1. Introduction
Ionizing radiation injuries (RI), often in conjunction with other injuries namely combined injury (CI), have historically and contemporarily manifested in both warfare scenarios, such as Hiroshima and Nagasaki, and nuclear accidents like Chernobyl and Fukushima. The persistent relevance of these events underscores the need for a deep understanding of the biological impacts of such injuries. RI and CI manifest across a spectrum of symptoms classified temporally into acute, consequential, and late effects. Acute effects occur during or immediately after exposure, consequential effects evolve from ongoing damage after the initial exposure, and late effects, such as kidney injury, may not surface until months or years later [
1,
2,
3,
4,
5,
6,
7].
The pathophysiology of radiation-induced kidney damage is complex and multi-faceted. Key mechanisms include direct DNA damage, inflammation, oxidative stress, microvascular damage, mitochondrial impairment, all of which culminate in renal tubular cell death [
3,
8,
9,
10,
11,
12]. There is no specific treatment for radiation-induced kidney injury or any types of acute kidney injury (AKI) [
13,
14].
Transcription factors like STAT1 and STAT3 are instrumental in regulating inflammatory mediators. They are activated by phosphorylation of certain amino acids, including tyrosine residuals. Protein tyrosine phosphatase SHP-1 modulates these pathways by dephosphorylating STAT proteins, thus acting as a critical regulatory checkpoint [
15,
16,
17]. Ionizing radiation induces reactive oxygen species (ROS), which include superoxide anion and hydrogen peroxide. The body's defense against ROS involves several antioxidative enzymes, including heme oxygenase-1 (HO-1), superoxide dismutase (SOD), catalase, and peroxidase [
18,
19,
20,
21,
22]. HO-1, in particular, is crucial for mitigating oxidative stress by breaking down heme, producing antioxidants like biliverdin and bilirubin, generating the protective gas carbon monoxide, and preventing iron from contributing to oxidative damage [
23,
24,
25]. The interplay between inflammation and ROS can establish a self-reinforcing cycle, where immune cells augment ROS levels, which then escalates further inflammation, thereby exacerbating organ injury [
20,
26]. AKT, also known as protein kinase B (PKB), is a serine/threonine kinase and plays a protective role in organ injury by promoting cell survival, reducing apoptosis, enhancing cell proliferation, suppressing inflammation, and preserving mitochondrial function [
27,
28]. AKT often inhibits p53 in order to exert its pro- cell survival and anti-apoptotic cell death function [
29,
30,
31].
Although the pathophysiology of RI and CI-induced kidney damage has become increasingly understood, the molecular pathways affected before any noticeable functional kidney damage remain largely unknown. Investigating these early alterations offers insight into the radiation's effects at the cellular and molecular levels, potentially elucidating why the kidney is resilient to immediate radiation-induced injury and identifying new therapeutic targets for intervention. In this study, we explored and contrasted the molecular pathways associated with inflammation, oxidative stress, and cell survival 30 days after mice were injured by radiation alone or in combination with a skin wound.
2. Results
Radiation alone or in combination with a skin wound has little immediate effect on renal function. As expected, RI did not significantly increase BUN levels (
Figure 1a), serum NGAL levels (
Figure 1b), or protein abundance of KIM-1 (
Figure 1c), a marker for renal tubular damage, in the mouse renal cortex compared to the sham group 30 days after mice were exposed to 9.5 Gy of
60Co radiation. CI did not significantly affect BUN or NGAL levels (
Figure 1d and e), but increased KIM-1 protein abundance (
Figure 1f) in the renal cortex when compared to the wound group. Compared with the sham group, wound alone had no significant effects on BUN (sham 42.7±2.8 mg/dL vs. wound 35.2±3.0 mg/dL) or serum NGAL levels (sham 3.2±0.7 ng/mL vs. wound 2.1±0.7 ng/mL) (
Figure 1a, b, d and e). These data suggest that neither RI nor CI reduce kidney function within 30 days after injury, but CI has a nuanced effect at the molecular level.
Radiation alone and combined injury upregulate inflammatory pathways, but they do so in distinct ways in the renal cortex. While RI had no significant effect on STAT1-Y701 phosphorylation (
Figure 2a), it significantly increased STAT3-Y705 phosphorylation compared with the sham group (
Figure 2b). In contrast, CI significantly increased STAT1-Y701 phosphorylation (
Figure 2c), but it had no significant effects on STAT3-Y705 phosphorylation (
Figure 2d). Wound alone significantly increased phosphorylation of both STAT1-Y701 and STAT3-Y705 compared with the sham group (
Figure 2e and f). Having found that RI, CI and wound alone activated STAT3 and STAT1, we sought to determine whether the insults affected the upstream signaling pathway of STATs. None of the insults had a significant effect on SHP-1-S591 phosphorylation or protein abundance of SHP-1 (
Figure 2g to i). However, RI decreased p38 phosphorylation (
Figure 2j), whereas CI tended to increase p38 protein abundance (
Figure 2k). Wound alone had no significant effects on phosphorylation of p38 or p38 protein abundance when compared with the sham group (
Figure 2l). These data indicate that RI and CI activate inflammatory pathways through different mechanisms without SHP-1 involvement in either pathway.
