Evolutionary dynamics of Erinaceus coronavirus in Italy: spike protein adaptation and variable CD200 ortholog insertion

1 Introduction

Coronaviruses (CoVs) classified within the genus Betacoronavirus (Beta-CoV) are divided into five subgenera: Embecovirus, Sarbecovirus, Merbecovirus, Nobecovirus, and Hibecovirus (1).

Betacoronavirus is the only genus within the family Coronaviridae that includes zoonotic strains known to cause severe diseases in humans. Within the Sarbecovirus are SARS-CoV and SARS-CoV-2 (Severe Acute Respiratory Syndrome Coronavirus) and within Merbecovirus is MERS-CoV (Middle East Respiratory Syndrome).

Phylogenetic analyses have established bats as the natural reservoirs of Beta-CoV (2) however, zoonotic CoVs typically require transmission through intermediate hosts, such as civets and raccoon dogs for SARS-related coronaviruses and dromedary camels for MERS-CoV (3).

The CoV genome consists of a single-stranded, positive-sense RNA and displays a mutation rate (10^-³ to 10^-6 substitutions per site per year) typical of RNA viruses, although lower than other respiratory RNA pathogens, such as those of influenza A virus (4), due to the proofreading activity of the nsp14 exonuclease (5). Despite this comparatively reduced substitution rate, CoV still accumulate substantial genetic diversity through point mutation, insertions, deletions, and the acquisition of novel genomic regions through recombination between strains. The combination of these evolutionary mechanisms and the structural plasticity of the spike (S) protein enable coronaviruses to cross species barriers, adapt to new hosts, and potentially pose a threat to human health (6, 7).

Since the emergence of zoonotic coronaviruses, the number of studies aimed at characterizing CoV strains and species circulating in wildlife has increased. Among wild animals, hedgehogs have been identified as potential reservoirs of coronaviruses. In 2013, a novel Beta-CoV species was discovered in the Western European hedgehog (Erinaceus europaeus) in Germany (8) and recognized as Erinaceus coronavirus (EriCoV). Subsequently, in 2019, the species HKU31 was reported in the Amur hedgehog (Erinaceus amurensis) in China (9). Both viruses belong to the group of MERS-like coronaviruses.

Hedgehogs are considered likely natural hosts for Merbecovirus, as multiple surveillance studies have consistently detected Merbecovirus RNA genome across European hedgehog populations with high detection rates, ranging from approximately 10% in the United Kingdom to around 60.0% in Germany and Italy, 20.0% in Poland, 24.5% in Portugal and 27.4% in Hungary (8, 10–16). However, no confirmed cross−species transmission events have been reported, and their role as true reservoirs capable of transmitting the virus to other species remains unproven.

EriCoVs possess a genome of approximately 30 kb, with a genomic organization similar to that of other Merbecovirus strains. The genome comprises ORF1ab, which encodes non-structural proteins (nsp1–16), followed by genes encoding the structural proteins spike (S), envelope (E), membrane (M), and nucleocapsid (N), as well as several accessory proteins encoded by additional ORFs located at the 3′ end of the genome (17). Notably, in several strains detected in Erinaceus europaeus in Italy, an additional open reading frame encoding a CD200 ortholog has been identified (17).

In Europe, the presence of EriCoV in hedgehogs has been reported in Germany, the United Kingdom, Poland, Italy, Russia, Hungary, and Portugal, with nucleotide identity at the genomic level ranging from 89.0% to 100% (8, 11–14, 17, 18). Sequence identity decreases to approximately 60.0% or lower when compared with MERS-CoV strains, particularly in the spike gene (17).

Hedgehogs are small insectivorous mammals that inhabit a wide range of environments. Increasing urbanization has led to closer contact between hedgehogs and humans increasing the likelihood of direct human exposure. The diversity of coronaviruses harboured by hedgehogs remains largely unknown and is likely underestimated. Therefore, investigating viral diversity in wildlife is essential for monitoring the emergence and spread of novel viral species that may pose a risk to human health.

