Fanconi anemia genes are highly expressed in primitive CD34⁺hematopoietic cells

Abstract

Background

Fanconi anemia (FA) is a complex recessive genetic disease characterized by progressive bone marrow failure (BM) and a predisposition to cancer. We have previously shown using the Fancc mouse model that the progressive BM failure results from a hematopoietic stem cell defect suggesting that function of the FA genes may reside in primitive hematopoietic stem cells.

Methods

Since genes involved in stem cell differentiation and/or maintenance are usually regulated at the transcription level, we used a semiquantitative RT-PCR method to evaluate FA gene transcript levels in purified hematopoietic stem cells.

Results

We show that most FA genes are highly expressed in primitive CD34-positive and negative cells compared to lower levels in more differentiated cells. However, in CD34^- stem cells the Fancc gene was found to be expressed at low levels while Fancg was undetectable in this population. Furthermore, Fancg expression is significantly decreased in Fancc -/- stem cells as compared to wild-type cells while the cancer susceptibility genes Brca1 and Fancd1/Brac2 are upregulated in Fancc-/- hematopoietic cells.

Conclusions

These results suggest that FA genes are regulated at the mRNA level, that increased Fancc expression in LTS-CD34⁺ cells correlates with a role at the CD34⁺ differentiation stage and that lack of Fancc affects the expression of other FA gene, more specifically Fancg and Fancd1/Brca2, through an unknown mechanism.

Background

Fanconi anemia (FA) is a genetic autosomal recessive disease characterised by bone marrow (BM) failure associated with cancer susceptibility and congenital defects [1]. Somatic cell fusion studies of FA cells have revealed eight complementation groups suggesting the existence of at least eight genes implicated in the disease [2]. Six FA genes have been cloned, FANCA, FANCC, FANCD2, FANCE, FANCF and FANCG and recently the Breast cancer susceptibility gene, BRCA2, has been assigned to complementation group D1 and possibly to group B although this requires further confirmation [3–10]. Knockout mouse models from complementation group A (Fanca -/-), C (Fancc -/-) and G (Fancg -/-) have been generated in order to study the molecular basis of this disease [11–14]. Bone marrow (BM) failure similar to that observed in FA patients was shown in Fancc -/- mice after treatment with a DNA damaging agent, mitomycin C, showing progressive pancytopenia and reduced BM cellularity [15]. Also, hematopoietic stem cells from Fancc -/- mice were shown to have a reduced long-term reconstitution ability in recipient mice concomitant with a reduced number of primitive (Lin^-) Thy1.2^lowc-kit⁺Sca-1⁺CD34⁺ hematopoietic cells [16, 17]. These results suggest that FA genes may be involved at specific stages of stem cell growth and/or differentiation. Since genes involved in stem cell differentiation are highly regulated at the RNA level [18–22], we evaluated FA genes expression pattern in murine hematopoietic cell populations more specifically in lineage-negative (Lin^-) Thy1.2-Sca-1⁺CD34⁺ and Thy1.2-Sca-1⁺CD34^- stem cells. Our results show that in wild-type BM, FA genes are highly expressed in the Lin^-Thy1.2^-Sca-1⁺CD34⁺ stem cell population and that Fanc gene expression is altered in Fancc -/- hematopoietic cells.

Methods

Hematopoietic cell purification

Bone marrow (BM) cells were obtained from 4 to 6 month old Fancc -/- and Fancc +/+ mice (C57BL/6J, 11^th generation of backcrosses) as previously described [17]. Bone marrow were collected from femurs, tibias and humeri, resuspended in PBS supplemented with 2% fetal bovine serum (FBS, Life Technologies, Burlington, ON, Canada) and depleted of red blood cells in ammonium chloride solution (StemCell Technologies Inc., Vancouver, BC, Canada) for 10 min at 4°C. Total BM cells were either used for RT-PCR analysis or further divided into subpopulations as follows. Total BM cells were depleted of lineage (CD5, CD45R, CD11b, myeloid differentiation antigen Gr-1, TER 119) and Thy1.2 positive cells using StemSep negative cell selection procedure as described by the manufacturer (StemCell Technologies Inc., Vancouver, BC, Canada). The lineage and Thy1.2 depleted (Lin^-) cell population obtained was either used for gene expression analysis or further sorted into Sca-1⁺CD34⁺ (LTS-CD34⁺) or Sca-1⁺CD34^- (LTS-CD34^-) populations as previously described [17]. Briefly, Lin^-Thy1.2^- cells were labeled with fluorochrome conjugated rat anti-mouse antibodies directed against CD34 and Sca-1 antigens (FITC-CD34 and PE-Sca-1, Ly-6A/E; PharMingen, Mississauga, ON, Canada). Rat IgG2a monoclonal antibodies conjugated to FITC or PE were used as isotype standards. Cells were labeled 30 min at 4°C, washed and resuspended in PBS supplemented with 2% FBS. Cells were sorted on an Epics Coulter cell sorter (BeckmanCoulter Canada, Mississauga, ON, Canada). Cells were defined by forward (FSC) and side (SSC) scatter and gated for Sca-1⁺ and CD34⁺ (LTS-CD34⁺) or CD34^- (LTS-CD34^-) populations (Figure 1A). Re-analysis of sorted Sca-1⁺CD34⁺ and Sca-1⁺CD34^- cells showed a purity of 85 to 95% in each fraction (Figure 1B). Purity of each cell fraction was also evaluated by RT-PCR as described below using specific murine CD34 primers (Table 1). No CD34 mRNA expression was detected in sorted LTS-CD34^- fraction (Figure 1C). Animal experiments were approved by the Animal Care Committee of Laval University, QC, Canada.

