Instituto de Medicina Molecular, Faculdade de Medicina, Universidade de Lisboa, Lisbon, Portugal
Address correspondence to: Miguel Castanho, Instituto de Medicina Molecular, Faculdade de Medicina, Universidade de Lisboa, Av. Prof. Egas Moniz, 1649-028 Lisboa, Portugal. Tel.: +351217999477. E-mail: firstname.lastname@example.org
Gene-targeted therapies have a huge potential to fight rare genomic diseases. Both gene silencing using siRNA or shRNA [1, 2] and gene editing tools using the CRISPR/Cas nuclease system [3-6] have proven promising innovative ways to treat severe genetic malignancies/disorders, from infections to cancer and hereditary diseases [7-9]. Only 10% of the genome can be targeted by the conventional rationale of the Lipinski rule-of-five for small molecule drug design . Moreover, gene therapies need to be tissue-specific, and nucleic acid drugs are required to penetrate the intracellular lumen and act on the nuclear machinery with no significant toxicity [11, 12]. Although, delivering nucleic acids into cells is a difficult task because they typically have high molecular weight, are vulnerable to enzymatic degradation, and are anionic, which makes them poor translocators of cell membranes. To overcome this technological limitation, among the wide variety of vectors that have been developed in the past decades (from lipid structures, nanoparticles and peptides), cell-penetrating peptides (CPP) have proven to be successful in delivering various therapeutics against multiple diseases [13-15]. During the last 2 decades, CPP have been widely studied as delivery vectors [16-18] and have been increasingly used to transport small RNA/DNA, plasmids, antibodies, and nanoparticles into cells [19-21].
However, despite the huge potential that these peptide-based approaches have proven to have in academic research, few advances have been put to pharmacological or clinical use. In the related field of peptide-based drugs and drug leads, advances in the use of peptides are tremendous, with some drugs reaching a status of major economic relevance. For instance, commercialization of Copaxon, Lupron, Zoladex, Sandostatin, and Velcade, have all surpassed the one-thousand million US$ global sales threshold, and around 100 peptide-based drugs have already reached the market [22, 23], with more than a billion dollars invested in therapeutic peptides in the last 10 years, although very limited resources were invested in developing peptide-based drug carriers . In this review, we will describe the origin, development, and usage of CPP as drug delivery systems (DDS), with special emphasis on their current application to gene-therapies.
Why Use DDS?
The human organism poses many difficulties for the safe delivery of effective pharmaceuticals to the target cell . Systemic administration and endovenous strategies allow for 100% bioavailability, obviating enteric, and hepatic metabolic degradation. Once in the bloodstream, the drug lead needs to overcome some challenges, avoid the immune system response (complement activation and phagocytosis), as well as serum protein aggregation and kidney filtration . Once near the target cell, the eukaryotic membrane will pose further obstacles, being an impermeable barrier for most of xenobiotics [11, 15, 24, 25]. The plasma membrane only allows the influx of small compounds, requiring transporters for most hydrophilic macromolecules .
The need of DDS for small molecules and gene-based tailored therapies also results from its biochemical characteristics such as poor stability, lack of cellular uptake and insufficient ability to reach targets [18, 19, 26, 27]. These are therefore the new key challenges in medicine – how to obtain effective therapeutic agents that act locally with limited adverse off-target effects , with no loss of pharmaceutical potency or need for increasing dosages.
Several DDS have been developed [14, 28, 29]. While envisioning such technology, one should consider the use of biocompatible materials for simple, yet strong assembly processes. Optimization and fine-tuning of the biophysical-chemical parameters to enhance the DDS pharmacokinetic and pharmacodynamic properties are also urgent demands in this field [14, 26]. A successful DDS must work on different tissues, have fast endosome release, be functional at low dosage, have no toxicity, and be easy to administer concerning therapeutical applications .
Several technologies have been developed and some examples of DDS approaches are: electroporation , ultrasound mediated plasmid delivery , viral delivery , nebulization , direct chemical modification , and association with nonviral delivery vehicle such as lipids , liposomes , dendrimers , cationic polymers , inorganic particles , carbon nanotubes , small molecules , receptor ligands , more recently, supercharged proteins [43, 44], and CPP [18, 45, 46]. This review will focus on one specific kind of DDS, CPP, mainly on its features and its application to clinical therapies. Figure 1 exemplifies the status of DDS technologies and the cargoes that were already delivered into cells using cellular endocytic pathways or by means of direct membrane translocation.
