Molecular characterization of four Helicobacter cetorum strains from dolphins compared to human Helicobacter pylori

ABSTRACT

Helicobacter species colonize the stomachs of many aquatic and terrestrial mammals, including Helicobacter pylori in humans and Helicobacter cetorum in dolphins. There are several H. cetorum genome sequences in databases, but a detailed molecular characterization of these bacteria is missing. Here, we compared four H. cetorum isolates from dolphins with H. pylori strains using electron microscopy as well as structural and functional analyses. All strains expressed similarly high urease activity and were hemolytic to erythrocytes. Western blots revealed conserved expression of flagellin-A, neutrophil-activating protein NapA, serine protease HtrA, γ-glutamyl-transpeptidase GGT and toxin VacA. In contrast, the virulence-associated cag pathogenicity island of H. pylori is missing in H. cetorum. 3D-modeling revealed similar structures of hexameric VacA from both species with minor differences. H. cetorum VacA expression was associated with vacuole formation in epithelial cells similar to that of s1/m2, but not as strong as H. pylori s1/m1 vacA strains, and complementation of H. pylori with H. cetorum vacA restored the s1/m2-like VacA phenotype. While H. pylori infection robustly activated toll-like receptors TLR1, TLR2, TLR4, TLR5, TLR9, and TLR10, H. cetorum only stimulated TLR1/2, TLR4, and TLR10, but much less pronounced than H. pylori. Accordingly, infection of epithelial cells with H. pylori induced strong DNA damage, NF-κB activation, and IL-8 secretion, but these responses were barely detectable in H. cetorum-infected cells. Activation of only few TLRs and significantly weaker pro-inflammatory responses than H. pylori suggest that H. cetorum is a commensal or only moderately virulent pathobiont in the stomach of dolphins, comparable to the less pathogenic cagPAI-negative H. pylori strains in humans. Since H. cetorum is evolutionarily older than H. pylori, we propose that H. cetorum represents a direct ancestor of H. pylori that arose after a host jump over 623,000 years ago, which is the coalescence time of the two species.

Introduction

Many Helicobacter species share a long co-evolution with their mammalian hosts. The well-characterized human stomach bacterium Helicobacter pylori has been associated with modern humans since their origin in Africa about 200,000 years ago.^1,2 This infection is usually acquired in childhood when the immune system is not yet fully developed and the acidity of the stomach is less pronounced than in adulthood. While transmission of the bacteria commonly occurs within families, extra-familial transmission between unrelated individuals has also been reported, particularly in the developing world.^3,4 During the long co-evolution, host and pathogen diversified, which resulted in the development of distinct biogeographic H. pylori populations in Africa (hpAfrica1, hpAfrica2, hpNEAfrica), Eurasia (hpEastAsia, hpAsia2, hpEurope), the Sahul (hpSahul), Siberia and Americas (hpNorthAsia).^5–8

Infection with H. pylori is widespread; about 50% of the human global population is estimated to be colonized by this bacterium.⁹ Colonization of the highly acidic stomach is facilitated by the release of urease and by flagella. Secreted urease buffers the immediate surrounding of the bacteria by hydrolysis of urea into ammonium and carbon dioxide.¹⁰ Flagella, that are encoded by a large set of flagellar genes, including the flagellin gene flaA, enable the bacteria to migrate through the mucus layer of the epithelium.¹¹ Adherence to the gastric mucosa is then mediated by bacterial adhesins such as the blood group antigen-binding proteins BabA and BabB,¹² the sialic acid-binding adherence protein SabA,¹³ and the HopQ adhesin that binds the human carcinoembryonic antigen-related cell adhesion molecule (CEACAM) receptors.^14,15 During the initial colonization, H. pylori causes acute gastric inflammation in all infected individuals.¹⁶ The inflammation then generally subsides in the majority of the infected people and the subsequent chronic infection remains asymptomatic. However, persistent infection and inflammation can lead to the development of gastric and duodenal ulcers in approximately 10–15% of the infected individuals and gastric adenocarcinoma or lymphoma of the mucosa-associated tissue (MALT) in about 1%, usually late in life.¹⁷ Conversely, H. pylori colonization was reported to be beneficial early in life by protecting the human host from autoimmune diseases such as asthma and inflammatory bowel disease.¹⁸

The development of malignant disease depends on several factors, including the genetic susceptibility, lifestyle and diet of the host, and the genotype of the pathogen, particularly of its virulence genes.¹⁹ Important H. pylori virulence factors include the cag pathogenicity island (cagPAI) and its effector protein CagA that are only present in the highly virulent type-I strains, but not in the less virulent type-II strains, the vacuolating cytotoxin VacA, serine protease HtrA, and a variety of outer membrane proteins that are involved in host cell binding.²⁰ The cagPAI encodes a type IV secretion system (T4SS) that injects the effector molecules CagA and ADP-heptose (ADPH) into gastric epithelial cells.^21,22 Once injected, ADPH activates pro-inflammatory responses by stimulating transcription factor NF-κB and the release of interleukin-8 (IL-8),²² and CagA interferes with various intracellular signaling pathways, thus inducing malignant cell changes.²³ T4SS assembly and CagA delivery occur at the basolateral side of the epithelial cells, which requires opening of the cell-to-cell junctions. Cleavage of junction proteins by H. pylori serine protease HtrA enables bacterial migration between the cells and subsequent T4SS functions.²⁴ A serine/leucine polymorphism (171S/L) affects the stability of proteolytically active HtrA trimers, and thus the disruption of cell junctions and subsequent CagA delivery. Strains with the 171 L-type HtrA, which are evolutionarily ancestral, cause stronger cleavage of the junction proteins and more frequent DNA damage in host chromosomes.^25–27 Another important virulence factor is the vacuolating cytotoxin VacA that induces the formation of large cellular vacuoles from lysosomes, alters various cellular signaling pathways, and interferes with immune cell functions.^28,29 In addition, VacA triggers membrane disruption in mitochondria and activates apoptosis of infected cells.³⁰ Interestingly, CagA and VacA display antagonistic effects on actin-cytoskeletal rearrangements and apoptosis of infected host cells.^31,32 Although the vacA gene is present in all H. pylori isolates, strains differ in VacA expression, secretion, and cytotoxic activity. VacA cytotoxicity is determined by allelic variations in heterogenic regions of the protein, most notably the signal peptide region (alleles s1, s2) that determines vacuole formation and the middle region (m1, m2) that is thought to be associated with cell binding.²⁸ Allele s1 facilitates the vacuole-forming activity by creating an anion channel across the lysosome membrane. In contrast, s2 VacA fails to form anion channels across the lipid bilayers because of a hydrophilic 12-amino-acid extension in the signal peptide region that prevents integration into the hydrophobic membrane. As a result, s1/m1 VacA and to a lesser degree s1/m2 VacA are cytotoxic to epithelial cells, but s2/m2 VacA does not display cytotoxicity.^29,33 The s2/m1 allele combination is extremely rare. The individual VacA alleles are associated with the presence of the cagPAI. Strains that contain the cagPAI (type I strains) usually contain s1/m1 or s1/m2 VacA variants, whereas cagPAI-negative (type II) strains contain the non-cytotoxic s2/m2 type VacA.²⁰