Figure 1

Sorting regions used for LTS-CD34 ⁺ and LTS-CD34 ^- cell purification. (A) Lineage and Thy1.2 positive cells were removed from murine BM cell preparation as described in materials and methods. Lin^-Thy1.2^- cells were sorted on the basis of CD34 and Sca-1 expression using R1 and R2 sorting windows. (B) Reanalysis of sorted Lin^-Thy1.2^-Sca-1⁺CD34^- and Lin^-Thy1.2^-Sca1⁺CD34⁺ cells. (C) Representative gel electrophoresis of RT-PCR products in sorted Lin^-Thy1.2^-Sca1⁺CD34^- and Lin^-Thy1.2^-Sca1⁺CD34⁺ cells from wild-type mice using either murine CD34 or Hprt primers.

Table 1 Oligonucleotides primers used for PCR analysis

RNA extraction and RT-PCR analysis

One to 5 × 10⁶ of total hematopoietic or Lin^-Thy1.2^- cells were used for total RNA purification and RT-PCR analysis. Total RNA was extracted using TRIZOL reagent (Life technologies) as described by the manufacturer and resuspended in diethylpyrocarbonate (DEPC)-treated water containing ribonuclease inhibitor Superase• in™ (1 U/μl, Ambion, Austin, TX, USA). For total BM and Lin^-Thy1.2^- cells, RNA concentration was quantified by spectrophotometry at 260 nm and 400 ng of total RNA was used for RT-PCR analysis. For LTS-CD34⁺ and LTS-CD34^- stem cell populations, the RNA extraction protocol was modified to improve yield from small cell numbers ranging from 5000 to 35000 cells. Briefly, total RNA was extracted using TRIZOL reagent and precipitated with glycogen as carrier. Each RNA preparation from LTS-CD34⁺ and LTS-CD34^- stem cells was used directly in RT-PCR reactions. The reverse transcription reaction was done with oligo (dT) primers using the SuperScript™II protocol as described by the manufacturer (Life technologies). For PCR amplification, 10 to 20 % of cDNAs from each cell preparation was directly added to a standard PCR amplification mixture. Sense and anti-sense primer sequences and the corresponding cDNA PCR product sizes are shown in Table 1. First, we determined optimal PCR conditions for linear amplification of gene transcripts and to avoid saturation conditions. Linear amplification of the hypoxanthine guanine phosphoribosyl transferase gene (Hprt) were obtained with cDNA obtained from 10 to 50 ng of total RNA or with RNA extracted from 6 000 to 60 000 hematopoietic cells (data not shown). Thirty cycles for BM or Lin^- cell RNA extract or 35 cycles for LTS-CD34⁺ or LTS-CD34^- cell RNA extract was carried out and found to be in the linear PCR reaction range. PCR products were separated on 8% polyacrylamide gel and visualized by ethidium bromide staining. Volumes of 5 or 10 ul of PCR products were deposited on gels according to the expression level of each gene. Signal intensities were analyzed using the BioRad Quantity one™ software. The relative expression level for each target gene was quantified by calculating the ratio of the target gene intensity signal to the Hprt intensity signal obtained from the same cDNA preparation and visualized on the same gel. Specificity of all primers was further confirmed by direct nucleotide sequencing analysis of PCR products. Each RNA extraction products were tested and found free of DNA contamination using PCR amplification on samples without RT reactions. Statistical analysis were performed using the Student t-test.

