Predictive Metagenomic Analysis Reveals a Role of Cutaneous Dysbiosis in the Development of Hidradenitis Suppurativa
2019; Elsevier BV; Volume: 140; Issue: 7 Linguagem: Inglês
10.1016/j.jid.2019.11.011
ISSN1523-1747
AutoresHans Christian Ring, Jonathan Thorsen, A.‐H.R. Jørgensen, Lene Bay, Thomas Bjarnsholt, Kurt Fuursted, Simon Francis Thomsen, Gregor B. E. Jemec,
Tópico(s)Anorectal Disease Treatments and Outcomes
ResumoWe read the recently published manuscript by Schneider et al., 2019Schneider A.M. Cook L.C. Zhan X. Banerjee K. Cong Z. Imamura-Kawasawa Y. et al.Loss of skin microbial diversity and alteration of bacterial metabolic function in Hidradenitis suppurativa [e-pub ahead of print].J Invest Dermatol. 2020; 140: 716-720Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar with great interest. The authors investigated the bacterial microbiome in patients with hidradenitis suppurativa (HS) and healthy controls using next-generation sequencing (NGS) of 16S rRNA marker genes. Overall, the authors demonstrated several factors that point to a potential involvement of the microbiome in the pathogenesis of HS. Indeed, the authors provide additional evidence for the hypothesis that the reduced presence of sebaceous glands and the subsequent reduced presence of Propionibacteria spp. in HS skin may contribute to the cutaneous dysbiosis. Moreover, the authors also found an abundance of Peptoniphilus spp. and Porphyromonas spp. in HS skin. Interestingly, the recent NGS studies on the HS microbiome appear to support these recent findings from Schneider et al., 2019Schneider A.M. Cook L.C. Zhan X. Banerjee K. Cong Z. Imamura-Kawasawa Y. et al.Loss of skin microbial diversity and alteration of bacterial metabolic function in Hidradenitis suppurativa [e-pub ahead of print].J Invest Dermatol. 2020; 140: 716-720Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar (Guet-Revillet et al., 2017Guet-Revillet H. Jais J.P. Ungeheuer M.N. Coignard-Biehler H. Duchatelet S. Delage M. et al.The microbiological landscape of anaerobic infections in Hidradenitis suppurativa: a prospective metagenomic study.Clin Infect Dis. 2017; 65: 282-291Crossref PubMed Scopus (66) Google Scholar; Nikolakis et al., 2017Nikolakis G. Liakou A.I. Bonovas S. Seltmann H. Bonitsis N. Join-Lambert O. et al.Bacterial colonization in hidradenitis suppurativa/acne inversa: A cross-sectional study of 50 patients and review of the literature.Acta Derm Venereol. 2017; 97: 493-498Crossref PubMed Scopus (20) Google Scholar, Ring et al., 2017aRing H.C. Bay L. Kallenbach K. Miller I.M. Prens E. Saunte D.M. et al.Normal skin microbiota is altered in pre-clinical hidradenitis suppurativa.Acta Derm Venereol. 2017; 97: 208-213Crossref PubMed Scopus (60) Google Scholar, Ring et al., 2017cRing H.C. Thorsen J. Saunte D.M. Lilje B. Bay L. Riis P.T. et al.The follicular skin microbiome in patients with hidradenitis suppurativa and healthy controls.JAMA Dermatol. 2017; 153: 897-905Crossref PubMed Scopus (136) Google Scholar). In particular, striking similarities exist between our NGS case-control study from 2017 (Ring et al., 2017cRing H.C. Thorsen J. Saunte D.M. Lilje B. Bay L. Riis P.T. et al.The follicular skin microbiome in patients with hidradenitis suppurativa and healthy controls.JAMA Dermatol. 2017; 153: 897-905Crossref PubMed Scopus (136) Google Scholar) and the study from Schneider et al., 2019Schneider A.M. Cook L.C. Zhan X. Banerjee K. Cong Z. Imamura-Kawasawa Y. et al.Loss of skin microbial diversity and alteration of bacterial metabolic function in Hidradenitis suppurativa [e-pub ahead of print].J Invest Dermatol. 2020; 140: 716-720Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar. Schneider et al., 2019Schneider A.M. Cook L.C. Zhan X. Banerjee K. Cong Z. Imamura-Kawasawa Y. et al.Loss of skin microbial diversity and alteration of bacterial metabolic function in Hidradenitis suppurativa [e-pub ahead of print].J Invest Dermatol. 