Evaluating cpn60 for high

ISME Communications volume 3, Article number: 69 (2023) Cite this article

257 Accesses

3 Altmetric

Metrics details

Despite being the most widely used phylogenetic marker for amplicon-based profiling of microbial communities, limited phylogenetic resolution of the 16S rRNA gene limits its use for studies of host-microbe co-evolution. In contrast, the cpn60 gene is a universal phylogenetic marker with greater sequence variation capable of species-level resolution. This research compared mammalian skin microbial profiles generated from cpn60 and 16S rRNA gene sequencing approaches, testing for patterns of phylosymbiosis that suggest co-evolutionary host-microbe associations. An ~560 bp fragment of the cpn60 gene was amplified with universal primers and subjected to high-throughput sequencing. Taxonomic classification of cpn60 sequences was completed using a naïve-Bayesian QIIME2 classifier created for this project, trained with an NCBI-supplemented curated cpn60 database (cpnDB_nr). The cpn60 dataset was then compared to published 16S rRNA gene amplicon data. Beta diversity comparisons of microbial community profiles generated with cpn60 and 16S rRNA gene amplicons were not significantly different, based on Procrustes analysis of Bray-Curtis and UniFrac distances. Despite similar relationships among skin microbial profiles, improved phylogenetic resolution provided by the cpn60 gene sequencing permitted observations of phylosymbiosis between microbial community profiles and their mammalian hosts that were not previously observed with 16S rRNA gene profiles. Subsequent investigation of Staphylococcaceae taxa using the cpn60 gene showed increased phylogenetic resolution compared the 16S rRNA gene profiles, revealing potential co-evolutionary host-microbe associations. Overall, our results demonstrate that 16S rRNA and cpn60 marker genes generate comparable microbial community composition patterns while cpn60 better facilitates analyses, such as phylosymbiosis, that require increased phylogenetic resolution.

Mammalian skin microbial communities have a direct influence on host health and disease and share a longstanding evolutionary history with their respective hosts. Initial predatory and nutritional interactions between ancestors of modern bacteria and eukaryotes are hypothesized to have led to multicellularity [1], further developing to include complex metabolic symbioses [2] and vertebrate innate immune responses [3, 4]. Given variations in physiology [5], hair and fur coverage [5, 6], geographic origin and habitat features [7], and evolutionary history and relatedness [8], the mammalian skin environment promotes instances of host-specific microbial co-evolution. Because of niche heterogeneity conferred by mammalian hosts, skin microbial community assemblage is thought to be deterministic (i.e., influenced by specific environmental or host factors) rather than stochastic (i.e., random assemblage and birth-death events) [9]. For specific mammalian orders, microbial evidence demonstrates that host phylogeny correlates with microbial community composition, manifested as “phylosymbiosis” [8].

Phylosymbiosis is a pattern wherein the microbial community composition of a host reflects the host’s environmental and phylogenetic history [10,11,12], with more distantly related host species showing greater differences in microbial community composition compared to those that are more closely related [8]. Initial microbial community assemblage from either stochastic or deterministic processes could foster close interactions between the host and associated microbiota over time, potentially leading to co-evolutionary relationships and increased associations [12]. Using the 16S rRNA gene phylogenetic marker, Ross et al. showed first evidence of phylosymbiosis within the Perissodactyla and Artiodactyla orders, which constitute odd-toed and even-toed ungulates, respectively, [8]. Their study also identified a core microbiome common to all sampled orders, represented by soil-associated bacteria, such as Agrobacterium and Arthrobacter, and taxa from the common skin bacterial genus Staphylococcus [8, 13, 14], among others. Within Primates, a core axillary microbiome containing Staphylococcus was identified as a dominant contributor to microbial beta diversity [5].

Microbiome studies, such as those mentioned above, most often use short-read amplicon 16S rRNA gene sequencing because of the costs associated with full-length 16S rRNA gene sequencing and metagenomics. However, limited phylogenetic “resolution” of the 16S rRNA gene, when used for short-read sequencing, results in sequence classification mostly to the genus-level [15], preventing detailed investigations of microbial species and specific populations, such as Staphylococcaceae. This limitation to phylogenetic resolution may also mask observations of phylosymbiosis perhaps only present at lower taxonomic levels. Investigating the mammalian skin microbiome using alternative gene markers with higher phylogenetic resolution, such as cpn60 [16], could reveal key microbe-host associations that may indicate existing or potential future co-evolutionary relationships within the class Mammalia, with implications for mammalian skin health and disease.

Several features of the cpn60 gene, alternatively known as hsp60 and groEL, make it useful for short-read amplicon microbial community profiling studies. The cpn60 gene is universal, encoding for a 60 kDa GroEL protein belonging to a family of type I chaperonins present in bacteria and chloroplasts, with equivalent type II chaperonins found in archaea and eukaryotes. Like the 16S rRNA gene, the presence of cpn60, and function of GroEL, is essential for cell survival and thus present in all prokaryotic life [17]. However, unlike the 16S rRNA gene, only a single copy of cpn60 is present in most prokaryotic genomes [16]. Additionally, the cpn60 gene contains a much greater sequence diversity compared to the 16S rRNA gene variable regions used in short-read amplicon studies, with a higher diversity in nucleotide identity and larger differences in sequence similarity between species [16]. Furthermore, the primers used to amplify the 274–828 “universal target” region of the cpn60 gene [16] produces an amplicon of ~554 bp in length. Although this length exceeds current short-read high-throughput sequencing technology, only 150–250 bases of the forward read is necessary to produce sufficient data for phylogenetic classification to the species level [18], allowing it to be used on current sequencing platforms (e.g., MiSeq, Illumina). Additionally, recent validation of naïve-Bayesian classification using the RDP classifier demonstrated a high degree of species-level classification using cpn60 reference sequences [19]. Because of intrinsic features of the cpn60 gene, the continued development of primers [20, 21] and reference databases [22], prior in silico and in situ validation [16, 18, 19], and history of use in both universal and targeted microbial profiling studies [23,24,25,26,27,28,29], the cpn60 gene is a powerful complement to the 16S rRNA gene as a universal phylogenetic marker within the context of microbial community profiling and phylosymbiosis/co-evolution research of the skin microbiome.

