Comparative analysis reveals molecular adaptation of mammalian HCA2 to microbial metabolites Authors: Franziska Bischof, Petra Krumbholz, Gunnar Kleinau, Patrick Scheerer, Claudia Stäubert Published: 2026 ABSTRACT SummaryMetabolite-sensing G protein-coupled receptors (GPCRs), such as hydroxycarboxylic acid receptor 2 (HCA2), translate endogenous and microbial signals into physiological responses, regulating metabolism and immunity, yet the extent of HCA2 functional diversification across mammals remain unclear. Here, comparative pharmacology, metabolomics, evolutionary analysis, and structural mapping of mammalian HCA2 orthologs reveal extensive functional diversification across mammals, especially in odd-toed ungulates. Notably, African rhinoceros HCA2 exhibits multiple HCA3-like substitutions, resulting in loss of responsiveness to HCA2 agonists and increased sensitivity to HCA3-specific ligands. Positive selection analyses and metabolomic profiling of fecal extracts implicate microbiome-derived metabolites, particularly phenylpropionic and trans-cinnamic acid, as potential drivers of this adaptive shift. Mutagenesis experiments identified key amino acid substitutions in extracellular and transmembrane regions that modulate ligand potency and efficacy. These findings demonstrate that mammalian HCA2 receptors have undergone lineage-specific molecular evolution shaped by host-microbe metabolic interactions, highlighting how ecological contexts drive receptor adaptation and functional diversification. Introduction G protein-coupled receptors (GPCRs) play a central role in cellular communication, enabling organisms to detect and respond to a wide range of environmental, metabolic, and microbial signals.1 Their ability to recognize diverse ligands and engage multiple signaling pathways forms the basis of both their physiological versatility and pharmacological importance.2,3 Subtle sequence variations can profoundly influence ligand binding and receptor activation, driving the diversification of receptor function across species to fine-tune physiological responses to their ecological niches.4 Among these, metabolite-sensing GPCRs play a crucial role in sensing a remarkable array of metabolites originating from both host cells and the surrounding microbial environment, coordinating key physiological processes at the interface of metabolism, immunity, and health.5,6 However, the molecular basis of their evolutionary adaptation remains poorly understood.7 In particular, it remains unclear how evolutionary sequence variation in metabolite-sensing GPCRs reshapes ligand recognition while preserving conserved signaling architecture across species. Hydroxycarboxylic acid receptors (HCAs) are prototypical metabolite sensors activated by endogenous and microbial metabolites. HCA2 (GPR109A) and its paralog HCA3 (GPR109B) modulate lipid metabolism, inflammatory processes, and barrier function through Gi/o protein-mediated signaling.8,9,10,11,12,13,14,15 In humans, HCA3 recognizes structurally related, lactic acid bacteria-derived aromatic acids, reflecting potential divergence from an ancestral metabolite-sensing function.16 This paralog divergence suggests that ligand-binding specificity within the HCA family is evolutionarily plastic, yet the molecular determinants of such diversification across mammals remain unresolved. All these HCA2-mediated effects make the receptor an attractive target for anti-inflammatory and immunomodulatory drugs and highlight its critical role in protecting the host against pro-inflammatory insults. Especially the skin and gastrointestinal mucosa, where HCA2 is highly expressed, act as key defense barriers preventing microbial invasion and inflammation.17,18 The close association between HCA2 signaling and host defense suggests that the receptor operates at the interface of immunity and metabolism, in systems closely associated with the resident microbiota. Symbiotic relationships between commensal microorganisms and mammalian hosts are thought to influence the evolution of metabolite-sensing receptors.19,20 HCA2 was identified in 2003 as the molecular target of nicotinic acid (NiAc), a pharmacological ligand used clinically to treat dyslipidemia.21,22,23 Although NiAc activates HCA2 at supraphysiological concentrations, endogenous activation of the receptor predominantly arises from metabolic and microbial intermediates. While the therapeutic lipid-lowering effects of NiAc occur independently of HCA2, the clinically limiting side effect of skin flushing results from pharmacological HCA2 activation in epidermal Langerhans cells and keratinocytes via prostanoid release.24,25,26,27,28 Beyond pharmacology, NiAc and its vitamin equivalent nicotinamide (Nam) are also produced by gut microbes, linking HCA2 activity to host-microbiome metabolism.6,29,30 Physiologically, HCA2 is primarily activated by endogenous metabolites such as the ketone body 3-hydroxybutyrate (3-OH-C4), produced during fasting, and the microbial short-chain fatty acid butyrate (C4), positioning the receptor as a sensor of metabolic and microbiota-derived states.31,32 In humans and other mammals, several synthetic HCA2 ligands, including monomethylfumarate (MMF), acipimox, acifran, MK-6892, GSK256073, MK-0354, and LUF6283, have been developed over the past two decades.33,34,35,36,37,38 Many of these compounds show species-dependent potency differences, with EC50 values for MK-6892 varying more than 100-fold between human and rodent HCA2.35 Such disparities highlight the importance of considering evolutionary variation when translating pharmacological data across species, particularly in the context of drug development and comparative physiology.39,40 Given its dual role in metabolic sensing and immune regulation, HCA2 continues to be investigated as a therapeutic target for inflammatory and metabolic disorders, including psoriasis, Parkinson’s disease, and multiple sclerosis.12,15,41,42 Recent advances in GPCR structural biology have produced numerous high-resolution 3D structures of HCA2 bound to diverse ligands.5,43,44,45,46,47,48,49,50,51,52,53,54,55 These studies revealed that different agonists interact with similar HCA2 amino acid residues, suggesting that ligand-receptor interactions along the entry path and residues outside the orthosteric pocket modulate ligand specificity and signaling outcomes. However, these structural studies have focused almost exclusively on human receptors and therefore do not explain how naturally occurring sequence variation across mammals translates into functional diversification of ligand recognition. Despite the wealth of available structural data for HCA2 and its role in regulating metabolic and immune cell processes, its role in mammalian evolution remains unclear, particularly in adapting to changing metabolic conditions and microbial exposure. We hypothesize that HCA2 has evolved under distinct selective pressures across mammalian lineages, potentially driving adaptive changes in ligand recognition. Such pressures can generate amino acid variation within ligand-binding sites, altering receptor pharmacology while preserving the conserved G-protein-coupling architecture. Comparative genomic analyses have revealed that mammalian GPCRs frequently evolve under selective pressures driven by diet, physiology, and symbiotic microbial communities.4 Importantly, phylogenetic relatedness does not necessarily predict shared metabolite exposure, as diet, habitat, and microbiome composition can diverge substantially among closely related species. However, how such molecular changes translate into species-specific receptor function and which ecological or metabolic factors drive them remain unresolved for most metabolite-sensing GPCRs, including HCA2. We therefore tested how evolutionary sequence variation in HCA2 relates to species-specific metabolite environments and receptor pharmacology across mammals. A comprehensive functional comparison of mammalian HCA2 orthologs, integrating biochemical signaling, evolutionary conservation, and metabolomic context, has so far been lacking. In this study, we combine pharmacological profiling, analysis of known 3D structural data, and evolutionary sequence analyses to elucidate the molecular determinants of HCA2 and HCA3 signaling diversity and investigate how cell culture conditions influence HCA2 signaling. To explore the evolutionary diversification of HCA2, we analyze over 200 mammalian orthologs and identify specific amino acid substitutions that correlate with altered ligand specificity. Structure-guided mutagenesis demonstrates that these substitutions causally reshape ligand selectivity. Functional assays across representative species reveal striking divergence, most notably in the African rhinoceros, where HCA2 has lost responsiveness to canonical HCA2 agonists but has gained sensitivity to HCA3-like ligands. Finally, using metabolomics of fecal extracts and site-directed mutagenesis, we show that HCA2 has undergone adaptive evolution, particularly at residues that shape the ligand-binding pocket, and suggest that these functional shifts are due to species-specific exposure to microbial metabolites. These findings define receptor-level evolutionary adaptation without invoking organismal fitness effects. Together, our work provides the first phylogenetically broad, structure-guided functional analysis of mammalian HCA2, linking evolutionary sequence variation to ligand recognition shaped by microbiome-derived metabolites. We establish a comprehensive framework for understanding how HCA2 sequence variation and ecological factors influence receptor evolution, demonstrating lineage-specific functional diversification through adaptive modification of ligand-binding residues while preserving conserved signaling architecture. These findings illustrate how metabolite-sensing GPCRs can be evolutionarily tuned to species-specific metabolic environments. Importantly, this functional diversification is defined at the receptor level and does not imply organismal fitness effects or direct in vivo causality. Results Substantial differences in ligand potency, receptor selectivity, and signaling bias comparing human HCA2 and HCA3 The closest paralog to human HCA2 is HCA3, both with known synthetic and natural ligands (Figure 1A). To compare their signaling profiles, we systematically examined the effects of various known ligands on cyclic adenosine monophosphate (cAMP) inhibition, extracellular signal-regulated kinases (ERKs) activation, and arrestin-3 recruitment (Figures 1B–1D).Figure 1Variable signaling induced by agonists of human HCA2 and HCA3(A) Structures and abbreviations of known HCA2 and HCA3 ligands.(B) Heatmaps showing percentage inhibition and logEC50 extrapolated from concentration-response curves in the presence of 2 μM forskolin measured in cAMP inhibition assays (mean, n = 4).(C) Phosphorylated and total ERK1/2 levels (n = 5) were determined.