SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The role of progenitor cells in liver repair and fibrosis has been extensively described, but their purification remains a challenge, hampering their characterization and use in regenerative medicine. To address this issue, we developed an easy and reproducible liver progenitor cell (LPC) isolation strategy based on aldehyde dehydrogenase (ALDH) activity, a common feature shared by many progenitor cells. We demonstrate that a subset of nonparenchymal mouse liver cells displays high levels of ALDH activity, allowing the isolation of these cells by fluorescence-activated cell sorting. Immunocytochemistry and qPCR analyses on freshly isolated ALDH+ cells reveal an enrichment in cells expressing liver stem cell markers such as EpCAM, CK19, CD133, and Sox9. In culture, the ALDH+ population can give rise to functional hepatocyte-like cells as illustrated by albumin and urea secretion and cytochrome P450 activity. ALDH1A1 expression can be detected in canals of Hering and bile duct epithelial cells and is increased on liver injury. Finally, we showed that the isolation and differentiation toward hepatocyte-like cells of LPCs with high ALDH activity is also successfully applicable to human liver samples. Conclusion: High ALDH activity is a feature of LPCs that can be taken advantage of to isolate these cells from untreated mouse as well as human liver tissues. This novel protocol is practically relevant, because it provides an easy and nontoxic method to isolate liver stem cells from normal tissue for potential therapeutic purposes. (HEPATOLOGY 2012)

Functional hepatocyte cultures are a crucial tool in the development of new therapeutic agents, providing the best available surrogate for liver toxicity testing. In addition, transplantation of freshly isolated human hepatocytes can be a cellular source for hepatic parenchymal replacement therapy, because the number of needed donor organs exceeds the demand. Additionally, for certain liver diseases even a partial correction of gene function can be curative, and thus a complete liver transplantation is not always necessary.1, 2 The major limiting factor in these applications is the availability of sufficient good-quality hepatocytes. One of the promising properties of stem cells is their potential to provide a renewable source of hepatocytes.3 From this perspective, much is expected from the use of embryonic stem cells, their induced counterparts, and liver progenitor cells (LPCs).4

LPCs are bipotential stem cells present in low numbers in several different niches in the healthy liver.5 When hepatocytes are overwhelmed by the severity of an injury, these LPCs expand and replenish the liver with new hepatocytes.6 The isolation and characterization of LPCs has been the subject of many studies aiming at a better understanding of LPC biology to apply in the generation of high-quality hepatocytes for toxicology studies, improving liver regeneration, or even to serve as a source for direct transplantation after liver injury.

Techniques for isolation of LPCs are often a combination of gradient centrifugation methods with cell separation techniques like fluorescence-activated cell sorting (FACS) or magnetic activated cell separation using antibodies against various stem cell markers.7-10 Functions attributed to stem cells have also been used for their enrichment, e.g., telomerase activity,11 selective attachment on matrices,12 resistance to hypoxia,13 or the side population technique, which is based on efflux of Hoechst-33342 through the ATP-binding cassette (ABC) pumps.14 Recently, aldehyde dehydrogenase (ALDH) activity has received considerable attention as a valuable functional strategy for isolating normal and cancerous stem cells from several tumor types.15, 16

Our objective was to develop an alternative strategy for isolating LPCs from healthy noninjured mouse livers based on high ALDH activity. The nonparenchymal (NP) ALDH+ cells are nonproliferative, range in size from 7 to 8 μm in diameter, and have a high nucleus-to-cytoplasm ratio. These ALDH+ cells express common LPC markers such as EpCAM, CD133, CK19, and Sox9 and can differentiate into functional hepatocyte-like cells in vitro. In vivo, ALDH1A1-positive cells can be identified in canals of Hering and their vicinity, whereas ALDH1A1 expression increases on liver injury in these cells.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

A detailed methods section is provided in the Supporting Materials.

Cell Isolations.

