Molecular Cloning and cDNA Sequencing of Endoxyloglucan ...

THEJOURNAL OF BIOLOXAL CHEMISTRY Q 1993 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 266, No. 34, Issue of December 5, pp. 25364-25368, 1993 Printed in U S A .

Molecular Cloning and cDNA Sequencing of Endoxyloglucan Transferase, a Novel Class of Glycosyltransferase That Mediates Molecular Grafting between Matrix Polysaccharides in PlantCell Walls* (Received for publication, June 30, 1993, and in revised form, July 28, 1993)

Kazuhide Okazawa, Yoshiko Sato, Tomoko Nakagawa, Kiyozo Asada, Ikunoshin Kato, Etsuko TomitaS, and Kazuhiko NishitaniSil From Biotechnology Research Laboratories, Takara Shuzo Co., Ltd. Otsu, Shiga 520-21 and the $Departmentof Biology, College of Liberal Arts, Kagoshima Uniuersity, Kagoshima890, Japan

Endoxyloglucan transferaseis a novel class of glycos- grove, 1989).The cell wall reconstruction processes are animVigna angularis yltransferaserecentlypurifiedfrom portant but poorly understood aspect of the cell wall metabo(Nishitani,IL,and Tominaga,R.(1992)J. Biol. Chem. 207, lism (Fry, 1988; Hayashi, 1989). 21058-21084). This enzymeis the first transferase idenRecently, we have purified endoxyloglucan transferase tified that catalyzes molecular grafting between poly(EXT)' from the extracellular space of epicotyls of Vigna angusaccharide cross-linksin the cell wall matrix and par- laris, a beanplant,andcharacterized its enzymereaction ticipates in reconstruction of the network structure in (Nishitani and 'Ibminaga 1991, 1992). The enzyme catalyzes se- transfer of a large segment of one xyloglucan molecule to anthe cell wall. Basedon the NH2-terminal amino acid quence informationof the purified transferase, we haveother xyloglucan molecule andtherebymediates molecular here cloned and sequenced cDNAs derived from five grafting dif- betweenthe polysaccharide molecules (Nishitani, ferentplantspecies, V; angularis, %ticum aestivum, 1992). Similar enzyme activity was detectedin different plant Arabidopsis thaliana, Lycopersicon esculentum, and Glycine max. In the five plant species, the amino acid se- species (Fry et al., 1992; Farkas et al., 1992; Talbott and Ray, quence of the mature proteins is conserved in the range 1992b). This suggests thatEXT mediates interchange of crossof 71-90% throughout their length. The consensus se- links in the cell wall matrix and functions as a key enzyme quence for N-linkedglycosylation,and four cysteine responsible for regulation of the cell wall reconstruction, an essential process in thecell growth anddevelopment in plants residues are all conserved in the five species. Thus, the endoxyloglucan transferase protein is ubiquitous (Nishitani and'Ibminaga, 1991). It is, therefore, of fundamenamonghigherplants.Thehighlyconserved DNA se- tal importance to isolatecDNA encoding EXT for understandquence will serveas a promising tool for exploring the ing the molecular basis underlying thecell growth and develwe report the cloning and sequencing of molecular process by which cell wall construction, and opment in plants. Here cDNAs coding for EXTs derived from five plant species. The hence cell growth,is regulated. Plant growth is characterized by a marked increase in cell volume. This process, which is called cell enlargement, is basically controlledby mechanical propertiesof the cell wall matrix (Taiz, 1984; Cleland, 1987). Xyloglucans are the major components of the matrix and are characterized as p-(l-4)-glucans with side chains attached at the 0-6 position of the glucosyl residues (Darvill et al., 1980). The polysaccharides are extensively hydrogen-bonded to cellulose microfibrils and serve as cross-links betweencellulose microfibrils and thereby form an interwoven network structure (Hayashi, 1989; Talbott and Ray, 1992a). Thus, growth in size of a plant cell requires the network structure of its cell wall to be continuously reconstructed or rearranged so that newly deposited cell wall polymers may be incorporatedintothepre-existingnetworkstructure (Ray, 1967;Albersheim,1976; Nishitani and Masuda, 1981; COS-

amino acid sequences deduced from the five cDNA sequences indicate that not only the primary structure but also thesecondary structure of the EXTs is highly conserved.