(D) For BRET analyses, cells were transfected with HCA2-mVenus and arrestin-3-Nluc (n = 6). BRET ratios were defined as acceptor emission/donor emission.(C and D) Data are shown as x-fold over vh (vehicle equals no agonist), mean ± SEM.(E) Data presented in (C and D) are plotted against each other to highlight the relationship of ERK signaling to arrestin-3 recruitment for the different ligands.(F) logEC50 values determined in cAMP inhibition assays (B) and BRET arrestin-3 recruitment assays (Table S1) are shown to visualize the shift in EC50 for all ligands for which HCA2-mediated arrestin-3 recruitment was determined.(G–I) Basal cAMP level determined for HCA2, HCA3, and empty vector-transfected CHO-K1 cells after 15 min incubation in HBSS with 2 μM forskolin (mean ± SEM, n = 3). (H) Cells were pre-incubated for 16 h with and without 100 ng/mL pertussis toxin (PTX) or cholera toxin (CTX). Statistical analyses were performed using unpaired two-tailed t tests. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001. (I) Cells were incubated overnight in nicotinamide-free custom-made medium without (no Nam) or with supplementation of Nam, NiAc, and/or D-PLA. Related to Table S1. Variable signaling induced by agonists of human HCA2 and HCA3 (A) Structures and abbreviations of known HCA2 and HCA3 ligands. (B) Heatmaps showing percentage inhibition and logEC50 extrapolated from concentration-response curves in the presence of 2 μM forskolin measured in cAMP inhibition assays (mean, n = 4). (C) Phosphorylated and total ERK1/2 levels (n = 5) were determined. (D) For BRET analyses, cells were transfected with HCA2-mVenus and arrestin-3-Nluc (n = 6). BRET ratios were defined as acceptor emission/donor emission. (C and D) Data are shown as x-fold over vh (vehicle equals no agonist), mean ± SEM. (E) Data presented in (C and D) are plotted against each other to highlight the relationship of ERK signaling to arrestin-3 recruitment for the different ligands. (F) logEC50 values determined in cAMP inhibition assays (B) and BRET arrestin-3 recruitment assays (Table S1) are shown to visualize the shift in EC50 for all ligands for which HCA2-mediated arrestin-3 recruitment was determined. (G–I) Basal cAMP level determined for HCA2, HCA3, and empty vector-transfected CHO-K1 cells after 15 min incubation in HBSS with 2 μM forskolin (mean ± SEM, n = 3). (H) Cells were pre-incubated for 16 h with and without 100 ng/mL pertussis toxin (PTX) or cholera toxin (CTX). Statistical analyses were performed using unpaired two-tailed t tests. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001. (I) Cells were incubated overnight in nicotinamide-free custom-made medium without (no Nam) or with supplementation of Nam, NiAc, and/or D-PLA. Related to Table S1. MK-6892, so far the most potent and efficient HCA2-specific agonist, induced cAMP inhibition, activated ERK, and caused recruitment of arrestin-3 (Figures 1B–1D). NiAc, which was about 10,000-fold less potent at HCA3, showed HCA2-mediated activation of all three analyzed signaling components. GSK256073 activated HCA3 with 5-fold lower potency and half the efficacy compared to HCA2, in contrast to a previously reported 100-fold lower potency.37 GSK256073 induced ERK activation via both HCA2 and HCA3 but arrestin-3 recruitment only via HCA2, a pattern also observed for acifran (Figures 1C and 1D). MMF and acipimox did not induce HCA3-dependent ERK signaling (Figures 1B–1D). MK-0354 and LUF6283 are both human HCA2-specific partial agonists; both recruited arrestin-3 (with a weaker effect for MK-0354) without ERK activation (Figures 1B–1D). In contrast to butyric acid (C4), the endogenous low-potent HCA2-specific ligand 3-OH-C4 slightly induced arrestin-3 recruitment (Figures 1B–1D). Human HCA3-specific ligands (3-hydroxyoctanoic acid (3-OH-C8), D-phenyllactic acid (D-PLA), D-phenylalanine (D-Phe), IPBT-5CA) showed no activity at human HCA2 (Figures 1A–1D). Both potent non-HCA2-specific ligands, GSK256073 and acifran, were inefficient in human HCA2-dependent ERK activation (Figures 1C and 1E). Concentration-response curves revealed that arrestin-3 recruitment potencies at HCA2 were about 50- to 70-fold lower than for cAMP inhibition (Figure 1F; Table S1). In summary, we comprehensively and functionally characterized the ligands that activate human HCA2 and HCA3. Standard cell culture medium modulates basal HCA2 signaling via Gi-dependent sensitization In forskolin-treated cells, HCA2-transfected cells showed increased basal cAMP levels compared to the control, while human HCA3 exhibited classical Gi protein-mediated reduction in cAMP levels (Figure 1G). Both effects were sensitive to the Gi protein inhibitor pertussis toxin (PTX), and not the Gs protein inhibitor cholera toxin (CTX) (Figure 1H). Dulbecco’s modified Eagle medium: nutrient mixture F12 (DMEM/F-12) medium contains 17 μM Nam, which itself has been reported not to bind to HCA2 or to act only as a very weak HCA2 agonist.21,22,23 Our data indicate that Nam, or trace NiAc present in the medium, accounts for the increased basal cAMP observed in HCA2-expressing cells via Gαi/o-coupled receptor-mediated heterologous sensitization of adenylyl cyclase.56 Using Nam-free medium abolished this effect, whereas adding 4 μM Nam restored it in HCA2 but not HCA3-transfected cells. Conversely, the addition of D-PLA selectively increased basal cAMP in HCA3-transfected cells (Figure 1I). These findings confirm that standard culture media can contain low levels of HCA2 ligands that can modulate basal signaling. This is particularly relevant for comparative pharmacology, as baseline receptor activation by media components may differ across orthologs and influence apparent species differences in ligand potency. Because similar media are commonly used for receptor expression in structural studies, residual ligand occupancy should also be considered when interpreting HCA2 structures. Positional conservation analyses uncover ligand-binding adaptations in mammalian HCA2 orthologs with distinct HCA3-like substitutions in African rhinoceros Human HCA2 and HCA3 differ by only 16 amino acids and an extended C terminus, yet display distinct ligand signaling profiles, indicating that limited sequence variation can reshape ligand recognition.16 We therefore performed a positional conservation analysis across more than 200 mammalian HCA2 orthologs to identify evolutionary substitutions within the ligand-binding region (Figure 2A; Tables S2 and S3).Figure 2Conservation of mammalian HCA2 orthologs and comparison to human HCA3(A) Phylogenetic tree with mammalian orders and the number of species (total, 220) included in conservation analyses visualized in the snake plot, and (B–D) mapped to a 3D structure using the software package Chimera.57 The color of the HCA2 backbone (PDB ID, 8j6q44) reflects the degree of conservation (red, highly conserved, blue, less/not conserved).(B) A brown circle indicates the principal ligand binding region, which shows high conservation (C), particularly for residues with side chains directed to the transmembrane domain core. Extracellular loop 3 (ECL3) and the N-terminal part of ECL2 are not conserved.(D) The G-protein interaction site (intracellular) is highly conserved, especially in intracellular loop 1 (ICL1) and ICL2. aa, amino acid; TMH, transmembrane helix; Ntt, N-terminal tail; Ctt, C-terminal tail. Related to Tables S2 and S3–S5. Conservation of mammalian HCA2 orthologs and comparison to human HCA3 (A) Phylogenetic tree with mammalian orders and the number of species (total, 220) included in conservation analyses visualized in the snake plot, and (B–D) mapped to a 3D structure using the software package Chimera.57 The color of the HCA2 backbone (PDB ID, 8j6q44) reflects the degree of conservation (red, highly conserved, blue, less/not conserved). (B) A brown circle indicates the principal ligand binding region, which shows high conservation (C), particularly for residues with side chains directed to the transmembrane domain core. Extracellular loop 3 (ECL3) and the N-terminal part of ECL2 are not conserved. (D) The G-protein interaction site (intracellular) is highly conserved, especially in intracellular loop 1 (ICL1) and ICL2. aa, amino acid; TMH, transmembrane helix; Ntt, N-terminal tail; Ctt, C-terminal tail. Related to Tables S2 and S3–S5. Notably, African rhinoceroses, white (Ceratotherium simum), and black rhinoceros (Diceros bicornis), exhibit substitutions at eight HCA3-specific sites, including key ligand-binding residues (Leu832.60, Met1033.28, and Leu1073.32) and multiple positions in the extracellular loop 2 (ECL2; Met167, Pro168, Gly173, and Leu176) (Figure 2A; Table S3). Thus, HCA2 may have evolved under changing constraints to recognize a diverse range of ligands. Mixed effects model of evolution (MEME) analysis revealed several sites under episodic diversifying selection, including Gly173 (Figure 2A). Additionally, Phe251.30, Arg90 (ECL1), and Lys164 (ECL2), located near divergent HCA2/HCA3 positions, also showed signatures of positive selection (Figure 2A). These patterns indicate lineage-specific diversification concentrated in ligand-binding regions. To place these variations in structural context, residue conservation across mammalian orthologs was mapped onto available HCA2/HCA3 structures (Figures 2B–2D; Table S4).5,43,44,45,46,47,48,49,50,51,52,53,54,55 The principal orthosteric pocket is highly conserved, particularly residues projecting into the transmembrane core (Figure 2B). In contrast, ECL3 and the N-terminal portion of ECL2 show substantial variability, whereas residues implicated in G-protein interaction (ICL1 and ICL2) remain strongly conserved, indicating preservation of signaling architecture (Figures 2C and 2D; Table S5). Based on this phylogenetic and structural analysis, we selected representative HCA2 orthologs spanning major mammalian clades and key ligand-binding substitutions for functional characterization, including marsupials, afrotherians, rodents, carnivores, odd- and even-toed ungulates, cetaceans, and primates, as well as human HCA2 and HCA3 (Figure S1A). This panel captures major evolutionary transitions in HCA2 ligand-binding sequence variation while maintaining conserved signaling interfaces. 