**6.3 Linking of two overlapping BACs**

There are many large human genes which are of the same order of size, or larger, than the average insert size of the BAC libraries and for these it is often difficult to find a single BAC spanning the entire gene with all its associated controlling elements. Gene therapy using the genomic loci of such genes would require the assembly of different sequences into a single BAC clone by linking together all available overlapping BAC clones spanning the desired region. Recombineering mediated by the Red system from the λ-prophage has been used to link two overlapping BACs (Kotzamanis & Huxley, 2004; Zhang & Huang, 2003) and linking has been shown to be precise without causing any rearrangements, including shifting of the reading frame of the therapeutic gene (Kotzamanis et al., 2009). As shown in Figure 3, the method comprises two rounds of homologous recombination to link the inserts of two overlapping BACs. In the first round, the inserts of the BACs are subcloned into modified BAC vectors (pBACLink vectors linearized by *Not*I) by homologous recombination at regions indicated as HomA, HomB and HomC (which are PCR amplified and cloned into the pBACLink vectors prior to their linearization). In the second round, one of the modified BACs is linearized by the rare cutting enzyme I-*Ppo*I and introduced into recombination efficient bacteria containing the other modified BAC, resulting in recombination at HomB and Cma (part of the chloramphenicol resistance gene present on all BACs) and linking of the two inserts in a single BAC. More overlapping BAC inserts can be added by alternating use of the two pBACLink vectors described in the study (Kotzamanis & Huxley, 2004).

Non-Viral Gene Therapy Vectors Carrying Genomic Constructs 17

In previous preclinical and clinical studies where *CFTR* cDNA-heterologous promoter systems and different viral vectors were used for the delivery and expression of the transgene, some expression has been shown in transgenic mice and low levels of transient correction of Cl- ion transport deficiency has been observed in patients but no significant

Due to the strict regulation of expression of the *CFTR* gene at specific developmental stages and in specific tissues, controlled by regulatory elements found before, after and within the gene (McCarthy & Harris, 2005), the use of constructs covering the whole genomic locus of the gene may have a better therapeutic potential for Cystic Fibrosis. To date, the only transgene that has fully restored transgenic mice, which did not express endogenous CFTR and would normally die, is the intact gene present on a YAC of approximately 300 kb in length (Manson et al., 1997). However, YAC vectors have the disadvantage of being difficult to shuttle between cells and are inherently unstable and therefore have been replaced by BACs. The *CFTR* gene is one of the large human genes that have not been found to be contained intact in any of the sequenced BACs available from the Human Genome Project. For this reason, the technology described in section 6.3 was developed and used to construct a BAC vector carrying the whole *CFTR* gene and the associated regulatory elements (Kotzamanis et al., 2009). Successful transcription of the gene to a correctly spliced mRNA has been demonstrated in a mouse cell line. This BAC is the only CFTR genomic system available on a convenient vector and may be the basis for non viral

Viral approaches to gene therapy for Cystic Fibrosis suffer from gene delivery barriers such as absence of viral receptors in the respiratory epithelium and safety concerns such as provocation of an inflammatory response. This makes either *in vivo* or *ex vivo* non viral gene therapy an attractive avenue of research. However, several issues need to be solved before any clinical application. For instance, the *in vivo* delivery of non viral vectors is limited by the low efficiency, which becomes lower when bigger constructs are used. The demonstration that bone marrow-derived MSCs were able to differentiate to several types of cells including airway epithelial (Wang et al., 2005) indicated a potential application in an *ex vivo* approach but this is limited by poor data on how the *ex vivo* corrected MSCs can be

Ideally, genetic manipulation with the CFTR BAC whether in the form of *in vivo* or *ex vivo* gene therapy would have to confer permanent transgene expression to avoid repeated gene or cell administration, respectively. In this regard, any of the systems that allow episomal maintenance or targeted integration at non-associated with carcinogenesis regions (described in sections 4 and 5) would have to be added to the CFTR BAC. The methodology required to add small sequences such as the *OriP*/EBNA-1 or the *S/MAR* elements, or large sequences such as the 70-kb alphoid array onto the CFTR BAC is available (see sections 6.1 and 6.2). Moreover, several methods for either *in vivo* or *ex vivo* delivery of the final construct to the respiratory epithelium have been developed and are available for use with

Non-viral gene therapy using the entire genomic locus of the therapeutic gene has two advantages over viral cDNA gene therapy; it is not associated with unwanted immune responses and can confer controlled levels of transgene expression in correct time and

clinical improvement has been achieved (Griesenbach & Alton, 2009).

administered and engrafted in the lung of Cystic Fibrosis patients.

the genomic CFTR-locus containing BAC (see section 3).

**8. Conclusion** 

gene therapy for Cystic Fibrosis.

Fig. 3. Linking of two overlapping BACs into a single larger BAC
