Mitogen-Activated Protein Kinase 14 Promotes AKI

Abstract

An improved understanding of pathogenic pathways in AKI may identify novel therapeutic approaches. Previously, we conducted unbiased liquid chromatography-tandem mass spectrometry–based protein expression profiling of the renal proteome in mice with acute folate nephropathy. Here, analysis of the dataset identified enrichment of pathways involving NFκB in the kidney cortex, and a targeted data mining approach identified components of the noncanonical NFκB pathway, including the upstream kinase mitogen-activated protein kinase kinase kinase 14 (MAP3K14), the NFκB DNA binding heterodimer RelB/NFκB2, and proteins involved in NFκB2 p100 ubiquitination and proteasomal processing to p52, as upregulated. Immunohistochemistry localized MAP3K14 expression to tubular cells in acute folate nephropathy and human AKI. In vivo, kidney expression levels of NFκB2 p100 and p52 increased rapidly after folic acid injection, as did DNA binding of RelB and NFκB2, detected in nuclei isolated from the kidneys. Compared with wild-type mice, MAP3K14 activity–deficient aly/aly (MAP3K14^aly/aly) mice had less kidney dysfunction, inflammation, and apoptosis in acute folate nephropathy and less kidney dysfunction and a lower mortality rate in cisplatin-induced AKI. The exchange of bone marrow between wild-type and MAP3K14^aly/aly mice did not affect the survival rate of either group after folic acid injection. In cultured tubular cells, MAP3K14 small interfering RNA targeting decreased inflammation and cell death. Additionally, cell culture and in vivo studies identified the chemokines MCP-1, RANTES, and CXCL10 as MAP3K14 targets in tubular cells. In conclusion, MAP3K14 promotes kidney injury through promotion of inflammation and cell death and is a promising novel therapeutic target.

The incidence of AKI is increasing.¹ However, there is currently no therapy that reliably prevents the progression to AKI or accelerates recovery of renal function.^2,3 Thus, reliable biomarkers and novel therapeutic approaches are needed.⁴ AKI is characterized by kidney inflammation and tubular cell death, dedifferentiation, and subsequent proliferation.^5–9 However, given the complexity and redundancies of the process, it is unlikely that targeting a single inflammatory molecule provides the kind of benefit that will be clinically relevant. Thus, attention has focused on upstream signaling pathways that may regulate the coordinated expression of an array of inflammatory molecules. The combination of unbiased protein expression profiling with focused data mining is a powerful tool to expand our knowledge of relevant pathways and key factors in disease. Liquid chromatography-tandem massspectrometry (LC-MS/MS) identified approximately 2000 proteins in murine renal cortex.¹⁰ However, its applications to the study of AKI has been limited and mainly concentrated in the analysis of biofluids such as urine or in the study of the metabolome rather than the proteome.^11–14 To identify novel pathways and mediator networks active in AKI in a comprehensive manner, we used tissue LC-MS/MS to assess changes in the renal proteome of experimental acute folate nephropathy. Bioinformatics analysis of 41,235 peptides in cortical kidney tissue by LC-MS/MS proteomics allowed the identification of 6516 unique proteins, of which 1480 were differentially expressed in samples from experimental nephrotoxic AKI as compared with controls.¹⁵ On this previously reported raw dataset, we have now performed novel pathway analysis in search of cell death or inflammatory pathways that are activated in acute folate nephropathy. This analysis indicated enrichment of proteins from the noncanonical activation pathway for transcription factor NFκB. NFκB promotes inflammation by modulating gene transcription.^16,17 Canonical NFκB activation is rapidly initiated through degradation of IκB proteins by the proteasome, thus releasing complexes that translocate to the nucleus to promote transcription of proinflammatory genes and downregulate the expression of anti-inflammatory molecules such as Klotho.^18,19 By contrast, noncanonical NFκB activation is a delayed response that is engaged by a limited number of stimuli and involves activation of the mitogen-activated protein kinase kinase kinase 14 (MAP3K14)/NFκB-inducing kinase, proteasomal processing of NFκB p100 to p52, and nuclear translocation of p52/RelB complexes.^16,20 The role and regulation of MAP3K14 in AKI is poorly understood.

The combined proteomic and bioinformatics approach enabled identification of evidence for MAP3K14 activation and the upregulation of several proteins of the noncanonical NFκB activation pathway in acute folate nephropathy that were confirmed by Western blot and immunohistochemistry. Functional studies identified chemokine expression and cell death and proliferation as novel MAP3K14-regulated processes in tubular cells. Furthermore, MAP3K14 was overexpressed during human AKI and genetically modified mice confirmed the key role of MAP3K14 in acute folate nephropathy and cisplatin-induced AKI.

