Introduction
Acute kidney injury (AKI) is a common complication of perioperative illness, and mortality is up to 70 %. In noncardiac surgical patients, the incidence is 1 %, conferring a seven- to eight-fold increment in mortality. The most common cause of hospital-acquired AKI is whole-body hypoperfusion [
1‐
5], which is also independently associated with perioperative AKI.
Overall, men are more likely to experience AKI than women [
6,
7]. Clinical data suggest that low estrogen states are associated with increased risk of AKI [
8,
9]. Data from animal models indicate that sex hormones mediate AKI; androgen and testosterone enhance the susceptibility to AKI, and estrogen is a protective factor [
10,
11]. We previously demonstrated in mice that females enjoy relative protection from AKI after whole-body ischemia-reperfusion injury, removal of physiologic estrogen worsens AKI, and preprocedural restoration of physiologic estrogen to ovariectomized female mice restores the protected state [
12,
13].
Although estrogens have been studied widely, whether estrogen is efficacious when administered after whole-body ischemia-reperfusion and whether such protection can be observed in males as well as females with intact ovaries have not been determined. Therefore, to gain insights into the role of estrogen as a treatment option, we determined whether a single injection of estrogen, administered on recovery from cardiac arrest and cardiopulmonary resuscitation (CA/CPR), improves AKI in both sexes and whether the effect is dependent on renal blood flow. It is relevant to administer estrogen in acute phase after ischemic insult in the aspect of investigating estrogen as a therapeutic agent since clinically CA is frequently unheralded, and therefore therapy is applied following resuscitation.
Clinical trials of ischemia-protective strategies proven in mice may have failed in the past because of the reliance on young mice [
14-
16]. Both experimental and clinical studies demonstrate the strong linkage between aging and AKI [
17-
19]. Clinical data suggest that women enjoy protection from AKI in general but not when undergoing cardiac and vascular surgery, settings in which patients are older than the overall surgical population [
20]. This bimodal risk distribution may be a phenomenon of the menopause and changing exposure to estrogen with age. However, few studies have addressed effect modifications of aging on the association between sex and AKI. Therefore, other important questions are whether the advantage of females against AKI after whole-body ischemia is lost with age and whether this is modulated by estrogen.
Together, the major aims of the present study were (1) to determine whether post-arrest administration of estrogen reduces AKI following CA/CPR in both sexes and (2) to investigate whether aged females, the clinical population with greatest AKI risk, lose the beneficial effect of female sex or benefit from post-arrest administration of estrogen or both. A further aim was to determine whether estrogen receptor (ER) populations in the kidney, unaffected by pre-arrest hormonal manipulation, might change with age.
Methods
All experimental protocols involving animals were conducted at Oregon Health & Science University in Portland, Oregon. All animal experiments were conducted in conformity with the National Institutes of Health guidelines for the care and use of animals in research and were specifically approved by the Oregon Health & Science University Institutional Animal Care and Use Committee.
Animals and experimental groups
Male and female C57BL/6 mice were obtained from Charles River Laboratories (Boston, MA, USA). We used young mice (10–15 weeks), middle-aged mice (43–48 weeks), and aged mice (78–87 weeks). Experimental animals were randomized to 17β-estradiol or vehicle treatment and subsequently underwent experimental procedures.
In vivo whole-body ischemia-reperfusion (CA/CPR)
We performed normothermic CA/CPR as previously described [
12,
13,
21,
22] Briefly, mice were anesthetized with 4 % isoflurane in a 2:1 air-to-oxygen mixture, and anesthesia was maintained with 1 %–2 % isoflurane. After tracheal intubation with a 22-guage Teflon catheter (BD, Franklin, NJ, USA), animals were mechanically ventilated and the electrocardiogram (EKG) was monitored continuously. For drug administration, a PE-10 catheter was inserted into the right internal jugular vein. Rectal temperature was maintained 37 ± 5 °C throughout the experiment with a digitally controlled lamp (Cole Parmer, Vernon Hills, IL, USA). CA was induced with 50 μl of 0.5 M potassium chloride and confirmed by isoelectric EKG signal and absence of visible cardiac contraction on the chest wall. At onset of CA, the endotracheal tube was disconnected. Seven and a half minutes after onset of CA, mechanical ventilation resumed. Eight minutes after onset of CA, chest compressions were delivered by a finger at a rate of 300 per minute, and epinephrine in 0.9 % sodium chloride solution, 11–16 μg (0.7–1.0 ml), was administered intravenously. Return of spontaneous circulation (ROSC) was confirmed by reappearance of electrical activity on the EKG and visible cardiac contractions on the chest wall. Animals were extubated when their spontaneous respiratory rate reached more than 60 breaths per minute and then were moved to a recovery cage.
