Regulation of Btk Function by a Major Autophosphorylation Site Within the SH3 Domain

Immunity, Vol. 4, 515–525, May, 1996, Copyright 1996 by Cell Press

Regulation of Btk Function by aMajor Autophosphorylation SiteWithin the SH3 Domain

Hyunsun Park,*# Matthew I. Wahl,*†# homology domains, SH3, SH2, and SH1 (catalytic) (Tsu-kada et al., 1993; Vetrie et al., 1993; Rawlings et al.,Daniel E. H. Afar,* Christoph W. Turck,‡

David J. Rawlings,* Christina Tam,* 1993; Thomas et al., 1993). Mutations in the Btk generesult inhuman X-linked agammaglobulinemia (XLA) andAndrew M. Scharenberg,§ Jean-Pierre Kinet,§

and Owen N. Witte*‖ murine X-linked immunodeficiency (Xid). XLA patientshave a deficit in mature B cells due to a developmental*Department of Microbiology and Molecular Genetics

University of California, Los Angeles arrest during pre-B cell expansion (Kinnon et al., 1993).In Xid, B cell development is selectively impaired andLos Angeles, California 90095-1662

†University of California, Los Angeles the number of B cells is reduced by 30%–50% (Wickerand Scher, 1986). Xid B cells do not respond normallyCedars Sinai Medical Genetics Training Program

Los Angeles, California 90048 to a wide variety of activating signals, including thymus-independent type 2 antigens, surface immunoglobulin‡Howard Hughes Medical Institute

Department of Medicine M (sIgM) cross-linking, and interleukin-5 (IL-5), IL-10,CD38, or CD40 receptor ligation (reviewed by WickerCardiovascular Research Instituteand Scher, 1986; Rawlings and Witte, 1995).University of California, San Francisco

Enhancement of catalytic activity in association withSan Francisco, California 94143-0724tyrosine phosphorylation is involved in the activation of§Laboratory of Allergy and Immunologymany tyrosine kinases (reviewed by Perlmutter et al,Beth Israel Hospital and Harvard Medical School1993). Btk catalytic activity and tyrosine phosphoryla-Boston, MA 02215tion increase in response to cross-linking or stimulation‖Howard Hughes Medical Instituteof the sIgM complex, the IL-5 receptor, or the IL-6 recep-Los Angeles, California 90095-1662tor of B cells, or the high affinity IgE receptor of mastMolecular Biology Institutecells (Saouaf et al., 1994; de Weers et al., 1994; Aoki etUniversity of California, Los Angelesal., 1994; Sato et al., 1994; Matsuda et al., 1995; Kawa-Los Angeles, California 90095-1662kami et al., 1994). Stimulation of the B cell receptor(BCR) by sIgM cross-linking increases binding of Srcfamily tyrosine kinases to the BCR and triggers activa-

Summary tion of Src family tyrosine kinases (reviewed by Cambieret al., 1994). The activation of Src family tyrosine kinases

Bruton’s tyrosine kinase (Btk) plays a crucial role in B precedes Btk activation after sIgM cross-linking, sug-cell development. Overexpression of Btk with a Src gesting that Src family tyrosine kinases may be regula-family kinase increases tyrosine phosphorylation and tors of Btk activation (Saouaf et al., 1994).catalytic activity of Btk. This occurs by transphosphor- Vaccinia virus–driven high level coexpression of Btkylation at Y551 in the Btk catalytic domain and the with Lyn results in an increase in both Btk catalyticenhancement of Btk autophosphorylation at a second activity and Btk tyrosine phosphorylation (Rawlings etsite. A gain-of-function mutant called Btk* containing al., 1996). Lyn/Btk coexpression induces Btk phosphor-E41 to K change within the pleckstrin homology do- ylation at two distinct tyrosine residues by two indepen-main induces fibroblast transformation. Btk* enhances dent steps. Phosphorylation at Y551 requires Lyn kinasethe transphosphorylation of Y551 by endogenous Src activity, indicating that Y551 is a transphosphorylationfamily tyrosine kinasesand autophosphorylation at the site. The sequence surrounding BtkY551 is an excellentsecond site. We mapped the major Btk autophosphor- match to the consensus Src family phosphorylation siteylation site to Y223 within the SH3 domain. Mutation (Hanks, 1988). This transphosphorylation at Y551 is fol-of Y223 to F blocks Btk autophosphorylation and dra- lowed by phosphorylation at a second site, which ismatically potentiates the transforming activity of Btk* dependent on Btk catalytic activity.in fibroblasts. The location of Y223 in a potential li- We had previously identified an activated form of Btk,gand-binding pocket suggests that autophosphoryla- which was isolated using a retroviral random mutagene-tion regulates SH3-mediated signaling by Btk. sis scheme (Li et al., 1995). Btk* contains a point muta-

tion (E41K) in the PH domain and leads to constitutiveactivation of Btk. Btk* can induce transformation of fi-broblasts and IL-5-independent growth of a pro-B cellline. The transforming activity of Btk* is associated withIntroductionan increase in tyrosine phosphorylation and increasedmembrane localization of Btk. The transforming activityBruton’s tyrosine kinase (Btk) is a member of the Btk/is dependent on Btk kinase activity and phosphorylationTec subfamily of cytoplasmic tyrosine kinases. Btk con-of the consensus phosphorylation site of Src family tyro-tains a pleckstrin homology domain (PH) and proline-sine kinases (Y551) within the Btk catalytic domain. Ki-rich sequences at its N terminus, in addition to the Srcnase-inactive Btk*/K430R lacks tyrosine phosphoryla-tion but the Btk*/Y551F double mutant accumulatestyrosine phosphorylation at a low level. These results#These authors contributed equally to this manuscript.

Immunity516

suggest that Btk*-mediated tyrosine phosphorylation Resultsresults from at least two phosphorylation events, which

The Btk* Mutation Leads to the Phosphorylationoccur at Y551 and at an unidentified tyrosine residue.of Two Distinct Tyrosine ResiduesBtk* cannot transform rat-2 fibroblasts to agar growthby Two Independent Stepsunless it is coexpressed with a weakly activated formPhosphorylation mapping studies by Rawlings et al.of c–src (srcE378G) (D. E. H. A. et al., submitted). Trans-(1996) have shown that theSrc family kinase Lyn inducesforming activity correlates with the enhancement of BtkBtk tyrosine phosphorylation by transphosphorylationtyrosine phosphorylation. Btk*-mediated transformationof Btk at Y551 and enhancement of autophosphorylationis suppressed by coexpression of the C-terminal Srcat a second site. Phosphorylation at these sites corre-kinase (Csk), which inactivates Src family tyrosine ki-lates with a greater than 5- to 10-fold increase in Btknases. These results provide genetic evidence for thecatalytic activity. To determine whether activation byrole of Src family kinases as upstream regulators of BtkBtk* correlated with phosphorylation of the same tyro-signaling. The transforming potential of Btk* is furthersine residues induced by Btk/Lyn coexpression, we ana-enhanced by a block deletion of the SH3 domain of Btklyzed the phosphorylation patterns induced by Btk*.(Li et al., 1995; D. E. H. A. et al., submitted). AnalogousWild-type Btk and Btk* were expressed in NIH 3T3 cellsto other cytoplasmic tyrosine kinases, the SH3 domainby retroviral infection. Following in vivo 32P-orthophos-of Btk is likely to act as a regulatory domain (Hirai andphate labeling, Btk proteins were immunoprecipitated

Varmus, 1990; Seidel-Dugan et al., 1992; Jackson et al.,from cell extracts and analyzed by two-dimensional

1993; Mayer and Baltimore, 1994; Okada et al., 1993).tryptic phosphopeptide mapping (Boyle et al., 1991).

