Aryl Hydrocarbon Receptor Pathway Augments Peritoneal Fibrosis in a Murine CKD Model Exposed to Peritoneal Dialysate

Abstract

Key Points

• CKD and high glucose–containing peritoneal dialysate alter peritoneal membrane contributing to peritoneal dialysis failure, with a poorly understood mechanism.

• CKD milieu activates the aryl hydrocarbon receptor pathway in the subperitoneal vasculature, increasing the peritoneal fibrosis and collagen deposition in humans and mice.

• An aryl hydrocarbon receptor inhibitor mitigates CKD and peritoneal dialysis–mediated peritoneal fibrosis, collagen deposition, and vasculogenesis in a mouse model.

Background

CKD is a proinflammatory and profibrotic condition and can independently alter the peritoneal membrane structure. Peritoneal dialysis (PD) results in profound alterations in the peritoneal membrane. The mechanisms contributing to the alterations of the peritoneal membrane structure in CKD milieu, along with PD, are poorly understood.

Methods

Here, we show that human CKD induces peritoneal membrane thickening, fibrosis, and collagen deposition and activates the aryl hydrocarbon receptor (AHR) pathway in the subperitoneal vasculature. Leveraging a novel model of PD in CKD mice, we confirm these CKD-induced changes in the peritoneal membrane, which are exacerbated on exposure to the peritoneal dialysate. Peritoneal dialysate further augmented the AHR activity in endothelial cells of peritoneal microvasculature in CKD mice.

Results

Treatment of CKD mice with an AHR inhibitor in peritoneal dialysate for 2 weeks resulted in a seven-fold reduction in AHR expression in the endothelial cells of subperitoneal capillaries, a five-fold decrease in subperitoneal space, and a nine-fold decrease in fibrosis and collagen deposition compared with vehicle-treated CKD mice. AHR inhibition reduced inflammation, subperitoneal neovascular areas, and its downstream target, tissue factor. The AHR inhibitor treatment normalized the peritoneal dialysate-induced proinflammatory and profibrotic cytokines, such as IL-6, monocyte chemoattractant protein-1, and macrophage inflammatory protein 1 levels, in CKD mice.

Conclusions

This study uncovers the activation of the AHR-cytokine axis in the endothelial cells of subperitoneal vessels in humans and mice with CKD, which is likely to prime the peritoneal membrane to peritoneal dialysate–mediated alterations. This study supports further exploration of AHR as a potential therapeutic target to preserve the structural and functional integrity of the peritoneal membrane in PD.

Visual Abstract

Introduction

Approximately 11% of patients with ESKD worldwide undergo peritoneal dialysis (PD).¹ PD offers several advantages over maintenance hemodialysis. PD performs two fundamental functions—volume and solute clearance to maintain fluid, electrolyte, and acid–base homeostasis and treat uremic toxicity.² The rate of ultrafiltration and solute clearance depends on the integrity of the peritoneal membrane. A prolonged period of PD is associated with structural and functional changes resulting in the loss of ultrafiltration and solute clearance.^3–6 These alterations compromise the PD prescription and predispose patients to inadequate dialysis and fluid overload. In fact, <12% of patients with PD in the United States use PD after 5 years.² The main causes leading to the cessation of PD are technique failure, PD–associated peritonitis, and acquired membrane function loss.

The peritoneum consists of a single layer of mesothelial cells resting on different cells and capillaries in the extracellular matrix.^3–6 The endothelium of capillaries in the subperitoneal space represents the main barrier regulating the fluid and solute clearances through different pores. The peritoneal membrane is thin and is devoid of inflammation. Peritoneal dialysate, especially higher concentrations of glucose, results in inflammation, neovascularization of the subperitoneal space, and fibrosis. The peritoneal mesothelial cells undergo epithelial–to–mesenchymal transition after exposure to dialysate to form fibroblasts.⁴

Although multiple factors are implicated in the development of ultrastructural changes with exposure to the peritoneal membrane to PD fluid, there is a dearth of literature on signaling processes influencing all the components of peritoneal membrane failure. This point is important as CKD is a profoundly proinflammatory and profibrotic condition. Patients with CKD show doubling of the submesothelial space and vascular alterations in patients even before exposure to the peritoneal dialysate.^7,8 Several studies have shown that the retained solutes (uremic toxins) activate aryl hydrocarbon receptor (AHR), a xenobiotic pathway in various cell types in CKD mice and humans.^9,10 However, its importance in peritoneal membrane integrity remains uninvestigated.

