A requirement for fibroblast growth factor in regulation of skeletal muscle growth and differentiation cannot be replaced by activation of platelet-derived growth factor signaling pathways.

The distinct effects of cytokines on cellular growth and differentiation suggest that specific signaling pathways mediate these diverse biological activities. Fibroblast growth factors (FGFs) are well-established inhibitors of skeletal muscle differentiation and may operate via activation of specific signaling pathways distinct from recently identified mitogen signaling pathways. We examined whether platelet-derived growth factor (PDGF)-activated signaling pathways are sufficient to mediate FGF-dependent repression of myogenesis by introducing the PDGF beta receptor into a mouse skeletal muscle cell line. Addition of PDGF-BB to cells expressing the PDGF beta receptor activated the PDGF beta receptor tyrosine kinase, stimulated mitogen-activated protein (MAP) kinase, and increased the steady-state levels of junB and c-fos mRNAs. Despite the activation of these intracellular signaling molecules, PDGF beta receptor activation elicited no detectable effect on cell proliferation or differentiation. In contrast to PDGF-BB, addition of FGF-2 to myoblasts activated signaling pathways that resulted in DNA synthesis and repression of differentiation. Because of the low number of endogenous FGF receptors expressed, FGF-stimulated signaling events, including tyrosine phosphorylation and activation of MAP kinase, could be detected only in cells expressing higher levels of a transfected FGF receptor cDNA. As the PDGF beta receptor- and FGF receptor-stimulated signaling pathways yield different biological responses in these skeletal muscle cells, we hypothesize that FGF-mediated repression of skeletal muscle differentiation activates signaling pathways distinct from those activated by the PDGF beta receptor. Activation of PDGF beta receptor tyrosine kinase activity, stimulation of MAP kinase, and upregulation of immediate-early gene expression are not sufficient to repress skeletal muscle differentiation.

Estrogen receptor affinity and location of consensus and imperfect estrogen response elements influence transcription activation of simplified promoters.

We have examined the ability of estrogen receptor (ER) to bind and bend DNA fragments containing the Xenopus laevis vitellogenin A2 estrogen response element (ERE), which contains a palindromic, consensus ERE sequence, the X. laevis vitellogenin B1 ERE2, which contains a 1-bp mismatch in the 5'-end of the half-palindrome, and the human pS2 ERE, which contains a 1-bp mismatch in the 3'-end of the half-palindrome. ER binding induced a 65 degrees bend in DNA fragments containing the consensus ERE, the vitellogenin B1 ERE2, or the pS2 ERE. However, ER affinity for the consensus ERE was 2-fold greater than for either the vitellogenin B1 ERE2 or the pS2 ERE. When Chinese hamster ovary (CHO) cells were transfected with reporter plasmids containing either the consensus ERE, the vitellogenin B1 ERE2, or the pS2 ERE separated from the TATA sequence by 26 helical turns, exposure to 10 nm 17 beta-estradiol increased transcription 12.7-, 2.4-, and 3.8-fold, respectively. Increasing the spacing between the ERE and TATA sequence to three helical turns decreased the ability of the consensus ERE to activate transcription by 55% and increased the ability of the pS2 ERE to activate transcription by 35% but had no significant effect on vitellogenin B1 ERE2 activity. Further increasing the distance between the ERE and TATA sequence to 3.6 helical turns restored the activity of promoters containing the consensus ERE and pS2 ERE but decreased the activity of the promoter containing the relatively weak vitellogenin B1 ERE2. These data support the idea that 1) the affinity of ER for the ERE, 2) the location of an ERE within the promoter, and 3) the magnitude and orientation of DNA bends induced by binding of ER or other proteins are important in transcription activation of estrogen-responsive genes.

Overproduction of full-length and truncated human estrogen receptors in Escherichia coli.

The full-length human estrogen receptor (hER) as well as two overlapping peptides of hER were overproduced in Escherichia coli JM109 cells, using the inducible pIC vector system. The N-terminal receptor peptide contains the DNA-binding domain as well as the hinge region, whereas the C-terminal peptide contains the same hinge region and the hormone-binding domain. Typically, 1-6 mg of estrogen receptor (ER) peptides can be recovered from 1 L E. coli cell cultures. The majority of the overexpressed proteins are found in inclusion bodies, which allow the isolation of ER peptides in high yields and of 50-80% purity. Induction for short time periods at 10 microM inducer yielded up to 50% of the ER peptides in soluble form with full biological activity. Both the intact receptor and the C-terminal fragment specifically bound estrogens and antiestrogens, whereas ER peptides that contained the DNA-binding domain were retained on a DNA-agarose resin.