Chronic renal failure (CRF), characterized by glomerulosclerosis and interstitial fibrosis, is the common end stage of all kinds of chronic kidney diseases (CKDs). CRF is usually associated with high mortality and the progression of CKD to CRF cannot be effectively intervened in clinic (Woo et al., 2015). Renal fibrosis, particularly tubulointerstitial fibrosis, is a typical feature in the progression of CKD which contributes to permanent loss of renal function (Iwano and Neilson, 2004; Inoue et al., 2015). Transforming growth factor-β1 (TGF-β1), an extensively studied profibrogenic cytokine, plays an important role in renal fibrosis in CKD (Wang et al., 2005; Lopez-Hernandez and Lopez-Novoa, 2012). It promotes fibrosis through several actions including increasing extracellular matrix (ECM) synthesis, decreasing ECM degradation, activating resident myofibroblast and inducing epithelial to mesenchymal transition (EMT) and inflammation (Lopez-Hernandez and Lopez-Novoa, 2012).

Mitochondria are essential eukaryotic cells organelles responsible for numerous important physiological processes such as energy generation, ROS production and cellular apoptosis (Newmeyer and Ferguson-Miller, 2003). Thus, mitochondrial dysfunction is detrimental to cells and organs because of the lack of energy supply and excessive ROS generation. As an organ with great demand of energy, kidney is abundant in mitochondria and mitochondrial dysfunction is one of the major factors in the pathogenesis of renal injury (de Cavanagh et al., 2007; Yuan et al., 2012). Clinical study also found that mitochondrial respiratory system was impaired in CKD patients (Granata et al., 2009). Treatments that improve mitochondrial functions have been proved to be beneficial for renal injury attenuation (Hui et al., 2017; Thomas et al., 2017; Zhao et al., 2017). Therefore, mitochondria may be a potent target for CKD treatment.

As a kind of effective antidiabetic drug, PPARγ agonists not only conquer diabetic nephropathy, but they are also considered potential therapeutic agents in non-diabetic CKD (Fogo, 2011). Previous studies have demonstrated that rosiglitazone could protect kidney from damage induced by several factors (Huang et al., 2013; Kumar et al., 2013; Korolczuk et al., 2014). Pioglitazone attenuated renal ischemia-reperfusion injury through its anti-inflammatory and antioxidant effects (Reel et al., 2013; Zou et al., 2013), reduced kidney damage in diabetic rats through increasing glomerular podocalyxin protein expression (Peng et al., 2014) and provided renoprotection by interfering with the renin-angiotensin system and profibrotic proteins (Ochodnicky et al., 2014). In addition, mitochondria are a target of PPARγ agonists. They can improve mitochondrial functions including mitochondrial respiration, ROS production and ATP generation in various diseases (Patel et al., 2017). Furthermore, PPARγ agonists can also modulate mitochondrial metabolism proteins such as dynamin related protein 1 (Drp1) and mitochondrial fission protein 1 (Fis1) (Zolezzi et al., 2013; Chuang et al., 2016). Particularly, rosiglitazone PPARγ-dependently protects podocyte from aldosterone-induced injury and restores mitochondrial function (Zhu et al., 2011) and pioglitazone protected against aging-related renal injury by improving mitochondrial functions (Yang et al., 2009).

Our previous study showed that pioglitazone benefited renal failure through activation of the antioxidative system and inhibition of angiogenesis in the remnant kidney (Sun et al., 2016). In the present study, we aim to continue to investigate the pharmacological mechanisms of pioglitazone in the 5/6 nephrectomized (Nx) rat model from the angle of fibrosis and mitochondrial functions.

The present study demonstrated that PPARγ agonist pioglitazone attenuated renal injury and tubulointerstitial fibrosis evidenced by the improved renal functions, reduced fibrotic areas and downregulated profibrotic proteins. In addition, pioglitazone alleviated mitochondrial dysfunction and modulated mitochondrial ETC and dynamic proteins in the remnant kidney.

Extracellular matrix deposition is a common feature of tissue fibrosis. As an early biomarker of fibrosis, TGF-β1 induces synthesis of ECM components such as collagen and fibronectin (Leask and Abraham, 2004). Previous study has demonstrated that pioglitazone inhibited TGF-β1-induced fibronectin mRNA expression in mouse mesangial cells (Guo et al., 2004). In agreement with this finding, we found that pioglitazone not only decreased the expression of TGF-β1 in the remnant kidney, but also downregulated the TGF-β1-induced collagen I and fibronectin expressions in HK2 cells, which indicated that pioglitazone may both target to TGF-β1 and its downstream profibrotic proteins. In addition, we found that PPARγ antagonist GW9662 completely inhibited pioglitazone-induced expression of PPARγ, but has not totally suppressed the effects of pioglitazone on all the profibrotic proteins we detected. This finding suggests that the anti-tubulointerstitial fibrosis effects of pioglitazone may be through both PPARγ-dependent and -independent pathways. The detailed mechanisms need to be investigated in further studies.

