P‐cresyl sulfate causes mitochondrial hyperfusion in H9C2 cardiomyoblasts

Abstract Increased circulating level of uraemic solute p‐cresyl sulphate (PCS) in patients with chronic kidney disease (CKD) is known to increase myocardial burden relevant to mitochondrial abnormalities. This study aimed at investigating mitochondrial response to PCS in H9C2 cardiomyoblasts. H9C2 cardiomyoblasts were treated with four different concentrations of PCS (3.125, 6.25, 12.5 and 25.0 µg/mL) to study the changes in cell proliferation, cell size and mitochondrial parameters including morphology, respiration, biogenesis and membrane potential. The lowest effective dose of PCS (6.25 µg/mL) induced mitochondrial hyperfusion with enhanced mitochondrial connectivity, mitochondrial oxygen consumption rates, mitochondrial mass, mitochondrial DNA copy number and increased volume of cardiomyoblasts. After PCS treatment, phosphorylation of energy‐sensing adenosine monophosphate‐activated protein kinase (AMPK) was increased without induction of apoptosis. In contrast, mitochondrial mass was recovered after AMPK silencing. Additionally, mitochondrial hyperfusion and cell volume were reversed after cessation of PCS treatment. The results of the present study showed that low‐level PCS not only caused AMPK‐dependent mitochondrial hyperfusion but also induced cell enlargement in H9C2 cardiomyoblasts in vitro.


| INTRODUC TI ON
The risk of premature death is fivefold to tenfold higher in patients with chronic kidney disease (CKD) compared with that of their progressing to end-stage kidney disease. Cardiovascular disease is found to be the major contributor to the exponentially increased risk of death as kidney function deteriorates. 1 One of the significant problems in patients with CKD is the accumulation of different toxic chemicals that cause uraemic symptoms and cardiac remodelling. 2 Uraemic solute p-cresyl sulphate (PCS) is a metabolized product originating from intestinal microbial fermentation and difficult to be removed through conventional haemodialysis. 3 Several studies have shown that PCS is able to induce oxidative stress in endothelial cells, vascular smooth muscle cells, renal tubular cells and cardiomyocytes. [4][5][6] Additionally, a recent study has reported that increased serum total PCS levels are associated with a worse prognosis in stable angina patients with early-stage renal failure. 7 Mitochondria are energy-generating organelles, the functions of which are intrinsically linked to their morphology and membrane structure. Mitochondria stay in a dynamic balance through switching between mitochondrial fusion and fission for maintaining a physiological cellular mitochondrial population. 8 Accumulated evidence has demonstrated that progression of cardiovascular diseases is accompanied by increased mitochondrial damage and dysfunction. 9 A previous study has demonstrated that mitochondrial hyperfusion with increased mitochondrial ATP production can be maintained as long as the cells are exposed to some specific stress stimuli. This phenomenon has been known as 'stress-induced mitochondrial hyperfusion' (SIMH). 10 The stress stimuli reported to trigger mitochondrial hyperfusion as a protective response against stress include starvation, UV irradiation and actinomycin D. 10 However, little is known about the mitochondrial defence mechanism against potential threat under mild uraemic conditions. To mimic the clinical scenario of CKD at different stages of development, the present study attempted to address the effects of increasing concentrations of PCS on mitochondrial functions in cardiomyoblasts.
During PCS treatment, different concentrations of PCS were added in fresh medium every three days.

| Assessment of cell proliferation by Ki67 staining and CCK8 assay
For Ki67 staining, H9C2 cardiomyoblasts were fixed with cold 4% paraformaldehyde followed by permeabilization with 0.1% Triton X-100. Fixed cells were incubated with murine anti-Ki67 monoclonal antibody (1:500, ab155801, Abcam) and goat anti-rabbit Alexa Fluor 594 secondary antibody, as well as counterstained with DAPI. The cells were examined by fluorescence microscopy. For CCK8 assay, H9C2 cardiomyoblasts growing in 96-well culture dish were treated with PCS in different concentrations for three days. At the end of the culture, 10 μL of the CCK8 reagent (96992, SIGMA-ALDRICH) was added to each well. After two hours of incubation at 37°C, the absorbance was determined by spectrophotometer at 450 nm.

| Analysis of mitochondrial membrane potential (ΔΨm)
For the analysis of ΔΨm in H9C2 cardiomyoblasts, the membrane-

| Statistical analysis
GraphPad Prism software (ver. 6) (GraphPad software) was used for statistical analyses and data plotting. Data are expressed as mean ± SD. The difference in means between two groups was evaluated using the t test. One-way ANOVA was used to compare multiple groups. Test for the linear trend between groups was calculated from left to right group order. P values of <.05 were considered statistically significant.