The tissue dissociation protocol used to isolate cells from normal liver was designed based on a protocol for the isolation of nonparenchymal fraction (NPF) as described earlier using an in situ pronase/collagenase perfusion protocol.17 Erythrocytes were lysed using ammonium chloride (red blood cell lysis buffer; Miltenyi Biotec). Cell viability was determined by trypan blue exclusion (>95%).

ALDH Activity and Sorting of ALDH-Positive and -Negative Cells.

The Aldefluor kit was used to isolate a population with high ALDH versus low ALDH enzymatic activity (hereafter referred to as ALDH+ and ALDH−) according to the manufacturer's instructions (StemCell Technologies). Dissociated single cells were suspended in Aldefluor assay buffer containing the ALDH substrate (Bodipy-aminoacetaldehyde [BAAA] 1 μmol/L per 3 × 106 cells) and incubated for 40 minutes at 37°C (Supporting Fig. 1). As negative control, to confirm the specificity of the Aldefluor assay, an aliquot of cells was treated with the ALDH inhibitor diethylaminobenzaldehyde (DEAB). Importantly, thereafter and for all subsequent procedures, samples were constantly maintained at 4°C to prevent efflux. The sorting gate of the ALDH+ cells was established using DEAB-treated cells as a reference. Cell sorting was conducted using a FACS-Aria (BD Biosciences, San Jose, CA), after excluding residual erythrocytes, debris, and dead cells by forward scatter, side scatter, and propidium iodide gating, respectively. Aldefluor fluorescence was excited at 488 nm, and fluorescence emission was detected using a standard fluorescein-isothiocyanate 530/30-nm bandpass filter.

Differentiation of ALDH+ Liver Cells.

Freshly sorted ALDH+ mouse cells were seeded on collagen I–coated dishes (BD Biosciences, 354236) and cultured with Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, 100 U/mL streptomycin, and 5 mM L-glutamine (all Biowhittaker), until they reached 75%-80% confluence. The ALDH+ population reached this confluency after 10-12 days. At that time, medium changes with supplements were carried out every 2 days as indicated in Fig. 4. The cultures were maintained at 37°C in a humidified incubator in a mixture of 95% air and 5% CO2. For the differentiation of human ALDH+ cells, we used the materials and methods described by Wang et al.18

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Isolation of ALDH+ Liver Cell Population.

ALDH activity has been used to isolate stem/progenitor cells from a plethora of normal16 and cancerous tissues.15 Stem cells seem to have a higher ALDH activity than cells in surrounding tissue. This activity can be detected using an artificial fluorescent substrate whose cleavage product can be used to separate cells with distinct activities using flow cytometry (Supporting Fig. 1A). Hepatocytes are the principal detoxifying cells in the liver, using a large arsenal of enzymes, like ALDHs, to cleave toxic products. To be able to use ALDH activity to isolate LPCs, we eliminated hepatocytes by using a standard two-step in vivo liver perfusion protocol commonly used to isolate hepatic stellate cells (HSCs).17 In addition to collagenase, this method uses pronase, an enzyme known to harm hepatocytes (PMP70 positive) and two short 50g centrifugation steps to eliminate most of the parenchymal cells (Fig. 1A,B). Subsequently, we removed any residual hepatocytes during the ALDH activity sorting procedure by excluding high side scatter and forward scatter cells. Before performing the ALDH activity assay, red blood cells were lysed, and a lineage-depletion MACS step was carried out after the cell sorting to eliminate any residual CD5-, CD11b-, CD19-, Ter119-, CD45R-, and Ly-6G/C-positive cells (±0.2%).

