t

Archives of Biochemistry and BiophysicsVol. 375, No. 1, March 1, pp. 161–164, 2000doi:10.1006/abbi.1999.1648, available online at http://www.idealibrary.com on

Acan125 Binding to the SH3 Domainof Acanthamoeba Myosin-IC1

Henry G. Zot,*,2 Venugopal Bhaskara,* and Lixia Liu†*Department of Biology, Eastern Michigan University, Ypsilanti, Michigan 48197; and†Department of Microbiology, UT Southwestern Medical School, Dallas, Texas 75235

Received August 25, 1999, and in revised form November 24, 1999

The domain organization of Acanthamoeba myo-sin-I, an oligomodular motor protein, includes a poten-tially important protein interaction module that ismostly uncharacterized. The Src homology 3, SH3, do-main of myosin-I binds Acan125, a protein containingat least two consensus ligand binding domains: C-ter-minal SH3 binding motifs (PXXP) and N-terminalleucine-rich repeats. We report the first affinities de-termined for an SH3 domain of any myosin, namely,Kd 5 7 mM for a 21-residue synthetic peptide based onhe PXXP domain sequence and Kd 5 0.15 mM for the

PXXP domain included in the C-terminus of Acan125.These values are consistent with affinities reportedfor peptides and proteins that associate with SH3. Bydeletional analysis we show that only the PXXP do-main is required for Acan125 to interact with the SH3domain of Acanthamoeba myosin-IC (AmyoC(SH3)).The synthetic peptide described above at a concentra-tion near the Kd for SH3 binding blocked the interac-tion between native AmyoC and Acan125, mapping theinteraction to the PXXP domain of Acan125 and theSH3 domain of myosin-I. These results are consistentwith prototypical SH3 binding and suggest that aPXXP module is both necessary and sufficient to inter-act with an SH3 module of myosin-I. © 2000 Academic Press

Key Words: myosin; Src homology; protozoa; amoeba;affinity.

BACKGROUND

The family of myosin-Is is noted for several domainswith interesting properties. Common to all myosin-Is isa head that moves actin filaments. Other domains rep-

1 This work was supported by NSF Grant MCB-9514248 (H.G.Z.).2

To whom correspondence should be addressed. Fax: 734-487-

9235. E-mail: hzot@online.emich.edu.

0003-9861/00 $35.00Copyright © 2000 by Academic PressAll rights of reproduction in any form reserved.

resented in the family of myosin-Is include a basicregion that associates with phospholipids, an IQ motifthat reversibly associates with calmodulin or calmod-ulin-like light chains, a proline-rich region that asso-ciates with actin filaments, and an SH33 domain re-lated to a homologous ligand binding domain of Src (1).SH3 domains have been proposed to form a linkagewith signal transduction pathways by associating withproteins that contain one or more consensus SH3 bind-ing sites. These PXXP motifs consist of a linear se-quence of seven residues that adopts a proline helixand that contains N-substituted residues at conservedpositions (2).

SH3 domains are present in many, but not all, of thegenes that code for myosin-I (1), yet little is knownabout interactions that take place with the SH3 do-main of myosin-I. Three ligands have been identifiedfor the SH3 domains of myosin-I, namely, the oncogenec-Cbl (3), the actin associated protein verprolin (4), andthe Acanthamoeba protein Acan125 (5). Both c-Cbl andAcan125 contain consensus PXXP motifs that are re-quired for interaction with SH3. The PXXP motifs ofAcan125 map to an SH3 binding domain of only 17residues near the C-terminus (6). This well-definedbinding site as well as results showing a functionalPXXP domain in the C-terminus (6) makes Acan125attractive for further characterization. Here we useAcan125 as a model ligand to probe the binding char-acteristics of AmyoC(SH3).

