Advertisement

Rescue of Lethal Hepatic Failure by Hepatized Lymph Nodes in Mice

  • Toshitaka Hoppo
    Affiliations
    McGowan Institute for Regenerative Medicine, Department of Pathology, University of Pittsburgh Medical School, Pittsburgh, Pennsylvania
    Search for articles by this author
  • Junji Komori
    Affiliations
    McGowan Institute for Regenerative Medicine, Department of Pathology, University of Pittsburgh Medical School, Pittsburgh, Pennsylvania
    Search for articles by this author
  • Rohan Manohar
    Affiliations
    McGowan Institute for Regenerative Medicine, Department of Pathology, University of Pittsburgh Medical School, Pittsburgh, Pennsylvania
    Search for articles by this author
  • Donna Beer Stolz
    Affiliations
    Cell Biology and Physiology, Center for Biologic Imaging, University of Pittsburgh Medical School, Pittsburgh, Pennsylvania
    Search for articles by this author
  • Eric Lagasse
    Correspondence
    Reprint requests Address requests for reprints to: Eric Lagasse, PharmD, PhD, McGowan Institute for Regenerative Medicine, University of Pittsburgh, 450 Technology Drive, Suite 300, Pittsburgh, Pennsylvania 15219. fax: (412) 624-5263
    Affiliations
    McGowan Institute for Regenerative Medicine, Department of Pathology, University of Pittsburgh Medical School, Pittsburgh, Pennsylvania
    Search for articles by this author
Published:November 10, 2010DOI:https://doi.org/10.1053/j.gastro.2010.11.006

      Background & Aims

      Hepatocyte transplantation is a potential therapeutic approach for liver disease. However, most patients with chronic hepatic damage have cirrhosis and fibrosis, which limit the potential for cell-based therapy of the liver. The development of an ectopic liver as an additional site of hepatic function represents a new approach for patients with end-stage liver disease. We investigated the development and function of liver tissue in lymph nodes in mice with liver failure.

      Methods

      Hepatocytes were isolated from 8- to 12-week-old mice and transplanted by intraperitoneal injection into 8- to 12-week-old fumarylacetoacetate hydrolase mice (Fah−/−), a model of the human liver disease tyrosinemia type I. Survival was monitored and the locations and functions of the engrafted liver cells were determined.

      Results

      Lymph nodes of Fah−/− mice were colonized by transplanted hepatocytes; Fah+ hepatocytes were detected adjacent to the CD45+ lymphoid cells of the lymphatic system. Ten weeks after transplantation, these mice had substantial improvements in serum levels of transaminases, bilirubin, and amino acids. Homeostatic expansion of donor hepatocytes in lymph nodes rescued the mice from lethal hepatic failure.

      Conclusions

      Functional ectopic liver tissue in lymph nodes rescues mice from lethal hepatic disease; lymph nodes therefore might be used as sites for hepatocyte transplantation.

      Keywords

      Abbreviations used in this paper:

      Fah (fumarylacetoacetate hydrolase), HGF (hepatocyte growth factor), NTBC (2-(2-nitro-4-trifluoromethylbenzoyl)-1,3 cyclohexanedione), OLT (orthotopic liver transplantation), SP (splenic), WT (wild type)
      Orthotopic liver transplantation (OLT) is currently the only curative treatment for severe liver disease. However, because of the shortage of donor organs, its application is greatly limited. Furthermore, patients with comorbidities and advanced age are either not considered candidates for OLT or are expected to have reduced post-transplant survival.
      • Perkins J.D.
      • Halldorson J.B.
      • Bakthavatsalam R.
      • et al.
      Should liver transplantation in patients with model for end-stage liver disease scores <or= 14 be avoided? A decision analysis approach.
      • Volk M.L.
      • Hernandez J.C.
      • Lok A.S.
      • et al.
      Modified Charlson comorbidity index for predicting survival after liver transplantation.
      • Lipshutz G.S.
      • Busuttil R.W.
      Liver transplantation in those of advancing age: the case for transplantation.
      Cell-based transplantation has been proposed as a therapeutic alternative to OLT or as a bridge for patients who are waiting for an organ to become available.
      • Fisher R.A.
      • Strom S.C.
      Human hepatocyte transplantation: worldwide results.
      • Strom S.C.
      • Chowdhury J.R.
      • Fox I.J.
      Hepatocyte transplantation for the treatment of human disease.
      • Ito M.
      • Nagata H.
      • Miyakawa S.
      • et al.
      Review of hepatocyte transplantation.
      Most cellular therapies for liver diseases have been directed at cell engraftment in the liver itself. This approach limits the possible efficacy of cellular therapy in the vast majority of patients with end-stage liver diseases, in whom cirrhosis and fibrosis are the common pathologic features.
      • Nussler A.
      • Konig S.
      • Ott M.
      • et al.
      Present status and perspectives of cell-based therapies for liver diseases.
      • Lorenzini S.
      • Andreone P.
      Stem cell therapy for human liver cirrhosis: a cautious analysis of the results.
      • Lorenzini S.
      • Gitto S.
      • Grandini E.
      • et al.
      Stem cells for end stage liver disease: how far have we got?.
      The development of an ectopic liver as an additional site of hepatic function represents a new therapeutic opportunity for patients with end-stage liver disease who would be at high risk for OLT. Transplantation of hepatocytes at several different extrahepatic sites has been shown in animal models, but engraftment of hepatocytes has been associated with variable results.
      • Gupta S.
      • Vemuru R.P.
      • Lee C.D.
      • et al.
      Hepatocytes exhibit superior transgene expression after transplantation into liver and spleen compared with peritoneal cavity or dorsal fat pad: implications for hepatic gene therapy.
      • Ohashi K.
      • Waugh J.M.
      • Dake M.D.
      • et al.
      Liver tissue engineering at extrahepatic sites in mice as a potential new therapy for genetic liver diseases.
      Here, we show that the development of life-supporting ectopic liver tissue is possible in lymph nodes after liver failure.

      Materials and Methods

       Animals

      Fumarylacetoacetate hydrolase (Fah)−/− mice (129sv), a kind gift from Dr Markus Grompe (Portland, OR), or Fah−/− mice backcrossed into C57bl were used for recipients, and 129S4 and green fluorescent protein–C57Bl mice (cat# 004353) obtained from The Jackson Laboratory (Bar Harbor, ME) were used for donors. Freshly isolated hepatocytes were obtained from 8- to 12-week-old mice and were transplanted into 8- to 12-week-old Fah−/− mice. The protocol followed National Institutes of Health guidelines for animal care and was approved by the University of Pittsburgh's Institutional Animal Care and Use Committee.

