Cocktail of chemical compounds robustly promoting cell reprogramming protects liver against acute injury

Tissue damage induces cells into reprogramming-like cellular state, which contributes to tissue regeneration. However, whether factors promoting the cell reprogramming favor tissue regeneration remains elusive. Here we identified combination of small chemical compounds including drug cocktails robustly promoting in vitro cell reprogramming. We then administrated the drug cocktails to mice with acute liver injuries induced by partial hepatectomy or toxic treatment. Our results demonstrated that the drug cocktails which promoted cell reprogramming in vitro improved liver regeneration and hepatic function in vivo after acute injuries. The underlying mechanism could be that expression of pluripotent genes activated after injury is further upregulated by drug cocktails. Thus our study offers proof-of-concept evidence that cocktail of clinical compounds improving cell reprogramming favors tissue recovery after acute damages, which is an attractive strategy for regenerative purpose.


INTRODUCTION
Tissue damage induces cells into reprogramming-like cellular state, which contributes to tissue regeneration (Jessen et al., 2015). Furthermore, genes related with cell reprogramming are activated and play important role during the regeneration process after damage (Bhave et al., 2011). It's also known that these genes involved are key factors to reprogramming cells into pluripotency in vitro and in vivo (Mosteiro et al., 2016). Thus it is very interesting to know whether factors promoting cell reprogramming, which are identified through in vitro assays, would help tissue regeneration in vivo. To date, reports already reveal that cell reprogramming efficiency can be enhanced by chemical compounds, even reprogramming factors can be partially or completely replaced by combination of small chemical compounds in pluripotent cell generation (Huangfu et al., 2008;Ichida et al., 2009;Esteban et al., 2010;Hou et al., 2013). In this paper we explore whether combination of chemical compounds, strongly promoting cell reprogramming but not generating pluripotent cells by themselves alone, would show beneficial effects for mouse liver regeneration after acute injuries.
Studies already demonstrate that inhibitors of histone deacetylases (HDACs), glycogen synthase kinase 3 beta (GSK-3β) or transforming growth factor beta (TGF-β) kinase contribute to efficient cell reprogramming . Thus we investigated the potential reprogramming-promoting activity via combinations of inhibitors of these pathways. . All data are presented as mean ± SD. *P < 0.05, **P < 0.01 vs. DMSO.

Drug cocktails robustly enhance cell reprogramming in vitro
Since inhibitors of HDACs, GSK-3β, and TGF-β kinase have been applied against diseases including cancer and mental illness in clinic (Johnstone, 2002;Avila and Hernández, 2007;Ikushima and Miyazono, 2010), we further accessed the reprogramming-promoting efficiency of pharmaceutical compounds from these pathways. VPA has been used clinically against cancer (Atmaca et al., 2004), thus it is remained for ongoing test. Li 2 CO 3 (Lithium carbonate, GSK-3β inhibitor) has been prescribed against manic and bipolar disorders (Solomon et al., 1997). Among inhibitors of TGF-β, LY2157299 has also been considered as an anti-cancer agent (Bueno et al., 2008), and Tranilast has been widely used as an antioxidant (Holmes et al., 2000), antiallergic (Kondo et al., 1992), antiangiogenesis (Isaji et al., 1997), and anti-inflammatory agent (Shiota et al., 2010). Our results showed that combination of these drugs showed robust reprogramming-enhancing effect ( Fig. 2A). Reprogramming efficiency promoted by cocktail containing 0.5 mmol/L VPA, 0.3 mmol/L Li 2 CO 3 , and 1 μmol/L LY2157299 or cocktail including 0.5 mmol/L VPA, 0.3 mmol/L Li 2 CO 3 , and 30 μmol/L Tranilast was comparable to that by VPA/CHIR99021/Repsox (Fig. 2B). Considering cell reprogramming inducted by injury in situ associates with microenvironment, the key factor involved is extracellular matrix (ECM). It is important to examine the effect of ECM on cell reprogramming. Matrigel containing multiple growth factors and cytokines was used to mimic ECM here. Cell reprogramming efficiency was almost twice-fold higher in Matrigel-treated group in comparison with that in control group, which could be calculated from the number of GFP + colony formation (Fig. 2C). Furthermore, the reprogramming efficiency by drug cocktails was also significantly enhanced after Matrigel treatment, which is even slightly higher than that under non-Matrigel treatment (Fig. 2D). Thus drug cocktails VPA/ Li 2 CO 3 /LY2157299 or VPA/Li 2 CO 3 /Tranilast could significantly improve cell reprogramming, which was even better in favor of extracellular matrix. However, none of the two drug cocktails alone could generate GFP + pluripotent cells from OG2-MEFs without introduction of exogenous Yamanaka factors, no matter combining with or without Matrigel (Fig. 2E).

