THP JOURNAL OF BIOI.OOIC~I. CHEUISTSY 6 lB93 by The American Society for Biochemistry and hkhcuhr Biology. Inc. Vol. 268. No. 32. Issue of November 15. pp. 24402-2UO7.1993 Printed in U.S.A. +Amylase from the Hyperthermophilic Archaebacterium Pyrococcus furiosus CLONING AND SEQUENCING OF THE GENE AND EXPRESSION IN ES'CHERICHIA COW (Received for publication, August 24, 1992, and in revised form, July 28,1993) Kenneth A. Ladermant, K. Asadap, T. UemoriQ, H. Mukait, Y. TaguchiQ, I. Katot, and Christian B. Anfinsen@l From the $Department of Biology, The Johns Hopkins University, Baltimore, Marylund 21218 and $Takura Shuzo Co., Seta 3-4-1, Otsu, Shiga 520-22, Japan A gene encoding a highly thermostable cr-amylase from the hyperthermophilic archaebacterium Pyro- coccus furiosus was cloned and expressed in Eeche- richia coli. The nucleotide sequence of the gene pre- dicts a 649-amino acid protein with a calculated mo- lecular mass of 76.3 kDa, which corresponds well with the value obtained from purified enzyme using dena- turing polyacrylamide gel electrophoresis. The NH2 terminus of the deduced amino acid sequence corre- sponds precisely to that obtained from the purified enzyme, excluding the NH*-terminal methionine. The amylase expressed in E. coli exhibits temperature-de- pendent activation characteristic of of the original en- zyme from P. furiosus, but has a higher apparent mo- lecular weight which is attributed to the improper formation of the native quaternary structure. No ho- mology was found with previously characterized pro- motor or termination sequences. The deduced amino acid sequence displayed strong homology to the a-am- ylase A of Dictyoglomus thermophilum, an obligately anaerobic, extremely thermophilic bacterium. Evolu- tionary implications of this homology are discussed. Hyperthermophilic archaebacteria provide an extraordi- nary opportunity to study the factors influencing protein thermostability. Unfortunately, due to the recentness of their discovery, the extent of the research completed in this area remains limited. Pyrococcus furiosus is an anaerobic marine heterotroph with an optimal growth temperature of 100 "C, isolated by Fiala and Stetter (1986) from solfataric mud off the coast of Vulcan0 island, Italy. cu-Amylase activity has been reported in the cell homogenate and growth medium of P. furiosus (Brown et aL, 1990, Koch et al., 1990), and the enzyme has been purified to homogeneity (Laderman et al., 1993). The amylase is a homodimer with a molecular mass of 130 kDa which exhibits optimal activity at the optimal growth temperature of the organism. In an attempt to better under- stand the mechanisms of the enzyme's inherent thermosta- bility the gene coding for the cu-amylase from P. furiosus was o The costa of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The nucieotide sequence(s) reported in thispaper has been submitted to the GenBankmlEMBL Data Bank with accession number(s) L.22346. n To whom correspondence should be addressed: Dept. of Biology, The Johns Hopkins University, 34th and Charles Sts., Baltimore, MD 21218. Tel.: 410-516-8552; Fax: 410-516-5213. cloned and expressed in Escherichiu coli, and the nucleotide sequence was determined, In addition the 3'- and Y-noncod- ing regions were analyzed in an attempt to identify sequences involved in transcriptional regulation, and a search for ho- mology between the deduced amino acid sequence and other ar-amylases was completed. MATERIAL8 AND METHODS Bacterial Strains-Cultures of P. furiosus (DSM 3638) were grown as described previously (Ladermanet al., 1993). For cloning and expression of the a-amylase gene, E. coli strain JM109 lrec Al, A&c- pro AB), end Al, gyr A 96, thi-1, ksd R 17, rel Al, sup E 44,F'tra D 36, pro AB+, lac PZAM 15) was used. Plasmids, Enzymes, and Chemicals-The vector pUC18 was used for cloning and DNA sequencing. For expression the vector pTV118N from Takara Shuxo Co. (Kyoto, Japan) was used. Restriction endonucleases, alkaline phosphatase, DNA Blunting Kit, Random Primer DNA Labeling Kit, DNA Ligation Kit, 7- DEAZA Sequencing Kit, Mutan-K site-directed mutagenesis system, and SeaKem Ultrapure Agarose, 5-bromo-4-chloro-3-indolyl-B-D-ga- lactopyranoside were obtained from Takara Shuzo Co. (Kyoto, Ja- pan). Geneclean II Kit was obtained from Bio 101 (La Jolla, CA). PCR' reagents and enzymes were from