Characterization and Differentiation of Sorted Human Fetal Pancreatic ALDH hi and ALDH hi /CD133 + Cells Toward Insulin-Expressing Cells

Abstract

The enzyme aldehyde dehydrogenase (ALDH) is found in developing and multipotent cell populations, and is important for the production and regulation of retinoic acid, which controls β-cell differentiation in the pancreas. The role of ALDH-expressing cells in the formation of endocrine-like cells and co-localization with the putative stem cell marker CD133 has not been examined during human pancreatic development. This study focuses on the co-expression of CD133 on ALDH⁺ cells from the human fetal pancreas (18–22 weeks of fetal age) with transcription factors (TFs) central to endocrine cell development. Fluorescence-activated cell sorting demonstrated that cells with high ALDH activity (ALDH^hi) had increased co-expression of CD133 and endocrine-lineage TFs when compared with cells with low ALDH (ALDH^lo) expression. Hormone-expressing (insulin, somatostatin) and ductal cells (CK19) were noted in the ALDH^hi population, while mesenchymal (vimentin) and endothelial (CD31) markers were predominantly found in ALDH^lo cells. Culture of sorted ALDH^hi or ALDH^hi/CD133⁺ cells resulted in loss of endocrine TF, insulin, and CK19 expression. The formation of cell clusters from cultured ALDH^hi or ALDH^hi/CD133⁺ cells led to restored CK19 expression and showed endocrine TFs and insulin expression. In summary, pancreatic ALDH^hi cells contain a heterogeneous CD133-enriched population with a subset of β-cell associated markers in the developing human pancreas.

Introduction

The decline in pancreatic β-cell mass and function is a critical factor that contributes to the progression of diabetes. Although islet cell transplantation is one promising treatment, long-term limitations in the supply of islets and the gradual loss of exogenous insulin production from donor islets [1] prevent widespread application of this approach. As a result, examining undifferentiated cell sources for their potential to generate functioning and insulin-responsive β-cells suitable for transplantation is needed. The enzyme family aldehyde dehydrogenase (ALDH) mediates NAD⁺-dependent aldehyde conversion and functions as a protective regulator against intracellular oxidative stress and cytotoxic treatments [2 –4]. Cells with increased ALDH activity share similarities with progenitor-like cells, including enhanced rates of proliferation and the capacity to form renewing spheroids [5 –8].

ALDH activity stimulates cellular differentiation by facilitating the conversion of retinaldehydes to retinoic acid (RA), a critical regulator of organogenesis during embryonic development [9 –11]. RA treatment also reestablishes differentiation in precursor populations after reversible ALDH inhibition through diethylaminobenzaldehyde (DEAB) [9,10,12]. During human pancreatic development, ALDH1A1⁺ cells correlate with neurogenin 3 (NGN3)- and insulin-expressing cells [10,13]. Thus, ALDH is a potential marker for examining and isolating cells that contribute to pancreatic endocrine cell development.

The presence of other reported adult stem cell markers have been identified within the developing and mature pancreas, and have been isolated using cell surface expression of CD34, CD117, and CD133 [14 –17]. Co-localization of these surface markers with ALDH activity has been used to analyze pancreatic cells with multilineage potential, as seen in ALDH⁺ adult murine centroacinar cells co-expressing CD133 [18]. Cells isolated from human bone marrow and umbilical cord blood demonstrated high ALDH activity and co-localization with other putative stem cell markers, and these cell populations promote effective regenerative properties in STZ-damaged islets [19]. Therefore, examining ALDH activity in conjunction with other stem cell markers in the developing human pancreas may lead to the identification of a population with the capacity to produce islet-associated cell types. We have previously characterized ALDH1 activity during different stages of human pancreatic development [10], yet no study to date has examined the co-expression of sorted ALDH⁺ cells with putative stem cell markers from the dissociated human fetal pancreas and their contribution to pancreatic endocrine cell differentiation. In this study, we hypothesized that cells with high ALDH activity (ALDH^hi) and CD133 co-expression in the developing human pancreas represent a pancreatic progenitor cell population with the potential to express transcription factors (TFs) and hormones present in pancreatic endocrine cells.

Materials and Methods

Preparation and dissociation of human fetal pancreata

Human fetal pancreata between 18 and 22 weeks of fetal age were obtained and approved by the Health Human Sciences Research Ethics Board at Western University, in accordance with the Canadian Council on Health Sciences Research Involving Human Subjects guidelines [10]. Pancreata were immediately digested with dissociation buffer containing DNase I (Roche, Mississauga, ON, Canada) and collagenase V (1 mg/mL; Sigma, Oakville, ON, Canada) [10], and further treated with trypsin-ethylenediaminetetraacetic acid (Invitrogen, Burlington, ON, Canada) to prepare single cell suspension. Cell suspensions recovered in CMRL-1066 media supplemented with 10% fetal bovine serum (FBS; Invitrogen) until the time of cell sorting at the London Regional Flow Cytometry Facility, Robarts Research Institute (London, ON, Canada) [10].

Fluorescence-activated cell sorting

Cell preparations were suspended in Aldefluor buffer (StemCell Technologies, Inc., Vancouver, BC, Canada) at a density between 6 × 10⁶ and 1 × 10⁷ viable cells. Aldefluor reagent (StemCell Technologies, Inc.) and anti-human fluorophore-conjugated CD34, CD117, and CD133 antibodies (Supplementary Table S1; Supplementary Data are available online at www.liebertpub.com/scd) were applied to detect the presence of ALDH and surface stem cell markers, respectively [10]. Fluorophore Minus One controls were prepared using a minimum of 1 × 10⁵ live cells, and DEAB (StemCell Technologies, Inc.) was used to inhibit ALDH-mediated conversion to the negatively charged anion BODIPY-aminoacetate.