The closest known relative of H. pylori is H. acinonychis (Figure 1), which has been isolated from large felines living in captivity, but not from wild animals. H. acinonychis infection and erosions of the gastric epithelium, including chronic gastritis and ulcers, have been reported in captive cheetahs, cougars, lions and tigers from zoos and circuses, indicating cross-infections between large felines.^34–38 Eradication of H. acinonychis with antibiotic therapy resulted in disappearance of the gastric lesions, showing a causal connection of H. acinonychis infection and gastric disease.³⁹ However, H. acinonychis strains do not contain the cagPAI, and the vacA gene is highly degenerated by mutations.^34,40 Similar to H. pylori, H. acinonychis isolates express urease, flagellin, serine protease HtrA, neutrophil-activating protein NapA, lytic transglycosylase Slt, and γ-glutamyltranspeptidase GGT, but several outer membrane protein (OMP) coding genes are not expressed due to frameshift mutations, including oipA, babB, hopQ, hopN, hopU, and sabA.^38,40

Figure 1.

16SrRNA gene and housekeeping gene-based phylogeny and presence of the vacA gene in gastric Helicobacter species. A) Neighbor-joining tree of 16SrRNA gene sequences. B) Neighbor-joining tree of concatenated housekeeping gene sequences atpA, efp, mutY, ppa, trpC, and ureI from H. pylori and its closest relatives H. acinonychis and H. cetorum. Species names are in black, and H. pylori populations are in gray. C) presence and absence of the vacA gene. D) phylogeny of VacA protein sequences from four H. pylori and four H. cetorum strains. Hc - H. cetorum; Hp - H. pylori.

The next closest relatives of H. pylori are Helicobacter cetorum and Helicobacter delphinicola. All other Helicobacter species are genetically much more distinct (Figures 1(A,B)). H. cetorum was originally isolated from the main stomach of two wild, stranded Atlantic white-sided dolphins⁴¹ and later from the feces of wild Atlantic bottlenose dolphins⁴² and of captive animals, including a Pacific white-sided dolphin, an Atlantic bottlenose dolphin, and a beluga whale.^43,44 Endoscopic examination of infected animals, which were reported to show lethargy and periodic regurgitation, revealed gastric inflammation and ulcers in the esophagus and the forestomach.⁴³ Genome sequences of H. cetorum isolates from a dolphin and a beluga whale were 1.8 and 1.95 Mb in size, slightly larger than the genomes from H. pylori.⁴⁵ These genomes showed that H. cetorum lacks the cagPAI and revealed divergent sets of OMP genes, the presence of the nickel-cofactored urease operon known from H. pylori plus an iron-cofactored urease also found in H. felis and H. mustelae, and several metabolic genes distinct from those in H. pylori. Moreover, the genome of the dolphin isolate MIT 99–5656 contained three additional vacA paralogs inserted five genes upstream of vacA.⁴⁵ H. delphinicola was isolated from captive common bottlenose dolphins held at the Port of Nagoya Public Aquarium in Japan.⁴⁶ Some of the animals were lethargic and showed signs of anorexia and dyspepsia. Endoscopy of the forestomachs and examination of gastric fluid samples revealed inflammation, bleeding, and ulcers.⁴⁶

Infection with H. pylori is widespread among humans, approximately half of the global population carries this bacterium.⁴⁷ An initial screening among wild Atlantic bottlenose dolphins revealed the presence of H. cetorum in about 50% of the examined dolphin population,⁴² suggesting that the prevalence of H. cetorum among dolphins might be as high as H. pylori among humans. However, besides the initial H. cetorum species description after the original isolation of the bacteria and the genome analysis of two isolates, very little is known about H. cetorum. Here, we performed a molecular characterization of four different H. cetorum isolates in comparison to four H. pylori laboratory strains. Based on the observed mild pathology in captive dolphins infected with H. cetorum,^41,43,44 we were interested in comparing bacterial pathogenicity factors in H. cetorum and H. pylori and to evaluate their role in gastric disease-associated processes, including induction of inflammation, hemolysis, and damage of epithelial cells by vacuole formation and induction of DNA double-strand breaks (DSBs).

Results

Phylogenetic analysis

Gastric Helicobacter species colonize the stomach of their respective hosts.^48,49 To analyze the relationship of H. cetorum with H. pylori and other gastric Helicobacter species, we generated a Neighbor-joining from their 16SrRNA gene sequences using Campylobacter concisus as an outgroup (Figure 1(A)). In agreement with previous analyses,^2,43,45,49 the gastric Helicobacter species clustered in two groups. The first cluster was composed of H. salomonis, H. bizzozeronii, H. felis, H. heilmannii and H. suis, and the second cluster contained H. cetorum, H. delphinicola, H. pylori and H. acinonychis. A Neighbor-joining tree based on concatenated housekeeping gene sequences (Figure 1(B)) confirmed the close relation of H. cetorum, H. pylori and H. acinonychis. Phylogenetic analyses of those sequences showed that the genetic diversity of the housekeeping genes (Figure 1(B)) was significantly larger in H. cetorum (Π₉₅ = 4.94–9.19%) than in H. pylori (Π₉₅ = 3.10–3.20%), indicating that H. cetorum is evolutionarily much older than H. pylori. Of all helicobacters, only those three species contained the vacA gene (Figure 1(C)). While present in full length in H. cetorum and H. pylori, vacA is present as 13 gene fragments and is thus highly degenerated in H. acinonychis.^34,40 Figure 1(D) shows the vacA phylogeny of four H. cetorum isolates that were originally obtained from a Pacific white sided dolphin (MIT 01–5903), an Atlantic white sided dolphin (MIT 01–6096), and two Atlantic bottlenose dolphins (MIT 99–5656, MIT 01–6202), and from four selected human H. pylori strains (N6, G27, 60190 and P1).