Results

FA genes are highly expressed in murine wild-type primitive hematopoietic cells

Several studies have surveyed changes in the expression profiles of many genes associated with hematopoietic cell growth and differentiation such as transcription factors, cytokine receptors and ligands, cell cycle related genes and integrins [18–22]. Since the hematopoietic defect in Fanconi anemia is thought to reside in the stem cell compartment [17], we sought to determine the expression profile of FA genes in murine hematopoietic stem and progenitor cells. We evaluated the mRNA expression levels of FA genes from complementation group A, C, D1/Brca2, D2, E, F and G in sorted hematopoietic stem cell fractions using a semiquantitative RT-PCR procedure. The relative expression of each FA gene was based on the housekeeping gene Hprt transcription level (equivalent to 1) in each sample tested allowing the comparison of Fanc gene expression levels between murine BM, lineage and Thy1.2-depleted, LTS-CD34⁺ and LTS-CD34^- hematopoietic cells. We found that Fanca, Fancc, Fancd2 and Fancg were highly expressed in LTS-CD34⁺ cells (Fanc/Hprt ratios ranging from 0.45 to 0.75) whereas Fancd1/Brca2, Fance, Fancf and the FA-associated cancer susceptibility gene Brca1 were expressed at low levels in these cells with ratios ranging from 0.09 to 0.30 (Figure 2 and 4). Expression of all Fanc genes tested was dramatically low in both BM and Lin^-Thy1.2^- cells with relative expression ranging from 0.029 to 0.36. Both Fance and Fancf were expressed at low levels in all BM cell populations tested with relative expression of 0.07 to 0.27 and 0.14 to 0.30 respectively.

Adult murine long-term repopulating stem cells are thought to reside in the CD34^- cell compartment while CD34⁺ cells have short-term reconstitution potential [23–25]. We measured Fanc gene mRNA expression in the LTS-CD34^- cell fraction. Our results show that Fanca is highly expressed in LTS-CD34^- cells (ratio of 0.875 ± 0.045) whereas other Fanc genes showed low mRNA expression levels (ratio ranging from 0.085 to 0.33). In addition, we were unable to detect Fancg expression in LTS-CD34^- stem cells (4 independent reactions). Our results suggest that most Fanc genes, and more dramatically Fancc and Fancg are upregulated in murine LTS-CD34⁺ cells correlating with a specific role at this hematopoietic cell differentiation stage.

Fanc genes are disregulated in Fancc-/- hematopoietic cells

We investigated the expression levels of Fanc genes in Fancc-/- hematopoietic cells in order to verify if absence of one FA gene, notably the murine Fancc, affects the expression profile of other Fanc genes. Our results show that Fanca, and Fancd2 are significantly upregulated in Fancc -/- total BM cells as compared to Fancc+/+ cells while Fance, Fancf and Fancg expression levels are either down regulated or unchanged (Figure 3). Similarly, Fanca showed a 7 fold significant increase in mRNA expression in Fancc-/- progenitors (Lin^-Thy1.2^-) compared to Fancc+/+ cells (Table 2). Surprisignly, Fancc-/- progenitor cells showed lower Fancd2 mRNA levels with an expression ratio of 0.18 compared to Fancc+/+ cells while Fance, Fancf and Fancg expression levels were similar between Fancc -/- and Fancc+/+ Lin^-Thy1.2^- cells.

Table 2 FA gene expression ratios in Fancc -/- hematopoietic cell population as compared to Fancc+/+ cells

We also evaluated Fanc mRNA expression in both CD34⁺ and CD34^- stem cells. We show that Fancg expression is significantly decreased in Fancc-/- LTS-CD34⁺ cells as compared to Fancc+/+ cells (ratio of 0.06). Again, Fancg expression could not be detected in Fancc-/- LTS-CD34^- cells as in Fancc+/+ cells. Other Fanc genes, Fanca, Fancd2, Fance or Fancf, did not show any significant changes in mRNA expression in Fancc-/- LTS-CD34⁺ and Fancc-/- LTS-CD34^- cells as compared to Fancc+/+ cells. These results suggest that absence of Fancc influences mRNA expression levels of other Fanc genes in progenitors and total BM cells.

