Review article
Challenges in the design of insulin and relaxin/insulin-like peptide mimetics

https://doi.org/10.1016/j.bmc.2017.09.030Get rights and content

Abstract

Peptidomimetics are designed to overcome the poor pharmacokinetics and pharmacodynamics associated with the native peptide or protein on which they are based. The design of peptidomimetics starts from developing structure-activity relationships of the native ligand-target pair that identify the key residues that are responsible for the biological effect of the native peptide or protein. Then minimization of the structure and introduction of constraints are applied to create the core active site that can interact with the target with high affinity and selectivity. Developing peptidomimetics is not trivial and often challenging, particularly when peptides’ interaction mechanism with their target is complex. This review will discuss the challenges of developing peptidomimetics of therapeutically important insulin superfamily peptides, particularly those which have two chains (A and B) and three disulfide bonds and whose receptors are known, namely insulin, H2 relaxin, H3 relaxin, INSL3 and INSL5.

Introduction

Currently available drugs in the market fall broadly into two categories. There are ‘small molecule’ drugs (molecular weight of <500 Da) with oral bioavailability and much larger ‘biologics’ (molecular weight of typically >5000 Da) with no oral bioavailability. Due to their small size, small molecule drugs often suffer from reduced target specificity and toxicity. Large biologics, on the other hand, are highly target-specific and thus less toxic than small molecules. Therefore, the compounds that fit between these two molecular weights (500–5000 Da) and possess the advantages of both the small molecule (e.g. bioavailability and stability) and larger biologics (e.g. highly target specific) are of great interest. Peptidomimetics are such compounds that fall into this category.

Peptidomimetics are designed to circumvent the problems associated with the native peptide or protein on which they are based (e.g. poor proteolytic stability and bioavailability).1 They can also be improved in terms of potency and target specificity. Hence peptidomimics have great potential as synthetic tools of drug discovery. The design process of mimics starts from developing structure-activity relationships (SAR) of the native ligand-target pair that identify the key residues that are responsible for the biological effect of the native peptide or protein. Then minimization of the structure and introduction of constraints (e.g. cyclization through hydrocarbon stapling, lactam, Disulfide Bridge, disulfide isostere, N-methylation) are applied to create the core active site that can interact with the receptor or enzyme with high affinity and selectivity.1, 2, 3, 4, 5, 6, 7, 8, 9 Developing peptidomimetics is not trivial and often challenging, particularly when peptides’ interaction mechanism with their target is complex. This review will discuss the challenges of developing peptidomimetics of therapeutically important insulin superfamily peptides.

In humans, there are ten insulin superfamily peptides that includes three insulin family peptides (insulin, insulin-like growth factor (IGF) I and II and seven relaxin family peptides. The latter consists of human relaxin 1 (H1 relaxin), human relaxin 2 (H2 relaxin), human relaxin 3 (H3 relaxin), insulin-like peptide 3 (INSL3 also known as relaxin-like factor or Leydig insulin-like peptide), insulin-like peptide 4 (INSL4 or placentin), insulin-like peptide 5 (INSL5) and insulin-like peptide 6 (INSL6).10, 11, 12, 13, 14, 15, 16 Except IGF-I and II, which are single-chain peptides, all others are proven or predicted to be mature hormones with two chains (A and B) linked by two inter-chain disulfide bonds and one intra-A-chain disulfide bond17, 18 [Fig. 1A and B]. Insulin acts on tyrosine kinase receptors [Fig. 1C], the relaxin and INSL peptides, on the other hands, act upon G protein-coupled receptors that are now known as relaxin family peptide (RXFP) receptors [Fig. 1D].19

The native receptors for INSL4 and INSL6 are yet to be identified.20 H2 relaxin is the ortholog of the relaxin-1 gene found with relaxin-3 in most other mammalian species21 and it is presumed that H1 relaxin is also a ligand of RXFP1. While the function of H1 relaxin is unknown, all the family members of relaxin peptides and their target receptors (RXFPs) have wide physiological roles and clinical significance (reviewed in: Bathgate et al.22).