Perk&Elmer Cetus. Hybond- N nylon hybridization filters and [-y-32P]ATP were obtained from Amersham Corp. Ingredients for E. coli media were from Difco. Isopropyl-B-D-thiogalactopyranoside (IPTG) and ampicillin were ob- tained from Sigma. Ultrapure Urea was obtained from Bio-Rad Assay of Amylnse Activity and Native Polyacrylamide Gel Eiectro- phoresis-The dextrinixing activity of a-amylase was determined at 92 `C using the IJKI method as described previously (Laderman et al., 1993). One unit of the enzyme activity was defined as the amount which hydrolyzed 1 mg of starch/min. Native gel electrophoresis and subbequent staining techniques were performed as described else- where (Laderman et aL, 1993). Preparation of Chromosomul DNA from P. fur&us-The cells were harvested and approximately 0.1 g of cells (wet weight) was suspended in 0.5 ml of 0.05 M Tris-HCl (pH 8.0) containing 25% sucrose. To the suspension was added 0.1 ml of lyeozyme (5 mg/ml). Incubation of the mixture for 1 h at 20 `C was followed by the addition 4 ml of SET solution (150 mM NaCl, 1 mM EDTA, and 20 mM Tris-HCl (pH 8.0)). 0.5 ml of 5% SDS and 100 ~1 of proteinase K (10 mg/ml) were added, and the mixture was incubated for 1 h at 37 `C. The solution was extracted with phenol-chloroform, and the DNA was precipitated with 2 volumes of ethanol. DNA was recovered by winding and it was rinsed in 80% ethanol. The yield of genomic DNA was approximately 1.3 mg from 0.1 g of cells. Preparation of a DNA Probe Using PCR-The NH&rminal se- quence of the intact a-amylase as well as a peptide fragment, deter- mined previously (Laderman et al., 1993), were used for the construc- tion of three degenerate oligonucleotide probes (Fig. 1). The probes were synthesized using an Applied Biosystems model 380B DNA synthesizer. The PCR reactions were performed using 500 ng of P. furiosus `The abbreviations used are: PCR, polymerase chain reaction; IPTG, isopropyl-8-D-thiogalactopyranoside; kb, kilobase( 24402 Cloning of a-Amyhse from P. furiosus 24403 m- of the - 5' GATAAAATPAATTlTATTTT CGCCCC A A TGGTATTCATAATCATCAACC CCCCCCG A A G 3' aim!= 1 Exim- 2 ThrLeuAsnAspMetArgGlnGluTYrlyrPheLys 3' TMCTATACTCAGTTCTT ATMTAAAATTT G G GG C C G G G C T C 5' &imPT 3 FIG. 1. Degenerate oligonucleotide primers based on the NHI-terminal amino acid sequence. Probes were designed for use in the PCR amplification of a portion of the P. furiosus a-amylase gene. The oligonucleotides were prepared as shown, based on the NH&.erminal sequence of the purified enzyme and a peptide frag- ment. chromosomal DNA as the template with 100 pmol of primer 1 and 100 pmol of primer 3 (see Fig. 1). The PCR profile was 94 `C for 0.5 min, 40 "C for 2 min, and 72 `C for 2 min. Amplification was contin- ued for 35 cycles in a total volume of 100 ~1. 1 pl of the reaction mixture was further reamplified with primer 2 (see Fig. 1) and primer 3. The PCR protile was same as above, but the number of cycles was 30. Five microliters of this mixture were analyzed by agarose gel electrophoresis. Cloning and Sequencing of the a-Amyhse Gene of P. furiasus-5- rg portions of P. furiosus chromosomal DNA were digested with PstI, HindIII, XhoI, or EcoRI. The resulting fragments were separated on a 1% agarose gel, transferred to a nylon membrane, and hybridized to the random primer labeled PCR product (Sambrook et al., 1986). Hybridization was carried out for 2 h at 65 `C in 6 X SSC, 0.1% SDS, 5 x Denhardt's solution, 100 pg/ml calf thymus DNA, and 1 X lo7 cpm/ml 3*P-labeled probe. The membrane was washed for 40 min in a solution of 2 x SSC and 0.1% SDS at 65 "C and then for 20 min in a solution of 0.5 X SSC and 0.1% SDS at 65 "C. Genomic DNA digested with PstI was size-fractionated on a 1% agarose gel, and a size-fractionated library, with an insert size of approximately 5-5.5 kb, was constructed in the PstI site of pTV118N and used to transform E. co/i JM109 cells. Recombinant plasmids containing the target sequence were screened by colony hybridization (Sambrook et aL. 1986) uainn the =P-labeled PCR nroduct nroduced as described above. The scr&ing yielded three poiitive clones, and one of these, pKENF, was used for further characterization. Double-stranded recombinant plasmids and single-stranded DNA were isolated following the protocol of Sambrook et al. (1986). Se- quencing was completed using the dideoxy