Introduction

Obstructive sleep apnea hypopnea syndrome (OSAHS) is the most common type of sleep-disordered breathing associated with excessive daytime sleepiness, snoring, recurrent oxyhemoglobin desaturations, and arousals from sleep [1]. It is characterized by recurrent episodes of airway collapse during sleep. The prevalence of OSAHS in children in the clinic is not unusual; an epidemiological survey in 2007 showed that the prevalence of OSAHS in children was 2% [2]. Another epidemiological survey in Hong Kong in 2010 showed that the OSAHS prevalence rate among school-age children was at 3.8% for girls and 5.8% for boys [3]. Children are a susceptible group, as they are in growth and development period; their tissues and organs demand more oxygen than adults, therefore they are more sensitive to hypoxia. So, the impact of OSAHS on children is of more importance; it can occur in newborns till adolescents at any stage, but more commonly in preschool children.

OSAHS involves multiple system dysfunctions such as the cardiovascular, respiratory, digestive, genitourinary, endocrine, vascular, neurodevelopmental, and muscular system and also participates in retinopathy of prematurity and impaired growth [4,5,6,7,8,9,10,11]. In our previous studies, we found that nervous system injury and cognitive dysfunction were more common in children with OSAHS [12, 13]. There are ample evidence that inflammatory responses to the perturbations associated with OSAHS trigger a variety of genes and signaling cascades that ultimately lead to end-organ injury and changes in the kidney function and protein expression. It can also cause renal damage which has been clinically confirmed [14, 15]. The kidney is a high blood flow, high-perfused organ; its oxygen supply and oxygen tension changes are more sensitive, thus vulnerable to hypoxic injury [16]. Renal function, proteinuria, and renal tubular dysfunction are mainly nocturnal, but its molecular mechanism remains unclear. Repeated hypoxia and re-oxygenation, similarly to ischemia and reperfusion, causes cell mitochondria to produce more reactive oxygen species (ROS), therefore resulting in oxidative stress which is an important pathophysiological mechanism of renal damage caused by OSAHS. The aim of this study was to test the hypothesis that oxidative stress played an important role in the pathogenesis of CIH-associated renal damage in growing rats.

Materials and methods

This study was authorized by the Ethics Committee of Wenzhou Medical University. SPF Sprague-Dawley (SD) rats were purchased from Experimental Animal Center of Wenzhou Medical University. The design, preparation of intermittent low oxygen stainless steel chamber, compressed oxygen (concentration > 99.5%), high purity compressed nitrogen (concentration > 99.99%), and compressed air were purchased from Wenzhou Medical oxygen plant filling.

Animal model of OSAHS and experimental groups

Animal grouping of SPF SD male rats, weighing between 90 and 110 g, and aged 3–4 weeks; 40 rats were randomly divided into 4 groups by the method of random number table so that the number in each group was 10 and named as follows: chronic intermittent hypoxia (IH) for 2 weeks and 4 weeks group (2IH, 4IH) and control group 2 and 4 weeks (group 2C, 4C). Rats excluding the control groups were laid in intermittent hypoxia cabin, an automated alternate nitrogen/oxygen gas delivery system (Scientific research center of Wenzhou Medical University, Zhejiang, China) to deliver hypoxia/re-oxygenation. There were two durations of chronic intermittent hypoxia (CIH) which were studied: 2 weeks (2IH) and 4 weeks (4IH).

Intermittent hypoxia exposure

Establishment of CIH animal model was according to Wang Y [17] with modifications. To generate IH and air control, a steel cabin was created with an automated nitrogen/oxygen gas delivery system to deliver hypoxia/re-oxygenation using our previously described protocol [18]. Briefly, the experimental parameters were set as follows: O2 concentration could be reduced to a nadir of 9% ± 1.5% in 30 s by infusion of 99.99% nitrogen with the pressure kept at 0.3 KPa, stabilized at that level for 30 s, and then gradually increased to 21.0 ± 0.5% over the next 12 s by infusion of 99.50% oxygen (25 L/min) into the cabin. The oxygen concentration in the tank was maintained at 9% ± 1.5% in the hypoxia phase and in the re-oxygenation was maintained at 21.0% ± 0.5%, and the concentration of CO2 in the tank was less than 0.01%.This process was computer controlled. This cycle was repeated every 90 s over 7.5 h (from 8:00 to 15:30) during the animals’ diurnal sleep period for certain days according to the experimental design. The control groups were placed in cabin filled with compressed air for 2 weeks as 2C group or for 4 weeks as the 4C group, respectively. The O2 concentration was kept at 21.0 ± 0.5% in the control cabin. Ambient temperature was kept at 22–24 °C and humidity 40~50%. At the end of each day, the rats were grouped in additional cages, which were illuminated with fluorescent lamps to simulate daytime conditions. The rat’s activity and diet were ad libitum.

