目的：肾小管间质炎症(TI)是肾脏病的重要病理特征，可导致肾小管间质纤维化(TIF)。作为肾损伤最常见的原因之一，缺氧是导致TI的关键原因。缺氧引起的肾小管上皮细胞(TECs)损伤可以诱导肾小管间质巨噬细胞浸润。但缺氧TECs引起巨噬细胞活化的分子机制仍不十分清楚。低氧诱导因子-1(HIF-1)是调节机体低氧适应的关键转录因子。在肾脏中，表达于TECs的HIF-1α是低氧应激的主要调节因子。而TECs HIF-1α是否通过TECs-巨噬细胞的相互作用介导低氧诱导的TI尚不清楚。外泌体是细胞分泌的、大小在30到150nm之间的细胞外囊泡。其可以通过转移miRNA将信息传递到受体细胞。因此，缺氧导致的TECs HIF-1α能否通过外泌体转移miRNA促进TI形成需要进一步探讨。 氧依赖性的脯氨酰羟化酶(HIF-PHD)作为调节HIF的关键酶成为治疗相关疾病的新靶点。HIF-PHD抑制剂(HIF-PHI)主要通过模拟HIF-PHD缺氧后功能抑制，稳定HIF表达发挥作用。证据表明，HIF-PHI可以剂量依赖性的活化HIF-α。尽管在治疗肾性贫血的临床试验中，HIF-PHIs显示出良好的有效性和安全性，但由于反复或持续的HIF活化，仍存在一些潜在的非红细胞生成效应。研究显示，HIF-1α的持续活化可导致TIF。MK-8617(MK)是最近发现的一种口服生物活性的HIF-PHI，能有效刺激红细胞生成，但长期使用是否会通过持续诱导HIF-1α活化而对CKD小鼠TIF产生影响以及相关机制有待阐明。 本研究目的在于：通过一系列体内外实验探讨缺氧导致的TECs HIF-1α表达能否通过外泌体转移miRNA促进TI形成；并进一步探讨长期使用不同剂量HIF-PHI(MK-8617)诱导的HIF-1α活化对CKD小鼠TIF的影响及机制。 本研究包括四个部分： 第一部分：肾小管间质炎症与缺氧所致TECs HIF-1α表达和外泌体miRNA-23a的关系研究 方法：构建肾脏缺血／再灌注(I／R)损伤和单侧输尿管梗阻(UUO)模型。模型构建后的第1天、第3天和第7天处死小鼠，收集尿液、血清和肾脏标本进行肾功能分析，病理和免疫染色以及分子生物学实验。使用差速离心法提取肾组织和小管组织分泌的外泌体。将miRNA-23a inhibitor导入I／R损伤小鼠肾脏，24小时后处死，获取其尿液、血清和肾脏标本进行肾功能分析，病理和免疫染色以及分子生物学实验。 结果：在I／R损伤和UUO模型中，观察到肾小管间质中F4／80+巨噬细胞浸润在第1天和第3天明显增加；同时，肾组织中炎症细胞因子(单核细胞趋化蛋白-1、肿瘤坏死因子-α和白细胞介素-1β)的mRNA和磷酸化的NF-κB p65(p-p65)表达亦显著增加，在第7天则显著降低。肾脏中HIF-1α的mRAN和蛋白表达在第1天和第3天显著增加，在第7天则显著降低。同时，在I／R损伤和UUO肾脏外泌体中，miRNA-23a在第1天和第3天高度富集，而在第7天，其表达则显著降低。此外，I／R损伤小管分泌的外泌体亦富含miRNA-23a。在I／R损伤模型中，敲降miRNA-23a可以显著缓解肾组织中F4／80+巨噬细胞浸润，炎症细胞因子mRNA和p-p65的表达。 结论：TI与缺氧所致TECs HIF-1α表达和外泌体miRNA-23a具有时间相关性。 miRNA-23a对缺氧引起的，TI起至关重要的作用。 第二部分：HIF-1a介导的外泌体miRNA-23a促进TI形成的机制探讨 方法：使用差速离心法提取TECs分泌的外泌体。使用siRNA干扰抑制HIF-1α表达水平。使用染色质免疫沉淀-PCR(ChIP-PCR)探讨HIF-1α与miRNA-23a表达的关系。使用miRNA-23a inhibitor或mimic抑制或过表达外泌体miRNA-23a水平。采用荧光报告素酶实验检测miRNA-23a与A20之间的相互作用。将缺氧TECs分泌的外泌体和miRNA-23a敲降的外泌体注射到小鼠肾脏内(10 μg外泌体溶于60μlPBS中)，获取肾脏进行病理和免疫染色以及分子生物学实验。 结果：缺氧处理的TECs中HIF-1α，p-p65和外泌体miRNA-23a表达显著增高。 ChIP-PCR分析显示，缺氧的TECs HIF-1α与miRNA-23a启动子结合。与对照组外泌体相比，缺氧TECs分泌的外泌体可显著活化巨噬细胞。此外，缺氧TECs表达的miRNA-23a可以通过外泌体转移到巨噬细胞。使用miRNA-23a mimic或inhibitor过表达或敲降miRNA-23a，外泌体miRNA-23a过表达可显著增加巨噬细胞浸润、p-p65和炎症细胞因子mRNA的表达，该现象可以被miRNA-23a敲降的外泌体显著缓解。此外，我们证明外泌体miRNA-23a直接通过靶向抑制A20导致巨噬细胞活化。体内研究发现，肾组织中miRNA-23a主要定位于CD68+巨噬细胞和TECs中。当注射富含miRNA-23a的外泌体时，小鼠肾脏的巨噬细胞浸润、p-p65和炎症细胞因子mRNA的表达显著增加，而注射敲降miRNA-23a的外泌体则可明显缓解肾脏巨噬细胞浸润、p-p65和炎症细胞因子mRNA的表达。 结论：缺氧引起的TECs HIF-1α可以转录性表达miRNA-23a，后者通过外泌体转移至巨噬细胞，靶向抑制A20、激活NF-κB通路，活化巨噬细胞而引起TI。 第三部分：HIF-1α活化对CKD小鼠TIF的影响 方法： 5／6肾大部切除建立CKD小鼠模型，8周后，给予小鼠[DMSO／PEG400／水(5:40:55，v／v／v)]或MK-8617(一种新型HIF-PHI，1.5、5或12.5 ㎎／㎏／d)灌胃，灌胃12周后处死小鼠，获取其尿液、血液和肾脏进行肾功能分析，血红蛋白(Hb)水平检测，病理和免疫染色以及分子生物学实验。 结果：MK可以剂量依赖性的稳定HIF-α表达。HIF-1α活化对CKD小鼠肾功能具有浓度依赖性的双向作用。与对照组CKD小鼠相比，HIF-1α低度(1.5mpk)或中度活化(5mpk)小鼠的血清SCr、BUN和ACR显著降低，而HIF-1α高度活化(12.5mpk)小鼠则显著增加。此外，HIF-1α活化对CKD小鼠TIF具有浓度依赖性双向作用。与对照组CKD小鼠相比，HIF-1α低度(1.5mpk)或中度活化(5mpk)小鼠的肾脏细胞外基质蛋白α-SMA、I型胶原蛋白和纤连蛋白表达显著降低，而HIF-1α高度活化(12.5mpk)小鼠则显著增加。 结论：HIF-1α活化对CKD小鼠TIF具有浓度依赖性的双向作用，即HIF-1α低中度活化可缓解CKD小鼠TIF，HIF-1α过度活化则促进CKD小鼠TIF。 第四部分：肾小管上皮细胞HIF-1α过度活化促进CKD小鼠TIF的机制探讨 方法：采用基因组测序探讨MK干预HK-2细胞的mRNA基因表达谱模式特征，使用SOMs分析转录组数据表达特征。使用ChIP-PCR探讨HIF-1α与KLF5表达关系。使用KLF5 siRNA抑制KLF5表达水平。尾静脉注射以KLF5为靶标的shRNA慢病毒以建立KLF5基因敲降小鼠。给予KLF5基因敲降小鼠MK-8617 (12.5 ㎎／㎏／d)灌胃，灌胃10周后处死小鼠，收集其尿液、血清和肾脏进行肾功能分析，病理和免疫染色以及分子生物学实验。 结果：HIF-1α过度活化(500和1000nM)可显著促进HK-2细胞的α-SMA、I型胶原蛋白和纤连蛋白表达。通过全基因组测序分析，HIF-1α过度活化的HK-2细胞中KLF5基因表达明显上调。