protein kinase activity required for an early step in ... fileprotein kinase activity required for...

Protein Kinase Activity Required for an Early Step in Interferon-a Signaling*

(Received for publication, June 19, 1991)

Daniel S. KesslerS and David E. Levy$ From the KaDlan Cancer Center and DeDartment of Pathology, New York University School of Medicine, New York, New York 10016

Interferon-a (IFNa) induces an immediate transcriptional response of a restricted set of genes in target cells. Specific transcription is mediated by the cytoplasmic activation of a transcription factor complex termed ISGF3. ISGF3 is a multimeric protein complex composed of a regulatory component (ISGF3a), which is activated following IFNa treatment, and a DNA- binding component ( ISGF~Y) , which recognizes the IFNa-stimulated response element (ISRE). Following activation, ISGF3a translocates to the nucleus where ISGF3 assembles as a high affinity complex on the ISRE. The biochemical basis for receptor-mediated activation of ISGF3 is unknown. We report that two potent protein kinase inhibitors, staurosporine and K- 252a, ablated the transcriptional response to IFNa treatment. These inhibitors prevented the activation of the ISGF3a component without affecting the ISGF3y component, resulting in no accumulation of mature ISGF3 in nuclei of treated cells. Although these agents are potent inhibitors of protein kinase C (PKC), PKC does not mediate ISGF3a activation. Down-regulation of PKC by chronic exposure of cells to 12-0-tetradecanoylphorbol- 13-acetate, which led to complete loss of PKC-immunoreactive material, failed to ablate the transcriptional response to IFNa or the activation of ISGF3a. The PKC-specific inhibitor calphostin C did not perturb activation or nuclear accumulation of ISGF3. We conclude that a novel, staurosporine/K- 252a-sensitive kinase is required for ISGF3 activity and may participate in receptor-mediated signal transduction.

Interferons (IFNs)’ are a family of polypeptide cytokines which induce an antiproliferative and antiviral response in susceptible cells. Type I interferons (IFNaIP) interact with a

* This work was supported in part by National Institutes of Health Grant R01-AI28900, an investigator grant from the Cancer Research Institute, and by the Life and Health Insurance Medical Research Fund. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported by a postdoctoral fellowship from the National Insti- tutes of Health.

I Kaplan Scholar of the Rita and Stanley H. Kaplan Cancer Center (National Institutes of Health Grant CA16087), Pew Scholar in the Biomedical Sciences, and recipient of Junior Faculty Research Award JFRA-278 from the American Cancer Society. To whom all corre- spondence should be addressed Kaplan Cancer Center, NYU School of Medicine, 550 First Ave., New York, NY 10016.

I The abbreviations used are: IFN, interferon; TPA, 12-0-tetradecanoylphorbol-13-acetate; ISRE, interferon-stimulated response element; ISG, interferon-stimulated gene; PKC, protein kinase C.

common cell surface receptor present on nearly all human cells, resulting in the synthesis of a group of new cellular proteins responsible for the resulting physiological changes (see De Maeyer and De Maeyer-Guignard (1988) for review). Many of the new proteins produced early in the response to IFNa are the products of a group of genes whose transcriptional activity is rapidly but transiently increased to very high levels (Friedman et al., 1984; Larner et al., 1984) in a response entirely dependent upon pre-existing cellular proteins (Fried- man and Stark, 1985; Levy et al., 1986; Larner et al., 1986). This transcriptional stimulation results from activation of a latent transcription factor which recognizes a conserved cis- acting DNA sequence (the IFN-stimulated response element (ISRE)) located within the regulatory sequences of target genes (Levy et al., 1988; Kessler et al., 198813; Porter et al., 1988; Rutherford et al., 1988; Cohen et al., 1988; Friedman and Stark, 1985; Kessler et al., 1988a; Reich and Darnell, 1989; Shirayoshi et al., 1988).

A fundamental question concerning the induction of transcription by IFNa is the nature of the signal transduction pathway activated by ligand binding to cell surface receptors. The initial stimulation of RNA transcription requires activation of a latent, cytoplasmic transcription factor, reducing the question of signaling to the regulatory mechanism im- pinging on this factor. This positive activator, termed ISGF3, is a multicomponent complex consisting of four polypeptides (Kessler et al., 1990; Fu et al., 1990). Following its activation in the cytoplasm of IFNa-treated cells, ISGF3 translocates to the nucleus (Levy et al., 1989; Dale et al., 1989) where it assembles on the ISRE and activates transcription.

In its latent state, ISGF3 most likely exists as two independent components which associate only following exposure of cells to IFNa (Levy et al., 1989; Bandyopadhyay et al., 1990). One of these components, termed ISGF37, consists of a single, 48-kDa polypeptide and possesses the DNA-binding domain of ISGF3 (Kessler et al., 1990). ISGF37, isolated from either unstimulated or IFNa-treated cells, is able to bind the ISRE and is thus probably not a direct target for regulation. The second, regulatory component of ISGF3, termed ISGF3a, may consist of as many as three distinct polypeptides (Kessler et al., 1990; Fu et al., 1990). ISGF3a is activated following IFNa treatment of cells, converting a latent, inactive form into a form capable of nuclear translocation, interaction with ISGF37, and stable complex assembly on the ISRE. One or more of the polypeptides in this component is likely to be directly modified in response to the liganded receptor. Acti- vation of ISGF3a is thus the most receptor-proximal event yet identified in the IFNa signaling pathway.

In attempts to further elucidate receptor-coupled signal transduction in the IFNa pathway, we and others have tested a series of compounds with known effects on intracellular signaling pathways for their ability to either stimulate or

23471

23472 IFNa Signal Transduction Pathway

inhibit transcriptional induction following IFNa treatment. As previously reported, agents which perturb the activities of conventional signaling mechanisms involving known protein kinases or ion fluxes have failed to elucidate the biochemical components of the IFNa signaling apparatus (Kusari and Sen, 1987; Larner et d., 1984; Yan et al., 1989; Hubbell et al., 1991; Hannigan and Williams, 1991; Reich and Pfeffer, 1990).

Recently, two reports have suggested a role for protein kinase C (PKC) in IFNa-stimulated gene expression. Pfeffer et al. (1990) demonstrated that IFNa causes an increase in membrane associated phorbol ester binding sites and a stimulation of diacylglycerol production due to increased phos- phatidylcholine hydrolysis. In addition, Reich and Pfeffer (1990) reported that the protein kinase inhibitor staurosporine blocked the antiviral action of IFNa as well as the activation of ISGF3 and of IFNa-stimulated transcription, leading them to postulate a pivotal role for PKC in signaling. The inherent difficulties associated with the use of inhibitors in the complex environment of the cell have prompted us to directly examine the role of kinases in IFNa signaling. The data reported here demonstrate no involvement of PKC. We show that staurosporine and a second inhibitor of protein kinase activity, K-252a, completely abolished transcriptional induction by IFNa by preventing activation of the ISGF3a regulatory component. However, prolonged treatment of HeLa cells with phorbol ester, which led to a quantitative depletion of PKC immunoreactive material, had no effect on IFNa signaling. Activation of ISGF3a and transcriptional induction of ISGs were not appreciably affected by the absence of detectable PKC, nor was ISGF3a activation sensitive to an inhibitor with high specificity for PKC, calphostin C, an inhibitor with no other known cellular targets. Even in the absence of PKC, activation of ISGF3a was still sensitive to inhibition by staurosporine and K-252a, indicating that ISGF3 activation and IFNa-induced transcription are mediated by a kinase distinct from the classical PKC isoforms.

EXPERIMENTAL PROCEDURES

Cell Culture-HeLa cells (clone S3) were obtained from ATCC, Rockville, MD and were maintained in Dulbecco’s modified Eagle’s medium supplemented with 5% calf serum as monolayer cultures. Recombinant human IFNn-2a was a kind gift of Dr. Michael Brunda (Hoffmann-La Roche) and was added to culture media a t 500 IU/ml for 30 min. Staurosporine and calphostin C, obtained from Kamiya Biomedical Co. (Thousand Oaks, CA), K-252a from Calbiochem, and TPA from Sigma were resuspended in dimethyl sulfoxide prior to addition to culture media. Kinase inhibitors were added 15 min prior to addition of IFNn and remained throughout the IFNn treatment. As control, dimethyl sulfoxide alone was added to untreated samples.

Transcription Measurements-Run-on transcription experiments (Weber et a/., 1977) were performed as previously described (Decker et al., 1989) using approximately 10 million cells/point. DNA probes were as follows: pGEMl (Promega Biotech); &tubulin cDNA (Cleve- land et al., 1980); ISG54 EcoRI-TaqI fragment from exon 2 (Levy et al., 1986); and ISG15 Tag1 fragment from exon 2 (Reich et al., 1987). Autoradiograms were quantitated by laser densitometry, and expression of ISGs was normalized relative to tubulin transcription.

Cell Extracts, Gel Retardation Assays, and Immunoblotting-Nu- clear and cytoplasmic extracts (Dignam et al., 1983) were prepared and analyzed by gel retardation as previously described (Levy et al., 1988 Levy et al., 1989; Kessler et al., 1990). Levels of ISGF3n and ISGF3y in crude extracts were quantitated by mixing with highly enriched cytoplasmic fractions of complementary activities (provided by S. A. Veals, New York University School of Medicine), as described (Kessler et al., 1990). Detergent extracts for immunoblotting were prepared by lysing cells in 0.5% nonidet P-40 followed by removal of insoluble material by centrifugation.

Immunoblots were prepared following sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Laemmli, 1970) and transfer to nitrocellulose (Towbin et al., 1979) using 200 pg of crude protein extract per lane, as described (Harlow and Lane, 1988). Blots were probed

with an affinity-purified rabbit antipeptide serum, prepared against the sequence KLLNQEEGEYYNVPVAD derived from amino acids 276-292 of PKC-y (Coussens et al., 1986), kindly provided by Dr. A. Czernik (Rockefeller University). Blots were developed using a horse- radish peroxidase-labeled second antibody and enhanced chemiluminescence according to the manufacturer’s instructions (Amersham Corp.). Purified rat brain PKC (Woodgett and Hunter, 198’ib), a kind gift of Dr. A. Nairn (Rockefeller University), was used as positive control.

RESULTS

Protein Kinase Inhibitors Block Transcriptional Stimulation by IFNa-We and others have previously shown that inhibitors and activators of a variety of intracellular signaling systems have little or no effect on transcriptional stimulation of ISGs. Under a variety of experimental regimens, pertur- bation of cAMP levels using either dibutyryl cAMP or 8- bromo-CAMP, of Ca’+ concentrations using ionophores and chelating agents, and of protein kinase activities using inhibitors such as H7, H8, HA1004, and 2-aminopurine were found to have little or no effect on transcriptional activation of ISGs in HeLa S3 cells (Larner et al., 1984; Lew et al., 1989; Reich and Pfeffer, 1990) or on oligoadenylate synthetase gene transcription in human Daudi cells (Faltynek et al., 1989) or in mouse Balb/c 3T3 cells (Yan et al., 1989). On the other hand, treatment of cells with NaF dramatically reduced ISG expression. This inhibition was not due to inhibition of ISGF3 activation or other receptor-proximal events but rather to a block of nuclear translocation of ISGF3a, a later step in signal transduction (Levy et al., 1989; Kessler et al., 1990).

Treatment of HeLa cells with the potent protein kinase inhibitors staurosporine and K-252a prevented transcriptional induction of ISGs in a dose-dependent manner (Fig. 1A). Transcriptional measurements were made by pulse label- ing RNA synthesized in nuclei in vitro, and labeled RNA was hybridized to excess cDNA clones immobilized on nitrocellu-

A rn - I

SIaubM) IFN(r 10 +

100 +

500 +

500 -

K252r lnM)rn

10 +

100 +

500 +

500 -

TPA IFNa ”

+ + a 1

+ - I

FIG. 1. Inhibition of IFNa-stimulated gene transcription. A, HeLa S3 cells (1 X 10’ cells/point) were untreated (-) or treated as indicated (+) with 500 IU/ml IFNn for 30 min either alone or following a 15-min pretreatment with the indicated concentrations of either staurosporine (Stau) or K-252a. Following these treatments, nuclei were isolated and processed for in vitro nuclear run-on transcription assays. Labeled RNA was hybridized to plasmid DNA immobilized on nitrocellulose as shown in the key at the bottom of panel B. B, HeLa cells were pretreated with 40 nM TPA for 48 h and then analyzed for induction of transcription in response to IFNn.

IFNO Signal Transduction Pathway 23473

lose. In the absence of IFNa treatment, the ISG54 and ISG15 genes were transcriptionally silent. However, following a 30- min exposure of cells to 500 IU/ml IFNa, these genes were induced by 10-20-fold (upper panel). Cells pretreated with staurosporine or K-252a in increasing doses from 10 to 500 nM followed by IFNa treatment showed a marked inhibition of ISG transcription with no apparent effect on constitutive transcription of housekeeping genes such as tubulin (lower panels). Staurosporine treatment substantially inhibited the induced level of ISG54 transcription a t 100 nM and showed near complete inhibition of ISG54 and ISG15 at 500 nM. K- 252a treatment resulted in a marked diminution in IFNa induction of ISG54 transcription a t 100 nM with again nearly complete inhibition of ISG54 and ISG15 transcription a t 500 nM (see Table I). Neither drug had any discernible effect on ISG transcription in the absence of IFNa.

Staurosporine and K-252a Prevent Activation of ISGF3a- Cytoplasmic activation of ISGF3 is the earliest detectable intracellular event in the IFNa signaling pathway leading to ISG induction. We therefore examined the effect of kinase inhibitors on the stimulation of this transcription factor. As expected, neither cytoplasmic nor nuclear extracts of untreated HeLa cells contained active ISGF3 as detected by gel retardation assays (Fig. 2A, lane I ) . Following a 30-min exposure to IFNa, a time when ISG transcription is high (see Fig. lA), substantial amounts of ISGF3 were detected in both cytoplasmic and nuclear extracts (lane 2). However, extracts from cells treated with both IFNa and either staurosporine or K-252a showed diminished amounts of ISGF3. Treatment with 10 nM staurosporine markedly reduced both cytoplasmic and nuclear levels of active ISGF3 (lane 3) , and 500 nM showed complete inhibition (lane 5 ) . Likewise 500 nM K-252a also severely inhibited ISGF3 (lane 9). This dose-dependent diminution of ISGF3 activation paralleled the decline in transcriptional inducibility of ISGs.

To determine a t which step in ISGF3 activation these kinase inhibitors were acting, the effects of staurosporine and K-252a on the individual components of ISGF3 were determined (Fig. 2B). Levels of cytoplasmic ISGF3y in cells treated with inhibitors were quantitated by mixing cytoplasmic extracts with a highly enriched fraction of active ISGF3a (Kes- sler et al., 1990) and were assayed for ISGF3 activity by gel retardation. Neither IFNa (lane 2) nor staurosporine or K- 252a in combination with IFNa (lanes 3 and 4 ) significantly affected the constitutive level of ISGF3y. To quantitate the levels of ISGF30, extracts were mixed with a highly enriched fraction of ISGF3-y in amounts sufficient to saturate active ISGF3a (Kessler et al., 1990) and were assayed for ISGF3 activity (Fig. 2B). As expected, no active ISGF3a was detected in extracts of untreated cells (lane 6), whereas abundant amounts were readily detectable following IFNa treatment (lane 7). Significantly, both staurosporine and K-252a a t 500 nM prevented the activation of ISGF3a (lanes 8 and 9), concentrations of these inhibitors which also blocked ISG transcription (Fig. 1). In all mixing experiments, enriched

TABLE I Transcriptional induction following inhibitor plus IFNn treatment

relative to IFNn treatment alone

ISG54 1.0 0.2 0.2 0.8

ISG15 1.0 0.2 0.3 1.4 ” Stau, staurosporine; averages of three independent determina-

’’ Averages of two independent determinations. tions.

A Stau (nM) K252a (nM) ” 10 10050050010 100500500

IFNa: - + + + + - + + + - .,... .

*& . . f lii Cytoplasm

*. -

a- (I, Nucleus

*mubut, -*I*) 1 2 3 4 5 6 7 8 9 1 0

B +ISGF3a

ISCF3a

Extract: ’ 0 I S + I l + I - ‘ ISiF3y

+ISGF3

1 2 3 4 5 6 7 8 9 1 0 1 1

FIG. 2. Kinase inhibitors block activation of ISGF3a. A, nuclear and cytoplasmic extracts were prepared from untreated HeLa cells or from cells treated with 500 IU/ml IFNn for 30 min following a 15-min pretreatment with the indicated concentrations of staurosporine (Stau) or K-252a. Extracts were assayed for the presence of active ISGF3 by gel retardation assays using an oligonucleotide containing an ISRE sequence from the ISG15 gene. Only the portions of the autoradiograms showing relevant complexes are shown. Cyto- plasmic ISGF3 resolves as a doublet while nuclear ISGF3 appears as a single band running slower than a band due to a nonspecific protein- DNA complex (Kessler et al., 1990). R, the presence of ISGF37 component (lanes 1-4) and ISGF3n component (lanes 6-9) in cytoplasmic extracts was quantitated by mixing extracts with a highly enriched fraction containing excess amounts of the complementing factor, as described under “Experimental Procedures.” Cell extracts were derived from untreated cells (lanes I and 6 ) or from cells treated with IFNn for 30 min (lanes 2 and 7), with 500 nM staurosporine for 45 min and with IFNa for the last 30 min (lanes 3 and 8 ) or with K- 252a for 45 min and with IFNn for the last 30 min (lanes 4 and 9). Lanes 5 and 10 show the lack of DNA binding activity produced by the complementing fractions alone, while lane I I shows the maximal amount of ISGF3 activity that could be produced by mixing these fractions.

fractions of ISGF3a or ISGF3y were sufficient to saturate a much greater amount of the complementing activity than was present in the crude extracts (lane I 1 ).

Chronic Exposure of HeLa Cells to Phorbol Ester Has NO Effect on ISG Induction-Staurosporine and K-252a are inhibitors of a variety of protein kinases, including calcium ion and phospholipid-dependent protein kinases (Tamaoki et al., 1991; Kase et al., 1987) such as PKC, enzymes which are also sensitive to phorbol esters. PKC can be down-regulated by chronic treatment with phorbol ester (Woodgett and Hunter, 1987a) by a mechanism which leads to the complete loss of PKC protein from treated cells (Diaz-Laviada et al., 1990). Because PKC has been proposed to mediate the action of IFNa (Yap et al., 1986; Pfeffer et al., 1990; Reich and Pfeffer, 1990), we measured the transcriptional response of ISGs in cells treated for 48 h with 40 nM 12-0-tetradecanoylphorbol 13-acetate (TPA) to induce down-regulation of PKC (Wood- gett and Hunter, 1987a; Diaz-Laviada et al., 1990). In two independent experiments, the transcriptional induction of ISG15 and ISG54 in response to IFNa was minimally affected by TPA treatment, and TPA alone did not induce ISG transcription (Fig. 1B and Table I).

The effectiveness of TPA down-regulation of PKC was determined by immunoblotting. Postnuclear supernatants

23474 IFNa Signal Transduction Pathway IFNa: - + + -

“PA: - - + + 3 ” . ,.,.

I”, so+

49.5+

1 2 3 4 5

FIG. 3. Chronic treatment of HeLa cells with TPA leads to a total loss of PKC. Detergent extracts (200 wg/lane) from untreated HeLa cells (lane I ) or from cells treated with 500 IU/ml IFNn for 30 min (lanes 2 and 3 ) or with 40 nM TPA for 48 h (lanes 3 and 4 ) were immunoblotted with an affinity-purified rabbit polyclonal antipeptide serum specific for the major PKC isoforms and developed using enhanced chemiluminescence. Arrows at left indicate the migration of prestained molecular weight markers. Lane 5 shows the mobility of 10 ng of purified rat brain PKC. The slight difference in mobility between HeLa cell and rat PKC was found to be an artifact of the differential amounts of total protein loaded per lane by electrophoresis of a mixed sample (not shown). The major 60-kDa species present in the HeLa extracts is due to an unknown protein cross- reactive with the antipeptide serum.

from total detergent lysates of untreated HeLa cells or of HeLa cells treated with TPA, IFNa, or TPA plus IFNa were fractionated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose. These blots were probed with an affinity-purified anti-peptide serum specific for a conserved region of PKC (Parker et al., 1986; Coussens et al., 1986) which has been found to cross-react with the a-, p- , and y-isoforms of PKC (Fig. 3, lane 5).’ As shown in Fig. 3, a 78-kDa protein, most likely a mixture of PKC-a and -p , was detected in extracts from untreated (lane I ) or IFNa-treated cells (lane 2). However, no protein corresponding to the size of purified PKC (lane 5 ) was detected in extracts from cells treated with TPA or with TPA and IFNa (lanes 3 and 4 ) . The slight difference in mobility between crude and purified PKC was due to differential amounts of total protein loaded per lane. Therefore, the phorbol ester treatment regimen led to undetectable levels of PKC while leaving intact the IFNa signaling pathway leading to ISG transcription.

Phorbol Ester Treatment Has a Minimal Effect on ISGF3a and ISGF3y Levels-Lack of an effect of the loss of PKC on the response of ISGs prompted us to examine the levels of ISGF3 in TPA-treated cells. Chronic treatment of HeLa cells with TPA did not lead to an activation of ISGF3 (Fig. 4A, lane 3 ) nor did it cause a diminution in the level of nuclear ISGF3 activated in response to IFNa treatment (lane 4 ) . In contrast, in the complete absence of detectable PKC in the TPA-treated cells, both staurosporine and K-252a completely abolished IFNa-mediated activation of ISGF3 (lunes 5 and 6 ) . Therefore, it seems highly unlikely that these inhibitors are exerting their effects on IFNa signaling through blocking the activity of a classical PKC isoform.

A modest reduction in active ISGF3 in TPA-treated cells was observed in cytoplasmic extracts even though nuclear levels of ISGF3 and transcription of ISGs were unaffected (Fig. 4A) . This reduction was due to a small decrease in the levels of both cytoplasmic ISGF3y (Fig. 4B) and of cytoplasmic ISGF3a (Fig. 4C). Gel retardation analysis following mixing with exogenous sources of the complementary component revealed a 2.5-fold decrease in the constitutive level of cytoplasmic ISGF3y (Fig. 4B, lane 3 ) , which was not further affected by IFNa, staurosporine, or K-252a (lanes 4-6) . Like-

A. Czernik, unpublished data.

A T T T

Extract 0 I T I S+IK+I .- T . . y . .c -.- R

+ + +

e -

I

Nucleus

B +ISCF3a I T T T I

Extract: 0 I T I S+IK+I - + + +

1 2 3 4 5 6 7

C +IsGF3y I T T T 1 I s m a

+ + + + Extract: 0 I T I S+I K+I - ISGF3y

1 2 3 4 5 6 7 8

FIG. 4. TPA down-regulation of PKC does not affect ISGF3 activation. Extracts from untreated HeLa cells (0) or from cells treated with 500 IU/ml IFNa for 30 min ( I ) , with 40 nM TPA for 48 h (T), with 500 nM staurosporine for 45 min (S), with 500 nM K- 252a for 45 min ( K ) , or with a combination of these agents, as indicated, were assayed for ISGF3 (panel A ), for ISGF3-y by addition of partially purified ISGF3a (panel B ) , or for ISGF3n by addition of partially purified ISGF3-y (panel C). For all treatments, kinase inhibitors were added during the final 45 min prior to cell harvest, and IFNa was added for the final 30 min. Panel A shows results for both cytoplasmic and nuclear extracts, as indicated, while panels B and C show results only for cytoplasmic extracts.

wise, TPA treatment resulted in a 2-fold reduction in activated cytoplasmic ISGF3a (Fig. 4C, lane 4 ) , which was completely abolished by both staurosporine and K-252a (lanes 5 and 6 ) .

Recently, a polycyclic compound isolated from Cladospor- ium cladosporioides has been shown to be a potent and highly specific inhibitor of PKC (Kobayashi et al., 1989). This compound, calphostin C, interacts with the regulatory domain of PKC and inhibits phorbol dibutyrate binding and subsequent activation of the kinase function. Treatment of HeLa cells with up to 500 nM calphostin C, a concentration known to be inhibitory in HeLa S3 cells (Kobayashi et al., 1989), showed no effect on nuclear ISGF3. However, in accord with an inhibition of PKC equivalent to complete down-regulation of the enzymes following TPA treatment, calphostin C treatment led to a similar partial reduction in cytoplasmic, but not nuclear, ISGF3a and ISGF3-y (data not shown). These modest effects on cytoplasmic ISGF3 levels by disruption of PKC

IFNa Signal Transduction Pathway 23475

function with no corresponding effects on nuclear ISGF3 or on ISG transcriptional induction may indicate a secondary role for PKC in regulating steady-state levels or subcellular distribution of ISGF3 precursor proteins.

DISCUSSION

The precise mechanisms of signal transduction following interaction of cell surface receptors with polypeptide ligands have remained elusive for many mammalian systems. The rapid transcriptional response of cells exposed to IFNa with no requirement for new protein synthesis suggests that RNA synthesis is the earliest anabolic response to IFNa signaling. This transcriptional response relies on the activation of a set of pre-existing, cytoplasmic proteins which subsequently translocate to the nucleus and bind IFN-responsive enhancer elements. This mechanism solves some of the problems inherent in signal transduction by providing a direct link for information transfer from the cytoplasmic compartment to the nucleus. One remaining question is how the receptor- ligand interaction brings about the activation of ISGF3.

An attractive hypothesis for transcription factor activation is phosphorylation. Many transcription factors have been found to be phosphoproteins, and in some cases phosphorylation has a direct effect on activity (Sorger et al., 1987; Larson et al., 1988; Prywes et al., 1988; Bagchi et al., 1989). For example, phosphorylation of the CAMP response element- binding protein by protein kinase A is believed to be the primary regulatory event leading to increased transcription of CAMP-activated genes (Yamamoto et al., 1988; Gonzalez et al., 1989; Mellon et al., 1989). In addition to kinase domains intrinsic to cell surface receptors (Ullrich and Schlessinger, 1990), cellular serine/threonine kinases have been described in many signaling pathways. The MAP2 kinase has been implicated in the signal transduction pathway linked to the insulin receptor (Boulton et al., 1990; Anderson et al., 1990), and a role for the c-raf kinase has been suggested for several pathways linking cell surface events with transcriptional re- sponses (Druker et al., 1989; Siegfried and Ziff, 1990). Cyto- plasmic activation of the transcription factor NFKB involves phosphorylation of an inhibitor (Shirakawa and Mizel, 1989; Ghosh and Baltimore, 1990). A kinase has also been implicated in IFNy-induced gene expression (Lew et al., 1989); and in some cell lines the drug 2-aminopurine, which may be acting as a kinase inhibitor, has been found to affect IFNa- induced genes (Tiwari et al., 1988; Wathelet et al., 1989).

We have found that certain inhibitors of cellular kinases prevent the transcriptional response to IFNa. Both staurosporine and K-252a in nanomolar concentrations led to abla- tion of IFNa-induced gene expression. Previously, other kinase inhibitors have been reported to affect accumulation of ISG mRNA without a direct effect on transcription rate, indicating that kinases may be involved in post-transcriptional events (Lew et al., 1989; Faltynek et al., 1989). Stauros- porine and K-252a, in contrast, directly affected transcriptional initiation. This inhibition of IFNa-induced transcription was associated with a lack of activated ISGF3, caused by inhibition of the activation of the regulatory component ISGF3a. The slight difference in dose-response observed between ISG15 and ISG54 transcription presumably reflects their relative affinities for activated ISGF3 (Kessler et al., 1988a).

Although effects of these kinase inhibitors are often consid- ered indicative of a role for PKC, an enzyme previously proposed to mediate IFNa signaling (Yap et al., 1986; Pfeffer et al., 1990; Reich and Pfeffer, 1990), we found no evidence for an involvement of PKC in IFNa-stimulated transcription.

First, chronic treatment of cells with TPA neither activated ISGs nor inhibited their activation in response to IFNa even though this TPA treatment led to a loss of immunoreactive PKC isoforms. In particular, the 0-isoform of PKC, previously suggested to be involved in IFNa signaling (Pfeffer et al., 1990), was depleted by TPA treatment. Second, calphostin C, a kinase inhibitor specific for the phorbol ester-binding site in the regulatory subunit of PKC, did not inhibit IFNa activation of ISGF3. It is unlikely that further members of the PKC family exist in HeLa cells which lack both a phorbol ester-binding site and the highly conserved epitope recognized by this antibody. PKC may indeed be involved in IFNa action, but it must act at a post-transcriptional step rather than directly on signal transduction. Several other kinases, in particular CAMP- and cGMP-dependent kinases, do not ap- pear to be involved in IFNa signaling (Lew et al., 1989; Larner et al., 1984; Yan et al., 1989). We have been unable to activate ISGF3a in vitro using the catalytic domain of purified CAMP- dependent protein kinase or using lipid-activated, purified PKC (data not shown). In addition, it is unlikely that kinases linked to other cell surface receptors whose ligands have no effect on the IFNa response (e.g., c-raf or MAP2 kinase) would mediate ISG induction (Levy and Darnell, 1990). These results suggest that a novel protein kinase is required for ISGF3 activation.

The involvement of a kinase in ISGF3 activation could result from two distinct mechanisms. The latent, cytoplasmic form of ISGF3a may require a critical phosphorylation for activation provided by an IFNa-regulated kinase. Although the IFNa receptor does not possess an intrinsic kinase domain (Uz6 et al., 1990), a kinase could directly associate with the IFNa receptor and could be activated by receptor aggregation or by some other ligand-induced allosteric event. Phosphoryl- ation of ISGF3a could lead to conformational changes, un- covering protein interaction domains required for nuclear targeting and for association with ISGF3y. By this model, phosphorylation would be the primary regulatory event linking receptor occupation with an active transcriptional response.

Alternatively, inhibition of ISGF3 activation by staurosporine and K-252a may reflect the requirement for a constitutive kinase to maintain ISGF3a in a receptive state. By this model, both latent and active ISGF3a would be phosphoproteins, and phosphorylation would be required for activity. The steady-state level of phosphorylation of these proteins would reflect a balance between constitutive kinases and phospha- tases (Cohen, 1989), with inhibition of kinase activity yielding largely dephosphorylated ISGF3a. This dephosphorylated form, since intrinsically nonfunctional, would no longer be responsive to receptor-mediated activation, which could be some other, possibly non-kinase, activity. Preliminary experiments have indicated that calf intestinal phosphatase and potato acid phosphatase inhibit the DNA-binding activity of ISGF3, suggesting that this factor contains phosphoamino acids critical for activity."

While this work was in progress, Hannigan and Williams (1991) reported effects of perturbing arachidonic acid metab- olism on subsequent ISG expression. Although they did not directly measure transcriptional effects, they reported inhibition of a nuclear, ISRE-binding factor likely to be the mouse equivalent of ISGF3. Therefore, if the kinase implicated by staurosporine and K-252a inhibition should mediate a constitutive rather than regulatory event, possibly an arachidonic acid metabolite could prove to be the ultimate activator of ISGF3a. Evidence of arachidonic acid metabolite involvement

I S. A. Veals and D. E. Levy, unpublished data.

23476 IFNa Signal ~ ~ a n s d u c ~ i o ~ Pathway

in IFNa signaling must be viewed with caution, however, since it has been shown that arachidonic acid analogues have direct effects on the number of IFNa cell surface receptors available for binding Iigand rather than on intracellular signaling pathways (Menon et al., 1990). The recent purification of the subunit polypeptides of ISGF3a should allow us to directiy ascertain the phosphorylatio~ state of latent and activated forms and thereby determine what post-transla- tional modific~tion mediates ISGF3 activation.

Acknowledgments-We thank Drs. A. Nairn and A. Czernik for gifts of purified PKC and specific antisera, respectively, and for helpful discussion; and Dr. M. Brunda for gifts of IFNa. We also thank M. Picciotto for helpful advice and members of the lab for comments on the manuscript.

Note Added in Proof-We have found that TPA treatment of HeLa cells also depleted the t-isoform of PKC.

REFERENCES

Anderson, M. G., Maller, J. L., Tonks, N. K., and Sturgill, T. W. (1990) Nature 343 , 651-653

Bagchi, S., Raychaudhuri, P., and Nevins, J. R. (1989) Proc. Natl. Acad, Sci. U. S. A. 86,4352-4356

Bandyopadhyay, S. K., Kalvakolanu, D. V. R., and Sen, G. C. (1990) Mol. Cell. Eiol. 10, 5055-5063

Boulton, T. G., Yancopoulos, G. D., Gregory, J. S., Slaughter, C., Moomaw, C., Hsu, J., and Cobb, M. H. (1990) Science 249,64-67

Cleveland, D. W., Lopata, M. A., MacDonald, R. J., Cowan, N. J., Rutter, W. J., and Kirschner, M. W. (1980) Cell 20,95-105

Cohen, P. (1989) Annu. Rev. Biochem. 58,453-508 Cohen, B., Peretz, D., Vaiman, D., Benech, P., and Chebath, J. (1988)

Coussens, L., Parker, P. J., Rhee, L., YangFeng, T. L., Chen, E., Waterfield, M. D., Francke, U., and Ullrich, A. (1986) Science 233 ,

Dale, T. C., Imam, A. M. A,, Kerr, I. M., and Stark, G. R. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 1203-1207

Decker, T., Lew, D. J., Cheng, Y.-S. E., Levy, D. E., and Darnell, J. E., Jr. (1989) EMBO J. 8,2009-2014

De Maeyer, E., and De Maeyer-Guignard, J. (1988) Interferons and Other Re~ulatory Cytokines, John Wiley & Sons, New York

Diaz-Laviada, I., Larrodera, P., Diaz-Meco, M. T., Corney, M. E., Guddal, P. H., Johansen, T., and Moscat, J. (1990) EMBO J. 9,

Dignam, J. D., Lebovitz, R. M., and Roeder, R. G. (1983) Nucleic

Druker, B. J., Mamon, H. J., and Roberts, T. M. (1989) N. En& J.

Faltynek, C. R., Princler, G. L., Gusella, G. L., Varesio, L., and

Friedman, R. L., and Stark, G. R. (1985) Nature 314,637-639 Friedman, R. L., Manly, S. P., McMahon, M., Kerr, I. M., and Stark,

G. R. (1984) Cell 38, 745-755 Fu, X.-Y,, Kessler, D. S., Veals, S. A,, Levy, D. E., and Darnell, J. E.,

Jr. (1980) Proc. Natl. Acad. Sci. U. S. A. 87,8555-8559 Ghosh, S., and Baltimore, D. (1990) Nature 344,678-682 Gonzalez, G. A., Yamamoto, K. K., Fischer, W. H., Karr, D., Menzel,

P., Biggs, W., 111, Vale, W. W., and Montminy, M. R. (1989) Nature

Hannigan, G. E., and Williams, B. R. G. (1991) Science 2 6 1 , 204-

Harlow, E., and Lane, D. (1988) Antibodies. A Laboratory Manual,

Hubbell, H. R., Boyer, 3. E., Roane, P., andBurch, R. M. (1991) Proc.

Kase, H., Iwahashi, K., Nakanishi, S., Matsuda, Y., Yamada, X., Takahashi, M., Murakata, C., Sato, A,, and Kaneko, M. (1987) Biochem. Biophys. Res. Commun. 142,436-440

EMBO J . 7, 1411-1419

859-866

3907-3912

Acids Res. 11 , 1475-1489

Med. 321,1383-1391

Radzioch, D. (1989) J. BioL Chem. 2 6 4 , 14305-14311

337 , 749-752

207

Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Natl. Acad. Sei. U. S. A. 88,906-910

Kessler, D. S., Levy, D. E., and Darnell, J. E., Jr. (1988a) Roc. Nutl.

Kessler. D. S., Pine. R., Pfeffer. L. M., Levy, D. E., and Darnell, J. Acad. Sci. U. S. A. 85,8521-8525

. - . E., Jrl (1988b) EA&O J. 7, 3779-3783

Kessler. D. S.. Veals. S. A.. Fu. X.-Y., and Levy. D. E. (1990) Genes , .

