Vol. 173, No. 14JOURNAL OF BACTERIOLOGY, JUlY 1991, p. 4464 44730021-9193/91/144464-10$02.00/0Copyright © 1991, American Society for Microbiology

Galactose Utilization in Lactobacillus helveticus: Isolation andCharacterization of the Galactokinase (galK) and Galactose-l-

Phosphate Uridyl Transferase (gaiT) GenesBEAT MOLLET* AND NATHALIE PILLOUD

Nestle Research Center, Nestle Ltd., Vers-chez-les-Blancs, P.O. Box 44, 1000 Lausanne 26, Switzerland

Received 29 January 1991/Accepted 13 May 1991

By complementing appropriate gal lesions in Escherichia coli K802, we were able to isolate the galactokinase(gaLK) and galactose-1-phosphate uridyl transferase (gall) genes of Lactobacillus helveticus. Tn1O transposonmutagenesis, together with in vivo complementation analysis and in vitro enzyme activity measurements,allowed us to map these two genes. The DNA sequences of the genes and the flanking regions were determined.These revealed that the two genes are organized in the order galK-galT in an operonlike structure. In an in vitrotranscription-translation assay, the galK and galf gene products were identified as 44- and 53-kDa proteins,respectively, data which corresponded well with the DNA sequencing data. The deduced amino acid sequenceof the galK gene product showed significant homologies to other prokaryotic and eukaryotic galactokinasesequences, whereas galactose-1-phosphate uridyl transferase did not show any sequence similarities to otherknown proteins. This observation, together with a comparison of known gal operon structures, suggested thatthe L. helveticus operon developed independently to a translational expression unit having a different geneorder than that in E. coli, Streptococcus lividans, or Saccharomyces cerevisiae. DNA sequencing of the flankingregions revealed an open reading frame downstream of the galKT operon. It was tentatively identified as galM(mutarotase) on the basis of the significant amino acid sequence homology with the corresponding Streptococcusthermophilus gene.

Lactobacillus helveticus and Lactobacillus delbrueckiisubsp. bulgaricus are used together with Streptococcusthermophilus in the dairy industry as starter cultures forcheese, yogurt, and other dairy product fermentations. As aprimary carbon and energy source, they utilize lactose, theonly sugar present in milk, and ferment it to lactate. Theuptake of lactose by these organisms occurs via a permease-type system and is followed by a ,3-galactosidase-catalyzedcleavage of the sugar into its monosaccharide glucose andgalactose moieties (18, 40).

In L. delbrueckii subsp. bulgaricus and S. thermophilus,the glucose moiety is metabolized, while the galactosemoiety is excreted stoichiometrically into the medium (18).In fact, it has been observed that galactose efflux from thesecells is coupled to uptake of lactose or methyl-p-D-thioga-lactopyranoside from the external medium and that thetransport system involved acts as a lactose-galactose an-tiporter (21, 39, 39a). However, L. helveticus, which isrelated to L. delbrueckii subsp. bulgaricus and difficult todistinguish taxonomically from it (22, 38), is able to fermentthe glucose and galactose moieties of lactose and does notaccumulate free galactose in the external medium (47). It canutilize galactose as an energy source. Hickey et al. (18)observed that there is no lactose or galactose phosphotrans-ferase system in L. helveticus and suggested that the galac-tose moiety of the lactose is metabolized via the Leloirrather than the tagatose 6-phosphate pathway. To substan-tiate this suggestion, we present here the isolation andcharacterization of the genes involved in galactose metabo-lism in L. helveticus.The genetics of L. helveticus and L. delbrueckii subsp.

bulgaricus are still very poorly developed (9). Efficient

* Corresponding author.

transformation or gene transfer systems do not exist yet (5,17, 25, 46). Furthermore, host strains with defined mutantgenotypes within the sugar-fermenting pathway have onlybeen reported so far for ,-galactosidase and not for enzymesinvolved in the galactose-metabolizing pathway (33, 34). AsEscherichia coli seems to possess the same Leloir pathwayas L. helveticus (2), we decided to use E. coli gal mutantstrains as initial host system for our work.

MATERIALS AND METHODS

Bacterial strains, plasmids, and phages. L. helveticus typestrain N2 is from our collection and is identical to strainATCC 15009 (American Type Culture Collection). N123 isL. delbrueckii subsp. bulgaricus type strain NCDO 1489(National Collection of Dairy Organisms, Food ResearchInstitute, Reading, Great Britain). The E. coli strains usedare listed in Table 1. E. coli plasmids pACYC184 and pUC19served as cloning vectors (8, 53). Phage X1105 carries ahigh-hopper, kanamycin-resistant variant of TnlO and wasobtained from Nancy Kleckner's laboratory (48).