Radiation alone and combined injury have different effects on pro- and anti-inflammatory gene mRNA levels in the renal cortex. We found that while RI and CI had no significant effects on the majority of pro-inflammatory gene mRNA levels that were examined, RI significantly up-regulated mRNA abundance of IFNγR1, IFNγR2, and heparanase compared with the sham group (
Figure 3a), whereas CI only tended to increase mRNA levels of heparanase (
Figure 3b). Likewise, RI and CI had different effects on mRNA levels of immune-regulatory genes. RI remarkably increased IL-10 mRNA levels by 287.2%, whereas CI had no significant effects on the IL-10 mRNA abundance, but reduced the IL-4 mRNA abundance by 61.5% compared with their respective controls (
Figure 3c and d). Wound alone did not notably impact the mRNA levels of most pro-inflammatory and immune-regulatory genes. However, it led to a significant decrease in IFNγR1 mRNA levels when compared to the sham group, as shown in
Figure 3e and f.
Radiation alone and combined injury upregulate HO-1. Both RI and CI increased the HO-1 protein abundance by 80.7% and 38.3%, respectively, in the renal cortex when compared to their respective control groups, indicating the presence of oxidative stress after injuries (
Figure 4a and b). However, RI had an additional effect of reducing the protein abundance of MnSOD and catalase (
Figure 4c and g), whereas CI did not exhibit this reduction (
Figure 4d and h). To gain a comprehensive understanding of how these two injuries affected the anti-oxidative stress pathways, we conducted additional experiments to assess their impact on CuZnSOD, PRDX1, and ferritin heavy chain. Our findings indicate that neither RI nor CI had a significant effect on the levels of these proteins (
Figure 4e, f, and i to l). When compared with the sham group, wound alone had no significant effects on these anti-oxidant protein levels except for reducing catalase protein abundance (
Figure 4p).
Radiation alone and combined injury upregulate cell survival and downregulate cell apoptosis pathways. Both RI and CI led to a remarkable increase in the phosphorylation of AKT-S473 (149.0% and 208.8% respectively,
Figure 5a and b), while CI reduced protein abundance of AKT1 and AKT2 compared with wound alone (
Figure 5b). However, RI and CI had different effects on the phosphorylation of GSK-3β-S9, one of AKT's downstream targets. While RI increased the phosphorylation of GSK-3β-S9 and concurrently decreased the overall GSK-3β protein levels, CI did not significantly affect either parameter, highlighting the complexity of AKT signaling pathways (
Figure 5c and d). Despite any changes in protein abundance of HSC-70 (heat shock cognate protein 70), a pro-survival molecule (
Figure 5e and f), both injuries resulted in a decrease in protein abundance of p53 when compared with their respective controls (
Figure 5g and h). Wounding by itself did not significantly alter the levels of any of these proteins in comparison to the sham group (
Figure 5i to l).
Inflammatory Pathways
While it is well known that inflammation plays a critical role in RI- and CI-induced organ injuries, specific studies of the effects of these injuries on STAT1 and STAT3 remain scarce. X-ray intraperitoneal irradiation induced structural and functional damage in the mouse kidney from 3 to 5 months after the injury. This radiation-induced nephropathy was associated with increased protein abundance of STAT1 and STAT3 in the kidney, coupled with elevated blood levels of multiple pro-inflammatory cytokines, such as TNF-α, IL-6, and IFN-γ [
33]. CI (radiation+hemorrhage) activated STAT3 in the mouse ileum, a radiation-sensitive organ, on day 1 after the injury, whereas RI did not [
34].