thumbnail image

Figure 1. Enrichment of LPCs using ALDH activity. (A) Schematic representation of the procedure. (B) By light microscopy and immunofluorescence for PMP70 we observe a hepatocyte contamination of ±5% in NPF, which is eliminated by the FACS gating. (C) Gating of Aldefluor-positive cells based on DEAB-sensitive ALDH enzymatic activity. (D) Microscopic examination of sorted ALDH+ cells; a relatively uniform population of small round ALDH+ cells (green) is seen, whereas ALDH cells are heterogeneous in size and are negative for ALDH activity. (E-G) Immediately after sorting, ALDH1A1 expression on RNA (E) and protein (F and G) levels in the two sorted-ALDH populations was compared with the NPF. (H) One day after sorting and plating, ALDH cultures are a heterogeneous composition of HSCs (1), hepatocytes (2), LSECs (3), (myo)fibroblasts (4), and Kupffer cells. After two washes, only HSCs and fibroblasts are present. The ALDH+ cultures consist of only two populations: round small cells and (myo)fibroblasts. (I) During cell culture of the ALDH+ population, ALDH activity in vitro indicates that only the small round cells maintain their ALDH activity.

Download figure to PowerPoint

In addition to hepatocytes, we observed some HSCs with high ALDH activity. During the FACS procedure, we thus eliminated HSCs by their capacity to autofluoresce due to their high content of retinol in lipid droplets19 (Supporting Fig. 2A). Using such a procedure, we obtained 2.24% ± 0.50% ALDH+ cells from healthy BALB/c livers (n = 78; Supporting Fig. 1C). ALDH+ was clearly confirmed by a shift in fluorescence of this population when the ALDH inhibitor DEAB was used (Fig. 1C). ALDH positivity was also illustrated by demonstration of the expression of ALDH1A1 enzyme in cells seeded immediately after sorting (Fig. 1D-G). The ALDH population contained nonparenchymal (NP) cells (liver sinusoidal endothelial cells [LSECs], HSCs, and Kupffer cells) and some hepatocytes, which were removed after a brief wash, leaving only HSCs and fibroblast-like cells after 1 day in culture. In contrast, at day 1 (after washing) the ALDH+ population remained phase-bright and seemed to consist of two types of cells, small round cells and fibroblast-like cells. During culture, only the small round cells kept their ALDH activity and grew on top of the fibroblast-like cells that had lost their ALDH activity earlier on in culture (Fig. 1I and Supporting Fig. 3).

The Sorted Nonparenchymal ALDH+ Population Specifically Displays Liver Progenitor Cell Markers.

The ALDH+ population was analyzed immediately after sorting by immunohistochemistry of cytospins, thereby avoiding cell culture-induced artifacts. The “surface marker footprints” of the two freshly isolated populations is displayed in Fig. 2 (for antibodies, see Supporting Table 1). These show that the two populations are clearly distinct, as demonstrated by the very low percentage of cells that expressed endothelial (CD31, CD146), Kupffer (CD11b, F4/80), LSEC (MRC1, CD32b), and HSC markers (Bodipy, GFAP) and the complete absence of hepatocyte markers (PMP70, Connexin-32, CYP1A1) in the ALDH+ population (Fig. 2A).

thumbnail image

Figure 2. Characterization of the NP ALDH+ population. (A) Freshly sorted ALDH+ and ALDH populations were stained and quantified for markers of hepatic cell populations, with CD31 for endothelial cells; CD146, CD32b, and MRC1 for LSECs; CD11b and F4/80 for Kupffer cells; and Bodipy and GFAP for HSCs; PMP70, Connexin32, and CYP1A1 for hepatocytes. (B) A panel of representative staining for markers of LPCs. (C) Antibodies frequently used in LPC isolations. (D) Markers often used to demonstrate the bipotential capacity of LPCs. (E) Markers already reported to be present on LPCs. See also Supporting Table 3. MRC1, mannose receptor C type 1; MRP2, multidrug-resistant–associated protein-2; ABCG2, ATP-binding cassette protein G2; ICAM, intercellular adhesion molecule-1; GFAP, glial fibrillary acidic protein.