Anti-human fluorophore-conjugated CD235a antibody (Supplementary Table S1) and viability marker 7-actinomycin (7-AAD; BD Biosciences, Mississauga, ON, Canada) were used to remove erythrocytes and nonviable cells from the final sorted populations, respectively. Cells were separated using multiparametric cell sorting on the FACSAria III Cell Sorter (BD Biosciences, San Jose, CA), and paired into two sorting groups for our primary focus: (1) ALDH^hi and ALDH^lo populations; and (2) ALDH^hi/CD133^hi and ALDH^lo/CD133^lo populations. All analyses were conducted using FACSDiva software (BD Biosciences).

Expansion and differentiation culture of sorted cells

To determine the extracellular matrix (ECM) composition that best promoted the viability of sorted ALDH^hi cells, matrix proteins of fibronectin, laminin, and collagen I and IV (BD Biosciences) were used as coating in 96-well plates at 5 μg/mL, with control wells coated using 1% BSA solution. ALDH^hi cells were plated at a density of 1 × 10⁴ cells per well in Dulbecco's modified Eagle's medium (DMEM)/F12 media supplemented with 10% FBS (Invitrogen). Cellular viability and metabolic activity was measured 24 h and 5 days after plating using a MTT assay (3-[4,5-dimethyl-2-thiazolyl]-2,5-diphenyl-2H-tetrazolium bromide; Sigma). MTT⁺ cells were imaged by selecting five consistent fields in each well and quantified. Data were normalized to the total number of adhered cells per experimental group (three replications per experimental group).

Based on our initial tests and our previously published reports [20,21], collagen I coating (5 μg/mL) was chosen for expansion of the sorted ALDH^hi, ALDH^lo, ALDH^hi/CD133^hi, and ALDH^lo/CD133^lo cell groups in culture. Cells were cultured with DMEM/F12 + 10% FBS media. Once wells were 90% confluent, expanded cells were passaged and harvested for analysis. For differentiation cultures, expanded cells were seeded in 12-well plates and cultured with serum-free differentiation media [22 –24]. After 7 days in differentiation media, clusters were collected for RNA extraction and histological analysis.

Immunofluorescence analyses of cell populations

Sorted, cultured, and cluster-forming cell pellets were fixed in 4% paraformaldehyde and suspended in 2% agarose gel before embedding in paraffin [10]. Sections were cut at 4 μm and stained with the appropriate dilutions of primary and secondary antibodies (Supplementary Table S2). Analysis and quantification was conducted using the Image ProPlus software (Media Cybernetics, Inc., Rockville, MD) by counting positive-staining cells over total 4′,6-diamidino-2-phenylindole (DAPI)-labeled cells.

Quantitative reverse transcription polymerase chain reaction (qRT-PCR) analyses of cell populations

RNA from sorted, cultured, and cluster-forming cells was extracted using the RNAqueous-4PCR kit (Invitrogen) [10]. SuperScript II Reverse Transcriptase (Invitrogen) was used alongside isolated RNA for the generation of complementary DNA (cDNA). Real-time polymerase chain reactions (PCRs) were performed using the iQ SYBR Green Supermix kit (Bio-Rad Laboratories Canada Ltd., Mississauga, ON, Canada) and designed primer sequences (Supplementary Table S3). Data were analyzed using the Bio-Rad CFX Connect real-time PCR detection system (Bio-Rad Laboratories Canada Ltd.). Relative gene expression was quantified using the comparative CT method (2^−ΔΔCT) and normalized to the internal control 18S rRNA [10,25].

Statistical analyses

Data are expressed as mean ± standard error of the mean, and statistical analyses using paired or unpaired Student's t-tests were conducted using GraphPad Prism program (Version 5.0c; GraphPad Software, Inc., La Jolla, CA). Data were considered statistically significant at P < 0.05.

Results

Sorted ALDH^hi cells highly co-expressed CD133 in the developing human pancreas

ALDH^hi cells from dissociated pancreata at 18–22 weeks of fetal age were isolated after gate establishment using a DEAB control (Fig. 1A) and comprised the majority (>60%) of viable cells (7-AAD⁻) retrieved (P < 0.001; Fig. 1B). Examination of stem cell–associated surface markers showed CD133 was prominently co-expressed in the ALDH^hi population (66% ± 5%) compared with the ALDH^lo sorted cells (8% ± 3%) (P < 0.001; Fig. 1C, F). In contrast, the majority of CD34⁺ cells were observed in ALDH^lo sorted cells (42% ± 3%) compared to the ALDH^hi cell population (24% ± 3%; P < 0.001; Fig. 1D, F). A low frequency of CD117⁺ cells was noted in both the ALDH^hi (13% ± 8%) and ALDH^lo (11% ± 4%) sorted groups (Fig. 1E, F).

FIG. 1.

Analysis of ALDH activity, cell viability, and surface stem cell markers in the human fetal pancreas. (A) Separation of sorted pancreatic populations using ALDEFLUOR fluorescence gating to isolate ALDH activity (left), and inhibitor DEAB to distinguish ALDH^hi (red) and ALDH^lo (blue) cells (right). (B) Sort plot (left) and quantification (right) of viability in the final sorted populations. In sort plots, ALDH^hi cells, red; ALDH^lo cells, blue; nonviable and excluded cells, gray population. Data are expressed as mean ± SEM (n = 15 sorted pancreata/group at 18–22 weeks). Sort plots of the co-localized surface stem cell markers CD133 (C), CD34 (D), and CD117 (E), and quantification of each marker (F) in ALDH cell populations of sorted fetal pancreatic cells (n = 4–8 sorted pancreata/group). ***P < 0.001 versus ALDH^lo sorted population. ALDH, aldehyde dehydrogenase; DEAB, diethylaminobenzaldehyde; SEM, standard error of the mean.

Using multiparametric analyses, overlapping co-localization for these surface markers in the ALDH^hi and ALDH^lo sorted cells was further examined. Co-expression of CD133⁺ cells with CD34 (Fig. 2A), CD117 (Fig. 2B), or CD34⁺ cells with CD117 (Fig. 2C) in both ALDH^hi and ALDH^lo sorted groups were observed at low levels (4%–10%). Thus, CD133 expression highly correlated with high ALDH activity, but not with CD34 or CD117 co-expression.

