Potential conflict of interest: Nothing to report.
Liver regeneration after surgical partial hepatectomy (PH) in retrorsine-exposed rats is accomplished through the outgrowth and expansion of small hepatocyte-like progenitor cells (SHPCs). The cells of origin for SHPCs and their tissue niche have not been identified. Nevertheless, some investigators have suggested that SHPCs may represent an intermediate or transitional cell type between oval cells and mature hepatocytes, rather than a distinct progenitor cell population. We investigated this possibility through the targeted elimination of oval cell proliferation secondary to bile duct destruction in retrorsine-exposed rats treated with 4,4′-diaminodiphenylmethane (DAPM). Fischer 344 rats were treated with 2 doses (30 mg/kg body weight) retrorsine (at 6 and 8 weeks of age) followed by PH 5 weeks later. Twenty-four hours before PH, select animals were given a single dose of DAPM (50 mg/kg). Treatment of rats with DAPM produced severe bile duct damage but did not block liver regeneration. Oval cells were never seen in the livers of DAPM-treated retrorsine-exposed rats after PH. Rather, liver regeneration in these rats was mediated by the proliferation of SHPCs, and the cellular response was indistinguishable from that observed in retrorsine-exposed rats after PH. SHPC clusters emerge 1 to 3 days post-PH, expand through 21 days post-PH, with normalization of the liver occurring by the end of the experimental interval. Conclusion: These results provide direct evidence that SHPC-mediated liver regeneration does not require oval cell activation or proliferation. In addition, these results provide strong evidence that SHPCs are not the progeny of oval cells but represent a distinct population of liver progenitor cells. (HEPATOLOGY 2007.)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The mammalian liver possesses enormous capacity and cellular flexibility in its ability to replace tissue lost to surgical resection, necrosis, or other injury.1–4 Several cell populations are capable of being activated to repair or regenerate the liver depending on the nature and extent of injury.5 In an otherwise normal liver, the restoration of damaged or resected liver tissue is accomplished through the proliferation of mature hepatocytes present in the remaining viable tissue.1, 2, 4, 6 However, in certain pathological conditions, mature hepatocytes are incapable of dividing, and reserve progenitor cell populations respond to restore liver mass and function.1–3, 7, 8 At least 2 reserve progenitor cell populations have been described for the adult rat liver: oval cells and small hepatocyte-like progenitor cells (SHPCs). Oval cells originate from cells located in the portal tracts (possibly from biliary epithelial cells) and proliferate to regenerate the liver in several well-characterized models of rat liver injury.2, 3, 5, 9 In contrast, SHPCs proliferate to regenerate the liver tissue of rats exposed to the pyrrolizidine alkaloid retrorsine.1, 10, 11 The participation of SHPCs in other models of liver injury and regeneration has not been determined. Likewise, the cells of origin of SHPCs and their precise tissue niche are not known. Some investigators have suggested that SHPCs may represent an intermediate or transitional cell type between oval cells and mature hepatocytes.2, 12 However, other possibilities for the cellular origin of SHPCs exist. Some evidence supports the suggestion that these cells may represent a distinct immature progenitor cell population.1, 10, 11, 13 A third possibility is that SHPCs represent a population of “retrorsine-resistant” mature hepatocytes.1, 10, 14
In the current study, we directly investigated the possible precursor–product relationship between oval cells and SHPCs by treating retrorsine-exposed rats with the toxin DAPM (4,4′-diaminodiphenylmethane). DAPM specifically destroys the bile ducts of rat liver, resulting in the complete abolishment of oval cell–mediated regenerative responses to liver deficit or injury.9 DAPM treatment results in the highly specific destruction of bile ducts in the livers of retrorsine-exposed rats, resulting in a complete blockade of oval cell proliferation after partial hepatectomy (PH). However, the DAPM-mediated inhibition of oval cell proliferation did not affect the emergence of SHPCs after PH. SHPCs appear at early times after PH in the livers of DAPM-treated retrorsine-exposed rats and proliferate in a manner that is indistinguishable from that observed after PH in retrorsine-exposed animals in the absence of DAPM.10 These results strongly suggest that SHPCs are not the progeny of oval cells or other periductular progenitor cell population, and that oval cell activation and proliferation are not required for SHPC-mediated liver regeneration Rather, this investigation provides strong evidence that SHPCs represent a distinct parenchymal progenitor cell population capable of restoring hepatocyte numbers and liver tissue mass in response to certain forms of liver injury in which mature hepatocytes are incapable of proliferating.