CPP [18, 45, 46] belong to a vast group of molecules, “membrane-active peptides” that comprise short amino acid (aa) sequence peptides that exert their action at the lipid membrane and are currently pursued for their multitude of biomedical applications [Drug Delivery , Anticancer [47, 48], Antiviral , Antimicrobial [50, 51]]. Assigned to different nomenclatures such as “protein transduction domains,” “membrane translocating peptides,” and “Trojan Horse Peptides,” the terminology CPP introduced by Pooga et al.  and is the most accepted nomenclature for the set of peptides capable of interacting with, getting into, and ultimately transposing membrane structures . CPP are typically short, usually 5–30 aa residues long, and can be isolated from naturally existing proteins, modified or designed de novo [16, 18]. As cellular, endosome, and other organelle membranes are formidable barriers in the quest for delivering drugs to potential intracellular therapeutic targets, CPP are now contemplated as frontrunners for the development of such reliable DDS [16, 18]. They can associate to molecules such as nucleic acids, proteins, drugs, or imaging agents by covalent or noncovalent binding, and deliver the cargo to the cytoplasm or nuclei of cells [16, 18, 27, 46]; Fig. 1A).
CPP research had its origins in the study of protein transduction domains, which demonstrated the ability to shuttle transcription factors within and between cells , together with the observation by Frankel and Pabo  that the HIV-1 transcription-transactivating protein (TAT) could enter cells and translocate into the nuclei . Later, in 1991, another peptide sequence demonstrated the ability to internalize cells, Drosophila Antennopedia homeodomains, which in 1994, yielded the first so called CPP–Penetratin [55, 56]. The following years reflected a boom in CPP research and several new sequences were designed and created such as MPG–the first CPP capable of non-covalent nucleic acid delivery [57, 58]; and Pep-1 for protein and peptide delivery [59, 60]. Afterwards, Lebleu and coworkers  identified Tat peptide, containing the minimum, nine residues long, sequence from TAT protein required for cellular uptake, and the first proofs-of-concept for the in vivo application of CPP were reported for the delivery of peptides and proteins in 1999 [46, 62]. Another important breakthrough was the demonstration by the groups of Wender and Futaki that octa-arginine (R8) was sufficient for eliciting cellular and in vivo peptide uptake [63, 64]. Ever since the discovery of Tat, researchers have been steadily increasing the pool of known CPP, with entries derived from natural, chimeric and later synthetic sources (Fig. 2; 65). Together with CPP, Homing peptides (HP) are molecules that deliver their cargo to the cell surface with cell specificity that are lacking in CPP by targeting specific markers on cell surface but with no membrane internalization. A class of CPP derivatives, CPHP (Cell-penetrating HP) that conjugate both translocation and cell specificity features of CPP and HP, respectively, was successfully conceived, combining specific delivery with cell internalization capability . To maximize the design of new CPP, recent advances in bioinformatics gave birth to tools that allow the identification of protein regions with membrane translocation domains (i.e., CPP; [66-68] to maximize the design of new effective CPP.
However, if one goes deep into CPP classification, one can find different categories under which they can be classified, by their physical chemical properties, by its origin or by its linkage to their cargo [16, 17, 65]. In addition, due to its chemical composition and acquired secondary structure, they can also be classified into different modes of uptake [16, 17, 65]. Chimeric and synthetic CPP can be obtained either by modification of functional groups of aa or the alteration of specific aa to increase its transfection efficiency [16, 17, 65]. Substitution of lysine by arginine increases the affinity for the cell surface , whereas using histidines helps the endosomal escape by creating a “proton sponge effect” .
CPP have various origins, from peptide libraries and random aa or peptide based combinations, to specific design from existing proteins/peptides, almost any origin is possible [16, 65]. They are usually subdivided into protein-derived peptides, like Tat and penetratin, which suffered minor modifications to enhance their efficacy; Chimeric peptides, such as transportan, composed by regions from wasp venom and galanin proteins ; and finally synthetic peptides, such as polylysine and octa-arginine (R8; 16, 65).