3 Abbreviations used: SH3, Src homology 3; LRR, leucine-rich re-peat; PXXP, consensus SH3 binding motif; AmyoC, Acanthamoebamyosin-IC; Acan125(AD3), residues 740–1121 of Acan125;Acan125(FL), residues 1–1121 of Acan125; Trx, thioredoxin; GST,glutathione S-transferase; DTT, dithiothreitol; PCR, polymerase

chain reaction, GT, glutathione.

161

viously (6).

wG

162 ZOT, BHASKARA, AND LIU

RESULTS AND DISCUSSION

Our previous results indicated that the C-terminusof Acan125, Acan125(AD3), interacts with the SH3domain of myosin-I (6). Here we test whether thisinteraction is saturable by measuring the amount ofsoluble Trx-AmyoC(SH3) that sediments with immobi-lized GST-Acan125(AD3) (Figs. 1 and 2). The amountof sedimented Trx-AmyoC(SH3), as determined by thedensity of a band on the gel, correlated positively withthe concentration of Trx-AmyoC(SH3) (Fig. 3). At thehigher concentrations, significant binding of Trx-SH3to control beads was also observed (Fig. 3). Because theamount of nonspecific binding might depend on themass of beads, we carefully maintained the sameweight of packed beads for all samples. After correctionfor nonspecific binding, a plot of Trx-AmyoC(SH3)binding to Acan125(AD3) revealed a typical saturationcurve with a Kd value of 151 nM (Fig. 3), which approx-

FIG. 1. (A) Schematic representation of constructs of Acan125 usedinclude leucine-rich repeats, LRR (diagonal lines), and PXXP motifs (sequence of Acan125; underlined residues match a consensus PXXGST-Acan125 (construct 1), GST-Acan125(D977–994) (construct 2),struct 4), and GST-Acan125(AD3) (construct 5), were prepared from1.1-kb cDNA corresponding to residues 740–1121 (Acan125(AD3)) thwith embedded restriction sites. The products, subcloned into complesis, were verified by automated sequencing. A BamHI site engcomplementary site in the native sequence near the 39 end of the 2cDNA was prepared from the 1.1-kb fragment by excising a 54-bp se994) cDNA was joined to the BamHI cut 2.4-kb fragment to geAcan125(FLD977–994), and Acan125(LRR) were excised from pBluepGEX-5X-3. cDNAs for Acan125(AD3) and Acan125(AD3D977–994)complementary sites of pGEX-KG (12). Plasmid constructs, verified bDomain organization showing the location of the SH3 domain in thetail containing a basic domain (B) and a proline-rich domain (C) thprotein GST-AmyoC(SH3) (construct 1) as described previously (5).with primers containing EcoRI and XhoI restriction sites. This proEcoRI/XhoI cut pET28b/Trx (6), yielding pET28b/Trx-AmyoC(SH3expression cassette codes for a fusion protein of thioredoxin and Aexpressed and purified by affinity chromatography as described pre

in this study. Sequences that match consensus ligand binding domainssolid black bar). Acan125(p977–996) is a synthetic peptide based on theP sequence. cDNAs for the glutathione-S-transferase fusion proteins,GST-Acan125(LRR) (construct 3), GST-Acan125(AD3D977–994) (con-a 2.4-kb cDNA corresponding to residues 1–777 (Acan125(LRR)) and aat were amplified from Acanthamoeba RNA by RT-PCR using primers

ementary sites of pBluescript for subsequent subcloning and mutagen-ineered at the 59 end of the 1.1-kb fragment enabled joining to a.4-kb fragment, yielding Acan125(FL) cDNA. Acan125(AD3D977–994)quence corresponding to residues 977–994 (6). The Acan125(AD3D977–nerate a cDNA for Acan125(FLD977–994). cDNAs for Acan125(FL),script with EcoRI and NotI and subcloned into complementary sites ofwere excised from pBluescript with XbaI and XhoI and subcloned intoy restriction analysis, transformed E. coli strains DH5a and BL21. (B)sequence of AmyoC. The domains of AmyoC include a motor (A) and aat surrounds the SH3 domain (solid black bar). We prepared a fusionTrx-AmyoC(SH3) (construct 2) was prepared by PCR amplification (5)duct, which was verified by automated sequencing, was ligated with), which transformed BL21 DE3 pLysS for protein expression. ThemyoC(SH3) with a hexahistidine tag at the N-terminus. Protein was