Preparation of 17β-estradiol and correspondent vehicle
A previous study performed in the same laboratory demonstrated that injection of 0.5 μg of 17β-estradiol resulted in a peak of serum estradiol levels shortly after administration followed by a decrease to physiologic levels within 1 h in male C57BL/6 mice [
23]. We therefore administered 0.5, 5, and 10 μg of 17β-estradiol to investigate whether any renoprotection is dose-dependent. Water-soluble cyclodextrin-encapsulated 17β-estradiol (E4389; Sigma-Aldrich, St. Louis, MO, USA) was dissolved in 0.9 % sodium chloride solution to be the desired dose in 100 μl. Each vehicle control was rendered so as to contain the same amount of hydroxypropyl cyclodextrin as its corresponding 17β-estradiol treatment.
Regional renal cortical blood flow measurement
We measured regional renal cortical blood flow (RRCBF) in the peri-arrest period. A laser Doppler flow probe (Moor Instruments, Wilmington, DE, USA) was placed perpendicular to the surface of the right kidney via a flank incision and optimally positioned for maximal signal. Laser Doppler flow was measured and recorded 5 min before CA, immediately before CA, every minute during CA, and every 5 min after CA for 45 min after ROSC for each animal treated with 17β-estradiol or vehicle at 15 min after ROSC. Fifty minutes after ROSC, arterial blood was aspirated from the left ventricle for blood gas assessment.
Transcardial perfusion and tissue sampling
Twenty-four hours after CA/CPR, thoracotomy was performed under deep anesthesia. After blood sampling, mice were perfused with saline, and the right kidney was snap-frozen in liquid nitrogen and stored at −80 °C until immunoblotting. Subsequently, mice were perfused with 4 % paraformaldehyde and then the left kidney was harvested for histological assessment.
Evaluation of serum urea nitrogen, serum creatinine, and serum estradiol levels
A point-of-care enzyme-coupled analyzer (Abaxis Medical Diagnostics, Union City, CA, USA) was used to measure serum urea nitrogen (sUN) and serum creatinine (sCr). 17β-estradiol concentrations were measured by using radioimmunoassay as previously described [
12]. The GDN 244 antibody was used for radioimmunoassay [
24]. The overall inter-assay variation for 17β-estradiol extraction radioimmunoassay was less than 15 %, and the intra-assay variation was less than 12 %.
Histological evaluation of tubular injury
An observer blinded to treatment and sex assessed histological damage by using rigorous unbiased stereology. After fixation, 6-μm sagittal sections from the kidney were sampled 160 μm apart and then stained with Flouro-Jade B (Histo-Chem, Jefferson, AK, USA), which stains necrotic cells bright green. Histological damage was determined in accordance with the Cavalieri principal of unbiased stereology by using Visiopharm Integrator Software (Visiopharm, Hørsholm, Denmark) and delineating the total kidney area and applying superimposed grids consisting of two marks in a fixed relation (4:64) randomly onto a field of view. The estimated volume fraction of necrotic tubules was calculated as follows:
$$ \mathrm{Volume}\ \mathrm{necrotic}\ \mathrm{tubules}\ \left(\mathrm{V}\mathrm{N}\mathrm{T}\right)={\sum}_{\mathrm{Marks}\ \left(\mathrm{necrotic}\ \mathrm{tubules}\right)}/\ {\sum}_{\mathrm{Marks}\ \left(\mathrm{kidney}\right)}\times 16 $$
The sampling intensity was decided according to our previous study to obtain precise estimation with coefficient of error of less than 0.1 by using the formula of Gundersen et al. [
25]. The coefficients of error for the necrotic tubules and the reference area in our study were 0.08 and 0.05, respectively.
