1. Analgesic and Disease-Modifying Effects of RgIA
RgIA is reported as a potent pain-relieving compound in rat models of neuropathy. A single intramuscular injection of RgIA (0.02 and 0.2 nmol) produced an acute, dose-dependent, antihyperalgesic effect in a chronic constriction injury (CCI) rat model. Repeated daily injections of 0.2 nmol of RgIA, for 5 days, resulted in a sustained analgesic effect and decreased inflammation at the sciatic nerve
[46][1]. Daily administration of RgIA (2 and 10 nmol) for 7 and 14 days provided acute antinociceptive effects on both days in CCI rats, with long-lasting effects evident only by day 14
[33][2]. Additionally, treatment with RgIA attenuated the degree of inflammation in the sciatic nerve and DRG, reducing edema and infiltrate. Both dosages of RgIA also significantly decreased the density of glial cells in the spinal cord and prevented morphological derangements in DRG
[33][2]. The same α-conotoxin doses also showed pain-relieving and neuroprotective properties in a rat model of oxaliplatin-induced neuropathy
[35][3]. The repeated administration (2 and 10 nmol, intramuscularly) reduced the oxaliplatin-induced neuropathic pain hyperalgesia and allodynia, and DRG damage. In the spinal cord, the numerical increase of astrocyte cell density present in oxaliplatin-treated rats was partially prevented by RgIA treatment. Nevertheless, the administration of the α-conotoxin was able, per se, to elicit a numerical increase and a morphological activation of microglia and astrocytes in specific brain areas
[35][3], suggesting that RgIA may modulate glial cells in order to promote neurorestoration and reduce pain. Supporting this evidence, the administration of RgIA (2 nmol) for two weeks in oxaliplatin-treated rats elicited similar antinociceptive effects, avoiding DRG morphological alterations
[47][4] These observations were consistent with a progressive disease-modifying effects in neuropathic context by RgIA. The presence of α9α10 nAChRs in immune cells
[42,48,49,50][5][6][7][8] has led to the hypothesis that the long-term effects on neuropathic pain produced by RgIA might be due to the modulation of immune cells infiltration and the consequent release of inflammatory mediators into the site of injury, which may sensitize nerves to nociceptive stimuli
[51,52][9][10].
However, RgIA was also proposed to exert its analgesic effects by modulating the N-type VGCC Ca
V2.2
[53,54,55,56][11][12][13][14]. This inhibition is mediated by the activation of GABA
BR, and block also occurs in DRG neurons of α9 knockout (KO) mice
[54][12], suggesting that the α9α10 subtype is not involved in the modulation of the calcium channels. Inhibition of N-type channels, in turn, decreases neurotransmitter release and synaptic transmission between DRG neurons and second-order neurons in the spinal cord and thus impairs the transmission of nociceptive signals to the brain
[57][15]. This is the mechanism of action of another well-known
Conus-derived compound, ω-conotoxin MVIIA, an FDA approved drug known as Prialt (Elan Pharmaceuticals, Dublin, Ireland), for the treatment of chronic pain
[58,59][16][17].
RgIA (0.1 μM) was shown to inhibit VGCC currents by 40–50% in >75% of DRG neurons isolated from either mice or rats
[54[12][13],
55], an effect that was prevented by competitive GABA
BR antagonists
[55][13] or the knockdown of the GABA
BR by siRNA
[56][14]. Consequently, GABA
BR seems to be necessary for the inhibition of currents by the conotoxin. RgIA (10 μM) does not displace the binding of the antagonist [3 H]-CGP54626 to human GABA
BR transiently transfected in HEK293T
[60][18], suggesting a non-competitive mechanism. The GABA
BR also functionally couples to the G protein coupled inwardly rectifying potassium (GIRK) channels to attenuate nociceptive transmission
[61][19]. Recently, it has been shown that RgIA (1 μM) potentiates inwardly rectifying potassium currents in HEK293T cells heterologously expressing human GABA
BR coupled to GIRK1/2 channels
[62][20], albeit it failed to elicit the same effects in
Xenopus oocytes in a previous study
[60][18]. In support of the latter evidence, Wright et al.
[63][21] demonstrated that RgIA (0.1–1 μM) also had an insignificant effect on VGCC currents (inhibition >10% in <20% of rat DRG neurons), showing contrasting data to those obtained by Callaghan et al.
