Pre-conditioning with nsPEFs enhances trilineage differentiation potential of pMSCs
Stem cell properties of MSCs can be assessed by assaying the potential to differentiate along the osteogenic, adipogenic, and chondrogenic lineages [25]. In order to optimize the treatment conditions of nsPEFs, five sets of nsPEFs parameters (i.e., 10 ns at 20 kV/cm, 60 ns at 5 kV/cm, 60 ns at 10 kV/cm, 60 ns at 20 kV/cm, and 100 ns at 10 kV/cm) were firstly screened by the differentiation assays of pMSCs (Fig. 1A). We found that pre-conditioning of nsPEFs with the stimulation levels of 10 ns at 20 kV/cm or 100 ns at 10 kV/cm could significantly increase the subsequent trilineage differentiation of pMSCs, compared with the non-nsPEFs-preconditioned control group, but not the other three sets of parameters (Table 1, Fig. 1B and C).
The expression levels of differentiation genes were also evaluated at day 14. Osteogenic transcription factor RUNX2 was upregulated by 10.53±1.91 and 9.03±1.77 fold by nsPEFs (10 ns at 20 kV/cm, and 100 ns at 10 kV/cm) (Fig. 1D); main regulating valves for adipogenic differentiation PPARγ was improved by 6.06±0.78 fold (10 ns at 20 kV/cm) and 9.93±1.42 fold (100 ns at 10 kV/cm) (Fig. 1E); chondrogenic transcription factor SOX9 was increased by 10.50±1.95 fold (10 ns at 20 kV/cm) and 10.82±1.09 fold (100 ns at 10 kV/cm) (Fig. 1F). The expression of other related functional genes (OCN, ALP; LPL, AP2; COLII, AGG) can be upregulated for 5-10 folds compared to the control group (Fig.1D-F and Fig. S1A-C). Taken together, these data suggest that the biological effects of nsPEFs depend on the time and energy levels of treatment. Only two sets of parameters, i.e., 10 ns at 20 kV/cm, and 100 ns at 10 kV/cm, could enhance the differentiation capacity of pMSCs.
Optimized nsPEFs do not influence the proliferation of pMSCs
Proliferation of pMSCs was evaluated with MTT assay over 7 days after pre-conditioning with nsPEFs, and nsPEFs treatments did not influence the proliferation of pMSCs (Fig. S2A). Moreover, cell cycle analysis and colony-forming units (CFU) assays were performed to evaluate the effects of nsPEFs. There were no significant differences in cell cycles (Fig. S2B) or CFU numbers (Fig. S3C) between nsPEFs treatments and control groups. These data indicate that our optimized nsPEFs parameters do not influence the clonogenicity and cell proliferation of MSCs.
nsPEFs enhance gene expression of OCT4 and NANOG via removing the methylation of their promoters
OCT4 and NANOG are critical transcriptional factors for stem cell pluripotency [26]. To further explore the cellular molecular mechanisms of the biological effects caused by nsPEFs, the expression levels of pluripotency genes OCT4 and NANOG were examined. Interestingly, an instant elevation of OCT4 and NANOG was found after 2 hours of nsPEFs treatment both in porcine MSCs (pMSCs) and human MSCs (hMSCs) (Fig. 2A). Expression of OCT4 increased significantly 2 hours after nsPEFs, with 2.89±0.30 fold changes in pMSCs (p=0.0029), and 4.82±0.97 fold in hMSCs (p=0.0044), in responsive of 10 ns at 20 kV/cm nsPEFs treatments; 3.56±0.30 fold in pMSCs (p=0.001), and 3.42±0.86 fold in hMSCs (p=0.0476), in responsive to 100 ns at 10 kV/cm of nsPEFs treatments (Fig. 2A). The expression levels of NANOG gene was also upregulated significantly (pMSCs: 1.68±0.27 fold, p=0.0396 and 1.7±0.16 fold, p=0.0044 ; hMSCs: 2.44±0.15 fold, p=0.0005 and 1.96±0.21 fold, p=0.0093) in both nsPEFs treatment groups (10 ns at 20 kV/cm, and 100 ns at 10 kV/cm) (Fig. 2A). We then tracked the gene expression levels of OCT4 and NANOG of pMSCs after 3 days and 7 days of nsPEFs, and found that the upregulated OCT4 subsequently decreased over 7 days (Fig. S3A and C), while the expression levels of NANOG remained the same after nsPEFs (Fig. S3B and D). In addition to the gene expression levels of OCT4 and NANOG, we further examined the epigenetic modification by using bisulfite sequencing analysis. With the precondition of nsPEFs, a clearly drop was found in the methylation levels of CpG sites of OCT4 and NANOG promoters, compared with non-treated pMSCs control group (Fig. 2B and C). Therefore, these data suggest that nsPEFs can directly function on MSCs by demethylating the promoter region of OCT4 and NANOG.
