Gut expression of CD71 and co-localisation with Dec-1 associated with high levels of serum SIgA and disease activity in SpA

Abstract

Spondyloarthritis (SpA) correlates with elevated serum secretory-IgA (sSIgA). Retrotranscytosis, mediated by CD71 and Dectin-1, is linked to high sSIgA levels. This research investigated retrotranscytosis in the gut of SpA patients and its relationship with sSIgA, ileocolonoscopic findings, and activity indices. A cohort of 82 was derived from 180 patients based on selection criteria; 41 consented to DCE colonoscopy assessment. Measurements included CD71, Dec-1, sSIgA, and disease scores. Samples exhibiting apical CD71, Dec-1, or both were analysed for receptor/SIgA interactions. Apical CD71 and Dec-1 were predominantly found in the ileum. Multivariate analyses indicated CD71’s association with heightened sSIgA (p = 0.05) and ASDAS-CRP (p = 0.036; OR 1.71; 95% CI 1.00–3.11), Dec-1’s correlation with ileal inflammation (p = 0.05; OR 3.06; 95% CI 1.00–10.01), and the CD71/Dec-1 relationship with ileal villi atrophy (p = 0.038; OR 4.24; 95% CI 1.00–18.88). Microscopic evaluations demonstrated interactions between CD71/SIgA but not between CD71/Dec-1 or Dec/SIgA. The localised expression of CD71 and its co-localisation with Dec-1 may contribute to systemic increases in serum SIgA and activity scores in SpA patients. The findings suggest that sSIgA levels may serve as a non-invasive biomarker for gastrointestinal involvement in SpA and that retrotranscytosis could be a key mechanism in the gut-joint axis of SpA.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-03532-4.

Introduction

Spondyloarthritis (SpA) is described as a group of chronic auto-inflammatory diseases characterised by shared clinical, radiographic, and immunogenetic features. One of the most important associations is with the histocompatibility molecule HLA-B*27, which seems to influence disease severity^1,2. This group comprises five subtypes, with ankylosing spondylitis (AS) as the proposed primary study model. However, other forms of these diseases, such as arthritis of inflammatory bowel disease (IBD), have gained significant interest in recent years³. The SpA subtype associated with IBD, besides affecting the axial and peripheral skeleton, can also present extra-articular manifestations, compromising ocular, intestinal, and cardiac function, the latter being shared across all SpA subtypes^4,5.

The literature shows a close relationship between joint inflammation in SpA and the chronicity of intestinal inflammation^6–9. De Keyser et al.¹⁰ demonstrated an association between histopathological changes typical of chronic intestinal disorders and joint inflammation. Years later, the author reports that more than 60% of AS patients may exhibit subclinical features of intestinal inflammation, which may influence disease activity and progression¹¹. Mielants et al. identified histologic inflammatory changes in the mucosa of the ileum and the ileocecal valve in SpA patients¹², which were also associated with joint inflammation^13,14. Likewise, Mielants et al. observed that intestinal inflammation in patients with AS is higher when peripheral arthritis is present^15,16. De Vos et al. also reported disease remission in patients when intestinal inflammation was reduced¹⁷.

Romero-Sanchez et al. showed a positive correlation between increased intestinal symptoms in patients diagnosed with SpA and autoantibodies anti-Saccharomyces cerevisiae (ASCA), anti-cytoplasm of the neutrophil (ANCA), and anti-transglutaminase (tTG). Additionally, they reported that HLAB27-positive patients have significantly more pronounced disease activity⁷. Arias et al. suggested that high serum levels of secretory immunoglobulin A (SIgA) in patients with peripheral SpA (p-SpA) and non-radiographical axial SpA (nr-axSpA) reflect a chronic microbial stimulation with higher concentrations in the early stages of the disease and a possible regulatory role in later stages⁹.

A few scholars currently endorse the immunopathological processes that link intestinal inflammation to the articular manifestations observed in spondyloarthritis (SpA). Van Bohemen et al.¹⁸ demonstrated that individuals with active spondyloarthritis exhibit elevated serum levels of SIgA against Yersinia spp., Shigella spp., and Klebsiella spp., implying a substantial role in disease exacerbation. Romero-Sanchez et al.⁷ and Salas-Cuestas et al.⁸ showed a strong association between the compound activity indices in SpA and serum concentrations of SIgA in theoretical models of linear regression and principal component analysis (PCA). The authors state that these SIgA titers were consistently higher than those found in healthy volunteers and even comparable to those found in patients with IBD, leaving the question of the role played by SIgA in the modulation of SpA activity^8,9.

SIgA is the principal antibody of the mucosal immune system. It arises from the polymerisation of multiple IgA monomers (pIgA), which is transported to the gut lumen via the polymeric immunoglobulin receptor (pIgR) through M cells in Peyer’s patches. This process is referred to as transcytosis¹⁹. Antigen/SIgA complexes can undergo retrograde/reverse transcytosis (retrotranscytosis) into the subepithelial dome, enabling intracellular neutralisation of pathogens and inflammatory products while aiding dendritic cell sampling of the intestinal microenvironment^20,21. The molecular mechanisms of retrotranscytosis through enterocytes are still largely unknown. However, it has been attributed to a predominant role in the modulation of immune responses and a critical role in the pathogenesis of gastrointestinal pathologies such as celiac disease^22,23.