Considering the above studies^7–10 and leveraging a recently described PD murine model,¹¹ we set out to examine the potential mechanisms underlying the peritoneal damage due to CKD with the hypothesis that the activation of aryl hydrocarbon signaling induces a proinflammatory profibrotic milieu, contributing to the peritoneal membrane fibrosis, inflammation, and vasculogenesis.

Methods

All the methods, including human samples, experimental models, cell culture, antibodies (Supplemental Table 1), reagents, and statistical analysis, are provided in the Supplemental Information.

Results

Peritoneum of Human CKD Shows Evidence of Fibrosis and Thickening

We first examined the peritoneum of three patients with CKD and sex, age, and ethnicity-matched controls. The control subjects were age 33–79 years and were similar to patients with CKD (48–76 years). Most patients were male in both groups. Both control and CKD groups consisted of a Black, a non-Hispanic White, and a Hispanic subject. The control group had BUN and creatinine of 16–19 and 0.73–0.75 mg/dl, respectively. The CKD group had BUN and creatinine of 35–149 and 1.72–10.52 mg/dl, respectively. Other characteristics are delineated in Supplemental Table 1.

The peritoneum from control (non-CKD) subjects showed a thin membrane represented by black arrowheads (Figure 1A). By contrast, CKD peritoneum showed strikingly different features and can be characterized by a thickening of the peritoneal membrane (Figure 1B). A modified Masson trichrome stain was used to characterize the fibrosis and skeletal muscles, which are stained in blue and red, respectively. The peritoneum from the non-CKD group showed minimal evidence of fibrosis, whereas the CKD group showed an increase in fibrosis in the subperitoneal area (the blue fibers marked by arrowheads, Figure 1, C and D). To further characterize the fibrosis, we used Sirus red dye that exclusively stains the collagen fibers. The CKD group showed prominent collagen deposition (black arrowheads, Figure 1, E and F) compared with the non-CKD group.

Figure 1.

CKD induces profound changes in the human peritoneum. Paraffin-embedded sections of the peritoneum of non-CKD patients (A) and CKD (B) were stained with H&E. Representative images taken at different magnifications are shown. The scale bar shows 100 μm. (C and D) Representative images of paraffin-embedded sections of the peritoneum from non-CKD (C) and CKD patients (D) stained with modified Masson trichrome stain at different magnifications are shown. The scale bar shows 100 μm. (E and F) Representative images of paraffin-embedded sections of the peritoneum from non-CKD (E) and CKD (F) patients stained with Sirius red stain at different magnifications are shown. The scale bar shows 100 μm. (G) The peritoneal membrane was measured on H&E-stained slides using ImageJ at six to seven random areas per patient. The average peritoneal membrane thickening is shown. Error bar=SEM. Student's t test was used to compare two groups. ****P < 0.001. (H) The subperitoneal fibrosis was measured using trichrome stained slides at six to seven random areas per patient by defining a region of interest, and the integrated density was measured using ImageJ. The average integrated density of fibrosis normalized to area is shown. Error bar=SEM. The Student t test was used to compare two groups. ****P < 0.001. (I) The collagen deposition was measured using Sirius red stained slides at six to seven random areas per patient by defining a region of interest and integrated density was measured using ImageJ. The average integrated density of collagen normalized to area is shown. Error bar=SEM. The Student t test was used to compare two groups. ****P < 0.001. (J) Immunofluorescence staining of the peritoneum from CKD and non-CKD patients stained with CD31 and AHR primary antibodies. Alex Fluro secondary antibodies were used. Representative images at ×400 magnification are shown. The scale bar shows 100 μm. (K) Averages of the integrated density of AHR from three randomly selected images per patient. Error bars=SEM. The Student t test was used to compare two groups. **P = 0.0025. AHR, aryl hydrocarbon receptor; H&E, hematoxylin and eosin.