Drug cocktails protects liver against acute injuries
Then we tested whether administration of these drug cocktails would be helpful for tissue recovery after damage. Partial hepatectomy (PHx) is well-established model for studying liver regeneration (Mitchell and Willenbring, 2008). Drug cocktail VPA/Li 2 CO 3 /LY2157299 or VPA/Li 2 CO 3 /Tranilast was administrated intraperitoneally 6 h after PHx. Drug dosage in each cocktail was according to previously reported data. 48 h after PHx, the ratio of liver weight to body weight was calculated and liver specimens were harvested and sectioned for immunofluorescent staining of proliferation marker Ki-67 to assess liver regeneration. As shown in Fig. 3A, cells in livers that received drug cocktails showed enhanced cell proliferation. Consistently, in drug cocktailtreated mice, the ratio of liver weight to body weight evaluated significantly higher than that in saline-treated mice ( Fig. 2B), indicating that both drug cocktails improved liver regeneration after physical damage.
Carbon tetrachloride (CCl 4 ) has been used to induce acute or chronic liver injury and leads to liver function failure (Lindroos et al., 1991). 6 h later following CCl 4 -administration, VPA/Li 2 CO 3 /LY2157299 or VPA/Li 2 CO 3 /Tranilast was injected intraperitoneally. 24 h after CCl 4 -treatment, serum alanine transaminase (ALT) and aspartate transaminase (AST) levels were monitored for the function of liver recovery. As shown in Fig. 3C and 3D, ALT and AST levels were significantly lower in study groups treated with drug cocktails comparing with that in saline-treated control group, suggesting CCl 4 -induced acute failure or hepatic dysfunction was attenuated by the drug cocktails. Liver histopathology analysis was then performed to assess the correlated hepatocellular damage. As shown in Fig. 3E, CCl 4 -induced acute liver injury including hepatocyte ballooning and necrosis was also attenuated by the drug cocktails. However, treatment with VPA/Li 2 CO 3 , LY2157299 or Tranilast alone showed no promotional effects for liver recovery from either PHx or CCl 4 -induced acute liver injury (Fig. S2).

Drug cocktails enhance expression of pluripotent genes
Expression of pluripotent genes is up-regulated instantly in regenerative liver after acute injury, which indicated that these factors play a role during liver regeneration (Bhave et al., 2011). To verify this phenomenon, 6 h after PHx without drug treatment, liver samples were harvested for quantitative real-time PCR analysis. The result revealed that expression level of c-Myc, Oct4, and Klf4 was significantly higher 6 h after PHx than that in sham, while expression level of Sox2 was slightly enhanced (Fig. 4A). Since combination of small chemical compounds with diverse targets highly promote cell reprogramming in vitro, it is very interesting to know whether these cocktails also enhance expression of pluripotent genes in this process. 6 h after treatment of VPA/CHIR99021/Repsox, or VPA/Li 2 CO 3 / Tranilast in OG2-MEFs transduced with Yamanaka factors, expression level of Oct4, Sox2, Klf4, and c-Myc was significantly elevated comparing with no chemical treatment, especially for Sox2. For cocktail VPA/Li 2 CO 3 /LY2157299, it significantly activated the expression of Oct4, Sox2, and Klf4 (Fig. 4B).
In vivo assays, expression levels of four pluripotent genes Sox2, Oct4, Klf4, and c-Myc were significantly elevated as early as 3 h after treatment of both drug cocktails in liver samples with PHx. The significant enhancement of Sox2  expression lasted for 6 h in both drug treated groups. Expression of c-Myc was still significantly higher 6 h after treatment of VPA/Li 2 CO 3 /LY2157299. Expression levels of all these genes were reduced 12 h after PHx in liver treated with either drug cocktail or saline (Fig. 4C). For CCl 4 -induced acute liver injury, 6 h after administration of drug cocktail, expression of Sox2, c-Myc, and Klf4 was significantly improved in VPA/Li 2 CO 3 /LY2157299 treated group, while VPA/Li 2 CO 3 /Tranilast only enhanced the expression of Sox2 and c-Myc (Fig. 4D).

