UNC5B Receptor Deletion Exacerbates Tissue Injury in Response to AKI

Abstract

Netrin-1 regulates cell survival and apoptosis by activation of its receptors, including UNC5B. However, the in vivo role of UNC5B in cell survival during cellular stress and tissue injury is unknown. We investigated the role of UNC5B in cell survival in response to stress using mice heterozygously expressing the UNC5B gene (UNC5B^−/flox) and mice with targeted homozygous deletion of UNC5B in kidney epithelial cells (UNC5B^{−/flox/GGT-cre}). Mice were subjected to two different models of organ injury: ischemia reperfusion injury of the kidney and cisplatin-induced nephrotoxicity. Both mouse models of UNC5B depletion had normal organ function and histology under basal conditions. After AKI, however, UNC5B^{−/flox/GGT-cre} mice exhibited significantly worse renal function and damage, increased tubular apoptosis, enhanced p53 activation, and exacerbated inflammation compared with UNC5B^−/flox and wild-type mice. shRNA-mediated suppression of UNC5B expression in cultured tubular epithelial cells exacerbated cisplatin-induced cell death in a p53-dependent manner and blunted Akt phosphorylation. Inhibition of PI3 kinase similarly exacerbated cisplatin-induced apoptosis; in contrast, overexpression of UNC5B reduced cisplatin-induced apoptosis in these cells. Taken together, these results show that the netrin-1 receptor UNC5B plays a critical role in cell survival and kidney injury through Akt-mediated inactivation of p53 in response to stress.

AKI is a life-threatening disorder that has been increasing in incidence. About 6% of hospitalized patients and more than 30% of patients in the intensive care unit develop AKI, which can lead to organ failure and death.^1–3 Currently, there are effective therapies available to treat AKI. Several in vivo studies suggest that administration of netrin-1 suppresses inflammation and apoptosis in the kidney and other organs caused by different insults.^4–10 Moreover, netrin-1 is highly induced in tubular epithelial cells and secreted into the tubular lumen within hours after ischemia reperfusion.¹¹ Overexpression of netrin-1, specifically in proximal tubular epithelial cells, suppressed ischemia reperfusion and nephrotoxin-induced kidney injury and dextran sulfate sodium-induced colon injury by suppressing apoptosis.^12,13 However, the receptor that mediates the protective effects of netrin-1 and the mechanism of protection against cell death are unknown.

Netrin-1 is a laminin-related secreted molecule identified as a neuronal guidance cue that directs neurons and their axons to targets during development of the nervous system.¹⁴ Although netrin-1 is primarily thought of as an axon guidance cue, this function is unlikely to be its only function, because expression studies have shown that netrins are widely expressed outside the nervous system, including in vascular endothelial cells as well as the liver, lung, colon, heart, and kidney.^4,15 In fact, the kidney was shown to have the highest level of netrin-1 mRNA compared with any other tissues studied so far.¹⁵ Netrin-1 mediates its biologic activity by binding to two vertebrate families of receptors: the Deleted in Colorectal Cancer (DCC) family comprising DCC and neogenin receptors and the uncoordinated 5 (UNC5) family UNC5A-D receptors.¹⁶ Both DCC and UNC5 family members are called dependence receptors because of the dependence of cell survival on the presence of the netrin-1 ligand, and the absence of these receptors is known to induce apoptosis.^17–19 However, the in vivo relevance of this dependence receptor hypothesis is not clear. For example, very little netrin-1 protein is expressed in normal kidney tubular epithelial cells, but UNC5B expression was found in the same cells. If dependence receptor hypothesis is functioning, then we would expect to see a large number of apoptotic cells in the normal kidney, which was not the case. Therefore, we hypothesized that UNC5B receptor signaling is critical for cell survival in response to injurious stimulus. In the absence of the UNC5B receptor, cells and tissues are highly prone to injury. Previous studies from our laboratory showed that the UNC5B receptor is highly expressed in proximal tubular epithelial cells of the kidney in vivo and in vitro, whereas DCC expression is very low or negligible. Moreover, UNC5C expression is restricted to distal tubules where UNC5A and UNC5D are not expressed in the kidney.²⁰ In addition, netrin-1–induced epithelial cell proliferation and migration in vitro were mediated through UNC5B, and in vivo antibody-based neutralization of UNC5B function abolished netrin-1 protective effects against ischemia reperfusion injury of the kidney.^10,21 These data led us to investigate the critical role of UNC5B in renal epithelial survival in response to AKI.

In this study, we investigated our hypothesis using tissue-specific UNC5B knockout animals as well as animal with partial deficiency of UNC5B receptors. We have generated kidney proximal tubular epithelial (TKPTS) cell-specific UNC5B knockout mice using Cre-lox technology. These mice were subjected to ischemia reperfusion injury and cisplatin-induced kidney injury. Here, we report that deletion of the netrin-1 receptor UNC5B rendered the mice highly susceptible to ischemic and toxic kidney injury. The exacerbated kidney injury is caused by increased apoptosis, p53 activation, and inflammatory response in the epithelial cells of the kidney compared with wild-type (WT) or heterozygous knockout mice. Moreover, this activation of p53 was caused by defective activation of protein kinase B (Akt) pathways in UNC5B knockout epithelial cells.

