Temporal dynamics of skin microbiota and immune correlates in psoriasis patients receiving systemic treatment

Su-Hsun Liu, Yu-Huei Huang, Hao-Jui Weng, Tsen-Fang Tsai, Huang-Yu Yang, Leslie Y Chen, Yen-Ling Chiu, Hsiao-Yun Yu, Yi-Chieh Chiu, Chao-Yu Ng, Ya-Ching Chang, Chung-Yee R Hui, Yhu-Chering Huang
doi: https://doi.org/10.1101/2023.11.27.23298999

ABSTRACT

Background How skin microbiota in psoriasis patients responded to systematic therapeutics remained unknown.

Objectives To profile temporal shifts in transcriptionally active skin microbiota in psoriasis patients receiving systemic therapies.

Methods We prospectively enrolled 61 psoriasis patients and 29 skin-healthy controls in 2015-2019. Using RNA-based 16S rRNA gene sequencing, we analyzed 969 samples from skin lesions and compared microbial abundance and diversity by therapeutic classes and disease severity.

Results Lesional microbiota in patients on conventional systemics and TNF-α inhibitor was different in relative abundances in Firmicutes (7.83% higher, adjusted P < 0.001) and Proteobacteria (6.98% lower, adjusted P < 0.01) from that in patients on anti-interleukin monoclonal antibodies (anti-ILAb) at baseline. The only difference during treatment was a 1.47% lower abundance in Bacteroides associated with nonbiologics use (adjusted P < 0.01). We identified no indicator taxa by disease severity at baseline yet noticed that a minor relative reduction in Corynebacterium sp. was associated with clinical responses to treatment.

Compared to anti-ILAb, TNF-α inhibitor and nonbiologics were associated with -0.21 lower Shannon Diversity (adjusted P < 0.01) and 0.03 higher Shannon Evenness (adjusted P < 0.01). Results of ordinated principal coordinates analysis revealed that, lesional microbiota from patients of these 3 therapeutic groups was compositionally distinct. Our work also demonstrated concurrent changes in clonal shifts in systemic T cell receptor clonotypes that were associated with systemic use of biologics.

Conclusions Community abundances and diversities of skin microbiota may be useful in distinguishing skin microbiota from patients receiving different systemic therapeutics. Specifically, use of anti-ILAb and TNF-α inhibitor was associated with sample-wise microbial abundances and diversities, but not richness, over time. These findings highlighted the potential utility of skin microbiota as biomarkers for personalized treatment plans in patients with moderate-to-severe psoriasis.

INTRODUCTION

Psoriasis vulgaris (psoriasis) is a chronic immune disease that manifests as well-defined scaly skin plaques, which may be associated with arthritis and systemic comorbidities [1]. Both innate immunity and T-cell response contribute to the disease pathogenesis, constituting major therapeutic targets for effective systemic therapies [2]. Recently, skin microbiota has emerged as novel candidate biomarkers for disease management [3, 4].

Despite the proven clinical effectiveness of anti-psoriatic therapies, few studies have examined therapy-specific influences on lesional microbiota longitudinally [5]. A recently-updated Cochrane review showed that biologics, including TNF-α inhibitor and anti-ILAb, were more effective in achieving clinical remission as measured by PASI 90 (90% reduction in Psoriasis Area and Severity Index relative to the baseline) than the small molecules and conventional systemics in the trial setting [2]. However, whether and how conventional therapies, TNF-α inhibitor, and anti-ILAb differentially alter the host microbiota in a non-trial setting remains unstudied. Besides, whether clinical responses were associated with any shifts in lesional microbiota has not been well-characterized [3]. As such, we sought to delineate relationships among systemic therapies, treatment responses, and characteristics of lesional microbiota in psoriasis patients via serial sampling. Particularly, to reduce environmental influences on the low biomass of skin samples [6] and to provide an in-depth insight into the relevant responses to a newly-prescribed regimens [7, 8], we sought to profile transcriptionally-active skin microbiota using bacterial 16S rRNA transcript sequencing. We also explored temporal changes in T cell receptor (TCR) repertoire diversity during treatment for a better understanding of treatment effects on the host immunity.

MATERIALS AND METHODS

Subject enrollment

Between May 2015 and November 2019, we enrolled and followed adults (aged 20 years or older) with psoriasis at Dermatology Outpatient Clinics of two tertiary hospitals (Chang Gung Memorial Hospital and National Taiwan University Hospital) in Taiwan. Patients were eligible for study entry when initiating new systemic therapy; majority of the washout period was > 4 weeks. We excluded those with a recent exposure to topical or systemic antibiotics, immunosuppressants, or chemotherapies within 12 weeks prior to study enrollment. In 2016- 2019, we also recruited skin-healthy adults who did not have diagnoses of chronic inflammatory skin disorders but either visited the same clinics for mole excision or alopecia treatment, or were care givers of the index patients. Center-specific Institution Review Boards reviewed and approved the study protocols, separately (CGMH 103-6837B, 104-9746B, 201702308B0C501; NTUH 201809061RINB).

