Structure-Based Functional Analysis of a Hormone Belonging to an Ecdysozoan Peptide Superfamily: Revelation of a Common Molecular Architecture and Residues Possibly for Receptor Interaction

The crustacean hyperglycemic hormone (CHH) peptide superfamily, consisting of a group of structurally and functionally diverse peptides, has long attracted research interest for the study of functional diversification through structural changes [1,2,3,4,5,6,7,8,9,10]. Peptides constituting the superfamily, once thought to be restricted to crustaceans, have now been clearly shown to be present across clades of Ecdysozoa [11].

Inferences through classical ablation and replacement experiments suggested the presence in the eyestalk of crustaceans of an array of humoral factors—presumptive hormones regulating metabolism, reproduction, molting and growth, cardiac activity, color changes, etc. [12]—attesting to the importance of the X-organ/sinus gland (XO/SG) complex, a neuroendocrine tissue in the eyestalk, in the regulation of crustacean physiology. Most of the presumptive factors have since been biochemically purified and characterized, and their biological activity confirmed; these include a group of sequence-related peptides—CHH (the prototypic member of the superfamily), molt-inhibiting hormone (MIH), gonad-inhibiting hormone (GIH) or vitellogenesis-inhibiting hormone (VIH), and mandibular organ-inhibiting hormone (MOIH) [11], that once constituted the CHH family [13,14,15]. Identification in the locust Schistocerca gregaria of ion transport peptide (ITP), which stimulates ileal salt and water reabsorption, as an insect member [16], expanded for the first time the existence of the superfamily peptides to other arthropod groups. Studies utilizing approaches of in silico data mining further extended its presence to several representative ecdysozoan clades [7,17,18]. Recent additions to the superfamily include latrodectin peptides and HAND (helical arthropod-neuropeptide-derived) toxins in the venoms of spiders and centipedes [8,10].

A structural signature of the CHH superfamily is the connectivity of six invariant cysteine residues forming a unique pattern of disulfide bonds (C1–C5, C2–C4, and C3–C6). Further, members of the CHH superfamily are placed into one of two groups, Type I or Type II, according mainly to the characteristics of the peptide precursors and its mature peptides. A Type-I peptide (CHH, ITP, or their long alternative splice form (CHH-L or ITP-L, see below)) is characterized by having a precursor consisting of a signal peptide, a precursor-related peptide, and a mature peptide, while a Type-II peptide (MIH, VIH, or MOIH) by having a precursor consisting of a signal peptide and a mature peptide, without an intervening precursor-related peptide [11] With regard to the mature peptide, the short-splice form of Type-I peptides (CHH and ITP), but not CHH-L or ITP-L, are C-terminally amidated, which is critically important for the biological activity, while both short- and long-splice forms are N-terminally pyroglutaminated. The majority of Type-II peptides are free at both ends, with some exceptional VIHs and MIHs being C-terminally amidated [11]. Analysis of sequence motifs revealed that two groups of the peptides (corresponding to the Type-I and Type-II peptides) can be differentiated by group-specific motifs found at the N- and C-termini of the peptides, while the motifs in the center part of the sequence are common to all peptides [1].

CHH-superfamily peptides expressed in crustacean tissues had initially been identified and characterized each using a distinct bioassay and thus named following the biological function the assay is based on. However, it is now widely recognized that these peptides are usually pleiotropic in biological activity and in some instances are active in the same assay, i.e., functional overlap. For example, while CHH is well established for being involved in metabolism regulation [19,20,21] and in particular in stress-induced hyperglycemia, it has also been shown to be active in repressing ecdysteroidogenesis in the Y-organ (i.e., MIH activity) and methyl farnesoate synthesis in the mandibular organ (i.e., MOIH activity) [11]. It has been proposed that evolution of the CHH-superfamily peptides involved events of duplication of an ancestral gene, which may have encoded a pleiotropic peptide similar to current CHHs, leading to two main paralogous lineages (Type-I and Type-II peptides). The two lineages during the course of evolution may have partitioned the ancestral functions via a process known as subfunctionalization, with the Type-II paralogous lineages evolved to peptides with more specialized functions such as MIHs or VIHs, while Type-I lineages to peptides retaining to a certain degree the functional pleiotropy of the ancestral peptide [7].

