Transplantation of bone marrow-derived mesenchymal stem cells (BMSCs) can repair acute kidney injury (AKI), but with limited effect. We test the hypothesis that CXCR4 overexpression improves the repair ability of BMSCs and that this is related to increased homing of BMSCs and increased release of cytokines. Hypoxia/reoxygenation-pretreated renal tubular epithelial cells (HR-RTECs) were used. BMSCs, null-BMSCs, and CXCR4-BMSCs were cocultured with HR-RTECs. The number of migrating BMSCs was counted. Proliferating cell nuclear antigen (PCNA) expression, cell death, and expressions of cleaved caspase-3 and Bcl-2 in cocultured HR-RTECs were measured. Cytokeratin 18 (CK18) expression and cytokine secretions of the BMSCs cultured with HR-RTEC supernatant were detected. BMSC homing, renal function, proliferation, and cell death of tubular cells were assayed in the AKI mouse model. CXCR4-BMSCs showed a remarkable expression of CXCR4. Stromal cell-derived factor-1 in the HR-RTEC supernatant was increased. Migration of BMSCs was CXCR4-dependent. Proportions of CK18+ cells in BMSCs, null-BMSCs, and CXCR4-BMSCs showed no difference. However, CXCR4 overexpression in BMSCs stimulated secretion of bone morphogenetic protein-7, hepatocyte growth factor, and interleukin 10. The neutralizing anti-CXCR4 antibody AMD3100 abolished this. In cocultured HR-RTECs the proportions of PCNA+ cells and Bcl-2 expression were enhanced; however, the proportion of annexin V+ cells and expression of cleaved caspase-3 were reduced. The in vivo study showed increased homing of CXCR4-BMSCs in kidneys, which was associated with improved renal function, reduced acute tubular necrosis scoring, accelerated mitogenic response of tubular cells, and reduced tubular cell death. The enhanced homing and paracrine actions of BMSCs with CXCR4 overexpression suggest beneficial effects of such cells in BMSC-based therapy for AKI.
- bone marrow-derived mesenchymal stem cells
- acute kidney injury
- renal tubular epithelial cells
acute kidney injury (aki) is a common clinical disease with high morbidity and mortality. Bone marrow-derived mesenchymal stem cells (BMSCs) can migrate to the injured kidney and take part in AKI repair by transdifferentiating into local cell types (21, 26, 28) or by paracrine mechanisms (17, 30, 36). However, it has been suggested that their repair efficiency is limited (14, 33), mainly because only a small proportion of the implanted BMSCs migrate to the injured kidney, whereas the majority of cells remain entrapped in other blood-rich organs (12, 39). Moreover, the differentiation potential and the paracrine ability of BMSCs migrating to the AKI area also affect the AKI repair effect.
Stromal cell-derived factor-1 (SDF-1) and its cellular receptor, CXCR4, have been demonstrated to direct the migration of stem cells associated with injury repair in many species and tissue types (2, 9, 16, 38, 41). Expression of SDF-1 is predominantly promoted under ischemic conditions including AKI (2, 19, 31, 35). However, the surface expression of CXCR4 in BMSCs is markedly reduced during ex vivo expansion (32, 40), which may lead to low efficiency of infused BMSCs homing toward the damaged tissues (18, 37). Regulation of CXCR4 expression is expected to influence the directional homing of infused BMSCs. In addition, the SDF-1/CXCR4 axis is also reported to be required for the paracrine actions of BMSCs (7, 23). Regulation of CXCR4 expression is also expected to contribute to the paracrine actions of BMSCs.
In this study, CXCR4-overexpressing BMSCs (CXCR4-BMSCs carrying CXCR4 and eGFP genes; eGFP as the tracer gene) were constructed through lentivirus infection, and a hypoxia/reoxygenation (HR) model of renal tubular epithelial cells (RTECs) was constructed to simulate the kidney cell damage of AKI in vitro. We demonstrated that overexpression of CXCR4 markedly augmented BMSC migration in response to HR-RTECs within the AKI microenvironment. We show increased paracrine actions of CXCR4 overexpressing BMSCs. The greater migration and paracrine actions of CXCR4-BMSCs both contribute to better AKI repair of CXCR4-BMSC transplantation. At the same time, we observed an increased proportion of PCNA+ cells, a decreased proportion of annexin V+ cells, decreased cleaved caspase-3, and increased Bcl-2 expression in cocultured HR-RTECs. Furthermore, CXCR4-BMSC-based therapy led to enhanced migration of BMSCs to the kidney area, improved renal function, and greater tubular repair in the AKI mouse models.
