Báo cáo khoa học: Endogenous expression and protein kinase A-dependent phosphorylation of the guanine nucleotide exchange factor Ras-GRF1 in human embryonic kidney 293 cells

Endogenous expression and protein kinase A-dependent phosphorylation of the guanine nucleotide exchange factor Ras-GRF1 in human embryonic kidney 293 cells Jens Henrik Norum1, Trond Me´ thi1, Raymond R. Mattingly2 and Finn Olav Levy1

1 Department of Pharmacology, University of Oslo, Norway 2 Department of Pharmacology, Wayne State University, Detroit, MI, USA

Keywords 5-HT7, cAMP, ERK, GEF, serotonin

Correspondence F. O. Levy, Department of Pharmacology, University of Oslo, PO Box 1057 Blindern, N-0316 Oslo, Norway Fax: +47 22840202 Tel: +47 22840237 or +47 22840201 E-mail: f.o.levy@medisin.uio.no

(Received 2 December 2004, revised 1 February 2005, accepted 10 March 2005)

doi:10.1111/j.1742-4658.2005.04658.x

We have previously reported the Ras-dependent activation of the mitogen- activated protein kinases p44 and p42, also termed extracellular signal- regulated kinases (ERK)1 and 2 (ERK1 ⁄ 2), mediated through Gs-coupled serotonin receptors transiently expressed in human embryonic kidney (HEK) 293 cells. Whereas Gi- and Gq-coupled receptors have been shown to activate Ras through the guanine nucleotide exchange factor (GEF) called Ras-GRF1 (CDC25Mm) by binding of Ca2+ ⁄ calmodulin to its N-terminal IQ domain, the mechanism of Ras activation through Gs-cou- pled receptors is not fully understood. We report the endogenous expres- sion of Ras-GRF1 in HEK293 cells. Serotonin stimulation of HEK293 cells transiently expressing Gs-coupled 5-HT7 receptors induced protein kinase A-dependent phosphorylation of the endogenous human Ras-GRF1 on Ser927 and of transfected mouse Ras-GRF1 on Ser916. Ras-GRF1 overexpression increased basal and serotonin-stimulated ERK1 ⁄ 2 phos- phorylation. Mutations of Ser916 inhibiting (Ser916Ala) or mimicking (Ser916Asp ⁄ Glu) phosphorylation did not alter these effects. However, the deletion of amino acids 1–225, including the Ca2+ ⁄ calmodulin-binding IQ domain, from Ras-GRF1 reduced both basal and serotonin-stimulated ERK1 ⁄ 2 phosphorylation. Furthermore, serotonin treatment of HEK293 cells stably expressing 5-HT7 receptors increased [Ca2+]i, and the sero- tonin-induced ERK1 ⁄ 2 phosphorylation was Ca2+-dependent. Therefore, both cAMP and Ca2+ may contribute to the Ras-dependent ERK1 ⁄ 2 acti- vation after 5-HT7 receptor stimulation, through activation of a guanine nucleotide exchange factor with activity towards Ras.

Introduction

Abbreviations 5-HT, 5-hydroxytryptamine (serotonin); CaM, calmodulin; EGF, epidermal growth factor; Epac, exchange protein directly activated by cAMP; ERK, extracellular signal-regulated kinase; GEF, guanine nucleotide exchange factor; GPCR, G protein-coupled receptor; H89, N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide dihydrochloride; HEK, human embryonic kidney; HRP, horseradish peroxidase; MAP, mitogen-activated protein; MEK, mitogen-activated protein ⁄ extracellular signal-regulated kinase kinase; PKA, protein kinase A; Sos1, son of sevenless 1.

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2304

the Raf extracellular signal-regulated kinase (ERK) – cascade. The serine ⁄ threonine kinases ERK1 and ERK2 are activated by dual phosphorylation by the MAP kinase kinase, MEK, which becomes phosphorylated and activated by MEK kinases of family. All three Raf isoforms [A-Raf, B-Raf and Raf-1 (C-Raf)] Signals mediated through receptor tyrosine kinases [1] and G-protein-coupled receptors (GPCRs) can induce the activation of intracellular cascades such as the mitogen-activated protein (MAP) kinase – also called

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

140 kDa), have been identified [17,19]. The smaller iso- forms correspond to N-terminal deletions of the full- length 140 kDa protein. The physiological role of the guanine nucleotide exchange activity of the truncated forms is not known as they are missing the Ca2+ ⁄ CaM-binding IQ domain that is involved in the activa- tion of Ras-GRF1.

receptor 5-hydroxytryptamine7

expressed in mammalian cells may become activated by members of the Ras family of small G proteins. The activity of Ras proteins is under tight control of several classes of guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins. Mammalian Son of sevenless 1 (Sos1) is a ubiquitous Ras GEF and activates Ras following the stimulation of receptor tyrosine kinases, e.g. the epidermal growth factor (EGF) receptor [1]. Various GPCRs can also induce Ras activation via several classes of GEFs [2,3]. Activation of phospholipase C through Gq-coupled receptors, with subsequent increased levels of inositol- 1,4,5-trisphosphate, diacylglycerol and free intracellular Ca2+, can activate Sos1 through a cascade that includes the proline-rich tyrosine kinase, Pyk2, Src and Grb2 [4,5], as well as Ras GEFs of the Ras-GRP (cal- DAG-GEF) family, through binding of Ca2+ ⁄ calmo- [6]. Ras-GRF1, also dulin (CaM) and diacylglycerol called CDC25Mm [7,8], is another major GEF with activity towards Ras. Ras-GRF1 mediates activation of Ras subsequent to the stimulation of Gi- and Gq- coupled receptors [8,9].

Stimulation of all the splice variants of the Gs-coupled (5-HT7) serotonin increases intracellular levels of the second messenger cAMP [20], resulting in several intracellular effects, e.g. activation of cAMP-dependent protein kinase (PKA) and exchange proteins directly activated by cAMP (Epacs), GEFs specific for Rap [21]. In rat adrenal glo- merulosa cells, stimulation of the 5-HT7 receptor also induces the increased free intracellular Ca2+ concentra- tion ([Ca2+]i) through the low-voltage-activated T-type Ca2+ channels [22,23]. We have recently shown that serotonin treatment of human embryonic kidney (HEK)293 cells transiently expressing either one of the Gs-coupled serotonin receptors 5-HT4(b) or 5-HT7(a) induces ERK1 ⁄ 2 phosphorylation [24]. Although both Ras and Rap1 were activated, only Ras was involved in the pathway inducing ERK1 ⁄ 2 phosphorylation, which also involved Raf1 and MEK downstream of Ras. How- ever, in PC12 cells, 5-HT7-mediated Ca2+-independent and N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinoline- sulfonamide dihydrochloride (H89)-insensitive ERK1 ⁄ 2 phosphorylation has been reported to be enhanced by the overexpression of Epac and mimicked by a cAMP analogue stimulating both PKA and Epac, but not by an Epac-specific cAMP analogue [25]. The differences in H89 sensitivity and possible signalling pathways involved may reflect cell-type variations in the ERK1 ⁄ 2 phosphorylation mediated through Gs-coupled sero- tonin receptors.

expression of show endogenous

the Ras-GEF activity of Ras-GRF1,

The main mechanism for the activation of Ras- GRF1 is the binding of Ca2+ ⁄ CaM to the N-terminal IQ motif [10]. We have previously shown that the treatment of NIH3T3 and COS-7 cells with carbachol [9] and lysophosphatidic acid [11], activating both Gq- and Gi-coupled receptors, induces the activation and phosphorylation of Ras-GRF1. Furthermore, Ras-GRF1 is also heavily phosphorylated upon agon- ist activation of GPCRs, but the exact role of these phosphorylations is not fully understood. Protein kin- ase A (PKA) is one of probably several kinases that can induce the phosphorylation of Ras-GRF1 [12,13]. The residues Ser916 and Ser898 in the mouse and rat sequences, respectively, are homologous PKA phos- phorylation sites [14]. Although forskolin-induced phosphorylation of Ser916 is not sufficient to activate wild-type Ras-GRF1, a recombinant version of Ras- GRF1, with a mutated phosphorylation site (Ser916- Ala), has been shown to have reduced activity towards Ras both in vitro [12] and in an assay of Ras-dependent outgrowth of neurites from PC12 cells [14]. These results indicate that even though phos- phorylation of Ser916 may contribute to stimulation of cAMP- dependent phosphorylation alone is not sufficient to activate Ras-GRF1.

