Thiol-derivatized minihepcidins retain biological activity

Abstract

Minihepcidins are small peptides that mimic biological activity of the iron-regulatory hormone hepcidin. Structurally, they contain thiol-free-cysteine residue in position 7 which is crucial for their bioactivity. Nonetheless, free sulfhydryl group is not desirable in pharmaceutical entities as it may lead to dermatological side effects. Moreover free thiol moiety is quite reactive and depending on conditions/reagents may be alkylated and/or oxidized giving various Cys-derivatives: S-alkyl cysteines, sulfoxides, sulfones, disulfides, cysteinesulfinic and cysteic acids. To limit such reactivity and maintain bioactivity of minihepcidin(s) we used thiol-protection strategy based on activated vinyl thioethers. Novel S-protected analogs of physiologically active minihepcidin PR73 were synthesized and tested in vitro showing activity comparable to parental molecule. The most active compound, PR73SH was also tested in vivo showing activity profile analogous to PR73. Collectively, our findings suggest that S-vinyl-derivatization of minihepcidin(s) may be a suitable approach in the development of physiologically active agonists of hepcidin.

Hepcidin is a key regulator of iron homeostasis¹ and its abnormal production has been associated with several common iron disorders². This 25 amino acids long peptide hormone is mainly produced in the liver³ although other tissues were also implicated^4;5. The iron-regulatory function of hepcidin is mediated through its receptor, ferroportin (Fpn)⁶, which is the sole known cellular iron exporter in vertebrates⁷. Mechanistically, hepcidin’s binding to Fpn causes the receptor internalization and subsequent proteasomal degradation, effectively reducing systemic iron availability, by decreasing both, the absorption of iron (duodenum) and release of recycled iron (macrophages)⁸.

Hepcidin and related pathways, have recently emerged as an attractive target for the development of novel therapeutics for iron disorders, with diverse group of leading compounds under development, such as monoclonal antibodies, proteins, peptides, oligonucleotides and small organic molecules^9–11.

We previously described minihepcidins, rationally designed peptide-based hepcidin agonists, with potent in vitro and in vivo bioactivity^12;13, which are currently under commercial development by Merganser Biotech LLC. Collectively, available data strongly suggest that minihepcidins may be useful for the prevention of iron overload, or as an auxiliary in combination with phlebotomy or chelation for the treatment of existing iron overload. Structurally, these analogs contain unprotected free-cysteine residue in position 7, which is crucial for their bioactivity. However, drugs containing/releasing free sulfhydryl group(s) may be problematic in pharmaceutical development due to either (1) decreased stability associated with inherent free-thiol reactivity (S-alkylation/oxidation), and/or (2) dermatological side effects (e.g. skin eruptions)^14;15. Therefore, we decided to test whether minihepcidins can be efficiently S-derivatized and retain their bioactivity. Based on previous findings we concluded that most obvious –S-S-R derivative(s) (e.g. S-tertbutylthio), although viable, may not represent an optimal protecting strategy, as such analog (C7-SStBut) showed significantly lower activity compared to parental free-SH peptide (hep9)¹². As an alternative protective moiety we decided to use 1,2-double substituted vinyl-sulfides which may be efficiently synthesized from corresponding electron-deficient alkynes and unprotected free-cysteine containing peptides in aqueous media¹⁶ (Figure 1). Notably, similar cysteine-protecting/labeling synthetic schemes employing, both mono-^17;18 and double-substituted alkynes^17–19, were recently reported. Analogous reaction(s) employing allenes were also successfully used to create vinylthioether-linkage^20;21. In our case, all employed electron-deficient alkynes were commercially available and symmetrical (X₁=X₂), except for PR73SE (X₁=phenyl, X₂=−SO₂NH₂). For structural details please see Figure 2¹⁶. Parental minihepcidin PR73 was synthesized as previously described¹². S-alkylation reaction was carried out in the mixture of 80% 1,4-dioxane in water due to physical properties of PR73 (e.g. relatively high hydrophobicity resulting from the presence of N-palmitylamide) and proceeded efficiently over 25 min (room temp.). The use of acetylenedicarboxylic acid as alkylating reagent lead to partially decarboxylated analog PR73SD, which is consistent with previously reported data²². Obtained analogs were purified by RP-HPLC and characterized by matrix-assisted laser desorption ionization spectrometry (MALDI-MS) as well as analytical RP-HPLC²³ (Table 1).

Figure 1.

General synthetic scheme for derivatization of minihepcidin PR73.

Figure 2.

General structure of S-derivatized PR73 analogs.

Table 1.

Analytical and in vitro activity data for S-alkylated PR73 analogs.

