1. Introduction

1. Introduction

In designing effective membrane-active peptides, not only membrane activity but also stability in biological fluids are crucial considerations. Many promising membrane-active peptide candidates have seen their progress toward the market hampered by excessively high susceptibility to protease degradation ^[1][2][99,100]. In peptides, chirality stems from the presence of a stereogenic center at the α-carbon of every amino acid monomer (except Gly) of any peptide sequence. In general, peptides partially or totally made up of D-amino acids are predictably more stable in biological fluids than their all-L counterparts [25–27]. Consequently, for curtailing peptide clearance the conversion of bioactive peptide structures into their topoisomer versions is a good strategy. Topoisomer term is used to describe all-D-amino acid counterparts of a peptide with either conserved (enantio version, e) or fully reversed (retroenantio version, abbreviated re) sequence relative to the canonical version. Additionally, when the re analogue in standard view (N-to-C-terminal from left) is flipped horizontally, i.e., a 180° rotation on an in-the-plane axis that gives a C-to-N (from left) arrangement, and then compared with the original peptide (all-L, N-to-C from left), the side chains in both structures have coincident orientations.

2. AMP Topoisomers for Facing the AMR Challenge

2. AMP Topoisomers for Facing the AMR Challenge

The persistent overuse of antibiotics in both preclinical and clinical settings has contributed to an alarming rise in antimicrobial resistance (AMR). The cursory prescription of broad-spectrum antibiotics to patients with suspected infections resulted in the unwarranted use of, e.g., over 30% of prescribed antibiotics in the USA alone in 2014, with adverse effects in up to 20% of patients [101,102]. The World Health Organization (WHO) warned that, without intervention, the global consequences of AMR may be devastating, with a projected annual death toll of up to 10 million people by 2050 [103]. This prediction surpasses the combined deaths caused by cancer (8.2 million) and diabetes (1.5 million) [104].

The COVID-19 pandemic has exacerbated this issue [105,106,107,108]. According to a USA antimicrobial resistance report, 80% of patients hospitalized with COVID-19 received antibiotics [109] even though only a minority (17.6%) had a confirmed bacterial infection [110]. The financial burden on governments in relation to eventually tackling the AMRs ensuing from these practices will likely be huge. For instance, 6 out of the 18 most-worrying AMR threats have fetched up a staggering USD 4.6 bn annual bill in the USA alone [109]. Not surprisingly, deaths attributed to AMR are projected to escalate even more rapidly due to the aforementioned factors [111].

A major challenge in the AMR struggle revolves around the six ESKAPE bacteria (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter spp.) [112,113,114], a group of highly virulent pathogens with an extraordinary ability to defeat antibiotic activity. In this context, AMPs are promising contenders [115,116,117] due to their rapid action at low micromolar concentrations; their broad spectrum, encompassing both Gram-positive and -negative bacteria, fungi, and viruses [92,97]; and their mechanisms of action, with a much lower tendency to develop resistance compared to conventional antibiotics [118,119].

AMPs were originally isolated from natural sources. Gramicidin S [120], extracted from the soil bacterium Bacillus brevis [121,122], was first reported in 1939. The finding of AMPs in prokaryotes raised the question of whether eukaryotes also produced AMPs against infections, particularly plants or insects lacking an immune system.

A substance in wheat flour found to be lethal to bread yeast was first described in 1896 [123], but it took some 80 years until it was isolated in a pure form (purothionin) from wheat endosperm in 1972 and shown to inhibit bacterial growth [124]. In 1962, a paper described antibacterial activity in the skin secretion of the Bombina variegata frog [125]; this activity was later identified as corresponding to the AMP bombinin [126]. Subsequent reports of eukaryotic AMPs included the cecropins (1981) of the Hyalophora cecropia moth related by Boman et al. [127,128] and, also in the 1980s, the α-defensins of rabbits [129,130,131] and humans [132] reported by Lehrer et al. and the magainins [133] from the Xenopus laevis frog reported by Zasloff et al. in 1987; this was followed by an ever-growing stream of AMPs from diverse sources such as β-defensins and θ-defensins from immune cells [134,135] or the first anionic (Asp rich) AMP reported in the mid-1990s [136]. At present, AMPs have been found in all types of organisms, including plants [93,137], animals [138,139,140,141,142], and bacteria [143], and they have been have widely acknowledged as ideal candidates for tackling AMR [144,145,146,147,148].

