Some ways to home CPPs

Read Also

Some particular situations have been exploited to bypass the absence of cell specificity of CPPs and to design molecular systems to improve targeting. These are based on the physiological or biological features of the targeted organ or cell type, such as its microenvironment or its enzymatic activity. In all cases, the CPP domain is first somehow hidden prior reaching the targeted area to avoid unspecific uptake. After reaching its target, CPP is then fully exposed to promote an efficient internalization. Cancer cells are very often targeted for a therapeutic purpose, and their metabolism, which is different from that of normal cells, can be used to improve the delivery specificity.

 Exploiting matrix metalloproteases

In 2004, the first targeting system which uses an “activable” CPP was described (Figure 1A) [93]. It exploited the ability of a specific CPP (the arginine homopolymer) to interact intra-molecularly with a polyanionic counterpart through ionic interactions. Both ionic parts of this construct were linked together via a cleavable matrix metalloprotease (MMP) sequence, forming a hairpin. When the construct was introduced in the blood stream (at a concentration of around 6mM), no cleavage occurred due to the insufficient level of circulating MMPs. Thanks to the neutralization of the cationic charges by the anionic counterparts, the cationic CPP could not interact anymore with the anionic charges of the cells of the blood vessels and, therefore, unspecific binding was avoided. Around a tumor, MMPs concentration increases significantly because they are secreted by tumor cells. Thus, the linker could be cleaved, the ionic parts of the chimera dissociated, and CPP ability to bind to the surrounding cells was restored. In this way, a preferential uptake by the tumor cells was made possible. Since a payload molecule is covalently attached to the CPP, this strategy is an alternative mean to concentrate CPP-transported molecules into tumor cells. In other words, this CPP-based molecular construct promotes an indirect, but more specific delivery system in vivo [93].

Exploiting the peritumoral acidic pH

More recently, ionic interactions between CPPs and their anionic counterparts have been exploited to temporally hide the cellular “sticky opportunism” of cationic CPPs during their transport towards the tumor site [94]. This delivery system consists of two components: a conventional hydrophobic core made of a polymer into which any chemotherapeutic molecule can be incorporated, and a peripheral hydrophilic layer composed of polyethylene glycol and the Tat peptide. An anionic and ultra-sensitive di-block copolymer is then complexed to the cationic Tat. Such ionic interactions are expected to shield the cationic charges during delivery up to when the slightly acidic microenvironment of the tumor triggers the protonation of the anionic moiety. This induces the ionic dissociation and the subsequent exposure of the Tat peptide sequence, allowing the preferential internalization of the drug-loaded polymer into the surrounding tumor cells [94] (Figure 1B). This strategy has been pursued using the very pH-sensitive sulfonamide (PSD) group [95]. The PSD compound is fully protonated at pH 7.4 (the pH in the blood stream) and becomes neutral at pH 6.8 [96]. The peritumoral pH has been evaluated to be about 6.9 +/- 0.14 at the tumor-host interface, rising from 7.10 to 7.15 at only 200 micrometers away from the tumor [97]. Indeed, this pH gradient strongly reduces the “effective volume” where the Tat peptide could be freed from its pro-drug structure. The main advantage is again that the Tat peptide is active only in close proximity of the tumor area. On the other hand, the main inconvenience is that the volume where the pro-drug becomes effective is very small as compared to the whole circulation. Therefore, the pro-drug peptide needs to reach rapidly the peritumoral area to fulfill its therapeutic effect prior to its renal and/or hepatic elimination. Further studies are indeed required to provide all the kinetic parameters of this interesting strategy.

Similarly, Kale et al. obtained an enhanced in vivo transfection of DNA using pH-sensitive Tat-modified pegylated (PEG) liposomes (Figure 1D) [98]. Basically, a number of PEG chains were attached to the liposome surface that contained also some Tat peptides of much smaller size. The steric hindrances created by the PEG coat were expected to shield the surface-attached Tat peptides, therefore preventing again the non-specific liposome/cell interactions. Since these long PEG chains were coupled to the liposome surface through a pH-sensitive hydrazone linker, cleavage of the linker could occur only in the acidic environment of the tumor [98]. In experiments performed in tumor-bearing mice, the liposomes were administered directly intra-tumorally, and allowed to obtain at least a three times more efficient transfection than with the corresponding pH-insensitive system. Whether the intra-tumoral injection is required to improve the efficacy of this approach is currently unknown.