H.cetorum electron microscopy analyses

All four H. cetorum isolates were analyzed by scanning electron microscopy (SEM) and transmission electron microscopy (TEM), yielding similar results. Representative pictures are shown from two strains, MIT 01–5903 and MIT 01–6096 (Figures 2, 3, S1 and S2). The SEM images showed typically curved, flagellated bacteria that are approximately 1.5–3 µm long and about 0.4 µm wide. The size and cell shape (curved or helical) varied slightly among the individual cells (Figure 2(A–C); Fig. S1A-B, Figure 3, blue arrows). Typical outer membrane vesicles (OMVs) are visible at the surface of many of these bacteria and in their immediate surroundings. These OMVs, which are initially formed at the bacterial surface and then secreted into the medium, are about 40–100 nm in diameter (Figure 2(C), red arrows), similar to OMVs from other Gram-negative bacteria. The bacteria that are usually cultured under microaerophilic conditions were very sensitive to oxidative stress. Exposure of H. cetorum to normal atmosphere for 10 min already resulted in the formation of coccoid bacteria (Fig. S2A-C, blue arrows). Similar coccoid forms of H. pylori are no longer cultivable and likely represent dead bacterial cells.⁵⁰ In addition, H. cetorum expresses flagella that were up to 3–5 µm long. In contrast to H. pylori that usually forms a bundle of several mono-polar flagella,¹¹ over 50% of the H. cetorum bacteria possessed a single polar flagellum (Figure 3(A,B) yellow arrows). About one-third of the bacterial cells had two polar flagella (Figures 3(C,D)), and the remaining cells had three (Figure 3(E)) or four flagella each (Figure 3(F)) or more. Thus, the number of flagella varied, even among individual bacteria from the same culture. Since the thickness of flagella in the SEM images seem to vary (e.g. in panels B and E), we additionally performed negative staining and TEM. The TEM images revealed the presence of 1–4 monopolar sheathed or non-sheathed flagella (Figure 3(G–F)), marked by red and black arrows, respectively. These sheaths had a diameter of about 35–50 nm compared to 20 nm of non-sheathed flagella.

Figure 2.

Scanning electron microscopy of H. cetorum. A) overview of multiple, curved H. cetorum strain MIT 01–5903 bacteria (blue arrows) with flagella (yellow arrows). B) enlarged view. C) image of a single bacterium (blue arrow) with a monopolar flagellum (yellow arrow) and multiple budding and released OMVs (red arrows). Representative pictures from two independent preparations are shown.

Figure 3.

Morphological investigation of H. cetorum by scanning (A-E) and transmission electron microscopy (G-I). Single H. cetorum strain MIT 01–5903 and MIT 01–6096 bacteria (blue arrows) were analyzed, exhibiting either one monopolar flagellum (yellow arrows, panels A and B), two monopolar flagella (yellow arrows, panels C and D), three monopolar flagella (yellow arrows, panel E) or four monopolar flagella (yellow arrows, panel F). Negative contrast and transmission electron microscopy of the same H. cetorum samples (panels G-I) also revealed 1–4 monopolar flagella per bacterium (black arrows), some of which were covered with a sheath (red arrows). Representative pictures are shown from two independent preparations each.

H.cetorum strains express and secrete a functional urease

Coomassie-stained gels of total protein extracts revealed similar protein profiles among the four analyzed H. pylori strains, and showed only minor differences between the strains (Figure 4(A)). The protein patterns of the four H. cetorum strains were distinct from those of H. pylori. While the patterns from the three H. cetorum strains MIT 99–5656, MIT 01–6096, and MIT 01–6202 were relatively similar to each other, strain MIT 01–5903 had a different protein profile. A Western blot of the same protein extracts using an antibody specific for the H. pylori urease B protein (α-Urease-B) revealed a specific band at approximately 60 kDa for all samples, which indicated similar expression levels of urease by both H. pylori and H. cetorum (Figure 4(B)). Moreover, these data showed loading of similar total protein amounts and that the urease proteins are conserved (Fig. S3) to be recognized by the same antibody. In order to test whether the urease is secreted and active by H. cetorum similar to H. pylori, bacteria were cultured on acidified GC agar plates supplemented with urea, which is the substrate of the urease enzyme (Figure 4(C)). The observed red color change indicated comparative levels of urease secretion and activity by both species. An isogenic H. pylori urease deletion mutant included as control failed to induce the color change, as expected. In agreement with previously published data,⁴⁵ the H. cetorum genomes of strains MIT 99–5656, MIT 01–6096 and MIT 01–6202 contained ureA and ureB genes coding for an iron-cofactored urease enzyme in addition to the 7-gene urease operon known from H. pylori (locus tags HP0067–HP0073 in H. pylori strain 26695) (Fig. S3). In strain MIT 01–5903, these extra urease genes were truncated and likely not functional.

Figure 4.

Protein profiles of H. pylori and H. cetorum and urease expression. A) Coomassie stain of total protein extracts of four H. pylori and four H. cetorum strains after separation by SDS-PAGE. B) Western blot with antibodies against the urease B subunit. C) positive urease test indicated by the red color during growth of four H. cetorum strains and H. pylori strain P1 on agar plates supplemented with urease. The isogenic H. pylori P1ΔureB deletion mutant served as a negative control.

H.cetorum is hemolytic against erythrocytes

Other virulence factors conserved in both H. pylori and H. cetorum were also analyzed for protein expression during infection (Figure 5). Western blots showed expression of the flagella protein flagellin A, neutrophil-activating protein NapA, serine protease HtrA, γ-glutamyl transpeptidase GGT, and vacuolating cytotoxin VacA. Alignments of these proteins are shown in Figs. S4-S8. Next, the strains were analyzed for their ability to trigger hemolysis of red blood cells embedded in agar plates as H. pylori is known to display high hemolytic activity for iron acquisition from the host.⁵¹ Both H. pylori and H. cetorum exhibited strong hemolytic activity toward erythrocytes during incubation on blood agar (Figure 6), indicating that H. cetorum has a similarly high hemolytic activity as H. pylori, which is consistent with the large number of iron acquisition genes in the genomes of both species.

Figure 5.