Test of the chronic intermittent hypoxia cabin

Validation of the CIH cabin was carried out before this experiment. Ten rats were anesthetized with 35 mg/kg pentobarbital (Sigma, USA) through intraperitoneal injection. The carotid artery was then catheterized using a catheter with heparin anticoagulation, which was inserted in the left carotid artery and sutured in place. When the rats recovered from surgery, 5 rats were randomly chosen and put individually in the CIH cabin and the other 5 in the control cabin. The experimental protocol was performed for 2 hours. The blood samples were collected at 22.5-s intervals during a single IH cycle continuous blood for 5 times, every <3 s, with the initial nitrogen gas input as the first sample, respectively. Arterial blood samples (0.5 ml) were collected in a 5-gauge needle at the end of each consecutive condition and immediately analyzed using a blood gas analyzer (GEM Premier 3000; America).

Specimen collection

Four rats in each group were anesthetized; thoracotomy with separation of the heart was performed to collect 5 mL of arterial blood which was refrigerated in a temperature of 4 °C for 2 h. Subsequently, the arterial blood was centrifuge at 4 °C 4000rpm for l5 min and the serum was preserved at − 80 °C in refrigerator. The serum was then used to detect total superoxide dismutase (SOD) activity. The remaining 6 rats were afterwards anesthetized by intraperitoneal injection of 3% phenobarbital sodium (40 mg/kg). Laparotomy was done to expose the left and right kidneys. The renal vein was injected with pre-cooled saline. The right and left kidneys were collected when it became pale. The left kidney was placed on an ice plate. The right kidney was fixed in 4% paraformaldehyde solution.

Renal histopathological staining

HE staining, PAS staining, and Masson staining

The tissue sections were paraffin-embedded and 4-μm sections were used. The sections were routinely dewaxed with xylene, washed with ethanol at all levels, and stained with HE, PAS, and Masson, respectively. Tissue samples from the kidneys were scored histopathologically. Pathological scoring (0–4) was used to assess the degree which was defined as glomerular swelling, renal tubular epithelial cell swelling, mesangial proliferation, and glomerular and interstitial fibrosis. Counts were performed in at least 10 different fields of square micrometers, using scores on a scale of 0 (< 5%), 1 (5–25%), 2 (25–50%), 3 (50–75%), and 4 (> 75%). The severity of each injury was assessed by 0~4 scores: 0 = minor damage, 1 + = mild damage, 2 + = moderate damage, 3 + = severe damage, and 4 + =serious damage. All the above four pathological scores were added into 0~16 points.

Real-time PCR

The expression of HIF-1α, Cu/Zn-SOD, and MnSOD mRNA in the left kidney was detected by real-time PCR. Total RNA was extracted from the renal tissue using the Trizol reagent according to manufacturer’s instructions. The RNA was purified and quantified by RNAi, reverse transcribed into cDNA, and amplified using a PCR amplification apparatus. PCR reaction conditions were 95 °C 5 min; 95 °C 10 s, 60 °C 10 s, 72 °C 10 s, and 45 cycles of amplification. β-actin was used as an internal control by previous work [12]. β-actin, HIF-1α, Cu/Zn SOD, and MnSOD primer sequences were designed and synthesized by Shanghai Shengong Bioengineering Co. Ltd., and the sequences were as follows: HIF-1 α: upstream 5′-TGAACATCAAGTCAGCAACG-3′, downstream: 5′-CACAAATCAGCACCAAGCAC-3′, Cu/ZnSOD: upstream 5′-GTGGTGGAGAACCCAAAGGA-3′, downstream 5′-GCGTGCTCCCACACATCAAT-3′, MnSOD: upstream 5′-ATGGGGACAATACACAAGGC-3′, downstream 5′-TCATCTTGTTTCTCGTGGAC-3′, β-actin: upstream 5′-TCACCAACTGGGACGATATG-3′, and downstream 5′-GTTGGCCTTAGGGTTCAGAG-3′. The relative expression of the target gene was calculated according to the formula 2−ΔCt (ΔCT = Ct value of the target gene−Ct value of the internal reference gene) using β-actin as the internal reference gene using Lightcycler48015.0 software.