While insulin family peptides share little sequence homology apart from six cysteines that form three disulfide bonds [Fig. 1A], the overall fold of these peptides is similar [Fig. 1B]. However, the differences in their amino acid sequences have resulted in subtle changes in the length of the conserved helices and their functional selectivity towards their receptors. Detailed SAR towards understanding the role of residues in each of these chains have been reported over the years. This review summarizes the current knowledge of the SAR of insulin, H2 relaxin, H3 relaxin, INSL3 and INSL5 and describes how this knowledge has been used to develop peptidomimetic agonists and antagonists of RXFP receptors.

Section snippets

Insulin and insulin receptor

Mature human insulin consists of two chains, an A chain of 21 residues, and a B chain of 30 residues (51 residues in total, molecular weight 5808). They are linked by two inter-chain disulfide bridges and one intra-A-chain bridge [Figs. 1A, 2A, B]. The crystal structure of insulin was solved in 1969.23 Insulin forms dimers and hexamers (3 dimers of insulin) in the presence of zinc in humans.

The insulin receptor (IR) is a transmembrane tyrosine kinase protein with momodimeric (αβ)2 structure

3.1. Binding sites within insulin molecule

Extensive biochemical and structural characterization of insulin demonstrated two binding sites within the molecule. Site 1 (classical or primary binding site) utilizes 13 residues (GA1, IA2, VA3, QA5, TA8, YA19, NA21, VB12, YB16, GB23, FB24, FB25, and YB26) which forms the surface of the insulin dimer [Fig. 2A].24, 25, 26 Site 2 (secondary binding site) comprises 6 residues (SA12, LA13, EA17, HB10, EB13, and LB17) which forms the surface of the insulin hexamer24, 25, 26 [Fig. 2B]. The

Insulin mimetics

Despite decades of endeavour and the recent elucidation of the structure of the insulin-bound IR site 1,37 the mechanism of insulin-induced signal transduction remains largely elusive. The extremely complex interaction between insulin and IR made it almost impossible to develop peptidomimetics as truncating or minimising the structure of insulin significantly alters its biological activity. Hence no small molecule agonist of the IR has been discovered to date. Thus all of the insulin analogues

Human relaxin 2 (H2 relaxin) and relaxin family peptide receptor 1 (RXFP1)

The native form of the H2 relaxin peptide has been isolated from human ovary49 and prostate50 and demonstrated to contain a 24 residue A-chain and a 29 residue B-chain (ie. 53 residues, molecular weight 5963)24, 51 [Fig. 1A]. Although this is a major circulating form of H2 relaxin, there are two other equally active isoforms found in vivo with longer B-chains, B1–31 and B1–33.52 The X-ray crystallography and solution structure of H2 relaxin were solved in 199153 and 2016 [Fig. 4A]54

Binding and activation mechanism

Studies on the mode of binding and activation of RXFP1 by H2 relaxin have demonstrated that both the A- and B-chain seem to be necessary for full activity. Interaction with RXFP1 requires several interactions involving the LRR domain,58 the linker between the LRR and LDLa and the extracellular TM loops.59, 60, 61 Importantly the linker and N-terminal LDLa module are essential for activation and may act together as a tethered agonist61, 62, 63, 64, 65 [Fig. 4D]. The specific aspects of binding

H2 relaxin peptidomimetics

The complex mechanism of H2-RXFP1 interaction, as discussed above, suggests considerable challenges for developing peptidomimetics of the H2 relaxin without altering its biological activity. However, as the A-chain is thought to possesses low affinity secondary binding residues74 and the mid-region of the B-chain contains the high affinity primary binding residues69 it was hypothesized that it might be possible to develop a minimised peptidomimetic or even a B-chain-only mimetic of H2 relaxin

Human relaxin 3 (H3 relaxin) and relaxin family peptide receptor 3 (RXFP3)