Using detailed phosphopeptide analysis, we demon-Btk* contained four major phosphopeptides that were

strate that the Btk* mutation or Lyn/Btk coexpressionnot present in wild-type Btk (Figures 1A and 1B). Phos-

induce indistinguishable patterns of tyrosine phosphor- phoamino acid analysis of individual peptides revealedylation on Btk. This suggests that alternative modes the presence of phosphotyrosine in the four peptidesof activation generate common signals leading to Btk (py1, py19, py29, and py299). We defined the py1 peptidestyrosine phosphorylation. We identified the major Btk and py2 peptides as two families of closely related spe-autophosphorylation site and mapped it to Y223 within cies, since they appear to be derived from peptidesthe SH3 domain. Mutation of Y223 to phenylalanine containing the same phosphorylation sites (see below).strongly potentiates Btk* transforming activity, sug- In addition, Btk* exhibited enhanced phosphorylation ofgesting that autophosphorylation serves to regulate Btk four peptides compared with wild-type Btk. Of theseactivation. The Y223 residue is highly conserved among peptides with enhanced phosphorylation, three con-the SH3 domains of a broad array of signaling proteins tained phosphoserine (ps1, ps2, and ps3) and one con-and is proposed to lie in the surface of the ligand-binding tained phosphotyrosine (py2). Analysis of wild-type Btkgroove (Koyama et al., 1993; Noble et al., 1993; Feng et showed a prominent phosphothreonine-containing pep-al., 1994; Erpel et al., 1995). Thus, Y223 of Btk is likely tide (pt1), which was significantly reduced in Btk*.to have a critical role in SH3 function of Btk by regulating To test which phosphopeptides represent autophos-

phorylation or transphosphorylation sites, we analyzedprotein–protein interactions.

Figure 1. Comparison of Two-DimensionalTryptic Peptide Map of Btk Proteins

(A) Wild-type Btk, (B) Btk*, (C) Btk*/Y551F, (D)Btk*/K430R Btk proteins were stably ex-pressed inNIH 3T3 cells by retroviral infection(see Experimental Procedures). Cells (1 3 107)were incubated with 1 mCi/ml 32P-orthophos-phate in 2 ml of phosphate free DMEM for3 hr. Btk proteins were immunoprecipitatedusing affinity-purified antibodies specific tothe N-terminal region of Btk (N2) (Tsukada etal., 1993; Li et al., 1995), separated by SDS–PAGE, blotted onto nitrocellulose, and di-gested with trypsin. The tryptic peptides wereoxidized with performic acid and separatedby thin-layer electrophoresis at pH 1.9 fol-lowed by chromatography (Boyle et al., 1991).Phosphopeptides were detected by autoradi-ography, eluted from the thin-layer celluloseplate, and subjected to phosphoamino acidanalysis (Boyle et al., 1991). ps, pt, and pyindicate peptides containing phosphoserine,phosphothreonine, and phosphotyrosine, re-spectively.

Regulation of Btk by Autophosphorylation517

Figure 2. IgA Protease Cleavage of Btk

The schematic representation indicates thestructural domains of Btk, the possible sitesof IgA protease cleavage, and the N-terminaland C-terminal fragments generated by theIgA cleavage. At the bottom of the figure, par-tially purified wild-type protein, expressed byvaccinia virus coinfection with Lyn (see theExperimental Procedures) was incubated inthe absence (lanes 1, 3, 5) or presence (lanes2, 4, 6) of IgA protease. The digested frag-ments were separated by SDS–PAGE, andWestern blot analysis was performed usingpolyclonal antisera against the N-terminalportion (N-term, lanes 1, 2) or C-terminal 15aa of Btk (C-term, lanes 3, 4), or monoclonalantibodies against phosphotyrosine (anti-PY,lanes 5, 6). Full-length (80 kDa), N-terminal(30 kDa), and C-terminal (50 kDa) fragmentsare indicated.

the phosphopeptide patterns of a set of Btk* double Lyn from Btk. Btk protein was eluted from the columnusing high salt (0.5 M NaCl), while Lyn was recoveredmutants. Mutants that are either kinase negative (K430R)

or that contain a mutation that abolishes the Src kinase from the flow-through fraction (see Experimental Proce-dures). Btk protein was digested with IgA protease,consensus phosphorylation site (Y551F) were generated

in conjunction with Btk*. Previous studies by Li et al. which cleaves at two adjacent proline-rich motifs C-ter-minal to the Btk PH domain, resulting in an N-terminal(1995) showed that either of these mutations reduces

the tyrosine phosphorylation associated with Btk* and and a C-terminal fragment of protein. Western analysisusing antibodies specific for phosphotyrosine, Btk Nsuppresses the transforming activity in fibroblasts. Btk*/

Y551F abolished phosphorylation of py1 but retained terminus, and Btk C terminus revealed that Btk tyrosinephosphorylation sites were all located in the C-terminalpy2, indicating that the py1 peptide was derived from a

peptide phosphorylated onY551 (Figure 1C). In contrast, fragment, which includes the SH3, SH2, and SH1 do-mains (Figure 2).the kinase-inactive Btk*/K430R lost phosphorylation of

py2, py29, and py299 but retained phosphorylation of py1 To test whether the Btk autophosphorylation site re-sides in the SH3 domain, a Btk*/delSH3 double mutant(Figure 1D). Thus, while phosphorylation of Y551 does

not require Btk kinase activity, phosphorylation of all (Li et al., 1995) was stably expressed in NIH 3T3 cells.Tryptic phosphopeptide mapping revealed that deletionpy2-related peptides is dependent on Btk kinase activ-

ity. This indicates that the three py2 peptides (py2, py29, of the SH3 domain abolished all three py2 peptidesinduced by Btk*. In contrast, phosphorylation at py1and py299) contain autophosphorylation sites. The py2

peptide migrates at an identical position to the peptide was unaltered by the SH3 domain deletion (Figure 3A).In addition, BtkdelSH3 coexpressed with Lyn in NIH 3T3containing the Btk autophosphorylation site following

Btk/Lyn coexpression. These results suggest that Btk* cells retained phosphorylation of py1 (Y551) but lostphosphorylation of py2. Interestingly, coexpression ofand Btk/Lyn coexpression induce highly related phos-

phorylationpatternscontaining twomajor tyrosinephos- BtkdelSH3 with Lyn resulted in the production of anotherpeptide containing phosphotyrosine (py3) (Figure 3B).phorylation sites.

The lack of py2 phosphorylation in the absence of theSH3 domain suggested that the Btk autophosphoryla-The Major Btk Autophosphorylation Site

Resides within the SH3 Domain tion site resides within the SH3 domain. Alternatively,deletion of the SH3 domain may cause a structuralTo identify the autophosphorylation site in Btk, wild-

type Btk protein was coexpressed with Lyn in NIH 3T3 change in the protein, which prevents autophosphoryla-tion at a site distal to the SH3 domain. To determinecells using vaccinia virus. Btk was partially purified by

phosphocellulose chromatography, which separates whether the autophosphorylation site resides in the SH3

Immunity518

Figure 3. Phosphopeptide Map of Btk Containing SH3 Deletion

(A) Btk*/delSH3 stably expressed in NIH 3T3.(B) BtkdelSH3 coinfected with Lyn in NIH 3T3 cells using vacciniavirus.(C) In vitro phosphorylated GST–BtkSH3 fusion protein.Vaccinia virus expressing Btk or Lyn was coinfected into NIH 3T3cells for 12–15 hr prior to in vivo 32P04 labeling. For in vivo labeling,1 3 107 cells were incubated with 1 mCi/ml of 32P-orthophosphatein 2 ml of phosphate-free DMEM for 3 hr. Btk proteins were immuno-precipitated by affinity-purified Btk antibodies (N2) and digestedwith trypsin as described in the Experimental Procedures. GST–BtkSH3 fusion proteins were expressed in Escherichia coli and iso-lated as described in Experimental Procedures. GST–BtkSH3 pro-tein (4 mg) was phosphorylated by 50 ng of purified Btk protein invitro in 20 mM PIPES (pH 7.0), 20 mM MnCl2, 10 mM ATP, and 100mCi of [g–32P]ATP at 308C for 1 hr. In vitro phosphorylated GST–BtkSH3 was resolved by SDS–PAGE, transferred onto nitrocellulosemembrane, and digested by trypsin. The tryptic peptides were sepa-rated by thin-layer electrophoresis at pH 1.9, followed by chroma-tography (Boyle et al., 1991). The locations of py2 and py29 peptidesare indicated.

