Background
As widely demonstrated, genetic and environmental factors interplay in the development of autoimmune diseases, such as systemic lupus erythematosus (SLE) [
1]. Several environmental factors have been implicated in the different pathological conditions, and great emphasis has been placed on the role of infection [
2].
In recent years there has been growing interest in the possible role of the microbiome in the development and course of disease. Of note, the gut microbiome has been widely investigated in autoimmune diseases, such as type 1 diabetes, inflammatory bowel diseases, rheumatoid arthritis (RA), and spondyloarthropathies [
3]. Conversely, few data are available on the skin microbiome and the relationship with autoimmune diseases.
Staphylococcus aureus (SA) is a commensal microorganism and represents one of the most important components of the human skin microbiome [
4]. SA is characterized by very heterogeneous pathogenic features, ranging from minor and self-limiting skin infections, such as impetigo, folliculitis, and furuncles, to invasive and life-threatening diseases, such as septic arthritis, osteomyelitis, meningitis, septicemia and staphylococcal toxic shock syndrome [
5].
The anterior naris is the most frequent carriage site for SA, due to specific anatomical and biochemical characteristics facilitating the persistence of SA [
6]. Data from the National Health and Nutrition Examination Survey 2001–2002 described a frequency up to 30 % of SA colonization in the general population in the USA [
7]. A large cohort constituted by nine European countries described a frequency of SA carriage of 21.6 %, with lower values in the older population [
5]. In the great majority of the cases, this colonization is intermittent and only in 20 % of cases is persistent [
6].
Very few studies have evaluated the prevalence of SA nasal carriage in patients affected by autoimmune diseases and its association with the specific disease phenotype. In 1996 Tabarya and colleagues described a prevalence of SA carriers of 50 % among patients with RA from a cohort of 88 individuals, compared with 33 % identified in a healthy control population [
8]. More recently, in 2005 Bassetti et al. did not identify any significant difference in SA carrier prevalence, between RA and a group of patients without, enrolled as controls (34.5 % versus 32.5 %). Moreover, concomitant treatment with tumor necrosis factor (TNF) antagonists and methotrexate appeared to be the only independent factor associated with carriage of nasal SA (OR 3.24) [
9]. Conversely, a relationship between SA and granulomatosis with polyangiitis (GPA) has been identified, suggesting the role of this specific bacterium in disease development and relapse [
10]. Moreover, the study conducted by Laudien and colleagues demonstrated a significantly higher rate of SA nasal carriage in patients with GPA compared to a cohort of patients with RA and staff members (72.0 %, 46 %, and 58 %, respectively). Notably, the risk of relapse was higher in patients with GPA who had evidence of nasal SA [
11].
Starting from the lack of studies in patients with SLE, in the present analysis we aimed at assessing the prevalence of SA nasal carriers in a monocentric SLE cohort and evaluated the association between SA nasal colonization and disease phenotype.
Discussion
In the present study, for the first time we evaluated the prevalence of SA nasal colonization in a cohort of patients affected by SLE. Despite a similar frequency of SA being observed in patients and a healthy control group, the SA colonization in patients with SLE was associated with a specific disease phenotype, characterized by renal and skin involvement, and a higher prevalence of a broad spectrum of autoantibodies.
SLE is an autoimmune disease characterized by very heterogeneous autoantibody production and clinical manifestations [
1]. Infectious agents seem to play an important role in disease pathogenesis due to their ability to activate B-cell-mediated and T-cell-mediated autoimmune responses leading to the production of autoantibodies [
2,
20]. More recently, the interferon (IFN) signature was shown as an additional mechanism involved in the disease, confirming the possible role of infection [
21]. Moreover, innate immunity through pathogen-associated molecular patterns and Toll-like receptors (TLRs) may also contribute to disease development [
22].
The microbiome is a novel and intriguing concept that has been reported to be involved in the pathogenesis of several autoimmune diseases. Most studies focused on the gut microbiome; nonetheless, the skin microbiome could play a role in human autoimmune conditions. Changes in the skin microbiome seem to influence the disease course through the modulation of the cutaneous immune system. Moreover, each individual has a unique skin microbiome influenced by pH, salinity, sebum content of the topographical body region, and by intrinsic (e.g. genotype, age, and sex) and extrinsic individual-dependent factors (e.g., occupation, geographical location, smoking, sun exposure, and use of antibiotics or cosmetics) [
4]. SA is one of the components of the skin microbiome that could potentially colonize all body surfaces including the gut and anterior nares [
5,
6].
It seems to induce an inflammatory response by exposing staphylococcal superantigen, molecular mimicry, causing increased TLR signaling in leukocytes and inducing neutrophil extracellular traps [
6,
10]. Moreover, SA seems to be able to interact with endothelial, B and T cells, leading to the activation of neutrophils and the production of pro-inflammatory cytokines [
23]. Age, sex, and ethnicity influence SA colonization. Indeed, significantly higher colonization rates are identified in younger people, in men, and in white populations [
6]. Several pathological conditions, such as diabetes mellitus, hemodialytic treatment, end-stage liver disease, obesity, and HIV infection, predispose to nasal SA [
6].