       Cell Preparation

      Hepatocytes were harvested using the 2-step collagenase perfusion technique introduced by Seglen.
      • Seglen P.O.
      Preparation of isolated rat liver cells.
      The number and viability of cells were determined by trypan blue exclusion. One million viable cells were suspended in 30 μL Hank's balanced salt solution and kept on ice until transplantation.

       Transplantation

      For intraperitoneal hepatocyte transplantation, 1 million viable liver cells were injected into the lower peritoneal cavity with a 28-gauge needle. For splenic hepatocyte transplantation, animals were anesthetized and a small surgical incision was made in the left flank. The spleen was exposed and 0.2 × 106 liver cells, suspended in 30 μL Hank's balanced salt solution, were injected into the inferior pole of the spleen using a 28-gauge needle. The injection site was ligated to prevent cell leakage and bleeding. All mutant mice were kept on 2-(2-nitro-4-trifluoromethylbenzoyl)-1,3-cyclohexanedione (NTBC) until transplantation. NTBC was discontinued just after transplantation. The weight of experimental animals was taken weekly to monitor their health. Generally, Fah−/− mice lose weight during the first few weeks after transplantation owing to the gradual loss of liver function and progressively regain their initial weight later when donor liver cells regenerate liver tissue and hepatic functions. Whenever the animals lost more than 25% of their initial body weight, the risk of these animals dying increased and NTBC was given back to restore liver function. It usually took 5–7 days for the mice to return to their initial weight and liver functions under NTBC. At that point, NTBC was discontinued again to induce liver failure. Such protocol was used previously to allow a low number of engrafting liver cells to selectively generate enough liver mass to finally rescue the animals from liver failure.
      • Lagasse E.
      • Connors H.
      • Al-Dhalimy M.
      • et al.
      Purified hematopoietic stem cells can differentiate into hepatocytes in vivo.

       Fah Enzyme Assay

      Fah enzyme assays were performed at 37°C as described previously.
      • Knox W.E.
      • Edwards S.W.
      Homogentisate oxidase of liver.
      The harvested tissues stored at −80°C were homogenized and sonicated in complete lysis M buffer (Roche, Mannheim, Germany). Protein concentrations were measured with the BCA protein assay kit (Pierce, Rockford, IL) and adjusted to 3 μg/mL. A total of 8 μL of fumarylacetoacetate (a gift from Dr Grompe), the substrate for this assay, was incubated with each protein solution and the attenuation of absorbance at 330 nm was measured spectroscopically every 10 seconds. Wild-type (WT) and Fah−/− livers were used as positive and negative controls, respectively. Fumarylacetoacetate is not commercially available and was prepared enzymatically from homogentisic acid.
      • Knox W.E.
      • Edwards S.W.
      Homogentisate oxidase of liver.

       Serial Transplantation of Hepatocytes From Hepatized Lymph Nodes

      Harvested hepatized lymph nodes were minced into small pieces and incubated in 0.1 mg/mL collagenase type II solution supplemented with 0.05 mg/mL DNase I (Sigma, St Louis, MO) at 37°C for 30 minutes. The isolated cells were collected by filtration through a 70-μm nylon mesh and washed 3 times with Hank's balanced salt solution. The number and viability of cells were determined by trypan blue exclusion. A total of 105 cells were suspended in 30 μL Hank's balanced salt solution and transplanted by splenic injection as described earlier. The repopulation of Fah-positive hepatocytes in the recipient liver was calculated by counting the number of Fah-positive cells in any 4 views randomly selected on Fah-stained sections.

