotides with a sulfur to create a PS linkage [79]. Extensive genotoxicity studies performed with Vitravene, a PS antisense oligonucleotide administered by IVT injection, indicate that oligonucleotides with a PS backbone do not pose genotoxic potential [80, 81]. These results are in line with those obtained in the analysis of over 30 compounds studied in the standard battery, all of which have yielded negative results. Other modified nucleotides could potentially be incorporated into nucleotide pools and be thereafter used to synthesize DNA. The standard battery of tests would detect eventual damaging potential of theses degradation products.

The EMA reflection paper also recommends assessing the potential of triplex formation with the DNA fiber. For this to happen, siRNA molecules would have to enter the nucleus of the cell and their structure should include an uninterrupted homopurine stretch of at least 10–12 base pairs that should be homologous a given region of the DNA. In silico design of siRNAs usually addresses these issues and candidates with the ability of forming triplex are avoided prior to lead selection.

#### **5.3. Carcinogenicity**

Standard carcinogenicity studies are generally not required for NBEs, but these studies may be required for siRNAs depending on their chemical structure, clinical dosing, patient population, or biological activity. If the in vitro test genotoxic studies indicate that there is cause of concern for carcinogenic potential, further studies should be required in relevant models.

For RNAi products under development for eye conditions, the systemic bioavailability of these products is usually very low, and a waiver to perform these studies may be justified. Strategies should be discussed in a case-by-case base with the competent health authorities.

#### **5.4. Reproductive and developmental toxicity**

**5.2. Genotoxicity**

258 RNA Interference

degradation products.

prior to lead selection.

**5.3. Carcinogenicity**

models.

As mentioned in the previous section, both the ICH M3 (R2) and the ICH S6 guidelines apply to programs developing siRNAs [73]. The ICH S6 states that the range and type of genotoxicity studies routinely conducted for NCEs are usually not applicable to NBEs; pointing out that performance of these studies is only required when there is a cause of concern. The European Medicines Agency (EMA) issued a reflection paper on the assessment of the genotoxic potential of antisense oligodeoxynucleotides in January 2005. This paper recommends addressing at least two issues in regards to oligonucleotides which may indicate a cause of genotoxic concern: (a) analyzing the potential of incorporation of phosphorothioated (PS) oligonucleotides into the DNA and (b) addressing the potential of triplex formation of oligonucleotides with the DNA fiber [78]. Several years of experience with siRNAs indicate that full-length molecules are very unlikely to interact with the DNA. Thus, the potential cause of concern may arise from the genotoxic potential of metabolites or chemical contaminants. The metabolism of nonmo‐ dified oligonucleotides yields naturally occurring nucleotides that are subsequently incorpo‐ rated to the natural degradation pathways of endogenous nucleic acids; thus, toxicities derived from these degradation products are not expected. Modified oligonucleotides, on the other hand, incorporate very frequently backbone modifications to reduce nuclease activity and improve other pharmaceutical properties of the molecule. The most commonly used backbone modification is the replacement of a nonbridging oxygen on the backbone between two ribonucleotides with a sulfur to create a PS linkage [79]. Extensive genotoxicity studies performed with Vitravene, a PS antisense oligonucleotide administered by IVT injection, indicate that oligonucleotides with a PS backbone do not pose genotoxic potential [80, 81]. These results are in line with those obtained in the analysis of over 30 compounds studied in the standard battery, all of which have yielded negative results. Other modified nucleotides could potentially be incorporated into nucleotide pools and be thereafter used to synthesize DNA. The standard battery of tests would detect eventual damaging potential of theses

The EMA reflection paper also recommends assessing the potential of triplex formation with the DNA fiber. For this to happen, siRNA molecules would have to enter the nucleus of the cell and their structure should include an uninterrupted homopurine stretch of at least 10–12 base pairs that should be homologous a given region of the DNA. In silico design of siRNAs usually addresses these issues and candidates with the ability of forming triplex are avoided

Standard carcinogenicity studies are generally not required for NBEs, but these studies may be required for siRNAs depending on their chemical structure, clinical dosing, patient population, or biological activity. If the in vitro test genotoxic studies indicate that there is cause of concern for carcinogenic potential, further studies should be required in relevant

The assessment of reproductive and developmental toxicity is required to support the use of a given pharmaceutical in pregnant women, women of childbearing potential, or children. These studies are regulated by the ICH S5 guideline [82], which recommends assessing the effect of drugs on all phases of the reproductive cycle. These recommendations apply to siRNAbased products. Nevertheless, due to the unique features of these compounds, a case-by-case approach should be followed for each product, and the requirements for these studies should be discussed with the competent authorities. The target, indication, chemical modifications, and systemic bioavailability of the RNAi-based drugs are features that may influence in the nature of the required studies.