Data and sample collection

After obtaining written informed consent, well-trained research assistants conducted in-person interviews to collect sociodemographic information, medical history, and behavioral habits.

Patients also reported quality of life and anxiety/depressive symptoms using Dermatology Life Quality Index (DLQI)[9] questionnaire and Center for Epidemiologic Studies Depression Scale (CESD-S)[10] at baseline and follow-up visits. According to European S3-Guidelines [11], we grouped patients by non-biologics (acitretin, cyclosporin, methotrexate); anti-interleukin monoclonal antibodies (anti-ILAb: ustekinumab, secukinumab, ixekizumab, and guselkumab); tumor necrosis factor-α antagonists (TNF-α inhibitor: adalimumab, golimumab, and etanercept). Treatment responses were absolute PASI scores and PASI 75 responses.

To profile skin microbiota from sequencing 16S rRNA gene transcripts, we used Tween20- moistened FLOQSwabs (Copan, Murrieta, CA, USA) to collect scrapes over an area of 2 x 2 cm2 on pre-specified surfaces with active lesions [12, 13], including bilateral scalp, post-auricular hairlines, extensor forearms, pretibial shafts, and palms; these sampling sites also applied to skin-healthy controls. To facilitate swab collection, we requested the study participants to refrain from bathing or body scrubbing at least 24 hours prior to a study visit. We also encouraged each participant to cover up sampling sites with cotton-made clothing to reduce environmental influences while in the study.

16S rRNA gene transcript isolation, amplification, sequencing, and processing

We performed RNA isolation by firstly disrupting bacterial cell walls in a chemical lysis buffer [7]. Reverse transcription of total RNA immediately followed the recovery of RNA samples using the SuperScript RT-PCR kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s manual. For each PCR sample, we confirmed the presence of amplicons by gel electrophoresis. We included 2-3 negative controls for each library preparation as suggested in the literature [14, 15], and mock community DNA (Zymo Research, Irvine, CA, USA) as positive controls [16].

We applied ‘mothur’ (version 1.43.0) to the raw sequence reads using a modified reference region for sequence alignment [17, 18] and clustered the filtered reads into operational taxonomy units (OTU) at 97% identify before subsequent statistical analysis and graphing in R (v 4.0.2) [19]. Using R packages ‘decontam’ and ‘phyloseq’[15, 20], we identified and removed potential contaminant taxa inferred from 63 negative control samples [21], singletons, and samples with low coverage (< 400 reads/ sample). In total, we included 969 samples (82.0% of sequenced samples) with a median of 43 577 reads per sample (interquartile range [IQR]: 33 987-55 972) in the following analysis.

TCR repertoire profiling

We collected peripheral blood samples in EDTA tubes from 9 anti-ILAb users before and 12 weeks after treatment began. We also collected 1 blood sample from 7 skin-healthy participant at baseline. Using flow cytometry, we sorted CD3+ T cells, from which RNA isolation was performed by employing RNeasy Mini Kit (Qiagen). Procedures of the library preparation included nested amplifications targeting the V-C regions of the complementarity-determining region (CDR3) of TCR using TCRαβ kits (iRepertoire, Huntsville, AL, USA).

Statistical Analysis

Patient data

We described and compared baseline participants’ characteristics by disease status and therapy class (patients only) using median tests for independent continuous variables and Chi-squared or Fisher’s exact test for categorical ones. We tested our hypothesis using multivariable quantile regression analysis with robust or bootstrapped (3 000 replicates) variance estimation to account for repeated measures. All longitudinal data analysis considered observations over the 12-15 weeks of follow-up period to allow for sufficient statistical power.

We used Stata-MP (Version 13.1) with a 2-tailed significance of 0.01 for comparative analysis [22], and employed R packages ‘ggplot2’ and ‘vegan’ for graphics production [23, 24].

Microbiota data

We calculated α-diversity measures of richness (observed number of unique OTUs, Chao1 index), Shannon diversity (ShannonD) and Shannon evenness (ShannonE) indices for quantifying within-sample variation. Based on non-metric and linear R2 values, we chose Bray-Curtis dissimilarity index to compare between-sample distance matrices [25]. We performed hypothesis testing via permutational multivariate analysis of variance (PERMANOVA, 10 000 permutations) [26]. We also repeated relevant analysis using read-based rarefied datasets; the results did not alter our conclusions.

TCR profiling

We performed descriptive analysis by sample and patient characteristics; compared differences in the normalized usage of V-and J-segment of the CDR3 gene, and in diversity metrics of CDR3 clonotypes using bootstrapped quantile regression models to estimate the empirical variance (3 000 replicates) in Stata-MP (Version 13.1) [22]. We applied a 2-tailed significance level of 0.05 for exploratory analysis on gene expression and TCR sequence data.