3-OH-C4 and C4 more potently activate minke whale and opossum HCA2, while rhinoceros HCA2 is unresponsive to HCA2 agonists but activated by HCA3 ligands Basal cAMP inhibition largely reflected HCA2 cell surface expression (Figure S1B). All orthologs were analyzed in the same cellular background, providing a common endogenous G-protein complement. Relative plasma membrane expression was quantified by cell-surface enzyme-linked immunosorbent assay (ELISA) under identical conditions for all orthologs analyzed, ensuring that pharmacological differences reflect receptor-intrinsic properties rather than variability in signaling context or expression levels. HCA2 from white rhinoceros, opossum, and cat showed lower surface expression and reduced basal cAMP levels, indicating higher basal Gi activation in these orthologs (Figure S1B). All HCA2 orthologs, except rhinoceros HCA2, responded to 3-OH-C4 and C4 with notable differences in potency and efficacy (Figures 3A and 3B; Table S6). Notably, 3-OH-C4 and C4 were ∼10-fold more potent at HCA2 of minke whale and opossum, and these were the only orthologs showing ERK activation at a concentration of 1 mM (Figures 3A–3C; Table S6). NiAc induced cAMP inhibition with similar potency and efficacy, as well as ERK activation in all HCA2 orthologs except rhinoceros HCA2 (Figures 3A–3C; Table S6). Testing HCA3 agonists revealed that rhinoceros HCA2 gained responsiveness to D-PLA and IPBT-5CA, while opossum HCA2 also responded to IPBT-5CA (Figures 3D and 3E; Table S6). These findings demonstrate species-specific ligand recognition, with rhinoceros HCA2 showing extreme functional divergence driven by positional variation (Figure S1A; Table S6).Figure 33-OH-C4 and C4 more potently activate minke whale and opossum HCA2, while rhinoceros HCA2 is unresponsive to HCA2 agonists but activated by HCA3 ligandsCells were transiently transfected with HCA2 orthologs or human HCA3.(A) Heatmaps showing logEC50 and percentage inhibition were extrapolated from concentration-response curves of 3-hydroxybutyric acid (3-OH-C4), butyric acid (C4), and nicotinic acid (NiAc) in cAMP inhibition assays performed in the presence of 2 μM forskolin (mean, n = 4).(B) Exemplary concentration-response curves. Data are shown as percentage of cAMP in the absence of agonist for each ortholog (w/o) as mean ± SEM.(C) Phosphorylated and total ERK1/2 levels (data shown as min to max, line at mean, n = 4).(D) Heatmap showing percentage inhibition determined in cAMP inhibition assays in the presence of 2 μM forskolin upon stimulation with indicated concentrations of 3-hydroxyoctanoic acid (3-OH-C8), D-phenyllactic acid (D-PLA), D-phenylalanine (D-Phe), or IPBT-5CA. Concentration-response curves upon stimulation with D-PLA. Data are shown as percentage of cAMP in the absence of agonist (w/o) for each ortholog as mean ± SEM, n = 4.(E) Heatmap showing phosphorylated over total ERK1/2 levels upon stimulation with indicated concentrations of 3-OH-C8, D-PLA, D-Phe, or IPBT-5CA. Comparison of pERK/total ERK for human HCA3 and rhinoceros HCA2 (data are shown as min to max, line at mean, n = 4). (C and E) Statistical analyses were performed using paired two-tailed t tests against w/o. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001. Related to Figure S1, Table S6. 3-OH-C4 and C4 more potently activate minke whale and opossum HCA2, while rhinoceros HCA2 is unresponsive to HCA2 agonists but activated by HCA3 ligands Cells were transiently transfected with HCA2 orthologs or human HCA3. (A) Heatmaps showing logEC50 and percentage inhibition were extrapolated from concentration-response curves of 3-hydroxybutyric acid (3-OH-C4), butyric acid (C4), and nicotinic acid (NiAc) in cAMP inhibition assays performed in the presence of 2 μM forskolin (mean, n = 4). (B) Exemplary concentration-response curves. Data are shown as percentage of cAMP in the absence of agonist for each ortholog (w/o) as mean ± SEM. (C) Phosphorylated and total ERK1/2 levels (data shown as min to max, line at mean, n = 4). (D) Heatmap showing percentage inhibition determined in cAMP inhibition assays in the presence of 2 μM forskolin upon stimulation with indicated concentrations of 3-hydroxyoctanoic acid (3-OH-C8), D-phenyllactic acid (D-PLA), D-phenylalanine (D-Phe), or IPBT-5CA. Concentration-response curves upon stimulation with D-PLA. Data are shown as percentage of cAMP in the absence of agonist (w/o) for each ortholog as mean ± SEM, n = 4. (E) Heatmap showing phosphorylated over total ERK1/2 levels upon stimulation with indicated concentrations of 3-OH-C8, D-PLA, D-Phe, or IPBT-5CA. Comparison of pERK/total ERK for human HCA3 and rhinoceros HCA2 (data are shown as min to max, line at mean, n = 4). (C and E) Statistical analyses were performed using paired two-tailed t tests against w/o. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001. Related to Figure S1, Table S6. African rhinoceros HCA2 shows strong positive selection, potentially driven by microbiome-derived metabolites, as HCA agonists are present in fecal samples Evolutionary analysis of odd-toed ungulate HCA2 orthologs revealed significant selection intensification in African rhinoceroses (K = 8.73, p ≤ 0.0001, LRT = 14.90), with 13 sites under positive selection (p ≤ 0.1), including Leu1073.32 (Figures 4A and 4B). We propose that microbiome-derived metabolites such as C4 and NiAc may be associated with these adaptations in rhinoceros HCA2. Therefore, fecal samples from rhinoceros, tapir, and zebra were collected, extracted with aqueous buffer to preferentially capture abundant soluble metabolites most likely to access luminal and systemic HCA2, and tested on human HCA2, HCA3, and rhinoceros HCA2 in cAMP assays (Figure 4C). PTX was used to determine the Gi protein dependence of the signal. Human HCA2 responded most weakly to rhinoceros extracts and most strongly to tapir extracts, consistent with the detected NiAc levels (Figures 4C and 4D). Only rhinoceros and zebra fecal extracts activated human HCA3 and rhinoceros HCA2, correlating with high levels of PLA and other metabolites (Figures 4C and 4D).Figure 4African rhinoceros HCA2 shows strong positive selection, potentially driven by microbiome-derived metabolites(A) Evolutionary history of odd-toed ungulate and human HCA2 was inferred using Maximum Likelihood (Kimura 2-parameter, MEGA 11). The tree (log likelihood = −2,483.78) shows branch lengths as substitutions per site, with bootstrap support indicated. RELAX revealed significant selection intensification in African rhinoceroses (K = 8.73, p ≤ 0.0001, LR = 14.90, highlighted in red).(B) Among the odd-toed ungulates HCA2 orthologs and human HCA2, shown in (A) FEL identified 13 sites (p ≤ 0.1) under diversifying positive selection in African rhinoceros HCA2 orthologs. Fecal samples from black rhinoceros (n = 4), Asiatic tapir (n = 2), and zebra (n = 3) were extracted, diluted, and (C) tested for cAMP inhibition in cells expressing human or rhinoceros HCA2, human HCA3, or empty vector, with or without pertussis toxin (100 ng/mL, 16 h). Data are shown as percentage of untreated controls (mean ± SEM, n ≥ 4). Statistical analyses were performed using unpaired two-tailed t tests, each against the respective empty vector control. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001.(D) HILIC-based metabolomics of all extracts identified relative levels of HCA2/HCA3 agonists and other metabolites; selected metabolites differing in concentrations are shown. 2-OH-MTBA, 2-hydroxy-4-(methylthio)butyrate; 2-OH-PAA, 2-hydroxy-phenylacetic acid; 4-OH-PPA, 4-hydroxy-phenylpropionic acid. Related to Figures S2 and S3. African rhinoceros HCA2 shows strong positive selection, potentially driven by microbiome-derived metabolites (A) Evolutionary history of odd-toed ungulate and human HCA2 was inferred using Maximum Likelihood (Kimura 2-parameter, MEGA 11). The tree (log likelihood = −2,483.78) shows branch lengths as substitutions per site, with bootstrap support indicated. RELAX revealed significant selection intensification in African rhinoceroses (K = 8.73, p ≤ 0.0001, LR = 14.90, highlighted in red). (B) Among the odd-toed ungulates HCA2 orthologs and human HCA2, shown in (A) FEL identified 13 sites (p ≤ 0.1) under diversifying positive selection in African rhinoceros HCA2 orthologs. Fecal samples from black rhinoceros (n = 4), Asiatic tapir (n = 2), and zebra (n = 3) were extracted, diluted, and (C) tested for cAMP inhibition in cells expressing human or rhinoceros HCA2, human HCA3, or empty vector, with or without pertussis toxin (100 ng/mL, 16 h). Data are shown as percentage of untreated controls (mean ± SEM, n ≥ 4). Statistical analyses were performed using unpaired two-tailed t tests, each against the respective empty vector control. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001. (D) HILIC-based metabolomics of all extracts identified relative levels of HCA2/HCA3 agonists and other metabolites; selected metabolites differing in concentrations are shown. 2-OH-MTBA, 2-hydroxy-4-(methylthio)butyrate; 2-OH-PAA, 2-hydroxy-phenylacetic acid; 4-OH-PPA, 4-hydroxy-phenylpropionic acid. Related to Figures S2 and S3. Hydrophilic interaction liquid chromatography (HILIC)-based metabolomics analyses identified 83 metabolites, with rhinoceros samples showing low 3-OH-C4 levels but elevated concentrations of PLA and other metabolites that may contribute to the observed differences in activation patterns (Figures 4D and S2). Therefore, we tested all these compounds and other microbiota-derived metabolites for their agonistic activity at human HCA2 and HCA3, which reflect the ligand specificity differences seen in rhinoceros HCA2 (Figure S3). Most compounds were inactive but trigonelline selectively activated HCA2, while 2-hydroxy-4-(methylthio)butyrate (2-OH-MTBA), indolepropionic acid, and trans-indoleacrylic acid specifically activated HCA3 (Figure S3). Several aromatic acids activated both receptors, with phenylpropionic acid (PPA) and trans-cinnamic acid (tCA) inducing the highest reduction in cAMP (Figure S3). These data indicate that such aromatic acids can act as ligands across mammalian HCA2 orthologs, including rhinoceros HCA2. Indeed, PPA and tCA activated rhinoceros and all other mammalian HCA2 orthologs, with tCA generally being more potent (Figure 5; Table S6). However, the potency and efficacy of both ligands also varied across species (Figure 5; Table S6).Figure 5Phenylpropionic acid and trans-cinnamic acid are evolutionarily conserved HCA2 agonistsCells were transiently transfected with mammalian HCA2 orthologs, human HCA3, or an empty vector control, and cAMP inhibition assays in the presence of 2 μM forskolin were performed upon stimulation with PPA and tCA. Data are shown as mean ± SEM (n = 3). Heatmaps are extrapolated from concentration-response curves and show percentage inhibition and logEC50 values. Related to Figures S3 and S4, Table S6. Phenylpropionic acid and trans-cinnamic acid are evolutionarily conserved HCA2 agonists Cells were transiently transfected with mammalian HCA2 orthologs, human HCA3, or an empty vector control, and cAMP inhibition assays in the presence of 2 μM forskolin were performed upon stimulation with PPA and tCA. Data are shown as mean ± SEM (n = 3). Heatmaps are extrapolated from concentration-response curves and show percentage inhibition and logEC50 values. Related to Figures S3 and S4, Table S6. Together, these findings suggest that microbiome-derived metabolites, particularly aromatic acids such as PPA and tCA, may be linked to the evolutionary diversification of HCA2 in odd-toed ungulates, with rhinoceros HCA2 representing a striking example of adaptation to species-specific metabolic environments. To better understand the physiological roles of HCA2 across mammals, we next examined its