EXPERIMENTAL,PROCEDURES Materials-Seeds of azuki bean (V angularis Ohwi et Ohashi cv. Takara) and tomato (Lycopersicon esculentun Mill cv. Kurihara) were obtained from Watanabe Seed Co., Ltd. (Miyagi, Japan). Restriction enzymes, modification enzymes, M13 universal primerM4, a competent cell of an Escherichia coli strain JM109, andthe cloningvectors pUC118 and pUC119, Bca Best DNA sequencing kit, and random primer labeling kit were products of Takara Shuzo (Kyoto, Japan). Oligonucleotide primers were synthesized on an Applied Biosystems 380B synthesizer. [d2P]dCTPwas purchased from Amersham (Amersham, U.K.). Purification of EXT Protein and the Amino Acid Sequencing-V angularis EXT protein was purified fromextracellular fluid obtained from epicotyls of V angularis seedlings accordingto the procedure described before (Nishitani and Tominaga, 1992). The purified EXT preparation was subjected to NH2-terminalamino acid sequencing by the Edman degradation procedureby use of the Applied Biosystems 475A protein * This work was supported in part by Grants-in-aid for Scientific sequenator. RNA Isolation and Construction of cDNA Libraries-Seeds of V anResearch 05276212 and 05640733 (toK. N.) from the Ministry of Education, Science and Culture.The costsof publication of this article were gularis and L. esculentum were sown on vermiculite bed and grown in defrayed in part by the payment of page charges. This article must darkness for 6 days at 25 "C. Epicotyls of V angularis seedlings or shoots of L. esculentum seedlings were collected and immediately frozen therefore beherebymarked "aduertisement" in accordancewith18 U.S.C. Section 1734 solelyto indicate this fact. in liquid nitrogen.RNA was extracted from plant tissues accordingto The nucleotide sequencefs) reported in this paper has been submitted the procedures described by Chomczynski and Sacchi (1987). Poly(A)+ to the GenBankTM/EMBLData Bank with accession numberfs) 016454. RNA was purified with the aid of oligo(dT)-cellulose chromatography. (A. thaliana), Dl6455 (G. nux), 016456 (L.esculentum), 016457 (I: aestiuum), and 016458 (V angularis). The abbreviations usedare: EXT, endoxyloglucantransferase;PCR, 5 To whom correspondence should be addressed. Fax: 0992-85-8946 polymerase chain reaction; RACE, rapid amplificationof cDNA ends. (Japan).