Results

Kidney Tissue Proteomics Bioinformatics Analysis Identifies Upregulation of MAP3K14 and Noncanonical NFκB Components in Folate Nephropathy–Associated AKI

Acute folate nephropathy is characterized by increased serum creatinine (0.53±0.25 versus 0.10±0.02 mg/dl at 24 hours; P<0.05), tubular cell death, and interstitial inflammation.¹⁵ As previously described, unbiased proteomics combined with focused data analysis was conducted in 24-hour kidney cortex control and folate nephropathy samples.¹⁵ LC-MS/MS identified 41,235 peptides in the kidney cortex corresponding to 6516 unique, nonredundant proteins (Supplemental Figure 1).¹⁵ This study represents a new complimentary analysis of this previously generated dataset. KEGG pathway analysis identified the enrichment of several pathways on the basis of the upregulation of key proteins in folate nephropathy samples (Table 1). NFκB was at the crossroads of several of these pathways. NFκB activation is regulated by MAPK, requires ubiquitination and proteasomal processing or degradation, and regulates apoptosis and chemokine secretion. Canonically activated NFκB signaling has long been implicated in kidney injury.¹⁶ However, there is much less information on noncanonical NFκB activation and its components. A targeted data mining approach searched for components of the noncanonical NFκB pathway. A KEGG-generated NFκB signaling pathway map (Supplemental Figure 2) summarizes the expression of noncanonical NFκB signaling pathway components and NFκB2 (p100/p52) ubiquitination and proteasomal activation. Upregulation was observed for MAP3K14, the essential upstream kinase activating the noncanonical NFκB pathway,^21,22 for proteins required for NFκB2 p100 ubiquitination and proteasomal processing to active NFκB2 p52, such as UBE2M/UBC12 (E2) and CULLIN-1 (E3), and for the two main components of noncanonical NFκB DNA-binding heterodimers, NFκB2 and RELB (Table 2).

Table 1.

Signaling pathways modulated in folate nephropathy–associated AKI samples and identified by pathway analysis using KEGG database searching

KEGG Map	Number of Hits	Name
mmu04151	46	PI3K-Akt signaling pathway
mmu04010	29	MAPK signaling pathwaya
mmu04910	27	Insulin signaling pathway
mmu04020	22	Calcium signaling pathway
mmu04120	21	Ubiquitin mediated proteolysisa
mmu04062	19	Chemokine signaling pathwaya
mmu04310	17	Wnt signaling pathway
mmu04630	16	Jak-STAT signaling pathway
mmu04660	15	T cell receptor signaling pathway
mmu03320	13	PPAR signaling pathway
mmu04064	13	NFκB signaling pathwaya
mmu04370	11	VEGF signaling pathway
mmu04912	11	GnRH signaling pathway
mmu04210	11	Apoptosisa
mmu04722	10	Neurotrophin signaling pathway

^aNFκB activation is at the crossroads of these pathways.

Table 2.

Noncanonical NFκB signaling pathway and ubiquitination and proteasomal degradation proteins significantly modulated in acute folate nephropathy

Name	Gene	Fold Change	P Value
Noncanonical NFκB pathway
Nuclear factor NFκB p100 subunit	Nfkb2	1.46	0.03
Mitogen-activated protein kinase kinase kinase 14 (NIK)	Map3k14	1000	0.05
Transcription factor RelB	Relb	3.33	0.01
Ubiquitination system
Cullin-1	Cul1	1000	0.02
NEDD8-conjugating enzyme Ubc12	Ube2m	3.01	0.01

Data represent focused data mining after nonbiased analysis of the significant dataset. NIK, NF-κ-B-inducing kinase.

Validation of Noncanonical NFκB Pathway Activation in Acute Folate Nephropathy

The proteomics findings of increased MAP3K14, RELB, and NFκB2 p100/p52 were validated by Western blot and immunohistochemistry and the mechanisms by which the system is upregulated were explored by assessing mRNA expression. Kidney Map3k14, RelB, and NFκB2 mRNA expression was increased in folate nephropathy, suggesting transcriptional upregulation (Figures 1A and 2, A and C). Western blot confirmed increased kidney MAP3K14, RELB, NFκB2 p100, and NFκB2 p52 in folate nephropathy (Figures 1B and 2, B, D, and E). Immunohistochemistry localized the increased expression of these proteins to tubular cells (Figure 1C). In addition, NFκB2 p52 and RELB DNA-binding activity was increased in nuclear extracts from folate nephropathy kidneys (Figure 2F). Thus, evidence for increased activation of MAP3K14 includes processing of NFκB2 p100 to NFκB2 p52 and nuclear translocation and increased DNA binding activity of the RelB/NFκB2 p52 transcription factor. The expression of the ubiquitination pathway CULLIN-1 protein was also confirmed to be increased in folate nephropathy (Supplemental Figure 3).

Figure 1.

Increased kidney mRNA and protein expression of MAP3K14 in acute folate nephropathy and human AKI. Kidney mRNA levels were assessed by quantitative RT-PCR and protein levels by Western blot. (A) MAP3K14 mRNA, *p<0.01 versus vehicle. (B) MAP3K14 protein, *p<0.005 versus vehicle. (C) MAP3K14 immunohistochemistry. Increased MAP3K14 expression was localized to tubular cells in nephropathy samples from WT mice at 24 hours. Original magnification ×40; n=6 animals per group. (D) MAP3K14 expression in human kidney tissue. Immunohistochemistry was performed in human control and AKI tissue. Increased tubular cell immunostaining for MAP3K14 was observed in AKI. Original magnification ×20; detail ×100.

Figure 2.

