Abstract
P21, a cyclin kinase inhibitor, is upregulated by renal ‘ischemic preconditioning’ (IPC), and induces a ‘cytoresistant’ state. However, P21-induced cell cycle inhibition can also contribute to cellular senescence, a potential adverse renal event. Hence, this study assessed whether: (i) IPC-induced P21 upregulation is associated with subsequent renal senescence; and (ii) preconditioning can be established ‘independent’ of P21 induction and avoid a post-ischemic senescent state?
CD-1 mice were subjected to either IPC (5–15 min) or to a recently proposed ‘oxidant-induced preconditioning’ (OIP) strategy (tin protoporphyrin-induced heme oxygenase inhibition +/− parental iron administration). P21 induction [messenger RNA (mRNA)/protein], cell proliferation (KI-67, phosphohistone H3 nuclear staining), kidney senescence (P16ink4a; P19Arf mRNAs; senescence-associated beta-galactosidase levels) and resistance to ischemic acute kidney injury were assessed.
IPC induced dramatic (10–25×) and persistent P21 activation and ‘downstream’ tubular senescence. Conversely, OIP did not upregulate P21, it increased, rather than decreased, cell proliferation markers, and it avoided a senescence state. OIP markedly suppressed ischemia-induced P21 up-regulation, it inhibited the development of post-ischemic senescence and it conferred near-complete protection against ischemic acute renal failure (ARF). To assess OIP’s impact on a non-P21-dependent cytoprotective pathway, its ability to activate Nrf2, the so-called ‘master regulator’ of endogenous cell defenses, was assessed. Within 4 h, OIP activated each of three canonical Nrf2-regulated genes (NQO1, SRXN1, GCLC; 3- to 5-fold mRNA increases). Conversely, this gene activation pathway was absent in Nrf2−/− mice, confirming Nrf2 specificity. Nrf2−/− mice also did not develop significant OIP-mediated protection against ischemic ARF.
OIP (i) activates the cytoprotective Nrf2, but not the P21, pathway; (ii) suppresses post-ischemic P21 induction and renal senescence; and (iii) confers marked protection against ischemic ARF. In sum, these findings suggest that OIP may be a clinically feasible approach for safely activating the Nrf2 pathway, and thereby confer protection against clinical renal injury.
INTRODUCTION
It has long been recognized that a bout of acute kidney injury (AKI) can elicit renal resistance to subsequent ischemic or toxic insults [1–3]. Characteristics of the phenomenon, commonly referred to as ‘preconditioning’, or ‘acquired cytoresistance’, include the following: (i) after application of a preconditioning stimulus (e.g. ischemia), a delay of ∼12–24 h is required before cytoresistance is fully expressed [4, 5]; (ii) during this period, increased synthesis and renal accumulation of diverse cytoprotective proteins result (e.g. heat shock proteins/heme oxygenase 1, ferritin, bilirubin reductase, hemopexin, haptoglobin, alpha 1 anti-trypsin, interleukin 10) [3, 5]; and (iii) given their diverse cytoprotective actions, these multifunctional proteins induce broad-based protection against subsequent toxic and ischemic renal damage [5–11].
P21 is a critical cell cycle regulatory protein that promotes cell cycle arrest by inhibiting multiple cyclin-dependent kinases, most notably, CDK2 [12–21]. In addition, P21 is dramatically upregulated in tubular cells in response to diverse forms of AKI [12–23]. At first consideration, an injury-induced P21 increase would seemingly be a maladaptive response, given that a blockade of cell proliferation could prevent replacement of lethally damaged tubular cells. Conversely, cellular protection can also result, given that P21 can activate multiple pro-survival (e.g. antiapoptotic, antioxidant) pathways [12–23]. Furthermore, transient cell cycle arrest may confer cellular protective effects [12, 20]. Supporting the concept of P21-mediated cytoprotection are well-established observations that P21-deficient (−/−) mice are highly susceptible to ischemic or toxic acute renal failure (ARF) [17, 20].
Given P21’s cytoprotective properties, Nishioka et al. [24] recently proposed that its upregulation, and resulting cell cycle inhibition, are responsible for the so-called ‘ischemic preconditioning’ (IPC) and the emergence of the above-noted renal cytoresistant state. Their conclusion was based on the following observations: (i) by 24 h post-induction of IPC (20 min of intermittent ischemia), marked P21 messenger RNA (mRNA) and protein increases were observed; (ii) a suppression of tubular cell proliferation resulted (as assessed by KI-67 nuclear histochemical staining); (iii) by 24 h post-IPC (a time corresponding to the marked P21 increases), dramatic protection against severe renal ischemia resulted; and (iv) when IPC was deployed in P21−/− mice, protection against AKI was no longer observed. Further supporting a role for P21/cell cycle inhibition in the emergence of cytoresistance comes from studies of a uranyl nitrite-induced preconditioning state [25, 26].
Despite the above observations, it remains unclear as to whether P21 upregulation is a cellular ‘prerequisite’ for all renal preconditioning-induced cytoresistant states. In this regard, we have recently reported that a variety of renal pharmacologic preconditioning agents, including tin protoporphyrin, iron sucrose (FeS) and nitrite-myoglobin, when used alone or in combination, induce broad-based protection against diverse forms of ischemic and nephrotoxic AKI, and prevent post-AKI progression to chronic kidney disease [6, 7]. Based on these studies, we have advanced the hypothesis that these forms of preconditioning emerge from the induction of mild, transient renal oxidative stress [27], which then activates the redox-sensitive Nrf2 pathway (the so-called ‘master regulator’ of endogenous cellular defenses) [27]. This culminates in the accumulation of diverse cytoprotective proteins, such as those noted above.
It has not been discerned whether the above oxidant-induced preconditioning (OIC) strategies also upregulate P21 expression which then helps induce a renal protected state. Noteworthy in this regard is that P21 activation is predominantly regulated via P53, not Nrf2. Thus, were P21 involved, this would provide additional mechanistic insights into OIC pathway(s). Alternatively, a potential adverse consequence of P21 upregulation/cell cycle inhibition is the onset of cellular senescence, a well-known pro-inflammatory state [28–32]. Whether ischemic or oxidant preconditioning impacts renal senescence is unknown.
There is a strong clinical interest in identifying preconditioning strategies to protect patients against AKI [33, 34]. Hence, the present study was undertaken to better define the mechanisms and consequences of the above-noted ischemic and OIP strategies.
MATERIALS AND METHODS
Animal use
CD-1 mice
Male CD-1 mice (30–40 g; Charles River Laboratories, Wilmington, MA, USA) were used for all experiments, excepting those that utilized Nrf2 deletion mice, as discussed below. All mice were maintained under routine vivarium conditions with free food and water access. All surgical protocols were performed under deep pentobarbital (∼50 mg/kg intraperitoneally) anesthesia. The protocols were approved by the Institutional Animal Care and Use Committee (IACUC).
Nrf2−/− mice
To assess the role of Nrf2 in select preconditioning responses, some experiments used male Nrf2−/− mice (B6.129X1-Nfe212/J), obtained from Jackson Laboratories, Bar Harbor, ME, USA. Male wild-type (Nrf2+/+) mice (C57 BL/6 J; Jackson Laboratories) were used as controls. The mice were maintained under routine vivarium conditions and studied ∼2 weeks after their receivership (10–12 weeks of age). No phenotypic differences (11 backcrosses) have been reported between these Nrf2−/− versus these wild-type controls (as per Jackson Laboratories; and observations during the present experiments).