By leveraging the increased phylogenetic resolution provided by the cpn60 gene, in combination with amplicon sequence variants (ASVs) capable of identifying individual amplicon sequences [30], skin microbiota can be more thoroughly profiled and evaluated for patterns such as phylosymbiosis and co-evolution. As well, core microbiota associated with the skin, such as Staphylococcaceae, can be profiled at increased phylogenetic resolution similar to multi- or single-locus sequence typing approaches [31], balancing high-specificity and universal detection of bacteria. Although previous studies have used the cpn60 gene to profile the human vaginal [23,24,25] and pig feces microbiomes [32], it has yet been applied to profile the broader mammalian skin microbiome. This study uses comparative microbial profiling of paired cpn60 and 16S rRNA genes datasets to provide additional support for using cpn60 to profile skin microbiota, confirm previous reports of phylosymbiosis on mammalian skin [8], and present novel insight into specific microbial populations of mammalian skin and their potential interactions with their respective hosts. This study further demonstrates how the cpn60 database, cpnDB [22], can be integrated into the full microbial sequence data QIIME2 pipeline for producing complete cpn60 gene amplicon datasets.

Genomic DNA from mammalian skin swabs was extracted as part of a previous 16S rRNA gene survey generated from amplification of the V3-V4 region with the universal prokaryotic Pro341F/Pro805R primers [8]. From these, a representative 95-sample subset was chosen for cpn60-based sequencing and microbial profiling. Both PCR amplification and high-throughput sequencing of samples were completed at the University of Saskatchewan as previously described [20, 21]. The cpn60 gene was amplified by PCR using a primer mix comprised of 100 µM M279 (5′ – GAIIIIGCIGGIGAYGGIACIACIAC – 3′), M280 (5′ – YKIYKITCICCRAAICCIGGIGC– 3′), M1612 (5′ – GAIIIIGCIGGYGACGGYACSACSAC– 3′), and M1613 (5′ – CGRCGRTCRCCGAAGCCSGGIGCCTT– 3′). All primers contained Illumina adapter sequences on the 5′ end. Primers were combined in a 1:3 molar ratio of M279 and M280 (3 µL each) and M1612 and M1613 (9 µL each), then diluted in Ultrapure water to a total volume of 300 µL. All PCR tubes, plates, and Ultrapure water used for PCR and sequencing were decontaminated prior to use by exposure to UV light for 20 min. A PCR master mix was prepared using 38.1 µL of UV-treated Ultrapure water, 0.4 µL of Invitrogen Platinum Taq (ThermoFisher Scientific), 5 µL of 10X Thermopol buffer, 1 µL of 10 mM dNTPs, and 1 µL of the 1:3 primer cocktail for a total reaction volume of 50 µL for 2 µL of template. The amplification reaction conditions were 95 °C initial denaturation for 5 min, followed by 40 cycles of denaturation at 95 °C for 30 s, annealing at 60 °C for 30 s, extension at 72 °C for 30 s, and a final extension at 72 °C for 2 min. All PCR amplifications were first visualized on a 1% ethidium bromide gel then extracted from the gel and purified using NucleoMag beads (Macherey-Nagel) as previously described [21]. The purified amplicon library was sequenced using a 401 × 101 cycle using TG MiSeq Reagent Nano Kit v2 (Illumina Canada, MS-103-1003) on a MiSeq System (Illumina).

Demultiplexed sequences were used to generate ASVs using QIIME2 version 2019.10.0 [33, 34]. Only forward reads containing a 400-nucleotide long amplicon were imported into QIIME2 and truncated to 200 nucleotides using DADA2 version 2019.10.0 [35]. The trimmed reads were then denoised and chimeric sequences removed prior to ASV generation using DADA2. A QIIME2 naïve-Bayes taxonomy classifier [33, 34] was constructed using a combination of reference sequences from the cpnDB reference database [22] and NCBI [36]. Representative sequences produced from DADA2 were used with nBLAST to query the NCBI non-redundant, cultured-only reference database, incorporating an e-value threshold of 1e−6, using Entrez Direct [37]. The top three results for each sequence query were collected to create a database containing 16,624 sequences. Replicate entries and sequences below 180 nucleotides were removed (4400) and the remaining sequences (12,224) were combined with the cpnDB_nr (5489) to create a final database of 17,713 cpn60 gene reference sequences. Taxonomic lineages for the representative sequences were obtained from NCBI using the reference sequence accession numbers and subsequently formatted to integrate within the QIIME2 environment.

All cpn60 sequences generated for the current study were deposited in the European Nucleotide Archive (ENA) under project accession number PRJEB43503. A cpn60 ASV table has been made available at https://figshare.com/articles/dataset/cpn60_ASV_table/14955753.

Cytochrome oxidase I (COXI) genes for the Cape eland, donkey, goat, horse, olive baboon, Przewalski’s horse, sheep, spotted hyena, and Sumatran orangutan were obtained from a previous study [8] and used to construct a mammalian COXI-based phylogeny. The mammalian COXI gene sequences were aligned using ClustalW [38] in MEGA X version 10.1.8 [39], with trimming and gap removal as appropriate. The optimal nucleotide substitution model was determined using JModelTest2 version 2.1.10 [40, 41]. For the mammalian COXI gene, a maximum likelihood dendrogram was constructed using MEGA X with a GTR + G + I substitutional model and with a confidence assessment of 1000 bootstraps. The mammalian phylogeny was then compared with literature and confirmed for accuracy.

Microbial dendrograms based on community composition were generated using 16S rRNA gene sequences obtained from a previously generated amplicon dataset [8] and the cpn60 gene sequences from this current study. For the 16S rRNA and cpn60 gene microbial dendrograms, the respective ASV tables were sample-collapsed based on mammalian host, and the ASV read counts were summed for each category. Bray-Curtis, unweighted UniFrac, and weighted UniFrac distance matrices for the ASV tables were generated using the QIIME2 diversity beta-rarefaction command, rarefied to 1000 reads, and used to construct UPGMA dendrograms with a confidence assessment of 1000 bootstraps.

To compare microbial community composition between 16S rRNA and cpn60 gene datasets, Procrustes analyses were performed on principal coordinate analysis (PCoA) ordinations generated using Bray-Curtis and weighted/unweighted UniFrac metrics. Distance matrices were generated using the qiime2 diversity core-metrics-phylogenetic command, then exported with qiime-tools export. Procrustes analyses were completed in R using the vegan package and protest command with 100,000 permutations and plotted using ggplot.

Mammalian COXI gene and microbial 16S rRNA and cpn60 gene dendrograms were compared and evaluated for congruence using the vegan, phangorn, and ape packages in R as previously described [8]. Phylosymbiosis was assessed with Robinson-Foulds scores for both the 16S rRNA and cpn60 gene dendrograms against the mammalian COXI dendrogram. The significance of the Robinson-Foulds metric was determined by comparing the mammalian COXI gene dendrogram against 100,000 randomly generated trees containing identical terminal nodes (i.e., taxa). Congruency between dendrograms were measured with the normalized Robinson-Foulds score, ranging between 0 and 1, with 0 representing perfect congruency. Random dendrograms were considered significant if they obtained Robinson-Foulds scores equal to or greater than those obtained from comparisons between the mammalian phylogeny and the microbial gene (16S rRNA or cpn60) dendrograms.