Matysiak-Budnik et al.²³ developed a model elucidating the pathogenesis of celiac disease, highlighting the aberrant apical expression of transferrin receptor (TfR), also named CD71. This IgA receptor facilitates the retrotranscytosis of gliadin peptides via SIgA²³. CD71 functions as an intermediate affinity receptor for IgA, with its physiological expression localised to the basolateral surface of enterocytes; however, the authors propose that individuals with celiac disease exhibit enhanced and mislocalised CD71 expression at the apical surface²⁴. The authors demonstrated that this ectopic expression of CD71 assists the retrotranscytosis of SIgA immune complexes, which are not typically managed by conventional effector pathways, thereby activating pathogenic clones of T and B lymphocytes²³. In contrast, Rochereau et al.²⁵ underestimated the role of CD71 for retrotranscytosis, attributing this mechanism to the interaction of SIgA-Dectin-1, a B-glucan receptor, demonstrating how the blockade of Dectin-1 significantly inhibits the retrotranscytosis of this antibody, an effect not seen when CD71 was inhibited.

The investigation of CD71 in SpA has predominantly focused on its presence within lymphoid cell populations, with implications for other biological aspects in patients with PsA and AS^26,27. Ruutu et al.²⁸ and Mikulic et al.²⁹ agreed that activation of genes associated with susceptibility/progression to SpA in humans (IL23R, IL12B, IL17R, and CARD9, among others) is conditioned by the downstream activation of Dectin-1 when there is exposure to B-glucans. This evidence does not describe studies where the expression of these receptors was determined in the gut tissues of patients with SpA. The retrotranscytosis of SIgA has not been evaluated in SpA; however, this model could explain the clinical manifestations and the correlations observed between gastrointestinal compromise, increase in systemic inflammatory biomarkers, and the clinical disease activity indices.

This study exposes the first approach to the evaluation of SIgA reverse transcytosis in intestinal tissue biopsies by identifying CD71 and Dec-1 expressed on the apical side of the epithelial barrier of the ileum and colon, their interactions receptor/receptor and receptor/SIgA complexes. Instead, that was assessed the association of retrotranscytosis receptors with levels of serum SIgA, gastrointestinal symptoms, microscopic and macroscopic gut changes, systemic biomarkers of inflammation, and activity of disease scores in patients with SpA without IBD who present subclinical features of gastrointestinal involvement.

Results

Sociodemographic description and rheumatological status of the study population

This study examined 41 patients with intestinal involvement, averaging 44.6 ± 10.2 years, predominantly male (56.1%). Table 1 outlines the sociodemographic characteristics. In rheumatologic findings, 87.8% exhibited axial/peripheral involvement, with 90.2% reporting inflammatory lumbar pain, 17.1% dactylitis, 80.5% enthesitis, and 75.6% arthritis. CRP levels exceeded 0.5 mg/L in 61.0% of patients (mean 1.68 ± 2.35 mg/L). Functional and activity scores indicated high BASDAI and BASFI > 4.0 in 58.5% and 53.7% of patients, respectively, with over 70% exhibiting mild/moderate activity in ASDAS indices. Among patients, 63.4% were on biological treatment, 88.5% with anti-TNF alpha and 11.5% with anti-IL17 therapy.

Table 1.

Sociodemographic conditions of patients with SpA and evaluated by colonoscopy.

	SpA n = 41
	n	%
Age*	44.6 ± 10.2
BMI*	26.2 ± 3.1
Sex
Male	23	56.1
BMI > 25	29	70.7
HLA B27 positive	16	39.0
Current smoker	2	4.9
Former smoker	12	29.3
Passive smoker	9	22.0
Economy activity
Home	7	17.1
Independent	9	22.0
Employee	17	41.5
Retired	6	14.6
Student	2	4.9
Marital status
Married	23	56.1
Single	9	22.0
Common-law marriage	9	22.0
Education level
Basic	1	2.4
High school	12	29.3
Technician	5	12.2
College	23	56.1

*Results expressed in mean ± standard deviation.

Gastrointestinal features and ileocolonoscopic changes of SpA patients

Although all patients reported more than two gastrointestinal symptoms, only 31.7% showed positive faecal calprotectin levels (≥ 120 ng/mL). Within gastrointestinal symptoms, the most prevalent were pain or abdominal distension (85.4%), 63.4% of patients had diarrhoea for more than 4 weeks, and 97.6% of the patients affirmed they were omnivorous. 63.4% of patients reported some food intolerance, with dairy as the most frequent (37.8%), followed by grains (18.9%) and meat (13.5%). Please see Supplementary Table 1 (Table 2).

Table 2.

Binomial logistic regression analysis.

Reference	B (SE)	OR	95% CI
Apical expression of Dectin-1
Intercept	− 2.217 (0.571)*
Inflammatory changes of ileum mucous	1.119 (0.605)	3.061*	1.000–10.013
Diarrhea + BASDAI > 4	0.852 (0.557)	2.344	0.721–7.617
Apical expression of CD-71
Intercept	0.510 (0.414)*
Serum SIgA	− 0.002 (0.013)	0.998	0.972–1.024
ASDAS-CRP > 2.1	0.860 (0.254)	1.710*	1.001–3.114
Apical expression of Dectin-1 & CD71
Intercept	− 9.430 (3.877)*
Ileum villi Atrophy	1.445 (0.762)	4.241*	1.003–18.88
ESR	0.420 (0.190)	1.009*	1.000–1.111

Model 1 to Dectin-1. R² = 0.077 (Cox&Snell). 0.118 (Nagelkerke). LR X² = 72.13, p = 0.007. Wald p value: *p < 0.05.