MATERIALS AND METHODS
Generation of lentivirus vector.
By using gateway technology, the attB1-CXCR4-attB2 (or attB1-eGFP-attB2) gene sequence was amplified by PCR. The specific primer sequences are listed in Table 1. After BP (a recombination between attB and attP sites) and LR (a recombination between attL and attR sites) reactions by using a BP Clonase II Enzyme Mix and LR Clonase II Enzyme Mix (both from Invitrogen, Carlsbad, CA), products were directly used for bacterial transformation and colony PCR screen to obtain the target plasmids (i.e., pLV.EX3d.P/puro-CMV>CXCR4>IRES/eGFP and pLV.EX2d.P/Neo-CMV>eGFP). The plasmids together with pLV/helper-SL3, pLV/helper-SL4, and pLV/helper-SL5 were cotransfected into 293FT cells (SIBCB, CAS, China) with lipofectamine 2000 (Invitrogen) to obtain lenti-CXCR4-eGFP/puro (lenti-CXCR4) or lenti-null-eGFP/neo (lenti-null).
Mouse BMSCs (ATCC, Manassas, VA) were transfected using dilutions of concentrated lenti-CXCR4 or lenti-null (1.5 ml/well) and then selected with 500 μg/ml puromycin (or neomycin). Because the lentivirus vectors both express eGFP, the transfection efficiency of BMSCs was evaluated with fluorescence microscopy. Both the CXCR4-BMSCs (carrying CXCR4 and eGFP genes) and null-BMSCs (carrying only the eGFP gene) were prepared for further use.
CXCR4 expression in transfected BMSCs.
RT-PCR was performed to detect the mRNA level of CXCR4 in CXCR4-BMSCs and null-BMSCs. Sequences of PCR primers are listed in Table 2.
The protein expression of CXCR4 was tested by Western blot. CXCR4-BMSCs and null-BMSCs were scraped in RIPA lysis buffer that included protease inhibitors. After SDS-PAGE, the protein was transferred to a polyvinylidene difluoride (PVDF) membrane and blocked with Tris-buffered saline-Tween containing 1% BSA (10 mM Tris-HCl pH 7.14, 150 mM NaCl, 1% Tween-20) at room temperature for 4 h. The PVDF membrane was incubated with rabbit anti-mouse CXCR4 monoclonal antibody (Biosynthesis Biotechnology, Beijing, China) at 4°C overnight. Horseradish-peroxidase (HRP)-labeled goat anti-rabbit IgG (Santa Cruz Biotechnology, Santa Cruz, CA) was added and incubated for another hour at room temperature. The mixture of reacted enhanced chemiluminescence was added to the PVDF membrane for 1–2 min, and then placed into a FluorChem HD2 gel image analysis system for observing. The intensity value of the band was measured. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the internal reference. The protein expression was the ratio of the two band intensity values.
For cell surface expression of CXCR4, cultured CXCR4-BMSCs and null-BMSCs were incubated with an Alexa Fluor 647 anti-mouse CD184 (CXCR4) or an isotype-matched control (Alexa Fluor 647 Rat IgG2b, κ Isotype Ctrl) (both from Biolegend, San Diego, CA) in a darkroom at 4°C for 30 min. Flow cytometric analysis of CXCR4 was performed using a MACS Quant flow cytometer (Miltenyi Biotech, Bergisch Gladbach, Germany).
Hypoxia/reoxygenation pretreated RTECs.
On the basis of previous experiments (24), 4 × 105 cells/well of mouse RTECs (ATCC) were seeded in 6-well plates and cultured under hypoxic conditions (92% N2, 3% O2, and 5% CO2) using the AnaeroPack System (Mitsubishi Gas Chemical, Tokyo, Japan) with glucose-free DMEM (Invitrogen) for 12 h. Then they were cultured under reoxygenation conditions (95% air and 5% CO2) with low-glucose DMEM for an additional 12 h. These RTECs were termed hypoxia/reoxygenation pretreated RTECs (HR-RTECs). Normoxia-preconditioned (for 24 h in 95% air, 5% CO2) RTECs (termed N-RTECs) in the 6-well plates were used as a control.
