Isolation and characterization of the gene encoding the starch debranching enzyme limit dextrinase from germinating barley

Short sequence-paper

Isolation and characterization of the gene encoding the starchdebranching enzyme limit dextrinase from germinating barley

Michael Kristensen 1;a, Finn Lok b, Veronique Planchot 2;a, Ib Svendsen a,Robert Leah 3;b, Birte Svensson a;*

a Department of Chemistry, Carlsberg Laboratory, Gamle Carlsberg Vej 10, DK-2500 Valby, Denmarkb Carlsberg Research Laboratory, Carlsberg Laboratory, Gamle Carlsberg Vej 10, DK-2500 Valby, Denmark

Received 22 December 1998; received in revised form 17 March 1999; accepted 23 March 1999

Abstract

The gene encoding the starch debranching enzyme limit dextrinase, LD, from barley (Hordeum vulgare), was isolated froma genomic phage library using a barley cDNA clone as probe. The gene encodes a protein of 904 amino acid residues with acalculated molecular mass of 98.6 kDa. This is in agreement with a value of 105 kDa estimated by SDS^PAGE. The codingsequence is interrupted by 26 introns varying in length from 93 bp to 825 bp. The 27 exons vary in length from 53 bp to 197bp. Southern blot analysis shows that the limit dextrinase gene is present as a single copy in the barley genome. Geneexpression is high during germination and the steady state transcription level reaches a maximum at day 5 of germination.The deduced amino acid sequence corresponds to the protein sequence of limit dextrinase purified from germinating malt, asdetermined by automated N-terminal sequencing of tryptic fragments coupled with matrix assisted laser desorption massspectrometry. The sequenced peptide fragments cover 70% of the entire protein sequence, which shows 62% and 77% identityto that of starch debranching enzymes from spinach and rice and 37% identity to Klebsiella pullulanase. Sequence alignmentsupports the multidomain architecture and identifies both secondary structure elements of the catalytic (L/K)8-barrelsubstrate, catalytic residues, and specificity associated motifs characteristic of members of the glycoside hydrolase family 13which cleave K-1,6-glucosidic bonds. A remarkable distribution of the secondary structure elements to individual exons isobserved. ß 1999 Elsevier Science B.V. All rights reserved.

Keywords: Single gene; Genomic sequence; Protein sequence; Glycoside hydrolase family 13; Limit dextrinase; Pullulanase; R-enzyme;(Hordeum vulgare)

The starch debranching enzyme limit dextrinase(LD; EC 3.2.1.41, pullulanase, R-enzyme) catalysesthe hydrolysis of K-1,6-glucosidic linkages in L-limitand K-limit dextrins, pullulan, and amylopectin. Fol-lowing an initial solubilization by K-amylase ofstarch granules in the endosperm of germinating ce-real seeds, the subsequent hydrolysis to dextrins, oli-gosaccharides, and glucose is achieved by the con-certed action of high and low pI K-amylases,L-amylase, K-glucosidase, and LD [1]. Of these en-

0167-4838 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 7 - 4 8 3 8 ( 9 9 ) 0 0 0 7 7 - 1

* Corresponding author. Fax: +45-33-274708;E-mail : [email protected]

1 Present address: Danish Pest Infestation Laboratory, Skov-brynet 14, DK-2800 Lyngby, Denmark.

2 Present address: Laboratoire de Biochimie et Technologiedes Glucides, I.N.R.A., Rue de la Geraudie©re, B.P. 1627, 44316Nantes Cedex 03, France.

3 Present address: Statens Serum Institut, Department of Vac-cine Development, Artillerivej 5, 2300 Copenhagen S, Denmark.

BBAPRO 30436 6-5-99

Biochimica et Biophysica Acta 1431 (1999) 538^546

www.elsevier.com/locate/bba

zymes only the LD has the capacity to hydrolyse K-1,6 linkages in branched maltooligosaccharides. LDis de novo synthesized in germinating barley seeds inresponse to gibberellic acid [2,3]. The rate of appear-ance of LD is lower than that of most other hydro-lases induced during germination [4]. LD, moreover,needs to undergo some kind of mobilization, beingeither synthesized in an inactive form [5] or found asa complex with an endogenous proteinaceous inhib-itor [6]. The highest level of LD activity is reached atday 7 of germination and is retained at this plateauuntil around day 12 of germination. Of the totalamount of barley LD present in the seed, the pro-portion of the bound LD which can be mobilized bytreatment with dithiothreitol is about 85% at day 3and decreases to about 35% at day 7 of germination[5].