termination method of Sanger et nf. (1977) with the `I-DEAZA sequencing kit. Overlapping subfragments were generated using the method of Yanisch-Perron et al. (1985). Expression of P. furiosus admyhse Gene in E. coli-To prepare a construct which expressed the P. fur&us a-amylase in E. coli an NcoI site was created at the translation initiation codon of the a- amylase gene using the site-directed mutagenesis system Mutan-K (Takara Shuzo, Kyoto), converting the original initiation codon from GTG to ATG. This plasmid (pKENF-2N) was digested with iVc01 and self-ligated to eliminate the 5'noncoding region and to position the gene at a suitable distance from the vector-derived lac promotor and ribosome binding site. The resulting plasmid (pKENF-N) was digested with HindHI and self-ligated to delete a 3'-noncoding region: the product was then designated pKENF-NH. E. coli JM 109 cells carrying this plasmid were designated E. coli JMlOS/pKENF-NH and deposited at the Fermentation Research Institute, Agency of Industrial Science and Technology, Japan, as FERM BP-3182. E. coli JM lOS/pKENF-NH cells were grown for 5 h at 37 `C in 5 ml of L broth containing 100 pg/ml ampicillin. The lac promotor was subseauentlv induced bv the addition of 1 rnt+i IPTG. The cells were allowed to grow for anadditional 12 h, with vigorous shaking, then collected by centrifugation (7000 x g for 10 min), suspended in 200 11 of 50 mM Tris-HCl, pH 7.0. The mixture was sonicated and centrifuged (30 min at 27,000 X g). The supematant was incubated for 10 min at 99 "C and centrifuged again. This supernatant was used as the crude cell extract. The ru-amylase activity was measured using the standard activity assay described above. Computer Analysis-The search for existing sequences displaying homology to the P. furiosus a-amylase gene was completed through GenBank" using FASTA searches. Predictions of secondary structure and physical characteristics based on deduced amino acid sequence were completed using PC/GENE (IntelliGenetics Inc., Mountain View, CA) or the Wisconsin Genetics Computer Group (WGCG) sequence analysis software package Version 6.0 (Genetics Computer Group, University of Wisconsin Biotechnology Center, Madison, WI). RESULTS Cloning and Sequencing of the admyhe Gene: Churacter- istics of Coding and Noncoding Regions-The preparation of a probe using nested PCR resulted in the amplification of a DNA fragment of approximately 1 kilobase. The amplified DNA fragment was blunted using the DNA Blunting Kit (Takara Shuzo, Kyoto) and subcloned into the H&II site of pUC18 (Sambrook et al., 1986).The cloned plasmid was se- quenced by the dideoxy method. When a Southern blot prepared with digested genomic DNA from P. fur&us was probed with the 32P-labeled PCR product, a 5.3-kb PstI fragment, a 3.1-kb HindI fragment, a 5.3-kb XhoI fragment, and two EcoRI fragments of 0.7 and 2.4 kb were found to specifically hybridize to the probe. Three clones carrying an identical 5.3-kb PstI fragment were identified by colony hybridization of an enriched gene bank of P. furiosw genomic DNA using the PCR product known to contain the coding region for the protein's NH,- terminal sequence. One of these clones, shown to contain the a-amylase coding region in the same orientation as the vector- derived lac promotor, was digested with HirulIII and allowed to self-ligate removing a portion of the 3'noncoding region. The nucleotide sequence of the resulting 3.1-kb insert was determined in both orientations. The restriction map and sequencing strategy are shown in Fig. 2. The complete nu- cleotide sequence of the 3.1-kb insert is given in Fig. 3. The a-amylase gene encompasses 1950 nucleotides, with the initiation codon GTG at position 715 (Fig. 3). There is no strong homology, preceding the coding region, with known archaebacterial, eukaryotic, or eubacterial concensus promo- tor sequences described previously. Immediately upstream of the coding region is the sequence GGTGGA, similar to the putative ribosome-binding site of the glyceraldehyde-3-phos- phate dehydrogenase gene of Pyrococcus woesei (GAGGT) (Zwickl et al, 1990). The G + C content of the a-amylase gene is 41.6%, slightly higher than the value reported for the total genome of 38% (Fiala and Stetter, 1986). As has been seen in other sequenced genes from extreme thermophiles, A and T are the preferred bases