Increased kidney RelB and NFκB2 expression and evidence for noncanonical NFκB activation in acute folate nephropathy. Kidney mRNA levels (A and C) were assessed by quantitative RT-PCR and protein levels by Western blot (B and D). (A) RelB mRNA, *p<0.01 versus vehicle. (B) RelB protein, *p<0.03 versus vehicle. (C) NFκB2 mRNA, *p<0.01 versus vehicle. (D) NFκB2 p100 and p52 proteins, representative Western blot. (E) NFκB2 p100 and p52 protein quantification, *p<0.03 and **p<0.05 versus vehicle. NFκB2 p100 is processed to NFκB p52 by the proteasome. (F) Increased nuclear DNA-binding activity of NFκB2 p52 and RelB in folate nephropathy. A DNA-binding ELISA was used to quantify DNA-binding activity of NFκB2 p52 and RelB in nuclei obtained from kidneys 24 hours after induction of folate nephropathy or vehicle administration. *p<0.01 versus vehicle; n=6 animals per group.

Given the poor understanding of its role in kidney injury and its upstream situation in the pathway, we further explored the role of MAP3K14 in folate nephropathy­–associated AKI. In this regard, extensive MAP3K14 immunostaining was also observed in kidney tubules in human AKI (Figure 1D).

MAP3K14 Deficient Mice Were Protected from Folate Nephropathy and Cisplatin-Induced AKI

To explore the role of MAP3K14 in folate nephropathy, we used MAP3K14 activity–deficient alymphoplasia (MAP3K14^aly/aly) mice, which carry a point mutation causing an amino acid substitution in the carboxy-terminal interaction domain of MAP3K14.^23,24 MAP3K14^+/aly heterozygote mice and MAP3K14^+/+ mice were used as controls.

MAP3K14^+/+ or MAP3K14^+/aly heterozygote mice developed folate nephropathy characterized by increased serum creatinine and urea levels (Figure 3, A and B, Supplemental Figure 4) and increased kidney NFκB2 activation (Figure 3, C and D, Supplemental Figure 4), expression of chemokines (Figure 3, E–G, Supplemental Figure 4), and interstitial macrophages.

Figure 3.

MAP3K14 deficient mice were protected from folate nephropathy–associated AKI. (A) Serum creatinine, *p<0.02 versus heterozygous mice. (B) Serum urea, *p<0.001 versus heterozygous mice. (C) NFκB2 p100 and p52 proteins (representative Western blot). (D) NFκB2 mRNA, *p<0.01 versus heterozygous mice with folate nephropathy. (E) Decreased whole kidney MCP-1, (F) RANTES, and (G) CXCL10 mRNA expression in MAP3K14 deficient mice with folate nephropathy compared with heterozygous mice. *p<0.02 versus heterozygous mice with folate nephropathy. (H) CCL21a mRNA expression. Mean±SD of six mice per group at the 72-hour time-point; *p<0.03 versus heterozygous mice with folate nephropathy.

MAP3K14 deficient mice were protected from folate nephropathy. Serum creatinine and urea (Figure 3, A and B, Supplemental Figure 4) and kidney expression of NFκB2 p100/52 protein and mRNA (Figure 3, C and D, Supplemental Figure 4), MCP-1 (Figure 3E, Supplemental Figure 4), RANTES (Figure 3F, Supplemental Figure 4), CXCL10 (Figure 3G, Supplemental Figure 4), and Ccl21a mRNA expression (Figure 3H, Supplemental Figure 4) were lower than in control MAP3K14^+/+ mice or MAP3K14^+/aly heterozygote mice with folate nephropathy.

Finally, we tested a different model, cisplatin-induced AKI. An initial study at 72 hours resulted in 5/5 (100%) mortality in wild-type (WT) mice, whereas mortality in MAP3K14 deficient mice was 0/5 (0%). Next we studied the outcome at 48 hours. Mortality was 1/7 (14%) in WT mice and 0/5 (0%) in MAP3K14 deficient mice. Furthermore, renal function was preserved in MAP3K14 deficient mice as compared with WT mice (Figure 4), suggesting that mortality was due to severe AKI.

Figure 4.

MAP3K14 deficient mice were protected from experimental cisplatin-induced AKI. (A) Serum creatinine. (B) Serum urea. *p<0.02, **p<0.04 versus MAP3K14^+/+ cisplatin-induced AKI mice. Mean±SD of 5–6 mice per group at the 48-hour time-point.

Immunohistochemistry confirmed the lack of RELB (Figure 5A) and NFκB2 P52 expression in MAP3K14 deficient mice with folate nephropathy (Figure 5B), and disclosed decreased F4/80 macrophages and CD3 T lymphocytes (Figure 6, A and B) and deoxynucleotidyl-transferase–mediated dUTP nick-end labeling (TUNEL) positive tubular cells representing dying cells (Figure 6C) in MAP3K14 deficient mice compared to heterozygous controls with folate nephropathy.

Figure 5.

MAP3K14 deficient mice are protected from tubular noncanonical NFκB pathway activation in acute folate nephropathy. (A) RelB and (B) p100/52 immunohistochemistry. Nuclear p52 is observed in renal tubules from heterozygous mice with folate nephropathy (arrows), whereas no staining was observed in MAP3K14 deficient mice with folate nephropathy. Immunohistochemistry does not discriminate between NFκB2 p100 and NFκB2 p52. However, Western blot shown in Figure 4C shows the presence of the active NFκB2 p52 protein. Images representative of six animals per group at the 72-hour time-point; original magnification ×40; detail ×400; n=6 animals per group.

Figure 6.

