**5. Engineering exosomes as therapeutics for breast cancer**

Exosomes are emerging as promising therapeutic agents because of their role in tumor-related processes and their ability to deliver their cargo, such as proteins, lipids, and nucleic acids, into the tumor sites. However, their full clinical applicability has not yet been realized. This is because of many factors, including low yield and relatively low percentage loading to the therapeutic moiety. As such, new approaches for mass production and enhancement of the percent loading need to be explored. In general, these approaches are divided into two categories: passive and active loading, which are discussed in detail in the following sections.

Direct modification, also known as non-cell-based loading or exogenous loading, refers to the direct loading of therapeutic moieties such as siRNA, miRNA, drugs, and proteins after the isolation and purification of exosomes from the cells. This may encompass a series of procedures such as incubation, freeze-thaw cycles, sonication, and electroporation, and thus can further be categorized into passive and active loading. Passive loading includes loading of therapeutic moiety into exosomes by diffusion; on the other hand, active loading includes disrupting the exosomal membrane by electroporation, sonication, or freeze thawing, thus allowing the therapeutic moiety to enter into these vesicles. In passive drug loading, exosomes are incubated with drugs and allowed to diffuse into vesicles along a concentration gradient. Because exosomes consist of a lipid bilayer, the drug loading efficiency depends largely on the hydrophobicity of the drugs. Dong et al. loaded curcumin into milk exosomes by incubating at 4°C overnight and reported 70.46% drug loading using an incubation method [51]. Similarly, Sun et al. incorporated curcumin into exosomes derived from a mouse lymphoma cell line by incubating in PBS at room temperature (22°C) for 5 min and showed a binding capacity of 2.9 g curcumin to 1 g of exosomes [52]. Sun et al. packaged Cho-miR159 (cholesterol-modified miRNA 159) along with doxorubicin into exosomes derived from the human monocytic cell line THP-1 by incubating in PBS at 37°C to deliver to triple-negative breast cancer cells [53]. Linezolid was incorporated into exosomes derived from the mouse macrophage cell line RAW 264.7, by mixing both and incubating at 37°C for 1 h, resulting in ~5% drug loading. The exosomal formulation of linezolid was more effective against MRSA infections than the free drug [54]. Although several studies have reported the use of incubation with exosomes for drug or any therapeutic agent loading, it often suffers from issues of low percent drug loading, urging a requirement for improved methods for higher drug loading percent. Another method (less common) of passive loading includes incubating the exosome donor cells with the drugs/therapeutic agents. First, the donor cells are exposed to drugs or therapeutic agents, followed by isolation of released exosomes (supposedly) containing the loaded drugs or therapeutic agents. This method was used in a study by Pascucci et al., wherein they exposed bone marrow-derived mesenchymal stromal cells (MSCs) with a very high concentration of paclitaxel followed by incubation at 37°C for 24 h. After incubation, the cells were washed twice with PBS, trypsinized, and seeded in a fresh flask for 48 h. After 48 h, cell-conditioned medium was collected to isolate exosomes containing paclitaxel. They found that MSC-PTX-derived exosomes had a greater inhibitory effect on tumor cell proliferation (**Figure 2**) [55].

For active cargo loading, the exosomal membrane is temporarily disrupted using different methods and then restored once the drug/therapeutic agent was loaded. These methods may include sonication, extrusion, freeze-thawing, electroporation, use of membrane permeabilizers, conjugation using click chemistry, and antibodies against exosomal surface proteins. Electroporation uses an electric field to generate small pores in the exosomal membrane to disturb the phospholipid bilayer of exosomes. Drug/therapeutic agents can enter these vesicles via the generated pores. Once they entered, the pores were closed to recover the exosomal membrane integrity. This method has mostly been used to encapsulate siRNA or miRNA into exosomes and has been reported to enhance the percent loading compared to the simple diffusion method. Jia et al. loaded exosomes derived from RAW 264.7 cells with curcumin and superparamagnetic iron oxide nanoparticles (SPIONs) synchronously using optimal electroporation conditions of 400 V, 150 μF, and 1 ms discharge time. They observed that electroporation had no effect on the membrane integrity of exosomes and

#### **Figure 2.**

*Illustration representing different methods of cargo loading in exosomes. (A) Passive cargo loading is achieved by incubating the therapeutic moiety directly with isolated exosomes or by exposing to the exosome secreting donor cells followed by isolation of loaded exosomes. (B) Active cargo loading methods include use of physical treatments to disrupt the membrane integrity thus allowing entry of cargo in the interiors of exosomes. These treatments include sonication, electroporation, freeze thawing cycles and extrusion method.*

efficiently encapsulated curcumin and SPIONs [56]. Similarly, Jia et al. incorporated doxorubicin into exosomes isolated from MDA-MB-231 and HCT-116 cell lines using an electroporation method, which resulted in ~1.5% drug loading [57]. According to published studies, although electroporation enhanced the percentage of drug loading in exosomes compared to the incubation method, it was still low. Therefore, scientists have employed sonication methods to load cargo more efficiently. The mechanical shear force of a sonicator/homogenizer probe is applied to disrupt the membrane integrity of exosomes, thus allowing the mixed drug/therapeutic agent to enter into the exosomes. In 2017, Kim et al. compared the incubation, electroporation, and sonication method of cargo loading in RAW264.7 cell derived exosomes to develop an exosomal formulation of paclitaxel (PTX). For the incubation method, the authors mixed and incubated PTX with exosomes at 37°C for 1 h. Using electroporation, exosomes and PTX were added to a pre-chilled electroporation cuvette and applied at 1000 kV for 5 ms followed by incubation at 37°C for half an hour to fully recover the exosome membrane. For sonication*,* the PTX-exosome mixture was sonicated at 20% amplitude, given 6 cycles of 30 s on/off for 3 min and a 2 min cooling period between each cycle. After sonication, the solution was incubated at 37°C for 1 h to fully recover the membrane of the exosomes. They showed the highest percent drug loading of ~28% using sonication followed by ~5% using electroporation and the lowest at ~1.4% with the incubation method [58].

In the extrusion method, exosomes mixed with the drug are passed through a syringe-based lipid extruder with a membrane ranging from to 10 to 400 nm pore size. In this process, the membrane of exosomes is disrupted by the extensive mechanical force of the extruder. In a study by Kim et al. when breast cancer cellderived exosomes loaded with porphyrin were extruded, it altered the surface charge of blank exosomes, leading to cytotoxic effects [59]. On the other hand, in another study by Fuhrmann et al., loading cargo in exosomes using the extrusion method did not render them cytotoxic [60]. In the freeze-thaw method, the drug was first incubated with exosomes at ambient temperature and then frozen at −80°C. The mixture was then repeatedly thawed at room temperature to ensure drug loading into these vesicles. The main disadvantage is that this method often leads to particle aggregation, resulting in a wide size distribution. This method has also been reported to result in a lower percent drug loading than other methods, such as sonication.