Western blots of SDS-PAGE separated cell extracts from H. pylori and H. cetorum strains. Antibodies against flagellin A, NapA, HtrA and GGT showed similar protein patterns between H. pylori and H. cetorum. Hybridization with α-VacA antibody revealed additional, smaller bands in H. cetorum VacA.

Figure 6.

Blood hemolysis. Hemolysis caused by indicated H. pylori and H. cetorum strains grown for 3 days on blood agar plates.

H.cetorum VacA induces cell vacuole formation similar to H.pylori s1/m2-type strains

Since the vacA gene is only present in H. pylori and H. cetorum, and is heavily degenerated in H. acinonychis, the evolutionary origin of vacA is unclear. Therefore, we compared the VacA cytotoxin from both species in greater detail. Analysis of VacA in Western blots revealed the same bands in both species (at about 98 kDa and 88 kDa), but also the presence of additional, smaller fragments (ca. 85 kDa and 80 kDa) in H. cetorum VacA compared to H. pylori VacA (Figure 5, bottom). This suggests that H. cetorum VacA may be further processed compared to the p95 and p88 fragments known from H. pylori. As next, all four H. cetorum strains were tested for their ability to induce the formation of vacuoles during infection of the human gastric epithelial cell line AGS (Figure 7). For comparison, four H. pylori strains were selected that differ in their respective VacA alleles: G27 with the highly cytotoxic allele combination s1/m1, strain P1 with the moderately cytotoxic alleles s1/m2, the recently isolated strain SBA-03 containing the unusual and very rare allele combination s2/m1, and the non-cytotoxic s2/m2 strain SBA-06.⁵² As expected, the s1/m1-type strain G27 triggered strong formation of vacuoles in the epithelial cells, and less pronounced host cell vacuolization was observed for the s1/m2-type strain P1. In contrast, cell vacuolization was nearly not observed in cells infected with s2/m1- or s2/m2-type H. pylori (Figure 7(A)). In addition to H. pylori, cell vacuolization was also induced by H. cetorum. By comparison, H. cetorum-triggered vacuole formation in host cells was less pronounced than vacuolization during infection with H. pylori s1/m1-type strains (Figure 7(B)). Visual inspection of the images (Figure 7(B)) as well as quantification of the number of vacuole-containing cells (Figure 7(C)) revealed that H. cetorum-induced vacuole formation was comparable to infection with H. pylori s1/m2-type strains. Therefore, we propose to designate VacA-triggered vacuolization by H. cetorum as s1/m2-like (Figure 7(B)).

Figure 7.

Vacuolization of epithelial cells during infection with H. pylori and H. cetorum. A) degree of vacuole formation during infection with VacA s1/m1 (G27), s1/m2 (P1), s2/m1 (SBA-03), and s2/m2 (SBA06). B) vacuole formation caused by H. cetorum was comparable to s1/m2 from H. pylori and termed s1/m2-like. Intracellular vacuoles are exemplarily marked with yellow arrows in two pictures (infection with G27 and MIT 01–5903). C) quantification of vacuolization.

Alignment of H. pylori and H. cetorum VacA sequences showed long protein stretches with high homology (Fig. S8). In addition, computer modeling predicted the VacA 3D-structures to be very similar (Figure 8(A)), with most amino acid stretches overlapping between H. pylori (red) and H. cetorum (blue) proteins. Just like the H. pylori toxin, the hexameric H. cetorum VacA was predicted to exhibit a topology in which the subunits are arranged around a central pore (Figure 8(B)). However, there were slight differences in the form of some H. cetorum VacA loops protruding from the H. cetorum VacA structure (Figure 8). Those protruding loops are likely accessible to proteases, which might explain the presence of at least two additional bands in the Western blots that are somewhat smaller in size than the p88 and p95 fragments observed from H. pylori VacA (Figure 5). To investigate H. cetorum VacA in more detail, the vacA gene from H. pylori strain Ca173 was complemented with vacA from H. cetorum strain MIT 01–6202. Western blots confirmed H. cetorum VacA re-expression in the H. pylori strain background (Figure 8(C)). Western blots of flagellin, CagA, and urease served as internal controls. Subsequently, the complemented mutant was assessed for formation of host cell vacuoles during infection of AGS cells. Phase contrast microscopy confirmed vacuole formation by the complemented mutant (Figure 8(D)). The images showed that vacuole forming was similar to that induced by H. cetorum wt bacteria (Figure 8(D,E)).

Figure 8.

VacA 3D-structure and function. A) computer model of H. cetorum VacA (blue) superimposed on the experimental 3D-structure of H. pylori VacA (red). Black arrows point to protruding loops of VacA from H. cetorum compared to H. pylori. B) model of the H. cetorum VacA hexamer with the individual subunits shown in different colors. C) complementation of vacA in H. pylori strain Ca173 with the vacA gene from H. cetorum shown by Western blot against VacA. Western blots against FlaA, CagA and UreB served as loading controls. D) vacuolization of host cells caused by H. pylori Ca173, an isogenic Ca173ΔvacA deletion mutant, the deletion mutant complemented with H. cetorum vacA, and the corresponding H. cetorum wt strain MIT01–6202. E) quantification of vacuole formation.

H.cetorum fails to induce NF-κB response and IL-8 secretion

H. pylori infection is known to trigger an inflammatory host response that involves activation of transcription factor NF-κB and subsequent secretion of inflammatory cytokines such as IL-8.¹⁹ Whether infection with H. cetorum induces a similar host inflammation response is unknown. Therefore, AGS gastric epithelial cells infected with either H. pylori or H. cetorum bacteria were assessed for activation of NF-κB using a Quanti-Blue reporter assay, and the amounts of secreted cytokines were determined by ELISA against IL-8 (Figure 9). Quantification of the data revealed strong activation of NF-κB (Figure 9(A)) and IL-8 expression (Figure 9(B)) by H. pylori, but only very weak, if any, activation of either of those inflammation markers by infection with H. cetorum.

Figure 9.

Pro-inflammatory host response during infection of epithelial cells by H. pylori and H. cetorum. A) activation of NF-κB relative to uninfected cells (mock) and (B) secretion of IL-8 into the supernatant. H. cetorum failed to induce a pronounced inflammatory response compared to H. pylori.