Functional and structural polymorphism of the CHH-superfamily is further diversified by a post-transcriptional process—RNA alternative splicing. Thus, for crustacean chh gene, there are long- and short-splice forms, which share the same amino acid sequence for the first forty residues but differ significantly in the rest of the sequences that are encoded by different exons [11]. The peptide of long-splice form derived from chh gene, the CHH-L peptide, is functionally less characterized than its short form counterpart, CHH. In several species including the mud crab Scylla olivacea, where CHH and CHH-L have been functionally tested and compared, CHH-L neither elicited in vivo hyperglycemic response nor exerted in vitro inhibitory effect on the Y-organ [22,23,24]. The specific functions of CHH-L, which is usually preferentially expressed in non-eyestalk tissues, remain to be characterized.

Studies using recombinant mutants of the CHH-superfamily peptides to pinpoint functionally critical residues have been carried out for MIH of Marsupenaeus japonicus [4], CHHs of S. olivacea [9], and ITP of S. gregaria [2,5], with a general consensus that functionally important residues are located at the N-terminal and C-terminal ends of the peptides. The tertiary structure of two crustacean CHH-superfamily peptides, Pej-MIH and Pej-SGP-I-Gly (a glycine-extended precursor of CHH from M. japonicus), has been resolved [25,26], as well as that of HAND toxins in the venom of spiders and centipedes [10]. Interestingly, despite Pej-MIH and HAND toxins being otherwise structurally similar, the structure of the latter peptides lack a C-terminal helix (α5), which is present in the Pej-MIH, a structural difference thought to be related to the functional differences between the two types of peptides [10]. Moreover, unexpected dissimilarity between the structure of Pej-MIH and Pej-SGP-I-Gly, particularly at the backbone fold at the C-terminus with Pej-SGP-I-Gly lacking an α5, was conspicuously noted [26].

Pericardial organs (POs) are a pair of neurohemal organs, located in the venous cavity surrounding the heart [27], synthesizing and secreting into hemolymph amine and peptide hormones [28,29]. One of the physiological processes regulated by POs is osmoregulation [28,30]. In the present study, Sco-CHH-L, a CHH-L peptide preferentially expressed in the POs of the mud crab Scylla olivacea [31], was functionally and structurally characterized. Sco-CHH-L transcript levels were significantly elevated when animals were facing a hypo-osmotic stress. We further report that Sco-CHH-L stimulated in vitro the Na⁺/K⁺-ATPase activity in the posterior gills, establishing a specific role for the peptide in the osmoregulation of marine brachyurans. The tertiary structure of Sco-CHH-L was resolved using nuclear magnetic resonance spectroscopy. Functional test of alanine-substituted Sco-CHH-L mutants was performed using the Na⁺, K⁺-ATPase activity assay showing that several functionally critical residues at both ends of the sequence presumably constituting the receptor binding/activation sites. Comparison of Sco-CHH-L with other crustacean peptides, especially the Pej-MIH, suggests a structural architecture likely common to crustacean members of the superfamily.

We demonstrated in the present study that a CHH-like peptide (Sco-CHH-L), preferentially expressed in the pericardial organs of the mud crab (Scylla olivacea) with its transcript levels being up-regulated by hypo-osmotic stresses, stimulated in vitro the Na⁺/K⁺-ATPase activity in the posterior gills, establishing a specific role for the peptide in osmoregulation. The solution structure of Sco-CHH-L was resolved, of which the comparison with the structure of other family members defines a molecular topology that is characteristic of the crustacean members of the CHH-superfamily. Pump-activating activity of ten alanine-substituted Sco-CHH-L mutants was tested in the Na⁺, K⁺-ATPase activity assay, showing that the functionally critical residues are present at both ends of the structure, which are sterically close to each other, presumably constituting the receptor binding/activation sites.