MAP3K14 deficient mice were protected from acute folate nephropathy–induced inflammation and cell death. (A) F4/80 macrophage and (B) CD3 immunohistochemistry. Macrophage infiltration is milder in MAP3K14 deficient mice with folate nephropathy than in heterozygous mice with folate nephropathy. * P<0.001; ** P<0.02; original magnification ×20. (C) TUNEL for fragmented DNA characteristic of apoptosis was frequently positive in tubular cells in heterozygous mice with folate nephropathy. The rate of apoptosis was lower in MAP3K14 deficient mice with folate nephropathy. * P<0.03; original magnification ×20; mean±SD of six mice per group at the 72-hour time-point.

We next induced folate nephropathy in bone marrow chimeras to test whether MAP3K14 deficiency in kidney cells or in bone marrow–derived cells was responsible for nephroprotection. Mice on a MAP3K14^aly/aly background were protected from folate nephropathy–induced death when compared with MAP3K14 ^+/+ mice and this was independent of the bone marrow genotype (Supplemental Figure 5).

MAP3K14 Function in Tubular Cells

After the findings of MAP3K14 upregulation and of evidence for MAP3K14 activation (NFκB2 P100 processing to P52) in tubular cells in vivo, and a beneficial effect of MAP3K14 deficiency in vivo, the function of MAP3K14 was explored in cultured murine tubular epithelial cells by small interfering RNA (siRNA) targeting (Figure 7, A and B). Because KEGG pathway analysis had identified chemokine signaling and apoptosis as enriched pathways and MAP3K14 deficiency indeed resulted in lower inflammation and tubular cell death in vivo, we explored the potential regulation of chemokine secretion and cell death by MAP3K14 in tubular cells. For this we took advantage of TWEAK, the only cytokine characterized to date to activate the noncanonical NFκB pathway in tubular cells.²⁵ In order to assess for further functions of MAP3K14 we explored canonical NFκB targets, including CXCL10, whose expression was recently related to MAP3K14 polymorphisms in human lymphoblastoid cells, but that had not previously been linked to MAP3K14 by functional studies.²⁶ MAP3K14 silencing by specific siRNAs prevented TWEAK-induced upregulation of Cxcl10 mRNA (Figure 7C) and protein (Figure 7D) as well as of canonical NFκB targets such as Mcp1 and Rantes²⁷ (Figure 7, E and F, Supplemental Figure 6). Differences in MCP-1 expression, which peaks earlier than RANTES, were more evident at earlier time points (Supplemental Figure 6). In this regard some genes are targeted by both canonical and noncanonical NFκB, whereas others such as CCL21 are specifically targeted by noncanonical NFκB activation in tubular and extrarenal cells.^25,28,29

Figure 7.

Functional characterization of MAP3K14 actions on cultured proximal tubular cells: chemokine expression. (A) MAP3K14 siRNA silencing in cultured murine proximal tubular cells suppressed MAP3K14 protein expression. Representative Western blot. (B) MAP3K14 siRNA silencing in cultured murine proximal tubular cells suppressed MAP3K14 mRNA expression. (C) MAP3K14 siRNA silencing prevents CXCL10 mRNA upregulation induced by a 24-hour stimulation by the noncanonical NFκB activator TWEAK (100 ng/ml). qRT-PCR. *p<0.01 versus control; **p<0.01 versus TWEAK alone. (D) MAP3K14 siRNA silencing prevents the increase in culture supernatants of the CXCL10 chemokine induced by exposure for 24 hours to 100 ng/ml TWEAK (ELISA). *p<0.001 versus control; **p<0.01 versus TWEAK alone. (E) MAP3K14 siRNA silencing prevents MCP1 mRNA upregulation induced by the noncanonical NFκB activator TWEAK. qRT-PCR. *p<0.001 versus scrambled; **p<0.001 versus TWEAK alone. (F) MAP3K14 siRNA silencing prevents RANTES mRNA upregulation induced by TWEAK. qRT-PCR. *p<0.002 versus scrambled; **p<0.003 versus TWEAK alone. Cells were treated with scramble or MAP3K14 siRNA before addition of 100 ng/ml TWEAK for 24 hours; mean±SD of three independent experiments.

Deprivation of the survival factors from serum is a classic inducer of apoptosis.³⁰ MAP3K14 targeting decreased apoptosis in tubular cells cultured in the absence of survival factors (Figure 8).

Figure 8.

Functional characterization of MAP3K14 actions on cultured proximal tubular cells: cell death. (A) MAP3K14 siRNA silencing decreases spontaneous apoptosis of serum-deprived tubular cells. Representative flow cytometry diagrams of cell DNA content. Hypodiploid cells consistent with apoptosis are indicated by a horizontal bar. (B) Quantification of hypodiploid apoptotic cells. *p<0.05 versus control; **p<0.03 versus TWEAK/TNFα/INFγ alone; mean±SD of three independent experiments.

Discussion

For the first time we have uncovered evidence that MAP3K14 is a therapeutic target in kidney injury. A nonbiased proteomics characterization of acute folate nephropathy kidneys disclosed enrichment for components of the noncanonical NFκB activation, chemokine, and apoptosis pathways. Further studies evidenced activation of the apical kinase of the noncanonical NFκB pathway, and showed that in kidney tubular cells MAP3K14 regulates the expression of chemokines not previously associated with noncanonical NFκB, such as CXCL10, and cell survival. In vivo MAP3K14 targeting protected from folate nephropathy and cisplatin-induced AKI, improving renal function and decreasing inflammation and tubular cell death.

