Abstract
Background
Persian cats are predisposed to chronic and severe dermatophytosis. Alterations to the cutaneous microbiota are one potential contributor to this predisposition.
Objectives
To characterise the cutaneous and environmental fungal microbiota of Persian cats with chronic, severe dermatophytosis, and to compare the fungal microbiota of cats with and without dermatophytosis.
Animals:
Thirty-six client-owned cats, including 26 Persian cats and 10 domestic long hair (DLH) cats.
Methods and materials
Skin and home environment swabs were collected from Persian cats with severe, chronic dermatophytosis as well as groups of healthy control cats (Persian and DLH). Sequencing of the internal transcribed spacer 1 (ITS1) region was performed in addition to ITS1 quantitative PCR and fungal culture.
Results
Next-generation sequencing (NGS) targeting the fungal ITS region detected Microsporum sp. DNA from all Persian cats diagnosed with dermatophytosis and from environmental samples of their homes. A significant difference in community structure was identified between cases and controls, largely due to the Microsporum sp. DNA in samples from affected cats. Persian cats with dermatophytosis do not exhibit decreased fungal diversity. NGS failed to identify dermatophyte DNA on two culture-positive asymptomatic Persian controls and identified Trichophyton rubrum DNA from a culture-negative asymptomatic Persian control.
Conclusions
Aside from M. canis, our results indicate that an underlying fungal dysbiosis is not likely to play a role in development of dermatophytosis in Persian cats. Other explanations for predisposition to this disease, such as a primary immunodeficiency, ineffective grooming or unique features of Persian cat hair should be investigated.
Introduction
Dermatophytosis is a common infection of the hair, skin and nails caused by a group of keratinophilic, filamentous fungi referred to as dermatophytes.1 This disease occurs in humans and animals worldwide, and is caused by several different species of dermatophytes that exhibit preferences for particular hosts.1 While dermatophytosis is typically mild and self-limiting, it is a zoonotic disease with an enormous impact on human and animal welfare.1 Although uncommon, severe infections, including deep, nodular infections, are reported in both humans and animals.2,3
Feline dermatophytosis is most commonly caused by Microsporum canis.4 In one study, Persian cats represented just 5% of the total cats seen by a small animal hospital yet accounted for 25% of the cats diagnosed with dermatophytosis.5 Additionally, dermatophytic pseudomycetomas (pyogranulomatous nodules caused by the fungus) occur almost exclusively in Persian cats.4 It has been suggested that asymptomatic carriage of dermatophytes (also referred to as “mechanical carriage” of spores or “fomite carriage”) is more common in Persian cats than in other breeds;6 however, this has not been investigated thoroughly outside contaminated catteries.
It is unknown why Persian cats are afflicted more commonly and more severely than other breeds. Prevailing theories include (i) ineffective grooming of the long hair coat, (ii) an altered cutaneous microenvironment, or (iii) a primary (genetic) immunodeficiency.7 A recent study of feline grooming mechanics provides some support for ineffective grooming in Persian cats.8 Only anecdotal evidence exists thus far for a potential genetic predisposition. Alterations to the feline cutaneous microenvironment during dermatophytosis, particularly alterations in the microbiota, have been examined only with culture-based methods. In one particular culture-based study, cats initially infected with M. canis had cultures with mixed populations of M. canis and saprophytic fungi; after 30 days, all of these cats had pure cultures of M. canis, suggesting decreased fungal diversity in the face of clinical dermatophytosis.9
Next-generation sequencing (NGS) methods have allowed for more in-depth examination of the cutaneous microbiome than culture-based methods and have revealed a wide variety of microbes on the skin surface.10 Several NGS studies have shown that disruption of microbial communities (dysbiosis) is associated with a variety of diseases/processes in humans and animals, including atopic dermatitis, acne vulgaris, psoriasis, wound healing, cutaneous malignancies, and white-nose syndrome in bats.11-14
Importantly, the microbiome has been found to protect against colonization by pathogens. For example, particular members of the cutaneous microbiota, including Staphylococcus epidermidis, were shown to activate components of host adaptive immunity including particular T-cell subsets.15 By secreting antimicrobial peptides and proteases or competing for nutrients, S. epidermidis also can directly harm pathogens.15 Therefore, the bacterial and fungal cutaneous microbiota could be important in protecting the host against dermatophyte infection.