[55][13]. Moreover, no correlation was found between the responses induced by RgIA and baclofen, a GABA
BR agonist. By activating the presynaptic GABA
BR in DRG neurons, baclofen inhibits excitatory post-synaptic currents (EPSCs) and prevents the release of glutamate, which blocks pain transmission between the primary nociceptors and second-order neurons in the dorsal horn of the spinal cord
[34,64][22][23]. α-Conotoxins are charged peptides, so they may not cross the blood–brain barrier and reach spinal neuron synapses, suggesting that the VGCC mechanisms are not involved in RgIA-induced analgesia
[34,60,63][18][21][22]. These findings are consistent with a study that demonstrated no inhibition of EPSCs in the dorsal horn neurons by Vc1.1, another peptide of the α-conotoxin family
[64][23].
There are several conflicting results that make it difficult to determine the molecular mechanism by which α-conotoxins exert their analgesic effect. Many studies suggested that RgIA and some other α-conotoxins do not relieve pain by the activation of GABA
BR, reinforcing the idea of the involvement of α9α10 nAChRs
[34,36,60,63,64][18][21][22][23][24]. As mentioned above, RgIA and other α9α10 nAChRs antagonists have been associated with disease-modifying effects in neuropathic pain conditions
[9,33,34,35,36,60][2][3][18][22][24][25]. Known GABA
BR agonists, such as baclofen, did not show similar properties. RgIA4 (see below for its amino acid composition), a RgIA derivative that lacks GABA
BR activity, maintains the capacity to prevent the development of neuropathic pain
[36,65,66][24][26][27]. Moreover, RgIA, which induces calcium transients in murine granulocytes, decreased reactive oxygen species production
[42,44][5][28] and increased the production of the anti-inflammatory interleukin-10 in these cells. Even in this case, the α9α10 nAChR mechanism is more suitable than the GABA
BR mechanism for explaining such regulation, given that the inhibition of VGCC currents mediated by GABA
BR might result in lowering intracellular calcium, not in its increase
[42][5]. Finally, evidence from α9 subunit KO mice also supports a role for α9α10 nAChRs in pain
[36,67,68][24][29][30].
2. Anti-Colitis Effects of RgIA
In addition to analgesic and disease-modifying activities in animal models of pain, AlSharari et al. reported for the first time the anti-inflammatory effects of RgIA in the dextran sodium sulfate (DSS) experimental mice colitis model. Briefly, the lower doses (0.02 and 0.1 nmol) of the RgIA treatment in DSS-treated mice were inactive, whereas the higher dose (0.2 nmol) reversed the disease activity index score, loss of body weight, total histological damage score, as well as the colonic increase of the TNF-α concentration compared to the control group. Moreover, 0.2 nmol of RgIA significantly prevented the colon length shortening in DSS-treated mice
[40][31]. The block of α9α10 nAChRs on gut immune cells as an anti-inflammatory signal was suggested as a pharmacodynamic mechanism.
3. Anticancer Effects of RgIA
As mentioned above, nAChRs also have a major role in tumorigenesis and cancer progression
[37,38,39][32][33][34]. The expression of α9 and/or α10 subunits has been reported in several cancer cell lines, human tumors, and immune cells that could promote cancer-related inflammation
[42,48,49,50,69,70,71,72][5][6][7][8][35][36][37][38]. On these bases, α9α10 nAChR blockers were studied as antitumor agents. Intraperitoneal injections of 0.1 nmol/kg of RgIA led to changes of a degenerative nature of both cancer cells and leukocytes infiltrating the tumor in Ehrlich carcinoma (EC)-bearing mice
[43][39]. Interestingly, the α-conotoxin (1 nmol/kg) also enhanced the antitumor activity of splenocytes and increased the survival rate of EC-bearing mice by impairing tumor growth
[73][40]. In vitro, RgIA increased the cytotoxic effects of the lipoxygenase pathway inhibitors nordihydroguaiaretic acid (24 h of treatment) and baicalein (48 h of treatment) on EC cells, which display an increased activity of the arachidonic acid cascade
[74][41]. By contrast, a recent report showed that while baicalein exerted a significant antiproliferative and cytotoxic activities against C6 glioma cells, RgIA enhanced the proliferation of these cells
[75][42]. No additional studies on the use of RgIA as an anticancer agent are reported. RgIA may have contrasting effects on cell proliferation, depending on the concerned cell line. Better anticancer activity was reported for synthetic derivatives of RgIA, as described below for RgIA4
[71][37].
