SEARCH

SEARCH BY CITATION

Keywords:

  • Bone marrow-derived progenitor cells;
  • Myofibroblasts;
  • Endothelial cells;
  • p38 mitogen-activated protein kinase;
  • Smad Renal fibrosis

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Recent evidence suggests that bone marrow (BM)-derived cells may integrate into the kidney, giving rise to functional renal cell types, including endothelial and epithelial cells and myofibroblasts. BM-derived cells can contribute to repair of the renal peritubular capillary (PTC) network following acute ischemic injury. However, the cell fate and regulation of BM-derived cells during the progression of chronic renal disease remains unclear. Using chimeric mice transplanted with enhanced green fluorescent protein (EGFP)-expressing BM, we demonstrate that the number of BM-derived myofibroblasts coincided with the development of fibrosis in a mouse adriamycin (ADR)-induced nephrosis model of chronic, progressive renal fibrosis. Four weeks after ADR injection, increased numbers of BM-derived myofibroblasts were observed in the interstitium of ADR-injected mice. Six weeks after ADR injection, more than 30% of renal α-smooth muscle actin (+) (α-SMA+) interstitial myofibroblasts were derived from the BM. In addition, BM-derived cells were observed to express the endothelial cell marker CD31 and the myofibroblast marker α-SMA. Blockade of p38 mitogen-activated protein kinase (MAPK) and transforming growth factor (TGF)-β1/Smad2 signaling was found to protect BM-derived PTC endothelial cells and inhibit the number of BM-derived von Willebrand factor (vWF)(+)/EGFP(+)/α-SMA(+) cells, EGFP(+)/α-SMA(+) cells, and total α-SMA(+) cells in ADR-injected mice. Inhibition of the p38 MAPK and TGF-β1/Smad signaling pathways enhanced PTC repair by decreasing endothelial-myofibroblast transformation, leading to structural and functional renal recovery and the attenuation of renal interstitial fibrosis. Investigation of the signaling pathways that regulate the differentiation and survival of BM-derived cells in a progressive disease setting is vital for the successful development of cell-based therapies for renal repair.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Tubulointerstitial fibrosis is a major determinant of the progression of chronic renal failure regardless of the nature of the initial insult. The process of tubulointerstitial fibrosis includes the loss of renal tubules, the accumulation of interstitial myofibroblasts, and extracellular matrix (ECM) deposition [1]. Renal hypoxia and oxidative stress [2] induce the accumulation of fibroblasts and the conversion of fibroblasts to active myofibroblasts, leading to ECM deposition and the development of interstitial fibrosis [3, 4]. The number of interstitial α-smooth muscle actin-positive [α-SMA(+)] cells, a putative maker for renal interstitial myofibroblasts, is a good prognostic indicator of progression of both human and experimental renal disease [5, [6], [7]8]. However, the origin of myofibroblasts in diseased kidneys remains largely unknown. Some studies have suggested that interstitial myofibroblasts may be derived from resident tubular epithelial cells [1, 9, 10], renal fibroblasts [11, 12], or renal perivascular smooth muscle cells [13, 14].

There is increasing evidence that bone marrow (BM)-derived cells contribute to regeneration following renal damage, although it is not clear whether this process is the result of cell differentiation, cell fusion, or a humoral response resulting in the production of reparative growth factors [15, [16], [17], [18]19]. The ability of BM-derived cells to respond to the inflammatory cues, move into the kidney, and express renal markers of tubular epithelial cells [20], mesangial cells [21], and endothelial cells [22] has been well-documented. Furthermore, fibroblasts or myofibroblasts may be derived from BM cells in fibrotic lesions [13, 14, 23, [24], [25]26]. However, it remains uncertain whether BM-derived cells are capable of differentiating into renal fibroblasts or myofibroblasts directly from stromal cells, hematopoietic progenitor cells, or circulating fibrocytes.

Despite the ability of BM-derived cells to engraft and differentiate into kidney cells in response to injury [15], the fate of BM-derived renal cells in injured kidneys and whether these BM-derived cells constantly undergo differentiation remains largely unknown. Recently, we demonstrated that p38 mitogen-activated protein kinase (MAPK) and transforming growth factor (TGF)-β/Smad signaling pathways are activated in both BM-derived endothelial cells in addition to intrinsic renal endothelial cells in adriamycin (ADR)-induced nephropathy, a chronic disease model [27]. More importantly, following BM cell migration, in response to inflammatory cues, BM-derived endothelial cells undergo apoptosis at the same rate as intrinsic renal cells in this progressive disease setting [27]. This suggests that the interaction between pathological circumstances and a reparative role of BM-derived cells should be carefully scrutinized when investigating a regenerative approach to organ therapy.

The p38 MAPK pathway has been implicated in TGF-β-induced biological responses, such as epithelial-mesenchymal transition [28], collagen type I synthesis in mesangial cells [29], and TGF-β1-stimulated migration of smooth muscle cells [30]. The p38 MAPK pathway is also required by members of the TGF-β superfamily for cell differentiation, including bone morphogenic protein (BMP)-induced cardiomyocyte differentiation [31] and neuronal differentiation of PC12 cells stimulated by BMP-2 [32]. We have previously demonstrated that coadministration of p38 MAPK inhibitor SB203580 and TGF-β receptor I (ALK5) inhibitor (ALK5I) reduced glomerulosclerosis and renal interstitial fibrosis [33]. However, the role of p38 MAPK and TGF-β/Smad signaling pathways in the contribution of BM-derived cells to the development of renal interstitial fibrosis remains unclear.

