DYRK1A controls the transition from proliferation to quiescence during lymphoid development by destabilizing Cyclin D3

Abstract

Pre-B and pre-T lymphocytes must orchestrate a transition from a highly proliferative state to a quiescent one during development. Cyclin D3 is essential for these cells' proliferation, but little is known about its posttranslational regulation at this stage. Here, we show that the dual specificity tyrosine-regulated kinase 1A (DYRK1A) restrains Cyclin D3 protein levels by phosphorylating T283 to induce its degradation. Loss of DYRK1A activity, via genetic inactivation or pharmacologic inhibition in mice, caused accumulation of Cyclin D3 protein, incomplete repression of E2F-mediated gene transcription, and failure to properly couple cell cycle exit with differentiation. Expression of a nonphosphorylatable Cyclin D3 T283A mutant recapitulated these defects, whereas inhibition of Cyclin D:CDK4/6 mitigated the effects of DYRK1A inhibition or loss. These data uncover a previously unknown role for DYRK1A in lymphopoiesis, and demonstrate how Cyclin D3 protein stability is negatively regulated during exit from the proliferative phases of B and T cell development.

Figure 1.

Conditional inactivation of the Dyrk1a gene. (A) Exons 5 and 6 were floxed in the targeted allele and excised in the conditional knockout (CKO) allele. (B) PCR from thymocyte genomic DNA was performed 2 wk after pI:pC treatment using the indicated primers in A (i and ii) and assessing the presence or loss of the targeted allele in Dyrk1a^f/f Mx1-Cre⁻, Dyrk1a^f/w Mx1-Cre⁺, and Dyrk1a^f/f Mx1-Cre1⁺,mice. (C) Dyrk1a mRNA expression measured by qRT-PCR using primers within the excised gene segment in bone marrow and thymus for the indicated mice. (D) Western blot shows DYRK1A protein expression in bone marrow and thymus after loss of 0, 1, or 2 Dyrk1a alleles. Densitometry values were normalized to HSC₇₀. Loss of DYRK1A expression was verified for all subsequent experiments by qPCR, Western blot, or both. Data are derived from 1 litter and are representative of over 20 cohorts that were analyzed by RT-PCR and/or Western blot.

Figure 2.

Loss of Dyrk1a adversely affects B cell development. (A) Mean bone marrow cellularity was assessed in Dyrk1a^f/f Mx1-Cre⁻ (Control) and Dyrk1a^f/f Mx1-Cre⁺ (CKO) mice 2 and 4 wk after pI:pC treatment; n = 6 mice per genotype, pooled from 3 independent cohorts of 2 mice per genotype. (B) Representative flow cytometry analysis of bone marrow for total pre–B (IgM⁻ gate, left) and large/small pre–B (pre–B gate, right) cells 4 wk after pI:pC treatment is shown. Numbers indicate percentages in each gate; mean percentages of each population among total bone marrow cells were quantified in (C); n = 6 mice per genotype, pooled from 3 independent cohorts of 2 mice per genotype. (D) Representative flow cytometry analysis of pre–B cell surface CD25 expression in bone marrow from Control and CKO mice 2 wk after pI:pC treatment is shown; n = 6 mice per genotype, pooled from 3 independent cohorts of 2 mice per genotype. (E) Mean number of colony forming units from total bone marrow 2 wk after pI:pC treatment as percent of Control for each colony type was calculated; n = 3 mice per genotype, with duplicate plates for each mouse. Data are representative of three independent experiments. (F) The mean numbers and phenotypes of colonies formed by WT bone marrow in the presence of indicated concentrations of EHT 1610 were calculated. Results depict duplicate plates for each condition, and represent 3 independent experiments. (G) Surface IL7R-α expression (H) and intracellular STAT5 phosphorylation (Y694, C) in small and large pre–B cells from the bone marrow of Control and CKO mice 2 wk after pI:pC treatment were assessed by flow cytometry. Data are representative of 3three mice per genotype. For all graphs, error bars depict SD. **, P < 0.01; ****, P < 0.0001.