Because the toxicity of siRNAs can be caused by exaggerated pharmacology whenever reproductive toxicity studies are required, they should be performed in a pharmacologically active species. Standard reproductive toxicity species in rodents or rabbits can give information on toxicity related to chemical structure. However, if the compound is not active in these species or if the biological activity is not deemed to be equivalent to the foreseen activity in humans, the assessment of reproductive risk may be conducted using an active analog or in a nonrodent species in which the compound has biological activity. If the former strategy is chosen, the toxicity and toxicokinetic profile of the surrogate should be taken into account when interpreting the results. If the compound is only active in NHPs, studies should only be performed in cases where there is cause for concern. In these particular cases, the number of animals should be optimized, and a combined enhanced pre-and postnatal developmental study can be performed as recommended for NBEs. In NHP studies, the assessment of reproductive toxicology is usually studied by histopathologic examination of the reproductive organs as part of the general toxicology studies of at least three months. The timing of reproductive and developmental studies depends on when women of childbearing potential are to be included in clinical trials. If NHPs are required for the assessment, the timing is more flexible due to the length and complexity of the studies [83].

#### **5.5. Local tolerance**

Local tolerance studies are required for all topically administered drugs. In most cases, the potential adverse events caused by local tolerance issues are evaluated in the single or repeated-dose toxicology studies, reducing the number of animals required for the program.

#### **5.6. Safety pharmacology**

According to ICH S7A, safety pharmacology studies can be reduced or eliminated for locally applied products as well as for NBEs that achieve highly specific receptor targeting [84]. For siRNAs under development for ophthalmology indications, separate safety studies are not usually required; instead, functional safety end points are incorporated into the repeated-dose toxicity studies. If the results of the toxicology studies indicate that there is cause of concern, separate safety pharmacology studies should be performed.

#### **6. Programs in development and future ahead**

Table 1 summarizes the status of siRNA-based therapies under development for ocular conditions. As mentioned in Section 1, the eye offers multiple advantages for developing innovative therapies; therefore, studies in the eye pioneered the field of siRNA therapeutics. The first siRNA to enter clinical development for an ophthalmology indication was bevasiranib in 2004 shortly followed by sirna-027. Bevasiranib targeted VEGFA, whereas sirna-027 targeted VEGFR1. These compounds were being developed for the treatment of AMD as both showed a dose-dependent inhibition of experimental CNV in animal models that correlated with knockdown of their respective target genes [10, 85]. As mentioned in Section 1, in 2008 Kleinman and coworkers published a study demonstrating that the effect of siRNAs targeting VEGF and VEGFR on CNV was not mediated by an on-target effect but by activation of TLR3 [12]. The results of these studies indicated that the effect of the compounds on CNV was sequence-independent and mediated by siRNAs of 21 base pairs or longer. The study also showed that the internalization of the siRNAs was not required for the inhibition of CNV as cells of the RPE abundantly express TLR3 on the cytoplasmic surface. The authors used several sequences, including those of the siRNAs undergoing clinical trials at the time, to point out that the inhibition of CNV by both bevasiranib and sirna-027 was mediated through an offtarget effect. A subsequent study by the same group showed that activation of TLR3 by IVT siRNAs led to caspase-3-mediated degeneration of the retinal pigment epithelium (RPE) questioning the safety of these compounds as therapeutics for back of the eye diseases [13].

The clinical development of bevasiranib was halted in 2007 and of sirna-027 in 2009 both as a result of not reaching or being unlikely to reach their respective efficacy end points in phase III trials.