The peritoneal membrane thickness in non-CKD controls was 77.56±29.57 μm and that of patients with CKD increased approximately three-fold (288.0±103.9 μm, P < 0.001, Figure 1G). The fibrosis and collagen deposition were quantitated using ImageJ and expressed as the integrated densities normalized to the surface area, as described previously.^12,13 The integrated density consists of pixel density and pixel number in a region of interest. The subperitoneal fibrosis increased by 2.8- to two-fold in CKD compared with non-CKD (P < 0.001, Figure 1H). The integrated density of collagen increased by 2.5-fold in CKD (305.0±155.5 integrated density/μm) compared with non-CKD (59.29±30.83 integrated density/μm, P < 0.001, Figure 1I).

Our previous work and that of others showed an induction of AHR signaling in the vasculature of patients with CKD because of the retention of protein-bound uremic solutes.^14,15 Therefore, we posited that the vasculature in the subperitoneal region of patients with CKD is likely to show AHR activation. Upregulation and nuclear translocation of AHR are considered pathognomonic features of AHR activation. The subperitoneal vessels of patients with CKD showed increased AHR signaling in the endothelium and nuclear AHR (white arrowhead, Figure 1J), whereas the expression of AHR was low in the non-CKD patients. The AHR signal was quantitated as integrated density and normalized to the surface area (Figure 1K). CKD peritoneum showed a 2- to 2.5-fold upregulation of AHR expression (P = 0.0025). These results strongly suggest increased fibrosis and collagen deposition in the subperitoneal space of patients with CKD and activation of AHR signaling within the subperitoneal vasculature of patients with CKD.

CKD Induces Striking Changes in the Peritoneal Membrane, Which Are Exacerbated by Peritoneal Dialysate in Mice

We examined the above findings and changes superimposed by the peritoneal dialysate in a CKD rodent model (Figure 2A). We chose this experimental design to emulate clinical course in patients with CKD on PD. By 7 days of exposure to a 0.2% adenine diet, mice showed high BUN (average 42.7+5.2 mg/dl compared with mice on a normal diet 23.1+3.1 mg/dl, P = 0.038) consistent with early-stage CKD, and the PD catheter was placed. We observed no peritoneal fluid leakage after 7 days of catheter insertion when dialysate infusion was initiated and continued for an additional 2 weeks along with the adenine diet (Figure 2A).

Figure 2.

CKD induces profound changes in the peritoneum of a rodent model, which are further exacerbated by dialysate. (A) A group of 8- to 12-week-old female C57BL/6J mice were randomized to three groups (n=5/group). (B) Paraffin-embedded sections of kidneys harvested from the mice from three different groups were stained with modified Masson trichrome stain. Representative images taken at ×100 magnification are shown. Scale bar=100 μm. (C) Representative images of paraffin-embedded sections of peritoneum stained with H&E are shown. The black arrowheads point to peritoneal inflammation, and the black asterisk denotes an increase in the subperitoneal space. Scale bar=100 μm. (D) Averages of subperitoneal space of different mice detected from H&E images at ×100 magnification. Two randomly selected images per mouse were used. Error bars=SEM. ANOVA was used to compare all the groups (P = 0.003). The Student t test was used to compare two groups. *P = 0.0020, **P = 0.0084, ****P < 0.001. (E) Representative images of paraffin-embedded sections of peritoneum stained with modified Masson trichrome stain at different magnifications are shown. Scale bar=100 μm. (F) Averages of integrated density of peritoneal fibrosis of mice detected using Masson trichrome stained images at ×100 magnification. Two randomly selected images per mouse were used. Error bars=SEM. ANOVA was used to compare all the groups (P < 0.001). The Student t test was used to compare two groups. ***P = 0.0055, ****P < 0.001, ###P < 0.001.

Both BUN and creatinine values were significantly higher in CKD and CKD+PD groups compared with the controls (Supplemental Figure 1, B and C). BUN levels were reduced in CKD mice on PD compared with CKD mice without PD infusion, whereas no difference was noted in the creatinine levels in CKD and CKD+PD groups. Compared with mice on a normal diet, mice on an adenine diet showed evidence of glomerular atrophy and sclerosis along with loss of brush border, tubular atrophy, and perivascular and interstitial fibrosis as evidenced by a blue stain that corresponds to fibrosis (Figure 2B). An integrated density (measured on ImageJ) is a composite index of the pixel intensity and pixel number and used to quantitate histological features in a heterogeneous mixture of tissue. The renal fibrosis was increased by three- to four-fold in mice with CKD compared with non-CKD controls. There was no significant difference in renal fibrosis between the CKD and CKD+PD groups (Supplemental Figure 1C).

