MicroRNA-34a perturbs B lymphocyte development by repressing the forkhead box transcription factor Foxp1

Abstract

MicroRNAs (miRNAs) can influence lineage choice or affect critical developmental checkpoints during hematopoiesis. We examined the role of the p53-induced microRNA miR-34a in hematopoiesis by gain-of-function analysis in murine bone marrow. Constitutive expression of miR-34a led to a block in B cell development at the pro-B-cell-to-pre-B-cell transition, leading to a reduction in mature B cells. This block appeared to be mediated primarily by inhibited expression of the transcription factor Foxp1. Foxp1 was a direct target of miR-34a in a 3'-untranslated region (UTR)-dependent fashion. Knockdown of Foxp1 by siRNA recapitulated the B cell developmental phenotype induced by miR-34a, whereas cotransduction of Foxp1 lacking its 3' UTR with miR-34a rescued B cell maturation. Knockdown of miR-34a resulted in increased amounts of Foxp1 and mature B cells. These findings identify a role for miR-34a in connecting the p53 network with suppression of Foxp1, a known B cell oncogene.

Figure 1. Constitutive expression of miR-34a in the bone marrow compartment leads to a decrease in mature B-lymphocytes

A. Schematic representation of the MSCV-based retroviral vector used to express miR-34a, based on the previously described MGP vector (O’Connell et al., 2009). B. Expression of pre-miR-34a (~60nt) and mature miR-34a (~20nt) was assayed by RNA blotting. The bottom panel shows Ethidium bromide staining (EtBr) to confirm equal loading. C. Analysis of GFP expression in bone marrow two months after retroviral transduction and transfer. D. Expression of miR-34a in bone marrow by RT-qPCR, normalized by 5S. Individual dots represent one mouse; the lines show the mean of miR-34a expression in individual mice from two independent experiments (n=8 for each group; T-test, p=0.009). E. Bone marrow mature B cells detected as a percentage of GFP⁺ cells by flow cytometry in mice expressing vector alone or miR-34a are shown. Individual dot represents one mouse; the lines show the mean percentage of B-cells in individual mice from three independent experiments (n=10 for each group; T-test, p=0.0093). F–H. Flow cytometric analyses for myeloid cells (defined as GFP⁺CD11b⁺), erythroid cells (GFP⁺Ter119⁺), and T cells (GFP⁺CD3ε⁺) are shown. Individual dots represent percent of myeloid cells (F), erythroid cells (G) and T cells (H) in the bone marrow of an individual mouse, and the bars show the mean percent of each cell type in mice from three independent experiments (n=9 for each group; T-test, n.s.).

Figure 2. miR-34a expression causes an increase in pro-B cells and a decrease in pre-B cells

A. Analysis of pro-B cells. Bone marrow cells stained with CD19, c-kit and IgM, analyzed by flow cytometry. Left hand panels show representative histograms of the GFP⁺IgM⁻ compartment in control (vector) and miR-34a expressing mice. In the right hand panel, individual dots represent the percent of pro-B cells in the bone marrow of an individual mouse, and the bars show the mean percent of pro-B cells in mice from three independent experiments (n=12 for control and n=10 for miR-34a mice; T-test, p=.0001). B. Analysis of pre-B cells. Bone marrow cells stained with B220, IgM and CD43 were analyzed by flow cytometry. Left hand panels show representative histograms of the GFP⁺IgM⁻ compartment in control and miR-34a expressing mice. In the right hand panel, individual dots represent the percent of pro-B cells in the bone marrow of an individual mouse, and the bars show the mean percent of pro-B cells in mice from three independent experiments (n=14 for control and n=10 for miR-34a mice; T-test, p=0.0002). C. Assessment of miR-34a expression in B-lineage cells at various stages of differentiation collected by cell sorting. Bar graph shows mean + s.d. of miR-34a expression (n=2 samples; three independent experiments were analyzed). D. Peripheral blood B-lymphocytes in control and miR-34a mice, with individual dots representing the percent of B-lymphocytes in individual mice and the bars showing the mean percent of B-cells from three independent experiments (n=11 for vector control and n=7 for miR-34a; p=0.02). E. Enumeration of splenic B cells in miR-34a expressing mice. Individual dots represent the percent of B-cells amongst total spenocytes in individual mice; bars represent the mean percent of B-cells in mice from 3 independent experiments (n=9 for vector control and n=9 for miR-34a).