Download figure to PowerPoint

In contrast, the ALDH+ population was enriched in cells expressing markers that have been reported to be suitable for the isolation and/or identification of LPCs like CK19, EpCAM, CD133, SOX9, ABCG2, MRP2, CD117, and CD49f (Fig. 2B-D). The complete absence of Sca-1-, CD34-, and CD45-positive cells in the ALDH+ population indicated the absence of any contaminating hematopoietic (stem) cells (data not shown). The ALDH+ population was also enriched in cells expressing ALB, AFP, and CK-7, -8, -14, -18, and -19, suggesting their bipotentiality (Fig. 2D). In addition, markers found to be expressed by a variety of LPCs isolated from different liver injury mouse models were also readily detected in the ALDH+ population (Connexin43, CD24, CD26, CD29, Claudin-3, and Integrin β4) (Fig. 2E).

The Yield of Isolated LPCs Through ALDH Activity Is Superior to the One Obtained by Percoll Gradient Centrifugation.

To test the overall efficiency of isolating LPCs using the ALDH strategy, we investigated whether this method could be improved by combining it with an established method like Percoll gradient enrichment of LPCs.20 Cells were isolated either by their ALDH activity, through a Percoll gradient, or by both methods combined (Fig. 3A,B). ALDH activity sorting resulted in a higher enrichment of CK19+, EpCAM+, CD133+, and E-cadherin+ cells compared with the Percoll gradient method. Moreover, ALDH activity could enrich for this population when applied subsequently to the Percoll gradient but did not result in a higher proportion of desired cells compared with the ALDH activity sorting method alone (Fig. 3C).

thumbnail image

Figure 3. Comparison of ALDH activity assay with Percoll gradient enrichment of LPCs. (A) Schematic representation of the three different approaches used: (1) ALDH activity alone; (2) Percoll gradient followed by ALDH activity; and (3) Percoll gradient alone. (B) Representative pictures of collected cells. (C) After the collection of the desired fraction, the cells were used for immunocytochemistry for the LPC markers CK19, EpCAM, CD133, and E-cadherin. Percentage of positive cells for each marker in the different approaches has been quantified (ns, not significant; *, P < 0.05; ***, P < 0.001).

Download figure to PowerPoint

ALDH+ Population Can Be Readily Differentiated Into Hepatocyte-Like Cells.

To demonstrate that our strategy selectively enriches for LPCs, we applied a differentiation protocol that follows the late stages of liver development and includes addition of specific cytokines/growth factors at defined time points during culture21 (Fig. 4A). Brightfield microscopy (Fig. 4B) revealed that at early stages the cells had round/ovoid nuclei and high nuclear/cytoplasm ratios. During the maturation and differentiation steps the ALDH+ cells were successively organized in cord-like structures (starting at stage I), proliferated, with a cobblestone appearance, and finally acquired morphological features similar to those of primary hepatocytes, i.e., binucleated and polygonal-shaped cells (Fig. 4B). When maintained in 10% FBS without additives, the cells neither acquired the above-mentioned morphological features nor showed any functional hepatocyte activity such as ALB secretion (Supporting Fig. 4). A clear down-regulation of CK19 and EpCAM at the RNA level indicated the loss of progenitor cells during in vitro differentiation (Fig. 4C).

thumbnail image

Figure 4. ALDH+ populations give rise to hepatocyte-like cells under defined culture conditions. (A) Schematic representation of the protocol used to induce hepatocyte differentiation of the ALDH+ cells. Black arrows indicate the isolation of RNA, and gray arrows are a sign of conditioned media collection. (B) Representative brightfield images of the ALDH+ cells during the differentiation procedure. (C) Relative mRNA levels of CK19 and EpCAM were quantified by qPCR at each stage of the differentiation protocol.

Download figure to PowerPoint

ALB secretion, urea synthesis, and CYP1A2 activity, all markers/indicators of hepatocyte function, were tested during the differentiation steps of the ALDH+ cells (Fig. 5). Although barely present during the maturation stages I and II, all activities were induced during the differentiation stage (stage III). Furthermore, periodic acid-Schiff and Bodipy staining demonstrated the capacity of the hepatocyte-like cells to accumulate glycogen and lipids, respectively (Fig. 5D). These data clearly demonstrate that the ALDH+ cell population is able to give rise to functional hepatocyte-like cells in vitro using a defined differentiation protocol, suggesting that this population comprises LPC capacities.

thumbnail image

Figure 5. The ALDH+ population can give rise to functional hepatocyte-like cells. At the end of the differentiation protocol with William's E medium, the conditioned media of ALDH+ cells were collected to quantify several hepatic function parameters: (A) albumin secretion, (B) CYP1A2 activity, (C) urea synthesis, and (D) glucose storage (indicated by periodic acid Schiff staining) and lipid metabolism (using Bodipy493/503). Freshly isolated hepatocytes, cultured for 1 day, were used as control.