In this study, we employed targeted NGS and quantitative (q)PCR to profile the fungal microbiome of Persian cats with severe dermatophytosis. We hypothesised that Persian cats with severe disease would exhibit a cutaneous dysbiosis characterised by decreased microbial diversity and a high relative abundance of M. canis. We further hypothesised that healthy Persian cats would harbour different microbial populations than non-Persian cats.
Methods and materials
Ethics
An animal use protocol (IACUC 2018-0256 CA) was approved by the Texas A&M University Institutional Animal Care and Use Committee to enroll clinical cases for this study. Informed written consent was obtained from each owner.
Participants
Study participants included 26 adult Persian cats and 10 adult domestic long hair (DLH) cats (Table 1). Detailed sample metadata can be found in Supporting information Table S1. Persian cats were divided into three groups according to their clinical history and veterinary exam findings. Group 1, referred to as ”naïve Persian controls”, included 11 Persian cats with no lesions or history of dermatophytosis; Group 2, referred to as ”resistant Persian controls”, included eight Persian cats with documented past exposure to dermatophytes and no history of severe/chronic infection; Group 3, referred to as ”Persian cases”, included seven Persian cats with severe infections, including disseminated, chronic and deep infections (cases). Ten healthy DLH cats with no history of dermatophytosis served as a nonbreed-matched control population.
Table 1.
Overview of cats enrolled in the study.
| Case no. |
Group | Samples collected |
Age (years) |
Sex | Recent antifungals |
|---|---|---|---|---|---|
| 1 | Naïve Persian control | G, E | 1–5 | M | No |
| 2 | Naïve Persian control | G, E | >10 | F | No |
| 3 | Naïve Persian control | G | >10 | F | No |
| 4 | Naïve Persian control | G, E | 6–10 | M | No |
| 5 | Naïve Persian control | G | 1–5 | M | No |
| 6 | Naïve Persian control | G, E | 1–5 | F | No |
| 7 | Naïve Persian control | G, E | >10 | M | No |
| 8 | Naïve Persian control | G, E | >10 | M | No |
| 9 | Naïve Persian control | G | 1–5 | M | No |
| 10 | Naïve Persian control | G, E | >10 | F | No |
| 11 | Naïve Persian control | G, E | >10 | F | No |
| 12 | Resistant Persian control | G | >10 | M | No |
| 13 | Resistant Persian control | G | >10 | M | No |
| 14 | Resistant Persian control | G, E | 1–5 | F | No |
| 15 | Resistant Persian control | G | 6–10 | M | No |
| 16 | Resistant Persian control | G | 6–10 | F | No |
| 17 | Resistant Persian control | G | 6–10 | F | No |
| 18 | Resistant Persian control | G, E | 1–5 | M | No |
| 19 | Resistant Persian control | G, E | 6–10 | M | No |
| 20 | DLH control | G | 1–5 | F | No |
| 21 | DLH control | G | 1–5 | F | No |
| 22 | DLH control | G | 6–10 | M | No |
| 23 | DLH control | G | >10 | F | No |
| 24 | DLH control | G | >10 | F | No |
| 25 | DLH control | G | 1–5 | M | No |
| 26 | DLH control | G | 1–5 | M | No |
| 27 | DLH control | G | 6–10 | M | No |
| 28 | DLH control | G | >10 | F | No |
| 29 | DLH control | G | 1–5 | M | No |
| 30 | Persian case | G, L, E | >10 | M | No |
| 31 | Persian case | G, L | 6–10 | F | No |
| 32 | Persian case | G, L | 1–5 | F | Yes |
| 33 | Persian case | G, L | 6–10 | F | No |
| 34 | Persian case | G, L | 6–10 | M | No |
| 35 | Persian case | G | 1–5 | F | Yes |
| 36 | Persian case | G, L, E | 6–10 | M | Yes |
G, general swab; E, environmental swab; M, male; F, female; DLH, domestic long hair; L, lesion swab.