Nrf2 breeding mice
In addition to the above, two breeding pairs of Nrf2−/− mice (Jackson Laboratories) were used to generate a small Nrf2−/− mouse colony, using a standard IACUC-approved breeding protocol. The offspring mice were studied at 12–14 weeks of age. They were paired by sex and litter for subsequent experiments.
Employed OIP protocols
Three variations of OIP were studied, as detailed in prior reports [6, 7, 27] and as follows:
SnPP alone: the heme oxygenase inhibitor, SnPP, induces a transient prooxidant state [27]. This experiment assessed whether P21 [as well as two other CDK inhibitors: P16ink4a (P16) and P19Arf (P19)] is induced by the SnPP preconditioning protocol. To this end, 12 male CD-1 mice received a tail vein injection of 1 µmol of SnPP (0.75 mg/mouse, in 150 µL of 0.1 N NaOH/saline; Frontier Scientific, Logan, UT, USA) [27]. In addition, 12 vehicle-injected mice served as controls. At either 4 or 18 h post-injection (six mice per group at each time point), the mice were deeply anesthetized with pentobarbital and the kidneys were resected through a midline abdominal incision. The kidneys were iced, and the renal cortices were dissected and extracted for total mRNA (RNeasy mini+; Qiagen, Germantown, MD, USA) and protein [6]. The 4-h samples were assayed for P21, P16 and P19 mRNAs by reverse transcription polymerase chain reaction (RT-PCR) using the primer pairs presented in Table 1. Results were factored by simultaneously determined glyceraldehyde 3-phosphate dehydrogenase (GAPDH) product, used as a housekeeping gene. The 18-h protein samples (i.e. the time at which cytoresistance is expressed) were also probed for P21 protein by western blotting (Abcam #AB109199; Cambridge, MA, USA). As a positive control for P21 protein upregulation, renal cortical protein samples obtained 1, 3 and 7 days post-induction of glycerol-induced rhabdomyolysis (50% glycerol; 9 mL/kg, injected intramuscular in equally divided doses in the hind limbs; n, three at each time point) were also probed for P21 by western blotting.
SnPP ± FeS preconditioning: because coadministration of FeS and SnPP has been demonstrated to confer greater cytoresistance than SnPP alone [7], the impact of combined SnPP + FeS on P21 (as well as on P16 and P19) expression was assessed. Ten mice were injected via the tail vein with SnPP (1 μmol) + 1 mg of FeS (in the form of Venofer; American Regent, Shirley, NY, USA; [7]). The SnPP is added to FeS stock. An equal number of vehicle-injected mice served as controls. At either 4 or 18 h post-injection (n, five per group at each time point), kidney tissues were harvested and probed for P21, P19 and P16 mRNAs, as noted above. The 18-h post-injection samples also underwent P21 western blotting.
SnPP ± nitrite myoglobin (N-Mgb) preconditioning: in this OIP protocol, N-Mgb is substituted for FeS as the source of catalytic Fe, as previously described [6]. To determine the impact of this oxidant preconditioning regimen on P21, P16 and P19 expression, 10 mice were injected via the tail vein with 1 µmol of SnPP + 1 mg of N-Mgb [6]. Vehicle-injected mice served as controls. At either 4 or 18 h post-injection, renal cortical mRNA and protein samples were prepared and assayed for P21, P16 and P19 mRNAs. The 18-h samples were also probed for P21 by western blotting.
Impact of IPC on P21 expression
To confirm previous observations that IPC upregulates P21 (e.g. [24]), 24 mice were subjected to either 0, 5, 10, 15 or 25 min of unilateral (left) renal ischemia (n, three each, obtained using the protocol detailed immediately below). At 18 h post-ischemia, the kidneys were resected and cortical RNA extracts were probed for P21 mRNA. Because P53 is the dominant P21 inducer, its mRNA levels were also assessed (primers shown in Table 1). For comparison, P21 and P53 mRNAs were determined 18 h post-SnPP/FeS injection. Values were compared with those observed in normal kidney samples.
Mouse primers for RT-PCR
| mRNA | Primer sequences | Product size |
|---|---|---|
| P21 | 5′-TAT TTA AGC CCC TCC CAA CC-3′ | 218 bp |
| 5′-AGC TGG CCT TAG AGG TGA CA-3′ | ||
| P16ink4a | 5′-TCG CAC GAT GTC TTG ATG TC-3′ | 394 bp |
| 5′-GAA AGC GAA CTC GAG AG-3′ | ||
| Nested P16ink4a | 5′-CGA ATC TGC ACC GTA GTT GA-3′ | 194 bp |
| 5′-ATC TGG AGC ATG GAG TC-3′ | ||
| P19Arf | 5′-CTT GGT CAC TGT GAG GAT TCA-3′ | 225 bp |
| 5′-CAC CGT AGT TGA GCA GAA GAG-3′ | ||
| P53 | 5′-CCA GAA GAT ATC CTG CCA TCA C-3′ | 728 bp |
| 5′-GCA CAA ACA CGA ACC TCA AAG-3′ | ||
| NQO1 | 5′-GAG GTA CTC GAA TCT GAC CTC TA-3′ | 254 bp |
| 5′-ACT CTC TCA AAC CAG CCT TTC-3′ | ||
| GCLC | 5′-AGG CTC TCT GCA CCA TCA CT-3′ | 203 bp |
| 5′-TGG CAC ATT GAT GAC AAC CT-3′ | ||
| SRXN1 | 5′-TCA CCA TTG TGG CAA CTC TT-3′ | 262 bp |
| 5′-GTG CCC ACA GAG CCT AGA AG-3′ | ||
| GAPDH | 5′-CTG CCA TTT GCA GTG GCA AAG TGG-3′ | 437 bp |
| 5′-TTG TCA TGG ATG ACC TTG GCC AGG-3′ |
| mRNA | Primer sequences | Product size |
|---|---|---|
| P21 | 5′-TAT TTA AGC CCC TCC CAA CC-3′ | 218 bp |
| 5′-AGC TGG CCT TAG AGG TGA CA-3′ | ||
| P16ink4a | 5′-TCG CAC GAT GTC TTG ATG TC-3′ | 394 bp |
| 5′-GAA AGC GAA CTC GAG AG-3′ | ||
| Nested P16ink4a | 5′-CGA ATC TGC ACC GTA GTT GA-3′ | 194 bp |
| 5′-ATC TGG AGC ATG GAG TC-3′ | ||
| P19Arf | 5′-CTT GGT CAC TGT GAG GAT TCA-3′ | 225 bp |
| 5′-CAC CGT AGT TGA GCA GAA GAG-3′ | ||
| P53 | 5′-CCA GAA GAT ATC CTG CCA TCA C-3′ | 728 bp |
| 5′-GCA CAA ACA CGA ACC TCA AAG-3′ | ||
| NQO1 | 5′-GAG GTA CTC GAA TCT GAC CTC TA-3′ | 254 bp |
| 5′-ACT CTC TCA AAC CAG CCT TTC-3′ | ||
| GCLC | 5′-AGG CTC TCT GCA CCA TCA CT-3′ | 203 bp |
| 5′-TGG CAC ATT GAT GAC AAC CT-3′ | ||
| SRXN1 | 5′-TCA CCA TTG TGG CAA CTC TT-3′ | 262 bp |
| 5′-GTG CCC ACA GAG CCT AGA AG-3′ | ||
| GAPDH | 5′-CTG CCA TTT GCA GTG GCA AAG TGG-3′ | 437 bp |
| 5′-TTG TCA TGG ATG ACC TTG GCC AGG-3′ |
Primers used for assessment of gene expression, utilizing RT-PCR. GAPDH was used as a housekeeping gene.