To investigate mammalian skin microbial communities using the cpn60 gene, 95 representative skin swab samples, already extracted for genomic DNA and sequenced from a previous project [8], were selected for additional amplification and sequencing of the cpn60 gene. This cpn60 gene amplicon dataset contained samples from 19 unique mammals, with varying representation with respect to number of samples and read proportions (Table S1). Of the 95 samples submitted for sequencing, 88 unique samples contained at least one read, with most samples associated with fewer than 500 total reads. The horse and Przewalski’s horse samples represented 36.6% and 25.4% of all sequenced cpn60 gene reads, respectively, followed by the olive baboon (15.8%) and Cape eland (10.0%). All other mammalian host groups contained reads representing less than 5% of total cpn60 gene reads. To avoid downstream analysis issues related to shallow sampling depths, all samples with fewer than 1000 reads were removed, resulting in a final dataset of 37 samples (Fig. 1). This subset contributed 97.6% (118,645) of all cpn60 gene reads (121,622). Although overall read loss was minimal, mammalian host representation was reduced from 19 to 9: the Cape eland, donkey, goat, horse, olive baboon, Przewalski’s horse, sheep, potted hyena, and Sumatran orangutan (Fig. 1).

The average number of amplicons per mammalian host is indicated with a box.

The loss of mammalian host representation due to shallow sequencing depth was unexpected and might have methodological cause. Microbial biomass on mammalian skin is variable and can be comparably low, with dry-swab sampling of the skin producing the least amount of biomass compared to other methods [42]. The samples used in this current study were collected via the dry-swabbing method [8] and therefore are likely to have relatively low biomass and associated DNA yields. Additionally, the samples were several years old (i.e., four years at time of sequencing) and stored at −20 °C instead of the recommended −80 °C for skin swab samples [42]. Combined, these factors may have affected the integrity of the genomic DNA and associated cpn60 templates, and subsequent sequencing depth. The cpn60 gene itself has a median copy number of one per genome [16] and thus samples could be more susceptible to DNA degradation impacting target amplification. Future investigations of specific communities of the mammalian skin microbiome using the cpn60 gene would benefit from using genomic DNA that has been recently extracted from mammalian skin samples, ideally using wet-swabbing or tape-stripping to increase biomass collection [42].

Despite a reduction in host representation, cpn60-based sequencing provided valuable insight into the mammalian skin microbiome. The olive baboon samples contained distinct and comparatively uniform microbial profiles (Fig. 2). These profiles were represented by Prevotella, Prophyromonas, and Butyricicoccus, which were most abundant in the olive baboon samples. The Cape eland, horse, and Przewalski’s horse samples had less consistent microbial profiles among samples. The Cape eland samples contained sequences associated with Jeotgalicoccus that were shared among the goat, horse, and sheep samples. For the horse samples, microbial profiles were variable among samples and contained unique horse-associated sequences associated with Moraxella. The microbial profiles of the Przewalski’s horse samples were more uniform and contained sequences affiliated with Planomicrobium and Macrococcus; these taxa were nearly absent in other mammalian host microbial profiles. Of mammalian hosts with single sample representation, the Sumatran orangutan microbial profiles included more unique genera (nine) than the donkey (four), goat (two), sheep (zero), and spotted hyena (one). Across all samples, Corynebacterium-associated sequences dominated and were represented in moderate relative abundance (>5%) within the goat, horse, olive baboon, Przewalski’s horse, and sheep, although in some cases represented as much as 75% of total community. Sequences affiliated with Acidobacteria were also present on the donkey, horse, olive baboon, and Przewalski’s horse at relative abundances ranging from 4 to 38%. The most observed sequences belonged to unclassified bacteria (Bacteria_394), which were present in 35 of the 37 samples in high abundance, as well as an unresolved Proteobacteria (Proteobacteria_383) in 30 of the 37 samples.

Generated ASVs were collapsed to the genus level and filtered at >3% relative abundance. Bubble sizes represent the relative abundances of taxa in each sample. Taxa unresolved to a genus level were labeled according to their next resolved taxonomic level.

Comparisons against the previously classified 16S rRNA gene dataset showed inconsistencies among taxonomic profiles (Fig. 3, Table S2). Although different gene markers will be subject to various biases such as gene copy number [43, 44], nucleotide GC content [45, 46], the region targeted and primers used for amplification [47, 48] and certain bacterial proportions [49], the primary cause for dissimilarity in taxonomic profiles is the separate reference database used for each dataset. The availability of curated cpn60 gene reference databases is limited. The only currently maintained cpn60 database is the cpnDB [22], which at the timing of this study contained ~7000 sequences in its non-redundant database. Although this was increased to 17,713 sequences by combining it with nucleotide BLAST results, the cpnDB is far surpassed by the SILVA 138.1 16S rRNA gene database, which contains over 510,000 non-redundant reference sequences [50]. A large proportion of the cpn60 gene reads were unclassified bacteria or remained unresolved to the genus level compared to the 16S rRNA gene data (Table S2). An increase in cpn60 gene database coverage would improve classification and facilitate direct comparisons of taxonomy between microbial profiles generated from separate phylogenetic markers.

The ASVs were collapsed to the genus level and filtered at >5% relative abundance. Bubble sizes represent the relative abundances of taxa in each sample. Taxa unresolved to a genus level were labeled according to their next resolved taxonomic level.