domain, GST–BtkSH3 fusion protein was used as a sub-strate for purified Btk in in vitro kinase assays. Full-length Btk containing a histidine tag at the C terminuswas purified to near homogeneity by phosphocellulosechromatography and a Ni21–NTA affinity column (seeExperimental Procedures). Btk was able to phosphory-late GST–SH3 but not GST protein alone (data notshown). Two tryptic phosphopeptides were recoveredfrom phosphorylated GST–SH3 protein whose migrationpatterns were indistinguishable from py2 and py29 pep- Figure 4. Physical Mapping of the Btk Autophosphorylation Sitetides derived from the Btk* protein (Figure 3C). Taken

(A) Strategy for mapping the Btk autophosphorylation site. NIH 3T3together, these results strongly suggest that a major Btk cells were coinfected with BTKY551F and Lyn vaccinia virus con-autophosphorylation site(s) is located within the Btk SH3 structs, labeled with 32P-orthophosphate, partially purified by phos-domain. phocellulose chromatography, and then immunoprecipitated by af-

finity-purified Btk antibodies (N2) as described in the ExperimentalProcedures. Immunoprecipitates were separated by SDS–PAGE,Y223 Is the Major Btk Autophosphorylation Siteblotted onto nitrocellulose, and digested with trypsin. The trypticRawlings et al. (1996) demonstrated that BtkY551Fpeptides containing phosphotyrosine were isolated using an anti-

coexpressed with Lyn retains the autophosphorylation phosphotyrosine column, separated by HPLC, and subjected tosite contained in py2 but loses py1 (containing Y551). automated Edman peptide sequence analysis for 10 cycles as de-To identify the tyrosine autophosphorylation site by scribed.

(B) HPLC profile and the amino acid sequences of phosphotyrosinephysical mapping, phosphorylated py2 peptide was en-containing peptides. At the top of the figure, the schematic diagramriched by coexpression of BtkY551F with Lyn by vac-of Btk indicates the domain structure of Btk, the location of tyrosinecinia virus in NIH 3T3 cells. Following in vivo labeling with223 within the SH3 domain, and the location of tyrosine 551 within32P-orthophosphate, purified Btk protein was digested the SH1 domain. A portion of the primary amino acid sequence of

with trypsin and phosphotyrosine-containing peptides Btk within the SH3 domain is shown. At the bottom of the picture,were isolated using an anti-phosphotyrosine antibody the HPLC elution profile demonstrates the separation of phosphoty-

rosine containing tryptic peptides. The anti-phosphotyrosine eluatecolumn. The peptides were separated by reverse phasewas separated using a C18 reverse phase HPLC and eluted with ahigh pressure liquid chromatography (HPLC) and radio-linear acetonitrile gradient of 5%–27% as described in the Experi-activity in each fraction was measured (Figure 4A). Onemental Procedures. Radioactivity in each HPLC fraction was mea-

major peak (B) and one minor peak (A) (Figure 4B) were sured (cpm). Two major peaks containing 32P were indicated as Aindependently subjected toautomated Edman degrada- and B. Peak fractions containing A or B were pooled, dried, andtion analysis for 10 cycles. Sequence analysis revealedthat peptides A and B contain overlapping sequences

Regulation of Btk by Autophosphorylation519

within the SH3 domain. Peptide B originated from apredicted tryptic digestion site. However, peptide A wasderived from a cleavage at L222, which probably wasgenerated by contaminating protease activity in the tryp-sin preparation. Independent microsequencing analysisof these two fractions showed preferential loss of posi-tion Y223, indicating it is the only phosphotyrosine-mod-ified residue in either peptide (Figure 4B).

Since the predicted tryptic peptide containing the Btkautophosphorylation site has another tyrosine residue,Y225, it is possible that generation of multiple py2 pep-tides arose from differential phosphorylation on two dif-ferent tyrosine residues in the same tryptic peptide. Totest this possibility, the tryptic peptides derived fromin vitro phosphorylated GST–SH3 protein by Btk weresequenced directly. Following trypticdigestion of invitrophosphorylated GST–BtkSH3 fusion protein, phospho-tyrosine containing peptides were isolated on an anti-phosphotyrosine antibody column, separated by HPLCfractionation, and subjected to 12 cycles of automatedEdman sequence analysis. The results showed that onlyY223 was phosphorylated in each peptide. Thus, thedifferent migration properties of py2 peptides do notresult from differential phosphorylation of the sametryptic peptide, but are likely due to the generation ofheterogeneous products by contaminating proteases.Conceivably, they could arise from additional proteinmodifications that occur within the same phospho-peptide.

Mutation of Y223 Potentiates Btk*Transforming ActivityBtk* is a weak transforming allele in rat fibroblasts com-pared with NIH 3T3 cells. However, Btk*-transformingactivity is dramatically potentiated by coexpression witha weakly activated form of c–src (srcE378G) in rat-2cells. srcE378G contains an activating mutation in thecatalytic domain (Levy et al., 1986) but does not trans-form rat-2 cells (D. E. H.A. et al., submitted). The synergy

Figure 5. Effect of Y223F Mutation in Btk* Transforming Activitybetween Btk* and srcE378G correlates with a significantParental rat-2 cells and cells expressing SrcE378G were infectedincrease in Y551 phosphorylation and enhanced Btkwith retroviruses encoding wild-type Btk, BtkY223F, Btk*, or Btk*/autophosphorylation. Btk*-transforming activity is fur-Y223F and were plated into soft agar in medium containing 20%ther enhanced by a block deletion of the SH3 domainfetal calf serum at 1 3 104 cells/plate. Colonies equal to and larger

when coexpressed with srcE378G. This supports the than 0.5 mm in diameter were counted after 3 weeks in soft agar.role of the Btk SH3 domain in regulating Btk activity The approximate locations of point mutations in the Btk subdomain(D. E. H. A. et al., submitted). are indicated by arrows in the schematic representations of the

different Btk forms. The number of colonies are indicated and repre-We tested whether the point mutation of Y223 to phe-sent averages from two different independent experiments. The pic-nylalanine, which cannot serve as a phosphate ac-tures were taken after colonies were grown for 3 weeks in agar.ceptor, stimulates or inactivates Btk*-mediated trans-

formation. If Btk autophosphorylation inactivates SH3function similar to a deletion of the SH3 domain, muta-

These Btk forms were introduced into either rat-2 cellstion of Y223 would activate Btk signaling leading toor rat-2 cells coexpressing srcE378G. Tryptic phospho-enhanced Btk* transformation. Alternatively, if Btk auto-peptide mapping of in vivo 32P-labeled Btk moleculesphosphorylation serves as an activation mechanism,showed that the Y223F mutation abolished phosphory-then mutation of Y223 would reduce Btk* transforma-lation of py2 peptides (data not shown).tion. To test these possibilities, the Y223F mutation was

Rat-2 cells expressing wild-type Btk or Btk* formedgenerated in the context of wild-type Btk and Btk*.only a low number (<10) of small colonies (<0.5 mmdiameter) per 104 cells plated in soft agar (Figure 5). Aspreviously shown, coexpression of Btk* with SrcE378G

subjected to 10 cycles of automated Edman degradation sequencein rat-2 cells resulted in the growth of numerous (>350),analysis. PH, pleckstrin homology domain; SH, src homology do-medium- to large-sized colonies (0.5–1.5 mm in diame-mains; pro, proline-rich sequences. Locations of Y223 and Y551 are

indicated. ter) in agar after 3 weeks in culture. Mutation of Y223

Immunity520

Table 1. Differences in Transforming Activity of Btk* and Btk*/Y223F

Number of Colonies in Soft AgarRat-2 cells 5% Serum Number of Cells Plates (3 103) 20% Serum Number of Cells Plates (3 103)plated insoft agar 2 5 10 2 5 10

Mock ,10 ,10 ,10 ,10 ,10 ,20BTK* 90 140 230 160 260 420BTK*Y223F 300 500 .800 560 .800 .800

Rat-2 cells expressing SrcE378G were infected with retroviruses encoding Btk* or Btk*/Y223F. Cells (2 3 103, 5 3 103, or 1 3 104) were platedin soft agar in either 5% or 20% fetal calf serum. The growth of colonies in soft agar was allowed to proceed for 3 weeks, at which time thecolonies larger than 0.5 mm in diameter were counted. The transforming strength was discernible between Btk* and Btk*/Y223F within 10days after plating in soft agar.

in Btk* resulted in a dramatic increase in number (>800) 3). Btk*/Y223F exhibited a small increase in kinase activ-ity (<2-fold) compared with Btk* in both autophosphor-and sizeof colonies (1.0–1.5 mm in diameter) when coex-

pressed with SrcE378G (Figure 5). Btk*/Y223F induced ylation and transphosphorylation reactions (Figure 6,lane 4).more rapidly the appearance of colonies and the acidifi-

cation of the media (within 2 weeks) than Btk* did (within We also compared the transphosphorylation activitybetween Btk* and Btk*/Y223F using a synthetic peptide3–4 weeks).