DISCUSSION
Pluripotent genes are activated in injured tissue cells and play important role in regeneration. By detecting orthologs of mammalian reprogramming factors in newts, expression of Sox2, Klf4, and c-Myc are upregulated during lens and limb regeneration (Maki et al., 2009). In the process of chick retinal pigmented epithelial (RPE) cell dedifferentiation and transdifferentiation after retinal injury, expression of these three pluripotent genes are also upregulated in injured RPE during short time comparing with non-injured one. Expression of these genes is further enhanced and cell reprogramming process is sustained after addition of FGF2 (Luz-Madrigal et al., 2014). In adult zebrafish, expression of Oct4 and Sox2 is upregulated in injured spinal cord as shown by time course analysis (Hui et al., 2015). Importantly, it is shown that chemical compound FM19G11 increased the expression of Sox2, Oct4, and Nanog in a dose-responsive manner, which can promote the regeneration after spinal cord injury (Rodriguez-Jimnez et al., 2012). Except for neural system, expression of the above three pluripotent genes is also detected at early time then decreased gradually in rat tracheal epithelium injured by 5-fluorouracil in vivo, which indicates the important role of these factors during tracheal regeneration (Song et al., 2010). During liver regeneration, Oct4, c-Myc, Klf4, and Nanog message are induced as early as 3 h after partial hepatectomy in rats. Western blot analysis demonstrates that the relative expression level of these genes is less than that in pluripotent stem cells, except for c-Myc (Bhave et al., 2011). Consistent with these results, our data show that expression level of pluripotent genes activated in liver can be further enhanced by cocktail of chemical compounds after acute liver injuries. Beyond the enhanced pluripotent gene expression, expression level of interleukin-6 (Il-6) is also significantly increased by drug cocktail treatment after PHx (Fig. S3). It has been reported that IL-6 mediates increase of cell reprogramming in vitro and helps enabling cellular reprogramming in vivo (Brady et al., 2013;Chiche et al., 2016;Mosteiro et al., 2016). Importantly, these enhancements are accompanied by improved efficiency of liver regeneration. Transient introduction or partial activation of pluripotent genes can convert somatic cells into transition state, which can then be inducted into various cell types under specific cell culture conditions with soluble lineage-specific signals, absent of establishment of pluripotent state (Zhu et al., 2015). This strategy serves as a new avenue for generating diverse functional cell types to cell-based regeneration with more advantageous comparing with induced pluripotent stem cell technology. Many functional cells including hepatocyte (Zhu et al., 2014), cardiomyocyte (Efe et al., 2011), neural stem cells (Wang et al., 2013), and endothelial cells have been generated via this method (Margariti et al., 2012). It also has been shown that combination of complete small molecules promoting cell reprogramming can convert somatic cells into intermediate cells. These cells then directly differentiate into desirable functional cells without pluripotent state (Cheng et al., 2014;Fu et al., 2015;Tian et al., 2016). Beyond these successful in vitro results, it is very interesting that short-term administration of exogenous pluripotency genes through hydrodynamic tail-vein injection only reprogram hepatocytes towards pluripotency transiently without disruption of host liver function and without teratoma formation (Yilmazer et al., 2013), although long-term induction of the factors in vivo still can generate induced pluripotent stem cells in situ (Abad et al., 2013). Thus the in vivo cell reprogramming processes is very similar to that in vitro. Beyond that, the in vivo microenvironment can provide tissue specific factors and three dimensional context for cell survival, which may be even better than in vitro cell culture condition. Many desirable cells generated through in vivo cell reprogramming, such as cardiomyocytes derived from in situ cardiac fibroblasts and hematopoietic stem cells isolated from teratoma (Qian et al., 2012;Amabile et al., 2013;Suzuki et al., 2013), acquire better cell function than their counterparts from in vitro assays. Accordingly, here we speculate that the chemical compounds may function together with endogenous pluripotency genes that were activated instantly by liver injury and promote reprogramming of parenchymal or non-parenchymal cells into "activated cells" with more plasticity. Then these cells would adaptively redifferentiate into new desirable cells under niche conditions in vivo, which contribute to liver regeneration (Yanger et al., 2013;Jessen et al., 2015).
It is still a big issue for introduction of exogenous genes by virus in clinical translation due to safety concern including tumorigenesis. However, small chemical compounds are much safer and have shown more advantageous in modulating cell function and cell fate changes (Xu et al., 2008). Here, each drug in our cocktails has been approved individually for their dosage safety as well as the tolerability, pharmacokinetics and pharmacodynamics. Some of those drugs have been considered as agents against cancer. Moreover, only the drug cocktails cannot induce pluripotent cell generation in vitro and short term administration of these drugs in vivo just activates transient enhancement of the pluripotent gene expression. Besides, there is no harmful effect of these drug cocktails on other organs, such as kidney (data not shown here). Thus our study provides proof-ofconcept evidence that clinical compounds identified through promoting cell reprogramming efficiency could be applied safely to facilitate tissue regeneration. During the preparation of this manuscript, an excellent work by Izpisua Belmonte and colleagues reported partial reprogramming in vivo not only ameliorates signs of aging but also improves regeneration of pancreas and muscle after injury (Ocampo et al., 2016). However, in this study, we only assess beneficial effects of the drug cocktails to liver tissue, whether these drug cocktails would also be applicable for other organs or anti-aging needs to be investigated in future.