Results

Proximal Tubular Epithelial Cell-Specific Deletion of UNC5B in Mice Exacerbates Ischemia Reperfusion Injury of the Kidney

To determine the role of tubular epithelial UNC5B receptor in ischemic injury of the kidney, mice with UNC5B^{−/flox/GGT-cre} and littermates without cre (UNC5B^−/flox) or WT mice were subjected to 26 minutes of ischemia followed by reperfusion. Both UNC5B^{−/flox/GGT-cre} mice and littermates without cre died within 24 hours after ischemia, but the WT mice survived, suggesting that deletion of a single allele of UNC5B alone increases susceptibility to ischemia reperfusion of the kidney. To determine the contribution of UNC5B in the proximal tubular epithelial cells, UNC5B^{−/flox/GGT-cre} and UNC5B^−/flox mice were subjected to 22 minutes of ischemia followed by 48 hours of reperfusion. Kidney function was assessed by measuring the levels of serum creatinine at different times after reperfusion. WT mice did not show any increase in serum creatinine levels with a mild form of ischemia, whereas mice with proximal tubular epithelial cell-specific UNC5B deletion showed a large increase in serum creatinine levels and BUN (Figures 1 and 2, A and B). Most of the animals died by 72 hours, which suggests UNC5B has a critical protective role against ischemic tubular injury. Consistent with kidney dysfunction, histology showed more necrotic tubules, cast formation, and dilated tubules in UNC5B^{−/flox/GGT-cre} mice compared with UNC5B^−/flox mice kidneys (Figure 2C). Netrin-1 is known to regulate immune cell infiltration^10,15 in the kidney. To determine whether UNC5B deletion in proximal tubular epithelial cells increased infiltration of neutrophils, immunolocalization was performed. As shown in Figure 2, C and D, UNC5B^{−/flox/GGT-cre} mouse kidneys showed a large number of neutrophils compared with UNC5B^−/flox mice kidneys.

Figure 1.

Characterization of mice with proximal tubular epithelial cell specific deletion of UNC5B receptor. Characterization of TKPTS-specific deletion of UNC5B. (A) Genomic PCR showing successful generation of three different genotypes: UNC5B^flox/flox, UNC5B^−/flox, and UNC5B^{−/flox/GGT-cre}. The band marked deleted represents whole-body deletion of a single allele for UNC5B. (B and C) Immunohistochemical localization of the UNC5B receptor in UNC5B^−/flox and UNC5B^{−/flox/GGT-cre} mice. Heterzygous knockout mice without cre showed UNC5B staining in the proximal tubular epithelial cells, whereas GGT-cre–positive mice showed no staining for UNC5B, suggesting successful deletion of the floxed UNC5B gene. Scale bar, 100 µM.

Figure 2.

Kidney injury and histopathological changes in kidney-specific UNC5B knockout mice. (A and B) UNC5B^−/flox and UNC5B^{−/flox/GGT-cre} mice were subjected to 22 minutes of ischemia followed by 48 hours of reperfusion. Kidney function was measured by quantifying serum creatinine. Mice with kidney-specific UNC5B deletion developed severe renal injury, which was seen by increased (A) serum creatinine and (B) BUN, whereas heterozygous knockout mice did not develop any injury. *P<0.001 versus UNC5B^−/flox. (C) Photomicrograph of periodic acid–Schiff (PAS)-stained and neutrophil-stained kidney tissue sections 6 hours after ischemia reperfusion. UNC5B^{−/flox/GGT-cre} mice kidney shows cast formation and brush border damage, whereas UNC5B^−/flox mice did not develop any injury. In addition, UNC5B^{−/flox/GGT-cre} mouse kidney showed a large number of infiltrated neutrophils compared with UNC5B^−/flox mouse kidney. (D) Quantification of tubular damage and neutrophil infiltration in UNC5B^−/flox and UNC5B^{−/flox/GGT-cre} mice subjected to ischemia reperfusion. n=8. Scale bar, 100 µM.

UNC5B Deletion in Proximal Tubular Epithelial Cells Increases Inflammatory Response in the Kidney

Because UNC5B is known to regulate inflammatory and apoptotic responses,¹⁰ we profiled inflammatory mediator genes and genes that regulate apoptosis expression in the kidney at 6 hours after ischemia reperfusion. Consistent with increased neutrophil infiltration in the UNC5B^{−/flox/GGT-cre} mouse kidney (Figure 2, C and D), expression of several cytokines, chemokines, and their receptors and genes known to regulate apoptosis was significantly upregulated after ischemia in heterozygous UNC5B knockout mice without cre, and it was also dramatically increased in UNC5B^{−/flox/GGT-cre} mice (Figure 3). This result suggested that deletion of UNC5B exacerbates the inflammatory response of the kidney.

Figure 3.

Ischemia reperfusion–induced expressions of proinflammatory and proapoptotic genes were enhanced in UNC5B knockout mouse kidney as compared to WT kidney. Regulation of genes that regulate inflammatory and apoptotic responses in UNC5B^−/flox and UNC5B^{−/flox/GGT-cre} mouse kidney. Gene expression was analyzed by real-time PCR. (A) Kidney-specific deletion of UNC5B induced a large number of cytokines, chemokines, and their receptor expression compared with sham-operated control and heterozygous knockout mice subjected to ischemia reperfusion (IR). *P<0.05 versus sham operated UNC5B^−/flox; ⁺P<0.001 versus all other groups. n=8. (B) Kidney-specific deletion of UNC5B induced many genes that regulated apoptosis compared with sham-operated control and heterozygous knockout mice subjected to ischemia reperfusion (I/R). n=5.