To quantify better the difference in transforming activ- of 12 aa containing sequences surrounding Y223 asa substrate. The transphosphorylation activity of Btk*/ity between Btk* and Btk*/Y223F, rat-2 cells coexpress-

ing these Btk mutants with SrcE378G were plated in Y223F was slightly elevated compared with Btk* (<2-fold), which was consistent with the enolase phosphory-soft agar at three different cell concentrations in either

5% or 20%serum (Table 1). Cells expressing Btk*/Y223F lation results (data not shown). This demonstrates thatthe Y223F mutation does not dramatically enhance Btkformed at least three times more colonies than cells

expressing Btk* under all conditions (Table 1). kinase activity. Thus, increase in transforming activityIn separate experiments, NIH 3T3 cells were infected

with retroviruses encoding wild-type Btk, BtkY223F,Btk*, or Btk*/Y223F and plated in soft agar at 1 3 104

cells/plate in 5% or 20% serum. Btk wild-type orBtkY233 did not induce agar colony formation. Cellsexpressing Btk*/Y223F formed numerous (>800), large-sized (1.0–2.0 mm indiameter) colonies within 3–4 weeksafter plating. Cells expressing Btk* gave rise to only afew (50–100), medium-sized colonies under these condi-tions (data not shown).

Y223F Mutation Does Not Significantly AffectBtk Kinase ActivityThe increase in biological activity of Btk*/Y223F may beexplained by a dramatic increase in Btk kinase activity,which is illustrated in the cases of transforming mutantsof abl and src oncogenes (Coussens et al., 1985; Iba etal., 1985; Konopka and Witte, 1985; Lugo et al., 1990).To test whether the Y223F mutation enhances tyrosinekinase activity, we compared the in vitro kinase activityof Btk in the presence or absence of Y223F mutation.Wild-type Btk, BtkY223F, Btk*, Btk*/Y223F, and kinase-inactive Btk*/K430R were coexpressed with SrcE378Gin rat-2 cells. Btk proteins were immunoprecipitated and

Figure 6. Tyrosine Kinase Activities of Btk* or Btk*/Y223F Coex-analyzed for autokinase and transphosphorylation activ- pressed with SrcE378Gity using enolase as a substrate. Btk*/K430R exhibited

Rat-2 cells expressing SrcE378G were infected with retrovirusesno autokinase and very low transphosphorylation activ-encoding wild-type Btk (lane 1), BtkY223F (lane 2), Btk* (lane 3),

ity (Figure 6, lane 5). This indicates that the immunopre- Btk*/Y223F (lane 4), and kinase-inactive Btk*/K430R (lane 5). Btkcipitates were largely free of contaminating kinases. proteins were immunoprecipitated from 1 3 107 cells using antibod-

ies directed against Btk (N2). Btk autokinase as well as transphos-When the extent of phosphorylation is normalized tophorylation activity using enolase as a substrate (2 mg/reaction) wasthe amount of protein present in each kinase assay,measured in 20 mM PIPES (pH 7.0), 20 mM MnCl2, 20 mM ATP, 10BtkY223F had similar autokinase and transphosphoryla-mCi of [g–32P]ATP at 308C for 10 min. Phosphorylated proteins weretion activities compared with wild-type Btk (Figure 6,resolved by SDS–PAGE and detected by autoradiography. The level

lanes 1 and 2). As previously shown by Li et al. (1995), of Btk proteins was analyzed by Western blot analysis of the immu-the in vitro autokinase activity of Btk* protein was not noprecipitates using anti-Btk antibodies (N2) (aBtk). The positions

of Btk and enolase are indicated by arrows.enhanced compared with wild-type Btk (Figure 6, lane

Regulation of Btk by Autophosphorylation521

of Btk*/Y223F is most likely due to an alteration in signal Kinetic studies by Saouafet al. (1994) showed sequen-tial activationof cytoplasmic tyrosine kinases upon sIgMtransduction mediated in some manner by the SH3

domain. cross-linking of B cells. Btk activation follows that ofLyn and Blk, while Syk becomes activated graduallyfollowing Btk activation. This suggested that Btk tyro-

Discussion sine phosphorylation in B cells may be dependent onactivation of Src family tyrosine kinases. Genetic analy-

A Two-Step Mechanism for Btk sis of animals carrying null alleles of Fyn or Lyn demon-Tyrosine Phosphorylation strate the specific role of these src family tyrosine ki-In this report, we show that two different modes of acti- nases in Btk signaling. Although Fyn knockout animalsvation lead to phosphorylation of the same two tyrosine do not display any severe abnormality in B cell develop-residues of Btk. The same tyrosine phosphorylations ment, B cells derived from these animals are defectivecan be induced upon sIgM cross-linking of a human B in IL-5 signaling (Appleby et al., 1995). B cells from Lyncell line (Rawlings et al., 1996), which correlates with knockouts show impaired response to sIgM stimulationenhancement of Btk autokinase activity. While phos- and lack of B-1 B cell production (Hibbs et al., 1995).phorylation of Btk at Y551 by Src family kinases is crucial Btk knockout animals display an Xid phenotype withfor activating Btk function, it is not yet clear how phos- impaired responses in IL-5 and sIgM signaling (Khan etphorylation at Y223 regulates Btk signaling in B cells. al., 1995; Kerner et al., 1995). Thus, Lyn and Fyn appear

Contrary to a recent report by Mahajan et al. (1995), to participate in distinct Btk-mediated signaling duringwe demonstrate that Y551 is a transphosphorylation B cell development.site. They argue that Y551 is a major Btk autophosphory-lation site, since a Y551F mutation abolishes Btk tyro-

SH3 Domain in Btk Signalingsine phosphorylation in their experiments. However, weAutophosphorylations of Src family and receptor tyro-have observed that the Y551F mutant immunoprecipi-sine kinases occur in their catalytic domains and en-tated from NIH 3T3 cells is still kinase active and be-hance their enzymatic activities (reviewed by Coopercomes autophosphorylated at othersites (Li et al., 1995).and Howell, 1993; Kazlauskas, 1994). In contrast, theThe discrepancy may be due to different expressionmajor Btk autophosphorylation site is mapped to Y223levels of catalytically active Btk proteins in two indepen-in the SH3 domain. The precise role of the Btk SH3dent experimental settings. Our mapping studies anddomain in signaling and its ligand specificity is notmutational analyses clearly illustrate that the major Btkknown. Mutations in the SH3 domain of c–src or c–ablautophosphorylation occurs at Y223, which resides inincrease their transforming potentials, suggesting thata distinct peptide from that containing Y551. Mahajantheir SH3 domains have a role in down-regulation ofet al. (1995) showed that CNBr cleavage of autophos-their kinase activity (Jackson and Baltimore, 1989; Franzphorylated Btk yielded a single-sized phosphopeptideet al., 1989; Hirai and Varmus, 1990; Jackson et al., 1993;that they thought was phosphorylated only at Y551.Mayer and Baltimore, 1994; Okada et al., 1993). DeletionHowever, CNBr cleavage of Btk protein can generateof the Btk SH3 domain enhances Btk*-mediated trans-two peptides of about 7 kDa containing either Y551 (61formation of fibroblasts upon stimulation by SrcE378Gaa from D510–M570) or Y223 (63 aa from G164–M226).(D. E. H. A. et al., submitted). Since deletion or mutationSince these two peptides could not be resolved by one-of the Btk SH3 domain does not appear to affect Btkdimensional SDS–PAGE, the peptides derived from au-kinase activity directly, it is likely to regulate Btk functiontophosphorylated Btk by CNBr cleavage in their reportby binding to regulatory molecules.likely contain species phosphorylated at Y551 and Y223.