In the present study, we found that both STAT1 and STAT3 were activated prior to the manifestation of kidney dysfunction (
Figure 2). However, they were activated distinctly. RI primarily increased STAT3-Y705 phosphorylation, while CI upregulated STAT1-Y701 phosphorylation in the mouse kidney cortex. In contrast to what was observed with X-ray intraperitoneal irradiation-induced nephropathy [
33], both RI and CI either significantly reduced or tended to reduce protein abundance of STAT1 and STAT3 (
Figure 2). Subsequently, these differential activations resulted in different expression patterns of inflammation genes in the kidney cortex. RI increased mRNA abundance of IFNγR1, IFNγR2, and heparanase, while CI primarily influenced heparanase (
Figure 3). Heparanase exacerbates inflammation by promoting cell migration, angiogenesis, the release of inflammatory mediators, and immune cell activation [
35,
36,
37]. Similarly, cytokines exhibited diverse responses, with RI significantly elevating IL-10 mRNA levels but CI reducing IL-4 mRNA abundance (
Figure 3). Both IL-10 and IL-4 are well known for their immune regulatory effects [
38,
39,
40].
It appears that RI induced a stronger immune response than CI, possibly influenced by the higher radiation dose used in RI (9.5 Gy compared to 9.0 Gy for CI) and differential activation of STAT3 and STAT1. STAT3 is known for its role in promoting inflammation, but it can also be involved in anti-inflammatory processes depending on the context. This dual role might allow STAT3 to more effectively initiate a strong but controlable immune response. In contrast, STAT1 is generally associated with promoting inflammation [
41,
42,
43]. Indeed, the increase in IL-10 mRNA levels following RI may indicate a regulatory response to the RI-induced inflammation, potentially contributing to the delayed renal function damage. It is known that activation of STAT1 inhibits expression of IL-4 [
44,
45]. The reduction of IL-4 mRNA following CI points to a suppression of anti-inflammatory signaling, which could exacerbate inflammatory damage. Since STAT1 could have transient effects on cytokine expression [
46,
47], it remains to be seen whether CI might induce pro-inflammatory cytokines at earlier time points. Our findings indicate that different mechanisms govern the inflammation between RI and CI, highlighting the specificity of their effects, which could influence long-term outcomes.
The role of p38 in inflammation is complex and context-dependent. The activation of p38 is generally regarded as pro-inflammatory, although in certain situations, p38 activation can have anti-inflammatory effects [
48,
49,
50]. Research indicates that p38 activation is linked to more severe radiation-induced injury. Male mice are more susceptible to such injuries than females, with males displaying increased p38 activation in the ileum following irradiation, a response not observed in females [
51]. Similarly, CI with hemorrhage induced activation of p38, whereas RI alone did not, in the ileum [
34]. Correspondingly, CI induced more pro-inflammatory cytokine expression and greater structural damage in the ileum than RI [
34]. Wang et al. demonstrated a causal effect of p38 on radiation-induced residual bone marrow injury by showing that inhibiting p38 in irradiated mice with SB203580 boosted the frequency of bone marrow hematopoietic stem and progenitor cells [
52]. We have found that RI reduced activation of p38, which was associated with a lack of expression of various pro-inflammatory cytokine observed in the ileum. This mechanism could add an explaination for the renal resilience to RI.
Cell Survival Pathways
Exposure to ionizing radiation can trigger apoptotic pathways in cells, leading to tissue damage in organs. Cells developed anti-apoptotic mechanisms such as activation of AKT and inhibition of p53 to inhibit apoptosis. AKT activation can promote cell survival and inhibit apoptosis by phosphorylating and inactivating pro-apoptotic proteins [
27,
30]. We previously demonstrated that in RI- and CI-induced-injuries in brain and ileum, activation of AKT was reduced associated with increased activation of caspase-3 [
57]. Ghrelin mitigated the injuries associated with increased activation of AKT and decreased activation of caspase-3 in the tissues [
57]. In the present study, we found that both RI and CI increased activation of AKT in the kidney (
Figure 5), suggesting a self-defensive mechanism for the delay in radiation-induced damage to the kidney compared to the brain and ileum.
p53 induces apoptosis by activating the transcription of pro-apoptotic genes, while repressing anti-apoptotic genes. In normal conditions, Mdm2 (Mouse double minute 2) sequesters p53 in the cytosol by binding to p53, preventing its transcriptional activity [
58,
59]. RI- and CI-induced brain hemorrhage reduced association of Mdm2 with p53, suggesting that more free form p53 was available in the brain [
57]. Ghrelin attenuated brain hemorrhage associated with increasing association of p53 with Mdm2 [
57]. We report that RI and CI reduced p53 protein abundance, another mechanism for inhibition of p53, in the kidney cortex (
Figure 5). It is important to point out that since inflammation, oxidative stress and apoptosis can intricately affect each other, HO-1 can inhibit inflammation and apoptosis, and activation of AKT and down regulation of p53 can inhibit inflammation and oxidative stress as well. Our research has suggested the complex interplay between inflammatory, oxidative and cell survival pathways in the kidney in early response to RI and CI.