In addition to limited database coverage, direct comparisons of taxonomic profiles between two phylogenetic markers requires that ASVs be collapsed into taxonomic levels (e.g., genus or species), which is dependent on the taxonomic structuring of the database itself. Much how ASVs with a single nucleotide difference may be interpreted as a unique ASV, taxonomic lineages that differ by as little as one character will be classified as separate taxa within QIIME2. In this study, the 16S rRNA gene dataset was assigned taxonomy using the SILVA database [50, 51], which obtains its taxonomic information from a combination of sources, including the NCBI and GTDB [52], and undergoes further manual annotation. In contrast, the cpnDB does not currently maintain a taxonomy reference database. Instead, the cpn60 gene dataset was assigned taxonomy based on an NCBI-derived taxonomy database generated for this study as a requirement for implementation into the QIIME2. As such, differences between the cpn60 and 16S rRNA gene profiles may have been affected by incongruous taxonomies between databases and may not accurately reflect true differences. For example, two Corynebacterium classifications (i.e., Corynebacterium_93, and Corynebacterium1_94) were present in the data (Fig. 3). A successful collapse to the genus level should have placed the associated ASVs into a single common Corynebacterium genus. In this case, the difference is the inclusion of a “1” at the end the Corynebacterium lineage within the 16S rRNA gene taxonomy file. Similarly, Massilia is separated into Massilia_1203 and Massilia_1347 (data not shown). Here, the differences are a result of higher taxonomic classification: Massilia has been reclassified under class Gammaproteobacteria within the 16S rRNA taxonomy based on the GTDB [52], whereas the NCBI taxonomy and SILVA database currently maintains the original class Betaproteobacteria lineage. As such, ASVs associated with the same genus are listed as two separate taxa instead of one, and the difference is “invisible” at the genus level. Should the cpn60 gene continue to be used as an alternative phylogenetic marker to the 16S rRNA gene, as well as implemented within the QIIME2 environment, it is important that a reference taxonomy database containing compatible taxonomic lineages with the SILVA database (or other routinely accessed reference databases) be maintained alongside the cpnDB so that direct comparisons can be made. Alternatively, a combination of BLAST and Smith-Waterman alignments (watered-BLAST) have been used previously to assign taxonomy to cpn60 gene datasets [25], although this method also uses the NCBI taxonomy database and would be subject to the same taxonomic incompatibility issues.

Microbiome data generated from two separate universal prokaryotic phylogenetic markers should, ideally, produce similar compositional profiles, despite any difference in taxonomic classification. PCoA and Procrustes analyses were used to compare microbial community composition dissimilarity between cpn60 and 16S rRNA gene datasets, independent of taxonomic classifications and bias imparted by chosen reference databases. The ordinations produced from both datasets showed significant correlations (p < 0.05), with correlation coefficients of 0.91, 0.69, and 0.66 for Bray-Curtis, weighted UniFrac, and unweighted UniFrac, respectively (Fig. 4). The olive baboon sample compositions were distinct, with samples grouping separate from other mammalian hosts for each tested metric. The Przewalski’s horse and Cape eland samples also grouped with their respective hosts, and all other samples grouped homogenously among mammalian hosts for each tested diversity metric.

Triangles (arrow-heads) and connecting lines indicate the change in ordination space of samples between datasets. The analyses were completed with 100,000 permutations to calculate significance.

Because Bray-Curtis only accounts for presence and abundance data, a high correlation of the Bray–Curtis metric indicates that both the cpn60 and 16S rRNA gene datasets contain a similar proportion and distribution of taxa within their respective samples. Including phylogenetic data (i.e., UniFrac) can unmask potential differences with respect to ASV relatedness and provide insight into whether microbial community composition between the cpn60 and 16S rRNA gene datasets are comparable. Indeed, weaker correlations observed using the UniFrac metrics could indicate that the cpn60 gene amplicon provides additional phylogenetic information that the V3-V4 region of the 16S rRNA gene cannot. As the V3-V4 fragment of the 16S rRNA gene often lacks sufficient nucleotide diversity to confidently resolve species [15], diversity metrics dependent on phylogenetic resolution will be similarly limited. In contrast, the cpn60 gene amplicons contain enough phylogenetic information to resolve species and radiate underlying phylogenetic trees [18, 53,54,55,56]. Additional validations using a subset of this current dataset support that cpn60 gene amplicons more clearly resolves species-level taxa compared to the 16S rRNA gene amplicons (Fig. S1). The differences in resolution will influence dissimilarity measurements resulting in changes in observed sample similarity, and ultimately differences in Procrustes correlation between datasets. Specifically for the unweighted UniFrac metric, the phylogenetic resolution provided by the cpn60 gene would have considerable impact because resolving of shallow-branch taxa contribute nearly 90% of the sample distance [53], resulting in lower correlations. For the weighted UniFrac, deep-branch taxa are largely responsible for sample distances [53] and therefore should be less influenced by an increase in phylogenetic resolution unless it results in changes to deep-branch topology (i.e., phylum or class-level changes). Nonetheless, all Procrustes tests (Fig. 4) indicate microbial community compositions generated from cpn60 and 16S rRNA gene datasets are significantly similar to each other and can be used to make similar conclusions regarding the mammalian host-associated microbiome and phylosymbiosis.

The phylogenetic resolution provided by the cpn60 marker gene should allow for additional observations of host-microbe associations, particularly if those associations are at finer taxonomic levels. Patterns of phylosymbiosis within cpn60 and 16 S rRNA gene amplicon microbial profiles were assessed by comparing microbial community composition dendrograms against a COXI mammalian dendrogram representing mammalian phylogenetic history. Significant (p = 6.73 × 10−3) patterns of phylosymbiosis were observed in the 16S rRNA gene Bray–Curtis microbial dendrogram for clades containing the Cape eland, goat, and sheep (Artiodactyla) and the donkey, Przewalski’s horse, and horse (Perissodactyla), although these observations were not significant for the cpn60-based microbial dendrogram (Fig. 5). In contrast, significant results were observed for the cpn60 gene amplicon unweighted (p = 4.36 × 10−2) and weighted (p = 4.43 × 10−2) UniFrac microbial dendrograms for Artiodactyla (Cape eland excluded) and Perissodactyla, but not within the 16S rRNA gene-based dendrograms. Evidence for phylosymbiosis was absent for the Primates (olive baboon and Sumatran orangutan) and Carnivora (spotted hyena). These observations further confirm that microbial communities of Perissodactyla and Artiodactyla are influenced by host evolutionary history, as observed previously using the 16S rRNA gene [8]. Because the mammalian hosts included in the study varied in location and age, and are potentially more influenced by “environmental” microorganisms, phylosymbiosis patterns could be masked when mammals from multiple mammalian orders are included together for analysis [8]. This current study relies on samples originally obtained from the same publication, thus similar confounding factors may also influence the current observations, and have been previously addressed [8]. Nonetheless, cpn60-based microbial dendrograms produced significant and congruent phylosymbiosis results for Artiodactyla and Perissodactyla using UniFrac measures, without requiring isolating them from other mammalian orders, as performed previously [8]. That phylogenetic-based UniFrac metrics produced significant phylosymbiosis results only with the cpn60-based microbial profiles suggests that the phylogenetic resolution provided by the cpn60 gene reveals subtle compositional differences and unmasks phylosymbiosis patterns not observed in 16S rRNA gene dataset. However, mammalian host representation is limited within this current study, and the inclusion of additional samples might result in re-masking of phylosymbiosis patterns.

Blue squares indicate identical clades between the mammalian phylogeny and microbial dendrogram. Congruency was tested for significance using a normalized Robinson-Foulds measure (nRF) which ranges between 0 and 1, with 0 representing perfect congruity. Significant observations of tree congruity are indicated with an “*”.