Considering CPP derived from proteins, they are found in heparin binding proteins from the GAG family , DNA and/or RNA binding proteins with their cationic motifs , some antimicrobial peptides (AMP), amphipathic peptides with positive net charge and hydrophobic residues , and finally viral proteins, such as HIV-TAT and Dengue virus Capsid protein [16, 44, 74]. Additionally, because CPP can acquire different secondary and tertiary conformations, they are arranged into three different major groups: cationic, amphipathic, and hydrophobic [16, 65].
Physical Chemical Properties
Cationic CPP's have a stretch of positive charges without an amphipathic helix in its tridimensional structure and are mainly composed by arginine, lysine, and histidine residues [16, 65]. Their translocation capability is concentration-dependent, for higher dosages it performs direct translocation, whereas for lower doses it is dependent on endocytic pathways. Tat and R8 are examples of cationic CPP being the uptake dependent on the amount of arginine residues in the sequence .
Amphipathic CPP's are also composed by arginine and lysine, although it is their lipophilic and hydrophilic tails that are responsible for the translocation feature. Generally, these are peptides that acquire α-helical secondary structure with the hydrophobic and hydrophilic domains faced in opposite directions using the charged region for cell membrane interaction and the hydrophobic region for membrane perturbation and translocation [16, 65]. Amphipaticity is an exquisite characteristic of this CPP family as single aa sequence mutation can greatly diminish their cellular uptake [76, 77]. Similar to α-helical CPP, β-sheet amphipathic CPP are composed by one hydrophobic and one hydrophilic stretch of aminoacids exposed to the solvent. Proline-rich amphipathic CPP are sequences enriched by this residue with the ability to translocate cells [16, 65]. Hydrophobic CPP are mainly composed of hydrophobic residues, resulting in a low global net charge, and their cell-penetrating feature arises from the high affinity that these sequences have for the hydrophobic regions of cellular membranes.
In addition, CPP can be classified considering as having viral vs. nonviral origin. Viral vectors are based on adenoviruses and retroviruses, and were the first to be developed, and are the most common (70%), due to their natural and efficient transfection abilities as well as genetic material transport [46, 78]. However, there are some risks of oncogenesis and immunogenicity; therefore, nonviral gene delivery is being studied as a safer alternative .
Covalent and Noncovalent Cargo
CPP can transport the demanded cargo by either being covalently bound to or by simply establishing a stable noncovalent complex with it . Covalent approaches may refer to chemically conjugating the CPP and cargo through crosslinking or by cloning the CPP sequence as a fusion-protein. A word of caution is advised as the type of linkage, chemistry, and nature of the spacer will influence the stability and efficiency of the cargo and CPP . Thus, covalent reactions that are reversible inside the cell are advised . Some advantages arise from this procedure since it is a reproducible method with controlled stoichiometry, enhanced cellular uptake [18, 79]. There is also a higher likelihood that CPP-cargo formulation is approved for clinical use [18, 79]. The formation of a noncovalent conjugate is generally dependent on the chemical nature of the CPP. Cationic CPP may form stable complexes with the cargo by means of electrostatic interactions, especially with genomic material. This type of linkage was developed for gene delivery of condensed DNA associated with peptides favoring endosome escape [18, 79].
Advantages Versus Limitations
DDS are useful tools to circumvent new intracellular-targeted therapeutic limitations, such as poor delivery, low bioavailability, and lack of specificity [17, 18]. CPP have a large cargo capacity, deliver payload with high yield, are stable in physiological conditions, lack sensitivity to serum, and have residual toxicity or immunogenicity [15, 18, 46, 78]. However, CPP have some disadvantages: they lack oral bioavailability, which requires other routes of administration such as intranasal or injectable; and have a short life-spam, mainly due to enzymatic degradation, though it can be surpassed by sequence modification with d-aminoacids . Another raised issue for CPP-based technologies is the potential humoral immune response toward the cargo carried by the CPP . Also, most of CPP are not cell-specific, which can be tackled by the combination of HP sequences yielding CPHP ; by an ACPP (activatable CPP prodrug) strategy, in which the CPP is fused with a polyanionic counterpart that can only be cleaved by proteases present in the target cell ; by CPP-modified ATTEMPTS (Antibody Targeted Triggered Electrically Modified Prodrug Type Strategy), where the CPP is combined with heparin, increasing antibody targeting and plasma stability, being released by the administration of protamine ; by PEG shielding sensitive to tumor specific proteases, since the drug will only be released when in contact with specific proteases ; and by reversal shielding by pH-sensitive PEG .