ChIP-PCR分析表明，HIF-1α过度活化的TECs HIF-1α与KLF5启动子结合。此外，HIF-1α过度活化(12.5mpk)d、鼠肾脏的TGF-β1表达亦明显升高，KLF5 siRNA可以通过抑制TGF-β1表达而减轻细胞外基质蛋白α-SMA、I型胶原蛋白和纤连蛋白表达。体内实验发现，KLF5基因敲降可显著缓解细胞外基质蛋白α-SMA、I型胶原蛋白和纤连蛋白表达。 结论：TECs HIF-1α过度活化通过激活HIF-1α-KLF5-TGF-β1信号通路促进CKD小鼠TIF。 全文结论： 1.缺氧是引起肾脏TI的重要原因，缺氧引起的TI与TECs HIF-1α和外泌体miR-23a有时间相关性，miRNA-23a在缺氧诱导的TI中发挥重要作用。靶向抑制miRNA-23是治疗肾脏缺氧损伤的新靶点。 2.肾脏缺氧损伤后，TECs HIF-1α通过调节外泌体miRNA-23a表达，促进巨噬细胞NF-κB通路活化而引起TI发生。 3.HIF-1α活化对CKD小鼠TIF具有浓度依赖性的双向作用。 4.HIF-1α-KLF5-TGF-β1信号通路激活是HIF-1α过度活化诱导TIF的重要机制。 本研究的创新之处： 1.发现缺氧引起的TI与TECs HIF-1α和外泌体miRNA-23a有时间相关性， HIF-1α通过调节外泌体miRNA-23a表达促进巨噬细胞NF-κB通路活化而引起TI发生。 2.首次阐明HIF-PHI MK-8617诱导的TECs HIF-1α活化呈现对CKD小鼠TIF浓度依赖性的双向作用。HIF-1α过度活化引起的CKD小鼠TIF加重与HIF-1α-KLF5-TGF-β1信号通路激活有关。 关键词：低氧诱导因子-1α；肾小管间质炎症；肾小管间质纤维化；miRNA-23a； Krüppel样因子5

 Objective: Tubulointerstitial inflammation (TI) is a critical pathological feature of kidney diseases and triggers the development of interstitial fibrosis (TIF). Renal hypoxia, which is one of the most common causes of kidney injury, is a key instigator of TI. Hypoxic TECs can facilitate TI by promoting macrophage. However, the molecular signals through which hypoxic TECs activate macrophages remain obscure. Hypoxia inducible factor-1 (HIF-1) is a key transcription factor mediating adaptive responses to hypoxia. In response to hypoxia in the kidney, HIF-1α is expressed predominantly in TECs and works as a master regulator of hypoxic stress. However, whether HIF-1α expressed in TECs mediates hypoxia-induced renal TI through TECs-macrophage interactions is a fundamental question for further study. Exosomes are small extracellular vesicles secreted by various cell types, with size ranging from 30 to 150 nm. Exosomes could transfer microRNA (miRNA) and thus shuttle information to target cells in the immediate vicinity of, or at a distance from, the parent cell. The exosomal transfer of miRNAs could be a novel mechanism for intercellular communication by exerting their regulatory effects on recipient cells. Therefore, whether HIF-1α in TECs promotes TI through exosomal miRNAs during hypoxia-related TI need further to be explored. The identification of oxygen-dependent prolyl hydroxylase domain (HIF-PHD) enzymes as regulators of HIF has led to the development of novel therapeutic agents for renal anemia. HIF-PH inhibitor (HIF-PHI) primarily functions by mimicking the hypoxia-driven expression of HIF. Convincing evidence suggests that HIF-α could be activated by HIF-PHI in a dose-dependent manner. Although HIF-PHIs were well-tolerated in clinical trials to treat CKD patients with anemia, there are several safety concerns related to the potential non-erythropoietic effects due to repeated or persistent HIF activation. However, the effects of HIF activation induced by HIF-PHI on TIF in CKD remain unclear. MK-8617 (MK), a recently identified, selective, orally bioavailable HIF-PHI, actively stimulates erythropoiesis. However, the effects and mechanisms of HIF-1α activation, which is induced by MK administration with long time, on TIF in CKD mice need to be elucidated. The present study aimed to investigate whether HIF-1α expressed in TECs promotes