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2305

Ras-GRF1 is mainly expressed in brain tissue [15– 17], but expression of Ras-GRF1 mRNA has also been reported in some other tissues and non-neural cell lines [18]. Several murine Ras-GRF1 cDNAs, encoding (from 54 to proteins of different molecular mass The mechanism of Ras activation through Gs-cou- pled receptors is not fully understood. In the present study, we the Ca2+-dependent 140 kDa and shorter isoforms of Ras- GRF1 in HEK293 cells, as well as cAMP ⁄ PKA- dependent phosphorylation of Ras-GRF1 associated with ERK1 ⁄ 2 phosphorylation following stimulation transfected 5-HT7 receptors. However, mutating of Ser916 of Ras-GRF1 to alanine, aspartic acid or gluta- mic acid did not alter the Ras-GRF1-induced ERK1 ⁄ 2 phosphorylation. We confirm 5-HT7-mediated [Ca2+]i increase and show Ca2+ dependence of serotonin- induced ERK1 ⁄ 2 phosphorylation and a mandatory role of the Ca2+ ⁄ CaM-binding IQ domain in Ras- GRF1-stimulated ERK1 ⁄ 2 phosphorylation. Thus, both cAMP and Ca2+ may contribute to Ras-depend- ent ERK1 ⁄ 2 activation following stimulation of the

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

5-HT7 receptor, by activating a guanine nucleotide exchange factor with activity towards Ras.

anti-(Ras-GRF1) Ig (Fig. 2A, right panel). Preabsorb- ing the Ras-GRF1 antibody with a blocking peptide prevented the antibody from recognizing any of the Ras-GRF1 isoforms (data not shown).

Results

cDNA to HEK293 cell mRNA was used as the sub- strate in PCR reactions, as described in the Experi- mental procedures. The primer pairs specific for the HEK293 cells express the guanine nucleotide exchange factor Ras-GRF1

A

B

C

The guanine nucleotide exchange factor Ras-GRF1 is mainly expressed in neurones of the central nervous system, although it has also been reported to be expressed in some other tissues [18,26]. To investigate whether Ras-GRF1 plays a role in the activation of Ras ⁄ ERK signalling in HEK293 cells, we first used immunoprecipitation analysis and RT-PCR to detect whether Ras-GRF1 protein and mRNA, respectively, were expressed in our HEK293 cells. The proteins im- munoprecipitated from HEK293 cell lysates, by using a polyclonal antibody raised against a peptide mapping to the C terminus of the rat Ras-GRF1 sequence, were separated on SDS ⁄ PAGE (6% gel) and visualized on western blots probed with another polyclonal antibody raised against the C terminus of the human Ras-GRF1 sequence. A protein of (cid:1) 140 kDa was detected in immunoprecipitates from HEK293 cells (Fig. 1A), but immunoprecipitations. In was not present in control whole-cell lysates from HEK293 cells, both full-length 140 kDa Ras-GRF1 and shorter isoforms, of (cid:1) 110, 95 and 60 kDa, were detected on western blots with

D

A

B

Fig. 1. Human embryonic kidney (HEK)293 cells express Ras-GRF1. (A) Paramagnetic beads coated with anti-(Ras-GRF1) Ig (# sc-224) were used to immunoprecipitate Ras-GRF1 from the HEK293 cell lysate. The precipitated proteins were separated on 6% SDS ⁄ PAGE and electroblotted over to poly(vinylidene difluoride) membranes before probing with polyclonal Ras-GRF1 antibodies (# sc-863). (B) cDNA produced from mRNA isolated from HEK293 cells was used as the substrate in RT-PCR with primer pairs specific for human Ras-GRF1. Primer pairs: lane 2, ON359 and ON360; lane 3, ON357 and ON358; and lane 4, ON361 and ON360. The PCR prod- ucts and a DNA size marker, lane 1, were separated on agarose gels. The expected sizes of the PCR products, in bp, are indicated to the right.

Fig. 2. Serotonin induces phosphorylation of Ras-GRF1 through the 5-hydroxytryptamine7(a) (5-HT7(a)) receptor, and HA-Ras-GRF1 indu- ces extracellular signal-regulated kinase (ERK)1 ⁄ 2 activation. Human embryonic kidney (HEK)293 cells cotransfected with 5-HT7(a) receptor and empty or HA-Ras-GRF1 vector, as indicated, were treated with vehicle or 10 lM serotonin for the indicated peri- ods of time. The control, C, was treated with vehicle (10 lM HCl) for 5 min. Proteins were separated on 6% (A, B and D) or 10% (C) SDS ⁄ PAGE and electroblotted over to poly(vinylidene difluoride) membranes before probing with antibodies. (A) Western blots were probed with phosphospecific Ras-GRF1 (pRas-GRF1; left panel) or anti-(Ras-GRF1) Igs (Ras-GRF1; right panel) to confirm equal load- ing. (B) Western blot of cell lysates of HEK293 cells cotransfected with the 5-HT7(a) receptor and HA-Ras-GRF1 were incubated with anti-(pRas-GRF1) Ig (upper panel) or anti-(Ras-GRF1) Ig (lower panel) to confirm equal loading. (C) The same samples as in (A) and (B), but separated on 10% SDS ⁄ PAGE, were probed with phosphospe- cific ERK1 ⁄ 2 antibodies (pERK1 ⁄ 2; upper panel) and subsequently with ERK1 ⁄ 2 antibodies (ERK1 ⁄ 2; lower panel), to confirm equal (D) Non-transfected HEK293 cells were treated with or loading. without 10 nM epidermal growth factor (EGF) for 5 min. The pro- teins were separated, blotted and probed with antibodies as in (A).

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2306

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

phosphorylation of endogenous Ras-GRF1 in rat brain [9].

Serotonin-induced phosphorylation of Ras-GRF1 is dependent on cAMP and PKA

cyclase activity increases

stimulated Ras-GRF1 human Ras-GRF1 nucleotide sequence (NM_002891, GI:24797098) gave PCR products of expected size (Fig. 1B). The primer sequences are located at the 5¢ end (ON361, ON360 and ON359) and in the middle (ON357 and ON358) of the human Ras-GRF1 nucleo- tide sequence. Sequencing of the purified PCR prod- ucts confirmed sequence identity with cDNA encoding the human 140 kDa Ras-GRF1 (data not shown). Taken together, these mRNA and protein data demon- strate that the full-length 140 kDa Ras-GRF1 protein is endogenously expressed in the HEK293 cells used for this study, and that truncated forms of Ras-GRF1 may also be present.