In previous studies conducted in Italy, our group sequenced and analysed eight complete EriCoV genomes obtained from 40 hedgehogs originating from urban areas in Northern Italy and admitted to wildlife rehabilitation centres (WRCs) (17). At that time, only three complete European EriCoV genomes were available on NCBI database, corresponding to strains from Germany (NC_039207.1/KC545383), the United Kingdom (KX022605), and Russia (MN067812). To expand current knowledge of coronaviruses circulating in hedgehogs, a new study was therefore undertaken, increasing the number of Italian EriCoV strains subjected to full-genome sequencing.

2 Materials and methods

2.1 Samples

A total of 55 EriCoV RNA–positive samples, collected between 2018 and 2022, previously identified as positive through diagnostic PCR testing, were subsequently sent to the Istituto Superiore di Sanità (ISS) for whole-genome sequencing. Eleven of them were characterized by sequencing short genomic regions in previous papers (19). Samples were collected from 7 Italian regions: Lazio (n=4), Umbria (n=5), Emilia-Romagna (n=19) Piemonte (n=6), Veneto (n=4), Lombardia (n=15), Puglia (n=2) (Figure 1).

2.2 Whole-genome sequencing

The protocol followed for Next Generation Sequencing was previously described by De Sabato, Di Bartolo (17). The libraries were prepared following the sequence independent single primer amplification (SISPA). The RNA was reverse transcribed by SuperScript III Reverse Transcriptase (Thermo Fisher Scientific, Monza, Italy). The second strand of cDNA was synthesized by DNA Polymerase I Large (Klenow) Fragment (Promega, Milan, Italy) and product amplified by the PCRBIO HiFi Polymerase (PCR BIOSYSTEM, Resnova s.r.l., Rome, Italy). Libraries were prepared by NEBNext Fast DNA Library Prep Set for Ion Torrent (New England BioLabs, Milan, Italy) and the NGS run performed using the Ion 520 & Ion 530 Kit-Chef using an Ion chip 530 on Ion Gene Studio S5 System.

2.3 Complete genome assembling

Reads obtained by NGS run were analysed using command line tools. The quality of the reads was checked using the FastQC Read Quality tool and the fastp tool for trimming. Reads were mapped against the EriCoV reference genome (ErinaceusCoV/2012-174/GER/2012, NC_039207.1) using Bowtie2 software and the BAM output visualized by IGV software. The consensus sequences were compared to the contigs built by Spades v. 3.1. The Open Reading Frames (ORFs) were predicted using the online tool ORF Finder (NCBI, http://www.ncbi.nlm.nih.gov/gorf/gorf.html).

The complete genome Accession numbers on NCBI database are: PV428872-PV428875, PX866705-PX866735.

2.4 Recombination analysis

To detect any recombination events, the complete genome of the detected CoVs and those published on NCBI database were analysed using different methods available in the RDP package v.4 (20) and RAT (Recombination Analysis Tool) (21). Only recombination events identified by at least three different methods with p-value below 1.0 × 10E-10 were taken into account.

2.5 Codon sites analysis

The alignment of complete S gene of European EriCoV strains were used to detect individual codon sites under positive or purifying selection pressure. Positive selection at individual codon sites was assessed using three methods implemented in the HyPhy package (www.datamonkey.org). Fixed-Effects Likelihood (FEL) estimates site−specific nonsynonymous (β) and synonymous (α) substitution rates and identifies positive selection when β > α with a significance threshold of p < 0.1. Fast Unconstrained Bayesian Approximation (FUBAR) infers selection using a Bayesian approach and considers a site positively selected when the posterior probability for β > α exceeds 0.9. Mixed Effects Model of Evolution (MEME) tests for episodic diversifying selection and identifies a site as positively selected when at least one branch shows β > α with a p−value < 0.1 and an empirical Bayes factor > 20. Only codon sites supported by all three methods were considered under positive selection.

2.6 CoV RdRp and CD200 ortholog detection by reverse transcription-PCR

For those RNA samples for which was not possible to obtain complete genomes, the RdRp region of ~430 nt bp was amplified using a RT-PCR assay designed to amplify all known mammalian and avian coronaviruses (22). The first step (RNA reverse transcription and DNA amplification) was performed using the Qiagen OneStep RT-PCR kit (Qiagen, Hilden, Germany) and the second step (nested DNA amplification) was carried out with the Platinum Taq DNA Polymerase kit (Thermo Fischer Scientific, Monza, Italy). A second RT-PCR was performed using primer pairs for CD200 ortholog insertion found in EriCoV genomes in Italy, as previously described (17). The length of the amplicon obtained ranged between 500 and >800 depending on the absence or presence of the insertion, respectively.