TI through exosomal miRNAs during hypoxia-related TI. Furthermore, the effects and mechanisms of HIF-1α activation, which is induced by MK administration with long time, on TIF in CKD mice were explored. The study Includes four parts: Part one: The study on relationship between TI and hypoxia-induced renal tubular epithelial cells HIF-1α and exosomal microRNA-23a Method: Bilateral ischemia/reperfusion (I/R) injury and unilateral ureteral obstruction (UUO) mice models were estabilshed. Mice were killed on days 1, 3, or 7 after I/R injury or UUO under general anesthesia, and their urine, serum and kidneys were harvested for renal function analysis, pathologic and immunohistochemistry staining and molecular biological analysis, respectively. Exosomes from hypoxic kidney and hypoxic tubules were isolated using differential centrifugation. The miRNA-23a inhibitor was transfected into I/R injury kidneys using the in vivo-jetPEI. Mice were killed 24 hours after injected, and their urine, serum and kidneys were harvested for renal function analysis, pathologic and immunohistochemistry staining and molecular biological analysis, respectively. Results: In I/R injury and UUO models, renal TI was observed on days 1, 3, and 7 associated with an increase in F4/80+ macrophages infiltration. Concomitantly, there were significant increases in mRNA expression of renal inflammatory cytokines (MCP-1, TNF-α, and IL-1β) and phosphorylation of NF-κB p65 (p-p65) at days 1 and 3,preceding a significant reduction at day 7. The expression of HIF-1α was significantly increased in the injured kidney at days 1 and 3, preceding a significant decrease at day 7. The hypoxic kidney secreted exosomes that were highly enriched in miRNA-23a at days 1 and 3, receding at day 7. Interestingly, tubular exosomes were highly enriched in miRNA-23a in I/R injured kidneys and kidneys with UUO at day 1 and day 3, respectively. Meanwhile, miRNA-23a inhibition represses F4/80+ macrophages infiltration, renal inflammatory cytokines and p-p65 expression in hypoxia-induced kidney injury. Conclusion: There are close temporal correlation between the TI and hypoxia induced TECs HIF-1α and exosomal miRNA-23a expression level. miRNA-23a plays a critical role in hypoxia-induced TI. Part two: The mechanism of TI induced by HIF-1α-mediated exosomal miRNA-23a Method: Exosomes from TECs were isolated using differential centrifugation. HIF-1α siRNAs were used to regulate the HIF-1α levels. The role of HIF-1α on miRNA-23a expression was studied using Chromatin immunoprecipitation-PCR (ChIP-PCR). Besides, exosomal miRNA-23a was inhibited or overexpressed using miRNA-23a inhibitor or miRNA-23a mimics. The luciferase reporter assay was performed to test the interaction between the miRNA-23a and 3'-UTR of A20. In vivo, TECs exosomes silenced miRNA-23a was transferred to mice via kidney parenchyma injection (10μg in 60μl of PBS). Their kidneys were harvested for pathologic and immunohistochemistry staining and molecular biological analysis. Results: Hypoxic TECs presented with higher HIF-1α and upregulation of p-p65 as well as exosomal miRNA-23a. ChIP-PCR assay showed that the