Aldehyde dehydrogenase and ATP binding cassette transporter G2 (ABCG2) functional assays isolate different populations of prostate stem cells where ABCG2 function selects for cells with increased stem cell activity

Abstract

Introduction: High expression of aldehyde dehydrogenase1A1 (ALDH1A1) is observed in many organs and tumors and may identify benign and cancer stem cell populations.

Methods: In the current study, the stem cell characteristics were determined in cells isolated from human prostate cell lines and clinical prostate specimens based upon the ALDEFLUOR™ assay. Cells isolated based on the ALDEFLUOR™ assay were compared to cells isolated based on ATP binding cassette transporter G2 (ABCG2) activity using the side population assay. To test for stem cell characteristics of self-renewal and multipotency, cells with high and low ALDH1A1 activity, based on the ALDEFLUOR™ assay (ALDHHi and ALDH Low), were isolated from prostate clinical specimens and were recombined with rat urogenital sinus mesenchyme to induce prostate gland formation.

Results: The percentage of ALDH Hi cells in prostate cell lines (RWPE-1, RWPE-2, CWR-R1, and DU-145) was 0.5 to 6%, similarly in non-tumor and tumor clinical specimens the percentage of ALDH Hi cells was 0.6 to 4%. Recombinants using ALDH Hi cells serially generated prostate tissue up to three generations with as few as 250 starting cells. Immunohistochemical analysis of the recombinants using ALDHHi cells contained prostatic glands frequently expressing androgen receptor (AR), p63, chromogranin A, ALDH1A1, ABCG2, and prostate specific antigen (PSA), compared to their ALDH Low counterparts. Inhibition of ALDH resulted in the reduction of sphere formation capabilities in the CWR-R1, but not in the RWPE-2 and DU-145, prostate cell lines. ABCG2 inhibition resulted in a more robust decrease of sphere formation in androgen sensitive cell lines, CWR-R1 and RWPE-2, but not androgen insensitive DU-145. ALDH1A1 expression was enriched in ALDH Hi cells and non-side population cells. ABCG2 expression was only enriched in side population cells.

Conclusions: The percentage of ALDHHi cells in prostate cell lines and prostate tissue was consistently higher compared to cells with high ABCG2 activity, identified with the side population assay. The expression of the stem and differentiation markers indicates the ALDH Hi recombinants contained cells with self-renewal and multipotency activity. When the two assays were directly compared, cells with the side population phenotype demonstrated more stem cell potential in the tissue recombination assay compared to ALDH Hi cells. The increased stem cell potential of side population cells in the tissue recombination assay and the decrease in sphere formation when ABCG2 is inhibited indicates that the side population enriches for prostate stem cells.

Figure 1

Isolation of ALDH^Hi and ALDH^Low cells from prostate cells by ALDEFLUOR™ assay. RWPE-1, RWPE-2, CWR-R1 and DU-145 prostate cells were stained with ALDEFLUOR™ reagent and analyzed using BD FACS Aria II sorter. ALDH^Hi cells were detected in the higher fluorescent region (A, C, E and G). ALDH^Hi population was inhibited with DEAB resulting in no ALDH^Hi cell detection (B, D, F and H). The ALDH^Low population is gated with a clear separation from the ALDH^Hi population.

Figure 2

Isolation of ALDH^Hi and ALDH^Low cells from clinical human prostate specimen by ALDEFLUOR™ assay. Cells isolated from clinical human prostate specimen were stained with ALDEFLUOR™ reagent and analyzed using BD FACS Aria II sorter. ALDH^Hi cells were detected in the higher fluorescent region (A, C and E). ALDH^Hi population was inhibited with DEAB resulting in no ALDH^Hi cell detection (B, D and F). ALDH^Low population is gated with a clear separation from the ALDH^Hi population.

Figure 3

Number of human prostate glands generated per recombinant and recombinant survival rate of recombinants grafted. A. The percentage of recombinants with human ductal growth in the first generation ALDH^Hi n = 42 and ALDH^Low n = 41, analyzed by Fisher’s Exact test P = 0.4692. B. The number of glands per recombinant in the first generation derived from ALDH^Hi cells (n = 6) and ALDH^Low (n = 5) was quantitated and analyzed by unpaired t-test analysis P = 0.2558. C. Kaplan-Meier plot of recombinant survival rate in multiple generations, Logrank test P = 0.3890.

Figure 4

IHC analysis of prostate differentiation markers in a recombinant using 2,000 ALDH^Hi cells. IHC analysis of prostate differentiation markers in a recombinant using 2,000 ALDH^Hi cells: A. H & E staining; B. p63 IHC; C. AR IHC; D. PSA IHC; E. ALDH1A1 IHC; F. Higher magnification of boxed area of ALDH1A1 IHC in E; G. ABCG2 IHC; H. Higher magnification of boxed area of ABCG2 IHC in G; I. Chromogranin A (Chr A) IHC; J. Higher magnification of Chromogranin A IHC in a different region of same recombinant. Arrow indicates positive cell. Scale bar = 50 μm.

Figure 5

Sphere formation assay. RWPE-2, CWR-R1, and DU-145 prostate cells were treated in the presence of DEAB at 5, 10 and 20 μM or Ko143 at 1 and 5 μM concentration or vehicle control. Spheres were counted in each well after 10 to 14 days of growth. Sphere forming ability of RWPE-2 cells did not decrease upon treatment with DEAB (A). Sphere formation was decreased in RWPE-2 cells upon treatment with Ko143 at 5 μM concentration (B). Sphere formation was decreased in CWR-R1 cells upon treatment with DEAB (C) or Ko143 (D). Sphere forming ability of DU-145 cells did not decrease upon treatment with DEAB (E) or Ko143 (F).

Figure 6

Gene expression in ALDH^Hi, ALDH^Low, side population, and non-side population cells isolated from CWR-R1 cells. ABCG2, ALDH1A1, ALDH4A1, ALDH7A1 and ALDH9A1 gene expression was detected by real-time PCR in CWR-R1 cells isolated by the side population and ALDEFLUOR™ assays. Gene expression was normalized to GAPDH and fold change in gene expression in each population was calculated based on the formula: 2^-(ΔΔdCt). ABCG2 expression was the highest in the side population with 5.4-fold increase compared to non-side population (A), whereas ALDH1A1 expression was highest in NSP (A) and ALDH^Hi cells (B). Approximately 26-fold and 30-fold increase in ALDH1A1 gene expression was observed in NSP (A) and ALDH^Hi cells (B) respectively as compared to their counter populations. Expression of ALDH4A1, ALDH7A1 and ALDH9A1 genes was not significantly different in any of the populations (A and B). NSP, Non-Side Population; SP, Side Population.

Figure 7

Isolation of ALDH^Hi, ALDH^Low, side population and non-side population cells from clinical human prostate specimen. Cells isolated from clinical human prostate specimen were stained with either DCV or ALDEFLUOR™ reagent. The ALDH^Hi cells were detected in the higher fluorescent region (A). ALDH^Hi population was inhibited with DEAB resulting in no ALDH^Hi cell detection (B). Side population was gated based upon its absence in the presence of FTC, a specific inhibitor of ABCG2 (C and D).