Fischer 344 (F344) male rats were either purchased (Charles River Laboratories, Wilmington, MA) or bred in-house, and maintained in the Association for Assessment and Accreditation of Laboratory Animal Care International–accredited animal facilities of the University of North Carolina at Chapel Hill. All procedures involving animals were carried out in accordance with federal and state guidelines put forth by the National Institutes of Health and the Institutional Animal Care and Use Committee of the University of North Carolina at Chapel Hill.
Retrorsine, DAPM Administration, and Partial Hepatectomy.
Male 6-week-old littermate Fischer 344 rats were randomized into retrorsine treatment (n = 72) and control (n = 91) groups at the outset of the experiment. Rats in the retrorsine treatment group received 2 treatments of retrorsine (30 mg/kg intraperitoneally) 2 weeks apart (at 6 and 8 weeks of age). Retrorsine (12,18-dihydroxysenecionan-11,16-dione; β-Longilobine, Sigma Chemical Co., St. Louis, MO) was added to distilled water at 10 mg/mL and titrated to pH 2.5 with 1 N hydrochloric acid to completely dissolve the solid. Subsequently, the solution was neutralized using 1 N sodium hydroxide, and sodium chloride was added for a final concentration of 6 mg/mL retrorsine and 0.15 M sodium chloride (pH 7.0). The working solution was used immediately after preparation. Four weeks after the second retrorsine treatment, experimental and age-matched control rats were randomized into the following experimental groups: retrorsine/PH (n = 27), DAPM/retrorsine/PH (n = 45), DAPM/PH (n = 29), DAPM only (n = 20), dimethylsulfoxide (DMSO)/PH (n = 23), and DMSO only (n = 19) (Fig. 1). Five weeks after the second retrorsine treatment, two-thirds surgical PH was performed on selected experimental rats and control rats of the same age (13 weeks old), essentially as originally described.15 Selected animals were treated with DAPM (50 mg/kg intraperitoneally) or DMSO (vehicle) 24 hours before PH. DAPM (4,4′-diaminodiphenylmethane, Sigma Chemical Company, St. Louis, MO) was prepared by dissolving the solid directly into DMSO (Sigma Chemical Co.) at a concentration of 50 mg/mL. The working solution was used immediately after preparation. Very little mortality was associated with these experimental treatments, with 83% (n = 135) of animals surviving the experimental protocol. At 3, 7, 10, 14, 21, and 30 days post-PH (n = 3-6 rats per time point), liver tissues were harvested and fixed in 10% neutral buffered formalin. Body weights and liver weights were recorded at the time of PH, and tissue harvest and liver/body ratios were calculated from these data.
Colorimetric immunoperoxidase analysis was carried out on paraffin-embedded tissue sections using standard procedures. Tissue sections were deparaffinized in xylene, washed through a gradient of ethanol solutions (100%, 95%, and 70%), and rehydrated in a phosphate-buffered saline solution (1.54 mM KH2PO4, 155.17 mM sodium chloride, 2.71 mM NaH2PO4-7H2O; pH 7.2). Endogenous peroxidase activity was quenched for 10 minutes using a 0.3% hydrogen peroxide solution diluted in 100% methanol. Antigen retrieval was achieved by placing tissue sections in a heated 10 mM citrate antigen retrieval buffer (DakoCytomation, Carpinteria, CA) and subsequently placing the slide chamber (containing citrate buffer) in a steamer for 30 minutes. Nonspecific antibody binding was blocked with a serum-free protein block (DakoCytomation), and primary antibodies were detected with the DakoCytomation Labelled Streptavidin-Biotin2, Horseradish Peroxidase (LSAB2, HRP) system (DakoCytomation). The secondary antibody was detected using a substrate containing diaminobenzidine (DakoCytomation). Tissue sections were counterstained using Mayer's hematoxylin (Sigma Chemical Co.). Primary antibodies were diluted using antibody diluent with background reducing components (DakoCytomation). Mouse anti-human cytokeratin 19 antibody (DakoCytomation) was used at a dilution of 1:25. Goat anti-rat α-fetoprotein antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was used at a dilution of 1:100. Rabbit anti-bovine albumin antibody (Cappel, Solon, OH) was used at a dilution of 1:200.
Histology and Morphometry.