All Group 3 cats had at least two different diagnostic tests positive for dermatophytosis at the time of sample collection, including culture and one or more of the following: Wood’s lamp examination, skin cytological investigation, trichogram, PCR or biopsy examination. These cats exhibited clinical signs including alopecia over a large area of the body, scaling, erythema, pruritus and easily epilated hair. Five of these cats had chronic or recurrent infections despite previous treatment with oral itraconazole or terbinafine and topical lime sulfur treatments. Two cats had biopsy-confirmed pseudomycetoma in addition to the lesions described previously. Three of the Persian cases received oral and/or topical antifungal treatment within one month of sample collection.
Sample collection and sequencing
For each animal, two skin swabs (Isohelix, Cell Projects Ltd.; Harrietsham, UK) were rubbed on five total skin sites: interscapular, dorsal tail-base, axilla, groin and facial folds. These two swabs immediately were stored together in a single MoBio PowerBead tube (MoBio Laboratories, Inc.; Carlsbad, CA, USA), and this sample was referred to as the ”general” swab. For each of six Group 3 Persian cats with lesions, two additional swabs were rubbed on a single lesion. These two swabs immediately were stored together in a single MoBio PowerBead tube, and this sample was referred to as the ”lesion” swab. During swabbing, the hair was spread with gloved hands and swabs were rubbed 10 times on each side of the swab on the skin within a 2.54 cm2 area at each site. All swabs were stored within the MoBio PowerBead tube at 4°C for no more than two weeks before DNA extraction. Owners were instructed to use the same swabbing method – using the same type of swab – to collect samples from the household environment (specifically a site of carpet or bedding frequented by the cat).
DNA was extracted from skin and household environmental swabs using the MoBio PowerSoil DNA Isolation Kit (MoBio Laboratories) with a modified protocol. Modifications to the manufacturer’s protocol were (i) a reduction in volume of Solution C4 from 1.2 mL to 900 μL, and (ii) a reduction in volume of Solution C6 from 100 μL to 50 μL. Negative extraction controls were performed, including an unused Isohelix swab and a MoBio PowerBead tube without a swab.
Extracted DNA was sequenced at the University of Minnesota Genomics Center (Minneapolis, MN, USA) on an Illumina MiSeq (Illumina, Inc.; San Diego, CA, USA). The internal transcribed spacer 1 (ITS1) region was targeted using primers forward ITS1F_Nextera: TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCTTGGTCATTTAGAGGAAGTAA and reverse ITS2_Nextera: GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGCTGCGTTCTTCATCGATGC. The bacterial 16S region was sequenced as well; however, owing to the low number of bacterial reads in many of these samples and lack of amplification of Staphylococcus spp. from the mock community (positive control), these methods and results are relegated to Appendix S1. The sequences analysed are available in the NCBI Sequence Read Archive under BioProject ID PRJNA704990.
The qPCR targeting the ITS1 region was performed as described in section 1A.II of the protocol developed by Gohl et al.16 The qPCR was performed for 35 cycles on undiluted as well as eight-fold and 64-fold diluted template using the same primers as above. The resulting Cq values provided an estimate of the absolute abundance of fungal DNA in each sample.