Media. L. helveticus was grown in lactobacillus MRSbroth at 42°C (13). E. coli strains were grown in LB (0.5%NaCl, 1% tryptone, 1% yeast extract) or minimal mediumM9 (70 mM KP04 buffer [pH 7.0], 10 mM NaCl, 20 mMNH4Cl, 1 mM MgSO4, 0.1 mM CaCl2, 0.002 mM FeC6H5O7)supplemented with 1 mM each essential amino acid andeither 0.4% glucose or 0.4% galactose (27). Media weresolidified for plating by the addition of 1% agar. Chloram-phenicol, kanamycin, and ampicillin were added to yieldfinal concentrations of 30, 50, and 100 ,ug/ml, respectively.X1105 was propagated in E. coli NK5012 in Xym broth (1%tryptone, 0.25% NaCl, 0.2% maltose, 0.1% yeast extract)and on tryptone plates (1% tryptone, 0.5% NaCl, 1% agar).

Cloning of the gal system. Chromosomal DNA of L.

4464

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GALACTOSE UTILIZATION IN L. HELVETICUS 4465

TABLE 1. E. coli strains used

Strain Relevant features Reference ororigin

K802 hsdR2 galK2 galI22 supE44 52HB101 hsdS20 galK2 supE44 recAJ3 6MM294 hsdR17 supE44 3MM152 galK2 recA3 16N23-53 galT23 recA41 37NK5012 supE44 49NK5019 su' (=W3110) N. KlecknerJM105 gal' 31PL225 A(nadA-gal)35 recAl 32

helveticus type strain N2 was prepared as described byDelley et al. (12) and checked on a 0.7% agarose gel forquality and for estimation of the DNA concentration. TheDNA was partially digested with Sau3A and size fraction-ated on an 0.8% low-melting-temperature agarose gel, andthe 6- to 12-kb fraction was ligated into the BamHI-digested(unique site), dephosphorylated vector pACYC184. Theligation mixture was transformed into E. coli K802, platedonto M9 plates supplemented with methionine, proline,galactose, and chloramphenicol (30 ,ug/ml), and grown at37°C for 3 days. Colonies were pooled and grown in theappropriate minimal medium M9 to saturation, and plasmidDNA was extracted as described by Maniatis et al. (28). E.coli K802 was transformed with the plasmid pool and platedonto identical minimal medium plates as described above.After growth at 37°C, single colonies were picked andpurified by being restreaked on agar plates, and plasmidDNA was extracted and analyzed by restriction site map-ping. Three such independently isolated clones were namedpLHgall, pLHgal2, and pLHgal3.

Plasmid constructions. The 7-kb internal BamHI fragmentof plasmid pLHgall (Fig. 1) was isolated from the agarosegel, ligated into the unique BamHI site of pACYC184, and

ASOB VHI I

A59 B38V HIVE

transformed into E. coli K802. The transformation mixturewas plated onto minimal medium plates supplemented withmethionine, proline, and galactose and incubated at 37°C for2 days. Single colonies were isolated and purified on identi-cal minimal medium plates as described above, and plasmidDNA was extracted and analyzed by restriction site map-ping. Plasmids carrying the 7-kb BamHI fragment in eitherorientation were identified and named pNP1 and pNP2. Asynthetic SalI-XmaI-ApaI adaptor was introduced at theunique SalI site of pNP1 and pNP2, located within thepACYC184 sequence, close to one end of the insertedBamHI fragment. The resulting clones were named pNP4and pNP5, respectively. Both plasmids pNP4 and pNP5were used for exonuclease III deletion mapping by firstprotecting the vector by cutting at the ApaI site and thenstarting the exonuclease digestion at the SalI site. PlasmidspNP7 and pNP8 are two representatives of the plasmidsobtained (Fig. 1). The 1.4- and 2.4-kb EcoRI fragments ofpNP5 were cloned into the EcoRI site of pUC19, resulting inplasmids pNP16 and pNP17, respectively.Transposon mutagenesis. Plasmids pLHgall, pLHgal3,

and pNP5 were transformed into NK5019 and grown ontryptone plates supplemented with 30 ,ug of chloramphenicolper ml. Single colonies were picked and placed in Xym brothwith chloramphenicol and grown at 37°C. We chose akanamycin-resistant variant of TnJO as a transposable ele-ment, on a A phage as the delivery system, X1105 (48). Thecells were infected at a multiplicity of infection of 0.5 andincubated for 30 min at room temperature and 90 min at 37°Cto allow for transposition and expression of the kanamycingene. The cells were spread onto LB plates containing 50 ,gof kanamycin per ml and 1.25 mM sodium PP, and incubatedovernight at 42°C. The 1,000 to 2,000 kanamycin-resistantcolonies per plate were taken up in 3 ml of LB, used toinoculate fresh LB supplemented with kanamycin, andgrown at 37°C. Plasmid DNA was extracted from suchcultures, transformed into HB101 or K802, and grown on