NFκB is a family of structurally homologous proteins, including NFκB1, NFκB2, RELA, RELB, and C-REL, which form homo- or hetero-dimers that bind to κB enhancers in DNA to promote or inhibit gene transcription.¹⁶ Two main pathways for NFκB activation are known. Canonical NFκB activation is usually a rapid, protein synthesis–independent and transient response to a wide range of stimuli that involves proteasomal degradation of cytosolic IκB inhibitory proteins leading to the release of RELA/P50 and other dimers that then migrate to the nucleus. NFκB-driven IκBα resynthesis contributes to a fast turn-off of the response. By contrast, noncanonical NFκB activation requires MAP3K14 activation and NFκB2 P100 processing to P52 by the proteasome, resulting in a delayed nuclear translocation of RELB/P52 heterodimers and in prolonged activation of NFκB target genes.^31–33 Increased transcription of NFκB2 and RELB may contribute to persistence of the response.³⁴ In contrast to the canonical pathway, only a limited number of stimuli are known to activate the noncanonical NFκB pathway. These include advanced glycosylation end products³³ and TNF receptor superfamily members such as lymphotoxin-β receptor, B-cell activating factor, CD40, receptor activator for NFκB, CD27, and the TWEAK receptor Fn14.^20,25,35 None of these receptors or their ligands was identified in the proteomic analysis of kidney cortex. However, some of them had been previously shown to contribute to AKI. A literature search revealed a role in AKI for CD27 and TWEAK/Fn14.^36–38 CD27 was localized to sloughed cells in tubular lumens postischemia and CD27-deficient mice were protected from AKI and tubular cell apoptosis.^36,37 In contrast to this single report, multiple studies from several institutions have provided evidence for a role of TWEAK/FN14 in kidney injury.³⁹ Furthermore, TWEAK targeting decreased the expression of the noncanonical NFκB target CCL21 in tubular cells.²⁵ Thus for cell culture studies we chose TWEAK as a pathophysiologically relevant activator of the noncanonical NFκB pathway.

There is evidence for a role of canonical activation of NFκB in kidney injury.¹⁶ However, no therapeutic approach specifically targeting NFκB systemically is undergoing clinical trials, suggesting a fundamental lack of understanding of the system. In this regard, the role of MAP3K14 and noncanonical NFκB activation in AKI has not been well characterized. There is very little and scattered information on activation of this pathway in kidney disease. MAP3K14 was phosphorylated in tubular cells during kidney ischemia-reperfusion⁴⁰ and levels were increased in experimental diabetic nephropathy and human delayed graft function.^40,41 TWEAK-dependent nuclear translocation of RELB and p52 was observed in tubular cells in nephrotoxic AKI.²⁵ In this report, kidney tissue proteomics identified upregulation of several proteins in the noncanonical NFκB pathway, upregulation of these proteins was confirmed and localized to tubular cells, the contribution of transcriptional regulation was identified, and the role of MAP3K14 in tubular cell injury was characterized. In this regard, MAP3K14 activity–deficient mice were protected from folate nephropathy and cisplatin-induced AKI. Thus, MAP3K14 represents a key regulated step promoting AKI that may potentially be subject to therapeutic manipulation, although at present there are no satisfactory MAP3K14 inhibitors.⁴²

MAP3K14 is the essential upstream serine/threonine kinase of the noncanonical NFκB pathway that binds to TRAF2 and participates in NFκB signaling in response to the TNF superfamily and IL-1 receptors.²² MAP3K14 protein concentrations are low in quiescent cells as a result of rapid degradation. Cytokines and oxidative stress may increase MAP3K14 protein stability, leading to MAP3K14 activation.²⁰ In addition to this universal mechanism of MAP3K14 regulation, we found increased steady-state Map3k14 mRNA levels to be an additional regulatory mechanism of MAP3K14 expression in tubular epithelium that also takes place in vivo during AKI. MAP3K14 induces IκB kinase-α–mediated phosphorylation of NFκB2 p100, a prerequisite for p100 ubiquitination and subsequent proteasomal processing to active NFκB2 p52.³²

Ubiquitination is required for targeting of specific proteins to the proteasome. F-box proteins provide specificity for substrate recognition in the S-phase kinase associated protein 1-cullin 1-F-box protein (SCF) family of the Cullin-RING ligases E3 ubiquitin ligase superfamily.⁴³ The F-box protein β-transducin repeat containing (β-TrCP; FBW1A) provides substrate specificity for MAP3K14-phosphorylated p100, allowing ubiquitination by SCF^β-TrCP.^44–47 Efficient NFκB2 p100 ubiquitination requires UBA3 (ube1c) and UBE2M (UBC12), UBE2D3 (UBCH5c), and intact CULLIN-1 in SCF^β-TrCP.⁴⁵ Interestingly, the UBE2M E2 ubiquitin conjugating enzyme and CULLIN-1 of the SCF^β-TrCP E3 ubiquitin ligase were found to be upregulated in folate nephropathy.