SDF-1 level in the RTEC culture supernatant.
The concentration of SDF-1 was measured in supernatants obtained from N-RTECs and HR-RTECs using an SDF-1 ELISA Detection Kit (R&D Systems China, Shanghai, China) according to the manufacturer's instructions.
Chemotaxis assay in vitro.
BMSC migration was analyzed using the Transwell chamber with an 8-μm pore polycarbonate filter (Corning, Corning, NY). After completion of HR-RTEC and N-RTEC preparations, the 6-well plates with HR-RTECs or N-RTECs were removed from the incubator and the Transwell chambers were inserted into the plates. Mouse BMSCs, null-BMSCs, and CXCR4-BMSCs resuspended in the medium were placed in the upper chambers. The plating density for BMSCs was 1.8 × 105 cells/chamber. The chemotaxis assay was used for the B/N group (BMSCs/N-RTECs), B/H group (BMSCs/HR-RTECs), N/H group (null-BMSCs/HR-RTECs), C/H group (CXCR4-BMSCs/HR-RTECs), and C/H + AMD3100 group [CXCR4-BMSCs were first preincubated with 5 mg/ml AMD3100, a neutralizing anti-CXCR4 antibody (Sigma-Aldrich, St Louis, MO) for 30 min on ice and then added to the upper chamber]. All groups were then incubated for 12 h at 37°C. Cells that had migrated beneath the surface of the filters were stained with 1% crystal violet and counted microscopically.
Determination of renal differentiation of BMSCs in vitro.
BMSCs, null-BMSCs, and CXCR4-BMSCs were all cultured with the culture supernatant obtained from HR-RTECs at 37°C for 12 h. For the inhibition experiment, CXCR4-BMSCs were first preincubated with 5 mg/ml AMD3100 for 30 min on ice. Immunofluorescent staining was performed to detect whether BMSCs express cytokeratin 18 (CK18), the specific marker of RTECs. The cells were fixed in 4% paraformaldehyde for 30 min. The sections were incubated with rabbit anti-mouse CK18 polyclonal antibody (Biogot Technology) at 37°C for 1 h. Sections were then incubated with Cy3-conjugated goat anti-rabbit secondary antibody (Jackson ImmunoResearch, West Grove, PA) at 37°C for another 1 h in the darkroom. Hoechest 33342 (Sigma-Aldrich) was used to redye the nucleus for 30 min in the darkroom. Orange-red fluorescence in the cytoplasm represented the CK18-positive (CK18+) cells. The nucleus of all cells showed blue. Fifteen nonoverlapping view fields (200× magnification) were randomly chosen for each section to determine the proportion of CK18+ cells using the Image-Pro Plus 6.0 program (Media Cybernetics, Rockville, MD). The mean value of the proportion of CK18+ cells for each section was used for statistical analysis.
After culture with the HR-RTEC supernatant, the concentrations of bone morphogenetic protein-7 (BMP-7), hepatocyte growth factor (HGF), and interleukin 10 (IL-10) in the supernatants of BMSCs, null-BMSCs, and CXCR4-BMSCs were determined by ELISA using a commercially available ELISA kit (R&D Systems, China) according to the manufacturer's instructions.
Immunofluorescent staining for HR-RTECs in the lower chambers.
After the chemotaxis assay, the HR-RTECs in the lower chambers of the B/H, C/H, and N/H groups were subjected to immunofluorescent staining for proliferating cell nuclear antigen (PCNA), a marker of mitogenesis. HR-RTECs cultured for 12 h at 37°C with only low-glucose DMEM added into the upper chamber were used as the control group. The procedure, described above, was performed with rabbit anti-mouse PCNA polyclonal antibody (Biogot Technology) as the primary antibody. Orange-red fluorescence in the nucleus represented the PCNA-positive (PCNA+) cells. The cell death of RTECs plays an important role in the pathophysiological course of AKI (22). In our experiment, we also tested the cell death of cocultured HR-RTECs using an Annexin V-FITC Detection Kit (Beyotime Biotechnology, Shanghai, China) according to the manufacturer's instructions. Green fluorescence of the cell membrane represented annexin V-positive (annexin V+) cells. Fifteen nonoverlapping view fields (200× magnification) were randomly selected for each section, and the proportion of PCNA+ cells and annexin V+ cells was analyzed by an image analysis system using Image-Pro Plus 6.0 (Media Cybernetics).