Data Availability

We have deposited raw sequences of 16S rRNA transcripts and TCR transcripts to the Short Read Archive (SRA) database under the BioProject accession number PRJNA662768.

RESULTS

Study population

We prospectively enrolled and followed 119 participants for 12 weeks or longer between 2015 and 2019. Ten patients received an extended observation for additional 6 months while receiving maintenance biologics therapy (Fig. S1). We included 743 and 226 quality swabs from 61 psoriasis patients and 29 controls who returned at least once within 16 weeks in the current analysis (follow-up rate: 87% and 83%). The patient population comprised mainly of men (87%) (median age: 43 years) with an average disease duration of 23 years (IQR: 19-33). There were no statistical differences in patient characteristics associated with therapeutic classes at study entry. In total, 79% of patients had moderate-to-severe psoriasis; all patients of the 3 therapeutic groups began with comparable disease severity at study entry (Table 1). Patients excluded based on low-quality sequencing samples (N=5) were comparable to those included in the current analysis (Table S1).

View this table:
Table 1. Characteristics of psoriasis patients and skin-healthy control participants at baseline, 2015-2019

At 12-15 weeks after treatment began, disease activity in all patients declined significantly, with the median PASI score decreasing from 15 (IQR: 11-20) to 4 (IQR: 1-8) (adjusted P for trend < 0.001). Patients in each therapeutic class had comparable trends in disease activities over time; there were no significant differential group effects (all P for interaction tests > 0.01). Self-reported DLQI also decreased substantially from 9 (IQR: 4-15) at baseline to 2 (0-8, adjusted P for trend < 0.001) whereas self-reported anxiety symptoms as quantified by CESD scores did not vary over time (Table S2).

Normal versus lesional microbiota

Microbial abundances

At study entry, phylum abundances in normal skin microbiota were statistically comparable to those in lesional microbiota with or without accounting for participants’ age, sex, absence of prior treatment, and year of enrollment (Table 2). A relative dominance of Proteobacteria in lesional microbiota by 3.35% (95%CI: 1.44%, 5.25%) emerged over time (adjusted P < 0.001); simultaneous comparisons across disease status and sampling sites revealed similar results (adjusted P=0.002, Fig. S2a). Likewise, we examined abundances of the most common 6 genera as suggested in the literature [27] (Fig. S2b), and found no evidence for indicator taxa between lesional and normal skin (Table S3).

View this table:
Table 2. Differences in phylum-level relative abundance between lesional and healthy skin microbiota at baseline and during 12-15 weeks of systemic treatment in adult psoriasis patients overall, by sex and absence of prior treatment, 2016-2019a

Despite an over-representation of men in the patient group, results of sex-specific comparisons between lesional and normal microbiota were alike at baseline but not longitudinally, during which a relative over-abundance of Proteobacteria was significant in samples from male patients by 4.13% (95%CI: 1.88%, 6.38%) as compared to normal skin (Table 2). Formal tests for disease status-by-sex interaction were nonsignificant. Additional stratified analysis by the absence of prior treatment found no significant subgroup effects either.

Microbial alpha-& beta-diversity

When compared to healthy skin, psoriatic lesions had resembling microbial richness and diversity before and during treatment; only the degree of community evenness (all unique taxa showing equal abundances) was lower in lesional than normal microbiota (adjusted P=0.01; Table S4). Correspondingly, results of Bray-Curtis distance-based PCoA showed an overall community similarity between the two groups (PERMANOVA P > 0.05, Fig. S2c).

Lesional microbiota, clinical treatment, and responses

Microbial abundances

Before initiating new therapies, lesional microbiota of patients using non-biologics had 7.83% more Firmicutes (95%CI: 4.26%, 10.5%) than those using anti-ILAb (adjusted P< 0.001) whereas TNF-α inhibitor was associated with a 6.98% lower abundance (95%CI: -12.2%, -1.73%) in Proteobacteria than anti-ILAb (Table 3). Over time, psoriatic lesions in non-biologics users showed a relative 1.47% reduction (95%CI: -2.49%, -0.45%) in Bacteroidetes as compared to anti-ILAb users while losing the dominance of Firmicutes (adjusted P> 0.01, Fig. S3). Abundances in the four major phyla did not differ by initial disease severity or patients’ clinical responses as indicated by PASI 75 at follow-up (Table 3).

View this table:
Table 3. Adjusted associations between phylum-level relative abundances of lesional microbiota and therapeutic class or clinical outcome at baseline and over time in psoriasis patients, 2015-2019a

Stratified analysis by prior treatment history revealed a higher abundance of Firmicutes in nonbiologics-associated samples from patients with previous treatment history (6.93%, 95%CI: 5.84%, 8.02%) but a lack thereof in those who were treatment-naïve. Similarly, subgroup analysis by sex showed a significant reduction in Proteobacteria (-17.5%, 95%CI: -29.0%, -6.1%) in women using TNF-α inhibitor as compared to anti-ILAb but no statistical differences by therapy group in men.