A proposed benefit of using the cpn60 gene for microbiome profiling is the ability to universally detect and resolve specific microbial populations and their associations with an environment or host. To test this advantage, family Staphylococcaceae was selected for additional analysis, given the ubiquity of affiliated taxa among mammalian hosts [8] and the relevance of certain members to skin health and disease. Of the 37 samples, 33 (89%) contained Staphylococcaceae-associated reads. The proportion of reads associated with Staphylococcaceae varied among mammalian hosts and ranged from 0.08% (2/2307 reads) to 26.0% (1339/5,142 reads), with most samples (23/33, 69.9%) below 10% relative abundance (Fig. 6). The spotted hyena sample contained no Staphylococcaceae-associated reads and was therefore removed from further analysis. The olive baboon and Sumatran orangutan samples contained the fewest number of reads associated with Staphylococcaceae, with no sample exceeding 1% relative abundance. The distribution of specific Staphylococcaceae species varied among mammalian hosts. Reads associated with Jeotgalicoccus halophilus were present among all mammalian hosts excluding the donkey and Przewalski’s horse, and unresolved Staphylococcaceae species were similarly present, though absent from the Przewalski’s horse and Sumatran orangutan. For the Przewalski’s horse samples, Macrococcus carouselicus was the predominant Staphylococcaceae species in most samples, followed by Macrococcus equipercicus, which was absent from the horse samples. The horse and Przewalski’s horse samples both contained Salinicoccus species, but otherwise had minimal overlap between hosts. The horse samples were the most variable with overlap between many other mammalian hosts, although contained sequences uniquely associated with Staphylococcus fleurettii, Salinicoccus halodurans, and Macrococcus brunensis. The Cape eland samples contained Staphylococcaceae populations that evenly split between an unresolved Staphylococcaceae and Jeotgalicoccus halophilus.

Taxa resolved to a species level are indicated in blue. Bubble sizes represent the relative abundances of specific taxa in each sample, calculated from the total number of Staphylococcaceae reads. The proportion of Staphylococcaceae reads in the cpn60 gene amplicon dataset is indicated for each sample within the bottom bar chart.

Comparisons made between the cpn60 and 16S rRNA gene amplicon Staphylococcaceae profiles show improved species classification of Staphylococcaceae populations when using the cpn60 gene (Fig. 6, Fig. S1). Most of the profiles produced using the 16S rRNA gene contained large relative abundances of unresolved Staphylococcaceae sequences, with only two resolved species present in the dataset (i.e., Macrococcus brunensis and Salinicoccus roseus). In comparison, the cpn60 gene profiles contained mainly species-resolved populations, with few exceptions. The benefit of improved phylogenetic resolution is most evident for Prezwalski’s horse samples, wherein most Staphylococcaceae-associated sequences resolved to Macrococcus carouselicus or Macrococcus equipercicus between the 16S rRNA gene and cpn60 gene dataset. The distribution and relative abundances of these taxa, in combination with previous observations of phylosymbiosis in Perissodactyla, suggests a host-specificity for Przewalski’s horses and potential co-evolutionary influence. The Przewalski’s horse is considered a true “wild” horse, having remained undomesticated and relatively isolated to the Asian steppes [57], although these current samples are sourced from the Toronto zoo (Ontario, Canada). The skin environment of Przewalski’s horses could differ from the common domesticated horses include in this study, and thus might influence microbial community assemblage. Importantly, although the horses included in this analysis were regularly brushed (i.e., daily to weekly), the Przewalski’s horses were left ungroomed. As such, the population of Staphylococcaceae observed on Przewalski’s horses could represent a more “natural” population, in contrast to domesticated horses where their microbiota is continually disturbed by human interaction. Differences or disruption in Staphylococcaceae populations on the Prezewalski’s horse and domesticated horse could impact host susceptibility to disease given that Staphylococcaceae has strong associations in the development of equine pastern dermatitis [58].

The most prominent resolved Staphylococcaceae representative affiliated with the cpn60 dataset was Jeotgalicoccus halophilus. The comparable 16S rRNA gene dataset [8] also contains the genus Jeotgalicoccus with similar proportions for the Cape eland and several horse samples, although most Staphylococcaceae affiliated sequences belong to the Macrococcus genus (83.7%). None of the Jeotgalicoccus associated ASVs within the 16S rRNA gene dataset were resolved to J. halophilus. The genus Jeotgalicoccus was originally isolated from a traditional Korean fermented seafood [59], with other representatives captured as aerosols from pig [60] and turkey [61] farms. Specifically, J. halophilus was first isolated from a salt lake [62] and has since been detected as an airborne bacterium in hatcheries [63] and in association with marine corals [64]. At this writing, there is no additional literature that mentions J. halophilus in association with mammalian skin, although this does not exclude their previous detection. Querying the NCBI database returned sequences obtained from environmental studies previously mentioned, but also included a bovine mastitis study in which J. halophilus was detected [65].

Although Jeotgalicoccus was detected previously on mammalian skin using the 16S rRNA gene [8], the cpn60 gene has enabled resolution of J. halophilus from other species therein. Within the Primates, J. halophilus represents the totality of the Staphylococcaceae associated reads (Fig. 6), although this is due to low Staphylococcaceae read depth. However, the Cape eland samples have a considerably higher Staphylococcaceae read depth, with J. halophilus representing at least half the total reads, and are the predominant species-resolved taxon. Similarly, the sheep sample, which contained the highest proportion of Staphylococcacea-associated reads, also had a high proportion of J. halophilus, suggesting that the relative abundance and prevalence of this taxon is not an artifact of limited read depth. Additionally, given its absence in several samples and two mammalian hosts (i.e., donkey and Przewalski’s horse) as well as controls, it is unlikely to be a cross-contaminant introduced from within the lab during sample extraction and processing.

The importance or role of J. halophilus in the context of mammalian skin is unknown. As a facultative anaerobe with basic metabolic requirements, a growth range of 4–40 °C, and salt tolerability of 0.1 to 16% w/v [62], it is well adapted for survival on mammalian skin. As well, it is coagulase and oxidase positive and resistant to several natural antibiotics [62], which could indicate its ability to act as an opportunistic pathogen. It is difficult to make conclusions about the influence of detected host associated Staphylococcaceae with respect to skin function, health, and disease. For example, Staphylococcus fleuretti, found in a single horse sample, is a coagulase-negative organism associated with various animal diseases and has been indicated as a contributor to methicillin resistance within the environment [66]. However, the horse from which the sample was taken within this current study had no reported skin health issues, although did have a mild respiratory infection [8]. Even well-established pathogens, like S. aureus or S. epidermidis, can exist in non-disease states within the skin microbiome, only causing pathology when the skin barrier is broken or when the community is disrupted [67, 68]. Thus, the detection of disease-associated genera or species on mammalian skin can provide only limited insight into host-microbe dynamics. Ultimately, more work is necessary to further characterize the interactions these microbial species might have with their mammalian hosts. Regardless, the fact that these specific Staphylococcus spp. have been detected on mammalian skin demonstrates the potential for cpn60 to resolve species from vague genus classifications and for its application in universal high-resolution profiling of the mammalian skin microbiome.