Model 2 to CD71. R² = 0.093 (Cox&Snell). 0.132 (Nagelkerke). LR X² = 83.20, p = 0.05. Wald p value: *p < 0.05.

Model 3 to Dectin-1&CD71. R² = 0.132 (Cox&Snell). 0.237 (Nagelkerke). LR X² = 49.20, p = 0.014. Wald p value: *p < 0.05.

The ileocolonoscopic assessment revealed that 17.1% of patients exhibited rectal alterations, primarily characterised by vascular pattern loss. Macroscopic inflammatory changes were noted in 14.6% of patients, while 19.5% displayed features of microscopic inflammation. The sigmoid colon evaluation indicated that the predominant mucosal alteration was vascular pattern loss or erosions, occurring in 9.8% of cases. Histologically, 14.6% of sigmoid colons displayed cryptitis/atrophy. Additionally, 43.9% exhibited an inflammatory pattern, with chronic infiltrate being the most prevalent (26.8%). Mucosal alterations in the ileum occurred in 41.5% of patients, highlighting villus atrophy (36.6%), loss of vascular pattern (29.3%), ulcers (12.2%), and erosions (9.8%). Histological analysis indicated that 65.9% of patients had ileal mucosal changes, with 47.4% showing inflammatory infiltrate. Among these, 72.2% primarily exhibited a chronic pattern. Furthermore, alterations in ileal architecture were noted in 31.7% of cases, alongside 24.4% showing cryptitis/atrophy.

CD71 and Dec-1 expression in the gut from SpA patients without IBD

From 41 patients in this study, 79 gut biopsies were collected (38 ileal and 41 colonic). However, only 74 samples underwent IHC processing. The apical expression of CD71 was identified in 31.1% (23/74) of samples, predominantly in the ileum (see Fig. 1) (27.0% vs. 4.1% in colon, see Supplementary Figs. 1 and 2). Similarly, Dec-1 expression was found in 23.0% (17/74) of biopsies (Please see Supplementary Figs. 3 and 4), with 20.3% in the ileum and 2.7% in the colon. CD71 and Dec-1 co-expressions were associated with biopsies exhibiting significant inflammatory infiltrate and villous atrophy, particularly in the ileum (see Fig. 2). Individual patient analysis revealed apical CD71 expression at 48.8%, Dec-1 at 36.0%, and CD71/Dec-1 co-location at 26.3% in ileal samples (Fig. 3), while colon tissue co-expression was absent. Supplementary Tables 2 and 3 provide a bivariate analysis correlating clinical rheumatology and gastrointestinal variables with retrotranscytosis receptor expressions in the ileum.

Fig. 1.

Apical expression of CD71 in ileum biopsy from a SpA patient (20× and 63×). (A) Negative control (20×). (B) Ileum villi with positive apical staining to CD71 20× (Green), note the evident loss and atrophy of the intestinal villi. (C) Close-up of positive apical staining to CD71 on intestinal villi at 63×.

Fig. 2.

Dectin-1 expression in an ileum biopsy from a SpA patient (20×). (A) Negative control (20×). (B) A marked positive reactivity (Green) is evident on the apical side of the epithelial barrier.

Fig. 3.

Co-localisation of CD71 and Dec-1 in the ileum of a patient diagnosed with SpA (20×). Immunolabeling by IHC technique for the identification of CD71 (First column—green labelling), Dec-1 (Second column—red labelling) and the co-localisation of both simultaneous staining (third column—yellow labelling due to the overlapping of fluorescence). Row (A) Internal control of the IHC technique without adding primary antibody (20×). Note the autoreactivity of the red blood cells and the absence of fluorescence in the epithelial barrier. Row (B) Note the positive reactivity for immune response cells present in the inflammatory infiltrate. However, the epithelial reactivity is absent. Row (C) Observe positive reactivity for the co-location of both receptors in the apical face of the epithelial barrier in the ileum (20×).

Most patients (80.0%) exhibiting apical CD71 in the ileum had elevated sSIgA levels (p = 0.05). An inverse relationship exists between CD71 and both CRP > 3.0 mg/L and ESR > 20 mm/h. Over 70.0% of patients with apical CD71 expression displayed mild/moderate ASDAS-CRP (p = 0.036) and ASDAS-ESR (p = 0.006) activity scores (Supplementary Table 2). In terms of ileocolonoscopic findings, 52.9% of patients with ileal mucosal changes tested positive for apical CD71 (p = 0.573), 45.5% of those with vascular pattern loss (p = 0.364), and 64.3% with villi atrophy (p = 0.267). Histological observations in the ileum mirrored these patterns. The retrotranscytosis receptor CD71 was found in 80.0% of patients with cryptitis/atrophy (p = 0.053) and 61.5% of those with chronic inflammatory infiltrates (p = 0.379), as well as 69.2% of patients with altered ileal architecture (p = 0.126).