Cell culture
MEFs were isolated from E13.5 embryos heterozygous for the Oct4-GFP transgenic allele as previously described . Briefly, OG2 mice carried Oct4 promoter driving GFP expression mated with 129 mice. Internal organs and gonads were removed before processing for cell isolation. Isolated MEFs in early passages (less than 3 passages) were used for further experiment and cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mmol/L L-glutamax, 0.1 mmol/L nonessential amino acids (NEAA), 100 units/mL penicillin and 100 μg/mL streptomycin.

Generation of mouse induced pluripotent stem cells
Retrovirus expressing Oct4, Sox2, Klf4 or c-Myc was produced in Plat-E cells as previously described . Briefly, 80% confluent Plat-E cells were transfected with retroviral vectors pMXs containing coding sequences of mouse Oct4, Sox2, Klf4 or c-Myc by FugeneHD transfection reagent (Promega, Cat. No. E2311). 48 h later, supernatant containing virus was added into plates of MEFs cell culture with 4 μg/mL polybrene then spun at 2,000 rpm for 90 min to ensure infection. MEFs were infected with virus twice, one time per day. Two days after virus infection, infected MEFs were digested into single cell and seeded at 5,000 cells per well into 96-well plates pre-seeded with irradiated MEFs or pre-coated with Matrigel (BD biosciences,Cat. No. 356234). Medium was changed into KSR medium (knockout-DMEM supplemented with 15% knockout serum replacement, 2 mmol/L L-glutamax, 0.1 mmol/L NEAA, 0.1 mmol/L β-mercaptoethanol, 1000 U/mL LIF, 100 units/mL penicillin and 100 μg/mL streptomycin, N2 and Vitamin C) with or without chemical compounds. One week later, medium was replaced with mouse embryonic stem cell medium (DMEM supplemented with 15% FBS, 2 mmol/L L-glutamax, 0.1 mmol/L NEAA, 0.1 mmol/L β-mercaptoethanol 1000 U/mL leukemia inhibitory factor, 100 units/ mL penicillin and 100 μg/mL streptomycin).

Screening chemical compounds to promote cell reprogramming
Small chemical compounds were added into KSR medium 1 day after reseeding of virus infected MEFs into 96-well plate. Chemical treatment lasted for 3 days, then medium was changed without chemical compounds every two days. 10 days later from initial chemical treatment, each well of 96-well plate was photographed by Acumen eX3 microplate scanner (TTP Labtech) and the number of GFP + colonies was automatically counted by ImageXpress ® Micro  Selleckchem, Cat. No. S2230, 10 mmol/L stored in DMSO. Tranilast, Selleckchem Cat. No. S1439, 100 mmol/L stored in DMSO.