Download figure to PowerPoint

ALDH1A1 Expression in LPC Niches.

In adult healthy mice livers, ALDH1A1 is predominantly expressed by hepatocytes in the centrilobular region (Supporting Fig. 5). Analysis of bile ducts and canals of Hering (by CK19 staining) also confirmed ALDH1A1 positivity in two well-known niches of LPCs22 (Fig. 6A-D; for confocal images, see Supporting Fig. 6). We hypothesized that, if high ALDH activity is associated with LPC activation, the expression of ALDH1A1 should increase in different liver injury models, known to activate the LPC niche. ALDH1A1 expression was rapidly induced in bile ducts, i.e. after 3 days in CDE (choline deficient-ethionine supplemented) and DDC (3,5-diethoxycarbonyl-1,4 dihydrocollidine) treated mice, after 12h and 24 hours in respectively APAP (N-acetyl-paraaminophen) and CBDL (Common Bile Duct Ligation) mice and after 2 weeks in AAF/PH (2-acetylaminofluorene/ partial hepatectomy) treated rats (Fig. 7 and Supporting Figs. 7 and 8). However, ALDH1A1 expression rapidly returned to control levels after the initial increase of expression, suggesting that ALDH1A1 up-regulation in these cells is an early response to injury.

thumbnail image

Figure 6. In noninjured liver ALDH1A1 is found in LPC niches: the canals of Hering and their vicinity. Localization of bile ducts (A) and canals of Hering (C) in normal liver evidenced by CK19 immunocytochemistry. CK19+ and ALDH1A1+ cells are present in the bile ducts (B4) and canals of Hering (D4). DAPI (B3,D3) is used to counterstain the nuclei and the colocalization of ALDH1A1 (B1,D1) and CK19 (B2,D2) observed in merged pictures (B4,D4). (E-H) Autofluorescence and negative controls (no primary antibody) were used to validate the staining. BD, bile duct; DAPI, 4′,6-diamidino-2-phenylindole.

Download figure to PowerPoint

thumbnail image

Figure 7. ALDH1A1 expression is induced in different liver injury models commonly used for oval cell activation. Mice and rats fed control diet were used to determine basal ALDH1A1 expression. Either for control animals or those treated or submitted to hepatic injury, a large magnification of the portal tract is shown to provide a better view of ALDH1A1 expression, alone or merged with DAPI staining. The exposure time for the green fluorescence has been calibrated on the injured livers not on the normal liver to avoid overexposition.

Download figure to PowerPoint

Isolation of ALDH-Positive Cells From Human Liver Tissue.

To investigate whether the ALDH strategy is applicable to human liver tissue, we sorted two different human NP samples: a cell fraction obtained after centrifugation of collagenase-digested human liver tissue for hepatocyte transplantation purposes and an in situ digested liver lobe by a pronase/collagenase/Dnase1 digestion (Supporting Fig. 9A). Approximately 4% and 6% of the respective human NPFs had high ALDH activity (Supporting Fig. 9B). As in mouse, bile ducts and canals of Hering were found to be positive for ALDH1A1 in healthy human tissue (Supporting Fig. 9C). The ALDH+ fraction was enriched for Krt7+, Krt19+, EpCAM+, and CD133+ cells (Fig. 8A), albeit less than in healthy mouse livers (Fig. 2). Together, these data indicate that the ALDH activity assay can also be used to enrich for human LPCs.

thumbnail image

Figure 8. Differentiation of human ALDH+ cells to hepatocyte-like cells. (A) Freshly sorted cell populations were spun down, fixed, and stained for human stem cell markers Krt7, Krt19, EpCAM, and CD133. The table represents the quantification of each marker. (B) Schematic representation of the differentiation protocol of human ALDH+ cells using Biomatrix. (C,D) Phase-contrast images of human ALDH+ cells in culture. (E-H) Differentiated cultures display a hepatocyte-like phenotype; strong albumin and Krt18 expression, high glycogen storage, and albumin and urea secretion. Freshly isolated hepatocytes, cultured for 1 day, were used as control.