tissue-specific RNA expression patterns, which are available and accessible for several species in the Bgee database, allowing for the comparison of mammalian transcriptomic data for some species (Figure S4).58,59 However, the white rhinoceros is a critically endangered species for which no RNA expression data are available. mRNA expression of HCA2 in different species We investigated the distribution of HCA2 expression in immune, digestive, and other organ systems to assess conserved and species-specific functions. HCA2 is expressed in immune-associated tissues, including lymph nodes, spleen, bone marrow, tonsils, and leukocytes (Figure S4). Furthermore, high expression is observed in the tongue of cattle, in the salivary glands of humans, sheep, and dogs, and in the esophagus of several species (Figure S4). Notably, while HCA2 expression is low in the stomach, small, and large intestine of non-ruminants, it is elevated in the rumen of cattle, sheep, and goats, consistent with the role of microbial fermentation in plant digestion of ruminant animals (Figure S4).60 Other tissues for which relatively high HCA2 expression is found in several species of different orders include the olfactory bulb, the lung, skin, adipose tissue, and female reproductive tissues (Figure S4). These expression patterns highlight a conserved role of HCA2 in immune-associated tissues across mammals. At the same time, the elevated levels in the rumen of ruminants point to a potential specialization in host-microbe interactions during plant digestion. Moreover, its presence in diverse epithelial and reproductive tissues suggests broader functions beyond immunity, possibly linking metabolic and barrier-related processes. We further analyzed ligand preferences across mammalian HCA2 orthologs using available synthetic ligands. Divergent agonist responses of mammalian HCA2 orthologs reflect sequence variability in the ligand-binding pockets Systematic profiling of mammalian HCA2 orthologs revealed striking species-specific differences in responses to synthetic agonists (Figure 6A; Table S6). MK-6892 activated rhinoceros HCA2 with ∼500-fold lower potency than human, and was also less potent at elephant HCA2, which is also reflected in a lack of MK-6892-induced ERK activation (Figure 6B). GSK256073 displayed higher potency at opossum HCA2 but was markedly weaker at polar bear, cat, minke whale, cattle, and rhinoceros HCA2 (Figure 6A; Table S6). Consequently, no GSK256073-mediated ERK activation was detected for these orthologs (Figure 6B). Acifran, although least potent at polar bear HCA2, strongly activated rhinoceros HCA2, while MMF and acipimox showed reduced potency at polar bear, cat, minke whale, horse and elephant orthologs, and did not activate rhinoceros HCA2 (Figure 6A; Table S6). Similarly, acifran stimulation resulted in ERK activation of all HCA2 orthologs, while MMF and acipimox failed to activate rhinoceros HCA2 (Figure 6B). MK-0354 activated rhinoceros HCA2 with high potency but low efficacy, whereas LUF6283 displayed high potency at opossum HCA2 but low potency at polar bear HCA2 (Figure 6A). MK-0354 and LUF6283 failed to induce ERK activation in any ortholog (Figure 6B).Figure 6Synthetic HCA2 ligands exhibit variable ligand profiles at mammalian HCA2 orthologsCells were transiently transfected with HCA2 orthologs or human HCA3.(A) Ligand-induced cAMP inhibition was determined in the presence of 2 μM forskolin. Exemplary concentration-response curves (data shown as percentage of cAMP in the absence of agonist for each ortholog (w/o) (mean ± SEM, n = 4), as well as heatmaps of logEC50, and % inhibition extrapolated from these curves, are shown.(B) Phosphorylated and total ERK1/2 levels (data shown as min to max, line at mean, n = 4) upon stimulation with MK-6892, GSK256073, acifran, monomethylfumarate (MMF), acipimox, MK-0354, and LUF6283. Statistical analyses were performed using paired two-tailed t tests against w/o. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001. Related to Table S6. Synthetic HCA2 ligands exhibit variable ligand profiles at mammalian HCA2 orthologs Cells were transiently transfected with HCA2 orthologs or human HCA3. (A) Ligand-induced cAMP inhibition was determined in the presence of 2 μM forskolin. Exemplary concentration-response curves (data shown as percentage of cAMP in the absence of agonist for each ortholog (w/o) (mean ± SEM, n = 4), as well as heatmaps of logEC50, and % inhibition extrapolated from these curves, are shown. (B) Phosphorylated and total ERK1/2 levels (data shown as min to max, line at mean, n = 4) upon stimulation with MK-6892, GSK256073, acifran, monomethylfumarate (MMF), acipimox, MK-0354, and LUF6283. Statistical analyses were performed using paired two-tailed t tests against w/o. ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001. Related to Table S6. Sequence conservation analysis across 220 mammalian orthologs identified hot spots in the ligand-binding region that may contribute to the observed differences in pharmacological properties (Figure S5; Table S7). Specifically, a comparison of cattle with human HCA2 revealed several differing positions located in the ligand-binding region (Figure 7A; Table S7). Thus, we generated the exemplary human HCA2 mutants F277Y7.40 and L34V1.39, as these substitutions occur only in minke whale, cattle, and opossum HCA2, of all orthologs functionally tested (Figure S1A). As for minke whale, cattle, and opossum HCA2, 3-OH-C4, and LUF6283 activated the mutant F277Y7.40 more potently (Figures 3A, 6A, and 7B). Similarly, GSK256073 less potently activated L34V1.39, as observed for minke whale and cattle HCA2, both of which have this amino acid exchange (Figures 6A and 7B; Figure S1A). However, neither mutant differed in potency for NiAc (Figure 7B).Figure 7Conservation analyses at the active state 3D structure, combined with mutational analyses, identified key residues in the ligand-binding region that explain some of the species-specific functional differences(A) The human HCA2 sequence was compared with cattle HCA2, chosen due to its variation in responses to several ligands (Figure 6A). In addition, conservation was mapped onto the active-state structure (PDB-ID 8j6q), with red indicating high and blue low conservation (top). For analyses and representation of conservation degree, the software package Chimera57 was used. Variants chosen for experimental substitution and functional tests are highlighted with gray labels. Non-conserved residues between human and cattle HCA2 were visualized with exemplarily bound MMF from superimposed complexes. While many substitutions (e.g., M84, H184, and T185) are distant from the ligand, others, such as F277, are in direct proximity.(B and D) Cells were transiently transfected with human HCA2 wild-type (wt) or mutants, and ligand-induced cAMP inhibition was determined in the presence of 2 μM forskolin as indicated. Data are shown as percentage of cAMP in the absence of agonist for HCA2 mutant (w/o) as mean ± SEM (n = 3).(C) Active (PDB: 8j6q) versus inactive (PDB: 7zL9) HCA2 structures show marked differences in the C-terminal ECL2 and the adjoining TMH5 (helix-kink at P2005·50), while most of the extracellular parts overlap. Side-chain orientations differ, highlighting how substitutions or ligand binding can reshape this flexible region and impact orthosteric, allosteric, or bitopic ligand interactions.(A and C) Structural representations were generated using PyMol (Molecular Graphics System v.2.5.5, Schrödinger, LLC, New York, NY). ECL, extracellular loop; ICL, intracellular loop; TMH, transmembrane helix; Ntt, N-terminal tail; 3-OH-C4, 3-hydroxybutyric acid, NiAc, nicotinic acid. Related to Figure S5, Table S7. Conservation analyses at the active state 3D structure, combined with mutational analyses, identified key residues in the ligand-binding region that explain some of the species-specific functional differences (A) The human HCA2 sequence was compared with cattle HCA2, chosen due to its variation in responses to several ligands (Figure 6A). In addition, conservation was mapped onto the active-state structure (PDB-ID 8j6q), with red indicating high and blue low conservation (top). For analyses and representation of conservation degree, the software package Chimera57 was used. Variants chosen for experimental substitution and functional tests are highlighted with gray labels. Non-conserved residues between human and cattle HCA2 were visualized with exemplarily bound MMF from superimposed complexes. While many substitutions (e.g., M84, H184, and T185) are distant from the ligand, others, such as F277, are in direct proximity. (B and D) Cells were transiently transfected with human HCA2 wild-type (wt) or mutants, and ligand-induced cAMP inhibition was determined in the presence of 2 μM forskolin as indicated. Data are shown as percentage of cAMP in the absence of agonist for HCA2 mutant (w/o) as mean ± SEM (n = 3). (C) Active (PDB: 8j6q) versus inactive (PDB: 7zL9) HCA2 structures show marked differences in the C-terminal ECL2 and the adjoining TMH5 (helix-kink at P2005·50), while most of the extracellular parts overlap. Side-chain orientations differ, highlighting how substitutions or ligand binding can reshape this flexible region and impact orthosteric, allosteric, or bitopic ligand interactions. (A and C) Structural representations were generated using PyMol (Molecular Graphics System v.2.5.5, Schrödinger, LLC, New York, NY). ECL, extracellular loop; ICL, intracellular loop; TMH, transmembrane helix; Ntt, N-terminal tail; 3-OH-C4, 3-hydroxybutyric acid, NiAc, nicotinic acid. Related to Figure S5, Table S7. Further, position T18545.48 in human HCA2, critical for allosteric ligand binding in ECL2 (Figure 7C; Table S7), is substituted by alanine in cattle and serine in elephant HCA2 (Figure S1A). The corresponding mutants T185A45.58 and T185S45.48 showed reduced MK-6892 potency, while GSK256073 and NiAc remained unaffected (Figure 7D). Acipimox potency was reduced in elephant but not cattle HCA2, consistent with our findings for T185S45.48 but not T185A45.48, which shows that position 185 affects the orthosteric binding site by modulating ECL2 conformation and dynamics (Figures 6A and 7D). In summary, mutagenesis confirmed that substitutions at several positions fine-tune ligand potency and efficacy. These findings highlight that subtle sequence variations in the ligand-binding pocket are responsible for the striking functional diversity of HCA2 across mammals, shaping species-specific drug responses and revealing evolutionary adaptation at the receptor-ligand interface. Thus, the context-dependent interplay between HCA2 sequence variation and ligand activity offers a structural basis for the divergent pharmacological profiles observed across species. Discussion Our study provides a functional, pharmacological, and evolutionary characterization of HCA2 across mammals, revealing species-specific differences in ligand potency, selectivity, and signaling