A42

A34A37A41 A39v V E SH HIv

I I....E BI I

ASS

ASH XI

pNP5

A7A23 AS A9 A17 A13AS A2 A6 AlAll A24 A4 A12

E P Vy E HV H W IT v EI I I lIIII

1 kb

BpNP8:

pNP7:

ga/K gaiT ga/M

FIG. 1. Restriction and transposon insertion mapping of the gal region. (A) L. helveticus fragment of pLHgall. (B) Enlargement of thefragment in panel A. Bars underneath the maps mark the extents of the indicated subclones. Restriction sites are as follows: B, BamHI; H,HindIII; E, EcoRI; and V, EcoRV. TnlO insertion sites within the galK gene (V), within the galT gene (v), and outside the galK and galTgenes (V) are indicated and labeled with isolation numbers. Transposon insertions shown in panel A are within pLHgall and those shown inpanel B are within pNP5. The arrows underneath the maps indicate the extent and orientation of the galK and galT genes and part of the galMgene, as determined by DNA sequencing.

B

I-

VOL. 173, 1991

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

4466 MOLLET AND PILLOUD

LB-kanamycin plates. Single colonies were picked andplaced on LB-kanamycin and M9-galactose plates containingleucine, proline, and thiamine to screen for TnJO insertionswithin the plasmids which destroyed the relevant, host-complementing gal function. To screen for lost complement-ing gal functions in hsdR+ strains like MM152 and N23-53,we passed the plasmid pool through MM294. This experi-ment followed the standard A hop procedure described byWay et al. (48). Phages were propagated and titrated onNK5012. Plasmid isolations and transformations were doneas described by Maniatis et al. (28). The transposon inser-tions were mapped by restriction analysis with EcoRI,HindIII, and HindIII-XhoI digests.DNA sequencing and computer analysis. For DNA se-

quencing, we used plasmid pNP5, the exonuclease III dele-tion derivatives of pNP4 and pNP5, and subclones pNP16and pNP17. The dideoxy chain termination method ofSanger et al. (43) was used with a T7 DNA polymerasesequencing kit from Pharmacia LKB. As sequencing prim-ers, we used vector- or fragment-internal 18-mer oligonucle-otides synthesized with an Applied Biosystems 380B DNAsynthesizer and purified on NAP-10 gel filtration columnsfrom Pharmacia LKB. The DNA sequences of both DNAstrands were determined with [ca-35S]dATP and by readingan average of 350 bp per primer and sequencing reaction.TnJO insertions were sequenced with a TnJO-internal se-quencing primer hybridized to agarose gel-purified DNAfragments containing only one transposon end or to appro-priate subclones of such DNA fragments in pUC19. ForDNA sequence and resulting protein sequence analyses, asoftware package from Genetics Computer Group Inc.,Madison, Wis., was used (14).Enzyme assays. Lactobacilli were grown in MRS broth and

E. coli was grown in LB supplemented with chloramphenicolto the mid-log phase, and 12-ml portions of these cultureswere centrifuged and resuspended in 500 ,ud of 10 mMpotassium phosphate buffer (pH 6.9). The cell suspensionswere cooled on ice, and the cells were lysed by ultrasonica-tion. The lysates were cleared by a 15-min centrifugation at13,000 x g, and the supernatants were used for determininggalactokinase and galactose-1-phosphate uridyl transferaseactivities as described by McKenney et al. (30) and Maxwellet al. (29). Total protein concentrations were estimated by adye binding assay with a kit from Bio-Rad (7).

In vitro transcription-translation assay. As an E. colicell-free coupled transcription-translation system, we useda prokaryotic DNA-directed translation kit from Amershamas recommended by the supplier. The template was 5 ,ugof CsCl-gradient purified, covalently closed circular DNAper assay (28). As a radioactive label, we incorporatedL-[35S]methionine. The proteins were separated on a 10%sodium dodecyl sulfate (SDS)-polyacrylamide gel (24). Thegel was fixed for 1 h in a solution containing 30% methanoland 10% acetic acid, impregnated for 10 min with a fluoro-graphic reagent solution (Amplify; Amersham), dried, andexposed to an X-ray film for 12 h with an intensifying screenat -70°C. The prestained, low-range SDS-polyacrylamidegel electrophoresis standards (Bio-Rad) were used as proteinmolecular weight markers.