The findings of Kleinman and coworkers boosted research on alternative designs that were not able to activate TLR3, and as result, a new generation of compounds is currently under‐ going clinical trials. Currently, the most advanced siRNAs-based programs for ocular indica‐ tions are Quark's QPI-007 and Sylentis' bamosiran (SYL040012). QPI-007 is a 19-nt modified siRNA-targeting caspase 2 currently in phase II/III for the treatment of nonarteritic anterior ischemic optic neuropathy (NAION) [86]. QPI-1007 has shown to be safe when IVT injected to animal models and humans. The ongoing phase II/III trial for this compound analyzes the potential of multiple IVT doses to improve visual acuity in patients suffering NAION [87, 88]. Bamosiran is a canonical-designed naked siRNA-targeting β2-adrenergic receptor (ADRB2) under development for the treatment of increased IOP associated to glaucoma [64, 89–91]. Glaucoma is a degenerative, chronic disease of the optic nerve that can lead to blindness if left untreated [92]. The mechanistic details of optic nerve degeneration observed in glaucoma are yet to be fully detailed, but it is well established that reduction of intraocular pressure avoids development of the disease. ADRB2 controls the production and release of aqueous humor. The aqueous humor is responsible for maintaining optimal IOP. Treatment with topic betablockers has shown to efficiently reduce intraocular pressure, but currently approved betablockers are small molecules and are thus able to reach systemic circulation and systemic organs were they cause unwanted effects. The rationale behind bamosiran is developing a locally active compound that efficiently knocks down ADRB2 in the eye but that is not able to reach systemic tissues reducing the likelihood of side effects. The compound is administered in eye drops and has been shown to be well tolerated in animal models and humans [64, 93]. Three different doses of bamosiran are currently being studied in an active controlled phase IIb trial. Previous clinical trials with this compound have shown promising results in healthy individuals and patients with ocular hypertension [91, 93].

usually required; instead, functional safety end points are incorporated into the repeated-dose toxicity studies. If the results of the toxicology studies indicate that there is cause of concern,

Table 1 summarizes the status of siRNA-based therapies under development for ocular conditions. As mentioned in Section 1, the eye offers multiple advantages for developing innovative therapies; therefore, studies in the eye pioneered the field of siRNA therapeutics. The first siRNA to enter clinical development for an ophthalmology indication was bevasiranib in 2004 shortly followed by sirna-027. Bevasiranib targeted VEGFA, whereas sirna-027 targeted VEGFR1. These compounds were being developed for the treatment of AMD as both showed a dose-dependent inhibition of experimental CNV in animal models that correlated with knockdown of their respective target genes [10, 85]. As mentioned in Section 1, in 2008 Kleinman and coworkers published a study demonstrating that the effect of siRNAs targeting VEGF and VEGFR on CNV was not mediated by an on-target effect but by activation of TLR3 [12]. The results of these studies indicated that the effect of the compounds on CNV was sequence-independent and mediated by siRNAs of 21 base pairs or longer. The study also showed that the internalization of the siRNAs was not required for the inhibition of CNV as cells of the RPE abundantly express TLR3 on the cytoplasmic surface. The authors used several sequences, including those of the siRNAs undergoing clinical trials at the time, to point out that the inhibition of CNV by both bevasiranib and sirna-027 was mediated through an offtarget effect. A subsequent study by the same group showed that activation of TLR3 by IVT siRNAs led to caspase-3-mediated degeneration of the retinal pigment epithelium (RPE) questioning the safety of these compounds as therapeutics for back of the eye diseases [13].

The clinical development of bevasiranib was halted in 2007 and of sirna-027 in 2009 both as a result of not reaching or being unlikely to reach their respective efficacy end points in phase

The findings of Kleinman and coworkers boosted research on alternative designs that were not able to activate TLR3, and as result, a new generation of compounds is currently under‐ going clinical trials. Currently, the most advanced siRNAs-based programs for ocular indica‐ tions are Quark's QPI-007 and Sylentis' bamosiran (SYL040012). QPI-007 is a 19-nt modified siRNA-targeting caspase 2 currently in phase II/III for the treatment of nonarteritic anterior ischemic optic neuropathy (NAION) [86]. QPI-1007 has shown to be safe when IVT injected to animal models and humans. The ongoing phase II/III trial for this compound analyzes the potential of multiple IVT doses to improve visual acuity in patients suffering NAION [87, 88]. Bamosiran is a canonical-designed naked siRNA-targeting β2-adrenergic receptor (ADRB2) under development for the treatment of increased IOP associated to glaucoma [64, 89–91]. Glaucoma is a degenerative, chronic disease of the optic nerve that can lead to blindness if left untreated [92]. The mechanistic details of optic nerve degeneration observed in glaucoma are yet to be fully detailed, but it is well established that reduction of intraocular pressure avoids

separate safety pharmacology studies should be performed.

**6. Programs in development and future ahead**

III trials.

260 RNA Interference

SYL1001 is a naked 19-bp siRNA-targeting transient receptor potential vanilloid-1 (TRPV1) for the treatment of ocular pain. TRPV1 is a cation channel permeable to calcium activated by heat, low pH, and capsaicin among other signals. This receptor is present in several structures of the eye where it has been related, among other roles, to nociception [94]. SYL1001 has shown to be safe when administered in eye drops to animals and humans and to have analgesic effect in the capsaicin-induced eye pain model. The compound is currently undergoing a phase I/II for the treatment of ocular pain associated to dry eye disease, a condition for which no specific treatment currently exists [89].