binding of HIF-1α was enriched in the miRNA-23a promoter in TECs by hypoxia stimulation. Exosome from hypoxic TECs could develop more severe macrophage activation compared to exosome released from normal TECs. Furthermore, miRNA-23a secreted from hypoxic TECs can be transferred to macrophages via exosomes. More interestingly, exo(Hypo)-miRNA-23a-mimic increased the release of inflammatory factors from macrophages, an effect that was abrogated by exo(Hypo)-miRNA-23a-inhibitor. Furthermore, we demonstrated that exosomal miRNA-23a directly suppressed its target A20, leading to macrophage activation. Inhibition of miRNA-23a reversed macrophage activation. In vivo, the uptake of exo(Hypo) by intrarenal CD68+ macrophages, and infiltration of these cells was significantly increased. And miRNA-23a predominantly located in the cells of CD68 staining macrophages. F4/80+ macrophages infiltration, inflammatory cytokines and p-p65 expression of kidney were significantly increased when mice transferred with miRNA-23a enriched exosomes, which was not shown by miRNA-23a silenced exosomes. Conclusion: Transcriptionally regulated miRNA-23a by TECs HIF-1α, which can be transferred to macrophages via exosomes, could promote NF-κB signaling through inhibiting A20, activating macrophages to induce TI. Part three: The effect of HIF-1α activation on TIF in CKD mice Method: CKD mice models were estabilshed in mice by 5/6 nephrectomy (5/6Nx). After 8 weeks, the CKD mice were dosed with vehicle [DMSO/PEG400/water (5:40:55, v/v/v)] or MK (a novel HIF-PHI, 1.5, 5 or 12.5 ㎎/㎏/d in vehicle) for 12 weeks. Mice were killed on week 20 after 5/6Nx under general anesthesia, and their urine, blood and kidneys were harvested for renal function analysis and molecular analysis, pathologic and immunohistochemistry staining and molecular biological analysis, respectively. Results: Western blotting showed that HIF-α could be stabilized by MK dose-dependently. HIF-1α activation has concentration-dependent biphasic effects on kidney function. Compared to vehicle-administered CKD mice, serum creatinine, blood urea nitrogen and albuminuria are significantly decreased in mice with HIF-1α mild activation (1.5mpk) or HIF-1α moderate activation (5mpk) and increased in those with HIF-1α excessive activation (12.5mpk). HIF-1α activation has concentration-dependent biphasic effects on TIF in vivo. Notably, compared to vehicle-administered CKD mice, α-SMA, collagen-I and fibronectin was significantly attenuated in mice with HIF-1α mild activation (1.5mpk) or HIF-1α moderate activation (5mpk) and increased in those with HIF-1α excessive activation (12.5mpk). Conclusion: HIF-1α activation has concentration-dependent biphasic effects on TIF. TI was significantly attenuated in mice with HIF-1α mild or moderate activation and markedly increased in those with HIF-1α excessive activation. Part four: The mechanism of TECs HIF-1α excessive activation promoting TIF in CKD mice Method: RNA-sequencing was used to characterize the expression patterns of mRNA genes in HK-2 cells treated with MK. Self-organizing maps (SOMs) were intuitively used to analyze the structure and