Serotonin induces phosphorylation of Ras-GRF1 through 5-HT7 receptors

in reduced phosphorylation was

The serine residue at position 916 in mouse Ras-GRF1 is a PKA phosphorylation site both in vitro [12] and in vivo [14], and the corresponding human residue is serine 927. We therefore used a polyclonal antibody that was generated against a synthetic phosphopeptide analogous to the Ser916 phosphorylation site, and which has previously been shown to recognize mouse and rat Ras-GRF1 when they are phosphorylated at this residue [14], to test whether serotonin may stimu- late phosphorylation of Ras-GRF1 in HEK293 cells that express 5-HT7 receptors. HEK293 cells transfected with 5-HT7(a) receptors alone, or cotransfected with the HA-tagged mouse Ras-GRF1 (HA-Ras-GRF1), were treated with 10 lm serotonin for the indicated periods of time (Fig. 2). Serotonin treatment increased the phosphorylation of the endogenous 140 kDa and (cid:1) 60 kDa isoforms of Ras-GRF1 (Fig. 2A, left panel) and of recombinant HA-Ras-GRF1 (Fig. 2B, upper panel). Phosphorylation of ERK1 ⁄ 2 in the same sam- ples was fully induced after 3 min of treatment with 10 lm serotonin (Fig. 2C). Furthermore, both basal and serotonin-induced phosphorylation of ERK1 ⁄ 2 was increased in cells cotransfected with HA-Ras- GRF1 and 5-HT7(a) receptor, compared to cells trans- fected with receptor only (Fig. 2C). Serotonin in adenylyl HEK293 cells expressing the human Gs-coupled sero- tonin receptor 5-HT7 [27]. Forskolin increases aden- ylyl cyclase activity and induces the phosphorylation of Ser916 in the mouse Ras-GRF1 sequence [12] and of Ser898 in the rat sequence [14]. To test whether phosphorylation serotonin through PKA, HEK293 cells were cotransfected with 5-HT7(a) receptors, HA-Ras-GRF1 and either empty vector or the human phosphodiesterase hPDE4D2, which indirectly reduces PKA activity by reducing cAMP levels. The serotonin-induced phosphorylation essentially abolished and of HA-Ras-GRF1 was ERK1 ⁄ 2 cells cotransfected with hPDE4D2 (Fig. 3A). The phos- phorylation of overexpressed HA-Ras-GRF1 was also eliminated in cells incubated with 20 lm H89, an inhibitor of PKA but also of other kinases [28], for to treatment with 10 lm serotonin 25 min prior (Fig. 3B). The serotonin-induced ERK1 ⁄ 2 phosphory- lation was concomitantly reduced, as expected based on the results of our previous publication [24]. Cotrans- fection of HEK293 cells with the PKA inhibitor protein kinase inhibitor, in addition to 5-HT7(a) receptors and HA-RasGRF1, also reduced the serotonin-induced phosphorylation of recombinant HA-Ras-GRF1, as well as ERK1 ⁄ 2 phosphorylation (not shown). Phos- phorylation of the endogenously expressed 140 kDa and (cid:1) 60 kDa isoforms of Ras-GRF1 was increased following stimulation with serotonin. The (cid:1) 60 kDa iso- form of Ras-GRF1 seems to be expressed at a higher level than the 140 kDa isoform. The serotonin-induced increase in phosphorylation of both isoforms was reduced by the coexpression of hPDE4D2 with 5-HT7 receptors (Fig. 3C). This was also the case for ERK1 ⁄ 2 phosphorylation (Fig. 3D). The EGF receptor

Phosphorylation of Ser916 is neither necessary nor sufficient for Ras-GRF1-mediated phosphorylation of ERK1 ⁄ 2

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2307

To investigate the potential role of phosphorylation at Ser916 ⁄ Ser927 of Ras-GRF1 in 5-HT7(a) receptor- dependent ERK1 ⁄ 2 activation, we compared the activities of wild-type Ras-GRF1 to proteins that had single amino acid substitutions at Ser916. We also used the mutants to verify the specificity of the induces activation of Ras through a GEF, called Sos1, in a Ca2+-independent manner. Treatment of HEK293 cells with 10 nm EGF for 5 min resulted in ERK1 ⁄ 2 phosphorylation (Fig. 6D) but not in phosphorylation of endogenous Ras-GRF1 at the site recognized by the antibody directed against Ras-GRF1 phosphorylated at Ser916 ⁄ 927 (Fig. 2D). This indicates that ERK1 ⁄ 2 activation induced by EGF does not increase the phosphoryla- tion of endogenously expressed Ras-GRF1 on Ser927. to increase similarly been reported not EGF has

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

A

B

C

D

Fig. 3. Serotonin-induced Ras-GRF1 and ext- racellular signal-regulated kinase (ERK)1 ⁄ 2 phosphorylation is dependent on protein kin- ase A (PKA) ⁄ cAMP. (A) Human embryonic kidney (HEK)293 cells cotransfected with the 5-hydroxytryptamine7(a) (5-HT7(a)) recep- tor, HA-Ras-GRF1, and either with or with- out hPDE4D2, were treated with 10 lM 5-HT for 5 min. (B) HEK293 cells cotrans- fected with 5-HT7(a) receptor and HA-Ras- GRF1 were treated with or without 20 lM N-[2-(p-bromocinnamylamino)ethyl]-5-isoquin- olinesulfonamide dihydrochloride (H89) for 25 min prior to treatment with or without 10 lM serotonin for 5 min. (C) HEK293 cells cotransfected with the 5-HT7(a) receptor and empty vector or hPDE4D2, as indicated, were treated with 10 lM serotonin for 5 min. (D) The same samples as in (C) were assayed for ERK1 ⁄ 2 phosphorylation by SDS ⁄ PAGE (10% gel) and the western blot was probed with phosphospecific ERK1 ⁄ 2 antibodies (pERK1 ⁄ 2; upper panel) and then with ERK1 ⁄ 2 antibodies (ERK1 ⁄ 2; lower panel), to confirm equal loading. The pro- teins were separated by SDS ⁄ PAGE (6% gel) for Ras-GRF1 and by SDS ⁄ PAGE (10% gel) for ERK1 ⁄ 2 and electroblotted to poly(vinylidene difluoride) membranes. The membranes were probed with antibodies, as indicated.

An intact N-terminal region is required for Ras-GRF1 to potentiate ERK1/2 activation

The role of calcium in the phosphorylation of ERK1 ⁄ 2 induced by Ras-GRF1 was addressed by cotransfecting HEK293 cells with 5-HT7(a) receptors and Ras-GRF1- D1-225 (i.e. lacking the PH1-, coiled-coil and IQ domains). Cotransfection of HEK293 cells with this truncated form of Ras-GRF1 did not increase the basal or serotonin-induced phosphorylation of ERK1 ⁄ 2 com- pared to cells transfected with the receptor only (Fig. 4B). Serotonin treatment did, however, increase the phosphorylation of Ras-GRF1-D1-225 on Ser916 (Fig. 4A). phosphorylation to

that Serotonin increases [Ca2+]i through 5-HT7 receptors

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2308

phosphoRas-GRF1 antibody. The antibody to phos- phoSer916-Ras-GRF1 was developed against a syn- thetic phosphopeptide corresponding to the residues surrounding Ser916 of mouse Ras-GRF1 and had previously been shown to be unreactive with a Ras- GRF1 Ser916Ala mutant protein that was expressed in COS-7 or PC12 cells [14]. The antibody did not recognize HA-Ras-GRF1 proteins mutated at the Ser916 residue to alanine, aspartic acid or glutamic acid and expressed in HEK293 cells (Fig. 4A). Inter- estingly, neither inhibiting phosphorylation of Ser916 by mutating the amino acid to alanine, nor poten- tially mimicking it by mutation to aspartic acid or influenced the ability of recombinant glutamic acid, HA-Ras-GRF1 of induce ERK1 ⁄ 2 in HEK293 cells (Fig. 4B). These results suggest the phosphorylation of Ras-GRF1 at this residue may be neither necessary nor sufficient to mediate stimulation of ERK1 ⁄ 2 activation in HEK293 cells. We have previously shown that the Gs-coupled sero- induce phos- tonin receptors 5-HT4(b) and 5-HT7(a)

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

A

B

the 5-HT7(b)

Fig. 4. Mutation of Ser916 of mouse Ras-GRF1 does not alter the activation of extracellular signal-regulated kinase (ERK)1 ⁄ 2 but dele- tion of amino acids 1–225 blocks the stimulatory effect of Ras- GRF1. Human embryonic kidney (HEK)293 cells transfected with the 5-hydroxytryptamine7(a) (5-HT7(a)) receptor and empty vector or with Ras-GRF1, Ras-GRF1Ser916Ala, Ras-GRF1Ser916Asp, Ras- GRF1Ser916Glu or Ras-GRF1-D1-225, were treated with or without 10 lM serotonin for 5 min. (A) Western blots of SDS ⁄ PAGE (6% gel) of lysates of cells, transfected as indicated, were probed with anti-(pRas-GRF1) immunoglobulin (upper panel) and anti-HA-probe immunoglobulin (lower panel), to confirm equal loading. (B) Western blots of SDS ⁄ PAGE (10% gel) of lysates of cells, transfected as indicated, were probed with anti-pERK1 ⁄ 2 immunoglobulin (upper panel) and anti-ERK1 ⁄ 2 immunoglobulin (lower panel), to confirm equal loading.

the phorylation of ERK1 ⁄ 2 through a Ras-dependent mechanism [24]. The two other known human 5-HT7 receptor splice variants (5-HT7(b) and 5-HT7(d)) also induce phosphorylation of ERK1 ⁄ 2 through a Ras- dependent mechanism (data not shown). Therefore, in this respect we consider the different 5-HT7 splice vari- ants to behave similarly when expressed in HEK293 cells. HEK293 cells stably expressing the 5-HT7(b) receptor (KB1 cells) were used to determine whether serotonin can increase [Ca2+]i through human 5-HT7 receptors. Treatment of the KB1 cells with 10 lm sero- tonin resulted in a rapid, transient increase in [Ca2+]i, level, with a maximum of 40–60% above the basal whereas there was no effect of vehicle (10 lm HCl; the effect was mediated Fig. 5). To establish that nontransfected receptors, through HEK293 cells were subjected to the same treatment; no effect of serotonin on [Ca2+]i was detected. The serotonin-induced increase in [Ca2+]i was abolished by the calcium influx inhibitor, carboxyamido-triazole (CAI) (20 lm), but not by vehicle control (dimethyl- inset). These results indicate that sulfoxide) (Fig. 5, serotonin (10 lm) can increase [Ca2+]i through the human Gs-coupled 5-HT7 receptors in HEK293 cells. The serotonin-mediated exact mechanism for increase in Ca2+ levels is not known.