Routine paraffin-embedding, sectioning, and preparation of hematoxylin-eosin (HE) stained tissue sections from formalin-fixed liver tissue was performed by Histo-Scientific Research Laboratories (Mount Jackson, VA). Morphometric analysis was performed by digitally scanning HE-stained slides using an Aperio Scanscope T2 Virtual Microscope System (Vista, CA) at a resolution of 0.4667 μm/pixel. Images were analyzed using Aperio Imagescope v6.25 software. SHPC clusters were identified in the H&E sections based upon cell morphology and arrangement. Using the various tools provided by the Imagescope v6.25 software, SHPC clusters were enumerated and measured (area). All data obtained using Imagescope v6.25 were normalized to the cross-sectional area of the tissue section as determined using Image J v1.36 software (National Institutes of Health, Bethesda, MD).
The numerical data presented in figures and tables represent the mean ± standard error of the mean. All statistical analyses were performed using Kaleidograph v4.0 (Synergy Software, Reading, PA). Mean SHPC cluster sizes were calculated using each individual cluster as a separate event. All additional analyses were performed using individual animals as separate events. Significance of quantitative data was calculated using the Student two-tailed t test with unequal variance among groups. A statistically significant difference in data was defined by a P < 0.05. Where applicable, NS denotes no significant difference in data presented.
DAPM Treatment Produces Severe Bile Duct Injury in Retrorsine-Exposed Rats.
Treatment of rats with a singe dose of DAPM results in the targeted destruction of bile ducts (Fig. 2), as previously reported.9, 16 The destruction of bile duct structures is rapid, with evidence of severe damage by 24 hours after DAPM treatment (Fig. 2A). DAPM-induced bile duct damage persists through 3 days post-DAPM treatment before recovery of biliary cell types is observed (data not shown). For several days after DAPM administration (in DAPM-only animals), remnants of bile duct structures are observed in damaged portal tracts, and immune cells recruited in areas of injury populate the necrotic lesions (data not shown). Morphologically normal bile duct structures are found in portal tracts 7 days after DAPM injection in DAPM-only treated animals (data not shown). In DAPM/PH animals, DAPM-induced bile duct damage occurs rapidly, with evidence of bile duct injury at 24 hours post-DAPM treatment (Fig. 2). However, recovery of biliary epithelial cells and bile duct structures is achieved over a shorter time interval compared with DAPM-only animals. In DAPM/PH rats, restoration of morphologically normal bile duct structures is observed at 3 days post-PH, corresponding to 4 days post-DAPM administration (Fig. 3I). Likewise, in DAPM/retrorsine plus partial hepatectomy (RP) rats bile duct damage is noted 24 hours after DAPM administration (Fig. 2), evidence of bile duct repair is observed at 3 days post-PH (4 days post-DAPM treatment), and normalization of bile duct structure occurs by 7 days post-PH (Fig. 3J-K). No bile duct injury is seen in DMSO-treated control animals at any time after injection (Fig. 2 and data not shown). Similarly, retrorsine treatment alone (in the absence of DAPM) does not produce notable bile duct injury.10
Liver Regeneration After Partial Hepatectomy in Retrorsine-Exposed Rats After DAPM Treatment.
At 3 days post-PH, animals in all treatment groups exhibit similar body weights, liver weights, and liver/body weight ratios (Fig. 4). These similarities persist through 7 days post-PH among animals in the RP and DAPM/RP treatment groups. At 7 days post-PH, these animals have liver weights (P = 0.561 for RP versus DAPM/RP) and liver/body weight ratios (P = 0.275 for RP versus DAPM/RP) that are not significantly different. However, control (DMSO/PH and DAPM/PH) animals have significantly increased (P < 0.05) liver weights and liver/body weight ratios compared with RP and DAPM/RP animals at this time point (Fig. 4B, C), reflecting a robust regenerative response in these animals. This trend continues through 10 and 14-days post-PH (Fig. 4B, C). Liver regeneration in DMSO/PH and DAPM/PH animals is mediated by proliferation of mature hepatocytes and is completed over a short period. However, at 10 and 14 days post-PH, the liver weights and liver/body weight ratios for DAPM/RP and RP rats are lower than those observed in control animals (Fig. 4B, C). The low liver weights are not attributable to a lack of cellular regeneration, but reflect loss of tissue mass related to apoptosis of retrorsine-injured hepatocytes (megalocytes) in retrorsine-exposed liver.17 By 21 days post-PH, an increase in the liver weights of retrorsine-exposed animals is observed (Fig. 4B). At this point, animals in the DAPM/RP group have liver weights comparable to those observed in RP rats (P = 0.453) but have liver/body weight ratios that are significantly (P = 0.032) lower than that observed in RP rats (Fig. 4C). The difference in liver/body weight ratio between RP and DAPM/RP animals reflects higher average body weights observed among DAPM/RP animals (versus RP animals), whereas the liver weights observed in these 2 treatment groups are indistinguishable at this point (Fig. 4). Control animals at 21 days post-PH have significantly higher (P < 0.05) body weights, liver weights, and liver/body weight ratios than those observed in the retrorsine-treated animals (Fig. 4). At the end of the 30-day post-PH period, animals in the RP and DAPM/RP groups have significantly (P < 0.05) lower liver weights, and liver/body weight ratios than those observed in DAPM/PH and DMSO/PH control groups, suggesting that regeneration is not complete and may be ongoing in these retrorsine-exposed animals (Fig. 4B, C).