H.cetorum bacteria do not or only moderately activate toll-like receptors

In addition, H. pylori is known to be recognized by a wide range of toll-like receptors (TLR1, TLR2, TLR4, TLR5, TLR9, and TLR10)^53,54, but whether H. cetorum is also recognized by any of those TLRs is unknown. Therefore, the activation of TLRs was analyzed during exposure of Hek293 TLR reporter cells to H. cetorum in comparison to exposure to H. pylori (Figure 10). As expected, infection with cagPAI-positive H. pylori strains (N6, 60190) activated all analyzed TLRs, and increased the reporter activity relative to the uninfected and parental controls by approximately 22-fold (TLR1/2), 20-fold (TLR2), 4-fold (TLR4), 12-fold (TLR5), 11-fold (TLR9) and 6-fold (TLR10). The two cagPAI-negative H. pylori strains SBA-03 and SBA-06,⁵² which were included as controls, stimulated TLR1/2, TLR2, TLR4, and TLR10 similar to cagPAI-positive H. pylori. However, in agreement with previous data^55–57, cagPAI-deficient H. pylori cannot activate TLR5 and TLR9. In contrast, H. cetorum were not recognized by TLRs (TLR4, TLR5, TLR9) or activated TLRs only moderately, one-third to half as strong as H. pylori (TLR1/2, TLR2, TLR10). Together, those data show that exposure of epithelial cells to H. cetorum activates only a few TLRs and causes a significantly weaker inflammatory response than H. pylori.

Figure 10.

Recognition of H. pylori and H. cetorum by host cell TLRs. A-F) infection of HEK cells expressing the individual TLRs with H. pylori induced between 4-fold and 20-fold higher NF-κB reporter activity compared to the uninfected mock control. H. cetorum was not recognized by TLR4 (C), TLR5 (D) and TLR9 (E), and triggered only moderate activation of NF-κB through TLR1/2 (A), TLR2 (B) and TLR10 (F).

H.cetorum does not induce DNA double-strand breaks (DSBs) in host cell chromosomal DNA

Finally, various earlier reports have shown that H. pylori infection induces substantial DNA damage in the host chromosome in a cagPAI-dependent fashion.^58–61 This is triggered through the introduction of DSBs that arise during induced transcription in RNA/DNA-hybrid structures (R-loops) upon ADPH-stimulated NF-κB activation, followed by the onset of DNA repair processes.⁶¹ These DSBs result in genomic fragility and accumulation of DNA mutations, which enhance cancerogenic gene modifications in the host. Pulsed-field gel electrophoresis (PFGE) revealed strong fragmentation of host chromosomal DNA upon infection with cagPAI-positive H. pylori isolates, but not with cagPAI-negative H. pylori and H. cetorum strains (Figure 11).

Figure 11.

Induction of DNA double-strand breaks during infection with cagPAI-positive H. pylori, but not cagPAI-negative isolates or H. cetorum. (A) PFGE-analysis of host chromosome damage. (B) DNA fragmentation (bottom) was normalized against intact chromosomal DNA retained in the loading well (top), and against the uninfected control (lane 1) that was set to 100%. All data are represented as mean ± SEM. p < 0.001 (***) was considered statistically significant.

Discussion

Helicobacters are known to comprise enterohepatic species that inhabit the intestinal tract and/or the gall bladder, and gastric species that colonize the stomach of their respective hosts.² While there is a small possibility that a species more closely related to the human stomach bacterium H. pylori exists, the closest currently known relatives are H. acinonychis from large felines and H. cetorum and H. delphinicola from dolphins (Figure 1(A)). Genome comparisons showed that H. acinonychis arose in Southern Africa by a host jump of H. pylori of the biogeographic population hpAfrica2 from humans to large cats likely when an H. pylori-infected human was eaten by a large feline. This jump between hosts, which happened approximately 100,000 years ago, indicates that H. acinonychis directly descendent from H. pylori.⁴⁰ Colonization of lions, cheetahs, cougars, and tigers with H. acinonychis indicated that cross-infection occurred later among other big cats that were kept together in captivity.^35–38 An analysis of pathogenicity-related factors revealed H. acinonychis expression of flagellin, urease, serine protease HtrA, neutrophil activating protein NapA, γ-glutamyl transpeptidase, and lytic transglycosylase Slt.³⁸ In contrast, the cagPAI, the cagA gene and the babA gene were absent from the genome, and other H. pylori virulence factors such as VacA and OipA were not expressed because of several frameshift mutations.^34,40

The next closest known relatives of H. pylori are H. cetorum and H. delphinicola (Figure 1(A)). H. cetorum has been isolated from several species of wild and captive dolphins and from a beluga whale held in the Mystic Aquarium in Connecticut, USA.^41–44 Whether wild whales also carry H. cetorum is questionable because H. cetorum infection of the beluga whale had likely occurred in captivity. This particular animal was reported to have shared an enclosed basin with a bottlenose dolphin in which gastric ulcers were detected during a health checkup.⁴⁴ Thus, similar to H. acinonychis among large felines in zoos and circuses, cross-infection of H. cetorum appears to occur among captive cetaceans. H. delphinicola was first identified in bottlenose dolphins at the Nagoya Public Aquarium, Japan, and subsequently found among dolphins in several animal facilities in Japan.⁴⁶ Apparently, simultaneous infection with both H. delphinicola and H. cetorum seems frequent in these facilities, as 79 of the 82 dolphins tested were positive for H. cetorum and 45 were positive for H. delphinicola.⁵⁷ However, we propose that H. delphinicola might simply represent a different phylogenetic clade of H. cetorum, similar to the hpAfrica2 population of H. pylori that is distinct from other H. pylori groups (Figure 1(A)).^1,2 In agreement, a BLAST nucleotide search of the published H. delphinicola 16SrRNA gene sequence yielded hits for both H. delphinicola and H. cetorum with 100% sequence identity, with H. cetorum isolated from a common dolphin (Delphinus delphis) at the southwest coast of England (Genbank accession FN565165). Unfortunately, there is currently no publically available genome for H. delphinicola that would allow a detailed analysis. Future studies should clarify this question.