The non-CKD mice had a thin monolayer mesothelium and minimal subperitoneal space. CKD mice showed several areas of subperitoneal space expansion (black asterisk and black arrowhead, Figure 2C). Most prominent changes were in CKD mice on PD, including an increase in the subperitoneal space with foci of lymphocytic and polymorphonuclear inflammatory infiltrate and a disruption of the peritoneal membrane (black arrowhead, Figure 2C), foci of neovascularization in the subperitoneal space, and intramuscular compartment of mice. The subperitoneal space was significantly increased, and the peritoneal membrane was thickened (Figure 2B). Peritoneum of CKD mice and peritoneal dialysate (22.70±10.78 μm) showed 2.1-fold and 1.4-fold increases in the peritoneal membrane thickness compared with normal (P < 0.001) and CKD (P = 0.0084) mice, respectively (Figure 2D). Compared with the non-CKD mice, CKD mice showed 60% increase in the subperitoneal space (P = 0.0020). CKD mice exposed to peritoneal dialysate showed approximately 2.7-fold higher peritoneal membrane thickness than CKD mice (Figure 2D) (P = 0.0020).

The non-CKD mice showed a very thin peritoneal membrane and a nonsignificant linear demarcation of collagen deposition below the peritoneal mesothelium using a modified Masson trichrome stain (upper Panel Figure 2E). CKD peritoneum showed prominent fibrosis in the subperitoneal area. CKD mice on PD showed areas of striking hyalinosis/fibrosis interspersed with several folds of the peritoneum (black arrowhead, lower panel Figure 2E). Fibrosis was increased by three- to four-fold in mice with CKD compared with non-CKD controls. There was no significant difference in renal fibrosis between the two groups (Supplemental Figure 1A). Compared with normal peritoneum, CKD mice showed 1.8-fold and 5.7-fold increases in the integrated peritoneal fibrosis in CKD (P < 0.001). The exposure to peritoneal dialysate in CKD mice increased the peritoneal membrane thickness by 3.4-fold (P = 0.0055) compared with CKD mice without PD (Figure 2F).

CKD Stimulates Subperitoneal Membrane Collagen Deposition

Masson trichrome stain detects collagen, elastin fiber, and various proteoglycans. Next, we specifically examined the peritoneal membranes with Sirius red staining. The peritoneum from non-CKD mice showed a uniform thin lining of loosely organized collagen (black arrowhead, Figure 3A). CKD mice showed prominent collagen deposition made of tight bundles in the peritoneal membrane (yellow arrowhead, Figure 3A). CKD mice exposed to peritoneal dialysate showed areas of neovascularization and striking collagen deposition was noted in subperitoneal and intramuscular regions, especially around the areas of neovascularization (marked by yellow asterisk). The neovascularization was not observed in other groups. This collagen was deposited along the circumferential axis of the intestine encircled by muscle fibers (yellow asterisk, Figure 3A lower panel). Compared with normal peritoneum, CKD mice had approximately two-fold (P < 0.001) and CKD+peritoneal dialysate showed approximately five-fold (P = 0.0071) upregulation of collagen deposition (Figure 3B). Compared with CKD mice, peritoneal dialysate exposure increased the integrated density of collagen by two-fold (P = 0.0355). These data suggest that CKD mice show significantly increased subperitoneal space and collagen deposition; both these features are further enhanced with peritoneal dialysate exposure. Intriguingly, the neovascularization in the subperitoneal space was only observed in CKD mice on exposure to peritoneal dialysate.

Figure 3.