FIG. 2.

Co-localization between cell surface markers CD133, CD34, and CD117 in ALDH^hi and ALDH^lo subpopulations. Sorting gates (left) and quantification (right), in separated ALDH^hi (red) and ALDH^lo (blue) populations, of cells that co-localize markers CD133⁺/CD34⁺ (A), CD133⁺/CD117⁺ (B), and CD117⁺/CD34⁺ (C) (n = 4 sorted pancreata/group). White bar, ALDH^hi group; black bar, ALDH^lo group.

Sorted ALDH^hi cells showed higher expression of endocrine TFs

In addition to high CD133 co-localization, sorted ALDH^hi pancreatic cells also demonstrated higher expression of endocrine cell-associated TFs (Fig. 3). Increased expression of SOX9, PDX-1, NGN3, and NKX6.1 mRNA was observed in ALDH^hi cells, but statistical significance was not reached due to high variability between sample pancreata collected (Fig. 3A). To verify the nuclear localization of these TFs, immunohistochemical analyses were performed on sorted cells (Fig. 3B).

FIG. 3.

Pancreatic endocrine-committing TF expression in ALDH^hi and ALDH^lo sorted populations of the human fetal pancreas. (A) mRNA analysis of SOX9, PDX-1, NGN3, and NKX6.1 expression in sorted ALDH^hi and ALDH^lo cells (n = 4 sorted pancreata/group). (B) Representative immunofluorescence images of nuclei staining (green) for SOX9, PDX-1, NGN3, and NKX6.1. (C) Quantification of the number of nuclei TF-positive cells in sorted ALDH^hi and ALDH^lo cell populations (n = 3–5 sorted pancreata/group). Data are expressed as percent of total number of cells (DAPI labeled nuclei in blue) counted per group (mean ± SEM). Scale bar: 15 μm. White bars, ALDH^hi group; black bars, ALDH^lo group. *P < 0.05, **P < 0.01, ***P < 0.001 versus ALDH^lo sorted population. DAPI, 4′,6-diamidino-2-phenylindole; NGN3, neurogenin 3; NKX6.1, NK6 homeobox 1; PDX-1, pancreatic and duodenal homeobox 1; SOX9, sex-determining region Y-box 9; TF, transcription factor.

Sex determining region Y-box 9 (SOX9) expression was found in the nucleus of approximately half (55%) of sorted ALDH^hi cells, while only 10% of ALDH^lo cells contained nuclear SOX9 staining (P < 0.001 vs. ALDH^lo; Fig. 3C). In addition nuclear pancreatic and duodenal homeobox 1 (PDX-1) and NK6 homeobox 1 (NKX6.1) expression were both significantly higher in the ALDH^hi sorted cells (Fig. 3C). A high percentage of neurogenin 3⁺ (NGN3) cells were also noted in the ALDH^hi group, but statistical significance was not detected (Fig. 3C). Finally, islet 1 (ISL1) and paired box gene 6 (PAX6) expression were also significantly increased in sorted ALDH^hi cells compared with ALDH^lo cells (P < 0.05; Fig. 3C). Similar to NGN3 expression, there was no significant difference in NK2 homeobox 2 (NKX2.2) expression between ALDH^lo and ALDH^hi cells. These data indicate that ALDH^hi pancreata examined immediately after sorting contained a high percentage of cells expressing endocrine TFs necessary for islet cell differentiation.

ALDH^hi cells represent a heterogeneous pancreatic population showing low proliferation

The presence of hormone⁺ cells and markers of nonendocrine pancreatic cell lineages were also examined in the sorted cells using immunofluorescence staining. A significantly higher frequency of insulin⁺ cells was observed in the ALDH^hi population (5% ± 1%), while insulin⁺ cells were nearly undetectable in ALDH^lo cells (0.8% ± 0.2%; P < 0.05 vs. ALDH^hi; Fig. 4A). A high percentage of somatostatin⁺ cells were also present in the ALDH^hi group (P < 0.05 vs. ALDH^lo; Fig. 4A), yet both sorted populations contained a similar percentage of glucagon⁺ cells (Fig. 4A).

FIG. 4.

Characterization of cell phenotypes and cell proliferation in ALDH^hi and ALDH^lo sorted populations. Quantitative analysis for endocrine (A) and exocrine (B) cell markers, phenotypic markers of ductal CK19, epithelial E-cadherin; β-catenin cells, and mesenchymal marker vimentin (C), and proliferation marker Ki67 (D) in ALDH^hi and ALDH^lo sorted cells. White bars, ALDH^hi group; black bars, ALDH^lo group. Data are expressed as mean ± SEM (n = 2–4 sorted pancreata/group). *P < 0.05, **P < 0.01 versus ALDH^lo sorted populations. Representative images for HES1, amylase (Amyl), and Ki67 staining (green) in the sorted cell populations (B, D). Nuclei staining labeled by DAPI (blue). Scale bar: 15 μm.

Exocrine cell markers HES1 and amylase showed a similar distribution within both ALDH^hi and ALDH^lo groups (Fig. 4B). Expression of the pancreatic ductal epithelial cell marker CK19 correlated with the ALDH^hi population, while no correlation was observed with E-cadherin and β-catenin expression (Fig. 4C). Sorted ALDH^lo cells contained a high percentage of cells positive for the mesenchymal cell marker vimentin (P < 0.01 vs. ALDH^hi; Fig. 4C). Furthermore, Ki67 detection in actively proliferating cells was higher in ALDH^lo cells compared with the ALDH^hi population (P < 0.05; Fig. 4D), indicating decreased cell cycle activity in the ALDH^hi population at the time of selection.