Evidence for MAP3K14 activation in vivo in folate nephropathy included increased MAP3K14 levels, NFκB p100 processing to NFκB p52, increased nuclear localization and DNA binding activity of P52/RELB, and decreased kidney inflammation and cell death and preserved renal function in MAP3K14 activity–deficient mice. Protection from folate nephropathy may depend on systemic MAP3K14 deficiency (e.g., leukocyte MAP3K14 deficiency) and/or kidney MAP3K14 deficiency. Bone marrow transplantation experiments results suggest that MAP3K14 deficiency in marrow-derived cells is not required for nephroprotection, and are consistent with, but do not confirm, the hypothesis that renal cell MAP3K14 targeting is important for nephroprotection. Alternatively, MAP3K14 targeting in additional, nonrenal, nonmarrow-derived cells may play a role. The fact that nonrenal cells also express MAP3K14 may result in undesired side effects when targeting MAP3K14 with small molecules. MAP3K14 and noncanonical NFκB gene targets have been characterized in the immune system, but there is little information on kidney cells.¹⁶ We now provide evidence of a role of MAP3K14 in the regulation of the inflammatory and cell death/proliferation responses in cultured tubular cells that together with the upregulation of MAP3K14 in tubular cells in folate nephropathy suggest at least a partial contribution of MAP3K14 targeting in tubular cells to the therapeutic responses.

The P52/RELB heterodimers characteristic of MAP3K14-initiated noncanonical NFκB activation share a number of gene targets with RELA-containing, classically activated NFκB complexes.^16,29 Because canonical NFκB activation is an early transient response peaking at around 1–3 hours, whereas noncanonical NFκB activation is delayed and peaks at around 24 hours, noncanonical NFκB activation may contribute to sustained NFκB-dependent gene expression.^16,28,48 In this regard, the CC genotype of the MAP3K14 SNP rs7222094 was recently associated with increased mortality and renal dysfunction in septic shock patients.²⁶ CXCL10 was the gene with the greatest difference in expression between major and minor MAP3K14 genotypes. The rs7222094 genotype strongly associated with decreased CXCL10 levels in lymphoblastoid cell lines and in septic shock patients.²⁶ Urinary CXCL10 is increased in AKI patients^49,50 and, as shown here, in folate nephropathy kidney tissue. We now provide for the first time direct functional evidence that persistent CXCL10 expression in response to TWEAK is regulated by MAP3K14. CXCL10 (IP-10) had long been associated with kidney injury in animals and humans.^51,52 MCP-1 and RANTES, which are targets of canonical NFκB activation,⁵³ were also found to be MAP3K14-dependent in tubular cells. CCL21a was previously shown to be MAP3K14-dependent in this cell system.²⁷ Thus, a wide spectrum of chemokines, commonly considered as both canonical NFκB targets and noncanonical NFκB targets, is regulated by MAP3K14 in cultured tubular cells and during folate nephropathy.

KEGG pathway analysis also disclosed apoptosis pathways as overrepresented in the folate nephropathy–associated AKI proteome. MAP3K14 had been identified as a cell death regulator in cancer cells. Indeed, Map3k14 siRNA targeting reduced serum deprivation–induced death in tubular cells. These results were consistent with decreased tubular cell apoptosis in vivo in MAP3K14 activity–deficient mice during folate nephropathy. These results are also consistent with observations targeting another component of the noncanonical NFκB pathway, RELB. Thus, RelB targeting by siRNA protected mice against lethal kidney ischemia,⁵⁴ and in cultured proximal tubular cells knockdown of RELB abrogated the excess apoptosis induced by TNF in combination with cisplatin.⁵⁵

In conclusion, preclinical functional studies in cell culture and in vivo identified MAP3K14 as a promising therapeutic target in kidney injury. In this regard MAP3K14 was upregulated during human kidney injury, suggesting that experimental findings may be applicable to the clinical settings. This information sets the stage for the exploration of the potential of MAP3K14 as a therapeutic target in humans.

Concise Methods

Animal Model

Studies were conducted in accord with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Acute folate nephropathy causes reversible increase in serum creatinine and urea, tubular cell death, compensatory tubular cell proliferation, activation of an inflammatory response, and eventual progression to mild fibrosis.^27,56–58 Indeed, folic acid nephropathy has been reported in humans.⁵⁹ C57/BL6 female mice (12- to 14-week-old) from the Instituto Investigacion Sanitaria-Fundacion Jimenez Diaz animal facilities received a single intraperitoneal (ip) injection of folic acid (Sigma-Aldrich, St. Louis, MO), 250 mg/kg in 0.3 mol/L sodium bicarbonate or vehicle, and were euthanized 24 hours or 72 hours after injection (n=6 per group). The kidneys were perfused in situ with cold saline before removal. One half-kidney from each mouse was fixed in buffered formalin, embedded in paraffin, and used for immunohistochemistry, and the other half was snap-frozen in liquid nitrogen for RNA and protein studies. The cortex from one kidney obtained 24 hours after folic acid or vehicle injection was carefully separated and snap-frozen for proteomics analysis.

To assess the role of MAP3K14 in folate nephropathy, MAP3K14^aly/aly mice deficient in MAP3K14 and MAP3K14^+/aly heterozygote or MAP3K14 ^+/+ littermate control mice received a single ip injection of folic acid (Sigma-Aldrich), 250 mg/kg in 0.3 mol/L sodium bicarbonate or vehicle, and were euthanized 72 hours after injection (n=5 per group). MAP3K14^aly/aly mice deficient in MAP3K14 were from the Centro Biologia Molecular (Madrid, Spain) animal facilities.²³

A different model of AKI was induced by the ip injection of a single dose of 25 mg/kg cisplatin (Sigma-Aldrich) dissolved in 0.9% saline solution. The cisplatin dose was based on literature analysis and results of preliminary experiments, showing renal function impairment at day three after cisplatin injection. MAP3K14^aly/aly mice (n=10) and MAP3K14^+/+ littermate control mice (n=12) were used in these experiments and sacrificed at 48 and 72 hours.