Cellular Responses After PH in DAPM-Treated Retrorsine-Exposed Rats.
Oval cells originate from cells located in the portal tracts (possibly biliary epithelial cells) and expansively proliferate in several models of liver injury and regeneration.2, 3, 5, 9 DAPM-induced destruction of bile ducts effectively eliminates the oval cell reaction after PH in these models of liver injury secondary to bile duct injury.9 In the current study, oval cells were never observed in H&E-stained liver sections of DAPM-treated animals (Fig. 3). To verify the absence of an oval cell response after DAPM treatment in experimental rats, tissues from DAPM/RP animals were immunostained for ck19. At 3 and 7 days post-PH, residual (DAPM-damaged) biliary tracts are decorated by antibodies to ck19 in DAPM/RP animals, but proliferating oval cells are not observed (Fig. 5A, B). Likewise, the biliary tracts are ck19-positive at 3 and 7 days post-PH in RP animals, but oval cells are not observed (Fig. 5C, D). In contrast, liver harvested from a rat that was treated with 2-acetamidofluorene (2-AAF) shows a robust ck19-positive oval cell response after PH, as reported by others.18–20 At 7 days post-PH in 2-AAF–treated animals, oval cells are clearly present in HE-stained liver tissue (Fig. 5E). Ck19 immunostaining of liver tissue from these animals clearly labels proliferating oval cells (Fig. 5F). Together, these results demonstrate that oval cells are not present after PH in DAPM-treated retrorsine-exposed animals, suggesting strongly that oval cells are not the source of the SHPCs.
Examination of HE-stained liver tissues from DAPM/RP rats shows an SHPC-mediated regenerative response similar to that observed in RP rats (Fig. 3). Livers harvested from RP animals contain SHPC clusters that are easily recognized at 3 days post-PH (Fig. 3E). These cell clusters expand through 7 and 10 days post-PH (Fig. 3F, G), reach lobule size by 14 days post-PH (Fig. 3H), and eventually replace the injured hepatocytes and restore the normal liver structure (data not shown). Similarly, SHPC clusters are seen in liver tissues from DAPM/RP animals at 3 days post-PH (Fig. 3A). These cell clusters expand through 7 days and 10 days post-PH (Fig. 3B, C) and reach lobule size by 14 days post-PH (Fig. 3D). SHPC clusters in the livers of DAPM/RP rats ultimately replace damaged hepatocytes, resulting in the normalization of the liver parenchyma by 30 days post-PH (data not shown). Immunohistochemical analysis of livers from DAPM/RP–treated animals at 7 days post-PH demonstrated that these SHPC clusters are positive for both α-fetoprotein and albumin (Fig. 6), as reported.10, 11 Retrorsine-injured hepatocytes (megalocytes) express α-fetoprotein 7 days post-PH in reaction to the DAPM/RP injury (Fig. 6A, B). This reaction is similar to that observed when the liver is exposed to certain necrotic agents such as carbon tetrachloride.21, 22 These observations combine to suggest that the cellular regenerative response observed in DAPM/RP animals is indistinguishable from that seen in RP animals. SHPC clusters are never observed in DAPM/PH, DMSO/PH, DAPM-only, or DMSO-only control rats (Fig. 3 and data not shown). In addition, SHPCs are not observed in animals treated with retrorsine only in the absence of PH or after PH in control animals that do not receive retrorsine.10 Together, these observations suggest that liver regeneration in DAPM/RP rats is mediated by an oval cell–independent SHPC response and is histologically identical to that observed in RP rats.
Morphometric Analysis of SHPC Clusters in DAPM/RP Rats.