Immunoblotting
Expression of ERα and G protein-coupled estrogen receptor (GPER) in homogenates of the entire right kidney obtained 24 h after CA/CPR were measured by using rabbit polyclonal anti-ERα (clone HC-20; Santa Cruz Biotechnology, Inc., Dallas, TX, USA) and anti-GPER (clone N-15; Santa Cruz Biotechnology, Inc.) antibodies. To eliminate any effect of exogenously administered estradiol, this assay was performed only in vehicle-treated mice. Mouse monoclonal anti-vinculin (clone cp74; CalBiochem/EMD Millipore, Billerica, MA, USA) was used as loading control. Secondary antibodies were anti-mouse horseradish peroxidase (HRP) and anti-rabbit-HRP from GE Healthcare (Little Chalfont, Buckinghamshire, UK) and Cell Signaling Technology (Danvers, MA, USA), respectively. Tissue was homogenized in ice-cold radioimmunoprecipitation assay buffer (ThermoFisher Scientific, Waltham, MA, USA) containing protease inhibitor cocktail (Sigma-Aldrich) and PhosSTOP phosphatase inhibitor cocktail (Roche, Indianapolis, IN, USA). Protein concentration was measured by using the BCA protein assay (ThermoFisher Scientific). Western blots were performed as previously described [
26]. Primary antibodies were used at a dilution of 1:50 (ERα), 1:500 (GPER), or 1:2000 (vinculin) in 5 % bovine serum albumin (BSA) in tris-buffered saline with Tween-20 (TBST) (25 mM Tris, 1 M NaCl, 1 % Tween 20). Secondary antibodies were diluted 1:2000 (ERα), 1:5000 (GPER), or 1:15,000 (vinculin) in 5 % BSA in TBST.
Statistical analysis
Statistical significance was inferred from a P value of less than 0.05. Two-group two-treatment analyses were performed by using two-way analysis of variance with post hoc Sidak’s test. Categorical outcomes were assessed by using Fisher’s exact test or chi-squared test, as appropriate. Analysis of differences in estrogen expression between young and aged mice was performed via unpaired t test. All results are presented as mean ± standard deviation. Analysis was performed with Prism 6.0 software (GraphPad Software, Inc., La Jolla, CA, USA).
Discussion
There are three major findings of the present study. First, the advantage of female sex against AKI after whole-body ischemia, which is consistently observed both in our previous study and in the present study, weakens with age. Second, estrogen exerts renoprotective effects even when administered after resuscitation from CA. To the best of our knowledge, this is the first study to demonstrate that exogenous estrogen acts as a potent renoprotective agent when administered in a clinically relevant manner, after whole-body ischemia-reperfusion. Third, estrogen injection improves renal outcome in males, but reproductively intact young female are not affected because injury is small, whereas in aged mice, estrogen is beneficial to both sexes. Taken together, these three findings indicate that post-arrest estrogen administration is renoprotective in a sex- and age-dependent manner.
We confirmed that females enjoy relative renoprotection after CA as previously reported [
12,
13,
21] and extend this finding with the finding that this advantage weakens in middle age as tubular necrosis becomes equivalent to that of males. Functional measures of injury are still reduced in middle-aged females. Previous studies have demonstrated that female mice enter reproductive senescence by 44–64 weeks [
29]. We chose to assess sex difference in middle age precisely because some females, but not all, may have ceased reproductive cycling, thus the mixed finding was expected. To further evaluate the effect of age and reproductive cycling, we used mice of extreme age (78–87 weeks). These aged females lose their advantage both in renal function and in tubular cell death. Serum estradiol levels in aged females decreased to the same level as physiologic level of male mice. Taken together, these data suggest that females lose renoprotection according to age because of changing exposure to estrogen. This would recapitulate our previous findings using ovariectomized young mice [
12,
13]. However, whether intact females lose their advantage according to reproductive senescence remains a vital translational question, and we directed this study at this query. Our findings are in line with previous literature studying age-dependent progression of renal damage in female mice. Urbieta-Caceres et al. suggested that aging-induced renal interstitial fibrosis in female mice was mediated by loss of endogenous estrogen [
30]. Zheng et al. reported that female mice developed progressive glomerulosclerosis after menopause [
31]. Although we did not find any difference between young and aged female surgically naïve animals, those structural abnormalities caused by aging may increase susceptibility to acute ischemic insult in aged females.
