In the present study, we demonstrate the aptitude of Pdx-1 or Pdx-1-VP16 protein transduction to reprogram hepatic stem-like cells into β-like cells in vitro.
First, we verified that the PTD of Pdx-1 and TAT allow transduction into hepatic WB and HepG2 cell lines. Intracellular localization of fusion proteins revealed by confocal microscopy analysis indicate a true uptake of proteins and not mere adherence to the cell surface. Furthermore, transduced recombinant Pdx-1, TAT-Pdx-1, Pdx-1-VP16 and TAT-Pdx-1-VP16 proteins exert biological activity on an insulin promoter reporter system. Pdx-1 fusion to VP16 activation domain does not further enhance insulin promoter activity consistent with previous findings [10
]. Our results confirm reports of Pdx-1 transduction [15
] and are the first demonstration of Pdx-1-VP16 protein transduction.
Pancreatic differentiation experiments were conducted in an HG environment. In fact, long-term HG culture furthers liver cell commitment towards a pancreatic fate [4
]. Previous studies of Pdx-1(-VP16)
expression in WB cells do not distinguish between respective contributions of transgene expression and HG culture on differentiation. Here, we show that HG culture alone converts WB cells into pancreatic endocrine precursor cells. In contrast to LG cultures, HG cultures express four pancreatic endocrine genes: Ngn3, Nkx2.2, Pdx-1
. These results concur with Yang et al.
study, where confluent culture of hepatic oval stem cells for 2 months in HG medium induces conversion into insulin-producing cells [21
expression is at odds with previous reports where Kir6.2
was detected in WB cells overexpressing Pdx-1
genes only (i) after 2–3 months in HG culture [14
], (ii) 40 days post-transplantation into diabetic mouse [13
], or (iii) after Pax4
Pdx-1, TAT-Pdx-1, Pdx-1-VP16, or TAT-Pdx-1-VP16 proteins further induce expression of NeuroD, Pax4, Nkx6.1
, insulin 1
after two weeks of treatment. Some samples express just a part of these markers or display RT-PCR expression patterns similar to HG controls. Quantitative analysis would help to further nuance pancreatic gene expression in-between these samples, in particular up-regulation of the endogenous Pdx-1 gene. The heterogeneity between samples may be a consequence of: (i) infrequent liver to endocrine pancreas conversion, leading to few pancreatic gene positive cells which may be difficult to detect, (ii) different kinetics of gene expression between wells. According to Tang et al.
study, our experiments do not reveal more efficient differentiation following Pdx-1-VP16 protein treatment compared to Pdx-1 protein treatment [14
]. TAT-mediated transduction does not lead to more advanced differentiation suggesting that containing two PTDs (PTDPdx-1
and TAT) does not increase transduction efficiency. Surprisingly, insulin 1
, two of the pancreatic genes expressed after our protein treatments, are not detected in a previous in vitro
study on WB cells transduced with Pdx-1
genes, even after 3 months of HG culture [14
]. Here, despite insulin 1
expression, neither insulin 2
are detected pointing at the possible need for long-term culture in HG medium to obtain mature β-like cells [14
]. Moreover, Ngn3
expression in wells scoring positive for all other pancreatic genes, including insulin 1
, suggests the presence of remaining subpopulations of immature pancreatic precursor cells.
Overexpression of Pdx-1
genes in hepatic cells leads to exocrine as well as a range of endocrine cell types [4
]. In our study, Pdx-1(-VP16) or TAT-Pdx-1(-VP16) protein treatments result in expression of PP
, but not of glucagon
suggesting that Pdx-1(-VP16) transduction in hepatic stem-like cells may lead to endocrine β-cell and non-β-cell phenocopies. Amylase
, expressed in rat liver or exocrine pancreatic cells, is detected in all conditions tested here [23
Recently, Koya et al.
demonstrated that Pdx-1 protein delivery into diabetic mice restores euglycemia mainly through pancreatic β-cell regeneration. The authors observed β-cell gene expression and insulin synthesis in the pancreas and in the liver of treated mice suggesting hepatic insulin contribution to euglycemia [20
]. In complement, our findings provide the first direct evidence that Pdx-1(-VP16) protein transduction in conjunction with HG culture reprograms hepatic stem-like cells into cells displaying similarities with β-cells in vitro
. At a point where strategies for targeted β-cell differentiation begin to surface, our study illustrates how simple exposition to Pdx-1(-VP16) protein in the surrounding medium triggers short-term pancreatic endocrine conversion. This study may contribute to the development of protein transduction therapy, a new concept to induce β-cell differentiation.