Phosphorylation of ERK1/2, mediated through 5-HT7 receptors, is dependent on Ca2+

Fig. 5. Serotonin increases intracellular Ca2+ concentration through 5-hydroxytryptamine7(b) (5-HT7(b)) receptors. Non-transfected or stably transfected human embryonic kidney (HEK)293 cells expressing the 5-HT7(b) receptor, KB1 cells, were cultured, washed and loaded with 5 lM FURA-2-AM for 20 min. The fluorescence intensity in single cells was recorded at 340 nm and 380 nm for up to 300 s on an inverted microscope. The cells were treated with 10 lM serotonin 30 s subsequent to the start of the recordings, as indicated with an arrow. Inset, in addition to treatment with FURA-2-AM, as described above, the cells were treated with carboxyamido-triazole (CAI) (20 lM) or vehicle con- trol (dimethylsulfoxide) for 25 min prior to treatment with 10 lM serotonin.

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2309

Transiently transfected HEK293 cells were incubated with CAI (20 lm) for 25 min prior to 5 min of treat-

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

of ERK1 ⁄ 2 in HEK293 cells was inhibited by pretreat- ment with 20 lm CAI for 25 min, demonstrating that CAI inhibited the calcium-mediated phosphorylation of ERK1 ⁄ 2 under these conditions (Fig. 6C).

ment with 10 lm serotonin. Serotonin-induced phos- phorylation of ERK1 ⁄ 2 was markedly reduced in the presence of 20 lm CAI (Fig. 6A). Serotonin-induced phosphorylation of ERK1 ⁄ 2 was also reduced in cells incubated with the Ca2+ chelator, BAPTA-AM (40 lm), for 25 min prior to 5 min of treatment with 10 lm serotonin (Fig. 6B). Increasing the free intracel- lular levels of Ca2+ by treatment of HEK293 cells with thapsigargin induced phosphorylation of ERK1 ⁄ 2 (Fig. 6C). Previously, CAI has been shown to inhibit the thapsigargin-induced activation of ERK1 ⁄ 2 in Rat1 cells [29]. Thapsigargin-induced phosphorylation To determine whether the effect of CAI on the sero- tonin-induced ERK1 ⁄ 2 phosphorylation was specific, HEK293 cells were treated with 20 lm CAI for 25 min prior to treatment with 10 nm EGF for 5 min. EGF- induced phosphorylation of ERK1 ⁄ 2 was not influ- enced by the presence of CAI (Fig. 6D), demonstrating that CAI does not have a general suppressive effect on the Ras-dependent activation of ERK1 ⁄ 2.

A

Increased basal ERK1 ⁄ 2 phosphorylation in the presence of HA-Ras-GRF1 is reduced by CAI and RasN17

B

In HEK293 cells transfected with the 5-HT7(a) recep- tor, cotransfection with HA-Ras-GRF1 increased basal ERK1 ⁄ 2 phosphorylation (Fig. 7A, lanes 5 and 6 vs. lane 1). Serotonin-induced ERK1 ⁄ 2 phosphorylation in these cotransfected cells was abolished by pretreat- ment with CAI (Fig. 7A, lanes 5–12), as in cells trans-

A

C

B

D

C

(5-HT7(a))

Fig. 6. Serotonin-induced extracellular signal-regulated kinase (ERK)1 ⁄ 2 phosphorylation is dependent on Ca2+. (A) Human embry- onic kidney (HEK)293 cells transiently transfected with the 5-hy- droxytryptamine7(a) (5-HT7(a)) receptor were treated with or without 20 lM carboxyamido-triazole (CAI) for 25 min prior to treatment with or without 10 lM serotonin for 5 min, as indicated. (B) HEK293 cells, transiently transfected with the 5-HT7(a) receptor, were treated with or without 40 lM BAPTA-AM for 25 min prior to incubation for 5 min with or without 10 lM serotonin. (C) and (D) HEK293 cells were treated with or without 1 lM thapsigargin (C) or 10 nM epidermal growth factor (EGF) (D) for 5 min subsequent to treatment with or without 20 lM CAI for 25 min, as indicated. (A), (B), (C) and (D) show representative western blots of proteins sep- to arated by SDS ⁄ PAGE (10% gel) and electroblotted over poly(vinylidene difluoride) membranes before probing with antibod- ies, as indicated.

Fig. 7. Phosphorylation of extracellular signal-regulated kinase (ERK)1 ⁄ 2, induced by recombinant HA-Ras-GRF1, is dependent on Ca2+ and Ras. (A) Human embryonic kidney (HEK)293 cells cotrans- fected with the 5-hydroxytryptamine7(a) receptor and empty vector or HA-Ras-GRF1 were treated with or without 10 lM serotonin for 5 min subsequent to treatment with 20 lM carbox- yamido-triazole (CAI) or vehicle for 25 min, as indicated. (B) HEK293 cells transiently cotransfected with the 5-HT7(a) receptor and HA-Ras-GRF1 were treated with or without 20 lM CAI for 25 min prior to treatment with or without 10 lM serotonin for 5 min. (C) HEK293 cells were cotransfected with 5-HT7(a) receptor and empty vector, HA-Ras-GRF1 or RasN17, as indicated. The transfected cells were treated with or without 10 lM serotonin for 5 min. (A), (B) and (C) show representative western blots of 10% (A and C) and 6% (B) SDS ⁄ PAGE, probed with antibodies as indicated.

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2310

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

fected with the 5-HT7(a) receptor alone (Figs 6A and 7A). These results indicate that the serotonin-stimula- ted ERK1 ⁄ 2 phosphorylation is Ca2+ dependent. There was also a slight inhibitory effect of CAI on the increased basal phosphorylation of ERK1 ⁄ 2 observed upon cotransfection with HA-Ras-GRF1 (Fig. 7A). On the other hand, the serotonin-induced phosphory- lation of HA-Ras-GRF1 was not affected by CAI, this phosphorylation is not Ca2+ indicating that dependent (Fig. 7B). sequence

To determine whether the increased ERK1 ⁄ 2 phos- phorylation in cells transfected with HA-Ras-GRF1 was mediated through Ras, HEK293 cells were cotransfected with plasmids encoding the 5-HT7(a) receptor, HA-Ras-GRF1 and a dominant-negative construct of Ras, RasN17. RasN17 essentially elimin- ated the increase in ERK1 ⁄ 2 phosphorylation (both basal and serotonin-stimulated) induced by the overex- pression of HA-Ras-GRF1 (Fig. 7C), indicating that the effect of Ras-GRF1 on basal and serotonin-stimu- lated ERK1 ⁄ 2 phosphorylation is Ras-dependent.