25364

Molecular Cloning of Endoxyloglucan Dansferase

25365

Double-stranded cDNA was synthesized by reverse transcription of the RESULTS AND DISCUSSION poly(A)+ RNAwith oligo(dT) primers by use of Amersham's cDNA synCloning and Analysis of the cDNA Encoding V: angularis thesis kit according to the procedures of the supplier. A A g t l O cDNA library of V angularis was constructed by use of a packaging kit pur- EXT-We analyzed the NHz-terminal amino acid residues of chased from Stratagene (La Jolla, CA) and propagated in E.coli C6OOM the EXT protein purified from extracellular fluid, or cell wall space, of epicotyls of V: angularis seedlings by the Edman degfor screening. AAEXlox library of L.esculentum was constructedby use of a AEXlox cloning kit from Novagen (Madison, WI). A AgtlO cDNA radation procedure (see "Experimental Procedures") and deterlibrary constructed from shoots of 4.5-week-old seedlings ofArabidopsis mined 30 aminoacid residues of sequence (boxed in Fig. 1).The thaliuna and a A g t l l cDNA library constructed from either 6-day-old amino terminus of the protein was not blocked. seedlings of Glycine mas or 13-day-old seedlings of Diticum uestiuum To clone the cDNA encoding the EXT weemployed PCR with were purchased from Clontech (Palo Alto, CAI. the nested RACE method according to Frohman et al. (1988) Preparation of PCR Primers-Based on the NH,-terminal amino acid sequence of the purified V angularis EXT, two 29-base pair degenerate using two PCR primers, P-VE1 and P-VE2, which had been designed based on the NHZ-terminal amino acid sequence of V: primers (P-VE1and P-VE2) were synthesized. These two primers have sense strand sequences corresponding to a sequence from the 7th to angularis EXT. A 1.1-kilobase pair cDNA fragment was pro16th aminoacid residues (P-VE1)and a sequence from the 17th to 26th duced by the two-step PCR and was subcloned into the HincII residues (P-VE2) of the NH,-terminal amino acid sequence of the puri- site of the pUC119 plasmid vector and sequenced. A deduced fied EXT. Nucleotide sequences of the two primers are as follows: amino acid sequence from the DNA sequence of the PCR fragP-VE1, 5'-ATCGA(C/T)GTGCCGTl'(C/TH3GG(NC)G(AIC/G/T)AA(C/T) ment included the 30-amino acid sequence identified in the TA(CPT)GT-3'; P-VE2, 5'-CCGACGTGGGCGTT(C/T)GA(CPT)CA(C/T) AT(NC/T)AA(NG)TA-3'. PCR Amplification of EXT cDNA for Probe Preparation-The rapid -175 ATTGARACACAAAATACCAGTTCTCARATACAATGAACATTATTAATTATAATTCAGTTAA -115 amplification of cDNA ends (RACE) method developed by Frohman et -114 MGTCATTGATCAGAACAGCAGTGAAGGTTAGCTATAAGCGCGTTATAGGTGCAGGCAGA - 5 5 al. (1988) was employed to produce a cDNA probe fragment. A 37-base pair oligonucleotide primer (5'-Gl"CCCAGTCACGAC(T)zo-3'), 6 -54 GTGTCGTGCCTATATATACCCTTTGGAATGCACAAGTTGAAACACAAAAGAAAAAAGGT which was composed of oligo(dT)20sequence linked tothe 3'-end of the 2 M G M13 universal primer M4, was synthesized and designated as P-TM4. 66 7 TCTTCTTTGTGGACTTGTCTGATTCTGTTATCACTGGCTTCTGCTTCTTTCGCTGCCAAC Two pg of poly(A)+RNA prepared from epicotyls of V angularis seed22 S S L W T C L I L L S L A S A S F A L 3 lings were annealed to50 pmol of P-TM4 in 20 1.1 of a reaction mixture t 126 67 CCAAGAACTCCAATTGATGTACCATTTGGCAGARACTATGTGCCTACTTffiGCCTTTGAT containing 10 m~ Tris-HC1 (pH 8.3), 5 nm MgC12, 50 nm KCl, 0.01% 42 P R T P I D V P F G R N Y V P T W A F O 23 gelatin, 1m~ dNTP, 20 units of RNase inhibitor, and 50 units of reverse 186 127 CATATCARATATCTCAATGGAGGTTCTGAGATTCAGCTTCATCTCGATAAGTACACTGGT transcriptase from Rous avian virus. The reverse transcription was 62 H X K Y L N G G I S E X Q L H L D K Y T G 43 performed at 42 "C for 40 min. To this tube were added 4 pl of 25 nm MgCC12, 8 pl of 10 x PCR buffer (500 m~ KCl, 100 nm Tris-HC1(pH 8.3), TG 187 A C T G G A T T C C A G T C C A R A G G T C A T A C T T G T T T ~ ~ ~ T ~ A C T T C A G C A T G T A C A T ~ T246 T G F Q S K G S Y L F G H F S M Y I K L 82 63 0.1% gelatin), 65.5 pl of sterilized distilled water, 20 pmol each of two primers P-VE1 and M4, and 2.5 units of AmpliTaq DNA polymerase 241 GTTCCTGGTGATTCAGCTGGCACAGTCACTGCTTTCTATTTATCGTCCAC~CGCAGAA 306 (Perkin-Elmer). The firstPCR was carried outat 94 "C for 30 s, at 45 "C V P G D S A G T V T A F Y L S S T N A E 102 83 for 2 min, and at 72 "C for 1min for 30 cycles and maintained at 72 "C 366 307 CATGATGARATAGACTTCGAGTTCTTGGGAAACAGAACTGGGCAACCATACATTTTACAA for 7 min. One pl of the amplification reaction product was used a s 122 H D E I D F