Mature H3 relaxin is predicted to consist of two chains, an A chain of 24 residues, and a B chain of 27 residues (51 residues, molecular weight 5500)19, 51, 81 [Figs. 1A, 6A] based on the native porcine relaxin-3 sequence isolated from pig brain.18 The chains are linked by two inter-chain disulfide bridges and one intra-A-chain bridge [Figs. 1A, 6A]. The solution structure of H3 relaxin was solved in 2008 [Fig. 6A].82 The cognate receptor for H3 relaxin is the GPCR, RXFP3 [Fig. 1E].18 However

Binding and activation mechanism

There are distinct differences in the residues and modes of binding and activation of the different receptors – RXFP1, RXFP3 and RXFP4 by H3 relaxin. Our studies demonstrate that deletion of the N-terminal residues of the B-chain does not affect the ligand binding activity for RXFP1, RXFP3 or RXFP4.85, 86, 87 H3 relaxin contains an RXXXRXXI motif (RB12, RB16 and IB19) and mutation studies85, 86 confirm that, like H2 relaxin, this is the driver of high affinity RXFP1 binding.88 The binding of H3

Receptor-specific H3 relaxin mimetics

From SAR studies it is clear that the H3 relaxin B-chain is responsible for binding and activation of RXFP3 and RXFP4. Other studies suggest that the A-chain of H3 relaxin may play a role in interacting with RXFP1. Therefore, in an attempt to develop a receptor-specific agonist, the A-chain of H3 relaxin was replaced with that from INSL5 and the resulting chimeric R3/I5 analogue exhibited selectivity for RXFP3 and RXFP4 over RXFP1.92 Additionally, by truncating the C-terminus of the B-chain of

Human insulin 3 (INSL3) and relaxin family peptide receptor 2 (RXFP2)

The predicted structure of human INSL3 involves an A-chain of 26 residues and a B-chain of 31 residues (57 residues, molecular weight 6292). Both chains are connected by three disulfide bridges [Figs. 1A, 7A]. The solution structure of INSL3 determined by our lab in 2006 [Fig. 7A]97 suggests insulin like folding of the peptide. The cognate receptor for INSL3 is the G protein-coupled receptor, RXFP2.57 The receptor is closely related to RXFP1 with a complex structure containing the 7 TMD, 10

Primary binding site

Like H2 relaxin, INSL3 binds to the LRR domain of RXFP2 using B-chain-specific residues: HB12, RB16, VB19, RB20 and WB2798, 99 [Figs. 1A, 7A]. The complementary binding site in the LRR of RXFP2 that interact with the B-chain of INSL3 was also determined and characterised by receptor mutation studies and demonstrated to be distinct from the H2 relaxin-RXFP1 site.100

Secondary binding site

While no single amino acid within the INSL3 A-chain was found to be critical for RXFP2 binding and activity, the potency of the

Minimised INSL3 mimetics

As part of the B-chain is important for RXFP2 binding and the A-chain is necessary for activation,98, 99, 101, 102 the minimised INSL3 peptidomimetic INSL3:A5–26/B7–27 was developed as an RXFP2 agonist [Figs. 7B, 5].101 This peptide contains truncated A- and B-chains and is almost one-quarter smaller than the native peptide and thus easier and cheaper to make and represents a potential lead peptide for further development to probe biological function of RXFP2. A high-affinity RXFP2 antagonist

Insulin-like peptide 5 (INSL5) and relaxin family peptide receptor 4 (RXFP4)

Human insulin-like peptide 5 (INSL5) has a predicted structure with 2 chains (A-chain 21 amino acids and B chain 24 amino acids; Molecular weight 5048) connected by three disulfide bridges [Figs. 1A, 8A]. INSL5 was originally identified as a novel peptide hormone with high expression in the gut.14 The solution NMR structure of human INSL5108 determined by our laboratory (2009) indicates that the B-chain has an extended C-terminal α-helix compared with H3 relaxin Fig. 8A].109 Subsequent studies