One interesting feature of the crustacean peptides belonging to the CHH superfamily is that they are often characterized as being pleiotropic in regulatory function, and, in some instances, different peptide members are both active in the same bioassay (although with different degree of potency) [11,34]. For example, CHH is best characterized for its role in stress-induced hyperglycemia, but has also been shown to exert inhibitory effect on ecdysteroid synthesis and secretion by the Y-organ, for which MIH is considered the principle regulator [35,36,37,38]. On the other hand, CHH and CHH-L peptides are alternative splice forms of chh gene [11]. CHH-L peptides from several species have been shown possessing neither CHH (hyperglycemic) nor MIH (ecdysteroidogenesis-inhibiting) activity [22,23,24]. In this regard, it is noted that early studies suggested the presence of putative factor(s) in the neuroendocrine tissues, including the eyestalk X-organ/sinus gland complex and pericardial organs, which are involved in osmoregulation [28,30,39]. More recent studies have clarified to some extent the identity and molecular target of action of these factors. Thus, CHH, specifically a CHH isoform containing a D-Phe³ residue, when in vivo injected, was able to significantly increase the eyestalk ablation-induced decrease in hemolymph osmolality, Na⁺ concentration, or both, in the freshwater crayfish Astacus leptodactylus [40] and the American lobster Homarus americanus [41]; CHH purified from the crab Pachygrapsus marmoratus was able to increase in vitro trans-epithelial electrical potential difference and Na⁺ influx in the posterior gills [42]; CHH derived from the freshwater crayfish Cherax quadricarinatus was effective in restoring stress-induced decrease in hemolymph Na⁺ and K⁺ levels to the pre-stress levels [43]. A gut-derived CHH in the crab Carcinus maenas was demonstrated to be involved in regulating water and ion uptake during molting, allowing the swelling necessary for successful ecdysis and the subsequent increase in size during postmolt [44]. Interestingly enough, apart from the identified CHH, other factors of unknown identity in the sinus glands were also found to be effective in certain osmoregulatory parameters measured [41,42]. In addition, the molecular target(s) on which CHH acts to achieve its osmo-ionic regulation were not directly addressed by these studies, although it has been suggested that an increase in energy availability as a result of CHH-stimulated glycogen mobilization is a probable mode of action [43], a suggestion compatible with the data showing that exposing the shore crabs C. maenas to dilute seawater increased the levels of glucose in the gills and hemolymph CHH [45]. In this regard, it is relevant to note that in the Christmas Island blue crab Discoplax celeste, two CHHs were found effective in stimulating Na⁺ transport across the gill epithelia in a seasonally dependent manner, but have no effect on gill Na⁺/K⁺-ATPase or V-ATPase activity [46]. On the other hand, Pt-CHH2, a CHH-L peptide of the crab Portunus trituberculatus, was suggested to be a regulator for gill Na⁺/K⁺-ATPase and carbonic anhydrase activity, as CHH dsRNA treatment that decreased Pt-CHH2 transcript levels significantly reduced the enzyme activity in the gills [47].

In the present study, we showed that the transcript levels of Sco-CHH-L were increased by exposing animals to a hypo-osmotic stress (25ppt to 5ppt transfer), but not to a hyper-osmotic stress (25ppt to 45ppt transfer). Many marine decapods, including Scylla spp., actively regulate hemolymph osmolality in diluted seawater, but comply as an osmo-conformer in concentrated seawater [48,49,50]. Thus, the Sco-CHH-L expression pattern, in response to osmotic challenges, suggested that Sco-CHH-L is involved in osmoregulation when S. olivace faces hypo-osmotic stresses. Corroboratively, data showing that Sco- CHH-L peptide, but not Sco-CHH, stimulated in vitro the gill Na⁺/K⁺-ATPase, indicated that its osmo-regulatory functions are mediated, at least in part, via stimulating the Na⁺/K⁺-ATPase pump activity. Our data and those mentioned above collectively indicate that CHH and CHH-L peptides are both important factors for regulating water and ionic balance, but these peptides likely do so via acting on distinct molecular targets, perhaps in a concerted manner.

Structure of Sco-CHH-L was compared and contrasted with those of two other CHH-superfamily peptides, namely, Pej-MIH and Pej-SGP-I-Gly [25,26]. The six invariant cysteine residues with a unique disulfide bridge connectivity (C1-C5, C2-C4, and C3-C6, Figure 3) is a signature of the superfamily that shapes the overall fold of the peptides [11]. The common core of all three of the structures is composed of three helices, α2, α3, and α4 (π4, in Sco-CHH-L), that have similar topological relationships, with α2 and α3 running in anti-parallel direction and α4 (π4) running orthogonally (Pej-SGP-I-Gly) or obliquely (Pej-MIH and Sco-CHH-L) to α2/α3 (Figure 2). Three aromatic residues in the hydrophobic core of a venom toxin (Ta1a), a venom peptide of the CHH-superfamily, have been suggested to play a role in the packing of the structure [10]. The three aromatic residues are conserved in the Pej-MIH and Sco-CHH-L, respectively as Tyr¹⁷, Phe⁴⁵, Phe⁵⁰, and Phe¹⁶, Phe⁴⁴, and Phe⁴⁹, and only the second (Tyr⁴⁴) and third (Phe⁴⁹) aromatic residues are present in SGP-I-Gly (Figure 3). It appears that for these peptides, only the second and third aromatic residues, being buried in the hydrophobic core, are involved in the packing of the structure, while the side chain of the first conserved residue (in Pej-MIH and Sco-CHH-L) is lying on the surface of the structure (Figure 4). There are other hydrophobic residues that are conserved across the CHH-superfamily peptides and have been suggested to stabilize the structure (Figure 3) [11,25]; among those residues, only Ile¹⁵, Val³⁵, and Val⁵³ are located inside the Sco-CHH-L structure and likely contribute to the stabilization of the structure.