The PCR products were analysed on 2% agarose gel. The DNA amplicons were sequenced using the same PCR primers by Eurofins Genomic (Milan, Italy). The RdRp and CD200 ortholog sequences were submitted to the NCBI database. Accession numbers: RdRp: PX666428-PX666446; CD200: PX971016- PX971020.

2.7 Phylogenetic analysis

To perform phylogenetic analysis 4 different datasets were built.

The first dataset included 64 complete genomes: 35 Italian genomic sequences from this study, 8 Italian sequences from previous studies, 14 international sequences from NCBI database, 5 HKU31 and 2 MERS-CoV sequences used as outgroup.

The second datasets included 139 nucleotide sequences from short genomic fragment within the RdRp: 54 sequences from this study, 26 sequences from previous Italian studies, 46 sequences from international studies obtained from NCBI database, 11 HKU31 and 2 MERS-CoV sequences used as outgroup.

Third and fourth datasets were built using amino acid sequences to account for functional conservation and to enable reliable comparisons between highly divergent sequences.

The third dataset included 65 amino acidic sequences from complete spike proteins: 35 sequences from this study, 8 Italian sequences from previous studies, 14 international sequences from NCBI database, 5 HKU31, 1 from HKU4 and 2 MERS-CoV sequences used as outgroup.

The fourth dataset included 63 CD200 amino acidic sequences: 20 sequences from this study, 16 from mammals, 15 CD200 orthologs from viruses and the 12 CD200 reported in Italian EriCoVs from a previous study.

Nucleotide sequence analysis was performed using BlastN on NCBI database. Sequence alignments were performed by Aliview tool. The phylogenetic analysis and Maximum Likelihood tree reconstruction were performed using IQTREE 2 tool using the best model suggested by Modeltest and with 1000 bootstrap replication.

3 Results

3.1 EriCoV sequences

Thirty-five out of 55 samples were fully sequenced as a result of NGS. The NGS runs for those samples produced between 514,227 and 1,684,424 reads per sample (mean 996,439 ± 345,846) with a number of reads aligned on EriCoV reference sequence (NC_039207.1) ranging between 8,579 and 1,372,557 (mean 136,101 ± 312,089) with coverage between 95X and 3.242X (mean 719 ± 845).

For the remaining 20 samples, although NGS runs yielded similar total reads counts, the number of reads mapping to the EriCoV reference genome was extremely low (<250 reads), which was insufficient to build either the entire genome or partial genomic regions.

The genome organization of all strains was consistent with that of other European EriCoV, comprising 11 open reading frames (ORFs): 5’UTR-1ab- S-3a-3b-4a-4b-5-E-M-8b-N-3′UTR. Fifteen complete genomes contained the CD200 ortholog located between Spike and ORF3a.

Samples for which the complete genome was not obtained and therefore could not be included in whole-genome phylogenetic analyses, partial amplification of the RdRp was performed and the presence of CD200 was investigated by RT-PCR.

The RdRp sequences were obtained for 19 of the 20 samples, whereas one sample failed to produce RdRp amplicons by either NGS or Sanger. The CD200 was detected in 5 samples out of the 20 samples.

3.2 Sequence and phylogenetic analysis

BLAST analysis revealed that the complete genomes obtained in this study matched more closely with the EriCoV reference genome reported from an E. europaeus in Germany in 2012 (NC_039207.1) with nucleotide identity (nt.id.) ranging from 91.2% to 93.5%. Pairwise comparison among Italian sequences showed nt.id. ranging between 91.6%-99.9%. The phylogenetic analysis (Figure 2) based on complete genome sequences revealed the separation of European strains into two major clusters: one represented by Russian strains and the second comprising sequences from Germany, Italy, Poland, France, the United Kingdom and Hungary.

Within the second cluster, non-Italian sequences formed 2 sub-clusters, whereas Italian sequences were distributed across multiple sub-clusters.