A few primary structures of plant starch debranch-ing enzymes have been reported: the amino acid se-quence of LD from rice endosperm [7] and spinachleaf, and of isoamylase from maize endosperm [8] areavailable at EMBL/DDBJ/GB. The gene sequencesof a rice LD and maize sugary1 isoamylase havealso recently been published [9,10]. Interestingly,these genes contains 26 and 17 exons, respectively,which is an unusual high number found in plantgenes. Starch and glycogen debranching enzymesare multidomain proteins containing a catalytic (L/K)8-barrel typical of a large group of amylolytic andrelated enzymes [11^13] generally referred to as gly-coside hydrolase family 13 [14]. In an unrooted evo-lutionary tree calculated for the di¡erent familymembers, the strictly debranching enzymes, i.e.,those hydrolysing only K-1,6-glucosidic bonds, andthose with dual bond-type speci¢city hydrolysingboth K-1,6- and K-1,4-linkages are grouped sepa-rately [11]. Furthermore, certain sequence motifs ofthe (L/K)8-barrel domain, which have been correlatedwith enzyme speci¢city, and two di¡erent putativestructural domains located N-terminally to the barreldomain are characteristic of the endo-acting K-1,6-bond cleaving enzymes [12].

In this paper we report the isolation and sequenc-ing of a barley LD gene composed of 27 exons, witha high degree of similarity to the recently publishedrice LD gene, but also remarkable di¡erences, whichare discussed. The complete deduced barley LD pro-tein sequence is compared with those of other starch

debranching enzymes and enzymes of known crystalstructure from the glycoside hydrolase family 13.

A barley genomic library (cv. Igri, Stratagene) wasscreened at high stringency with a 600 bp barleycDNA fragment, LOK-PS473 from a barley germi-nation speci¢c EST library, encoding a polypeptidewith homology to a Klebsiella pullulanase (Gen-Bank: P07811). One clone, of 1.2U106 plaquesscreened, was identi¢ed and characterized by restric-tion mapping and sequencing [15]. The restrictionmap of the 15 kb genomic insert was consistentwith data obtained by a Southern blot hybridizationanalysis of barley genomic DNA from leaves (cv.Alexis) and using the cDNA as probe (not shown).One hybridizing fragment ranging in size from 20 kbto less than 1 kb was found when the DNA wasdigested with BamHI, HindIII, EcoRI, ApaI, SacI,SalI, HincII, and PvuII, respectively. PstI gave twohybridizing bands, which is due to a PstI-site posi-tioned in an intron interrupting the coding sequenceof the cDNA probe. This pattern of hybridizationsuggests the presence of a single LD gene in thebarley genome. This is in agreement with data forthe rice starch debranching enzyme [7].

The steady-state levels of LD transcripts were fol-lowed in developing and germinating seeds (cv.Alexis) from 30 days post-anthesis (DPA) to 7 dayspost-germination. RNA samples (20 Wg) were trans-ferred to Biodyne A nylon membranes using a Mini-fold II slot-blot manifold (Schleicher and Schuell)and probed with LOK-PS473. The LD transcriptsare detected from day 2^3 of germination. Thesteady-state LD transcript level was highest ataround day 4, 50% of the maximum level was atdays 2 and 6, and the level decreased at days 7^8(data not shown), in excellent agreement with thede novo synthesized LD appearing during maltingof cv. Alexis barley as monitored by activity meas-urements in extracts [5]. LD transcripts were testedfor but not detected in developing seeds at 30 DPA.However, LD protein and enzymatic activity wasdetected by Western blotting and Red-pullulan hy-drolysis already at 7 DPA and observed to increaseduring seed development [16,17], which could suggesta role of this LD in the starch biosynthesis. The peakof barley LD transcription is observed at day 5 ofgermination. Therefore LD seems not to be involvedin the early stages of starch mobilization, but may

BBAPRO 30436 6-5-99

M. Kristensen et al. / Biochimica et Biophysica Acta 1431 (1999) 538^546 539

play an important role in the degradation of dextrinsproduced by the prior action of K-amylases and L-amylases, and K-glucosidases. This is in contrast towhat is observed about K-amylase where expressionis observed immediately after onset of germinationand peaks at days 2 to 3.