The present study demonstrates for the first time a beneficial effect of estrogen on AKI after CA/CPR in male mice. To assess critical translational benefit of estrogen in male mice, we chose the timing of injection at 15 min after ROSC and evaluated three doses. We chose CA/CPR as our model for ischemia-reperfusion injury because it replicates key features of clinical ischemia-reperfusion, making it clinically relevant. First, it replicates the most common cause of clinical AKI, whole-body hypoperfusion [
4]. Second, as can be seen from our data, CA/CPR produces mild-moderate renal injury similar to that seen in clinical ischemia-reperfusion injury. Third, other groups have found that CA/CPR may more closely model ischemic inflammatory, immune, and renal arterial flow changes than occlusion of the renal pedicle, a widely used model of renal ischemia [
32,
33]. Our data indicate a dose-dependent reduction of renal injury by post-ischemic administration of estrogen. In previous experiments using young female mice exposed to CA/CPR, we found that ovariectomy significantly increased injury and that estradiol treatment 7 days prior to ischemic insult, designed to replicate physiologic levels, restored protection [
12,
33]. Müller et al. found that increased mortality after focal renal ischemia in male rats compared to female was reversed with estrogen treatment 7 days prior to the insult [
10]. Park et al. also reported that estrogen treatment 15 days before renal artery occlusion reduces the post-ischemic increase of sUN and sCr in male mice [
11]. Takaoka et al. [
34] and Tanaka et al. [
35] demonstrated with the renal pedicle occlusion model that male rats injected with estrogen 15 min prior to reperfusion demonstrated improved post-ischemic renal outcomes. On the other hand, a recent study which employed intramuscular injection of estrogen 3 days prior to renal ischemia-reperfusion surgery found that estrogen’s protective effect was limited to ovariectomized female rats and did not protect male rats [
36]. The discrepancy of these findings is probably due to the heterogeneity of the experimental model and the length of time of administration. Estrogen dose varied according to the study (25 μg/kg × 7 days [
10], 40–500 μg/kg × 14 days [
11], 100 μg/kg × single dose [
35] and 500 μg × 3 days [
36]); however, no prior study has evaluated the effect of estrogen administration
after reperfusion, either whole-body ischemia/reperfusion or renal vascular occlusion. In the present study, only higher dose of estrogen (400 μg/kg) injected during the early reperfusion phase improved the renal outcome in male mice. Estrogens may exert a different mechanism of action depending on the timing of injection, dose, and extent of overall injury. Thus, it is notable that not all studies of estrogen in CA have demonstrated a protective effect. A recent study using a male rat model of ventricular fibrillation and resuscitation reported that estrogen administration did not result in beneficial myocardial effect or impact survival. The survival rates 3 h after resuscitation in this severe-injury model are 0 % versus 33 %,
P = 0.32, estrogen versus saline, for bolus injection and 63 % versus 100 %,
P < 0.05, estrogen versus saline, for continuous injection [
37]. The present study uses a CA/CPR model which demonstrates a mild injury and survival of more than 65 % in order to assess renal outcome 24 h after ROSC, and this may explain the apparently discordant result.
Pre-ischemic estrogen alters outcome of focal renal ischemia via nitric oxide and endothelin-1 [
10,
34]. Because both of these mediators are believed to act on renal blood flow, our findings that neither RRCBF nor systemic metabolic parameters are affected by estrogen during the post-arrest period suggest that the mechanism of acute renoprotection by post-arrest estrogen may be distinct from that of chronic or pre-ischemic administration. Since no prior study addresses post-treatment with estrogen in the kidney, we note that after experimental stroke estrogen exerted a protective effect due to enhanced cerebral blood flow [
38,
39]. In our whole-body ischemia model, we found that estrogen does not alter RRCBF or systemic metabolic parameters in the recovery period up to 45 min following CA/CPR. We cannot eliminate the possibility that renal medullary but not renal cortical blood flow is affected by estrogen, as it is reported that even though cortical blood flow fully recovers in 30–45 min after bilateral occlusion of the renal arteries and veins and reperfusion, medullary blood flow remains compromised for prolonged periods that are associated with a long-term decline in renal function [
40,
41]. However, this investigation is focused on acute, 24-h outcomes, and estrogen mediation of long-term outcome was not assessed in this study. Another possibility is that the use of isoflurane in our study confounds the evaluation of blood flow, as it is a vasodilator. An additional possible limitation regarding the use of isoflurane is its well-characterized renoprotective effect [
42]. It is not possible to perform our study without anesthesia; we carefully designed experiments to exclude differential effect of either vasodilatory or renoprotective properties of isoflurane. As all animals received the same concentration of isoflurane, no observed effect can be due to the influence of isoflurane alone.