Mouse primers for RT-PCR
| mRNA | Primer sequences | Product size |
|---|---|---|
| P21 | 5′-TAT TTA AGC CCC TCC CAA CC-3′ | 218 bp |
| 5′-AGC TGG CCT TAG AGG TGA CA-3′ | ||
| P16ink4a | 5′-TCG CAC GAT GTC TTG ATG TC-3′ | 394 bp |
| 5′-GAA AGC GAA CTC GAG AG-3′ | ||
| Nested P16ink4a | 5′-CGA ATC TGC ACC GTA GTT GA-3′ | 194 bp |
| 5′-ATC TGG AGC ATG GAG TC-3′ | ||
| P19Arf | 5′-CTT GGT CAC TGT GAG GAT TCA-3′ | 225 bp |
| 5′-CAC CGT AGT TGA GCA GAA GAG-3′ | ||
| P53 | 5′-CCA GAA GAT ATC CTG CCA TCA C-3′ | 728 bp |
| 5′-GCA CAA ACA CGA ACC TCA AAG-3′ | ||
| NQO1 | 5′-GAG GTA CTC GAA TCT GAC CTC TA-3′ | 254 bp |
| 5′-ACT CTC TCA AAC CAG CCT TTC-3′ | ||
| GCLC | 5′-AGG CTC TCT GCA CCA TCA CT-3′ | 203 bp |
| 5′-TGG CAC ATT GAT GAC AAC CT-3′ | ||
| SRXN1 | 5′-TCA CCA TTG TGG CAA CTC TT-3′ | 262 bp |
| 5′-GTG CCC ACA GAG CCT AGA AG-3′ | ||
| GAPDH | 5′-CTG CCA TTT GCA GTG GCA AAG TGG-3′ | 437 bp |
| 5′-TTG TCA TGG ATG ACC TTG GCC AGG-3′ |
| mRNA | Primer sequences | Product size |
|---|---|---|
| P21 | 5′-TAT TTA AGC CCC TCC CAA CC-3′ | 218 bp |
| 5′-AGC TGG CCT TAG AGG TGA CA-3′ | ||
| P16ink4a | 5′-TCG CAC GAT GTC TTG ATG TC-3′ | 394 bp |
| 5′-GAA AGC GAA CTC GAG AG-3′ | ||
| Nested P16ink4a | 5′-CGA ATC TGC ACC GTA GTT GA-3′ | 194 bp |
| 5′-ATC TGG AGC ATG GAG TC-3′ | ||
| P19Arf | 5′-CTT GGT CAC TGT GAG GAT TCA-3′ | 225 bp |
| 5′-CAC CGT AGT TGA GCA GAA GAG-3′ | ||
| P53 | 5′-CCA GAA GAT ATC CTG CCA TCA C-3′ | 728 bp |
| 5′-GCA CAA ACA CGA ACC TCA AAG-3′ | ||
| NQO1 | 5′-GAG GTA CTC GAA TCT GAC CTC TA-3′ | 254 bp |
| 5′-ACT CTC TCA AAC CAG CCT TTC-3′ | ||
| GCLC | 5′-AGG CTC TCT GCA CCA TCA CT-3′ | 203 bp |
| 5′-TGG CAC ATT GAT GAC AAC CT-3′ | ||
| SRXN1 | 5′-TCA CCA TTG TGG CAA CTC TT-3′ | 262 bp |
| 5′-GTG CCC ACA GAG CCT AGA AG-3′ | ||
| GAPDH | 5′-CTG CCA TTT GCA GTG GCA AAG TGG-3′ | 437 bp |
| 5′-TTG TCA TGG ATG ACC TTG GCC AGG-3′ |
Primers used for assessment of gene expression, utilizing RT-PCR. GAPDH was used as a housekeeping gene.
Delayed (2 weeks) post-IPC effects on P21 expression and senescence
P21 mRNA
To ascertain whether IPC might have an ongoing impact on P21 expression, the above IPC experiment was repeated except that P21 mRNA levels were assessed 2 weeks post-ischemia induction (n, three each for the 5, 10 and 15 min IPC groups). The delayed impact of IPC on the senescence inducer (and biomarker) P16 was also assessed (mRNA; using primers in Table 1). Results were compared with those from normal kidneys and kidneys obtained 2 weeks post-SnPP + FeS injection (n, three each).
Loss of renal mass
To assess whether structural renal injury resulted by 2 weeks post-IPC, a potential loss of renal mass (i.e. reduction in post-ischemic kidney weights) [35, 36] was determined at the time of kidney harvesting. Results were contrasted to those observed in normal mice and mice 2 weeks following SnPP + FeS injection (n, three each).
Senescence-associated beta-galactosidase localization
Freshly prepared cross-sections of the above kidneys were flash frozen in liquid nitrogen and 5-μm sections were cut and placed on glass slides. The slides were then stained for the senescence marker senescence-associated beta-galactosidase (SA-β-gal) using a commercially available kit (Biotium, Fremont, CA, USA; cat. #30031) as per manufacturer’s instructions. As a positive control, kidney sections from two aged male CD-1 mice (24 and 28 weeks of age) were also probed for SA-β-gal localization.
Induction of severe ischemic AKI ± OIP—impact on P21, P16, P19 expression and AKI severity
The following experiment assessed whether previously noted OIP-induced protection against ischemic AKI [6, 7, 27] is associated with altered P21 expression. To this end, 10 mice were deeply anesthetized, the abdominal cavities were opened and then bilateral renal ischemia × 25 min was induced by occluding both renal pedicles with microvascular clamps. Half of the mice had been preconditioned with SnPP 18 h prior to ischemia induction. Eighteen hours post-ischemia, the mice were re-anesthetized, the kidneys were removed and renal cortical protein and RNA were extracted. The severity of kidney injury was determined by BUN and plasma creatinine levels. The RNA cortical extracts were probed P21, P19 and P16 mRNAs. P21 levels were assessed by western blotting. A potential correlation between P21 mRNA and BUN levels (r value) was assessed. The results from SnPP preconditioned mice were compared with those observed in non-preconditioned post-ischemic mice and with values found in normal mice/kidney samples. The P21 western blot signals were evaluated by densitometry, thereby providing numerical values for Student’s t-testing.