At genus levels, Staphylococcus sp. was more abundant in samples from the non-biologics group than the anti-ILAb group (adjusted P< 0.001) at baseline. There were no significant differential abundances by baseline disease severity (PASI > 10). Longitudinally, classes of systemic therapies were not associated with any indicator genera considered. Instead, lesions of PASI 75 responders had a 0.38% decrease in Corynebacterium sp. but nonsignificant differences in other genera relative to poor responders (Table S5).

Microbial alpha-& beta-diversity

Before treatment began, α-diversity measures of skin microbiota appeared comparable among samples from the three therapeutic groups, or by initial degrees of disease severity (Table 4). During the follow-up period, we noted a significant decrease (-0.21; 95%CI: -0.37, -0.06) and increase (0.03; 95%CI: 0.01, 0.06) in Shannon D and Shannon E of the TNF-α inhibitor group and the non-biologics group relative to the reference group, separately (Fig. 1a). Results of distance-based analysis of variance suggested significant pair-wise dissimilarities between samples from healthy skin microbiota and lesional microbiota of patients receiving anti-ILAb or TNF-α inhibitor (both permuted P < 0.001, Fig. 1b). On the other hand, samples from non-biologics users and healthy controls were compositionally similar as shown in unconstrained PCoA (Fig. S4a). Again, sample-wise α-diversity measures in PASI 75 responders and poor responders were indistinguishable during treatment (Table 4), wherefore these results were compatible to results of β-diversity analysis (Fig. S4b).

Fig 1a.
Fig 1a.

Alpha diversity measures for quantifying within-sample microbial richness (OTU, Chao1), diversity (Shannon D), and compositional evenness (Shannon E) in samples from healthy controls and psoriasis patients in the 3 therapeutic classes. Over the study period, significant differences included a relative decrease and increase in Shannon D and Shannon E in samples associated with TNF-α Inhibitor and nonbiologics as compared to anti-ILAb (adjusted P<0.01).

Fig 1b.
Fig 1b.

Principal coordinates analysis of genus-level community data from patients of different therapeutic class. Based on 104 permutations, pairwise comparisons of distances between group centroids (solid circles) revealed significant community dissimilarities between TNF-α Inhibitor and anti-ILAb patients (P <0.001) or nonbiologics patients (P < 0.001) as well as between non-biologics and anti-ILAb users (P <0.01).

View this table:
Table 4. Adjusted associations between genus-level alpha-diversity measures of lesional microbiota and therapeutic class or clinical assessment in adult psoriasis patients based at baseline and longitudinally, 2015-2019a

Clonal shifts of TCR during treatment

In TCR analysis, we included sequencing results of 8 samples with 4157 (amino acid sequence) unique CDR3 clonotypes (IQR: 2282 - 5642) identified per sample. The average D50 value was 2.8% (IQR: 2.1%-3.2%), suggesting that the most abundant 131 clonotypes accounted for 50% of the total sample reads; there were no statistical differences in D50 values by disease status at baseline (median difference: 0.9%, 95%CI: -4.5%, 6.3%). Shannon Diversity index of TCR repertoire in psoriasis patients varied little after treatment (Table S6). We also showed that, among 10 out of 13 J-and 48 V-segment genes, treatment-associated changes were significant for segment hTRBJ1-6 (+0.9%, 95%CI: 0.4%, 1.4%), hTRBJ2-6 (+1.4%, 95%CI: 0.3%, 2.5%), and hTRBV3-1 (-0.7%, 95%CI: -1.4%, -0.03%) in after-versus before-treatment comparisons among psoriasis patients (Fig. 2).

Fig 2.
Fig 2.

Boxplots of selected hTRB-V and -J segments with significant differential usage in peripheral CD3+ T-cells between skin-healthy controls and psoriasis patients at baseline (T0) and those at 12-15 weeks after anti-ILAb treatment began (T2). Normalized usage was the proportion of each segment being identified in all CDR3 clonotypes of a given sample; the copy number of each clonotype was not taken into account in order to avoid the influence of clonal expansion on estimating sample-specific segment usage. Usage of segment hTRBJ1-2 (difference: 2.2%, 95%CI: 1.8%, 2.6%) and hTRBV9 (difference: 4.3%, 95%CI: 1.0%, 7.6%) were significantly higher in healthy controls than in psoriasis patients before treatment. In contrast, the normalized usage of hTRBJ2-7 (difference: 11.5%, 95%CI: 2.1%, 20.9%), hTRBV11-2 (difference: 1.5%, 95%CI: 1.1%, 1.9%), and hTRBV27 (difference: 1.3%, 95%CI: 0.8%, 1.8%) were statistically higher in psoriasis patients than in controls. NS, non-significant; *: P< .05; **: P< .01; ***: P< .001.