Sco-CHH-L shares the most significant topological similarity with Pej-MIH. Besides both having α2, α3, and α4 (or π4), they are the two structures that form a topologically equivalent α5 helix, running in an anti-parallel fashion to α4 (or π4), with the C-terminal end being directed towards the N-terminal tail (Figure 2 and Figure 3). Towards the N-terminus, Pej-MIH additionally has a short α1, while the relatively longer α2 in Sco-CHH-L encompasses the region corresponding to that of the α1 in Pej-MIH (Figure 3). Interestingly, one of the sequence features distinguishing Type I and Type II peptides is that the former group (e.g., CHH-L peptide) lacks a Gly¹² residue, which is present in the latter (e.g., MIH) [11]. Glycine is known to be a helix breaker, prone to disrupt the regularity of α helical backbone conformation [51]. Thus, it is suggested that the presence or absence of the Gly¹² residue is related to the distinct conformation of the Type I and II peptides in that Sco-CHH-L forms a continuous and long α2, and the corresponding region of Pej-MIH, having a Gly¹², breaks into two helices (α1 and α2), providing a topological signature that differentiates Type I and Type II peptides (Figure 3), which probably bears functionally important significance as experimental insertion of a Gly¹² residue into the shrimp CHH (CHH-Gly¹²) significantly decreased its hyperglycemic activity [4].

Superposing Sco-CHH-L on Pej-SGP-I-Gly, an M. japonicas CHH precursor, revealed a lower level of resemblance with a higher backbone RMSD. It is somewhat surprising as both structures, as mentioned above, are constrained and shaped by the same disulfide bridge connectivity. The two structures differ significantly in that Pej-SGP-I-Gly, lacking an α5, has a relatively long α4 with the C-terminal tail being kept away from the N-terminal end, while Sco-CHH-L forms, as mentioned above, an α5 which runs in an anti-parallel fashion against α4, with the C-terminal end being close to the N-terminus (Figure 2). It is interesting to note that HAND toxins, Ta1a and Ssm6a, characteristically lack the C-terminal α5, a structural change thought to be leading to “weaponization” of the ancestral hormonal peptides, by losing the C-terminus of the ancestral sequence that contained α5 [10]. Pej-SGP-I-Gly, however, has a rather long C-terminal tail following the α4. As CHH and CHH-L peptides are alternatively spliced forms of chh gene having the same sequence up to the first 40 residues but differing significantly after the 40th residues [31], it is possible that the topological differences between Pej-SGP-I-Gly and Sco-CHH-L could be attributed to the differences in the post-40th-residue sequence (Figure 3). It should also be noted however that Pej-SGP-I-Gly is a non-amidated, glycine-extended precursor of the Pej-SGP-I CHH [26]. CHH is post-translationally amidated at the C-terminal end, which is required for full hyperglycemic activity and the α-helical content of CHH increased after the peptide was amidated [24,52,53], indicating that C-terminal amidation of the peptide leads to structural changes. Whether the structure of Pej-SGP-I-Gly will change, particularly at the C-terminus and its topological relationship with the other areas of the structure, as a result of amidation of the peptide, is an interesting issue awaits interrogation.