UNC5B Deletion in Proximal Tubular Epithelial Cells Increases Tubular Epithelial Cell Apoptosis

Netrin-1 was shown to be a survival factor for intestinal epithelial cells, and its administration or overexpression suppressed ischemia-induced tubular epithelial cell apoptosis. Netrin-1 is also highly induced after ischemia and may act to limit the apoptosis. To determine whether UNC5B deletion in proximal tubular epithelial cells alters ischemia reperfusion-induced apoptosis, apoptotic cells were quantified using the terminal deoxynucleotidyl transferase-mediated digoxigenin-deoxyuridine nick-end labeling (TUNEL) assay. Sham-operated UNC5B^−/flox (Figure 4A) and UNC5B^{−/flox/GGT-cre} (Figure 4C) mouse kidneys did not show any apoptotic cells. Ischemia reperfusion induced a small increase in apoptosis in UNC5B^−/flox mice (Figure 4B). However, ischemia reperfusion-induced apoptosis was drastically increased in UNC5B^{−/flox/GGT-cre} mice (Figure 4D). Quantification of apoptotic cells showed a significant increase in UNC5B^{−/flox/GGT-cre} mice compared with UNC5B^−/flox mice (Figure 4E). Increased apoptosis was associated with increased p53 activation and expression in UNC5B^−/flox and UNC5B^{−/flox/GGT-cre} mice compared with WT mice. To determine whether the increased netrin-1 expression seen in proximal tubular epithelial cells after reperfusion¹⁵ is associated with UNC5B receptor expression, colocalization of netrin-1 and UNC5B was carried out in kidneys from WT mice that were subjected to 26 minutes of ischemia followed by 24 hours of reperfusion. As shown in Supplemental Figure 1, both netrin-1 and UNC5B expressions are colocalized in the cortical tubular epithelial cells, suggesting that netrin-1 mediates its activity through UNC5B in proximal tubular epithelial cells.

Figure 4.

UNC5B deletion exacerbates tubular epithelial cells apoptosis in response to ischemia reperfusion. Effect of UNC5B deletion in proximal tubular epithelial cells in ischemia reperfusion (IR)-induced apoptosis. (A–D) Apoptotic cells were identified using TUNEL staining. Sham-operated (A) UNC5B^−/flox and (C) UNC5B^{−/flox/GGT-cre} mouse kidneys did not show any apoptotic cells. Six hours after reperfusion, the number of apoptotic cells (blue staining) was increased dramatically in (B) UNC5B^{−/flox/GGT-cre} mice compared with (D) UNC5B^−/flox mice. Scale bar, 100 µM. (E) Quantification of apoptotic cells. Apoptotic-positive cells were counted in five ×40 magnification fields. *P<0.05 versus sham-operated; ^#P<0.001 versus UNC5B^−/flox subjected to IR. n=4–6. (F) Western blot analysis of p53 expression and phosphorylation in kidney from sham-operated mice and mice that are subjected to IR. UNC5B deletion increased p53 expression and phosphorylation (P-p53) in response to IR compared with sham operation or flox/flox (WT) mice subjected to IR.

Partial or Proximal Tubular Epithelial Cell-Specific Deletion of UNC5B in Mice Exacerbates Cisplatin-Induced AKI

To determine whether UNC5B has the protective role in other models of tissue injury, we treated WT, heterozygous knockout, and proximal tubular epithelial cell-specific knockout mice with cisplatin. Kidney function was assessed by measuring serum creatinine levels at different times after cisplatin administration. WT mice developed AKI as expected by 72 hours after cisplatin administration. However, UNC5B^−/flox developed much more severe renal injury, which was further exacerbated in proximal tubular epithelial cell-specific UNC5B knockout mice (UNC5B^{−/flox/GGT-cre}) (Figure 5). Consistent with kidney dysfunction, histology showed more necrotic tubules, cast formation, and dilated tubules in UNC5B^−/flox and UNC5B^{−/flox/GGT-cre} mice kidneys compared with UNC5B^flox/flox mice kidneys (Figure 5, B–F).

Figure 5.

UNC5B deletion exacerbates cisplatin-induced AKI and inflammatory response of the kidney. (A) Cisplatin was administered intraperitoneally at a dose of 25 mg/kg body wt. Kidney function was assessed by measuring serum creatinine. Cisplatin-induced kidney dysfunction in WT (UNC5B^flox/flox) was exacerbated with partial deletion of UNC5B. The cisplatin-induced kidney dysfunction was worse in UNC5B whole-body partial deletion with proximal tubular epithelial-specific deletion. *P<0.05 versus corresponding time in other groups; **P<0.05 versus corresponding time point in WT and UNC5B^{−/flox/GGT-cre} mice; ***P<0.001 versus saline treatment (0 hours). n=6. (B) Cisplatin administration induced the expression inflammatory cytokines and chemokines in WT, which was increased further in mice with both partial (heterozygous) and kidney-specific deletion of UNC5B. *P<0.05 versus saline treatment; ^#P<0.05 versus all other groups. n=4. ICAM-1, intercellular adhesion molecule 1; KC; MCP-1; TLR4. (C–F) Periodic acid–Schiff-stained section showing tubular injury, including dilated tubules, cast formation, and necrotic tubules. (C) Saline-treated UNC5B^flox/flox mice kidneys. (D) Cisplatin-treated UNC5B^flox/flox mice kidneys. (E) Cisplatin-treated UNC5B^−/flox mice kidneys. (F) Cisplatin-treated UNC5B^{−/flox/GGT-cre} mice kidneys. (G) Quantification of tubular injury score. *P<0.001 versus saline treatment; ^#P<0.05 versus all other groups. n=6.