Italian sequences shared 90.5%-92.4% nt.id. with sequences from cluster with Russian strains and 91.0%-93.5% with sequences from Germany, Poland, France, the United Kingdom and Hungary.

No clear geographical distribution was observed, as sequences from Italian regions were interspersed.

Sequence analysis of the RdRp region showed nucleotide identities of the 93.7%-97.4% relative to the reference sequence NC039207.1.

Among Italian RdRp sequences, nucleotide identity ranged between 92.6% and 99.9%, while identity with non-Italian sequences ranged from 91.6% to 97.6%. Phylogenetic reconstruction (Figure 3) based on the RdRp sequences confirmed the topology observed in the whole genomic tree, resulting in a not defined geographical clustering. Italian sequences were interspersed with the other non-Italian sequences. Sequences from Central Italy (Lazio and Umbria regions) formed one single cluster with other Italian sequences reported in the same regions in 2021-2022, except for EriCoV/Italy_PG144a/2021 sequence that clustered with sequences detected in Northern Italy. Sequences from Puglia region (Southern Italy) formed one cluster with sequences from the Lombardia region (Northern Italy) and one sequence from Hungary (PX460032.1).

3.3 Spike protein analysis

Analysing the spike gene, at both the nucleotide and amino acid level, showed that sequences from this study had the highest nucleotide and amino acidic identities to previously described Italian EriCoV sequences (15, 16). Lower sequence identity was observed with non-Italian EriCoV sequences, followed by HKU31 and MERS-CoV species as reported in Table 1.

Table 1. Nucleotide and amino acidic identities of the spike protein between EriCoVs sequences from this study and previously published Italian strains, non-Italian strains, MERS-CoV and HKU31.

Accession numbers: NC_019843.3.

Accession numbers: MK907286, MK907287, OM451212, MN879331, OM451213

Accession numbers: KC545386, NC_039207.1, MK679660, PP721362, PP721361, PQ869634, PQ869635, PX460032, PV359183, PV359184, PV359185, PP421220, PP421221, PV359182

Accession numbers: MW246796, MW246798, MW246802, MW246800, MW246797, MW246801, MW246795, MW246799

Phylogenetic analysis of the spike protein confirmed the results observed with the analyses of whole genomes, with the EriCoV sequences grouped independently of their geographical origin and no evident geographic structuring (Supplementary Figure 1).

The selection pressure profiles of the S protein of European Eri-CoV strains were analysed. As expected, the calculated dN/dS ratio was < 1 (0.172), indicating that the S protein of these strains evolved predominantly under purifying (negative) selection. Conversely, 30 individual codon sites were identified as being under positive selection, as supported by three independent methods (MEME, FUBAR and FEL) used. Of these sites, 10 codons were detected within the N-terminal domain of the S1, 3 within the receptor binding domain (RBD) and 17 within S2 region. Among the positively selected sites, two were located near the fusion peptide region (864-866 nt) and two downstream of the Heptad 2 repeat (1145-1147 nt) (Table 2).

Table 2. Selection pressure profiles of the S protein.

The alignment of the receptor binding domain (RBD, 18 aminoacidic residues) of the Italian strains with the other European, HKU4 MERS-CoV and HKU31 strains is shown in Table 3. Among the 18 RBD residues, the Italian EriCoV strains shared 3 positions identical to MERS-CoV. Fifteen positions showed identical residues with the non-Italian EriCoV strains group, however, 7 showed high heterogeneity with non-synonymous changings (463, 506, 535, 536, 537, 538, 555).

Table 3. Multiple sequence alignments showing variations in key amino acid binding residues within the Spike protein.

Accession numbers: NC_019843.3.

Accession numbers: NC_009019.1

Accession numbers: MK907286, MK907287, OM451212, MN879331, OM451213

Accession numbers: MW246796, MW246798, MW246802, MW246800, MW246797, MW246801, MW246795, MW246799

Accession numbers: KC545386, NC_039207.1, MK679660, PP721362, PP721361, PQ869634, PQ869635, PX460032, PV359183, PV359184, PV359185, PP421220, PP421221, PV359182

Compared to the MERS-CoV, the remaining 12 amino acid positions showed non-synonymous substitutions and 3 deletions.