Chemicals. Lactobacillus MRS broth, tryptone, yeast ex-tract, and agar were purchased from Difco Laboratories;chemicals were purchased from E. Merck Chemicals Inc.;and antibiotics were purchased from Sigma Chemical Co.Restriction endonucleases, polymerases, T4 DNA ligase,exonuclease III, and alkaline phosphatase were purchasedfrom Boehringer Mannheim Co. and New England BioLabs

Co. and used as recommended by the suppliers.[a-35S]dATP, L-P5S]methionine, and D-[1-14C]galactosewere purchased from Amersham, and the enzymes andsubstrates used for the galactokinase and galactose-1-phos-phate uridyl transferase assays (29, 30) were purchased fromSigma.

Nucleotide sequence accession number. The EMBL DNAData Library sequence accession number for the gal regionnucleotide sequence is X57248.

RESULTS

Cloning of the L. helveticus gal genes. By shotgun cloning ofsize-fractionated L. helveticus DNA, we were able to isolateseveral clones complementing the gal lesion in E. coli K802.Three independent clones, pLHgall, pLHgal2, and pLH-gal3, were analyzed. To test further the complementation ofgal lesions in other E. coli strains, we transformed the threeclones into HB101 and, after passage through MM294 forappropriate DNA methylation, into N23-53 and MM152. Alltested strains were able to grow after transformation onM9-galactose plates, indicating sufficient complementationof the galK and galTE. coli mutant genes by the correspond-ing L. helveticus genes carried on pLHgall, pLHgal2, andpLHgal3. The three clones were analyzed by restriction sitemapping. The results indicated that we had isolated threefragments of between 8 and 10 kb. They all showed the samerestriction pattern within the overlapping region of thecloned inserts. Southern blot analysis (44) confirmed theDNA homology of these clones (data not shown). Therefore,the three plasmids are independent clones of the samegenome region of L. helveticus N2. For subcloning studies,we concentrated on pLHgall. The 7-kb BamHI fragmentwas subcloned into a pACYC184 derivative vector, resultingin plasmid pNP5 (Fig. 1). Further deletion cloning withexonuclease III resulted in pNP7 and pNP8. These threeplasmids were tested for complementation of the gal lesionsin HB101, N23-53, and MM152 as described above. PlasmidpNP5 was able to complement these galK and galT lesions,whereas pNP7 and pNP8 could only complement the galTlesion. Thus, we could narrow down the galT-complement-ing factor of L. helveticus to a 2.3-kb stretch of DNA. Toidentify further the region carrying the galT- and galK-complementing genes, we decided to use transposon muta-genesis.Mapping of the gal-complementing genes by transposon

mutagenesis. Plasmids pLHgall and pLHgal3 were subjectedto TnJO transposon mutagenesis as described in Materialsand Methods. Total plasmid DNA from such transpositionexperiments was transformed into HB101 and screened forTnlO-carrying plasmids by growth of the transformed cellson LB-kanamycin plates. Single colonies were picked andscreened on galactose plates for loss of their ability tocomplement the galK lesion of HB101. Seventeen suchcolonies were isolated, and their plasmid DNAs were ana-lyzed by restriction site mapping and compared. We foundthat in all cases, a TnWO element was inserted within thesame small region of about 1 kb on pLHgall or pLHgal3.The integration sites of four representative isolates, A34,A37, A41, and A42, are shown in Fig. 1A. The integrationsites of a few characterized isolates with no loss of theirability to complement the HB101 galK lesion are alsoshown. They are all clearly located outside the small regionof the above-described isolates. We therefore localized theL. helveticus galK-complementing gene to a defined smallstretch of the clones.

J. BACTERIOL.

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GALACTOSE UTILIZATION IN L. HELVETICUS 4467

The same transposition experiment as that describedabove was repeated with pNP5 as a recipient plasmid.Plasmids carrying TnJO were analyzed for loss of the galK-or gaIT-complementing capability with HB101, N23-53, orMM152 as a host. Several plasmids were isolated, and theirTnlO insertion sites were mapped. The results are summa-rized in Fig. 1B. All characterized TnJO insertion sitesaffecting the galK-complementing factor were located withinthe same small DNA region as that observed with pLHgalland pLHgal3. Transposon insertions affecting galT comple-mentation were also localized to a defined small region ofabout 1 kb situated next to the galK-complementing region.Thus, we have identified two regions within pNP5, oneresponsible for galK complementation and the other respon-sible for galT complementation. These results are consistentwith the observed complementation characteristics of thedeletion subclones pNP7 and pNP8.

