The residence of adult stem cells was found in various tissues such as skin and intestine. These tissue-specific stem cells are able to generate differentiated cells in the particular tissue. However, liver stem cells have not been precisely defined in adult livers since they can only be detected under limited conditions. Liver is characterized by the regenerative ability of mature hepatocytes, which can undergo hypertrophy and proliferation to recover the loss of liver mass. The emergence of liver stem cells, or frequently be termed LPCs or oval cells, are exclusively shown under chronic liver injury.
Several models including the administration of DDC diet or CDE diet have been extensively used to activate LPCs in mice (Itoh and Miyajima, 2014). We previously conducted CCl4 injection to induce liver fibrosis and found that LPCs were also activated.
The result demonstrated that LPCs could appear and proliferate under chronic liver injury, regardless of the model used. Additionally, hypoxic condition was created by fibrosis-induced oxygen deficiency, leading to HIF-1α expression. Surprisingly, those HIF-1α+ cells corresponded to LPCs. Therefore, I hypothesize that hypoxic condition could be a major factor which promotes LPCs activation.
The origin of LPCs during regeneration is of great interest. On account of previous findings, LPCs often emerge from the Canal of Hering, which is near the periportal area.
emergence of LPCs was detected around the central vein after CCl4-induced liver injury which seemed different from those earlier researches. Since mature hepatocytes are the major cell type near the central vein, I assumed mature hepatocytes as the cell origin of LPCs in our model. In other words, hepatocytes may possess the plasticity to de-differentiate into LPCs, which could further bi-potentially de-differentiate into hepatocytes or cholangiocytes. To prove this hypothesis, I cultured primary hepatocytes under hypoxic condition to simulate chronic liver injury and investigated whether they could serve as LPCs origin. Results showed that the expression of mature hepatocyte marker, HNF4α, was downregulated while the LPC markers, EpCAM and CK19 elevated remarkably.
Moreover, stemness markers including Sox2 and Nanog also increased significantly.
Taken together, mature hepatocytes could regain stemness and de-differentiate to LPCs under hypoxic stress. Recent studies provided some clues that are consistent with our finding. For instance, Tarlow et al. used lineage tracing system to illustrate that mature hepatocytes could contribute to progenitor cells and re-differentiate into hepatocytes to restore liver function (Tarlow et al., 2014). Besides, another group showed that ICC could originated from hepatocytes through Notch-Hes1 pathway (Sekiya and Suzuki, 2012).
These works as well as our results challenged the former viewpoint and established a new insight that hepatocytes hold great plasticity that could give rise to LPCs and even proceed lineage conversion.
Opposing results from different researches make this topic controversial. It is possible that diverse approaches of cell lineage tagging and various injury models may make difference to the characteristics of LPCs among these studies. Therefore, LPCs would arise in periportal or pericentral region depending on different models. A review even pointed out that cholangiocytes could serve as the source of LPCs emerging in periportal area while hepatocytes could be the origin of LPCs emerging in pericentral region (Michalopoulos, 2014). Although culturing cells under hypoxia could not recapitulate liver injury in vivo, I still prove that hypoxia could enhance the stemness of hepatocytes and transform them into LPC-like cells. In view of our former results, since hypoxia was induced during liver fibrosis and LPCs emergence was shown in pericentral area, I concluded that mature hepatocytes could de-differentiate into LPCs in CCl4 injury model in order to perform liver regeneration.
Next, I examined the molecular mechanisms of hepatocyte de-differentiation to see how HIF-1α affected the fate of hepatocytes. HIF-1α could maintain undifferentiated status of a variety of stem cells and stimulate cancer cells de-differentiation into progenitor-like cells (Heindryckx et al., 2012; Mathieu et al., 2011). Recently, Wnt signaling was also shown to be crucial for LPCs activation (Apte et al., 2008). Therefore, Wnt signaling and HIF-1α may synergistically promote hepatocyte de-differentiation and
after cultured under hypoxia. The downstream target genes of Wnt signaling, Axin2 and Lef1 were increased. Although the association between HIF-1α and Wnt signaling has been discussed in HCC, it was the first time to know their relationship during LPCs activation. Moreover, direct interaction between HIF-1α and β-catenin was found in HCC,
which could enhance EMT features significantly (Zhang et al., 2013). Furthermore, EMT could drive hepatocytes de-differentiation towards progenitor-like cells (Chen et al., 2012). Thus, I suggested that during LPCs activation, HIF-1α and β-catenin would interact with each other and induce EMT, which could provoke hepatocytes to gain stemness and drive the process of de-differentiation. However, more evidences should be
explored in my study to support this hypothesis.
EpCAM is a LPC marker and is rarely detected in mature hepatocytes. It could also maintain stem cell population and promote cell proliferation (Dolle et al., 2015). In my study, I found that 1α would increase the expression of EpCAM, indicating that HIF-1α may somehow activate LPCs through modulating EpCAM expression. Next, I further hypothesized that HIF-1α would bind onto HRE of EpCAM promoter in order to enhance its transcription. From luciferase assay, EpCAM transcriptional activities of hepatocytes were elevated when cultured under hypoxia. Also, direct interaction of HIF-1α and EpCAM promoter was observed using ChIP assay. Previous studies seldom discussed the relationship between 1α and EpCAM. There is only a study demonstrated that
HIF-1α expression was correlated with EpCAM in HCC, and claimed that HIF-HIF-1α increment
would subsequently induce the expression of EpCAM (Yamada et al., 2014). Taken
together, HIF-1α could regulate EpCAM expression by interaction with its promoter, which led to hepatocyte de-differentiation and LPCs emergence.
The intracellular domain of EpCAM, EpICD, was found to form a complex with components of Wnt signaling to initiate its functions (Maetzel et al., 2009). Although I have not investigated the interaction between EpCAM and Wnt signaling during LPCs activation, I supposed that they would have close connections with each other. From all of the findings above, HIF-1α, Wnt signaling and EpCAM would have complex interactions which could facilitate hepatocyte de-differentiation and help liver regeneration.
In vitro culture of primary hepatocytes under hypoxia could simulate chronic liver
injury in vivo and examine how HIF-1α modulated the characteristics of hepatocytes.
However, this model may neglect other possible factors that would together regulate LPCs activation. For example, microenvironment is a critical component that should take into account. Several signaling pathways triggered by niche cells could contribute to LPCs activation and proliferation. Therefore, whether LPCs were derived from hepatocytes and how hypoxic stress induce LPCs emergence should be further confirmed
Additionally, whether LPCs could differentiate back into hepatocytes is another topic that should be elucidated. Earlier study showed that Notch and Wnt signaling could drive LPC lineage specification into cholangiocytes and hepatocytes respectively. When hepatocytes encountered damages, surrounding macrophages could engulf the cellular debris and produce Wnt ligand, which then induce Wnt signaling in LPCs. Afterwards, LPCs were stimulated to form hepatocytes (Boulter et al., 2012). Interestingly, Wnt signaling was upregulated in my study. Thus, I suggested that Wnt signaling would not only transform hepatocytes into LPCs, but also direct these LPCs to form new hepatocytes, leading to proper regeneration.
In conclusion, mature hepatocytes could serve as the cellular origin of LPCs under hypoxic condition in order to facilitate liver regeneration. During the fate transition process, HIF-1α, Wnt signaling and EpCAM would have complex interactions which play important roles in stemness acquisition and cellular proliferation. Although the mechanisms modulating liver regeneration is not clear, it is certain that liver has great plasticity which could repair itself by various signaling pathways.