| PCS-induced mitochondrial hyperfusion
To investigate the effect of p-cresyl sulphate (PCS) on H9C2 cardiomyoblast proliferation, the cells were treated with DMSO (ie vehicle control) and four different concentrations of PCS (ie 3.125, 6.25, 12.5 and 25.0 µg/mL) that were categorized into five groups: Ctrl, PCS3, PCS6, PCS12 and PCS25, respectively. We found that cell proliferative potential was significantly lower in groups of PCS6, PCS12 and PCS25 than that in group of Ctrl through the assessments of Ki67 staining and CCK8 assay (Figure 1A-C; P < .05). To evaluate whether PCS triggered abnormal mitochondrial dynamics in a dose-dependent manner, the same grouping was applied in H9C2 cells for the comparisons of mitochondrial morphology. Among these groups, PCS25 group was removed because of the lowest cell proliferation rate after treatment.
By comparing vehicle control to PCS-treated H9C2 cells, there were significant progressive increases in the structural complexities of mitochondria with increasing PCS concentrations as reflected by the mitochondrial parameters of branch length (P for trend = .0013), network number (P for trend = .0011) and footprint area (P for trend = .0006) ( Figure 1D,E). Both PCS6 and PCS12 groups displayed significantly higher network number and larger footprint area compared to those in the Ctrl group (P < .05). Also, PCS12 group showed a significantly longer branch length than that of Ctrl group (P < .05).

| PCS-triggered changes in mitochondrial bioenergetics
To estimate whether PCS impaired mitochondrial respiration, groups of Ctrl, PCS3, PCS6 and PCS12 were subjected to Seahorse XF24

| PCS-induced alterations of mitochondrial biogenesis
To estimate the effect of PCS treatment on mitochondrial biogenesis, mitotracker green staining and quantitative PCR were applied in Ctrl, PCS3 and PCS6 groups for the measurements of mitochondrial mass and mitochondrial DNA, respectively. After PCS treatment, the amount of mitochondrial mass in PCS3 and PCS6 groups was significantly increased on Day 2 and Day 3 compared to that in Ctrl group ( Figure 3A,B). Furthermore, increased

| PCS-induced mitochondrial hyperfusion mediated by AMPK
Both dynamin-related protein-1 (Drp1) and adenosine monophosphate-activated protein kinase (AMPK) mediate the process of mitochondrial fission in response to stress. 17,18 To determine whether both Drp1 and AMPK were associated with PCS-induced mitochondrial hyperfusion, protein expression and phosphorylation of Drp1 and AMPK were examined ( Figure 4A). After three days of PCS treatment, we found that AMPK phosphorylation (Thr172) was increased in PCS6 group compared to those in Ctrl group (P < .5) but Drp1 phosphorylation (S616) did not show significant difference.
To clarify the importance of AMPK involved in PCS-induced mitochondrial hyperfusion, knockdown of AMPK protein expression was conducted by siRNA transfection. After two days of AMPK siRNA transfection with H9C2 cardiomyoblasts, treatment of low concentration PCS (6.25 µg/mL) for three days was continuously performed ( Figure 4B). We observed that the decrease of PCS-induced mitochondrial hyperfusion regarding to the recovery of mitochondrial mass ( Figure 4C) was demonstrated in AMPK-silenced H9C2 cardiomyoblasts. Therefore, we assumed that AMPK protein expression plays an important role in PCS-induced mitochondrial hyperfusion.

| PCS-induced cell enlargement
To assess the cellular alternation in hypertrophy after PCS treatment, the appearance of treated H9C2 cardiomyoblasts was quantitated for the cellular morphologic parameters such as cell area and perimeter. After three weeks of PCS treatment, both cell area and parameter in PCS6 group were higher than those in Ctrl group (P < .05) (Figures 4B, 5A). Furthermore, myocardial hypertrophyrelated gene expression, atrial natriuretic peptide (ANP) was also determined. We found that higher ANP mRNA expression in PCS6 group was revealed compared to Ctrl group (P < .05) ( Figure 5C), suggesting that hypertrophic responses should be noted after PCS treatment.

| Mitochondrial hyperfusion reversed after cessation of PCS treatment
To verify the degree of mitochondrial recovery after cessation of PCS treatment, we implemented a five-week PCS treatment before cessation of PCS for two weeks ( Figure 6A). After PCS cessation for two weeks, cell proliferation rate returned from 83% to 100% ( Figure 6B). Also, cellular mitochondrial content reversed from 175% to 117% ( Figure 6C). During the period of PCS treatment, persistently high mitochondria mass compared with that in the control group was noted from Day 2 to Day 28, and the mitochondrial mass returned to normal one week after cessation of PCS treatment ( Figure 6D). After four weeks of PCS treatment, PCS-treated H9C2 cells displayed an average 105% increase in cell size compared to that in the controls, and the increase was reversed in the absence of PCS ( Figure 6E).
Taken together, PCS-induced increment in mitochondrial mass was associated with an increase in cell size, whereas mitochondrial hyperfusion was reversed after elimination of PCS stimulus in H9C2 cardiomyoblasts. concentrations in normal subjects have been reported to vary in range from 2.8 ± 1.7 µg/mL 19 to 6.6 ± 3.7 µg/mL, 20 determined by different methods, respectively. After renal injury, the mean plasma concentration of PCS accumulated in patients with end-stage renal disease is up to 106.9 ± 44.6 µg/mL. 21  SIMH is a compensatory response through mitophagy suppression 23 in response to external stress that mandates increased cellular energy (ie ATP) production. 10 If the ATP demand is not promptly satisfied, it will lead to initiation of senescence or induction of apoptosis. 24 Our findings indicate that increases of mitochondrial mass and mitochondrial DNA copy number after PCS treatment may allow physiological elongation and interlacing between mitochondria.