Mitochondria are important cellular organelles that are responsible for numerous physiological processes such as energy production, oxidative phosphorylation and Ca²⁺ homeostasis. In this study, 5/6 nephrectomy induced renal injury in rats, accompanied by mitochondrial dysfunction, as evidenced by the increased ROS generation, decreased MMP and ATP production, reduced activities of complexes I and III, and downregulated Cyt C expression in mitochondria in the remnant kidney. These observations suggest the damage to the mitochondrial membrane, respiratory chain complex inhibition and leakage of Cyt C from the mitochondria. Pioglitazone restored these changes. In vitro studies provide compelling evidence that TGF-β1 induced mitochondrial dysfunction in HK-2 cells. Energy supply is the major mission of mitochondria. In mitochondria, ATP is produced by ETC using the proton gradient via oxidative phosphorylation (Bertram et al., 2006). In this study, three typical ETC proteins, ATPβ, complex I subunit NDUFB8 and COX I were examined. NDUFB8 and COX I are the enzymes in mitochondrial ETC. ATPβ, also called complex V, is important in ATP production through oxidative phosphorylation. We found that the expression levels of all the three proteins in the whole cell were downregulated both in the remnant kidney and TGF-β1-induced HK-2 cells. However, in the isolated mitochondria, two nuclear DNA-encoded protein NDUFB8 and ATPβ were unchanged. Unfortunately, we cannot make exact conclusion from our data, this discrepancy may due to the auto-regulation of the transport mechanisms or may be attributed to the time point we detected as discussed in a previous study (Hui et al., 2017). Anyhow, pioglitazone protected mitochondrial function from ETC damage, thereby restoring ATP production, and this effect is, at least partly, PPARγ-dependent. These results indicate that the mitochondria protective effects may contribute to the renoprotection of pioglitazone.

Maintaining homeostasis of cells and mitochondrial functions require a dynamic balance between mitochondrial fission and fusion. The imbalance of the mitochondrial dynamics largely contributes to the pathogenesis of the tissues (Archer, 2013). Mitochondrial fission and fusion are regulated by several dynamin-related GTPase proteins. Drp1, a GTPase localized in the cytosol, is the key mediator in mitochondrial fission in a variety types of cells (Smirnova et al., 2001). During fission, Drp1 is recruit by its receptors to outer membrane and oligomerized to promote membrane constriction and scission (Smirnova et al., 2001). For mitochondrial fusion, the inner membrane fusion and the outer membrane fusion are regulated by different proteins. Mfn1 and 2 are responsible for outer membrane fusion and Opa1 is responsible for inner membrane fusion (Santel and Fuller, 2001). In this study, the mitochondrial dynamics were found to be pathologically remodeled in the remnant kidney and TGF-β1-induced HK-2 cells, as evidenced by downregulated expression of mitochondrial fusion proteins Mfn2 and Opa1 and increased expression of fission protein Drp1. This finding is in agreement with Hui et al. (2017) study. However, Aparicio-Trejo OE et al reported an opposite finding in the same model (Aparicio-Trejo et al., 2017). This discrepancy may be attributed to the different time point to detection–we and Hui et al. (2017) performed the detection 8 and 4 weeks after the model establishment, respectively, whereas Aparicio-Trejo et al. (2017) did the examination 24 h after the nephrectomy. In their study, the shift of mitochondrial dynamics to fusion was as a response to energy demand, which may be a compensatory mechanism to complement mitochondrial dysfunction. In our studies, after long time damage, the mitochondria lost its auto-compensation as evidenced by the increased expression of fission protein. Excessive fission is a key factor of mitochondrial dysfunction and inhibitor of fission has been developing as a therapeutic strategy for mitochondrial diseases (Reddy, 2014). Here we found that pioglitazone PPARγ-dependently normalized the expression levels of mitochondrial dynamic proteins, which indicates that pioglitazone exerts its mitochondrial protective effects through modulating mitochondrial dynamics.

Although the present study did not reveal exact molecular target of pioglitazone, we can deduce that the modulation of TGF-β1expression and mitochondrial dysfunction may be reciprocal. The interaction between TGF-β1 signaling and mitochondria has been demonstrated in the previous studies (Willaert et al., 2012; Jain et al., 2013). Corcoran et al demonstrated that mitochondria were involved in the amplification of TGF-β1 signaling in renal epithelial cells (Corcoran et al., 2013), suggesting mitochondria may contribute to the TGF-β1-induced renal fibrosis. In our findings, TGF-β1 expression and mitochondrial functions were both recovered after pioglitazone treatment, which indicates that pioglitazone may protects kidney from fibrosis through improving mitochondrial functions.

In summary, pioglitazone attenuates mitochondrial dysfunction in the kidney of CKD rat model. This effect may contribute to its effects on renal function improvement and renal fibrosis inhibition. Our study provides new evidence of the anti-fibrotic mechanisms of pioglitazone and suggests that pioglitazone may be effective in preventing renal fibrosis in CKD treatment.

The study was designed by LS and DW. The article was written by LS. The experiments was performed by LS, QY, TX, LY, JF, JM, LW, and CL. The data was analyzed by LS, QY, and TX.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.