The present study investigated the recruitment and contribution of BM-derived cells during the time course of the development of renal interstitial fibrosis in a progressive renal disease model of ADR-induced nephropathy. A reduction in the number of BM-derived endothelial cells was associated with a gradual increase in the number of α-SMA interstitial fibroblasts in ADR-injected mice compared with control animals. The number of BM-derived cells expressing the endothelial cell markers CD31 and von Willebrand factor (vWF) and the myofibroblast marker α-SMA was significantly increased in ADR-injected mice compared with the normal saline (NS) vehicle-treated group. Moreover, cultured BM-derived CD31(+)/Flk-1(+) cells were found to demonstrate reduced expression of CD31 and obtained expression of α-SMA, suggesting that BM-derived endothelial cells may lose their endothelial cell phenotype and acquire a myofibroblastic phenotype in a pathological setting. In addition, p38 MAPK and TGF-β1/Smad signaling pathways were activated in BM-derived PTC endothelial cells and α-SMA(+) cells in ADR-injected mice. Furthermore, blockade of the p38 MAPK and TGF-β1/Smad signaling pathways by SB203580 and ALK5I reduced the number of BM-derived CD31(+)/α-SMA(+) cells, reduced BM-derived myofibroblast accumulation, and retarded the progression of end-stage renal disease (ESRD). These studies suggest that the p38 MAPK and TGF-β1/Smad signaling pathways can modulate the contribution of BM-derived cells to the development of renal interstitial fibrosis.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Experimental Animals

To investigate the role of BM-derived cells in renal fibrosis, we generated chimeric mice using BM transplanted from constitutively expressed-EGFP donor mice. The EGFP gene, driven by the chicken β-actin promoter, is expressed in all tissues and organs except hair and red blood cells [34]. Six-week-old BALB/c male mice (20–25 g body weight) were irradiated with two doses of 5 Gy γ-irradiation separated by 4 hours. Whole BM was isolated from the femur and tibia of male BALB/c mice of the same age expressing EGFP by flushing with Iscove's minimal essential medium (IMEM). Irradiated mice were injected with 0.1 ml of IMEM with or without 1 × 106 BM cells via tail vein injection. Twelve weeks after BM transplantation, mice received a single i.v. injection of ADR (10.5 mg/kg; Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com). Control mice were treated with an equivalent i.v. volume of saline vehicle (NS). Mice in all treatment groups (n = 6/time point) were killed at 72 hours and 1, 2, 4, and 6 weeks following ADR or NS injection.

To block the p38 MAPK and TGF-β1/Smad signaling pathways, mice were treated with a p38 MAPK inhibitor (SB203580; Calbiochem, San Diego, http://www.emdbiosciences.com) and a TGF-β receptor I (ALK5) inhibitor (ALK5I; Calbiochem) 2 weeks after ADR injection, via implantation of ALZET osmotic pumps (DURECT Corporation, Cupertino, CA, http://www.durect.com/) until the experimental endpoint. A preliminary experiment was carried out to determine the effective dose range of SB203580 and ALK5I in ADR-induced nephropathy. Doses of SB203580 and ALK5I from 0.25–2 mg/kg were administered respectively to 10 groups of three mice. Based on the results, a larger experiment was performed to confirm that ADR-injected mice receiving 1 mg/kg/day SB203580 + 1 mg/kg/day ALK5I can achieve maximal renoprotective effects without obvious side effects compared with vehicle alone, 1 mg/kg/day SB203580 alone, or 1 mg/kg/day ALK5I alone. Animals received either the vehicle or a combination of 1 mg/kg/day SB203580 (SB) + 1 mg/kg/day ALK5I. Mice were killed 4 weeks after the initiation of vehicle or SB203580+ALK5I treatment (n = 6/group/time point). All experiments were performed with the approval of a Monash University Animal Ethics Committee, which adheres to the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes.

Measurement of Proteinuria and Creatinine

Mice were housed in metabolic cages with free access to food and water on the days of urine collection. Protein from 24-hour urine samples and serum creatinine levels were measured by the detergent-compatible protein assay kit (Bio-Rad, Hercules, CA, http://www.bio-rad.com) and creatinine assay kit (Cayman Chemical, Ann Arbor, MI, http://www.caymanchem.com), respectively, according to instructions supplied.

Antibodies

The following antibodies were used for immunofluorescence: rabbit anti-phospho p38 MAPK (p-p38) raised against the dual phosphorylated tyrosine and threonine residues of the p38 peptide (catalog number P1491) and mouse anti-α-SMA conjugated with Cy3 (catalog number C6198) were from Sigma-Aldrich. Rat anti-CD31 (catalog number 550274) was from Becton, Dickinson and Company (Franklin Lakes, NJ, http://www.bd.com), and rabbit anti-von Willebrand factor (catalog number A0082) was from DakoCytomation (Glostrup, Denmark, http://www.dako.com). Rabbit anti-phospho-Smad2 (catalog number AB3849), rabbit anti-GFP (catalog number AB3080), rat anti-mouse Flk-1 (catalog number MAB1669), and rat anti-CD31 conjugated with phycoerythrin (PE) (catalog number CBL1337P) were from Chemicon (Temecula, CA, http://www.chemicon.com).

