PU.1 regulates the commitment of adult hematopoietic progenitors and restricts granulopoiesis

Abstract

Although the transcription factor PU.1 is essential for fetal lymphomyelopoiesis, we unexpectedly found that elimination of the gene in adult mice allowed disturbed hematopoiesis, dominated by granulocyte production. Impaired production of lymphocytes was evident in PU.1-deficient bone marrow (BM), but myelocytes and clonogenic granulocytic progenitors that are responsive to granulocyte colony-stimulating factor or interleukin-3 increased dramatically. No identifiable common lymphoid or myeloid progenitor populations were discernable by flow cytometry; however, clonogenic assays suggested an overall increased frequency of blast colony-forming cells and BM chimeras revealed existence of long-term self-renewing PU.1-deficient cells that required PU.1 for lymphoid, but not granulocyte, generation. PU.1 deletion in granulocyte-macrophage progenitors, but not in common myeloid progenitors, resulted in excess granulocyte production; this suggested specific roles of PU.1 at different stages of myeloid development. These findings emphasize the distinct nature of adult hematopoiesis and reveal that PU.1 regulates the specification of the multipotent lymphoid and myeloid compartments and restrains, rather than promotes, granulopoiesis.

Figure 1.

Conditional deletion of PU.1 in adult hematopoiesis. (A) The genomic locus of PU.1. Exons are represented as boxes; introns are represented as black lines; coding regions are in gray; nontranslated regions are in white; arrows indicate the direction of translation. The alleles derived from the integration of the targeting vector and subsequent manipulations are shown. pA, polyadenylation signal sequence; circles, frt sites; triangles, loxP sites; stop, stop codons in all reading frames; splice acc., splice acceptor. Targeted PU.1 ^gfp alleles can be converted by flp-mediated deletion to PU.1 ^fl or Cre deletion to PU.1 ^Δ. The position and direction of the genotyping primers (a–e) and Eco RI sites are indicated. (B) Southern blot analysis of ES cell Eco RI-digested DNA showing the wild-type (6.6 kb) and targeted (5.3 kb) alleles (C) PCR genotyping of tail DNA using the primer set a/b/c showing the correct amplification of the wild-type, PU.1 ^gfp, and PU.1 ^fl alleles. (D) Flow cytometric analysis of E18.5 fetal liver from PU.1 ^{Δ gfp/ +} intercrosses showing loss of myeloid cells in the absence of PU.1 and the GFP fluorescence in progenitors (c-kit⁺). (E) PCR showing the relative polyIC-inducible deletion of the PU.1 ^fl allele in vivo. Mice were injected with polyIC on days 0 and 3 and analyzed on day 10. DNA was amplified with the primer combination b/c/d. (F) Western blot analysis of PU.1 of mice of the indicated genotypes. Equal loading was confirmed using β-actin. PU.1 is not expressed in most thymocytes and demonstrates nonspecific antibody binding. (G) GFP fluorescence in adult BM. PU.1 ^{Δ gfp/ Δ gfp} (gfp/gfp MxCre ⁺) and control (gfp/+ MxCre ⁻) mice are shown at 2 wk after polyIC injection.

Figure 2.

PU.1 inactivation results in relative expansion of granulopoiesis. Representative analysis of the (A) myeloid, (B) B lymphoid, and (C) erythroid lineages from the BM of PU.1 ^{Δ / Δ} (fl/fl MxCre ⁺) and control (fl/fl MxCre ^-) mice. Mice were injected on days 0 and 3 with polyIC and analyzed after 2 or 8 wk. (D) PCR analysis showing the relative PU.1 deletion of total BM and sorted cells of the indicated lineages 2 wk after polyIC treatment. B cells were identified as CD19⁺B220⁺, granulocytes as Gr1⁺. PU.1 ^fl/Δ tail DNA was used as a control.

Figure 3.