Western blot for cleaved caspase-3 and Bcl-2 in HR-RTECs.
The HR-RTECs in the B/H, C/H, N/H, and control groups were also used for Western blot analysis. After coculture, the Transwell chambers were removed and the lysis buffer was added into the 6-well plates for the full lysis of HR-RTECs. The protocols, described above, were performed with rabbit anti-mouse cleaved caspase-3 monoclonal antibody and rabbit anti-mouse Bcl-2 monoclonal antibody (both from Cell Signaling Technology, Danvers, MA) as the primary antibodies. GAPDH was used as the internal reference and the protein expression was the ratio of the two band intensity values.
Induction of ischemia/reperfusion-AKI (I/R-AKI).
Models of I/R-AKI were created in 6- to 8-wk-old C57BL/6J mice (Experimental Animal Center of the Second Military Medical University; 20 ± 2 g) by clamping both renal pedicles for 30 min followed by clamp release to allow reperfusion. At 24 h after surgery, BMSCs (2 × 106 cells suspended in 0.2 ml of low-glucose DMEM) were injected into the tail vein. To be able to detect the homing of BMSCs to the injured kidney, BMSCs, null-BMSCs, and CXCR4-BMSCs were all labeled with 5-bromo-2′-bromodeoxyuridine (BrdU; Sigma-Aldrich) as described earlier (27). Preliminary experiments showed that the viability and growth of these labeled BMSCs were not adversely affected by this labeling procedure (data not shown).
The mice were randomly assigned to four experimental groups (n = 6 for each group): the model group, the BMSC group, the null-BMSC group, and the CXCR4-BMSC group. All mice were kept at favorable temperatures and humidity with an unlimited supply of water and food. All procedures were performed in accordance with the Guidelines for Animal Experimentation of The Second Military Medical University (Shanghai, China).
Blood biochemical indicators.
Seven days after BMSC transplantation, mice were killed and blood samples were collected. Concentrations of blood urea nitrogen (BUN) and serum creatinine (Scr) were measured using a Beckman Automatic Biochemistry Analyzer (Beckman Coulter).
Acute tubular necrosis scoring.
The killed mice were first transcardial-flushed with PBS, and the excised kidneys were flushed with PBS again. Kidneys were fixed in 10% phosphate-buffered formalin, sectioned, and then stained with periodic acid-Schiff (PAS) by standard methods. A renal pathologist conducted the histological examinations in a blinded manner. Acute tubular necrosis (ATN) scoring was quantified by counting the percentage of tubules that displayed cell necrosis, brush border loss, naked basement membrane, and vacuolar degeneration as follows: 0, none; 1, ≤10%; 2, 11–25%; 3, 26–45%; 4, 46–75%; 5, ≥76%. Fifteen nonoverlapping view fields (200× magnification) were reviewed for each section. Semiquantitative analysis was performed and the mean value was used to indicate the ATN scoring.
In vivo migration of BMSCs.
The kidney tissue sections were subject to immunohistochemical staining for BrdU. Briefly, after deparaffination, the sections were dehydrated with graded ethanol and incubated with a peroxidase-blocking reagent for 30 min. The sections were incubated with rabbit anti-mouse BrdU monoclonal antibody (Sigma-Aldrich) at 4°C for 16 h. After washing with PBS, the sections were incubated with HRP-labeled anti-mouse lgG (Santa Cruz Biotechnology) at 37°C for 1 h. The sections were then incubated in diaminobenzidine chromogenic substrate liquid for 15 min, and the lining was dyed with hematoxylin. After sealing with resin, the sections were stored at room temperature for use. The number of BrdU-positive (BrdU+) cells was carried out by counting the number of positive nuclei that appeared brown. Fifteen nonoverlapping view fields (200× magnification) were selected, and the mean value of the proportion of BrdU+ cells for each section was used for statistical analysis.
Tubular repair and assessment of cell death.