Mechanism of Action
CPP have to overcome several barriers in order to achieve its purpose of delivering cargo into a cell. After establishing a tight protective system with the cargo (Fig. 1); it must avoid organism elimination . CPP are required to cross the vascular endothelium, which usually does not allow the passage of particles larger than 5 nm, before it reaches each individual target cell. Various cellular membrane translocation mechanisms have been proposed for CPP uptake (Fig. 1B; [15, 16, 84], none of which can be ubiquitously applied. In fact, one of the major limitations of studying peptide-membrane interactions is that they seem dependent on system parameters, such as CPP concentrations, pH, peptide sequence, and conformation, lipid membrane composition, cell line of choice, etc . This fact is further complicated when studying CPP, because it is accepted that the payload can significantly impact internalization routes and internalization yield [85, 86].
Once close to the cell membrane, the first CPP/Cell membrane contact is promoted by electrostatic interactions with cellular proteoglycans GAG [87, 88]. Such molecules can mediate membrane fluidity, and initiate actin remodeling signaling cascades, thus their contact with peptides has been hypothesized to promote the initial step for CPP entry by direct translocation, macropinocytosis, clathrin-dependent endocytosis or other mechanisms [15, 16, 65, 84, 89], by which both can take part in one CPP entry mechanism, at low and higher CPP concentration, respectively.
In nonendocytotic mechanisms or direct translocation, the peptide physical-chemical properties are compatible with the lipid bilayer or are able to disturb its structural integrity allowing it to translocate the membrane via an energy and temperature independent process [15, 84, 90]. It is based on the continuous internalization of molecules with high affinity to the membrane into the cell . However, this mechanism is highly affected by the cell membrane heterogeneity, leading to different modes of uptake for different tissues. There are several models explaining this process such as the barrel stave, inverted micelle, carpet model, adaptive translocation, transient-pore formation, electroporation-like translocation, and entry by microdomain boundaries [27, 84, 91]. This kind of internalization is favored for being energy-independent and nonsensitive to endocytic inhibitors. Cellular uptake by means of nonendocytic routes is mainly dependent on the chemical nature and secondary structure of the CPP, its ability to interact with the cell surface and lipid moiety, the nature, type and active concentration of the cargo, and also the cell type [27, 84, 91]. However if the translocation occurs by a physical mechanism only, it also depends on promoting forces such as pH gradients  and transmembrane potentials .
Endocytosis and Endosomal Escape
Although many studies have demonstrated the contribution of endocytosis in the internalization of several CPP and their conjugates, the exact pathways that contribute to these entries is still under discussion, and may involve concerted action of various entry routes. Endocytosis mechanisms  comprehend the endocytic entry and endosome escape, by which escape constitutes the critical step for an effective delivery . Endocytic entry is subdivided in phagocytosis, reserved for specialized cells in which extracellular molecules are encapsulated into lipid vesicles, and pinocytosis, which occurs in all types of cell and includes clathrin-dependent and -independent mechanisms . In the former, the CPP is recognized and packed into clathrin-coated vesicles and in the latter, other mechanisms are activated such as macropinocytosis, caveolae, and/or lipid raft mediated endocytosis [15, 27, 84, 91].
Distinct pinocytosis mechanisms have been associated with the internalization of Tat. Whereas in a set of experiments the uptake of glutathione S-transferase (GST)-Tat-green fluorescent protein (GFP) fusion proteins was found to derive mainly from caveolae-mediated endocytosis, other groups found that the fusion Tat-influenza virus HA peptide is internalized mostly through macropinocytosis, as well as by clathrin coated vesicles [91, 95]. Furthermore, studies in genetically engineered cells lacking functional clathrin- or caveolae-mediated pathways suggested that Tat cell uptake is also endocytosis independent . An abundant number of studies in the field show similar conclusions, which illustrates the difficulty in inferring behavior patterns in CPP cellular uptake.
