Advancing Gastrointestinal Microbiota Research in Systemic Sclerosis

Dysbiosis is a feature of patients with systemic sclerosis (SSc). While a causal relationship between the gastrointestinal (GI) microbiota and SSc pathogenesis has not been established, alterations in the GI microbiota are appreciated early in the SSc disease course. Moreover, recent research has illuminated specific microbial signatures that define SSc phenotypes. This review summarizes new research on the GI microbiome in SSc with a focus on technical advancements and the emerging study of the GI metabolome. This review also addresses diverse modalities for manipulating the GI microbiome with the hope of developing preventative treatment strategies to avert progression of SSc.

Key points

•

Alterations of the gastrointestinal (GI) microbiota (ie, dysbiosis) occur relatively early in systemic sclerosis (SSc) disease course.
•

Specific bacterial species are associated with clinical features of SSc, including GI symptoms and interstitial lung disease.
•

An integrative analysis of GI microbiome-metabolome will further our knowledge of disease mechanisms underlying specific SSc phenotypes.
•

Modulating of the GI microbiome via targeted therapies, supplements, and fecal microbial transplant is promising, but these interventions are still under investigation.
•

Nutritional and mind–body interventions offer hope for improving homeostasis of the GI microbiome in SSc.

Abbreviations

ACHIM	anaerobic cultivated human intestinal microbiota
FMT	fecal microbial transplant
FODMAP	fermentable oligosaccharides, disaccharides, monosaccharides, and polyols
GI	gastrointestinal
HC	healthy control
IBD	inflammatory bowel disease
IBS	irritable bowel syndrome
ILD	interstitial lung disease
LC	liquid chromatography
MS	mass spectrometry
OUT	operational taxonomic unit
SCFA	short-chain fatty acid
SIBO	small intestinal bacterial overgrowth
SSc	systemic sclerosis
VEDOSS	very early diagnosis of SSc
WGS	whole genome shotgun sequencing

Introduction

A wealth of evidence supports an association between alterations in gastrointestinal (GI) microbiota and human health and disease. Between 10 and 100 trillion microorganisms comprise the human microbiota, among which bacteria are the most abundant with a density of 10 ¹¹ to 10 ¹² cells/ml. Within the GI tract, over 1000 bacterial species play a critical role in host homeostasis, from nutrition, immune function, and metabolism to defense against pathogenic organisms. Major shifts in the GI microbiota membership and function, which are linked to disease states (ie, dysbiosis), are commonly observed in patients with autoimmune diseases. For example, numerous studies have demonstrated that patients with various autoimmune diseases exhibit reduced diversity of GI microbiota and enrichment of pathobiont species.

Dysbiosis is a recognized feature of systemic sclerosis (SSc), ^, an autoimmune disease with diverse clinical manifestations and high morbidity and mortality. For example, studies have demonstrated that patients with early SSc exhibit alterations in GI microbiota compared with unaffected controls. Furthermore, the abundance of certain bacterial species is associated with the severity of organ involvement in SSc. While dysbiosis is common among patients with SSc, the significance of these changes remains unclear. For instance, it is unknown whether dysbiosis is a cause or consequence of immune dysregulation in SSc, or both. Moreover, while the GI microbiota is fairly stable through adulthood, and the most common time period for developing SSc is between the age of 30 and 50 years; a number of external factors can perturb the balance of the GI microbiota, including dietary alterations, environmental shifts, medications, lifestyle changes, and infections.

Unlike other autoimmune rheumatic diseases, SSc has a profound predilection for the GI tract. Over 90% of patients with SSc have signs and symptoms of GI involvement. Since there are currently no disease-modifying therapies for the GI manifestations of SSc, this clinical dimension of SSc typically progresses (worsens) throughout the disease course, representing a major source of morbidity. Patients with SSc-GI manifestations report worse health-related quality of life overtime than those without SSc-GI manifestations, ^, and severe malabsorption is an independent predictor of mortality in patients with SSc. ^, Interestingly, manipulation of the GI microbiota with broad-spectrum antibiotic therapy offers many patients temporary relief from certain GI symptoms, including distension and bloating. This fact is what led our group to embark on the first exploration of the GI microbiota in SSc.