in the third position of the codons (Zwickl et al., 1990). The five transcripts of the Sulfolobus virus-like particle SSV-1 (Reiter et al., 1988b) and the glyceraldehyde-3-phos- phate dehydrogenase gene of P. woesei (Zwickl et al., 1990) include the sequence TTTTTT in a pyrimidine-rich region directly downstream of the termination codon. A pyrimidine- rich region exists 34 bases 3' of the termination codon in the 24404 Cloning of Lu-Amylase from P. furiosus Base pairs 1000 I I I pc S S ORF 1 I ---a----- -- - c-e -----0 - -- -- -- 4 P. Pst I E EcoRl t-l Hind Ill l-k. Hint II s sac1 FIG. 2. Restriction map and sequencing strategy of the cloned P&I-HindIII insert carrying the a-amylase gene of P. furioeus. Arrows indicate the individual sequence runs. ORF, open reading frame. P, PstI; E, EcoRI; H, HindIII; Hc, HincII; S, SacI. ar-amylase gene, but unlike other archaebacterial sequences examined, there is no pyrimidine-rich region immediately downstream from the TAG stop codon. Expression of P. fur&us a-Amylase Gene in E. coli and Comparison with the Enzyme Purified from P. furiosus-For expression of the P. furiosccs cy-amylase in E. coli an insert containing the gene flanked by archaebacterial noncoding regions was inserted in the expression vector pTV118N. This construct, with the P. furiosus 5' sequence intact, was found to not express any thermophilic cY-amylase activity. To pre- pare a more stream-lined construct for expression in E. coli, a unique NcoI restriction site was created at the initiation codon, converting the initiation codon from GTG to ATG, and the P. furiosus noncoding region was removed. The re- sultant expression plasmid, denoted pKENF-NH, placed the gene in the correct reading frame at an appropriate distance downstream of the vector promotor (Fig. 4). IPTG-induced E. coli JM109 cells transformed with the plasmid pKENF-NH were found to produce 0.2 unit/ml of thermophilic amylase activity in the crude extract. The het- erologously expressed amylase was compared with the enzyme purified from P. fur&us regarding molecular weight and temperature dependence of activity. When the apparent mo- lecular weights of the recombinant protein and the isolated enzyme were compared on an activity stained native gel the protein produced in E. coli displayed a higher apparent mo- lecular weight. The comparison of the temperature depend- ence of the cY-amylase activity between the purified and re- combinant proteins displayed virtually identical relationships of relative activity as a function of temperature (Fig. 5). When the complete amylase gene was analyzed, no protein or nucleotide sequence was found which displayed complete homology with the deduced NHz-terminal sequence of the peptide fragment. This suggests that the peptide sequence, on which the synthesis of primer 3 was based, represents a contaminant and not a portion of the P. fur&us cr-amylase. It was therefore fortuitous that the degenerate primer pre- pared was able to act as a random primer for the PCR reaction. To confirm that the a-amylase expressed in E. coli was the enzyme purified from the bacterium, not a unique enzyme with a similar activity profile and a shared amino terminal Cloning of cu-Amylase from P. furiosus 24405 FIG. 3-Continued sequence, the physical characteristics of the protein were examined. The gene product and the P. furiosus cr-amylase were found to have comparable isoelectric points, specific activities, and pH of optimal activity (data not shown). Primary Structure of P. furiosus Lu-Amylase and Computer Aided Comparison with Enzyme Homologs from Mesophilic and Thermophilic Sources-The deduced amino acid sequence of the P. fur&us a-amylase comprises 649 amino acids, with a calculated molecular mass of 76.3 kDa. This agrees well with the apparent molecular mass of the protein, determined by gel electrophoresis under denaturing conditions, of 66 kDa (Laderman et al., 1993). It is known that the amylase of P. furiosus is secreted into the growth medium under native conditions (Koch et al., 1990). When the primary sequence of the protein was analyzed using the PC/GENE PSIGNAL program, no typical eubac- terial or eukaryotic NH2-terminal signal sequences were found. Using the standard activity assay, it was not possible to detect any thermophilic amylase activity in the extracel- lular media of JMlOS/pKENF-NH