How to cite this article: Krishan A, Sharma D, Sharma S, Hamelik RM, Ganjei-Azar P, Nadji M. ALDH+/CD44+/CD24− expression in cells from body cavity fluids. Cytometry Part B 2010; 78B: 176–182.
Enhanced expression of aldehyde dehydrogenase 1 (ALDH1) and phenotypic markers (CD44+/CD24−) in stem cells from breast tumors has been reported. This study was undertaken to monitor expression of these markers in cells from body cavity fluids of female patients suspected to have a malignancy.
Cells from peritoneal and pleural fluids of 100 female patients were examined by diagnostic cytology and analyzed by laser flow cytometry for enhanced ALDH1 expression. Cells from 36 body cavity fluids with ALDH1bright fluorescence were then analyzed for the expression of CD44 and CD24 markers.
In samples positive for malignancy, ALDH1bright cells with both SSClow and SSChigh were seen. In 15 body cavity fluids positive for malignancy, the percentage of ALDH1bright cells ranged from 0.26 to 6.34% of the total cells. The percentage of ALDH1bright cells with CD44+/CD24− expression in these samples ranged from 0.02 to 3.66%. ALDH1bright cells with CD44+/CD24− expression were also present in body cavity fluids of patients in whom diagnostic cytology could not detect any malignancy. However, the percentage of ALDH1bright and CD44+/CD24− cells amongst the 21 body cavity fluids with negative cytology was lower than that of samples with malignancy.
Several recent studies have focused on the expression of stem cell markers in human solid tumors and in metastatic tumor cells in bone marrow (1–4). In one of the earliest studies, Al-Hajj et al. (5) identified and isolated the tumorigenic breast tumor stem cells as CD44+/CD24−/Lineage- in eight of nine breast cancer patients. In Brca1 positive breast tumors, distinct CD44+/CD24−/CD133+ cells with cancer stem cell characteristics and increased expression of stem cell associated genes (e.g., Oct4, Notch1, ALDH1, Fgfr1, and Sox1) has been reported by Wright et al. (6). Honeth et al. (7) reported that CD44+/CD24− cells (ranging from only a few to close to 100% of tumor cells) were detected in 31% of basal-like breast tumors.
Besides the CD44+/CD24−/CD133+ markers, over-expression of aldehyde dehydrogenase 1 (ALDH1) has been reported in hematopoietic and solid tumor stem cells (8). Ginestier et al. (2) have shown that normal and malignant human mammary epithelial cells with increased ALDH1 activity have stem/progenitor like properties and ALDH1 expression correlated with poor prognosis. In colon cancer patients, Huang et al. (9) reported that progression from normal to mutant epithelium and adenoma was accompanied by significant increase in the number of ALDH1bright/ CD133+/ CD44+ cells.
Although expression of these markers has been described in both hematopoietic and solid tumor stem cells, there are no published reports on their expression in cells from body cavity fluids of patients suspected to have a malignancy. In the present study, we have analyzed cells from body cavity fluids of female patients for the coexpression of ALDH1, CD44 and CD24 markers.
MATERIALS AND METHODS
Processing of Body Cavity Fluids
Residual peritoneal and pleural fluids were collected under University of Miami and Jackson Medical Centre approved IRB protocols. After coarse filtration thru gauze, cell pellets were prepared by centrifugation (relative centrifugal force x 220, 5 min at 4°C) and washed once with 5 ml of phosphate buffered saline (PBS). Trypan blue was used to count the number of dye excluding cells. Most of the samples were analyzed within 24 hours of collection and storage in a refrigerator. A total of 100 samples were collected and out of these, 36 had distinct ALDH1bright cells as determined by comparison of cells incubated with or without the ALDH1 inhibitor, diethylaminobenzaldehyde (DEAB). Fifteen out of the 36 samples were confirmed by diagnostic cytology to have malignant cells. Seven of the 36 samples were described by diagnostic cytology to contain atypical epithelial or inflammatory cells while the remaining 14 samples did not contain any recognizable malignant cells.
Slides prepared from cells in the body cavity fluids were stained and examined by the diagnostic pathologist for the presence of malignant cells. Where ever feasible, immunohistochemisty was used to stain the cells for the presence of markers such as estrogen and thyroid transcription factor-1.