that arise from sequence variation within ligand-binding sites. Integration of phylogenetic breadth with structure-guided functional analysis demonstrates that evolutionary divergence directly reshapes ligand recognition at HCA2, establishing a mechanistic link between microbiome-derived metabolites and receptor pharmacology across mammals. Differences in HCA2 and HCA3 ligand potency and selectivity Consistent with previous structural and pharmacological studies, human HCA2 and HCA3 differ markedly in ligand preferences and downstream signaling bias (e.g., G-protein versus arrestin-3 or ERK pathways). Arrestin-3 recruitment was therefore assessed only for the human HCA2 and HCA3 reference receptors to define ligand signaling bias. NiAc is endogenously present in the plasma of healthy individuals at concentrations of ∼80 nM, primarily derived from dietary sources or microbial metabolism.61,62,63,64,65,66 The anti-dyslipidemic actions of NiAc at pharmacological concentrations of up to 200 μM are HCA2-independent, whereas the clinically limiting flushing response is mediated by HCA2 activation in skin cells.27 Synthetic HCA2 agonists remain of therapeutic interest for treating inflammatory conditions, and our analysis refines their signaling profiles. As previously shown, MK-0354 and LUF6283 acted as HCA2-selective partial agonists, primarily activating Gi protein signaling.67 However, in contrast to previous reports, neither ligand induced detectable ERK activation in our assays.36,67 Notably, MK-0354 promoted arrestin-3 recruitment despite prior reports suggesting a lack of receptor internalization, indicating that HCA2 can engage arrestin-dependent signaling at the plasma membrane without obligatory internalization (Figure 1). MK-6892 and GSK256073 have been proposed to favor Gi over arrestin pathways52 but our data did not support pronounced bias. Instead, MK-6892 showed the highest potency at HCA2 in both the Gi protein and the arrestin pathway (Figure 1).50 GSK256073 displayed lower HCA2 selectivity than previously described and, together with acifran, produced robust arrestin recruitment but comparatively weak HCA2-dependent ERK activation (Figure 1). Overall, these results refine ligand-specific signaling assignments and align with structural determinants of HCA2/HCA3 ligand recognition. Impact of culture conditions on HCA2 signaling Our findings show that standard insect or eukaryotic cell culture media containing Nam or NiAc induce HCA2-mediated and Gi-dependent heterologous sensitization of adenylyl cyclase (Figure 1).56 This effect can alter estimates of basal receptor activity or ligand potency, indicating that media-derived ligands can influence basal signaling and apparent pharmacology, particularly in cross-species comparisons. Media-derived NiAc may also occupy the receptor during protein expression, potentially influencing ligand occupancy in structural studies. NiAc binds HCA2 with nM affinity, and structural differences have primarily been observed only for higher affinity ligands such as MK-6892 and GSK256073. These considerations underscore the importance of controlling culture conditions when comparing HCA2 pharmacology and structure–function relationships across species. Species variation, adaptation, and ligand specificity Across mammals, ligand-binding residues varied substantially while G-protein-interacting residues remained conserved. Consistently, HCA2 orthologs from cattle, minke whale, and opossum displayed increased potency for the endogenous ketone body 3-OH-C4 and microbial C4, indicating physiological activation at lower ligand concentrations (Figure 3). These findings demonstrate species-specific differences in HCA2 physiology at the interface of host metabolism and immunity that arise from evolutionary diversification of ligand recognition rather than from alterations in downstream signaling architecture. Unique evolutionary adaptation of HCA2 in the white rhinoceros Rhinoceros HCA2 is unresponsive to canonical HCA2 agonists and instead recognizes HCA3-like ligands, representing the most divergent ortholog identified (Figures 3, 4, and 6). Because HCA3 is restricted to great apes, the HCA3-like pharmacological features observed in rhinoceros HCA2 likely reflect convergent evolution toward a similar ligand-recognition profile rather than coordinated diversification between paralogs, highlighting independent evolutionary solutions to similar metabolite environments. This shift correlates with HCA3-like substitutions under positive selection and with microbiome-derived metabolites detected in rhinoceros fecal extracts, supporting receptor-level adaptation to species-specific metabolite environments without invoking organismal fitness effects. Two evolutionary scenarios could explain this divergence. First, rhinoceros may no longer require HCA2 to sense fasting-associated metabolites such as 3-OH-C4 levels, potentially because this function has shifted to another GPCR, allowing HCA2 to specialize toward microbiome-derived ligands such as PLA. Alternatively, a conserved, yet unidentified endogenous ligand may activate HCA2 across mammals, while residues not required for its recognition diversified under species-specific dietary and microbial pressures. Consistent with this view, fecal metabolite profiles differ markedly among odd-toed ungulates (rhinoceros, tapir, and zebra) and correlate with distinct activation patterns of rhinoceros HCA2, human HCA2, and HCA3 (Figures 4, S2, and S3). These features are consistent with the rhinoceros hindgut-fermenting herbivorous physiology, which shapes a distinct microbial metabolite environment rich in plant- and fermentation-derived aromatic compounds, which represent plausible selective drivers of the HCA3-like ligand recognition observed in rhinoceros HCA2.68,69,70 Similar diet- and microbiome-associated diversification of ligand-binding residues has been reported for other metabolite- and chemosensory-GPCR families, supporting evolutionary tuning of receptor specificity to ecological niche.71 Microbial aromatic amino acid metabolites as conserved HCA2 agonists Phenylpropanoid metabolites such as PPA and tCA activated all HCA2 orthologs, including rhinoceros HCA2 and human HCA3, indicating that aromatic microbial metabolites represent conserved ligands across mammalian HCA2 orthologs (Figure 5). Both ligands differed in potency and efficacy, with PPA generally less potent and efficacious than tCA across species. PPA is not produced by mammalian metabolism but by Firmicutes and Bacteroidetes.72 In humans, circulating levels are in the micromolar range, and fecal concentration reaches ∼515 μM.73,74,75 Although previously reported as HCA3-selective, we show that PPA also activates HCA2 (EC50 ∼ 250 μM), albeit ∼25-fold less potently than at HCA3 (Figure 5), suggesting that the reported PPA-mediated protection against acetaminophen hepatotoxicity may involve both receptors.6,72 Furthermore, tCA, a plant- and microbiome-derived compound with documented antioxidant and anti-inflammatory properties, is likewise abundant in intestinal environments and activated both HCA2 and HCA3 across species (Figure 5).74,76,77,78,79,80,81,82,83,84 Together, these data identify phenylpropanoid microbial metabolites as evolutionarily conserved HCA2 ligands linking microbiome metabolism to receptor diversification. Molecular basis of functional divergence of mammalian HCA2 orthologs To link the observed species-specific differences in ligand responses to underlying sequence variation, we next examined substitutions within the ligand-binding pocket and tested their functional impact. Using structure-guided mutagenesis, we identified ligand-binding pocket substitutions that substantially alter ligand potency and efficacy. These residues map to regions previously implicated in HCA ligand recognition and align with structural studies distinguishing HCA2 and HCA3 ligand selectivity. Because these substitutions were experimentally tested in a common receptor background, the observed potency shifts establish direct causal relationships between sequence variation and pharmacological divergence. While these data establish causal effects on ligand potency, the precise structural mechanisms by which these substitutions reshape the binding pocket remain to be resolved and will require dedicated structural or computational analyses. Although available structures focus largely on human receptors, our analysis shows that naturally occurring mammalian HCA2 orthologs harbor substitutions at the same or analogous positions that correlate with altered pharmacology. Notably, the F277Y7.40 substitution in human HCA2 shifted ligand preferences toward those of cattle, minke whale, and opossum (Figures 6 and 7). Its fixation in even-toed ungulates and whales, together with independent occurrence in marsupials, indicates recurrent selection in distant mammalian clades, consistent with convergent pressures on HCA2 function imposed by diet, metabolism, or microbiome environments. Concordance between phylogenetic distribution of substitutions and functional effects across representative orthologs further indicates that the sampled panel captures key evolutionary transitions in mammalian HCA2 diversification. Similarly, the T185S45.48 (ECL2) substitution, also present in African elephant HCA2, reduced the potencies of MK-6892 and acipimox in the human receptor background (Figures 6 and 7). Together, these results identify positions within the orthosteric and extended binding pockets as recurrent “hot spots” for adaptive diversification, whereas residues that mediate G-protein coupling remain highly conserved across mammals. This pattern mirrors broader GPCR evolution, in which ligand-binding regions diversify under selective pressure while signaling architecture is preserved. Thus, species-dependent differences in surrogate ligand potency likely reflect evolutionary tuning of HCA2 toward distinct endogenous metabolite repertoires shaped by diet and microbiome context. Physiological and translational implications Our findings indicate that species-specific HCA2 pharmacology must be considered in preclinical ligand evaluation. Compounds potent at human HCA2 can differ substantially in potency, efficacy, or selectivity across mammalian orthologs, creating risks for cross-species translation and interpretation of efficacy or safety. Similarly, structural or pharmacological studies assuming conserved ligand recognition across species may not reflect in vivo receptor function. The data further support the concept that microbiome-derived metabolites shape GPCR ligand specificity. While lactic acid bacteria-derived ligands regulate HCA3 signaling in humans, our results extend this principle to mammalian HCA2, identifying aromatic microbial metabolites as plausible selective pressures underlying rhinoceros receptor divergence. These observations are consistent with co-adaptation of diet, microbiome composition, and metabolite-sensing receptor specificity