Generation of Bone Marrow Chimera

Recipient MAP3K14^+/+ littermate control mice and MAP3K14^aly/aly mice at age 6 weeks were γ-irradiated with two doses of 5 Gy for ablation of endogenous bone marrow cells. For bone marrow transplantation, bone marrow cells were isolated (donor) by flushing the femurs and tibias using a 25 g needle with DMEM (Invitrogen, Carlsbad, CA). After resuspension, bone marrow cells were centrifuged (300 × g, 5 minutes, 4 °C). After resuspension with ice-cold DMEM, bone marrow cells were filtered through a 35 μm filter. Irradiated recipient MAP3K14^+/+ littermate control and MAP3K14^aly/aly mice were injected intravenously with 4 × 10⁶ donor bone marrow cells (in 100 μl per recipient) within 4 hours after the last irradiation dose. Eight weeks after bone marrow transplantation, bone marrow chimeric mice (four groups of five mice: recipient MAP3K14^+/+ with donor MAP3K14^+/+, recipient MAP3K14^+/+ with donor MAP3K14^aly/aly, recipient MAP3K14^aly/aly with donor MAP3K14^+/+, and recipient MAP3K14^aly/aly with donor MAP3K14^aly/aly) were subjected to folic acid nephropathy and euthanized at 72 hours.

Sample Preparation and Mass Spectrometry Analysis

Tissue samples were weighed out and extracted using the filter aided sample preparation method,⁶⁰ as described previously.¹⁵ Briefly, tissue samples were homogenized in SDS-lysis buffer (1:10 sample to buffer ratio) (0.1 M Tris-HCl pH 7.6 supplemented with 4% SDS and 0.1 M DTT) using an Ultra-Turrax T 25 (IKA, Staufen, Germany), incubated at 95°C for 3 minutes, and clarified by centrifugation at 16,000 × g for 5 minutes at room temperature. An aliquot of the supernatant was taken and placed in a Micron YM-30 filter device (EMD Millipore, Watford, UK). Eight M urea buffer was added to the protein extract, centrifuged at 14,000 × g for 15 minutes, and then repeated. The protein extract was then mixed gently for 1 minute with 0.05 M iodoacetamide buffer and incubated for a further 20 minutes before centrifugation. Eight M urea buffer was again added and centrifuged (twice). Ammonium bicarbonate buffer (50 mM NH₄HCO₃, pH 8) was added and centrifuged (twice) before incubating overnight with trypsin. The trypsin homogenate was centrifuged and washed with the ammonium bicarbonate buffer before acidification with 10% formic acid. Sample volumes were adjusted to match final concentration of protein before analysis by LC-MS/MS.

Tissue extracts were separated on a Dionex Ultimate 3000 RSLS nano flow system (Dionex, Camberly, UK). A 5 µl sample was loaded in 0.1% formic acid and acetonitrile (98:2) onto a Dionex 100 µm × 2 cm, 5 µm C18 nano trap column at a flow rate of 5 µl/min. Elution was performed on an Acclaim PepMap C18 nano column 75 μm × 50 cm, 2 μm, 100 Å with a linear gradient of solvent A, 0.1% formic acid and acetonitrile (98:2), against solvent B, 0.1% formic acid and acetonitrile (20:80), starting at 1% B for 5 minutes rising to 30% at 400 minutes then to 50% B at 480 minutes. The sample was ionized in positive ion mode using a Proxeon nano spray ESI source (Thermo Fisher Scientific, Vernon Hills, IL) and analyzed in an Orbitrap Velos FTMS (Thermo Finnigan, Bremen, Germany). The MS was operated in a data-dependent mode (top 40) to switch between MS and MS/MS acquisition and parent ions were fragmented by collision-induced dissociation. Data files were searched against the IPI mouse nonredundant database using SEQUEST with enzyme specified as trypsin. A fixed modification of carbamidomethylation was set and oxidation of methionine and proline as variable modifications were selected. Mass error windows of 20 ppm and 0.8 Da were allowed for MS and MS/MS, respectively. In SEQUEST, only peptides that showed mass deviation of less than 10 ppm were passed, and the peptide data were extracted using high peptide confidence and top one peptide rank filters. Statistical P value analysis was performed using the Mann–Whitney U test.

Bioinformatics Analysis

Protein identification and a significant dataset of 1480 entries with P values <0.05 and fold changes of >2 have been previously described.¹⁵ This dataset was used for metabolic and signaling pathway analysis using the KEGG web-resource (www.genome.jp/kegg-bin/) or with PathVisio (www.pathvisio.org). Focused data mining was then amplified to all molecules with a P value ≤0.05.

Cells and Reagents

MCT cells are a cultured line of proximal tubular epithelial cells harvested originally from the renal cortex of SJL mice and have been extensively characterized.⁶¹ MCT cells were cultured in RPMI 1640 (Gibco, Carlsbad, CA), 10% decomplemented FBS, 2 mM glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin, in 5% CO₂ at 37°C.⁶¹ Recombinant human soluble TWEAK (EMD Millipore, Billerica, MA) was used at 100 ng/ml.