Aldefluor® staining was performed according to the protocol described by Storms et al. (10). An aliquot of 2 × 106 cells was centrifuged and washed with 2 ml of PBS. The cell pellet was resuspended in 1 ml of Aldefluor® buffer containing 5 μl of activated Aldefluor® reagent (StemCell Technology, Durham, NC). One half of the sample was transferred to a second tube containing 5 μl of the ALDH1 inhibitor, DEAB. Samples were mixed and incubated in a 37°C water bath for 1 h with intermittent shaking.
After 60 min of incubation with the Aldefluor® reagent, 20 μl of CD44-RPE (catalogue no 555479, clone-G44-26, BD Pharmingen™, San Jose, CA) and 5 μl of CD24-PE-Alexa Fluor® 610 conjugate, (catalogue no. MH CD2422, clone SN3, Invitrogen, Camarillo, CA.) were added to the incubation mixture for 20 min at room temperature in the dark. After centrifugation and washing, the cell pellets were suspended in 400 μl of Aldefluor® buffer.
Samples were analyzed on a Beckman Coulter XL flow cytometer. Threshold and electronic gates were used in Forward Scatter versus Side Scatter plots to exclude debris and red blood cells. Electronic gates were used to identify cells with ALDH1 positive expression by comparing dot plots of cells incubated with Aldefluor® with or without the ALDH1 inhibitor, DEAB. Cells with ALDH1 positive expression and high and low side scatter were subsequently analyzed for CD44 and CD24 expression. List mode data analysis and color compensation was performed with WinList software (Verity Software, Topsham, ME).
Tables 1 and 2 list the diagnostic cytology evaluation and percentage of ALDH1bright cells with CD44+/CD24− phenotype in the 36 body cavity fluids (19 peritoneal and 17 pleural fluids) with ALDH1bright cells.
Table 1. Characteristics of the 19 Peritoneal (ascites) Fluids Analyzed
No tumor cells detected, (history of squamous cell carcinoma)
Body Cavity Fluids with Malignant Cells
Figure 1 shows dot plots of cells from body cavity fluid of a patient with primary breast adenocarcinoma in which diagnostic cytology confirmed the presence of estrogen receptor positive malignant cells (Table 2, PL-63). In Side Scatter vs. Forward Scatter plots, threshold and electronic gates were used to exclude debris (data not shown). Dot plots 1A and 1B are respectively of cells incubated with Aldefluor® and Aldefluor® plus ALDH1 blocker (DEAB). Comparison of these two plots identifies the cells with high ALDH1 expression in gates R2 and R3 of SSClow and SSChigh, respectively. In dot plot 1B of cells incubated with the ALDH1 inhibitor (DEAB), very few cells with ALDH1bright expression were seen in these two gates. Dot plots 1C and 1D show the expression of CD44 and CD24 in ALDH1bright cells (gates R2 and R3 in Fig. 1A) with SSClow and SSChigh. Amongst the ALDH1bright/SSClow, (2.66% out of 4.1%) or 64% were CD44+/CD24− (Fig. 1C). In contrast 30.7% (1.3% out of 4.1%) of the ALDH1bright/SSClow were double positive for the expression of CD44 and CD24. CD44+/CD24− and CD44+/CD24+ cells amongst the ALDH1bright/SSChigh cells were 44.6% and 52.6% (1.0% and 1.18% out of 2.24%), respectively as shown in Figure 1D.
In contrast to data shown in Figure 1 where the dot plot1A shows two distinct populations of ALDH1bright cells with low and high side scatter, some of the samples from patients with malignant cells had a diffuse cluster of ALDH1bright cells (Fig. 2A, gate R2). These cells were not present in the sample incubated with the DEAB inhibitor and thus were presumed to be ALDH1bright cells. Figure 2B shows that most of the ALDH1bright cells in this sample had CD44+/CD24− expression.
In Figure 2C, cells from body cavity fluid of a patient with metastatic breast carcinoma shows the presence of ALDH1 bright cells with low (gate R2) and high side scatter. In Figure 2D, gate R3 shows that most of the ALDH1bright cells with low sided scatter in this sample had CD44+/CD24− expression. A small population (0.57%) were CD44+ with mid level CD24 expression (gate R4).