On the other hand, although Dec-1 was found in 86.7% of patients with sSIgA levels above 45 mg/mL, there was no significant association. Regarding functional and activity indices, Dec-1 was noted in 66.7% of patients with BASFI > 4.0 (p = 0.253), 73.3% of those with BASDAI > 4.0 (p = 0.209), 73.3% with ASDAS-CRP > 2.1 (p = 0.412), and 80.0% with ASDAS-ESR > 2.1 (p = 0.468). However, the combination of diarrhoea and BASDAI > 4.0 was significantly associated with Dec-1 presence (p = 0.05). From the ileoscopic approach, Dec-1 expression was observed in 66.7% with inflammatory changes in the ileum mucous (p = 0.05), without associations with the histological changes. Among patients showing CD71/Dec-1 co-location, 70.0% exhibited villi atrophy (p = 0.029), and 60.0% had histological cryptitis/atrophy in the ileum (p = 0.016). Mucosal changes were observed in 41.2% of patients (p = 0.109), 36.4% exhibited vascular pattern loss (p = 0.380), and 46.2% showed histological alterations in ileal architecture (p = 0.098).

CD71 and Dec-1 gene expression

The mean CD71 mRNA expression in IBD patients was 0.36 ± 0.04 in the ileum and 0.49 ± 0.19 in the colon. Consistent with IHC findings, CD71 was overexpressed in ileum biopsies from SpA patients, exhibiting an eightfold increase compared to the reference (p = 0.001) and IBD ileum biopsies (p = 0.0001) as illustrated in Supplementary Fig. 5. No significant variations in CD71 expression were noted in colon biopsies (0.90 ± 0.1). Additionally, Dec-1 expression showed no amplification across any samples assessed.

Evaluation of interactions among retrotranscytosis receptors and SIgA

The findings indicated co-localisation of CD71 and Dec-1 at the luminal enterocyte surface. Nevertheless, insufficient evidence remains for physical interactions among receptors or their engagement with SIgA in SpA patients. A Duolink PLA technique was utilised on ileum samples to evaluate potential interactions of retrotranscytosis receptors with SIgA. Results demonstrated positive PLA staining for CD71/SIgA interactions at the apical surface, implying direct associations. Conversely, no interactions between Dec-1 and CD71 or SIgA were observed (see Fig. 4).

Fig. 4.

CD71/SIgA physical interaction by Duolink PLA. To assess CD71/SIgA physical interactions, Duolink® Proximity Ligation Assay (PLA) was applied to those ileum samples that showed apical expression of CD-71. Duolink® PLA allows in situ detection of protein interactions with high specificity and sensitivity. Protein targets using two primary antibodies raised in different species (mouse anti-human CD71 and rabbit anti-human SIgA). A pair of oligonucleotide-labeled secondary antibodies (PLA probes) then binds to the primary antibodies. Next, hybridising connector oligos join the PLA probes only if they are < 40 nm distance. A ligase forms a closed, circle DNA template required for rolling-circle amplification in which the PLA probe acts as a primer for a DNA pol. This allows up to 1000-fold amplified signal that is still tethered to the PLA probe, allowing signal localisation. Lastly, labelled oligos hybridise with the complementary sequences within the amplicon, which are then visualised and quantified as red signals by microscopy image analysis.

Multivariate analysis for CD71, Dec-1 and its co-location

A multivariate binomial regression model was performed for each retrotranscytosis receptor and their co-location, including those variables that showed some association level in bivariate analysis (see Table 2). The findings indicate that CD71 was exclusively associated with ASDAS-CRP (OR 1.71, 95% CI 1.00–3.11), while the regression analysis for Dec-1 revealed associations solely with inflammatory alterations in the Ileum (OR 3.06, 95% CI 1.00–10.01), and the co-localisation of CD71/Dec-1 was linked to Ileum-villi Atrophy (OR 4.24, 95% CI 1.00–18.88). Two models of MCDA were performed, one including CD71 expression and the other based on the CD71/Dec-1 co-location. The first model demonstrated a Cronbach’s Alpha coefficient of 0.666, revealing two dimensions: Dimension 1 (Dim1) included ileum-chronic inflammation (CC 0.687), ileum-architecture alteration (CC 0.698), and ileum-Vascular Pattern Loss (CC 0.613), while Dim 2 was associated with CD71 (CC 0.398), ESR (CC 0.546), and CRP (CC 0.441); in contrast, the second model exhibited a Cronbach’s Alpha of 0.597, where Dim1 aggregated ileum-chronic inflammation (CC 0.546), ileum-villi atrophy (CC 0.577), and ileum-Vascular Pattern Loss (CC 0.673), and Dim 2 correlated with CD71/Dec-1 co-location (CC 0.329), sSIgA (CC 0.456), and ASDAS-CRP (CC 0.577) (see Supplementary Fig. 6).

Discussion

The evidence has widely supported the relationship between joint disease and gastrointestinal involvement in patients with SpA^{7–17,30–37}. Mielants et al. observed that the prevalence of intestinal inflammation in patients with AS is higher when peripheral arthritis is present^15,16. The population included in this study has revealed that 87.8% of the patients present a combination of axial and peripheral compromise, with a high prevalence of more than signs of an intestinal condition associated with their state, accompanied by mild/moderate disease activity scores.