CKD induces collagen deposition in the subperitoneal space and activates AHR signaling in the peritoneal vessels. (A) Representative images of paraffin-embedded sections of peritoneum stained with Sirius red stain at different magnifications are shown. Scale bar=100 μm. (B) Averages of integrated density of collagen deposition of mice detected using Masson trichrome stained images at ×100 magnification. Two randomly selected images per mouse were used. Error bars=SEM. ANOVA was used to compare all the groups (P = 0.004). The Student t test was used to compare two groups. **P < 0.001, ***P = 0.0355, ****P = 0.0071. (C) Paraffin-embedded sections of the peritoneum from two groups were stained with an AHR and CD31 (a marker for endothelial cells). Alex Fluro secondary antibodies were used. Representative images at ×400 magnification are shown. The scale bar shows 100 μm. (D) Averages of integrated density of AHR from two randomly selected images per mouse. Error bars=SEM. The Student t test was used to compare two groups. *P < 0.1, ***P = 0.0236.

AHR Signaling Is Activated in the Peritoneal Vessels of CKD Mice, Which Is Further Augmented with Peritoneal Dialysate

CKD is a state of AHR signaling hyperactivation (upregulation and nuclear translocation of AHR) in the vasculature of CKD mice and humans.^9,10,16,17 AHR signaling plays a significant role in organ fibrosis.⁹ We posited that vessels in the peritoneal membrane are likely to show higher AHR expression supporting AHR activation in them compared with the control peritoneum.

CKD mice exposed to peritoneal dialysate alone exhibited a notable increase in AHR expression in the endothelium and vessel walls, which coincided with CD31 when compared with the control group (Figure 3C). Compared with the AHR expression of 189±97.91 integrated density/μm in the normal peritoneum, CKD peritoneum showed 85% increase in AHR expression (372.4±89.26 integrated density/μm, P = 0.0148) and to 3.2-fold increase in the CKD with exposure to dialysate (Figure 3D) (698.6±246.1 integrated density/μm P = 0.0026). Exposure to peritoneal dialysate doubled the AHR expression in CKD mice compared with the group without CKD (P = 0.0236). These data suggest that CKD mice show AHR activation in the endothelial cells of peritoneal vessels, which is further augmented with exposure to peritoneal dialysate.

AHR Inhibitor Suppresses PD-Induced Inflammation and Fibrosis

The above results demonstrate the activation of AHR signaling in the endothelial cells of peritoneal vessels, which is further augmented with exposure to peritoneal dialysate. Based on these results, we hypothesized that pharmacological manipulation of the AHR may mitigate profibrotic processes in the peritoneum of CKD mice exposed to peritoneal dialysate. A group of C57BL/6J mice were randomized into two groups: vehicle and experimental (n=7 in each group). Both vehicle and experiment groups had a PD catheter inserted and then were subjected to a 0.2% adenine diet as shown in Figure 1A. Mice were subjected to a 0.2% adenine diet for 2 weeks, after which a PD catheter was inserted. One week post-PD catheter insertion, mice were initiated on peritoneal dialysate injection for 2 additional weeks. In parallel, the experimental group was cotreated with CH223191 10 mg/kg of body weight once daily injection with peritoneal dialysate for 2 weeks. The peritoneum and peritoneal fluid were harvested for further examination.

The vehicle group revealed findings consistent with Figure 1B characterized by peritoneal membrane thickening, peritoneal inflammation, neovascularization (yellow asterisk), and expansion of subperitoneal space (Figure 4A). All these findings were reduced in mice that were exposed to dialysate along with CH223191. In this group, the peritoneum was characterized by a thin membrane with the subperitoneal space reduced (black arrowhead, Figure 4A). In addition, higher magnification images showed an intact peritoneal membrane with minimal infiltration in the subperitoneal space and no hyalinosis. The Masson trichrome stain showed fibrosis in subperitoneal space, which was reduced in the CH223191 mice (Figure 4B). Peritoneum from the experimental mice which received treatment with CH223191 showed an intact peritoneal membrane with reduced fibrosis in the subperitoneal area. However, some fibrous deposition was still noted in some focal areas. A Sirius red stain confirmed reduced collagen in the subperitoneal and around the areas of neovascularization in the group treated with CH223191 compared with the vehicle-treated group (black arrowhead, Figure 4C).

Figure 4.