PF-655 is a chemically stabilized siRNA-targeting RTP801, a stress-induced adaptor protein that inhibits mTOR function upstream to TSC1/TSC2 complex in response to a variety of stresses. Expression of RTP801 is upregulated in response to ischemia, hypoxia, and/or oxidative stress. Intravitreal injection of PF-655 in preclinical animal models of laser-induced CNV leads to silencing of RTP801 via a RNAi mechanism without TLR activation and reduction of CNV volume, vessel leakage, and infiltration of inflammatory cells into the choroid [95–97]. This compound has undergone phase II clinical trials for the treatment of diabetic macular edema and wet AMD. Treatment with PF-655 of patients with diabetic macular edema over a period of 12 months caused a dose-dependent improvement in visual acuity compared to the visual acuity observed in patients treated with laser photocoagulation [98]. A subsequent phase IIb trial was conducted with a new set of doses but terminated because the primary end point was not likely to be achieved. The compound was thereafter tested in combination with ranibizumab, a monoclonal antibody fragment that targets VEGF and is the current gold standards for treatment of the disease. The results of this study have not yet been disclosed. PF-655 has also been studied in patients suffering wet AMD. In this indication, the compound did not show improvement as a single agent or in combination with ranibizumab in mean visual acuity after 3 months of dosing.

Self-delivery rxRNAs (sd-rxRNAs) incorporate 2'-F and 2-'O-Me modifications and a sterol conjugate on the sense strand with the goal of improving stability and cellular uptake. These compounds have a 19-nt antisense strand and a sense strand usually shorter than 15 nt resulting in an asymmetric duplex with a phosphorothioated single-stranded tail on the antisense strand [99]. These compounds have been tested in vitro where they have shown to be able to induce target knockdown in different cell lines. In vivo analysis of their activity showed these compounds are readily taken-up by retinal cells and that the compound is evenly distributed throughout the mouse retina. Several of these compounds are under development for different eye conditions and are expected to enter clinical development shortly.

SYL116011 is a naked 19-bp siRNA targeting the calcium release-activated calcium modulator 1 (ORAI1). Store-operated Ca2+ entry (SOCE) is activated in response to depletion of endo‐ plasmic reticulum Ca2+ pools. Activation of SOCE induces Ca2+ entry from extracellular compartments, and this is mediated by store-operated CRAC channels. CRAC channels are composed of calcium sensing proteins called STIM (stromal interaction molecule) and poreforming subunits named ORAI [100]. Mammalian cells have three ORAI isoforms: ORAI1, ORAI 2, and ORAI3; although ORAI2 and 3 fulfill the same role as ORAI1, the Ca2+ currents generated by these proteins are around two- to threefold smaller than the ones generated by ORAI1 [101]. There is growing evidence that indicates that short-term and long-term activation of immune cells in allergic responses is mediated by influx of Ca2+ to immune cells from the extracellular compartment. Short-term responses include the degranulation of mast cells and the activation of effector cytolitic T cells. Indeed, mast cells lacking either STIM1 or ORAI1 show a considerable defect in degranulation [102, 103]. Long-term responses involve the modulation of gene expression that controls B and T cell proliferation and differentiation. SYL116011 is being developed for the treatment of ocular allergies and has shown to reduce immediate clinical signs in a mouse model of ragweed pollen-induced ocular allergy. The decrease in clinical signs was accompanied by a reduction in the number of infiltrating eosinophils in the conjunctiva and reduction of allergy biomarkers.

TT-211 is an AAV‐encapsidated construct that expresses a single shRNA modeled into a miRNA backbone that inhibits the expression of VEGF‐A for the treatment of wet AMD. VEGFa protein is responsible initiating a signaling cascade that stimulates the growth of new blood vessels, a hallmark of wet AMD. TT-231 is a second-generation candidate designed to express three shRNAs, which target three different genes, VEGF receptor 2, PDGF-β, and human complement factor B, proteins that play a major role in the progression of wet AMD. Both these compounds are yet in a preclinical phase; IND filing is planned for 2017.

STP601 is a multitargeted siRNA cocktail nanoparticle formulation administered by IVT injection under development for treatment of wet AMD, proliferative diabetic retinopathy, and herpetic stromal keratitis. The cocktail includes three 25-mer siRNA duplexes targeting VEGF, VEGFR1, and VEGFR2. Inhibiting this clinically validated pathway at the endothelial cells lining the interior of the growing blood vessels is thought to halt the progression of AMD. This product is currently in preclinical stage.