Discussion

the report endogenous expression of

GPCRs Ras-GRF1 becomes phosphorylated on several sites, with incompletely understood effects. The Ser916 residue of mouse Ras-GRF1 becomes phosphorylated by PKA in vivo and in vitro [12]. This phosphorylation is insufficient for activation but may enhance the activ- ity of Ras-GRF1 towards Ras [12,14]. The phospho- specific antibody that selectively recognizes mouse and rat Ras-GRF1, which are phosphorylated at Ser916 ⁄ 898, respectively, also recognizes human phos- phorylated Ras-GRF1. The surrounding Ser927 in human Ras-GRF1 is homologous to that surrounding Ser916 in mouse Ras-GRF1, with three amino acid substitutions. In addition, several other putative phosphorylation sites have been identified in Ras-GRF1. Baouz and colleagues, for example, did not find Ser916 as an in vitro PKA phosphorylation site [13], but rather identified Ser745 and Ser822 as the two most heavily phosphorylated residues. However, compared with the human Ser927 sequence, the sequences surrounding these two serine residues do not align as well with the mouse Ser916 sequence. There- fore, the phosphospecific antibody developed against mouse phosphoSer916-Ras-GRF1 probably recognizes human Ras-GRF1 phosphorylated at Ser927. The anti- body is highly specific for the phosphorylated residue, as mutations of Ser916 (in the mouse sequence) to alanine, aspartic acid or glutamic acid were not recog- nized by the antibody. Our finding, that reactivity of the endogenous Ras-GRF1 in HEK293 cells to the phospho-Ras-GRF1 antibody is stimulated by the acti- vation of 5-HT7 receptors, is also consistent with the selective recognition of human Ras-GRF1 by this anti- body when Ras-GRF1 is phosphorylated at Ser927. The

inhibited in the presence of We several isoforms of the guanine nucleotide exchange factor Ras-GRF1 in HEK293 cells. Serotonin treatment of HEK293 cells, transiently transfected with the Gs-cou- pled 5-HT7 receptors, induced cAMP ⁄ PKA-dependent phosphorylation of endogenous Ras-GRF1 at Ser927 and recombinant mouse HA-tagged Ras-GRF1 at Ser916. However, mutation of the Ser916 PKA phos- phorylation site did not alter the increased basal or serotonin-induced ERK1 ⁄ 2 phosphorylation induced by the overexpression of HA-Ras-GRF1. A truncated version of Ras-GRF1, lacking the Ca2+ ⁄ CaM-binding IQ domain, did not increase the basal or serotonin- induced ERK1 ⁄ 2 phosphorylation. The ERK1 ⁄ 2 phos- phorylation was the calcium influx inhibitor, CAI. phosphorylation of Ras-GRF1

serotonin-induced phosphorylation of both endogenous and recombinant Ras-GRF1 shows that Ras-GRF1 is modified by stimulation with serotonin, but is not direct evidence that Ras-GRF1 contributes serotonin-induced activation of Ras and to the ERK1 ⁄ 2. Pretreatment with H89 eliminated the sero- tonin-induced at Ser916 ⁄ 927. Transfection with the human phosphodi- esterase PDE4D2 also reduced the serotonin-induced Ras-GRF1 phosphorylation. In both cases, the sero- tonin-induced ERK1 ⁄ 2 phosphorylation was lowered concomitant with the reduced Ras-GRF1 phosphoryla- tion, but ERK1 ⁄ 2 phosphorylation was only partially reduced compared to the more substantial reduction of Ras-GRF1 phosphorylation. The endogenous expression of 5-HT6 and 5-HT7 receptors has been reported in some HEK293 cells in the current study, serotonin treat- [30]. However, ment of nontransfected HEK293 cells did not result in ERK1 ⁄ 2 phosphorylation or increased [Ca2+]i (data not shown), indicating that the HEK293 cells used did not show endogenous expression of functional 5-HT7 or other Gs-coupled serotonin receptors.

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2311

Ras-GRF1 contains several protein motifs that are presumably involved in numerous regulatory mecha- nisms. Binding of Ca2+ ⁄ CaM to the N-terminal IQ motif is considered to be the main mechanism for Ras-GRF1 activation [10]. Upon stimulation of Neither preventing PKA-mediated phosphorylation of mouse Ras-GRF1 Ser916 by mutating this residue to alanine nor mutating the residue to either aspartic or glutamic acid to potentially mimic the phosphoryla- serotonin- tion, influenced the increased basal or

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

this site was insufficient

induced ERK1 ⁄ 2 phosphorylation. Taken together, these data indicate that the PKA-mediated phosphory- lation of Ser916 of mouse Ras-GRF1, and presumably Ser927 of human Ras-GRF1, does not have a central role in ERK1 ⁄ 2 activation. The small differences in Ras activation observed between wild-type Ras-GRF1 and the Ser916Ala mutant, both in vitro [12] and in an assay of Ras-dependent neurite outgrowth from PC12 the level of cells [14], may not be detectable at ERK1 ⁄ 2 phosphorylation owing to amplification of the signal through the kinase cascade. These results are also in agreement with our previous report that phos- to activate phorylation at Ras-GRF1 in the absence of other signals [12]. It is probable that phosphorylation at this site is only one of several regulated phosphorylation events that occur on Ras-GRF1 to regulate its activity in coordination with increases in Ca2+, and so an effect from the mutation of a single site may not be apparent. The importance of phosphorylation of Ras-GRF1 at this residue is underlined by the demonstration that it is a physiologically relevant phosphorylation event which occurs at the equivalent site (Ser898) in the dendritic tree of rat prefrontal cortical neurones [14]. In addition to regulation of the Ras GEF activity of Ras-GRF1, other phosphorylation events, particularly on tyrosine residues, may regulate its activity as a GEF for another small G-protein, Rac [31].

Expression of recombinant, murine, HA-tagged Ras- GRF1 (HA-Ras-GRF1) in HEK293 cells increased the basal ERK1 ⁄ 2 phosphorylation compared to that of nontransfected cells. Serotonin caused additional phos- phorylation of ERK1 ⁄ 2 in HEK293 cells cotransfected with the 5-HT7(a) receptor and HA-Ras-GRF1, but the combined effect of 5-HT7(a) activation and HA-Ras- GRF1 expression was not much higher than the sum of the separate effects on ERK1 ⁄ 2 phosphorylation. If endogenous Ras-GRF1 was the limiting factor in the cascade from the 5-HT7(a) receptor to ERK1 ⁄ 2 phos- phorylation, one might expect that the overexpression of HA-Ras-GRF1 would elicit greater effects than observed on ERK1 ⁄ 2 phosphorylation. On the other hand, if endogenous Ras-GRF1 was not the limiting factor in the cascade, one could hypothesize that the effect of Ras-GRF1 overexpression on ERK1 ⁄ 2 phosphorylation would be similar to the sum of the receptor-induced effect and increased basal phosphory- lation, mediated from overexpressed Ras-GRF1, poss- ibly localized in different cellular compartments from the receptor.

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2312

only slightly reduced by the Ca2+ influx inhibitor, CAI. Both interventions prevented the serotonin- induced phosphorylation of ERK1 ⁄ 2. A truncated ver- sion of Ras-GRF1 (Ras-GRF1-D1-225) lacking the PH1-, coiled-coil and IQ domain and thus not expec- ted to bind Ca2+ ⁄ CaM, did not increase the basal or serotonin-induced ERK1 ⁄ 2 phosphorylation. The to induce reduced ability of Ras-GRF1-D1-225 the lost ERK1 ⁄ 2 activation may be a result of Ca2+ ⁄ CaM-binding site of the IQ domain, but the missing PH1- and coiled-coil domains may also change the subcellular localization of this version of Ras- GRF1. These domains have been shown to contribute to the regulation of Ras GEF activity [32]. The sero- tonin-induced phosphorylation of Ras-GRF1-D1-225 at Ser916 indicates that the protein is located in cellu- lar compartments within the reach of kinases activated upon serotonin treatment. We have previously shown that while increased intracellular Ca2+ is required for the stimulation of Ras-GRF1 activation by a Gi-cou- pled pathway [12], Ca2+ does not stimulate Ras-GRF1 phosphorylation at Ser916 [14]. It is probable that Ras-GRF1 can serve to integrate signals from the cAMP and Ca2+ second messenger cascades to deter- mine activation of the ERK1 ⁄ 2 cascade. In addition to the influence of second messengers and phosphoryla- tion events on its activities, Ras-GRF1 can also be regulated by interaction with another small GTPase, Cdc42 [33], and can serve a scaffolding function that directs signalling downstream of Ras activation [34,35]. In rat adrenal glomerulosa cells, 5-HT7 receptors were shown to increase [Ca2+]i through T-type Ca2+ channels in a cAMP ⁄ PKA-dependent manner [22,23]. Increase in [Ca2+]i following stimulation of over- expressed 5-HT7(a) receptors in HEK293 cells has pre- viously been shown [36] and no evidence of coupling to Gq or Gi was found. We showed that serotonin stimu- lation of HEK293 cells stably expressing 5-HT7(b) resulted in increased [Ca2+]i. Serotonin- receptors induced ERK1 ⁄ 2 severely phosphorylation was reduced in the presence of CAI, but the PKA-depend- ent phosphorylation of HA-Ras-GRF1 was not influ- enced by the presence of CAI. In nonexcitable cells, CAI can specifically inhibit store-operated calcium channels and may thereby reduce the serotonin- induced sustained increase in [Ca2+]i, as has been shown for endothelin-1-induced Ca2+ increase in Rat1 cells [29]. Whether HEK293 cells express T-type Ca2+ channels, or whether the increase in [Ca2+]i is mediated through a different mechanism, has not been addressed further in this study, and the results obtained with the calcium influx inhibitor, CAI, do not provide conclu- sive data concerning the nature of the calcium increase. The increased ERK1 ⁄ 2 phosphorylation in the pres- ence of HA-Ras-GRF1 was essentially eliminated in the presence of dominant-negative Ras, RasN17, but

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

tamine were from Cambrex (Vervierse, Belgium). Supersignal West Dura extended-duration chemiluminescent substrate was from Pierce Biotechnology (Rockford, IL, USA), and the BC assay protein quantification kit was from Uptima (Monticon, France). BAPTA-AM was from Calbiochem (La Jolla, CA, USA).