Data analysis
Raw ITS sequencing data were adapter- and quality-trimmed using Trim Galore!17 Quality trimming was performed to a quality threshold of 18 as recommended by Mohsen et al.18 ITS sequences were further processed with QIIME 2 v2020.2,19 including denoising with DADA2,20 taxonomy assignment using the UNITE (dynamic_04.02.2020) ITS database for fungi,21 and a scikit-learn classifier.22 In the statistical software program R,23 the decontam package24 was used to remove two contaminant taxa using the “prevalence” method with the isContaminant command at a probability threshold of 0.5. Manual filtering was performed for one particular amplicon sequence variant (correlating with Malassezia sp.) that was not removed by decontam and was highly abundant in the swab control sample. Alpha diversity metrics including Shannon, Pielou’s evenness and observed ASVs (amplicon sequence variants) were calculated to uncover species richness and evenness within samples. Distance matrices were generated using Bray–Curtis dissimilarity and Jaccard distance metrics to enable evaluation of community structure and were visualised using principle coordinates analysis (PCoA) plots generated with Emperor.25,26
Dermatophyte culture
Toothbrushes were used to collect hair samples and to inoculate dermatophyte test media (DTM) plates (Hardy Diagnostics; Santa Maria, CA, USA) using the Mackenzie toothbrush method.4 Plates were incubated at room temperature and evaluated daily for fungal growth for up to 21 days. If a potential dermatophyte could not be definitively identified by standard morphological methods, direct colony PCR targeting the fungal ITS2 region was performed followed by Sanger sequencing to confirm the identity of the colony (Appendix S1). The number of dermatophyte colonies was used to help differentiate asymptomatic/fomite carriers and infected individuals, as described previously.4
Statistical methods
The statistical significance of qPCR results and alpha diversity results was analyzed using the Wilcoxon rank sum test or Kruskal–Wallis test in JMP Pro 15 (SAS Institute; Cary, NC, USA). For beta diversity results, ANOSIM (analysis of similarities) was performed on the distance matrices using the vegan package in R.27 The linear discriminant analysis (LDA) effect size (LEfSe) algorithm28 was used to analyse differential taxa abundance with an LDA cut-off score of 2.5 and significance cut-off of P < 0.01. The Benjamini–Hochberg procedure for P-value correction29 was used to correct for multiple comparisons where appropriate.
Results
For NGS targeting the ITS region, 55 samples were collected from Group 1, Group 2 and DLH healthy control cats (29 general swabs, 11 environment swabs) and Group 3 Persian cats with severe dermatophytosis (seven general swabs, six lesion swabs and two environment swabs; Table 1). Of these samples, three were excluded from downstream analysis owing to low sequence counts: one severe dermatophytosis general swab (sample 30G) and two healthy DLH general swabs (samples 23G and 26G).
After filtering, 1,018 ASVs were identified from four fungal phyla: Ascomycota, Basidiomycota, Chytridiomycota and Mucoromycota. The mean number of reads per sample was 10,648. The feature table was rarefied to 1,800 sequences/sample. Results of NGS targeting the bacterial microbiota can be found in Appendix S1.
Species richness and diversity
Alpha diversity metrics (Shannon, observed ASVs and Pielou’s evenness) showed no significant differences in fungal species evenness or richness between healthy Persian cats and healthy DLH cats, or between Persian cats with dermatophytosis and all healthy cats (Table 2). Likewise, no differences were observed between lesion swabs and the general swabs collected from the same affected cats (Shannon, P = 0.505; observed ASVs, P = 0.360; Pielou’s evenness, P = 0.627).
Table 2.
Statistical analyses of alpha and beta diversity metrics.
| Cases (general + lesion swabs) versus all control |
Lesion swabs only versus all control |
Healthy Persian versus DLH |
|
|---|---|---|---|
| Alpha diversity | |||
| Observed ASVs | 0.763 | 0.726 | 0.625 |
| Shannon | 0.505 | 0.926 | 0.926 |
| Pielou’s evenness | 0.342 | 0.898 | 0.912 |
| Beta diversity | |||
| Bray Curtis | R = 0.14, P = 0.05 | R = 0.22, P = 0.05 | R = −0.05, P = 0.84 |
| Jaccard | R = 0.14, P = 0.06 | R = 0.23, P = 0.06 | R = −0.07, P = 0.87 |
All alpha diversity values are Wilcoxon rank sum P-values. Beta diversity values are ANOSIM R-values and P-values. Bold values are significant at P ≤ 0.05.
Microbial community structure
Fungal community structure (beta diversity) was mildly and significantly different between all healthy cats and Persian cats with dermatophytosis, as demonstrated by the Bray–Curtis dissimilarity index (Fig. 1 and Table 2). However, the Jaccard distance metric did not indicate significance (Table 2). No differences in fungal community structure were observed between healthy Persian cats and healthy DLH cats (Table 2).
Figure 1. Principal coordinate analysis (PCoA) plot of Bray–Curtis distance matrix comparing fungal communities on Persian cats with severe dermatophytosis to healthy control cats (Persian cats and domestic long hair cats).