      Results

       Intraperitoneal Injection of Hepatocytes Rescues Mice From Liver Failure

      We transplanted Fah−/− mice to explore the feasibility of functional ectopic liver in a model of highly efficient liver regeneration.
      • Grompe M.
      • Lindstedt S.
      • al-Dhalimy M.
      • et al.
      Pharmacological correction of neonatal lethal hepatic dysfunction in a murine model of hereditary tyrosinaemia type I.
      • Overturf K.
      • al-Dhalimy M.
      • Ou C.N.
      • et al.
      Serial transplantation reveals the stem-cell-like regenerative potential of adult mouse hepatocytes.
      Fah−/− tyrosinemic mice have progressive and fatal liver failure unless treated with 2-(2-nitro-4-trifluoromethylbenzyol)-1,3-cyclohexanedione (NTBC, nitisone, Orphadin; Rare Disease Therapeutics, Inc, Franklin, TN).
      • Grompe M.
      • Lindstedt S.
      • al-Dhalimy M.
      • et al.
      Pharmacological correction of neonatal lethal hepatic dysfunction in a murine model of hereditary tyrosinaemia type I.
      We and others have shown that WT hepatocytes have a strong selective growth advantage when transplanted in the liver of Fah−/− mice after NTBC removal, resulting in near-complete regeneration of the liver.
      • Lagasse E.
      • Connors H.
      • Al-Dhalimy M.
      • et al.
      Purified hematopoietic stem cells can differentiate into hepatocytes in vivo.
      • Overturf K.
      • al-Dhalimy M.
      • Ou C.N.
      • et al.
      Serial transplantation reveals the stem-cell-like regenerative potential of adult mouse hepatocytes.
      To evaluate a possible ectopic location for liver cell transplant, 106 liver cells from congenic WT mice were transplanted in Fah−/− tyrosinemic mice intraperitoneally (IP) (n = 50). Splenic (SP) injection was used as a positive control, indirectly delivering the cells to the liver
      • Rhim J.A.
      • Sandgren E.P.
      • Degen J.L.
      • et al.
      Replacement of diseased mouse liver by hepatic cell transplantation.
      • Overturf K.
      • Al-Dhalimy M.
      • Tanguay R.
      • et al.
      Hepatocytes corrected by gene therapy are selected in vivo in a murine model of hereditary tyrosinaemia type I.
      (n = 21). NTBC was removed to induce progressive liver failure in all the transplanted animals and their weight was monitored weekly as an indicator of liver function. SP-injected mice initially lost weight and then spontaneously regained weight with donor hepatocytes repopulating the entire diseased liver and reversing lethal tyrosinemia, as described previously
      • Overturf K.
      • al-Dhalimy M.
      • Ou C.N.
      • et al.
      Serial transplantation reveals the stem-cell-like regenerative potential of adult mouse hepatocytes.
      (19 of 21 mice transplanted; 90.4% survival) (Figure 1A). IP transplantation of liver cells resulted in long-term survival of these animals (Figure 1A). Long-term survival was successful with one period of selection (4 of 11 mice transplanted; 36% survival) but with less efficiency than after 2 periods of selection (42 of 50 mice transplanted; 84% survival) (Figure 1B).
      Figure thumbnail gr1
      Figure 1Fah−/− mice are rescued from lethal hepatic failure by IP injection of hepatocytes. (A) Body weight after SP and IP transplantation indicates hepatic regeneration. Body weight of the transplanted mice was monitored weekly after liver cell transplantation (time 0), to follow hepatic engraftment and rescue from tyrosinemia. Fah−/− mice transplanted by either single SP or IP injections lost weight during the first 3 weeks. Weight loss is indicative of a decline in liver function. SP-injected mice spontaneously regained weight. IP-injected mice required 2 periods of selection before regaining weight for efficient survival, a protocol previously described for engrafting low levels of liver cells.
      • Lagasse E.
      • Connors H.
      • Al-Dhalimy M.
      • et al.
      Purified hematopoietic stem cells can differentiate into hepatocytes in vivo.
      (B) Spontaneous weight gain after a single period of selection was possible but with a lower survival rate. Fah−/− mice were transplanted with 106 WT liver cells followed by NTBC removal from their diet (blue and red line). n = number of mice analyzed. (C) Anatomic location of enlarged nodules 10 weeks after transplantation. Left panel: many enlarged nodules around the stomach region and on the mesenterium are observed in a mouse transplanted IP with WT hepatocytes (circles). Middle upper panel: native liver of the IP-injected Fah−/− mouse and a control WT liver. The native liver of the IP-injected Fah−/− mouse was atrophic with a couple of small regenerative nodules containing WT hepatocytes on its surface (arrows). Middle lower panel: isolated enlarged nodules from mouse in the left panel with diameters from 1 to 10 mm. Right upper and lower panels: mesenteric lymph nodes (mln) repopulated with green fluorescent protein–positive liver cells. Blood vessels (bv) and small intestine (si) are green fluorescent protein negative.
      Ten weeks after transplantation and after apparent rescue of tyrosinemia, laparotomies were performed on the experimental mice. Twenty to 40 enlarged nodules were observed around the stomach region and along the mesenterium in all the IP transplanted animals (Figure 1C). None of these enlarged nodules were found in SP-transplanted Fah−/− mice. The distribution of nodules matched the expected distribution of lymph nodes present in these regions. The presence of enlarged nodules and the reversal of tyrosinemia was a long-lasting effect. More than 6 months after transplantation, Fah−/− animals were still alive and healthy.

       Lymph Nodes Are Colonized by Hepatocytes

      At various times after IP transplantation of hepatocytes, Fah−/− mice were killed to determine the origin of the hepatic nodules. Within a few days, Fah+/CK18+ hepatocytes were detected adjacent to the CD45+ lymphoid cells of the lymphatic system. CK18+ hepatocytes co-localized with Meca79 (CD62L ligand), a marker of high endothelial venules present in lymph nodes (Figure 2A). Several weeks after transplantation, Fah+ hepatocytes had entirely colonized several lymph nodes (Figures 2B and 3). Furthermore, bromodeoxyuridine-labeling experiments indicated that donor hepatocytes in the lymph nodes proliferate for 2–3 weeks after transplantation and cease to proliferate by 8 weeks postinjection (Figure 2B). Immunofluorescent analyses of hematopoietic markers (CD45) with T-cell markers (CD3, CD4, and CD8), B-cell marker (B220), and myeloid markers (Gr-1 and CD11b) suggest the transformation of the lymph nodes from a lymphoid organ to a hepatic organ with the presence of CK18+ hepatocytes (Figure 2C) 10 weeks after transplantation. By using green fluorescent protein–labeled donor liver cells, we found that no other organs were colonized except visceral lymph nodes and occasionally native liver with small colonies. This result suggests that hepatocytes rapidly migrate into the lymphatic system through afferent lymph vessels, colonize lymph nodes, proliferate, and then passively or actively eliminate lymphocytes from the lymph nodes.
      Figure thumbnail gr2
      Figure 2Immunohistochemistry of lymph nodes from the gastric and common hepatic arteries. (A) Two and 3 days after IP injection of hepatocytes in Fah−/− mice. On day 2, some WT Fah+CK18+ hepatocytes could be detected in the lymphatic system near lymphocytes. On day 3, clusters of CK18+ hepatocytes were seen in association with CD45+ hematopoietic cells. (B) Meca79 (CD62L ligand), a marker of high endothelial venules found in lymph nodes, is co-localized with CK18+ donor hepatocytes 2 and 3 weeks after IP injection, Fah+ hepatocytes (green) have colonized the lymph nodes and have a high index of proliferation, as shown by the high ratio of bromodeoxyuridine (BrdU) incorporation (red nuclei). Eight weeks after IP injection, few liver cells are proliferating in lymph nodes. (C) Immunofluorescence analysis of hematopoietic markers in hepatized lymph nodes 10 weeks after IP injection of hepatocytes in Fah−/− mice. Both sections of hepatized lymph nodes and control (WT) lymph nodes were stained with hematopoietic markers. Each staining has 2 panels, the upper panel represents the lymph nodes engrafted with hepatocytes (Hepatized LN) and the lower panel is normal WT mouse lymph node (Control LN). Stainings were performed on serial sections. Bar: 100 μm.
      Figure thumbnail gr3
      Figure 3Hepatized lymph nodes 10 weeks after IP transplantation. (A) Sections were immunostained with an anti-Fah antibody (brown, horseradish-peroxidase staining) then counterstained with hematoxylin. Fah+ hepatocytes and several islands of small hematopoietic cells were present, but no biliary structures were observed. Bar: 100 μm. (B) Immunofluorescence of lymph nodes engrafted with hepatocytes (Hepatized LN), control liver, and control lymph nodes (Control LN). Frozen sections were stained with hepatocyte marker CK18 and the endothelial marker CD31. CD26, dipeptidyl-peptidase-IV, was used as a hepatocyte maker and E-cadherin was used as an epithelial maker. Most cells in the hepatized LN were CK18+ hepatocytes, with expression patterns similar to those of control liver. These cells also were albumin positive (brown cells, insert in left upper panel) with CK18 and CD26 co-localized (insert in CD26 staining panel). Hepatocyte and epithelial markers were negative in normal lymph node. In the hepatized LN, CD31+ endothelial cells corresponding to vessels were similar in size and morphology to those found in normal (control) liver. In contrast, CD31+ cells indicative of high endothelial venules found in normal (control) lymph nodes differ in morphology. Bar: 100 μm. (C) The ratio ± standard deviation of the weight of liver and enlarged nodules to body weight. The ratio of hepatic tissues to body weight was determined in Fah−/− mice transplanted IP and both liver (atrophic) and enlarged nodules (hypertrophic) were collected and compared with normal WT liver. n = number of mice analyzed. (D) Transmission electron microscopy of the hepatized lymph nodes. Ultrastructure of a hepatized lymph node (upper panels) and control liver (lower panels). Left upper panel: hepatocytes present in lymph nodes have large prominent nuclei (N), bile canaliculi (BC), mitochondria (M), peroxisomes (P), and rough endoplasmic reticulum (RER). Bar: 2 μm. Center upper panel: higher magnification of the bile canaliculus, containing microvilli (MV) with tight junctions (arrowheads) and adherent junctions (AJ). A lipid vacuole is seen within the canaliculus. Bar: 500 nm. Right upper panel: vessels in hepatized lymph node consisted of nonfenestrated sinusoidal endothelial cells (SECs). Bar: 1 μm. Left lower panel: hepatocytes in control liver showing fenestrations (arrows) in SECs. Bar: 2 μm. Center lower panel: higher magnification of bile canaliculus showing tight junctions (arrowheads), lipid vacuoles, and space of Disse (SD). Bar: 500 nm. Right lower panel: organization of hepatic plates in control livers with bile canaliculi at the apical surface and fenestrated sinusoids (S) at the basolateral surface. Bar: 2 μm. (E) Immunofluorescence analysis with nonhematopoietic liver cell markers in hepatized lymph nodes, normal liver, and normal lymph node (LN). Staining was performed with the nonparenchymal cell markers F4/80, desmin, Glial Fibrillary Acidic Protein, CK19, and ER-TR7. F4/80 (Kupffer cells), CK19 (biliary cells), and ER-TR7 (reticular fibroblasts) stainings were negative in hepatized lymph nodes. Bar: 100 μm.