Nucleotide sequence analysis of the L. helveticus gal region.The DNA sequence of the galK- and gaiT-complementingregions of pNP5 was determined and is shown in Fig. 2.Computer analysis revealed the presence of two open read-ing frames (ORF) spanning 1,167 and 1,467 bp, respectively.They are both oriented in the same direction and use thesame reading frame. They are only separated by 21 bp,suggesting an operonlike structure. Proposed translationalstart and stop codons and putative ribosome binding sites areindicated in Fig. 2. In addition, E. coli-like promoter se-quences in which the promoter sequence of the second ORFoverlapped the 3' end of the first ORF were found upstreamof each ORF. Several sites of insertion of TnJO elementsaffecting the galK or galT complementation competence ofpNP5 were determined by sequencing across the integrationjunctions. For isolate A3, the integration junctions at bothends of the element were determined and confirmed theformation of a 9-bp target duplication of TnJO (23). For allother isolates, only one side of the integration site wasdetermined. A comparison of the integration sites with theresults of complementation study showed that all the TnJOinsertions affecting galK function were located within thefirst ORF; galT function-affecting TnJO insertions werelocated within the second ORF. We therefore denote the firstORF as being the L. helveticus galK gene and the secondORF as being the L. helveticus galT gene.The galK and galT genes were translated into amino acid

sequences and compared by dot plot analysis with thecorresponding E. coli genes (11, 26). The results are shownin Fig. 3. The two galK genes are of similar size and showclusters of significant homology distributed throughout theirentirety. This homology confirms the related functions of thetwo genes and the position of the proposed start codon.There is no apparent amino acid sequence homology be-tween the two galT genes. The L. heiveticus gene is morethan 100 amino acids longer than the E. coli gene. However,the TnlO insertion in isolate Al, which is close to theproposed beginning of the gene, and the presence of aribosome binding site may confirm the position of theproposed start codon for this gene.DNA sequence analysis of the regions flanking the galK

and galT genes revealed a third ORF, reading in the samedirection as the two identified genes (Fig. 2). Furthermore,computer analysis of codon preference, based on the L.helveticus codon usage of the galK and galT genes, identifiedthis ORF as a potential protein coding sequence (4). ThisORF starts 100 bp downstream of the end of the galT geneand reads beyond the pNP5 clone. It is preceded by an E.coli-like promoter sequence and putative ribosome binding

site. The DNA sequence of this ORF was determined onlypartly. The resulting amino acid sequence was comparedwith those of known proteins of the galactose pathway fromother organisms and showed significant homology to theamino-terminal part of the sequence of the galM (mu-tarotase) gene of S. thermophilus (41). An alignment of thesesequences is shown in Fig. 4. The amino acid sequenceidentity is 31% (35). We assume that both proteins have ananalogous or identical function and, therefore, refer to thisthird ORF as the L. helveticus galM gene.

Transcription and translation of the gal genes. We intro-duced plasmid pNP5 and its TnJO-carrying derivatives intoan E. coli cell-free transcription-translation system andseparated the resulting proteins on an SDS-polyacrylamidegel (24). As shown by an autoradiogram of such a gel in Fig.5, pNP5 produced several different polypeptides which werenot present with vector pACYC184. The three largest pep-tides expressed were about 53 kDa (a), 44 kDa (,B), and 41kDa (-y), respectively. The two isolates from pNP5 transpo-son mutagenesis which have the galT gene disrupted by aTnJO insertion, Al and A4, lost the expression of the 53-kDaprotein. We therefore identified this protein as the galT geneproduct. Its apparent size is in agreement with its calculatedmass of 55.6 kDa, as deduced from the nucleotide sequence.The two pNP5::TnJO isolates which have the galK genedisrupted, A5 and A6, did not express the 44- and 41-kDapeptides. The 44-kDa protein corresponds to the galK geneproduct, which has a calculated mass of 43.3 kDa. The41-kDa protein is about 25 amino acids shorter than the galKgene product. It is encoded by the same ORF as the galKgene product and is either a specific degradation product ofit or a result of premature protein termination. As the 41-kDapeptide is not present on isolate A5, which has TnlO insertedclose to the ATG starting codon, the presence of a possiblesecondary start site is excluded. Isolates A4 and A6 expressadditional peptides of about 38 kDa (5). They correspond insize to the cryptic galT and galK gene products, which wereinterrupted prematurely by the inserted TnWO element. Thecorresponding peptides of isolates Al and A5 were too shortto be identified. Furthermore, several other peptides below33 kDa were present with pNP5 and its TnlO-carryingderivatives but not with vector pACYC184 on the autoradio-gram (Fig. 5). They have not been identified and may partlyrepresent degradation products of the above-mentioned pro-teins as well as polypeptides expressed from smaller ORF,including that of the partial putative galM gene product,flanking the galKT region.