The 15 kb genomic fragment was sequenced usingDye Terminator Cycle Sequencing kit and analysedby automated sequencing using an ABI 373 DNASequencer (Perkin^Elmer). The LD gene is approx.10 kb and the open reading frame (ORF) is scatteredthroughout 27 exons interrupted by 26 introns ofvarying length. The exons ranged from 53 bp to197 bp, and the introns were from 93 bp to 825 bpin size. The exon/intron boundaries were consistentwith the consensus GT/AG rule of the donor andacceptor sites of RNA splicing. The gene is namedHvLDX99, and the nucleotide sequence is availablethrough the GenBank database accession numberAF022725.

A genomic LD clone from rice cv. Norin 8 hasrecently been published by Nakamura and co-work-ers [7,9]. The ORF of this rice LD is found on 26exons compared to that of 27 exons in barley. Align-ment of the deduced amino acid sequence of thesetwo cereal LDs reveals the presence of 28 additionalamino acids in the rice LD, encoded within exon 8 ofthe rice LD ORF, which are not found in barley LDprotein (Fig. 1). The nucleotide sequence encodingthis additional peptide sequence in rice LD is circum-scribed by an intron/exon splice site junction with theGT/AG consensus sequence. The isolation of a riceLD cDNA retaining this coding sequence indicatesthat this intron is not spliced out in the japonica ricecv. Norin 8 [7]. However, a rice LD cDNA from anindica variety IR 36 has recently been characterizedby Bower [18], from which the putative intron wasapparently spliced out. The homologous nucleotidesequence of the barley LD ORF, located at the junc-tion between exon 8 and 9, is recognized as an intronand is spliced out during transcript maturation.

The LD protein was puri¢ed from the barley cul-tivar Natasha as earlier described [19]. After reduc-tion and alkylation with 4-vinylpyridine, and hydrol-ysis by trypsin or GluC cleaving enzyme, the peptidefragments obtained were separated by RP-HPLC ona Vydac C18 column using 0.1% tri£uoroacetic acidand a linear gradient of acetonitrile from 2% to 60%

for 2 h for elution at a £ow rate of 1 ml min31 (Fig.2A). The molecular mass of the individual peptideswere determined by mass spectrometry on a MAL-DI-TOF (Lasermat 2000, Finnegan). Amino acid se-quencing was performed on an amino acid sequencerequipped with an on-line PTH-amino acid analyser(Applied Biosystems, model 470A). The experimen-tally determined peptide sequences cover approx.70% of that deduced from the nucleotide sequenceand the coding region and identi¢cation of exon/in-tron boundaries of the barley LD gene were resolvedby aligning the sequences of the puri¢ed tryptic LDfragments (Fig. 2B), and by comparison with thestarch debranching enzyme sequences of spinach(GenBank: X83969) and rice [7]. The peptide sequen-ces matched the sequence predicted from the genesequence with the exception of the single aminoacid substitutions in the tryptic fragments T12(Gly452 is substituted by Gln), and T22 (Thr673 issubstituted by Gly). In fragment T29, Leu279 is sub-stituted by Val, and IleArg282 is substituted byValLys (Fig. 3). Such substitutions may be explainedby slight variation among cultivars and are typicalfor the degree of variation seen for examples forbarley K-amylase.

Barley LD consists of a single polypeptide chain of904 amino acids (Fig. 1) with a calculated Mr of 98.6kDa. The N-terminal sequence of mature LD wasrecently determined to ATQAFMPDARAYWVTS-DLIAGNVGE by automated sequence analysis[19]. The mature LD N-terminus as deduced fromthe nucleotide sequence starts at Ala19. The transla-tion product thus contains only a short hydrophobicleader peptide MAVGETGASVSAAEAEAE18, how-ever, computer-based sequence analysis using theSignalP algorithm (performed on a WEB server athttp://www.cbs.dtu.dk/services/SignalP [20]) indicatesthat this peptide is not recognized as a signal peptidedirecting the LD polypeptide to the lumen of theendoplasmic reticulum. Moreover, varying fractionsof the LD molecules lacked the three residuesAlaThrGln21 from the N-terminus of the mature pro-tein, as determined by N-terminal amino acid se-quence analysis [19].