The pcDNA3.1(–) vector (Invitrogen), encoding the human 5-HT7(a) receptor, was as described previously [27]. The pKH3 mammalian expression plasmids encoding the full- length murine wild-type HA-Ras-GRF1 and the Ser916Ala mutant were as described previously [9,12,47]. HA-Ras- GRF1-Ser916Asp, HA-Ras-GRF1-Ser916Glu and HA-Ras- GRF1-D1-225 were constructed by PCR using appropriate mutagenic primers and the protocol previously described [12] and then confirmed by DNA sequencing. The pCMV vector encoding dominant-negative Ras, RasN17, was from Clontech (Palo Alto, CA, USA). The pCMV5 vector enco- ding the human phosphodiesterase 4D2, hPDE4D2, was (Department of Obstetrics and provided by M. Conti Gynaecology, Stanford, CA, USA).

Plasmids

Cell culture and transfection

empty

necessary,

protocol. When

Ras-GRF1 is implicated in signalling from various neurotransmitter receptors [9,12,37]. The downstream target of Ras-GRF1, Ras, may help to regulate expres- sion of specific genes involved in processes such as memory. In Aplysia, the activation of MAP kinases by Gs-coupled serotonin receptors is implicated in mem- ory formation [38,39]. There is increasing evidence for the biological importance of the Ras ⁄ MAP kinase cas- cade in human learning and memory [40]. Gs-coupled serotonin receptors are found in the hippocampus [41,42], and 5-HT7 receptors activate ERK1 ⁄ 2 in cul- tured neurones [43]. Ras-GRF1 is highly expressed in hippocampal and other neurones, and Ras-GRF1-defi- cient mice have memory defects [44,45]. Therefore, a possible involvement of Ras-GRF1 in the Ca2+- and Ras-dependent activation of ERK1 ⁄ 2 through 5-HT7 receptors may be of physiological relevance. Since the original manuscript was submitted for publication, Johnson-Farley and colleagues have shown interplay between Gs- and Gq-coupled serotonin receptors in the activation of ERK1 ⁄ 2 and PKB (Akt) in PC12 cells [46]. They found that PKA activation through Gs-cou- pled serotonin receptors was Ca2+ dependent, whereas ERK1 ⁄ 2 phosphorylation was Ca2+ independent. Considering all the different pathways reported for the activation of Ras and ERK1 ⁄ 2 downstream of GPCRs, Ras-GRF1 could be one of possibly several GEFs involved in the activation of Ras and subse- quently ERK1 ⁄ 2 downstream of Gs-coupled serotonin receptors. This remains a challenge for future research.

Experimental procedures

HEK293 cells were cultured in DMEM containing 10% (v ⁄ v) fetal bovine serum and supplements (2 mm l-gluta- mine, 100 UÆmL)1 penicillin, 100 lgÆmL)1 streptomycin), at 37 (cid:1)C in a humidified atmosphere of 5% CO2 in air, and indicated transfected at 60–70% confluence with the cDNA(s) using Lipofectamine 2000, according to the manu- facturer’s vector [pcDNA3.1(–)] was included in the transfection to ensure that each dish received the same amount of DNA (1.0 or 2.9 lg of plasmid DNA per 35 or 60 mm dish, respectively). Cells expressing 5-HT7 receptors were cultured in UltraCUL- TURETM serum-free medium with supplements, as described above, prior to starvation in DMEM without serum for the last 16–20 h before serotonin treatment and lysis ((cid:1) 48 h after transfection for transiently transfected cells). Non- transfected cells were similarly starved in DMEM without serum before treatment (with EGF or thapsigargin) and lysis. Where indicated, cells were preincubated with 20 lm H89, 20 lm CAI or 40 lm BAPTA-AM for 25 min prior to treat- ment with agonist. Cells were stimulated for 5 min if not indicated otherwise. All experiments were carried out in duplicate at least three times, if not otherwise indicated.

Materials

Equal amounts of cell lysate proteins were separated by SDS ⁄ PAGE and electroblotted onto poly(vinylidene difluo- ride) membranes. The membranes were incubated with pri-

HEK293 cells were from the American Type Culture Collec- tion (Manassas, VA, USA). Mouse monoclonal antiphos- pho-ERK1 ⁄ 2 and rabbit polyclonal anti-(phosphoSer916- Ras-GRF1) Ig (#3321) were from Cell Signaling Technology (Beverly, MA, USA), sheep polyclonal antimouse immuno- globulin–horseradish peroxidase conjugate (Ig-HRP) and sheep anti-(rabbit IgG)–HRP were from Amersham Pharma- cia Biotech (Little Chalfont, Bucks, UK), rabbit polyclonal anti-ERK1 ⁄ 2 Ig was from Upstate Biotechnology (Lake Placid, NY, USA), and rabbit polyclonal anti-(Ras-GRF1) Ig (human, rat) was from Santa Cruz Biotechnology (Santa Cruz, CA, USA). 5-HT, EGF, H89 and Dulbecco’s modified Eagle’s medium (DMEM) were from Sigma (St Louis, MO, USA). Hybond-P [poly(vinylidene difluoride)] membrane was from Amersham. LipofectamineTM 2000 was from Invi- trogen (Carlsbad, CA, USA). Fetal bovine serum was from EuroClone (Milano, Italy). UltraCULTURETM general pur- pose serum-free medium, penicillin ⁄ streptomycin and l-glu-

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2313

Western Blotting

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

secondary

mary antibodies [anti-(phospho-ERK1 ⁄ 2), 1 : 2000, v ⁄ v; anti-ERK1 ⁄ 2, 1 : 10 000, v ⁄ v; anti-(phospho-Ras-GRF1), 1 : 1000, v ⁄ v; and anti-(Ras-GRF1), 1 : 1000, v ⁄ v] in NaCl ⁄ Pi containing 5% (w ⁄ v) nonfat dry milk and 0.05% (v ⁄ v) Tween 20, and thereafter incubated with the corres- ponding HRP-conjugated antibodies. The immobilized HRP-conjugated secondary antibodies were visualized with SuperSignal Dura West extended-duration substrate and analysed with a UVP chemiluminescent BioChemie system.

isolate mRNA from the pool of total RNA. The purified mRNA was used for oligo-dT primed cDNA synthesis. The cDNA was treated with RNase H to remove RNA comple- mentary to the cDNA. The purified cDNA was used as substrate in PCR with the following Ras-GRF1 gene-speci- fic primers: ON357, 5¢-TGAAACATCACCAACTAAATC TCCAA-3¢; ON358, 5¢-GACGACTCCATTGTTATAGG AAAAGAGT-3¢; ON359, 5¢-GCCGCTGGAGAAACAG 5¢-GCCACCCATTCGTCACAATC-3¢; CAT-3¢; ON360, and ON361, 5¢-ATGCAGAAGGGGATCCGG-3¢.

The cells were cultured in 35 mm dishes, transfected and stimulated with agonist as described, lysed in ice-cold cell [1% (w ⁄ v) SDS, 1 mm Na3VO4, 50 mm lysis buffer Tris ⁄ HCl, pH 7.4, at room temperature], scraped with a Teflon cell scraper, sheared through a 25 GA syringe and immediately frozen in liquid nitrogen. The thawed cell ly- sates were cleared by centrifugation (13 000 g at 4 (cid:1)C) and the protein concentrations in the supernatants were quanti- fied by the BC assay quantification kit (Uptima) using BSA as the standard. Equal amounts of protein were prepared for separation by SDS ⁄ PAGE.

