Aneuploidy impairs hematopoietic stem cell fitness and is selected against in regenerating tissues in vivo

Abstract

Aneuploidy, an imbalanced karyotype, is a widely observed feature of cancer cells that has long been hypothesized to promote tumorigenesis. Here we evaluate the fitness of cells with constitutional trisomy or chromosomal instability (CIN) in vivo using hematopoietic reconstitution experiments. We did not observe cancer but instead found that aneuploid hematopoietic stem cells (HSCs) exhibit decreased fitness. This reduced fitness is due at least in part to the decreased proliferative potential of aneuploid hematopoietic cells. Analyses of mice with CIN caused by a hypomorphic mutation in the gene Bub1b further support the finding that aneuploidy impairs cell proliferation in vivo. Whereas nonregenerating adult tissues are highly aneuploid in these mice, HSCs and other regenerative adult tissues are largely euploid. These findings indicate that, in vivo, mechanisms exist to select against aneuploid cells.

Keywords: aneuploidy; chromosomal instability; hematopoiesis; population flush hypothesis; single-cell sequencing.

Figure 1.

Aneuploidy decreases HSC competitive fitness in vivo. (A) The percentage of HSCs (CD150⁺ CD48⁻ Sca-1⁺ lin⁻ cells) found in trisomy 16 and trisomy 19 fetal livers was quantified by flow cytometry. Data are shown as mean ± standard deviation. (B) Schematic of competitive reconstitution experiments. (C–E) CD45.2 fetal liver cells from wild-type or aneuploid E14.5 littermates were coinjected into a lethally irradiated CD45.1 recipient with an equal number of fetal liver cells from a CD45.1 common wild-type donor of the same age derived from a separate mating. Peripheral blood was sampled at the indicated times. The percentage of the white blood cell population contributed by each donor was quantified by flow cytometry with isoform-specific antibodies against CD45.1 and CD45.2 for recipients of common wild-type cells and trisomy 16 fetal liver cells (C, left graph) (n = 17), trisomy 19 fetal liver cells (D, left graph) (n = 10), and Bub1b^H/H fetal liver cells (E, left graph) (n = 10). (C–E, right graphs) The contribution of wild-type littermates when competed to the common wild type for all aneuploidies was quantified at the same time in C (right graph) (n = 20), D (right graph) (n = 8), and E (right graph) (n = 6). Data are represented as mean ± standard deviation for each time point. (F) Ratios of the average percentage of the peripheral blood reconstituted by the aneuploid fetal liver cells to the average percentage of the peripheral blood reconstituted by wild-type littermate fetal liver cells at the indicated times are shown. (G) Single-cell sequencing of white blood cells from a mouse competitively reconstituted with CD45.2 Bub1b^H/H and CD45.1 euploid FL-HSCs at 16 wk after transplantation (Fig. 1E) revealed that seven of 18 CD45.2 Bub1b^H/H cells analyzed (∼39%) were aneuploid. Karyotypes of the seven aneuploid cells are shown with chromosome gains in red, chromosome losses in blue, and euploidy in black. Segmentation plots of all sequenced cells are shown in Supplemental Figure S7A.

Figure 2.

Proliferation but not homing ability is reduced in trisomy 16 and trisomy 19 reconstitutions. (A) DiI-labeled fetal liver cells were injected into irradiated recipient mice. The percentage of DiI-positive cells in the bone marrow of recipient mice was measured 24 h after injection. Data are shown as mean ± standard deviation. (B) Representative images of sections of spleens isolated from mice transferred with wild-type or trisomy 16 fetal liver cells 8 d after reconstitution. Bar, 1 mm. (C) Quantification of colony-forming unit spleen (CFU-S) colonies from spleen sections of recipient mice of trisomy 16 or wild-type littermate fetal liver cells 8 d after injection and trisomy 19, Bub1b^H/H, or wild-type littermate fetal liver cells 7 d after injection. The bar represents the mean value for each condition. (D) Quantification of the mean size of each colony in C as determined by the percent of total spleen area. Measurements from all individuals for each condition were pooled. The bar represents the mean value for each population. Populations were compared by Student's t-test. (*) P < 0.05. (E) Trisomy 16 or Bub1b^H/H fetal liver cells or cells from their wild-type littermates were injected into lethally irradiated recipients. Mice were injected with EdU 6 d later, and the level of EdU incorporation in CD45.2-positive donor-derived bone marrow cells was evaluated by flow cytometry 24 h later. The bar represents the mean value for each population. (*) P < 0.05 by Student's t-test.