Certain plant starch debranching enzymes showhigh homology. Barley LD thus exhibits 62% and77% amino acid sequence identity with the spinachand rice debranching enzyme, respectively (Fig. 1).

BBAPRO 30436 6-5-99

M. Kristensen et al. / Biochimica et Biophysica Acta 1431 (1999) 538^546540

The maize debranching enzyme sugary1 is signi¢-cantly less similar to these enzymes, but has closerresemblance to Pseudomonas isoamylase [8]. Se-quence comparison of LD and amylolytic enzymesof glycoside hydrolase family 13 [11,13,14,21] showsthat LD has the catalytic (L/K)8-barrel domain char-acteristic of that family. This fold pattern consists ofalternating L-strands and K-helices where an innercylindrical L-sheet is surrounded by an outer cylinder

made from the K-helices and the active site is acrossthe C-terminal end of the L-barrel [11,13,22^28].

The structural relationship between the membersof the glycoside hydrolase family 13 allows tentativeidenti¢cation of the secondary structure elements inthe catalytic domain of the LD molecule (Fig. 3) asguided by the known three-dimensional structures ofdi¡erent family members [22^28]. The predicted sec-ondary structure elements of LD are encoded by in-

Fig. 1. Comparison of the barley limit dextrinase amino acid sequence, with starch debranching enzymes from higher plant and bacte-ria. The alignment was based on data from the following GenBank entries: OsLD = d50602 for Oryza sativa ; HvLD = AF022725 forHordeum vulgare ; KaPL = P07811 for Klebsiella aerogenes ; SoLD = X83969 for Spinacea oleacera.

BBAPRO 30436 6-5-99


BBAPRO 30436 6-5-99


dividual exons from exon 7 to exon 24, with theexception of the H1 and E2, which are both presentin exon 9, and the E5 and H5 both present in exon16 (Fig. 4). No secondary structure elements of thecatalytic barrel domain were predicted in the super-secondary (L/K)8-structure motif for three of the 18exons spanned by the (L/K)8-domain. Among the as-signed structure selements, H2, E5, and E7 cannot beexcluded, however, to occur in two adjacent exons(Fig. 4). In the known structures of di¡erent familymembers, the C-terminal residue of the helix H2 isseparated by three residues from the N-terminus ofE3 which shows high sequence conservation andtherefore may be of conserved three-dimensionalstructure. It is possible, however, that H2 in the

LD is shorter by four residues and thus terminatesto match the end of the exon 12 (Fig. 4). The E5 hasthe predicted N-terminal Tyr525 to terminate exon 15and the three succeeding residues in exon 16. In thiscase high conservation among the three-dimensionalstructures supports that E5 may indeed be encodedby two exons. Finally, the E7 may well be very shortand comprise only residues IleAsnTyr656 of exon 20.This corresponds to the secondary structure in Taka-amylase A [27] and Bacillus licheniformis K-amylase[24], whereas in isoamylase [23], the enzyme mostclosely related to LD of those with known three-di-mensional structure, the L-structure element starts atthe position equivalent to the residue preceding Ile438

of LD (Fig. 3).

Fig. 3. Alignment of sequence motifs at L-strands of the catalytic (L/K)8-domain which are characteristic for members of glycoside hy-drolase family 13 of di¡erent speci¢city. Barley LD, BAMY2 = barley K-amylase 2 [22], oligo-1,6 = Bacillus thermoglucosidasius oligo-1,6-glucosidase [26], PPA = pig pancreatic K-amylase [25], TAA = Aspergillus oryza Taka K-amylase [27], BLA = B. licheniformis K-amy-lase [24], CGTase = B. circulans cyclodextrin glucosyltransferase [29], PsIA = Pseudomonas amyloderamosa isoamylase [23], Bpul = B.stearothermophilus pullulanase [30], Neopul = B. stearothermophilus neopullulanase [31], Amypul = Clostridium thermohydrosulfuricum K-amylase-pullulanase [32], E. coli BE = glycogen branching enzyme [33]. L-Strands identi¢ed in crystal structures are underlined. Sidechains are marked as involved in: t, transition state stabilization; c, catalysis ; b, substrate binding.