Body Cavity Fluids with Atypical or Inflammatory Cells
Figure 3 shows two examples of body cavity fluids in which diagnostic cytology did not find any malignant cells. In dot plot 3A (pleural fluid with atypical, epithelial cells), 1.8% of the cells had ALDH1bright expression as determined by comparison with dot plots of cells incubated with ALDH1 inhibitor, DEAB (data not shown). Gated analysis of the ALDH1bright cells in this sample shown in Figure 3B shows that out of the 1.8% ALDH1bright cells, 1.17% had CD44+/CD24− expression.
In cells from another diagnostic cytology negative sample (pleural fluid with inflammatory cells) shown in Figures 3C and 3D, a distinct population of ALDH1bright cells (1.66%) was seen. As shown in Figure 3D, most of these cells (0.93%) had positive expression of CD44 and CD24. A small population (0.67%) of the ALDH1bright cells was CD44+/CD24− as shown in gate R3 of Figure 3D.
As shown in Table 3, the percentage of ALDH1bright cells in the 15 body cavity fluids in which presence of malignant cells was confirmed by diagnostic cytology, ranged from 0.26 to 6.34% of the total cells. The percentage of ALDH1bright/ cells with CD44+/CD24− expression ranged from 0.02 to 3.66%. Amongst the 21 body cavity fluids in which diagnostic cytology did not detect any malignant cells, the percentage of ALDH1bright cells was 0.06–2.6% and the percentage of CD44+/CD24− cells was 0.02–1.7%.
Table 3. Percentage of ALDH1bright/CD44+/CD24− Cells in Body Cavity Fluids of Patients with Benign and Malignant Cytology
Bold values are of the percentage of cells with ALDH1/CD44 positive and CD24 negative expression. The values in brackets are of standard deviation and values at 95% confidence level.
0.26–6.34% (±1.3, 95% CI = 0.84–2.3)
0.02–3.66% (±0.8, 95% CI = 0.36–1.29)
0.06–2.60% (±0.8, 95% CI = 0.95–1.7)
0.02–1.70% (±0.51, 95% CI = 0.33–0.81)
Detection and characterization of stem cells in solid tumors offers the possibility that therapeutic modalities to selectively target tumor stem cells can be developed resulting in possible control of tumor growth and metastasis. However, it is important that markers specific to tumor stem cells be identified so as to provide a rationale for selective destruction of the tumor stem cells while sparing those in the normal tissues. Expression of cell surface markers (e.g., CD44, CD133), certain detoxifying enzymes such as ALDH1 and drug efflux characteristics of stem cells [for identification of side population (SP) phenotype] have been used to identify and sort putative tumor stem cells for further study in vitro and for testing of tumorigenicity in immunodeficient mice (1, 6, 11–15).
Aldehyde dehydrogenases are involved in metabolism of aldehydes to their corresponding carboxylic acids. ALDH1 may have a role in early stem cell differentiation by oxidizing retinol to retinoic acid (16). Increased ALDH1 activity has been reported in murine and human hematopoietic and cord blood stem cells and in solid tumor such as those of pancreas and lung (17–22).
Flow cytometric detection of ALDH1 in single cells was made possible by development of a fluorescent substrate, BODIPY® aminoacetaldehyde (BAAA), which after intracellular diffusion is converted by ALDH1 into BODIPY® aminoacetate (BAA), a green fluorescent marker (10). By blocking ALDH1 with DEAB and comparing cells incubated with and without the blocker and stained with the Aldefluor® dye, one can rapidly identify the ALDH1bright cells.
Several studies have used the Aldefluor® staining method for identification and sorting of the ALDH1bright cells and shown that ALDH1bright cells with low side scatter (SSClow) are capable of forming tumors in NOD-SCID mice (2, 4, 9). Hess et al. (23) showed that ALDH1bright cells from human cord blood express primitive stem cell markers such as CD34 and CD133. Pearce and Bonnet (24) purified ALDH1positive lineagenegative cells from murine bone marrow and showed that they overlapped with the SP cells. Gentry et al. (25, 26) showed that sorted ALDH1bright cells from cord blood were highly enriched in hematopoietic colony forming cells and ALDH1bright cells from human bone marrow formed endothelial and fibroblast colonies in culture.
Ginestier et al. (2) reported that in normal human and breast tumors 3–10% of the cells are ALDH1positive and are more efficient in forming mammospheres in vitro than the ALDH1negative cells. In NOD-SCID mice implantation of the sorted ALDH1positive cells (but not the ALDH1negative cells) led to development of tumors.