       Characterization of Hepatized Lymph Nodes

      Immunohistologic analyses confirmed the presence of donor hepatocytes in all the analyzed enlarged nodules (Figure 3A and B). The newly generated hepatized lymph nodes had a hepatic mass representing more than 70% of expected normal liver mass or 1.5 × 107 liver cells (Figure 3C). This massive ectopic engraftment and expansion of liver cells subsequently rescued the animal from lethal liver failure. Analysis of the hepatized lymph nodes showed that not only had the lymphocytes almost completely disappeared (Figure 2C), but high endothelial venules, the specialized postcapillary venules found in lymphoid tissue,
      • Girard J.P.
      • Springer T.A.
      High endothelial venules (HEVs): specialized endothelium for lymphocyte migration.
      also were absent after hepatocyte colonization (Figure 3B). The high endothelial venules were replaced with large vessels that have a histology and size similar to that found in normal liver (Figure 3B), but lacked the characteristic fenestrations (Figure 3D). These vessels appeared to be abundant in hepatized lymph nodes, indicating a possible adaptation of the vasculature to the newly generated hepatic tissue.
      On the other hand, ER-TR7, a marker for reticular fibroblasts and reticular fibers, was missing in the hepatized lymph nodes, as well as F4/80, a marker for macrophage/Kupffer cells (Figure 3E). CK19, a marker for biliary epithelial cells, also was absent, but the presence of bile canaliculi around desmosomes and tight junctions between hepatocytes was confirmed by electron microscopy (Figure 3D). Desmin and Glial Fibrillary Acidic Protein, 2 markers of stellate cells, were detected. However, because these markers also were present in normal lymph nodes, definitive identification of the donor's hepatic stellate cells was inconclusive (Figure 3E).

       Biochemical Liver Functions Are Restored by Hepatized Lymph Nodes

      We assessed the biochemical liver function of Fah−/− mice transplanted IP or SP, normal WT donor mice, Fah−/− mice under NTBC, and untreated Fah−/− mice (under liver failure) by measuring serum levels of transaminases, bilirubin, and amino acids (Figure 4A). Ten weeks after transplantation, IP-injected Fah−/− mice showed substantial improvement in all parameters. They differed from SP mice by a slight decrease in some of the liver functions. Interestingly, concentrations of both total and direct bilirubin were abnormally increased in the serum of IP mice, but were several-fold lower than untreated tyrosinemic mice in hepatic failure (Figure 4A). The higher concentration of bilirubin is explained by the absence of CK19+ biliary cells observed in the hepatized lymph nodes (Figure 3E), even though biliary canaliculi containing bile were present between hepatocytes (Figure 3D).
      Figure thumbnail gr4
      Figure 4Biochemical liver functions are restored by hepatized lymph nodes. (A) Biochemical measurement of liver function in blood. Tyrosinemic (Fah−/−) mice were rescued by IP or by SP injection of WT liver cells. Ten weeks after transplantation, the mean biochemical measurements of various liver functions ± SD were compared between littermate WT controls, Fah−/− mice under NTBC, and untreated Fah−/− mice (NTBC withheld for 5 weeks and experiencing hepatic failure). All animals were 3–6 months old. The number of mice (serum) analyzed is indicated in parentheses. ALT, alanine aminotransferase. (B) Plasma concentration of albumin, fibrinogen, and HGF after IP injection of WT hepatocytes in Fah−/− mice at 1, 3, 4, 6, and 10 weeks. Controls correspond to normal WT mice. Plasma samples were tested by enzyme-linked immunosorbent assay. Each open circle represents the value from one mouse. Prism (GraphPad Software, Inc, La Jolla, CA) was used to run t tests to determine significant differences between particular groups. HGF increased at 6 weeks after the second and final selection (off NTBC), reflecting the massive expansion of hepatocytes in lymph nodes necessary for the Fah−/− survival. Bars indicate mean values. For HGF, *Pcontrol & 6wk = .0079 by the Mann–Whitney test. (C) Serum concentration of blood urea nitrogen (BUN), total cholesterol, and triglyceride levels in Fah−/− mice over 10 weeks after IP injection and rescue of the animals, and compared WT mice with P values. (D) Glycogen storage in hepatized lymph nodes. Glycogen storage was determined by periodic acid–Schiff (PAS) staining. Bar: 100 μm. The black and white electron microscopy panel identified glycogen rosettes (arrowheads) in hepatocytes. Bar: 500 nm. (E) Fah enzyme assay. A standard curve to measure enzyme activity was established using WT liver (100% activity), Fah−/− liver (0% activity), and WT/Fah−/− mixes to achieve 15%, 25%, and 80% enzyme activity, respectively. Fah enzyme activities in engrafted lymph nodes (LN) ranged from 80% to almost 100% of WT liver levels. In contrast, Fah enzyme activity in native livers of Fah−/− mice rescued by hepatized LN had Fah activity ranging from 25% of WT liver activity to 0% (mean Fah activity of 15% of WT liver levels). n = number of mice analyzed.
      Furthermore, the levels of serum albumin, fibrinogen, and hepatocyte growth factor (HGF), as well as blood urea nitrogen, total cholesterol, and triglyceride, were analyzed in the IP-injected Fah−/− mice and compared with normal mice (Figure 4B and C). Serum albumin and fibrinogen plasma levels were restored to normal levels. Interestingly, serum HGF level showed an increase at 6 weeks after transplantation. This increase in serum HGF level correlates with the massive expansion of hepatocytes expected in lymph nodes around 6 weeks after transplantation. Serum blood urea nitrogen level was normal but total cholesterol was slightly increased and triglyceride level was significantly lower than WT normal mice. The slight discrepancy in the lipid analyses found in the IP transplanted animals may be explained by the variability observed between normal male and female mice. The transplantations were not sex-matched, which could explain some differences observed in the experimental animals. Glycogen storage was determined by periodic acid–Schiff staining of hepatized lymph node in liver sections, and appeared normal. In addition, electron microscopy was used to identify the glycogen rosettes in a lymph node–derived hepatocyte (Figure 4D).