Since the first study of the GI microbiota in SSc was published in 2016, our understanding of the role of GI microbiota in this disease state has evolved. Considering the considerable phenotypic heterogeneity of SSc, it seems increasingly less plausible that a single microbiota signature will define all patients with SSc. Instead, new research points to species-specific organ system associations. This review provides an in-depth coverage of original research conducted over the last 6 years on the GI microbiota in SSc. By comparing and contrasting the methodological approaches to characterize the GI microbiota in SSc, this review offers expert insight into sequence analytical techniques and their capacity to provide sufficient species-level resolution for describing interindividual variations among patients living with SSc. This review also describes emerging research on functional analyses and the application of machine-learning platforms that may propel discovery in this area. Finally, this review summarizes the theoretic basis for personalized approaches to modify the gastrointestinal tract (GI) microbiota in SSc with the hope of identifying preventative treatment strategies for this condition.

Microbiota analysis

Sequencing Approaches

Since the majority of bacterial species cannot be readily cultured, high-throughput sequencing facilitates culture-independent analysis. Initial research on the GI microbiota in SSc relied on low-resolution 16S ribosomal RNA (rRNA) amplicon sequence analytics. In this method, a 16S rRNA region undergoes amplification by polymerase chain reaction (PCR) with primers that recognize the highly conserved regions of the gene. While this approach provides important compositional information on microbial community diversity, its output is limited to phyla or genus-level data. This is due to the fact that this method is based on putative associations between the 16S rRNA gene with taxa defined as an operational taxonomic unit (OTU). Due to the existence of numerous bacterial strains, OTUs are less precise at the species level.

Whole genome shotgun sequencing (WGS) is an alternative approach to 16S rRNA amplicon sequencing ^, ( Table 1 ). This method uses sequencing with random primers to sequence overlapping regions of a genome, as opposed to just one specific region of DNA. A major advantage of WGS method is that taxa can be more precisely defined at the species level. While WGS is more costly than 16S sequencing, studies have demonstrated that WGS identifies significantly more bacterial species per read than the 16S method. Moreover, unlike 16S sequencing, WGS can identify nonbacterial organisms, including viruses, fungi, and protozoa, which may contribute to the SSc disease state. As described further below, most of literature on the GI microbiota in SSc employed the 16S method; however, new research uses the WGS approach.

Table 1

Key differences in 16S rRNA amplicon sequence method and whole genome method of characterizing the gastrointestinal microbiota

	16S Sequencing	Whole Genome Sequencing
Methodological Differences	Sequences only a single region of the bacterial genome	Sequences broad regions of the genome
Taxa resolution	Poor species-level resolution; Increased risk of species misclassifications	Excellent species-level resolution with lower misclassifications
Organisms identified	Bacteria	Bacteria, Viruses, Protozoa, Fungi
Functional information	Not provided	Provided
Archived data	More archived data for reference	Less archived data for reference as this is a newer method
Cost	Less costly	More costly

Study Designs

Study design is an important consideration in SSc-GI microbiota research. Early studies investigating the GI microbiota in SSc were insufficiently powered to perform subgroup analyses or create multivariable models that adjusted for confounders ( Fig. 1 ). The majority of studies were also cross-sectional, with the exception of one small study that described changes in microbiota over the course of 1 year in patients with SSc. These design issues have delayed our progress in understanding the role of the GI microbiota in the pathogenesis of SSc in contrast to other disease states, such as inflammatory bowel disease (IBD). For example, a study of patients with ulcerative colitis with ileal pouch-anal anastomosis who underwent serial endoscopy and sampling of host mucosa and pouch microbiomes identified pathobionts that were observed prior to endoscopic changes or clinical symptoms. ^, Longitudinal studies that characterize changes in species abundance relative to the evolution of organ involvement/progression may transform our understanding of the role of the GI microbiota in SSc.

Studies aiming to understand alterations in the GI microbiota between SSc and unaffected controls have largely matched groups by age and gender. However, given the complexity of external factors that shape the GI microbiota from the time of birth (eg, birth order, mode of delivery, breastfeeding, antibiotics predelivery) to the time of sampling, it may be impossible to identify the perfect control population for an SSc study. Some have proposed to use household members as controls to help adjust for environmental impacts (eg, pets, diet); however, studies have demonstrated that healthy first-degree relatives of patients with IBD exhibit dysbiosis that may predispose to disease or subclinical inflammation. Thus, selecting the ideal control population is an area of ongoing controversary in SSc microbiota research.

Sampling: Endoscopic versus Fecal Samples

Most of the SSc GI microbiota studies have utilized fecal samples, which represent a composite of distal GI microbiota and are not representative of region-specific GI microbiota. Endoscopic evaluations can facilitate the acquisition of region-specific samples. Our first study demonstrated microbial community differences between the cecum and sigmoid colon of patients with SSc who underwent colonoscopy with lavage sampling. However, unlike IBD, where colonoscopies are routinely done to evaluate disease activity, they are not performed as standard of care in SSc, rendering it a costly research procedure. Furthermore, patients with SSc are often reluctant to undergo colonoscopy, particularly when they suffer from chronic constipation. For these reasons, collecting fecal samples represents a more cost-effective and less invasive approach for sampling the GI microbiota with the caveat that region-specific information is lost.