imates the Kd determined for the interaction betweenAT

FIG. 2. Gel showing (1–5 mg) GST fusion proteins. Gel samplesere taken from beads used for binding studies: GST-Acan125 (A);ST-Acan125(D977–994) (B); GST-Acan125(LRR) (C); GST-can125(AD3D977–994) (D); and GST-Acan125(AD3) (E). GST and

rx fusion proteins were prepared as described previously (6).

Ac2FdpF

s

a

Apa

163BINDING PROPERTIES OF SH3

the PXXP motifs of HIV-1 Nef and the SH3 domains ofHck and Fyn (7).

Having established a binding constant for the C-terminus of Acan125, we searched for other potentialbinding sites in the Acan125 structure (Fig. 1). Variousconcentrations of Trx-AmyoC(SH3) near the Kd werecombined with beads containing the following fusionproteins: GST-Acan125, GST-Acan125(D977–994),

FIG. 3. Trx-AmyoC(SH3) domain of myosin-I binds saturably toimmobilized GST-Acan125(AD3). Shown are silver-stained bands ofprotein corresponding to Trx-AmyoC(SH3) that sedimented withGST-Acan125(AD3) (upper) and GST alone (lower). To measure sol-uble Trx-AmyoC(SH3) sedimenting with GST fusion proteins boundto GT-beads, packed beads (25 ml) were resuspended in KMEM7 (100mM 3-(N-morpholino)propanesulfonic acid (Mops), pH 7.0, 100 mMKCl, 2 mM MgCl2, 2 mM ethylene glycol bis(2-aminoethyl ether)N,N,N9,N9-tetraacetic acid) supplemented with 1% BSA. Equalamounts of Trx-AmyoC(SH3) were added to sample pairs consistingof beads coupled with GST alone (control) and beads coupled withGST fusion proteins; the concentrations of Trx-AmyoC(SH3) addedin nM were 100, 250, 500, 750, 1000, 2000, and 5000. All samplesadjusted to a total volume of 100 ml were incubated 10 min at 4°Cnd the suspensions transferred to 0.65-mm Durapore centrifugation

filters. Beads were collected on the filter by centrifugation at 14,000gfor 2 min and were washed twice by suspension in 1 ml of KMEM7followed immediately (,60 s) by centrifugation to remove unboundprotein. Bead samples were transferred with 0.5 ml of KMEM7 toclean Eppendorf tubes and centrifuged; supernatants were aspi-rated. Trx-AmyoC(SH3) bound to the beads was detected on gelsstained with silver following manufacturer’s instructions (Novex).Proteins on the gel were stained with silver, developed sufficiently toview bands corresponding to Trx-AmyoC(SH3). Absorbance of Trx-AmyoC(SH3) bands was measured by densitometry and converted tomass using the standards on the same gel. For each pair of samples,the mass of Trx-AmyoC(SH3) sedimenting with GST was subtractedfrom the mass of Trx-AmyoC(SH3) sedimenting with GST-Acan125(AD3). [Trx-AmyoC(SH3)]free was calculated from the differ-ence in masses of Trx-AmyoC(SH3) added and bound. [Trx-AmyoC(SH3)]bound versus [Trx-AmyoC(SH3)]free was fit with the rela-tionship [Trx-AmyoC(SH3)]bound/Bmax 5 1/(1 1 Kd/[Trx-

myoC(SH3)]free). A best fit of the data converged with thearameters and associated standard errors, Bmax 5 0.32 6 0.02 mgnd Kd 5 151 6 56 nM.

