Characterizing the virally encoded 3Clike serine protease motif

Currently available astrovirus protease motif sequences (Jiang et al 1993, Lewis et al 1994, Willcocks et al 1994a, GenBank accession AF141381 and Protein accessions AAF18462 and BAA92848) can be aligned with the motifs defined for other viruses (Fig. 3). Prominent features include a putative catalytic triad of His461, Asp489 and Ser551 (numbering according to serotype 1, Oxford strain)

HASTl

447

GTGFFSG

ND

IV.

. TAAHW

. . GNNTFVNV

CY.

.EGLMYE

AK. .

. . V. R

YM

HAST2

447

GTGFFSG

ND

IV.

. TAAHW

. . GNNTFVNV

CY.

.EGLMYE

AK. .

. . V. R

YM

HAST3

447

GTGFFSG

ND

IV.

.TAAHW

. . GNNTFVNV

CY.

.EGLMYE

AK. .

. .V.R

YM

ANV

510

GVGFRLG

NY

IY.

. TAGHW

..GEAKIAKI

TW.

.KGLTSQ

AK. .

. . VLG

HI

TAST

586

GVGFRFM

NY

IL.

. TAEHW

. .QGSDIATL

KN.

.GSVSVK

SK. .

. .VIK

TI

EAV

1089

..VWTRN

NE

VWLTASHW

GRANMATLKI

..G...D

AMLTLTFK.K

HCV

1075

NG

.VCWTVYH..

GAGTRTIASP

.KGPVIQ

M

495 495 495

1136 1110

TEV 2071 GIGF..G.PF II. -TNKHLF RRNNGTLLVQ SL..HG.VFK VKNTTTLQ.Q HL I...DGRDM.

MHV 31 WLD DKVYCPRHVI CSSAD M.T DPDYPNLLCR V...TSSDF.

HAV 1551 GVKD DWLLVPSHAY KFEKDYEMME FYFNRGGTYY SISAGNV VIQSLDVG FQDV.

HRV14 1564 GLGIHD RVCVIPTHAQ PGD.DVLVNG QKIRVKDKYK LVDPENIN.L ELTVLTLDRN E...KFRDI.

1609 1627

HASTl

496

A. FVTCPGDL

HP. TARLKLS

KNP. .

DYS.

CVTVMAYVN.

.EDL...

. . . WS

HAST2

496

A.FITCPGDL

HP.TARLKLS

KNP. .

DYS.

YVTVMAYVN.

.EDL...

. . . WS

HAST3

496

A.FITCPGDL

HP. TARLKLS

KNP. .

DYS .

CVTVMAYVN.

.EDL...

. . . WS

ANV

560

A.RLEIPKPF

QQ.LPVFRLA

KSS. .

END.

YVQMVCFDNQ

LQNV...

. . . VTF

TAST

636

A.VLKLPPEL

NS.VKPIKLA

KKV. .

QSD.

YLTLTAYDPN

FQHA...

. . . ATF

EAV

1137

AEAVTTQSEL

PGNWPQLHFA

Q.PTTGPAS.

WCTATG

..DEE..

. . G LLS

HCV

1111

QDLVGWPA. .

.PQGSR...S

LTPCT

CGSSDLYLV.

.TRHA..

. .D VIP

TAAAMVHGNT 54 6

EVC 1188

RRRGDSRG.S 1158

TEV 2114 I.IIRMPKDF PPFPQKLKF. REPQREERI. CLVTTNFQT.

MHV 61 CVM SGRMSLT VMSYQMQGCQ LVLTVTLQN.

HAV 1600 VLMKVPTIPK FRDITEHFIK KGDVPRALNR LATLVTTVN. HRV14 1618 RGFIS E.DL.EGVD. .ATLWHSN.

