3.6. The change of AMPK phosphorylation and the expression of SP1 and PGAM5 were recovered in Empa treated HK2 cells.
We investigated the change of phosphorylated AMPK, PGAM5 and SP1 protein in vitro. Similarly, we found a remarkable downregulation of phosphorylated AMPK and a significant upregulation of PGAM5 and SP1 expression in HK2 cells subjected to HG, and the Empa recovered these changes significantly (Fig. 4A, F). In order to detect the causality between the AMPK phosphorylation and the SP1-PGAM5 changes, we introduced AMPK activator (AICAR) and inhibitor (Compound C) into our experiments. We found that both HG and Compound C exposure result in decreased phosphorylated AMPK, DRP1S637 and increased SP1 and PGAM5 expression. Both the Empa and the AICAR could alleviate the changes of AMPK, DRP1S637, SP1 and PGAM5 in HK2 cells (Fig. 4C, D, E). Meanwhile, this alleviation effect of Empa diminished again when the Compound C was administrated (Fig. 4C, D, E).
In order to consolidate the involvement of SP1 and PGAM5 in the AMPK regulated mitochondrial fission, we examined the fission related protein in HG treated HK2 cells that subjected to si-SP1 or si-PGAM5. We found that both si-PGAM5 and si-SP1 significantly reversed the decrease of DRP1S637 in HK2 cells subjected to HG (Fig. 4B, H), but neither si-PGAM5 or si-SP1 could affect the AMPK phosphorylation (Fig. 4B, G). Meanwhile, the si-SP1 obviously reduced the PGAM5 expression but the si-PGAM5 had no effect on SP1 (Fig. 4B, G). The Empa could not only upregulate the phosphorylated AMPK and DRP1S637, but also decrease the expression of SP1 and PGAM5 in HG treated HK2 cells (Fig. 4B, G, H).
3.7. The SP1 could bind to the promotor region of PGAM5 and increased its transcription in HK2 cells.
Finally, we further testify our deduction that the SP1 could bind to the PGAM5 promoter region directly. The predicted binding sites within the SP1 sequence are illustrated in Figure 5A. JASPER analysis revealed 3 putative SP1 binding motifs within the promoter of PGAM5 that located at 713, 539, 320 bp upstream of the transcription start (Fig. 5B). We found that overexpression of SP1 significantly increased the protein levels of PGAM5 in HK2 cells (Fig. 5C, D), which confirmed that SP1 can activate the expression of PGAM5. As shown in Fig. 5E, the WT promoter construct of PGAM5 had transcriptional activity in HK2 cells as compared to the promoterless pEZX-PG02. The SP1 remarkably increased the transcriptional activity of the WT promoter. Moreover, the mutant of the SP1 binding site at -539 nt significantly decreased the transcriptional activity of the PGAM5 promoter. By contrast, the mutant of the SP1 binding site at either -713 or -320 nt did not alter the promoter activity significantly (Fig. 5E). We next testified the physical interacts between SP1 and PGAM5 promoter by chromatin immunoprecipitation (ChIP) assay using HK2 cells transfected with either SP1 expression construct or its empty vector. DNA from the immunoprecipitates was then amplified by PCR with primers located in the promoter region near the putative SP1 binding site. SP1 was found to physically bind to this region, whereas no binding was detected in these samples with the negative control (NC) primers located far from the -539 nt region (Fig. 5F).