Fig. 2. (A) RP-HPLC of tryptic peptides of barley limit dextrinase. In the chromatogram peptides collected and sequenced are num-bered T1^T46. (B) Alignment of barley limit dextrinase tryptic (C) and Glu-C (- - C) peptide sequences with the amino acid se-quence deduced from the genomic clone.6

BBAPRO 30436 6-5-99


The glycoside hydrolase family 13 contains atpresent more than 20 di¡erent amylolytic and relatedenzymes, including starch and glycogen debranchingand branching enzymes [11]. The (L/K)8-barrel sec-ondary structure elements of this family are recog-nized in spite of very few amino acid residues being

invariant throughout the entire family. The catalysisof glycoside bond hydrolysis involves three carbox-ylic acid residues in LD identi¢ed by the alignmentto be Asp576, Glu613, and Asp745 (Fig. 3). At motifsthat re£ect classi¢cation of speci¢cities, the align-ment reveals that typical debranching enzymes have

Fig. 4. The amino acid sequence of limit dextrinase, with indication of putative secondary structure elements of the catalytic (L/K)8-barrel domain. The exon boundaries are indicated by S. Putative L-strands, indicated by E, and K-helices, indicated by H, are inbold and numbered according to the order of occurrence from the N-terminus of the catalytic domain.

BBAPRO 30436 6-5-99


Gly at the extension of the E4 L-strand (Fig. 3). Forthe barley LD GlyHis580 is identi¢ed, which is equiv-alent to a LysHis motif found in most K-amylases,e.g., LysHis210 in Taka-amylase A [27], except in K-amylases from higher plants which have either Arg-Gly or LysGly [22]. Three-dimensional structures ofenzyme substrate^analogue complexes are resolvedonly for K-1,4-speci¢c enzymes of the glycoside hy-drolase family 13. The LysHis and LysGly dipeptidesin these structures are involved in binding to the ¢rstand second glucosyl residues from the bond to becleaved, i.e., at the glucosyl residue binding subsites+1 and +2 that accommodate the substrate aglyconpart [13]. The binding of the substrate branch pointoccurs at a catalytic site which is geometrically iden-tical to those involved in cleavage of the K-1,4 glu-cosidic bonds by K-amylases. Therefore, residuesbound at subsites +1 and +2 must orient di¡erentlyin the K-1,4 acting and the debranching enzymes.Although this obviously is in support of the sequencemotif di¡erences, only future three-dimensionalstructure determination of an K-1,6 debranching en-zyme in complex with an appropriate substrate ana-logue can provide insight into the structural detailsof the enzyme substrate interactions involved in thisbond type speci¢city. The E2 L-strand of the cata-lytic barrel domain is £anked in barley LD byGly403- - -Pro410 (Taka-amylase A, Gly56- - -Pro64),which recurs in the sequences of all the enzymes ofvarying speci¢city. Sequence similarity is also ob-served for the E3 L-strand that precedes the pre-dicted barley LD His508, corresponding to His122 inTaka-amylase A. This histidine and LD His744 thatcorresponds to His296 in Taka-amylase A and suc-ceeds the E7 L-strand, are conserved and both areessential for the transition state stabilization throughhydrogen-bond formation to OH6 and OH2/OH3,respectively, of the glucosyl residue in the substrateglycon moiety adjacent to the catalytic site, i.e., insubsite-1 [28]. The geometry of the enzyme^substrateinteractions for that particular glucosyl residue willbe very similar or identical in the K-1,4 and K-1,6speci¢c amylolytic enzymes. Also the E7 and E5 L-strands are easily aligned on the basis of the invari-ant £anking residues involved in catalysis, substratebinding, and transition state stabilization. Side-chains from these selected sequence motifs are prob-ably involved in the K-1,6-hydrolase substrate specif-

icity, since LD and bacterial pullulanases show dis-tinct similarities not present in the members of theglycoside hydrolase family 13 having di¡erent specif-icity (Fig. 3).

The increasing number of sequences of plant de-branching enzymes and the corresponding genes willenable the understanding of the transcriptional reg-ulation. Moreover, the sequence and future correla-tion of structure and substrate speci¢city will be use-ful for the determination of three-dimensionalstructures and improve the insight into structureand function of the entire glycoside hydrolase family13.