GST-Acan125(LRR), GST-Acan125(AD3), and GST-

Acan125(AD3D977–994) (Fig. 2). Only the control,GST-Acan125(AD3), and the full-length Acan125 con-struct, GST-Acan125, bound Trx-AmyoC(SH3) signifi-cantly (Fig. 4, panel 4). We did not detect significantTrx-AmyoC(SH3) bound to the C-terminal truncationGST-Acan125(LRR) or to the PXXP mutant GST-Acan125(D977–994) (Fig. 4, panels 2 and 3). These

FIG. 5. AmyoC(SH3) binds the synthetic peptide Acan125(p977–996). The tryptophan emission spectrum with 290-nm excitation ofGST-AmyoC(SH3) is enhanced by the addition of 100 mM

can125(p977–996) (1) (inset). With constant 290-nm excitation, thehange in 325-nm intensity of 1 mM GST-AmyoC(SH3) in KMEM7 at5°C was titrated by sequential additions of Acan125(p977–996).luorescence intensity is plotted as a function of added peptide. Theata were fit with the following relationship that corrects for boundeptide assuming all of the GST-AmyoC(SH3) is competent to bind:/Fmax 5 1/(1 1 Kd/L9), where L9 5 [peptide added] 2 [peptide bound]

and [peptide bound] 5 F/Fmax[GST-AmyoC(SH3)]. A nonlinear least-quares fit of the data converged with K

FIG. 4. Trx-AmyoC(SH3) binding to constructs of Acan125. GSTfusion proteins immobilized on beads were combined with solubleTrx-AmyoC(SH3) at concentrations that bracket the Kd: 50, 150, and750 nM. Silver-stained gels show the amount of Trx-AmyoC(SH3)(arrowhead) that sediments with GST-Acan125(AD3D977–994) (1),GST-Acan125(LRR) (2), GST-Acan125(D977–994) (3), and GST-Acan125 (4). On the same gel, GST-Acan125(AD3) combined with750 nM Trx-AmyoC(SH3) serves as the positive control (C). Bindingconditions replicate those of Fig. 3.

d 5 6.8 6 0.86 mM standarderror.

1

1

164 ZOT, BHASKARA, AND LIU

results demonstrate that consensus PXXP motifs ofAcan125 are required for high-affinity interaction withthe SH3 domain of myosin-I and that alternative sitesfor binding SH3 are too weak to measure.

Previously characterized SH3 domains have beenshown to bind peptide sequences based on native PXXPmotifs (8). To test AmyoC(SH3) for peptide binding, wesynthesized a peptide, Acan125(p977–996), based onthe PXXP sequences of Acan125. Binding ofAcan125(p977–996) was detected as a change in fluo-rescence emission spectrum brought about by the per-turbation of one or more tryptophans of GST-AmyoC(SH3) (Fig. 5, insert). Though both SH3 andGST contain two tryptophans, no change in emissionspectrum was detected when a saturating concentra-tion of Acan125(p977–996) was added to GST alone(data not shown). The change in the fluorescence emis-sion spectrum of GST-AmyoC(SH3) as a function of thepeptide concentration fit a binding curve with a Kd ;7mM (Fig. 5). This is close to binding constants reportedfor peptides binding to the SH3 domains of many pro-teins, including Src, Fyn, Lyn, PI3K, and Abl (9, 10).These results are consistent with residues 977–994 ofAcan125 serving as the core binding site for interactionwith the SH3 domain of myosin-I.

Though a direct interaction is shown between PXXPand SH3 in isolation, native myosin-I and Acan125may have an indirect interaction in a cellular setting.To investigate this, we tested whether the SH3 bindingpeptide, Acan125(p977–996), would inhibit an interac-tion between Acan125 and myosin-IC in a cellular ex-tract. We previously showed that the two proteins co-precipitate from a soluble lysate of Acanthamoeba us-ing M1.7, a monoclonal antibody to myosin-I (5). Theanalysis of the immunoprecipitate by Western blotshows that the amount of Acan125 declined in directrelation to the concentration of Acan125(p977–996)(Fig. 6). Half-maximal inhibition of Acan125 sedimen-tation was approximately 5 mM Acan125(p977–996),which compares favorably with the estimated Kd ofAmyoC(SH3). Similar results were reported for Src,namely, a peptide that associates with the SH3 domainof Src inhibited the sedimentation of Src binding pro-teins from a cell lysate in a dose-dependent manner(11). Thus, the inhibition by Acan125(p977–996) ar-gues for a direct interaction in the cell lysate betweenAcan125 and the SH3 domain of myosin-I.