SH3 domains bind to peptides containing proline-richsequences. The SH3 domain core is composed of twoperpendicular and antiparallel three-stranded b-pleatedRole of Src Family Tyrosine Kinases

in Btk Activation sheets. The hydrophobic surface contains shallowgrooves formed by highly conserved aromatic aminoThe Btk* mutation increases membrane localization of

Btk (Li et al., 1995). A recent structural study shows that acids that contact proline-rich peptides. Recent struc-tural and mutational analyses of SH3 domains from aPIP2 binds to the PH domain at the site of the Btk*

mutation (E41K) (Ferguson et al., 1995). This suggests wide range of signaling molecules identified importantresidues involved in these protein–protein interactionsthat Btk* may alter the binding affinity of the Btk PH

domain to specific phosphoinositol lipids that may aid (Knudsen et al., 1995; Maignan et al., 1995; Goudreauet al., 1994; Terasawa et al., 1994; Yu et al., 1994; Koy-Btk association with membranes. We have demon-

strated that activation of Src family tyrosine kinases ama et al., 1993; Kohda et al., 1993; Lim et al., 1994;Feng et al., 1994). Based on structural comparison withpotentiate Btk* transforming activity, while suppression

of Src family tyrosine kinases by Csk diminishes Btk* the Fyn SH3 domain, Btk Y223 represents one of thehighly conserved aromatic residues forming the surfaceactivation (D. E. H. A. et al, submitted). This data sup-

ports the idea that Src family tyrosine kinases act as of the peptide-binding groove (Noble et al., 1993). Muta-tional analysis of the c–src SH3 domain demonstratedupstream regulators of Btk. In this assay, Src kinases

show synergy only with Btk* not with wild-type Btk. Btk* that Y90 of c–src (equivalent to Btk Y223) is critical forSH3-mediated protein–protein interactions (Erpel et al.,is presumably more accessible to activated Src kinases,

which results in enhanced tyrosine phosphorylation and 1995). Thus, Y223 of Btk is highly likely to play a crucialrole in ligand binding.Btk activation.

Immunity522

prepared by transient transfection of 293T cells (Muller et al., 1991;Function of Phosphorylation at Y223Pear et al, 1993) and were used to infect NIH 3T3 cells. Stable cellin the SH3 Domainlines expressing Btk were generated by selection in G418. Recombi-Phosphorylation of a tyrosine residue in the SH3 domainnant vaccinia viral constructs expressing Lyn or BtkdelSH3 were

of c–src has been recently implicated as a mechanism produced as described previously (Rawlings et al., 1996).of regulating SH3-mediated signaling. The Y138 residueof the c–src SH3 domain (equivalent to Y263 of Btk) has Construction of GST–BtkSH3 Fusion Protein

Polymerase chain reaction was performed to amplify the regionbeen shown to be phosphorylated by the PDGF receptorencoding the SH3 domain (nucleotides 787–951) from murine Btk(M. Broome and T. Hunter, personal communication).cDNA using the primers SH3-59 (59-GTGGATCCAAGGTCGTGGCCCThe Y138F mutation acts as a dominant negative inhibi-TTTAT-39) and SH3-39 (59-CCGAATTCTGTAGCCTTCCTGCCCAT).

tor of PDGF-stimulated mitogenesis without inhibiting The polymerase chain reaction product was cloned into pGEX-2TSrc catalytic activity (M. A. Broome and T. Hunter, sub- (Pharmacia Biotech). The fusion protein was induced and purifiedmitted). Interestingly, phosphorylation at Y138 signifi- as described previously (Frangioni and Neal, 1993).cantly diminishes SH3 binding affinity for peptide li-

Site-Directed Mutagenesisgands containing proline-rich sequences, suggesting aAn EcoRI fragment of Btk (nucleotides 136–1457) containing therole for this phosphorylation in SH3-mediated ligandBtk* mutation was cloned into M13mp18 vector and single-stranded

binding. DNA was generated. Site-directed mutagenesis was performed onWe have shown that mutation of Y223 of Btk potenti- single-stranded DNA template using an oligomer (59-ATGTAATCGA

ates Btk* transforming activity. A block deletion of the AAAGGGCCA-39) containing phenylalanine at theY223 position (Am-ersham). DNA sequencing was performed to verify the presence ofSH3 domain of Btk* induces a similar phenotype, sug-the Y223F mutation. Constructs containing BtkY223F or Btk*/Y223Fgesting that mutation of Y223 is equivalent to inactiva-were generated in the pSRaMSVtk–neo vector (Muller et al., 1991)tion of SH3 function. This leaves us with two alternativeand were transfected into 293T cells to generate helper-free retro-interpretations. First, activation of Btk by phosphory-virus.

lation at Y551 leads to autophosphorylation at Y223,which would down-regulate Btk activity. In this case, In Vivo 32P04-Labeling, Two-Dimensional Trypticmutation of Y223 would prevent down-regulation of Btk Phosphopeptide Mapping, and Phosphoamino

Acid Analysisby autophosphorylation and result in activation of BtkCells (1 3 107) were used for 32P-orthophosphate labeling. To allowfunction. Alternatively, both mutation of Y223 or phos-sufficient protein expression, vaccinia viruses encoding Btk werephorylation at Y223 may alter SH3 function and lead toinfected into 1 3 107 NIH 3T3 cells 12–15 hr prior to in vivo 32P-labeling

activation of Btk signaling. Since the Y223F mutation (Rawlings et al., 1996). Cells were washed with phosphate-freedoes not significantly change Btk catalytic activity, it is DMEM and labeled with 1 mCi/ml of 32P-orthophosphate (New En-likely that in either case Y223 in the SH3 domain regu- gland Nuclear) in 2 ml of phosphate-free DMEM for 3 hr. Cell extracts

were prepared in lysis buffer (20 mM HEPES [pH 7.4], 100 mM NaCl,lates Btk activity by modulating binding to cellular regu-3% Triton X-100, 10% glycerol, 2 mM EDTA, 1 mM Na3V04, 10 mMlatory molecules. One must certainly include the possi-NaF, and protease inhibitors including 1 mM PMSF, 5 mg/ml aproti-bility that phosphorylation of Y223 creates a dockingnin, and 5 mg/ml leupeptin). Btk was immunoprecipitated with 20

site for a SH2 containing protein or proteins.mg of affinity-purified antibody (N2), which specifically recognizes

A genetic study in C. elegans implicates the SLI-1 the N-terminal portion of Btk (Tsukada et al., 1993; Li et al, 1995).gene product, a homolog of mammalian c–cbl, as a Immunoprecipitates were separated by SDS–PAGE (6% polyacryl-

amide), blotted onto nitrocellulose membrane, and visualized bynegative regulator of a receptor tyrosine kinase pathwayautoradiography. Btk protein was excised from the membrane and(Yoon et al., 1995). This suggests a similar role for c–cbldigested with 15 mg TPCK–trypsin (Sigma) twice for 3 hr at 378C inin down-regulating tyrosine kinase signaling pathways50 mM NH4HCO3, lyophilized, and treated with performic acid at

in mammalian cells. Recently, it was reported that a GST 48C for 1 hr (Luo et al., 1991). Two-dimensional analysis of trypticfusion protein containing the SH3 domain of Btk can peptides was performed by thin-layer electrophoresis at pH 1.9,bind to the c–cbl protooncogene product from B cell followed by chromatography in phosphochromatography buffer

(Boyle et al., 1991). For phosphoamino acid analysis, peptides wereextracts (Cory et al., 1995). It is not known whether full-extracted from thin-layer cellulose plates using pH 1.9 electrophore-length Btk is able to interact with c–cbl. Autophosphory-sis buffer and hydrolyzed for 1 hr at 1108C in 6 Na–HCl. 32P-labeledlation at Y223 is predicted to increase negative electro-amino acids were mixed with phosphoamino acid standards and

static potential and perturb the hydrophobic nature of separated by thin layer electrophoresis at pH 3.5 (Boyle et al., 1991).the ligand-binding surface of the SH3 domain. This is Standards were developed by ninhydrin staining and 32P-labeledlikely to attenuate binding of proline-rich peptides to amino acids were identified by autoradiography.the SH3 domain. Btk autophosphorylation substantially