The exploitation of the acidic pH has been also described for a peptide derived from the HA2 glycopolypeptide of the influenza virus hemagglutinin [99]. The aim was to improve cytoplasmic delivery of cargo molecules performed with CPPs. In this case, the lyzosome acidification, down to 5 pH units, induced a conformational change of the HA2 derived-peptide, leading to a structural change and exposure of a fusion property. More recently, a Tat-HA chimeric peptide has been used by Dowdy’s group to improve the escape of the Tat-Cre recombinase fusion construct co-incubated in the experiment [59]. The Tat peptide from the Tat-HA chimera probably caused a higher membrane adsorption of the HA peptide itself, thus increasing its cellular uptake followed by lyzosomal destabilization. Subsequently, the improvement of the cytoplasmic escape of the HA-Cre fusion construct could be recorded [59]. The problem with this strategy is that there is no possibility to target a specific cell line as the endocytotic pathway is ubiquitous.

Exploiting the biological state of targeted cells

Another approach is to take advantage exclusively of the internal biological status of the targeted cell, as caused by an infection, or of a metabolic change induced by a pathological disorder. For instance, one can imagine that a toxic molecule, coupled to a CPP as a harmless pro-drug, could be activated only once inside a specific type of cells. Such a strategy has been elegantly achieved by Dowdy's group in 1999 (Figure 1C) [100]. In this study, they fused the Tat peptide to a caspase-3 protein precursor that could be activated only upon cleavage by the HIV protease. This chimerical construct is processed into its active form only by the HIV protease that is exclusively present in HIV-infected cells. This results in caspase-3 induced apoptosis of these cells. This construct was shown to transduce efficiently about 100% of cells and remained indeed fully inactive in healthy cells. Despite these very promising results a well-controlled in vivo experiment in humanized T cell SCID mice is required to definitively validate this approach (S.F. Dowdy, personal communication).

The same group also used the Tat peptide fused to the p27 tumor suppressor to evaluate the effect of its transduction on tumor proliferation in vitro and in vivo [101]. Cell cycle arrest of tumor cells was obtained in culture and an inhibition of the tumor growth was observed in mouse models. These encouraging data have, however, to be considered with the “pharmacist’s eye” for two reasons: first, the experimental model had peritoneal tumors and the chimerical molecule was injected directly into the peritoneal cavity. Therefore, this strategy can be considered as a local therapy, thus strongly reducing the risks of a large spreading of the Tat-fusion construct in the circulation. Second, two injections of 300mg of fusion protein during four days were necessary to obtain the expected biological activity. Since the Tat-p27 chimerical molecule has a molecular weight of 25kDa, this dose corresponds, for one single injection, to a molar excess of more than 140 folds compared to Trastuzumab®, a monoclonal antibody used in therapy in current standard treatments against some cancers. The need of such an important dose could be the direct consequence of the low quantity of Tat-derived construct which enters effectively the cell. Indeed, ongoing work in our laboratory indicates that less than one percent of the extracellular Tat peptide enters effectively the cell (Vivès et al.; in preparation). However, the use of CPPs to mediate the anti-tumoral response is very attractive and it has been reported in the literature. Garcia-Echeverria and colleagues designed a short peptide to disrupt the interaction between p53 and hDM2 [102] in order to restore p53 activity in tumor cells. It was then coupled to a CPP (the Antennapedia peptide), but apparently, the biological evaluation in vivo was not further developed, despite preliminary in vitro tests showing that the chimera did not impair the activity of the interfering peptide [103]. Other groups have obtained in vivo responses following mainly IP, but also with IV or local administration of CPP-peptides at doses ranging from some milligrams to some fractions of microgram per kg (for a recent review [104]). However, although IP injection is commonly used in mouse models, IV or subcutaneous injections would be preferred for a clinical development.