CONCLUSIONS

Results of this study are threefold. Myosin-I is shownto interact with Acan125 exclusively through its SH3

domain. Interaction with the SH3 domain of myosin-I 1

requires only that the ligand have a cognate PXXPmotif. AmyoC(SH3) binds a protein ligand (Kd 5 0.15mM) about 10-fold better than a peptide ligand (Kd 5 7mM). Thus, the SH3 domain of Acanthamoeba myo-sin-I, a novel SH3 domain from a protozoan, sharesbinding characteristics with SH3 domains of Src andother metazoan protein families.

ACKNOWLEDGMENTS

We thank Dr. Pin Xu for preparation of a full-length cDNA ofAcan125 and Drs. Anita Zot and Wei-Lih Lee for critical reading ofthe manuscript.

REFERENCES

1. Coluccio, L. M. (1997) Am. J. Physiol. 273, C347–C359.2. Mermall, V., Post, P. L., and Mooseker, M. S. (1998) Science 279,

527–533.3. Anderson, B. L., Boldogh, I., Evangelista, M., Boone, C., Greene,

L. A., and Pon, L. A. (1998) J. Cell Biol. 141, 1357–1370.4. Roberson, H., Langdon, W. Y., Thien, C. B. F., and Bowtell,

D. D. L. (1997) Biochem. Biophys. Res. Commun. 240, 46–50.5. Xu, P., Zot, A., and Zot, H. G. (1995) J. Biol. Chem. 270, 25316–

25319.6. Xu, P., Mitchelhill, K. I., Kobe, B., Kemp, B. E., and Zot, H. G.

(1997) Proc. Natl. Acad. Sci. USA 94, 3685–3690.7. Lee, C. H., Leung, B., Lemmon, M. A., Zheng, J., Cowburn, D.,

Kuriyan, J., and Saksela, K. (1995) EMBO J. 14, 5006–5015.8. Ren, R., Mayer, B. J., Cicchetti, P., and Baltimore, D. (1993)

Science 259, 1157–1161.9. Viguera, A. R., Arrondo, J. L. R., Musacchio, A., Saraste, M., and

Serrano, L. (1994) Biochemistry 33, 10925–10933.0. Rickles, R. J., Botfield, M. C., Weng, Z., Taylor, J. A., Green,

O. M., Brugge, J. S., and Zoller, M. J. (1994) EMBO J. 13,5598–5604.

1. Weng, Z., Thomas, S. M., Rickles, R. J., Taylor, J. A., Brauer,A. W., Seidel-Dugan, C., Michael, W. M., Dreyfuss, G., andBrugge, J. S. (1994) Mol. Cell. Biol. 14, 4509–4521.

FIG. 6. The peptide Acan125(p977–996) inhibits the coprecipita-tion of Acan125 and myosin-I. Myosin-I was immunoprecipitatedfrom Acanthamoeba high-speed supernatant (20 mg/ml in KMEM7held at 4°C) with antibody M1.7. Separate blots showing Acan125and myosin-I reactivities were performed from equal aliquots. Theamount of Acan125 that coprecipitated with myosin-I in the absenceof peptide is shown on the blot (0 mM). This is compared with therelative amount of Acan125 obtained from samples that containedpeptide as shown. Preparation of the high-speed supernatant andthe procedure for immunoprecipitation are as described previously(5).

2. Guan, K., and Dixon, J. E. (1991) Anal. Biochem. 192, 262–267.