Large Scale Purification of Btk from Vacciniaincreases upon B cell activation but the role of Btk auto-Virus–Infected Cellsphosphorylation in B cell signaling is not known. A fur-NIH 3T3 cells (2 3 109) were coinfected with vaccinia viruses ex-ther search for B cell proteins that interact with subdo-pressing BtkY551F and Lyn. After a 12–18 hr infection, 5 3 107 cells

mains of Btk should provide candidate regulators of Btk were washed with phosphate-free DMEM and incubated with 10–20function. mCi of 32P-orthophosphate in 5–10 ml of phosphate-free DMEM plus

0.1% dialyzed fetal calf serum for an additional 4–6 hr. Cells weretreated with 200 mM of Na3VO4 1 hr prior to harvesting and lysed inExperimental Proceduresthe lysis buffer as described in the previous section. Cell extractswere applied to a phosphocellulose column (Whatman) and washedCell Culture and Retrovirus or Vaccinia

Virus–Mediated Transfer with buffers containing low salt and detergent (50 mM Tris [pH 7.4],150 mM NaCl, 0.1% Triton X-100, 10% glycerol, 1 mM Na3V04),NIH 3T3 cells were grown in Dulbecco’s modified Eagle’s medium

(DMEM) supplemented with 10% calf serum. Recombinant Btk was medium salt anddetergent (300mM NaCl), low salt with nodetergent(150 mM NaCl), and then eluted with high salt buffer without deter-cloned into the retroviral expression vector, pSRaMSVtk–neo, as

described by Li et al., (1995). Helper-free retroviral stocks were gent (500 mM NaCl). By immunoblot analysis, Btk was observed to

Regulation of Btk by Autophosphorylation523

elute with 500 mM NaCl, while Lyn was recovered in the initial flow- samples were plated in duplicate in medium containing either 5%or 20% fetal calf serum. Colonies equal to or larger than 0.5 mm inthrough fraction. The phosphocellulose eluate was concentrated

through a Centricon C-30 filter (Amicon) and stored at 2208C with diameter were counted 2–4 weeks after plating.25% glycerol. Btk was immunoprecipitated with affinity-purified an-tibodies (N2) from the phosphocellulose eluate, separated by SDS– In Vitro Btk Kinase AssayPAGE, blotted onto nitrocellulose, and visualized by autoradiogra- Rat-2 cells expressing SrcE378G were infected with retrovirusesphy. Btk protein was digested with 15–25 mg of TPCK–trypsin twice encoding different forms of Btk. Cell extracts were prepared in lysisfor 90 min at 378C, dried, and resuspended in ddH2O. The tryptic buffer (20 mM Tris [pH 7.4], 150 mM NaCl, 1% Triton X-100, 2 mMpeptides were denatured at 908C for 10 min and then applied to EDTA, 1mM Na3V04, 10 mM NaF, and protease inhibitors includingan anti-phosphotyrosine column (1G2-Sepharose). The column was 1 mM PMSF, 5 mg/ml aprotinin, and 5 mg/ml leupeptin) and Btk waswashed with 50 mM ammonium acetate (pH 7.0), and then eluted immunoprecipitated by antibodies recognizing Btk (N2). Immuno-with 50 mM ammonium acetate and 500 mM acetic acid (pH 3.7). precipitates were washed four times in lysis buffer and the kinase

activity of Btk proteins was measured in kinase buffer (20 mM PIPESHPLC Purification of Tryptic Peptides [pH 7.0], 20 mM MnCl2, 20 mM ATP, 10 mCi of [g–32P]ATP) includingand Peptide Sequencing 2 mg of enolase as an exogenous substrate for 15 min at 308C. TheThe eluate from the anti-phosphotyrosine column was concentrated reaction mixtures were boiled in SDS sample loading buffer andto 250 ml and applied to a Betasil C18 250 mm 3 4.6 mm reverse separated by SDS–PAGE (8%). Phosphorylated proteins were de-phase HPLC column (Keystone Scientific), using a flow rate of 1.25 tected by autoradiography. The level of Btk protein was measuredml/min. After a 10 min wash at 5% acetonitrile, the phosphopeptides by Western blot analysis using antibodies against Btk (N2).were eluted with a linear acetonitrile gradient of 5%–27%. Fractionswere obtained at 1 min intervals and radioactivity in each fraction Acknowledgmentswas measured. Peak fractions containing 32P were pooled, dried,and subjected to ten cycles of automated Edman degradation se- Correspondence should be addressed to O. N. W. We thank D.quence analysis. Blank cycles at the appropriate position indicated Saffran, and A. Satterthwaite for their helpful comments, and J.the presence of a phosphotyrosine residue. Shimaoka and J. White for assistance in preparation of the manu-

script and figures. We are grateful to Dr. K. Faull for provision ofIgA Protease Cleavage HPLC facilities. We are grateful to Drs. L. Zipursky, S. Smale, andBtk proteins expressed by vaccinia virus in 1 3 107 NIH 3T3 cells D. Black for their critical comments on the manuscript. We thank T.were partially purified by phosphocellulose chromatography as de- Li for providing us with Btk retroviruses and R. M. Kato for excellentscribed above. The Btk proteins were incubated with 1 mg of IgA technical assistance. This work is supported by a fellowship fromprotease (Boehringer Mannheim) in 50 mM Tris (pH 7.4) at 168C for the Cancer Research Institute (H. P.), Human Genetics Training grant16 hr, lyophilized, separated by SDS–PAGE (15%), and blotted onto GM08243 (M. W.), Special Fellowship of the Leukemia Society ofnitrocellulose. Western blot analysis was performed using poly- America (D. E. H. A.), National Institutes of Health Physician Scientistclonal antibodies that specifically recognize either the N-terminal Award AR01912 (D. J. R.), Pediatric Scientist Training Grant fel-(N2) or the C-terminal 15 aa of Btk, ora monoclonal antibody directed lowship (A. S.), and National Cancer Institute Grants (O. N. W.).against phosphotyrosine (4G10, Upstate Biotechnology, Incorpo- O. N. W. is an Investigator of the Howard Hughes Medical Institute.rated). Secondary antibody was either goat anti-rabbit or goat anti-mouse conjugatedto horse radish peroxidase (Bio-Rad). The protein Received November 5, 1995; revised April 2, 1996.was detected using chemiluminescent signal (ECL, Amersham).

ReferencesPurification of Histidine-Tagged Btk ProteinWild-type Btk containing a six histidine tag at the C terminus was Aoki, Y., Isselbacher, K.J., and Pillai, S. (1994). Bruton tyrosine ki-constructed in the pCD-SRa296 expression vector (Takee et al., nase is tyrosine phosphorylated and activated in pre-B lymphocytes1988). Btk was expressed in 293T cells by transient transfection. and receptor-ligated B cells. Proc. Natl. Acad. Sci. USA 91, 10606–Cells were lysed in lysis buffer 2 days after transfection (see the 10609.previous sections). Cell extracts were applied to a phosphocellulose Appleby, M.W., Kerner, J.D., Chien, S., Maliszewski, C.R., Bondada,column (Whatman) in 50 mM Tris (pH 7.4), 150 mM NaCl, 5 mM S., and Perlmutter, R.M. (1995). Involvement of p59 fynT in IL-5EDTA, 1% Triton-X 100, 10% glycerol, and protein inhibitors (1 mM receptor signaling. J. Exp. Med 182, 811–820.PMSF, 5 mg/ml aprotinin, and 5 mg/ml leupeptin). The column was

Boyle, W.J., van der Geer, P., and Hunter, T. (1991). Phosphopeptidewashed with 50 mM Tris (pH 7.4), containing 250 mM NaCl and Btkmapping and phosphoamino acid analysis by two-dimensional sep-protein was eluted with 300–500 mM NaCl in 50 mM Tris (pH 7.4).aration on thin-layer cellulose plates. Meth. Enzymol. 201, 110–149.The eluted proteins were applied to a Ni21–NTA agarose columnCambier, J.C., Pleiman, C.M., and Clark, M.R. (1994). Signal trans-(Qiagen). The column was washed with 5 mM imidazole (pH 7.0)duction by the B cell antigen receptor and its coreceptors. Annu.and the Btk protein was eluted with 25 mM imidazole (pH 7.0) inRev. Immunol. 12, 457–486.15% glycerol. Btk protein was concentrated using Centricon C-30

filters (Amicon) and stored at 2208C. Cooper, J.A., and Howell, B. (1993). The when and how of src regula-tion. Cell 63, 1051–1054.

