A quantitative trait locus on chr.4 affects cycling of hematopoietic stem and progenitor cells through regulation of transforming growth factor-beta 2 responsiveness

Mice

4-6-week-old C57BL/6J mice were purchased from Jackson Laboratories (Bar Harbor, ME). B6.D2-chr.4 mice were generated through the Speed Congenic Service of the Jackson Laboratory (Bar Harbor, ME), using 150 single nucleotide polymorphism (SNP) markers (39, 40). After the initial C57BL/6 and DBA/2 mating, 80-160 animals of each generation after the F1 crossbreeding were scanned using 108 Mit markers, five of which, d4mit155, d4mit37, d4mit251, d4mit234 and d4mit190, were on chr.4. Of these, 20-25 samples with the telomeric chr.4 genotype of DBA/2 were then scanned using 150 SNP markers to select the animals with the highest percent of C57BL/6 in the remaining part of the genome and mated those to C57BL/6 mice to produce the next generation. After six generations (N6) animals were found to be 100% C57BL/6 for all SNP markers except in the telomeric 20cM of chr.4 which was DBA/2. N6 animals were mated to produce homozygous DBA/2 chr.4 telomeric region at generation N6F1. The sex chromosomes were fixed via mating. Animals were housed in a specific pathogen-free facility. Experiments and animal care were performed in accordance with the Mount Sinai Institutional Animal Care and Use Committee (IACUC).

Antibodies and cytokines

Unconjugated anti-CD2, -CD3ε, -CD8α, -CD4, -B220, -Gr1, -Mac1, PE-conjugated anti-CD45.1, biotinylated anti-Thy1 and FITC-conjugated goat anti-rat antibody were purchased from Southern Biotechnologies (Birmingham, AL). FITC-conjugated anti-CD2, -CD3ε, -CD8α, -CD4, -CD19, -B220, -Gr1, and -Mac1, PE-conjugated anti-flt3, PECy7-conjugated streptavidin and APC-AlexaFluor®750-conjugated anti-c-kit were purchased from eBiosciences (San Diego, CA). Unconjugated anti-Ter119, biotinylated anti-Sca1 and -CD34, PE-conjugated anti-Sca1 and -CD34, PerCP-Cy5.5-conjugated streptavidin and anti-CD45.2, APC-conjugated anti-c-kit and goat anti-rat antibody, PerCP-conjugated streptavidin, PECy7-conjugated anti-CD19, APC-Cy7-conjugated streptavidin and anti-CD19 were purchased from Pharmingen (San Diego, CA). Pacific blueconjugated anti-Sca1 were purchased from BioLegend (San Diego, CA). Recombinant cytokines were purchased from R&D Systems (Minneapolis, MN). Recombinant human FLT3 ligand was received from Amgen (Seattle, WA).

Cell sorting and flow cytometry

Bone marrow (BM) cells were prepared by flushing the femora and tibia of mice with cold DMEM (Cellgro, Mediatech, Manassas, VA) containing 2% FBS (Hyclone, Logan, UT) and penicillin/streptomycin (Cellgro). Mononuclear cells were obtained after gradient centrifugation using lymphocyte separation medium (Cellgro). Low-density BM cells were stained with antibodies for lineage antigens (CD2, CD3ε, CD8α, CD4, CD19, B220, Ter119, Gr1, Mac1), Sca1 and c-kit, and isolated by cell sorting, as shown in Figure S1, using FACSVantage SE (Beckon Dickinson, San Diego, CA), Cytopeia (Advanced Cytometry Systems, Seattle, WA) or MoFlo (CytoMation, Fort Collins, CO) sorters, to obtain lineage^-Sca1⁺c-kit⁺ (LSK) cells. Flow cytometric analysis was performed on a three-laser LSRII or a Special Order five-laser LSRII with DiVa software (Beckon Dickinson). Data were analyzed using FlowJo software. Doublets were excluded by plotting with forward scatter pulse area versus side scatter both for sorting and FACS analysis.

Culture of lineage ^-Sca1⁺c-kit⁺ (LSK) cells

Sorted LSK cells were cultured in triplicate at 20 to 60 cells per well in flat bottom 96-well plates in StemPro34 medium (Invitrogen), 10% FCS (Hyclone, Logan, UT), penicillin/streptomycin, in the presence of various cytokines, as mentioned for each experiment. Within an hour after plating, the exact number of cells per well was determined by visually counting the cells at 200x magnification. After 5 days of liquid culture in a humidified incubator with 5% CO₂ at 37°C the cells were again counted.