Fungal community structure of healthy control cats is mildly and significantly different from that of Persian cats with dermatophytosis (R = 0.14, P = 0.05). Persian cats with dermatophytosis are labelled with case number so that general swabs (G) and lesion swabs (L) from the same cat can be visualised relative to each other.
Microbial community composition
The composition of fungal communities in Persian cats with dermatophytosis and healthy cats were similar, with the exception of Microsporum sp., which was identified on the skin of all clinically affected cats and not on the controls (Figure 2). Microsporum sp. relative abundance in cases ranged from 3% to 88%, with an average of 8% in general skin swabs and 29% in lesion swabs. In addition to a higher relative abundance of Microsporum sp. and its parent taxa, LEfSe analysis identified a higher relative abundance of Bjerkandera sp. in the general and lesion swabs from Persian cats with dermatophytosis. However, this taxon was identified in only three cats enrolled at the same clinic and may therefore be specific to that clinic or geographical location. While not significantly differentially abundant, Candida sp. and Debaryomyces sp. also appeared to have higher relative abundance in these three cats (Figure 2).
Figure 2. Relative abundance of fungal taxa on the skin of all cats included in the study.
DLH control, domestic long hair controls; naïve Persian, Group 1 Persian cats; resistant Persian, Group 2 Persian cats. In each case of Group 3, severe dermatophytosis is identified by sample name so that lesion and general swabs from the same cat can be compared. Cases 32, 33 and 34 were enrolled at approximately the same time from the same clinic in Italy.
Across all cats, the most common fungal phyla, Ascomycota and Basidiomycota, were present at 47% and 46% average relative abundances, respectively. The most abundant families across all cats were Malasseziaceae, Didymosphaeriaceae and Aspergillaceae. The most common genus on both affected and healthy cats was Malassezia, with an average relative abundance of 15%.
Absolute abundance of fungal DNA
Quantitative PCR targeting the ITS1 region revealed that swabs from Persian cats with dermatophytosis exhibited higher estimated absolute abundance of fungal DNA than healthy cats, as evidenced by significantly lower Cq values (Figure 3). Significance was maintained over a range of sample dilutions (undiluted, P = 0.018; eight-fold dilution, P = 0.016; 64-fold dilution, P = 0.010).
Figure 3. Plot of quantitative PCR results from undiluted sample providing an estimate of absolute abundance of fungal DNA in samples from Persian cats with severe dermatophytosis (n = 12) and healthy control cats (Persian cats and domestic long hair cats, n = 28).
Lower Cq values indicate higher estimated absolute abundance of fungal DNA. Persian cats with dermatophytosis exhibit significantly higher estimated absolute abundances of fungal DNA.
Culture results and detection of asymptomatic carriers
Microsporum canis was cultured from all of the Group 3 Persian cases. It also was cultured from two asymptomatic Persian cats that were housed in an environment known to be contaminated with dermatophytes. Less than five colonies of M. canis were observed on the culture plates of these two cats, consistent with asymptomatic carriage of spores due to environmental contamination.4 NGS failed to detect M. canis DNA from skin swabs of these two cats. No other control cats were positive for dermatophyte growth in culture. NGS did detect Trichophyton rubrum on one culture-negative, asymptomatic Persian control at a relative abundance of 9%; T. rubrum was found at 4% relative abundance in this cat’s environment.
Dermatophytes in the environment
Swabs of the home environment exhibited significantly higher fungal alpha diversity than samples obtained from feline skin (P < 0.0001 for Shannon, observed ASVs and Pielou’s evenness). Of the two environmental samples obtained from Persian cats with dermatophytosis, both contained Microsporum DNA at relative abundances of 2% and 3%, respectively. Of the 11 environmental swabs obtained from Persian control cats, none had detectable Microsporum DNA, and two had T. rubrum DNA at 4% and <1% relative abundance.