Mielants et al.¹⁴ reported a strong association between gut inflammation in biopsies and persistent joint inflammation. They observed chronic inflammatory lesions in 83% of enterocolitic-associated reactive arthritis (ReA) patients and 57% of AS subjects, most with peripheral arthritis¹⁴. The same author suggested that chronic ileal inflammation may contribute to the pathogenesis of some cases of reactive synovitis and even peripheral joint involvement, frequently seen in AS³⁸. Intestinal chronic lesions are characterised by mucosal architecture alterations, crypt distortion, villous atrophy, and the presence of basal lymphoid aggregates^14,39. Van Praet et al.³⁹ found microscopic gut inflammation in 46.2% of patients with axial and peripheral SpA, affecting the ileum in 50% of cases and the colon in 23.3%. Similarly to the reports above, microscopic inflammatory changes in the sigmoid colon (43.9%) and ileum (47.4%) of SpA patients were observed in this study, presenting both axial and peripheral involvement. Chronic inflammation was predominantly observed in the ileum (72.2%) and colon (26.8%) biopsies. In the ileum, the most frequent changes included architectural alterations (31.7%), cryptitis/atrophy (24.4%), microscopic villus atrophy (36.6%) and loss of vascular pattern (29.3%) at the macroscopic level. Some of these changes were associated with the apical expression of CD71 or its co-localisation with Dec-1 (p > 0.05).

Previous findings in our group and others have shown that high levels of SIgA could determine the intestinal-articular relationship^7–9,16. Salas-Cuestas et al.⁸ reported high levels of sSIgA in patients with SpA compared with healthy subjects, and these increases are independent of the presence of HLA-B*27. Likewise, Arias et al.⁹ showed elevated serum SIgA in a cohort of SpA patients, which seemed comparable to patients with IBD. However, the authors state that the mechanism by which SIgA produced on mucosal surfaces is transported into the circulation remains to be explained. Its presence in serum could be proposed as a promising biomarker for gastrointestinal disorders in SpA patients with subclinical intestinal disorders^8,9.

Some hypotheses have been proposed to explain increases in SIgA in the serum of patients with different gastrointestinal diseases^22–25,40. There are study models in celiac disease in which retrotranscytosis of SIgA through CD71 and Dec-1 has shown to be the mechanism of the gut-associated lymphoid tissues, considered the best mechanistic model that explains the pathogenic pathways in this disease^23–25,40. Wei and Wang⁴¹ showed how increased serum SIgA in different pathologies such as Crohn’s disease might be related to the significant reduction in PIgR expression in the basal epithelial face of mesenteric lymphoid organs and Peyer’s patches. However, none of these mechanisms has been evaluated in SpA.

The expression of CD71 and Dec-1 have previously been related to the clinical manifestations and inflammatory processes associated with celiac disease, although these receptors have never been evaluated in SpA. Papista et al.²², Matysiak-Budnik et al.²³, Abed et al.²⁴, Lebreton et al.⁴⁰, and Rochereau et al.^25,42 showed that the expression of CD 71 in the intestinal lumen is directly related to reverse transcytosis of SIgA and, therefore, to the pathogenesis of celiac disease. This study is the first approach exploring the retrotranscytosis mechanism in patients with SpA. Here, we show that SpA patients without IBD with a high prevalence of subclinical manifestations of intestinal involvement present an ectopic expression of transcytosis receptors CD71 and Dec-1 in distal ileum but not in sigmoid colon tissues, showing that the expression of retrotranscytosis receptors is associated with local macroscopic and microscopic findings derived from chronic inflammation in the ileum. The increase in sSIgA and the activity scores in SpA patients might reflect the local expression of these receptors. Our results suggest a possible role of retrotranscytosis in the intestinal and systemic inflammatory involvement of SpA patients without IBD independently of HLA-B*27 expression. On the other hand, the patients included in this study did not present clinical features of celiac disease nor show anti-gliadin antibodies (Data not shown).

The molecular evaluation of the studied receptors was also conducted, revealing that their relative expression, consistent with IHC results, shows overexpression of CD71 mRNA in ileum biopsies of patients diagnosed with SpA, even with an eightfold increase from the reference (p = 0.001). This receptor is constitutively expressed in the intestinal tissue, and the measurement method does not discriminate the distribution and location of the molecule in the enterocytes. Therefore, molecular exploration of CD71 does not determine any association with retrotranscytosis since it is known that this receptor is constitutively located on the basolateral side of human enterocytes and has been described as the leading marker of trafficking to trans-Golgi⁴³. Consequently, the IHC findings must be considered the gold standard to demonstrate the aberrated localisation of the transferrin receptor in patients with SpA.

The study of CD71 and Dec-1 in SpA has been limited to their identification in myeloid and lymphoid mononuclear subpopulations as a marker of cell proliferation and activation^28,44. As in this study, the evidence shows an increase in the expression of both receptors in gut samples from SpA patients. Konttinen et al.⁴⁴ demonstrated a significant increase in CD71 expression on the surface of mononuclear cells in synovial fluid from active ReA patients. Bos et al.²⁶ found that patients with psoriatic arthritis had a significantly higher percentage of circulating CD3+ cells expressing CD71 as an early activation marker expressed by proliferating T cells. For their part, Ruutu et al.²⁸ mentioned that the activation of genes associated with susceptibility and progression to SpA in humans (IL23R, IL12B, IL17R, and CARD9, among others) is conditioned by an increase in the expression of Dec-1 and its activation downstream when there is exposure to β-glucans.