PCNA+ cells in the tubular tissues were tested by immunohistochemical staining. The protocols were described above, albeit with rabbit anti-mouse PCNA polyclonal antibody as the primary antibody. Fifteen nonoverlapping view fields (200× magnification) were selected, and the mean value of the proportion of PCNA+ cells for each section was used for statistical analysis. Terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) was performed using an in situ cell apoptosis detection kit (Sigma-Aldrich) according to the manufacturer's instructions to detect the cell death of tubular cells (10, 11, 22). The TUNEL-positive (TUNEL+) nuclei appeared brown under the microscope, whereas the normal nuclei appeared blue. Fifteen nonoverlapping random fields (200× magnification) were selected, and the number of TUNEL+ nuclei per field was counted. The results were expressed as proportion of TUNEL+ cells.
The fate of CXCR4-BMSCs migrating to the injured kidney.
Mice of the BMSC group and the CXCR4-BMSC group (n = 6 each) were included for this test. Fourteen days after transplantation, the mice were killed and the kidney sections were prepared as described above. The retention of BrdU+ cells in the kidney tissues was analyzed by immunohistochemistry (protocols described above).
Results are expressed as means ± SD. A Student's t-test was performed to analyze the differences between the two groups. Multiple-group comparison was performed using one-way ANOVA followed by a Student-Newman-Keuls test. SPSS 17.0 statistical software was used for analysis. P < 0.05 was considered statistically significant.
CXCR4-BMSCs and null-BMSCs.
After 3 or 4 days of drug selection, CXCR4-BMSCs or null-BMSCs were successfully obtained (Fig. 1A). The mRNA level and protein expression of CXCR4 in CXCR4-BMSCs were both high, but they were low or undetectable in null-BMSCs (Fig. 1, B and C). To examine the cell surface expression of CXCR4, flow cytometry was performed, revealing that a proportion of CXCR4-positive cells was significantly higher in CXCR4-BMSCs (Fig. 1D).
CXCR4-BMSCs are more responsive to HR-RTECs in vitro.
Mouse RTECs were exposed to HR to generate cell damage to simulate AKI in vitro. This led to an increased concentration of SDF-1 in the culture supernatant as determined by ELISA (Fig. 2A). With the increased SDF-1 level in the lower HR-RTEC chamber, BMSCs in the upper chamber showed enhanced chemotaxis to the HR-RTECs than to the N-RTECs. Migration of BMSCs transfected with lenti-null showed no difference from BMSCs, however, CXCR4 overexpression significantly increased BMSC migration. Blocking CXCR4 with a neutralizing anti-CXCR4 antibody, AMD3100, abolished this effect (Fig. 2, B and C).
Differentiation of BMSCs in vitro.
To test whether CXCR4 overexpression has an effect on the renal differentiation of BMSCs, we examined the expression of CK18, a specific marker of RTEC, in BMSCs cultured with the HR-RTEC culture supernatant (1, 5). Orange-red fluorescence in the cytoplasm represented CK18+ cells (Fig. 3A). The immunofluorescent staining results showed that the proportion of CXCR4-BMSCs that expressed CK18 was not different from that of BMSCs, and neutralization with anti-CXCR4 antibody resulted in no additional change. In addition, the proportion of CK18+ cells was very low in each group, thereby suggesting that most BMSCs within the AKI microenvironment did not differentiate into RTEC-like cells (Fig. 3, A and B).
Overexpression of CXCR4 promotes BMSC paracrine actions in vitro.
We know paracrine mechanisms play an important role in the effect of BMSCs homing toward AKI repair. To test whether overexpression of CXCR4 in BMSCs enhances the cellular release of paracrine cytokines, we examined the expression of several cytokines by ELISA. Within the AKI microenvironment, CXCR4-BMSCs produced significantly higher levels of BMP-7, HGF, and IL-10 protein compared with BMSCs and null-BMSCs. In addition, blocking CXCR4 with an anti-CXCR4 antibody reduced the secretion of these cytokines (Fig. 3C).
CXCR4 overexpression improves the protective effect of BMSCs on HR-RTECs.
To compare with the control group, BMSC coculture increased the proportion of PCNA+ cells, a specific marker of mitogenesis, in HR-RTECs. No difference was found between the N/H and B/H groups. CXCR4-BMSC coculture further enhanced this proportion (Fig. 4A). Many annexin V+ RTECs were observed in the control group, and the proportion of annexin V+ RTECs decreased after BMSC coculture. CXCR4 overexpression in BMSCs further decreased the proportion of annexin V+ cells in HR-RTECs (Fig. 5B). Western blot also showed that CXCR4-BMSC coculture significantly decreased cleaved caspase-3 expression in HR-RTECs; however, increased Bcl-2 expression (Fig. 4, C and D). These results suggest that BMSC coculture is helpful for the survival of injured RTECs and that CXCR4 overexpression enhances this protective effect.