To our knowledge, no data on SA colonization in patients with SLE are available so far. Only two studies have been performed in patients with RA. In both analyses, a higher prevalence of SA carriers was found compared to our SLE cohort, and also the prevalence in their control groups was higher than that observed in our HC [
9,
11]. Thus, we could reasonably exclude that disease
per se or differences in the immunosuppressive treatment could justify such an observation. It is more likely that the conditions in which the swab was obtained or differences between the studied populations in age, sex, and ethnicity, could justify such a result.
Intriguingly, we observed that the presence of SA was associated with a specific phenotype of SLE, namely characterized by high frequency of different autoantibodies (anti-dsDNA, anti-Sm, anti-SSA, anti-SSB, and anti-RNP) and by a more frequent renal and skin involvement. Indeed, SA carriers had a significantly higher prevalence of anti-SSA and anti-SSB antibodies, which are known to be associated with cutaneous involvement. In addition, anti-dsDNA and anti-Sm antibodies, which are associated with renal involvement, were present in SA+ patients [
24].
It could be hypothesized that SA carriage status induces the production of autoantibodies. No data are available on this topic in patients with SLE. It is possible that SA, by stimulating the type-I IFN pathway, leads to increased production of autoantibodies and therefore to the development of the previously mentioned clinical manifestations [
25]. Indeed, dendritic cells (DC), able to recognize pathogens, to activate T cells and to product the type I IFNs, could take up SA through an endocytic mechanism, resulting in the activation of TLR9 signaling. As known, TLR9 is localized at the endosomal level and is involved in autoimmune responses to DNA-associated proteins [
22,
26]. Moreover, SA is able, independently of TLR2, to activate human plasmacytoid DC and subsequent IFN-α secretion [
27,
28]. The study published by Viau in 2005, evaluating the effects of repeated injection of SA protein A on the (NZBxNZW) F (1) mice lupus model, demonstrated the reduction of anti-DNA IgG production and of proteinuria. The authors suggested that this result could be related to the depression of B-cell response induced by the protein A [
29]. As widely demonstrated, SA could interact with both the innate and adaptive immune responses by different virulence factors, among these, the SA protein A, characterized by the presence of immunoglobulin-binding domains, able to bind the Fab of VH3 idiotype antibodies [
30,
31].
It should be considered that SA could influence immune response also by the activation of T cells. Data from the literature demonstrated that Staphylococcal enterotoxins (SEs) could bind directly the major histocompatibility complex (MHC) class II of antigen-presenting cells. The presentation to T cells leads to massive non-specific activation of the immune system, by stimulating around 20 % of the naïve T-cell population [
32].
The higher prevalence of SA nasal colonization was not associated with any treatment except glucocorticoids. Data from the literature demonstrate that cortisol status can influence susceptibility to infection and that glucocorticoids seem able to reduce the release of pro-inflammatory cytokines, the activation of anti-inflammatory genes, the upregulation of cell adhesion molecules and the downregulation of neutrophil adhesion molecules, thus facilitating the onset of an infective process [
33]. Furthermore, Van den Akker and colleagues suggested an association between SA carrier status and polymorphisms of the glucocorticoid receptor gene. Those subjects homozygous for the haplotype 3, which is associated with relative glucocorticoid resistance, had 68 % decreased risk of persistent nasal carriage. Conversely, the genotype combination of the haplotype 5 and the haplotype 1 allele was associated with 80 % increased risk of persistent nasal carriage [
34]. It would also be of interest to assess the genotype of the glucocorticoid receptor gene in a population of patients with SLE.
Moreover, it should be considered that glucocorticoid treatment could determine skin abnormalities. In particular, permeability barrier homeostasis and stratum corneum integrity and cohesion could be modified by glucocorticoid treatment, also when performed for a brief period. This could be related to inhibition of the synthesis of epidermal lipid exerted by glucocorticoids [
35].
We could not find any significant difference in disease activity nor in the number of flares between SA+ and SA- patients with SLE and only a trend towards higher frequency of persistently active disease was identified in SA+ patients. This result could be related to the single SA assessment performed in the study. SA colonization can vary during the time and it would be of interest to link SA colonization with the occurrence of disease flares. On the other hand, the identification of persistent carriers should be evaluated in relation to the development of more severe chronic damage. In this view, we evaluated disease activity modifications after treatment. The lack of a significant improvement in the disease course evaluated 12 months after the successful eradication with muropicin could be due to the follow up being too short, or to the weak influence of nasal SA on disease activity. Finally, we evaluated only the anterior nares, despite possible colonization in different body sites. However, the primary reservoir for SA in humans is the anterior nares, probably due to the high affinity for nasal epithelial cells. Moreover, nasal secretions also seem to improve the bacterium adherence, in particular, thanks to clumping factor B and iron-regulated surface determinant A. Therefore, in this study we decided to evaluate the colonization of SA exclusively in the anterior nares [
36].
A limitation of the present study is the SA identification by a classical morphological evaluation, without molecular characterization. Moreover, the cross-sectional design and the evaluation of nasal SA at a single time point did not allow the exclusion of transient carriers.