Figure 3.

Loss of Dyrk1a adversely affects T cell development. (A) Mean thymocyte numbers in Control and CKO mice were assessed 2 and 4 wk after pI:pC treatment; n = 6 mice per genotype, pooled from 3 independent cohorts of 2 mice per genotype. (B) Representative flow cytometry analysis of total (left) and CD4⁻/CD8⁻ double-negative (DN, right) thymocytes is shown. Numbers indicate the percentages in each gate; mean absolute numbers were quantified in (C); n = 6 mice per genotype, pooled from 3 independent cohorts of 2 mice per genotype. (D) Mean percentage of Annexin V⁺ cells was assessed from CD4⁺CD8⁺ double-positive (DP) thymocytes 2 wk after pI:pC treatment; n = 4 mice per genotype. Data are representative of 2 independent cohorts of n = 4 mice per genotype. (E) Mean total thymocyte numbers and representative flow cytometry analysis (F) of thymi from Dyrk1a^wt/wt Lck-Cre⁺ (Control) and Dyrk1a^f/f Lck-Cre⁺ (CKO) mice are shown; n = 6 mice per genotype, pooled from 3 independent cohorts of 2 mice per genotype. **, P < 0.01; ****, P < 0.0001.

Figure 4.

Genetic deletion of Dyrk1a results in cell autonomous loss of competitiveness in lymphoid cells. Total bone marrow cells from either Dyrk1a^f/f Mx1-Cre⁻ (Control CD45.2) or Dyrk1a^f/f Mx1-Cre⁺ (CKO CD45.2) were mixed with an equal number WT CD45.1 bone marrow cells and transplanted into lethally irradiated recipients, which were treated with pI:pC 4 wk later. (A) Mean peripheral blood chimerism (pre–pI:pC treatment) of transplanted mice analyzed by flow cytometry for surface CD45 isoform expression 4 wk after transplant is shown; n = 5 mice per group; error bars depict SD. (B and C) Donor chimerism in thymocytes (B) and bone marrow B cell populations (C) 2 wk after pI:pC treatment was assessed. Each triangle represents one mouse. n = 5 mice per group; error bars depict SD. Data represent two independent transplant experiments. *, P < 0.05; **, P < 0.01; ****, P < 0.0001.

Figure 5.

Loss of Dyrk1a results in failed cell cycle exit during lymphoid development. Cell cycle status of bone marrow and thymocytes from Dyrk1a^f/f Mx1-Cre⁻ (Control) and Dyrk1a^f/f Mx1-Cre⁺ (CKO) was assessed 4 wk after pI:pC treatment. (A and B) Representative flow cytometry plots depicting steady-state cell cycle status using DNA versus RNA content (A) and in vivo BrdU incorporation 24 h after BrdU injection (B) are shown. Numbers depict percentages in each gate. (C) The corresponding defined cell cycle phases and mean percentages in G0 are shown from the indicated Control and CKO cell types shown in A; n = 6 mice per genotype, pooled from 2 independent cohorts of 3 mice per genotype. (D) The mean percentages of cells in each phase of the cell cycle from the indicated Control and CKO cell types in B are shown; n = 4 mice per genotype. Data are representative of 2 independent experiments. Significant differences are denoted for each phase to the right of each CKO bar. (E) Flow cytometry was performed using CellTrace dye dilution in cultured FACS-purified Control and CKO large pre–B cells. Data are representative of two independent experiments. (F) Thymocytes from Control and CKO mice (Lck-Cre) were stained for DNA and RNA content to assess cell cycle status as in A. Representative flow cytometry plots for DP thymocytes from Control and CKO mice are shown; n = 4 mice per genotype, pooled from 2 independent cohorts of 2 mice per genotype. (G) Mean percentages of DP thymocytes in G0 are depicted; n = 4 mice per genotype, pooled from 2 independent cohorts of 2 mice per genotype. For all graphs, error bars depict SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; NS, not significant.