E F L G N R T G Q P Y I L Q 103 * template in the second PCR, which was carried out using primers P367 ACARATGTGTTCACCGGAGGCARAGGTGACAGAGAGCAGAGAATCTACCTCTGGTTTGAC426 VE2 and M4 under the samecondition as the firstPCR except that the T N V F T G G K G D R E Q R I Y L W F D 142 123 cycle was repeated 25times. AcDNA fragment amplified in thesecond 427 CCTACGACTCAATACCACAGATATTCAGTGCTATGGAACATGTACCAGATTGTATTCTAT 486 step PCR was subcloned into the HincII site of pUC119 (Vieira and P T T Q Y H R Y S V L W N M Y Q I V F Y 162 143 Messing, 1987), and the resulting plasmid was sequenced to confirm the presence of a cDNA coding for the NH2-terminalamino acid sequence of 487 GTGGATGACTACCCAATAAGGGTGTTCAAGAACAGCAATGACTTGGGAGTGAAGTTCCCC 54 V D D Y P I R V F K N S N D L G V K F P 163 182 the V angularis EXT by use of the Bca Best DNA sequencing kit. This cDNA fragment was labeled with [a-32PldCTP by use of the random 547 TTCAATCAACCAATG~TATACAACAGTTTGTGGAATGCAGATGACTGGGCTACAAGG 606 primer labeling kit and used as a probe for screening of the A cDNA 183 F N Q P M K I Y N S L W N A D D W A T R 202 libraries. 607 GGTGGTTTGGAGAARACAGATTGGTCCARAGCCCCCTTCATAGCCTCTTACAAGGGCTTC 666 Screening of h cDNA Clones fiom Various Plants-Hybridization was 203 G G L E K T D W S K A P F I A S Y K G F 222 performed in a solution containing approximately 60 MBq/ml of labeled 667 CACATTGATGGGTGTGAGGCCTCAGTGAATGCCAAGTTCTGTGACACAC~GCAAGAGG 726 DNA, 6 x SSC, 5 x Denhardt'ssolution, 0.1% SDS, and 0.1 mg/ml 223 H I D G O E A S V N A K F O D T Q G K R 242 salmon sperm DNA at 65 "C. Nylon membranes were washed at the 727 TGGTGGGATCAACCAGAGTTTCGTGACCTTGATGCTGCTCAGTGGCAAAA4CTGGCTTGG final step either at 65 "C (for V angularis and G. mar) or at 37 "C (for 786 243 W W D Q P E F R D L D A A Q W Q K L A W 2 62 A. thaliann, L. esculentum, and T.aestiuum) for 30 min with washing solution composed of 0.5% SDS and either 0.1x SSC (V angularis), 1x 787 GTACGCAACARATACACCATCTACAACTACTGCACTGATCGCARACGCTACTCTCAAGTC 846 263 V R N K Y T X Y N Y O T D R K R Y S Q V 282 SSC (G. mas),or 2 x SSC (A. thnliana, L.esculentum, and 2: aestiuum). Sequence Analysis of EXT cDNAs fiom Various Plants-Phage clones 847 CCTCCAGAGTGCACCAGAGACCGTGACATTTAATCATTTTAACCTATACTTTAAGGCCTC 906 were isolated from A g t l O and hgtll cDNA libraries, and their cDNA 283 P P E O T R D R D I 292 inserts were subcloned into theEcoRI site of the plasmid vector pUC119 907 ATTATTTTCMTATCACACTCCCATARAGTTCCCACCTAAGTGCTGATACAGATTTCATT 966 for DNA sequence analysis. cDNA inserts in the AEXlox vector were automatically subcloned according to Palazzolo et al. (1990). Briefly, the 9 67 TGAGCTTGCTATTGCTTCCTGTATTGTTGATAATCATATCATCATTTATTGCTGCTACTG 1026 AEXlox phage was allowed to adsorbto E. coli strain BM25.8 for 30 min, 1027 TTTGGCTATGTACCAGATATGGCTCTGTGCCTTAATTTGTATCTGTAATTCTGTTTCACT 1086 and theE. coli cells were spread on an agar platecontaining 50 pg/ml ampicillin followed byincubation at 37 "C overnight. A plasmid contain- 1087 C T G A A T C W C T A G T M T A C T A C T A G T T T A T A C T G G T T P 1144 ing theEXT cDNA was extracted from a transformant obtained on the Nucleotide and FIG. l . plate. The cDNA insert obtained was substituted for the EcoRI-Hind111 predicted amino acid sequence of Y; region of the pUC119. DNA sequencing was performed by the dideoxy anguzuris EXT. Sequences are numbered from the postulated initiachain termination procedure (Sanger et al., 1977). The DNA sequence tion codon of the EXT protein and thecorresponding nucleotide residue. was assembled and analyzed using a DNA analysis software DNASIS Negative numbers are given to the residues upstreamof the initiation (Hitachi Software Engineering Co., Ltd. Yokohama, Japan). EXT amino codon. The initiation codon, termination codon, and the putative polyadenylation signal are underlined. The NH2-terminal amino acid seacid sequences deducedfrom the corresponding cDNA sequences of five quence identified by protein sequencing of the mature protein is deplants werealigned to give maximum match to each otherby use ofthe noted by a box. An arrow between Ala20 and Ala21indicates the cleavage software and visual inspection. Harr plot analysis was done by the site for the signal peptide. The potential N-linked glycosylation site is Homology Prot program in the software to comparethe deduced amino indicated with a n asterisk. Four cysteine residues at the COOH-termiacid sequence of V angularis EXT to that of other plants. nal region are circled.