cell, confirming the ab- sence of a signal sequence which would make the protein competent for export in E. cofi. When the NH*-terminal sequence of the native enzyme purified from P. furiosus was compared with the predicted NHP-terminal sequence they were found to be identical, suggesting there is no NH,-ter- minal processing. The GenBank" FASTA searches resulted in the acquisition of only a single strongly homologous protein sequence. The query with the protein sequence of the P. furiosus cY-amylase, using the method of Pearson and Lipman (1988), identified a 41.9% identity in a 537-amino acid overlap in the sequence of one of the a-amylases from the extremely thermophilic bac- terium Dictyoglomus thermophilum designated cu-amylase A (Fukusumi et al., 1988). The complete sequence of the D. thermophilum cY-amylase and a number of additional cY-amy- lases from various sources were aligned using the PC/GENE CLUSTAL multiple sequence alignment program, and none displayed the high homology noted above. When a search for homology with previously identified consensus sequences, c + Ncol site (Mutan K, Takara) (Circular maps are not drawn in scale.) FIG. 4. The construction scheme for the recombinant plas- mid pKENF-NH, used for the expression of the P. furiosus a- amylase gene in E. coli. The expression plasmid was constructed by the introduction of a NcoI site at the initiation codon of the amylase gene, followed by restriction digestion with this enzyme and subsequent self ligation. The resulting construct positions the initia- tion codon adjacent to the Shine-Dalgarno sequence of the plasmid lac promotor. The construct was completed by digestion with Hind111 followed by self-ligation, to remove a portion of the 8'-noncoding region. known to be located in the active center and participate in substrate binding in a number of amylases from a variety of sources (Bahl et al. 1991; Tsukagoshi et al., 1985), was per- formed no significant homology was found. The codon usage of cY-amylases from three thermophilic sources, P. furiosus, D. thermophilum (Fukusumi et al., 1988), and Bacillus stearothermophilus (Tsukagoshi et al., 1985), were compared. As noted previously, the higher the optimal temperature of activity the more extensive the bias against the usage of the dinucleotide CG (Zwickl et al., 1990). This trend is apparent not only for the arginine codons, but for the serine, proline, threonine, and alanine codons as well. NO other anomalies in codon usage attributable to thermostability are apparent other than shifts inherent to the changes in amino acid composition. The D. thermophilum cY-amylase A displays physical char- acteristics similar to those observed with the P. furiosus a- amylase. The D. thermophilum enzyme exhibits optimal ac- tivity at 90 "C with approximately 70% residual activity fol- lowing incubation for 1 h at this temperature (Fukusumi et 24406 Cloning of a-Amylase from P. furiosus FIG. 5. Comparison of the tem- perature-dependent amylase activ- ity of the P. furioeus cr-amylase and the amylase activity of the heat- treated crude extract from pKENF- NH-transformed JM109 E. coli cells. Amylase activity was determined at a variety of temperatures, using the standard technique. P. fUtiSus JMlOS/pKENF-NH al., 1988). In contrast, the P. furiosus amylase is optimally active at 100 "C and exhibits a substantially higher thermo- stability: 85% after 3 h at 100 "C (Laderman et al., 1992). The variance in temperature-dependent activity and thermosta- bility between two proteins displaying a high level of homol- ogy provides a unique opportunity to investigate the aspects of the primary sequence which confer enzyme thermostability. Computer analysis of the primary structures and the pre- dicted secondary structures were prepared for the cu-amylases from P. furiosus and D. thermophilum in an attempt to identify possible factors effecting thermostability. When hy- dropathy of the two sequences was plotted, using the PC/ GENE SOAP program, no apparent increase was found in the overall hydrophobicity associated with the increase in the thermostability of the P. furiosus amylase. When regions of high primary sequence homology were compared, no specific trend in hydropathy was noted. When the predicted secondary structures of the two pro- teins (obtained using the PC/GENE GARNIER program) were compared, little