Furthermore, some studies revealed artifacts associated with microscopy studies, which are a main data source in this type of research and can often lead to erroneous conclusions. For instance, initial reports stated that CPP internalization occurred even under endocytosis pathway-blocking conditions, suggesting an energy-independent uptake attributed to direct membrane translocation. However, a bias towards the opposite conclusion can also happen, because live cells recycle their membranes at quite high rates; hence, all peptides can potentially be randomly internalized by flip-flop mechanisms following static absorption into cellular surfaces. For such reasons, interpretation of optical microscopy information should integrate knowledge acquired from membrane models, biophysical studies and other assays [44, 74]. For those molecules that indeed require cell energy-dependent mechanisms for internalization, it should be emphasized that endosomal escape is an obligatory feature in delivery applications, as peptides or peptide-cargo complexes retained within mature endocytosis vesicles will be degraded by acidic peptidases in lysosomes before reaching their intended targets. In fact, like the cytoplasmic membranes from which they are originated, these organelles constitute one of the most notable pharmaceutical barriers, and their transposition arouses great interest among scientists. In this regard, several lines of evidence have been provided to support capacity for endosomal escape by CPP . An interesting model integrating these considerations was presented for Tat peptide by Mishra et al. who defend that it combines cytoskeletal remodeling with membrane translocation to escape endosomal compartments, and deliver cargoes intracellularly .
Influence of Cargo on Mode of Uptake
CPP investigation mainly derives from the necessity of developing new drug delivery applications. Therefore studying the efficacy and effects of CPP-drug complexes is as important as analysing CPP per se. With respect to this point, currently available data suggest cargo parameters like size, structure, charge or other biophysical properties can exert a deep influence on cellular uptake and cytosolic distribution [85, 86]. In general, high molecular weight cargoes, like CPP-fused proteins and nanoparticles, cannot be internalized by direct transposition of cell surface membranes and must follow an endocytic route, probably because large cargo transport would imply an extensive destabilization of cell surfaces and consequent cytotoxicity [85, 91]. For compounds of lower molecular weight, although, like small peptides coupled with CPP sequences it is not very clear which mechanisms are most important for biological activity, and it cannot be excluded that both direct and endocytic transport pathways may, in some cases, equally contribute to internalization . There are no common patterns to describe CPP and CPP-Cargo uptake by cells, which poses a clear setback in the development of CPP-based therapeutics. Every particular peptide-cargo system will have to be extensively studied and optimized de novo before market translation can even be envisioned [18, 27]. There are two main methodologies to ascertain such drawback: in vivo/in cellulo or in vitro [18, 27]. The first follows labeled versions of CPP or Cargo (if the CPP are associated with one), by indirect or direct techniques [27, 74, 89]. More direct techniques can be enviosioned, electron microscopy studies of peptides localization and in-cell Raman spectroscopy study of the peptides secondary structure in cellular compartments. Following the cellular uptake, the cargo must reach its target, which may be monitored by cell tracking by fluorescence microscopy [27, 74, 89].
CPP Therapeutical Applications: Challenges and Hopes
Although showing remarkable properties and constituting promising vectors for drug delivery, CPP currently encounter serious obstacles to clinical application. Toxicity and lack of selectivity, are perhaps the most obvious ones, but problems like reduced in vivo efficacy, poor bioavailability, instability and a short life span of action, must also be surpassed . Unfortunately, these are standard issues, common to most DDS under development, and usually call for compromises between different parameters. For instance, one regular dichotomy concerns the choice of increasing drug potency or reducing its toxicity, since there is generally a strong correlation between the first and the second. A number of studies have already assessed the toxicity of CPP in vitro, while in vivo information from animal assays remains fairly limited and the data are difficult to correlate, because of different CPP sequence, payload, and administration route . In general, it is assumed that CPP show low levels of cytotoxicity. A recent study on the impact of different cargoes on the profiles of penetratin, Tat and transportan also suggested that the choice of CPP has to be adapted to different cargoes, and that the uptake and toxicological profile of one peptide-cargo combination does not necessarily apply to other formulations .
Other than toxicity, the kinetics and biodistribution of some CPP have also been studied in vivo confirming that CPP therapeutics can potentially be safe and target most tissues in the body . Rapid blood clearance and preferable distribution to the liver, kidneys, spleen, bowels, and lungs were common results, which is expected for most drug delivery systems . Substitutions of biological l-aa for d-analogs or other unnatural residues, stabilization of the 3D structure, or linkage to shielding molecules, such as using polyethylene glycol, are common strategies for trying to overcome those shortcomings .