Peptide	Composition	MW Calc/Found	RT [min]	EC50 [nM] TREX-hFpn-GFP cells
PR73	C86H133N21O15S	1733.19/1734.34	47.11	4.2±0.3
PR73SA	C98H151N21O19S	1959.46/1959.80	52.47	6.3±1.2
PR73SB	C94H143N21O19S	1903.35/1904.58	49.44	10.4±1.2
PR73SC	C92H139N21O19S	1875.30/1876.60	48.32	12.6±1.8
PR73SD	C89H135N21O17S	1803.24/1803.66	46.60	218.1±13.4
PR73SE	C94H140N22O17S2	1914.40/1915.02	48.52	34.0±5.4
PR73SF	C96H151N21O19S3	1999.56/1999.80	52.89*	10.0±3.4
PR73SG	C90H137N23O17S	1845.28/1846.59	49.33*	8.4±2.5
PR73SH	C96H153N21O21P2S	2031.40/2031.33	52.18*	1.1±0.1

Analytical RP-HPLC was performed using an analytical reversed-phase C4 XBridge™ BEH300 column, 4.6×150 mm, 3.5 μm (Waters, Milford, MA), or ⁽*⁾ an analytical reversed-phase C18 SymmetryShield™ column, 4.6×250 mm, 5 μm (Waters, Milford, MA).

Novel S-protected analogs were tested in vitro using a previously described cellular assay based on Fpn degradation²⁵ (results summarized in Table 1). Generally, all new analogs showed high potency in the low nanomolar range. However only one analog (PR73SH, EC₅₀=1.1±0.1 nM) showed bioactivity higher than parental PR73 (EC₅₀=4.2±0.3 nM). Interestingly, chemical making of S-substituent doesn’t appear to have clear, defined impact on bioactivity, rather overall steric hindrance plays the major role, with the most bulky substituents having favorable properties. The hydrophobicity may also play some role, as the activity increases in the carboxy-esters-substituent(s) order:

$$-{\mathrm{CH}}_3<-{\mathrm{C}}_2{\mathrm{H}}_5<-{\mathrm{C}}_4{\mathrm{H}}_9(\mathrm{PR}73\mathrm{SC}<\mathrm{PR}73\mathrm{SB}<\mathrm{PR}73\mathrm{SA})$$

Moreover, the geometry of the vinyl substituents (planar versus tetrahedral) does not appear to significantly influence activity, as planar analog PR73SA has fairly similar potency to its tetragonal counterpart (PR73SF). Considering that remaining tetragonal analog PR73SH shows highest activity, and the fact that all 3 analogs (e.g. PR73SA, PR73SF and PR73SH) are chemically fairly similar having the same number of substituent(s)-carbon-atoms (2×4=8), overall volume/space occupied by S-attached moiety appears again as important factor, with the activity increasing from most compact (PR73SF) to most bulky (PR73SH) substituent(s). Consistently, PR73SD which has the most hydrophilic and least bulky substituent shows the lowest potency (EC₅₀=218.1±13.4 nM).

Based on in vitro results, we selected PR73SH as suitable candidate for animal studies²⁶, which were carried out as previously described^12;13;25. We compared directly PR73SH and PR73 in vivo activity by assaying serum iron levels at 3 time points: (6, 24 and 48 hours) and concentrations that were previously shown to be sufficient for PR73 to exert potent bioactivity (50–100 nmoles/mouse). Because disulfide bond (–S-S-) crosslinking between minihepcidin C7 and Fpn C326 seems to be essential for their interaction¹², we considered that protection of free –SH group in PR73SH might results in a “delayed” onset of activity, analogous to previously observed effects¹² and/or extended time of bioactivity. Nonetheless, PR73SH activity mirrored the parental PR73 activity profile, with decreased serum iron observed at 6 and 24 h time points, but not at 48 h time point (Figure 3B). Since no significant activity difference between PR73 and PR73SH was observed in either, in vitro or in vivo experiments, we conclude that proposed free-cysteine derivatization scheme of minihepcidins is viable option in future development of these promising drug candidates. Notably, PR73SH appears also to be remarkably stable in mildly oxidizing conditions as prolonged storage of the compound in DMSO (10 mM solution) at room temperature for 30 days shows very limited levels of decomposition or sulfide oxidation (99.5±0.5% of stability, determined by LC/MS/MS experiments). Interestingly, compounds containing similar functional group(s) described to date^17–19 show diverse bio-stability levels.

Figure 3.

Comparison of in vitro and in vivo activity of PR73 and PR73SH: (A) the representative examples of in vitro dose response curves obtained for PR73 and PR73SH analogs using ferroportin degradation assay, (B) in vivo activity of PR73 and PR73SH at 6, 24 and 48h time-points.

In conclusion, a new scheme of free-thiol derivatization in peptides was described and applied in synthesis of S-substituted analogs of potent minihepcidin PR73. Various 1,2-double substituted vinyl-sulfides of the peptide (PR73SA-PR73SH) were tested in vitro showing activity comparable to the parental compound (PR73), with the most potent analog PR73SH being slightly more active then original peptide. PR73SH was additionally tested in vivo showing activity profile similar to the parental free-thiol containing analog (PR73). Considering its simplicity, the protocol described in our study may be useful as a general method of free-thiol derivatization in pharmaceutically relevant peptides.