       Analyses of Fah Activity in Hepatized Lymph Nodes and Native Liver

      Occasionally, small intrahepatic nodules of donor hepatocytes were identified in native tyrosinemic livers of IP-injected mice (Figure 1C). We hypothesized that hepatocytes might have drained from the lymphatic system into the bloodstream via the subclavian vein and, subsequently, into the liver. Fah enzyme activity was measured
      • Knox W.E.
      • Edwards S.W.
      Homogentisate oxidase of liver.
      to estimate the number of donor hepatocytes in the native liver vs hepatized lymph nodes, and, more importantly, to determine if they contribute to the restoration of liver function (Figure 4E). Fah enzyme activities in hepatized lymph nodes ranged from 80% to almost 100% of WT liver levels. In contrast, Fah enzyme activities in native tyrosinemic liver had a mean activity close to 15% of WT liver levels. There was a complete lack of Fah enzyme activity in the native tyrosinemic liver of 1 of 5 mice, which is comparable with untreated Fah−/− mice. No significant correlation was found between the level of serum bilirubin (an indicator of liver function) and the Fah activity level in native livers of the transplanted Fah−/− mice (an indicator of WT hepatocytes engrafted in the liver) (Supplementary Figure 1). These results indicate that the presence of WT hepatocytes sometimes found in the liver of the Fah mice could not explain the massive expansion of liver cells in lymph nodes and restoration of liver functions observed with the survival of the animals.

       Expansion of the Hepatized Lymph Nodes After Partial Hepatectomy

      Partial hepatectomy is known to lead to regeneration of the remnant liver.
      • Michalopoulos G.K.
      Liver regeneration.
      Here we asked whether hepatized lymph nodes would respond to the same regenerative triggers after partial hepatectomy. Fah−/− mice were transplanted IP with 106 WT hepatocytes. Ten weeks after hepatic engraftment in lymph nodes, a partial hepatectomy was performed in 3 of the 6 transplanted Fah−/− mice. Hepatic regeneration was induced by surgically removing the median and left lateral hepatic lobes, representing two thirds of the liver mass. Three weeks later, all 6 mice were killed and their livers and hepatized lymph nodes were harvested. Partial hepatectomy showed that further resection of the native liver results in an expansion of the hepatized lymph nodes with survival of the animal (Figure 5). This result provides additional evidence that hepatized lymph nodes are responding to homeostatic mechanisms regulating the maintenance of liver tissue mass and liver function after injury.
      Figure thumbnail gr5
      Figure 5Expansion of hepatized lymph nodes after hepatectomy. (A) Anatomic location of hepatized lymph nodes in IP-injected Fah−/− mice after hepatectomy. Enlarged nodules were found around the stomach region and on the mesenterium (yellow circles). (B) Native liver and extrahepatic nodules after hepatectomy from the mouse on the left panel. The native tyrosinemic liver of the IP-injected Fah−/− mouse was atrophic and the enlarged nodules had a diameter of 3–15 mm. (C) The ratio ± standard error of the mean of the weight of native liver and hepatized lymph nodes to body weight. The ratio of hepatic tissues to body weight was determined between transplanted Fah−/− mice with or without hepatectomy. The hepatized lymph nodes show a significant increase in their weight after hepatectomy (P = .0052). n = number of mice analyzed.

       Hepatocytes Derived From Hepatized Lymph Nodes Are Not Tumorigenic

      Hepatocyte migration and invasion into the lymph nodes represent a profound change in the morphology and behavior of epithelial cells reminiscent of the metastatic process. However, such profound changes in behavior of epithelial cells are not always correlated with tumor progression and have been observed during embryonic development.
      • Yang J.
      • Weinberg R.A.
      Epithelial-mesenchymal transition: at the crossroads of development and tumor metastasis.
      Even in adult life, lymph nodes can contain benign inclusions of epithelial cells without malignant disease. To rule out malignant transformation, hepatocytes were isolated from hepatized lymph nodes and retransplanted via splenic injection into secondary Fah−/− recipients (n = 5). Three to 6 months after transplantation, none of the rescued animals showed the presence of either hepatocytes in the lymphatic system or tumors. The mean hepatocyte repopulation in the liver was 85.3%, indicating the similar transplantability and therapeutic effect of lymph node–derived hepatocytes when compared with liver-derived hepatocytes (Figure 6).
      Figure thumbnail gr6
      Figure 6Serial transplantation of lymph node–derived hepatocytes. Hepatocytes isolated from hepatized lymph nodes were serially transplanted into Fah−/− mice by SP injection. (A) Body weight of Fah−/− mice after splenic transplantation. The body weight lost and spontaneous gain after lymph node–derived hepatocyte transplantation is very similar to the change observed when liver-derived hepatocytes are transplanted. Two selections were necessary owing to the low number of hepatocytes transplanted. (B) Fah+ hepatocytes were observed only in the repopulated liver of Fah−/− mice 8 weeks after transplantation. Counterstaining was performed with eosin. Bar: 100 μm.