The breast cancer susceptibility genes, Brca1 and Brca2/Fancd1 are upregulated in Fancc -/- hematopoietic cells

We evaluated mRNA expression levels of the newly identified FA gene Brca2/Fancd1, and the FA-associated gene, Brca1, in Fancc+/+ and Fancc-/- hematopoietic cells. We found that both Brca2/Fancd1 and Brca1 genes were expressed at low levels in all Fancc+/+ hematopoietic cell populations tested (ratios ranging from 0.29 to 0.36 and 0.09 to 0.24, respectively; Figure 4). We also evaluated mRNA levels in Fancc-/- hematopoeitic cells and found that both Brca2/Fancd1 and Brca1 are upregulated in Fancc-/- total BM, progenitors and CD34⁺ cells. Of all the FA genes tested, Fancd1/Brca2 and the FA associated gene Brca1 seemed to be the most affected from absence of Fancc with 5 to 7 fold increase in gene expression levels in Fancc-/- differentiated BM cells and progenitors (Lin-Thy1.2-), respectively (Table 2).

Figure 2

Fanc gene expression profiles in murine hematopoietic cells. (A) Representative gel electrophoresis of 10 ul (* represents 5 ul) of RT-PCR products from wild-type BM (BM), lineage and Thy1.2-depleated (Lin-) and sorted LTS-CD34+ (CD34+) and LTS-CD34- (CD34-) cells using specific Fanc and Hprt primers. (B) Mean relative expression of Fanc genes in wild-type hematopoietic cell populations from 2 to 10 separate determinations with the exception of Fancf in CD34- cells where values represent one determination. Fancg expression in CD34- cells was undetectable (n= 4).

Figure 3

Comparative analysis of Fanc gene expression in hematopoietic cells from Fancc -/- mice. (A) Representative gel electrophoresis of 10 ul (* represents 5 ul) of RT-PCR products from Fancc -/- BM cells, BM; lineage and Thy1.2-depleated, Lin^-; sorted LTS-CD34⁺, CD34⁺; and LTS-CD34^-, CD34^- cells using specific Fanc and Hprt primers. (B) Quantitative comparison of the expression levels of Fanc genes in hematopoietic cell populations from Fancc+/+ and Fancc-/- mice. Bars represent the mean value from 2 to 10 separate determinations with the exception of Fance and Fancf in CD34^- cells where values represent one determination. Fancg expression in CD34^- cells was undetectable in both Fancc+/+ (n= 4) and Fancc-/- cells (n= 4). Fancc +/+ expression ratios are identical to those in Figure 2. Significant differences between Fancc -/- and Fancc +/+ cells where * p < 0.05, ** p < 0.005 and † p < 0.001.

Figure 4

Comparative analysis of Brca1 and Brca2/Fancd1 gene expression in hematopoietic cells from Fancc -/- mice. (A) Representative gel electrophoresis of 10 ul and 5 ul of RT-PCR products using specific Brca1 or Brca2/Fancd1 and Hprt primers, respectively, from Fancc -/- BM cells, BM; lineage and Thy1.2-depleated, Lin^-; sorted LTS-CD34⁺, CD34⁺; and LTS-CD34^-, CD34^- cells. (B) Quantitative comparison of the expression levels of Brca1 and Brca2/Fancd1 genes in hematopoietic cell populations from Fancc+/+ and Fancc-/- mice. Bars represent the mean value from 4 to 10 separate determinations. Absence of SEM bars represents values to low to appear. Significant differences between Fancc -/- and Fancc +/+ cells where *p < 0.05 and ** p < 0.005.

It is difficult to know if the observed increase in Brca gene expression in Fancc-/- hematopoietic cells is also reflected at the protein level. Since there are no commercially available antibodies directed against murine Brca proteins, we used anti-human BRCA1 and BRCA2 antibodies in Western blot and immunoprecipitation techniques and could not detect murine Brca1 and Brca2 proteins (data not shown). Thus, we could not evaluate protein expression levels in Fancc-/- BM cells to determine if the observed mRNA expression profiles correlated with protein expression levels.

Nevertheless, these results suggest that absence of Fancc affects the regulation of other Fanc gene expression more dramatically, Brca2/Fancd1 and Brca1, through an unknown mechanism.

Discussion

Gene expression profiling and transcription regulation of different genes can be surveyed using several techniques such as Northern blots, differential display or microarray technology. Some may provide information on numerous genes simultaneously such as microarrays, but these tend to generate large and complex data. We chose a simple direct semiquantitative RT-PCR procedure to characterize specific gene expression levels in sorted hematopoietic stem and progenitor cell fractions. Since primitive hematopoietic stem cells are only obtainable in very limited amounts, PCR based methods are more sensitive and allow the detection of low abundance transcripts in small cell populations as compared to Northern blot analysis. In addition, we used single-round PCR amplifications followed by direct detection on polyacrylamid gels to keep relative abundance relationship between transcripts and to reduce variability between samples. The reverse transcription was performed using oligo (dT) primers and target gene expression ratios were based on the housekeeping gene, Hprt, transcription level (equivalent to 1.0) in the same sample, thus, allowing the comparison between different genes, between cell populations and between wild-type and Fancc -/- cells.