Utilizing topoisomers as natural mimics of AMPs to develop improved candidates is a sensible strategy with which to combat AMR. Among AMP topoisomers, re versions have a high degree of structural resemblance to natural AMPs and are therefore promising candidates.

In the 1990s, the Merrifield laboratory [42] described the e version of the natural AMPs cecropin A, melittin, and magainin 2 amide, as well as cecropin–melittin hybrids, and showed that the antimicrobial activity of the D-enantiomeric versions was equivalent to that of L-parental peptides. They also found that the mechanisms of action did not require a specific chiral receptor [42]. More recently, Kumar et al. studied the e and re forms of peptide 73, a derivative of aurein2.2 (GLFDIVKKVVGAL) [149,150]. Both e73 and re73 versions exhibited activity similar to that of peptide 73, including efficacy against S. aureus in a cutaneous infection model [150].

In a study by Lynn et al., the e and re forms of BMAP-28, a bovine cathelicidin AMP, were tested for their activity against Leishmania parasites. Both topoisomers effectively reduced both promastigote and amastigote forms of L. major [151,152]. In contrast, canonic BMAP-28 was ineffective due to degradation by the parasite metalloproteinase GP63.

Another bovine AMP, the 13-amino-acid indolicidin, isolated from neutrophil granules [153], was also studied in its r, e, and re versions [44]. While all the peptides exhibited antimicrobial activity comparable to that of natural indolicidin, those incorporating D-amino acids were advantageous due to their lower hemolytic activity.

Crotalicidin (Ctn), a cathelicidin AMP derived from South American pit vipers, exhibits both antibacterial and anticancer properties. Falcão et al. dissected Ctn and showed that a Ctn [15-34] fragment had similar activity but much better serum stability when compared to the parent peptide [154,155,156].

Neubauer et al. investigated the activity of the AMPs aurein 1.2, CAMEL, citropin 1.1, omiganan, pexiganan, and temporin A along with their r analogues. With the exception of r-omiganan, the retro analogues exhibited reduced activity compared to their native counterparts. The authors attributed the lower antimicrobial efficacy observed to a relatively higher hydrophilicity in comparison to the natural peptides.

The last two examples serve to emphasize the need for caution when proposing and/or implementing topoisomeric approaches in AMPs (or, for that matter, in bioactive peptides of any type), as the biological outcomes of sequence inversion (r versions), D-amino acid replacement (e versions), or the combination of both (re versions) are arguably non-innocuous given their impact on D structure and ultimately activity [31,99].

3. CPP Topoisomers and Drug Delivery Challenges

3. CPP Topoisomers and Drug Delivery Challenges

CPPs, also known as Trojan horse, protein translocation domains (PTD), or membrane translocation sequences (MTS), are peptides with the ability to traverse membranes —including barriers such as the gastrointestinal barrier [172], the blood–placental barrier [173], or the highly restricted blood–brain barrier (BBB) [174]—and deliver therapeutically active payloads across the boundary. CPPs are about 5–30 residues long, usually linear, (although cyclic versions have been described [27,69]), and mostly composed of L-amino acids, with some sequences including D-residues or L,D-combinations [69]. CPPs have been classified as cationic [175,176,177,178], amphipathic [179,180,181], hydrophobic [182,183,184], or anionic [185,186], with cargoes that include small molecules (drugs and dyes), proteins, nanoparticles, or genetic material [71].