Histological examination of stomach mucosa samples from stranded, H. cetorum-infected dolphins showed gastric lesions, including signs of mild multifocal gastritis.⁴¹ Likewise, captive animals with clinical signs such as vomiting and inappetence were shown to have ulcers in their forestomachs,^44,46 and histology images of the main stomach revealed mild mucosal erosion with mononuclear cell infiltrates in the lamina propria.⁴³ Immediately adjacent to those areas of gastritis, including pseudolymphoid follicles in the main glandular portion of the stomach, spiral bacteria were found on the gastric epithelium. Given that such lymphoid follicles are often seen in glandular stomach of Helicobacter infection in mice, this suggested a causal relationship between H. cetorum and the observed pathological changes.⁴³ Besides the initial species description, including the relatively mild pathology, and the description of two H. cetorum genomes isolated from a dolphin and the beluga whale, very little was known about H. cetorum. Therefore, we analyzed four H. cetorum isolates in comparison to H. pylori strains. In agreement with previous data,⁴³ electron microscopy images of H. cetorum showed flagellated curved bacteria that were approximately 1.5–3 by 0.4 µm in size. Our scanning electron microscopy images taken from over 600 cells from each of the four analyzed strains showed bacteria with one, two, three, and even four flagella protruding from one end of the bacterial cell (Figures 2, 3). Thus, all bacteria had mono-polar flagella, which is in contrast to a previous transmission electron micrograph of H. cetorum strain MIT99–5656 that showed bipolar flagella with a single flagellum at each cell end.⁴³ Just like other Helicobacter bacteria, H. cetorum was shown to apply a run-reverse-reorient mechanism for its motility in liquid and semi-solid media, and also in gastric mucin.⁶² Moreover, our results revealed that at least part of the H. cetorum flagella was sheathed, similar to flagella reported from H. pylori⁶³ and H. acinonychis.³⁸ Indeed, the gene encoding the flagella sheath protein HpaA was identified in the genomes of all three species.

Similar to H. pylori and H. acinonychis, H. cetorum expresses and secretes urease to buffer the acidic surrounding in the stomach (Figure 4). In this regard, urease expression appears to be a common trait as numerous gastric Helicobacter species secrete nickel-cofactored urease to buffer the acidic pH.^64,65 A tblastn search against the NCBI RefSeq Genome Database revealed that, in addition to H. pylori, H. acinonychis and H. cetorum, those include H. bizzozeronii, H. salomonis, H. felis, H. heilmannii, H. vulpis, H. labacensis, H. mehlei and H. suis¸ and even several enterohepatic species such as H. mustelae and H. hepaticus. In addition, H. cetorum possesses the ureA/ureB-encoded iron-cofactored urease also present in H. mustelae, H. felis, H. salomonis, H. cynogastricus, H. baculiformis and H. acinonychis. Recently, both urease gene clusters were also found in a set of H. pylori isolates from indigenous Siberians and native Americans.^8,66

Functional VacA is only expressed by H. pylori and H. cetorum. In addition, H. acinonychis possesses a highly degenerated gene with multiple frameshifts.^40,45 Since VacA is not present in other species, including any other gastric and enterohepatic helicobacters, the origin of VacA is an enigma. Therefore, we looked at VacA in more detail and compared the predicted structure of H. cetorum VacA to the published cryo-EM structure of H. pylori VacA.⁶⁷ The VacA monomers as well as the hexameric pore-forming toxins were very similar (Figure 8(A,B)), but several H. cetorum VacA loops stuck out from the protein overlay. We hypothesize that those protruding loops can be accessed by proteases because Western blots showed two smaller bands in the H. cetorum protein extracts compared to H. pylori (Figures 5, 8(C)). Given that these smaller extra bands were present in all four analyzed H. cetorum samples, this suggests additional cleavage sites in VacA rather than post-translational processing of VacA encoded by the three additional vacA paralogs found in the genome of strain MIT 99–5656, because those paralogous gene copies are only present in strain MIT 99–5656, but not in any of the other three H. cetorum genomes. Interestingly, two of those paralogs (vacA2 and vacA3) were also found in several of the above mentioned H. pylori strains isolated from indigenous Siberians and native Americans.⁶⁶ In addition, epithelial cells infected with H. cetorum exhibited vacuolization that was comparable to cells infected with H. pylori containing s1/m2-type VacA (Figures 7, 8(D)). Thus, this s1/m2-like VacA from H. cetorum triggered a relatively mild vacuolization. In addition, both species were hemolytic against erythrocytes, as indicated by halos surrounding the bacteria on blood agar plates (Figure 6). While H. pylori was known to sequester iron from host cells for its metabolism,^68,69 which can lead to iron deficiency anemia diseases in children and adults,^70,71 it is unclear whether iron acquisition by H. cetorum similarly affects the health of dolphins. Taken together, the two species are not only evolutionarily closely related, but also display a similarly high urease (Figure 4), hemolytic (Figure 6), and VacA activity (Figures 5, 7, 8).

In contrast to H. pylori, incubation of epithelial cells with H. cetorum induced hardly any inflammatory response in cultured gastric epithelial cells (Figure 9), which is probably associated with the absence of the cagPAI-encoded T4SS and cagA oncogene in H. cetorum,⁴⁵ as well as with the inferior recognition by TLRs in reporter cells (Figure 10). TLR4, TLR5, and TLR9 failed to detect H. cetorum altogether, and recognition by TLR1, TLR2 and TLR10 was poor. While stimulation of a TLR2-based reporter system was about 10-fold compared to the baseline, it was still only half of that induced by H. pylori (Figure 10). Likewise, cancer-associated DSBs in host chromosomal DNA were strongly pronounced during infection with cagPAI-positive H. pylori strains, but not cagPAI-negative H. pylori and H. cetorum isolates (Figure 11). Taken together, incubation of epithelial cells with H. cetorum induced a very mild overall reaction. This leads to the hypothesis that H. cetorum represents a commensal or only moderately pathogenic bacterium in the stomach of dolphins that is comparable to the significantly less pathogenic cagPAI-negative (also known as type II) H. pylori strains in humans. However, we note that all infection experiments were performed with the human gastric epithelial cell line AGS. Human cells may differ from dolphin epithelial cells in terms of bacterial adherence and overall pathogen-host interaction, which may have contributed to the observed low pathogenicity, particularly given the reported ulcers and gastritis in dolphins.^41,43,46