To further investigate if the instant upregulation of pluripotency genes was an universal effect for all stem cell types, we also evaluated the OCT4 and NANOG changes in human embryonic stem cells (hESCs, details are in supplementary documents) after 2 hours of nsPEFs preconditioning. Interestingly, we found that only nsPEFs with parameter of 100 ns at 10 kV/cm can enhance the gene expression of OCT4 (4.92±1.00 fold, p=0.0097) and NANOG (4.63±1.16 fold, p=0.0223) of hESCs significantly, but not with 10 ns at 20 kV/cm (Fig. S3C and D).
nsPEFs temporally decrease DNMT1 expression
We next aimed to gain insights into how the hypomethylation of the OCT4 and NANOG promoters was regulated by nsPEFs. DNA methylation of CpG dinucleotides is catalyzed by at least three different DNA methyltransferases (DNMTs), including DNMT1, DNMT3a and DNMT3b. And DNMT3a and DNMT3b function primarily as de novo methyltransferases that establish DNA methylation patterns, while DNMT1 is a key enzyme that maintains methylation patterns following DNA replication [27]. The DNMTs are essential for maintaining the methylation pattern in stem cells and for regulating their self-renewal and differentiation [24, 28]. The protein expression levels of DNMT1 substantially dropped by 0.58±0.11 and 0.27±0.05 fold respectively after 2 hours of nsPEFs treatment (10 ns at 20 kV/cm; 100 ns at 10 kV/cm) in pMSCs, while declined to 0.69±0.02 and 0.56±0.06 fold in hMSCs (Fig. 3A). Gene expression of DNMT1 decreased significantly to 0.3±0.07 and 0.3±0.06 fold in pMSCs, and to 0.52±0.03 and 0.41±0.06 fold in hMSCs (Fig. 3B). However, the levels of DNMT3a and DNMT3b did not change in both pMSCs and hMSCs (Fig. S4A and B). To confirm the function of elevated DNMT1, the 5-methylcytosine contents which reflect global DNA methylation level were measured after 2 hours of nsPEFs. The global DNA methylation analysis revealed a 0.39±0.06 or 0.51±0.05 decrease in nsPEFs-preconditioned groups compared with control group (Fig. 3C).
To investigate how long the effects can last, protein expression levels of DNMT1 in pMSCs at 2, 12, 24 and 72 hours after nsPEFs were examined. After nsPEFs treatment, the expression of DNMT1 gradually increased from a lower level after 2 hours, and peaked at 24 hours, which was greatly higher than control groups, and then entered the end point of a dynamic equilibrium to the levels of control groups after 72 hours (Fig. 3D).
Overexpression of DNMT1 blocks the upregulation of OCT4 and NANOG induced by nsPEFs
To further justify if nsPEFs-reduced DNMT1 directly affect the expression of OCT4 and NANOG, as well as the subsequent differentiation of pMSCs, we established a tet-on system to drive the DNMT1 expression in pMSCs (GFP as system control) (Fig. 4A). Because there were no significantly differences between the two sets of nsPEFs parameters (10 ns at 20 kV/cm v.s. 100 ns at 10kV/cm) in terms of the biological effects in earlier experiments, nsPEFs at the levels of 100 ns at 10 kV/cm were used in this section. Overexpression of DNMT1 by the tet-on system increased the protein expression of DNMT1 by 1.33±0.09 fold (p=0.0138), which indicated that we successfully established the DNMT1 overexpression model. Treatment of nsPEFs lowered the protein expression of DNMT1 by 0.34±0.06 fold in GFP control group (Fig. 4B), which matched with the earlier results in pMSCs and hMSCs (Fig. 3A, 3C). Notably, the enhanced expression levels of OCT4 (3.50±0.77 fold, p=0.0309, nsPEFs+ group) and NANOG (1.95±0.22 fold, p=0.0121, nsPEFs+ group) were blocked by overexpression of DNMT1, and the gene expression levels of OCT4 and NANOG stayed unchanged after 2 hours of nsPEFs treatment (Fig 4C, DNMT1+/nsPEFs+ group). We then evaluated the percentage of CpG demethylation of OCT4 and NANOG promoters with bisulfite sequencing analysis in this DNMT1 overexpression model (Fig. 4D), and the results were consistent with the genes expression levels (Fig. 4C). Taken together, these data show that overexpression of DNMT1 can block the effects of nsPEFs on gene expression of OCT4 and NANOG in pMSCs.
Overexpression of DNMT1 blocks the subsequent effects of nsPEFs on stem cell differentiation
To further investigate if DNMT1 erased the subsequent differentiation performance of MSCs enhanced by nsPEFs, both trilineage differentiation and related functional genes were evaluated with DNMT1 overexpression (Fig. 5A). Osteogenic differentiation, which was indicated by the quantification of alizarin red staining intensity (Fig. 5B), was increased by 1.37±0.09 fold (p=0.0071, GFP+/nsPEFs+ group) by nsPEFs (100 ns, 10 kV/cm), and decreased by 0.78±0.06 fold by overexpression of DNMT1 (p=0.0068, DNMT1+ group). Meanwhile, there was a significant difference between control group (GFP+ group) and nsPEFs stimulated DNMT1 overexpression group (p=0.4912, DNMT1+/nsPEFs+ group). The differentiation performance of pMSCs into adipogenic lineage (oil-red O staining, Fig. 5B) and chondrogenic lineage (Alcian blue staining, Fig. 5B) shared the same trends as osteogenic differentiation. The expression levels of trilineage differentiation related key genes (osteogenic: RUNX2, OCN; adipogenic: PPARγ, LPL; chondrogenic: SOX9, COLII) showed similar trends with the differentiation assays, that all functional genes of trilineage differentiation were upregulated in GFP+/nsPEFs+ groups, and had no significant change in DNMT1+/nsPEFs+ groups (Fig. 5C).