Figure 3.

Trisomy 16 causes peripheral blood defects and decreases HSC reconstitution potential. (A) Schematic of serial reconstitution experiments. (B–D) For primary reconstitutions, fetal liver cells from a CD45.2 trisomic embryo or its wild-type littermate were injected into lethally irradiated CD45.1 recipients. Bone marrow cells from primary recipients were injected into secondary CD45.1 recipients to assess serial reconstitution capacity. Peripheral blood of primary recipients of trisomic fetal liver cells or their wild-type littermates was sampled at the indicated times. The percentage of CD45.2-positive cells in the blood of trisomy 16 or wild-type primary recipients was determined by flow cytometry (B), and white blood cell count (C) and hematocrit (D) were determined by complete blood cell counts. The bar represents the mean, and asterisks indicate that the trisomy 16 values are significantly different from the values of wild-type littermates at the indicated time by Student's t-test. P < 0.05. (E) Survival of recipients of trisomy 16 or wild-type fetal liver cells after transfer. (F) Survival of secondary recipients of trisomy 16 or wild-type bone marrow cells from primary recipients.

Figure 4.

Trisomy 19 HSC reconstitution potential is reduced upon serial reconstitution. (A–E) For trisomy 19 or wild-type primary recipients, the percentage of CD45.2-positive cells in the peripheral blood (A), white blood cell count (B), and hematocrit (C) was determined. The percentage of CD45.2-positive cells in the blood of trisomy 19 or wild-type secondary recipients (D), tertiary recipients (E), and quaternary recipients (F) was also evaluated. The bar represents the mean value for each condition, and asterisks indicate that the trisomy 19 values are significantly different from the values of wild-type littermates at the indicated time by Student's t-test. P < 0.05.

Figure 5.

Bub1b^H/H BM-HSCs undergo stem cell exhaustion upon serial reconstitution. (A) Schematic of serial reconstitution experiments. (B–D) CD45.2 adult Bub1b^H/H bone marrow cells were serially transplanted into lethally irradiated CD45.1 recipients. The percentage of the peripheral blood that is CD45.2-positive was determined in primary (B), secondary (C), and tertiary (D) recipients at the indicated times. The bar represents the mean value for each condition.

Figure 6.

Bub1b^H/H adult regenerative tissues show evidence of selection against aneuploid cells. (A) The percent aneuploidy over time during hematopoietic reconstitution with Bub1b^H/H HSCs was determined by single-cell sequencing of peripheral blood cells derived from primary recipient mice of Bub1b^H/H bone marrow (open circles) or Bub1b^H/H fetal liver (closed circles) cells at the indicated times after transfer. Peripheral blood cells from a mouse reconstituted with Bub1b^H/H FL-HSCs 16 wk after competitive reconstitution (shown as the percent of total peripheral blood; gray triangle) (Fig. 1G) and from a Bub1b^H/H secondary bone marrow recipient mouse (black triangle) (Fig. 5C) were also sequenced. Baseline aneuploidy was determined by single-cell sequencing of FL-HSCs and BM-HSCs. Segmentation plots of all sequenced cells are shown in Supplemental Figures S7 and S8. (B) Percentage of euploid cells found in different adult Bub1b^H/H cell types. BM-HSCs, peripheral blood cells (PB), keratinocytes, and intestines (in blue) are from ∼4-mo-old Bub1b^H/H mice. Data from hepatocytes and brains (in red) are from Knouse et al. (2014) and ∼6-mo-old Bub1b^H/H mice. Segmentation plots of all newly sequenced cells are shown in Supplemental Figure S9. (C) The number of aneuploid chromosomes per cell in all adult cells analyzed in A and B. The bar represents the mean value for each population. t-tests were performed for significance. (***) P < 0.001. (D) Summary of chromosome gain and loss events observed in each cell from transplantation peripheral blood cells and adult nonproliferative neurons and hepatocytes. “Multiple gains” and “multiple losses” describe cells that have gained or lost two or more chromosomes. “Both loss and gain” describes cells that have gained at least one chromosome and lost at least one chromosome. (E,F) Frequency of chromosome gain (red) or chromosome loss (green) by chromosome observed in all peripheral blood cells after transplantation (E) and in all adult neurons and hepatocytes (F).