Given that H. cetorum is evolutionarily much older than H. pylori, we propose that H. pylori may have originated by a host jump of an H. cetorum-like ancestor from dolphins to early humans. Hypothetically, early humans ate stranded H. cetorum-infected dolphins, and thus contracted the bacteria. Subsequently, the bacteria adapted to the new host and eventually evolved into H. pylori, which was followed by spread of the bacteria in the human population. This scenario would be similar to the origin of H. acinonychis that supposedly arose from a host jump of H. pylori from early humans to large cats after H. pylori-infected humans were eaten by lions.^40,72 When did the proposed host jump of H. cetorum from dolphins to humans happen? Based on mitochondrial (mt) DNA analyses human matrilineal diversity coalesces to a “mitochondrial Eve” approximately 200,000 years ago when mtDNA haplogroups L1-L6 split from haplogroup L0 that is associated with the former hunter-gatherer people in southern Africa known as the Khoi and San (Khoisan).^73,74 At the same time, the H. pylori phylogeny split into two major lineages: 1) the population hpAfrica2 from southern Africa, which is characteristic for the Khoisan, and 2) all other biogeographic H. pylori populations in Africa and elsewhere in the world.² On average, housekeeping gene sequences from H. pylori of the hpAfrica2 population show a genetic distance of 0.049 to H. pylori from any other population (Figure 1(B)). The average genetic distance of H. pylori to H. cetorum is 0.1527, and thus roughly three times as large (Figure 1(B)). Given the 200,000-year coalescence estimate for H. pylori, this results in a time of the most recent common ancestor (TMRCA), and thus, split of H. cetorum and H. pylori approximately 623,000 years ago. This coalescence date is in agreement with a previous study,⁴⁹ in which the divergence between H. pylori and H. cetorum was calculated to be 610,000 years ago.

H. pylori contains the cagPAI, but H. cetorum does not, indicating that H. pylori must have acquired it after the divergence from H. cetorum. In addition, the average GC content of the cagPAI genes (35%) is lower than the GC content of the rest of the H. pylori genome (39%), which suggests lateral (from another bacterium) or horizontal (from archaea or eukaryotes) gene transfer into H. pylori.⁷⁵ However, the evolutionary source of the cagPAI import is still unknown. The cagPAI is not present in all biogeographic H. pylori populations.^20,76,77 In Africa, the cagPAI is entirely missing in the hpAfrica2 population, but is present in all hpAfrica1 strains, and is variably present in strains of the hpNEAfrica population (Figure 1(B)). Outside of Africa, the cagPAI is (variably) present in all H. pylori populations. This indicates that the cagPAI was imported in Africa after the divergence of the hpAfrica2 population from all other H. pylori that was estimated to 200,000 years ago.² But the acquisition of cagPAI happened before the Out-of-Africa migration that was estimated to having occurred approximately 60,000 years ago,^1,78 and thus between 200,000 and 60,000 years ago.

Taken together, in this study we have performed a detailed molecular characterization of four different H. cetorum strains from dolphins compared to H. pylori laboratory strains. Since H. cetorum is evolutionarily older than H. pylori, we propose that H. cetorum represents a direct ancestor of H. pylori that arose after a host jump over 623,000 years ago, which is the time of the coalescence of the two species. Remarkably, both H. pylori and H. cetorum express a highly active urease and trigger hemolysis, which is in agreement with the assumption that both bacteria must counteract the acid environment in the stomach and must acquire iron for their metabolism. However, H. cetorum induced only weak VacA-dependent vacuole formation and no visible DNA damage compared to highly virulent H. pylori. In addition, H. cetorum activated only a few TLRs and caused significantly weaker pro-inflammatory responses than H. pylori, suggesting that H. cetorum is a commensal or an only moderately pathogenic bacterium in the stomach of dolphins, comparable to the less pathogenic cagPAI-negative H. pylori strains in humans. These new studies are in well agreement with the mild pathology observed in captive dolphins infected with H. cetorum as discussed above.^41,43,46 Future studies should inaugurate the genetics in H. cetorum for in depth analyses of the virulence factors, combined with establishing a suitable animal model system for in vivo infection studies.

Materials and methods

Bacterial culture

H. pylori laboratory wt strains N6, G27, 60190, P1, and Ca173 (cagPAI-positive), and the recently isolated SBA-03 and SBA-06 (cagPAI-negative) were described.^{15,21,52,55,56} The urease mutant was generated as reported previously.⁷⁹ All H. pylori variants were cultured on GC agar plates supplemented with 10% horse serum, 1% vitamin mix and antibiotics and incubated at 37°C for 7 days in a microaerophilic atmosphere generated by CampyGen gas-generation sachets (Oxoid) in anaerobic jars.⁸⁰ H. cetorum strains MIT99–5656, MIT01–5903, MIT01–6202⁴³ and MIT01–6096⁴² were grown on BHI blood agar plates under microaerophilic conditions as described.⁴⁵ For the urease test, agar culture plates containing 600 μg/mL urea and 100 μg/mL phenol red were acidified to pH 5 using 1 M HCl.³⁸ For the analysis of hemolytic activity, the bacteria were grown on Colombia blood agar plates (Fisher Scientific, Cat# 10463833). Hemolysis, i.e. the formation of a pink zone around the culture, which indicates bacteria-induced hemolysis of erythrocytes in the culture medium, was assessed by visual inspection. Chromosomal DNA was isolated using the Gene Jet DNA Purification Kit (Invitrogen) following the protocol of the supplier.

Genetic analyses

16SrRNA and housekeeping genes were downloaded from the Bacterial Isolate Genome Sequence database (BIGSdb) hosted at the pubmlst.org website,⁸¹ or were extracted from genome accessions (JAHGXX000000000, JAXTOX000000000, VAPN00000000, CP001173, NC_017735, FZMW00000000, FZMU00000000, FZMR00000000). For validation, a 1.2 kb DNA fragment of the 16SrRNA gene was PCR amplified and sequenced from the four H. cetorum strains as described.³⁸ The 16SrRNA sequences were identical to the sequences from the published genomes. The Neighbor-joining trees were generated in MEGA.⁸² Nucleotide diversity (Π) within housekeeping genes from H. pylori and H. cetorum were estimated in DnaSP,⁸³ 95% confidence limits (Π₉₅) were estimated using an online confidence limit calculator (https://www.statskingdom.com/confidence-interval-calculator.html). For the dating analyses, the four-gamete criterion implemented in DnaSP⁸³ was used to identify stretches of DNA sequences that were likely to be the result of recombination. These recombinant nucleotides were removed from the dataset before the genetic distances between the groups were calculated in MEGA.⁸²