similarity in the proposed structures was noted. This dissimilarity was seen both overall and in areas of high primary structure homology. It is not possible to deduce, with confidence, the protein's secondary structure based exclusively on computer analysis. These results indicate that the primary sequence motifs upon which the computer secondary structure predictions are based differ sufficiently between the two proteins. DISCUSSION A number of unusual characteristics of the P. furiosus a- amylase gene set it apart from a majority of the genes char- acterized to date. The gene utilizes the relatively rare initia- tion codon GTG, as does the glyceraldehyde-3-phosphate dehydrogenase gene from P. woesei (Zwickl et al., 1990). It is possible that this represents a tendency in the usage of this initiation codon in hyperthermophilic archaebacteria. Unfor- tunately the number of structural genes isolated from these sources are too limited to allow an accurate assessment of whether the initiation codon GTG is in fact a preferred initiation codon in these organisms, although this may be an intriguing possibility. It is known that the production of a-amylase from P. furiosus is increased by the presence of starch (data not T*mp C shown) indicating that the gene possesses an inducible pro- motor, a feature previously uncharacterized in hyperthermo- philic archaebacteria. When the 5'-noncoding region of the amylase gene was compared with the promotor sequences of other archaebacteria, no homology with the consensus se- quences previously identified in Sulfolobus and Methanococcus (Reiter et al., 1988a) was found. The lack of homology with the putative ribosome-binding site of the P. woesei glyceral- dehyde-3-phosphate dehydrogenase gene suggests either a difference in the 3' terminus of the 16 S rRNA between the two closely related species, or a different promotor mechanism or both. In addition the P. furiosus amylase gene lacks the pyrimidine-rich region found immediatly downstream of the proteins 3' termini in the forementioned archaebacterial genes. This lack of homology with previously investigated archaebacterisl promoters and termination sequences may be a result of the local environment of the gene. The P. woesei glyceraldehyde-3-phosphate dehydrogenase gene is flanked closely by a number of open reading frames, which would benefit from a translational coupling mechanism. In two instances with the SSVl genes in Sulfolobus, shown to share similar termination sequence characteristics with the P. woe- sei gene, this linkage between termination and re-initiation was observed. The P. furiosus gene exhibits no flanking open reading frames within the insert which was sequences, there- fore precluding the necessity for translational coupling. This characteristic as well as the inducibility of the gene are unique among the archaebacterial genes investigated thus far. The a-amylase from P. furiosus is one of a number of thermophilic proteins which have been expressed, in meso- philic hosts, in an active form. The three forms of a-amylase from D. thermophilum (Fukusumi et al., 1988), xylan-degrad- ing enzymes from Caldocellum saccharolyticum (Luthi et al., 1990), and the glyceraldehyde-3-phosphate dehydrogenase from P. woesei (Zwickl et al., 1990), produced under in vivo conditions at 73, 70, and 100 "C, respectively, have all been successfully expressed in E. coli. In these instances, the pro- teins produced by the transformed bacteria remained inactive until they were heated to the temperature appropriate to the enzyme's native conditions. This suggests that the proper folding of the protein into a structure which can be tempera- ture-activated is possible at temperatures other than those at which the protein is produced under native growth conditions. Cloning of a-Amylase from P. furious 24407 The P. furiosus cY-amylase which is a dimer differs from the aforementioned examples in that it additionally requires the formation of an appropriate quaternary structure. Although temperature-dependent amylase activity was observed in E. coli, the apparent native molecular weight of the enzyme was higher than the form purified from P. furiosus, suggesting improper subunit assembly. It is not possible, however, to determine whether this improper assembly is due to transla- tion at lower temperature or to unidentified