AHR inhibition suppresses dialysate-induced peritoneal membrane alterations in CKD mice. (A) Representative images of paraffin-embedded sections of peritoneum stained with H&E stain from two groups. Scale bar=100 μm. Averages of integrated density of AHR from two randomly selected images per mouse. Error bars=SEM. The Student t test was used to compare two groups. (B) Representative images of paraffin-embedded sections of peritoneum stained with a Masson trichrome stain from two groups. Scale bar=100 μm. (C) Representative images of paraffin-embedded sections of peritoneum stained with a Sirius red stain at different magnifications are shown. Note that the skeletal muscles are cut in coronal and sagittal sections. Scale bar=100 μm. (D) Averages of the integrated density of subperitoneal fibrosis of mice detected using Masson trichrome-stained images at ×100 magnification. Two randomly selected images per mouse were used. Error bars=SEM. The Student t test was used to compare two groups. ***P = 0.003. (E) Averages of the integrated density of collagen deposition of mice detected using Sirius red stained images at ×100 magnification. Two randomly selected images per mouse were used. Error bars=SEM. The Student t test was used to compare two groups. **P = 0.00216.

Integrated density analysis on the trichrome-stained slides revealed that the level of peritoneal fibrosis decreased by five-fold (P = 0.003, Figure 4D) and collagen reduced by 2.5-fold (P = 0.00216, Figure 4E) in the group treated with the AHR inhibitor compared with the vehicle group. These data suggest that exposure to 4.25% peritoneal dialysate-induced fibrosis along with neovascularization are suppressed with an AHR inhibitor.

AHR Inhibitor Downregulates Expression of AHR and Its Downstream Mediator, Tissue Factor

Consistent with our data in Figure 2C, prominent AHR expression and nuclear AHR were found in the vehicle group, which was reduced to approximately 90% with CKD mice treated with CH223191 compared with the vehicle-treated CKD mice (P < 0.001) (Figure 5, A and B). Tissue factor expression, a downstream target of AHR in the CKD milieu, was almost normalized with CH223191 compared with the controls (P < 0.001) (Figure 5, C and D). These results suggest that AHR inhibitor suppresses AHR activity and its downstream target in the endothelium of peritoneal capillaries.

Figure 5.

CH223191 suppresses AHR activity and its downstream target TF in the endothelial cells of peritoneal vascular of CKD mice exposed to peritoneal dialysate. Paraffin-embedded sections of the peritoneum from two groups (n=5 mice/group) were stained with AHR and CD31 antibodies (A) and TF and CD31 antibodies (C). Alex Fluro secondary antibodies were used. Representative images at ×400 magnification are shown. The scale bar shows 100 μm. (B and D) Averages of integrated density of AHR and TF from two randomly selected images per mouse. Error bars=SEM. The Student t test was used to compare two groups. ****P < 0.001. TF, tissue factor.

AHR Inhibitor Suppresses a Set of Proinflammatory and Profibrotic Cytokines

AHR regulates several cytokine pathways. We observed an approximately 95% suppression in IL-6 (P = 0.0006) (Figure 6A), a significant reduction in macrophage inflammatory protein (P < 0.0009) (Figure 6B), and a reduction in monocyte chemoattractant protein-1 (MCP-1) (P = 0.0018) (Figure 6C) with the AHR inhibitor in CKD mice compared with the controls. A similar pattern was noted in IL-5 (P < 0.002) (Figure 6D), approximately 3.5-fold suppression of IL-4 (P < 0.0087) and approximately three-fold decrease in IL-1α (P = 0.0009) in mice treated with CH223191 (Figure 6, E and F). No statistically significant difference was found in granulocyte-macrophage colony-stimulating factor, TNFα, regulated upon activation, normal T cell expressed and secreted (CCL5), and several other cytokines. These data suggest that exposure to an AHR inhibitor reduces highly specific, prothrombotic, and proinflammatory cytokines.

Figure 6.

CH223191 downregulates the level of a host of proinflammatory and profibrotic cytokines. Averages of specific cytokine from the peritoneum of mice from both groups are shown. Error bars=SEM. The Student t test was used to compare the groups. (A) ***P = 0.0006. (B) ***P < 0.0009. (C) ***P = 0.0018. (D) ***P < 0.002. (E) ***P < 0.0087. (F) *P = 0.0009.