Sorted ALDH^hi cells lose expression of endocrine transcript factors during in vitro expansion

Due to the lower proliferative capacity identified in sorted ALDH^hi cells, selected ECM proteins were used to promote ALDH^hi cell viability and growth at 24 h and 5 days in expansion culture. At 24 h, viable adherent ALDH^hi cells were scattered throughout the well in both control and ECM-coated wells (Fig. 5A). However, viable monolayers of expanded cells were significantly higher on laminin and collagen I-coated wells at day 5 (Fig. 5A). Both ALDH^hi and ALDH^lo cells displayed a spindle-shaped morphology during the expansion culture. Gene expression analysis, measured in the expanded ALDH^hi and ALDH^lo cells, showed decreased SOX9, PDX-1, NKX6.1, and INS mRNA in ALDH^hi cells, yet levels of NGN3 appeared to be retained when compared with ALDH^lo cells (Fig. 5B).

FIG. 5.

Analysis of cell phenotypes in ALDH^hi and ALDH^lo cells during expansion culture. (A) Quantification of viable adhered ALDH^hi sorted cells cultured on either control or selected ECM coated conditions for 24 h and 5 days, determined by MTT assay. Black bars, 24 h; gray bars, 5 days. Data are expressed as mean ± SEM (n = 3 sorted ALDH^hi cells/group). *P < 0.05, **P < 0.01 versus viable cells at 24 h. (B) Gene expression analysis of TFs and INS mRNA in ALDH^hi and ALDH^lo cells during expansion culture in vitro. White bars, ALDH^hi group; black bars, ALDH^lo group. (C) Immunofluorescence analysis of cell markers E-cadherin, vimentin, SOX9, and insulin (green) in expanded ALDH^hi and ALDH^lo cells. Nuclei are detected with DAPI (blue). Scale bar: 25 μm. Data are expressed as mean ± SEM (n = 4 cultured passages/group). ECM, extracellular matrix; MTT, 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide.

Immunostaining determined that expanded ALDH^hi and ALDH^lo cells positively labeled for both E-cadherin and vimentin (Fig. 5C), with loss of CK19 staining in both populations (images not shown). Further examination of nuclei staining for endocrine TFs revealed that SOX9 labeling was retained in ALDH^hi and ALDH^lo cells, but PDX-1⁺ and NGN3⁺ nuclear labeling was lost (images not shown) along with insulin expression after culture expansion of both cell population (Fig. 5C).

Cultured ALDH^hi cells restore ductal phenotype during the formation of cell clusters

Treatment of cultured ALDH^hi and ALDH^lo populations with a modified differentiation media led to the formation of cell clusters within 36 h of administration (Fig. 6A). Quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis of TF expression showed equivalent expression of PDX-1, NGN3, and INS mRNA between differentiated ALDH^hi and ALDH^lo clusters (Fig. 6B). Immunofluorescence staining of fixed clusters showed the continued expression of vimentin and restoration of CK19 detection (Fig. 6C). Similar to cells expanded under nondifferentiated conditions, nuclear SOX9 staining was maintained in ALDH^hi clusters. However, clusters derived from ALDH^hi and ALDH^lo cell populations showed weak staining for insulin (Fig. 6C).

FIG. 6.

The presence of pancreatic TFs and hormones in clusters differentiated from expanded sorted populations. (A) Representative phase-contrast images of expanded ALDH^hi and ALDH^lo cells treated with cluster-forming differentiation medium at 0 h, 36 h, and 7 days post-treatment. Scale bar: 200 μm. (B) Gene expression analysis of cluster-forming cells derived from expanded ALDH^hi and ALDH^lo cell populations after 7 days in serum-free differentiation medium for expression of TFs and insulin mRNA. White bars, ALDH^hi group; black bars, ALDH^lo group. Data are expressed as mean ± SEM (n = 3–5 cultured passages/group). (C) Immunofluorescence analysis of clusters for CK19, vimentin, SOX9, and insulin (green). Nuclei are stained with DAPI (blue). Scale bar: 25 μm.

Examination of ALDH^hi/CD133⁺ sorted cells in vitro

Since sorted ALDH^hi cells contained CD133⁺ cells with high SOX9, PDX-1, and NKX6.1 co-expression (Fig. 7A), we focused on expansion of the rare ALDH^hi/CD133⁺ subpopulation to determine whether this cell type could produce an endocrine cell population in vitro. Although qRT-PCR analysis of the gene expression for TFs in expanded ALDH^hi/CD133⁺ and ALDH^lo/CD133⁻ cells showed no significant differences in expression, higher levels of HES1 and AMYLASE (AMLY) genes were detected in the expanded ALDH^hi/CD133⁺ cells (Fig. 7B).

FIG. 7.

Analysis of ALDH^hi/CD133⁺ and ALDH^lo/CD133⁻ expanded populations. (A) Representative immunofluorescence image of CD133 (green) co-localization with SOX9, PDX-1, and NKX6.1 (red) in ALDH^hi sorted cells. Nuclei are detected with DAPI (blue). Scale bar: 25 μm (insert: 10 μm). The mRNA expression analysis for TFs, INS, and AMYL in expanded ALDH^hi/CD133⁺ and ALDH^lo/CD133⁻ cultured populations (n = 4–5 passaged cells/group) (B); and in ALDH^hi/CD133⁺ and ALDH^lo/CD133⁻ clusters after 7 days in differentiation culture (n = 7–9 passaged cells/group) (C). White bars, ALDH^hi/CD133⁺; black bars, ALDH^lo/CD133⁻. Data are expressed as mean ± SEM. ***P < 0.001 versus ALDH^lo/CD133⁻ cells.

After the 7-day differentiation culture of ALDH^hi/CD133⁺ and ALDH^lo/CD133⁻ cells, SOX9, PDX-1, NGN3, and NKX6.1 mRNA expression was increased in clusters derived from ALDH^hi/CD133⁺ cells (Fig. 7C). Elevated INS gene expression was also detected in ALDH^hi/CD133⁺ clusters (P < 0.001; Fig. 7C), but there was no change in AMLY gene expression.