Five out of 12 substitutions and all the 3 deletions were located in the two main contact interfaces between RBD and human DPP4 (Patch1 and Patch2 interfaces).

In Patch 1, corresponding to the hydrophilic interface, two substitutions were observed: residues E536 and D537 were replaced by 536PQKST in a change from negatively charged aa to positively, polar or non-polar aa. The position D537, was conserved or changed to 537N a polar aa.

In Patch 2, characterized by a hydrophobic core surrounded by a hydrophilic periphery, three substitutions and the 3 deletions were detected. Specifically, aa L506, W553, W555 were replaced by non-conservative aa such as: 506TS, 553Y, 555LV, resulting in loss of hydrophobic amino acids or conservative as 506I.

In addition, the hydrophilic interface built by D510, R511, E513, Y540 is disrupted by the deletion of D510, R511, and E513 leaving only the h-bond formed by Y540 and the residue Q344 of receptor DPP4.

Of the last 7 substitutions, 2 were conservative (N501S, K502R), 3 aa substitutions were non-conservative substitutions (D455T, P463E, R542G), G538A/S is a relatively conservative substitution while G538V was non-conservative.

Table 4 reports the multiple sequence alignment of the RBD region needed to bind the Aminopeptidase N (APN), a novel potential receptor for EriCoV (23). The Italian EriCoV sequences and non-Italian were aligned to the sequences of strains ErinCoV-12-19 and ErinCoV-Ger12 used to determine the structure of the binding region to hedgehog APN by Cryo-electron microscopy (cryo-EM). The alignment showed a certain heterogeneity at key binding positions.

Table 4. Multiple sequence alignment showing variations in key amino acid binding residues of the spike protein to the APN receptor.

Accession numbers: MW246800.

Accession numbers: KC545383

Accession numbers: MW246796, MW246798, MW246802, MW246800, MW246797, MW246801, MW246795, MW246799

Accession numbers: KC545386, NC_039207.1, MK679660, PP721362, PP721361, PQ869634, PQ869635, PX460032, PV359183, PV359184, PV359185, PP421220, PP421221, PV359182

Only three amino acid positions out of 10 are conserved (527, 529, 534). The remaining 7 positions showed the presence of non-synonymous substitutions leading to non-conservative substitutions, except the S531T and T536N residues.

3.4 Analyses of the CD200 ortholog sequences

By reconstruction of full genome sequences, 15 complete genomes showed the presence of CD200, and an additional 5 were retrieved from partial sequences obtained by RT-PCR. The length of the CD200 insertion varied between 462 and 687 bp, corresponding to 154-229 amino acids.

BLAST comparison against the NCBI database showed that the CD200 orthologs shared the highest amino acid identity (50.0% to 81.0%) with the host Erinaceus europaeus CD200 protein (also named OX-2 membrane glycoprotein) (XP_007516410,2).

The CD200 sequences from this study showed 79.5% to 99.4% nt.id., and 50.3%-88.7% amino acidic identities (aa.id.) among themselves. Only two strains were identical at both aminoacidic and nucleotide level (100%, EriCoV/Italy_BO61/2021, EriCoV/Italy_BO64/2021).

Comparing sequences with those reported online, and previously described in Italy, showed 76.8%-97.0% nt.id. and 51.3% to 99.0% aa.id. Notably, the strain ErinaceusCoV/Italy/50265-30/2018 was identical to ErinaceusCoV/Italy/50265-13/2019 (MW294203), both reported in Northern Italy.

The phylogenetic tree confirmed the clustering of the CD200 sequences from EriCoV genome with the genomic Erinaceus europaeus CD200 and no geographical clustering was observed (Supplementary Figure 2).

4 Discussion

The first report of EriCoV circulating in Italy date back to 2020 (16) from European hedgehogs sampled in Northern Italy. Two subsequent studies confirmed its circulation in Central and Northern Italy, and based on sequencing of short genomic regions, such as RdRp, partial phylogenetic analyses were conducted (15, 19). In 2021, a total of 8 complete genomes were obtained from Italian strains (17), which were insufficient to comprehensively assess the heterogeneity of circulating strains. In the present study, we expanded the available dataset by analysing EriCoV in 55 European hedgehog samples across Italy.