Another way to tailor a drug delivery system depends on the cargo-binding approach as covalent and noncovalent binding are possible. Although versatile, crosslinking reactions are labor intensive and difficult to control, originating heterogeneity in final products, and may interfere in cargo functionality. It is also quite difficult to combine strongly anionic molecules, such as oligonucleotides, with positively charged peptides, due to aggregation phenomena. This nonspecific electrostatic-driven aggregation absorption is, however, the basis for non-covalent linkage alternatives [58-60], and as such, is associated with gene delivery and other similar applications. However, these noncovalent complexes are usually less stable than those based on covalent bonds.
Another important topic for CPP is selectivity. Because these molecules can potentially transpose membranes of any cell type and tissue, CPP might deliver drugs to locations where there might be detrimental effects (Fig. 3). As such, CPP therapeutic development is often connected to the field of targeted drug delivery and the development of CPHP.
As CPP-based delivery strategies have gained considerable popularity, more than a thousand attempted applications, have already been reported, accounting for both in vitro and in vivo drug delivery attempts. Although these molecules have been shown to favor a great variety of cargoes (Fig. 1), most studies have focused on the transport of nucleic acids and their analogs [11, 17], as well as peptides and proteins [59, 60]. Since the first proofs-of-concept for the in vivo potentialities of CPP, these molecules have been used to target a myriad of pathologies such as asthma, ischemia, diabetes, inflammatory, and oxidative stress-related diseases [17, 91]. Therapies for cancer are also intensively studied, and CPP are equally investigated for treatment of brain-related disorders because of their ability to locally disrupt and cross the highly selective, nearly impermeable, blood-brain barrier . For this end, gene-targeted therapies are a very promising approach given the prospect of modulating genetic profiles of disease related tissues.
Over the last decades, the perspective of gene-therapies as the next generation therapeutics has faced an initial optimistic endeavor, due to their potential as a permanent cure of several diseases. However, it was followed by global disappointment because of its lack of efficacy and inability to reach its target; DDS challenged this initial disappointment and offered safe routes to optimize such innovative therapies. According to the cargo carried or diseased-targeted, CPP biomedical application can be categorized into different groups .
Protein delivery is a valuable tool in the treatment of cancer or cancer proliferation, asthma, apoptosis, ischemia, diabetes, and CPP-mediated protein delivery is a very interesting possibility to ascertain such therapies [17, 100]. It is a challenging prospect because proteins are large, bulky molecules, which must nonetheless retain their 3D conformation and maintain its biological activity during while crossing natural barriers such as the blood-brain barrier and the testicular barrier [15, 17, 100]. CPP are able to carry molecules from 30 to 120–150 kDa; however, the size of the payload may also pose an issue to an effective treatment.
In this regard, CPP have been shown to aid and mediate delivery of several proteins, such as β-galactosidase, GFP, Bcl-xL, human catalase, and glutamate dehydrogenase. Most of the available CPP are covalently linked to these cargoes, with the exception of Pep-1, which is able to noncovalently form protein complexes . As such, CPP have been deemed one of the most promising alternatives for the in vitro and in vivo delivery of these and other molecular drugs in pharmaceutical or imaging applications; either directly as delivery vectors, or possibly even as active therapeutic ingredients [17, 91]. A specific example of this application is the noninvasive insulin delivery , in which the CPP increases the insulin uptake from the intestinal lumen leading to an increase in its oral bioavailability, overcoming current issues such as invasive injectable formulations and low bioavailability from other administration routes. Similar technologies are currently under clinical trials, Amgen KAI-9083 and KAI-1678, in which a selective protein kinase C inhibitor combined with Tat peptide is being tested for angioplasty.
Nucleic Acid Delivery
Gene therapy has had many limitations due to the poor permeability of plasma membrane to DNA and the low efficiency of DNA to reach the target within the cells [11, 17, 18]. Many other DDS were attempted to surpass these difficulties but failed due to their toxicity and poor efficiency related to endosomal entrapment [11, 17, 18]. Success was obtained when peptide-based carriers were developed, which combined DNA-binding with membrane-destabilizing properties [17, 100]. Examples of CPP-nucleic acid delivery range from oligonucleotide analog delivery with gene-antisense application or mRNA splicing correction strategies, or delivery of short interfering RNA (siRNA), a biomedical tool to control protein activation and/or expression. The latter technique is much more effective when delivered by non-covalent strategies to deliver larger nucleic acid molecules such as plasmids or viral genes [11, 17, 100].