CXCR4 overexpression increases the in vivo homing of transplanted BMSCs to ischemic kidneys.
To validate the homing of BMSCs to the target tissue, the cells were labeled with BrdU and then systemically administered via tail vein into I/R-AKI mice. Seven days after transplantation, mice were killed, and the BrdU+ cells were quantitated. Some cells in the kidneys showed a brown nucleus in the BMSC group, indicating that the BrdU+ cells (BMSCs) were homing to the ischemic kidneys. The proportion of BrdU+ cells in the null-BMSC group was similar to that in the BMSC group. However, transplantation of CXCR4-BMSCs further increased the proportion of BrdU+ cells in ischemic kidneys (Fig. 5, A and B).
CXCR4 overexpression improves treatment with BMSCs in I/R-AKI mice.
The renal function of I/R-AKI mice was assessed by BUN and Scr levels. Administration of BMSCs and null-BMSCs both improved the renal function of mice at 7 days after transplantation. CXCR4-BMSC-treated mice had significantly lower BUN and Scr levels than mice in the BMSC group and the null-BMSC group (Fig. 6, A and B).
To further substantiate these results, histological examination of the tubular tissues, including ATN scoring, PCNA, and TUNEL staining were also evaluated at 7 days after transplantation. In the model group, the PAS staining (Fig. 6C) revealed that some tubules displayed brush border loss, naked basement membrane, vacuolar degeneration, and high ATN score, whereas PCNA expression of tubular cells was low (Fig. 6D), and at the same time, many TUNEL+ nuclei were observed by TUNEL assay (Fig. 6E). As expected, compared with the model group, both BMSC-treated mice and null-BMSC-treated mice had a significantly reduced ATN score with regenerated tubular cells observed in the PAS staining section (Fig. 6C), an increased proportion of PCNA+ tubular cells, and a decreased proportion of TUNEL+ cells. Interestingly, changes in these indicators in CXCR4-BMSCs-treated mice were more significant (Fig. 6, C–E).
CXCR4 overexpression does not prolong retention of BMSCs migrating to the kidney.
The experimental results described above suggest that CXCR4-BMSC transplantation is more effective in AKI repair than BMSC transplantation. We know that the poor survival of transplanted BMSCs in targeted damaged tissue also reduces the therapeutic efficacy (25, 29, 44). Therefore, we wanted to know whether overexpression of CXCR4 has an effect on the retention of transplanted BMSCs in the injured kidney. We examined BrdU+ cells at 14 days after transplantation in the BMSC group and the CXCR4-BMSC group. Only a few BrdU+ cells were observed in both groups. There was no difference in the proportion of BrdU+ cells between the two groups (Fig. 5C).
Numerous studies (17, 21, 26, 30, 36, 37) have reported that BMSC transplantation may be a feasible therapy for AKI repair. However, the repair efficiency is limited (14, 33). For cell transplantation therapy, it is important that cells migrate specifically to the damaged tissues; however, many transplanted BMSCs can still remain in other blood-rich organs, which should be one of the main reasons of their limited repair effect.
A number of studies (2, 9, 20, 38, 41) have shown that SDF-1 (also known as CXCL12) is critical for the process involving stem/progenitor cell chemotaxis and organ-specific homing to ischemic tissues through interaction with its receptor, CXCR4. SDF-1 is usually upregulated in injured tissues (2, 19, 31, 35). In our study, we constructed an HR model of RTECs to simulate AKI in vitro. The SDF-1 concentration was increased in the HR-RTEC culture supernatant. CXCR4 is highly expressed in BMSCs within the bone marrow; however, culture-expanded BMSCs progressively downregulate CXCR4 expression and lose their ability to migrate toward the SDF-1 gradient in ischemic tissue (3, 32, 40). It was observed in our study that null-BMSCs expressed only small amounts of CXCR4. Transplantation of BMSCs that overexpressed CXCR4 led to significant protection against ovariectomy-induced bone loss (4), augmented myoangiogenesis in infarcted myocardium (3, 16, 43), and induced early liver regeneration of small-for-size liver grafts (7). These observations suggest that overexpression of CXCR4 contributes to enhanced efficacy of BMSC therapy due to the increased homing ability of BMSCs.