Although wound alone did not significantly increase BUN and serum NGAL levels, it significantly increased phosphorylation of STAT1-Y701 and STAT3-Y705 (
Figure 2), and decreased IFNγR1 mRNA (
Figure 3) and catalase protein (
Figure 4) abundance in the renal cortex. Nonetheless, the significance of these findings remains unknown.
4. Methods and Materials
Animal and experimental design. All procedures were approved by the Institutional Animal Care and Use Committee of Uniformed Services University (MED-22-099). B6D2F1/J female mice (12 weeks old, approximately 20-26 g) purchased from Jackson Laboratory (Bar Harbor, ME) were maintained in a facility accredited by AAALAC in plastic microisolator cages with hardwood chip bedding and allowed to acclimate to their surroundings for at least 3 days prior to the study. Male mice were not used in this study because of potential problems associated with male mouse aggression, such as fight wounds which were not desirable during the experimental period. Previous injury studies [
57] also used female mice for this reason. As such, we continued to conduct this study with female mice so that data collected could be compared with previous ones. These mice were provided with commercial rodent chow (Rodent Diet #8604, Harlan Teklad, Madison, WI) and acidified tap water (pH=2.5-2.8)
ad libitum. Rooms holding animals were maintained at 22°C ± 2°C with 50% ± 20% relative humidity using at least 10-15 air changes/h of 100% conditioned fresh air with a 12-h 0600 (light) to 1800 (dark) full-spectrum lighting cycle. Mouse tails were tattooed for individual identification during acclimation. B6D2F1/J female mice were randomly divided into 4 groups: 1) sham (N=10); 2) wound (N=10); 3) Radiation (N=20) ; 4) Radiation+wound (N=20).The sham-irradiated animals were treated in the same manner as the irradiated animals but not exposed to the radiation source. The experiments were performed twice with a half number of mice in each experiment. However, some mice in both RI and CI groups did not survive to day 30, resulting in fewer than 20 mice remaining in these groups by the study's end. The mice were euthanized 30 days after exposure to irradiation by exsanguination under isoflurane, and cervical dislocations were performed as a secondary physical method.
Radiation. Animals were subjected to irradiation as previously described [
57]. Briefly, mice in the RI group received 9.5 Gy of
60Co gamma photons, delivered at 0.4 Gy/min in the Uniformed Services University
60Co facility. For the CI group, mice were exposed to 9.0 Gy of
60Co, delivered at 0.4 Gy min
-1 because other insults are known to increase radiation sensitivity and the use of 9.5 Gy in the CI group resulted in excessive lethality (data not shown). Sham mice and wounded mice were handled the same way, but remained in the staging room outside the
60Co facility. During irradiation, mice were kept awake in ventilated, acrylic plastic boxes with four compartments. The accuracy of the delivered doses was confirmed using an ionization chamber calibrated to measure the radiation dose to the midline soft tissue of the mice [
57].
Wounding. Within 1 hr after irradiation, animals were anesthetized with isoflurane, and a wound was created 19±1.3 mm from the occipital bone and between the scapulae, using a punch [
32]. Wounded and CI mice were injected i.p. with 0.5 ml saline containing 150 mg/kg acetaminophen (OFIRMEV injection, NDC 43825-102-01; Mallinckrodte Pharmaceuticals, Hazelwood, MO, USA) immediately after wounding to alleviate the pain. Sham mice and irradiated only mice received 0.5 ml saline i.p. The animal procedures are summarized in
Figure 7.
Blood urea nitrogen (BUN) assay. Blood was collected via a cardiac puncture under isoflurane on day 30. BUN levels in serum were measured using the diacetyl monoxime method, as previously described [
60,
61]. Briefly, serum samples were diluted tenfold with deionized water. Proteins in the samples were precipitated by adding a 0.61 M trichloroacetic acid solution, and then separated by centrifuging for 10 minutes at room temperature, at 10,000 g. The resulting supernatants were mixed with a chromogen and incubated for 5 min at 80°C, and absorption measurements were taken with a microplate reader (Elx800, BioTek, Winooski, Vermont, United States) at 490 nm. The chromogen was created by freshly mixing Reagent A and Reagent B in a 2:1 ratio. Reagent A was prepared by dissolving 10 mg of ferric chloride in 10 ml of 1.48 M phosphoric acid, which was then mixed with 60 ml of distilled water and 30 ml of 5.52 M sulfuric acid. Reagent B consisted of 50 mg of diacetyl monoxime with 1 mg of thiosmicarbazide dissolved in 10 ml of distilled water [
60,
61].