Gene delivery, siRNA and steric blocking DNA mimics, such as PNA and PMO, can thus be envisioned for CPP-driven gene therapy, antisense knockdown or mRNA splicing correction. CPP can provide an alternative, or a complement for viral vectors and liposomes strategies, current gold standards for gene therapy, which nonetheless present toxicity and efficacy-related issues. Coupling of nuclear localization signal (NLS) sequences to CPP is also common for gene delivery, since these nucleic acid sequences must be delivered to nuclei [17, 18]. siRNAs, as well as other nucleic acid molecules, can be covalently or non-covalently combined with CPP. MPG has been used to deliver siRNA for in vivo silencing of OCT-4 and cyclin B1, to prevent tumor growth when injected intravenously in mice .
Some explorations of these features are currently under clinical trials, Sarepta AVI-5126 and AVI-5038, which consists of antisense targeted therapy to inhibit c-Myc expression combined with an arginine rich CPP, in order to prevent tissue damage from internal hyperplasia following endovascular injury .
CPP share a multitude of features with AMP starting with their chemical structure, which explains their relationship with cell membranes and why some AMPs can be used as CPP, and vice-versa; CPP have AMP usages . Figure 4 briefly describes the multitude of applications that CPP may acquire due to the plethora of cargoes that can be conjugated with.
Inflammation plays a role in a variety of diseases, from infections to cutaneous, alergological, and rheumatologic diseases, and is mediated by mostly intracellular pathways of cells of the innate system. Inactivating these pathways has been the objective of many therapies such as the suppression of TNF α and IL 6, proinflammatory cytokines, with the KAFAK peptide, delivered by nanoparticles ; or the inhibition of NF-κB by the AIP6 peptide, which has both anti-inflammatory and CPP features .
One of the major clinical CPP applications has been cancer therapy . CPP are used as delivery of small molecule drugs used in chemotherapy, delivery of peptides, or proteins usually associated with apoptotic pathways, delivery of genes, affected on the course of oncogenesis, and delivery of nanoparticles, used in radiodiagnosis methods. Cancer cells develop a variety of mechanisms to become resistant to current therapeutics usually by the development of efflux pumps. The association of the anticancer molecule with CPP overcomes this issue, as it was demonstrated with the association of Taxol with octa-arginine in the treatment of ovarian cancer, both in vitro and ex vivo . Moreover, the association of chitosan/doxorubicin/Tat inhibited the growth of CT26 xenigraft in mice  and the PEGA-pVEC-Chlorambucil complex was effective against breast cancer by targeting the blood vessels . Tumor suppressor proteins, like p53, induce apoptosis and tend to be “shut-down” during oncogenesis. By delivering them, p53-regulated pathways, such as p21, BAX, PTEN, or AKT [107, 108] can be restored. The delivery of p53 has been carried out in association with Tat, which resulted in an increase in survival time in peritoneal carcinomatosis mouse models ; with penetratin in pancreatic tumor ; and with histone H4 N-terminal tail . Another tumor suppressor protein, p16 was restored as “Trojan p16” in a pancreatic tumor animal model, in association with penetratin, improving the survival rate . However, there are other ways to activate apoptosis: the complex Tat-SMAC, a mitochondrial caspase, sensitized cells to apoptosis stimuli ; and the delivery of toxic enzyme drugs such as gelonin-Tat, or caspase 3, asparaginase and KLA associated with peptides .
Gene therapy using RNA delivery, most frequently small interfering RNA (siRNA), is used to modulate gene expression  and has several usages in cancer therapy: VEGF can be silenced by siVEGF coupled with a R9-cholesterol carrier; cyclin B, essential for mitosis, can be down-regulated, leading to inhibition of angiogenesis, when silenced by a siRNA against it, complexed with MPG-8; also EGFR can be inhibited, causing a decrease in blood vessel proliferation, when a specific EGFR immunotoxin is delivered coupled with Tat [17, 115].
Tumor imaging, like cancer therapeutics, is a growing branch, and it has benefited from CPP, by their ability to deliver nanoparticles. Usually, these nanoparticles are fluorescent, which allows the identification of tumor cells, together with body scan techniques (for example MRI): Oligoguanadinium, for instance, has been conjugated with a peptide sequence that is cleaved at the cell surface, allowing it to enter specific cells, a tumor imaging technology developed by Tsien and coworkers . Another example is the Cy5-Cy7 fluorophore connection, which allows the detection of tumor cells when delivered by CPP that are cleaved by tumor-associated matrix metaloproteases.