We are grateful to Suksawad Vongvisuttikun fornucleic acid sequencing, and Bodil Corneliussen andLone SÖrensen for peptide sequence analyses. Wethank Jens Kossmann and Michael Emerson, Insti-tute fu«r Genbiologische Forschung, Berlin, for accessto the spinach sequence prior to database release andYoshiki Matsuura, Institute of Protein Research,Osaka University, for information of the three-di-mensional structure of isoamylase prior to publica-tion. Drs. JÖrgen Larsen and Verena Cameron-Millsare thanked for their continued encouragement dur-ing this work. This project was supported by theNordic Fund for Technology and Industrial Devel-opment in the programme Plant Cell Biotechnology;analysis of starch synthesis for improvement ofstarch composition by biotechnology.

References

[1] A.W. MacGregor, K-Amylase, limit dextrinase, and K-gluco-sidase enzymes in barley and malt, CRC Crit. Rev. Biotech-nol. 5 (1987) 117^128.

[2] D.G. Hardie, Control of carbohydrase formation by gibber-ellic acid in barley endosperms, Phytochemistry 14 (1975)1719^1722.

[3] W.J. Lee, R.E. Pyler, Barley limit dextrinase: varietal, envi-ronmental and malting e¡ects, J. Am. Soc. Brew. 42 (1984)11^17.

[4] G.B. Fincher, Molecular and cellular biology associated withendosperm mobilization in germinating cereals, Annu. Rev.Plant Physiol. 40 (1989) 305^346.

[5] M. Kristensen, B. Svensson, J. Larsen, Puri¢cation andCharacterization of Barley Limit Dextrinase During Malt-ing, Proceedings of the 24th Congress, European BreweryConvention, Oslo, Norway, IRL Press, 1993, pp. 37^43.

[6] L.J. Macri, A.W. MacGregor, S.W. Schroeder, S.L. Bazin,

BBAPRO 30436 6-5-99


Detection of a limit dextrinase inhibitor in barley, J. CerealSci. 18 (1993) 103^106.

[7] Y. Nakamura, T. Umemoto, N. Ogata, Y. Kuboki, M.Yano, T. Sasaki, Starch debranching enzyme (R-enzyme orpullulanase) from developing rice endosperm: puri¢cation,cDNA and chromosomal location, Planta 199 (1996) 209^218.

[8] M.G. James, D.S. Robertson, A.M. Myers, Characterizationof the maize gene sugary1, a determinant of starch compo-sition in kernels, Plant Cell 7 (1995) 417^429.

[9] P.B. Francisco, Y. Zhang, S. Park, N. Ogata, H. Yamanou-chi, Y. Nakamura, Genomic DNA sequence of a rice genecoding for a pullulanase-type of starch debranching enzyme,Biochem. Biophys. Acta 1387 (1998) 469^477.

[10] M.K. Beatty, A.M. Myers, M.G. James, Genomic nucleotidesequence of a full-length wild-type allele of the maize sug-ary1 (Su1) gene, Plant Phys. 115 (1997) 1731.

[11] H.M. Jespersen, E.A. MacGregor, B. Henrissat, M.R.Sierks, B. Svensson, Starch- and glycogen-debranching andbranching enzymes. Prediction of structural features of thecatalytic (L/K)8-barrel domain and evolutionary relationshipto other amylolytic enzymes, J. Protein Chem. 12 (1993)791^805.

[12] H.M. Jespersen, E.A. MacGregor, M.R. Sierks, B. Svensson,Comparison of the domain level organization of starch hy-drolases and related enzymes, Biochem. J. 280 (1991) 51^55.

[13] B. Svensson, Protein engineering in the K-amylase family:catalytic mechanism, substrate speci¢city, and stability, PlantMol. Biol. 25 (1994) 141^157.

[14] G. Davies, B. Henrissat, Structures and mechanisms of gly-cosyl hydrolases, Structure 3 (1995) 853^859.

[15] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning:A Laboratory Manual, 2nd ed., Cold Spring Harbor Labo-ratory Press, Cold Spring Harbor, NY, 1989.