Download figure to PowerPoint

To illustrate the hepatic differentiation capacities of the sorted human ALDH+ cells, we had to use another in vitro differentiation protocol based on the use of Biomatrix scaffolds and hormonally defined medium (Fig. 8B,C).18 In stage I, the ALDH+ cells transitioned to cells with an increased cytoplasmic/nuclear ratio and formed colonies with marked ALB and Krt18 expression and glycogen storage. These colonies also displayed channels reminiscent of bile canaliculi (Fig. 8D-F). Although the colonies were relatively homogeneous and densely packed, we noticed the presence of two different phenotypes: hepatoblast-like and hepatocyte-like cells (Fig. 8D,F). Finally, the differentiated human ALDH+ cultures exhibited ALB and urea secretion, illustrating a successful differentiation of human ALDH+ cells into hepatocyte-like cells (Fig. 8G,H).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

LPCs exist only in low numbers in healthy livers, which forces most scientists to use “two-hit liver injury models” in order to increase their overall yield (see references in Dollé et al.20 and Gaudio et al.23). The use of different liver injury models, however, has not facilitated the comparison of LPCs isolated from these models using diverse antibodies directed against LPC markers such as EpCAM, CD133, and Sca-1.9, 10, 24

Here we report, for the first time, the use of high ALDH activity for the enrichment of functional LPCs. This assay is highly reproducible, nontoxic, and easy to use, does not involve antibody recognition or the use of DNA intercalating dyes, and is also applicable to human material. ALDH1A1-positive cells can be located in the well-described LPC niches,22 the canal of Hering and its vicinity. Additionally, the expression level of ALDH1A1 is induced during various forms of liver injury. Although in healthy mice the ALDH+ population only accounts for ±2.24% of the NPF, the use of healthy livers ensures the isolation of resident LPCs and not additional progeny of LPCs induced by the injury. This is illustrated by the lack of expression of Sca-1, CD34, Dlk-1, Foxl1, and Trop2, all markers expressed in progenitor cells on liver injury.7, 9, 25, 26

When we compared the surface marker expression of the NP ALDH+ cells with other LPC populations described in the literature (Supporting Table 3), we found three populations that were closely related to ALDH+ cells because of their positivity for EpCAM, CD13/CD49f/CD133, and CD24 on freshly isolated material.26-28 ALDH+ cells show some similarities to oval cells isolated from liver injury models, e.g., the expression of c-Kit and CD90, although they are negative for CD34, CD45, Dlk-1, and Sca-1. Like oval cells, ALDH+ cells express bipotential markers like CK18, CK19, ALB, and AFP and differentiate to hepatocyte-like cells in vitro, suggesting that ALDH+ cells might also be bipotential progenitors.6

To further illustrate the identity of NP ALDH+ cells, we demonstrate that these ALDH+ cells are very small (8 μm) and have a scant, lightly basophilic cytoplasm (large nuclear-cytoplasmic ratio), as well as oval-shaped pale blue–staining nuclei, all features attributed to LPCs/oval cells (Supporting Fig. 10). In addition, ALDH+ cells are in a nonproliferative state (Ki-67) at the time of isolation and are located in canals of Hering and their vicinity in normal liver. Furthermore, the NP ALDH+ cells express genes attributed to a liver stem cell phenotype, i.e., Sox9, EpCAM, CD133, and genes identifying three important cell signaling axes involved in the activation of oval cells, i.e., SCF/c-Kit,29 SDF1/CXCR4,30 and TWEAK/Fn1431 (Supporting Table 4).

