Peritoneal Expression of SGLT-2, GLUT1, and GLUT3 in Peritoneal Dialysis Patients

Peritoneal dialysis (PD) is an effective method of renal replacement therapy for patients with end-stage renal disease providing a high level of patient autonomy [1, 2]. However, long-term PD is associated with peritoneal changes such as peritoneal fibrosis and, at worst, complications such as peritoneal fibrosis and encapsulating peritoneal sclerosis (EPS) eventually leading to ultrafiltration or method failure. The application of PD is limited due to permanent exposure of human peritoneal mesothelial cells in the peritoneal membrane to glucose-based PD fluids and glucose degradation products [3-6]. Generally, the transport of glucose in the body occurs through glucose transporters belonging to a large family of membrane proteins that is divided into 2 families: the solute carrier family 2 (SLC2A) gene family (also known as glucose transporter [GLUT]) [7, 8] and the so-called SLC5A gene family (SGLT) encoding for the Na⁺/glucose cotransporters [9, 10]. GLUTs are predominantly expressed in enterocytes, in the intestine and kidney, as well as in numerous other organs. In contrast, SGLTs are mainly restricted to the small intestine, proximal tubules of the kidney, and as recently reported, to mesangial cells [11-13]. Presently, inhibitors of glucose transporters and especially SGLT-2 inhibitors are of great interest due to their various applications in the clinics including substantial cardio- and renoprotective effects [14-16]. SGLT-2 inhibitors were initially used to reduce blood glucose levels in diabetic patients [17-21]. However, new studies showed that SGLT-2 inhibitors also reduce the risk of hospitalization from heart failure and of composite renal endpoints in patients regardless of the presence or absence of diabetes [15, 16, 22]. Since a new study suggested that the SGLT-2 inhibitor dapagliflozin mitigated peritoneal fibrosis and ultrafiltration failure in a PD mouse model [23], we hypothesized that SGLT-2 inhibitors may also be potential drugs for PD patients to reduce glucose absorption and potentially functional membrane deterioration. To date, most studies on glucose transporters, especially SGLT-2, have been performed in rats or in vitro models by molecular and biochemical analysis [24-26]. Therefore, we now performed expression studies in a larger cohort of human peritoneal biopsies. Here, we aimed to evaluate if and how the major glucose transporters (SGLT-2, GLUT1, and GLUT3) are expressed in the human peritoneum and if there are changes in transporter expression in association with different peritoneal conditions.

Chemicals were obtained from Sigma-Aldrich unless otherwise indicated.

The study was conducted after consultation with the Ethics Committee of the University of Tübingen, Germany (Process No. 322/2009BO1). For the biopsy registry at the Robert-Bosch Hospital in Stuttgart, peritoneal biopsies were collected from 2009 to 2017. Biopsies were taken from PD patients at the time of catheter removal, correction of a catheter malposition, or as diagnostic biopsies for suspected EPS. With regard to non-PD patients, biopsies were taken at the time of various surgical interventions. Table 1 gives an overview about the subgroups included. All patients had given their informed consent regarding a scientific workup of tissues taken during routine procedures. All clinical data as well as descriptive statistical data of the patient cohort are summarized in Table 2.

Human peritoneal biopsies were incubated in RNAlater or prechilled in liquid nitrogen and were then stored at −80°C. For RNA isolation, 50–100 mg peritoneal biopsy material was shredded in 600 μL lysis buffer (mirVana^TM miRNA Isolation Kit; Thermofisher) with an ultra-turrax (16,000 g, 1 min). Thereafter, the supernatant was transferred to tubes filled with ceramic balls and homogenized using the FastPrep system (3 times, 6 m/s, 20 s; MP Biomedicals). Last, the supernatant was centrifuged at 16,000 g for 1 min. For Western blot analysis, total protein lysates were prepared in ice-cold RIPA lysis buffer (50 mM Tris-HCl, pH 8.0, 120 mM NaCl, 1% NP-40, 0.5% Na-desoxycholat, 0.1% SDS, 2 mM EDTA, 25 mM NaF, 0.2 mM NaVO4, 1 mM DTT, and complete protease inhibitor cocktail from Roche) using a micro pestle (Roth, Germany). Peritoneal homogenates were lysed for 30 min on ice and then centrifuged at full speed for 15 min at 4°C. The protein content was determined using a BCA assay.