SDS-PAGE and Western blot

Bacteria were harvested and subsequently lysed by incubation in hot SDS buffer for 10 min. Protein extracts were separated on 8–12% SDS-PAGE gels and either Coomassie stained or blotted by Semi-dry blotting. Blotted PVDF membranes (Carl Roth, Karlsruhe, Germany) were analyzed using the following primary antibodies: rabbit α-Urease B (#21,987, BioGenes GmbH, Berlin, Germany),³⁸ rabbit α-HtrA (#26823, BioGenes GmbH),⁸⁴ rabbit α-VacA (#8036 BioGenes GmbH), rabbit α-GGT, rabbit α-FlaA, and rabbit α-NapA.³⁸ Goat α-rabbit antibodies conjugated to horseradish peroxidase (Thermo Fisher Scientific) was used as secondary antibody.⁵² Chemiluminescence was detected using 1.41 mM luminol in 0.1 M Tris – HCL (pH 8.6) mixed with 0.61 mM p-coumaric acid in DMSO and 0.02% hydrogen peroxide. Bands were visualized using a ChemiDoc XRS + Gel Imaging System (Bio-Rad, Hercules, CA, USA).⁵²

Cultivation and infection of AGS cells and phase contrast microscopy

The epithelial cell line AGS (human gastric adenocarcinoma cells, ATCC #CRL-1739™) was cultured at 37°C in RPMI1640 medium containing 10% fetal bovine serum (Thermo Fisher Scientific) and antibiotics (1% penicillin/streptomycin and 0.2% normocin) in a humidified atmosphere supplemented with 5% CO₂ as described.⁵² Before infection, cells were washed twice with phosphate-buffered saline (PBS, Sigma-Aldrich) and resuspended in fresh medium without antibiotics. The cells were then seeded into 6-well plates, grown to 70–80% confluency, and infected with H. pylori or H. cetorum at a multiplicity of infection (MOI) of 25. After 12 h, the samples were harvested and processed or analyzed by phase contrast microscopy with 10x objective to assess cell vacuolization. The frequency of cells containing vacuoles were expressed as mean values ± SEM after quantification of at least 100 cells per infection.

NF-κB, IL-8 ELISA, activation of TLRs and PFGE

AGS cells cultured to confluency were transfected with 5 µg of the pNF-κB-SEAP reporter plasmid (http://www.addgene.org.) for 48 h using the TurboFect transfection reagent (Thermo Fisher Scientific), following the protocol of the supplier. NF-κB-dependent production of the secreted embryonic alkaline phosphatase (SEAP) was quantified before and after infection by incubation of 20 μL cell culture supernatant with 180 μL Quanti-Blue reagent (InvivoGen) for 30 min at 37°C. SEAP levels were estimated at OD₆₂₀ in an Infinite F200 Pro microplate reader (Tecan, Grödig, Austria). IL-8 secretion from the infected cells was estimated by analyzing 20 µL of the supernatant in the IL-8 Human Uncoated ELISA Kit (Invitrogen, #88–8086). To monitor the activation of various TLRs (TLR1/2, TLR2, TLR4, TLR5, TLR9 and TLR10) by H. pylori, we used the corresponding HEK293 TLR/NF-κB/SEAP reporter cells (Invitrogen) in comparison to the HEK-Blue-Null1 cells as control. These 8 cell lines were infected for 24 h with either H. pylori or H. cetorum bacteria at a MOI of 25. The amount of NF-κB-dependent SEAP production was estimated as above by incubation of 180 µL Quanti-Blue reagent with 20 µL cell supernatant and subsequent measurement of the at OD₆₂₀ in an Infinite F200 Pro microplate reader. To study the induction of chromosomal DSBs, AGS cells were infected for 8 h with H. pylori or H. cetorum strains at an MOI of 25. Cells were harvested, and total DNA was prepared and analyzed by PFGE as described.⁵⁸

Electron microscopy

For scanning electron microscopy (SEM), cultured H. cetorum bacteria were directly fixed in the growth medium by the addition of formaldehyde (final conc. 5%) and glutaraldehyde (final conc. 2%). Samples were centrifuged, washed twice with TE buffer (10 mM TRIS, 2 mM EDTA, pH 6.9), and dehydrated with stepwise increasing acetone concentrations (10, 30, 50, 70, 90%) on ice for 10 min each, followed by two incubation steps with 100% acetone for 10 min each. Samples were further processed via critical point drying with a CPD300 (Leica Microsystems), and sputter coated with gold/palladium using a SCD500 (Bal-Tec). The bacteria were examined in a field emission scanning electron microscope Merlin (Zeiss) using an acceleration voltage of 5 kV and both Inlens and an Everhart – Thornley SE detector. For negative staining, sample drops were pipetted on formvar coated grids and incubated for 45s. After soaking the liquid, the grids were washed twice in dH₂0 followed by incubation on a droplet with uranyl-acetate for 30 sec and heat-drying on a bulb after removal of excessive liquid. The grids were examined in a transmission electron microscope Libra120 (Zeiss) at calibrated magnifications.

Molecular modelling

The structure of hexameric H. cetorum VacA (MIT 99–5656) was predicted using AlphaFold-3,⁸⁵ which resulted in a reliable model for residues I84-H896, with the exception of the amino acid residues D371-S408. The latter sequence stretch was also not resolved in the experimental structure of H. pylori VacA (PDB:6NYF)⁶⁷ and was therefore excluded from further structural analysis. Structure comparison was done with PDBeFold,⁸⁶ and RasMol⁸⁷ was used for structure analysis and visualization.

Statistics

All data were generated in at least three independent experiments and were evaluated by a two-tailed unpaired Student’s t-test and two-way ANOVA using the statistical analyses package implemented in Graph Pad Prism 8.0 (GraphPad Software, San Diego, CA, USA). Data are shown as mean values, and error bars indicate the standard error of the mean (SEM). The p-values of p ≤ 0.05 (*), p ≤ 0.01 (**) and p ≤ 0.001 (***) are considered to designate statistically significant results.

Funding Statement

The work of Steffen Backert has been funded by a Deutsche Gesellschaft für Zahn-, Mund- und Kieferheilkunde grant project number [953000-016] and German Science Fondation (DFG) grants [BA 1671/15-1] and [BA 1671/16-1]. Nicole Tegtmeyer is supported by DFG grant TE776/3-1 and the work of Bodo Linz is supported by the Wilhelm Sander-Stiftung grant [R2024.120.1].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All sequence sources are listed in the Materials and Methods section and are publicly available (https://www.ncbi.nlm.nih.gov/datasets/genome/?taxon=209), sequence alignments are provided in the supplementary material. The authors confirm that all data supporting the findings of this study are shown in the figures and are described in the text.

Supplementary Information

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19490976.2025.2557982