AQA001 is a single-stranded long chain nonmodified ribonucleotide connected by a prolinederived linker that self-anneals to form a shot-hairpin structure within the molecule. The compound targeting periostin acts through an RNAi mechanism and is being developed for the treatment of diabetic retinopathy. The compound has shown positive result in a proof-ofconcept study of CNV [104].


**Table 1.** siRNAs in development for ocular indication.

#### **7. Conclusion**

showed these compounds are readily taken-up by retinal cells and that the compound is evenly distributed throughout the mouse retina. Several of these compounds are under development

SYL116011 is a naked 19-bp siRNA targeting the calcium release-activated calcium modulator 1 (ORAI1). Store-operated Ca2+ entry (SOCE) is activated in response to depletion of endo‐ plasmic reticulum Ca2+ pools. Activation of SOCE induces Ca2+ entry from extracellular compartments, and this is mediated by store-operated CRAC channels. CRAC channels are composed of calcium sensing proteins called STIM (stromal interaction molecule) and poreforming subunits named ORAI [100]. Mammalian cells have three ORAI isoforms: ORAI1, ORAI 2, and ORAI3; although ORAI2 and 3 fulfill the same role as ORAI1, the Ca2+ currents generated by these proteins are around two- to threefold smaller than the ones generated by ORAI1 [101]. There is growing evidence that indicates that short-term and long-term activation of immune cells in allergic responses is mediated by influx of Ca2+ to immune cells from the extracellular compartment. Short-term responses include the degranulation of mast cells and the activation of effector cytolitic T cells. Indeed, mast cells lacking either STIM1 or ORAI1 show a considerable defect in degranulation [102, 103]. Long-term responses involve the modulation of gene expression that controls B and T cell proliferation and differentiation. SYL116011 is being developed for the treatment of ocular allergies and has shown to reduce immediate clinical signs in a mouse model of ragweed pollen-induced ocular allergy. The decrease in clinical signs was accompanied by a reduction in the number of infiltrating

TT-211 is an AAV‐encapsidated construct that expresses a single shRNA modeled into a miRNA backbone that inhibits the expression of VEGF‐A for the treatment of wet AMD. VEGFa protein is responsible initiating a signaling cascade that stimulates the growth of new blood vessels, a hallmark of wet AMD. TT-231 is a second-generation candidate designed to express three shRNAs, which target three different genes, VEGF receptor 2, PDGF-β, and human complement factor B, proteins that play a major role in the progression of wet AMD. Both these

STP601 is a multitargeted siRNA cocktail nanoparticle formulation administered by IVT injection under development for treatment of wet AMD, proliferative diabetic retinopathy, and herpetic stromal keratitis. The cocktail includes three 25-mer siRNA duplexes targeting VEGF, VEGFR1, and VEGFR2. Inhibiting this clinically validated pathway at the endothelial cells lining the interior of the growing blood vessels is thought to halt the progression of AMD.

AQA001 is a single-stranded long chain nonmodified ribonucleotide connected by a prolinederived linker that self-anneals to form a shot-hairpin structure within the molecule. The compound targeting periostin acts through an RNAi mechanism and is being developed for the treatment of diabetic retinopathy. The compound has shown positive result in a proof-of-

for different eye conditions and are expected to enter clinical development shortly.

eosinophils in the conjunctiva and reduction of allergy biomarkers.

compounds are yet in a preclinical phase; IND filing is planned for 2017.

This product is currently in preclinical stage.

concept study of CNV [104].

262 RNA Interference

RNA interference is on the verge of becoming a new class of therapeutics [105]. The field of ophthalmology has played a major role in advancing siRNAs from laboratory tools to the clinic. In the last few years, significant advances have been made in the understanding of how these molecules enter and exert its action in the eye and in the identification of the main hurdles that still need to be addressed. The introduction of chemical modifications as well as the under‐ standing of the immune activation in the eye has significantly improved the pharmaceutical properties of compounds for eye conditions. However, the following years will tell whether improvements on these molecules are enough to be of therapeutic value in the field of ophthalmology or not.

#### **Author details**

Tamara Martínez# , Maria Victoria González# , Beatriz Vargas, Ana Isabel Jiménez and Covadonga Pañeda\*

\*Address all correspondence to: cpaneda@sylentis.com

Sylentis SAU, Tres Cantos, Madrid, Spain