Binding and activation mechanism

Human INSL5 is one of the most difficult peptides to produce chemically as both A and B chains are unusually resistant to standard synthesis protocols and require highly optimised conditions for their acquisition.110 Therefore, preliminary SAR studies were carried out using mouse INSL5 whose sequences were predicted to be easier to assemble. Interestingly, mouse INSL5 was found to be almost 10 times more potent than hINSL5 on human RXFP4.111 We identified a residue, KA15, in the A-chain of

INSL5 peptidomimetics

These SAR studies led to the design and synthesis of the simplified peptidomimetics, mINSL5:A8–21118 and hINSL5:A8–21K15 [Figs. 8B, 5] 112. The latter analogue contains one mutation where human residue TA15 was replaced with equivalent mouse residue KA15.112 These analogues have 2 disulfide bridges and a truncated A-chain structure. They exhibited native hINSL5-like RXFP4 affinity and potency and were potent full agonists of both cAMP inhibition and pERK activation. They are good template for

Conclusion

This review has highlighted the distinct mechanisms utilized by insulin and relaxin-insulin-like peptides for binding and activation of their native receptors. Knowledge of these mechanisms has led the development of peptidomimetics which act as agonists or antagonists. However, the distinct differences in peptide structure and receptor interaction means that each ligand-receptor pairing has required a different approach for mimetic design. Studies on this unique peptide family therefore

Acknowledgements

This research was partly funded by NHMRC (Australia) project grants (1122170, 1023321, 1065481, and 1100676) and ARC linkage grant (LP120100654) to M.A.H. and R.A.D.B. Research at The Florey Institute of Neuroscience and Mental Health is supported by the Victorian Government Operational Infrastructure Support Program. Some of the figures were prepared by Praveen, a PhD student at the Florey.

Conflict of interest

The authors declare no conflict of interest.

References (118)

  • T. Kurose et al.

    Cross-linking of a B25 azidophenylalanine insulin derivative to the carboxyl-terminal region of the alpha-subunit of the insulin receptor. Identification of a new insulin-binding domain in the insulin receptor

    J Biol Chem

    (1994)
  • S.J. Chan et al.

    Complementation analysis demonstrates that insulin cross-links both alpha subunits in a truncated insulin receptor dimer

    J Biol Chem

    (2007)
  • E. Ciszak et al.

    Role of C-terminal B-chain residues in insulin assembly: the structure of hexameric LysB28ProB29-human insulin

    Structure

    (1995)
  • G.M. Williams et al.

    Replacement of the CysA7-CysB7 disulfide bond with a 1,2,3-triazole linker causes unfolding in insulin glargine

    Org Biomol Chem

    (2015)
  • Q.X. Hua et al.

    Design of an active ultrastable single-chain insulin analog: synthesis, structure, and therapeutic implications

    J Biol Chem

    (2008)
  • R.C. Pillutla et al.

    Peptides identify the critical hotspots involved in the biological activation of the insulin receptor

    J Biol Chem

    (2002)
  • C.F. Lawrence et al.

    Insulin mimetic peptide disrupts the primary binding site of the insulin receptor

    J Biol Chem

    (2016)
  • C. Eigenbrot et al.

    X-ray structure of human relaxin at 1.5 A. Comparison to insulin and implications for receptor binding determinants

    J Mol Biol

    (1991)
  • J. Kumagai et al.

    INSL3/Leydig insulin-like peptide activates the LGR8 receptor important in testis descent

    J Biol Chem

    (2002)
  • E.E. Büllesbach et al.

    The Trap-like Relaxin-binding Site of the Leucine-rich G-protein-coupled Receptor 7

    J Biol Chem

    (2005)
  • S. Sudo et al.

    H3 relaxin is a specific ligand for LGR7 and activates the receptor by interacting with both the ectodomain and the exoloop 2

    J Biol Chem

    (2003)
  • N.A. Diepenhorst et al.

    Investigation of Interactions at the Extracellular Loops of the Relaxin Family Peptide Receptor 1

    J Biol Chem

    (2014)
  • D.J. Scott et al.