On the basis of these findings, we hypothesize that expression of CXCR4 in BMSCs is necessary for their homing to injured kidneys. In our study, we accomplished remarkable CXCR4 expression in BMSCs by gene transfection. This was accompanied by enhanced migration of CXCR4-BMSCs to the HR-RTEC culture chamber in vitro, and the effect was entirely abolished by CXCR4 inhibition, which probably suggests a better AKI repair effect of CXCR4-BMSC transplantation in vivo.
The chemokine effect of the SDF-1/CXCR4 axis is regarded as an important mechanism for facilitating the therapeutic effects using BMSCs in several tissues, including the kidney. Here we show that overexpression of CXCR4 in BMSCs also results in a significant increased release of the cytokines BMP-7, HGF, and IL-10. This may contribute to the beneficial effect of homing of BMSCs.
The effect of homing BMSCs toward AKI repair is partly due to their direct differentiation process (21, 26, 28). Studies also suggest involvement of paracrine mechanism (8, 13, 17, 30, 34, 36). Homing BMSCs can secrete some cytokines that exert antiapoptotic, anti-inflammatory, and kidney regeneration-promoting effects. Therefore, in our in vitro study, we also tested whether overexpression of CXCR4 has a direct effect on differentiation or paracrine actions of BMSCs. BMSCs cultured by the HR-RTEC culture supernatant were used to imitate the homing of BMSCs in vivo. No difference was found in the proportion of CK18+ cells between CXCR4-BMSCs and BMSCs. In addition, the proportion of CK18+ cells in each group was very low. However, overexpression of CXCR4 in BMSCs resulted in significant increases in the release of BMP-7, HGF, and IL-10, all of which are beneficial in the repair of injured kidneys (6, 15, 42). Additionally, blocking CXCR4 with an anti-CXCR4 antibody abolished the secretion of these cytokines. All of these results demonstrated that the complete cell differentiation of BMSCs is a rare event in AKI repair, and that the paracrine effect of cytokines secreted by the homing of BMSCs may be the major pathway of AKI repair. Moreover, the SDF-1/CXCR4 axis is also required for the BMSC paracrine actions within the AKI microenvironment, and overexpression of CXCR4 is beneficial for the paracrine actions of homing BMSCs.
As a result of more migration and greater paracrine actions of CXCR4-BMSCs, improved survival of damaged RTECs was observed in our study, including accelerated mitogenic response, and reduced annexin V+ cells. Furthermore, cleaved caspase-3 expression was decreased and Bcl-2 expression was increased in HR-RTECs. In agreement with observations from the in vitro study, the in vivo study also showed that significantly greater numbers of intravenously infused CXCR4-BMSCs migrated to ischemic kidneys, together with significantly improved renal function, reduced ATN scoring, more PCNA+ tubular cells, and fewer TUNEL+ tubular cells. However, CXCR4 overexpression could not prolong the retention of homing BMSCs toward ischemic kidneys.
In conclusion, we found that overexpression of CXCR4 in BMSCs was effective for their migration to the injured kidney and for their paracrine actions. This finding may open a new approach in cell-based therapy of AKI.
This work was supported by Shanghai Rising-Star Program Grant 12QA1405000, Shanghai Municipal Natural Science Foundation Grant 11ZR1449600, Shanghai Key Projects of Basic Research Grant 12DJ1400203, Deutsche Forschungsgemeinschaft PA Grant 479/10-1, and Else Kröner-Fresenius-Stiftung Grant P40/09/A29/09.
No conflicts of interest, financial or otherwise, are declared by the author(s).
Author contributions: N.L. and J.Z. conception and design of research; N.L. performed experiments; N.L. and A.P. analyzed data; N.L., A.P., and J.Z. interpreted results of experiments; N.L. prepared figures; N.L., A.P., and J.Z. drafted manuscript; N.L., A.P., and J.Z. edited and revised manuscript; N.L., A.P., and J.Z. approved final version of manuscript.
We thank Dr. A. Bondke Persson for helpful discussions.
- Copyright © 2013 the American Physiological Society