Discussion

Peritoneal membrane failure is the predominant cause of discontinuation of PD in most patients with CKD/ESKD.^2,18 Of three categories of membrane dysfunction, such as low intrinsic ultrafiltration and fast peritoneal transport,² our study focuses on the acquired membrane dysfunction primarily contributed by CKD and peritoneal dialysate. Our results suggest that CKD milieu in humans and mice induces thickening of the peritoneal membrane and subperitoneal fibrosis. These changes are further augmented on exposure to the peritoneal dialysate, which increases inflammation, areas of vasculogenesis (new blood vessel formation), and thickness of the peritoneal membrane. Importantly, the CKD milieu activates AHR signaling in the endothelium of peritoneal vessels, which is further enhanced by peritoneal dialysate. AHR inhibitor significantly mitigates all these changes pathognomonic of the peritoneal membrane failure and suppresses proinflammatory and profibrotic cytokines in the peritoneal fluid. These findings are clinically important because the endothelium in the peritoneal vessels is the major determinant of solute kinetics regulating volume homeostasis and uremic toxin clearance.² These observations uncover the pathogenic role of AHR signaling in peritoneal changes in CKD mice exposed to peritoneal dialysate. Because AHR is the receptor for several protein-bound uremic solutes,^14,19 this observation links uremic toxins to peritoneal alterations in patients with CKD on PD.

Our previous study in CKD mice has shown activation of AHR-responsive promoter-β gal response element in the vasculature and different organs.¹⁷ Activation of AHR signaling in endothelium is also shown in large cohorts of patients with CKD.⁹ Therefore, it stands to reason that AHR is also activated in the peritoneum of CKD mice and patients. AHR signaling is also activated by high glucose.²⁰ It is conceivable that CKD, by activating AHR pathway in the endothelial cells of peritoneal vessels, primes the peritoneum to high glucose-mediated damage.

AHR signaling is linked to fibrosis, neovascularization, and inflammation. In a model of cardiac fibrosis, AHR signaling upregulates TGF-β activity, myofibroblast activation, and extracellular matrix degradation.²¹ AHR activation also contributes to pulmonary and renal fibrosis.^22–25

The role of AHR signaling in angiogenesis is context dependent. In CKD mice with an ischemic hind limb angiogenesis model (growth of blood vessels from the existing vasculature), AHR activation by uremic solutes suppresses the Wnt/β-catenin pathway to downregulate vascular endothelial growth factor (VEGF) and IL-6 in skeletal muscles.¹² It is likely that in a nonhypoxic milieu, AHR activation augments vasculogenesis (de novo production of blood vessels), especially in the presence of a glucose-rich environment in the subperitoneal space and skeletal muscle beneath the peritoneal membrane. This possibility is further supported by the observations that, in normoxia, AHR regulates VEGF in the endothelial cells through STAT1 signaling.²⁶ In fact, local production of VEGF is observed within the peritoneal cavity.²⁷ Local VEGF is causally linked to the loss of peritoneal ultrafiltration function^28,29 and may explain vasculogenesis in the subperitoneal and muscular layer of CKD mice exposed to peritoneal dialysate.

Depending on the context and ligand composition, AHR can attenuate or exacerbate inflammation.^22–24 Among several mechanisms postulated, AHR regulates a host of cytokines, including IL-6.³⁰ Many studies implicate IL-6 as a key mediator of inflammation, fibrosis, and peritoneal membrane failure. IL-6 levels correlated with vasculogenesis and inflammatory responses in the peritoneal dialysate.³¹ In support, others have found that chronic intraperitoneal inflammation is correlated with IL-6 levels in peritoneal fluid and the peritoneal membrane functional decline.^32,33 Stimulation of human peritoneal mesothelial cells with the IL-6 and soluble IL-6 receptor complex promotes the epithelial–to–mesenchymal transition.³⁴ Blockade of IL-6 trans-signaling prevented vascular permeability, glucose–induced peritoneal fibrosis, angiogenesis, and macrophage infiltration in a mouse model. Our analysis also found upregulation of IL-1α, IL-4, IL-5, MCP-1, and macrophage inflammatory protein-α, which can potentially contribute to peritoneal fibrosis. Rather than modulating individual cytokine levels, AHR inhibition downregulated a host of profibrotic and proinflammatory cytokines, making it a superior target to prevent peritoneal membrane inflammation and fibrosis in response to CKD and peritoneal dialysate.