Consistent with histologic damage, both UNC5B^−/flox and UNC5B^{−/flox/GGT-cre} mouse kidneys treated with cisplatin showed increased expression of inflammatory cytokines and chemokines (Figure 5G). Because netrin-1 binding to UNC5B receptor is known to suppress apoptosis, the extent of apoptosis in the absence of UNC5B was quantified by TUNEL assay and proapoptotic gene expression analysis using PCR array. As shown in Figure 6, either partial deletion (Figure 6C) or proximal tubule–specific deletion of the UNC5B receptor (Figure 6D) exacerbated apoptosis and proapoptotic gene expression (Figure 6F) compared with littermate controls (Figure 6B). No apoptosis was seen in untreated control mice (Figure 6A). Quantitative data of apoptosis are shown in Figure 6E. Our previous studies showed that netrin-1 overexpression suppressed p53 activation in response to cisplatin. Moreover, UNC5B signaling is known to suppress p53 expression and activation. Therefore, we determined p53 activation in UNC5B knockout mice. As shown in Figure 6, G–J, cisplatin increased phopho-p53 levels in WT kidneys (Figure 6H), which was further increased in partial (Figure 6I) and proximal tubule-specific (Figure 6J) UNC5B knockout mice. This finding suggests that UNC5B may regulate cisplatin-induced apoptosis through suppression of p53 activation; p53-positive nuclei were counted, and quantitative data are shown in Figure 6K.

Figure 6.

UNC5B deletion in proximal tubular epithelial cells increased epithelial cells apoptosis and p53 activation in response to cisplatin administration. (A–D) Apoptotic cells were identified using TUNEL staining. Saline-treated (A) UNC5B^flox/flox (UNC5B^−/flox and UNC5B^{−/flox/GGT-cre} are not shown) mouse kidney did not show any apoptotic cells; 72 hours after cisplatin administration, the number of apoptotic cells (blue staining) was increased dramatically in (D) UNC5B^{−/flox/GGT-cre} mice compared with (C) UNC5B^−/flox or (B) UNC5B^flox/flox mice. Scale bar, 100 µM. (E) Quantification of apoptotic cells. Apoptotic-positive cells were counted in five ×40 magnification fields. *P<0.05 versus saline treatment; ^#P<0.001 versus all other groups. n=4–6. (F) Regulation of genes that regulate inflammatory and apoptotic responses in UNC5B^−/flox and UNC5B^{−/flox/GGT-cre} mouse kidney. Gene expression was analyzed by real-time PCR. Kidney-specific deletion of UNC5B induced a large number of proapoptotic genes and their receptor expression compared with cisplatin-treated UNC5B^flox/flox and heterozygous knockout mice. *P<0.05 versus cisplatin-treated UNC5B^flox/flox; **P<0.001 versus all other groups. n=4. (G–J) Immunohistochemical localization of phospho-p53 in (G) saline-treated UNC5B^flox/flox, (H) cisplatin-treated UNC5B^flox/flox, (I) cisplatin-treated UNC5B^−/flox, and (J) cisplatin-treated UNC5B^{−/flox/GGT-cre} kidney tissue sections. (K) Quantification of p53-positive cells in five ×40 magnification fields. *P<0.05 versus saline treatment; ^#P<0.001 versus all other groups. n=4–6. Scale bar, 100 µM.

Short Hairpin RNA-Mediated Knockdown of UNC5B Receptor Expression in Kidney Epithelial Cells Increased Sensitivity to Cisplatin-Induced Cell Death

To determine whether UNC5B knockdown in epithelial cells increases sensitivity to cell stress (such as cisplatin), WT and UNC5B knockout cells were treated with cisplatin for 24 hours. Knockdown of UNC5B was confirmed with quantitative RT-PCR (Figure 7A) and Western blot analyses (Figure 7B). The expression of UNC5B is reduced over 80% compared with WT cells (Figure 7, A and B). Western blot analysis of p53 showed increased accumulation of both total and phospho-p53 in UNC5B knockout cells compared with WT cells, suggesting enhanced activation of p53 in the absence of the UNC5B receptor (Figure 7C). To determine whether enhanced activation of p53 may contribute to increased cell death in UNC5B knockout cells, cell death was quantified as described in Concise Methods. As shown in Figure 7D, cisplatin treatment induced cell death in WT mice, which was significantly higher in UNC5B knockout cells. The addition of p53 inhibitors suppressed cisplatin-induced exacerbation of cell death in UNC5B knockout cells, suggesting that p53 activation may play a critical role in cell death. Consistent with exacerbated apoptosis in the absence of UNC5B in kidney epithelial cells, the expression of several apoptotic genes was also significantly increased in UNC5B knockout cells compared with WT cells (Supplemental Table 1). Consistent with exacerbation of cisplatin in UNC5B knockout renal epithelial cells, hydrogen peroxide-induced apoptosis and necrosis were also significantly higher in UNC5B knockout cells compared with WT cells (Supplemental Figure 2).

Figure 7.