Assessment of whether OIP induces growth arrest
To determine whether OIP inhibits tubular cell proliferation (as previously documented for IPC; e.g. [12, 20, 24]), mice were subjected to either SnPP + FeS, SnPP + N-Mgb or vehicle injection, as noted above (n, three each). Eighteen hours later, the kidneys were removed and fixed in 10% formalin, and embedded in paraffin. Four-micron sections were cut and probed for phosphohistone H3 (PHH3), a specific marker of cells within the S phase of replication [37]. Additional sections were stained for KI-67, which, unlike PHH3, is expressed during G1, S, G2 and mitosis, but not in quiescent cells (G0) [38]. These evaluations were conducted by the Shared Resource Immunohistopathology Laboratory at the authors’ institution. In brief, the cut kidney sections were stained with ‘Leica Bond Rx’ (Leica Biosystems, Buffalo Grove, IL, USA). Slides undergoing PHH3 analyses were pretreated with H2 antigen retrieval buffer for 20 min. Endogenous peroxidase was blocked with 3% hydrogen peroxide for 5 min. A TCT protein block was applied for 10 min (0.05 M Tris, 0.15 M NaCl, 0.25% Casein, 0.1% Tween-20, pH 7.6). Anti-PHH3 (Millipore, Burlington, MA, USA; catalog# 06-570) was used at 1:250 dilution × 60 min. Anti Ki67 antibody (Cell Signaling, Danvers, MA, USA; catalog # 12202) was used at 1:500 dilution and was applied to the tissue × 60 min. Both antibodies were then detected using Leica Power Vision HRP rabbit specific polymer (Leica Biosystems; catalog# PV6119) × 12 min. Nuclear-specific staining was visualized with ‘Refine DAB’ (Leica Biosystems; catalog# D59800). A hematoxylin counterstain was then applied. Isotype-negative control antibody exposed slides were included as controls.
The total sections were scanned (HaloTM) and renal cortical + outer medullary regions were outlined on computer-generated images (see the ‘Results’ section). Total numbers of tubular nuclei and of PHH3-positive or KI-67-positive tubular nuclei were assessed. Results were expressed as percentage of positive PHH3 or KI-67 nuclei (per total numbers of nuclei in the scanned tissue sections).
Determination of whether ischemia–reperfusion injury leads to delayed (2 weeks) cellular senescence and whether OIP mitigates this response
Given that P21, P16 and P19 have each been implicated in inducing cellular senescence, we assessed whether the latter emerges in a model of post-ischemic progressive renal disease [35, 36]. To this end, five mice were subjected to 25 min of unilateral (left) renal ischemia without or with prior oxidant preconditioning (SnPP + N-Mgb). The mice were then sutured and allowed to recover from anesthesia. Two weeks post-surgery, they were re-anesthetized and both kidneys were resected. Renal cortical mRNA and protein samples were extracted and assayed for P21, P16 and P19 mRNAs and P21 protein (western blotting). SA-β-gal was determined by tissue enzymatic activity [39–41], as assessed with a commercially available SA-β-gal assay (Biovision; #K821-100; Milpitas, CA, USA). The results found in post-ischemic kidneys, contralateral (CL) uninjured kidneys and kidneys from normal mice were compared (n, five each). Because the loss of renal tubular cell mass post-AKI is reflected by parallel reductions in the amount of tissue lactate dehydrogenase (LDH) [42], LDH enzyme activity (units/mg renal tissue) in the above kidneys was compared with each other and with values in normal kidneys. Lastly, renal injury was assessed by NGAL mRNA expression by RT-PCR [42] in preconditioned versus non-preconditioned post-ischemic versus normal kidneys [42].
Assessment of whether canonical Nrf2 genes are activated by SnPP/Fe preconditioning
As an alternative to P21 induction as a mediator of OIP-induced cytoresistance, activation of the Nrf2 cytoprotective pathway was assessed [27]. To this end, Nrf2+/+ and Nrf2−/− mice were injected with SnPP/FeS, as noted above (three mice per group). Four hours later, the mice were deeply anesthetized, the kidneys were resected and renal cortical RNA was isolated as noted above. The impact of SnPP/FeS on the mRNAs of three canonical Nrf2-regulated genes (NQO1, GCLC and SRXN1) was assessed by RT-PCR using the primers listed in Table 1. The results were factored by simultaneously determined GAPDH product.
P21, P16 and P19 expression in Nrf2−/− versus Nrf2+/+ mice
To assess whether Nrf2 knockout alters P21, P16 or P19 expression (potentially predisposing to senescence), their mRNAs were evaluated in Nrf2−/− and Nrf2+/+ mice under baseline conditions (n, three each). To determine whether Nrf2−/− leads to altered P21, P16 or P19 responses to FeS/SnPP injection, Nrf2−/− and Nrf2+/+ mice were injected with FeS/SnPP and renal cortical RNA extracts were prepared 4 or 18 h later.
Assessment of whether Nrf2−/− mice are resistant to FeS/SnPP-induced protection against AKI
Nrf2−/− mice were treated with FeS/SnPP or received vehicle injection (n, seven; five males, two females in each group). The following day they were subjected to 18 min of bilateral ischemic injury. The severity of AKI was assessed 18 h later (BUN, creatinine levels).
Calculations and statistics
All values are presented as mean ± SEMs. All RT-PCR mRNA results were factored by simultaneously obtained GAPDH product, used as a housekeeping gene. Statistical comparisons were performed by unpaired or paired Student’s t-testing, as noted below. Statistical significance was judged by P < 0.05. If multiple comparisons were made the Bonferroni correction was applied.
RESULTS
Evaluation of SnPP preconditioning effects on P21 expression
To evaluate the impact of OIP strategies on P21, P16 and P19 expression, the independent effect of SnPP was first assessed. As shown in Figure 1, SnPP had no discernible effect on P21 mRNA levels, assessed at either 4 or 18 h post-SnPP injection. Furthermore, no changes in either P16 or P19 mRNA levels were observed. Western blotting confirmed a lack of a SnPP effect on P21 protein levels, compared with control tissue samples (Figure 1), given that virtually no P21 band was observed in either group. In contrast, a time-dependent stepwise increase in renal P21 protein expression was observed in kidneys harvested from mice at 1, 3 and 7 days post-induction of glycerol-induced AKI (Figure 1), thereby serving as a positive P21 control. Post-glycerol BUN concentrations that corresponded with these progressive P21 increases were 106 ± 5, 50 ± 8 and 61 ± 4 mg/dL (at 0, 1, 3 and 7 days post-glycerol injection; controls, 28 ± 3 mg/dL).
SnPP-mediated oxidant preconditioning: assessment of P21, P16 and P19 expression. At either 4 or 18 h post-SnPP injection, the mRNAs for P21, P16 and P19 levels were assessed and no SnPP-mediated changes were observed (C = control, normal, kidneys). Furthermore, no increases in P21 protein levels were observed at 18 h post-SnPP injection (western blot: ‘Rx’) Kidney samples obtained from mice from 1, 3 and 7 days post-glycerol-induced AKI showed progressive P21 protein increases, serving as a positive P21 western blot control.
Evaluation of SnPP + FeS preconditioning effects on P21 expression
As shown in Figure 2, the addition of FeS to the SnPP preconditioning protocol did not alter the results observed with SnPP alone, with no upregulation of renal cortical P21, P16 or P19 mRNA, or P21 protein levels being observed.