The evidence above shows the expression of CD71 and Dec-1 receptors directed toward other biological phenomena in patients with SpA. Here, we demonstrate that apical expression of these receptors in intestinal tissues from patients with SpA could contribute to understanding the gut-joint axis in this group of diseases. The findings described here show the expression of CD71 on the apical side of the enterocytes (which is abnormal) related to features of intestinal inflammation. The apical expression of CD71 and Dec-1 could suggest a retrotranscytosis mechanism in the gut. However, their presence does not reveal if the mechanism is executed. The co-location of Dec-1 seems to generate a synergistic effect that enhances the inflammatory response in these patients (see Fig. 3). Rochereau et al.²⁵ reported the Siglec-5 molecule as a SIgA transcytosis receptor at the intestinal level; for this reason, the expression of this receptor was evaluated in the samples obtained. However, its apical expression was not observed in patients with SpA (Data not shown). Additionally, this investigation assessed the expression levels of CD71 and Dec-1 in biopsy specimens obtained from individuals diagnosed with inflammatory bowel disease (IBD) as well as a control cohort; nonetheless, the sample size was limited, and no statistically significant disparities were detected when performing a comparison between the groups (data not presented). Figure 5 represents a preliminary hypothetical model based on the results derived from this research.

Fig. 5.

Hypothetical model of reverse transcytosis of SIgA in the intestinal barrier of SpA patients. Dysbiotic processes in the intestinal mucosa of SpA patients could generate SIgA increases in the intestinal lumen and stimulate an inflammatory response in the gut-associated lymphoid tissues (GALT). Previously, it has been reported that pro-inflammatory molecules increase in the gut could generate the differentiation of M-like cells, which could express on its apical side either an abnormal expression of the CD71 molecule, an overexpression of Dec-1, or its co-location. Lebreton et al. and Rochereau et al. reported these molecules as reverse transcytosis receptors for SIgA in the gut. That phenomenon might explain the increases in this immunoglobulin at the systemic level and the subclinical intestinal involvement in SpA.

Likewise, this study used some biopsies to establish an ex-vivo organotypic culture model on Transwell inserts to evaluate the retrotranscytosis phenomenon. Although the tissues were viable and metabolically active, the integrity of the epithelial barrier was lost entirely (see Supplementary Fig. 7). In addition, the size of the biopsies was insufficient to cover the surface of the insert, which makes the SIgA transport measurement unreliable. Our findings indicate the possibility of SIgA reverse transcytosis in SpA, necessitating validation through an in vitro model. Therefore, we are developing a model similar to those suggested by Rochereau et al.^25,45 and Beloqui et al.^25,45, tailored explicitly to the SpA microenvironment, which aims to demonstrate or dismiss reverse transcytosis in the pathogenesis of intestinal involvement in SpA patients.

This study initially aimed to elucidate the expression of reverse transcytosis receptors on the apical surface of the intestinal mucosa in patients diagnosed with spondyloarthritis (SpA) and to corroborate this phenomenon in the same patient population. The primary constraint of this investigation is that the size of the sample analysed was notably limited. However, the study commenced with a considerable cohort of patients diagnosed with SpA (n = 180); only 22.8% (41/180) satisfied the stringent inclusion criteria, explicitly requiring a colonoscopy assessment and a comprehensive histological examination. A larger sample size could lead to greater statistical validity. However, despite this limitation, the research provides important contributions to understanding SIgA retrotranscytosis in SpA. Ideally, the explanatory power of this model would be enhanced by including a sizable control group of healthy individuals. However, bioethical considerations underscore that those invasive procedures, such as ileocolonoscopy, are contraindicated in healthy subjects, except in cases of suspected malignancy. Consequently, due to the aforementioned justification, biopsy collection from healthy participants was restricted (n = 4) and was insufficient for statistical analysis. Supplementary Figs. 1, 2, 3 and 4 illustrate the variances in the expression of CD71 and Dec-1 between healthy participants and SpA patients, both in the ileum and colon.

Methods

Study design and population

A cross-sectional study was performed, and non-probability sampling was used. A total of 180 adult patients ≥ 18 years old with SpA diagnosis according to the European Spondyloarthropathy Study Group (ESSG)⁴⁶ and Assessment of Spondyloarthritis International Society (ASAS) classification criteria^47,48 were evaluated in the rheumatology services of Hospital Militar Central and Clinicos IPS in Bogotá, Colombia. Exclusion criteria were pregnancy, breastfeeding malignancies, other autoinflammatory or autoimmune diseases, immunodeficiency, chronic pancreatitis, or chronic liver disease and antibiotic treatment in the past 3 months, and those patients with concomitant IBD. The medical record was retrieved to find information about gastrointestinal symptoms: diarrhoea (> 3 depositions per day), stools with mucous or blood, number of stools per day, abdominal pain, abdominal bloating, food intolerances, and weight loss. Of the total, 82 patients met the selection criteria. Those with 2 or more gastrointestinal symptoms (n = 58, 70.7%) were referred to the GastroAdvanced IPS/gastroenterology service. Of them, 54 (93.1%) required a digital chromoendoscopy (DCE) colonoscopy with histological analysis, but only 41/58 (70.7%) attended the procedure.