      Discussion

      Organ transplantation is too often the last resort for patients suffering from terminal disease. It is thought that tissue engineering and regenerative medicine have the potential to solve some of the problems associated with organ transplantation. Although the liver is an extraordinary organ because of its regenerative properties, engineered liver organogenesis is not yet a viable therapeutic option. Our goal was to identify an in vivo location where ectopic liver organogenesis would be feasible. In this study, we show that hepatocytes survive in lymph nodes and generate functional hepatized lymph nodes in an animal model of type I tyrosinemia. Transplantation of liver cells in the peritoneal cavity allowed the hepatocytes to migrate into the lymphatic system, enter the lymph nodes, and expand under homeostatic mechanisms driven by the liver at the expense of lymph node lymphocytes. When ectopic liver tissue reached the required balance for hepatic function, proliferation ceased, resulting in 20 to 40 hepatized lymph nodes that represented 70% of the original liver mass. In addition, we did not observe possible complications such as ascites or lower-extremity edema expected by the intra-abdominal lymphadenopathy. Partial hepatectomy of the diseased liver further expanded the mass of hepatized lymph nodes, indicating that tight homeostatic control of liver mass was retained at the ectopic site.
      We speculate that the highly vascularized nature of the lymph nodes supports the efficient engraftment and massive expansion of the ectopic tissue. It has been proposed that an inadequate vascular supply leads to hepatocyte death as a result of hypoxia within a few days of ectopic transplantation.
      • Gupta S.
      • Vemuru R.P.
      • Lee C.D.
      • et al.
      Hepatocytes exhibit superior transgene expression after transplantation into liver and spleen compared with peritoneal cavity or dorsal fat pad: implications for hepatic gene therapy.
      • Smith M.K.
      • Mooney D.J.
      Hypoxia leads to necrotic hepatocyte death.
      Because lymph nodes are intended for the support of lymphocyte proliferation and expansion,
      • von Andrian U.H.
      Intravital microscopy of the peripheral lymph node microcirculation in mice.
      they may be better suited to the immediate survival of engrafting hepatocytes. Therefore, lymph nodes could be compared with a well-designed in vivo bioreactor originally built for the rapid expansion of lymphocytes but now retasked for colonization by hepatocytes. Although major advances have been made in scaffold and cell technology, the vascularization and nourishment of large tissues remains an engineering problem that could be resolved when hepatocytes engraft in lymph nodes.
      The mechanism of expansion of hepatocytes in lymph nodes of Fah−/− mice appeared similar to the very complex phenomenon of the regenerative process after partial hepatectomy. In this process, well-orchestrated signaling cascades, characterized over many years, direct the restoration of the lost hepatic mass.
      • Michalopoulos G.K.
      Liver regeneration.
      However, in our experiments it remains unclear how hepatocytes or subpopulations of hepatocytes enter the lymph nodes. The mechanism of active or passive infiltration, neutralization of lymphocytes, and reorganization of the architecture of the nodes to mimic functional hepatic tissue remains to be determined. We do know that the expansion of hepatocytes in lymph nodes is associated with essential factors of hepatic growth such as HGF. We have detected HGF during the development of ectopic liver tissue in the lymph nodes at 6 weeks post-transplant, when the massive expansion of liver cells in lymph nodes rescues animal survival. We postulate that hepatocytes in the lymph nodes are susceptible to a mitotic stimulus much like hepatocytes after partial hepatectomy, responding to similar growth factors and expressing similar transcriptional cascades, resulting in the restoration of liver homeostasis. Additional mechanisms for hepatocyte invasion and engraftment in the lymph nodes remain to be determined.
      Although biliary canaliculi with microvilli could be observed on transmission electron microscopy, biliary morphogenesis revealed by CK-19 staining could not be detected in hepatized lymph nodes. Surprisingly, IP injection of hepatocytes resulted in a decrease in conjugated bilirubin, but not to normal levels. Our proposed mechanism for the improved hyperbilirubinemia is the excretion of conjugated bilirubin by the diseased liver. This hypothesis is based on several observations. The presence of bile in the gallbladder of the animals indicates excretion by the native tyrosinemic liver (Supplementary Figure 1). The histology of the native tyrosinemic liver after rescue by IP injection supports these findings; the biliary cells appear normal and preserved whereas hepatocytes have abnormal morphology. Finally, we show no significant correlation between serum bilirubin levels and some of the low engraftment of the donor hepatocytes in the native livers, indicating that low levels of WT hepatocytes sometimes found in the liver of the Fah mice could not explain the improved hyperbilirubinemia (Supplementary Figure 1). All these observations indicate that the ectopic liver tissue complements some of the functions of the native liver, which results in the improvement in hyperbilirubinemia observed in IP-injected mice. Alternatively, conjugated bilirubin and bile salts could be excreted by the kidneys, especially in view of the very high glomerular filtration rate in mice, compared with that in other laboratory animals or human beings.
      Furthermore, both ER-TR7, a marker for connective tissue fibers, and F4/80, a marker for Kupffer cells, were absent in the hepatized lymph nodes, but present in a normal liver. These results indicate that the architecture and cellular content of the ectopic liver tissues in lymph nodes differ from normal liver. However, hepatized lymph nodes will normalize most of the liver functions, including albumin, fibronectin, urea, and lipid metabolism, and, most importantly, provide liver functions necessary for long-term survival of the tyrosinemic mouse.
      Patients with end-stage chronic liver disease have progressive hepatic failure that precludes any possible repair by the native liver. Liver disease in the Fah−/− model also results in progressive hepatic failure, preventing any possible repair by the native liver. Only long-term survival of functional donor hepatic tissue can contribute to survival of the Fah−/− mouse. Our successful approach of generating functional ectopic liver tissue suggests that it may be possible one day to apply this methodology in patients suffering from liver diseases when liver transplantation or regeneration of the liver by cell-based therapy are not possible. Targeting lymph nodes for liver cell transplantation could be an approach to limit growth of hepatocytes to specific lymphatic sites. However, it still needs to be shown that human hepatic insufficiency, in general, or in particular conditions such as cirrhosis, can lead to a selective advantage for the transplanted liver cells. Although conditioning protocols for hepatocyte repopulation after liver cell transplantation have been reported by other investigators,
      • Gupta S.
      • Vemuru R.P.
      • Lee C.D.
      • et al.
      Hepatocytes exhibit superior transgene expression after transplantation into liver and spleen compared with peritoneal cavity or dorsal fat pad: implications for hepatic gene therapy.
      • Laconi E.
      • Oren R.
      • Mukhopadhyay D.K.
      • et al.
      Long-term, near-total liver replacement by transplantation of isolated hepatocytes in rats treated with retrorsine.
      • Guha C.
      • Sharma A.
      • Gupta S.
      • et al.
      Amelioration of radiation-induced liver damage in partially hepatectomized rats by hepatocyte transplantation.
      • Guha C.
      • Parashar B.
      • Deb N.J.
      • et al.
      Normal hepatocytes correct serum bilirubin after repopulation of Gunn rat liver subjected to irradiation/partial resection.
      no effective protocol has been established and it is difficult to use any conditioning protocols reported for clinical application at present. Immunologic barriers may limit allogenic transplantation of hepatocytes. Allogenic hepatocytes, in the context of lymphatic tissue, should be investigated for the treatment of human hepatic failure. Recently, induced pluripotent stem cells have been established from fibroblasts and other mature somatic cells, indicating that immunologic issues affecting allogenic transplantation may be circumvented in the future.
      • Nakagawa M.
      • Koyanagi M.
      • Tanabe K.
      • et al.
      Generation of induced pluripotent stem cells without Myc from mouse and human fibroblasts.
      • Yamanaka S.
      Induction of pluripotent stem cells from mouse fibroblasts by four transcription factors.
      In conclusion, the therapeutic efficacy of hepatized lymph nodes in restoring liver function may represent a unique opportunity to treat certain patients with end-stage liver disease.