Cell cycle analysis of LSK cells

BM mononuclear cells were stained with lineage markers (CD2, CD3ε, CD8α, CD4, B220, Gr1, Mac1, Ter119, unlabeled antibodies), followed by goat anti-rat polyclonal IgG-FITC, PE-conjugated anti-Sca1 and APC-conjugated anti-c-kit. The cells were fixed in 1% paraformaldehyde (Sigma, St Louis, MO) and 0.2% Nonidet P40 (Sigma) for 1 hour at 4°C. Fixed cells were washed in PBS and labeled with Hoechst 33342 (Sigma), final concentration of 5μg/mL, for 1 hour at 37°C. Analysis was performed on a triple laser LSRII flow cytometer with DiVA software (Becton Dickinson). Double exclusion using pulse shape was done by plotting area versus height of the UV-excited Hoechst fluorescence. Cell cycle analysis was done using up to 3000 singlet LSK cells.

5-Fluorouracil administration

5-Fluorouracil (5-FU, Sigma) was dissolved in PBS at concentration of 20mg/mL for survival and of 11.25mg/mL for recovery studies, and filtered through 0.2μm SFCA filters (Corning). The effect of sublethal doses of 5-FU (150mg/kg, IP) was followed in 6 mice per experiment of which 3 were bled alternately every other day. Blood counts were performed using a Beckman-Coulter Ac·Tdiff automated hematology machine (Beckman Coulter, Inc., Miami, Florida, USA).

Competitive repopulation assays

LSK cells from donor mouse strain (C57BL/6 or B6.D2-chr.4, CD45.2⁺) were injected into lethally (700cG followed by 500 cG 3 hours later) irradiated recipient strain (C57BL/6, CD45.1⁺). Donor cells were mixed with 2·10⁵ bone marrow cell from CD45.1⁺CD45.2⁺ C57BL/6 mice in case of primary competitive repopulation transplantations. Peripheral blood cells were analyzed for the expression of CD45.1, CD45.2, and lineage antigens (Thy-1, Gr-1/Mac1, and CD19) 12 weeks after transplantation. Competitive serial transplantations were performed at least 16 weeks after reconstitution by injecting 2·10⁶ bone marrow cells from primary recipients into lethally irradiated secondary recipients (CD45.1⁺). Three secondary recipients were reconstituted with bone marrow from one primary donor (n=9 primary donors per donor strain). After 3 months, the ratio between the competing cell populations was measured in the peripheral blood of the secondary recipients (n=27 total recipients per donor strain). The data were presented as ratios between the two donor populations determined by the expression CD45 alleles. Changes in reconstitution ratios between primary and secondary were analyzed by evaluating the difference in the logarithm of the reconstitution ratios in the respective recipients. The reason for this strategy was described previously (25). Briefly, the change in the reconstitution ratio between primary and secondary recipients is a better measure of a shift in reconstitution capacity upon serial transplantation than a change in the percentage contribution of CD45.2⁺ cells between primary and secondary recipients, as the latter method does not weigh the same percent change at high or low contribution level equally (i.e., a 5% change from 20 to 25% is a change in reconstitution ratio from 0.25 to 0.33, whereas a 5% change from 90 to 95% represents a much larger change in reconstitution ratio from 9 to 19). A more accurate measure is the difference of CD45.2⁺(donor)/CD45.1⁺CD45.2⁺(competitor) ratios in primary recipients and secondary recipients. With ratiometric data, the difference between log(CD45.2⁺/CD45.1⁺CD45.2⁺) can be used which added the advantage of log transformation where a ratio smaller than one will give a negative value, and negative ratios will extend over the same numerical ranges as positive ones (e.g., a ratio of 0.01 gives a log ratio of 2, a ratio of 100 gives a log ratio of 2). Thus, serial repopulation data are presented as the difference between log(CD45.2⁺/CD45.1⁺CD45.2⁺) of primary transplantation recipients and the log(CD45.2⁺/CD45.1⁺CD45.2⁺) of secondary transplantation recipients, and will be referred to as Δ log ratio.

Statistical analysis

Student’s 2-tailed t-test for paired samples was used in the calculation of all P values unless indicated otherwise. Data represent mean ±SD.