6

25366

Molecular Cloning

of Endoxyloglucan Transferase

EXT protein by the Edman degradation procedure (boxed in Fig. 1). The PCR fragment was then used as a probe to screen a A g t l O cDNA library prepared from epicotyls of V angularis. About 300 signals with various intensity were identified on x-ray films out of 4 x lo4 phages screened (0.8%). Five cDNA clones having signals with the highest intensity were arbitrarily selected, and a clone with the longest cDNA(1.3kilobase pairs) insert was chosen for further analysis. Our sequencing of the V angularis cDNA fragment has disclosed an open reading framecoding for a protein. The deduced amino acid sequence includes the NH2-terminal sequence determined by the protein sequencing of the EXT (boxed in Fig. 1).Two in-frame ATG sequences (position -27 to -25 and 1-3) are present in the upstream of the codon (position 61-63) for the NH2-terminalamino acid. Since the ATG codon at position 1-3 possesses a flanking sequence that fits Kozaks (1986) rules better, we have postulated that the ATG sequence at position 1-3 serves as the initiation codon. A sequenceAATATA found at position 1094-1099 (indicated by an underline in Fig. 1)is a candidate for the polyadenylation signal. We did not find a consensus signal sequence for polyadenylation, AATAAA, between the stop codon (underlined in Fig. l) and poly(A) sequence in this open reading frame (Birnstiel et al., 1985). Based on these results we have concluded that the translational region is from position 1 to 876 in Fig. 1. This open reading frameencodes an additional 20 amino acid residues upstream of the NH2 terminus of the mature EXT protein. The polypeptide could bea signal sequence destinedfor initial transfer to the endoplasmic reticulum because its sequence has typical characteristics for that: a high content of Ser, Leu, and Ala, no acidic or basic residues in thepeptide, and five hydrophobic residues atthecentralpart. When the "(-3,-1)" rule is applied, the cleavage site for the signal sequence is assigned between Ala at position 20 and Ala at position 21 (von Heijne, 1986). It is consistent with the fact that Alaz1is theNH, terminus of the matureEXT protein. Probably the mature EXT is transported via vesicles derived from the Golgi apparatus andsecreted by exocytosis at the plasma membrane intocell wall space. In supportof this hypothesis we have demonstrated thatEXT is present in the extracellular space of V angularis cell walls and acts in the wall space (Nishitani and Tominaga, 1992). The deduced amino acid sequence contains one potential attachment sitefor N-linked oligosaccharide at position 113-115 in the deduced amino acid sequence. The mature EXT protein binds to concanavalin A-Sepharose 4B and is eluted with a buffer containing 0.5 M a-methylmannoside(Nishitaniand Tominaga, 1992).These results indicate that EXT is glycosylated at the asparagine residueat position 113. The deduced amino acid sequence contains 4 cysteine residues at positions 1

1319

V. angularis

1

975

G. max

227,236,273, and286 (circled in Fig.11, all ofwhich arelocated near the carboxyl terminus. Comparison of the Amino AcidSequences among Five EXTs Derived from Various Plants-To clone EXT cDNAs from other plant species, cDNA libraries of G. max, A. thaliana,L. esculentum, and T! aestivum were screened with the same probe DNA as used for isolation of the V angularis cDNA. Conditions for washing of membranes after hybridization were selected according to the phylogenetic distance between V angularis. For isolation of cDNA from G. max, which belongs to the same family as V angularis, membranes were washed with relatively for the V stringent conditions, i.e. at the same temperature as angularis clones (65 "C) but with a solution containinga higher concentration of salts (1 x SSC). On the other hand, less stringent conditionswereused for screening ofcDNAs from A. thaliana, L. esculentum, and T! aestivum, which belong to a different infraclass or subclass from V angularis; membranes were washed at 31"C in a solution containing 2 x SSC. Two or three A phage clones for respective plant species were isolated out of 5 x lo4 to 1 x lo5 phages screened. We selected a single positive clone containing the longest cDNA insert for each plant species and analyzed by DNA sequencing. The four DNA sequences thus obtained were compared with the EXT coding sequence of V angularis cDNA by Harr plot analysis (Fig. 2). The data indicate that thefour cDNAs are highly homologous to the V angularis cDNA throughout the entire coding sequence. The data also show that the homology slightly decreases asthe phylogenetic distance from V angularis increases. We analyzed each DNA sequence of the four plant species for its respective open reading frame. Since only one of the three reading frames ineach cDNA shows striking sequence similarity to that of V angularis EXT, we postulate that these open reading framesencode their respective cDNAs. Fig. 3 shows the five amino acid sequences of putative proteins deduced from respective cDNAs. The sequence identities between these proteins in percentage are shown in Table I. Clearly, the amino acid sequences of the mature proteins are highly conserved among the five plant species. This resultstrongly suggests that the five cDNAs isolated from the five different plant species encode their respective EXT proteins and indicate the high degree of sequence conservation of EXT during evolution. We also observed similar signal peptide regions in the other four EXT proteins. Highest similarity was observed between V angularis and G. max; both belong to the same family. The consensus sequence for the N-linked glycosylation site, Asn-X-Sed Thr (Xrepresents any residues except proline), found at the middle region (amino acid residues 113-115 indicated by a star in Fig. 3) in the V angularis EXT, is conserved in thefour plant species. Furthermore, 4 cysteine residues located at theCOOH1320 1

1

A. thaliana

1137

L. esculentum

1

1088

T. aestivum

FIG.2. Ham plot analysis of EXT cDNAs between V. angularis and four other plants. EXT cDNA sequences of V. angularis, G. max, A. thaliana, L. esculentum, and ?: aestiuum, which are plotted at the horizontal axis, are compared with the EXT coding sequence of V. angularis plotted at the vertical axis. Parameters are set so that dots are pointed when more than 9 residues in a stretch of 10 residues are identical. Nucleotide residues are numbered from the 5' end of the cDNAs and given at theleft and the top of each panel.

Molecular Cloning Vig LYC Tri

. . . . .. . . .

of Endoxyloglucan Dansferase

25367

, , ,~, , ,, .. .. ........ ...............

. .!I, I...! ,!.,

[., .;. ... .. .. . .;. .. . .. . .

..

..

Vig QPYILQTNVFTGGKGDREQRIYLWFDPTTQ H YSVLWNYQIYVDDPIRFKNS Gly QPYILQTNVFTGGKGDREQRIYLWFDPTKE H YSILWN YQI FVDE PIR FKNS Ara QPAILQTNVFTGGKGNREQRIYLWFDPSKA H YSILWN YQI FVDN PIR FKNA QPYILQTNVFTGGKGNREQRIYLWFDPTKGYH YSVLWN YLI IFVDD PIR FKNS Lyc T r lP Y I LT N V F S G G K G D R ER I Y L W F D P T K D Y HY S V L W N L Y M IF V D DP I RF K N S

Vig Gly Ara Lyc Tri

I

0 .

-9RDE6RC K R D RRDLA -9RDE6 KC T K D RRDLI RL

Vig Gly Ara LYC Tri

!.

.,i.

176 175 180 178 177

.,{ . . . . ? [.

,. .. .. .... ., .. ........., .. . ,. LGVKFPFNQPMKIYNSLWNADDWATRGG LGVKFPFDQPMKIYNSLWNADDWATRGG LGVRFPFNQPMKLYSSLWNADDWATRGG LGVKFPFNQPMKIYSSLWDADDWATRGG LGVRYPFDQPMKLYSSLWNADDWATRGG

D D D D D

EKTDWSKAPFIASY EKTDWSKAPFIAAY EKTNWANAPF ASY EKTNWANAPF ASY EKTDWSKAPF ASY

I !. ......, 0

GFHIDGCE GFHIDGCE GFHIDGCQ SFHVDGCE GFHVDGCE

AS NA-K 234 AS NA-K 233 AS EA-K 238 ATPQEVQ 238 AS EA-K 235

., ...,

0

KLWVRKYTIYNYCTDRRYSQVPP-ECTRDRDI292 RLWVRKYTIYNYCTDTRYPHISPECKRDRDI292 EHTIYNYCTDHDRYAAMAP-ECKRDRDV

293

FIG.3. Deduced aminoacid sequences alignment of five EXT proteins.Amino acidresidues of V. angularis (Vig),G. mar (Gly),A. thaliana (Ara),L. esculentum (Lyc),and I: aestivum (ni) are numbered starting from respective postulated NH2-terminal residues (Met)except that of G. m m , which are numbered from the deduced residue at the 5’ end of the longest cDNA obtained. Five methionine residues are upstream of the homologous part to the V. angularis mature protein in theA. thaliana EXT. The methionine residue that could have the longest signal peptide is tentatively postulated as theNH,-terminal residue, from whichthe sequence is numbered. Amino acidresidues are classified in five groups; group I: alanine (A), serine (S), threonine (T),proline (P),and glycine (G); group 11: asparagine (N), aspartic acid (Dl, glutamic acid (E), and g l u t a e n e (Q); group 111: histidine (H), arginine(R), and lysine (K); group IV methionine (M), leucine (L), isoleucine (I), and vahne(V);group V: phenylalamne (F), tyrosine (Y), and tryptophan (W). “Similar“ residues are defined as those that belong to the same group. Similar amino acid sequences in the five EXTs are enclosed by squares, whereas dots indicate the position at which the amino acid residues are identical among the five plant species. A solid horizontal bar indicates a sequence conserved in several glycanases.