Statistical Analyses

A central goal of SSc microbiota analyses is to identify key species that explain differences between groups of samples (eg, diffuse cutaneous disease vs limited cutaneous disease). The risk of type I error is high in these analyses given the high number of bacterial species or genera under investigation. Most studies employ specific methods to correct for multiple hypothesis testing. The use of the traditional Bonferroni method is considered too conservative and may lead to missed findings. Various other multiple-comparison procedures have been utilized to control the false discovery rate, including a widely used procedure introduced by Benjamini and Hochberg. This procedure identifies the expected fraction of non-differentially abundant taxa among taxa deemed significant (the discoveries) and adjusts individual P -values.

Another important analysis issue in SSc microbiota research is sample filtering. Typically, samples are filtered to retain species with at least 10% to 25% non-zero counts. This is particularly important in cross-cohort integrative analyses that are aiming to identify consistent associations between species and disease features within different cohorts. Collaborating with experienced microbiome statisticians can improve the rigor of this research.

Microbiota alterations in early systemic sclerosis

Due to the abundance of residing lymphocytes within the GI tract, the GI tract itself is considered an immunologic organ. ^, While few studies have investigated the direct relationship between the immune system and the GI microbiota, a number of studies have demonstrated that patients with SSc exhibit alterations in GI microbial composition compared with healthy controls (HCs; Fig. 2 ). In the first microbiome study performed in SSc, Volkmann and colleagues demonstrated decreased commensal genera and increased pathobiont genera in the lower GI tract of SSc compared with HCs using colonic lavage specimens. Soon after this original study was published, Andreasson and colleagues demonstrated that severe dysbiosis was observed in patients with SSc, particularly in patients with esophageal dysmotility. Several other investigations have confirmed that dysbiosis occurs in SSc ; however, these early studies did not investigate confounding factors that can modify the microbial composition such as medications, disease duration, SSc subsets, or the geographic region of the study patients.

Early studies on the GI microbiome in SSc aimed to understand whether dysbiosis is associated with GI symptoms. Using the UCLA GIT 2.0 questionnaire as a tool to quantify the severity of GI dysfunction, low abundance of Bacteroides fragilis was associated with worse GIT scores (more GI symptoms). ^, Another study found that SSc patients with GI symptoms had alterations in the number and distribution of taxa with an increased relative abundance of Desulfovibrio . These formative studies helped shaped our initial understanding of how the GI microbiome was altered in SSc and whether it could be linked with clinical manifestations of SSc.

A fundamental question that arises from all of these studies is if dysbiosis is a consequence of autoimmunity and chronic inflammation of SSc or, whether dysbiosis itself perpetuates inflammation. While a bidirectional relationship between the GI microbiota and immune system likely exists, the study of patients with SSc in very early stages (ie, preclinical phases) of their disease could help further our insight into this relationship and ultimately improve our understanding of the molecular mechanisms that drive the pathogenesis of SSc.

For the early stages of SSc, we refer to the first phases of SSc with an already definite, but recent SSc diagnosis (<3 years), while a very early diagnosis of SSc (VEDOSS) is based on the 2011 European League Against Rheumatism Scleroderma Trial and Research proposed criteria where patients experience Raynaud’s phenomenon, the presence of puffy fingers and harbor antinuclear antibodies, as well as SSc-specific autoantibodies (anti-centromere antibody or anti-topoisomerase I) and SSc pattern on nail fold capillaroscopy as necessary features. Based on LeRoy and Medsger criteria, subjects who present with raynaud phenomenon (RP), SSc-specific antibodies, and/or SSc pattern at capillaroscopy without any features of fibrosis (puffy fingers included) are defined as very early/preclinical SSc.

To test the hypothesis that early dysbiosis could directly influence the development of autoimmunity and fibrosis, a preclinical study demonstrated that early life dysbiosis induced by antibiotic administration to mice exacerbated the development of skin and lung fibrosis. However, human studies in early and very early SSc are lacking. One study of 106 patients with early SSc (median disease duration of 2 years) from Sweden study showed that dysbiosis was present in this early stage of disease, with a notable increase in the abundance of the pathobiont genera Desulfovibrio . In another study, Natalello and colleagues investigated the GI microbiome in 19 patients SSc with early disease (mean disease duration of 2.3 years) and 29 patients with SSc with longer disease duration (mean disease duration 10.8 years). In this study, patients with early SSc exhibited greater richness in bacterial species and differences in beta diversity compared with those patients with longer standing disease. This study also demonstrated a relationship between the GI microbiota and the cutaneous subset of SSc, as well as body mass index.

