Aldehyde dehydrogenase 1a1 is dispensable for stem cell function in the mouse hematopoietic and nervous systems

Abstract

High levels of aldehyde dehydrogenase (ALDH) activity have been proposed to be a common feature of stem cells. Adult hematopoietic, neural, and cancer stem cells have all been reported to have high ALDH activity, detected using Aldefluor, a fluorogenic substrate for ALDH. This activity has been attributed to Aldh1a1, an enzyme that is expressed at high levels in stem cells and that has been suggested to regulate stem cell function. Nonetheless, Aldh1a1 function in stem cells has never been tested genetically. We observed that Aldh1a1 was preferentially expressed in mouse hematopoietic stem cells (HSCs) and expression increased with age. Hematopoietic cells from Aldh1a1-deficient mice exhibited increased sensitivity to cyclophosphamide in a non-cell-autonomous manner, consistent with its role in cyclophosphamide metabolism in the liver. However, Aldh1a1 deficiency did not affect hematopoiesis, HSC function, or the capacity to reconstitute irradiated recipients in young or old adult mice. Aldh1a1 deficiency also did not affect Aldefluor staining of hematopoietic cells. Finally, Aldh1a1 deficiency did not affect the function of stem cells from the adult central or peripheral nervous systems. Aldh1a1 is not a critical regulator of adult stem cell function or Aldefluor staining in mice.

Figure 1

Aldh1a1 expression increases with age in HSCs, but Aldh1a1 deficiency is not compensated by increased transcription of other ALDHS. (A) Gene expression profiles for ALDHs in Thy-1.1^lowSca-1⁺Lineage⁻c-kit⁺ HSCs, Thy-1.1^lowSca-1⁺Mac-1^lowCD4^lowB220⁻ non–self-renewing multipotent progenitors and CD45⁺ bone marrow cells from young adult mice. These data were extracted from genome-wide data published as supplementary material from an earlier study. (B) Aldh1a1 transcript levels were compared by qPCR in CD150⁺CD48⁻CD41⁻Sca1⁺c-Kit⁺ HSCs or whole bone marrow cells independently isolated from three 2.5-month-old mice and three 22-month-old mice. cDNA content was normalized between samples based on ß-actin expression (data not shown). Aldh1a1 transcript levels are expressed in terms of fold change relative to old bone marrow (set to 1). Aldh1a1 was expressed at significantly (P < .01) higher levels in old HSCs compared with old bone marrow or young HSCs. (C) A schematic representation of Aldh1a1 showing the position of exons (numbered black boxes), 4 sets of qPCR primer binding sites (arrows), and the neomycin resistance cassette in the targeted allele. Bone marrow cells (D) or CD150⁺CD48⁻CD41⁻Sca1⁺c-Kit⁺ HSCs (E) from 3 littermate pairs aged 2.5 to 6 months were tested for expression of Aldh1a1 with the above primers, and all other Aldh family members by quantitative RT-PCR. cDNA content was normalized between samples based on ß-actin expression (data not shown) and data are expressed as fold change relative to the littermate control. #P < .1; *P < .05. Error bars represent SD. The mutant Aldh1a1 transcript was present at greatly reduced levels, and we did not detect any Aldh transcript with a compensatory increase in expression.

Figure 2

Aldefluor staining is not determined by Aldh1a1 function, and HSCs cannot be distinguished from bone marrow cells based on Aldefluor staining. (A) whole bone marrow (left) and CD150⁺CD48⁻CD41⁻Sca1⁺c-kit⁺ HSCs (right; cells highlighted in orange) were (B) analyzed for Aldefluor fluorescence in the presence and absence of the ALDH inhibitor DEAB. Aldh1a1 deficiency had no effect on Aldefluor staining. (C) Mean Aldefluor fluorescence intensity minus background fluorescence in the presence of DEAB (at least 2 independent experiments per age involving a total of 3 to 5 mice per treatment). Error bars represent standard deviation. (D) Representative plots of whole bone marrow cells (black dots) and HSCs (orange dots) from (A) show SSC versus Aldefluor staining. Note that mouse HSCs did not exhibit high levels of Aldefluor staining relative to whole bone marrow. Similar results were obtained using c-kit⁺Flk-2⁻Lineage⁻Sca1⁺ HSCs (Figure S1).