Confirmation of qRT-PCR analyses was conducted using immunofluorescence staining of fixed cell clusters, and increased nuclear labeling for SOX9, PDX-1, and NKX6.1 in the ALDH^hi/CD133⁺ clusters was detected with a few NGN3 nuclei positive cells present (Fig. 8A). Although these TFs were also present in the ALDH^lo/CD133⁻ clusters, staining was localized in their cytoplasm (Fig. 8A). Insulin staining was detected in both groups after treating cells with differentiation media, and quantification of the cell clusters found a higher percentage of insulin⁺ cells within the ALDH^hi/CD133⁺ population (P < 0.01 vs. ALDH^lo/CD133⁻; Fig. 8B). C-peptide staining was comparatively stronger in the ALDH^hi/CD133⁺ clusters, while no difference was observed for glucagon and somatostatin staining (Fig. 8C). Both CK19⁺ and desmin⁺ cells were observed in the ALDH^hi/CD133⁺ and ALDH^lo/CD133⁻ clusters (Fig. 8).

FIG. 8.

Immunofluorescence staining evaluation of ALDH^hi/CD133⁺ islet-like clusters. Immunofluorescence labeling (green) of endocrine TFs and ductal marker CK19 (A). (B) Quantification of insulin⁺ cells in islet-like clusters derived from expanded ALDH^hi/CD133⁺ and ALDH^lo/CD133⁻ cells (n = 4 passaged cells/group). Data are expressed as mean ± SEM. **P < 0.01 versus ALDH^lo/CD133⁻ cells. (C) Immunofluorescence staining of islet endocrine markers c-peptide, glucagon, and somatostatin, and mesenchymal marker desmin in islet-like clusters derived from expanded ALDH^hi/CD133⁺ (top row) and ALDH^lo/CD133⁻ (bottom row) clusters. Nuclei are labeled with DAPI (blue). Scale bar: 25 μm.

Discussion

In this study, we examined sorted ALDH^hi cells from the developing human pancreas at 18–22 weeks of fetal age, found that ALDH^hi cells highly co-expressed CD133 and early endocrine-associated TFs, and insulin and somatostatin. During expansion under nondifferentiative culture conditions, sorted ALDH^hi and ALDH^hi/CD133⁺ cells lost ALDH, endocrine TFs, insulin, and CK19 expression, with an increase in vimentin⁺ cells. In contrast, culture of these cells under differentiation-promoting conditions induced cluster formation and led to increased expression of endocrine TFs and INS mRNA in ALDH^hi and ALDH^hi/CD133⁺ clusters. These data suggested that ALDH^hi and ALDH^hi/CD133⁺ populations represent a heterogeneous pancreatic pool with an increased subset of cells expressing TFs required for the generation of endocrine cells in the developing human pancreas.

Sorted ALDH^hi cells highly co-expressed CD133, which is found on ductal cells that acquire NGN3 expression during human pancreatic development [26]. CD133 also co-localized with ALDH⁺ centroacinar cells in adult murine pancreata that can form insulin-producing cells [18], suggesting that this marker can be used to isolate cells with the potential to form endocrine hormone-producing cells. Sorted ALDH^hi/CD133⁺ cells show increased SOX9, PDX-1, and NKX6.1 expression, which reflects the association of CD133 as a pancreatic ductal cell marker and with the developing endocrine lineage [26 –29].

Sorted cells with high ALDH activity and CD133 expression did not correlate with CD34⁺ cells. CD34⁺ cells, which were found prominently in the ALDH^lo subset, have been shown to co-localize with the vasculature in developing pancreata [30,31] and indicates a potentially separate pancreatic cell fate than ALDH^hi and CD133⁺ cells.

Although both sorted ALDH^hi and ALDH^lo populations were heterogeneous and formed multiple pancreatic cell lineages, endocrine-associated TFs were increased in the ALDH^hi cell subset. The pancreatic progenitor TF SOX9 [32] was highly expressed in ALDH^hi cells and has been previously associated with cells demonstrating high ALDH activity in isolated human fetal islet-epithelial clusters [10]. NGN3⁺ cells were increased in purified ALDH^hi cells, which require early SOX9 expression and eventual downregulation to commit toward an endocrine cell fate [29], but also indicates that the expression of both TFs in the ALDH^hi group represent a pool of cells more closely related to early pancreatic progenitor fate and not of mature endocrine cells. NKX6.1 and ISL1, two TFs that are essential for β-cell maturation and regulation [33 –36], were also expressed in ALDH^hi cells.

PDX-1 is expressed throughout the pancreatic epithelium during early development (8–11 weeks) and co-localize with insulin⁺ and CK19⁺ cells at 19–21 weeks of development, indicating islet neogenesis and subsequent endocrine cell commitment at this stage [33,37]. The presence of increased PDX-1⁺, CK19⁺, insulin⁺, and somatostatin⁺ cells in the sorted ALDH^hi cells indicates a population that were composed predominantly of cells with ductal markers and early endocrine-associated TFs, with a small population of cells that are positive for endocrine hormones. Exocrine-linked markers HES1 and amylase expression were equivalent in both sorted cell groups. Because exogenous RA treatment has been shown to increase [38] and decrease [39] the number of amylase⁺ cells in the developing murine pancreas, enhanced RA production within ALDH^hi cells may also regulate exocrine commitment during the second trimester of pancreatic development.

Under expansion conditions, both vimentin and E-cadherin were expressed in ALDH^hi and ALDH^hi/CD133⁺ populations. Epithelial-to-mesenchymal transition (EMT), previously observed in insulin-expressing cells migrating from human fetal islets in vitro [22], is one hypothesized mechanism for β-cell formation in vivo [40]. Similar to our cultured ALDH^hi and ALDH^hi/CD133⁺ cell subsets, human pancreatic ductal cells and nonendocrine PDX-1⁺ human pancreatic cells demonstrate increased vimentin expression with repeated passages [41,42]. Interestingly, both of these studies achieved endocrine-like fate and upregulation of insulin⁺ cells when exposed to differentiation conditions. However, conflicting results refute EMT from β-cells in murine islets in vitro [43], and it should be taken into consideration that adherent vimentin⁺ cells were initially present in the heterogeneous sorted populations.