.KSMS S MVS D TSCTFPSSDG 2180

. PNTPKYSFG WKPGE.TFT VLAAYNGRPQ 124

■G.TP H LISEGPLKME EKATYVHKKN 1666

.NFTN T ILEVGPVTM. ..AGLINLSS 1668

.YGRV LAVHQTN 569 . YGRV LAVHQTN 569 .YGRV LAVHQTN 569

.DGRM LGVHFGS 639 . DGRL VGIHLGT 715

.GDAV VGVHTGS 1201

TEV 2181 IFWK HWI QTK DGQCGSP.LV STR...DGFI VGIHSAS 2207

MHV 125 GAFHVTLRSS H.■TIKGSFL CGSCGSVGYV LTGDSV..RF VYMHQLE 166 HAV 1667 DGTTVDLTVD QAWRGKGEGL PGMCGGA.LV SSNQSIQNAI LGIHVAG 1712 HRV14 1669 TPTNRMIRYD Y ATK TGQCGGV. LC ATGK I FGIHVGG 1700

HCV 1159 LLSP RPI.

.RTQ DGMSGAP..V CDK. .RTQ DGMSGAP. .V CDK. ■RTQ DGMSGAP..V CDK. .ETY AGTSGSP..1 INR. .DTK FGNSGAP..Y CDH. ■ WTT SGDSGSA.W ..Q. .SYL KGSSGGP . LL CPA.

FIG. 3. Alignment of the astrovirus 3C-like serine protease motif with those of other viruses. The predicted 3C-like serine protease motif of human astrovirus serotypes 1—3 ([HAST1, 2 and 3], GenBank accession numbers L23513, Z25771; L13745; and AR141381) were aligned with the 3C-like protease motifs of avian astroviruses (turkey astrovirus [TAST], AF206663; avian nephritis virus [ANV], protein accession number BAA92848), hepatitis C virus (HCV, P26664), tobacco etch virus (TEV Nia, P04517), murine hepatitis virus (MHV, 453423), hepatitis A virus (HAV, p26580), and human rhinovirus type 14 (HRV14, P03303) using the Pileup program. Motifs with Ser in the catalytic triad are shown in the upper panel and those with Cys (instead of Ser) are shown in the lower panel. Asterisks indicate residues that are conserved in all motifs compared. A filled circle indicates the amino acids (also shown in bold) predicted to form the catalytic triad, an inverted triangle represents predicted substrate binding amino acids (also underlined) and periods denote gaps. The single letter amino acid code is used.

and substrate binding residues (Fig. 3). Given that this alignment strongly suggests that a protease is encoded within ORFla, our studies sought to prove the functional significance of the protease motif. ORFla constructs (Fig. 4A) were expressed in an in vitro, cell-free, transcription-translation system (Kiang & Matsui 2001). Derived products were immunoprecipitated with antibodies specific for various epitopes encoded by ORFla (Fig. 4A), separated by SDS-PAGE, and visualized by autoradiography.

FIG. 4. (A) ORFla-specific antibodies and constructs. The upper panel shows a schematic representation of ORFla of human astrovirus serotype l, with epitopes for the 3C-like serine protease (SP), nuclear localization signal (NLS) and immunoreactive epitope (IRE) indicated. The predicted catalytic triad residues His46l, Asp489 and Ser55l are shown as HDS. The regions to which antibodies anti-laC2 and anti-laCl are directed are shown below the top bar. In the lower panel, the construct l (pAVla/Aprot) product contains a nine amino acid substitution by an unrelated nonapeptide (HA) in the region around the predicted catalytic serine. Constructs 2—6 encode progressively larger N-terminal deletions. The distance (in nucleotide length) from the 5' end at which each of these constructs begins is indicated to the left of the respective bar.

FIG. 4. (A) ORFla-specific antibodies and constructs. The upper panel shows a schematic representation of ORFla of human astrovirus serotype l, with epitopes for the 3C-like serine protease (SP), nuclear localization signal (NLS) and immunoreactive epitope (IRE) indicated. The predicted catalytic triad residues His46l, Asp489 and Ser55l are shown as HDS. The regions to which antibodies anti-laC2 and anti-laCl are directed are shown below the top bar. In the lower panel, the construct l (pAVla/Aprot) product contains a nine amino acid substitution by an unrelated nonapeptide (HA) in the region around the predicted catalytic serine. Constructs 2—6 encode progressively larger N-terminal deletions. The distance (in nucleotide length) from the 5' end at which each of these constructs begins is indicated to the left of the respective bar.