The complete genomic sequences were obtained from 35 animals. Despite the information on the complete genomes of European EriCoV strains is still limited, a comparative phylogenetic analysis was performed, using genomes reported from strains detected in Poland, Russia, France, Hungary and Germany and a partial genome from Portugal (8, 12–14, 18, 24).

The phylogenetic analysis of complete genomes showed the higher evolutionary correlation among the Italian sequences and the absence of a clear geographical clustering. In fact, sequences from EriCoVs detected in different Italian regions were interspersed, showing high nucleotide identities >93.0%. Despite the high distances between the areas where animals were collected, and since hedgehogs are not migratory animals with limited mobility and a restricted home range, the results suggest the circulation of similar strains within the country and that similar strains are fixed within the population. Conversely, non-Italian strains showed a clear clustering and were not interspersed among Italian sequences.

The phylogenetic tree based on the RdRp region showed the presence of non-Italian strains within the cluster formed by Italian strains. This can be explained by the use of a short genomic fragment, which results in lower phylogenetic resolution compared with analyses based on complete genomes. Despite the presence of one cluster formed by sequences from Central Italy (Lazio and Umbria regions), another sequence from the Centre (Umbria) and one from two from the South (Puglia) were closer to those from Northern Italy, confirming the absence of geographical clustering. Recently, another CoV species was reported in hedgehogs, but in Amur hedgehogs (Erinaceus amurensis) and only in China, suggesting to date a species specificity for EriCoV for European hedgehogs and the HKU31 for Amur hedgehogs. In turn, a study from Portugal reported the detection of an Alphacoronavirus in two European hedgehogs in Portugal. The sequences recovered were identical to M. schreibersii bats from the same area (14). However, since hedgehogs came from the same rehabilitation centre, the transmission from bats to hedgehogs cannot be ruled out.

In 2021, our group reported the presence of an additional ORF encoding a CD200 ortholog in some EriCoV strains from Northern Italy (17). The CD200 is a host immune-modulatory protein involved in the regulation of immune response, and similar host-derived immunomodulatory genes have been reported in other viruses, such as poxviruses and herpesviruses, where viral CD200 orthologs contribute to immune evasion and enhanced persistence (25, 26). Although host-gene capture is rarely documented in coronaviruses, recombination with host transcripts has been proposed in SARS-related CoVs (27). If functional, a CD200 ortholog could potentially influence immune evasion during intra−host circulation and, theoretically, contribute to host−jump dynamics, although this remains unproven for EriCoV.

In this study, the CD200 ortholog was reported in 19 coronaviruses detected in animals from Northern Italy (Lombardia, Veneto, Emilia-Romagna) and in one animal from the South (Puglia). Considering the geographical origin of the samples positive for the insertion, which are mainly from Northern, strains with the insertion may circulate frequently in the North. However, due to the lower number of samples collected in the Central and Southern area, we cannot exclude the presence of the same insertion also in more samples from this area. More studies and samples, in particular from regions never sampled before, are required to clarify the prevalence and circulation of strains with CD200 ortholog. No clinical sign of disease attributable to EriCoV infection was observed in the animals included in this study, regardless of whether the CD200 ortholog was present. Among EriCoV−positive hedgehogs, some survived and were released after rehabilitation, whereas others died due to trauma or poor initial condition; in none of these cases were signs or lesions suggestive of coronavirus−associated disease detected.

Analysis of the structural protein S, confirmed previously observation, showing lower aminoacidic identity within the RBD region, compared to the full-length protein when EriCoV strains were compared with MERS-CoV strains, while higher similarity among hedgehog-derived strain. The tree topology obtained using the S sequences was the same for the whole genome phylogeny, confirming the absence of geographical clustering.

Following the results of a recent Polish study (12), only 3 out of the 18 amino acids, known to be necessary for MERS-CoV binding to the human hDPP4 receptor were conserved in EriCoV strains. A similar amino acid sequence was also identified in German or British coronaviruses (8, 11) whereas the HKU4 strains from bat retain 7 amino acids conserved with respect to MERS-CoV and were able to use the hDPP4 (28).

