Background
Acute kidney injury (AKI) is an underestimated, yet important, risk factor for the development of chronic kidney disease (CKD) [
1]. Long-term follow-up studies (4 months to 6 years) report that between 35 and 71% of patients surviving an episode of AKI had incomplete recovery of renal function as assessed by creatinine clearance or serum creatinine measurements [
2]. These patients are more likely to progress to end-stage renal disease (ESRD) as compared to patients without a history of either AKI or CKD [
1] and contribute to the growing population of CKD patients. Currently, there are no therapeutic interventions targeting disease progression after AKI [
3] highlighting the urgent need for novel therapeutic approaches that aim at preventing and/or reversing the pathophysiologic sequelae of AKI [
4].
Renal ischemia-reperfusion injury, due to hypoperfusion after surgery, bleeding or dehydration, is a major aetiology in human AKI and is of particular importance in the setting of kidney transplantation [
5]. We previously optimized a mouse model of AKI to CKD by unilateral ischemia-reperfusion (UIRI) without contralateral nephrectomy, with development of moderate renal fibrosis and significant long-term inflammation [
6].
Inflammation plays a major role in the pathophysiology of ischemic AKI [
7]. Post-ischemic tissue infiltration by neutrophils, macrophages, and different subtypes of T-cells is a hallmark of acute renal ischemic injury, both in patients and experimental models [
7]. Persistence of inflammation may contribute to maladaptive cellular repair after acute injury and may be an intrinsic component of progression of renal injury [
3]. In view of the above, attenuation of inflammation after acute ischemic kidney injury may be a suitable therapeutic strategy in the attenuation or prevention of progressive renal injury. Dexamethasone is a glucocorticoid, widely used in renal diseases as an anti-inflammatory and immunosuppressive agent [
8]. Corticosteroids inhibit the synthesis of chemokines and cytokines resulting in protection against inflammation, and at high doses inhibit the immune response [
9]. It was already shown in experimental models that dexamethasone has a protective effect against ischemic damage in liver and hart [
10,
11]. In the kidney, pre-treatment with dexamethasone has been demonstrated to ameliorate the severity of an acute ischemic insult [
8]. These studies, however, particularly covered the acute injury phase up to 24 h after the ischemic insult. In the light of the recently appreciated link between AKI and CKD, we here evaluated for the first time the long-term (up to 6 weeks) impact of temporary (3 weeks) inflammatory suppression by dexamethasone on pathology progression and development of fibrosis in ischemia-reperfusion injured kidneys.
Discussion
It is becoming increasingly clear that incomplete recovery from severe AKI is an important pathway to persistent and progressive CKD with underlying fibrosis. A recent meta-analysis reported that patients surviving an episode of AKI have an 8.8-fold increased risk for CKD and a 3.3-fold increased risk for ESRD [
3]. Understanding the mechanisms underlying the progression from acute-to-chronic renal injury is the focus of recent research in the field [
15]. Since renal fibrosis is nearly always preceded by and closely associated with inflammation, both in patients [
16] and experimental models of fibrosis [
17], it is thought to be one of the major processes that contributes to progression of renal disease [
18]. Also, experimental studies demonstrated that even when renal function recovers after AKI, pro-inflammatory and pro-fibrotic pathways remain active [
19]. Several laboratories demonstrated that suppression of the inflammatory response can reduce post-ischemic injury 24–48 h after the ischemic insult [
20]. Yet, few studies investigated the long-term effects of inflammatory modulation on development of CKD, which, in the context of the recently appreciated AKI-to-CKD link, is of major therapeutic interest. Therefore, we here evaluated whether temporary treatment (3 weeks) with immune-suppressive dexamethasone is able to attenuate the development of post-ischemic renal fibrosis and avert the progression from acute to chronic renal injury.
In the current study, renal atrophy is pronounced and progressive, with loss of renal mass up to 44 and 64% within 3 and 6 weeks after the ischemic insult, respectively (Fig.
