Limits...
The side population of ovarian cancer cells defines a heterogeneous compartment exhibiting stem cell characteristics.

Boesch M, Zeimet AG, Reimer D, Schmidt S, Gastl G, Parson W, Spoeck F, Hatina J, Wolf D, Sopper S - Oncotarget (2014)

Bottom Line: In a systematic screen of 13 ovarian cancer cell lines we show that cells with stem cell properties are reliably detectable as a minor population, characterized by ABC transporter expression resulting in the side population (SP) phenotype.In different cell lines, either ABCG2 or ABCB1 was found to be responsible for this effect.Thus, our study confirms previous results showing that CSC are contained within the SP.

View Article: PubMed Central - PubMed

Affiliation: Internal Medicine V, Innsbruck Medical University, Innsbruck, Austria. Tyrolean Cancer Research Institute, Innsbruck, Austria.

ABSTRACT
Cancer stem cells (CSC) are believed to be involved in tumor evasion of classical antitumor therapies and have thus become an attractive target for further improvement of anticancer strategies. However, the existence and identity of CSC are still a matter of controversy. In a systematic screen of 13 ovarian cancer cell lines we show that cells with stem cell properties are reliably detectable as a minor population, characterized by ABC transporter expression resulting in the side population (SP) phenotype. In different cell lines, either ABCG2 or ABCB1 was found to be responsible for this effect. Purified SP cells featured virtually all characteristics of bona fide CSC, including clonogenicity, asymmetric division and high tumorigenicity in vivo. Using in-depth phenotyping by multicolor flow cytometry, we found that among the investigated ovarian cancer cell lines the SP compartment exhibits tremendous heterogeneity and is composed of multiple phenotypically distinct subpopulations. Thus, our study confirms previous results showing that CSC are contained within the SP. However, the exact identity of the CSC is still disguised by the high complexity of the CSC-containing compartment. Further functional studies are needed to determine whether a single cellular subset can unambiguously be defined as CSC or whether multiple stem cell-like cells with different properties coexist. Moreover, the observed heterogeneity may reflect a high level of plasticity and likely influences tumor progression, escape from immune-surveillance and development of resistance to anticancer therapies and should therefore be considered in the development of new treatment strategies.

Show MeSH

Related in: MedlinePlus

SPICE analysis of ovarian cancer heterogeneityDefined mixtures of corresponding SP and NSP fractions were stained for different markers (staining 1 of Suppl. Table 4) and analyzed by multicolor flow cytometry. (A) Definition of populations positive for individual markers (cell line: A2780). These gates were combined by Boolean operations to obtain proportions of cells within all possible combinations, which were then imported into SPICE for final data analysis. (B) SPICE analysis of ovarian cancer heterogeneity after class-division into SP (upper panel) and NSP (lower panel). Subset distributions are presented in the weighted category mode, and only subsets above a threshold of ≥0.1% are shown. Data are representative examples of at least two independent experiments. SP, side population; NSP, non-SP.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4196181&req=5

Figure 5: SPICE analysis of ovarian cancer heterogeneityDefined mixtures of corresponding SP and NSP fractions were stained for different markers (staining 1 of Suppl. Table 4) and analyzed by multicolor flow cytometry. (A) Definition of populations positive for individual markers (cell line: A2780). These gates were combined by Boolean operations to obtain proportions of cells within all possible combinations, which were then imported into SPICE for final data analysis. (B) SPICE analysis of ovarian cancer heterogeneity after class-division into SP (upper panel) and NSP (lower panel). Subset distributions are presented in the weighted category mode, and only subsets above a threshold of ≥0.1% are shown. Data are representative examples of at least two independent experiments. SP, side population; NSP, non-SP.