In Vitro Phosphorylation of GST–BtkSH3 Fusion Protein Cory, G.O.C., Lovering, R.C., Hinshelwood, S., MacCarthy-Morrogh,For two-dimensional peptide mapping analysis, 4 mg of GST– L., Levinsky, R.J., and Kinnon, C. (1995). The protein product ofBtkSH3 protein was phosphorylated by 50 ng of purified histidine- the c–cbl protooncogene is phosphorylated after B cell receptortagged Btk protein in 20 mM PIPES (pH 7.0), 20 mM MnCl2, 20 mM stimulation and binds the SH3 domain of Bruton’s tyrosine kinase.ATP, and 100 mCi of [g–32P]ATP for 1 hr at 308C. Phosphorylated J. Exp. Med. 182, 611–615.proteins were separated by SDS–PAGE (12%), visualized by autora-

Coussens, P.M., Cooper, J.A., Hunter, T., and Shalloway, D. (1985).diography, transferred onto nitrocellulose membrane, visualized byRestriction of the in vitro and in vivo tyrosine protein kinase activitiesautoradiography, and subjected to two-dimensional tryptic peptideof pp60c–src relative to pp60v–src. Mol. Cell. Biol. 5, 2753–2763.mapping as described above.de Weers, M., Mensink, R.G.J., Kraakman, M.E.M., Schuurman,R.K.B., and Hendriks, R.W. (1994). Mutation analysis of the Bruton’sSoft Agar Transformation Assaytyrosine kinase gene in X-linked agammaglobulinemia: identificationSoft agar transformation assays were performed as described pre-of a mutation which affects the same codon as is altered in immuno-viously (D. E. H. A. et al., submitted). Parental rat-2 cells or rat-2deficient xid mice. Human Mol. Genet. 3, 161–166.cells expressing SrcE378G were infected with retroviruses encoding

different Btk forms, grown for 48–72 hr, and then plated in soft agar Erpel, T., Superti-Furga, G., and Courtneidge, S.A. (1995). Mutationalanalysis of the src SH3 domain: the same residues of the ligandat different cell densities (2 3 103–1 3 104 cells/6 cm dish). The

Immunity524

binding surface are important for intra- and intermolecular interac- Schreiber, S.L. (1993). Structure of the PI3K SH3 domain and analy-sis of the SH3 family. Cell 72, 945–952.tions. EMBO J. 14, 963–975.

Levy, J.B., Hanafusa, H., and Iba, H. (1986). Activation of the trans-Feng, S., Chen, J.K., Yu, H., Simon, J.A., and Schreiber, S.L. (1994).forming potential of p60c–src by a single amino acid change. Proc.Two binding orientations for peptides to the Src SH3 domain: devel-Natl. Acad. Sci. USA 83, 4228–4232.opment of a general model for SH3–ligand interactions. Science

266, 1241–1247. Li, T., Tsukada, S., Satterthwaite, A., Havlik, M.H., Park, H., Takatsu,K., and Witte, O.N. (1995). Activation of Bruton’s tyrosine kinaseFerguson, K.M., Lemmon, M.A., Schlessinger, J., and Sigler, P.B.(Btk) by a point mutation in its pleckstrin homology (PH) domain.(1995). Structure of the high affinity complex of inositol triphosphateImmunity 2, 451–460.with a phospholipase c pleckstrin homology domain. Cell 83, 1037–

1046. Lim, W.A., Richards, F.M., and Fox, R.O. (1994). Structural determi-nants of peptide-binding orientation and of sequence specificity inFrangioni, J.V., and Neal, B.G. (1993). Solubilization and purificationSH3 domains. Nature 372, 375–379.of enzymatically active glutathione S-transferase (pGEX) fusion pro-

teins. Anal. Biochem. 210, 179–187. Lugo, T.G., Pendergast, A.-M., Muller, A.J., and Witte, O.N. (1990).Tyrosine kinase activity and transforming potency of bcr–abl onco-Franz, W.M., Berger, P., and Wang, J.Y.J. (1989). Deletion of angene products. Science 247, 1079–1082.N-terminal regulatory domain of the c–abl tyrosine kinase activates

its oncogenic potential. EMBO J. 8, 137–147. Luo, K., Hurley, T.R., and Sefton, B.M. (1991). Cyanogen bromidecleavage and proteolytic peptide mapping of proteins immobilizedGoudreau, N., Cornille, F., Duchesne, M., Parker, F., Tocque, B.,to membranes. Meth. Enzymol. 201, 149–152.Garbay, C., and Roques, B.P. (1994). NMR structure of the N-terminal

SH3 domain of GRB2 and its complex with a proline-rich peptide Mahajan, S., Fargnoli, J., Burkhardt, A.L., Kut, S.A., Saouaf, S.J.,from Sos. Nature Struct. Biol. 1, 898–907. and Bolen, J.B. (1995). Src family protein tyrosine kinases induce

autoactivation of Bruton’s tyrosine kinase. Mol. Cell. Biol. 15, 5304–Hanks, S.K., Quinn, A.M., and Hunter, T. (1988). The protein kinase5311.family: conserved features and deduced phylogeny of the catalytic

domains. Science 241, 42–52. Maignan, S., Guilloteau, J.P., Fromage, N., Arnoux, B., Becquart, J.,and Ducruix, A. (1995). Crystal structure of the mammalian Grb2Hibbs, M.L., Tarlinton, D.M., Armes, J., Grail, D., Hodgson, G., Mag-adaptor. Science 268, 291–293.litto, R., Stacker, S.A., and Dunn, A.R. (1995). Multiple defects in the

immune system of Lyn-deficient mice, culminating in autoimmune Matsuda, T., Takahashi-Tezuka, M., Fukada, T., Okuyama, Y., Fuji-disease. Cell 83, 301–311. tani, Y., Tsukada, S., Mano, H., Hirai, H., Witte, O.N., and Hirano, T.

(1995). Association and activation of Btk and Tec tyrosine kinasesHirai, H., and Varmus, H.E. (1990). Site-directed mutagenesis of theby gp130, a signal transducer of the IL-6 family of cytokines. BloodSH2- and SH3-coding domains of c–src produces varied pheno-85, 627–633.types, including oncogenic activation of p60c–src. Mol. Cell. Biol.

10, 1307–1318. Mayer, B.J., and Baltimore, D. (1994). Mutagenic analysis of theroles of SH2 and SH3 domains in regulation of the Abl tyrosineIba, H., Cross, F.R., Garber, E.A., and Hanafusa, H. (1985). Low levelkinase. Mol. Cell. Biol. 14, 2883–2894.of cellular protein phosphorylation by nontransforming overpro-Muller, A.J., Young, J.C., Pendergast, A.-M., Pondel, M., Landau,duced p60c–src. Mol. Cell. Biol. 5, 1058–1066.N.R., Littman, D.R., and Witte, O.N. (1991). BCR first exon sequencesJackson, P., and Baltimore, D. (1989). N-terminal mutations activatespecifically activate Bcr/Abl tyrosine kinase oncogene of Philadel-the leukemogenic potential of the myristoylated form of c–abl.phia chromosome positive human leukemias. Mol. Cell. Biol. 11,EMBO J. 8, 449–456.1785–1792.