Deletion of UNC5B in TKPTS exacerbates cisplatin-induced apoptosis through increased p53 activation. Cisplatin-induced cell death in WT and UNC5B knockout mouse proximal tubular epithelial cells was analyzed by flow cytometry after staining with annexin V–FITC and propidium iodide. (A) UNC5B knockout was confirmed by real-time RT-PCR. Over 80% knockdown was observed with short hairpin RNA (shRNA) to UNC5B in transfected cells compared with WT cells. *P<0.0001 versus WT. n=4. (B) Western blot analysis of UNC5B receptor expression in WT and UNC5B shRNA-transfected tubular epithelial cells. UNC5B expression reduced to a large extent with UNC5B shRNA transfection. (C) Western blot analysis of p53 proteins in WT and UNC5B knockout (KO) cells. Cisplatin induced accumulation of both total and phospho-p53 (P-p53) in WT cells, which was highly enhanced in UNC5B KO cells. The basal level of p53 was also higher in UNC5B KO cells compared with WT cells. Bands are from the same gel but rearranged for better presentation. (D) A 10 µM cisplatin addition induced a significant increase cell death, which was exacerbated in UNC5B KO cells. The addition of p53 inhibitors suppressed cisplatin-induced cell death exacerbation in UNC5B KO cells. *P<0.001 versus all other groups; ^#P<0.05 versus WT cisplatin treatment. n=4. PI, propidium iodide.

Phosphatidylinositol 3 Kinase-Akt Pathway Is a Critical Mediator of UNC5B-Regulated Cell Survival in Response to Stresses.

Because our previous study showed that netrin-1 activates Akt pathways through UNC5B²¹ and Akt is known to suppress p53 activation, we hypothesized that UNC5B regulates cell survival through phosphatidylinositol 3 kinase (PI3K) -mediated activation of Akt in response to stresses. We determined the activation of Akt and extracellular signal-regulated kinase (ERK) pathways by Western blot analysis in WT and UNC5B knockout cells treated with either cisplatin or netrin-1. The addition of cisplatin or netrin-1 induced a significant increase in Akt phosphorylation in WT epithelial cells compared with vehicle-treated controls. However, both netrin-1– and cisplatin-induced increases in Akt phosphorylation were completely blunted in UNC5B knockout cells (Figure 8, A and B) compared with WT epithelial cells. Moreover, the basal level of Akt phosphorylation was also blunted in the absence of UNC5B, suggesting the critical role of UNC5B in the activation of the PI3K-Akt pathway. Interestingly, a significant increase in ERK phosphorylation was seen in WT epithelial cells at 3 hours but not 1 hour after the treatment of cisplatin and netrin-1. The increase in ERK phosphorylation with cisplatin and netrin-1 was not blunted in UNC5B knockout cells, suggesting that stress-induced activation of PI3K-Akt is specific for UNC5B (Figure 8, A and C).

Figure 8.

Inhibition of PI3K exacerbates cisplatin-induced cell death in mouse proximal tubular epithelial cells. (A) Cisplatin- and netrin-1–induced Akt activation (phosphorylation [P-Akt]) was blunted, but ERK phosphorylation (P-ERK) in UNC5B knockout (KO) cells was analyzed by Western blot analysis. (B and C) Densitometric quantification of Western blot analysis of (B) phospho-Akt (1 and 3 hours) and (C) ERK (3 hours). *P<0.05 versus vehicle-treated and UNC5B KO cells. n=4. (D) Cell death was analyzed by flow cytometry after staining with annexin V-FITC and propidium iodide (PI). A 10 µM cisplatin addition induced a significant increase in cell death (red box), which was exacerbated with specific inhibition of PI3K with 50 µM LY294002. LY294002 alone did not increase cell death significantly without cisplatin. *P<0.001 versus all other groups; ^#P<0.05 versus WT cisplatin treatment. n=4.

To determine directly whether blunted activation of the Akt pathway plays a critical role in the increased apoptosis seen in UNC5B knockout cells, PI3K activation in WT cells was inhibited with the specific pharmacological inhibitor LY294002 as described in Concise Methods. As shown in Figure 8B, cisplatin addition significantly increased cell death over controls, which was further increased in PI3K inhibitor-treated cells, suggesting that PI3K-mediated Akt activation plays a critical role in cell survival in response to stresses and that UNC5B is a major activator of Akt pathways under stresses.

Overexpression of UNC5B in TKPTS Cells Reduced Cisplatin-Induced Apoptosis

To determine if forced overexpression of UNC5B in kidney epithelial cells reduced apoptosis in response to stress, TKPTS cells were transfected with the rat UNC5B expression plasmid as described in Concise Methods. Expression was confirmed by quantitative RT-PCR and immunohistochemistry. Cisplatin induced a significant increase in apoptosis in mock-transfected cells, which was significantly reduced in UNC5B-transfected cells (Supplemental Figure 3). Consistent with reduced apoptosis in UNC5B-transfected cells, caspase-3 activity was also reduced compared with mock-transfected cells.