Figure 3. Foxp1 is a bona fide target of miR-34a

A. Schematic representation of Foxp1 cDNA and 3′UTR showing the conserved miR-34a seed region in its 3′UTR. Human and mouse miR-34a mature sequences with proposed base pairing to the 3′UTR are presented as is the mutant 3′UTR used in the luciferase assays depicted in B. B. Targeting of Foxp1 by miR-34a is via direct 3′UTR interactions. Graphed is the relative luciferase activity of 293T extracts transfected with the designated luciferase-3′UTR construct and mir-34a compared to cells transfected with the luciferase-3′UTR and empty vector, normalized for transfection efficiency by a β-galactosidase reporter. Designations are for the various 3′UTRs attached to the luciferase reporter: a.s.2-mer, two copies of the miR-34a anti-sense sequence; FOXP1mt, FOXP1 3′UTR with mutated mir-34a seed site; BCL2mt, BCL2 3′UTR with mutated mir-34a seed site; No UTR, luciferase gene without a 3′UTR. C. FOXP1 expression, measured by RT-qPCR in NALM6 clones that express control vector (FUGW) or miR-34a (FUGW/34a). Individual dots represent FOXP1 RNA levels in individual clones, while the bars represent the mean values in each case (T-test, p=0.0086). D. miR-34a expression, measured by RT-qPCR, in NALM6 clones as in (C). E. Immunoblot analysis of of murine Foxp1 in 70Z/3 cells infected with MGP-34a versus vector alone. The numbers below the blot indicate the relative expression level of the protein as compared with the control. This experiment was repeated twice with similar results. F. Assessment of FOXP1 expression in B-lineage cells at various stages of differentiation collected by cell sorting. Bar graph shows mean + s.d. of FOXP1 expression (n=2 samples; two independent experiments were analyzed). G. and H. Foxp1 (G) and L32 (H) expression measured by RT-qPCR in the bone marrow of mice reconstituted with either vector- or miR-34a (n=7 for vector, n=8 for miR-34a). T-test, p=0.0225 for comparison of Foxp1 expression levels.

Figure 4. Knockdown of Foxp1 recapitulates miR-34a-induced B-lineage abnormalities

A. Schematic diagram showing the MGP-based construct used to express the Foxp1 siRNA. B. Left hand panel shows Foxp1 RNA, measured by RT-qPCR, in 70Z/3 cells infected with either empty vector (Vector) or MGP-Foxp1si (Foxp1si). Right hand panel shows immunoblot analysis of Foxp1 in these cells. The numbers represent the fold expression of Foxp1 at the protein level compared to vector control. C. GFP expression in the bone marrow of mice (vector and Foxp1si) 2 months after transplant, measured by flow cytometry. D. Foxp1 expression in the bone marrow of mice (vector and Foxp1si), as measured by RT-qPCR. Individual dots represent Foxp1 RNA amounts in individual mice, while the bars show the mean amount for a representative experiment (n=4 for each group; T-test, p=0.0026). Experiments were repeated three times, with similar trends in all three experiments. E. Analysis of pro-B cells. Left hand panels show representative histograms of the GFP⁺IgM⁻ negative compartment in control (vector) and Foxp1si mice. The right hand panel shows pro-B cells as a percentage of total GFP+ cells for a representative experiment, with individual mice represented by individual dots and the bars indicating the mean (n=4 per group; T-test, p=0.0004). Three independent experiments showed similar trends. F. Analysis of pre-B cells. Left hand panels show representative histograms of the GFP+ IgM− compartment in control and Foxp1si mice, below which the percentage of cells in each of the four quadrants is shown. The right hand panel shows pre-B cells as a percentage of total GFP+ cells for a representative experiment, with individual mice represented by individual dots and the bars indicating the mean (n=4 per group; T-test, p=0.005). Three independent experiments showed similar trends.