Discussion
Previous research has not explained the increased incidence and severity of dermatophytosis in Persian cats. The influence of the cutaneous microbiota was examined here given that changes to the host microbiota are associated with several skin diseases in humans and animals. NGS successfully identified Microsporum sp. DNA from all cases of Persian cat dermatophytosis and from none of the healthy control cats. We did not identify an underlying fungal dysbiosis associated with dermatophytosis in this breed. Furthermore, no significant differences were identified in the bacterial microbiota between cases and controls, although only a few samples had sufficient bacterial sequences to use for downstream analysis. This suggests that alterations in the cutaneous microbiota, aside from M. canis itself, do not play a role in pathogenesis. The idea that the microbiome can be a host defence mechanism is well-established;15 however, the microbiota of these cats does not appear to play a role in defence against dermatophytes. These results do indicate that NGS (ITS amplicon sequencing) is capable of consistently identifying dermatophytes on the hair and skin of clinically affected cats.
We expected to see decreased alpha diversity from lesions of dermatophytosis, as a previous culture-based study showed that saprophyte growth decreased to zero as lesions progress over time.9 Additionally, two studies evaluating the microbiota of humans with tinea pedis (dermatophyte infection of the feet) showed decreased fungal diversity on lesional skin.30,31 However, our findings indicate that fungal diversity was similar on dermatophyte lesions and normal skin. Overall, healthy control samples had low fungal diversity and often were composed of a few predominant taxa, which may explain the lack of significant difference in diversity between cases and controls.
The fungal microbiota of healthy Persian cats is not significantly different from that of the DLH cats included in this study. This was investigated owing to previously described breed-related differences in feline microbiota compositions.32 There do not appear to be any unique features of the normal Persian cat fungal microbiota that might explain the increased incidence and severity of dermatophytosis. Importantly, the normal Persian cat microbiota does not contain M. canis.
Microsporum canis was cultured from the hair coat of two Group 2 asymptomatic Persian cats that occupied the same environment as another Persian cat with severe dermatophytosis. Few colonies of M. canis grew on the culture plates from these cats, and these findings are consistent with asymptomatic carriage of dermatophyte spores on the hair coat resulting from environmental contamination.4 NGS did not detect Microsporum DNA on these cats; this is most likely to the consequence of a combination of low spore abundance and the sample collection method. While the Mackenzie toothbrush method samples the hair coat of the entire cat, the microbiome swabbing method used in this study samples only a small surface area of the skin, avoiding the hair as much as possible. This swabbing method commonly used for NGS therefore is not likely to consistently detect low numbers of spores on the hair coat of asymptomatic carrier cats.
Regarding the environmental microbiota of Persian cats, the findings are as expected, with Microsporum DNA detected only in the environment of clinically infected cats.4 Trichophyton rubrum DNA was detected in the environment of two Persian cats and on the hair coat of one of those cats. This is an anthropophilic dermatophyte typically associated with human infections (including foot and nail fungal infections), and it has been shown previously that cats may carry this dermatophyte asymptomatically.4 In these two Persian cat households, we suspect that T. rubrum was spread to the environment and subsequently the hair coat of a cat by an infected human.
A key limitation of this study is the limited number of cases, with three of the seven cases in Group 3 enrolled from one clinic in Italy. The microbiota of these cats contained unique taxa, and effects of geography or presence of environmental fungi in that particular clinic may explain these differences rather than an association with dermatophytosis. Inclusion of more cases from various locations would have been ideal. An additional limitation was the treatment of three cases with oral and topical antifungal drugs within one month of sample collection. This may have altered the diversity and structure of microbial communities, although dermatophytes were still detected on all three cats via NGS and culture. Given the low sample size and lack of high-quality bacterial microbiota data, further work may be needed to confirm whether the Persian skin microbiome plays a role in development of dermatophytosis.
In summary, the microbiota of Persian cats afflicted with dermatophytosis is different from healthy cats only in the increased abundance, absolute and relative, of Microsporum sp. No other significant differences were identified to indicate that an underlying dysbiosis exists. This study further supports that M. canis is not a component of the normal cat microbiota or of the normal Persian cat microbiota specifically. Other contributors to the increased incidence and severity of dermatophytosis in Persian cats should be sought, such as primary immunodeficiency, ineffective grooming or unique features of Persian cat hair.
Supplementary Material
Sample metadata file including information on all cats included in the study.
Statistical analysis of bacterial alpha and beta diversity metrics.
Figure S1. Relative abundance of bacterial taxa on the skin of all cats included in the study.