      Acknowledgments

      The authors would like to thank Dr Markus Grompe for providing the substrate for the fumarylacetoacetate hydrolase enzymatic assay, and Lynda Guzik, Dr Ira Fox, Dr Lindsey Boone, and Dr Aaron DeWard for editorial assistance.

      Supplementary Materials and Methods

       Antibodies

      The antibodies used in the immunohistochemistry were as follows: the primary antibodies used were phycoerythrin-conjugated rat anti-mouse CD31, CD45, CD45/B220, Gr-1, CD3, CD4, CD8, and CD11b; fluorescein isothiocyanate–conjugated rat anti-mouse CD26, rat anti-mouse peripheral lymph node addressin (MECA-79, CD62L ligand) (BD Biosciences Pharmingen, San Jose, CA); purified goat anti-mouse albumin (Bethyl Laboratories, Inc, Montgomery, TX), mouse anti-mouse CK18, goat anti-mouse CK19, goat anti-desmin, anti-bromodeoxyuridine (Santa Cruz Biotechnology, Inc, Santa Cruz, CA), rabbit anti-GFAP (Dako North America, Inc, Carpinteria, CA), rat anti-mouse F4/80, rat anti-mouse E-Cadherin (Invitrogen Co, Carlsbad, CA), rat anti-mouse ER-TR7 (Abcam, Cambridge, MA). The secondary antibodies used were fluorescein isothiocyanate–conjugated chicken anti-rat immunoglobulin G (IgG), phycoerythrin-conjugated chicken anti-rat IgG (Santa Cruz Biotechnology), goat anti-rabbit IgG AlexaFluor 594, chicken anti-goat IgG AlexaFluor 594 (Molecular Probes, Inc, Eugene, OR), biotinylated anti-rat IgG, and streptavidin-phycoerythrin (BD Biosciences Pharmingen). All of the antibodies were diluted to the optimal concentration with phosphate-buffered saline.

       Biochemical Analysis

      Animals were anesthetized by tribromoethanol and the laparotomy was performed to expose the inferior vena cava. Blood was collected from the inferior vena cava and immediately mixed with 10 μL of Na-heparin for anticoagulation. Red blood cells were removed by a brief centrifugation and the plasma was frozen at −80°C. ALT, total/direct bilirubin, and amino acid analysis were performed by The Mouse Metabolic Phenotyping Center at the Yale University School of Medicine. Blood urea nitrogen, total cholesterol, and triglyceride analyses were performed by Antech Diagnostics (Morrisville, NC).
      Figure thumbnail gre1
      Supplementary Figure 1Phenotypically normal biliary cells in the Fah−/− mice after IP transplantation. (A) Portal triad of the native liver in Fah−/− mice 10 weeks after IP injection, SP injection, and WT mouse. Upper panel: H&E staining, lower panel: anti-Fah staining with Fah-positive hepatocytes (brown, horseradish-peroxidase) counterstained with eosin. Arrows in the upper left panel (IP injection) highlight the bile ducts that appear normal in the Fah−/− mouse transplanted IP, however, hepatocytes seen in both the upper and lower left panels have clearly abnormal morphology (enlarged cells with pyknotic nuclei), which is indicative of a diseased tyrosinemic liver. WT mice are control animals. (B) Presence of bile in gallbladder (blue circle) of 2 independent Fah−/− mice 10 weeks after IP injection. This result indicates that bile is processed in the native liver of these mice. The presence of hepatized lymph nodes is highlighted by the yellow circles in both mice. (C) Analysis of the correlation between low hepatic engraftment in the native liver of Fah−/− mice and total serum bilirubin levels. Native livers of Fah mice (n = 5) were isolated 10 weeks post-IP injection of 106 hepatocytes, and Fah enzyme activity was determined as described previously in the Materials and Methods section. Total serum bilirubin levels were obtained for each mouse. The sample Pearson correlation coefficient (−0.401) was obtained from SPSS (Predictive Analytics Software, IBM, Somers, NY) by entering total serum bilirubin as the dependent variable and Fah enzyme levels as the independent variable. These data indicate a weak or poor correlation between Fah enzyme activity in the native liver and total bilirubin levels. Therefore, there is a low statistical likelihood that total bilirubin levels in the native liver of IP-injected Fah−/− mice are affected by residual Fah enzyme activity and consequently the low WT hepatocyte engraftment in the native tyrosinemic liver. The partial restoration of the biliary function in IP-injected mice raises the possibility that the bile produced by hepatocytes in the engrafted lymph nodules was excreted into the blood via bile canaliculi and the lymphatic system. We postulate that the native tyrosinemic liver partially processes and excretes bile into the intestine via the common bile duct, thereby keeping serum bilirubin levels lower than in untreated tyrosinemic animals. These data suggest that the atrophic tyrosinemic liver complements the function of the ectopic liver in the lymph nodes.