4. Derivatives of RgIA
Owing to its high specificity towards α9α10 nAChRs, RgIA also provided a promising lead for developing novel ligands that selectively target these receptor subtypes. Unfortunately, several α-conotoxins, including RgIA, were initially tested in rodent nAChRs-expressing systems or native receptors as well as preclinical pain studies were most often carried out on rodent models. The crystal structure of hα9 nAChR extracellular domain with RgIA revealed that RgIA also interacted with hα9α10 nAChRs, involving its Asp5-Pro6-Arg7 triad of loop I and Arg11 of loop II
[76][43]. However, there is a 300-fold difference (IC
50 rat 1.5 nM vs. IC
50 human 490 nM) of RgIA potency between the rat and human receptors, due to the residue at position 56 (Ile in humans vs. Thr in rats) of the α9 subunit
[77,78][44][45]. These species differences likely led to the failure of the clinical development of a potential hα9α10 antagonist based on Vc1.1
[79][46]. Hence, the derivatives of native α-conotoxins are needed to improve specificity and potency against hα9-containing nAChRs together with pharmacokinetics features (
Figure 31).
RgIA4 is the most reported RgIA-analogue in the literature data. This peptide has 5 of the 13 total amino acids modified but shows an increased potency and selectivity towards the hα9α10 subtype, retaining a high affinity for both rat and human receptors (IC
50 rat 0.9 nM vs. IC
50 human 1.5 nM)
[36][24]. With a 1200-fold lesser potency than that for α9α10, RgIA4 inhibits the hα7 subtype (IC
50 1.8 μM). Interestingly, RgIA4 lacks activity on GABA
BR, allowing the dissection of the involvement of α9α10 vs. GABA
BR in pain relief
[36][24]. Daily subcutaneous injections of RgIA4 (0.128, 16 and 80 μg/kg) prevented cold allodynia and mechanical hypersensitivity in oxaliplatin-treated rats, after only one week of treatment
[36][24]. The same treatment with 40 μg/kg of RgIA4 reversed oxaliplatin-induced cold allodynia in mice but only after 3 weeks of treatment. Interestingly, the effects lasted up to 3 weeks post-treatment, indicating a disease-modifying effect like that of the native peptide
[65][26]. Long-lasting effects have also been reported in a paclitaxel-induced neuropathy rat model, where animals subcutaneously injected with RgIA4 (80 μg/kg) showed a significant reduction of mechanical allodynia by day 12 post-treatment
[66][27]. These data support the effects of RgIA4 in multiple chemotherapy-induced neuropathy models. As anticipated above, RgIA4 (1 μM) completely abolished the proliferation of A549 adenocarcinoma cell line induced by 48 h treatment with 100 nM nicotine, blocking the nicotine-induced activation of Akt and ERK
[71][37].
However, RgIA4 is not an ideal drug candidate because of its high in vivo degradation and poor serum stability
[21][47]. Synthetic cyclization, with appropriately sized linkers between the N- and the C-termini, could be useful to overcome this issue. In fact, in a previous study, the insertion of a linker of 6 or 7 amino acids in the native RgIA (cRgIA-6 and cRgIA-7) caused an improvement in stability compared to the linear peptide, while maintaining a high potency towards the α9α10 nAChR
[28][48]. By contrast, incorporating smaller linkers does not make significant improvements, suggesting that they are likely to force the N- and the C-termini together, putting a strain on the peptide conformation
[28][48]. Similarly, the side chain cyclization of RgIA4 (analogue 6) led to an increased serum stability over linear RgIA4, while exhibiting a similar affinity with RgIA4 at the hα9α10 nAChR and analgesic effects in oxaliplatin-treated rats
[80][49]. Structurally, the lactam linkage introduced in RgIA4 stabilized the globular conformation and suppressed disulfide scrambling, leading to stability improvements.