Figure 5. Foxp1 expression rescues the miR-34a induced B cell developmental abnormality

A. Schematic representation of the constructs used to rescue miR-34a-mediated block in B cell development. B. Immunoblot analysis of extracts from 70Z/3 cells infected with retroviruses depicted in A. The numbers represent relative expression of Foxp1, compared with control. C. Expression of miR-34a as assayed by RT-qPCR in 70Z/3 cells infected with retroviruses [vector control only, MGP-34a (miR-34a), MIG-Foxp1 (Foxp1), or MIG-Foxp1+34a (Foxp1+34a)]. Individual dots represent measurements in individual mice, and the bars indicate the means for each group (n=4 per group; two independent experiments were performed). Statistically significant differences were found in the expression of miR-34a for the following comparisons: vector vs. miR-34a (T-test, p=0.0001) and vector vs. Foxp1+miR-34a (T-test, p=0.0049). D. Expression of Foxp1 in bone marrow, as assayed by RT-qPCR, in mice receiving bone marrow transduced with the same constructs as in C, represented as fold-overexpression over control and normalized to L32. Individual dots represent measurements in individual mice, and the bars indicate the means for each group (n=4 per group; two independent experiments were performed). Statistically significant differences were found in the expression of Foxp1 for the following comparisons: vector vs. Foxp1 (T-test, p=0.036) and vector vs. Foxp1+miR-34a (T-test, p=0.015). E. Expression of miR-34a in bone marrow, as assayed by RT-qPCR, in mice receiving bone marrow transduced with the same constructs as in C, shown as fold-overexpression over control and normalized to L32. Individual dots represent measurements in individual mice, and the bars indicate the means for each group (n=4 per group; two independent experiments were performed). Statistically significant differences were found in the expression of miR-34a for the following comparisons: vector vs. miR-34a (T-test, p=0.03) and vector vs. Foxp1+miR-34a (T-test, p=0.0056). F. Enumeration of pro-B cells (CD19+IgM-c-kit+) as a percentage of total GFP+ cells for a representative experiment (n=4 per group; T-test, p=0.0048). Individual dots represent measurements in individual mice, and the bars indicate the means for each group. Two independent experiments were performed. G. Enumeration of pre-B cells (B220+CD43−IgM−) as a percentage of total GFP+ cells for a representative experiment (n=4 per group; T-test, p=0.0084). Individual dots represent measurements in individual mice, and the bars indicate the means for each group. Two independent experiments were performed.

Figure 6. Knockdown of miR-34a using a “sponge” strategy is effective and results in increased mature B cells in the bone marrow

A. Schematic representation of the construct used to knockdown miR-34a expression. The backbone is the MGP vector described previously with the insertion of “spacer miR-Target sites” downstream of the GFP. Abbreviation, 34a-T: miR-34a synthetic target site. B. Immunoblot analysis of extracts from 70Z/3 cells infected with retroviruses prepared from construct depicted in A. The numbers represent relative expression of Foxp1, compared with control. C–E. Luciferase assays demonstrate that the knockdown of miR-34a by the construct designated in A is specific for miR-34a target-containing 3′UTRs, graphed as described for Fig 3B. Statistically significant differences were found in relative luciferase activity for the following comparisons: 34a versus 34a+anti-34as (Fig 6C, T-test, p=0.0053); control vs. anti-34as (Fig 6D; T-test, p=0.0001) and 34a vs. 34a+anti-34as (Fig 6D; T-test, p=0.0033). F. Bone marrow flow cytometry plot showing that transduction results in the expression of GFP. G–H. Bone marrow flow cytometry plot showing GFP+ cells plotted for expression of CD19 and IgM. Bone marrow is from a mouse transduced with MGP (Control; G), and a mouse transduced with MGP-anti-34as (H). I. Enumeration of mature B cells (CD19+IgM+) as a percentage of total GFP+ cells in mice receiving either control or the knockdown construct for miR-34a (n=4 for control MGP; n=8 for anti-34as; T-test, p=0.0285). Individual dots represent measurements in individual mice, and the bars indicate the means for each group. Two independent experiments were performed. J. Foxp1 expression in bone marrow transductant mice reciveing marrow with either MGP (Control) or anti-34as marrow (n=4 for control MGP; n=8 for anti-34as; T-test, p=0.0022). Individual dots represent measurements in individual mice, and the bars indicate the means for each group. Two independent experiments were performed. K. Enumeration of mature B cells (CD19+IgM+) as a percentage of total GFP+ cells in mice recieivng either control or constitutive expression contruct for Foxp1 (n=8 for control; n=8 for MIG/Foxp1; T-test, p=0.0383). Individual dots represent measurements in individual mice, and the bars indicate the means for each group. Two independent experiments were performed.