Determination of SOD

The hydroxylamine oxidation method was used to detect the activity of SOD. All experimental procedures were performed according to the manufacturer’s instructions (kit assay, Jiancheng Limited Company, Jiangsu China). UV spectrophotometric colorimetry was used to detect the absorbance of the sample at 550 nm. Enzyme values are presented as U/ml.

Statistical analysis

In addition to the renal pathological score, all groups were normal measurement data; with the mean + standard deviation (\( \overline{X} \) ±SD), the renal pathological scores were expressed as a median, with SPSS 21 statistical software processing. Renal pathological scores were examined by Kruskal-Wallis test. Other data were compared using multiple factor analysis of variance (two-way ANOVA), if the variance was homogeneous the LSD test was used, and if not, Dunnett’s T3 test was used to test the variance. The differences were considered significant if P values were < 0.05.

Results

Blood gas analysis

Blood gas analysis in rats was made in order to validate the CIH cabin. The result showed that over the course of the IH event for 2 h, the PaO2 fluctuated from 44 ± 3 to 80 ± 9 mmHg and SaO2 from 75 ± 3 to 95 ± 1 mmHg, respectively, in a cycle of 90 s. The PaO2 and SaO2 in control group exhibited no significant difference among the five time points. The magnitude of oxygen saturation that was induced in our model was consistent with the degree of hypoxia that occurs in moderate to severe OSAHS [19, 20].

Histopathological changes in each group

HE

The structure of glomerular and tubular epithelial cells was normal in 2C and 4C groups. There was mild hyperplasia of glomerular mesangial cells in IH groups. Compared to the 2IH group, the 4IH group had significant changes such as glomerular and renal cysts and renal tubular epithelial cell swelling (Fig. 1).

Fig. 1
figure 1

HE staining of renal tissue in each group (× 400 times). a and c Represent the 2C and 4C groups. b and d Represent the 2IH and 4IH groups. b and d The white arrows show mild glomerular mesangial hyperplasia and black arrows show swelling and structural disorders of renal tubules

Full size image

PAS staining

There was no significant thickening of the glomerular basement membrane, mesangial and no loss of the tubular epithelial cell brush border in the 2C and 4C groups. In the IH groups, glomerular basement membrane was slightly thickened, the brush border structure of renal tubular epithelial cells were incomplete, and the damage in the 4IH group was significantly higher than that of the 2IH group, as shown in Fig. 2.

Fig. 2
figure 2

PAS staining of renal tissue in each group (× 400 times). a and c Represent the 2C and 4C groups. b and d Represent the 2IH and 4IH groups. b and d The black arrows show glomerular mesangial proliferation

Full size image

Masson staining

No blue collagen fibers were identified in the glomerular and tubulointerstitial tissues neither in the IH groups nor the C groups as shown in Fig. 3.

Fig. 3
figure 3

Masson staining of renal tissue in each group (× 200 times). a and c Represent the 2C and 4C groups. b and d Represent the 2IH and 4IH groups. a, b, c, d Glomerular and renal tubules were not stained blue for collagen fibers

Full size image

Pathological score

In the four groups (n = 6), it showed that the IH group (4IH group was 5, 2IH 3) had more severe pathological damage compared to air control group (2C and 4C group were 0.5) and the difference was statistically significant (P < 0.05).The pathological damage of the 4IH group was more obvious than that of the 2IH group (P < 0.05), and the pathological damage was serious; the difference was statistically significant (P < 0.05) (refer to Table 1).

Table 1 Analysis of the histological assessment among 4 groups (n = 6)
Full size table

The mRNA expression of HIF-1α, MnSOD, and Cu/ZnSOD in the kidney tissues of each group was observed

Effect of IH on expression of HIF-1α (Fig. 4 and Table 2)

IH had significant effect on the expression of HIF-1αmRNA (F = 151.683. P < 0.001) and the time effect (F = 42.693, P < 0.001). The interaction between the treatment and the time effect was also significant (F = 52.212, P < 0.001). The expression of HIF-1αmRNA in the renal tissue of 2IH group was increased (P < 0.05) compared to the 2C group, similarly for the 4IH group compared to the 4C group (P < 0.05). The expression of the 4IH group was significantly higher than the 2IH group (P < 0.05).