in host signaling networks. Finally, we show that common culture media components can act as low-level HCA2 agonists, potentially introducing baseline receptor activation in functional or structural assays. This is particularly relevant for metabolite-sensing GPCRs whose endogenous ligands are present in standard media formulations. Together, these findings emphasize that species-specific GPCR pharmacology arises from evolutionary tuning of ligand recognition and should be considered in animal models and cross-species translation of receptor-targeted therapeutics. In conclusion, our data demonstrate that HCA2 exhibits greater evolutionary functional diversification in receptor-level ligand recognition than previously appreciated. Sequence variation concentrated in the ligand-binding pocket, likely shaped by evolutionary pressures such as exposure to microbiome-derived metabolites, drives remarkable divergence in ligand potency and selectivity across mammals while preserving conserved G-protein coupling interfaces. These findings demonstrate that metabolite-sensing GPCRs can undergo lineage-specific functional tuning through adaptive modification of ligand recognition without altering core signaling architecture. This evolutionary plasticity has important implications for receptor biology and cross-species pharmacology and highlights how host-microbiome metabolic interactions can shape GPCR function. Importantly, these findings define receptor-level functional diversification without implying direct effects on organismal fitness or direct in vivo causality. Limitations of the study While our ortholog panel spans major mammalian orders, additional species occupying specialized metabolic niches, such as highly specialized herbivores or animals adapted to extreme environments, may reveal further HCA2 divergence. Although we observe correlations between microbiome metabolite profiles and receptor properties, direct in vivo evidence linking ligand exposure to receptor function in species such as rhinoceros remains to be established. Fecal metabolite profiling was based on aqueous extraction, which enriches soluble metabolites most relevant to intestinal and systemic HCA2 exposure but may underrepresent less soluble microbiome products. The present study, therefore, focuses on receptor-level molecular mechanisms, for which comparative pharmacology and structure-guided mutagenesis provide sufficient resolution to infer functional diversification. Structural data for non-human HCA2 orthologs, along with a broader analysis of downstream signaling bias, would further refine the understanding of ligand-receptor co-adaptation across mammals. Resource availability Lead contact Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Claudia Stäubert (claudia.staeubert@medizin.uni-leipzig.de). Materials availability All unique plasmids generated in this study are available from the lead contact. Data and code availability This article analyzes existing, publicly available data. Therefore, for all sequences of mammalian HCA2 orthologs, NCBI GenBank database accession numbers are listed in Table S2. This article does not report original code. Any additional information required to reanalyze the data reported in this article is available from the lead contact on request. Acknowledgments This work was supported by 10.13039/501100001659Deutsche Forschungsgemeinschaft through CRC1423, project number 421152132, subproject C06 (to C.S.) and subprojects A01/Z03 (to P.S.) and through Germany’s Excellence Strategy – EXC 2008/1 (UniSysCat) – 390540038 (to G.K. and P.S.). C.S. was supported by the 10.13039/501100001659Deutsche Forschungsgemeinschaft – project number 407707190. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. We want to thank the contributors of species samples (Table S9) and Frank Meyer from Zoo Leipzig for the valuable information on the lifestyle and specialties of rhinoceroses. The Swedish Metabolomics Center, Umeå, Sweden (www.swedishmetabolomicscentre.se), Anders Nordström, Annika Johansson, Elin Näsström, and Hans Stenlund are acknowledged for metabolic profiling by HILIC-LC-MS. We acknowledge support from Leipzig University for open access publishing. Author contributions Conceptualization, formal analysis, data curation, and writing – original draft, F.B. and C.S.; methodology, F.B., P.K., G.K., P.S., and C.S.; validation, P.K. and C.S.; investigation, F.B., P.K., and C.S.; resources, C.S.; writing – review and editing and visualization, F.B., G.K., P.S., and C.S.; structural investigations, validation, conclusions, and visualizations, G.K. and P.S.; supervision and project administration, C.S.; funding acquisition, P.S. and C.S. All authors discussed the results and implications and commented on the manuscript at all stages. All authors read and approved the final manuscript. Declaration of interests The authors declare no competing interests. STAR★Methods Key resources table REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesAnti-HA-Peroxidase, High Affinity from rat IgG1 monoclonalSigma-Aldrich12013819001; RRID: AB_390917Biological samplesFecal samples RhinocerosZoo LeipzigN/AFecal samples TapirZoo LeipzigN/AFecal samples ZebraZoo LeipzigN/AGenomic DNA see Table S9 for speciessee Table S9 for sourcesN/ABacterial and virus strainsNEB 5-alpha competent E.coliNEBC2987Chemicals, peptides, and recombinant proteins10x HBSSThermo Fisher Scientific140650561x D-PBSThermo Fisher Scientific141901442-aminophenolSigma-Aldricha71301 (CAS: 95-55-6)2-hydroxy-4-(methylthio)butyric acidMedChemExpressHY-116688 (CAS: 583-91-5)2-hydroxy-phenylacetic acidSigma-AldrichH49804 (CAS: 614-75-5)3-hydroxybenzoic acidSigma-AldrichH20008 (CAS: 99-06-9)3-hydroxybenzyl alcoholSigma-AldrichH20601(CAS: 620-24-6)3-hydroxybutyrateSigma-Aldrich298360 (CAS: 13613-65-5)3-hydroxyoctanoateSanta Cruzsc-214136 (CAS: 88930-08-9)3-hydroxy-phenylacetic acidSigma-AldrichH49901 (CAS: 621-37-4)3-isobutyl-1-methylxanthine (IBMX)Sigma-AldrichI5879 (CAS: 28822-58-4)3-phenylpropionic acidSigma-AldrichW288918 (CAS: 501-52-)4-hydroxybenzoic acidSigma-AldrichH20059 (CAS: 99-96-7)4-hydroxy-phenylacetic acidSigma-AldrichH50004 (CAS: 156-38-7)4-hydroxy-phenylpropionic acidSigma-AldrichH52406 (CAS: 501-97-3)AcifranHölzel Diagnostika Handels GmbHHY-107579 (CAS: 72420-38-3)AcipimoxSigma-AldrichA7856 (CAS: 51037-30-0)Bovine serum albumin (BSA)Sigma-AldrichA3294 (CAS: 9048-46-8)ButyrateSigma-AldrichB5887 (CAS: 156-54-7)Cholera toxinSigma-AldrichC8052 (CAS: 9012-63-9)Citric acidSigma-Aldrich251275 (CAS: 77-92-9)CreatinineSigma-AldrichC4255 (CAS: 60-27-5)Deoxynucleotide (dNTP) Solution SetNEBN0446D-phenylalanineCarl Roth7886.1 (CAS: 673-06-3)D-phenyllactic acidSigma-Aldrich376906 (CAS: 7326-19-4)Dulbecco’s Modified Eagle Medium (DMEM)Thermo Fisher Scientific41966029Dulbecco’s Modified Eagle Medium (DMEM)/ F-12Thermo Fisher Scientific21041025Fetal bovine serum (FBS)Thermo Fisher Scientific10270106FormaldehydeSigma-AldrichF8775 (CAS: 50-00-0)ForskolinSigma-AldrichF3917 (CAS: 66575-29-9)GSK256073Hölzel Diagnostika Handels GmbHHY-119222 (CAS: 862892-90-8)H2O2Sigma-Aldrich1.08600 (CAS: 7722-84-1)HClSigma-Aldrich258148 (CAS: 7647-01-0)HEPES (1M)Thermo Fisher Scientific15630056Imidazole propionic acidSigma-Aldrich77951 (CAS: 1074-59-5)Indolepropionic acidMolportMolport-000-139-896 (CAS: 830-96-6)IPBT-5CABiomol GmbHCay15223 (CAS: 306935-41-1)L-arginineCarl Roth1655.2 (CAS: 74-79-3)Lipofectamine 2000Thermo Fisher Scientific11668019LUF6283Hölzel Diagnostika Handels GmbHHY-100976 (CAS: 92933-48-7)MK-0354Hölzel Diagnostika Handels GmbHHY-13008 (CAS: 851776-28-8)MK-6892Hölzel Diagnostika Handels GmbHHY-10680 (CAS: 917910-45-3)MonomethylfumarateSigma-Aldrich651419 (CAS: 2756-87-8)N-Acetyl-L-prolineSigma-AldrichA0783 (CAS:68-95-1)Na2HPO4 · 2 H2OSigma-AldrichS9763 (CAS: 7558-79-4)Na2SO3Sigma-AldrichS0505 (CAS: 7757-83-7)NicotinamideSigma-Aldrich47865-U (CAS:98-92-0)Nicotinic acidSigma-AldrichN0761 (CAS: 59-67-6)o-phenylenediamineSigma-AldrichP23938 (CAS: 95-54-5)Opti-MEMThermo Fisher Scientific31985070Oxalic acidSigma-Aldrich194131 (CAS: 144-62-7)Penicillin-Streptomycin (5.000 U/ml)Thermo Fisher Scientific15140122Pertussis toxinSigma-Aldrich516560 (CAS: 70323-44-3)Phenylacetic acidSigma-AldrichP16621 (CAS: 103-82-2)Poly-L-lysine solution (0.1% (w/v) in H2O)Sigma-AldrichP8920 (CAS: 25988-63-0)pyridoxineSigma-AldrichP5669 (CAS: 65-23-6)pyridoxic acidSigma-AldrichP9630 (CAS: 82-82-6)Q5 High-Fidelity DNA PolymeraseNEBM0491Roswell Park Memorial Institute (RPMI) custom-made niacin-, glutamin-free mediumCell Culture technologiesN/ATrans-cinnamic acidSigma-AldrichC80857 (CAS: 140-10-3)Trans-indoleacrylic acidSigma-AldrichI3807 (CAS: 29953-71-7)Trans-urocanic acidMolportMolport-006-833-341 (CAS: 3465-72-3)Trigonelline hydrochlorideSigma-Aldricht5509 (CAS: 6138-41-6)Tris/HClSigma-Aldrich10812846001 (CAS: 1185-53-1)Triton X-100Sigma-AldrichT8787 (CAS: 9036-19-5)Trypsin-EDTA (0.25%), phenol redThermo Fisher Scientific25200056Tween 20Sigma-AldrichP1379 (CAS: 9005-64-5)UltraPure 0.5 M EDTA, pH 8Thermo Fisher Scientific15575020VerseneThermo Fisher Scientific15040033Critical commercial assaysAlphaScreen cAMP assay kitPerkinElmer Life Sciences6760635MAlpha SureFire Ultra Multiplex Phospho/Total ERK1/2 Assay KitPerkinElmer Life SciencesMPSU-PTERK-M10KNanoBRET NanoGlo substratePromegaN1572Deposited dataNCBI database accession numbers of mammalian HCAR2 orthologssee Table S2NCBI Nucleotide, RRID: SCR_004860Experimental models: Cell linesCHO-K1 (passage #3 to max. #40)ATCCCCL-61; RRID: CVCL_0214HEK293-T (passage #3 to max #40)ATCCCRL-3216; RRID: CVCL_0063Oligonucleotidessee Table S8 for sequencesSeqLabN/ARecombinant DNAHA-/FLAG-tagged human-HCAR2 in pcDpsPeters et al.16N/AHA-/FLAG-tagged human-HCAR2 L34V in pcDpsThis paperN/AHA-/FLAG-tagged human-HCAR2 T185A in pcDpsThis paperN/AHA-/FLAG-tagged human-HCAR2 T185S in pcDpsThis paperN/AHA-/FLAG-tagged human-HCAR2 F277Y in pcDpsThis paperN/AHA-/FLAG-tagged human-HCAR3 in pcDpsPeters et al.16N/AHA-/FLAG-tagged gorilla-HCAR2 in pcDpsPeters et al.16N/AHA-/FLAG-tagged orangutan-HCAR2 in pcDpsPeters et al.16N/AHA-/FLAG-tagged rhesus macaque-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged mouse-HCAR2 in pcDpsPeters et al.16N/AHA-/FLAG-tagged polar bear-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged cat-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged minke whale-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged cattle-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged horse-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged white rhinoceros-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged African elephant-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged opossum-HCAR2 in pcDpsThis paperN/ApcDpsOkayama et al.85N/Ahuman-HCAR2-mVenus in EGFP-N1This paperN/Ahuman-HCAR3-mVenus in EGFP-N1This paperN/Ahuman arrestin-3-nanoluciferase in EGFP-N1Liebing et al.86N/ASoftware and algorithmsaBSREL (adaptive Branch-Site Random Effects Likelihood)Smith et