[16] M.J. Sissons, R.C.M. Lance, D.H.B. Sparrow, Studies onlimit dextrinase in barley. 3. Limit dextrinase in developingkernels, J. Cereal Sci. 17 (1993) 19^24.

[17] M. Kristensen, J. Abe, J. Larsen, J. Mundy, B. Svensson,Barley limit dextrinase during malting and seed develop-ment, in: F. Meuser, D.J. Manners, W. Seibel (Eds.), Prog-ress in Plant Polymeric Carbohydrate Research, Behrs,Hamburg, 1995, pp. 19^20.

[18] P. Bower, Cloned Pullulanase from Rice, Patent no. WO95/09922 (1995).

[19] M. Kristensen, V. Planchot, J. Abe, B. Svensson, Large-scalepuri¢cation and characterization of barley limit dextrinase, amember of the K-amylase structural family, Cereal Chem. 75(1998) 473^479.

[20] H. Nielsen, J. Engelbrecht, S. Brunak, G. von Heijne, Iden-ti¢cation of prokaryotic and eukaryotic signal peptides andprediction of their cleavage sites, Protein Eng. 10 (1997) 1^6.

[21] B. Henrissat, A. Bairoch, New families in the classi¢cationof glycosyl hydrolases based on amino acid sequence simi-larities, Biochem. J. 316 (1996) 695^696.

[22] A. Kadziola, J. Abe, B. Svensson, R. Haser, Crystal andmolecular structure of barley K-amylase, J. Mol. Biol. 239(1994) 104^121.

[23] Y. Katsuya, Y. Mezaki, M. Kubota, Y. Matsuura, Three-dimensional structure of Pseudomonas isoamylase at 2.2 Aî

resolution, J. Mol. Biol. 281 (1998) 885^897.[24] M. Machius, G. Wiegand, R. Huber, Crystal structure of

calcium-depleted Bacillus licheniformis K-amylase at 2.2 Aî

resolution, J. Mol. Biol. 246 (1995) 545^559.[25] M. Qian, R. Haser, F. Payan, Structure and molecular mod-

el re¢nement of pig pancreatic K-amylase at 2.1 Aî resolution,J. Mol. Biol. 231 (1993) 785^799.

[26] K. Watanabe, Y. Hata, H. Kizaki, Y. Katsube, Y. Suzuki,The re¢ned crystal structure of Bacillus cereus oligo-1,6-glu-cosidase at 2.0 Aî resolution: structural characterization ofproline-substitution sites for protein thermostabilization,J. Mol. Biol. 269 (1997) 142^153.

[27] Y. Matsuura, M. Kusunoki, W. Harada, M. Kakudo, Struc-ture and possible catalytic residues of Taka-amylase A,J. Biochem. (Tokyo) 95 (1984) 697^702.

[28] M. SÖgaard, A. Kadziola, R. Haser, B. Svensson, Site-di-rected mutagenesis of histidine 93, aspartic acid 180, gluta-mic acid 205, histidine 290, and aspartic acid 291 at theactive site and tryptophan 279 at the raw starch bindingsite in barley K-amylase 1, J. Biol. Chem. 268 (1993)22480^22484.

[29] C. Klein, G.E. Schulz, Structure of cyclodextrin glucosyl-transferase re¢ned at 2.0 Aî resolution, J. Mol. Biol. 217(1991) 737^750.

[30] T. Kuriki, J. Park, T. Imanaka, Characteristics of thermo-stable pullulanase from Bacillus stearothermophilus and thenucleotide sequence of the gene, J. Ferm. Bioeng. 69 (1990)204^210.

[31] T. Kuriki, T. Imanaka, Nucleotide sequence of the neopul-lulanase gene from Bacillus stearothermophilus, J. Gen. Mi-crobiol. 135 (1989) 1521^1528.

[32] H. Melasniemi, M. Paloheimo, L. Heimo, Nucleotide se-quence of the K-amylase-pullulanase gene from Clostridiumthermohydrosulfuricum, J. Gen. Microbiol. 136 (1990) 447^454.

[33] P.A. Baecker, E. Greenberg, J. Preiss, Biosynthesis of bacte-rial glycogen, J. Biol. Chem. 261 (1986) 8738^8743.

BBAPRO 30436 6-5-99


Documents

Isolation and characterization of the gene encoding the starch debranching enzyme limit dextrinase from germinating barley