DLH control, domestic long hair controls; naïve Persian, Group 1 Persian cats; resistant Persian, Group 2 Persian cats. In each case of Group 3, severe dermatophytosis is identified by sample name so that lesion and general swabs from the same cat can be compared.
Figure S2. Bacterial taxa that are more differentially abundant on Persians with severe dermatophytosis than on all healthy control cats.
Plot created using LEfSe (LDA > 2.5, P < 0.01).
Figure S3. Bacterial taxa that are more differentially abundant on healthy Persian cats compared with healthy control cats.
Plot created using LEfSe (LDA > 2.5, P < 0.01).
Acknowledgements
We would like to thank Amanda Friedeck, Mary McCaine, Cynthia McManis, Cheryl Stanley, Rendina McFadden, Massimo Beccati, Silvia Auxilia, Olga Sjatkovskaja, Sasha Gibbons and Warren Joubert for their immense help with sample collection. We acknowledge the University of Minnesota Genomics Center for performing the library preparation and sequencing, and Texas A&M High Performance Research Computing for access to computing resources.
Source of Funding:
This project was funded by a Winn Feline Foundation Miller Trust grant (MT16-015). Alexandra N. Myers was supported by an NIH-funded postdoctoral fellowship (T32OD011083).
Footnotes
Conflicts of Interest: No conflicts of interest have been declared.
REFERENCES
- 1.Gnat S, Nowakiewicz A, Łagowski D et al. Host- and pathogen-dependent susceptibility and predisposition to dermatophytosis. J Med Microbiol 2019; 68: 823–836. [DOI] [PubMed] [Google Scholar]
- 2.Lanternier F, Pathan S, Vincent QB et al. Deep dermatophytosis and inherited CARD9 deficiency. N Engl J Med 2013; 369: 1,704–1,714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Chang S-C, Liao J-W, Shyu C-L et al. Dermatophytic pseudomycetomas in four cats. Vet Dermatol 2011; 22: 181–187. [DOI] [PubMed] [Google Scholar]
- 4.Moriello KA, Coyner K, Paterson S et al. Diagnosis and treatment of dermatophytosis in dogs and cats: Clinical Consensus Guidelines of the World Association for Veterinary Dermatology. Vet Dermatol 2017; 28: 266–e268. [DOI] [PubMed] [Google Scholar]
- 5.Lewis DT, Foil CS, Hosgood G. Epidemiology and clinical features of dermatophytosis in dogs and cats at Louisiana State University: 1981–1990. Vet Dermatol 1991; 2: 53–58. [Google Scholar]
- 6.Nitta CY, Daniel AGT, Taborda CP et al. Isolation of dermatophytes from the hair coat of healthy Persian cats without skin lesions from commercial catteries located in São Paulo metropolitan area, Brazil. Acta Sci Vet 2016; 44: 0. [Google Scholar]
- 7.Nuttall TJ, German AJ, Holden SL et al. Successful resolution of dermatophyte mycetoma following terbinafine treatment in two cats. Vet Dermatol 2008; 19: 405–410. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Noel AC, Hu DL. Cats use hollow papillae to wick saliva into fur. Proc Natl Acad Sci U S A 2018; 115: 12,377–12,382. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Moriello KA, Deboer DJ. Fungal flora of the haircoat of cats with and without dermatophytosis. J Med Vet Mycol 1991; 29: 285–292. [DOI] [PubMed] [Google Scholar]
- 10.Findley K, Oh J, Yang J et al. Topographic diversity of fungal and bacterial communities in human skin. Nature 2013; 498: 367–370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sander MA, Sander MS, Isaac-Renton JL et al. The cutaneous microbiome: Implications for dermatology practice. J Cutan Med Surg 2019; 23: 436–441. [DOI] [PubMed] [Google Scholar]
- 12.Grisnik M, Bowers O, Moore AJ et al. The cutaneous microbiota of bats has in vitro antifungal activity against the white nose pathogen. FEMS Microbiol Ecol 2020; 96: fiz193. [DOI] [PubMed] [Google Scholar]
- 13.Meason-Smith C, Diesel A, Patterson AP et al. Characterization of the cutaneous mycobiota in healthy and allergic cats using next generation sequencing. Vet Dermatol 2017; 28: 71–e17. [DOI] [PubMed] [Google Scholar]
- 14.Langan EA, Künstner A, Miodovnik M et al. Combined culture and metagenomic analyses reveal significant shifts in the composition of the cutaneous microbiome in psoriasis. Br J Dermatol 2019; 181: 1,254–1,264. [DOI] [PubMed] [Google Scholar]