A locus on chr.4 regulates the stimulatory effect of TGF-β2 on HSPC growth, number and cycling, and 5-fluorouracil lethality

The isoform-specific proliferative effect of TGF-β2 on HSPCs is subject to genetically determined variation in inbred mouse strains, and a suggestive QTL for this trait was mapped to the telomeric region of chr.4 (25). To confirm the location of this QTL, we constructed congenic mice by introgressing 20cM of the telomeric region of chr.4 from DBA/2 into C57BL/6 mice. Although there was wide variation in TGF-β2 responsiveness of HSPCs among various inbred mouse strains, the TGF-β2 dose response was similar in DBA/2 and C57BL/6 mice (26). Constructing congenic mice from these strains may therefore seem ill justified at first sight. However, this QTL was mapped using BXD recombinant inbred mice, which are derived from the allele pool of C57BL/6 and DBA/2 mice. In BXD recombinant inbred strains, where the genome is made up of a patchwork of segments that are homozygously inherited from either progenitor strain, significant variation in TGF-β2 responsiveness was observed (25,26). This phenomenon is explained by the fact that if multiple loci are involved in determining a phenotype, some BXD strains can accumulate predominantly ‘high’ or ‘low’ alleles for this trait and acquire more extreme phenotypes than either of the progenitor strains. In other BXD strains and in the progenitor C57BL/6 and DBA/2 strains, the phenotypic effects of ‘high’ and ‘low’ alleles balance each other. These observations therefore proved that TGF-β2 responsiveness is a quantitative trait, and justified generating congenic mice. As only one or a limited number of QTL are transferred from the DBA/2 to the C57BL/6 background in the congenic mice, we anticipated that if the QTL mapping was correct, the phenotype of the congenic mice and the background C57BL/6 strain would differ. Using C57BL/6 as the acceptor strain also facilitates analysis of in vivo transplantation assays by taking advantage of allelic variation at the CD45 locus to track donor and host contribution to hematopoiesis.

Construction of congenic mice was accomplished by repeated backcrossing of DBA/2 onto C57BL/6 mice and selecting offspring where the telomeric region of chr.4 was heterozygous (see Material and Methods). These congenic mice will be referred to as B6.D2-chr.4 mice hereafter. The transition between the C57BL/6 and DBA/2-derived genome occurred between D4Mit37 (57 cM) and D4Mit251 (66cM) or between chromosome base pairs 04-114064127 and 04-135804867 (Figure 1A). Any difference in the effect of TGF-β2 on the proliferation of HSPCs between B6.D2-chr.4 and C57BL/6 mice must be caused by one or more alleles in the introgressed region of chr.4. If this is the case and if the antiproliferative effect of higher concentration of TGF-β2 is similar in B6.D2-chr.4 and C57BL/6 mice, then any phenotype shared by B6.D2-chr.4 mice and Tgfb2^+/- mice can with a very large likelihood be assigned to the stimulatory effect of low concentrations of TGF-β2 on HSPC proliferation. To test this, we cultured Lineage^negSca⁺c-kit⁺ (LSK) cells (Figure S1) from C57BL/6 and B6.D2-chr.4 mice in serum-containing media for 5 days in the presence of flt3 ligand (flt3L), kit ligand (KL), thrombopoietin (TPO) and increasing concentrations of TGF-β2. While the dose response of exogenously added TGF-β2 on the proliferation of LSK cells from parental C57BL/6 mice was biphasic with a stimulatory effect at low concentrations, this stimulatory effect was absent in B6.D2-chr.4 LSK cells (Figure 1B). Importantly, the inhibitory effect of higher concentrations of TGF-β2 (Figure 1B) and the dose response of TGF-β1 (data not shown) on the proliferation of LSK cells were similar in B6.D2-chr.4 and C57BL/6 mice. We have shown that TGF-β2 specifically enhances flt3 signaling in HSPCs (submitted for publication). Similar to Tgfb2^+/- mice, in the absence of any exogenously added TGF-β2, the response of LSK cells to increasing concentrations of flt3L was approximately 50% lower in B6.D2-chr.4 than in C57BL/6 mice (Figure 1C). In contrast, the response to IL-3 was similar in both mice (Figure 1D). Together, these data indicate that only the predominantly, though not exclusively cell autonomous enhancing effect of TGF-β2 on flt3L responsiveness of HSPCs and not the generic antiproliferative effect of TGF-β2 are regulated by a QTL on chr.4. These results in fact suggest that most genetic variation in the stimulatory effect of TGF-β2 is regulated by this QTL. As our observation provides unequivocal confirmation for the location of this QTL, we call this QTL, tb2r1 (for TGF-β2 responsiveness 1).