Discussion

The pathophysiological role of UNC5B receptors in kidney disease is unknown. In this work, we describe for the first time the critical role of the UNC5B receptor in renal tubular epithelial cell survival using a conditional knockout mouse model with targeted UNC5B deletion from renal proximal tubules. The conditional knockout mice did not show overt defects in kidney development, histology, or function. Interestingly, these animals are highly sensitive to ischemic AKI. We show for the first time the role of tubular epithelial cell UNC5B in ischemic AKI. Consistent with our earlier studies,²⁰ UNC5B expression was localized only in the proximal tubular epithelial cells in heterozygous knockout animal kidneys. In the conditional knockout mice, UNC5B deletion was complete and specific to the proximal tubular epithelial cells in the renal cortex (Figure 1). Whole-body knockout of UNC5B or endothelial cell-specific conditional knockout of UNC5B in mice is lethal to embryos, and mice die in utero because of defective angiogenesis.²² γ-Glutamyl transferase (GGT)-cre expression is known to occur in late kidney development (beginning around postpartum day 14) when nephrogenesis is virtually complete.²³ Moreover, bone marrow cells and other organs did not express GGT-cre.²³ However, it is possible that GGT-cre expression may start during early stages of development. Our studies show that the proximal tubular epithelial cell-specific knockout is not lethal and that the kidney shows normal histology. Therefore, UNC5B is not required for tubular epithelial cell development. During the observations for the past 3 years and the period of 6–8 months of age, these mice do not develop any abnormal histopathology in the kidney. The absence of developmental defects in these mice enables us to study the role of UNC5B during adult life and in response to ischemia reperfusion. Mice with cre expression in the tubular epithelial cells do not show any defect in kidney morphology and function (data not shown). Moreover, they respond to ischemia or cisplatin just like WT mice (data not shown).

In this study, we showed a remarkable sensitivity of UNC5B^{−/flox/GGT-cre} mice to ischemic AKI compared with UNC5B^−/flox mice. Our initial results showed that deletion of one allele (UNC5B^+/−) made mice more sensitive to ischemia reperfusion injury compared with WT littermates. All mice with whole-body deletion of one allele and deletion of both alleles in proximal tubules died with longer ischemic time. With shorter ischemic time, both WT and heterozygous UNC5B knockout mice did not develop kidney injury, but heterozygous mice showed an increased number of apoptotic cells in the cortex. In contrast, deletion of both UNC5B alleles in proximal tubular epithelial cells exacerbated tissue damage and increased the number of apoptotic cells, resulting in significantly worse renal function than their heterozygous littermates. Our PCR array analysis further revealed exacerbated expression of inflammatory mediators. We show for the first time that tubular epithelial cell-specific deletion of specific receptors enhanced the inflammatory response. Taken together, these observations support a critical role for UNC5B in protecting tubular epithelial cells against ischemic AKI.

Consistent with ischemia reperfusion injury, we see similar effects in cisplatin- induced acute injury with UNC5B deletion. Cisplatin-induced tissue injury and inflammatory response exacerbated with either partial or proximal tubular epithelial cell-specific deletion of the UNC5B receptor. Moreover, there is increased cell death and accumulation of phospho-p53 in mice with partial and kidney-specific deletion of UNC5B receptors. UNC5B is a target for p53.²⁴ UNC5B signaling was shown to actively suppress p53 expression and activation through an Akt-dependent manner.^24,25 Moreover, p53 can be stably expressed in cancer cells but does not induce apoptosis in the presence of netrin-1. It was shown that netrin-1 binding to UNC5B activates Akt pathways to inhibit p53 activation, which accumulates in an inactive form.^24,26,27 It was proposed that netrin-1 inhibits post-translational modification of p53, thereby suppressing p53-dependent apoptosis. Consistent with this view, our previous data suggest that cisplatin-induced p53 phosphorylation is inhibited in netrin-1 transgenic mice⁹ and netrin-1–activated Akt in a time-dependent manner in renal epithelial cells in vitro, which was mediated through UNC5B receptor.²¹ Therefore, it is possible that the absence of UNC5B inhibitory signaling on the p53 may lead to excess activation by phosphorylation and increase of p53-dependent apoptosis. This idea was further supported by our in vitro studies, where UNC5B knockdown increased the accumulation of both total and phospho-p53 in response to cisplatin and inhibition of p53 suppressed cisplatin-induced exacerbation of apoptosis in kidney epithelial cells. Moreover, Akt activation by cisplatin or netrin-1 was completely blunted in UNC5B knockout cells, suggesting the importance of UNC5B in activating cell survival pathway. The critical role of UNC5B-activated Akt in cell survival was further supported by our in vitro studies (Figure 8), where addition of a specific PI3K inhibitor exacerbated cisplatin-induced cell death.

Necrosis is a common pathologic observation in many forms of AKI.^28–31 Although inhibition of p53 activation suppresses apoptosis, it may have no effect on primary necrosis. It should, however, reduce secondary necrosis, which is known to occur after apoptosis. Our data show that both apoptosis and necrosis increased in the UNC5B knockout kidney. Although the extent of primary necrosis is not clear in our studies, our in vitro studies with hydrogen peroxide (Supplemental Figure 2) show that UNC5B deletion could increase both apoptosis and necrosis. These data suggest that UNC5B may regulate both apoptosis and necrosis in renal epithelial cells. However, the underlying mechanisms remain to be determined.

In conclusion, organ injury caused by toxic chemical and ischemia reperfusion is widespread in human disease, which includes AKI caused by vascular surgery, cardiac bypass surgery, sickle cell disease, anticancer drug-induced kidney injury, colon injury in Crohn’s disease, and inflammatory bowel disease.^{9,12,13,32–34} Cell death by apoptosis is a major contributor of organ dysfunction in these diseases. Currently, there are no effective treatments available to treat ischemia reperfusion injury or chemical-induced organ injury. Our studies shed light on the protective mechanisms and possible new ways to treat AKI based on increasing UNC5B receptor-mediated survival signaling.