To date, only one study, by Russo and colleagues, has evaluated the microbiota in patients meeting VEDOSS criteria. This study also investigated fecal metabolites and found a decrease of pro-tolerogenic bacterial strains and alterations in short-chain fatty acids in both VEDOSS and established patients with SSc when compared to HCs. In contrast to the prior study, which demonstrated microbiota differences between patients with early versus established SSc, the study evaluating patients with VEDOSS demonstrated no differences in the extent of dysbiosis between patients with very early SSc and those with established disease. While the reason for this difference is unclear, future studies are needed to understand when dysbiosis occurs during the SSc disease course and how it evolves overtime. Future studies are also needed to understand how individual microbiota signatures predict progression and distinct clinical phenotypes of SSc.

Microbiota correlates of systemic sclerosis features

Microbiota research in SSc has transitioned from attempting to discover a single microbiota signature that applies to all patients with SSc to understanding how specific taxa and their metabolic products associate with SSc features. There is also a growing appreciation that grouping microbes into categories based on their health-promoting versus disease-promoting facilities may lead to misclassifications of the individual microbes’ capacity to affect the host under different conditions. For example, Bacteroides , which was historically deemed a commensal genus, increased the incidence of colitis in mice whose mothers had received peripartum antibiotics. Therefore, individual microbes can transition from commensals to pathobionts in certain conditions that alter their functional capacities.

Gastrointestinal Tract Involvement

As described earlier, GI involvement occurs in most patients with SSc. One common GI manifestation of SSc is small intestinal bacterial overgrowth (SIBO). In a systematic review and meta-analysis, the prevalence of SIBO was 10 fold higher in the patients with SSc (N = 1112) compared with controls (N = 335). Treatment of SIBO is largely based on antibiotic therapy, and rifaximin is often the drug of choice. ^, Despite the fact that modulation of the microbiota through antibiotic therapy ameliorates SSc-GI symptoms, few studies have endeavored to understand the relationship between the microbiota and SSc-GI involvement. ^, One early study found that SSc patients with moderate/severe total GI symptoms had reduced abundance of B fragilis and increased abundance of Fusobacterium compared to those with none/mild symptoms. This study measured GI symptoms using the UCLA GIT 2.0, a valid questionnaire for assessing GI symptom burden in SSc. Patients in this study underwent serial assessment of their microbiome and GI symptoms every 3 months for 1 year, and low abundance of Bacteroides was associated with increased GI symptoms overtime.

However, in this longitudinal study, GIT 2.0 scores did not change significantly over the course of the 1 year. This finding may reflect stable GI disease in this cohort; however, one cannot exclude the possibility that GIT 2.0 scores cannot distinguish disease activity from disease damage in SSc. Research is underway to define new GI disease activity measures in SSc, which will undoubtedly accelerate our progress in identifying microbial biomarkers of SSc-GI involvement that correlate with disease activity.

Additional studies have identified bacterial genera and species associated with specific GI symptoms, such as constipation, bloating/distension, fecal incontinence, and malnutrition. However, the findings of these individual studies require replication in other cohorts. Research is also underway to understand the impact of diet on SSc-GI symptoms and the microbiota, as dietary manipulation can cause immediate shifts in GI microbial composition without the long-term safety concerns associated with repeated courses of antibiotic therapy, including antibiotic resistance and worsening dysbiosis.

Interstitial Lung Disease

The gut-lung axis refers to the ability of the GI microbiota to influence the course and outcome of underlying lung disease, and vice versa. Gut dysbiosis is associated with pulmonary diseases, including asthma and chronic obstructive pulmonary disease. ^, Moreover, accumulating evidence links the Western diet with shifts in the microbiota and predisposition to inflammatory respiratory diseases.

The role of the GI microbiota in interstitial lung disease (ILD) is an evolving area of research ( Fig. 3 ). In a murine model of hypersensitivity pneumonitis, Bacteroidetes phylum was enriched in streptomycin-treated mice during the perinatal period. In SSc, patients with ILD were found to have higher fecal calprotectin levels than patients without ILD. In another study, which included 2 geographically distinct SSc cohorts (the United States and Sweden), patients with ILD had marked differences in beta diversity compared with those without ILD. Beta diversity represents the degree of difference in community membership or structure between 2 samples.