DISCUSSION

In this prospective cohort study, we showed that, transcriptionally-active skin microbiota in psoriasis patients of 3 major therapeutic classes displayed distinctive community abundances and diversities before and after treatment. Over the past decade, the well-debated observations that TNF-α inhibitor could increase patients’ risks for severe infections [28, 29], such as reactivation of latent tuberculosis [30, 31], and the less-debated low risks of similar adverse events in patients receiving IL-17A or IL-12/23p40 inhibitor [32, 33], have led to our hypothesized differential effects of systemic therapies on lesional microbiota. Since both biologics lead to a common, downregulated secretion of cytokines, chemokines, and AMPs, such β-defensins and S100 proteins [12], we anticipated starker contrasts between lesional microbiota in biologics and nonbiologics users than between samples from patients receiving different biologics. The observation that only results of β-analysis conformed with our conjectures, alluded the multi-dimensional complexity in quantifying (dis)similarities among groups of surveyed human microbiota [21].

Besides specific immune perturbations inflicted on lesional microbiota by different therapies, the initial community structure may partially account for the group differences observed longitudinally. For instance, non-biologics users had the highest relative abundance of Firmicutes (25.8%, IQR: 22.1%, 29.4%) at baseline (anti-ILAb: 20.3%; TNF-α inhibitor: 21.8%), yet their lesional microbiota also followed a significant decline temporally (in Firmicutes, - 0.06% per day, 95%CI: -0.09%, -0.02%). Rates of changes in the other 2 treatment groups remained nonsignificant, leading to a consensus post-treatment state across the comparison groups. Contrary to the other two groups, samples from anti-ILAb patients showed evident stability in community abundances, richness, and diversity over the treatment period, rendering the group as an optimal reference for within-patient comparisons.

Our findings that the relative ranking in normalized usage of TCR segments TRBV3-1, V9, and V11-2 examined in the current study, agreed with an early investigation by Tsuda et al. in evaluating effects of ustekinumab on the presence or absence of TRB V-segments in psoriasis patients [34]. These transcriptional shifts in selective CDR3 clonotypes (31% of 13 J-segments, 13% of 48 V-segments) of circulating T cells in patients receiving IL-12/-23 or IL-17A monoclonal antibodies suggested host systemic responses in T cell differentiation after 12-week exposure to blockade against the Th1/Th17 axis. Whether these transitions in TCR clonality resulted from direct depletion of associated lineages or indirect responses to a re-shaping of skin microbiota, was beyond the study’s scope.

Conversely, the notion that characteristics of lesional microbiota were not linked to pre-treatment disease severity, was contrary to a previously-reported negative correlation found between baseline absolute PASI score and Shannon Diversity. In a pre-post study on effects of UVB treatment, Assarsson and colleagues found that there were no similarly positive relationships between disease severity and microbial diversity after treatment [35]. Although Corynebacterium sp. has been implicated in IL-1β production and expansion of γδ17 cells in inflammatory skin conditions of model animals [36], our finding that a minor decrease in Corynebacterium sp. Was associated with clinical responses during treatment, warrants additional research to confirm its potential clinical utility.

Although the observation that normal skin harbored fewer Proteobacteria than lesional microbiota, was inconsistent with previous findings [13, 37], such cross-study inconsistency has been well-recognized in the literature [27]. Potential sources of disagreements may include heterogeneous patient characteristics under different investigations; choices of analytic approaches; intrinsic discrepancies in DNA-vs. RNA-based profiling of bacterial 16S rRNA genes of skin microbiota. Particularly, Belheouane and others compared the murine ear skin microbiota of standing and active microbial communities using 16S rRNA genes and transcripts sequence data, respectively, and concluded that, RNA-was better than DNA-based profiling in reducing environmental contaminations [6]. Given that comparisons in microbial abundances and a diversity measures between DNA-and RNA-based methods may not always agree with each other even within the same study [8, 14, 38, 39], between-study heterogeneities can be expected; nonetheless, comparative results obtained by either technique shall remain valid within each study [14].

Limitations

A few limitations are noteworthy in the current analysis. First, as an observational study, we did not specify group sizes for each treatment group a priori; restrictions on funding also precluded extended recruitment. The impacts of unequal group sizes and the broad-ranging grouping of patients receiving conventional systemics or anti-ILAb, were uncertainties and possibly attenuated associations. As such, results of comparisons by therapeutic class required careful interpretation and generalization. According to a population-based patient registry in Sweden, female patients generally had a lower PASI score than male patients across the age groups [40]. Since the current study sought to enroll patients with moderate-to-severe psoriasis, the predominance of men in our patient population was expected and the sex-adjusted comparison results within the patient population remained relevant and valid, particularly to those receiving systemic treatment. Although proportions of patients receiving different systemic therapeutics did not differ substantially across the recruitment waves, microbial structure, and diversity of lesional microbiota may be conditioned by patients’ prior treatment, of which we used a proxy variable (the absence of prior treatment) for statistical adjustment; the reported associations were indeed weighted averages of a heterogeneous patient population. Lastly, we recognized the challenges of low biomass even on skin lesions and attempted to sequence cDNA sequences of 16S amplicons instead of the corresponding DNA sequences directly. How this unique approach could have resulted in biased representations of certain Phyla has not been well characterized [41]. Future studies that collect detailed information on serial medications, may provide novel insights on the long-term trajectories of therapeutic effects on the lesional microbiota.