Disulfide bond replacement with a dicarba unsaturated bridge is a further strategy to enhance the serum stability, as well as to modulate the specificity of conotoxins
[81,82][50][51]. Additionally,
[3,12][52][53]-dicarba RgIA analogues retain inhibition at the α9α10 nAChR but lack GABA
BR activity, whereas
[2,8][54][55]-dicarba analogues display reverse target selectivity, suggesting that the C
1-C
3 bridge is important for inhibiting the α9α10 nAChR and the C
2-C
4 bridge for GABA
BR modulation
[81][50]. Molecular dynamics simulations suggested that substitution at Cys2 and Cys8 abolishes the RgIA activity at α9α10, mainly by the reduction of contacts between RgIA-Tyr10 and α9-Arg138 and between RgIA-Arg9 and α10-Trp81
[81][50]. The importance of the C
1-C
3 bridge on the activity of RgIA has also been demonstrated by two recent works. The substitution of the Cys3 with the Cys surrogate L-penicillamine, together with amino acids replacements and addition of L-Arg in position 14, resulted in RgIA-5474, a peptide with a 9000-fold increased potency towards the hα9α10 nAChR (IC
50 0.05 nM) and improved serum stability compared to RgIA
[83][56]. To date, RgIA-5474 is one of the most potent RgIA-derivatives against hα9α10 nAChR. A fluorescently tagged derivative of this peptide has been synthesized using click chemistry and can be used to visualize native α9α10 nAChRs
[84][57]. Furthermore, the replacement of the C
2-C
4 bridge with unreducible methylene thioacetal and the same above-mentioned amino acids modifications led to the generation of RgIA-5524, a further compound with highly selectivity for hα9α10 nAChR (IC
50 0.9 nM), reduced serum degradation and pain-relieving effects in oxaliplatin-treated mice (40 μg/kg)
[68][30]. Hence, disulfide loop modifications could be helpful to improve biophysical properties of RgIA, by stabilizing its globular structure and avoiding disulfide scrambling.
In addition to the disulfide scrambling that makes its structure unstable, RgIA is more susceptible to proteolysis than other conotoxins due to its arginine-rich loop II and Tyr10, which provide cleavage sites for proteases. Ren et al. provided the solution by generating three RgIA D-amino acid scanning analogues (called peptides 13, 14, and 15) that are more resistant in serum and intestinal fluid than native RgIA
[85][58]. D-amino acids are less prevalent in nature, thus they are unlikely to be susceptible to enzymatic degradation. Notably, peptide 15 displayed a two-fold increase in the inhibition of hα9α10 nAChR, whereas peptide 13 had a two-fold more potency against rα9α10 nAChR, compared to RgIA
[85][58]. Peptide 13 retained its strong affinity because it contained D-enantiomer of Arg at the C-terminal, which is not involved in binding to α9α10
[31,85][58][59]. All the other synthesized D-enantiomer analogues showed decreased potency at the α9α10 receptor, indicating that amino acid chirality is important for the activity of these α-conotoxins. Separately, it was recently found that the substitution of Arg13 with Tyr ([R13Y]RgIA) significantly improved the potency of RgIA for hα9α10 nAChRs by 240-fold, nullifying the difference between rats and humans
[45][60]. Thus, scan strategies are useful in revealing the critical amino acid involved in the peptide-receptor interaction. Positional-scanning synthetic combinatorial libraries (PS-SCL) consist of a mixture of peptides where one or more of the positions are individually fixed at a specific amino acid, while the remaining positions are comprised of an equimolar mixture of amino acids
[86][61]. PS-SCL of RgIA# ([∆R13]RgIA) based on positions 9, 10, and 11 provided 10,648 possible conotoxin sequences, but, in the end, only three lead compounds (5, 17, and 26) displayed increased antinociception compared to the native peptide when intraperitoneally injected in mice (10 mg/kg)
[87][62].
Finally, a further way to enhance the activity of α-conotoxins is dimerization. PEG9-dimeric RgIA# resulted in increased inhibitions at hα7 (IC
50 dimeric RgIA# 63.1 nM vs. IC
50 RgIA# >1000 nM) and hα9α10 (IC
50 dimeric RgIA# 38.5 nM vs. IC
50 RgIA# 248.7 nM), by 17- and 7-fold, respectively, compared to RgIA#
[88][63]. Considering the role of these nAChRs in the development of cancer
[71][37], dimeric RgIA# could be a useful pharmacological tool in cancer research. The PEG-linker also allows the conotoxin to simultaneously bind two adjacent binding sites, and further elongation of the linker should not significantly affect the activity of the dimer. Indeed, the PEG6-dimeric RgIA# showed comparable potency to the PEG9-dimeric RgIA#
[88][63].
Figure 31. Sequences of the RgIA analogues reported in this
entr
eviewy. *, citrulline; ‡, 3-Iodo-Tyrosine; [], side chain cyclization; X, positional sequence insertion of L-allylglycine residues prior to ring-closing olefin metathesis cyclization; §, L-penicilammine; β, 3-homo tyrosine; £, methylene thioacetal replacement; #, amidated C-terminal; r, D-arginine; $, norleucine
[11,28,36,45,65,68,81,83,85,87,88][24][26][30][48][50][56][58][60][62][63][64].
Overall, chemical modifications of α-conotoxins may not only be helpful to improve their pharmacokinetic and pharmacodynamic features, but also to elucidate the molecular mechanisms underlying their activities, in order to facilitate their use as drug leads.