To determine whether the magnitude of SHPC responses observed in DAPM/RP and RP rats are similar, morphometric analyses on HE-stained liver sections from these animals at 7 and 14 days post-PH were performed. At 7 days post-PH, there is no significant difference in the number (P = 0.359) or size (P = 0.286) of the SHPCs clusters observed in DAPM/RP and RP animals (Table 1). Likewise, at 14 days post-PH there is no significant difference in the number (P = 0.122) or size (P = 0.268) of SHPC clusters observed in RP and DAPM/RP animals (Table 1). These data are consistent with the suggestion that the SHPC-mediated regenerative responses observed in RP rats and DAPM/RP rats are indistinguishable.
Table 1. Number and Size of SHPC Clusters in Liver Sections From DAPM-Treated Retrorsine-Exposed Animals
Numbers of SHPC clusters represent mean ± standard error of the mean (n = number of animals/tissue sections analyzed). The number of SHPC clusters in each section was normalized based on the total area of tissue section analyzed, using the area of the smallest section analyzed as the reference value.
Size of clusters (given in μm2) represents mean ± standard error of the mean (n = number of clusters analyzed per animal, between 3–5 animals analyzed per treatment group).
P = 0.359 (NS) for RP versus DAPM/RP at 7 days post-PH.
P = 0.286 (NS) for RP versus DAPM/RP at 7 days post-PH.
P = 0.122 (NS) for RP versus DAPM/RP at 14 days post-PH.
P = 0.268 (NS) for RP versus DAPM/RP at 14 days post-PH.
SHPCs were originally identified as the regenerative cell population in retrorsine-exposed rats after PH in the absence of transplanted hepatocytes.10 In this liver injury model, retrorsine exposure renders the mature hepatocyte population incapable of responding to liver deficit,23 resulting in the activation of progenitor cell populations. SHPCs emerge 1 to 3 days after PH, proliferate and expand during the subsequent interval, and restore the liver mass and structure by 30 days post-PH.10 SHPC-mediated liver regeneration in retrorsine-exposed rats is accompanied by a modest oval cell response. However, the observed magnitude of oval cell proliferation in retrorsine-exposed rats is significantly diminished compared with other well-characterized oval cell proliferation models of liver injury and regeneration (such as 2-AAF/PH), in which oval cells represent the regenerative cell population.24 SHPCs represent an interesting population of liver cells that exhibit an intermediate phenotype characterized by expression of markers of both immature and differentiated cell types. SHPCs observed 5 days after PH in retrorsine-exposed rats transiently express oval cell markers, including OV6, OC.2, and OC.5.10 Concurrent with the expression of these oval cell markers, emergent SHPCs express markers and characteristics of mature hepatocytes, including transferrin, albumin, H.4, glycogen, and bile canaliculi.10 Although the phenotype of SHPCs overlaps with that of oval cells and mature hepatocytes, it is distinct from each of these cell populations,10, 11 suggesting that SHPCs are not simply an intermediate cell type reflecting differentiating oval cells. Furthermore, emergent SHPC clusters are spatially separated from the portal tracts, which are known to be the niche for the progenitors of the oval cell reaction. In fact, emerging SHPC clusters are observed in all zones of the liver parenchyma in retrorsine-exposed rats,10 suggesting that these cells may represent a progenitor cell population that is sited in the parenchyma intimately associated with the hepatic plates. Despite the extensive characterization of the SHPC response in the retrorsine/PH model of rat liver injury and regeneration, the progenitors of SHPCs and their tissue niche are not yet established. Several known cell populations could represent the cells of origin of SHPCs, including (1) oval cells, (2) some other periductular cell population, (3) retrorsine-resistant (but otherwise mature) hepatocytes, or (4) some other parenchymal (hepatocyte-like) cell type.