Serum-free differentiation media have been used in early β cell differentiation protocols [22 –24,44 –46]. The treatment of expanded mesenchymal-like precursor cells from adult islets with serum-free media induced cluster formation and increased expression of markers associated with endocrine cells [22,45,46]. When mesenchymal-like expanded ALDH^hi and ALDH^hi/CD133⁺ cells were treated with a similar media, both populations demonstrated cell cluster aggregation, increased expression of endocrine TFs, and the presence of insulin⁺ and c-peptide⁺ cells. Although this was also achieved in ALDH^lo and ALDH^lo/CD133⁻ populations, the gene expression of endocrine TFs and INSULIN was lower than in ALDH^hi and ALDH^hi/CD133⁺ clusters.

The findings in this study are supported by similar experiments using NGN3⁺ cells from the adult pancreatic exocrine compartment with high CD133 co-localization, where pancreatosphere formation and upregulation of associated endocrine TFs were seen upon serum-free expansion in vitro [47]. However, it is important to note that increased expression of insulin⁺ cells in the ALDH^hi and ALDH^hi/CD133⁺ clusters may not indicate fully differentiated β cell–like cells. The upregulation of early endocrine TFs were observed in these clusters, and it has been previously established that insulin⁺ cells are detected as early as 9–10 weeks of fetal age in the developing pancreas [37].

Another limitation in this study is that current culture conditions in this experiment could only produce a small percentage of insulin⁺ cells in the ALDH^hi/CD133⁺ group, which limits the potential to test clusters for glucose responsive and insulin secreting β cell–like cells. Both ALDH^lo and ALDH^lo/CD133⁻ clusters were also able to yield hormone-expressing cells with this protocol, and indicates that ALDH activity is not mandatory for the in vitro differentiation of pancreatic precursor cells. Additional factors used previously to stimulate insulin production [48,49] may be necessary to overcome this limitation and bolster glucose responsiveness from formed ALDH^hi and ALDH^hi/CD133⁺ β cell–like clusters.

Adapting multiple step protocols that are able to induce insulin-producing cells from cells with a pancreatic progenitor cell type similar to ALDH^hi and ALDH^hi/CD133⁺ populations can also be ideal for further efficient maturation of the generated clusters. This has recently been shown in progenitor stem cell-derived NGN3⁺/NKX6.1⁺ cells, where the use of a guided differentiation protocol resulted in the production of glucose-sensitive insulin-secreting cells [50].

In conclusion, we have shown that sorted pancreatic ALDH^hi cells that expressed the CD133 surface marker receptor represented a pool of TFs associated with the endocrine precursor population, and demonstrated the capacity to increase endocrine-associated TFs in vitro. Characterizing factors in human pancreata that form islets and drive β cell commitment is necessary for understanding the generation of β cells in vivo, and will aid in furthering therapies that target β cell restoration in diabetic patients. Additional studies are also required to examine optimized culture environments for efficient generation of functional insulin-producing islet cells in vitro.

Footnotes

Acknowledgments

The authors would like to thank Dr. Kristin Chadwick from the London Regional Flow Cytometry Facility for her assistance with cell sorting of pancreatic samples. This work was supported by grants from the Canadian Diabetes Association (CDA, grant no. OG-3-11-3318-RW).

Portions of the data in this article were presented at the 51st European Association for the Study of Diabetes (EASD) Annual Meeting, held September 14–18, 2015.

Author Disclosure Statement

No competing financial interests exist.

References

Ryan

, Paty

, Senior

, Bigam

, Alfadhli

, Kneteman

, Lakey

and Shapiro

. (2005). Five-year follow-up after clinical islet transplantation. Diabetes, 54:2060–2069.

Hammen

, Allali-Hassani

, Hallenga

, Hurley

and Weiner

. (2002). Multiple conformations of NAD and NADH when bound to human cytosolic and mitochondrial aldehyde dehydrogenase. Biochemistry, 41:7156–7168.

Jean

, Laoudj-Chenivesse

, Notarnicola

, Rouger

, Serratrice

, Bonnieu

, Gay

, Bacou

, Duret

and Carnac

. (2011). Aldehyde dehydrogenase activity promotes survival of human muscle precursor cells. J Cell Mol Med, 15:119–133.

, Isaac

, Schott

, Huard

and Weiss

. (2013). Rapamycin inhibits ALDH activity, resistance to oxidative stress, and metastatic potential in murine osteosarcoma cells. Sarcoma, 2013:480713.

Burger

, Gupta

, Xiong

, Ontiveros

, Salm

, Moscatelli

and Wilson

. (2009). High aldehyde dehydrogenase activity: a novel functional marker of murine prostate stem/progenitor cells. Stem Cells, 27:2220–2228.

Ginestier

, Hur

, Charafe-Jauffret

, Monville

, Dutcher

, Brown

, Jacquemier

, Viens

, Kleer

, et al. (2007). ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome. Cell Stem Cell, 1:555–567.

, Su

, Mei

, Leng

, Liu

, Stass

and Jiang

. (2010). ALDH1A1 is a marker for malignant prostate stem cells and predictor of prostate cancer patients' outcome. Lab Invest, 90:234–244.

Lindgren

, Böstrom

, Nilsson

, Hansson

, Sjölund

, Möller

, Jirström

, Nilsson

, Landberg

, Axelson

and Johansson

. (2011). Isolation and characterization of progenitor-like cells from human renal proximal tubules. Am J Pathol, 178:828–837.

Chute

, Muramoto

, Whitesides

, Colvin

, Safi

, Chao

and McDonnell

. (2006). Inhibition of aldehyde dehydrogenase and retinoid signaling induces the expansion of human hematopoietic stem cells. Proc Natl Acad Sci U S A, 103:11707–11712.

10.