Jackson, P.K., Paskind, M., and Baltimore, D. (1993). Mutation of aNoble, M.E.M., Musacchio, A., Saraste, M., Courtneidge, S.A., andphenylalanine conserved in SH3-containing tyrosine kinases acti-Wierenga, R.K. (1993). Crystal structure of the SH3 domain in humanvates the transforming ability of c–abl. Oncogene 8, 1943–1956.Fyn; comparison of the three-dimensional structures of SH3 do-

Kawakami, Y., Yao, L., Miura, T., Tsukada, S., Witte, O.N., and Kawa- mains in tyrosine kinases and spectrin. EMBO J. 12, 2617–2624.kami, T. (1994). Tyrosine phosphorylation and activation of Bruton

Okada, M., Howell, B.W., Broome, M.A., and Cooper, J.A. (1993).tyrosine kinase (Btk) upon FceRI cross-linking. Mol. Cell. Biol. 14,Deletion of the SH3 domain of Src interferes with regulation by5108–5113.the phosphorylated carboxyl-terminal tyrosine. J. Biol. Chem. 268,

Kazlauskas, A. (1994). Receptor tyrosine kinases and their targets. 18070–18075.Curr. Opin. Genet. Dev. 4, 5–14.

Pear, W.S., Nolan, G.P., Scott, M.L., and Baltimore, D. (1993). Pro-Kerner, J.D., Appleby, M.W., Mohr, R.N., Chien, S., Rawlings, D.J., duction of high titer helper-free retroviruses by transient transfec-Maliszewski, C.R., Witte, O.N., and Perlmutter, R.M. (1995). Impaired tion. Proc. Natl. Acad. Sci. USA 90, 8392–8396.expansion of mouse B cell progenitors lacking Btk. Immunity 3,

Perlmutter, R.M., Levin, S.D., Appleby, M.W., Anderson, S.J., and301–312.Ila-Alberola, J. (1993). Regulation of lymphocyte function by protein

Khan, W.N., Alt, F.W., Gerstein, R.M., Malynn, B.A., Larsson, I., Rath- phosphorylation. Annu. Rev. Immunol. 11, 451–499.bun, G., Davidson, L., Muller, S., Kantor, A.B., Herzenberg, L.A.,

Rawlings, D.J., and Witte, O.N. (1995). The Btk subfamily of cyto-Rosen, F.S., and Sideras, P. (1995). Defective B cell developmentplasmic tyrosine kinases: structure, regulation, and function. Sem.and function in Btk-deficient mice. Immunity 3, 283–299.Immunol. 7, 237–246.

Kinnon, C., Hinshelwood, S., Levinsky, R.J., and Lovering, R.C.Rawlings, D.J., Saffran, D.C., Tsukada, S., Largaespada, D.A., Gri-(1993). X-linked agammaglobulinemia: gene cloning andfuture pros-maldi, J.C., Cohen, L., Mohr, R.N., Bazan, J.F., Howard, M., Cope-pects. Immunol. Today 14, 554–558.land, N.G., Jenkins, N.A., and Witte, O.N. (1993). Mutation of the

Knudsen, B.S., Zheng, J., Feller, S.M., Mayer, J.P., Burrell, S.K., unique region of Bruton’s tyrosine kinase in immunodeficient XIDCowburn, D., and Hanafusa, H. (1995). Affinity and specificity re- mice. Science 261, 358–361.quirements for the first Src homology 3 domain of the Crk proteins.

Rawlings, D.J., Scharenberg, A.M., Park, H., Wahl, M.I., Lin, S., Kato,EMBO J. 14, 2191–2198.R.M., Fluckiger, A.C., Witte, O.N., and Kinet, J.P. (1996). Activation

Kohda, D., Hatanaka, H., Odaka, M.,Mandiyan, V.,Ullrich, A., Schles- of BTK by a phosphorylation mechanism initiated by Src familysinger, J., and Inagaki, F. (1993). Solution structure of the SH3 do- kinases. Science 271, 822–825.main of phospholipase C-gamma. Cell 72, 953–960.

Saouaf, S.J., Mahajan, S., Rowley, R.B., Kut, S., Fargnoli, J., Burk-Konopka, J., and Witte, O.N. (1985). Detection of c–abl tyrosine hardt, A.L., Tsukada, S., Witte, O.N.,and Bolen, J.B. (1994). Temporalkinase activity in vitro permits direct comparison of normal and differences in the activation of three classes of non-transmembranealtered abl gene products. Mol. Cell. Biol. 5, 3116–3123. protein tyrosine kinase following B cell antigen receptor surface

engagement. Proc. Natl. Acad. Sci. USA 91, 9524–9528.Koyama, S., Yu, H., Daigarno, D.C., Shin, T.B., Zydowsky, L.D., and

Regulation of Btk by Autophosphorylation525

Sato, S., Katagiri, T., Takaki, S., Kikuchi, Y., Hitoshi, Y., Yonehara,S., Tsukada, S., Kitamura, D., Watanabe, T., Witte, O., and Takatsu,K. (1994). IL-5 receptor-mediated tyrosine phosphorylation of SH2/SH3-containing proteins and activation of Btk and JAK2 kinases. J.Exp. Med. 180, 2101–2111.

Seidel-Dugan, C., Meyer, B.E., Thomas, S.M., and Brugge, J. (1992).Effects of SH2 and SH3 deletions on the functional activities ofwild-type and transforming variants of c–Src. Mol. Cell. Biol. 12,1835–1845.

Takee, Y., Seiki, M., Fujisawa, J.I., Hoy, P., Yokota, K., Ara, K.I.,Yoshida, M., and Arai, N. (1988). SRa promoter: an efficient andversatile mammalian cDNA expression system composed of thesimian virus 40 early promoter and the R-U5 segment of human Tcell leukemia virus type I long terminal repeat. Mol. Cell. Biol. 8,466–472.

Terasawa, H., Kohda, D., Hatanaka, H., Tsuchiya, S., Ogura, K.,Nagata, K., Ishii, S., Mandiyan, V., Ullrich, A., Schlessinger, J., andInagaki, F. (1994). Structure of the N-terminal SH3 domain of GRB2complexed with a peptide from the guanine nucleotide releasingfactor Sos. Nature Struct. Biol. 1, 891–897.

Thomas, J.D., Sideras, P., Smith, C.I.E., Vorechovsky, I., Chapman,V., and Paul, W.E. (1993). Colocalization of X-linked agammaglobuli-nemia and X-linked immunodeficiency genes. Science 261, 355–358.

Tsukada, S., Saffran, D.C., Rawlings, D.J., Parolini, O., Allen, R.C.,Klisak, I., Sparkes, R.S., Kubagawa, H., Mohandas, T., Quan, S.,Belmont, J.W., Cooper, M.D., Conley, M.E., and Witte, O.N. (1993).Deficient expression of a B cell cytoplasmic tyrosine kinase in humanX-linked agammaglobulinemia. Cell 72, 279–290.

Vetrie, D., Vorechovsky, I., Sideras, P., Holland, J., Davies, A., Flinter,F., Hammarstrom, L., Kinnon, C., Levinsky, R., Bobrow, M., Smith,C.I.E., and Bentley, D.R. (1993). The gene involved in X-linked agam-maglobulinaemia is a member of the src family of protein–tyrosinekinases. Nature 361, 226–233.

Wicker, L.S., and Scher, I. (1986). X-linked immune deficiency (xid)of CBA/N mice. Curr. Topics Microbiol. Immunol. 124, 87–101.

Yoon, C.H., Lee, J., Jongeward, G.D., and Sternberg, P.W. (1995).Similarity of sli-1, a regulator of vulval development in C. elegans,to the mammalian proto-oncogene c–Bl. Science 269, 1102–1105.

Yu, H., Chens, J.K., Feng, S., Dalgarno, D.C., Brauer, A.W., andSchreiber, S. (1994). Structural basis for the binding of proline-richpeptides to SH3 domains. Cell 76, 933–945.

Note Added in Proof

The data reported in the text as D.E.H.A. et al., submitted, and M.A.Broome and T. Hunter, submitted, are now in press:

Afar, D.E.H., Park, H., Howell, B.W., Rawlings, D.J., Cooper, J., andWitte, O.N. (1996). Regulation of Btk by Src family tyrosine kinases.Mol. Cell. Biol. 15, in press.

Broome, M.A., and Hunter, T. (1996). Requirement for c-Src catalyticactivity and the SH3 domain in PDGF BB and EGF mitogenic signal-ing. J. Biol. Chem., in press.