Although most CPPs are considered safe, their (mostly) cationic nature may pose toxicity issues to some organs and tissues; hence it is important to develop CPPs safer for humans. Moreover, many CPPs degrade fast in biological fluids, proteolytic stability thus being critical for their efficacy as drug delivery vehicles [187,188]. The design of improved CPP platforms has addressed these issues by means of in vivo toxicity screens [189] and/or resorting to topoisomers (e or re versions) to avoid protease degradation.

The first CPP identified was a fragment of the trans-activator of transcription (Tat) protein [190]. A detailed study of the protein defined Tat [48–60] as the most effective fragment [69]. Wender et al. studied a shortened version, Tat [49–57], as well as the corresponding e and re versions and found that these topoisomers were more effective than the canonic L-version in terms of entering Jurkat cells [191]. Similarly, Seisel et al. investigated the e and re versions of the Tat [48–60] sequence using iCal36, a peptide intended for patients with cystic fibrosis, as cargo, with the re topoisomer found to be most appropriate [192].

Another CPP discovered soon after Tat was penetratin, derived from the DNA binding domain of the Antennapedia protein [69]. Nielsen et al. showed that oral coadministration of insulin with L- or D-penetratin lowers blood glucose levels significantly more than insulin alone. D-penetratin was most effective due to its lower level of degradation by proteases [193]. Similar results in terms of blood glucose reduction were reported by Kamei et al. [194] with respect to D-PenetraMax, another e penetratin topoisomeric analogue.

Following the discovery of Tat and penetratin, oligoarginines such as R6 or R8 have also been recognized as effective CPPs [177]. Studies of e/re topoisomeric poly-Arg peptides again showed their effectiveness in penetration. For instance, Garcia et al. showed that r6 and r8 in combination with lauric acid enhanced insulin transport through the gastrointestinal tract by around 30–40% in Caco-2/HT-29 cells [195]. Similar work conducted by Kamei et al. [196] showed that r8 improved on R8 with regard to intravenous co-administration with insulin.

A rather interesting example of CPP topoisomerism is

Angiopep, the re version of Angiopep-2, an artificial peptide that crosses the BBB [197]. By combining

Angiopep with nanoparticles and a dye, effective BBB penetration and glioma targeting within the brain was demonstrated. This approach could prove particularly valuable in facilitating tumor identification during surgical procedures [198]. Another relevant BBB-crossing CPP, the THR peptide, discovered through phage display, can successfully interact with the human transferrin receptor. The protease vulnerability of the THR peptide [199] has been successfully overcome by its re version [28], used to transport nanoparticles inside the brain [200].

Another example worth mentioning is mastoparan (MP), a 14-residue peptide isolated from the venom of Vespula lewisii. Topoisomers of MP and its MitP analogue (with α-aminoisobutyric acid instead of alanine in position 10) were shown to be effective CPPs, with the e version of MP displaying the highest translocation efficacy and protease resistance [201].

To investigate D- and L-CPP entrance in cells, Verdurmen et al. compared the effect of three peptides (hLf, penetratin, and nonaarginine) in their canonic and e forms on three different cell lines (HeLa, Mc57 fibrosarcoma, and Jurkat T) [202]. They observed distinct differences in uptake efficiency between the two enantiomers at low concentrations, which could be attributed to a two-step internalization process. A first step was binding to heparan sulphates (HS) [203], the receptors of Arg-rich CPPs (especially of L-versions [202,204]), followed by internalization via endocytosis. Notably, the presence of HS appeared to hinder the efficiency of the second step for D-CPPs at lower concentrations. In contrast, at higher concentrations, D-enantiomers became more efficient, as the dominant mechanism was direct penetration [187]. These findings underscore the importance of stereochemistry, mechanisms of action, and applied concentration when studying CPPs.

Another interesting example is

CDX, the re version of CDX, a 16-residue peptide derived from the II loop of snake neurotoxin candoxin [205], with higher transcytosis observed in BBB models compared to the protease labile L-version [206]. Han et al. proved that

CDX combined with liposomes crosses the BBB in vitro, following an energy-dependent lipid raft/caveolae- and clathrin-dependent pathway [207].

