Data from a published microarray reported differential expression of genes in human HSC populations that were sorted and isolated according to their cell cycle status (Graham et al., 2007). More specifically, CD34+ cells were isolated using flow cytometry and a
combination of DNA and RNA stains (Hoechst 33342 and Pyronin Y). Due to
technicalities, including the lack of a UV laser to detect Hoechst staining, this approach could not be replicated. An alternative approach to sorting cells according to cell cycle status was required. Initially, alternative DNA stains were tested which can be used in viable cells and do not require a UV laser, however these were shown to be unreliable (data not shown). As an alternative approach, cell populations were sorted using cell surface markers CD34 and CD38. CD34 is a cell surface marker known to be expressed on a heterogeneous population of cells including primitive stem cells and progenitor cells and including CD38 allows a generally accepted discrimination between more primitive, and therefore more quiescent cells (Civin et al., 1996, Bhatia et al., 1997, Paz et al., 2007).
Cells used in the original microarray study were derived from normal controls in which the stem cells had been mobilised (Graham et al., 2007). It is possible that this treatment could alter the gene expression signature, in particular for chemokine expression. Indeed,
CXCR2 binding ligands are modulated in response to G-CSF treatment (Richards et al., 2003, Eash et al., 2010). Therefore CD34+CD38- and CD34+CD38+ cells were isolated from normal BM samples derived from the BM to examine whether high levels of
chemokine expression is inherent to a primitive stem cell population regardless of cellular location and cytokine treatment.
Normal BM samples were enriched for cell surface marker CD34, stained with antibodies against CD34 and CD38 and sorted using flow cytometry into different cell populations according to their cell surface expression of CD34 and CD38. RNA was extracted, RT and examined for mRNA levels of CXC chemokines identified in the microarray; CXCL1, CXCL2 and CXCL6 using Q-PCR analysis. It can be seen in Figure 3-1 that all CXC
ligands were down regulated in the CD34+CD38+ fraction in comparison to the
CD34+CD38- fraction at the mRNA level however CXCL6 was not statistically significant.
It is likely this has arisen from a smaller sample size and variation between individuals.
From this result it can be inferred that CXC chemokines CXCL1 and CXCL2 are up
regulated in quiescent HSC derived from BM and mobilised PB samples as carried out in the original microarray study. It can also be inferred that sorting of CD34+ cells into CD34+CD38- and CD34+CD38+ fractions separates cells which are in different stages of the cell cycle as the results corroborate the findings from the original microarray.
However, it should be noted that a huge variation between samples was noted in both populations as seen by relative expression. It is possible chemokine levels vary greatly between individuals and information regarding gender and age would have been useful, however this information was not available.
As a control to validate that the CD34+CD38- fraction represents a more quiescent fraction than the CD34+CD38+ population, normal BM samples were examined for gene expression of genes associated with cell cycle status. CDC6 is only transcribed during the G1 phase of the cell cycle therefore it is predicated this will be at higher levels in CD34+CD38+ cells (Pelizon, 2003). In addition, gene expression analysis of cell surface marker CD38 was carried out to demonstrate the sorting efficiency. In one representative BM sample, CDC6 and CD38 mRNA levels were examined in CD34+CD38- and CD34+CD38+ sorted
populations. It can be seen from Figure 3-2 that both CDC6 and CD38 showed an increase in expression in the CD34+CD38+ fraction in comparison to the CD34+CD38- fraction. The higher expression of CD38 in the CD34+CD38+ fraction confirms that the sorting was efficient. Differential expression of CDC6 confirms that CD34+CD38+ are more
proliferative than CD34+CD38- cells, which justifies the use of these populations in this study. However due to one sample used, no significant differences are noted.
Figure 3-1 CXCL1 and CXCL2 are up regulated in CD34+CD38- compared to CD34+CD38+ cells derived from normal BM samples.
Normal BM samples were freshly isolated or recovered from cryogenic storage. Cells were enriched for CD34 and recovered overnight in medium supplemented with GF for cell survival. Cells were stained for antibodies against CD34 and CD38 and sorted for CD34+CD38- and CD34+CD38+ populations. RNA was extracted, RT and Q-PCR was carried out for CXCL1, CXCL2 and CXCL6 mRNA expression (A). Fold change was calculated relative to housekeeping control GAPDH according to the DeltaDeltaCT method. Data are presented as the mean fold change of expression in the CD34+CD38+ fraction using the CD34+CD38- fraction as a calibrator which is set to the value of 1. The chemokines CXCL1áCXCL2 and CXCL6 showed a mean 50 (P <0.05), 81 (P <0.05) and 84 (n.s) percent reduction in expression levels in the CD34+CD38+ fraction in comparison to the CD34+CD38- fraction. The panels in B demonstrate relative expression (2-DeltaCT) in both populations for each gene tested (B). Each dot displays an average of technical triplicates from independent samples. Statistical differences were analysed using the Wilcoxon matched paired test (n = 4-7) (* P <0.05). Details of age and gender from samples were not available.
A
B
Figure 3-2 CDC6 and CD38 show up regulation in CD34+CD38+ compared to CD34+CD38- cells derived from one normal, representative BM sample.
Normal BM samples were freshly isolated or recovered from cryogenic storage. Cells were enriched for CD34 and recovered overnight in medium supplemented with GF for cell survival. Cells were stained for antibodies against CD34 and CD38 and sorted for CD34+CD38- and CD34+CD38+ populations. RNA was extracted, RT and Q-PCR was carried out for CDC6 (A) and CD38 (B) mRNA expression. Fold change was calculated relative to housekeeping control GAPDH according to the DeltaDeltaCT method. Data are presented as the mean fold change of expression in the CD34+CD38+ fraction using the CD34+CD38- fraction as a calibrator which is set to the value of 1. Both CDC6 and CD38 showed an increase in expression in the CD34+CD38+ fraction with a 25.1 and 31.0 fold increase in expression levels in comparison to the CD34+CD38- fraction. No statistical analysis was used due to the sample size of 1. Details of sample age and gender were not available.
A B