Mammalian skin microbial profiles generated using the cpn60 marker gene were shown to be comparable with 16S rRNA gene datasets and supports previous 16S rRNA gene-based observations of phylosymbiosis. Evidence for phylosymbiosis was observed in the mammalian orders Perissodactyla and Artiodactyla using the cpn60 gene amplicon dataset and phylogeny-based UniFrac distance metric; this evidence was absent in the 16S rRNA gene amplicon dataset. Resolving species from within vague genus classifications provides insight into the distribution, presence, and potential influence of specific host-associated taxa and their influence on skin health and disease. The cpn60 gene amplicon dataset revealed previously unobserved associations between mammalian hosts and specific taxa, such as Jeotgalicoccus halophilus, that otherwise would have remained undetected. Amplification of the cpn60 gene does not exclude specific bacterial communities over others (e.g., archaea-specific or species-specific primers) and can thus be used for both whole microbial community and species-specific profiling. Although the 16S rRNA gene is likely to remain as the dominantly used phylogenetic marker for amplicon-based studies, the cpn60 gene is complementary to microbiome studies where universal low-level taxonomic resolution is desired. However, if cpn60 amplicon studies are to be compared with those produced using the 16S rRNA gene, it is imperative that a standardized taxonomy database be maintained alongside the cpnDB sequence reference database. This study integrated cpnDB into the QIIME2 environment using the NCBI taxonomy database, allowing for more rapid analysis of cpn60 gene amplicon datasets in the future. However, in lieu of a separate taxonomy database, progress towards integrating the cpnDB with existing taxonomy databases, such as SILVA [50, 51] or the ribosomal database project [69] would facilitate future cpn60-based research and improving taxonomic assignment, both within and external to the QIIME2 environment.

All data generated during this current study are available in the European Nucleotide Archive (ENA) repository under project accession number PRJEB43503. The complete cpn60 ASV table used for this study is available at https://figshare.com/articles/dataset/cpn60_ASV_table/14955753.

Telford MJ, Littlewood DTJ (eds). Animal evolution: genomes, fossils, and trees. Oxford, New York: Oxford University Press; 2009.

McFall-Ngai M, Hadfield MG, Bosch TCG, Carey HV, Domazet-Lošo T, Douglas AE, et al. Animals in a bacterial world, a new imperative for the life sciences. Proc Natl Acad Sci USA. 2013;110:3229–36.

Article CAS PubMed PubMed Central Google Scholar

Sanford JA, Gallo RL. Functions of the skin microbiota in health and disease. Semin Immunol. 2013;25:370–7.

Article CAS PubMed PubMed Central Google Scholar

Hooper LV, Littman DR, Macpherson AJ. Interactions between the microbiota and the immune system. Science. 2012;336:1268–73.

Article CAS PubMed PubMed Central Google Scholar

Council S, Skene JHP, Platt ML, Dunn RR, Horvath JE. Diversity and evolution of the primate skin microbiome. Proc Royal Soc B. 2016;283:20152586.

Article Google Scholar

Ross AA, Rodrigues Hoffmann A, Neufeld JD. The skin microbiome of vertebrates. Microbiome. 2019;7:79.

Article PubMed PubMed Central Google Scholar

Lavrinienko A, Tukalenko E, Mappes T, Watts PC. Skin and gut microbiomes of a wild mammal respond to different environmental cues. Microbiome. 2018;6:209.

Article PubMed PubMed Central Google Scholar

Ross AA, Müller KM, Weese JS, Neufeld JD. Comprehensive skin microbiome analysis reveals the uniqueness of human skin and evidence for phylosymbiosis within the class Mammalia. Proc Natl Acad Sci USA. 2018;115:E5786–E5795.

Article CAS PubMed PubMed Central Google Scholar

Fargione J, Brown CS, Tilman D. Community assembly and invasion: An experimental test of neutral versus niche processes. Proc Natl Acad Sci USA. 2003;100:8916–20.

Article CAS PubMed PubMed Central Google Scholar

Brooks AW, Kohl KD, Brucker RM, van Opstal EJ, Bordenstein SR. Phylosymbiosis: relationships and functional effects of microbial communities across host evolutionary history. PLoS Biol. 2016;14:e2000225.

Article PubMed PubMed Central Google Scholar

Davenport ER, Sanders JG, Song SJ, Amato KR, Clark AG, Knight R. The human microbiome in evolution. BMC Biol. 2017;15:127.

Article PubMed PubMed Central Google Scholar

Lim SJ, Bordenstein SR. An introduction to phylosymbiosis. Proc Royal Soc B. 2020;287:20192900.

Article Google Scholar

Foster T. Staphylococcus. In: Baron S (ed). Medical microbiology, 4th ed. Galveston, Texas: University of Texas Medical Branch at Galveston; 1996.

Rich M. Staphylococci in animals: prevalence, identification and antimicrobial susceptibility, with an emphasis on methicillin-resistant Staphylococcus aureus. Br J Biomed Sci. 2005;62:98–105.

Article CAS PubMed Google Scholar

Zeigler DR. Gene sequences useful for predicting relatedness of whole genomes in bacteria. Int J Syst Evol Microbiol. 2003;53:1893–1900.

Article CAS PubMed Google Scholar

Links MG, Dumonceaux TJ, Hemmingsen SM, Hill JE. The chaperonin-60 universal target is barcode for bacteria that enables de novo assembly of metagenomic sequence data. PLoS One. 2012;7:e49755.

Article CAS PubMed PubMed Central Google Scholar

Fayet O, Ziegelhoffer T, Georgopoulos C. The groES and groEL heat shock gene products of Escherichia coli are essential for bacterial growth at all temperatures. J Bacteriol. 1989;171:1379–85.

Article CAS PubMed PubMed Central Google Scholar

Vancuren SJ, Santos SJD, Hill JE, The Maternal Microbiome Legacy Project Team. Evaluation of variant calling for cpn60 barcode sequence-based microbiome profiling. PLoS One. 2020;15:e0235682.

Article CAS PubMed PubMed Central Google Scholar

Ren Q, Hill JE. Rapid and accurate taxonomic classification of cpn60 amplicon sequence variants. ISME Commun. 2023; In Press.