& Del. 4 , 1753-1765

Biochem. BioBhys. Res. Commun. 169,548-553

- .

Kobayashi, E., Nakano, H., Morimoto, M., and Tamaoki, T. (1989)

Kusari, J., andsen, G. C. (1987) Mol. Cell. Biol. 7 , 528-531 Laemmli, U. K. (1970) Nature 227,680-685 Larner, A. C., Jonak, G., Cheng, Y.-S. E., Korant, B., Knight, E., and

Darnell, J. E., Jr. (1984) Proe. Natl. Acad. Sci. U. S. A. 8 1 , 6733- 6737

Larner. A. C.. Chaudhuri. A.. and Darnell. J. E., Jr. (1986) J. Biol. I .

Chem. 261; 453-459 . .

Larson, J. S., Schuetz, T. J., and Kingston, R. E. (1988) Nature 335, 372-375

Levy, D. E., and Darnell, J. E., Jr. (1990) New Biol. 2,923-928 Levy, D. E., Larner, A. C., Chaudhuri, A., Babiss, L. E., and Darnell,

J. E., Jr. (1986) Proc. Natl. Acad. Sei. U. S. A. 83,8929-8933 Levy, D. E., Kessler, D. S., Pine, R., Reich, N., and Darnell, J. E., Jr.

(1988) Genes & Dev. 2,383-393 Levy, D. E., Kessler, D. S., Pine, R., and Darnell, J. R., Jr. (1989)

Genes & Deu. 3,1362-1371 Lew, D. J., Decker, T., and Darnell, J. E., Jr. (1989) Mol. Cell. Biol.

Mellon, P. L., Clegg, C, H., Correll, L. A., and McKnight, G. S. (1989)

Menon, S. D., Yap, W. H., Lim, A., and Tan, Y. H. (1990) Lipids 25,

Parker, P. J., Coussens, L,, Totty, N., Rhee, L., Young, S., Chen, E., Stabel, S., Waterfield, M. D., and Ullrich, A. (1986) Science 2 3 3 ,

Pfeffer, L. M., Struloviei, B., and Saltiel, A. R. (1990) Proc. Natl.

Porter, A. C. G., Chernajovsky, Y., Dale, T. C., Gilbert, 6. S., Stark,

Prywes, R., Dutta, A., Cromlish, J. A,, and Roeder, R. G. (1988) Proc.

Reich, N., and Darnell, J. E., Jr. (1989) Nucleic Acids Res. 17,3415-

Reich, N. C., and Pfeffer, L. M. (1990) Proc. Natl. Acad. Sci. U. S. A.

Reich, N., Evans, B., Levy, D. E., Fahey, D., Knight, E., and Darnell,

Rutherford. M. N.. Hanniean. G. E., and Williams, B. R. G. (1988)

9,5404-5411

Proc. Natl. Acad. Sci. U. S. A. 8 6 , 4887-4891

321-327

853-859

Acad. Sci. U. S. A. 8 7 , 6537-6541

G. R., and Kerr, I. M. (1877) EMBOJ. 7,85-92

Natl. Acad. Sci. U. S. A. 85,7206-7210

3424

87,8761-8765

J . E., Jr. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,6394-6398 ~. " .

EMBO J: 7,751"759 Shirakawa, F., and Mizel, S. B. (1989) Mol. Cell. Biol. 9,2424-2430 Shiravoshi. Y.. Burke, P. A., Apella, E., and Ozaro, K. (1988) Proc.

Nail. Acad. Sci. U. S. A. 85, 5884-5888 Siegfried, Z., and Ziff, E. B. (1990) Mol. Cell. Biol. 10 , 6073-6078 Sorger, P. K., Lewis, M. J., and Pelham, H. R. B. (1987) Nature 329,

Tamaoki, T., Nomoto, H., Takahashi, I., Kato, Y., Morimoto, M., and Tomita, F. (1986) Biochem. Biophys. Res. Commun. 135,397- 402

Tiwari, R. K., Kusari, J., Kumar, R., and Sen, G. C. (1988) Mol. Cell.

Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad.

Ullrich, A., and Schlessinger, J. (1990) Cell 61,203-212 Uz6, G., Lutfalla, G., and Gresser, I. (1990) Cell 6 0 , 225-234 Wathelet, M. G., Claws, I. M., Paillard, F, C., and Huez, G. A. (1989)

Weber, J., Jelinek, W., and Darnell, J. E., Jr. (1977) 10, 611-616 Woodgett, J . R., and Hunter, T. (1987a) Mol. Cell. Biol. 7, 85-96 Woodgett, J. R., and Hunter, T. (1987b) J. Biol. Chem. 262,4836-

Yamamoto. K. K.. Gonzatez. G. A.. Bigas, W. H., 111, and Montminy,

81-84

Bid. 8,4289-4294

Sci. U. S. A. 76,4350-4354

Eur. J. Biochem. 184,503-509

4843 ~.

M. R. (1988) Nhture 334; 494-498" Yan, C., Sehgal, P. B., and Tamm, I. (1989) Proc. Natl. Acad. sei.

Yap, W. H., Teo, T. S., and Tan, Y. H. (1986) Science 234,355-358 U. S. A. 86,2243-2247

protein kinase activity required for an early step in ... fileprotein kinase activity required for...

Documents