2020; 140: 716-720Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar also used a metagenomic prediction tool (PICRUSt, Phylogenetic Investigation of Communities by Reconstruction of Unobserved States [Langille et al., 2013Langille M.G. Zaneveld J. Caporaso J.G. McDonald D. Knights D. Reyes J.A. et al.Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences.Nat Biotechnol. 2013; 31: 814-821Crossref PubMed Scopus (4785) Google Scholar]). In our NGS case-control study, we recently also applied such a predictive metagenomic analysis to uncover and investigate potential metabolic profiles of HS and healthy controls. We performed the analysis on our previously published data from HS biopsies (lesional and non-lesional) and healthy controls (Ring et al., 2017cRing H.C. Thorsen J. Saunte D.M. Lilje B. Bay L. Riis P.T. et al.The follicular skin microbiome in patients with hidradenitis suppurativa and healthy controls.JAMA Dermatol. 2017; 153: 897-905Crossref PubMed Scopus (136) Google Scholar). Institutional approval or patient consent was not necessary for this additional analysis. We examined the group-wise taxonomic differences by computing predicted metagenomes using PICRUSt, which outputs predicted gene content in samples based on the 16S data by extrapolating the known genetic composition of identified taxa. We used FASTQ files from BIONs quality filtering as input for closed reference operational taxonomic unit picking in QIIME (Caporaso et al., 2010Caporaso J.G. Kuczynski J. Stombaugh J. Bittinger K. Bushman F.D. Costello E.K. et al.QIIME allows analysis of high-throughput community sequencing data.Nat Methods. 2010; 7: 335-336Crossref PubMed Scopus (21478) Google Scholar), based on Greengenes 13.5 (Second Genome, Inc., South San Francisco, CA). These operational taxonomic units were then normalized for copy number variation and fed into PICRUSt (Langille et al., 2013Langille M.G. Zaneveld J. Caporaso J.G. McDonald D. Knights D. Reyes J.A. et al.Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences.Nat Biotechnol. 2013; 31: 814-821Crossref PubMed Scopus (4785) Google Scholar) for metagenome prediction as described in the tutorial (http://picrust.github.io/picrust/tutorials/metagenome_prediction.html#metagenome-prediction-tutorial). The predicted gene orthologs were analyzed using principal component analysis and tested for systematic differences between groups with "adonis" from the R-package "vegan." Genes were tested between groups using linear models and Wilcoxon rank sum tests, after log-transforming relative gene abundances. The predicted metagenomes were analyzed with the HUMAnN2 software (https://bitbucket.org/biobakery/humann2/wiki/Home) using the legacy KEGG database as described in the tutorial for PICRUSt output. Pathways were tested between the groups using linear models and Wilcoxon rank sum tests after log-transforming relative pathway abundances. Overall, the PICRUSt-predicted metagenomes showed vast systematic differences between the three groups (P < 0.0001), primarily owing to the lesional samples (lesional vs non-lesional, F = 5.64, P < 0.0001; lesional vs healthy controls, F = 3.63, P = 0.002), whereas the non-lesional and healthy control samples were only moderately different (F = 2.04, P = 0.02). We found 420 differentially abundant genes between the lesional samples and healthy controls and 1,120 between the lesional and non-lesional. However, only eight genes were distinguished between the non-lesional and healthy controls. Furthermore, the associated functional pathways of the predicted metagenomes demonstrated significant differences between the lesional samples and two control groups, with 45 pathways