As Tgfb2^+/- mice display a hematopoietic defect, we further analyzed the hematopoietic phenotype of the B6.D2-chr.4 mice. We had previously mapped a QTL regulating the frequency of LSK cells to the telomeric region of chr.4 (20). Consistent with these mapping data, the frequency of LSK cells was significantly lower in B6.D2-chr.4 than in C57BL/6 mice (Figure 2A, Table I), confirming that the same region on chr.4 also contains a QTL regulating LSK frequency. This finding cannot be explained by intrinsic variation in c-kit expression, as the overall c-kit expression level in the bone marrow (BM) was similar in C57BL/6 mice and congenics. The frequency of a subpopulation of LSK cells containing the more primitive stem cells, the CD34^-flt3^-LSK population, was also decreased in B6.D2-chr.4 mice BM (Figure 2B). BM cellularity, peripheral blood counts and spleen weight were similar in B6.D2-chr.4 and C57BL/6 mice (Table I). Tgfb2^+/- mice have a more slowly cycling HSPC compartment than wt littermates (25). Therefore, we examined cell cycle activity of LSK cells in B6.D2-chr.4 mice. Similar to Tgfb2^+/- mice, fewer LSK cells in B6.D2-chr.4 mice than in control C57BL/6 mice were in cell cycle, as measured by staining fixed BM cells with DNA-binding dye Hoechst 33342 (Figure 2C), while the fraction of S/G2/M phase cells in the total BM cell population did not differ between B6.D2-chr.4 and C57BL/6 mice (data not shown). We hypothesized that a slower cycling HSPC compartment may render Tgfb2^+/- and B6.D2-chr.4 mice more resistant to the effects of cell cycle specific cytotoxic agents, such as 5-FU. The lethal effects of this drug are due to hematopoietic failure as mice are rescued from high doses of 5-FU by BM transplantation (41). Consistent with this notion, the survival of both Tgfb2^+/- and B6.D2-chr.4 mice after an intravenous (IV) injection of highly lethal dose of 5-FU (500mg/kg, IV) was better than that of control mice (P < 0.0001 and P=0.0001, respectively; Figure 3A). However, recovery from a sublethal dose of 5-FU was similar in B6.D2-chr.4 and C57BL/6 mice (Figure 3B), and in wt and Tgfb2^+/- mice (Figures 3C). We conclude that B6.D2-chr.4 mice and Tgfb2^+/- mice share several hematopoietic phenotypes, including the frequency and cycling activity of HSPCs in vivo, and their response to flt3L in vitro. Furthermore, both mouse strains show enhanced resistance to the lethal effects of 5-FU. These observations suggest that these phenotypes of B6.D2-chr.4 mice are likely due to regulation of TGF-β2 responsiveness by tb2r1.

Decreased serial reconstitution capacity is not shared between B6.D2-chr.4 and Tgfb2^+/- mice

HSC from Tgfb2^+/- mice show a defect in serial repopulation capacity compared to HSC from wt mice (25). Therefore, we tested the serial competitive repopulation capacity of LSK cells from B6.D2-chr.4 mice. 500 purified CD45.2⁺ LSK cells from C57BL/6 or B6.D2-chr.4 mice were competed with 2·10⁵ CD45.1⁺CD45.2⁺ C57BL/6 bone marrow cells in lethally irradiated CD45.1⁺ C57BL/6 recipients. In the primary transplantation, LSK cells from B6.D2-chr.4 mice competed equally well compared to LSK cells from C57BL/6 mice as determined by the donor CD45.2⁺ contribution to peripheral blood mononuclear cells after 12 weeks (P > 0.1; Figure 4A). This was not surprising, as the content of most primitive CD34^-flt3^- stem cell fraction was the same in the LSK population of C57BL/6 and B6.D2-chr.4 mice (Figure S2). For serial transplantation, we injected 2·10⁶ bone marrow cells from competitively repopulated primary recipients into lethally irradiated secondary CD45.1⁺ recipients, 4 months after the primary transplantation. 12 weeks later, the contribution of CD45.2⁺ and CD45.1⁺CD45.2⁺ cells to hematopoiesis in peripheral blood was assessed in secondary recipients. The data are presented as the logarithm of the ratio of the two donor populations determined by the expression CD45 alleles. The difference in the log reconstitution ratios (Δ log ratio; see Materials and Methods) between primary and secondary recipients was significantly higher for LSK cells from B6.D2-chr.4 mice than for LSK cells from C57BL/6 mice (Figure 4B). These data indicate that HSC from B6.D2-chr.4 mice performed significantly better than HSC from C57BL/6 mice in serial transplantation assays. This phenotype of B6.D2-chr.4 mice is not shared with Tgfb2^+/- mice, which show a subtle defect in competitive repopulation capacity that increases with serial transplantation.