**4. Peptide-derived vectors for gene therapy**

Peptide chains can be fabricated from any of the 20 naturally occurring L-amino acids, which are referred to by their single- or three-lettered code (e.g., R or arg for arginine). Peptides are thus biocompatible and often degradable. Peptide synthesis also does not involve harmful catalysts, which is a concern in the synthesis of some polymeric carriers. Furthermore, synthesis can now be automated, courtesy of advances in solid phase peptide synthesis, which makes the manufacturing process amendable to up-scaling.

Before peptides were seen as proper DNA carriers, studies conducted in the 1960s with histones had already suggested that cationic amino acids such as lysine and arginine can be useful (Akinrimisi et al., 1965). Today, peptide vectors are given fanciful names such as "cell penetrating peptides" (CPP) or "protein transduction domains" (PTD) to celebrate their ability to efficiently penetrate the plasma membrane and mediate the entry of nucleic acids or other macromolecules. Such CPP can be derived from proteins existing in nature (e.g., viral proteins or venom proteins of bees and wasps) or designed *de novo*. There is little in common among CPP and the only unifying theme seems to be the significant presence of cationic residues (mainly lysine and arginine). Amphipathicity has been suggested to be another common feature. However, this is only true for most CPP – oligoarginine being an exception. There are two methods by which peptide vectors carry their cargoes. Nucleic acids are usually non-covalently (electrostatically) attached, while proteins and other macromolecules are typically covalently coupled (chemical crosslinking or by plasmid fusion). The advantages of electrostatic attachment include convenience and the largely unaltered chemical properties of the cargo. However, charge interactions are non-specific and excess peptides are usually needed to completely bind the plasmid. On the other hand, covalent attachment requires chemical modification of the cargo and usually results in stable complexes which, as discussed earlier, may not be desirable. In the following sections, we will review the key classes of peptide vectors and provide non-exhaustive examples of strategies that have been used to improve their efficiency as vectors.

1995) and is frequently used in its branched, high molecular weight (usually 25 kDa) form. Together with liposomes, they are widely acknowledged to be the best non-viral vectors currently available (Putnam, 2006) and frequently serve as standards to which other novel carriers are referenced. A key feature of PEI is that nitrogen (in a mixture of primary, secondary and tertiary amines) accounts for a third of its molecular weight. Since different classes of amines possess characteristic pKa, this ensures that not all the nitrogen will be protonated at a given pH. Furthermore, the proximity of the nitrogen atoms also means that a protonated amine can suppress the protonation of its neighbours due to the energetic penalty that gets incurred by situating like charges adjacent to one another (Suh et al., 1994). This blend of protonated and yet-to-be protonated amines is then suggested to endow PEI with its strong DNA binding and buffering abilities. The intracellular trafficking properties of PEI/DNA complexes has been studied (Godbey et al., 1999) and targeting moieties such as mannose (Diebold et al., 1999) and transferrin (Kircheis et al., 1997) were also coupled onto PEI. However, two of the main problems of PEI are its toxicity and nonbiodegradability. In response to the latter, PEI was functionalised with reducible disulfide bonds (Lee et al., 2007) or hydrolysable ester bonds (Liu et al., 2008) to facilitate

biodegradation and at the same time, to aid its intracellular plasmid release.

synthesis, which makes the manufacturing process amendable to up-scaling.

Peptide chains can be fabricated from any of the 20 naturally occurring L-amino acids, which are referred to by their single- or three-lettered code (e.g., R or arg for arginine). Peptides are thus biocompatible and often degradable. Peptide synthesis also does not involve harmful catalysts, which is a concern in the synthesis of some polymeric carriers. Furthermore, synthesis can now be automated, courtesy of advances in solid phase peptide

Before peptides were seen as proper DNA carriers, studies conducted in the 1960s with histones had already suggested that cationic amino acids such as lysine and arginine can be useful (Akinrimisi et al., 1965). Today, peptide vectors are given fanciful names such as "cell penetrating peptides" (CPP) or "protein transduction domains" (PTD) to celebrate their ability to efficiently penetrate the plasma membrane and mediate the entry of nucleic acids or other macromolecules. Such CPP can be derived from proteins existing in nature (e.g., viral proteins or venom proteins of bees and wasps) or designed *de novo*. There is little in common among CPP and the only unifying theme seems to be the significant presence of cationic residues (mainly lysine and arginine). Amphipathicity has been suggested to be another common feature. However, this is only true for most CPP – oligoarginine being an exception. There are two methods by which peptide vectors carry their cargoes. Nucleic acids are usually non-covalently (electrostatically) attached, while proteins and other macromolecules are typically covalently coupled (chemical crosslinking or by plasmid fusion). The advantages of electrostatic attachment include convenience and the largely unaltered chemical properties of the cargo. However, charge interactions are non-specific and excess peptides are usually needed to completely bind the plasmid. On the other hand, covalent attachment requires chemical modification of the cargo and usually results in stable complexes which, as discussed earlier, may not be desirable. In the following sections, we will review the key classes of peptide vectors and provide non-exhaustive examples of strategies that have been used to improve their

**4. Peptide-derived vectors for gene therapy** 

efficiency as vectors.


Table 1. The amino acid sequences of all the peptides discussed in this chapter.