Effects of alanine-substituted mutant on the gill Na⁺/K⁺-ATPase revealed several residues that are functionally important in the Sco-CHH-L (Figure 5). Side chains of these residues are located on the surface of the structure, and these residues are located at either end of the peptide but in steric proximity (Figure 6). It is suggested that the parts of the structure, consisting of these residues, play important roles in forming the binding site for receptor interaction, a suggestion that is applicable to other members of the family, as functionally critical residues at either end of the MIH, CHH, and ITP sequences have also been experimentally demonstrated [2,4,5,9]. With respect to MIH, the suggestion is additionally supported by the observation that the two ends of the Pej-MIH are, similar to those of Sco-CHH-L, sterically close to each other (Figure 2), and that the functionally critical residues of each peptide occupy topologically similar positions in the respective structure (Figure 6). It is further noted that, although the presumed binding sites of the two peptides are topologically similar, differences in the side-chain property exist. Thus, whereas both peptides have in common two hydrophobic residues, Ile²/Phe³ in Sco-CHH-L and Phe²/Ile³ in Pej-MIH, a negatively charged residue (Asp¹²) in the former peptide is replaced by an uncharged residue (Asn¹³) in the latter (Figure 6). Asp¹² in Sco-CHH and Asn¹³ in Pej-MIH have been shown to be functionally important [4,9]. Similarly, there are two negatively charged residues (Glu⁴⁵ and Asp⁶⁹) in Sco-CHH-L that are characteristically absent in Pej-MIH (Figure 6). It is relevant to mention that receptor binding assay using radiolabeled ligands and membrane preparations derived from potential target tissues suggested the existence of separate binding sites (receptors) for distinct members of the superfamily [54,55]. Differences in the property of the residues presumably constituting the binding sites could contribute to the observed receptor binding specificity. Nonetheless, cross binding and activation of receptor by a non-cognate ligand is possible to certain degree due to high similarity in the overall fold of the peptides. How the functional specificity and oftentimes the functional overlap as observed for the crustacean members of the CHH-superfamily (see [56]) are achieved through interaction of peptide ligands of a common fold with their specific receptors represents an endeavor awaiting to be pursued for better understanding of functional diversification through structural changes and endocrine regulation of crustacean physiology.

The authors thank the NMR Instrumentation Center at National Tsing Hua University, Hsinchu, and Academia Sinica High-Field NMR Center (HFNMRC) for technical support. HFNMRC is funded by Academia Sinica Core Facility and Innovative Instrument Project (AS-CFII-108-112).

Supplementary materials can be found at https://www.mdpi.com/article/10.3390/ijms222011142/s1↗, Figure S1: Preferential expression of Sco-CHH-L in the pericardial organs of the mud crab Scylla olivacea, Figure S2: Pattern of changes in Sco-CHH and Sco-CHH-L transcript levels in the pericardial organ of the mud crab Scylla olivacea in response to osmotic stresses, Figure S3: Far-UV circular dichroism spectra of the wild-type rSco-CHH-L and ala-nine-substituted mutants, Figure S4: NMR data confirming disulfide bridge connectivity, Figure S5: Assignment of amide resonances of Sco-CHH-L on a 1H-15N HSQC spectrum, Table S1: NMR restraints and structural statistics for crustacean hyperglycemic hormone-like peptide (Sco-CHH-L) from Scylla olivacea in solution, Table S2: Synthetic oligonucleotides used in this study.

Conceptualization, Y.-R.C., C.-Y.L., N.-W.H. and P.-C.L.; investigation, Y.-R.C., S.-S.H. and C.-C.C.; validation, Y.-R.C. and Y.-Z.L.; writing—original draft, Y.-R.C. and C.-Y.L.; writing—review and editing, C.-Y.L., N.-W.H. and P.-C.L.; resource, J.-R.T., H.-C.L. and J.-Y.T.; supervision, C.-Y.L. and P.-C.L.; and funding acquisition, C.-Y.L. and P.-C.L. All authors have read and agreed to the published version of the manuscript.

This research was funded by the Ministry of Science and Technology (MOST), Taiwan (109-2311-B-007-004-MY3 to P.-C.L.; 108-2311-B-018-001 and 109-2311-B-018-001 to C.-Y.L.).

Experimental protocols involving animals employed by the present study were approved by the Review Committee of National Changhua University of Education (Permit number: NCUE-097123101), in full accordance with the recommendations (Guidelines for Management and Use of Experimental Animals) set by the Council of Agriculture, Taiwan.

Not applicable.

The coordinates of the ensemble of structures have been deposited into the Protein Data Bank (PDB) as entry 5XS1.

The authors declare that they have no conflict of interest with the contents of this article.