significantly different between the lesional and healthy controls, 80 between lesional and non-lesional, and none between non-lesional and healthy controls. Intriguingly, 37 of these pathways were significant in both comparisons (Figure 1) among which functions, such as DNA replication, cell cycle, and mismatch repair, as well as the biosynthesis of peptiodglycan and ansamycins, were significantly associated with the lesional samples, and the metabolism of several amino acids was inversely associated with the lesional samples. The PICRUSt data identified cell growth and division, such as DNA replication, cell cycle-caulobacter, nucleotide excision repair and mismatch repair, highly associated with lesional samples. As suggested by Schneider et al., 2019Schneider A.M. Cook L.C. Zhan X. Banerjee K. Cong Z. Imamura-Kawasawa Y. et al.Loss of skin microbial diversity and alteration of bacterial metabolic function in Hidradenitis suppurativa [e-pub ahead of print].J Invest Dermatol. 2020; 140: 716-720Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar, this may indicate a higher proliferation and/or turnover of the local bacterial community in the lesional samples. In that perspective, it appears attractive to suggest that the abscesses in HS have developed owing to infectious processes. Moreover, the significant upregulation of biosynthesis of peptidoglycan in the lesional samples may be early signs of bacterial biofilm formation in the active nodules (Bucher et al., 2015Bucher T. Oppenheimer-Shaanan Y. Savidor A. Bloom-Ackermann Z. Kolodkin-Gal I. Disturbance of the bacterial cell wall specifically interferes with biofilm formation.Environ Microbiol Rep. 2015; 7: 990-1004Crossref PubMed Scopus (41) Google Scholar). However, a recent study investigating bacterial biofilm in early acute lesions of HS did not find signs of bacterial biofilms (Okoye et al., 2017Okoye G.A. Vlassova N. Olowoyeye O. Agostinho A. James G. Stewart P.S. et al.Bacterial biofilm in acute lesions of hidradenitis suppurativa.Br J Dermatol. 2017; 176: 241-243Crossref PubMed Scopus (12) Google Scholar). Furthermore, the significantly higher level of ansamycin biosynthesis in lesional samples may reflect previous treatment periods of rifampicin. It may thus be hypothesized whether the standard duration of the washout period (one month) is inadequate to clear any residues of rifampicins. Considering that the recent NGS studies in HS demonstrate a significant reduced presence of Propionibacterium spp. in HS skin, it appears attractive to suggest that this potential dysbiosis is related to a reduced presence of sebaceous glands in HS (Figure 2). The reduced number or the atrophic sebaceous glands in intertriginous HS regions may lead to a shift in the local microbial community of the hair follicles, owing to a reduced presence of sebum, which has also previously been demonstrated in HS (Jemec and Gniadecka, 1997Jemec G.B. Gniadecka M. Sebum excretion in hidradenitis suppurativa.Dermatology. 1997; 194: 325-328Crossref PubMed Scopus (48) Google Scholar). This may result in two different pathogenic events: (i) The impairment of antimicrobial properties of local commensal bacteria such as P. acnes or an increased expression of pathogenic factors (e.g., biofilm formation) in otherwise harmless bacteria (e.g., Peptoniphilus spp. and Porphyromonas spp.). The deposition of scattered intradermal corneocytes and widely scattered fragments of hair reported in the HS lesions may provide foreign bodies acting as a nidus and promoting the formation of biofilm by commensal bacteria, thereby potentially intensifying their pathogenic properties (Bjarnsholt et al., 2009Bjarnsholt T. Tolker-Nielsen T. Givskov M. Janssen M. Christensen L.H. Detection of bacteria by fluorescence in situ hybridization in culture-negative soft tissue filler lesions.Dermatol Surg. 2009; 35: 1620-1624Crossref PubMed Scopus (0) Google Scholar; Ring et al., 2017bRing H.C. Bay L. Nilsson M. Kallenbach K. Miller I.M. Saunte D.M. et al.Bacterial biofilm in chronic lesions of hidradenitis suppurativa.Br J Dermatol. 