Figure 3

Aldh1a1 is not required for normal hematopoiesis. (A) Cellularity (left) and HSC frequency (right) were determined in young (2-3 months) and old (21-27 months) adult Aldh1a1^+/− and Aldh1a1^−/− mice (n = 6-14 mice per treatment in at least 6 independent experiments). (B) The frequency of colony-forming progenitors in bone marrow (top row) and spleen (second row) of young (left column) and old adult (right column) Aldh1a1^+/− and Aldh1a1^−/− mice (n = 6-7 mice per treatment in 5 independent experiments). (C) The frequency of various subsets of B (B220⁺), T (CD3⁺), erythroid (Ter119⁺), and myeloid (Mac-1⁺Gr-1⁺) lineage cells in the bone marrow (top row) and spleen (bottom row) from young (left column) and old (right column) Aldh1a1^+/− and Aldh1a1^−/− mice (n = 5 mice per treatment in 5 independent experiments). No significant differences were observed between Aldh1a1^−/− and control mice. Error bars represent SD.

Figure 4

Aldh1a1 is not required for HSC maintenance or function. Irradiated CD45.1⁺ recipient mice were competitively reconstituted with 3 × 10⁵ whole bone marrow cells from 2-month-old, 13-month-old, or 26-month-old Aldh1a1^+/− (black lines) or Aldh1a1^−/− (red lines) CD45.2 donor cells along with a radioprotective dose of 3 × 10⁵ CD45.1⁺ bone marrow cells. Similar levels of long-term multilineage reconstitution were observed at all ages from Aldh1a1^+/− and Aldh1a1^−/− cells. At least 3 independent experiments were performed at each age, from which one representative experiment is shown. Error bars represent SD.

Figure 5

Aldh1a1 is non–cell-autonomously required to reduce the toxicity of cyclophosphamide. (A) Aldh1a1^−/− and littermate control mice were treated with 250 μg G-CSF/kg body mass/day for 4 days to induce HSC mobilization. Aldh1a1 deficiency did not significantly affect bone marrow or spleen cellularity or HSC content (n = 2-5 mice per treatment in 2 independent experiments). (B) Littermates were treated with 200 mg/kg cyclophosphamide (CY) followed by 4 days of 250 μg/kg G-CSF. Aldh1a1^−/− mice had significantly (*P < .05) fewer HSCs in their spleen compared with littermate controls but did not exhibit a significant difference in overall bone marrow or spleen cellularity (n = 3-6 mice per treatment in 2 independent experiments). (C) Donor-cell chimerism in wild-type mice reconstituted with Aldh1a1^−/− (red lines) or littermate control (black lines) bone marrow cells that were treated weekly with 200 mg/kg cyclophosphamide. Because cyclophosphamide did not generally reduce the levels of Aldh1a1^−/− donor cells compared with control donor cells these results indicate that the toxic effect of cyclophosphamide on HSCs is primarily non–cell-autonomously attenuated by Aldh1a1. (D) Survival of Aldh1a1^−/− and littermate control mice treated weekly with 150 mg/kg 5-fluorouracil. Data represent the combined results from 2 independent experiments (n = 20-22 mice per treatment). Error bars represent SD.

Figure 6

Aldh1a1 is not required for central nervous system stem cell maintenance or function. (A) Lateral ventricle SVZ cells from young (2-3 months) and old (24-27 months) Aldh1a1^+/− () and Aldh1a1^−/− () mice were cultured. Aldh1a1 deficiency did not affect the frequency of neurospheres (A), neurosphere diameter after 8 to 10 days in culture (B), neurosphere differentiation (C,D), or the capacity of primary neurospheres to form large (> 100 μm; these are almost always multipotent) secondary neurospheres upon subcloning (E). Colonies were assessed for the presence of neurons (N), astrocytes (A) and oligodendrocytes (O). The size of secondary neurospheres was also not affected (F). All data represent a total of 3 to 7 mice per treatment, in at least 3 independent experiments except for F which represents 2 independent experiments with 1 mouse per experiment. (G, H) Aldh1a1 deficiency also did not affect the rate of neurogenesis (frequency of BrdU + NeuN + neurons) in the olfactory bulb of 18-month-old adult mice in vivo (n = 2 mice with 25 sections per mouse). Error bars represent SD.

Figure 7

Aldh1a1 is not required for peripheral nervous system stem cell maintenance or function. Outer muscle/plexus layer cells from the guts of young (2-3 months) and old (24-27 months) Aldh1a1^+/− () and Aldh1a1^−/− () mice were cultured. Aldh1a1 deficiency did not affect the frequency of neurospheres (A), neurosphere diameter after 9 to 11 days in culture (B), neurosphere differentiation (C,D), or the capacity of primary neurospheres to form multipotent secondary neurospheres upon subcloning (E). Colonies were assessed for the presence of neurons (N), glia (G), and myofibroblasts (M). All data represent a total of 3 to 8 mice per treatment, in at least 3 independent experiments. Error bars represent SD.