F I G U R E 5
Consequently, hyperfused structures may meet the sudden increase in energy demand through producing extra amount of ATP through oxidative phosphorylation.
Recent evidence also suggests that hyperfused mitochondrial networks suppress cell proliferation through the interruption of cell division. 25 In our study, treatment of PCS with low concentration did not cause cell apoptosis. Therefore, we speculate that reduced cell number following consecutive PCS treatment may be caused by cell cycle arrest rather than apoptosis. On the other hand, AMPK not only maintains metabolic homoeostasis through the regulation of mitochondrial dynamics 18 but also plays a crucial role in mitochondrial biogenesis. 26 Consistently, our findings indicate that increased AMPK activity may be associated with hyperfused mitochondria and irregular mitochondrial metabolism following PCS treatment.
Besides, the results of the current study suggested that mitochondria may not only control the cellular growth rate but also determine the cell size. Because overcrowded organelles (eg mitochondria) and cytosolic particles could slow down the rate of intracellular diffusive transport for oxygen and nutrients, there may be a compensatory increase in cell size to facilitate intracellular diffusion. 27 With regard to the control of mitochondrial networks, Drp1 is a critical mediator in mitochondrial dynamics. Loss of Drp1 impairs mitochondrial fission and leads to the formation of tubular morphology composed by elongated mitochondria. 22,30 However, Drp1 expression and activity in response to PCS treatment did not display significant differences. We assume that PCS-induced mitochondrial hyperfusion may be mainly regulated by AMPK rather than Drp1mediated mitochondrial dynamics.
It has been reported that H9C2 cardiomyoblasts and primary neonatal cardiomyocytes could display similar hypertrophic responses in vitro. 31 Expression of ANP has been reported to be closely associated with the progression of myocardial hypertrophy 32 and is considered to play an protective role in compensation. 33 Markedly elevation of ANP expression in heart tissue has been commonly used as a hypertrophic marker. 34 During long-term PCS treatment, we observed an association between an increase of mitochondrial mass and cell enlargement. Notably, cell size returned to normal after cessation of PCS treatment. This response is similar to physical exercise-induced physiological cardiac hypertrophy in response to the increased workload. This increase in cell size cannot be sustained unless the exercise is maintained. 35 Unlike physiological cardiac hypertrophy, pathological cardiac hypertrophy induced by disease conditions is less reversible. 36 Therefore, because low concentration PCS-induced cell hypertrophy was reversible in the present study, we assume that it was more likely a physiological rather than a pathological response.
Recently, it has been reported that metformin is able to raise the survival rate in older patients with moderate CKD. 37 Metformin is widely used as an antiglycemic drug to treat diabetes. Beyond the beneficial outcome of glucose reduction, administration of metformin may extend lifespan through the release of a small amount of reactive oxygen species (ROS) from mitochondria to trigger a cellular process against oxidative stress. 38 Furthermore, metformin may be able to lead the protection of myocardium and preservation of heart function through the activation of AMPK signalling pathway. 39 We speculate that low-level PCS in CKD may act as a kind of cellular stimuli, such as metformin, which may activate AMPK signal and generate excess ROS through mitochondrial hyperfusion.
The present study has its limitations. Despite the findings of the In conclusion, the results of the current study demonstrated that low-level p-cresyl sulphate (PCS) caused AMPK-dependent mitochondrial hyperfusion in H9C2 cardiomyoblasts with increased mitochondrial biogenesis, enhanced mitochondrial respiration and enlarged cell size.

ACK N OWLED G EM ENTS
This study was supported by programme grants from Chang Gung Memorial Hospital, Chang Gung University (No. CMRPG8F1311).

CO N FLI C T S O F I NTE R E S T
The authors report no relationships that could be construed as a conflict of interest.
AU TH O R CO NTR IB U TI O N THH and YLC conceived the study and participated in the design of the study, data acquisition and analysis, and drafting the manuscript.
YLC and THH were responsible for the laboratory assay and troubleshooting. HKY, CKS, CCY and FYL participated in interpretation. All authors read and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.