Histology and Immunofluorescence Microscopy

Renal histology was assessed in 10% buffered formalin-fixed, paraffin-embedded tissue sections (4 μm) stained with periodic acid-Schiff (PAS). For immunofluorescence, tissues were fixed in 4% paraformaldehyde (Sigma-Aldrich) for 8 hours, transferred to phosphate-buffered saline (PBS) containing 30% sucrose for overnight incubation at 4°C, embedded in O.C.T. (TissueTek; Sakura Finetechnical Co., Tokyo, http://www.sakura-finetek.com/), and stored at −80°C. Frozen sections were cut (5 μm) using a cryostat (Leica, Wetzlar, Germany, http://www.leica.com/) and blocked with 2% bovine serum albumin in PBS and incubated with rabbit anti-p-p38 MAPK (1:100) or rabbit anti-p-Smad2 (1:100) antibodies, respectively, overnight at 4°C. Sections were probed with Alexa Fluor 647 goat anti-rabbit conjugate (1:2,000; Molecular Probes Inc., Eugene, OR, http://probes.invitrogen.com) or Alexa Fluor 549 goat anti-rabbit conjugate (1:2,000; Molecular Probes) and mounted with fluorescent mounting medium (DakoCytomation). Kidney endothelial cells in EGFP chimeric mice were identified by the expression of CD31 (1:100) or von Willebrand Factor (vWF; 1:400), respectively. Sections were incubated with rat anti-CD31 or rabbit anti-vWF for 60 minutes followed by goat anti-rat or goat anti-rabbit, Alexa Fluor 647 conjugate (1:2,000; Molecular Probes), or Alexa Fluor 594 conjugate (1:2,000; Invitrogen Corp., Carlsbad, CA, http://www.invitrogen.com). To detect myofibroblasts in kidney tissue, sections were probed with mouse anti-α-SMA conjugated with Cy3 (1:800) for 20 minutes at room temperature. Sections were analyzed with an Olympus Fluoview 1000 confocal microscope (Olympus, Tokyo, http://www.olympus-global.com), FV10-ASW software (version 1.3c; Olympus), oil UPLFL 60× objective (NA 1.25; Olympus) at 2× or 3× digital zoom and step size of 0.5 μm if z-series confocal microscopy analysis was applied. Channels were acquired sequentially. Contrast and brightness of the images were further adjusted in Adobe Photoshop 7.0 (Adobe Systems, Inc., San Jose, CA, http://www.adobe.com/).

Evaluation of PTC Endothelial Cell Injury

In each sample group, 40 randomly selected microscopic fields were examined in an unbiased fashion under ×400 magnification for assessment of PTC changes [35]. PTC changes were expressed the numbers of CD31-positive PTC lumina per mm2 [35].

Isolation and Culture of BM-Derived CD31(+)/Flk-1(+) Cells

For BM-derived CD31(+)/Flk-1(+) cell isolation [36], a single cell suspension of BM was labeled with rat anti-Flk-1 (1:50) followed by goat anti-rat Alex 647 antibody (1:2,000; Invitrogen) and then a PE-conjugated CD31 antibody (1:100; Chemicon). BM-derived CD31(+) cells were enriched by sorting CD31(+)/Flk1(+) cells by using FACS-Diva (BD Biosciences, San Jose, CA, http://www.bdbiosciences.com). Dead cells were excluded by a combination of scatter gates and 4,6-diamidino-2-phenylindole (DAPI) staining. Cells were plated at 10 × 106 cells/cm2 on human fibronectin-coated eight-well chamber slides (BD Biosciences). Cells were incubated in EGM-2 Bullet kit system (Clonetics, San Diego, http://www.cambrex.com) with 20% fetal bovine serum.

Histological Assessment

At week 6, kidneys from NS, ADR+Vehicle, and ADR+SB+ALK5I group mice were removed (n = 6/group), fixed in 4% paraformaldehyde, and embedded in paraffin. Tissue was cut at 5 μm and stained with hematoxylin, PAS, and Masson's trichrome. The degree of glomerulosclerosis and interstitial fibrosis were measured using Image J software (http://rsb.info.nih.gov/ij/) as previously reported [37]. The percentage of glomerulosclerosis was calculated by dividing the total area of PAS-positive staining in the glomerulus by the total area of the glomerulus. Interstitial fibrosis was quantified by dividing the area of trichrome-stained interstitium by the total cortical area. The mean value of 20 randomly selected glomeruli or five cortical fields was determined for each section. Five sections were selected and analyzed from each kidney.