Hill JE, Town JR, Hemmingsen SM. Improved template representation in cpn60 polymerase chain reaction (PCR) product libraries generated from complex templates by application of a specific mixture of PCR primers. Environ Microbiol. 2006;8:741–6.

Article CAS PubMed Google Scholar

Fernando C, Hill JE. cpn60 metagenomic amplicon library preparation for the Illumina Miseq platform. 2021. Protocol Exchange.

Hill JE. cpnDB: a chaperonin sequence database. Genome Res. 2004;14:1669–75.

Article CAS PubMed PubMed Central Google Scholar

Chaban B, Links MG, Jayaprakash TP, Wagner EC, Bourque DK, Lohn Z, et al. Characterization of the vaginal microbiota of healthy Canadian women through the menstrual cycle. Microbiome. 2014;2:23.

Article PubMed PubMed Central Google Scholar

Albert AYK, Chaban B, Wagner EC, Schellenberg JJ, Links MG, van Schalkwyk J, et al. A study of the vaginal microbiome in healthy Canadian women utilizing cpn60-based molecular profiling reveals distinct gardnerella subgroup community state types. PLoS One. 2015;10:e0135620.

Article PubMed PubMed Central Google Scholar

Schellenberg J, Links MG, Hill JE, Dumonceaux TJ, Peters GA, Tyler S, et al. Pyrosequencing of the chaperonin-60 universal target as a tool for determining microbial community composition. Appl Environ Microbiol. 2009;75:2889–98.

Article CAS PubMed PubMed Central Google Scholar

Goh SH, Facklam RR, Chang M, Hill JE, Tyrrell GJ, Burns ECM, et al. Identification of Enterococcus species and phenotypically similar Lactococcus and Vagococcus species by reverse checkerboard hybridization to chaperonin 60 gene sequences. J Clin Microbiol. 2000;38:3953–9.

Article CAS PubMed PubMed Central Google Scholar

Oliver A, Chase AB, Weihe C, Orchanian SB, Riedel SF, Hendrickson CL, et al. High-fiber, whole-food dietary intervention alters the human gut microbiome but not fecal short-chain fatty acids. mSystems. 2021;6:e00115–21.

Article CAS PubMed PubMed Central Google Scholar

Sakamoto M, Ohkuma M. Usefulness of the hsp60 gene for the identification and classification of Gram-negative anaerobic rods. J Med Microbiol. 2010;59:1293–302.

Article CAS PubMed Google Scholar

Ishikawa D, Sasaki T, Takahashi M, Kuwahara-Arai K, Haga K, Ito S, et al. The microbial composition of bacteroidetes species in ulcerative colitis is effectively improved by combination therapy with fecal microbiota transplantation and antibiotics. Inflamm Bowel Dis. 2018;24:2590–8.

Article PubMed Google Scholar

Callahan BJ, McMurdie PJ, Holmes SP. Exact sequence variants should replace operational taxonomic units in marker-gene data analysis. ISME J. 2017;11:2639–43.

Article PubMed PubMed Central Google Scholar

Scholz CFP, Jensen A, Lomholt HB, Brüggemann H, Kilian M. A novel high-resolution single locus sequence typing scheme for mixed populations of propionibacterium acnes in vivo. PLoS One. 2014;9:e104199.

Article PubMed PubMed Central Google Scholar

Hill JE, Seipp RP, Betts M, Hawkins L, Kessel AGV, Crosby WL, et al. Extensive profiling of a complex microbial community by high-throughput sequencing. Appl Environ Microbiol. 2002;68:3055–66.

Article CAS PubMed PubMed Central Google Scholar

Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. 2019;37:852–7.

Article CAS PubMed PubMed Central Google Scholar

Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 2010;7:335–6.

Article CAS PubMed PubMed Central Google Scholar

Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: high-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13:581–3.

Article CAS PubMed PubMed Central Google Scholar

NCBI Resource Coordinators. Database resources of the national center for biotechnology information. Nucleic Acids Res. 2016;44:D7–D19.

Article Google Scholar

Kans J. Entrez direct: e-utilities on the UNIX command line. Entrez Programming Utilities Help [Internet]. https://www.ncbi.nlm.nih.gov/sites/books/NBK179288/.

Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994;22:4673–80.

Article CAS PubMed PubMed Central Google Scholar

Kumar S, Stecher G, Li M, Knyaz C, Tamura K. MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol Biol Evol. 2018;35:1547–9.

Article CAS PubMed PubMed Central Google Scholar

Darriba D, Taboada GL, Doallo R, Posada D. jModelTest 2: more models, new heuristics and parallel computing. Nat Methods. 2012;9:772–772.

Article CAS PubMed PubMed Central Google Scholar

Guindon S, Gascuel O. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol. 2003;52:696–704.

Article PubMed Google Scholar

Kong HH, Andersson B, Clavel T, Common JE, Jackson SA, Olson ND, et al. Performing skin microbiome research: a method to the madness. J Investig Dermatol. 2017;137:561–8.

Article CAS PubMed Google Scholar

Větrovský T, Baldrian P. The variability of the 16S rRNA gene in bacterial genomes and its consequences for bacterial community analyses. PLoS One. 2013;8:e57923.

Article PubMed PubMed Central Google Scholar

Louca S, Doebeli M, Parfrey LW. Correcting for 16S rRNA gene copy numbers in microbiome surveys remains an unsolved problem. Microbiome. 2018;6:41.

Article PubMed PubMed Central Google Scholar

Pinto AJ, Raskin L. PCR biases distort bacterial and archaeal community structure in pyrosequencing datasets. PLoS One. 2012;7:e43093.

Article CAS PubMed PubMed Central Google Scholar

Laursen MF, Dalgaard MD, Bahl MI. Genomic GC-content affects the accuracy of 16S rRNA gene sequencing based microbial profiling due to PCR bias. Front Microbiol. 2017;8:1934.

Article PubMed PubMed Central Google Scholar

Bukin YS, Galachyants YP, Morozov IV, Bukin SV, Zakharenko AS, Zemskaya TI. The effect of 16S rRNA region choice on bacterial community metabarcoding results. Sci Data. 2019;6:190007.

Article CAS PubMed PubMed Central Google Scholar

Meisel JS, Hannigan GD, Tyldsley AS, SanMiguel AJ, Hodkinson BP, Zheng Q, et al. Skin microbiome surveys are strongly influenced by experimental design. J Investig Dermatol. 2016;136:947–56.

Article CAS PubMed Google Scholar

Brooks JP, Edwards DJ, Harwich MD, Rivera MC, Fettweis JM, Serrano MG, et al. The truth about metagenomics: quantifying and counteracting bias in 16S rRNA studies. BMC Microbiol. 2015;15:66.