Concise Methods

Mouse Strains

Generation of the UNC5B alleles has been described previously.^22,35 GGT-cre mice were generously provided by Eric Nelson at Vanderbilt University School of Medicine. Genotypes were determined by PCR analysis of genomic DNA isolated from tail clips; amplification primers and conditions used for UNC5B mice were 5′-TAGCCTCAGGGTCTACTGTCTG, 5′-CTCTCAGACTTCTCAAAGAGATTC, and 5′-CCACTGTATGCCAGACGACATG under the conditions of 94°C for 30 seconds, 62°C for 20 seconds, and 72°C for 40 seconds. GGT-cre mice were genotyped using the primers 5′-AGGTGTAGAGAAGGCACTTAGC and 5′- CTAATCGCCATCTTCCAGCAGG-3′ under the conditions of 94°C for 30 seconds, 58°C for 20 seconds, and 72°C for 40 seconds. The Institutional Animal Care and Use Committee of the Georgia Health Sciences University approved all of the protocols and procedures for using animals (approval number BR10–10–369).

Proximal Tubular Epithelial Cell-Specific Deletion of UNC5B Shows Normal Phenotype in Mice

Creation and characterization of the UNC5B^flox allele were described previously.^22,35 To systematically inactivate UNC5B in a tissue-specific fashion, animals homozygous for the conditional allele (UNC5B^flox/flox) were mated with animals heterozygous for the UNC5B null allele (UNC5B^+/−). The UNC5B ^flox/− progeny were mated with mice carrying the Cre gene under the control of GGT promoter, with activity that was restricted to proximal tubular epithelial cells. The tubular-restricted GGT promoter was shown to be active only in adulthood but not during embryo development.²³ Deletion was confirmed by PCR and immunohistochemical staining of UNC5B in kidney sections (Figure 1). Deletion of UNC5B in proximal tubular epithelial cells resulted in mice that survived to adulthood and seemed to be normal.

Renal Ischemia Reperfusion

Eight- to nine-week-old UNC5B^{−/flox/GGT-cre}, UNC5B^−/flox, and UNC5B^flox/flox mice were anesthetized with sodium pentobarbital (50 mg/kg body wt intraperitoneally) and placed on a heating pad to maintain body temperature at 37°C. Both renal pedicles were identified through dorsal incisions and clamped for a period of 26 or 22 minutes. Reperfusion was confirmed visually on release of the clamps. As a control, sham-operated animals were subjected to the same surgical procedure, except that the renal pedicles were not clamped. Surgical wounds were closed, and mice were given 1 ml warm saline intraperitoneally. The mice were kept in a warm incubator until they regained consciousness.

Cisplatin-Induced AKI

Cisplatin was dissolved in saline at a concentration of 1 mg/ml. Mice were given a single intraperitoneal injection of either saline or cisplatin (25 mg/kg body wt). Animals were euthanized 72 hours after cisplatin injection, and blood and kidney tissues were collected. Kidney tissues were processed for histology, TUNEL assay, neutrophil and p53 staining, and RNA isolation.

Renal Function

Renal function was assessed by measurements of serum creatinine (DZ072B; Diazyme Laboratories, Poway, CA) and BUN (BioAssay Systems, Hayward, CA).^10,15

Histology and Immunostaining

Kidney tissue was fixed in buffered 10% formalin for 12 hours and then embedded in paraffin wax. For assessment of injury, 5-µm sections were stained with periodic acid–Schiff followed by hematoxylin. Tubular injury was assessed in periodic acid–Schiff-stained sections using a semiquantitative scale,^12,15 in which the percentage of cortical tubules showing epithelial cell necrosis, brush border loss, cast formation, and apoptotic bodies in the cortex was assigned a score: 0, normal; 1, <10%; 2, 10%–25%; 3, 26%–75%; 4, >75%. Sections were scored independently by two investigators who were blinded to the treatment of the animal. Ten fields of ×40 magnification were examined and averaged. The individual scoring of the slides was blinded to the genotype of the animal. To quantify leukocyte infiltration, sections were stained with rat anti-mouse neutrophil antibody (Abcam, Cambridge, MA) (1:200 dilution) followed by goat anti-rat biotin conjugate. Color was developed after incubation with ABC reagent (Vector Laboratories). Stained sections were photographed, and five ×40 fields of neutrophils were examined for quantification of leukocytes. To determine endogenous mouse netrin-1 and UNC5B protein expression, sections were stained with goat antinetrin-1 polyclonal antibody (Santa Cruz Biotechnology, Inc., Dallas, TX) and rabbit anti-UNC5B polyclonal antibody (MD Millipore Corporation, Billerica, MA) followed by secondary antibody conjugated with fluorophors or goat anti-rabbit peroxidase polymer (Vector Laboratories). To determine phospho-p53–positive cells, sections were stained with rabbit antiphospho-p53 (Ser15; Cell Signaling Technologies, Inc.). Color was developed after incubation with ABC reagent (Vector Laboratories). Stained sections were photographed using an Olympus inverted microscope with a color charge-coupled device camera.