Statistical Analysis

Data are mean ± SD with statistical analyses performed using one-way analysis of variance (ANOVA) from GraphPad Prism 3.0 or two-way ANOVA if appropriate (GraphPad Software, Inc., San Diego, www.graphpad.com) and post-test Tukey's analysis when appropriate. A p value less than 0.05 was considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Characteristics of ADR-Induced Nephropathy

The number of renal PTCs in kidneys from ADR-injected mice (Fig. 1B) compared with NS-injected animals (Fig. 1A) decreased with time, as demonstrated by CD31 immunostaining (Fig. 1A–1C; NS 42 days [d], 828 ± 43/mm2; ADR 3d, 688 ± 26/mm2; ADR 7d, 654 ± 44/mm2; ADR 14d, 594 ± 58/mm2; ADR 28d, 430 ± 31/mm2; ADR 42d, 296 ± 34/mm2). In the NS-treated group, glomeruli and tubulointerstitium showed normal histology (Fig. 1D). However, in the ADR-injected group, PAS staining showed severe histopathological changes including glomerular and tubulointerstitial damage, massive cast formation, glomerulosclerosis, and tubulointerstitial fibrosis (Fig. 1E). Proteinuria was prominent 7 days after ADR injection and was elevated thereafter (Fig. 1F; NS 42d, 1.48 ± 0.67 mg/24 hours; ADR 7d, 6.35 ± 0.53 mg/24 hours; ADR 14d, 10.02 ± 2.35 mg/24 hours; ADR 28d, 9.90 ± 2.93 mg/24 hours; and ADR 42d, 10.27 ± 1.68 mg/24 hours). Serum creatinine continuously increased following ADR injection and peaked at 6 weeks (Fig. 1G; NS 42d, 0.0262 ± 0.0024 mg/dl/g; ADR 3d, 0.0469 ± 0.0044 mg/dl/g; ADR 14d, 0.1176 ± 0.0185 mg/dl/g; ADR 28d, 0.1526 ± 0.0105 mg/dl/g; and ADR 42d, 0.1943 ± 0.01776 mg/dl/g). These results demonstrated that ADR administration leads to progressive renal fibrosis that by 6 weeks resembles chronic renal failure with marked functional impairment and severe histopathological alterations.

thumbnail image

Figure Figure 1.. Morphological alterations showing the loss of PTCs in ADR-induced nephropathy. NS-treated kidney (A) and a kidney from an ADR-injected mouse (B) 42 days after injection stained with CD31 (red) and 4,6-diamidino-2-phenylindole (blue). Magnification, ×600. Time course of the number of CD31-positive PTCs in ADR- and NS-treated mice (C). PAS staining demonstrates NS- (D) and ADR- (E) treated kidney. Magnification, ×400. ADR injection resulted in significant proteinuria (F) and increased serum creatinine (G). n = 6/group, values are means ± SD. Values with different letters are significantly different, p < .05. Abbreviations: ADR, adriamycin; d, day(s); hrs, hours; NS, saline vehicle; PTC, renal peritubular capillary.

Download figure to PowerPoint

BM-Derived PTC Endothelial Cells

We generated chimeric mice following BM transplantation from donor BM expressing EGFP. To investigate whether BM-derived cells are able to transdiffferentiate into PTC endothelial cells, 12 weeks after BM transplantation, a chronic, irreversible, and progressive nephropathy mouse model was induced by a single tail vein injection of ADR. Sections from recipient kidneys at days 3, 7, 14, 28, and 42 following ADR injection were examined. In PTCs, numerous EGFP(+) cells also expressed vWF, a marker of endothelial cells (Fig. 2A–2D). The colocalization of vWF(+)/EGFP(+) was confirmed by serial confocal microscopy (supplemental online Fig. 2(D)Video1). The percentage of vWF(+)/PTC endothelial cells that were vWF(+)/EGFP(+) peaked at 14 days (4.5% ± 1.2%) after ADR injection and decreased dramatically thereafter (ADR 28d, 0.88% ± 0.29%; ADR 42d, 0.48% ± 0.19%). Compared with the NS-treated group, ADR-injected mice showed a significantly higher proportion (%) of vWF(+)/EGFP(+) cells in PTCs (Fig. 2E). These findings suggest that BM-derived cells can differentiate into PTC endothelial cells in ADR-induced nephropathy.

thumbnail image

Figure Figure 2.. Bone marrow (BM)-derived PTC endothelial cells in kidneys from mice with ADR-induced nephropathy. Confocal microscopy demonstrating images of vWF(+) (A), EGFP(+) (B), DAPI(+) (C) cells, and a merged image (D) in the renal cortex from 14-day ADR-injected EGFP chimeric mice. Arrows show the coexpression of vWF in an EGFP(+) cell. Magnification, ×1,800. (E): Quantification of BM-derived cells expressing vWF in PTC endothelial cells over 42 days after ADR injection. n = 6/group, values are means ± SD. Values with different letters are significantly different, p < .05. Abbreviations: ADR, adriamycin; d, day(s); DAPI, 4,6-diamidino-2-phenylindole; EGFP, enhanced green fluorescent protein; PTC, renal peritubular capillary; NS, saline vehicle; vWF, von Willebrand factor.