Besides ALDH1, several other studies have reported on the expression of markers such as CD44, CD24 and CD133 in tumor cell lines and solid tumors. The importance of these markers was shown by Al-Hajj et al. (5) who reported that as few as 100 CD44+/CD24− and Lin− cells isolated from eight out of nine breast tumor patients were capable of tumor formation in NOD-SCID mice. Wright et al. (6) reported that expression of stem cell associated genes (Oct4, Notch1, ALDH1, Fgfr1, and Sox1) was increased in CD44+/CD24− and CD133+ cells. Croker et al. (4) used breast tumor cell lines to show that cells with ALDH1high/CD44+/CD24− and CD133+ expression demonstrated increased colony formation in vitro and tumorigenicity and metastasis in NOD-SCID mice as compared to cells with low ALDH1 and CD44+ expression. Similarly, Huang et al. (9) reported that ALDH1positive colorectal cells implanted in NOD-SCID mice generated tumors while the ALDH1negative cells did not form tumors. In this study, coexpression of CD44 and CD133 with ALDH1 had a modest effect on tumor initiation.
In contrast to these studies, other workers have reported heterogeneity in expression of these markers in normal and malignant tissues. For example, Honeth et al. (7) reported that 35 and 31% of breast tumors were CD44−/CD24− and CD44+/CD24−, respectively. In HER2+ tumors, the CD44+/CD24− phenotype was scarce as they had predominantly CD24+ expression. In breast tumor cell lines, Fillmore et al. (3) reported that percentage of CD44+/CD24− cells did not correlate with the tumorigenicity, but as few as 100 CD44+/CD24−/ESA+ cells could form drug resistant tumors. Mylona et al. (27) reported that CD44+/CD24− and CD44−/CD24+ expression was seen in 58.7 and 82.6% of paraffin embedded breast tumors, respectively. The CD44+/CD24− phenotype had no effect on prognosis but had a tendency towards increased disease free survival. In contrast, the CD44−/CD24+ patients had worse disease free period and overall survival.
Review of published data on ALDH1, CD44, CD24, and CD133 expression in cell lines based on growth in vitro and in immunodeficient mice indicate that these markers are expressed by both normal and tumor stem cells and show extensive heterogeneity and variation in their individual expression and coexpression with other markers.
In the present study, only 36 out of the 100 body cavity fluids analyzed, had ALDH1bright cells as detected by flow cytometric comparison of cells incubated with Aldefluor® reagent with or without the ALDH1 inhibitor, DEAB. In rest of the 64 samples, we did not see any distinct populations of ALDH1bright cells. Most of our samples were analyzed within 24 hours as residual body cavity fluids from diagnostic cytology lab can not be legally released before the diagnosis is completed. It is possible that in some of samples, ALDH1 activity was lost during storage.
In our dot plots, we saw a range of ALDH1 expression from tight clusters to diffuse distribution of ALDH1bright cells. In some of the samples such as shown in Figure 1A, ALDH1bright cells had two distinct populations with SSClow and SSChigh expression. In other samples, such as Figures 2A and 2C, the ALDH1bright cells were more diffuse in their distribution. The percentage of ALDH1bright cells in body cavity fluids in which diagnostic cytology had reported the presence of malignant cells was higher (0.06–6.34%, Table 3). In contrast, the samples which contained only atypical or inflammatory but no malignant cells had a lower percentage of ALDH1bright cells (0.06–2.6%). Similarly, the body cavity fluids with malignant cells had a higher percentage of CD44+/CD24− cells (0.02–3.66%) than the benign body cavity fluids (0.02–1.7%). These observations suggest that correlation of CD44+/CD24− expression with malignancy in body cavity fluids may not be valid as some of the benign samples with no evidence of malignancy show cells with high expression of this phenotype.
It should be noted that, even though ALDH1bright/CD44+/CD24− phenotype may be characteristic of breast cancer stem cells, ALDH1 activity alone might not be a marker for “stemness.” As demonstrated by Estes et al. (28) in human adipose-derived adult stem cells, ALDH1positive cells did not have increased chondrogenic potential than that of ALDH1negative cells. Thus, it is critical to carefully characterize the populations with ALDH1 activity and expression of the cancer stem cell phenotype markers such as CD44+/CD24−/CD133+.
The authors thank Mr. Alfredo F. Cordoves for collecting, storing, and providing the body cavity fluids for this study.