Determination of galactokinase and galactose-l-phosphateuridyl transferase activities. The enzymatic activities of ga-lactokinase and galactose-l-phosphate uridyl transferase incrude cellular extracts of Lactobacillus and E. coli strainswere determined. The results are shown in Table 2. L.helveticus N2 showed a galactokinase specific activity sim-ilar to that of E. coli JM105, whereas the galactose-l-phosphate uridyl transferase specific activity was increased.For L. delbrueckii subsp. bulgaricus (N123), a close relativeto L. helveticus, no such specific activities were detectable,indicating that these genes either are missing or are notexpressed.We determined the galactokinase and galactose-l-phos-

phate uridyl transferase activities encoded by plasmids pLH-gall, pNP5, and their derivatives carrying TnWO insertions.As a host, we used E. coli PL225, which has its own galoperon deleted. The results can be summarized as follows(Table 2). Clones pLHgall and pNP5 code for both enzymefunctions. The initial activity for each enzyme was similar in

VOL. 173, 1991

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

4468 MOLLET AND PILLOUD

(-36) (-IS)MG AT"TTTCTTCAAGOGTATTATCAGTTTTBTATTGCAATAAACTAAATAATATAAAT

_Z3-.16ATTTATTTTAAATAAACTATTACTAAATACGTT TTATTTT

SD start gIK1020 AAAcGArrACAGBATAAATTTBAGGTAAAAC1^AT AAAGAAGAACTACTTAAGAAA

N N K E E L L K K

.r--A .1 TACT

Y E A T S N E K A K D V F F S P G R I N

114 CTAATTGGTGAACATACTGACTACAACGGTGGACAGT|CCA ATV I B E H T D Y N N VH V F P A P I S L

1200 GGTGMTrATGGTCTATBGTCCA CAGGCAACB V Y B V Y B P R D D K K V R L Y SG N

1MG GTTil CGTTGTGATGACATTrGATACACAAATGTTGAAAAAGoATGAAGmACV D G D I V E F D I D D T N V E K D E D

1320 CG CCTGTAACTACTTTAAGATGATCACIATTTCGTGAAAAGTATGATGTR F V A N V F K G N I T Y L R E K Y D G

1355 ATTGATCATGGCTTTAACTTGTACATCGAAGCTAACTTGCCATCAGGTTCAMTCTTTCAI D H G F N L Y I E A N L P SGGS C L S

S SAA I E N L N G I I L K D E F N L D

1555

1S"

1GM

IM

1740

1500

SG

1920

1l85

204

2100

34M

25"

252

2M"

2a"

27"

2M

25"

21b

CCCCTCTC-A ^CTWATATATTGTGV DR V S L A K N C Q R T E N E F I G L

AATAGlTTATTCTrGCTCA CTTAT"AAAATGGTACN S G I NDF ACI N G K K N S A I F

CTAITTCAACATCAACTATGTACfCCCTGCATTAGTTAGATCL D C N T L K Y E Y L P L A L D Y E I

I I N A T N N P H T L A D S A Y N N R V

.~~~~~~~GGATACGA E C G R A L K K L a a K L D I K A L G

A._E L D N D T F D E VS Y L I N D E T E I

r-kkCTC^ ~~~~~~~~~~ATCAGTATG

K R A R H AY S E N RR T L R A T C A N

* *AG*l_K D G D L E K L ft L I N ASNH E S L H

Y D Y E V T G K E L D T L A E A S V K C

P G V L G A R N I G G G F G G SA I A I

V K K S E A E N F KK N BG K I Y R D A

21G5 GTTTATGAGTATA TTTAA

V C Y D A S F Y D A E I V D G T K R I

*J- start gaIl220 TrAGTGTbCGAAGGGAAG AATTAATTGAAAAATTTGCGATGAAGTTATAAAT

NiK I I E K F A D E V I N

2285 G_ ACA_C CTT T

S A Y E P L D R V Y V I N K I R A LV

Al-2340 AATAGTTATTTA

G D H D E E EE D PA A K L V D L

320"

S31

3185

3365

3485

3584

36"0

3?"