Download figure to PowerPoint

BM-Derived Myofibroblasts

To determine whether BM-derived cells can differentiate into renal interstitial myofibroblasts, α-SMA was used as a myofibroblast marker, and sections from recipient kidneys were examined by serial confocal microscopy. In ADR-injected mice, the number of α-SMA(+)/EGFP(+) cells increased in a time-dependent fashion. The number of α-SMA(+)/EGFP(+) cells per mm2 of cortex increased with time, together with the percentage of α-SMA(+) cells, which were EGFP(+) (Fig. 3A–3E, supplemental online Fig. 3(D)Video2). By 6 weeks after ADR injection, 27.4% ± 3.4% of myofibroblasts were found to be derived from the BM (Fig. 3E). In the NS-treated group, renal interstitial myofibroblasts were rarely observed (0.04% ± 0.05%).

thumbnail image

Figure Figure 3.. Bone marrow (BM)-derived myofibroblasts in ADR-induced nephropathy. Confocal microscopy demonstrating images of α-SMA(+) (A), EGFP(+) (B), DAPI(+) (C) cells, and a merged image (D) in the renal cortex from 42-day ADR-injected EGFP chimeric mice. Magnification, ×600. (E): Quantification of BM-derived cells expressing α-SMA in tubulointerstitium in ADR-induced nephropathy. n = 6/group, values are means ± SD. Two-way analysis of variance: time, p < .001; treatment, p < .001; interaction, p < .001. Abbreviations: α-SMA, α-smooth muscle actin; ADR, adriamycin; d, day(s); DAPI, 4,6-diamidino-2-phenylindole; EGFP, enhanced green fluorescent protein; NS, saline vehicle; wks, weeks.

Download figure to PowerPoint

BM-Derived Cells Coexpressing Endothelial and Myofibroblast Markers

Confocal microscopy demonstrated triple-positive vWF(+)/α-SMA(+)/EGFP(+) cells in ADR-induced nephropathy in recipient mice (Fig. 4A–4D, supplemental online Fig. 4(H)Video3). The number of vWF(+)/α-SMA (+)/EGFP(+) in the renal cortical area was significantly increased (5.6 ± 0.5/mm2 vs. 0.45 ± 0.2/mm2, p < .0001) at 4 weeks after ADR injection compared with the NS-treated group. Furthermore, CD31(+)/α-SMA(+)/EGFP(+)-colocalized cells (Fig. 4E–4H) were also evident 4 weeks after ADR injection and thereafter (NS 4 weeks, 0.3 ± 0.5/mm2; ADR 4 weeks, 5.7 ± 1.2/mm2; NS 6 weeks, 0.4 ± 0.5/mm2; ADR 6 weeks, 14.5 ± 2.4/mm2). The above data suggest that the number of BM-derived cells coexpressing endothelial cell and myofibroblast markers is dramatically increased during the development of fibrosis and renopathology.

thumbnail image

Figure Figure 4.. Evidence of bone marrow (BM)-derived endothelial-myofibroblast transformation in ADR-induced nephropathy. Confocal microscopy demonstrating images of vWF(+) (A), EGFP(+) (B), and α-SMA(+) (C) cells, and a merged image (D) in the renal cortex from 28-day adriamycin-injected EGFP chimeric mice. BM-derived endothelial-myofibroblast transformation was also demonstrated by the triple expression (arrows) of CD31 (E), EGFP (F), α-SMA (G), and a merged image (H). Magnification, ×1,800. The inset box (H) shows the triple-labeled cell at higher magnification (×3,600). Abbreviations: α-SMA, α-smooth muscle actin; EGFP, enhanced green fluorescent protein; vWF, von Willebrand factor.

Download figure to PowerPoint

BM-Derived CD31(+)/FLK-1(+) Cells Acquire the Myofibroblast Marker α-SMA

To investigate whether BM-derived endothelial cells can express the myofibroblast marker α-SMA, CD31(+)/Flk-1(+) cells were isolated from BM and cultured in endothelial cell growth medium with 20% fetal bovine serum. In support of the in vivo observations, endothelial cells gradually lost the expression of the endothelial cell marker CD31 and acquired the myofibroblast marker α-SMA in an autoinduction fashion (Fig. 5A–5Q). This evidence suggests that BM-derived endothelial-myofibroblast transformation can occur in an in vitro setting.

thumbnail image

Figure Figure 5.. CD31(+)/Flk-1(+) cells were isolated from bone marrow (BM) and incubated in the EGM-2 Bullet kit system with 20% fetal bovine serum. After 7, 14, 21, and 42 days in culture, the cells were immunostained with anti-CD31 and anti-α-SMA antibodies. Confocal microscopy demonstrated that cultured BM-derived CD31(+)/Flk-1(+) cells gradually lose CD31 expression (B, F, J, N) and acquire α-SMA expression (C, G, K, O) in an autoinduction fashion. (Q): Quantification of percentage of CD31(+) and α-SMA(+) cells. Values are means ± SD. Two-way analysis of variance: time, p < .001; treatment, p < .001; interaction, p < .001. Magnification, ×600. Abbreviations: α-SMA, α-smooth muscle actin; d, days; DAPI, 4,6-diamidino-2-phenylindole.