Mentions: As exemplified in the right two panels of Fig. 4A, several of the markers investigated exhibited biphasic expression, consistent with the presence of distinct subpopulations. Thus, in an attempt to further refine the putative CSC pool, we determined by multicolor flow cytometry whether these markers were co-expressed or expressed on mutually exclusive subpopulations (Fig. 5A). To this end, a combination of eight markers (i.e., SP/NSP discrimination plus seven additionally selected markers, Suppl. Table 4) was used, which theoretically results in 128 distinct cell populations of both SP and NSP. To exclude cell populations with limited biological significance, the cut-off for the definition of a bona fide cell population was set to 0.1% of total cells (corresponding to >85 cells). As results for such a vast number of parameters are difficult to visualize using common strategies we used SPICE, a software specifically designed for this purpose (Fig. 5B). The individual subpopulations are color-coded and their relative abundance is reflected by the size of sectors in the pie charts or the height of bars in the bar charts below. By simply comparing the color pattern of the marker combination CD24, CD49d, CD90, CD95, CD140a, CD184 and HLA-ABC, it becomes clear that the difference between cell lines is bigger than that between SP (upper panels) and NSP (lower panel) of the individual cell lines. Moreover, the level of heterogeneity as measured by the number of subsets is different between the cell lines, with A2780 being the most heterogeneous (19 populations above the threshold of 0.1% in SP and 18 in NSP), but similar when SP and NSP are compared. It is also obvious that most of the subpopulations existent in the bulk of NSP can also be found in the minor SP compartment, albeit at different proportions. For example in the IGROV1 cell line, the CD24+CD95+ (all other markers negative, blue) subset accounts for more than 50% of NSP but less than 25% of SP, where the CD24+CD95+HLA-ABC+ subset (green) predominates. Despite these differences between SP and NSP, no specific signature of SP subsets common to all cell lines could be detected. Interestingly though, the presence and proportion of the various subsets was stable over a period of several weeks (Suppl. Fig. 5). Importantly, when modifying the marker panel by exchanging two or three markers, additional subpopulations within the SP and the NSP became evident (Suppl. Fig. 6-12), although due to technical limitations in the number of fluorescence parameters, these subsets were not amenable to co-detection with the other subsets. Moreover, ALDH was also shown to contribute to SP and NSP heterogeneity by defining distinct subsets in both cell populations (data not shown). Finally in a proof-of-concept study in primary ovarian tumor tissue, we were able to detect a similar degree of intratumoral heterogeneity (Suppl. Fig. 4B), which corroborates our findings obtained from cell line models. Together, these data provide evidence for a high cellular complexity of both established ovarian cancer cell lines and freshly isolated primary ovarian tumor cells. Remarkably, even the stem-like SP compartment is composed of a plurality of distinct cell subsets.


The side population of ovarian cancer cells defines a heterogeneous compartment exhibiting stem cell characteristics.

Boesch M, Zeimet AG, Reimer D, Schmidt S, Gastl G, Parson W, Spoeck F, Hatina J, Wolf D, Sopper S - Oncotarget (2014)

SPICE analysis of ovarian cancer heterogeneityDefined mixtures of corresponding SP and NSP fractions were stained for different markers (staining 1 of Suppl. Table 4) and analyzed by multicolor flow cytometry. (A) Definition of populations positive for individual markers (cell line: A2780). These gates were combined by Boolean operations to obtain proportions of cells within all possible combinations, which were then imported into SPICE for final data analysis. (B) SPICE analysis of ovarian cancer heterogeneity after class-division into SP (upper panel) and NSP (lower panel). Subset distributions are presented in the weighted category mode, and only subsets above a threshold of ≥0.1% are shown. Data are representative examples of at least two independent experiments. SP, side population; NSP, non-SP.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4196181&req=5