Article PubMed PubMed Central Google Scholar

Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. 2013;41:D590–D596.

Article CAS PubMed Google Scholar

Pruesse E, Peplies J, Glöckner FO. SINA: accurate high-throughput multiple sequence alignment of ribosomal RNA genes. Bioinformatics. 2012;28:1823–9.

Article CAS PubMed PubMed Central Google Scholar

Parks DH, Chuvochina M, Waite DW, Rinke C, Skarshewski A, Chaumeil P-A, et al. A standardized bacterial taxonomy based on genome phylogeny substantially revises the tree of life. Nat Biotechnol. 2018;36:996–1004.

Article CAS PubMed Google Scholar

Fukuyama J. Emphasis on the deep or shallow parts of the tree provides a new characterization of phylogenetic distances. Genome Biol. 2019;20:131.

Article PubMed PubMed Central Google Scholar

Hill JE, Albert AYK, the VOGUE Research Group. Resolution and cooccurrence patterns of Gardnerella leopoldii, G. swidsinskii, G. piotii, and G. vaginalis within the vaginal microbiome. Infect Immun. 2019;87:e00532–19.

Article CAS PubMed PubMed Central Google Scholar

Verbeke TJ, Sparling R, Hill JE, Links MG, Levin D, Dumonceaux TJ. Predicting relatedness of bacterial genomes using the chaperonin-60 universal target (cpn60 UT): application to Thermoanaerobacter species. Syst Appl Microbiol. 2011;34:171–9.

Article CAS PubMed Google Scholar

Shukla I, Hill JE. cpn60 barcode sequences accurately identify newly defined genera within the Lactobacillaceae. Can J Microbiol. 2022;68:457–64.

Article CAS PubMed Google Scholar

Der Sarkissian C, Ermini L, Schubert M, Yang MA, Librado P, Fumagalli M, et al. Evolutionary genomics and conservation of the endangered Przewalski’s horse. Curr Biol. 2015;25:2577–83.

Article Google Scholar

Kaiser-Thom S, Hilty M, Axiak S, Gerber V. The skin microbiota in equine pastern dermatitis: a case-control study of horses in Switzerland. Vet Dermatol. 2021;32:646–e172.

Article PubMed PubMed Central Google Scholar

Yoon J-H, Lee K-C, Weiss N, Kang KH, Park Y-H. 2003. Jeotgalicoccus halotolerans gen. nov., sp. nov. and Jeotgalicoccus psychrophilus sp. nov., isolated from the traditional Korean fermented seafood jeotgal. Int J Syst Evol Microbiol. 2003;53:595–602.

Article CAS PubMed Google Scholar

Glaeser SP, Kleinhagauer T, Jäckel U, Klug K, Kämpfer P. Jeotgalicoccus schoeneichii sp. nov. isolated from exhaust air of a pig barn. Int J Syst Evol Microbiol. 2016;66:3503–8.

Article CAS PubMed Google Scholar

Kämpfer P, Busse H-J, Glaeser SP, Clermont D, Criscuolo A, Mietke H. Jeotgalicoccus meleagridis sp. nov. isolated from bioaerosol from emissions of a turkey fattening plant and reclassification of Jeotgalicoccus halophilus Liu et al. 2011 as a later heterotypic synonym of Jeotgalicoccus aerolatus Martin et al. 2011. Int J Syst Evol Microbiol. 2019;71:004745.

Article Google Scholar

Liu W-Y, Jiang L-L, Guo C-J, Yang SS. Jeotgalicoccus halophilus sp. nov., isolated from salt lakes. Int J Syst Evol Microbiol. 2011;61:1720–4.

Article CAS PubMed Google Scholar

Brauner P, Klug K, Jäckel U. Eggshells as a source for occupational exposure to airborne bacteria in hatcheries. J Occup Environ Hyg. 2016;13:950–9.

Article CAS PubMed Google Scholar

Peng J. Diversity and chemical defense role of culturable non-actinobacterial bacteria isolated from the South China Sea gorgonians. J Microbiol Biotechnol. 2013;23:437–43.

Article CAS PubMed Google Scholar

Bautista-Trujillo GU, Solorio-Rivera JL, Rentería-Solórzano I, Carranza-Germán SI, Bustos-Martínez JA, Arteaga-Garibay RI, et al. Performance of culture media for the isolation and identification of Staphylococcus aureus from bovine mastitis. J Med Microbiol. 2013;62:369–76.

Article CAS PubMed Google Scholar

Bhargava K, Zhang Y. Characterization of methicillin-resistant coagulase-negative staphylococci (MRCoNS) in retail meat. Food Microbiol. 2014;42:56–60.

Article CAS PubMed Google Scholar

Kobayashi T, Glatz M, Horiuchi K, Kawasaki H, Akiyama H, Kaplan DH, et al. Dysbiosis and Staphylococcus aureus colonization drives inflammation in atopic dermatitis. Immunity. 2015;42:756–66.

Article CAS PubMed PubMed Central Google Scholar

Williams MR, Gallo RL. Evidence that human skin microbiome dysbiosis promotes atopic dermatitis. J Investig Dermatol. 2017;137:2460–1.

Article CAS PubMed Google Scholar

Cole JR, Wang Q, Fish JA, Chai B, McGarrell DM, Sun Y, et al. Ribosomal database project: data and tools for high throughput rRNA analysis. Nucleic Acids Res. 2014;42:D633–D642.

Article CAS PubMed Google Scholar

Download references

This research was supported by a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada (NSERC).

Department of Biology, University of Waterloo, Waterloo, Ontario, Canada

Alexander K. Umbach & Josh D. Neufeld

Department of Veterinary Microbiology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

Champika Fernando & Janet E. Hill

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

AKU: Conceptualization, Investigation, Methodology, Formal analysis, Visualization, Writing—Original draft preparation. CF: Methodology, Validation. JEH: Supervision, Conceptualization, Methodology, Writing—Reviewing and Editing. JDN: Supervision, Conceptualization, Methodology, Funding acquisition, Writing—Reviewing and Editing.

Correspondence to Josh D. Neufeld.

The authors declare no competing interests.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

Umbach, A.K., Fernando, C., Hill, J.E. et al. Evaluating cpn60 for high-resolution profiling of the mammalian skin microbiome and detection of phylosymbiosis. ISME COMMUN. 3, 69 (2023). https://doi.org/10.1038/s43705-023-00276-y

Download citation

Received: 20 March 2023

Revised: 19 June 2023

Accepted: 21 June 2023

Published: 07 July 2023

DOI: https://doi.org/10.1038/s43705-023-00276-y

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative