Mammalian transcription factor LSF is a target of ERK signaling

Abstract

LSF is a mammalian transcription factor that is rapidly and quantitatively phosphorylated upon growth induction of resting, peripheral human T cells, as assayed by a reduction in its electrophoretic mobility. The DNA-binding activity of LSF in primary T cells is greatly increased after this phosphorylation event (Volker et al. [1997]: Genes Dev 11:1435-1446). We demonstrate here that LSF is also rapidly and quantitatively phosphorylated upon growth induction in NIH 3T3 cells, although its DNA-binding activity is not significantly altered. Three lines of experimentation established that ERK is responsible for phosphorylating LSF upon growth induction in both cell types. First, phosphorylation of LSF by ERK is sufficient to cause the reduced electrophoretic mobility of LSF. Second, the amount of ERK activity correlates with the extent of LSF phosphorylation in both primary human T cells and NIH 3T3 cells. Finally, specific inhibitors of the Ras/Raf/MEK/ERK pathway inhibit LSF modification in vivo. This phosphorylation by ERK is not sufficient for activation of LSF DNA-binding activity, as evidenced both in vitro and in mouse fibroblasts. Nonetheless, activation of ERK is a prerequisite for the substantial increase in LSF DNA-binding activity upon activation of resting T cells, indicating that ERK phosphorylation is necessary but not sufficient for activation of LSF in this cell type.

Fig. 1

Mitogenic stimulation of human peripheral T cells leads to LSF phosphorylation, increased LSF-DNA binding activity, and ERK activation. A: Western blot analysis of nuclear extracts from either unstimulated or stimulated human peripheral T cells. Approximately 40 µg of extract were loaded per lane. The blot was probed with α-LSF pep1-1 antibody. Lane 1: unstimulated cells, lanes 2–4: cells treated with PHA, ionomycin, and 74 ng/ml of PMA, lanes 5–6: cells treated with 74 ng/ml of PMA, and lanes 7–9: cells treated with PHA. The lengths of time of treatment in min are indicated. Note the slight curvature across the lanes in the relative mobilities of the proteins. B: Kinetics of ERK activation by mitogenic stimuli in T cells. Twenty micrograms of the same samples used in panel A were loaded per lane. The blot was probed with antiphospho-ERK antibody (upper panel) and reprobed with JNK antibody (lower panel) to detect differences in the loading of total protein. Lanes 1–9 are as in panel A. The positions of phosphorylated p44 ERK1, phosphorylated p42 ERK2, and p54 JNK are indicated. Note the slight curvature across the lanes in the relative mobilities of the proteins. C: LSF DNA-binding activity is enhanced in response to mitogenic stimulation in T cells. EMSA of nuclear extracts prepared in a separate experiment from that in panels A and B, either from quiescent cells or from cells stimulated with various mitogens. The positions of migration of LSF/DNA complexes and free DNA are indicated. The asterisk denotes a complex with non-specific DNA-binding proteins. In this experiment, increases in LSF DNA-binding activity were 2.3-fold (lane 3), 1.9-fold (lane 6), and 1.4-fold (lane 9), as compared with lane 1.

Fig. 2

LSF is quantitatively phosphorylated without altering DNA-binding activity in NIH 3T3 cells. A: Western blot analysis of LSF in total cell lysates from serum-deprived or mitogenically stimulated NIH 353 cells. Cell lysate from 1 × 10⁶ cells was loaded per lane. The membrane was probed with α-LSFpep1-1 antibody. Lane 1: serum-deprived cells, lanes 2–4: cells treated with EGF, lanes 5–7: cells treated with 10% serum, and lane 8: 250 ng His-LSF as a marker. The lengths of time of treatment are indicated. Note the slight curvature across the lanes in the relative mobilities of the proteins. B: ERK is rapidly activated by mitogenic stimuli in NIH 3T3 cells. An in gel kinase assay was performed on the same samples described in panel A. The mobilities of active p44 and p42 ERKs are indicated. C: Nuclear extracts prepared from serum-deprived cells or from cells stimulated for 5, 30, and 60 min with mitogens were subjected to EMSA. The positions of migration of LSF/DNA complexes, free DNA, and antibody/LSF/DNA complexes (“supershifted LSF”) are indicated. Lane 1: 50 ng His-LSF, lane 2: no added protein, lane 3: untreated cells; lanes 4–6 and 10–12: cells treated with EGF; lanes 7–9: cells treated with 10% serum. Lane 11: incubation with preimmune rabbit serum; lane 12: incubation with polyclonal antibody α-LSFpep1–2 [Volker et al., 1997].

Fig. 3

In vitro phosphorylation of LSF by ERK2 does not alter its DNA-binding activity. A: Western blot analysis of LSF phosphorylated by GST-ERK2 in vitro. Lane 1: 1 µg of His-LSF incubated only with ATP, lane 2: 1 µg of His-LSF incubated with activated GST-ERK2 and ATP, lane 3: 1 µg of His-LSF incubated with activated GST-ERK2 and AMP-PNP. The separated products of the kinase reactions were probed with α-LSFpep1-1 antibody. B: Increasing amounts of either unphosphorylated (lanes 1–14) or GST-ERK2 phosphorylated (lanes 15–28) His-LSF were analyzed by EMSA. Positions of free DNA and LSF/DNA complexes and concentrations of LSF monomer in nM are indicated. C: The averaged data of two experiments, including that shown in panel B, are plotted as a function of concentration of LSF monomer in the reaction. DNA binding is expressed as the percentage of DNA that formed a complex with LSF in the reaction. (Note: LSF binds DNA as a homotetramer.)

Fig. 4

The ERK pathway targets LSF for phosphorylation after mitogenic stimulation of NIH 3T3 cells. A: Western blot analysis of LSF, as described in Fig. 2. Lane 1: extract from serum-deprived NIH 3T3 cells; lanes 2–4: extracts from NIH 3T3 cells stimulated with EGF for 5 min. Lanes 3,4: cells were pre-treated with the MEK1/2 inhibitor PD98059, at the indicated concentrations. Note the slight curvature across the lanes in the relative mobilities of the proteins. B: Western blot analysis of the samples in panel A with polyclonal antibody against ERK1. This antibody recognizes both inactive and active (slower migrating) forms of both ERK1 and ERK2.

Fig. 5

The MEK 1/2 inhibitor PD98059 diminishes the rapid, quantitative phosphorylation of LSF in PMA-induced human peripheral T cells. A: T cells stimulated with 10 ng/ml PMA for the indicated amounts of time were either mock treated (lanes 1–4), or preincubated with 50 µM (lanes 5–8) or 100 µM (lanes 9–12) MEK 1/2 inhibitor. The blot of separated nuclear extracts was probed with α-LSFpep1-1 antibody. The position of LSF is bracketed. Note the curvature across the lanes in the relative mobilities of the proteins. B: The blot from panel A was reprobed with anti-phospho ERK antibody. The positions of phosphorylated p44 and p42 are indicated.

Fig. 6

The MEK 1/2 inhibitor PD98059 prevents both phosphorylation of LSF and activation of ERK activity in PHA-stimulated human peripheral T cells. A: T cells stimulated for the indicated amounts of time with PHA were either mock treated (lanes 1–4) or preincubated with 50 µM PD98059 (lanes 5–8). The blot of separated nuclear extracts (40 µg per lane) was probed with α-LSFpep1–2 antibody. The position of LSF is bracketed. Note the slight curvature across the lanes in the relative mobilities of the proteins. B: The blot from panel A was reprobed with anti-phospho ERK antibody. The positions of phosphorylated p44 and p42 are indicated. The asterisk denotes a non-specific, cross-reacting band.

Fig. 7

MEK 1/2 inhibitor prevents activation of LSF DNA-binding activity in human peripheral T cells. A: Nuclear extracts from T cells stimulated with PHA, which were either mock treated (lanes 1–4), or preincubated with 50 or 100 µM PD98059 (lanes 5–8 or 9–12, respectively), were subjected to EMSA. The positions of migration of LSF/DNA complexes and of free DNA are indicated. The asterisk denotes a non-specific DNA-binding protein. B: A different donor set of T cells from that in panel A were either preincubated with 100 µM MEK 1/2 inhibitor (lanes 2,4,6) or mock treated (lanes 1,3,5), followed by stimulation with PHA for the indicated periods of time. Extracts were analyzed as in panel A.