CPP are also useful for neurologic disorders treatment like Alzheimer's Disease, Epilepsy, Down syndrome, traumatic and ischemic brain injury among others . To inhibit the neurodegenerative glial protein S100B a thermally responsive S100B inhibitory peptide was designed– Synb1-ELP-TRTK, and to decrease the neuronal loss AP-1 inhibitory peptides were combined with Tat, providing neuroprotective effects .
Small Molecules Drugs
CPP's enhance properties, which previously limited the usefulness of certain drugs (deliverability and toxicity). For example, they are able to tackle cancer cells acquired resistance to some chemotherapeutic agents by the presence of efflux pumps .
Delivery of Imaging Agents
There has been a huge effort towards the development of noninvasive imaging techniques. CPP can be useful for the delivery of imaging agents, which usually do not cross the cell membrane, for example fluorophores, semiconductor nanocrystals/quantum dots or MRI contrast. Adding to its delivery ability CPP are of interest in this area due to the possibility of tissue and/or cellular specificity .
Enhancement of Activity for Peptide-Based Drugs
CPP can be used to improve the activity of certain drugs. For example, AMP used in cancer cells often have problems of specificity, which can be surpassed by the use of CPP addition, CPP amplify the apoptotic properties of certain peptides by increasing their cellular uptake. CPP can also be attached to inhibitors of transcription factors diminishing the expression of certain genes, such as the hypoxia-inducible factor 1, increased in renal cancer. Finally, the combination with doxorubicin with CPP increases its uptake and contributes to improving doxorubicin efficacy in killing cancer cells [17, 100].
The possibility of targeting specific organelles would allow us to study biological processes but also to deliver therapeutics to these organelles, with higher disease specificity. Nuclear target can be achieved, as stated previously, but the attachment of a NLS to CPP. Mitochondrial targeting is achieved by the insertion of artificial sequences like tetrapeptide renaming CPP to mitochondria penetrating peptides [16, 17, 100].
Conclusion: A Fruitful Endeavor
Despite the abundance in preclinical data, very few CPP have actually reached the clinical setting, and none has yet achieved commercial viability. Cell Gate Inc conducted the first evaluation of CPP usage in humans. They successfully reached phase II clinical trials with their technology in 2003, but have abandoned further trials. From this point on, other companies have been working on the pharmaceutical development of CPP. For instance, Auris Medical/Xigen recently completed phase IIb trials for a proprietary CPP (XG-102/AM-111) with favorable results, having received funding for phase III testing, and ReVance Therapeutics are currently pursuing CPP technology for cosmetic purposes (RT001) undergoing phase III evaluation. Still, the number of companies currently investing in preclinical to early clinical trials for CPP-based DDS is increasing as the first proofs of concepts for in vivo applications in therapeutical approaches emerge–CPP are on the threshold of clinical application. CPP present a vast repertoire of possibilities. These molecules might be derived from natural sources, or from synthetic approaches either due to random/guided chemical, or in silico assaying.
We can state that CPP history follows a parallel with that of biomedical nanotechnology, in a sense that only now, after decades of intense research and characterization by the academic community, they have been able to reach a state of “intellectual” maturity. Indeed, both fields comprise great potential in DDS and have found their way into clinical settings on the past few years. Such events are important because these technologies present great prospects not only for pharmaceuticals, great and small, but for the general public as well. Medical companies might thus benefit from a financial standpoint, due to the expansion of drugable targets and the recycling of old impermeable, although effective, formulations. Additionally, the general citizen is expected to get access to new products and safer formulations, with lower toxicological profiles and enhanced efficacy. For these reasons, CPP-based technology continues to draw the attention of several groups around the world so that one day their full potential might be realized.
The authors thank Fundação para a Ciência e Tecnologia–Ministério da Educação e Ciência (FCT-MEC, Portugal) [PTDC/QUI-BIQ/112929/2009] and Gabinete de Apoio à Investigação Científica, Tecnológica e Inovação (GAPIC–20130028/PEC/BG, 20120006, and 20110007). JMF acknowledges FCT-MEC for fellowship SFRH/BD/70423/2010.