Total RNA was isolated with the mirVana^TM miRNA Isolation Kit (Thermofisher) according to the manufacturer’s instructions. Total RNA (1 μg) was reverse transcribed with an RT-PCR kit (Applied Biosystems) and afterward diluted 1:1 with water. qPCR was performed using Power SYBR Green PCR Master Mix (Applied Biosystems). Gene expression was analyzed using the 7500 Real-Time PCR Systems and Sequence Detection Software, version 2.3 (Applied Biosystems). Gene expression was evaluated using the ΔΔCT method and human GAPDH as the reference gene. The following primer sequences were used: GAPDH: forward 5′-3′ GCA TCT TCT TTT GCG TCG, reverse 5′-3′ TGT AAA CCA TGT AGT TGA GGT; SGLT-2: forward 5′-3′ GAC ACG GTA CAG ACC TTC GTC AT, reverse 5′-3′ CTC CCA GGT ATT TGTCGAAGAGA; GLUT1: forward 5′-3′ AGG TGA TCG AGG AGT TCT AC, reverse 5′-3′ TCA AAG GAC TTG CCC AGT TT; GLUT3: forward 5′-3′ TTC GTC TCT AGC CTG CAC TG, reverse 5′-3′ ACA CAA CTT CTC CGG GTG AC.

For Western blot analysis, 20 μg total protein/lane was loaded on 10% SDS gels. Subsequently, separated proteins were transferred onto a nitrocellulose membrane (Amersham) using a wet blot-transfer system (90 V for 1.5 h). After blotting, membranes were blocked with 5% BSA in TBS-T for at least 30 min at RT. The following antibodies were used: anti-SLC2A1 (HPA031345, 1:1,000; Sigma), anti-SLC2A3 (HPA006539, 1:1,000; Sigma), anti-SGLT-2 (clone D6, 1:2,000; Santa Cruz), and anti-β-actin (clone AC-15, 1:20,000; Sigma). For quantitative analysis of transporter expression, signals were quantified with ImageJ 1.52v.

Biopsies from the peritoneum were formalin-fixed in 4% buffered formalin and paraffin-embedded following routine protocols [26]. Immunohistochemistry was performed as previously described [27, 28]. Dewaxed and rehydrated tissue sections were incubated in peroxidase blocking solution (S 2023; Agilent) to block endogenous peroxidases. Pretreatment was performed in a steamer, using an antigen retrieval solution (for SGLT-2: pH 6, S 1699, for SLC2A1 and SLC2A3: pH 9, S 2367; Agilent). The staining method used a dextran-coated peroxidase coupled polymer system (EnVisionTM Detection Kit, Peroxidase/DAB, Rabbit/Mouse, K 5007; Agilent). As positive control, the following tissues were used: kidney for SGLT-2, placenta for GLUT1, and testis for GLUT3 expression. The following antibodies were used: anti-SLC2A1 (HPA031345, 1:1,500; Sigma), anti-SLC2A3 (HPA006539, 1:400; Sigma), and anti-SGLT-2 (clone D6, 1:50; Santa Cruz). The sections were examined with the Olympus VS120 automated slide scanner equipped with a BX61VS microscope (objective: UPLSAPO 40x; Olympus).

Immunohistochemistry results were evaluated in a semiquantitative approach (Histo-Score) similar to tumor samples. Percentage of stained structures (1 point/10% structures) as well as staining intensity (no staining [0], weak [1], moderate [2], or strong [3]) was determined [27]. In our study, the following structures were analyzed: mesothelial structure, submesothelial structure, adipocytes, and vessels.

The final Histo-Score was assigned using the following formula: Histo-Score = (percentage mesothelial structure × staining intensity mesothelial structure) + (percentage submesothelial structure × staining intensity submesothelial structure) + (percentage adipocytes × staining intensity adipocytes) + (percentage vessels × staining intensity vessels).

To compare the different subgroups, a Kruskal-Wallis test and Dunn’s post hoc analysis was used. p values were calculated using GraphPad Prism 9 (GraphPad Software Inc., San Diego, CA, USA).