al.87https://www.datamonkey.org/absrelBioedit Sequence Alignment Editor 7.0.9Hall et al.88BiorenderBiorenderhttps://biorender.com/DatamonkeyWeaver et al.89https://www.datamonkey.orgDNASTAR LasergeneDNASTAR Lasergenehttps://www.dnastar.com/software/lasergene/FELKosakovsky et al.90https://www.datamonkey.org/felGraphPad Prism Version 10Graphpad Softwarehttps://www.graphpad.com/scientific-software/prism/MEGA11: Molecular Evolutionary Genetics Analysis Version 11Tamura et al.91https://megasoftware.net/homeMEME (Mixed Effects Model of Evolution)Murrell et al.92https://www.datamonkey.org/memePyMol Molecular Graphics System Version 2.5.5Schrödinger, LLC, New York, NYhttps://www.pymol.orgRELAXWertheim et al.93https://www.datamonkey.org/relaxSnapGene ViewerSnapGenehttps://www.snapgene.com/snapgene-viewerOtherMaxi-Sorp flat bottom 96-well platesThermo Fisher Scientific44-2404-21OptiPlate-384, White Opaque 384-well MicroplatePerkinElmer Life Sciences6007290Multimode multilabel plate reader able to detect Perkin Elmer Alpha TechnologyPerkinElmer Life SciencesEnVision 2104 multilabel plate reader with barcode #444 mirror (cat#2101-4010), barcode #244 AlphaScreen 570/100 filter (cat #2100-5710) equipped with laser (excitation: 680 nm) by default)Plate Reader to measure absorbance at 492 nm and 620 nmTecanTecan-Reader Infinite M NanoOrbital shakerHeidolphHeidolph Titramax 1000Shaker with 37 °C incubator hoodEdmund BühlerEdmund Bühler shaker SM30 with incubator hood TH30Overhead shakerHeidolphHeidolph Reax 2Multi-dispensing multi-channel pipettesGilson /Brandt / ThermoPipetman G Multichannel P12x200G (12 channel)Tranferpette S-12, (0.5 – 10 μl)Finnpipette Novus (16 channel)Aspiration systemBiovacBiovac 106 with 2 l glass bottle REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesAnti-HA-Peroxidase, High Affinity from rat IgG1 monoclonalSigma-Aldrich12013819001; RRID: AB_390917Biological samplesFecal samples RhinocerosZoo LeipzigN/AFecal samples TapirZoo LeipzigN/AFecal samples ZebraZoo LeipzigN/AGenomic DNA see Table S9 for speciessee Table S9 for sourcesN/ABacterial and virus strainsNEB 5-alpha competent E.coliNEBC2987Chemicals, peptides, and recombinant proteins10x HBSSThermo Fisher Scientific140650561x D-PBSThermo Fisher Scientific141901442-aminophenolSigma-Aldricha71301 (CAS: 95-55-6)2-hydroxy-4-(methylthio)butyric acidMedChemExpressHY-116688 (CAS: 583-91-5)2-hydroxy-phenylacetic acidSigma-AldrichH49804 (CAS: 614-75-5)3-hydroxybenzoic acidSigma-AldrichH20008 (CAS: 99-06-9)3-hydroxybenzyl alcoholSigma-AldrichH20601(CAS: 620-24-6)3-hydroxybutyrateSigma-Aldrich298360 (CAS: 13613-65-5)3-hydroxyoctanoateSanta Cruzsc-214136 (CAS: 88930-08-9)3-hydroxy-phenylacetic acidSigma-AldrichH49901 (CAS: 621-37-4)3-isobutyl-1-methylxanthine (IBMX)Sigma-AldrichI5879 (CAS: 28822-58-4)3-phenylpropionic acidSigma-AldrichW288918 (CAS: 501-52-)4-hydroxybenzoic acidSigma-AldrichH20059 (CAS: 99-96-7)4-hydroxy-phenylacetic acidSigma-AldrichH50004 (CAS: 156-38-7)4-hydroxy-phenylpropionic acidSigma-AldrichH52406 (CAS: 501-97-3)AcifranHölzel Diagnostika Handels GmbHHY-107579 (CAS: 72420-38-3)AcipimoxSigma-AldrichA7856 (CAS: 51037-30-0)Bovine serum albumin (BSA)Sigma-AldrichA3294 (CAS: 9048-46-8)ButyrateSigma-AldrichB5887 (CAS: 156-54-7)Cholera toxinSigma-AldrichC8052 (CAS: 9012-63-9)Citric acidSigma-Aldrich251275 (CAS: 77-92-9)CreatinineSigma-AldrichC4255 (CAS: 60-27-5)Deoxynucleotide (dNTP) Solution SetNEBN0446D-phenylalanineCarl Roth7886.1 (CAS: 673-06-3)D-phenyllactic acidSigma-Aldrich376906 (CAS: 7326-19-4)Dulbecco’s Modified Eagle Medium (DMEM)Thermo Fisher Scientific41966029Dulbecco’s Modified Eagle Medium (DMEM)/ F-12Thermo Fisher Scientific21041025Fetal bovine serum (FBS)Thermo Fisher Scientific10270106FormaldehydeSigma-AldrichF8775 (CAS: 50-00-0)ForskolinSigma-AldrichF3917 (CAS: 66575-29-9)GSK256073Hölzel Diagnostika Handels GmbHHY-119222 (CAS: 862892-90-8)H2O2Sigma-Aldrich1.08600 (CAS: 7722-84-1)HClSigma-Aldrich258148 (CAS: 7647-01-0)HEPES (1M)Thermo Fisher Scientific15630056Imidazole propionic acidSigma-Aldrich77951 (CAS: 1074-59-5)Indolepropionic acidMolportMolport-000-139-896 (CAS: 830-96-6)IPBT-5CABiomol GmbHCay15223 (CAS: 306935-41-1)L-arginineCarl Roth1655.2 (CAS: 74-79-3)Lipofectamine 2000Thermo Fisher Scientific11668019LUF6283Hölzel Diagnostika Handels GmbHHY-100976 (CAS: 92933-48-7)MK-0354Hölzel Diagnostika Handels GmbHHY-13008 (CAS: 851776-28-8)MK-6892Hölzel Diagnostika Handels GmbHHY-10680 (CAS: 917910-45-3)MonomethylfumarateSigma-Aldrich651419 (CAS: 2756-87-8)N-Acetyl-L-prolineSigma-AldrichA0783 (CAS:68-95-1)Na2HPO4 · 2 H2OSigma-AldrichS9763 (CAS: 7558-79-4)Na2SO3Sigma-AldrichS0505 (CAS: 7757-83-7)NicotinamideSigma-Aldrich47865-U (CAS:98-92-0)Nicotinic acidSigma-AldrichN0761 (CAS: 59-67-6)o-phenylenediamineSigma-AldrichP23938 (CAS: 95-54-5)Opti-MEMThermo Fisher Scientific31985070Oxalic acidSigma-Aldrich194131 (CAS: 144-62-7)Penicillin-Streptomycin (5.000 U/ml)Thermo Fisher Scientific15140122Pertussis toxinSigma-Aldrich516560 (CAS: 70323-44-3)Phenylacetic acidSigma-AldrichP16621 (CAS: 103-82-2)Poly-L-lysine solution (0.1% (w/v) in H2O)Sigma-AldrichP8920 (CAS: 25988-63-0)pyridoxineSigma-AldrichP5669 (CAS: 65-23-6)pyridoxic acidSigma-AldrichP9630 (CAS: 82-82-6)Q5 High-Fidelity DNA PolymeraseNEBM0491Roswell Park Memorial Institute (RPMI) custom-made niacin-, glutamin-free mediumCell Culture technologiesN/ATrans-cinnamic acidSigma-AldrichC80857 (CAS: 140-10-3)Trans-indoleacrylic acidSigma-AldrichI3807 (CAS: 29953-71-7)Trans-urocanic acidMolportMolport-006-833-341 (CAS: 3465-72-3)Trigonelline hydrochlorideSigma-Aldricht5509 (CAS: 6138-41-6)Tris/HClSigma-Aldrich10812846001 (CAS: 1185-53-1)Triton X-100Sigma-AldrichT8787 (CAS: 9036-19-5)Trypsin-EDTA (0.25%), phenol redThermo Fisher Scientific25200056Tween 20Sigma-AldrichP1379 (CAS: 9005-64-5)UltraPure 0.5 M EDTA, pH 8Thermo Fisher Scientific15575020VerseneThermo Fisher Scientific15040033Critical commercial assaysAlphaScreen cAMP assay kitPerkinElmer Life Sciences6760635MAlpha SureFire Ultra Multiplex Phospho/Total ERK1/2 Assay KitPerkinElmer Life SciencesMPSU-PTERK-M10KNanoBRET NanoGlo substratePromegaN1572Deposited dataNCBI database accession numbers of mammalian HCAR2 orthologssee Table S2NCBI Nucleotide, RRID: SCR_004860Experimental models: Cell linesCHO-K1 (passage #3 to max. #40)ATCCCCL-61; RRID: CVCL_0214HEK293-T (passage #3 to max #40)ATCCCRL-3216; RRID: CVCL_0063Oligonucleotidessee Table S8 for sequencesSeqLabN/ARecombinant DNAHA-/FLAG-tagged human-HCAR2 in pcDpsPeters et al.16N/AHA-/FLAG-tagged human-HCAR2 L34V in pcDpsThis paperN/AHA-/FLAG-tagged human-HCAR2 T185A in pcDpsThis paperN/AHA-/FLAG-tagged human-HCAR2 T185S in pcDpsThis paperN/AHA-/FLAG-tagged human-HCAR2 F277Y in pcDpsThis paperN/AHA-/FLAG-tagged human-HCAR3 in pcDpsPeters et al.16N/AHA-/FLAG-tagged gorilla-HCAR2 in pcDpsPeters et al.16N/AHA-/FLAG-tagged orangutan-HCAR2 in pcDpsPeters et al.16N/AHA-/FLAG-tagged rhesus macaque-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged mouse-HCAR2 in pcDpsPeters et al.16N/AHA-/FLAG-tagged polar bear-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged cat-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged minke whale-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged cattle-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged horse-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged white rhinoceros-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged African elephant-HCAR2 in pcDpsThis paperN/AHA-/FLAG-tagged opossum-HCAR2 in pcDpsThis paperN/ApcDpsOkayama et al.85N/Ahuman-HCAR2-mVenus in EGFP-N1This paperN/Ahuman-HCAR3-mVenus in EGFP-N1This paperN/Ahuman arrestin-3-nanoluciferase in EGFP-N1Liebing et al.86N/ASoftware and algorithmsaBSREL (adaptive Branch-Site Random Effects Likelihood)Smith et al.87https://www.datamonkey.org/absrelBioedit Sequence Alignment Editor 7.0.9Hall et al.88BiorenderBiorenderhttps://biorender.com/DatamonkeyWeaver et al.89https://www.datamonkey.orgDNASTAR LasergeneDNASTAR Lasergenehttps://www.dnastar.com/software/lasergene/FELKosakovsky et al.90https://www.datamonkey.org/felGraphPad Prism Version 10Graphpad Softwarehttps://www.graphpad.com/scientific-software/prism/MEGA11: Molecular Evolutionary Genetics Analysis Version 11Tamura et al.91https://megasoftware.net/homeMEME (Mixed Effects Model of Evolution)Murrell et al.92https://www.datamonkey.org/memePyMol Molecular Graphics System Version 2.5.5Schrödinger, LLC, New York, NYhttps://www.pymol.orgRELAXWertheim et al.93https://www.datamonkey.org/relaxSnapGene ViewerSnapGenehttps://www.snapgene.com/snapgene-viewerOtherMaxi-Sorp flat bottom 96-well platesThermo Fisher Scientific44-2404-21OptiPlate-384, White Opaque 384-well MicroplatePerkinElmer Life Sciences6007290Multimode multilabel plate reader able to detect Perkin Elmer Alpha TechnologyPerkinElmer Life SciencesEnVision 2104 multilabel plate reader with barcode #444 mirror (cat#2101-4010), barcode #244 AlphaScreen 570/100 filter (cat #2100-5710) equipped with laser (excitation: 680 nm) by default)Plate Reader to measure absorbance at 492 nm and 620 nmTecanTecan-Reader Infinite M NanoOrbital shakerHeidolphHeidolph Titramax 1000Shaker with 37 °C incubator hoodEdmund BühlerEdmund Bühler shaker SM30 with incubator hood TH30Overhead shakerHeidolphHeidolph Reax 2Multi-dispensing multi-channel pipettesGilson /Brandt / ThermoPipetman G Multichannel P12x200G (12 channel)Tranferpette S-12, (0.5 – 10 μl)Finnpipette Novus (16 channel)Aspiration systemBiovacBiovac 106 with 2 l glass bottle Experimental model and study participant details Cell lines The Chinese hamster ovary cell line CHO-K1 (ATCC CCL-6, female) and the human embryonic kidney cell line HEK293-T (ATCC CRL-3216, female) were acquired from the American Type Culture Collection. CHO-K1 cells were grown in Dulbecco’s Modified Eagle Medium: Nutrient Mixture F-12 (DMEM/F12) and HEK293-T cells in Dulbecco’s Modified Eagle medium (DMEM), both supplemented with 10 % fetal bovine serum (FBS), 10,000 U/mL penicillin, and 10,000 μg/ml streptomycin. Cells were maintained at 37 °C in a humidified 5 % CO2 incubator. For transient transfection, Lipofectamine 2000 (Thermo Scientific) was used according to the manufacturer’s instructions. Ethics statement All fecal samples were collected noninvasively from the ground during routine animal husbandry procedures by the professional zookeepers at Zoo Leipzig. No animal activities or interventions were involved. Method details HCAR2 ortholog identification HCAR2 sequences of various mammalian species were obtained using NCBI Blast (accession numbers are listed in Table S2). To generate expression plasmid encoding selected mammalian HCAR2 full-length orthologs, primer pairs (Table S8) were used to amplify specific HCAR2 sequences from genomic DNA samples (Table S9).16,94 Human, gorilla, orangutan, and mouse HCAR2 have been previously described.16 PCR reactions were performed with Q5 polymerase under variable annealing and elongation conditions following the manufacturer’s protocol. The full-length HCAR2 