SnPP + FeS-mediated oxidant preconditioning: assessment of P21, P16 and P19 expression. As with SnPP alone, the combination of SnPP + FeS did not alter P21, P16 or P19 mRNA levels at either 4 or 18 h following their injection. They also failed to induce a discernible increase in P21 protein levels, as assessed by western blot.
Evaluation of SnPP + N-Mgb preconditioning effects on P21 expression
By 4 h post-SnPP + N-Mgb injection, an ∼50% increase in P21 mRNA expression was observed (Figure 3). However, this was a transient finding, given that the P21 mRNA values returned to control levels by 18 h post-SnPP + N-Mgb injection (i.e. the time at which cytoresistance is observed) [6, 7]. Furthermore, there was no increase in P21 protein at 18 h post-SnPP + N-Mgb injection. P16 mRNA and P19 mRNA values remained at control levels at both the 4 and18 h post-agent administration (Figure 3).
SnPP + nitrite-myoglobin-mediated oxidant preconditioning: P21, P16 and P19 expression. At 4 h post-SnPP-nitrite myoglobin (N-Mgb) injection, a modest increase in P21 mRNA was observed. However, this normalized by 18 h, and no 18-h increase in P21 protein was observed. Neither P16 nor P21 mRNA responded to N-Mgb injection.
IPC effects on renal P21 expression
Unlike the above OIP strategies, ‘IPC’ evoked stepwise P21 increases (up to 25-fold), as assessed by its mRNA at 18 h of reflow. This was apparent even with a 5 min of IPC protocol (∼5-fold mRNA increases). Of note was that IPC also upregulated P53, the dominant P21 regulator (Figure 4, right). Conversely, neither P21 nor P53 mRNA elevations were increased by the above-noted oxidant preconditioning protocols (e.g. FeS/SnPP).
Increasing degrees of IPC induce stepwise increases in P21 and P53 mRNAs. Even a 5-min unilateral ischemic insult was sufficient to induce a 5-fold increase in P21 mRNA, assessed 18 h later. This was not a maximal response, given that progressive increases in ischemia times caused further P21 mRNA elevations. The P21 inducer, P53, showed parallel increases in its mRNA. Conversely, OIP did not alter either P21 or P53 mRNA.
Induction of severe ischemic AKI ± OIP: impact on P21, P16 and P19 expression and AKI severity
By 18 h post-induction of 25 min of bilateral renal ischemia in control mice, marked increases in both P21 and P16 mRNAs were observed (Figure 5), whereas P19 levels remained at control levels. Corresponding with the P21 mRNA increases was an approximate 25-fold increase in P21 protein levels, as determined by western blotting (Figure 6). The ischemia protocol induced severe renal failure, as denoted by marked increases in both BUN and plasma creatinine concentrations (Figure 7).
Bilateral 25-min-induced ischemic AKI causes marked P21 mRNA increases that are attenuated by SnPP preconditioning. Twenty-five minutes of bilateral renal ischemia–reperfusion (I/R) markedly increased P21 mRNA, as assessed 18 h later. Administration of SnPP 18 h prior to inducing renal ischemia approximately halved this post-ischemic P21 mRNA increase. Although the 25-min ischemic insult also raised P16 mRNA, it was not impacted by prior SnPP administration. No P19 induction was observed.
Post-ischemic P21 protein increases are markedly attenuated by SnPP preconditioning. The day following induction of 25 min of ischemia–reperfusion (I/R) injury, a 25-fold increase in renal cortical P21 protein levels were observed. This was not due to surgical stress, given that sham surgery did not alter P21 levels. Preconditioning with SnPP markedly reduced the post-ischemic P21 increases.
SnPP protects against ischemic AKI, and the degree of protection inversely correlates with P21 mRNA levels. By 18 h post-25 min of bilateral ischemia, severe renal failure resulted, as denoted by BUN and plasma creatinine increases. SnPP preconditioning markedly reduced the severity of AKI, with the plasma creatinine levels post-ischemia–reperfusion (I/R) + SnPP preconditioning not differing from normal (control; C) values. As shown in the right-hand panel, the severity of AKI, as denoted by BUN levels, directly correlated with P21 mRNA levels.
Preconditioning with SnPP conferred marked protection against ischemic AKI, with BUN and creatinine concentrations remaining at, or near, normal levels (Figure 7). Associated with this SnPP-mediated protection were marked reductions in post-ischemic P21 mRNA and protein levels (Figures 5 and 6). As shown in the right-hand panel of Figure 7, the post-ischemic BUNs in these mice strongly correlated (r = 0.73; P < 0.01) with each mouse’s P21 mRNA level. Of note, in previously published data [7], it has been demonstrated that this oxidant-induced functional protection is strongly associated with concomitant, significant reductions in renal histologic injury, as denoted by marked decreases in the extent of proximal tubule necrosis (in cortex and most notably the outer medullary stripe) and decreased cast formation within distal tubule lumina.
KI-67 and PHH3 cell-cycle markers following treatment with OIP strategies
A representative outline of renal cortical/outer medullary areas that were used for computer scanning is depicted at the top left panel of Figure 8. Examples of KI-67-positive staining nuclei are depicted by arrows (bottom left-hand panel). Approximately 100 000 nuclei per outlined renal cortical/outer medullary area of each kidney were counted by the Halo program. In normal kidney, ∼1% of nuclei were KI-67 positive. No evidence of cell-cycle suppression, as assessed by KI-67 staining, was observed. Rather, % KI-67 positivity was slightly increased with the oxidant preconditioning strategies. This stands in marked contrast to IPC results where suppressed KI-67 nuclear staining is observed (e.g. [24]).
KI-67 staining of proximal tubule nuclei following either FeS + SnPP or N-Mgb + SnPP-induced oxidant preconditioning. The top left panel demonstrates the cortical–outer medullary kidney regions that were scanned for proximal tubule nuclear staining. The bottom left panel illustrates KI-67-positive nuclei (see arrows). The right-hand panel presents the percentage of KI-67 positive nuclei compared with control, C, kidney values (∼1%). FeS + SnPP slightly raised % KI-67 positivity, which was not statistically different from that seen in controls. However, N-Mgb + SnPP did cause a statistically significant increase in KI-67 staining, affecting ∼3% of cells.
As shown in Figure 9, these KI-67 results were recapitulated with the S phase marker PHH3: significantly increased, rather than decreased, nuclear staining was observed. Mitotic figures are depicted by arrows in Figure 9, upper and lower left-hand panels.
PHH3 staining of proximal tubule nuclei following either FeS + SnPP or N-Mgb + SnPP-induced preconditioning. In control kidneys, ∼0.2% of proximal tubule nuclei demonstrated PHH3 positivity, as illustrated by mitotic figures (arrows) in the upper left and lower left-hand panels. The degree of positivity was increased by ∼2× and ∼3× for FeS/SnPP and N-Mgb preconditioning treatments, respectively.