    Characterization of novel splice variants of LGR7 and LGR8 reveals that receptor signaling is mediated by their unique low density lipoprotein class A modules

    J Biol Chem

    (2006)
  • E.J. Hopkins et al.

    The NMR solution structure of the relaxin (RXFP1) receptor lipoprotein receptor class a module and identification of key residues in the N-terminal region of the module that mediate receptor activation

    J Biol Chem

    (2007)
  • R.C.K. Kong et al.

    The relaxin receptor (RXFP1) utilizes hydrophobic moieties on a signaling surface of its N-terminal low density lipoprotein class a module to mediate receptor activation

    J Biol Chem

    (2013)
  • E.E. Bullesbach et al.

    Total synthesis of human relaxin and human relaxin derivatives by solid-phase peptide synthesis and site-directed chain combination

    J Biol Chem

    (1991)
  • E.E. Bullesbach et al.

    The receptor-binding site of human relaxin II. A dual prong-binding mechanism

    J Biol Chem

    (1992)
  • E.E. Bullesbach et al.

    The relaxin receptor-binding site geometry suggests a novel gripping mode of interaction

    J Biol Chem

    (2000)
  • E.E. Bullesbach et al.

    The trap-like relaxin-binding site of the leucine-rich G-protein-coupled receptor 7

    J Biol Chem

    (2005)
  • M.A. Hossain et al.

    The A-chain of human relaxin family peptides has distinct roles in the binding and activation of the different relaxin family peptide receptors

    J Biol Chem

    (2008)
  • M.A. Hossain et al.

    Chimeric relaxin peptides highlight the role of the A-chain in the function of H2 relaxin

    Peptides

    (2012)
  • J.I. Park et al.

    Regulation of receptor signaling by relaxin A chain motifs: derivation of pan-specific and LGR7-specific human relaxin analogs

    J Biol Chem

    (2008)
  • L.J. Chan et al.

    Identification of key residues essential for the structural fold and receptor selectivity within the a-chain of human gene-2 (H2) relaxin

    J Biol Chem

    (2012)
  • M.A. Hossain et al.

    The minimal active structure of human relaxin-2

    J Biol Chem

    (2011)
  • M.A. Hossain et al.

    A single-chain derivative of the relaxin hormone is a functionally selective agonist of the G protein-coupled receptor, RXFP1

    Chem Sci

    (2016)
  • S.A. Marshall et al.

    B7–33 replicates the vasoprotective functions of human relaxin-2, (serelaxin)

    Eur J Pharmacol

    (2017)
  • K.J. Rosengren et al.

    Solution structure and novel insights into the determinants of the receptor specificity of human relaxin-3

    J Biol Chem

    (2006)
  • C. Liu et al.

    INSL5 is a high affinity specific agonist for GPCR142 (GPR100)

    J Biol Chem

    (2005)
  • C. Kuei et al.

    R3(BDelta23 27)R/I5 chimeric peptide, a selective antagonist for GPCR135 and GPCR142 over relaxin receptor LGR7: in vitro and in vivo characterization

    J Biol Chem

    (2007)
  • M.A. Hossain et al.

    The A-chain of human relaxin family peptides has distinct roles in the binding and activation of the different relaxin family peptide receptors

    J Biol Chem

    (2008)
  • E.E. Büllesbach et al.

    The relaxin receptor-binding site geometry suggests a novel gripping mode of interaction

    J Biol Chem

    (2000)
  • K.J. Rosengren et al.

    Solution structure and characterization of the LGR8 receptor binding surface of insulin-like peptide 3

    J Biol Chem

    (2006)
  • E.E. Büllesbach et al.

    The mode of interaction of the relaxin-like factor (RLF) with the leucine-rich repeat G protein-activated receptor 8

    J Biol Chem

    (2006)
  • K.J. Rosengren et al.

    Solution structure and characterization of the LGR8 receptor binding surface of insulin-like peptide 3

    J Biol Chem

    (2006)
  • C. Recio et al.