      References

        • Perkins J.D.
        • Halldorson J.B.
        • Bakthavatsalam R.
        • et al.
        Should liver transplantation in patients with model for end-stage liver disease scores <or= 14 be avoided?.
        Liver Transpl. 2009; 15: 242-254
        • Volk M.L.
        • Hernandez J.C.
        • Lok A.S.
        • et al.
        Modified Charlson comorbidity index for predicting survival after liver transplantation.
        Liver Transpl. 2007; 13: 1515-1520
        • Lipshutz G.S.
        • Busuttil R.W.
        Liver transplantation in those of advancing age: the case for transplantation.
        Liver Transpl. 2007; 13: 1355-1357
        • Fisher R.A.
        • Strom S.C.
        Human hepatocyte transplantation: worldwide results.
        Transplantation. 2006; 82: 441-449
        • Strom S.C.
        • Chowdhury J.R.
        • Fox I.J.
        Hepatocyte transplantation for the treatment of human disease.
        Semin Liver Dis. 1999; 19: 39-48
        • Ito M.
        • Nagata H.
        • Miyakawa S.
        • et al.
        Review of hepatocyte transplantation.
        J Hepatobiliary Pancreat Surg. 2009; 16: 97-100
        • Nussler A.
        • Konig S.
        • Ott M.
        • et al.
        Present status and perspectives of cell-based therapies for liver diseases.
        J Hepatol. 2006; 45: 144-159
        • Lorenzini S.
        • Andreone P.
        Stem cell therapy for human liver cirrhosis: a cautious analysis of the results.
        Stem Cells. 2007; 25: 2383-2384
        • Lorenzini S.
        • Gitto S.
        • Grandini E.
        • et al.
        Stem cells for end stage liver disease: how far have we got?.
        World J Gastroenterol. 2008; 14: 4593-4599
        • Gupta S.
        • Vemuru R.P.
        • Lee C.D.
        • et al.
        Hepatocytes exhibit superior transgene expression after transplantation into liver and spleen compared with peritoneal cavity or dorsal fat pad: implications for hepatic gene therapy.
        Hum Gene Ther. 1994; 5: 959-967
        • Ohashi K.
        • Waugh J.M.
        • Dake M.D.
        • et al.
        Liver tissue engineering at extrahepatic sites in mice as a potential new therapy for genetic liver diseases.
        Hepatology. 2005; 41: 132-140
        • Seglen P.O.
        Preparation of isolated rat liver cells.
        Methods Cell Biol. 1976; 13: 29-83
        • Lagasse E.
        • Connors H.
        • Al-Dhalimy M.
        • et al.
        Purified hematopoietic stem cells can differentiate into hepatocytes in vivo.
        Nat Med. 2000; 6: 1229-1234
        • Knox W.E.
        • Edwards S.W.
        Homogentisate oxidase of liver.
        J Biol Chem. 1955; 216: 479-487
        • Grompe M.
        • Lindstedt S.
        • al-Dhalimy M.
        • et al.
        Pharmacological correction of neonatal lethal hepatic dysfunction in a murine model of hereditary tyrosinaemia type I.
        Nat Genet. 1995; 10: 453-460
        • Overturf K.
        • al-Dhalimy M.
        • Ou C.N.
        • et al.
        Serial transplantation reveals the stem-cell-like regenerative potential of adult mouse hepatocytes.
        Am J Pathol. 1997; 151: 1273-1280
        • Rhim J.A.
        • Sandgren E.P.
        • Degen J.L.
        • et al.
        Replacement of diseased mouse liver by hepatic cell transplantation.
        Science. 1994; 263: 1149-1152
        • Overturf K.
        • Al-Dhalimy M.
        • Tanguay R.
        • et al.
        Hepatocytes corrected by gene therapy are selected in vivo in a murine model of hereditary tyrosinaemia type I.
        Nat Genet. 1996; 12: 266-273
        • Girard J.P.
        • Springer T.A.
        High endothelial venules (HEVs): specialized endothelium for lymphocyte migration.
        Immunol Today. 1995; 16: 449-457
        • Michalopoulos G.K.
        Liver regeneration.
        J Cell Physiol. 2007; 213: 286-300
        • Yang J.
        • Weinberg R.A.
        Epithelial-mesenchymal transition: at the crossroads of development and tumor metastasis.
        Dev Cell. 2008; 14: 818-829
        • Smith M.K.
        • Mooney D.J.
        Hypoxia leads to necrotic hepatocyte death.
        J Biomed Mater Res A. 2007; 80: 520-529
        • von Andrian U.H.
        Intravital microscopy of the peripheral lymph node microcirculation in mice.
        Microcirculation. 1996; 3: 287-300
        • Laconi E.
        • Oren R.
        • Mukhopadhyay D.K.
        • et al.
        Long-term, near-total liver replacement by transplantation of isolated hepatocytes in rats treated with retrorsine.
        Am J Pathol. 1998; 153: 319-329
        • Guha C.
        • Sharma A.
        • Gupta S.
        • et al.
        Amelioration of radiation-induced liver damage in partially hepatectomized rats by hepatocyte transplantation.
        Cancer Res. 1999; 59: 5871-5874
        • Guha C.
        • Parashar B.
        • Deb N.J.
        • et al.
        Normal hepatocytes correct serum bilirubin after repopulation of Gunn rat liver subjected to irradiation/partial resection.
        Hepatology. 2002; 36: 354-362
        • Nakagawa M.
        • Koyanagi M.
        • Tanabe K.
        • et al.
        Generation of induced pluripotent stem cells without Myc from mouse and human fibroblasts.
        Nat Biotechnol. 2008; 26: 101-106
        • Yamanaka S.
        Induction of pluripotent stem cells from mouse fibroblasts by four transcription factors.
        Cell Prolif. 2008; 41: 51-56