Fig. 4
figure 4

The level of HIF-1α, Cu/ZnSOD, and MnSOD mRNA expression. a, b, c, d Real-time quantitative PCR was used to detect the expression of HIF-1α, Cu/ZnSOD, MnSOD, and SOD in kidney mean±SD. Statistically significant differences are indicated by ▲P < 0.05 vs 2C group; ■P < 0.05vs 4C group; ★P < 0.05 vs 2IH group

Full size image
Table 2 The relative expression of HIF-1α, MnSOD, and Cu/ZnSOD mRNA (n = 6) and serum SOD (n = 4) in the kidney of each group were compared. (\( \overline{\mathrm{X}} \)±SD)
Full size table

Effect of IH on expression of MnSOD (Fig. 4 and Table 2)

IH had significant effect on the expression of MnSOD mRNA (F = 108.613 P < 0.001). The time effect (F = 1.390, P > 0.05) and the interaction between treatment and the time effect was not significant (F = 2.551, P > 0.05). The expression of MnSOD mRNA of the 2IH was lower than that of 2C (P < 0.05); likewise, the 4IH group compared to 4C was significantly decreased (P < 0.05), but the 4IH compared to 2IH had no significant difference (P > 0.05).

Effect of IH on expression of Cu/ZnSOD (Fig. 4 and Table 2)

IH had significant effect on the expression of Cu/ZnSOD mRNA (F = 107.511 P < 0.001). The time effect (F = 3.912, P > 0.05) and the interaction between treatment and the time effect was not significant (F = 0.676, P > 0.05). The expression of Cu/ZnSOD mRNA in 2IH and 4IH groups were lower than those in 2C and 4C groups (P < 0.05), but there was no significant difference between 4IH group and 2IH group (P > 0.05) and 2C and 4C (P > 0.05).

Determination of serum SOD activity (Fig. 4 and Table 2)

2IH and 4IH groups’ serum SOD activity were lower than the control groups 2C and 4C (P < 0.05); 4IH group was significantly lower than 2IH group (P < 0.05); 2C and 4C groups were not statistically significant (P > 0.05).

Discussion

OSAHS is a multi-system functional disorder with intermittent hypoxia and sleep fragmentation as its main pathophysiological mechanism resulting in nocturnal apnea. The process of hypoxia-re-oxygenation is similar to the pathological process of ischemia-reperfusion injury (IRI), which can result in oxidative stress and free radical production. Long-term accumulation leads to multi-system organ damage. In the study of the molecular mechanism of cardiovascular disease and cognitive impairment in OSAHS, oxidative stress is considered to be one of the major mechanisms of injury [21,22,23]. In IRI, oxidative stress is the most common molecular mechanism of cell destruction and relatively low concentration of antioxidant enzymes makes it more susceptible to oxidative stress injury [16].

OSAHS-associated renal damage has been clinically confirmed [24, 25] mainly for chronic kidney disease, nocturnal polyuria, renal functional changes, proteinuria, renal tubular dysfunction, etc. In animal model of CIH [26, 27], it has been found that OSAHS can cause kidney tissue structure, ultrastructure, and proteomics changes, but the studies were performed on adult animals. It is still unclear whether OSAHS can cause similar changes in the children; therefore, we design a self-developed computer-controlled intermittent hypoxic oxygen chamber to study the effects of OSAHS on renal tissue in 3- to 4-week-old SD rats. The test for the hypoxia cabin confirmed that the magnitude of oxygen saturation that was induced in our model of IH was equivalent to the pathophysiological changes of moderate to severe intermittent hypoxia exposure.

In our study, we did not investigate mild OSAS but Butchner et al. [28] and Wissing et al. [29] investigated mild to moderate OSAS and its treatment on renal hemodynamics assessed by the renal resistance index. They found in multivariate analyses that the renal resistance index was independent of hypertension, diabetes mellitus, age, and baseline renal function. Their study demonstrated an impairment of renal hemodynamics in OSAS. These changes in renal blood flow may identify OSAS patients who are at risk of declining renal function. Further studies are warranted to determine OSAS’s direct influence on renal impairment in children.

Previous studies demonstrated that proximal tubule epithelial cells were one of the most vulnerable cells in IRI in the kidney [30, 31]. In this study, there were pathological changes of renal tissue; mainly in the IH group, mild hyperplasia of glomerular mesangial cells, edema of the renal tubular epithelial cells, and loss of complete brush border structure were identified. At the same time, glomerular changes were confirmed by PAS staining. Chronic hypoxia is a key factor in renal interstitial fibrosis. In this study, renal tissue fibrosis was not identified in neither 2 weeks nor 4 weeks IH group in Masson staining, but Sun et al. [32] and other studies in adult mice showed that after 8 weeks of IH exposure, there were significant decrease of antioxidant levels and significant increases of renal inflammation, oxidative damage, cell death, and renal fibrosis, suggesting the severity of renal fibrosis by IH depending on the length of time and the degree of hypoxia.