Serum neutrophil gelatinase-associated lipocalin (NGAL) assay. Serum NGAL was detected using a kit from Abcam (ab199083) according to the manufacturer’s protocol.
Western blot analysis. The kidney cortex was homogenized in a lysis buffer consisting of 10 mM triethanolamine, 250 mM sucrose, pH 7.4 plus a protease inhibitor cocktail tablet (Roche, Catolog # 11784500), 2mM of NaF (MilliporeSigma, Catalog# S6776) and 2 mM of Na
3VO
4 (MilliporeSigma, Catalog# 450243) to inhibit phosphatases. The homogenates were centrifuged at 13000 rpm at 4
0C for 8 min [
62]. The protein concentrations in supernatants were measured with the BCA assay (Thermo Scientific, Product# 23228 and 1859078). After assay, samples were dissolved in an SDS loading buffer for immediate use or frozen at -80
0C for future use. Equal quantities of protein samples (between 25-60 μg protein/lane) were loaded into 15 well or 17 well 4%–12% Bis-Tris gels (ThermoFisher, Catalog# NP0336BOX or NW04127BOX). After gel electrophoresis, the proteins were transferred onto nitrocellulose membranes (ThermoFisher, Catalog# LC2001). The membranes were soaked in blocking buffer (Odyssey, Part # 927-90001) at room temperature for 1 hour and the gels were stained with SimplyBlue
tm SafeStain (ThermoFisher, Catalog # LC6060) overnight to examine sample loadings. After soaking, the membranes were probed overnight with primary antibodies typically at 1:500 or 1:1000 dilution at 4°C. The membranes were then washed with PBS+0.1% Tween-20 and probed with an Alexa Fluor® secondary antibody for 1 hour at room temperature then examined with Odyssey, an infrared imaging scanner (Li-Cor). The data is presented without normalization, as stained gels were used to confirm equal sample loading. Reserving the first lane for protein markers, the maximum number of samples we could accommodate was 14 or 16. Protein samples were loaded into gels according to the randomly assigned numbers of mice, without any selection bias. If a significant difference in a protein abundance was observed in the first Western blot analysis, the experiment was repeated once with reproducible results presented. If there was no clear trend for a significant difference of protein levels in the first Western blot analysis, the experiment was stopped. The inconsistencies in sample numbers could arise from several factors. Occasionally, a bubble formed in the first well, or a distortion occurred in the last well, preventing their inclusion in the analysis. The data presented in the manuscript represents only a subset of the analyses we conducted. At times, we faced a shortage of samples for certain groups but proceeded with the Western blot analysis regardless. These results are included in the manuscript. The primary antibodies against proteins are listed as follows: Kidney injury molecule-1 (KIM-1, Invitrogen, PA520244), STAT1-Y701-P (Cell Signaling, 9167), STAT1 (Cell Signaling, 9176), STAT3-Y705-P (Cell Signaling, 9145S), STAT3 (Cell Signaling, 9139S), SHP-1-S591(ECM Biosciences, SP1531), SHP-1 (Santa Cruz, sc-287), p38-P (Cell Signaling, 9215S), p38 (Upstate, 05-454), MnSOD (Upstate, 06-984), Cu/ZnSOD (Upstate, 07-403), Peroxiredoxin 1 (PRDX) (Cell Signaling, 8732S), HO-1 (Cell Signaling, 43966S), p53 (Cell Signaling, 2524), AKT-S473 (Cell Signaling, 4060S), AKT (Cell Signaling, 2920S), GSK-3β-S9 (Cell Signaling, 9336), and GSK-3β (Cell Signaling, 9832).
qPCR. Quantitative PCR (qPCR) analysis of mRNA abundance were performed as described previously [
60]. Briefly, total RNA was isolated with the TRIzol kit (ThermoFisher, Ref# 15596026) and measured with NanoDrop 8000 Spectrophotometer (ThermoFisher). cDNAs were synthesized using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Part# 4368814). mRNAs were measured using the Fast SYBR Green Master Mix (Applied Biosystems, Ref# 4385612) in a Cycler 480 (Roche) and normalized to a housekeeping gene, L32. Fold difference was calculated as previously described [
60]. The primers used are listed in
Table 1.
Statistical analysis. Data are expressed as means ± standard errors. All analyses were normalized to the average value of the respective control groups. Shapiro-Wilk normality test and Q-Q plot revealed normal distributions of our data. Statistical analyses were performed using a non-paired two-tailed t-test with Microsoft Excel. Two-way ANOVA could not be performed because two different doses of radiation were used in the RI and CI groups: 9.5 Gy and 9.0 Gy, respectively. A p value less than 0.05 was considered significant.