Delayed post-ischemia–reperfusion cellular senescence: impact of prior oxidant preconditioning
P21, P16 and P19 mRNA levels
Given that P21, P16 and P19 are widely accepted mediators, as well as biomarkers, of cellular senescence, their mRNA levels were assessed 2 weeks post-induction of unilateral ischemic injury. Marked increases in each were observed in post-ischemic kidneys (Figure 10), compared with values seen in either uninjured CL kidneys or normal controls. The 2-week post-ischemic P21 mRNA increases were paralleled by increases in P21 protein levels (Figure 10). Each of these post-ischemic mRNA and P21 protein increases were markedly attenuated by prior OIP (Figure 10).
![Persistent upregulation of P21, P16 and P19 expression 2 weeks post-unilateral ischemic–reperfusion (I/R) injury. Unilateral ischemia (25 min) is well known to cause progressive renal damage [35, 36], and as per Figure 11. As illustrated here, this is associated with persistent up-regulation of P21, P19, and P16 mRNAs [assessed in control, post-ischemic kidneys (I/R), and uninjured CL uninjured kidneys]. Each of these changes was abrogated by prior OIP.](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/ndt/33/11/10.1093_ndt_gfy029/1/m_gfy029f10.jpeg?Expires=1747852214&Signature=dt2xXnxswaDmynMc4YP5KJ47gML4SiRcfx38ZD~6mUk48p5NZkACrErt697YfG8N1WeDU8jCz~R-FT0IakQOlklY67QfJEgMeEFi9byXoMFKzj~te80Rq2Qf4yFlY~-r0CNa-L~yW7sGYpffNe9JGSuAqp64JhQnxGvrCnYpi0b7jRPbBLc5Rje7bK~ikhklJOB86M7nnS9ensEIj9rb~R6H0FxfJJiSvSx7pQf5QXH9dYL67HFY0QNCrgNGqiQxrYqwpzz2ybKblOgEHzZtQlR63RyhpsTepAAnI2gXW0NYD5L5abWVSd0sBUl-wFPAVnP4-EdJlfPRIfI6-Cs~5w__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Persistent upregulation of P21, P16 and P19 expression 2 weeks post-unilateral ischemic–reperfusion (I/R) injury. Unilateral ischemia (25 min) is well known to cause progressive renal damage [35, 36], and as per Figure 11. As illustrated here, this is associated with persistent up-regulation of P21, P19, and P16 mRNAs [assessed in control, post-ischemic kidneys (I/R), and uninjured CL uninjured kidneys]. Each of these changes was abrogated by prior OIP.
SA-β-gal levels
As shown in Figure 11, left, assay of renal cortical tissues obtained 2-week post-ischemic injury demonstrated a doubling of the senescence marker SA-β-gal (versus values in either normal or uninjured CL kidneys), further supporting the existence of cellular senescence post-AKI. OIP almost completely blocked this increase in SA-β-gal levels.
Post-ischemic SA-β-gal expression and associated loss of renal mass at 2 weeks post-ischemic damage with and without prior OIP. Corresponding with the changes observed in Figure 10, a significant increase in SA-β-gal was observed at 2 weeks post-unilateral ischemic injury. The SA-β-gal increases were almost completely prevented by OIP preconditioning (left panel). Associated with this SA-β-gal reduction was a marked preservation in post-ischemic renal mass (45% versus 12% loss of renal mass without and with oxidant preconditioning) (middle panel). Further confirming OIP-induced protection was a marked reduction in NGAL mRNA values in OIP preconditioned post-ischemic kidneys (factored by GAPDH) (right-hand panel).
Post-ischemic loss of renal mass
At 2 weeks post 25 min of unilateral ischemia, a 45% reduction in post-ischemic renal mass (renal weight) was observed (Figure 11, middle panel). These renal mass reductions were largely blocked by OIP. This preservation was underscored by the observation that the OIP-mediated protection was associated with 10 ± 3% reduction in tissue LDH content (units/mg of residual tissue versus control tissues). Conversely, a 43 ± 3% LDH reduction was observed in unconditioned post-ischemic kidneys (P < 0.01) (not depicted). As previously noted [42], LDH loss in post-injured kidney corresponds with the severity of AKI/tissue damage. Preconditioning was also associated with an approximate 3-fold reduction in NGAL mRNA levels (Figure 11, right-hand panel), further evidence of decreased tissue damage.
In sum, OIP markedly attenuated 2 weeks post-ischemic P21, P19, P16 and SA-β-gal increases and preserved post-ischemic renal mass (weight, LDH content) and reduced NGAL mRNA levels.
IPC is associated with sustained P21 increases and corresponding reductions in renal mass
Given the above results demonstrating that 25 min of renal ischemia induces sustained P21 increases, evidence of senescence and corresponding reductions in renal mass, we questioned whether similar, albeit more modest, changes could result from mild ischemic injury, such as used in IPC protocols. This appeared to be the case, given that both 5 and 15 min of unilateral ischemia each induced sustained and comparable P21, as well as P16, mRNA increases. These changes corresponded correlated with modest, but significant, 5–10% reductions in renal mass (P < 0.01), as assessed at the 2-week time point (Figure 12, right). FeS + SnPP preconditioning did not affect the above injury markers (not depicted).
Unilateral IPC is associated with persistent increases in P21 and P16 expression and reductions in renal mass. Even a 5-min IPC episode was associated with persistent P21 and P16 mRNA increases, assessed 2 weeks later (*P < 0.01 versus no preconditioning ‘0’). The P21 and P16 increases were associated with corresponding reductions in post-ischemic kidney mass (expressed as a percentage reduction in renal weights compared with the weights of normal mouse kidneys).
SA-β-gal staining
As shown in Figure 13, kidneys from young (∼8-week-old) mice showed no evidence of SA-β-gal tissue staining (Figure 13A). Conversely, widespread but patchy SA-β-gal staining was observed in kidney sections from aged (∼2-year-old) mice (panel B), consistent with age-associated senescence.
Histochemical localization of SA-β-gal in young (A), aged (B) and young post-ischemic preconditioned kidneys (C, D). (A) This panel demonstrates an absence of SA-β-gal staining in a normal kidney from an 8-week-old mouse. In contrast, a normal aged kidney (B; 28 weeks) demonstrates widely scattered foci of SA-β-gal activity throughout renal cortex. Two weeks post-5 min of renal ischemia imposed on an 8-week-old mouse was associated with scattered foci of SA-β-gal activity. As shown in (D), this staining was localized to proximal tubules.
By 2 weeks post-5 min of IPC, sporadic tubular segments manifested tissue SA-β-gal positivity (Figure 13C). As shown in Figure 13D, this staining could be localized to proximal tubule cells. Thus, even 5 min of IPC resulted in P16, P19 mRNA and SA-β-gal increases, consistent with a mild post-ischemic senescent state.
P21, P16 and P19 expression in Nrf2−/− versus Nrf2+/+ mice
No significant differences in P21 or P19 expression were observed in Nrf2−/− versus Nrf2+/+ mice under control conditions or either 4 or 18 h post-oxidant preconditioning (with SnPP + FeS) (n, three each). Only very weak and non-quantifiable P16 signals were observed in any of these six black mice, whether Nrf2 positive or negative.