Figure 6.

Dyrk1a-deficient quiescent DP thymocytes and small pre–B cells fail to repress E2F target gene transcription. (A and B) Up- and down-regulated transcripts were identified by RNA-sequencing of FACS-purified quiescent DP thymocytes (CD4⁺/CD8⁺/DNA content^2N/Pyronin Y^low), cycling DP thymocytes (CD4⁺/CD8⁺/DNA content^>2N/Pyronin Y^high), small pre–B cells (IgM⁻/B220⁺/CD43⁻/FSC^low), large pre–B cells (IgM⁻/B220⁺/CD43^low/FSC^high), and granulocytes (Gr-1^high/Mac-1/CD11b^high) from Dyrk1a^f/f Mx1-Cre⁻ (Control) and Dyrk1a^f/f Mx1-Cre⁺ (CKO) mice. Venn diagrams depict shared up-regulated (A) and down-regulated (B) transcripts identified by RNA-sequencing in the three quiescent cell populations. (C–F) mRNA expression of E2F target genes was assessed by qRT-PCR in FACS-purified small pre–B cells (C), quiescent DP thymocytes (D), large pre–B cells (E), and cycling DP thymocytes (F) from Control and CKO mice 2 wk after pI:pC treatment. Transcript levels were normalized to Actb expression. PCRs were performed using 3 independently sorted pairs of samples (each with 1 mouse per genotype) for pre–B cells, and pooled samples from 3 mice per genotype for thymocytes. Error bars depict SD of triplicate wells for representative samples. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, not significant (CKO vs. Control).

Figure 7.

Loss of Dyrk1a alters pre–B cell differentiation markers without affecting pre–BCR signaling. (A and B) mRNA expression was analyzed by qRT-PCR to assess transcripts dynamically regulated during pre–B differentiation in FACS-purified large and small pre–B cells from the bone marrow of Dyrk1a^f/f Mx1-Cre⁻ (Control) and Dyrk1a^f/f Mx1-Cre⁺ (CKO) 4 wk after pI:pC treatment. PCRs were performed using cDNA from 2–3 independently sorted pairs of mice, with 1 mouse per genotype in each pair. Error bars depict SD of triplicate wells for representative samples. (C) Vκ-Jκ1 light chain gene rearrangement was assessed by qPCR using genomic DNA from the same purified cell populations as in A and B; graph depicts mean recombination frequency as a percentage of Control small pre–B from two independent cohorts of mice, with cells pooled from two mice per genotype in each cohort; error bars depict SD. (D and E) Representative flow cytometry plots depicting surface pre–BCR (D) and intracellular μ heavy chain expression (E) in pre–B cells from the bone marrow of Control and CKO mice are shown. Data are representative of three independent cohorts of mice, each with two to three mice per genotype. (F) Western blots were performed to assess the levels of signaling proteins in FACS-purified large and small pre–B cells from the bone marrow of Control and CKO mice as indicated. Data are representative of two independently sorted samples, each pooled from three mice per genotype. Densitometry values were normalized to Actin. (G) ERK activation was assessed by intracellular staining for phospho-ERK in large and small pre–B cells from the bone marrow of Control and CKO mice as indicated in the flow cytometry plot (left); (right) individual (dots) and mean (bars) percentages of phospho-ERK⁺ cells in each population are shown; n = 4 mice per genotype. Data are representative of three independent experiments. (H and I) qRT-PCR analysis of Ccnd3 mRNA expression (H) and pre–BCR induced transcripts (I) from the same purified populations as in A and B were measured. All qRT-PCRs were performed using cDNA from two to three independently sorted pairs of mice, with one mouse per genotype in each pair. Error bars depict SD of triplicate wells for representative samples. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; NS, not significant.

Figure 8.