    The potential therapeutic application of peptides and peptidomimetics in cardiovascular disease

    Front Pharmacol

    (2016)
  • A. Brik

    Metathesis in peptides and peptidomimetics

    Adv Synth Catal

    (2008)
  • M.G. Bursavich et al.

    From peptides to non-peptide peptidomimetics: design and synthesis of new piperidine inhibitors of aspartic peptidases

    Org Lett

    (2001)
  • D. Goyal et al.

    Rationally designed peptides and peptidomimetics as inhibitors of amyloid-beta (Abeta) aggregation: potential therapeutics of Alzheimer's disease

    ACS Comb Sci

    (2017)
  • X. Zhang et al.

    Antibody mimetics, peptides, and peptidomimetics

    Methods Mol Biol

    (2017)
  • M. Vodnik et al.

    Ghrelin receptor ligands reaching clinical trials: from peptides to peptidomimetics; from agonists to antagonists

    Horm Metab Res

    (2016)
  • Cited by (16)

    • Encapsulation of an insulin-modified phosphatidylcholine complex in a self-nanoemulsifying drug delivery system (SNEDDS) for oral insulin delivery

      2020, Journal of Drug Delivery Science and Technology
      Citation Excerpt :

      In both forms of DM, types 1 or 2, blood glucose levels can be modulated through the controlled dosing of insulin [4,5]. Human insulin is a hormone that consists of 2 peptide chains with a total of 51 amino acids [6]. The conventional administration method for insulin is via subcutaneous injection [7].

    • Transdermal Drug Delivery: Concepts and Application

      2020, Transdermal Drug Delivery: Concepts and Application
    • Single chain peptide agonists of relaxin receptors

      2019, Molecular and Cellular Endocrinology
      Citation Excerpt :

      There are seven insulin-like relaxin family peptides in humans that includes human relaxin 1 (H1 relaxin), human relaxin 2 (H2 relaxin), human relaxin 3 (H3 relaxin), insulin-like peptide 3 (INSL3 also known as relaxin-like factor or Leydig insulin-like peptide), insulin-like peptide 4 (INSL4), insulin-like peptide 5 (INSL5) and insulin-like peptide 6 (INSL6) (Hossain and Wade, 2014; Bathgate et al., 2002, 2013; Bathgate et al., 2013a,b; Patil et al., 2017; Hossain and Bathgate, 2017; Adham et al., 1993; Chang et al., 2000; Chassin et al., 1995; Conklin et al., 1999; Crawford et al., 1984; Hudson et al., 1983).

    • Targeting the relaxin/insulin-like family peptide receptor 1 and 2 with small molecule compounds

      2019, Molecular and Cellular Endocrinology
      Citation Excerpt :

      Hence, it will be useful to design small molecules that target the relaxin receptor, RXFP1 or RXFP2 for clinical applications in various diseases. The complex binding mode of endogenous peptide ligands to both large N-terminal ectodomain and extracellular loops of RXFP1 and RXFP2 (Hossain and Bathgate, 2018) makes it difficult to design small molecules replicating such complex interactions in orthosteric sites. As the high-resolution structure of the seven transmembrane domain (7TM) of these receptors transmitting ligand-induced conformational changes into activation of G proteins remains unresolved, the structure mapping design of synthetic allosteric agonists is only possible using modeling based on homology with better characterized GPCRs.

    • Use of Antidiabetic, Antihypertensive, and Psychotropic Drugs in Pregnancy

      2018, Side Effects of Drugs Annual
      Citation Excerpt :

      Moreover, the current insulin preparations have not been shown to cross the placenta [2c, 3M, 4c]. Currently, insulin analogs used in clinical practice are mature two-chain peptides which possess different pharmacokinetic profiles [5H]. Metformin is a biguanide that work by inhibiting glucose production in the liver, decreases the absorption of glucose in the intestine, and stimulates glucose uptake in the peripheral tissues.

    View all citing articles on Scopus
    View full text