Do LPCs require high ALDH activity to fulfill their role as (liver) progenitor cells or is this activity only a convenient way to isolate a population that includes cells with stem cell capacities? ALDHs have important functions in the development of epithelial homeostasis, and, as a result, deregulation of this class of enzymes has been implicated in multiple cancers.32 Aldehydes are organic compounds that are widespread in nature and arise endogenously during the metabolism of alcohols, amino acids, vitamins, retinoids, steroids, and lipid peroxidation, or are exogenously generated from the metabolism of drugs (e.g., acetaminophen, cyclophosphamide) and environmental agents (e.g., cigarette smoke, motor vehicle exhaust). Aldehydes are strong electrophilic compounds with terminal carbonyl groups that can form adducts with cellular targets (proteins and nucleic acids) thereby initiating adverse biological effects, i.e., loss of protein activities and mutation of nucleic acids, making their removal a priority. Cells deploy strategies to eliminate these toxic molecules by use of ALDHs yielding less toxic metabolites. In addition to self-renewal, multipotency, and proliferative capacities one can imagine that resistance to these aldehyde metabolites is also a requirement for a progenitor cell to prevail during harsh conditions,32, 33 a scenario recently played out for ABCG2.34

In addition to their role in cell defense, ALDHs metabolize retinaldehyde to retinoic acid, which is a strong morphogen initiating the programs of cellular differentiation and proliferation that are important during development. One can imagine that some of these functions should be maintained throughout the life of an organism to regulate cell fates and/or differentiation of stem cell populations. Recently, Yanes and colleagues35 have shown that the activation of oxidation is a metabolic signature of stem cells, which might include induction of ALDH activity.

The differentiation protocol we employed for mouse ALDH+ cells is similar to those used to generate hepatocyte-like cells out of adult LPCs (or iPS/ESCs).21 Although the final cultures exhibited several hepatic functions, including ALB secretion, urea synthesis, and CYP activity, the cultures still retain some immature characteristics, such as expression of AFP (60%) and only low CK18 (12%) and ALB (10%) positivity (data not shown), which could explain the relatively low performance of these cultures in ALB, urea, and CYP assays. Using the Biomatrix culture conditions,18 we did manage to have high percentages of ALB+/AFP colonies derived from the human ALDH+, cells indicative of a more advanced differentiation stage.

We describe, for the first time, the enrichment of LPCs based on high ALDH activity, offering a proof of principal for the existence of ALDH+ liver cells. Although this is a straightforward strategy, there is still room for improvement, e.g., avoiding pronase-I digestion and red blood lysis buffer, procedures that could damage LPCs. This strategy yielded a high proportion of cells with hepatic stem cell properties that were successfully differentiated into functional hepatocyte-like cells in vitro. This was possible without the need for prior manipulations to the donor organism, making the technique broadly applicable for the prospective isolation of therapeutically useful cell populations. Although studies in hematopoietic stem cells attribute a stem cell function to ALDHs,36 we do not yet know whether ALDH activity is important for LPC maintenance or function. Finally, our data suggest that ALDH activity can be added to the list of acknowledged LPC markers such as CK19, CD133, EpCAM, and Sox9.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We thank Danielle Blyweert, Nathalie Eysackers, and Tom Schouteet for excellent technical assistance and we express our warmest thanks to Jean-Marc Lazou and Kris Derom for their administrative assistance. We are grateful to Marsha Roach (GigaCyte) and Prof. Lola Reid for generously providing us with Giga-Matrix Liver Biomatrix and Giga-ESP media and helpful discussions. We thank Noémi Van Hul, Ange-Clarisse Dusabineza, and Kunal Chaudhary for providing tissue samples. The TROMA-III antibody developed by Rolf Kemler was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of NICHD and maintained by the Department of Biological Sciences, University of Iowa (Iowa City, IA).

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_24693_sm_SuppFigsandTabs.pdf8175KSupporting Information Figures and Tables

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.