, Feng

, Yeung

, Wong

, Oakie

, Fellows

, Goodyer

, Hess

and Wang

. (2014). Aldehyde dehydrogenase 1 activity in the developing human pancreas modulates retinoic acid signalling in mediating islet differentiation and survival. Diabetologia, 57:754–764.

11.

Duester

. (2008). Retinoic acid synthesis and signaling during early organogenesis. Cell, 134:921–931.

12.

Muramoto

, Russell

, Safi

, Salter

, Himburg

, Daher

, Meadows

, Doan

, Storms

, et al. (2010). Inhibition of aldehyde dehydrogenase expands hematopoietic stem cells with radioprotective capacity. Stem Cells, 28:523–534.

13.

Öström

, Loffler

, Edfalk

, Selander

, Dahl

, Ricordi

, Jeon

, Correa-Medina

, Diez

and Edlund

. (2008). Retinoic acid promotes the generation of pancreatic endocrine progenitor cells and their further differentiation into beta-cells. PLoS One, 3:e2841.

14.

Joanette

, Reusens

, Arany

, Thyssen

, Remacle

and Hill

. (2004). Low-protein diet during early life causes a reduction in the frequency of cells immunopositive for nestin and CD34 in both pancreatic ducts and islets in the rat. Endocrinology, 145:3004–3013.

15.

Wang

, Li

and Yashpal

. (2004). Phenotypic analysis of c-kit expression in epithelial monolayers derived from postnatal rat pancreatic islets. J Endocrinol, 182:113–122.

16.

Zhang

W-J

, Xu

S-Q

, Cai

H-Q

, Men

X-L

, Wang

, Lin

, Chen

, Jiang

Y-W

, Liu

H-L

, et al. (2013). Evaluation of islets derived from human fetal pancreatic progenitor cells in diabetes treatment. Stem Cell Res Ther, 4:141.

17.

Ghazalli

, Mahdavi

, Feng

, Jin

, Kozlowski

, Hsu

, Riggs

, Tirrell

and Ku

. (2015). Postnatal pancreas of mice contains tripotent progenitors capable of giving rise to duct, acinar, and endocrine cells in vitro. Stem Cells Dev, 24:1995–2008.

18.

Rovira

, Scott

, Liss

, Jensen

, Thayer

and Leach

. (2010). Isolation and characterization of centroacinar/terminal ductal progenitor cells in adult mouse pancreas. Proc Natl Acad Sci U S A, 107:75–80.

19.

Bell

, Meschino

, Hughes-Large

, Broughton

, Xenocostas

and Hess

. (2012). Combinatorial human progenitor cell transplantation optimizes islet regeneration through secretion of paracrine factors. Stem Cells Dev, 21:1863–1876.

20.

Krishnamurthy

, Li

, Fellows

, Rosenberg

, Goodyer

and Wang

. (2011). Integrin α3, but not β1, regulates islet cell survival and function via PI3K/Akt signaling pathways. Endocrinology, 152:424–435.

21.

Saleem

, Li

, Yee

, Fellows

, Goodyer

and Wang

. (2009). β1 Integrin/FAK/ERK signalling pathway is essential for human fetal islet cell differentiation and survival. J Pathol, 219:182–192.

22.

Gershengorn

, Hardikar

, Wei

, Geras-Raaka

, Marcus-Samuels

and Raaka

. (2004). Epithelial-to-mesenchymal transition generates proliferative human islet precursor cells. Science, 306:2261–2264.

23.

, Li

, Saleem

, Yee

, Hardikar

and Wang

. (2010). c-Kit and stem cell factor regulate PANC-1 cell differentiation into insulin- and glucagon-producing cells. Lab Invest, 90:1373–1384.

24.

Jiang

, Cram

, DeAizpurua

and Harrison

. (1999). Laminin-1 promotes differentiation of fetal mouse pancreatic beta-cells. Diabetes, 48:722–730.

25.

Schmittgen

and Livak

. (2008). Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc, 3:1101–1108.

26.

Sugiyama

, Rodriguez

, McLean

and Kim

. (2007). Conserved markers of fetal pancreatic epithelium permit prospective isolation of islet progenitor cells by FACS. Proc Natl Acad Sci U S A, 104:175–180.

27.

Kania

, Corbeil

, Fuchs

, Tarasov

, Blyszczuk

, Huttner

, Boheler

and Wobus

. (2005). Somatic stem cell marker prominin-1/CD133 is expressed in embryonic stem cell-derived progenitors. Stem Cells, 23:791–804.

28.

Lardon

, Corbeil

, Huttner

, Ling

and Bouwens

. (2008). Stem cell marker prominin-1/AC133 is expressed in duct cells of the adult human pancreas. Pancreas, 36:e1–e6.

29.

Shih

, Kopp

, Sandhu

, Dubois

, Seymour

, Grapin-Botton

and Sander

. (2012). A notch-dependent molecular circuitry initiates pancreatic endocrine and ductal cell differentiation. Development, 139:2488–2499.

30.

Piper

, Brickwood

, Turnpenny

, Cameron

, Ball

, Wilson

and Hanley

. (2004). Beta cell differentiation during early human pancreas development. J Endocrinol, 181:11–23.

31.

Sarkar

, Kobberup

, Wong

, Lopez

, Quayum

, Still

, Kutchma

, Jensen

, Gianani

, et al. (2008). Global gene expression profiling and histochemical analysis of the developing human fetal pancreas. Diabetologia, 51:285–297.

32.

McDonald

, Li

, Krishnamurthy

, Fellows

, Goodyer

and Wang

. (2012). SOX9 regulates endocrine cell differentiation during human fetal pancreas development. Int J Biochem Cell Biol, 44:72–83.

33.

Lyttle

, Li

, Krishnamurthy

, Fellows

, Wheeler

, Goodyer

and Wang

. (2008). Transcription factor expression in the developing human fetal endocrine pancreas. Diabetologia, 51:1169–1180.

34.