Functional and disease activity assessment of SpA patients

The BASFI index was used to evaluate patients’ degree of functional limitation. BASDAI and ASDAS scores were used to assess the disease activity, as they are validated scores for axSpA⁴⁹. Unfortunately, in peripheral SpA, there are currently no validated scores assessing disease activity and functionality^1,50.

Determination of faecal calprotectin-FC

Faecal calprotectin levels were obtained using a commercially available enzyme-linked immune sorbent assay (ELISA) according to the manufacturer’s instructions (DiaSource-Quantitative Fecal Calprotectin KAPEPKT849®, a cut-off of > 120 ng/mL, > 250 ng/mL was considered as high levels-HFC).

Determination of high-sensitivity C reactive protein-CRP

The chemiluminescence (Immulite 1000, Siemens®-LKCRP1 kit) assay was performed according to the manufacturer’s instructions (reference values were 0–3 mg/dL).

Determination of erythrocyte sedimentation rate-ESR

ESR was determined through the quantitative Wintrobe method, according to the manufacturer’s instructions. A value higher than 20 mm/h was considered high.

HLA-B27 sequencing

The HLA was performed with DNA Next-Generation Sequencing (NGS): Illumina/PacBio Sequencing with analysis of the second and third exons. The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below:

https://www.ncbi.nlm.nih.gov/, PRJNA843059.

Colonoscopy by digital chromoendoscopy with NBI

A low-volume preparation for colonoscopy was performed. Achieving adequate cleaning of all colonic and distal ileum pathways, measured with the Boston scale (9/9). The diagnostic and therapeutic endoscopy specialist performed the colonoscopies under intravenous sedation, assisted by an anesthesiologist. Submucosal vascular pattern and stromal alterations were evaluated both in the villi of the ileum and the crypts pattern of the colon by the electronic magnification of the mucosa and DCE NBI (Olympus) and BLI (FUJI) adapted to an Olympus EVIS EXERAIII(CF-HQ190L/I)/FUJI EC-760ZP-V/L Zoom, ELUXEOS700 equipment.

Histological evaluation

The paraffin-embedding fine cutting (4 µm) and routine staining with Hematoxylin–Eosin were used on the distal ileum, sigmoid colon, and rectum to evaluate acute inflammatory changes based on polymorphonuclear neutrophil cells presence, cryptitis, cryptic abscesses, ulcers, gut fissures, and erosions. Chronic inflammatory changes were based on architecture alterations, either metaplastic, stromal (or both changes), immune infiltrates with lymphocyte predominance, and eosinophilia.

Total secretory immunoglobulin A (SIgA) antibody quantitation by ELISA

An ELISA in-house sandwich assay previously described^8,9. The sSIgA cut-off point was established based on the median distribution of serum levels of this antibody in a healthy volunteer group. Thus, SIgA values greater than 45 mg/mL were considered positive.

Transferrin receptor (CD71) and dectin 1 (Dec-1) assessment by immunohistochemistry

Detection of CD71 and Dec-1 was performed on intestinal tissue biopsies fixed with 4% paraformaldehyde (Cat#P6148, Sigma Aldrich). Immunolabeling was performed after a deparaffinisation and antigen unmasking protocol by heat at 95 °C in 10 mM sodium citrate buffer (Cat#C8532 Sigma Aldrich). The tissues were treated with 50 mM ammonium chloride (Cat#A9434 Sigma Aldrich) to reduce basal autofluorescence. A classical indirect immunoassay was performed using 1 ug/mL of MAb mouse anti-human CD71 (Cat#ab9179-100, Abcam, Cambridge, MA) or anti-human Dec-1 (Cat#ab82888, Abcam, Cambridge, MA) after permeabilisation in PBS-TritonX-100 0.3% and blocking with goat serum (Gibco Cat. No 16210064) at 10% in PBST 0.1%. A secondary FITC-coupled goat anti-mouse IgG conjugated antibody was used (Cat#A11007, Invitrogen). The nuclei were counterstained using DAPI dye (Cat#62247 Thermofisher), and the coverslips were mounted using ProLong™ Gold Antifade Mountant (Cat#P36930 Cell Signalling). All slides were observed in an Axio Imager M2 fluorescence microscope using the Colibri Illumination System (Carl Zeiss, Germany) and analysed with Zen-lite 2012 software. Only those tissues that showed apical expression of retrotranscytosis receptors on enterocytes were considered positive.