The full-length ORFla construct, designated pAVla, was mutated in several ways (Fig. 4A). In pAVla/Aprot, the region encoding amino acids (aa) 546—554 that contained Ser55l of the predicted catalytic triad and two probable substrate-binding residues (Thr546 and Gln547) was substituted by a cassette encoding an unrelated nonapeptide. Three additional constructs were made in which each residue of the catalytic triad was mutated individually by site-directed mutagenesis (His46lLeu, Asp489Val and Ser55lAla).

Immunoprecipitation of the pAVla product by C-terminal antibodies (anti-laCl and anti-laC2) yielded two major products, plOl (a lOl kDa full-length product) and p38 (a 38 kDa fragment) (data not shown). Products of 48 and 4l kDa (p48 and p4l, respectively) were also immunoprecipitated. Similar evaluation of the pAVla/Aprot product yielded the plOl product primarily and only a faint band at 38 kDa, suggesting that mutation of at least one residue of the catalytic triad and nearby substrate binding residues interferes with the

FIG. 4. (B) Proteolytic processing of the ORFla product. ORFla constructs pAVla (wildtype) and pAVla/Aprot (with mutated protease motif) were modified to include an eight amino acid FLAG epitope at the N-terminus (pAVlaFLG and pAVlaFLG/Aprot, respectively). Products from pAVlaFLG lysates (lanes l, 3 and 5) and pAVlaFLG/Aprot (lanes 2, 4 and 6) were immunoprecipitated with anti-FLAG antibody (lanes l and 2), anti-laC2 (lanes 3 and 4) and anti-laCl (lanes 5 and 6). Upper arrow points to the full-length ORFla product (plOl), middle arrow to N-terminal cleavage fragment p64 and lower arrow to C-terminal cleavage fragment p38. Molecular size markers in kDa are indicated on the left.

processing of plOl. The intensities of p48 and p41 did not change, suggesting that these products were not derived by 3C-like protease-dependent processing of plOl. Altering individual residues of the catalytic triad gave similar results, suggesting the importance of each of the residues in proteolytic processing of plOl.

The marked hydrophobicity of the N-terminus of the ORFla product made it difficult to synthesize adequate quantities of substrates for antibody production. To circumvent this problem, a cassette encoding an 8 aa FLAG epitope (Sigma) was inserted at the 5' end of ORFla and ORFla/Aprot (pFLGAVla and pFLGAVla/Aprot, respectively). Anti-FLAG monoclonal antibody (Sigma) was used to detect N-terminal cleavage products of plOl catalysis. Immuno-precipitation of pFLGAVla with anti-FLAG antibody and anti-laC2 yielded two major products, plOl and a 64 kDa protein (p64) (Fig. 4B). Substantial reduction in the intensity of p64 was observed when pFLGAVla/Aprot was expressed and immunoprecipitated with anti-FLAG antibody. The intensity ofboth p64 and p38 was significantly reduced when the pFLGAVla/Aprot product was immuno-precipitated with anti-laC2 antibody.

Five N-terminal deletion constructs were designed to estimate the N-terminal boundary required to preserve proteolytic activity, as indicated by wild-type levels of p38. N-terminal truncations that did not impinge on the predicted protease motif (deletions of up to 42O N-terminal residues) yielded p38 levels comparable to wild-type (data not shown). However, p38 could no longer be detected with further truncations of 474 and 5l8 N-terminal residues, suggesting that an intact catalytic triad is required for processing to occur.

In summary, these data suggest that the full length ORF-la product, plOl, is processed in vitro to two fragments: p64 at the N-terminus that includes the protease motif and p38 at the C-terminus that includes the nuclear localization signal (NLS) (Jiang et al l993, Willcocks et al l999) and an immunoreactive epitope (Matsui et al l993). The precise cleavage site has not been determined, but is predicted to be between aa 55O and 6OO. The predicted catalytic triad appears to be an essential component of the protease motif and plays a role in the processing of plOl. Baseline levels of processing are maintained with constructs containing N-terminal deletions of up to 42O aa that do not impinge on the predicted protease motif. However, processing is significantly diminished when constructs with larger N-terminal deletions or site-specific mutation of His46l, Asp489 or Ser55l were studied.

0 0

Post a comment