Nrf2−/− mouse responses to SnPP/FeS preconditioning
Nrf2+/+ mice showed marked increases in NQO1, GCLC and SRXN1 mRNAs at 4 h post-SnPP/FeS administration. (Table 2) In contrast, these changes were essentially absent in Nrf2−/− mice (Figure 14). Furthermore, the Nrf2−/− failed to demonstrate OIP-mediated protection against 18 min of bilateral ischemic injury (BUNs: 58 ± 10 versus 48 ± 5 mg/dL, NS; plasma creatinines: 0.38 ± 0.03 versus 0.34 ± 0.04 mg/dL, NS; control versus FeS/SnPP preconditioned mice, respectively). Control BUN and creatinine values for Nrf2−/− mouse litter mates were 24 ± 2 mg/dL and 0.18 ± 0.02 mg/dL, respectively.
P21, P16 and P19 mRNA expression in Nrf2−/− and Nrf2+/+ mice
| Assessed mRNA | Nrf2+/+ | Nrf2−/− | 4-h OIP Nrf2+/+ | 4-h OIP Nrf2+/+ | 18-h OIP Nrf2+/+ | 18-h OIP Nrf2−/− |
|---|---|---|---|---|---|---|
| P21 | 0.84 ± 0.06 | 1.01 ± 0.12 | 1.3 ± 0.09 | 1.13 ± 0.02 | 0.97 ± 0.03 | 1.14 ± 0.17 |
| NS | NS | NS | ||||
| P19 | 0.01 ± 0.01 | 0.02 ± 0.02 | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.02 ± 02 | 0.020.02 |
| NS | NS | NS | ||||
| P16 | <0.1 | <0.1 | <0.1 | <0.1 | <0.1 | <0.1 |
| Assessed mRNA | Nrf2+/+ | Nrf2−/− | 4-h OIP Nrf2+/+ | 4-h OIP Nrf2+/+ | 18-h OIP Nrf2+/+ | 18-h OIP Nrf2−/− |
|---|---|---|---|---|---|---|
| P21 | 0.84 ± 0.06 | 1.01 ± 0.12 | 1.3 ± 0.09 | 1.13 ± 0.02 | 0.97 ± 0.03 | 1.14 ± 0.17 |
| NS | NS | NS | ||||
| P19 | 0.01 ± 0.01 | 0.02 ± 0.02 | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.02 ± 02 | 0.020.02 |
| NS | NS | NS | ||||
| P16 | <0.1 | <0.1 | <0.1 | <0.1 | <0.1 | <0.1 |
P21, P16 and P19 mRNA expression in Nrf2−/− and Nrf2+/+ mice under control conditions or either 4- or 18-h post-oxidant preconditioning (with SnPP + FeS) (n = 3–4 per group). No significant differences in P21 or P19 mRNAs were observed. Of note, only weak, nonquantifiable, P16 signals were observed either under baseline conditions or with FeS/SnPP in either Nrf2+/+ or Nrf2−/− mice. NS, not significant.
P21, P16 and P19 mRNA expression in Nrf2−/− and Nrf2+/+ mice
| Assessed mRNA | Nrf2+/+ | Nrf2−/− | 4-h OIP Nrf2+/+ | 4-h OIP Nrf2+/+ | 18-h OIP Nrf2+/+ | 18-h OIP Nrf2−/− |
|---|---|---|---|---|---|---|
| P21 | 0.84 ± 0.06 | 1.01 ± 0.12 | 1.3 ± 0.09 | 1.13 ± 0.02 | 0.97 ± 0.03 | 1.14 ± 0.17 |
| NS | NS | NS | ||||
| P19 | 0.01 ± 0.01 | 0.02 ± 0.02 | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.02 ± 02 | 0.020.02 |
| NS | NS | NS | ||||
| P16 | <0.1 | <0.1 | <0.1 | <0.1 | <0.1 | <0.1 |
| Assessed mRNA | Nrf2+/+ | Nrf2−/− | 4-h OIP Nrf2+/+ | 4-h OIP Nrf2+/+ | 18-h OIP Nrf2+/+ | 18-h OIP Nrf2−/− |
|---|---|---|---|---|---|---|
| P21 | 0.84 ± 0.06 | 1.01 ± 0.12 | 1.3 ± 0.09 | 1.13 ± 0.02 | 0.97 ± 0.03 | 1.14 ± 0.17 |
| NS | NS | NS | ||||
| P19 | 0.01 ± 0.01 | 0.02 ± 0.02 | 0.01 ± 0.01 | 0.01 ± 0.01 | 0.02 ± 02 | 0.020.02 |
| NS | NS | NS | ||||
| P16 | <0.1 | <0.1 | <0.1 | <0.1 | <0.1 | <0.1 |
P21, P16 and P19 mRNA expression in Nrf2−/− and Nrf2+/+ mice under control conditions or either 4- or 18-h post-oxidant preconditioning (with SnPP + FeS) (n = 3–4 per group). No significant differences in P21 or P19 mRNAs were observed. Of note, only weak, nonquantifiable, P16 signals were observed either under baseline conditions or with FeS/SnPP in either Nrf2+/+ or Nrf2−/− mice. NS, not significant.
Nrf2+/+ mice, but not Nrf2−/− mice, demonstrate significant up-regulation of canonical Nrf2-regulated genes. NQO1, GCLC and SRXN1 mRNAs were assessed 4 h post-administration of SnPP + FeS (OIP). Nrf2+/+ mice responded with an increase in all three mRNAs. Conversely, no significant responses were seen in Nrf2−/− mice. These findings clearly indicate that SnPP + FeS activate the Nrf2 pathway.
DISCUSSION
We have recently advanced the concept that the combined administration of an HO-1 inhibitor (SnPP) + a source of catalytic Fe (either FeS or N-Mgb) can induce profound renal preconditioning, and thereby confer marked protection against diverse forms of ARF [6, 7, 27]. The hypothesized mechanism has been as follows: (i) renal delivery of SnPP + catalytic Fe induces transient proximal tubule oxidative stress [27]; (ii) the latter frees cytoplasmic Nrf2 from its inhibitor, KEAP1, allowing for Nrf2 nuclear translocation [27]; (iii) Nrf2 binds to and activates a host of cytoprotective genes [6, 7, 27]; and (iv) with the accumulation of their cognate cytoprotective proteins (e.g. haptoglobin, hemopexin, hepcidin, heme oxygenase, ferritin), renal protection against diverse forms of renal injury results [6, 7].
Given P21’s well-known cytoprotective properties [12–23], and in light of Nishioka’s report that ‘ischemic’ preconditioning results from P21 induction [24], we questioned whether P21, and associated cycle inhibition, act in concert with Nrf2 to induce the SnPP/Fe-mediated cytoprotective state. In contrast to the above-noted protective proteins, P21 is a P53-inducible, not an Nrf2-inducible, gene. Nevertheless, it has recently been suggested that a P21–Nrf2 interaction exists [43]. For example, by inhibiting KEAP1–Nrf2 binding, P21 can facilitate Nrf2 nuclear translocation, and hence facilitate Nrf2 induction of cytoprotective genes [43]. Indeed, it has been suggested that P21’s cytoprotective activities may stem, at least in part, from increased Nrf2 activity.