The analysis also suggests that the EriCoV strains were evolving under negative selection pressure, but some sites of the S protein are positively selected. Among the 30 sites, 3 were located in RBD region. Previous studies on the adaptive evolution of the S of coronaviruses indicated that purifying selection is the dominant evolutionary force driving coronavirus evolution (29, 30). Similar results were previously obtained from the analysis of the European hedgehog by a Polish study. However, the variability found in this study is probably due to the high number of complete genome sequences produced. Despite this, positive selection positions were always found in RBD around the Fusion peptide region and the Heptad repeat 2. The changes in these positions increase cell to cell fusion activity and change in the virulence of the virion as seen in mutant studies of SARS-CoV and SARS-CoV-2 (31, 32). In turn, the Heptad repeat 2 region was previously thought to determine host expansions and therefore, facilitate virus cross-species transmission (33). In fact, the positive selection signal identified in RBD may reflect continuous attempts by coronaviruses to adapt to a different host species or tissue (34).

In addition, within the RBD region of strains from this study, the presence of three deletions and 6 positions showed a high variability, up to 5 different amino acids in the same position. Most of the changes reported were non-conservative positions, confirming that the EriCoV spike very likely does not bind the human receptor as reported by in vitro studies (23).

Another recent study by Jin, Jefferson (23), identified the APN as a novel potential receptor for EriCoV. Although EriCoV belongs to the genus Betacoronavirus its utilization of APN mirrors the entry mechanism of several alphacoronaviruses such as the Transmissible Gastroenteritis Coronavirus (TGEV) (35), the human coronavirus HCoV-229E43 (36), porcine respiratory coronavirus (PRCV) (37), the canine (CCoV) and feline (FIPV and FECV) coronaviruses (38), and the canine-feline recombinant coronavirus (CCoV-HUPn-2018) (39).

Jin, Jefferson (23) further reported that despite sharing a similar RBD structure for APN docking, different EriCoV strains exhibit varying binding affinity. The study revealed that within the 10 residue binding interface to the APN, only 3 positions are strictly conserved, and the other showed up to 4 different amino acids. This sequence variability, occurring within a conserved structural framework, confirms that subtle amino acid differences may lead to significant fluctuations in affinity to the receptor as previously reported observed in related viruses.

In conclusion, this study confirms the widespread circulation of EriCoV in Italian hedgehogs, with phylogenetically related strains detected across the country. The presence of a CD200 insertion in both Northern and Southern Italian strains indicates that the insertion is not fixed within the viral population and does not represent a conserved genomic trait. The uneven distribution of the insertion among phylogenetically related strains suggests that it may be a dynamic genomic feature, that is acquired or lost through genomic rearrangements, rather than being stably maintained. In the spike protein, 25 positions were under positive selection, 3 of which in RBD, which may allow the acquisition of favourable features for cross-species transmission. Further studies are necessary to uncover the diversity and abundance of viruses, since the number of species circulating in wildlife in unknown, in particular, uncharacterized strains with novel genomic features that can potentially transmit coronaviruses to other species.

5 Conclusion

In this study, complete genome sequences of EriCoV were characterized. The identified strains were closely related to European EriCoVs, and no geographical clustering was observed among the Italian sequences.

In addition, the presence of a CD200 ortholog was reported for the first time in a strain sampled in Southern Italy (Puglia), confirming that strains carrying this insertion can circulate in other Italian areas.

Analysis of the spike (S) protein revealed several sites under positive selection, three of which were located within the receptor-binding domain (RBD). High variability was also observed in residues involved in receptor binding, highlighting the potential acquisition of novel features that may facilitate cross-species transmission.

Despite the high variability of the analysed strains, further studies are needed to determine whether the CD200 insertion is restricted to specific Italian areas and to assess whether additional, yet uncharacterized, strains are circulating in other regions.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fviro.2026.1808060/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Humberto J. Debat, Instituto Nacional de Tecnología Agropecuaria, Argentina

Reviewed by: Kasia Domanska-Blicharz, National Veterinary Research Institute (NVRI), Poland

Patrick Berche, Université Paris Cité, France