- 15.Stacy A, Belkaid Y. Microbial guardians of skin health. Science 2019; 363: 227–228. [DOI] [PubMed] [Google Scholar]
- 16.Gohl D, MacLean A, Hauge A et al. An optimized protocol for high-throughput amplicon-based microbiome profiling. Protoc Exch 2016; 10: 978–971. [Google Scholar]
- 17.Krueger F Trim Galore! Available at: http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/. Accessed July 7, 2020. [Google Scholar]
- 18.Mohsen A, Park J, Chen Y-A et al. Impact of quality trimming on the efficiency of reads joining and diversity analysis of Illumina paired-end reads in the context of QIIME1 and QIIME2 microbiome analysis frameworks. BMC Bioinformatics 2019; 20: 581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bolyen E, Rideout JR, Dillon MR et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol 2019; 37: 852–857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Callahan BJ, McMurdie PJ, Rosen MJ et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods 2016;13: 581–583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Nilsson RH, Larsson K-H, Taylor AFS et al. The UNITE database for molecular identification of fungi: handling dark taxa and parallel taxonomic classifications. Nucleic Acids Res 2018; 47: D259–D264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Abraham A, Pedregosa F, Eickenberg M et al. Machine learning for neuroimaging with scikit-learn. Front Neuroinform 2014; 8: 14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.R Core Team. R: A language and environment for statistical computing, R Foundation for Statistical Computing, Vienna, Austria, 2019, https://www.R-project.org/. [Google Scholar]
- 24.Davis NM, Proctor DM, Holmes SP et al. Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome 2018; 6: 226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Vázquez-Baeza Y, Pirrung M, Gonzalez A et al. EMPeror: a tool for visualizing high-throughput microbial community data. GigaScience 2013; 2: 16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Vázquez-Baeza Y, Gonzalez A, Smarr L et al. Bringing the dynamic microbiome to life with animations. Cell Host Microbe 2017; 21: 7–10. [DOI] [PubMed] [Google Scholar]
- 27.Oksanen J, Blanchet FG, Friendly M et al. vegan: Community Ecology Package. R package version 2.5-6. 2019. Available at: https://CRAN.R-project.org/package=vegan. [Google Scholar]
- 28.Segata N, Izard J, Waldron L et al. Metagenomic biomarker discovery and explanation. Genome Biol 2011; 12: R60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol 1995; 57: 289–300. [Google Scholar]
- 30.Liu X, Tan J, Yang H et al. Characterization of skin microbiome in tinea pedis. Indian J Microbiol 2019; 59: 422–427. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Wang R, Song Y, Du M et al. Skin microbiome changes in patients with interdigital tinea pedis. Br J Dermatol 2018; 179: 965–968. [DOI] [PubMed] [Google Scholar]
- 32.Older CE, Diesel AB, Lawhon SD et al. The feline cutaneous and oral microbiota are influenced by breed and environment. PLoS One 2019; 14: e0220463. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Sample metadata file including information on all cats included in the study.
Statistical analysis of bacterial alpha and beta diversity metrics.
Figure S1. Relative abundance of bacterial taxa on the skin of all cats included in the study.
DLH control, domestic long hair controls; naïve Persian, Group 1 Persian cats; resistant Persian, Group 2 Persian cats. In each case of Group 3, severe dermatophytosis is identified by sample name so that lesion and general swabs from the same cat can be compared.
Figure S2. Bacterial taxa that are more differentially abundant on Persians with severe dermatophytosis than on all healthy control cats.
Plot created using LEfSe (LDA > 2.5, P < 0.01).
Figure S3. Bacterial taxa that are more differentially abundant on healthy Persian cats compared with healthy control cats.
Plot created using LEfSe (LDA > 2.5, P < 0.01).