1). Treatment with dexamethasone was unable to attenuate or prevent this loss of renal mass. However, adaptive growth of the contralateral kidney to compensate for the loss of functional renal tissue [
21] of the ischemic kidney did not occur immediately after dexamethasone treatment. This could indicate that dexamethasone was able to rescue a certain degree of renal function (however not renal mass) of the ischemic kidney. Although the treatment regimen and dosing was the same as in the experiment of Zager et al. (2011), who observed 50% loss of renal mass in dexamethasone treated animals as compared to 66% in untreated animals [
12] we can only speculate as to why dexamethasone did not influence renal atrophy in our study. Most likely this is due to differences in the severity of the ischemic insult, which is very subjective to inter-laboratory variation as well as mouse strain dependent susceptibility to ischemic AKI [
22]. The presumption that the loss of function of the ischemic kidney was attenuated by dexamethasone treatment, is further supported by the mitigated development of post-ischemic fibrosis, as evidenced by a reduction in collagen I gene- and protein expression, and
Ccn2 gene expression (Figs.
2b-c and
3b). Although we did not assess the effect of dexamethasone treatment on collagen I expression of fibroblasts within the kidney ex vivo or in vitro, among others, Zern et al. [
23] and Waqar et al. [
24] previously reported that treatment of fibroblasts with dexamethasone lead to decreased collagen I and collagen IV expression accompanied by decreased collagen synthesis in vitro. Moreover, gene expression of
Ccn3, that has been shown to have anti-fibrotic properties [
25], was significantly elevated upon dexamethasone treatment (Fig.
3d). Furthermore, although gene expression of the pro-fibrotic
Pai-1, an inhibitor of matrix degradation, appeared to be significantly elevated after dexamethasone treatment in comparison to the untreated condition, it did not differ from the vehicle-treated group. Similar results were observed in mercury chloride-induced nephropathy [
26]. Despite these molecular observations, dexamethasone treatment was not able to attenuate expansion of the tubulointerstitial area after UIRI.
To further examine the effect of dexamethasone treatment on the fibrotic response after an acute ischemic insult, protein expression of α-SMA, a marker for activated fibroblasts, was determined. In normal kidney tissue, α-SMA staining is only found in smooth muscle cells, mostly in blood vessels [
27]. In fibrotic diseases, α-SMA expression of myofibroblasts is recognized as a hallmark of their emergence and an indicator of disease severity [
28]. In our study, dexamethasone did not have an effect on the degree of α-SMA expression as compared to vehicle-treatment. In this respect it is worthwhile to note that α-SMA in myofibroblasts appears to have a suppressing role in tissue fibrosis progression, and forced expression in α-Sma
−/−animals ameliorates fibrosis in the model of ureter obstruction and mesangioproliferative glomerulonephritis [
28].
In addition to fibrosis, the post-ischemic period is characterized by an active inflammatory response, resulting from both activation of resident inflammatory cells and recruitment of circulating inflammatory cells [
16]. Baeck et al. (2015) have shown, by means of a fluorescent double stain (Ki67 and F4/80), that significant proliferation of monocyte-derived macrophages occurs in the ischemic kidney, both in the acute (day 3, day 5) and early chronic phase (day 20) [
29], thereby amplifying and prolonging the local inflammatory response [
30]. In accordance with this, we observed significant infiltration of inflammatory cells in the ischemic kidney, quantified by protein expression of the F4/80 glycoprotein, which is expressed by murine monocytes/macrophages. Also, significantly elevated gene expression of the inflammatory cytokines
Tnfα and
Tgfβ supports ongoing post-ischemic inflammation. Increased gene expression of the inflammatory
Tnfα was observed in the ischemic kidney after dexamethasone treatment as compared to untreated animals. Although unexpected at first sight, it has been shown that
Tnfα can modulate the expression of the glucocorticoid receptor isoforms in such a manner that glucocorticoid resistance may occur [
31]. Consistent with this is the fact that, steroid insensitivity has been described in renal epithelial cells [
32] and macrophages in the lung [
33]. Thus, it might not be surprising that dexamethasone treatment did not have an effect on the amount of F4/80 macrophage protein in the ischemic kidney. Also, it was shown by Castano et al. (2009) that it is possible to achieve attenuation of fibrosis without affecting the number of interstitial macrophages, quantified by F4/80 protein expression, which is also in line with our observations [
34].