interrogate transcriptome data. The role of HIF-1α on KLF5 expression was studied using ChIP-PCR. Besides, KLF5 was inhibited using siRNA to explore its functional role in TECs. Furthermore, the KLF5 knockdown mice were established by Lentiviruses expressing short-hairpin RNAs (shRNAs) targeting KLF5. These mice were administrated with MK (12.5㎎/㎏/d) for 10 weeks. Mice were killed under general anesthesia, and their urine, blood and kidneys were harvested for renal function analysis, pathologic and immunohistochemistry staining and molecular biological analysis, respectively. Results: As indicated by the levels of α-SMA, collagen-I and fibronectin, fibrogenesis was promoted in HK-2 cells with HIF-1α excessive activation (500 and 1000 nM). Intriguingly, Krüppel-Like Factor 5 (KLF5) expression appeared to be markedly upregulated in HK-2 cells with HIF-1α excessive activation by genome-wide sequencing analysis. ChIP assay showed that the binding of HIF-1α was enriched in the KLF5 promoter in TECs with HIF-1α excessive activation. The role of KLF5 in fibrogenesis was validated in in vivo and in vitro. Furthermore, TGF-β1 is significantly increased in mice with HIF-1α excessive activation (12.5mpk). KLF5 stimulates α-SMA, collagen-I and fibronectin expression by upregulating TGF-β1. Knockdown of KLF5 markedly diminished the increase of α-SMA, collagen-I and fibronectin expression in TECs. And KLF5 knockdown also diminished the increase of α-SMA, collagen-I and fibronectin expression induced by HIF-1α excessive activation in CKD mice. Conclusion: Excessive activation of HIF-1α in TECs could promote TIF in CKD mice by activating HIF-1α-KLF5-TGF-β1 signalling. Conclusions for the full text: 1.Hypoxia is a key instigator of TI. There are close temporal correlation between the TI and hypoxia induced TECs HIF-1α and exosomal miRNA-23a expression level. miRNA-23a plays a critical role in hypoxia-induced TI. Interrupting this broad, vertically integrated miRNA-23a signaling cascade represents a promising therapeutic target for hypoxia-induced TI. 2.We have discovered an important role for tubular HIF-1α in inciting TI in the hypoxic kidney via exosomal miRNA-23a mediated intercellular communication between TECs and macrophages. 3.HIF-1α activation has concentration-dependent biphasic effects on TIF. 4.Excessive activation of HIF-1α could promote TIF by activating HIF-1α-KLF5-TGF-β1 signaling. Innovations of this study: 1.We find that there are close temporal correlation between the hypoxia-induced TI and TECs HIF-1α and exosomal miRNA-23a expression. Transcriptionally regulated miRNA-23a by TECs HIF-1α, which can be transferred to macrophages via exosomes, could promote NF-κB signaling activation in macrophages to induce TI. 2.We firstly demonstrate that TECs HIF-1α activation, which is induced by HIF-PHIMK-8617, has concentration-dependent biphasic effects on TIF in CKD mice. The pro-fibrotic effect of HIF-1α excessive activation on TIF is mediated by activating HIF-1α-KLF5-TGF-β1 signaling. Keywords: hypoxia inducible factor-1α; tubulointerstitial inflammation; tubulointerstitial fibrosis; miRNA-23a; Krüppel-like factor 5