Mouse models of acute liver injury
Male C57BL/6 mice, 10-12 weeks old, were purchased from Shanghai Laboratory Animal Center (Shanghai, China) and maintained in pathogen-free condition. 2/3 partial hepatectomy (PHx) was performed according to previous published protocol (Mitchell and Willenbring, 2008). Briefly, mice were anesthetized with pentobarbital sodium. After restraining the mouse and exposing the xiphoid, the left lateral lobe and median lobe were tied using 4-0 silk surgical thread, then the lobes were removed surgically. Sutured the abdomen and waited for the mice recovering. 6 h after operation, drug cocktail containing diverse chemical compounds as indicated was injected intraperitoneally (i.p.) into the mice. Dosage for each compounds in cocktails referenced from reported data is as follows: 100 mg/kg VPA (Hunan Xiangzhong Pharmaceutical Co., Ltd), 60 mg/kg Li 2 CO 3 (Yangzhou Pharmaceutical Co., Ltd.), 30 mg/kg LY2157299 (Synthesized by BioChemPartner), and 25 mg/kg Tranilast (CPU-Pharma). Injection of saline is considered as control. VPA/Li 2 CO 3 / LY2157299 treated group contained 12 mice. VPA/Li 2 CO 3 /Tranilast treated group contained 12 mice. Control group contained 10 mice. 48 h after PHx, mice were sacrificed for evaluation.
Fulminant hepatic failure was induced by i.p. injection of CCl 4 (5% v/v CCl 4 in olive oil; 10 mL/kg body weight). 6 h after injection, mice were administrated i.p. with drug cocktails (Combination and dosage of each compound was same as the above treatment) or saline. Each group contained 12 mice. 24 h after initial CCl 4 treatment, mice were anesthetized with sodium pentobarbital (50 mg/kg) (Sigma) by i.p. injection. Blood samples were harvested from mouse orbital sinus by capillary tube then spun down at 12,000 rpm for serum collection. Serum samples were sent to clinical laboratory of Eastern Hepatobiliary Surgery Hospital for ALT and AST analysis. Liver samples were fixed and embedded in paraffin then sectioned for immunohistochemical analysis.
Mice went through PHx were anesthetized with sodium pentobarbital and perfused with saline. Liver tissues were harvested and fixed in 4% PFA overnight then in 30% sucrose solution for 24 h. Mounted the tissues in O.C.T. embedding compound and froze at −80°C. Frozen tissues then were dissected into 10 μm thick sections. After blocking the liver sections with 0.5% bovine serum albumin in 1× phosphate buffered saline for 30 min, incubated the sections with primary antibody Ki-67 (Santa Cruz Biotechnology, Cat. No. sc-7846, 1:500 dilution) at 4°C overnight then with donkey anti-goat secondary antibody Alexa Fluor 568 (ThermoFisher, Cat. No. A-11057, 1:1000 dilution) at RT for 2 h. Images were taken under a fluorescent microscope (Zeiss Axio Observer Z1) and Ki-67 positive cells were automatically counted and analyzed by software ImageJ.

H&E staining and histopathological analysis
For histological analysis, liver tissue specimens from CCl 4 -injuried mice were fixed in 4% PFA overnight, then dehydrated by graded ethanol at RT and embedded in paraffin. Tissue sections at 5 μm were deparaffinized by xylene and stained with hematoxylin and eosin (H&E). Histopathological analysis of liver was performed according to previous reports (Feng et al., 2010). The grades of liver damage in different groups were assigned in numerical scores (scale from 0 to 3).

Quantitative real-time PCR
Total RNAs were extracted from cells or liver tissues using TRI Reagent ® (Sigma Aldrich, Cat. No. T9424) according to the manufacture's protocols. To obtain cDNA, RNA was reverse-transcribed by M-MLV reverse transcriptase and random hexamers. cDNA, 2× PCR Mix and Eva Green were mixed and analyzed with MX3000P Stratagene PCR machine. The relative mRNA expression values were normalized against the inner control (Hprt). Primer sequences are listed in Table S1.

Statistical analysis
Unless otherwise indicated, all experiments were repeated three times and all data were presented as mean ± standard deviation (SD) and analyzed using Student's t-tests. A value of P < 0.05 was considered statistically significant. The following significance values are noted throughout the text: *P < 0.05; **P < 0.01.