five Bacillus subtilis (1-3)-p;(l-4)-P-glucanases (Borriss et al., 1990). The most homologous sequences to the EXT were the A. thaliana meri-5 and I: majus glucanase. The alignment of the three proteins shown in Fig. 4 indicates that the three proteins share several blocks of amino acid residues, particularly in their central regions. Interestingly, the Z! majus glucanase is an endo-(1-4)-P-glucanase specific for xyloglucans (Edwards et al., 1985, 1986) and mediates transglycosylation as a side reaction under specific conditions (Fanutti et al.,1993). On the other hand,EXT is an endo-type glycosyltransferase that cataG. m A. thnliana L. lesculentum l! aestivum lyzes both 1)endo-type splitting of a xyloglucan molecule and 2) V. angularis 79 73 90 79 linking of a newly generated reducing end of the xyloglucan G. mar 74 79 80 molecule, thereby mediating the transferof a large segment of A. thaliann 76 72 the polysaccharide (Nishitani and Tominaga, 1992). Although L. lesculentum 71 EXT exhibits no glycosidase or glycanase activity at all, the transglycosylation reaction itself involves a cleavage of the terminal region (residues 227,236,273, and 286 as indicated by p-glucosyl linkage in xyloglucan backbone. Accordingly, the two open circles in Fig. 3), which could be involved in two disulfide enzymes require xyloglucan structure, i.e. a p-( 1-4)-glucosyl linkages, are also conserved in the five plant species. These backbone with xylosyl side chains, for their respective catalytic data clearly indicate that the general secondary structure, as activities. In otherwords, they sharea common molecular procwell as the primary structure of the EXT protein, is highly ess in their enzymatic processes. It is, therefore, quite likely conserved, at least among theangiosperms. The fact indicates that the central region of EXT proteins, which is highly conthat EXT is ubiquitous among higher plants and suggests that served in the xyloglucan-specific glucanase, is responsible for the EXT-mediated molecular grafting between xyloglucans is a recognition and endo-type splitting of xyloglucan molecules, critical process that is common to higher plants. donor molecules essential for the endo-type transglycosylation. Structural Features for the Active Site of EXT-The The five EXTs have a conserved sequence Asp-Glu-Ile-AspGenBanMEMBL data base search hasrevealed that EXT pro- Phe-Glu-Phe-Leu-Gly (indicated by a solid horizontal bar in teins are homologous to several proteins: a A. thaliana men-5 Fig. 3) at the NH2-terminalside of the asparagine residue that protein of unknown function (Medford et al., 1991), a Dopaeo- may be attached by the N-linked oligosaccharide. A similar lum majus L. (1-4)-P-D-glucanase (de Silva et al., 1993), and sequence to the EXT sequence is conserved in several B. sub-

TABLEI Percent identities of amino acid sequences of mature EXT proteins derived from five different plant species The NH, terminus of the mature V. angularis EXT was identified by protein sequencing. The other four amino acid sequences of their mature EXT proteins were predicted by comparing the V. angularis EXT to its counterpart sequences of the other four EXTs. Figures in the table indicate coincident residues between plant species shown in thecolumn and row. Amino acid sequences used for comparison are as follows: V. angularis, 21-292; G. m a , 20-292;A. thaliana, 25-296; L. esculentum, 23-296; I: aestivum, 22-293.

25368

Molecular Cloning of Endoxyloglucan Dansferase Nas:

Mer: Ara : Nas:

Mer: Ara: Nas:

Mer: Ara: Nas:

Mer: Ara: Nas: Mer: Ara:

FIG.4. Deduced amino acid sequence homology among three proteins. Alignment of the deduced amino acid sequences of T. mujus xyloglucan-specificendo-p-(l-l)-D-glucanase (Nus),A. thulium men-5 protein (Mer),and A. thuliunu EXT ( A m ) .Conserved amino acids among the three proteins are boxed. Shading in the sequence of A. thuliunu EXT indicates conserved amino acidresidues in the five EX” proteins.