orthologs were inserted into the mammalian expression vector pcDps85 and epitope-tagged with an N-terminal hemagglutinin (HA) epitope and a C-terminal FLAG-tag by a PCR-based overlapping fragment approach. In addition, HCAR2 was C-terminally tagged with mVenus. The identity of all constructs and the correctness of all PCR-derived sequences were confirmed by sequencing (Seqlab). All primers used for cloning are listed in Table S8. Sequence alignments and datamonkey analyses HCAR2 nucleotide sequences of 220 mammalian species were aligned to infer a phylogenetic tree. All nucleotide alignments were generated with the Clustal W algorithm (Bioedit Sequence Alignment Editor 7.0.9),88,95 followed by manual trimming where gaps were deleted. The phylogenetic evolutionary history was inferred by the Maximum Likelihood method based on the General Time Reversible model using MEGA11.91 The resulting tree is shown in Figure S6. In addition, this nucleotide alignment was uploaded to the Datamonkey web server to be analyzed for evolutionary signatures.89 MEME (Mixed Effects Model of Evolution) was performed to detect episodic/positive selection at individual sites.92 aBSREL (adaptive Branch-Site Random Effects Likelihood) was applied to detect positive selection at individual branches.87 RELAX tests were used to analyze for selection relaxation or intensification along the specified set of branches93 and FEL (Fixed Effects Likelihood)90 were used to analyze whether individual sites were subject to pervasive (across the whole phylogeny) positive or purifying selection. Transient transfection for functional assays For transient transfection, Lipofectamine 2000 (Thermo Scientific) was used according to the manufacturer’s instructions. For cAMP assay and ELISA analyses, CHO-K1 cells were split into 25 cm2-cell culture flasks (0.9 x 106 cells/flask) and transfected with a total amount of 3 μg of plasmid the following day. For arrestin-3 recruitment analyses, HEK293-T cells were seeded into 75 cm2-cell culture flasks (3.5 x 106 cells/flask) and transfected the following day with 3.3 μg plasmid encoding C-terminally mVenus-tagged human HCAR2 and 0.7 μg arrestin-3-Nanoluciferase (Nluc) construct. ALPHAScreen cAMP assay and cell-surface ELISA ALPHAScreen cAMP inhibition assays and ELISA-based analyses of cell surface expression of N-terminal HA-tagged receptor constructs were performed as exhaustively described in this protocol.96 Cyclic AMP content of cell extracts was determined using the ALPHAScreen cAMP Detection Kit (Revvity). CHO-K1 cells were incubated in serum-free media overnight. Cells were stimulated for 15 min at 37 °C in HBSS with 1 mM 3-isobutyl-1-methylxanthine and 2 μM forskolin containing different agonists in different concentrations. Reactions were stopped by aspiration of the stimulation buffer. Cells were lysed in 50 μl/well of lysis buffer. For measurement, 5 μl lysate of each well were transferred into a 384-well plate and proceeded as described before in detail.96 Alpha SureFire Ultra Multiplex pERK 1/2 & total ERK assay The kit measures the phosphorylation and total levels of endogenous ERK1/2 in cellular lysates. The signal at 615 nm (Eu) corresponds to the pERK level, while the signal at 545 nm (Tb) corresponds to the total ERK level. The pERK/total ERK ratio of cell extracts was determined using the Alpha SureFire Ultra Multiplex p-ERK 1/2 (Thr202/Thr204)/ Total ERK assay according to the manufacturer (Revvity). CHO-K1 cells were split in 96-well plates 24 h after transfection and serum-starved overnight before the assay. 48 h after transfection, cells were stimulated with agonists in HBSS buffer (with 20 mM HEPES) for 15 min at 37 °C. Stimulation was stopped by aspiration of the media, and cells were lysed in 50 μl/well of the supplied lysis buffer with 250 μM protease inhibitor AEBSF. From each well, 10 μl of lysate were transferred into a 384-well plate, and acceptor beads and donor beads were added according to the manufacturer’s protocol. Measurements were taken using the EnVision Microplate reader (Revvity). Bioluminescence resonance energy transfer (BRET) assay 24 h after transfection, HEK293-T cells were harvested and plated in white 96-well culture plates coated with poly-L-lysine. 16 h before the assay, cells were cultured in serum and niacin-free media. BRET assays were performed 72 h after transfection. Cells were washed with HBSS and incubated for 60 min in 32.5 μl HBSS/well at room temperature. 12.5 μl of Nluc substrate (Promega, diluted 1:250 in HBSS) were added, and three consecutive baseline reads were taken using the Victor Microplate reader (Revvity). Luminescence was measured at 460 ± 80 nm (acceptor) and 645 ± 75 nm (donor) for Nluc. BRET ratio was calculated as the ratio between acceptor and donor emission.97 Cells were stimulated by adding 5 μl of 10x ligand solution, and the measurement was continued for 20 min. The basal BRET ratio before ligand stimulation was defined as the average of three consecutive BRET values. To quantify ligand-induced changes, ΔBRET was calculated for each well as % over basal [(Ratiostim - Ratiobasal)/Ratiobasal] x 100. Subsequently, the average ΔBRET of vehicle control was subtracted.97 Liquid chromatography-mass spectrometry (LC-MS) measurement Fecal samples were collected non-invasively from the ground during routine animal husbandry procedures by the professional zoo keepers of the Zoo Leipzig and frozen at -20 °C. Before initial extraction, fecal samples were thawed, and 3-4 g were weighed in a 50 ml tube and filled up with 1xPBS to 30-40 ml, respectively (equals 0.1 g/ml). Samples were vortexed and shaken at 37 °C for 20 min. Subsequently, samples were centrifuged (5000 g) for 10 min. 12 ml of the supernatant was transferred to a 15 ml tube and again centrifuged (5000 g) for 10 min. Subsequently, the supernatant was sterile filtered (0.45 μM filter) and aliquoted (500 μl, 1:10 diluted fecal-PBS). Sample preparation of 100 μl of these fecal extracts was performed as previously described.98 In detail, 900 μl of extraction buffer (90/10 v/v methanol: water), including internal standards (see supplementary), were added to the samples. The samples were shaken at 30 Hz for 2 min in a mixer mill, and proteins were precipitated at -20 °C for 2 h. The samples were centrifuged at 4 °C, 14,000 rpm (18620 g), for 10 min. 200 μL supernatant was transferred to microvials and evaporated to dryness in a speed-vac concentrator, and the samples were stored at -80 °C until analysis. Small aliquots of the remaining supernatants were pooled and used to create quality control (QC) samples. Metabolic profiling by HILIC-LC-MS was performed at the Swedish Metabolomics Center in Umeå, Sweden. Solvents and eluent additives: Methanol and Acetonitrile, HPLC-grade, were obtained from Fischer Scientific (Waltham, MA, USA). Ammonium formate and Medronic acid (Methylenediphosphonic acid) were obtained from Sigma (St. Louis, MO, USA). Reference and tuning standards: Purine, HP-0921 (Hexakis(1H, 1H, 3H-tetrafluoropropoxy)phosphazine), 4 μM, 1 μM Calibrant, ESI-TOF, ESI-L Low Concentration Tuning Mix, HP-0321 (Hexamethoxyphosphazine), 0.1 mM were obtained from Agilent Technologies (Santa Clara, CA, USA). Stable isotope internal standards: 13C9-Phenylalanine, D4-Cholic acid, Salicylic acid-D6, L-glutamic acid-13C5,15N, and D-sucrose-13C12 were obtained from Sigma (St. Louis, MO, USA). L-proline-13C5 and alpha-ketoglutarate-13C4 were obtained from Cil (Andover, MA, USA). Before the LC-MS analysis, the samples were re-suspended in 10 + 10 μl methanol and elution solvent A. The samples were analyzed according to a randomized run order. All samples were first analyzed in positive ionization mode, followed by a second injection in negative ionization mode. The chromatographic separation was performed on an Agilent 1290 Infinity UHPLC system (Agilent Technologies, Waldbronn, Germany). 2 μL of each sample were injected onto an Atlantis Premier BEH-Z-HILIC VanGuard FIT (1.7 μm, 2.1 x 50 mm) column (Waters Corporation, Milford, MA, USA) held at 40 °C. The HILIC gradient elution solvents were A) H2O, 10 mM ammonium formate, 5μM Medronic acid, pH 9, B) 90:10 Acetonitrile: H2O, 10 mM ammonium formate, pH 9. Chromatographic separation was achieved using a linear gradient (at a flow rate of 0.4 ml/min) starting at 90 % B and decreasing to 80 % B over 6 min; B was reduced to 20 % over 3.5 min and held at 20 % for 1.5 min; B was increased to 90 % for 0.5 min and the flow-rate was increased to 0.7 ml/min for 2 min; these conditions were held for 0.5 min, after which the flow-rate was reduced to 0.4 ml/min for 0.5 min before the next injection. The compounds were detected with an Agilent 6546 Q-TOF mass spectrometer equipped with a jet stream electrospray ion source operating in positive or negative ion mode. A reference interface was connected for accurate mass measurements; the reference ions purine (4 μM) and HP-0921 (Hexakis(1H, 1H, 3H-tetrafluoropropoxy)phosphazine) (1 μM) were infused directly into the MS at a flow rate of 0.05 ml/min for internal calibration, and the monitored ions were purine m/z 121.05 and m/z 119.03632; HP-0921 m/z 922.0098 and m/z 966.000725 for positive and negative mode respectively. The gas temperature was set to 150°C, the drying gas flow was set to 8 l/min, and the nebulizer pressure was 35 psi. The sheath gas temp was set to 350 °C, and the sheath gas flow was 11 l/min. The capillary voltage was set to 4000 V in both positive and negative ion mode. The nozzle voltage was 300 V. The fragmentor voltage was 120 V, the skimmer 65 V, and the OCT 1 RF Vpp 750 V. The collision energy was set to 0 V. The m/z range was 70 - 1700, and data was collected in centroid mode with an acquisition rate of 4 scans s-1. MS/MS analysis was run on the QC samples for identification purposes. All data pre-processing was performed using the Agilent MassHunter Profinder version B.10.0.2 (Agilent Technologies Inc., Santa Clara, CA, USA). Data preprocessing was conducted in a targeted manner. A pre-defined list of metabolites commonly found in plasma and serum was searched for using the Batch Targeted feature extraction in MassHunter Profinder. An in-house HILIC-LC-MS library built from authentic standards and run on the same system with the same chromatographic and mass spectrometry settings was used for targeted processing. Quantification and statistical analysis All experiments were performed at least three times, as indicated in the figure legends, and, unless otherwise indicated, they are presented as mean ± the standard error of the mean (SEM). EC50 values were determined in GraphPad Prism by fitting dose-dependence responses to a three-parameter sigmoidal curve (assuming a Hill slope of 1). Statistical analyses were performed using GraphPad Prism as indicated in the figure legends (Figures 1H, 3C, 3E, 4C, and 6B). P values with P ≤ 0.05 were considered statistically significant and are defined as follows: ∗P ≤ 0.05, ∗∗P≤ 0.01, ∗∗∗P ≤ 0.001.