This study demonstrates the potential role of AHR as an inhibitor of a host of cytokines, peritoneal inflammation, and fibrosis in CKD in response to peritoneal dialysate. This point is relevant as AHR inhibitors are currently in advanced clinical trials and can be repurposed in patients with CKD on peritoneal dialysate.

A recent study explored the role of an sodium-glucose cotransporter-2 (SGLT2) inhibitor in peritoneal membrane dysfunction induced by prolonged exposure to PD. Balzer et al. demonstrated that SGLT2 in non-CKD mice reduced peritoneal membrane thickness and subperitoneal space.³⁵ The SGLT2 inhibitor increased the resorption of glucose, resulting in the loss of ultrafiltration, and had no effect on vasculogenesis, IL-6, MCP-1, and TNFα levels. Intriguingly, SGLT2 increased the levels of peritoneal macrophages in mice. Although SGLT2 reduced the levels of inflammatory cytokines in vitro cell culture model, it exerted this effect at a high concentration of 100 μM, which is difficult to achieve in vivo.

A larger study with a longer duration, as well as functional peritoneal membrane tests, is required. It is necessary to investigate both sexes to understand and characterize any potential biological sex differences. Our study does not rule out the contribution of PD catheters or inflammation related to potential infection in the peritoneal cavity. Peritoneal fibrosis was noted to be local because of catheter damage,³⁶ not generalized, as observed in this study. While the AHR-inducing activity of sera from mice was not examined in this study, our previous work and that of others have shown a consistent increase in AHR activity in endothelial cells of CKD humans and mice.^9,10 We used an adenine-induced model of CKD, which results in a high uremic solute load. Further investigation using different CKD models will strengthen our findings.

This is the first study demonstrating the involvement of the AHR pathway in the peritoneal endothelium driven by two factors—uremic state and exposure to high glucose containing peritoneal dialysate. This work paves the way to explore the therapeutic targetability of AHR inhibitors to preserve the structural and functional integrity of the peritoneal membrane and prolong the PD in patients with ESKD.

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/KN9/A610.

Funding

V.C. Chitalia: NHLBI Division of Intramural Research (R01HL166608), National Institute of Diabetes and Digestive and Kidney Diseases (DK119740 and DK132784). S. Lotfollahzadeh: NHLBI Division of Intramural Research (T32HL125232) and American Heart Association (857078). W. Yin: American Heart Association (850917).

Author Contributions

Conceptualization: Vipul C. Chitalia, Lauren Stern.

Data curation: Janelle Clovie, Isaac Hoekstra, Saran Lotfollahzadeh, Arjun Patel, Herreet Paul, Isaac E. Sellinger, Aniket Vazirani, Wenqing Yin.

Formal analysis: Vipul C. Chitalia, Janelle Clovie, Saran Lotfollahzadeh, Aniket Vazirani.

Funding acquisition: Vipul C. Chitalia.

Investigation: Vipul C. Chitalia, Saran Lotfollahzadeh.

Methodology: Vipul C. Chitalia, Saran Lotfollahzadeh, Aniket Vazirani.

Project administration: Vipul C. Chitalia.

Resources: Vipul C. Chitalia.

Visualization: Vipul C. Chitalia.

Writing – original draft: Vipul C. Chitalia.

Writing – review & editing: Vipul C. Chitalia, Jean Francis, Marina Malikova, Abbas Brahim Malloum, Luise I. Pernar, Jeffrey Siracus, Lauren Stern, Pranav Yadati, Wenqing Yin.

Data Sharing Statement

All data are included in the manuscript and/or supporting information.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/KN9/A611.

Supplemental Methods

Supplemental Table 1. Immunofluorescence antibodies.

Supplemental Figure 1. (A and B) Averages of BUN (A) and creatinine levels (B) in sera from mice from three different groups at the end of the experiment are shown. ANOVA was performed to compare all the groups. The Student t test was performed to compare different groups. For BUN: ANOVA <0.0001, ****P < 0.0001, and *P = 0.023. For creatinine, ANOVA <0.0001, ****P < 0.0001, and ****P < 0.0001. (C) Averages integrated density of kidney fibrosis determined on the trichrome-stained slides from three randomly selected images per mouse in each group are shown. Error bars=SEM. The Student t test was used to compare the two groups. ****P < 0.0001.