Phospho-ERK1 ⁄ 2 and phospho-Ras-GRF1 assay Direct measurements of [Ca2+]i in single cells

increases, whereas

in the

change

Non-transfected HEK293 cells and HEK293 cells stably transfected with the 5-HT7(b) receptor (KB1 cells) were plated and cultured in specially designed glass-bottomed wells [48] coated with 12.7 lgÆcm)2 collagen type VII from rat tail. At 70–80% confluence, the cells were washed at 37 (cid:1)C with Hepes-buffered salt solution (HSS; 136 mm NaCl, 5 mm KCl, 1.2 mm MgCl2, 1.2 mm CaCl2, 11 mm Bacto-dextrose, 10 mm Hepes, pH 7.35) and incubated with 5 lm FURA-2-AM in HSS for 20 min at room tem- perature. The cells were washed once at 37 (cid:1)C with HSS buffer prior to mounting the cell culture dish on an inverted Nikon microscope equipped with digital record- ing facilities. Recordings of the fluorescence intensities started 30 s prior to the addition of vehicle or serotonin solution. The fluorescence intensity of FURA-2 at excita- tion wavelength 340 nm (F340) the fluorescence intensity at 380 nm (F380) decreases upon the binding of Ca2+. The ratio of F340 ⁄ F380 determined the change in Ca2+ concentration inside the cell.

Immunoprecipitation

Acknowledgements

HEK293 cells were cultured in 60 mm dishes and grown as described above, lysed in ice-cold lysis buffer and the lysate cleared by centrifugation (13 000 g at 4 (cid:1)C). Samples con- taining 200 lg of protein were diluted in ice-cold 1 · IP buf- fer [10 mm Tris ⁄ HCl, pH 7.4 at room temperature, 150 mm NaCl, 1 mm EDTA, 1 mm EGTA, 0.2 mm Na3VO4, 0.2 mm phenylmethanesulfonyl fluoride, 1% (v ⁄ v) Triton X-100, 0.5% (v ⁄ v) Nonidet P-40) to a final volume of 1 mL and incubated for 2 h at 4 (cid:1)C with or without (control) poly- clonal anti-Ras-GRF1 immunoglobulin and subsequently with sheep anti-rabbit paramagnetic beads (Dynal, Oslo, Norway) overnight at 4 (cid:1)C on a tumbler. The beads with the bound proteins were washed three times with cold 1 · IP buffer. The proteins were eluted from the beads by boiling for 5 min in 1· SDS ⁄ PAGE loading buffer [31 mm Tris ⁄ HCl, pH 6.8 at room temperature, 1% (w ⁄ v) SDS, 2.5% (v ⁄ v) glycerol, 0.025% (w ⁄ v) bromophenol blue, 2.5% (v ⁄ v) b-mercaptoethanol]. The protein samples were loaded and resolved on 6% (w ⁄ v) SDS ⁄ polyacrylamide gels. Western blotting was performed as described above.

This work was supported by The Norwegian Council on Cardiovascular Diseases, The Norwegian Cancer Soci- ety, The Research Council of Norway, The Novo Nor- disk Foundation, Anders Jahre’s Foundation for the Promotion of Science, The Blix Family Foundation and grants from the University of Oslo and the National Cancer Institute of the USA (R01-CA81150). The experiments were performed in accordance with all regu- lations concerning biomedical research in Norway. We thank Dr Marco Conti for generating and providing the plasmid encoding hPDE4D2 and thank Dr Jens-Gustav Iversen for help with the calcium measurements. Isolation of mRNA, and PCR

References

1 Chardin P, Camonis JH, Gale NW, Van Aelst L,

Schlessinger J, Wigler MH & Bar-Sagi D (1993) Human

Total RNA from HEK293 cells, cultured as described above, was isolated by using the SV total RNA isolation system (Promega, Madison, WI, USA), according to the manufacturer’s protocol. Paramagnetic beads were used to

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2314

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

signaling in the cerebral cortex. J Biol Chem 278, 13278–13285.

Sos1: a guanine nucleotide exchange factor for Ras that binds to GRB2. Science 260, 1338–1343.

2 Della RGBT, Daaka Y, Luttrell DK, Luttrell LM &

15 Shou C, Farnsworth CL, Neel BG & Feig LA (1992) Molecular-cloning of cDNAs encoding a guanine- nucleotide-releasing factor for Ras P21. Nature 358, 351–354.

Lefkowitz RJ (1997) Ras-dependent mitogen-activated protein kinase activation by G protein-coupled recep- tors. Convergence of Gi- and Gq-mediated pathways on calcium ⁄ calmodulin, Pyk2, and Src kinase. J Biol Chem 272, 19125–19132.

3 Luttrell LM (2002) Activation and targeting of mitogen- activated protein kinases by G-protein-coupled recep- tors. Can J Physiol Pharmacol 80, 375–382.

4 Lev S, Moreno H, Martinez R, Canoll P, Peles E,

Musacchio JM, Plowman GD, Rudy B & Schlessinger J (1995) Protein tyrosine kinase Pyk2 involved in Ca2+- induced regulation of ion channel and MAP kinase functions. Nature 376, 737–745.

16 Wei W, Mosteller RD, Sanyal P, Gonzales E, McKinney D, Dasgupta C, Li P, Liu BX & Broek D (1992) Identifi- cation of a mammalian gene structurally and function- ally related to the CDC25 gene of Saccharomyces cerevisiae. Proc Natl Acad Sci USA 89, 7100–7104. 17 Cen H, Papageorge AG, Zippel R, Lowy DR & Zhang K (1992) Isolation of multiple mouse cDNAs with cod- ing homology to Saccharomyces cerevisiae CDC25: iden- tification of a region related to Bcr, Vav, Dbl and CDC24. EMBO J 11, 4007–4015.

18 Guerrero C, Rojas JM, Chedid M, Esteban LM,

5 Dikic I, Tokiwa G, Lev S, Courtneidge SA & Schles- singer J (1996) A role for Pyk2 and Src in linking G- protein-coupled receptors with MAP kinase activation. Nature 383, 547–550.

6 Toki S, Kawasaki H, Tashiro N, Housman DE &

Zimonjic DB, Popescu NC, deMora JF & Santos E (1996) Expression of alternative forms of Ras exchange factors GRF and SOS1 in different human tissues and cell lines. Oncogene 12, 1097–1107.

19 Martegani E, Vanoni M, Zippel R, Coccetti P,

Graybiel AM (2001) Guanine nucleotide exchange fac- tors CalDAG-GEFI and CalDAG-GEFII are coloca- lized in striatal projection neurons. J Comp Neurol 437, 398–407.

7 Wolfman A & Macara IG (1990) A cytosolic protein

Brambilla R, Ferrari C, Sturani E & Alberghina L (1992) Cloning by functional complementation of a mouse cDNA encoding a homologue of CDC25, a Saccharomyces cerevisiae RAS activator. EMBO J 11, 2151–2157.

catalyzes the release of GDP from p21ras. Science 248, 67–69.

8 Shou C, Wurmser A, Ling K, Barbacid M & Feig LA

20 Heidmann DE, Metcalf MA, Kohen R & Hamblin MW (1997) Four 5-hydroxytryptamine7 (5-HT7) receptor iso- forms in human and rat produced by alternative spli- cing: species differences due to altered intron-exon organization. J Neurochem 68, 1372–1381.

(1995) Differential response of the Ras exchange factor, Ras-GRF to tyrosine kinase and G-protein mediated signals. Oncogene 10, 1887–1893.

21 de Rooij J, Zwartkruis FJ, Verheijen MH, Cool RH,

9 Mattingly RR & Macara IG (1996) Phosphorylation- dependent activation of the Ras-GRF ⁄ CDC25Mm exchange factor by muscarinic receptors and G-protein bc subunits. Nature 382, 268–272.

Nijman SM, Wittinghofer A & Bos JL (1998) Epac is a Rap1 guanine-nucleotide-exchange factor directly acti- vated by cyclic AMP. Nature 396, 474–477.

10 Farnsworth CL, Freshney NW, Rosen LB, Ghosh A,

Greenberg ME & Feig LA (1995) Calcium activation of Ras mediated by neuronal exchange factor Ras-GRF. Nature 376, 524–527.

22 Lenglet S, Louiset E, Delarue C, Vaudry H & Contesse V (2002) Involvement of T-type calcium channels in the mechanism of action of 5-HT in rat glomerulosa cells: a novel signaling pathway for the 5-HT7 receptor. Endocr Res 28, 651–655.