CONCLUSIONS

The current study showed that, classes of systemic therapies were independently associated with characteristics of lesional microbiota in adult psoriasis patients enrolled in a non-trial setting.

These findings may support for the clinical utility of transcriptionally-active skin microbiota as biomarkers for tailoring systemic therapies to patients with moderate to severe psoriasis in addition to the absolute PASI scores.

CONFLICTS OF INTEREST

The authors declared no conflicts of interests.

FUNDING SOURCES

The study was supported by Ministry of Science and Technology, Taiwan [MOST 104-2314-B- 182-002, 105-2628-B-182-006, 107-2314-B-418-015 to S Liu]; Chang Gung Medical Foundation [CMRPG3E1971-72 and BMRPE29 to S Liu]; Far Eastern Memorial Hospital [FMRP107-2314B418015 to S Liu]. The funders had no role in preparing or submitting the current work.

AUTHOR CONTRIBUTIONS

Dr. S Liu had full access to all the data in the study and takes responsibility for the integrity of the data and accuracy of the reported findings.

Concept and design: S Liu, YH Huang, T Tsai

Acquisition, analysis, and interpretation of data: S Liu, YH Huang, H Weng, T Tsai, H Yu, YC Chiu

Drafting of the manuscript: S Liu, H Weng

Critical revision of the manuscript for important intellectual content: S Liu, H Weng, T Tsai, YH Huang

Statistical analysis: S Liu

Administrative, technical, or material support: HY Yang, L Chen, YL Chiu, HY Yu, YC Chiu, C Ng, Y Chang, C Hui, YC Huang

Supervision: S Liu

Data Availability

All data produced in the present study are available upon reasonable request to the authors.

Footnotes

  • Funding sources: The current study was supported by Ministry of Science and Technology, Taiwan [MOST 104-2314-B-182-002, 105-2628-B-182-006, 107-2314-B-418-015 to S Liu]; Chang Gung Medical Foundation [CMRPG3E1971-72 and BMRPE29 to S Liu]; Far Eastern Memorial Hospital [FMRP107-2314B418015 to S Liu]. The funders had no role in preparing or submitting the current work.