In this study, we determined the expression of the glucose transporter by analyzing peritoneal biopsies. All clinical data as well as descriptive statistical data of the patient cohort are summarized in Table 2. In general, more male than female patients were included into the study. As expected, time on PD was longer in patients with signs of EPS. We isolated RNA as well as protein lysates from the human peritoneum and analyzed the expression of SGLT-2, GLUT1, and GLUT3 by qPCR and Western blot analysis. Furthermore, we determined the localization of the different glucose transporters in the peritoneum by immunohistochemistry using specific antibodies. We could show that SGLT-2, GLUT1, and GLUT3 are expressed in the human peritoneum (shown in Fig. 1-3). RNA analysis showed no significant differences in the expression of the transporters between the different subgroups analyzed (shown in Fig. 1A). In general, we observed large variations within subgroups as well as variations between the different subgroups analyzed. Total protein expression studies using Western blot showed no significant differences in protein expression of SGLT-2, GLUT1, or GLUT3 (shown in Fig. 1C) between the subgroups analyzed. We also performed localization studies and semiquantitative evaluation of protein expression between the different subgroups using a Histo-Score. In general, we observed that the expression of glucose transporters in the peritoneal membrane is very weak and not consistent. Furthermore, many sections showed a negative staining depending on the glucose transporter analyzed (34–46 out of 67 sections). We demonstrated that all 3 transporters are predominantly located adjacent to the vessel walls (shown in Fig. 2A, 3A). However, we observed also stained structures in the submesothelium for GLUT1 and SGLT-2 especially in EPS sections (shown in online suppl. Fig. S1; see www.karger.com/doi/10.1159/000520894 for all online suppl. material). Adipocytes were only positively stained for SGLT-2 in a few biopsies of patients with EPS (shown in online suppl. Fig. S1). A proportion of mesothelial cells expressed SGLT-2 (Fig. 2A) and GLUT1 weakly (shown in Fig. 3A). Nevertheless, the majority of biopsy samples containing mesothelial cells were unstained for SGLT-2 and GLUT1 expression (shown in Fig. 2A second row and Fig. 3A second and fourth rows). Our semiquantitative Histo-Score demonstrated a differentiated pattern of SGLT-2 protein. We observed a significant upregulation in SGLT-2 protein expression in patients with signs of EPS compared to predialysis patients (uremic group) or compared to PD patients (PD >12 months) (shown in Fig. 2B). However, for the GLUT transporters, no significant change in protein expression could be observed (shown in Fig. 3B).

Previous studies have described the presence of glucose transporters – particularly SGLT-2 – in single, isolated cells of the peritoneal membrane; however, a systematic description of peritoneal expression of glucose transporters at a relevant study size in humans has not yet been presented. Therefore, we analyzed the expression of SGLT-2, GLUT1, and GLUT3 in human peritoneal tissue of peritoneal biopsy samples. Here, we present data on the expression of glucose transporters from a large registry of human peritoneal biopsies. Moreover, we could mostly exclude diabetic patients from our analysis (see Table 2) since previous studies showed that glucose transporters were upregulated in diabetic patients compared to nondiabetic controls [28, 29]. Furthermore, diabetic rats showed an increased expression of SGLT-2 in peritoneal mesothelial cells compared to peritoneal mesothelial cells of nondiabetic rats [25].

We showed that SGLT-2, GLUT1, and GLUT3 are expressed in the human peritoneum (shown in Fig. 1-3). So far, most studies on glucose transporters have been performed in rats or in primary cells [24, 26, 30, 31]. Beyond such models, SGLT-2 expression in the human peritoneum has rarely been reported [23, 26]. On mRNA level as well as on total protein level, we observed no significant differences for SGLT-2, GLUT1, and GLUT3 expression upon exposure to glucose-based peritoneal dialysis fluids even though it has been shown that mRNA of GLUT1 is upregulated by high glucose dialysates in rats [31]. Notably, we detected large variations within as well as variations between the different subgroups (shown in Fig. 1-3). In general, glucose transporters are expressed in the peritoneal membrane, although not consistently, especially on biopsy sections. The expression is weak and not consistent (shown in Fig. 2A, 3A, negative scores). The expression is probably affected by the influence of peritoneal dialysis and the associated changes. This is particularly evident in the increased expression in the context of EPS (shown in Fig. 2). Furthermore, we observed that these transporters are predominantly, but not exclusively, located adjacent to the vessel walls (shown in Fig. 2A, 3A; online suppl. Fig. S1). Commonly, the majority of mesothelial cells showed no staining for SGLT-2. Therefore, we could partially reproduce the results observed using immunofluorescence by Balzer and colleagues [32]. They also showed expression of SGLT-2 in the pericapillary region but also the mesothelial cell layer [32]. Furthermore, we observed a significant upregulation in SGLT-2 protein expression in patients with signs of EPS compared to the uremic subgroup or compared to PD patients performing PD >12 months (shown in Fig. 2B).

