Molecular differences of adipose-derived mesenchymal stem cells between non-responders and responders in treatment of transphincteric perianal fistulas

Abstract

Background: Injection of autologous adipose tissue (AT) has recently been demonstrated to be an effective and safe treatment for anal fistulas. AT mesenchymal stem cells (AT-MSCs) mediate the healing process, but the relationship between molecular characteristics of AT-MSCs of the injected AT and fistula healing has not been adequately studied. Thus we aimed to characterize the molecular and functional properties of AT-MSCs isolated from autologous AT injected as a treatment of cryptogenic high transsphincteric perianal fistulas and correlate these findings to the healing process.

Methods: 27 patients (age 45 ± 2 years) diagnosed with perianal fistula were enrolled in the study and treated with autologous AT injected around the anal fistula tract. AT-MSCs were isolated for cellular and molecular analyses. The fistula healing was evaluated by MRI scanning after 6 months of treatment. AT-MSC phenotype was compared between responders and non-responders with respect to fistula healing.

Results: 52% of all patients exhibited clinical healing of the fistulas as evaluated 6 months after last injection. Cultured AT-MSCs in the responder group had a lower short-term proliferation rate and higher osteoblast differentiation potential compared to non-responder AT-MSCs. On the other hand, adipocyte differentiation potential of AT-MSCs was higher in non-responder group. Interestingly, AT-MSCs of responders exhibited lower expression of inflammatory and senescence associated genes such as IL1B, NFKB, CDKN2A, TPB3,TGFB1.

Conclusion: Our data suggest that cellular quality of the injected AT-MSCs including cell proliferation, differentiation capacity and secretion of proinflammatory molecules may provide a possible mechanism underlying fistula healing. Furthermore, these biomarkers may be useful to predict a positive fistula healing outcome.

Trial registration: NTC04834609, Registered 6 April 2021. https://ichgcp.net/clinical-trials-registry/NCT04834609.

Keywords: Adipose-derived mesenchymal stem cells; Autologous adipose tissue graft injection; Fistula healing; Stem cell potency; Transsphincteric perianal fistula.

Conflict of interest statement

The authors have nothing to disclose.

© 2021. The Author(s).

Figures

Fig. 1

The flow chart of the enrollment of the patients and outcome of the study 6 months after last AT injection

Fig. 2

Cellular characteristics of AT-MSCs from responders and non-reponders in fistula treatment. AT-MSCs were established from non-responders (n = 15) and responders (n = 12) in AT graft fistula treatment. The cells were examined in undifferentiated state in passage 1. A Short-term proliferative rate, area under the curve (AUC) and cell proliferation rate measured by MTT assay (from left to right) of AT-MSCs in non-responder and responder group (n = 12–15). *p < 0.05, non-responders versus responders. Screening of stem cell surface marker expression, such as B CD44, CD90, CD105 and C CD49a, LEPR, and SOX2 measured using flow cytometry in AT-MSCs isolated from non-responder and responder subjects (n = 12–15). Data are presented as means ± SEM; *p < 0.05, non-responders versus responders, (two-tailed unpaired Student’s t test)

Fig. 3

Differentiation potential and inflammatory profile of AT-MSCs from responders and non-responders in fistula treatment. AT-MSCs were established from non-responders (n = 15) and responders (n = 12) in AT graft fistula treatment. The cells were examined in undifferentiated and differentiated state in passage 2. Osteoblast differentiation potential of AT-MSCs evaluated by A Alizarin S staining and using quantification of alkaline phosphatase (ALP) activity represented as fold change (F.C.) over non-induced cells (day 7); B and gene expression of RUNX2 and BGALP mRNA levels (n = 12–15); *p < 0.05: non-responders versus responders (two-tailed unpaired Student’s t test). Adipocyte differentiation potential of AT-MSCs evaluated by C Oil red O staining of mature adipocytes (magnification 10x, scale bar 100 μm) and gene expression of PPARG and LPL (n = 12–15); D Gene expression profile of pro-inflammatory (NFKB, IL1B and TNFA) and anti-inflammatory genes (IL10) in non-responder and responder AT-MSCs (n = 12–15). E Gene expression profile of senescence associated secretory phenotype (SASP) (CDKN2A, TPB3, TGFB1, VEGFA, IFNG, IL6) in non-responder and responder AT-MSCs (n = 5); F Gene expression profile of matrix metalloproteinases (MMP2, MMP9) in non-responder and responder AT-MSCs (n = 5). Data are presented as means ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001: non-responders versus responders (two-tailed unpaired Student’s t test)