Download figure to PowerPoint

p38 MAPK and TGF-β1/Smad Signaling Pathways Are Activated in BM-Derived Endothelial Cells and Myofibroblasts

Myofibroblasts.

Sections from recipient kidneys were examined by confocal microscopy to investigate whether the p-p38 MAPK and TGF-β1/Smad signaling pathways are activated in bone marrow-derived myofibroblasts. Triple-positive p-p38 MAPK/α-SMA(+)/EGFP(+) cells (Fig. 6A–6D) and p-Smad2 (+)/α-SMA(+)/EGFP(+) cells (Fig. 6E–6H, supplemental online Fig. 4(H)Video4) were detected in myofibroblasts. p-p38MAPK and p-Smad2-positive cells were confirmed by DAPI nuclear staining. Quantitative analysis (Fig. 6I) demonstrated that 0.7% ± 0.5% of BM-derived myofibroblasts coexpressed p-p38MAPK, α-SMA, and EGFP in NS-treated kidneys compared with 40.0% ± 4.9% in kidneys from 14-day post-ADR-injected mice, respectively. However, there was no significant difference between the percentage of EGFP(+) cells coexpressing p-p38 MAPK and α-SMA and the percentage of EGFP(−) cells coexpressing p-p38 MAPK and CD31.

A similar level was observed in the expression of p-Smad2 in NS-treated and ADR-treated animals (Fig. 6J). Taken together, these results suggest that the p-p38 MAPK and TGF-β1/Smad signaling pathways are activated in BM-derived myofibroblasts. There were no significant differences in the activities of p-p38 MAPK or TGF-β1/Smad2 between BM-derived myofibroblasts and endogenous resident myofibroblasts in ADR-induced nephropathy.

PTC Endothelial Cells.

The activity of the p38 MAPK and TGF-β1/Smad signaling pathways in BM-derived PTC endothelial cells was also investigated by confocal microscopy. p38 MAPK and TGF-β1/Smad signaling pathways (data not shown) were activated in BM-derived myofibroblasts. More importantly, there was no significant difference between the percentage of p-p38MAPK(+)/α-CD31(+)/EGFP(+) and p-p38MAPK(+)/CD31(+)/EGFP(−) cells or between the percentage of p-Smad2(+)/CD31(+)/EGFP(+) and p-Smad2(+)/CD31(+)/EGFP(−) cells (Fig. 6). These findings indicate that the p38 MAPK and TGF-β1/Smad signaling pathways are activated in BM-derived PTC endothelial cells in a similar manner to endogenous resident PTC endothelial cells in ADR-induced nephropathy.

thumbnail image

Figure Figure 6.. p-p38 MAPK and Smad signaling pathways are activated in BM-derived myofibroblasts in ADR kidneys. p38 MAPK (A–E) and transforming growth factor-β1/Smad (G–K) signaling pathway activation in bone marrow-derived interstitial cells. Confocal microscopy showing images of p-p38 MAPK (+) (red) (A), EGFP (+) (B, H), α-SMA (+) (blue) (C, I), p-Smad2 (+) (red) (G) cells, and merged images (D, J) within the renal interstitium 2 weeks after ADR injection (arrows). The positive staining of p-p38 and p-Smad2 was confirmed by DAPI nuclear staining (E, K). Quantification of p-p38 (F) and p-Smad2 (L) expression in EGFP(+) or EGFP(−)/α-SMA(+) myofibroblasts. Data are means ± SD. Values with different letters are significantly different, p < .05. Magnification, ×600. Abbreviations: α-SMA, α-smooth muscle actin; ADR, adriamycin; DAPI, 4,6-diamidino-2-phenylindole; EGFP, enhanced green fluorescent protein; MAPK, mitogen-activated protein kinase; NS, saline vehicle; p, phospho-.

Download figure to PowerPoint

The Effects of Blockade of p38 MAPK and TGF-β1/Smad Signaling Pathways on BM-Derived PTC Endothelial Cells and Myofibroblasts

We investigated the role of p38 MAPK and TGF-β1/Smad signaling pathways in BM-derived endothelial-myofibroblast transformation in ADR-induced nephropathy. Two weeks after ADR injection, recipient mice were treated with SB203580, a p38 MAPK inhibitor, and ALK5I, a TGF-β receptor I inhibitor by osmotic pump for a further 4 weeks. Coadministration of SB203580 and ALK5I not only reduced glomerular and tubulointerstitial injury, glomerulosclerosis, and tubulointerstitial fibrosis (Fig. 7A, 7B) but also decreased proteinuria (Fig. 7C) and serum creatinine (Fig. 7D) in ADR-treated mice. Compared with vehicle treatment, the coadministration of SB203580 and ALK5I reduced the loss of PTCs in ADR-induced nephropathy (Fig. 7E). Coadministration of SB203580 and ALK5I to ADR-injected mice increased the number of BM-derived PTC endothelial cells (Fig. 7F), decreased the number of BM-derived cells that coexpress endothelial and myofibroblast markers (Fig. 7G) and BM-derived interstitial myofibroblasts, and reduced the accumulation of total interstitial myofibroblasts compared with the vehicle-treated group (Fig. 7H–7J).