DYRK1A destabilizes Cyclin D3 by phosphorylating T283. (A and B) Western blots of FACS-purified large and small pre–B cells from Dyrk1a^f/f Mx1-Cre⁻ (Control) and Dyrk1a^f/f Mx1-Cre⁺ (CKO) mice 4 wk after pI:pC treatment were assessed for the indicated proteins as shown. The same membrane (cut into segments according to molecular weight) from Fig. 7 F was stripped and reprobed, hence the same loading control is shown here. Data are representative of three independent experiments. Densitometry values were normalized to actin. (C) The expression levels of the D-type Cyclin transcripts from RNA-seq data (from Fig. 6) are shown for Control and CKO mice. (D and E) Control and CKO thymocytes (D), or WT cultured pre–B cells ± EHT 1610 (E) were treated with cycloheximide (25 µg/ml) for the indicated times before Western blot analysis. Data are representative of three independent experiments. Densitometry values were normalized to actin. (F) An amino acid alignment of C-terminal phosphodegrons in the D-type Cyclins is shown. Conserved residues are highlighted in the blue box; phosphorylated threonines are highlighted in the red box. (G) FLAG-tagged Cyclin D3 (WT or T283A) was immunoprecipitated from transfected 293T cells and used for in vitro kinase assays with recombinant active DYRK1A and ATP-γS. The reaction products were alkylated with PNBM and analyzed for thiophosphate esters by Western blotting. Data are representative of three independent assays. Densitometry values were normalized to FLAG-Cyclin D3.

Figure 9.

Expression of Cyclin D3 T283A mimics loss of DYRK1A activity. (A) Protein expression levels of FLAG-tagged Cyclin D3 retroviral constructs in cultured WT pre–B cells 2 d after transduction were assessed. Densitometry values were normalized to Actin. (B and C) mRNA expression levels of E2F target genes (B) and differentiation markers (C) were assessed by qRT-PCR in FACS-purified WT small pre–B cells after transduction with the indicated constructs. RT-PCRs were performed from two independently sorted samples for each construct; error bars depict SD of triplicate wells for representative samples. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, not significant (indicated condition vs. empty vector). (D) Cell cycle analysis was performed in cultured WT pre–B cells after 2 d of growth conditions (5 ng/ml IL-7 and 10 ng/ml SCF) (dotted line), cell cycle exit conditions (0.05 ng/ml IL-7 and 0.1 ng/ml SCF; shaded gray), or exit conditions in the presence of 500 nM EHT 1610 (red line). (E) Percentages of cells from (D) in S-G2-M under the indicated conditions are shown; error bars depict SD of duplicate wells from a representative of four independent experiments, two of which included Palbociclib treatment. *, P < 0.05; ***, P < 0.001.

Figure 10.

Inhibition of CDK4/6 can rescue cell cycle exit and differentiation in Dyrk1a-defcient cells. (A) Mice of the indicated genotypes were treated with vehicle or 150 mg/kg Palbociclib (oral gavage) for 1 wk after completion of pI:pC treatment. Cell cycle analysis of DP thymocytes and bone marrow small pre–B cells was assessed. Flow cytometry plots represent two independent experiments, each with two to three mice per condition. (B and C) Quantifications of mean quiescent DP thymocyte and small pre–B cell populations are shown; n = 2–3 mice per genotype; error bars depict SD. Data represent two independent experiments, each with two to three mice per condition. (D) qRT-PCR analysis was performed for the indicated transcripts in FACS-purified small pre–B cells from the mice shown in B and C. Data show two pooled mice per condition; error bars depict SD of triplicate wells. Data are representative of two independent Palbociclib treatment experiments. (E) Flow cytometry plot depicts surface CD25 expression on bone marrow small pre–B cells from a representative mouse from each condition as indicated. Data represent two independent experiments, each with two to three mice per condition. (F) Light chain recombination frequencies were measured by qPCR from genomic DNA in FACS-purified bone marrow small pre–B cells from the indicated mice; each sample was pooled from 2 mice; data represent two independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001.