The Ras oncogene products regulate the expression of genes in transformed cells, and members of the Ets family of transcription factors have been implicated in this process. To determine which Ets factors are the targets of Ras signaling pathways, the abilities of several Ets factors to activate Ras-responsive enhancer (RRE) reporters in the presence of oncogenic Ras were examined. In transient transfection assay, reporters containing RREs composed of Ets-AP-1 binding sites could be activated 30-fold in NIH 3T3 fibroblasts and 80-fold in the macrophage-like line RAW264 by the combination of Ets1 or Ets2 and Ras but not by several other Ets factors that were tested in the assay. Ets2 and Ras also superactivated an RRE composed of Ets-Ets binding sites, but the Ets-responsive promoter of the c-fms gene was not superactivated. Mutation of a threonine residue to alanine in the conserved amino-terminal regions of Ets1 and Ets2 (threonine 38 and threonine 72, respectively) abrogated the ability of each of these proteins to superactivate reporter gene expression. Phosphoamino acid analysis of radiolabeled Ets2 revealed that Ras induced normally absent threonine-specific phosphorylation of the protein. The Ras-dependent increase in threonine phosphorylation was not observed in Ets2 proteins that had the conserved threonine 72 residue mutated to alanine or serine. These data indicate that Ets1 and Ets2 are specific nuclear targets of Ras signaling events and that phosphorylation of a conserved threonine residue is a necessary molecular component of Ras-mediated activation of these transcription factors.

PU.1 is a tissue-specific transcription factor that is expressed in cells of the hematopoietic lineage including macrophages, granulocytes, and B lymphocytes. Bone marrow-derived macrophages transfected with an antisense PU.1 expression construct or treated with antisense oligonucleotides showed a decrease in proliferation compared with controls. In contrast, bone marrow macrophages transfected with a sense PU.1 expression construct displayed enhanced macrophage colony-stimulating factor (M-CSF)-dependent proliferation. Interestingly, there was no effect of sense or antisense constructs of PU.1 on the proliferation of the M-CSF-independent cell line, suggesting that the response was M-CSF dependent. This was further supported by the finding that macrophages transfected with a sense or an antisense PU.1 construct showed, respectively, an increased or a reduced level of surface expression of receptors for M-CSF. The enhancement of proliferation seems to be selective for PU.1, since transfections with several other members of the ets family, including ets-2 and fli-1, had no effect. Various mutants of PU.1 were also tested for their ability to affect macrophage proliferation. A reduction in macrophage proliferation was found when cells were transfected with a construct in which the DNA-binding domain of PU.1 was expressed. The PEST (proline-, glutamic acid-, serine-, and threonine-rich region) sequence of the PU.1 protein, which is an important domain for protein-protein interactions in B cells, was found to have no influence on PU.1-enhanced macrophage proliferation when an expression construct containing PU.1 minus the PEST domain was transfected into bone marrow-derived macrophages. In vivo, PU.1 is phosphorylated on several serine residues. The transfection of plasmids containing PU.1 with mutations at each of five serines showed that only positions 41 and 45 are critical for enhanced macrophage proliferation. We conclude that PU.1 is necessary for the M-CSF-dependent proliferation of macrophages. One of the proliferation-relevant targets of this transcription factor could be the M-CSF receptor.

PU.1 recruits the binding of a second B cell-restricted nuclear factor, NF-EM5, to a DNA site in the immunoglobulin kappa 3' enhancer. DNA binding by NF-EM5 requires a protein-protein interaction with PU.1 and specific DNA contacts. Dephosphorylated PU.1 bound to DNA but did not interact with NF-EM5. Analysis of serine-to-alanine mutations in PU.1 indicated that serine 148 (Ser148) is required for protein-protein interaction. PU.1 produced in bacteria did not interact with NF-EM5. Phosphorylation of bacterially produced PU.1 by purified casein kinase II modified it to a form that interacted with NF-EM5 and that recruited NF-EM5 to bind to DNA. Phosphopeptide analysis of bacterially produced PU.1 suggested that Ser148 is phosphorylated by casein kinase II. This site is also phosphorylated in vivo. Expression of wild-type PU.1 increased expression of a reporter construct containing the PU.1 and NF-EM5 binding sites nearly sixfold, whereas the Ser148 mutant form only weakly activated transcription. These results demonstrate that phosphorylation of PU.1 at Ser148 is necessary for interaction with NF-EM5 and suggest that this phosphorylation can regulate transcriptional activity.

We have isolated a cDNA clone, PU.1, that codes for a new tissue-specific DNA binding protein. Analysis of the binding site by methylation interference and DNA-ase I protection revealed that the PU.1 protein recognized a purine-rich sequence, 5′-GAGGAA-3′ (PU box). The PU.1 protein was shown to be a transcriptional activator that is expressed in macrophages and B cells. cDNA constructions used to generate proteins lacking portions of either the amino- or carboxy-terminal ends of the PU.1 protein placed the DNA binding domain in the highly basic carboxy-terminal domain of the protein. The amino acid sequence in the binding domain of PU.1 has considerable identity with proteins belonging to the ets oncogene family.