It embodies the re version of a transmembrane (TM5) motif of the CB1 cannabinoid receptor (CB

R) fused with the e version of a BBB shuttle. The co-administration of the peptide and a cannabinoid preserves THC-induced analgesia but minimizes side effects (i.e., cognitive impairment) by restricting, both in vitro and in vivo, the formation of a heterodimer between CB

R and the serotonin 5HT2A receptor responsible for the unwanted side effect [208].

4. ACP Topoisomers for Mitigating Side Effects in Cancer Treatments

4. ACP Topoisomers for Mitigating Side Effects in Cancer Treatments

Despite significant advances, the three main methods for cancer treatment, namely, chemotherapy [232,233,234], radiotherapy [235,236,237], and immunotherapy [235,238,239], suffer from low selectivity and serious side effects. In particular, in chemotherapy, the continued use of some antitumor drugs often gives rise to resistance [234]. Therefore, there is a clear need for new therapies that combine selective drug delivery with high toxicity against cancer cells [240].

In terms of successfully tackling the above-mentioned resistance and side effects [241], ACPs appear to be promising candidates. Currently, there are three FDA-approved ACPs, with revenues over USD 1 million [66]: goserelin (PyrHWSYs(tBu)LRP; Pyr—L-pyroglutamyl) and leuprolide (PyrHWSYlLRP), analogues of gonadotropin-releasing hormone (GnRH) [242,243]; and octreotide (fCFwLTCThre; Thre—L-threoninol), an analogue of somatostatin [244]. Since 2000, this list has expanded with entries such as ixazomib, thymalfasin, and mifamurtide [245,246].

ACPs are classified as direct- or indirect-acting based on their mechanism of action [247]. Direct-acting ACPs (DAAs) specifically target cancer cells, typically attaching to molecules that are either unique or overexpressed [248]. They are divided into five subclasses [241]: (a) CPPs acting as cytotoxic drug carriers [249], as discussed in Section 4.2.; (b) pore-forming peptides, inducing apoptosis or necrosis by interacting with phosphatidylserine anionic lipids exposed on the outer membrane of cancer cells [250]; (c) peptide inhibitors of signal transduction cascades, either inhibiting mitogenic signals or restoring the activity of tumor-suppressive proteins like p53 [251]; (d) cell-cycle-inhibitory peptides, modulating cyclin and cyclin-dependent kinase activity [252]; and (e) apoptosis-inducing peptides, inhibiting anti-apoptotic proteins from the Bcl-2 family [252].

Indirect-acting ACPs can influence the tumor environment or immune response in order to target cancer cells and are subdivided into two classes [241]: (a) immune-stimulating peptides, also referred to as peptide cancer vaccines [253], triggering immune cells such as T-cells to act as natural killers against cancer cells, and (b) analogues of hormone-releasing peptides, inhibiting the proliferation of hormone-stimulated tumor cells. These classes include the above-referred GnRH analogues goserelin and leuprolide [254,255] as well as octreotide and other somatostatin analogues [256].

Upon comparing the modes of action of pore-forming peptides, DAAs, and AMPs, it becomes clear that they exhibit similarities in terms of electrostatic interactions [257]; therefore, many AMPs tend to also be explored for their role as potential ACPs. An example, is Ctn and its fragment Ctn [15-34]. The anti-tumor activity of both peptides towards several tumor cells has been substantiated [154], and more recently, that of their topoisomers has been, too [31]. Furthermore, the mechanism by which these peptides combat tumor cells has been elucidated. Following initial accumulation on the tumor cell surface, Ctn and Ctn [15-34] enter the tumor cell via either an endocytic pathway or an energy-independent mechanism. Ultimately, Ctn and Ctn [15-34] induce cell death through necrosis or apoptosis [258].