In the current study, we tested the hypothesis that Fanc genes display a pattern of expression reminiscent of a role in hematopoietic stem cell differentiation and/or maintenance. Based on phenotypic analysis of Fancc -/- mouse bone marrow, targeted deletion of Fancc alleles was shown to affect the number of primitive CD34⁺ cells suggestive of a role at the CD34⁺ stem cell differentiation stage [17]. We found that expression of most FA genes, notably, Fanca, Fancc, Fancd2 and Fancg, are highly expressed in CD34⁺ stem cells as compared to more differentiated cell populations whereas Fance and Fancf are expressed at similar levels in all hematopoietic cell populations. We showed that Fanca was highly expressed in both long-term CD34^- and short-term CD34⁺ stem cells whereas Fancc and Fancg appear to be downregulated in LTS-CD34^- stem cells and upregulated in CD34⁺ cells consistent with a role of Fancc at the CD34^- to CD34⁺ differentiation stage[17]. The fact that Fancg is undetectable in CD34^- stem cells and highly expressed in CD34⁺ cells suggest that this gene may also function at the CD34⁺ stage. Taken together, these results suggest that Fanc genes are differently regulated in hematopoietic stem cells and suggest that each FA proteins may act separately from the FA complex during hematopoiesis similarly to what has been suggested for MMC sensitivity and cytokine signaling [26–29].

Our results showing increased Fanca and Fancc expression in primitive hematopoietic stem cells and reduced levels in differentiated BM cells are consistent with previous studies showing Fanca and Fancc expression in cells of mesenchymal origin that give rise to hematopoietic tissues during mouse development [30]. In addition, studies showing Fancc expression in areas undergoing osteogenesis, in regions containing dividing and/or differentiating cells, in progenitor/precursor cell populations and in primordial germ cell development are also consistent with our results showing elevated Fancc expression in primitive hematopoietic cells [20, 32–34].

Mouse models for Fanca, Fancc and Fancg have been generated [11–14] and were shown to display similar phenotypes, these include hypersensitivity to DNA crosslinking agents, increased spontaneous and induced chromosomal breakage and germ cell loss leading to reduced fertility. Studies using the Fancc -/- mice have shown that lack of Fancc resulted in a reduced number of CD34⁺ stem cells [17]. Since the murine CD34 marker has been associated not only with the differentiation state or short-term repopulating cells, but also with activation and/or cycling state [24, 35], the reduced number of CD34⁺ cells in Fancc-/- mice and the fact that Fancc-/- stem/progenitor cells (Lin^-c-Kit⁺Sca-1⁺) were shown to be less quiescent [36] may also reflect an altered activation state. Murine Brca1 and Brca2 mRNA expression was shown to be up-regulated in rapidly proliferating cells indicative of a role in proliferation and differentiation control [37]. Also, murine Fancg mRNA was shown to be expressed in highly proliferative and embryonic tissues [31]. Since Fancc-/- progenitor/stem cells were shown to be less quiescent [36], our results showing increased Brca1 and Brca2/Fancd1 mRNA expression levels in Fancc-/- BM cells would be consistent with an altered proliferation state. However, this is not the case with Fancg expression pattern, where mRNA levels were found to be dramatically reduced in Fancc -/- LTS-CD34⁺ cells. Thus, the increased expression levels of Brca1 and Brca2/Fancd1 in Fancc-/- cells and reduced Fancg expression in CD34⁺ cells may not be a consequence of an increase in proliferation but may reflect disregulated gene expression due to absence of Fancc. The role that Fancc might have on other FA gene expression has yet to be determined.

Conclusions

Together our results show that FA genes are regulated at the mRNA level that increased Fancc expression in LTS-CD34⁺ cells correlates with a role at the CD34⁺ differentiation stage and that lack of Fancc downregulates Fancg expression in CD34⁺ stem cells but increases Brca1 and Brca2/Fancd1 expression in all BM cell populations. Differential expression of FA genes in hematopoietic cell population suggest that FA genes may have distinct function at different stem cell stages but that Fancc and Fancg may function specifically at the CD34⁺ differentiation stage.