11 Mattingly RR, Saini V & Macara IG (1999) Activation of the Ras-GRF ⁄ CDC25Mm exchange factor by lyso- phosphatidic acid. Cell Signal 11, 603–610.

12 Mattingly RR (1999) Phosphorylation of serine 916 of Ras-GRF1 contributes to the activation of exchange factor activity by muscarinic receptors. J Biol Chem 274, 37379–37384.

13 Baouz S, Jacquet E, Accorsi K, Hountondji C,

23 Lenglet S, Louiset E, Delarue C, Vaudry H & Contesse V (2002) Activation of 5-HT7 receptor in rat glomeru- losa cells is associated with an increase in adenylyl cyclase activity and calcium influx through T-type calcium channels. Endocrinology 143, 1748–1760. 24 Norum JH, Hart K & Levy FO (2003) Ras-dependent ERK activation by the human Gs-coupled serotonin receptors 5-HT4(b) and 5-HT7(a). J Biol Chem 278, 3098– 3104.

Balestrini M, Zippel R, Sturani E & Parmeggiani A (2001) Sites of phosphorylation by protein kinase A in CDC25Mm ⁄ GRF1, a guanine nucleotide exchange factor for Ras. J Biol Chem 276, 1742–1749.

14 Yang H, Cooley D, Legakis JE, Ge Q, Andrade R &

25 Lin SL, Johnson-Farley NN, Lubinsky DR & Cowen DS (2003) Coupling of neuronal 5-HT7 receptors to activation of extracellular-regulated kinase through a protein kinase A-independent pathway that can utilize Epac. J Neurochem 87, 1076–1085.

Mattingly RR (2003) Phosphorylation of the Ras-GRF1 exchange factor at Ser916 ⁄ 898 reveals activation of Ras

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2315

J. H. Norum et al.

Ras-GRF1 and Ras-dependent ERK activation in HEK293

37 Kiyono M, Satoh T & Kaziro Y (1999) G protein bc subunit-dependent Rac-guanine nucleotide exchange activity of Ras-GRF1 ⁄ CDC25Mm. Proc Natl Acad Sci USA 96, 4826–4831.

26 Font de Mora J, Esteban LM, Burks DJ, Nunez A, Garces C, Garcia-Barrado MJ, Iglesias-Osma MC, Moratinos J, Ward JM & Santos E (2003) Ras-GRF1 signaling is required for normal beta-cell development and glucose homeostasis. EMBO J 22, 3039–3049. 27 Krobert KA, Bach T, Syversveen T, Kvingedal AM & Levy FO (2001) The cloned human 5-HT7 receptor splice variants: a comparative characterization of their pharmacology, function and distribution. Naunyn- Schmiedeberg’s Arch Pharmacol 363, 620–632.

38 Martin KC, Michael D, Rose JC, Barad M, Casadio A, Zhu H & Kandel ER (1997) MAP kinase translocates into the nucleus of the presynaptic cell and is required for long-term facilitation in Aplysia. Neuron 18, 899–912. 39 Michael D, Martin KC, Seger R, Ning MM, Baston R & Kandel ER (1998) Repeated pulses of serotonin required for long-term facilitation activate mitogen- activated protein kinase in sensory neurons of Aplysia. Proc Natl Acad Sci USA 95, 1864–1869.

28 Davies SP, Reddy H, Caivano M & Cohen P (2000) Specificity and mechanism of action of some com- monly used protein kinase inhibitors. Biochem J 351, 95–105.

29 Rodland KD, Wersto RP, Hobson S & Kohn EC

40 Weeber EJ, Levenson JM & Sweatt JD (2002) Molecu- lar genetics of human cognition. Mol Interv 2, 376–391.

41 Markstein R, Matsumoto M, Kohler C, Togashi H,

(1997) Thapsigargin-induced gene expression in nonexci- table cells is dependent on calcium influx. Mol Endocri- nol 11, 281–291.

Yoshioka M & Hoyer D (1999) Pharmacological char- acterisation of 5-HT receptors positively coupled to ade- nylyl cyclase in the rat hippocampus. Naunyn- Schmiedeberg’s Arch Pharmacol 359, 454–459.

30 Johnson MS, Lutz EM, Firbank S, Holland PJ & Mitchell R (2003) Functional interactions between native Gs-coupled 5-HT receptors in HEK-293 cells and the heterologously expressed serotonin transporter. Cell Signal 15, 803–811.

42 Thomas DR, Middlemiss DN, Taylor SG, Nelson P & Brown AM (1999) 5-CT stimulation of adenylyl cyclase activity in guinea-pig hippocampus: evidence for invol- vement of 5-HT7 and 5-HT1A receptors. Br J Pharmacol 128, 158–164.

43 Errico H, Crozier RA, Plummer MR & Cowen DS

31 Kiyono M, Kaziro Y & Satoh T (2000) Induction of Rac-guanine nucleotide exchange activity of Ras- GRF1 ⁄ CDC25Mm following phosphorylation by the nonreceptor tyrosine kinase Src. J Biol Chem 275, 5441– 5446.

(2001) 5-HT7 receptors activate the mitogen activated protein kinase extracellular signal related kinase in cul- tured rat hippocampal neurons. Neuroscience 102, 361– 367.

32 Buchsbaum R, Telliez JB, Goonesekera S & Feig LA (1996) The N-terminal pleckstrin, coiled-coil, and IQ domains of the exchange factor Ras-GRF act coopera- tively to facilitate activation by calcium. Mol Cell Biol 16, 4888–4896.

44 Brambilla R, Gnesutta N, Minichiello L, White G, Roylance AJ, Herron CE, Ramsey M, Wolfer DP, Cestari V, Rossi-Arnaud C et al. (1997) A role for the Ras signalling pathway in synaptic transmission and long-term memory. Nature 390, 281–286.

33 Arozarena I, Matallanas D & Crespo P (2001) Mainte- nance of Cdc42 GDP-bound state by Rho-GDI inhibits MAP kinase activation by the exchange factor Ras- GRF. Evidence for Ras-GRF function being inhibited by Cdc42-GDP but unaffected by Cdc42-GTP. J Biol Chem 276, 21878–21884.

45 Giese KP, Friedman E, Telliez JB, Fedorov NB, Wines M, Feig LA & Silva AJ (2001) Hippocampus-dependent learning and memory is impaired in mice lacking the Ras-guanine-nucleotide releasing factor 1 (Ras-GRF1). Neuropharmacol 41, 791–800.

34 Buchsbaum RJ, Connolly BA & Feig LA (2002) Inter- action of Rac exchange factors Tiam1 and Ras-GRF1 with a scaffold for the p38 mitogen-activated protein kinase cascade. Mol Cell Biol 22, 4073–4085.

35 Giglione C & Parmeggiani A (1998) Raf-1 is involved in

46 Johnson-Farley NN, Kertesy SB, Dubyak GR & Cowen DS (2005) Enhanced activation of Akt and extracellu- lar-regulated kinase pathways by simultaneous occu- pancy of Gq-coupled 5-HT2A receptors and Gs-coupled 5-HT7A receptors in PC12 cells. J Neurochem 92, 72–82.

the regulation of the interaction between guanine nucleotide exchange factor and Ha-Ras. Evidence for a function of Raf-1 and phosphatidylinositol 3-kinase upstream to Ras. J Biol Chem 273, 34737–34744. 36 Baker LP, Nielsen MD, Impey S, Metcalf MA, Poser

47 Mattingly RR, Sorisky A, Brann MR & Macara IG (1994) Muscarinic receptors transform NIH 3T3 cells through a Ras-dependent signalling pathway inhibited by the Ras-GTPase-activating protein SH3 domain. Mol Cell Biol 14, 7943–7952.

48 Røtnes JS & Iversen JG (1992) Thapsigargin reveals

SW, Chan G, Obrietan K, Hamblin MW & Storm DR (1998) Stimulation of type 1 and type 8 Ca2+ ⁄ calmo- dulin-sensitive adenylyl cyclases by the Gs-coupled 5-hydroxytryptamine subtype 5-HT7A receptor. J Biol Chem 273, 17469–17476.

evidence for fMLP-insensitive calcium pools in human leukocytes. Cell Calcium 13, 487–500.

FEBS Journal 272 (2005) 2304–2316 ª 2005 FEBS

2316