REFERENCES

  1. [1].↵
    Armstrong AW, Read C. Pathophysiology, Clinical Presentation, and Treatment of Psoriasis: A Review. JAMA : the journal of the American Medical Association. 2020;323; 1945–1960.
    OpenUrlCrossRefPubMed
  2. [2].↵
    Sbidian E, Chaimani A, Afach S, Doney L, Dressler C, Hua C, et al. Systemic pharmacological treatments for chronic plaque psoriasis: a network meta-analysis. Cochrane database of systematic reviews (Online). 2020;1; Cd011535.
  3. [3].↵
    Niemeyer-van der Kolk T, van der Wall HEC, Balmforth C, Van Doorn MBA, Rissmann R. A systematic literature review of the human skin microbiome as biomarker for dermatological drug development. Br J Clin Pharmacol. 2018;84; 2178–2193.
    OpenUrl
  4. [4].↵
    Harkins CP, Kong HH, Segre JA. Manipulating the Human Microbiome to Manage Disease. JAMA : the journal of the American Medical Association. 2019.
  5. [5].↵
    Loesche MA, Farahi K, Capone K, Fakharzadeh S, Blauvelt A, Duffin KC, et al. Longitudinal Study of the Psoriasis-Associated Skin Microbiome during Therapy with Ustekinumab in a Randomized Phase 3b Clinical Trial. The Journal of investigative dermatology. 2018;138; 1973–1981.
    OpenUrl
  6. [6].↵
    Belheouane M, Vallier M, Čepić A, Chung CJ, Ibrahim S, Baines JF. Assessing similarities and disparities in the skin microbiota between wild and laboratory populations of house mice. The ISME journal. 2020.
  7. [7].↵
    Yuan S, Cohen DB, Ravel J, Abdo Z, Forney LJ. Evaluation of methods for the extraction and purification of DNA from the human microbiome. PloS one. 2012;7; e33865.
    OpenUrlCrossRefPubMed
  8. [8].↵
    Knecht H, Neulinger SC, Heinsen FA, Knecht C, Schilhabel A, Schmitz RA, et al. Effects of beta-lactam antibiotics and fluoroquinolones on human gut microbiota in relation to Clostridium difficile associated diarrhea. PloS one. 2014;9; e89417.
    OpenUrlCrossRefPubMed
  9. [9].↵
    Finlay AY, Khan GK. Dermatology Life Quality Index (DLQI)--a simple practical measure for routine clinical use. Clin Exp Dermatol. 1994;19; 210–216.
    OpenUrlCrossRefPubMedWeb of Science
  10. [10].↵
    Radloff LS. The CES-D Scale: A Sef-Report Depression Scale for Research in the General Population. Applied Psychological Measurement. 1977;1; 385–401.
    OpenUrlCrossRef
  11. [11].↵
    Nast A, Jacobs A, Rosumeck S, Werner RN. Methods Report: European S3-Guidelines on the systemic treatment of psoriasis vulgaris--update 2015--EDF in cooperation with EADV and IPC. J Eur Acad Dermatol Venereol. 2015;29; e1–22.
    OpenUrlCrossRefPubMed
  12. [12].↵
    Nestle FO, Kaplan DH, Barker J. Mechanism of Disease: Psoriasis. New England Journal of Medicine. 2009;361; 496–509.
    OpenUrlCrossRefPubMedWeb of Science
  13. [13].↵
    Alekseyenko AV, Perez-Perez GI, De Souza A, Strober B, Gao Z, Bihan M, et al. Community differentiation of the cutaneous microbiota in psoriasis. Microbiome. 2013;1; 31.
  14. [14].↵
    von Rosenvinge EC, Song Y, White JR, Maddox C, Blanchard T, Fricke WF. Immune status, antibiotic medication and pH are associated with changes in the stomach fluid microbiota. The ISME journal. 2013.
  15. [15].↵
    Davis NM, Proctor DM, Holmes SP, Relman DA, Callahan BJ. Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data. Microbiome. 2018;6; 226.
  16. [16].↵
    Kong HH, Andersson B, Clavel T, Common JE, Jackson SA, Olson ND, et al. Performing Skin Microbiome Research: A Method to the Madness. The Journal of investigative dermatology. 2017.
  17. [17].↵
    Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. Development of a dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the MiSeq Illumina sequencing platform. Applied and environmental microbiology. 2013;79; 5112–5120.
    OpenUrlAbstract/FREE Full Text
  18. [18].↵
    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 research. 2013;41; D590–596.
    OpenUrlCrossRefPubMedWeb of Science
  19. [19].↵
    Team RC. R: A language and environment for statistical computing. 3.6.2 ed. Vienna, Austria. R Foundation for Statistical Computing, 2020.
  20. [20].↵
    McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PloS one. 2013;8; e61217.
    OpenUrlCrossRefPubMed
  21. [21].↵
    Knight R, Vrbanac A, Taylor BC, Aksenov A, Callewaert C, Debelius J, et al. Best practices for analysing microbiomes. Nature reviews Microbiology. 2018;16; 410–422.
    OpenUrlPubMed
  22. [22].↵
    StataCorp LP. Stata Statistical Software: Release 13. 13.1 ed. College Station, TX, 2013.
  23. [23].↵
    Wickham H. ggplot2: Elegant Graphics for Data Analysis. New York, USA. Springer-Verlag. 2009.
  24. [24].↵
    Oksanen J, Guillaume B, Friendly M, Kindt R, Legendre P, McGlinn D, et al. vegan: Community Ecology Package. R package version 2.5–6., 2019.
    OpenUrl
  25. [25].↵
    Magurran AE, McGill BJ. Biodiversity: frontiers in measurement and assessment. United States. Oxford University Press. 2011.