Perturbed granulocytic differentiation in the absence of PU.1. (A) Histologic section of day 21 BM from mice of the indicated genotype. Insets show cytocentrifuge preparations of corresponding BM. (B) Histologic section of relatively mild and severe granulocytic infiltrations in the spleen at day 21. (C) Severe PU.1 ^{Δ / Δ} (fl/fl MxCre ⁺) mice display splenomegaly and enhanced percentage of Gr-1⁺Mac-1⁻ cells by flow cytometry. PCR showing efficient deletion of PU.1 ^fl and PU.1^gfp alleles in 2-wk spleen of a PU.1 ^{Δ / Δ gfp} (fl/gfp MxCre ⁺) mouse. Total, total splenocytes; Gr, Gr-1⁺ cells; B, CD19⁺B220⁺ B cells; T, CD3⁺ T cells; Er, CD71⁺Ter119⁺ nucleated erythrocytes. B cells also were gated on GFP^high (B^Ghi) or GFP^low (B^Glo). The fl and gfp alleles were amplified with the primer combination b/c/d and e/c/d, respectively. (D) Control (fl/fl MxCre ^-) and PU.1 ^{Δ / Δ} (fl/fl MxCre ⁺) granulocytes were grown in G-CSF for 7 d before being analyzed by flow cytometry and morphology. PCR analysis showed the complete PU.1 deletion of cultured granulocytes. (E) BM Gr-1^hi cells were sorted and analyzed by RT-PCR for the expression of granulocytic enzymes, myeloperoxidase (MPO), LF, and gp91.

Figure 4.

PU.1 is essential for multipotent lymphoid and myeloid progenitors. (A) BM lin⁻ cells from 2-wk PU.1 ^{Δ / Δ} (fl/fl MxCre ⁺) and control (fl/fl MxCre ⁻) polyIC-treated mice were assayed for HSC and lymphoid and myeloid progenitors. For GFP analysis, PU.1 ^{Δ gfp/ Δ} (PU.1 ^gfp/flMxCre⁺) or control (PU.1 ^gfp/flMxCre^-) mice were treated as above. (B) Cytocentrifuge preparation and cell cycle analysis of lin⁻ cells. Inset: PCR showing complete deletion of exon-5 in lin⁻ cells. Range of cycling cells (S+G₂+M) observed from two experiments is indicated. (C) RT-PCR for key hematopoietic regulators and PU.1 target genes in lin⁻ and total BM. HPRT was used as a control for cDNA input. −RT, no reverse transcriptase was used in the reaction. Samples were taken at four cycle intervals. Ikaros isoforms 1 and 2 are indicated.

Figure 5.

PU.1 regulates granulocyte production from the GMP. BM lin⁻ cells from 36-h post-polyIC treatment PU.1 ^{Δ / Δ} (fl/fl MxCre ⁺) and control (fl/fl MxCre ⁻) mice were assayed for erythromyeloid progenitors. Sorted CMP and GMP using the indicated gates were subjected to PCR for PU.1 exon-5 deletion. Cytocentrifuge preparations were stained with May-Grunwald-Giemsa. Sorted cells were cultured in the indicated cytokines for 7 d, stained, and counted. 100 cells were plated in quadruplicate for each stimulus. Numbers represent the mean ± average deviation from three independent experiments.

Figure 6.

Competitive BM reconstitution. (A) PU.1 ^fl/fl MxCre⁺ BM was mixed with Ly5.1 BM (2:1 ratio) and used to reconstitute lethally irradiated Ly5.1 mice. The degree of reconstitution was assessed from peripheral blood after 6 wk, followed by the induction of PU.1 deletion by polyIC. Flow cytometric analysis of spleen from chimeric mice at 24 wk. Ly5.1⁻ cells were almost entirely PU.1-deficient granulocytes (Gr-1⁺Mac-1⁻), whereas Ly5.1⁺ competitors were Gr-1⁺Mac-1⁺ (macrophages/granulocytes) and CD19⁺Ly5.2⁻ (B cells). c-kit⁺CD71⁺ erythroid progenitors were present in both populations. (B) BM, thymus, spleen, and blood cells and sorted Ly5.1⁻ and Ly5.1⁺ BM cells were subjected to PCR for PU.1 exon-5 deletion. PU.1 ^fl/Δ, PU.1 ^fl/+, and PU.1 ^{+ / Δ} tail DNA was used as a control. (+) Competitor Ly5.1⁺ DNA. Note that the synthesis of + (WT) allele PCR product is more efficient that that of Δ allele and cannot be used for relative quantification of host/donor cellular contribution to chimerism. PU.1 ^{Δ / Δ} was amplified from the BM, whereas no PU.1-deleted cells were found in the thymus, spleen, or blood. (C) The percentage of PU.1 ^{Δ / Δ} cells in hematopoietic organs was determined at 6, 12, and 24 wk after polyIC injection. Total PU.1 ^{Δ / Δ} cells (Ly5.1⁻) are shown relative to Ly5.1⁺ cells. The numbers have been normalized for the predeletion reconstitution frequency. Because cells lose CD45 expression in some lineages, mutant cells were defined as Ter119⁻Ly5.1⁻.