3785

A V K N E K I P D D I T S R E V L N D C

CTTATATCTACrAC@CAACA A TA CTC TGL V D L A T P T P S K T N S I F V C K E

C K S S E E N T D V F Y K L C E D N N Y

BTTAAAAAAAAGMCTATTGCTAAGAATG1TGrTTAMTACTAGTTCTAAAGGCACV K K E A I A K N V V F SB T S S K C H

AM4

AGAATAAAlA TTACTTATCAAGTCAAAAGTCGAGCCATGCT6S L E I T I N L S K P E K D P K A I A A

A A H A T B K K Y P C C A L C L E N E G

TA1TffAGTGB1TATGTAAGAATGCTCTCTATTTAAGAATTATCCGCATGAACATTY LGBG BG K NM R S N L f I I R N N I

M,CTCCATGGTCTTCAATATTCACCTTATGCATACTTTAATGAACACTGCATTA B R P V G F C Y S P Y A Y F N E H C I

TTTC TTCGT_TCACC1TGATTACIK"TF L D C K H I P N V I N Ca T L I N L V

CAAATTGAAAAAATC1TTCCTCATTA1TGTGGTTATGrGTTACA1T1E I E K I F P H Y F V S N A D L P I V

G GB S L A H E H Y 0 G G R H T F P E N

IAAGCGMCAGAATJTTTTATCTACCTAATGGCBCTCK A G I K K N I N F D S Y P K V V A G I

V D V P N S D L R L T SD SN L D L I D

A13 A12

CTTGGCAAAGA1T1TATCTTCTACTBACTB .rAAGL B S K I I K F E D N Y S D T A R D I K

C CTrGTGGAATCCATACTGCACATCAGACTAAGTAAA Y E B E T R H H T V T P I N H R E G K

AATTTBGTcCrI"TCTTTTITFAG*TBTAACAATACTABCGAGAAATACCACTTGGN F V L D L V L R D N N T S E K Y P L B

I F H P H K Q L V H I K K E N I 4 L I E

B1TATTAWGATATTITrGCCBAGCTAATTAGTTGAACTTGGAAGAATAAAGAAAV N B RA I L P 4 f L K S E L EE V K K

TAC TlZTTZXY W L B E D N K N A A S H K E V A D C V

AGGTGAAAACAAATTAGCTTGATAATGTTBTCAAGTG TGGAACATCAC1TGTTK A E N E I S L D N V D Q V N E O S L V

BAAGTA1TTGAACAAGTTCrTCAAGAT CBTBTAITTAABAATAATBCAGATGGCGAGE V F E VV L C D A 4 V F K N N A D G E

E B V D K F I T A L T K E IK *

TGTCTC1TTrTBTTCBGACTSCTATOC1TT TAC

-IG . . . .D *tart gaiN

NAAATSAF

J. BACTERIOL.

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GALACTOSE UTILIZATION IN L. HELVETICUS 4469

NIY SR K D S K D L C E I V L E N D H

3,"

N I V K v L N Y C A T L E K V L L N D

E W E I L L NS P A D YS Q E R N Y L

4140

S S T v R I A S R v R K S QO R NB L

E T H Q L P I N D S E 1 H I HSS 1It T

D T E V E D F t P S C S E N S A R V D L

52" TAA1TIATIGATOC

T L L D

FIG. 2. Nucleotide sequence of the L. helveticus gal region.Start codons of the galK, galT, and galM genes are boxed andlabeled. The deduced amino acid sequences of these genes are givenbelow the DNA sequence. Putative E. coli promoters (-10 and -35)and ribosome binding sites (SD) are indicated by bars above thesequence. TnJO integration sites are labeled and shown as left-handbrackets (r- ) for the left and right-hand brackets (-I) for theright junction sites. The confirmed 9-bp TnlO target duplication ofisolate A3 is underlined.

both clones. In comparison with the enzyme activities inparent strain N2, the galactokinase activity was similar andthe galactose-1-phosphate uridyl transferase activity waselevated by a factor of 3. TnlO insertions within the firstORF, the galK gene, abolished the expressed galactokinaseactivity of the clones, whereas TnJO insertions within thesecond ORF, the galT gene, abolished the expressed galac-tose-l-phosphate uridyl transferase activity of the clones.We hence confirmed the E. coli complementation results. Itis worth mentioning that TnlO insertions in the galK gene notonly destroyed the galK function but also significantlyreduced the expression of the galT gene, as measured by theinitial activity (Table 2). This result indicates that the pro-moter in front of the galT gene is only weakly active andsuggests that the galT gene is normally expressed by thesame promoter as the galK gene.