In general, we observed differences between mRNA and total protein expression. However, these differences could be explained by posttranslational modifications. Similar differences between mRNA and protein levels were also observed in mice [33]. Beyond, the protein expression studies also observed differences in expression of the transporters. However, we observed protein expression in each biopsy using Western blot analysis, although we had several sections with a negative Histo-Score in the IHC. This discrepancy could be explained by the fact that the different glucose transporters showed an intense expression in erythrocytes, granulocytes, and leucocytes. In the total protein lysates of our fresh biopsy samples, we cannot exclude these cell types. However, for evaluation using our Histo-Score, such structures were not taken into account. The increased Histo-Score of SGLT-2 in long-term PD patients with signs of EPS suggests that long-term exposure to glucose-based peritoneal solutions may be a key factor for this upregulation. This finding could be underlined by the fact that no great difference has been observed between healthy controls and uremic patients with no exposure to PD fluids (Fig. 2B, white circles and blue squares). Beyond, these results are in accordance with previous in vitro studies or studies in mice or rats showing that expression of glucose transporters is enhanced after exposure to peritoneal dialysis solutions [26, 31].

Previous studies on mice or rats and also in vitro studies showed that different inhibitors of glucose transporters have positive effects on fibrosis and ultrafiltration failure. Hong and colleagues [31] described that rats in the PD group showed obvious signs of peritoneal fibrosis due to upregulation of glucose transporters such as GLUT1. However, peritoneal fibrosis could be mitigated upon treatment with phloretin, a potent GLUT1 inhibitor. An age-related study in mouse and human lung tissue showed that GLUT1 inhibition ameliorated fibrogenesis of the lung and thereby suggesting a potential therapeutic option in fibrosis-susceptible organ systems [34].

SGLT-2 inhibitors were examined for diverse clinical applications [21, 35, 36]. Currently, some studies in rats suggested potential beneficial effects of SGLT-2 inhibitors in peritoneal dialysis, too [23, 26, 37, 38]. In contrast to those findings, we found glucose transporters are expressed in the human peritoneal membrane, although not consistently. Since the expression is probably affected by the influence of peritoneal dialysis and the associated changes, it remains to be investigated if these observations can be extrapolated to the human peritoneal membrane.

We thank our study nurses Bianka Rettenmaier and Andrea Schwab and laboratory technician Dagmar Biegger for their great support.

The study was carried out according to the World Medical Association Declaration of Helsinki and was approved by the Ethics Committee of the University of Tübingen, Germany (Process No. 322/2009BO1). Written informed consent was obtained from all patients.

The authors have no conflicts of interest to declare.

This study was supported by the Robert-Bosch Foundation (Stuttgart, Germany) and the Berthold Leibinger Foundation (Ditzingen, Germany). The funding organization was not involved in the design of the study; the collection, analysis, and interpretation of the data; or the decision to publish the finished manuscript.

S.S. designed the study protocol, monitored data collection for the whole trial, cleaned and analyzed the data, and drafted and discussed the manuscript. T.O. designed and performed experiments, analyzed data, discussed results, and drafted the manuscript. P.F. analyzed data and discussed results. M.K. designed the study protocol, designed data collection tools, monitored data collection for the whole trial, guided the research, and edited the manuscript. M.D.A. initiated the collaborative project, designed the study protocol, and monitored the design of the data collection tools, the data collection for the whole trial, the draft, and the statistical analysis. He is the guarantor. M.S. designed the study protocol, designed data collection tools, monitored data collection for the whole trial, cleaned and analyzed the data, and drafted the manuscript. All contributors were substantially involved in the conception or design of the work and the acquisition, analysis, or interpretation of the data for the work. All authors critically revised the intellectual content and the drafted manuscript, approved the final version of this draft, and are accountable for all aspects of the work.

Deidentified participant data, used questionnaires, study protocol, etc., will be shared upon request and after review of institutional policies on data protection.