2017; 4: 993-1000Crossref Scopus (56) Google Scholar); and (ii) Locally aberrant immune responses, which may lead to the increased presence of proinflammatory cytokines. Importantly, several studies have demonstrated a dysregulated cytokine and antimicrobial peptides expression in HS (Hotz et al., 2016Hotz C. Boniotto M. Guguin A. Surenaud M. Jean-Louis F. Tisserand P. et al.Intrinsic defect in keratinocyte function leads to inflammation in Hidradenitis suppurativa.J Invest Dermatol. 2016; 136: 1768-1780Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar; Kelly et al., 2015Kelly G. Hughes R. McGarry T. van den Born M. Adamzik K. Fitzgerald R. et al.Dysregulated cytokine expression in lesional and nonlesional skin in Hidradenitis suppurativa.Br J Dermatol. 2015; 6: 1431-1439Crossref Scopus (137) Google Scholar, Moran et al., 2017Moran B. Sweeney C.M. Hughes R. Malara A. Kirthi S. Tobin A.M. et al.Hidradenitis suppurativa is characterized by dysregulation of the Th17:Treg cell axis, which is corrected by anti-TNF therapy.J Invest Dermatol. 2017; 137: 2389-2395Abstract Full Text Full Text PDF PubMed Scopus (84) Google Scholar). These recurrent findings are often suggested to be linked with the altered microbiota found in HS. Alternatively, the altered expression of commensal properties may trigger local aberrant immune responses. Overall, both events may hypothetically result in local inflammation of the hair follicle (dilation and subsequent rupture) causing inflamed nodules. In aggregate, further NGS studies are needed to explore the potential dynamic role of the HS microbiome over time. Moreover, the predicted functional differences that are illustrated merit further study in the form of full metagenomic shotgun sequencing of the microbiome in HS to further characterize potential mechanisms of microbial contributions to inflammation and pathogenesis. The sequencing dataset related to this article can be found at https://www.ebi.ac.uk/ena/data/view/PRJEB15266, hosted at the European Nucleotide Archive with the accession number PRJEB15266. Hans Christian Ring: https://orcid.org/0000-0002-6145-5549 Jonathan Thorsen: https://orcid.org/0000-0003-0200-0461 Astrid Helene Jørgensen: https://orcid.org/0000-0002-4256-116X Lene Bay: https://orcid.org/0000-0002-4164-5900 Thomas Bjarnsholt: https://orcid.org/0000-0002-8003-7414 Kurt Fuursted: https://orcid.org/0000-0002-3483-3145 Simon Francis Thomsen: https://orcid.org/0000-0002-4838-300X Gregor Borut Jemec: https://orcid.org/0000-0002-0712-2540 The authors state no conflict of interest. Conceptualization: HCR, JT, SF, GJ; Formal Analysis: HCR, JT, SF, GJ, TB, LB, KF, AHJ; Investigation: HCR, JT; Writing - Original Draft Preparation: HCR, JT, SF, GJ, TB, LB, KF, AHJ; Writing - Review and Editing: HCR, JT, SF, GJ, TB, LB, KF, AHJ Loss of Skin Microbial Diversity and Alteration of Bacterial Metabolic Function in Hidradenitis SuppurativaJournal of Investigative DermatologyVol. 140Issue 3PreviewHidradenitis suppurativa (HS) is a chronic scarring inflammatory skin disease affecting the pilosebaceous units of the axilla, inframammary folds, groin, and buttocks. Several previous microbiome studies consistently have shown that bacterial dysbiosis exists in HS compared with normal skin (Guet-Revillet et al., 2017; Nikolakis et al., 2017; Ring et al., 2017b; Ring et al., 2017a). The HS microbiome has been studied by invasive and noninvasive sampling of nonlesional and lesional skin using culture and next generation sequencing analyses. Full-Text PDF Open Archive
Referência(s)