Western Blot Analysis

Tissue and cell samples were homogenized in lysis buffer⁶² then separated by 10% or 12% SDS-PAGE under reducing conditions and transferred to PVDF membranes (EMD Millipore, Bedford, MA), blocked with 5% skimmed milk in PBS/0.5% v/v Tween 20 for 1 hour, and washed with PBS/Tween. Primary antibodies were rabbit polyclonal anti-p100/52 (1:500; Cell Signaling Technology, Danvers, MA), anti-RelB (1:500; Santa Cruz Biotechnology, Santa Cruz, CA), anti-MAP3K14 (1:1000; Cell Signaling Technology), anti-Cyclin D1 (1:1000; Cell Signaling Technology) and anti–cullin-1 (1:500; Santa Cruz Biotechnology). Antibodies were diluted in 5% milk PBS/Tween. Blots were washed with PBS/Tween and subsequently incubated with appropriate horseradish peroxidase–conjugated secondary antibody (1:2000; GE Healthcare, Waukesha, WI). After washing, the blots were developed with the chemiluminescence method. Blots were then reprobed with monoclonal anti-mouse α-tubulin antibody (1:2000; Sigma-Aldrich) and levels of expression were corrected for minor differences in loading.

Quantitative RT-PCR

One µg RNA isolated by Trizol (Invitrogen) was reverse transcribed with High Capacity cDNA Archive Kit and real-time PCR was performed on an ABI Prism 7500 PCR system (Applied Biosystems, Foster City, CA) using the DeltaDelta Ct method.⁶³ Expression levels are given as ratios to GAPDH. Predeveloped primer and probe assays were from Applied Biosystems, Foster City, CA.

Immunohistochemistry

Immunohistochemistry was carried out as previously described on paraffin-embedded 5 µm thick tissue sections.⁶² Primary antibodies were rabbit polyclonal anti-RelB (1:50; Santa Cruz Biotechnology), anti-NFκB2 p100/p52 (1:20; Santa Cruz Biotechnology), anti-MAP3K14 (1:100; Cell Signaling Technology, Danvers, MA), rat polyclonal anti-F4/80 antigen (1:50; Serotec, Oxford, UK), rabbit monoclonal anti-CD3 (1:100; Dako, Denmark), and anti–Cullin-1 (1:80; Santa Cruz Biotechnology). Sections were counterstained with Carazzi hematoxylin. Negative controls included incubation with a nonspecific immunoglobulin of the same isotype as the primary antibody.

Apoptosis was assayed by TUNEL (In Situ Cell Death Detection Kit; Roche, Basel, Switzerland) according to the manufacturer’s instructions.⁶³

For human kidney immunohistochemistry, control kidney tissue from nephrectomy specimens (n=4) and AKI tissue (n=7) diagnosed as “acute tubular necrosis” was studied. Mean age was 36.4±18.6 years, four patients were females, and serum creatinine ranged from 1.7 to 10.0 mg/dl (5.7±3.5 mg/dl). Immunohistochemistry was performed as described above by using anti-human MAP3K14 from Abcam.

Transfection with Small Interfering RNA

Cells were grown in six-well plates (Costar, Cambridge, MA) and transfected with a mixture of 20 nmol/ml MAP3K14 siRNA (Santa Cruz Biotechnology), Opti-MEM I Reduced Serum Medium, and Lipofectamine 2000 (Invitrogen).⁶⁴ After 18 hours, cells were washed and cultured for 6 hours in complete medium, and serum-depleted for 24 hours before addition of stimulus. This time point was selected from a time-course of decreasing MAP3K14 protein expression in response to siRNA. A negative control scrambled siRNA provided by the manufacturer did not reduce MAP3K14 protein.

Cell Death and Apoptosis

Cells were cultured to subconfluence in six-well plates and transfected with MAP3K14 siRNA as previously described.⁶⁵ Apoptosis was assessed by flow cytometry of DNA content. For assessment of the cell cycle and apoptosis, adherent cells were pooled with spontaneously detached cells, and stained in 100 µg/ml propidium iodide, 0.05% NP-40, 10 µg/mL RNAse A in PBS at 4°C for >1 hour. This assay permeabilizes the cells. Permeabilization allows entry of propidium iodide into all cells, dead or alive. Apoptotic cells are characterized by a lower DNA content (hypodiploid cells) because of nuclear fragmentation. Thus, this assay is not based on the known ability of propidium iodide to enter dead cells. The percentage of apoptotic cells with decreased DNA content (A_o) was counted.³⁰

ELISA

Cells were transfected with MAP3K14 siRNA and stimulated with 100 ng/ml TWEAK. Murine CxCL10 in the supernatants was determined by ELISA (BD Pharmingen, San Diego, CA).

NFκB DNA-Binding Activity

RelB and NFκB2 p52 subunits in nuclear extracts from kidney tissue were assessed by their binding to an oligonucleotide containing the NFκB consensus site using TransAM NFκB Family Kit (Active Motif, Carlsbad, CA).

Statistics

Statistical analysis was performed using SPSS 11.0 statistical software (IBM, White Plains, NY). Results are expressed as mean±SD. Significance at the P<0.05 level was assessed by t test for two groups of data and ANOVA for three or more groups.

Disclosures

H.M. is the cofounder and co-owner of Mosaiques Diagnostics (Hannover, Germany).