The first, noteworthy case is that of e PMI. PMI is a peptide recognized for its interaction with MDM2 and MDMX, two oncoproteins that negatively regulate the functionality and stability of tumor-suppressing protein p53. Active MDM2 and MDMX cause p53 inactivation and ensuing tumor proliferation [259]. The binding of PMI to MDM2 and MDMX prevents their inhibitory action toward p53, thus ensuring that PMI can exert its tumor suppressing role. Li et al. identified three e peptides,

PMI-α,

PMI-β, and

PMI-γ, capable of binding MDM2 and MDMX but unable to induce p53-dependent cell death due to their non-permeability with respect to the cell membrane [260]. They overcame this hurdle by encapsulating these peptides within liposomes decorated with an integrin-targeting cyclic-RGD peptide. This strategy allowed them to curb glioblastoma activity in vivo via the activation of the p53 pathway.

Another topoisomer ACP worth mentioning is re RPL (named

(LPR) by the authors). RPL is the minimal structural part of CPQPRPLC, a peptide obtained via phage display-library screening [261] that binds both VEGFR-1 (vascular endothelial growth factor) and NRP-1 (neuropilin-1), two essential contributors to angiogenesis whose inhibition can lead to a decrease in tumor size. Giordano et al. developed re RPL as an antiangiogenic drug with promising attributes in vitro and in vivo [262]. Also, Rezazadeh et al. linked re RPL to technetium-99m and showed that the conjugate was a good radioligand for imaging and targeting tumors in vivo [263].

The well-known tripeptide RGD binds specifically to integrin αvβ3, making it an antiangiogenic candidate and a molecular imaging probe [264]. Ramezanizadeh et al. showed that the re version of RGD, i.e., dGr, in either linear or cyclic form, was useful for tumor imaging and presented higher bioactivity than the natural version [265].

VAP is a seven-residue prostate-homing peptide that binds selectively to GRP78 (glucose-regulated protein 78) [266], which, in turn, regulates VEGF expression and is over-expressed in some tumor cells but remains absent in normal cells. Ran et al. tested e and re versions of VAP and found higher in vivo antitumoral efficacy when compared to the L-counterparts. Furthermore, tumor growth diminished with either e-VAP or re-VAP, concomitant with an increase in body weight, suggesting reduced side effects. The elevated activity of D-amino-acid-containing peptides was attributed to their resistance against proteolytic degradation [267].

As mentioned above, NRP-1 plays a pivotal role in tumorigenesis and is highly expressed within tumor cells. A library of peptides bearing the sequence motif R/K(X)

R/K (with the C-terminal R or K being particularly vital), named CendR, exhibited notable affinity for binding to NRP-1 [268] in the L-conformation. Upon binding, the peptides regulated vascular permeability (enhanced vascular permeability is indispensable for cancer metastasis [269]). Despite their promise, the susceptibility of CendR peptides to protease degradation resulted in low activity. However, Wang et al. [270] showed that topoisomeric CendRs retained functionality. Thus, using RGERPPR as an example, they showed that both topoisomers were superior, with the e version (rGerppr) displaying higher stability and stronger binding to NRP-1 than the L-peptide and the re version (rppreGr) demonstrating heightened tumor-penetrating prowess and stability. This outcome was interpreted using computational simulations revealing that the three D-Arg residues of the e version neatly aligned with the binding pockets of NRP-1, a phenomenon absent in the natural peptide [270].

A last noteworthy example of topoisomeric modulation is FP21, a 21-residue peptide (YTRDLVYGDPARPGIQGTGTF) corresponding to positions 33–53 of human follicle-stimulating hormone (FSH). The FSH receptor (FSHR) is selectively expressed in 50% to 70% of ovarian carcinomas; hence, it is a potential target in treating ovarian tumors [271,272]. Zhang et al. demonstrated that FP21, incorporated into nanoparticles, effectively bound to FSHR but suffered from a limited half-life [273]. The authors overcame this problem using the re version of FP21, which, again formulated as nanoparticles, achieved FSHR binding, improved biostability, and enabled a reduction in tumor size over the L-version, altogether positioning this topoisomer peptide as a promising candidate for treating ovarian cancer [274].