aspects of pro- duction in E. coli. Perhaps the most interesting facet of this research are the evolutionary implications of the sequence homology between the a-amylases of P. furiosu-s and D. thermophilum. Based on molecular phylogeny using rRNA sequences, existing orga- nisms are seen to fall into three coherent groups eukaryotes, eubacteria, and archaebacteria (Fox et al., 1980). Substantial physiological and structural differences exist between archae- bacteria and eubacteria, which is evidence of their deep evo- lutionary separation (Woese, 1985). The phylogenetic tree prepared by Pace et al. (1986) places archaebacteria closer to the common ancestor of all the kingdoms than one or both of the other primary kingdoms, suggesting that archaebacteria are more primitive than one or both of the other lines. D. thermophilum is a Gram-negative, obligately anaerobic, ex- tremely thermophilic bacterium (Saiki et al., 1985). It shares with P. furiosus a low G + C content and a tolerance for extreme thermal conditions, but is a member of a different phylogenetic kingdom. D. thermophilum has been shown to produce three different species of cu-amylase, which can be classified into two separate classes. First, amylase A, which displays a high degree of homology with the P. furiosus a- amylase, and second, amylase B and amylase C, which display homology with Taka-amylase A (Toda et al., 1982). No sig- nificant homology exists between these two classes, suggesting that they represent two independent gene families or a single family which diverged at a point so distant that no feature save enzyme activity remains as evidence of their relationship. It is possible, since archaebacteria are considered to represent a primitive kingdom, that the P. furiosus cY-amylase, and therefore the D. thermophilum amylase A, may be an example of an archaic form of the enzyme which is well suited to extreme temperatures. In contrast, the D. thermophilum amy- lases B and C contain regions known to be well conserved in several Bacillus species, hog, mouse, and human amylases (Fukusumi et al., 1988), and they represent the common form of the enzyme, various examples of which are active over a wide range of temperatures. REFERENCES BahI, H., Burchhardt, G., Spreinat, A., HaeckeI,,K., Weinecke, A., Schmidt, B., and Antranikian, G. (1991) Ap Brown, S. H., Constantine, H. 1. Enuimn. Mzcmbiol. 57, 1554-1559 fi Micmbiol 66,1965-1991 ., and KeIley, R. M. (1990) Appl. Enuiron. Fiala, G., and Stetter. K. 0. (1966) Arch. Microbial. 146, 56-61 FOX, G. E., Staekebrandt, E., Hespell, R. B., Gibson, J., Maniloff, J., Dyer, T. A., Wolfe, R. S., Balch, W. E., Tanner, R. S., Magnum, L. J., Zableno, L. B., Blakemore. R., Gupta, R., Bonen, L., Lewis, B. J., Stahl, D. A., Laehrsen, K. R., Chen, K. N., and Woese, C. R. (1980) Science 209.457-463 Fukusumi, S., Kamizono, A., Horinouchi, S., and Beppu, T. (1966) Eur. J. Biochem. 174,15-21 Koch, R., Zablowski. P., Spreinat. A., and Antranikian. G. (1990) FEMS Micmbtol. ht. 71, 21-26 Laderman, K. A., Davis, B. R., Krutzsch, H. C., Lewis, M. C., and Antinsen, C. B. (1993) J. Biol. Chem. 268.24394-24401 Luthi, E.. Love, D. R., McAnulty, J.! Wallace, C, Caughey, P. A., Saul, D., and Ber Pace R uist, P. L. (1990) Ap R , Olsen, G. J., an 8 1. Enucmn. Microbtol. 56, 1017-1024 Woese, C. R. (1986) Cell 45, 325-326 Pearson, W. R.. and Lipman, D. J. (1988) Pmt. Natl. Ad. Sci. U. S. A. 85, 2444-2448 Reiter, W., Palm, P.. and Zillig, W. (196&t) Nucleic Acids Res 16, 1-19 Reiter, W., Palm, P., and Zillig, W. (193Sb) Nucleic Acids Res. 16, 2445-2459 Saiki, T., Kobayashi, Y., Kawagoe, K., and Beppu, T. (1965) Inf. J. Syst. Bocteriol. 35, 253-259 Sambrook. J., Fritsch, E. F., and Maniatis, T. (1966) Molecular, CIonin : A e@rotory Manuel, Cold Spring Harbor Laboratory, Cold Sprmg Har 5 or, *. 1 Sanger. F., Nicklen, S., and Coulson, A. R. (1977) Proc. Notl. Ad. Sci. U. S. A. 74,5463-5467 Toda, H., Kondo. K.. and Narita, K. (1992) Proc. Jpn. Ad. 58.206-212 Tsukagoshi, N., Iritani, S., Sasaki, T., Takemura, T., Ihara, H.. Idota, Y.. Yama ate, H., and Udaka, S. (1965) J. Bacterial. 164, 1182-1167 Woese, 8. R (1965) T$e, Bacteria, Academic Press, New York Ya,;$ch-Perron, C., Vieira, J., and Messing, J. (1985) Gene (Amst.) 33, 103- ___ ZwickI, P., Fabry, S., Bogedain, C., Haas, A., and Hensel, R. (1990) J. Bocteriol. 172,4329-4336