The mechanism of hypoxia in OSAHS is similar to that of IRI, which is characterized by CIH and different durations of hypoxia, which is a more serious type of hypoxia. Gozal et al. exposed (pheochromocytoma-12, PC12) pheochromocytoma cells to continuous hypoxia (5%O2) and intermittent hypoxia (hypoxia 5%O2 35 min, 21% O2 21 min) 2 to 4 days; they found that 2 days exposure to IH can lead to cell apoptosis [33]. HIF-lα is a transcription factor that regulates oxygen balance. It is a heterogeneous two polymer structure composed of oxygen-sensitive alpha subunit and stably expressed beta subunit. The content of HIF-1α in the local tissue can indirectly reflect the hypoxia of tissue cells. Continuous hypoxia increased the expression of HIF-1α through mediating hypoxic adaptive response. IH also upregulated HIF-1α expression, but the signal transduction pathway was different from that of continuous hypoxia [34]. Da Rosa et al. [35] found that HIF-1 alpha protein was upregulated in lung tissue and liver in CIH. Therefore, the expression of HIF-1 alpha mRNA in renal tissue can reflect the renal anoxia in CIH condition. In this study, real-time quantitative PCR was used to detect hypoxia-related HIF-1αmRNA expression in renal tissue, which showed that the expression of HIF-1αmRNA was upregulated in a time-dependent manner, therefore indicating that hypoxia occurred in the kidney tissue.

A number of clinical studies show a decrease in oxidative stress [21, 36, 37] and anti-oxidative capacity in patients with OSAHS [38]. SOD is the first in vivo antioxidant enzyme reaction. Cu/Zn superoxide dismutase (copper-zinc superoxide, dismutase, Cu/ZnSOD) and manganese superoxide dismutase (manganese superoxide, dismutase, MnSOD) are the main two types. Cu/ZnSOD was mainly expressed in the cytoplasm and MnSOD was mainly located in the mitochondria. In our previous study, we found out that the level of 8-ISO-PGF2α, an in vivo oxidative stress marker, was significantly high in the CIH groups compared to the corresponding control groups, therefore suggesting that CIH induced oxidative stress injury [18]. In this study, the total SOD activity in serum of CIH rats was determined by chemical colorimetry; we found that SOD activity was decreased, which was negatively correlated with time, and the results showed the antioxidant capacity decreased in CIH SD rats. Real-time fluorescence quantitative detection showed that the expression of Cu/ZnSOD and MnSODmRNA in renal tissue was downregulated, which indirectly indicated that IH induced oxidative stress in renal tissue. Nanduri et al. [39] also showed that PC12 cells were mainly downregulated by IH after MnSOD expression, whereas other antioxidant enzymes were unchanged. These results suggest that IH exposure mainly results in downregulation of MnSOD expression. In this study, the expression of Cu/ZnSOD in the cytoplasm of IH group was lower than the control group. The hypoxia time was prolonged and the expression of MnSOD in the mitochondria was downregulated in CIH rats, which indicated oxidative stress played a critical role in renal damage CIH rats.

Our study had limitations. Firstly, our model of intermittent hypoxia does not entirely represent all of the events that occur during obstructive sleep apnea. Intermittent hypoxia does not cause negative intrathoracic pressure swings or obstruction of the airway. Second, we used 2 weeks and 4 weeks exposure to intermittent hypoxia which cannot predict if the changes in mRNA expression we found would be maintained or augmented with a longer exposure. Third, we only analyzed gene but it would have been better to analyze protein also in the kidney tissue. By immunohistochemistry, we could confirm that Cu/ZnSOD is mainly expressed in the cytoplasm, while MnSOD is mainly expressed in the mitochondria. Lastly, to better evaluate the kidney morphological changes, more stained sections should have been done.

Conclusion

In summary, oxidative stress played a critical role in renal damage by up regulating HIF-1α transcription and downregulating Cu/ZnSOD and MnSOD transcription after chronic intermittent hypoxia exposure in growing rats. Future research is needed to determine the clinical significance of these preliminary findings.