In the present study, efficacy of post-ischemic estrogen injection was observed amongst young mice only in males. On the other hand, estrogen was renoprotective in old mice of both sexes. In our previous studies, we found that deletion of ERs α, β, and GPER did not interfere with protection against renal injury from ischemia/reperfusion when estrogen was administered for a week prior to CA/CPR [
12,
13]. In evaluating these findings, we hypothesized that pre-ischemic and repeated dosing of estrogen could protect the kidney through multiple mechanisms mediated by more than one receptor; thus, deletion of a single receptor would not abrogate the protective effect. As acute and chronic administrations of estrogen are differentially protective in neurologic ischemia [
43], this might also be the case in renal ischemic injury. Thus, it is essential to evaluate involvement of renal ERs in the effect we observed. Since aged females are in a low-estrogen environment compared with young females, we hypothesized that ER expression levels differ between the two, and this might explain the different response to the exogenous estrogen. We therefore measured renal ERα and GPER levels in animals not treated with estrogen. We found that the expression of ERα in aged females was increased compared with young females 24 h after CA/CPR. This finding in animals after CA/CPR is similar to those of others: in surgically naïve animals, ERα expression increases with age in renal tissue [
44]. Our finding is limited in that the 24-h time point may not precisely replicate the levels present at the time of reperfusion; therefore, the inference of mechanistic association should be interpreted with caution. In heart [
45] and brain [
46,
47], ER activation reduces ischemic injury. In light of these findings and in light of recent data indicating ERα as a mediator of protective estrogen signaling in female rats with pulmonary hypertension-induced right ventricular failure [
48], we speculate (with the above limitations) that protective acute estrogen effects in aged females are mediated at least in part by ERα, and further investigation will test this hypothesis.
Sex differences in the expression of ER during aging [
44] and in neuroprotective effects of ER agonists have been reported [
49]. Broughton et al. found that GPER expression was increased in the brain of male mice but not of female mice after transient focal cerebral ischemia [
50] and that GPER activation worsens outcome in males but that an opposite protective effect occurs in females [
49]. Interestingly, the expression of GPER in kidney 24 h after CA/CPR was increased in aged male compared with young male mice in our study. The underlying mechanisms responsible for the estrogen’s protective effect to renal ischemic injury observed in our study require further investigation, particularly with regard to sex difference. Androgens, including testosterone, are reported to increase the susceptibility of males to ischemic renal injury [
10,
11]. Myocardial ischemia reperfusion injury in male rats aggravated cardiac damage through androgen receptor [
51]. The interaction of sex hormones besides estrogen and their change in the aging process might modify the effect of estrogen’s renoprotection in males. Together, these findings highlight the complex nature of endogenous estrogen signaling, which may vary between different physiological and disease settings. Sex should be considered in identification of the underlying mechanism responsible for the effect of estrogen.
In addition to the sex and age specificity of estrogen’s renoprotective mechanisms, it is possible that estrogen exerts its effect differently because of timing. We acknowledge that most previous studies are designed in the aspect of hormone replacement therapy, mostly involving long-term drug pretreatment before experimental ischemia, and that more work is required to fully evaluate the efficacy of estrogen as a potent therapeutic agent. Our study, designed to evaluate efficacy of acute and transient estrogen exposure, shows that the renoprotective effect is rapid, exerted within 24 h of ischemia reperfusion, perhaps sooner, suggesting one of the non-nuclear, rapid actions of estrogen [
45,
52]. The fact that serum estradiol levels have returned to physiologic levels within 24 h after estradiol injection reinforces the argument of estrogen’s rapid action. In our recent
in vitro study, we found that hyperpermeability of glomerular endothelial cells after ischemia was prevented by estrogen in part through GPER [
21]. Interestingly, in that study, the significant decrease in glomerular permeability by estrogen observed after 4 h of re-oxygenation was no longer evident at 8 h, suggesting that estrogen’s effect is acute and transient. Preserved glomerular barrier integrity during the acute phase of reperfusion might prevent subsequent tubular dysfunction and tubular cell death. We did not assess whether post-arrest estrogen injection alters glomerular permeability or GPER expression at an early phase of reperfusion in the present study, which was focused on whether estrogen exerted any acute beneficial effect in renal ischemia. Further investigation of glomerular permeability effects of estrogen
in vivo might delineate a mechanism of the protective effect we observed. We did not find any difference in 24-h survival rate between estrogen-treated animals and vehicle-treated animals in our potassium-induced CA model (76 % versus 66 %,
P = 0.30, estrogen versus vehicle), and estrogen-treated animals had less kidney injury.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
MI carried out the in vivo studies, analyzed data, and drafted the manuscript. TS performed in vivo studies and drafted the manuscript. AV carried out immunoblotting, analyzed data, and drafted the manuscript. TL conceived immunoblotting experiments, analyzed data, and drafted the manuscript. SA designed the study, drafted the manuscript, and critically revised the manuscript. MPH led the research team, designed the study, analyzed data, and drafted the manuscript. All authors read and approved the final manuscript.