Over the past several years, some investigators have suggested that oval cells and SHPCs represent the same (or a closely related) cell populations.2 This suggestion is primarily based on the observations that (1) SHPCs transiently express some oval cell markers,10, 11 and (2) differentiating oval cells form cell clusters that are histologically similar to proliferating SHPCs.1 However, the potential precursor–product relationship between oval cells and SHPCs has proved difficult to assess experimentally. In the past, elegant studies have been conducted using pulse labeling of proliferating oval cells to establish that these cells can give rise to differentiated hepatocyte progeny.25–28 However, this experimental approach cannot be applied to the retrorsine/PH model of liver injury and regeneration to identify the progenitors of SHPCs or their progeny. The inability to use this approach relates to the mechanism of mito-inhibition of mature hepatocytes by retrorsine. Retrorsine-injured hepatocytes synthesize DNA and attempt to undergo cell division in response to PH before growth arrest and megalocytosis.10 Thus, metabolic labeling studies cannot be effectively carried out in this model system. In an attempt to investigate the lineage relationship between oval cells and SHPCs, Vig et al.12 used the hepatitis B surface antigen transgenic mouse model of chronic liver injury29 in conjunction with retrorsine treatment. Using 3-dimensional mapping techniques, these investigators tracked oval cells streaming directly into α-fetoprotein–positive “SHPC clusters.”12 Based on their observations in this model system, Vig et al. concluded that SHPCs are the differentiated progeny of oval cells.12 However, there are several significant differences between the liver injury model employed by Vig et al. and the retrorsine model of liver injury and regeneration10, 23, 30 employed in the current study. The standard retrorsine model is based on treatment of F344 rats and generation of liver deficit through acute injury (usually surgical PH).23 In contrast, the study by Vig et al.12 employed a mouse model of chronic liver injury, in which liver progenitor cells are in a continual state of proliferation.12 A definitive SHPC lineage marker does not exist, making it difficult to assess whether the regenerative responses observed in retrorsine-treated hepatitis B surface antigen transgenic mice are identical to those observed in retrorsine-exposed F344 rats. Based on data from our studies and others,13, 31 it seems likely that the hepatocyte clusters observed by Vig et al. actually represent new hepatocyte progeny of oval cells rather than regenerative SHPCs. This supposition is bolstered by data in the current study that demonstrate that when oval cell–mediated regenerative responses are blocked through the use of DAPM, the SHPC-mediated regenerative response is unaffected, suggesting that oval cells are not the source of SHPCs.
Oval cells are known to proliferate robustly after PH in animals treated with the mito-inhibitory agent 2-AAF.1–3, 24, 31, 32 As such, oval cells are not susceptible to 2-AAF poisoning. Therefore, it would be expected that if SHPCs are the progeny of oval cells they too would be refractory to the toxic effects of 2-AAF (at least initially). However, data from our recent studies show that retrorsine-resistant SHPCs are susceptible to 2-AAF poisoning.13 In these studies, SHPCs were never found in the livers of 2-AAF–treated retrorsine-exposed animals at any time after PH, but clustering of new hepatocytes (oval cell progeny) occurred by 14 days post-PH.13 The cellular response in 2-AAF–treated retrorsine-exposed rats is identical to that observed in 2-AAF/PH animals (in the absence of retrorsine), suggesting that the presence of retrorsine has no effect on oval cell–mediated liver regeneration, but that 2-AAF treatment blocks the outgrowth of SHPCs.13 In addition, treatment of rats with 2-AAF 7 days after initiation of the RP protocol (at a time when SHPCs are evident) results in a blockade of SHPC proliferation.13 These observations demonstrate that SHPCs in early regenerative clusters are susceptible to 2-AAF poisoning, strongly suggesting that oval cells are not the progenitor cell of origin of the SHPCs.
In the current study, we directly addressed the potential precursor–product relationship between oval cells and SHPCs by treating retrorsine-exposed rats with the bile duct toxin DAPM. Oval cells (or their progenitors) reside in the periportal region of the liver and may be directly derived from biliary epithelial cells. Previous studies have demonstrated that elimination of bile ducts and biliary epithelial cells through exposure of rats to the biliary toxin DAPM results in a complete blockade of oval cell activation and proliferation.9 DAPM-mediated inhibition of oval cell proliferation is observed in well-characterized models of oval cell–mediated liver regeneration, including the 2-AAF/PH model.9 Based on these observations, DAPM provides a powerful tool for investigation of the potential oval cell origin of SHPCs. Our results show that the destruction of the biliary epithelium by DAPM did not block (or significantly modify) the SHPC-mediated regenerative response in retrorsine-exposed rats. SHPC-mediated liver regeneration in DAPM-treated retrorsine-exposed animals after PH was histologically indistinguishable from that of retrorsine-exposed animals in the absence of DAPM, suggesting that retrorsine-resistant SHPCs are also refractory to the effects of DAPM. These results provide the first direct evidence that SHPC-mediated liver regeneration in retrorsine-exposed rats does not require oval cell activation or proliferation. Based on this observation, we conclude that SHPCs are not the progeny of oval cells or similar periductular cell population. This leaves open the possibility that SHPCs are derived from an as yet unidentified parenchymal cell population that appears histologically similar to mature (fully differentiated) hepatocytes, but phenotypically less mature.