CD71 and Dec-1 gene expression by qPCR

Twenty-seven intestinal tissue samples were obtained and taken to 300 uL of RNAlater (Invitrogen Cat#AM7020), discriminated as follows: 12 intestinal tissue samples from SpA patients without IBD, 10 samples from IBD patients, and 5 samples from healthy donors. All samples were sonicated for 30 min/37 °C in Sonicator/Ultrasound Bath SERIES-CPXH and digested with proteinase-K 20 mg/mL for 12 h/56 °C. The mRNA was obtained by the TRIzol/chloroform/Isopropanol method and was purified using the Quick-RNA Kit (Zymo Research Cat#R1057) following the manufacturer’s instructions. The extracted product was quantified using a NanoDrop 2000 (Thermo). qPCR was performed using the Luna® Universal One-Step RT-qPCR kit #E3005, following the manufacturer’s instructions. The following primers were used: For CD71 FW: AAAATCCGGTGTAGGCACAG and RV: TTAAATGCAGGGACGAAAGG and for Dec-1 FW: ATCAACATTCTCTTCTAAC and RV: CCTTACTACCTCACATAT. Amplification was performed in a BioRad CFX 96 thermocycler as follows: one cycle of reverse transcription at 55 °C for 10 min, followed by initial denaturation at 95 °C for 10 min, followed by 45 cycles at 95 °C for 10 s and 60 °C for 10 s. 30 s. The relative quantification was performed considering the constitutive expression of ß-actin as the housekeeping gene.

Proximity ligation assay—PLA

To confirm the interaction between the receptors and SIgA, all samples that revealed either apical expression of CD71, Dec-1, or both were assessed using the Duolink® In Situ Red Starter Kit Mouse/Rabbit (Cat. No. DUO92101, Sigma Aldrich) and following the manufacturer’s instructions.

Statistical analysis

A descriptive analysis and frequency distributions were performed for clinical, laboratory, and demographic variables. Measures of central tendency and dispersion for continuous variables were made. The analysis of associations was done using the Chi-square or Fisher’s exact test and binomial logistic regression analysis. Comparisons were made with the t-student test. A multiple correspondence discriminant analysis (MCDA) was performed, including those variables with significant associations from bivariate analysis and some that were considered clinically relevant to explain the impact of SIgA and retrotranscytosis receptors in SpA.

Conclusion

This study confirms the ectopic expression in the gut of some receptors associated with retrotranscytosis in SpA patients without IBD. The findings here reported this expression to a local chronic inflammatory condition related to intestinal involvement in SpA patients without IBD but also demonstrate an evident association with high levels of serum SIgA and mild/moderate scores in activity indices, suggesting the measurement that this mucosal immune response-immunoglobulin could be a sufficient non-invasive biomarker to assess the intestinal condition in patients with SpA. Several mechanistic studies are required to confirm the reverse transcytosis at the intestinal level in SpA patients and to understand the pathogenic pathways involved in the intestine joint axis.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

The authors certify that they have participated sufficiently in the work to take public responsibility for the appropriateness of the design and method of review, as well as, the reviewing of the manuscript and approve it for publication: Conceptualization: De Ávila J. Methodology: De Ávila J, Ramos-Casallas A, Chila-Moreno L, Velandia-Romero ML, and Perdomo SJ. Validation: De Ávila J, and Romero-Sánchez C. Formal analysis: De Ávila J. Investigation: De Ávila J, Romero-Sánchez C, Ramos-Casallas A, Perdomo SJ, Velandia-Romero ML, Castellanos J, Chila-Moreno L, Flórez-Sarmiento C, Parra-Izquierdo V, Bautista-Molano W, Beltrán-Ostos A, Jaimes DA, Bello-Gualtero JM, Acero-Molina D, Chalem P, Pacheco-Tena C. Data curation: Ramos-Casallas A, Chila-Moreno L, De Ávila J. Writing – original draft: De Ávila J, Ramos-Casallas A, and Romero-Sánchez C. Writing – review and editing: De Ávila J, Romero-Sánchez, Ramos-Casallas A, Perdomo SJ, Velandia-Romero ML, Castellanos J, Chila-Moreno L, Flórez-Sarmiento C, Parra-Izquierdo V, Bautista-Molano W, Beltrán-Ostos A, Jaimes DA, Bello-Gualtero JM, Chalem P, Pacheco-Tena C. Funding acquisition: De Ávila J and Romero-Sánchez C,

Funding

This study was supported by The Ministry of Science, Technology, and Innovation—MinCiencias (Grants No. 68022 and 57442). Universidad El Bosque (PCI-2018-10300), Hospital Militar Central (Grant 2017-023), Clínicos IPS, Gastroadvanced, Fundación Instituto de Reumatología Fernando Chalem, in Bogotá, Colombia and Biomedicina de Chihuahua, México. The Asociación Colombiana de Inmunología -ACOI- and Universidad El Bosque provided the funding for publication fees.

Data availability

The datasets associated with NGS findings concerning HLA-B*27 that were produced and/or examined during the current study are available in the GenBank Sequence Read Archive (https://www.ncbi.nlm.nih.gov/sra) under the accession number, PRJNA843059.

Declarations

Competing interests

The authors declare no competing interests.

Ethics declarations

Before the commencement of the enrollment process, documented informed consent was obtained from all individuals participating in the study. All research methodologies involving human subjects were conducted in meticulous adherence to the relevant guidelines and regulations, including the Declaration of Helsinki of 1964 along with its subsequent amendments and national ethical standards (Resolution 8430 of 1993). Research Ethics Committee of Hospital Militar Central from Colombia, meticulously evaluated and approved the study, as well as the utilization of medical records in compliance with data privacy legislation applicable to clinical studies and scientific publications (Approval No. 2017-023).