On a technical side, the discrepancy we observed between Western blot and immunohistological quantitation of α-SMA and F4/80 expression is puzzling. A potential explanation could be that our tissue sampling strategy consistently lead to the use of tissue fragments that were harvested from the poles of the kidneys for Western blot whereas immunohistology was performed on complete transversal sections. Consequently, our Western blot samples could have contained less material originating from the inner cortex and outer medullary region which are the most affected by ischemia/reperfusion.
In normal physiological conditions, renal tubular cells have a low proliferation rate. The current consensus on renal tubular regeneration states that restoration of the tubular epithelium after an acute injury occurs predominately via proliferation of surviving epithelial cells that undergo dedifferentiation, primarily within the first 2 weeks [
35]. In our study, a significant increase in cell proliferation in the injured kidney is seen up to 6 weeks after UIRI. Since successful proliferation of proximal tubule cells, i.e. no G2/M cell cycle arrest, is associated with attenuation of fibrosis [
36], it was quite unexpected that temporary attenuation of renal decay immediately after dexamethasone treatment went along with an overall decrease in cell proliferation in the ischemic kidney. It should be noted, however, that quantification of proliferation in our study made no distinction between the renal cell types (epithelial cells, fibroblasts, inflammatory cells), nor cell cycle phase or their location (i.e. tubules vs. interstitium) due to severe distortion of the physiological tubulo-interstitial structure. However, independent studies reported decreased infiltration of lymphocytes [
26], T-cells [
37] and dendritic cells [
37] upon dexamethasone treatment. The question which proliferative cell type and cell cycle phase was primarily affected by dexamethasone treatment in the current setting lies outside the scope of the current report. However, we observed that a subset of interstitial cells were also positive for Ki67.
As a three-week dexamethasone treatment regimen attenuated the development of renal fibrosis after UIRI, we included an additional three weeks of follow-up without treatment, to evaluate whether the beneficial effects of dexamethasone treatment persist during the further course of the ischemic renal pathology. As compared to immediately after end of treatment, a further loss of renal mass was observed, indicating that the progressive nature of the pathological course could not be attenuated by dexamethasone. The fact that compensatory hypertrophy of the contralateral kidney emerged after the three-week follow-up period indicates that the positive effect of dexamethasone is temporary, and that longer or continuous treatment is necessary to reach persistent benefit from treatment. After the follow-up period, expression of the pro-fibrotic genes
collagen I and
Ccn2 was significantly increased as compared to the end-of-treatment time point (Figs.
2c and
3b), indicating that the beneficial effect of dexamethasone-treatment also on pro-fibrotic gene expression is transient. However, analysis of collagen I deposition by immunostaining showed no further progression of fibrosis during the follow-up period in the dexamethasone group, whereas vehicle treated animals displayed progressive renal fibrosis and increasing collagen I staining in the ischemic kidney (Fig.
2b).
Although the results of these experiments are consistent with respect to the long term effect of temporary immunosuppression, we must acknowledge some limitations to this study. Firstly, no functional assessment in serum or urine samples was included since the study setup was focussed on the effects on renal pathology. Secondly, as mentioned earlier, no distinction was made between the renal cell types, cell cycle phase or location of proliferating Ki67+-cells due to severe distortion of the physiological tubule-interstitial structure. Third, we observed for some parameters that vehicle treatment can induce similar effects as dexamethasone. Although we can only speculate on the reason for this, it emphasizes that comparative intervention studies cannot rely on untreated groups alone and should include vehicle treated groups to study the true potential of a particular compound.