PU.1-c-Jun interaction is crucial for PU.1 function in myeloid development

The Ets transcription factor PU.1 is essential for inducing the differentiation of monocytes, macrophages, and B cells in fetal liver and adult bone marrow. PU.1 controls hematopoietic differentiation through physical interactions with other transcription factors, such as C/EBPα and the AP-1 family member c-Jun. We found that PU.1 recruits c-Jun to promoters without the AP-1 binding sites. To address the functional importance of this interaction, we generated PU.1 point mutants that do not bind c-Jun while maintaining normal DNA binding affinity. These mutants lost the ability to transactivate a target reporter that requires a physical PU.1-c-Jun interaction, and did not induce monocyte/macrophage differentiation of PU.1-deficient cells. Knock-in mice carrying these point mutations displayed an almost complete block in hematopoiesis and perinatal lethality. While the PU.1 mutants were expressed in hematopoietic stem and early progenitor cells, myeloid differentiation was severely blocked, leading to an almost complete loss of mature hematopoietic cells. Differentiation into mature macrophages could be restored by expressing PU.1 mutant fused to c-Jun, demonstrating that a physical PU.1-c-Jun interaction is crucial for the transactivation of PU.1 target genes required for myeloid commitment and normal PU.1 function in vivo during macrophage differentiation.


Co-immunoprecipitations and Western Blotting Analysis
For the interaction screen by co-immunoprecipitation, truncated PU.1 wild-type and mutants without the N-terminal transactivation domain was cloned into pcDNA6 and transiently transfected into Cos-7 cells with pSV1-SPORT-c-Jun using Lipofectamine and Plus Reagent (Invitrogen). 24 hours after transfection, cells were harvested, washed in PBS and lysed in 150mM NaCl, 50mM Tris pH 7.6, 1mM EDTA, 0.1% 1mM PMSF, 1μg/μl leupeptin, 1μg/μl aprotinin, and 5% glycerol by incubating on ice for 15 minutes and passed 3 times through a 28-gauge syringe. After centrifugation at 16,000 x g for 10 minutes at 4°C, the lysate was collected and pre-cleared by incubation with 30μl blocked protein A agarose beads (Santa Cruz) and 10μg normal rabbit serum (Santa Cruz) for 1 hour at 4°C. After centrifugation to remove beads, the lysate was split into two samples and incubated with 5 μg of either normal rabbit IgG or PU.1 antibody (Santa Cruz sc-352) for 2 hours at 4°C. Then, 30μl of packed protein A agarose beads were added to each lysate and incubated overnight at 4°C. The beads were washed twice with lysis buffer, resuspended in 30μl SDS sample buffer, loaded on a 10% SDS-PAGE gel and processed for Western transfer. Western blots were probed with either anti-c-Jun (Santa Cruz sc-1694) or anti-PU.1 (sc-352) antibodies at 1:2000 dilution in TBS with 0.1% Tween-20 and 5% skim milk, incubated with anti-rabbit HRP-conjugated secondary antibody and developed with chemical-luminescent reagents (Perkin-Elmer).

qPCR Validation of Microarray
Total RNAs of both WT and KI GMP (c-kit + Sca-1 -CD16/32 hi CD34 + ) and ST-HSC (CD48 -/CD150 -LSK) from murine E14.5-16.5 fetal livers were extracted with the TRIzol reagent (Invitrogen, USA). cDNAs were then synthesized using QuantiTect ® Reverse Transcription Kit (RT) (QIAGEN, Germany) for RT-qPCR analysis. Quantitative real-time PCR (qPCR) analysis of the cDNAs was carried out with the iQ TM SYBR ® Green Supermix (Bio-Red, USA) and the products were detected by the Rotor-Gene 6000 real-time PCR machine (Corbett Research QIAGEN, Germany). The sequences of the primers used for the qPCR analysis were home-designed and tested in Raw and A20 cell line, as shown Supplementary Table1. Tables   Supplemental Table S1 Supplementary Fig S1: Identification of PU.1 mutants by yeast split-hybrid assay.
(a) Schematic of the screening system. Interacting proteins are fused separately to the LexA DNA binding domain (DBD) and to the VP16 transactivation domain, respectively. Interaction leads to the transcription of the tetracycline repressor (TetR) in the split-hybrid strain YI671, which can subsequently bind to a TetR site upstream of the HIS3 gene, thereby preventing growth in the absence of exogenous histidine. Non-interacting proteins do not synthesize TetR, allowing the yeast strain to synthesize histidine and to grow on histidine-free medium. The region of human c-Jun interacting with PU.1 (amino acids 251-334) attached to the LexA DBD was used as a bait, and the mutant human PU.1 library (containing amino acids 100-272 encompassing PEST and Ets domains) was cloned in-frame with the VP16 transactivation domain.
(b) The densitometry quantification of Fig. 1b. (b) Identification of three-point mutations by the split-hybrid assay. Vector control ("mock") and PU.1 wild-type and mutant vectors were used in yeast split-hybrid and twohybrid assay. The right panels show PU.1 mutants which score in the split-hybrid but not in the twohybrid assay, demonstrating that these mutants fail to interact with the c-Jun bait. Standard two-hybrid analysis allows growth upon interaction of bait and library and growth inhibition when no interaction is taking place.
PU.1 mutants localize to the nucleus.
Cos7 cells transfected with wild-type or mutant PU.1 expression vectors were stained with anti-PU.1 antiserum and analyzed by immunofluorescence microscopy. Wild-type and mutants localized equally efficiently in the nucleus, demonstrating that the failure of these mutants to transactivate their targets is not a result of loss of a nuclear localization function. Assay was repeated 3 times with representative blots shown.
FACS diagram of linc-kit + Sca-1progenitor cells from WT and KI E14.5-16.5 FL. WT GMP and the KI GMP-like cells are indicated by blue and red circles, respectively. (d) Western blot analysis of sorted cells of WT GMP and the GMP-like population (encircled in C). The sorted cell number of each population is indicated. (e) Wright-Giemsa-stained sorted CD34 lo CD16/32 hi (FcγRII/III hi ) population of cells resembling immature myeloid cells (population encircled in red in (c)) is shown. (f) Flow cytometry analysis of E16.5 fetal liver cells. The relative contribution of the listed cell populations to total live (DAPI -) fetal liver cells is shown as the mean of values from quadruplicates ± SD. Representative FACS profiles are shown for wildtype (WT), heterozygous (Het) and homozygous PU.1 Q202L knock-in cells (KI). Experiments were performed at least three times with similar results. Analysis of hematopoietic KSL (c-kit+ + Sca-1+ lin-) cells. For the data presented herein, three independent experiments were performed with representative results shown.