Figure 5: SPICE analysis of ovarian cancer heterogeneityDefined mixtures of corresponding SP and NSP fractions were stained for different markers (staining 1 of Suppl. Table 4) and analyzed by multicolor flow cytometry. (A) Definition of populations positive for individual markers (cell line: A2780). These gates were combined by Boolean operations to obtain proportions of cells within all possible combinations, which were then imported into SPICE for final data analysis. (B) SPICE analysis of ovarian cancer heterogeneity after class-division into SP (upper panel) and NSP (lower panel). Subset distributions are presented in the weighted category mode, and only subsets above a threshold of ≥0.1% are shown. Data are representative examples of at least two independent experiments. SP, side population; NSP, non-SP.
Mentions: As exemplified in the right two panels of Fig. 4A, several of the markers investigated exhibited biphasic expression, consistent with the presence of distinct subpopulations. Thus, in an attempt to further refine the putative CSC pool, we determined by multicolor flow cytometry whether these markers were co-expressed or expressed on mutually exclusive subpopulations (Fig. 5A). To this end, a combination of eight markers (i.e., SP/NSP discrimination plus seven additionally selected markers, Suppl. Table 4) was used, which theoretically results in 128 distinct cell populations of both SP and NSP. To exclude cell populations with limited biological significance, the cut-off for the definition of a bona fide cell population was set to 0.1% of total cells (corresponding to >85 cells). As results for such a vast number of parameters are difficult to visualize using common strategies we used SPICE, a software specifically designed for this purpose (Fig. 5B). The individual subpopulations are color-coded and their relative abundance is reflected by the size of sectors in the pie charts or the height of bars in the bar charts below. By simply comparing the color pattern of the marker combination CD24, CD49d, CD90, CD95, CD140a, CD184 and HLA-ABC, it becomes clear that the difference between cell lines is bigger than that between SP (upper panels) and NSP (lower panel) of the individual cell lines. Moreover, the level of heterogeneity as measured by the number of subsets is different between the cell lines, with A2780 being the most heterogeneous (19 populations above the threshold of 0.1% in SP and 18 in NSP), but similar when SP and NSP are compared. It is also obvious that most of the subpopulations existent in the bulk of NSP can also be found in the minor SP compartment, albeit at different proportions. For example in the IGROV1 cell line, the CD24+CD95+ (all other markers negative, blue) subset accounts for more than 50% of NSP but less than 25% of SP, where the CD24+CD95+HLA-ABC+ subset (green) predominates. Despite these differences between SP and NSP, no specific signature of SP subsets common to all cell lines could be detected. Interestingly though, the presence and proportion of the various subsets was stable over a period of several weeks (Suppl. Fig. 5). Importantly, when modifying the marker panel by exchanging two or three markers, additional subpopulations within the SP and the NSP became evident (Suppl. Fig. 6-12), although due to technical limitations in the number of fluorescence parameters, these subsets were not amenable to co-detection with the other subsets. Moreover, ALDH was also shown to contribute to SP and NSP heterogeneity by defining distinct subsets in both cell populations (data not shown). Finally in a proof-of-concept study in primary ovarian tumor tissue, we were able to detect a similar degree of intratumoral heterogeneity (Suppl. Fig. 4B), which corroborates our findings obtained from cell line models. Together, these data provide evidence for a high cellular complexity of both established ovarian cancer cell lines and freshly isolated primary ovarian tumor cells. Remarkably, even the stem-like SP compartment is composed of a plurality of distinct cell subsets.

Bottom Line: In a systematic screen of 13 ovarian cancer cell lines we show that cells with stem cell properties are reliably detectable as a minor population, characterized by ABC transporter expression resulting in the side population (SP) phenotype.In different cell lines, either ABCG2 or ABCB1 was found to be responsible for this effect.Thus, our study confirms previous results showing that CSC are contained within the SP.

View Article: PubMed Central - PubMed

Affiliation: Internal Medicine V, Innsbruck Medical University, Innsbruck, Austria. Tyrolean Cancer Research Institute, Innsbruck, Austria.

ABSTRACT
Cancer stem cells (CSC) are believed to be involved in tumor evasion of classical antitumor therapies and have thus become an attractive target for further improvement of anticancer strategies. However, the existence and identity of CSC are still a matter of controversy. In a systematic screen of 13 ovarian cancer cell lines we show that cells with stem cell properties are reliably detectable as a minor population, characterized by ABC transporter expression resulting in the side population (SP) phenotype. In different cell lines, either ABCG2 or ABCB1 was found to be responsible for this effect. Purified SP cells featured virtually all characteristics of bona fide CSC, including clonogenicity, asymmetric division and high tumorigenicity in vivo. Using in-depth phenotyping by multicolor flow cytometry, we found that among the investigated ovarian cancer cell lines the SP compartment exhibits tremendous heterogeneity and is composed of multiple phenotypically distinct subpopulations. Thus, our study confirms previous results showing that CSC are contained within the SP. However, the exact identity of the CSC is still disguised by the high complexity of the CSC-containing compartment. Further functional studies are needed to determine whether a single cellular subset can unambiguously be defined as CSC or whether multiple stem cell-like cells with different properties coexist. Moreover, the observed heterogeneity may reflect a high level of plasticity and likely influences tumor progression, escape from immune-surveillance and development of resistance to anticancer therapies and should therefore be considered in the development of new treatment strategies.

Show MeSH
Related in: MedlinePlus