tilis endo-~-1,3;~-1,4-g1ucanases (Borriss et al., 1990)as well as entiation in plants but also for exploring novel members of the xyloglucan-specific glucanase derived from I: majus (Fig. endoglycanotransferase and other xyloglucan-related proteins 4).The glucanase sequence, Asp-Glu-Ile-Asp-Ile-Glu-Phe-Leuinvolved in still unknown aspects of matrix polysaccharides. Gly, is the same as that found in EXTs exceptfor the 5th Acknowledgments-We thank 0. Takeda, M. Hiramatsu, and T. Miisoleucine residue, which is replaced in EXTs by phenylalanine. Based on the partialhomology with the sequence surrounding yamoto for help in preparation of the figures. We are grateful to memof Y. Kawase’s group and K. Nozawa’s group for preparation of the active site of phage T4 lysozyme (Borriss et al., 1990), the bers oligonucleotide primers and for help in DNA sequencing, respectively. conserved sequence of the B. subtilis glucanases has been pos- We also thank Dr. S. Okamoto for scientific discussions and for comtulated to serve a role as anactive center of the glucanase. This ments on drafts of this manuscript. line of evidence adds strong support for the suggestion that the REFERENCES conserved sequence in EXT proteins functions as a catalytic center for the endo-type splitting of xyloglucan molecule, a Albersheim, P. (1976) in Plant Biochemistry (Bonner, J., and Varner, J. F., eds) pp. 225-274, Academic Press, New York donor substrate for its transglycosylation reaction. Birnstiel, M. L., Busslinger, M., and Strub, K (1985) Cell 41,349-359 The central region of the sequence ofA. thaliana EXT is also Borriss, R., Buettner, K., and Maentsaelae, P.(1990)Mol. Gen. Genet. 222,278-283 highly homologous to that ofA. thaliana men-5, a protein with Chomczynski, P., and Sacchi, N. (1987)Anal. Biochem. 162, 156-159 unknown function. The homologous sequence suggests that Cleland, R. F. (1987) in Physiology of Cell Expansion during Plant Growth (Cosgrove, D. J., and Knievel, D. P., eds) pp. 18-27, The American Society of Plant men-5 protein has a domain structure that is responsible for Physiologists, Rockville, MD specific recognition andor splitting of xyloglucan molecules. Cosgrove, D. J. (1989) Planta 177, 121-130 A. G., McNeil, M., Albershiem, P., and Delmer, D. P. (1980) in The BioDarvill, Since meri-5 gene is exclusively expressed in shoot apical mechemistry of Plants (Stumpf, P. IC, and Conn, E. E., eds) Vol. 1, pp. 91-161, ristem (Medfordet al., 19911, it is quite likely that thegene may Academic Press, New York encode a receptor or an enzyme directed to xyloglucan molde Silva, J., Jarman, C. D., Arrowsmith, D. A,, Stronach, M. S., Chengappa, S., Sidebottom, C., and Reid, J. S . G. (1993) Plant J. 3, 701-711 ecules in themeristematic tissue. The presence of two homoloEdwards, M., Dea, I. C.M., Bulpin, P.V., and Reid, J. S . G . (1985) Planta 163, gous but distinct proteins in a single plant species also indi133-140 cates the presence of a gene family that contains a common Edwards, M., Dea, I. C. M., Bulpin, P.V., and Reid, J. S . G . (1986) J. B i d . Chem. 281,9489-9494 domain structure, which is responsible for recognitionor bindFanutti, C., Gidley, M. J., and Reid, J. S . G. (1993) Plant J. 3,691-700 ing to a xyloglucan molecule. Farkas, V., Sulova, Z., Stratilova, E., Hanna, R., and Maclachlan, G . (1992)Arch. Conclusion-In this study, we have identified nucleotide seBiochem. Biophys. 298,365-370 quences of cDNAs encoding EXTs derived from five plant spe- Frohman, M. A,, Dush, M. K., and Martin, G. R. (1988) Proc. Natl. had.Sci. u. s. A. as, 8 9 9 a w o z cies. EXT is a novel class of glycosyltransferase that is involved Fry,S . C. (1988) The Growing Plant Cell Wall, pp. 238-249, John Wiley & Sons, in molecular grafting between xyloglucan polymers, thereby New York mediating interchange of cross-links in the cell wall matrix. Fry, S . C., Smith, R. C., Renwick, K. F., Martin, D. J., Hodge, S . K., and Matthews, K. J. (1992) Biochem. J. 282,821-828 The transferase is clearly distinguishable from conventional Hayashi, T. (1989)Annu. Rev. Plant Physiol. Plant Mol. Biol. 40, 139-168 glycosyltransferase classified in EC 2.4.-.-., in that EXT medi- Kozak, M. (1986) Cell 44,283-292 ates endo-type transglycosylation between polysaccharide mol- Medford, J. I., Elmer, J. S., and Klee, H. J. (1991) Plant Cell 3, 359370 Nishitani, K. (1992)Plant Cell Physiol. 39, 1159-1164 ecules. EXT is also distinct from conventional glucanases with Nishitani, K., and Masuda, Y. (1981) Physiol. Plant. 62,482494 transferase activities as their side reactions, because EXT ex- Nishitani, K., and lbminaga, R.(1991) Physiol. Plant. 82,490-497 hibits no hydrolytic activity; EXT essentially requires xyloglu- Nishitani, IC, and lbrninaga, R. (1992) J . Biol. Chem. 267,21058-21064 Palazzolo, M. J., Hamilton, B. A,, Ding, D.,Martin. C. H., Mead, D. A., Mierendorf, can structure for both donor and acceptor activity. For this R. C., Raghavan, K. V.,Meyemwitz, E. M., and Lipshitz, H. D. (1990) Gene (Amst.) 8 8 , 2 5 3 6 novel class of endo-type transferase we have elsewhere proposed the general term “endoglycanotransferase” (Nishitani Ray,P.M. (1967) J. Cell Biol. 353,659474 Sanger, F., Nicklen,S., and Coulsen,A. R. (1977)Proc. Natl. Acad. Sci.U.S. A 274, and ‘Ibminaga, 1992). Here in this present paper we describe 5463-5467 first cloning and sequencing of the endoglycanotransferase. Taiz, L. (1984)Annu. Rev. Plant Physiol. 36,585457 Talbott, D., and Ray, P. M. (1992a) Plant Physiol. 98, 357368 The conserved amino acid sequence of EXTs offers future op- Talbott, L. L. D., and Ray, P. M. (1992b) Plant Physiol. 98,369-379 portunity not only for analyzing the molecular events respon- Vieira, J., and Messing, J. (1987) Methods Enzymol. 153,3-11 sible for wall reconstruction and hence cell growth and differ- von Heijne, G. (1986)Nucleic Acids Res. 14,4683-4690