Rezania

, Bruin

, Xu

, Narayan

, Fox

, O'Neil

and Kieffer

. (2013). Enrichment of human embryonic stem cell-derived NKX6.1-expressing pancreatic progenitor cells accelerates the maturation of insulin-secreting cells in vivo. Stem Cells, 31:2432–2442.

35.

Taylor

, Benthuysen

and Sander

. (2015). Postnatal beta cell proliferation and mass expansion is dependent on the transcription factor Nkx6.1. Diabetes, 64:897–903.

36.

Ediger

, Du

, Liu

, Hunter

, Walp

, Schug

, Kaestner

, Stein

, Stoffers

and Lee May

. (2014). Islet-1 is essential for pancreatic beta-cell function. Diabetes, 63:4206–4217.

37.

Riedel

, Asadi

, Wang

, Ao

, Warnock

and Kieffer

. (2012). Immunohistochemical characterisation of cells co-producing insulin and glucagon in the developing human pancreas. Diabetologia, 55:372–381.

38.

Kadison

, Kim

, Maldonado

, Crisera

, Prasadan

, Manna

, Preuett

, Hembree

, Langaker

and Gittes

. (2001). Retinoid signaling directs secondary lineage selection in pancreatic organogenesis. J Pediatr Surg, 36:1150–1156.

39.

Kobayashi

, Spilde

, Bhatia

, Buckingham

, Hembree

, Prasadan

, Preuett

, Imamura

and Gittes

. (2002). Retinoid signaling controls mouse pancreatic exocrine lineage selection through epithelial-mesenchymal interactions. Gastroenterology, 123:1331–1340.

40.

Cole

, Anderson

, Antin

and Limesand

. (2009). One process for pancreatic beta-cell coalescence into islets involves an epithelial-mesenchymal transition. J Endocrinol, 203:19–31.

41.

Corritore

, Dugnani

, Pasquale

, Misawa

, Witkowski

, Lei

, Markmann

, Piemonti

, Sokal

, Bonner-Weir

and Lysy

. (2014). β-Cell differentiation of human pancreatic duct–derived cells after in vitro expansion. Cell Reprogram, 16:456–466.

42.

Klein

, Álvarez-Cubela

, Lanzoni

, Vargas

, Prabakar

, Boulina

, Ricordi

, Inverardi

, Pastori

and Domínguez-Bendala

. (2015). BMP-7 induces adult human pancreatic exocrine-to-endocrine conversion. Diabetes, 64:4123–4134.

43.

Atouf

, Park

, Pechhold

, Ta

, Choi

and Lumelsky

. (2007). No evidence for mouse pancreatic beta-cell epithelial-mesenchymal transition in vitro. Diabetes, 56:699–702.

44.

Hardikar

, Marcus-Samuels

, Geras-Raaka

, Raaka

and Gershengorn

. (2003). Human pancreatic precursor cells secrete FGF2 to stimulate clustering into hormone-expressing islet-like cell aggregates. Proc Natl Acad Sci U S A, 100:7117–7122.

45.

Davani

, Ikonomou

, Raaka

, Geras-Raaka

, Morton

, Marcus-Samuels

and Gershengorn

. (2007). Human islet-derived precursor cells are mesenchymal stromal cells that differentiate and mature to hormone-expressing cells in vivo. Stem Cells, 25:3215–3222.

46.

Davani

, Ariely

, Ikonomou

, Oron

and Gershengorn

. (2009). Human islet-derived precursor cells can cycle between epithelial clusters and mesenchymal phenotypes. J Cell Mol Med, 13:2570–2581.

47.

Gomez

, O'Driscoll

, Sheets

, Hruban

, Oberholzer

, McGarrigle

and Shamblott

. (2015). Neurogenin 3 expressing cells in the human exocrine pancreas have the capacity for endocrine cell fate. PLoS One, 10:e0133862.

48.

Gopurappilly

, Bhat

and Bhonde

. (2013). Pancreatic tissue resident mesenchymal stromal cell (MSC)-like cells as a source of in vitro islet neogenesis. J Cell Biochem, 114:2240–2247.

49.

, Tai

, Linning

, Szabo

and Olson

. (2006). MafA expression and insulin promoter activity are induced by nicotinamide and related compounds in INS-1 pancreatic beta-cells. Diabetes, 55:742–750.

50.

Massumi

, Pourasgari

, Nalla

, Batchuluun

, Nagy

, Neely

, Gull

, Nagy

and Wheeler

. (2016). An abbreviated protocol for in vitro generation of functional human embryonic stem cell-derived beta-like cells. PLoS One, 11:e0164457.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

Characterization and Differentiation of Sorted Human Fetal Pancreatic ALDH hi and ALDH hi /CD133 + Cells Toward Insulin-Expressing Cells

Abstract

Introduction

Materials and Methods

Preparation and dissociation of human fetal pancreata

Fluorescence-activated cell sorting

Expansion and differentiation culture of sorted cells

Immunofluorescence analyses of cell populations

Quantitative reverse transcription polymerase chain reaction (qRT-PCR) analyses of cell populations

Statistical analyses

Results

Sorted ALDHhi cells highly co-expressed CD133 in the developing human pancreas

Sorted ALDHhi cells showed higher expression of endocrine TFs

ALDHhi cells represent a heterogeneous pancreatic population showing low proliferation

Sorted ALDHhi cells lose expression of endocrine transcript factors during in vitro expansion

Cultured ALDHhi cells restore ductal phenotype during the formation of cell clusters

Examination of ALDHhi/CD133 + sorted cells in vitro

Discussion

Footnotes

Acknowledgments

Author Disclosure Statement

References

Supplementary Material

Sorted ALDH^hi cells highly co-expressed CD133 in the developing human pancreas

Sorted ALDH^hi cells showed higher expression of endocrine TFs

ALDH^hi cells represent a heterogeneous pancreatic population showing low proliferation

Sorted ALDH^hi cells lose expression of endocrine transcript factors during in vitro expansion

Cultured ALDH^hi cells restore ductal phenotype during the formation of cell clusters

Examination of ALDH^hi/CD133⁺ sorted cells in vitro