  26. [26].↵
    Anderson MJ. Distance-based tests for homogeneity of multivariate dispersions. Biometrics. 2006;62; 245–253.
    OpenUrlCrossRefPubMedWeb of Science
  27. [27].↵
    Byrd AL, Belkaid Y, Segre JA. The human skin microbiome. Nature reviews Microbiology. 2018;16; 143–155.
    OpenUrlCrossRefPubMed
  28. [28].↵
    Grijalva CG, Chen L, Delzell E, Baddley JW, Beukelman T, Winthrop KL, et al. Initiation of tumor necrosis factor-alpha antagonists and the risk of hospitalization for infection in patients with autoimmune diseases. JAMA : the journal of the American Medical Association. 2011;306; 2331–2339.
    OpenUrlCrossRefPubMedWeb of Science
  29. [29].↵
    Dobry AS, Quesenberry CP, Ray GT, Geier JL, Asgari MM. Serious infections among a large cohort of subjects with systemically treated psoriasis. J Am Acad Dermatol. 2017;77; 838–844.
    OpenUrl
  30. [30].↵
    Sanchez-Moya AI, Garcia-Doval I, Carretero G, Sanchez-Carazo J, Ferrandiz C, Herrera Ceballos E, et al. Latent tuberculosis infection and active tuberculosis in patients with psoriasis: a study on the incidence of tuberculosis and the prevalence of latent tuberculosis disease in patients with moderate-severe psoriasis in Spain. BIOBADADERM registry. J Eur Acad Dermatol Venereol. 2013;27; 1366–1374.
    OpenUrlCrossRefPubMed
  31. [31].↵
    Garcia-Doval I, Cohen AD, Cazzaniga S, Feldhamer I, Addis A, Carretero G, et al. Risk of serious infections, cutaneous bacterial infections, and granulomatous infections in patients with psoriasis treated with anti-tumor necrosis factor agents versus classic therapies: Prospective meta-analysis of Psonet registries. J Am Acad Dermatol. 2017;76; 299–308.e216.
    OpenUrl
  32. [32].↵
    Gómez-García F, Epstein D, Isla-Tejera B, Lorente A, Vélez García-Nieto A, Ruano J. Short-term efficacy and safety of new biological agents targeting the interleukin-23-T helper 17 pathway for moderate-to-severe plaque psoriasis: a systematic review and network meta-analysis. The British journal of dermatology. 2017;176; 594–603.
    OpenUrl
  33. [33].↵
    Fowler E, Ghamrawi RI, Ghiam N, Liao W, Wu JJ. Risk of tuberculosis reactivation during interleukin-17 inhibitor therapy for psoriasis: a systematic review. J Eur Acad Dermatol Venereol. 2020;34; 1449–1456.
    OpenUrl
  34. [34].↵
    Tsuda K, Yamanaka K, Kondo M, Matsubara K, Sasaki R, Tomimoto H, et al. Ustekinumab improves psoriasis without altering T cell cytokine production, differentiation, and T cell receptor repertoire diversity. PloS one. 2012;7; e51819.
    OpenUrlCrossRefPubMed
  35. [35].↵
    Assarsson M, Duvetorp A, Dienus O, Soderman J, Seifert O. Significant Changes in the Skin Microbiome in Patients with Chronic Plaque Psoriasis after Treatment with Narrowband Ultraviolet B. Acta dermato-venereologica. 2018;98; 428–436.
    OpenUrl
  36. [36].↵
    Quan C, Chen XY, Li X, Xue F, Chen LH, Liu N, et al. Psoriatic lesions are characterized by higher bacterial load and imbalance between Cutibacterium and Corynebacterium. J Am Acad Dermatol. 2020;82; 955–961.
    OpenUrlPubMed
  37. [37].↵
    Fahlen A, Engstrand L, Baker BS, Powles A, Fry L. Comparison of bacterial microbiota in skin biopsies from normal and psoriatic skin. Arch Dermatol Res. 2012;304; 15–22.
    OpenUrlCrossRefPubMed
  38. [38].↵
    Rausch P, Basic M, Batra A, Bischoff SC, Blaut M, Clavel T, et al. Analysis of factors contributing to variation in the C57BL/6J fecal microbiota across German animal facilities. International journal of medical microbiology : IJMM. 2016;306; 343–355.
    OpenUrl
  39. [39].↵
    Twomey KB, Alston M, An SQ, O’Connell OJ, McCarthy Y, Swarbreck D, et al. Microbiota and metabolite profiling reveal specific alterations in bacterial community structure and environment in the cystic fibrosis airway during exacerbation. PloS one. 2013;8; e82432.
    OpenUrlCrossRefPubMed
  40. [40].↵
    Hagg D, Sundstrom A, Eriksson M, Schmitt-Egenolf M. Severity of Psoriasis Differs Between Men and Women: A Study of the Clinical Outcome Measure Psoriasis Area and Severity Index (PASI) in 5438 Swedish Register Patients. Am J Clin Dermatol. 2017;18; 583–590.
    OpenUrl
  41. [41].↵
    Chen Y, Knight R, Gallo RL. Evolving approaches to profiling the microbiome in skin disease. Front Immunol. 2023 Apr 4;14:1151527. doi: 10.3389/fimmu.2023.1151527.
    OpenUrlCrossRef
Back to top
Posted November 27, 2023.
Download PDF

Supplementary Material

Data/Code
Email
Temporal dynamics of skin microbiota and immune correlates in psoriasis patients receiving systemic treatment
Su-Hsun Liu, Yu-Huei Huang, Hao-Jui Weng, Tsen-Fang Tsai, Huang-Yu Yang, Leslie Y Chen, Yen-Ling Chiu, Hsiao-Yun Yu, Yi-Chieh Chiu, Chao-Yu Ng, Ya-Ching Chang, Chung-Yee R Hui, Yhu-Chering Huang
medRxiv 2023.11.27.23298999; doi: https://doi.org/10.1101/2023.11.27.23298999
Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area