TACS TdT In Situ Apoptosis Detection

To identify apoptotic cells, tissue sections were stained using the TACS TdT in situ Apoptosis Detection Kit (R&D Systems, Inc.) according to the manufacturer’s instruction. Briefly, tissue sections were deparaffinized, hydrated, and washed with PBS. Sections were digested with proteinase K for 15 minutes at 24°C. Slides were then washed, and endogenous peroxidase activity was quenched with 3% H₂0₂ in methanol. Slides were washed and incubated with TdT labeling reaction mix at 37°C for 1 hour and then streptavidin–horseradish peroxidase. Color was developed using TACS blue label substrate solution. Slides were washed, counterstained, and mounted with Permount. Sections were photographed, and labeled cells were counted and quantified.

Gene Expression Analysis by Real-Time RT-PCR

Real-time RT-PCR was performed in an Applied Biosystems, Inc. 7700 Sequence Detection System (Foster City, CA); 1.5 μg total RNA was reverse transcribed in a reaction volume of 20 µl using the Omniscript RT Kit and random primers. The product was diluted to a volume of 150 µl, and 6-µl aliquots were used as templates for amplification using the SYBR Green PCR Amplification Reagent (Qiagen) and inflammatory cytokine and chemokine PCR array and apoptosis PCR array (SABiosciences). The amount of DNA was normalized to a housekeeping gene, such as β-actin, glyceraldehyde-3-phosphate dehydrogenase, and hypoxanthine-guanine phosphoribosyl transferase 1, amplified in the same plate.

Apoptosis PCR Array

To determine apoptotic genes regulated in response to UNC5B deletion, PCR array (Realtimeprimers.com) was used to quantify apoptotic pathway genes.

Creation of Stable TKPTS Cell Lines Expressing Short Hairpin RNA for UNC5B

TKPTS cells stably expressing short hairpin RNA for UNC5B were created by transfecting lentiviral vectors containing short hairpin RNAs and the puromycin drug resistance selection marker (Sigma-Aldrich, St. Louis, MO). Colonies were selected based on drug resistance and propagated separately. Knockdown of UNC5B expression was determined by real-time RT-PCR. Fifteen clones for UNC5B were screened. Five clones for each gene showing ≥80% knockdown were selected and frozen in liquid nitrogen. All clones selected were shown to have normal expression of the other receptors.²¹

Cell Culture

Mouse TKPTS cells were cultured in advanced DMEM/F12 medium with 5% serum. To determine whether cisplatin-induced cell death is exacerbated in the absence of UNC5B receptor expression, WT TKPTS cells and UNC5B knockout cells were treated with vehicle, 10 µM cisplatin, or 100 µM hydrogen peroxide; 24 hours after the addition of cisplatin or hydrogen peroxide, cells were harvested, stained with PI and annexin V, and analyzed by flow cytometry or used for caspase-3 assay or p53 expression by Western blot analysis.

To determine the role of the PI3K pathway in UNC5B-mediated suppression of apoptosis and inactivation of p53, TKPTS cells were treated with 10 µM cisplatin with or without the PI3K inhibitor (10 µM LY294002; Cell Signaling Technologies) for 24 hours. Cells were harvested and used for Western blot analysis, PCR, and quantification of apoptosis by flow cytometry.

Western Blot Analysis

Protein was extracted by solubilizing cells in RIPA buffer (Sigma-Aldrich) containing a protease and phosphatase inhibitor cocktail. Protein concentration was quantified with the bicinchonic acid protein assay reagent (Pierce Biotechnology, Rockford, IL), and 50 μg total protein was loaded onto 4%–12% polyacrylamide gels, separated, and then transferred onto a polyvinylidene difluoride membrane. The membrane was probed with rabbit antiphospho-p53 (S15 and S392), antitotal-p53 (Cell Signaling Technologies), rabbit antiphospho-Akt and total Akt antibodies, rabbit antiphospho- and total ERK antibodies, rabbit anti-p38 (Cell Signaling Technologies), and rabbit anti-UNC5B antibodies (MD Millipore Corporation, Billerica, MA). Proteins were detected with enhanced chemiluminescence detection reagents (Amersham Pharmacia Biotech). Protein loading was normalized by probing the membrane with antiactin antibodies.

Quantification of Apoptosis by Flow Cytometry

To quantify the dead cells in WT and UNC5B knockout TKPTS cells, cells were harvested at 24 hours after treatment of cisplatin (10 µM) or hydrogen peroxide (100 µM). To quantify apoptosis, cells were washed and stained for Annexin V–FITC and propidium iodide (640914; Biolegend, San Diego, CA). Stained cells were immediately analyzed by flow cytometry (BD FACSCalibur), and the data were analyzed using Cyflogic V.1.2.1 software.

UNC5B Overexpression in TKPTS Cells

To determine the effects of UNC5B overexpression on renal epithelial cell apoptosis, TKPTS cells were transfected with 2 µg/well rat UNC5B expression construct (gift from Patrick Mehlen, Centre Leon Berard, Lyon, France) in a six-well plate; 48 hours after transfection, cells were treated with saline or 10 µM cisplatin, and 24 hours after cisplatin addition, cells were harvested and used for Western blot and flow cytometry to determine apoptosis and caspase-3 activity. Caspase-3 activity was quantified using an assay kit (BioAssay Systems, Hayward, CA).

Statistical Methods

All assays were performed in duplicate or triplicate. The data are reported as mean ± SEM. Statistical significance was assessed by an unpaired, two-tailed t test for single comparison or ANOVA for multiple comparisons. P<0.05 was considered significant.

Disclosures

None.