Given these considerations, we assessed whether P21 upregulation might also be triggered by SnPP-/Fe-mediated oxidative stress, and thus contribute to the resultant cytoresistant state. The answer appears to be no, given that P21 mRNA and protein levels remained at control levels at both 4 and 18 h post-SnPP/Fe administration. The finding of normal P21 expression at 18 h post-oxidant preconditioning is particularly noteworthy, given that this is the time at which SnPP/Fe-induced cytoresistance is fully expressed. Further supporting the conclusion that SnPP/Fe-induced preconditioning is independent of P21 activation were findings that: (i) P53, the normal inducer of P21, was not upregulated by SnPP/Fe injection (Figure 4) and (ii) OIP slightly ‘increased’, rather than ‘decreased’, nuclear localization of the cell proliferation markers KI-67 and PHH3 (possibly due to an upregulation of myc; data not shown). Finally, we considered the possibility that OIC might enhance P21 expression shortly after, rather than before, renal ischemia and thus exert a renal ‘rescue’ effect. However, this was clearly not the case, given that oxidant preconditioning markedly suppressed post-ischemic P21 protein levels (Figure 5). Indeed, the greater the degree of OIP-induced protection against ischemic AKI, the greater the suppression of renal cortical P21 content (Figure 7). Hence, these findings suggest that P21 might actually serve as a tissue ‘biomarker’ of post-ischemic injury severity. Indeed, given P21’s well-known cytoprotective actions, this post-ischemic P21 reduction, while reflective of less tissue injury, could potentially represent a maladaptive biomarker response.
Because the above data seemingly eliminated P21 upregulation as a critical determinant of oxidant-induced renal preconditioning, we sought to strengthen the alternative hypothesis, that is, that the Nrf2-mediated gene activation helps induce the oxidant-induced preconditioning state. Although we previously documented that SnPP/Fe increases Nrf2 nuclear translocation [27], and that this is associated with the activation of a number of Nrf2-sensitive genes (e.g. HO-1, haptoglobin, hemopexin, ferritin; [6, 7, 27]), it is notable that multiple transcription factors can overlap with Nrf2 to produce these results. For example, although HO-1 is a prototypic Nrf2-activated gene, it can also be upregulated by MAP kinase, NF-κB, Hif-1α, HSF1 and AP-1 as well as a number of growth factors [44, 45]. Hence, SnPP-/Fe-induced HO-1 gene induction does not equate with an Nrf2-dependent mechanism. Thus, to solidify the role of Nrf2 in SnPP/Fe preconditioning, we sought out genes whose induction are believed to be highly Nrf2 specific. NQO1, GCLC and SRXN1 are three such genes [46–49]. Indeed, when SnPP/FeS was administered to Nrf2−/− mice, no NQO1, GCLC or SRXN1 gene responsiveness was observed. Conversely, when SnPP/FeS was administered to their Nrf2+/+ counterparts, 3- to 5-fold NQO1, GCLC and SRXN1 mRNA increases occurred. Thus, these findings provide conclusive evidence that SnPP/Fe does, indeed, activate the Nrf2 pathway. It is notable that NQO1, GCLC and SRXN1 each possess cytoprotective properties, as follows: (i) NQO1 is a well-documented cytoprotectant, acting within quinone detoxification pathway; (ii) GCLC encodes gamma-glutamylcysteine synthetase, the first rate-limiting enzyme of glutathione synthesis; and (iii) SRXN1 detoxifies highly reactive peroxides, including hydrogen peroxide and peroxynitrite. Thus, documentation of SnPP-Fe-induced upregulation of NQO1, GCLC and SRXN1 expands upon the potential mechanisms by which oxidant-induced preconditioning agents may mediate their cytoprotective effects. Finally, the fact that oxidant preconditioning with SnPP/Fe did not significantly mitigate the severity of ischemic AKI in Nrf2−/− mice further implicates Nrf2 activation as a key mediator of the SnPP/Fe-induced preconditioning state.
Finally, we sought to determine whether a consequence of ischemic renal injury is the onset of cellular senescence, and if so, can this be abrogated by the above oxidant-induced preconditioning strategies? Both situations appear to be the case. In regards to senescence, at both 18 h and 2 weeks following a 25-min ischemic insult, dramatic increases in P21, P16 and P19 gene expression, and a doubling of 2-week SA-β-gal levels, were observed. These 2-week changes corresponded with a 45% loss of renal mass. Even the 5-min IPC protocol partially recapitulated these results (persistent P21/P16 upregulation; increased tissue SA-β-gal; a modest post-ischemic loss of renal mass). The ability of oxidant preconditioning to mitigate these changes was indicated by the following: (i) SnPP/Fe markedly blunted ischemia-induced P21, P19 and P16 increases (Figures 5 and 10); (ii) SnPP/Fe almost completely prevented post-ischemic SA-β-gal elevations (Figure 11); (iii) SnPP/Fe caused a 3-fold reduction in NGAL expression, a biomarker of AKI severity; and (iv) there was a dramatic preservation of post-ischemic renal mass, as assessed both by renal weight (Figure 11) and renal LDH retention. There are two likely scenarios by which oxidant preconditioning may have prevented this ischemia-induced senescent state. First, given that oxidant-induced DNA damage helps initiate the senescence process [50] an upregulation of Nrf2-mediated antioxidant defenses could blunt DNA damage, and hence, the onset of the post-ischemic senescence state; and second, senescence could simply be a ‘downstream consequence of’ post-ischemic AKI, rather than being a significant ‘mediator’ of it. Hence, the ability of oxidant-induced preconditioning to blunt AKI severity (Figure 7) and post-ischemic progression (Figure 10) could have been reflected by secondary reductions in senescence markers. Clearly, much additional work will be required to sort out these two possibilities. This is particularly true in light of conflicting observations that the INK4a gene locus, and its downstream products, P16 and P21, have been reported to have both protective [51, 52] as well as adverse [53] renal effects.
In conclusion, this study was undertaken to advance our understanding of the mechanisms by which transient, mild renal oxidative stress (as induced by SnPP/catalytic Fe) induces renal preconditioning and the emergence of a cytoresistant state. In sharp contrast to IPC, which is reportedly dependent on cytoprotective P21 upregulation, OIC is expressed in the absence of either P21 induction or associated cell cycle arrest. Also unlike IPC, OIC can confer an anti-senescence effect. The available data suggest that a non-P21-dependent mechanism, most notably Nrf2 pathway activation, is involved. This latter conclusion is based on findings of OIP upregulated Nrf2-dependent genes in Nrf2+/+, but not in Nrf2−/− mice, and observations that Nrf2−/− mice did not manifest significant OIP-mediated protection against ischemic ARF. In composite, these results add to a growing body of information that mild oxidant-induced renal preconditioning strategies could represent a sought-after clinical goal [54]: safely activating the Nrf2 pathway, and its concomitant cytoprotective effects.
ACKNOWLEDGEMENTS
The authors thank Savanh Chanthaphavong, PhD, Head of the Experimental Histopathology Laboratory, Fred Hutchinson Cancer Research Center, for her assistance with renal assessments of KI-67, PHH3 and SA-β-gal staining. Mr Stuart Tenney provided expert administrative support.
FUNDING
This work was supported by research funding from Renibus Therapeutics, Lantana, TX, USA.
CONFLICT OF INTEREST STATEMENT
R.A.Z. serves as a scientific consultant for Renibus.
Comments