thumbnail image

Figure Figure 7.. Blocking the p38 mitogen-activated protein kinase and transforming growth factor-β/Smad signaling pathways reduced ADR-induced nephropathy. Quantitation of glomerulosclerosis (A), interstitial fibrosis (B), proteinuria (C), serum creatinine (D), the number of CD31(+) PTC lumina (E), percentage of vWF(+) PTC endothelial cells expressing EGFP (F), the number of bone marrow-derived cells coexpressing CD31 and α-SMA (G), and percentage of α-SMA(+) cells expressing EGFP (J). Confocal microscopy demonstrates images of renal interstitium from the ADR+vehicle group (H) and ADR+SB+ALK5I group (I). In (H) and (I), 4,6-diamidino-2-phenylindole (blue), EGFP (green), and α-SMA (red). n = 6/group, values are means ± SD. Values with different letters are significantly different, p < .05. Magnification, ×600. Abbreviations: α-SMA, α-smooth muscle actin; ADR, adriamycin; EGFP, enhanced green fluorescent protein; PTC, renal peritubular capillary; NS, saline vehicle; SB, SB203580; vWF, von Willebrand factor.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

The present study provides evidence that BM-derived cells can coexpress endothelial cell and myofibroblast markers during the development of renal fibrosis in a chronic, irreversible, and progressive disease setting. BM-derived CD31(+)/Flk-1(+) cells were found to lose expression of the endothelial cell marker CD31 and acquire the myofibroblast marker α-SMA in vitro in an autoinduction fashion. In the in vivo setting, BM-derived cells contributed to more than 30% of renal interstitial myofibroblasts in mice with ADR nephropathy. Our data suggest that these BM-derived cells may contribute to the progression of renal interstitial fibrosis via endothelial-myofibroblast transformation in pathological circumstances. p38 MAPK and TGFβ-1/Smad2 signaling pathways were activated in BM-derived PTC endothelial cells and BM-derived myofibroblasts in ADR-induced nephropathy. The coadministration of SB203580 and ALK5I significantly restored BM-derived PTC endothelial cells, reduced the number of BM-derived cells coexpressing endothelial and myofibroblast markers, and reduced BM-derived myofibroblast accumulation, leading to a restoration of renal function. These results suggest that BM-derived cells may contribute to renal fibrosis through the process of endothelial-myofibroblast transformation and that the p38 MAPK and TGFβ-1/Smad2 signaling pathways may be involved in this process. This study raises the possibility that novel therapies to prevent or retard the progression of ESRD could target BM-derived endothelial-myofibroblast-transformation that accelerates renal interstitial fibrosis.

Myofibroblasts are the active form of fibroblasts that are thought to be the main source of increased ECM deposition in renal fibrosis [38, 39]. Renal interstitial myofibroblasts may be derived from local fibroblasts, tubular epithelial cells, or smooth muscle cells. Recent studies [13, 14, 24, 25] suggest that BM-derived cells may be a source of fibroblasts and myofibroblasts in fibrotic lesions. Iwano et al. [25] demonstrated that during experimental fibrosis, 15% of renal interstitial fibroblasts were derived from BM-derived cells in a mouse model of experimental hydronephrosis. In comparison, the present study used a more progressive and chronic mouse model of ADR nephropathy. We demonstrate that 28 days after ADR injection, the number of BM-derived myofibroblasts begins to dramatically increase. By 42 days, more than 30% of renal interstitial myofibroblasts were found to be derived from the BM.

The mechanisms controlling BM cell differentiation into myofibroblasts remains largely unclear. Arciniegas et al. [40] demonstrated that TGF-β1 can induce aortic endothelial cells to differentiate into a smooth muscle-like phenotype (α-SMA-positive) in vitro, suggesting a novel role for TGF-β1 in atherogenesis. Aortic valve endothelial cells can also undergo TGF-β2-mediated or non-TGF-β2-mediated mesenchymal transformation (α-SMA-positive) in a manner that resembles one of the early events during valve development. These studies suggest that aortic valve endothelial cells may serve to replenish the interstitial cells, which in turn synthesize the ECM [41]. Moreover, embryonic endothelial cells have been observed to transdifferentiate into mesenchymal cells expressing α-SMA in vitro and in vivo [42], and vascular endothelium-derived cells contain α-SMA in restenosis [43], inflammation, and hypertension [44]. These studies provide evidence to support the importance of endothelial-mesenchymal transformation in both development and various pathological processes [42]. In the present study, the number of BM-derived PTC endothelial cells peaked 14 days after ADR injection and then declined thereafter. In addition, the number of BM-derived myofibroblasts dramatically increased 28 days after ADR injection, and EGFP(+)/vWF(+)/α-SMA and EGFP(+)/CD31(+)/α-SMA cells were evident in the renal interstitium in mice with ADR nephropathy. In vitro analysis demonstrated that BM-derived CD31(+)/Flk-1(+) cells can gradually lose expression of the endothelial cell marker CD31 and acquire the myofibroblast marker α-SMA. These findings suggest that BM-derived cells may undergo endothelial-myofibroblast transformation under certain pathological process, and endothelial-myofibroblast transformation may contribute to renal interstitial fibrosis, thereby playing an important role in the progression of ESRD.

Previous studies [45, 46] suggest that BM stromal cells are progenitors for tissue fibroblasts that can shuttle through the circulation in the form of fibrocytes, characterized by a distinctive phenotype (collagen+/CD34+), to populate peripheral organs and mediate fibrosis. BM stromal cells also express α-SMA [47]. Furthermore, CD34 is also a marker for angioblasts [48]. In this study, we cannot rule out the possibility that renal interstitial EGFP(+)/vWF(+) cells and EGFP(+)/vWF(+)/α-SMA(+) cells were derived from BM-derived stromal cells that relocate into the renal interstitium through circulating fibrocytes.

Recently, Roufosse et al. [49] reported that 8.6% of SMA-positive interstitial cells were derived from BM in the hydronephrotic obstructed kidney; however, BM-derived cells did not significantly contribute to collagen I synthesis. However, Okada et al. [50] demonstrated that neither α-SMA(+) nor vimentin(+) fibroblasts predominately expressed α1(I) procollagen mRNA in Goodpasture syndrome rats, suggesting that the expression of α-SMA is not parallel to that of collagen I in renal interstitial fibroblasts or myofibroblasts.

The p38 MAPK and TGF-β1/Smad signaling pathways have been implicated in inflammation and fibrosis, respectively. The TGF-β1, TGF-β2 and TGF-β3 isoforms are good inducer of endothelial-mesenchymal transformation [40, 41, 51]. p38 MAPK is required for TGF-β1-induced epithelial-mesenchymal transformation [52, 53], thought to be another origin of renal interstitial myofibroblasts. The importance of p38 MAPK in epithelial-mesenchymal transformation suggests that p38 MAPK may play a role in endothelial-mesenchymal transformation. In our study, the inhibition of the p38 MAPK and TGF-β1/Smad signaling pathways increased the number of BM-derived PTC endothelial cells following injury, reduced BM-derived myofibroblasts and BM-derived cells coexpressing endothelial and myofibroblast markers. This data suggests that the p38 MAPK and TGF-β1/Smad signaling pathways may be involved in the contribution of BM-derived cells to the progression of renal interstitial fibrosis.

The contribution of BM-derived cells via a process of cell-cell fusion is not clear. Wang et al. demonstrated that cell-cell fusion is the major source of BM-derived hepatocytes [54]. BM-derived cells may also contribute to BM-derived Purkinje neurons and cardiac myocytes through cell-cell fusion rather than cell differentiation [55, [56], [57]58]. However, whether cell-cell fusion is the basic principle of BM-derived cell contribution to regenerating tissues is controversial in that cell-cell fusion events are observed to occur at a very low frequency (<1% of cardiomyocytes and <0.1% of hepatocytes or Purkinje cells, 56). In our study, engraftment of BM-derived cells was observed at relatively high frequencies depending on the cell type (endothelial cells 4.3% [29]; myofibroblasts 30%). This suggests that cell fusion is not the major contributor of BM-derived endothelial cells, myofibroblasts and BM-derived cells coexpressing endothelial and (myo)fibroblast markers in our fibrotic disease model.

This study provides evidence that BM-derived cells can differentiate into endothelial cells, (myo)fibroblasts, and cells coexpressing endothelial and (myo)fibroblast markers, together with the confirmation of BM-derived endothelium to myofibroblast transformation in vitro. However, to clarify this process in an in vivo setting, it is necessary to provide direct evidence of endothelium to (myo)fibroblast conversion by permanent genetic marking and lineage tracing of endothelial cells in a fibrotic disease model.

In summary, BM-derived cells may play an important role in renal interstitial fibrosis via two events, namely, differentiation into PTC endothelial cells and endothelial-mesenchymal transformation. The p38 MAPK and TGF-β1/Smad signaling pathways appear to be involved in these processes. The coadministration of SB203580 and ALK5I to mice with ADR nephropathy was found to preserve BM-derived PTC endothelial cells and reduce endothelial-mesenchymal transformation, thereby decreasing BM-derived interstitial myofibroblasts. The evidence of BM-derived endothelial-mesenchymal transformation and the involvement of p38 MAPK and TGF-β1/Smad signaling pathways in this transformation may be helpful not only for the design of novel therapies to prevent or retard the progression of renal fibrosis but also for manipulating adult stem cells for the treatment of renal disease.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

These studies were supported by a Kidney Health Australia Bootle Bequest. We acknowledge the members of the Renal Regeneration Consortium for their support. J.L. is the recipient of a Chinese Government Scholarship for Outstanding Oversees Students and a Monash Graduate Scholarship. Confocal imaging was performed at the Monash MicroImaging Facility at Monash University. This work was supported by the Australian Kidney Foundation, Australian National Health and Medical Research Foundation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
Ricardo_Supp_Legends.pdf19KSupplemental Legends
SC060133_SuppFile1.wmv76KSupplemental Video 1
SC060133_SuppFile2.wmv4654KSupplemental Video 2
SC060133_SuppFile3.wmv193KSupplemental Video 3
SC060133_SuppFile4.wmv229KSupplemental Video 4

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.