0

A

100 200 3000

I

L. helveticus ga/lK

1 WKTSFNKYGRKDGKDLCEIVLENDHGMIVKVLNYGATLEKVLLNDE L .he lv11 : 1: 1: 1:: I :11::I ::: 1111: 1:

1 MKISCEI IGKVDSGDVSKISMENNNGVVISTLTTGATLOEFLVPMETGAL S. term

47 NMILSLNSPADYSQERNYLGGTVGRIAGRVRKGOWRHGLETHOLPINDG L .he lv1::1:: 11 : : :11:111: 1:: I: 11 1:1

51 KNIVLGFSDFEDYYKNNLCACOSITnVAGRIGKASYTHNMVLYSLPKNEG S. term

96 ENHIHGGIGTDTEVWDFKPSCSENSARVDLTLL 128:1 :111 :1 :: 1: ::

101 DNCLHGG .PKGMQVONWNYVTNLNDDYVETKFI 132

L.helv

S. term

FIG. 4. Alignment of the amino acid sequences of the L. helve-ticus (L.helv) and S. thermophilus (S.term) galM gene products.Dots within the amino acid sequences represent gaps. Identical (I)and conserved (:) amino acids in the sequences are indicated. Thenumbers correspond to the positions of the amino acid sequences.

DISCUSSIONWe have described the isolation and analysis of the L.

helveticus galK and galT genes. This was possible becauseboth genes were expressed well in E. coli and were able tocomplement lesions of their corresponding E. coli genes.Biochemical analysis of crude protein extracts from appro-priate transformed E. coli strains confirmed the encodedenzymes of these genes as being galactokinase and galac-tose-1-phosphate uridyl transferase. A direct analysis of theisolated genes in L. helveticus or a related lactobacillus wasnot possible because of a lack of appropriate mutants and afeasible transformation system.

Organization of the L. helveticus gal region. The L. helve-ticus galK and galT genes are located very close togetherand use the same reading frame (Fig. 2). They are likely to beorganized as an operon and transcribed by a commonpromoter, located upstream of the galK gene. This organi-zation may explain the observed polar effect, i.e., thedecrease in galactose-l-phosphate uridyl transferase activityupon TnJO insertion into the galK gene (Table 2). However,an E. coli-like promoter sequence can also be found in frontof the galT gene, overlapping the end sequence of the galKgene. This would explain the weak residual activity ofgalactose-1-phosphate uridyl transferase in the above-men-tioned transposon mutants as well as the reduced galTprotein level in these mutants observed in the in vitrotranscription-translation assay (Fig. 5). Nevertheless, thelevel of galT expression was sufficient to complement theappropriate E. coli galT mutant strains. As complementationand promoter studies cannot yet be undertaken with L.

100 200 300 400. ..... ...... .. . . .. .. .. ...

- 23

-2C

.. . . . . .

Lhel.vet'i.cu,s' 'gaiT

B -,, ... ~

FIG. 3. Amino acid sequence comparison between L. helveticus and E. coli galactokinase (A) and galactose-1-phosphate uridyl transferase(B) (36). The comparison was done on a VAX computer with software from Genetics Computer Group Inc. (14). The window size was 6, andthe stringency was 5 amino acids. The numbers indicate the amino acid positions of the proteins.

34

a j 2

-IC

/: ,

.,*@@.... ,... ./tw .. -0

VOL. 173, 1991

I

00

00

DO

00

00

00

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

4470 MOLLET AND PILLOUD

6 5 4 3 2 1,Nor -Iwo NW -- M

t-., ...

-110

- 84

am47

65 --o

- 33

24

- 16

TABLE 2. In vitro enzyme activities

Activity of:

Strain Plasmid Galactose-1-(isolate) Galacto- phosphate

kinasea uridyl trans-feraseb

L. helveticus N2 NCc 1.5 3.0L. delbrueckii subsp. None d

bulgaricus N123E. coli JM105 F' 4.4 0.15E. coli PL225 None

pACYC184 -pLHgall 2.1 10.2pLHgall::TnJO (A39) 7.4pLHgall::TnJO (A41) 0.1 2.7pLHgall::TnlO (A42) 0.15pLHgall::TnlO (A55) 2.7 11.0pNP5 2.5 9.8pNP5::TnO (A3) 0.26pNP5::TnlO (AS) 0.78pNP5::TnlO (A6) 0.55pNP5::TnlO (A23) 0.29pNP5::TnJO (Al) 5.1pNP5::TnlO (A4) 5.1pNP5::TnJO (All) 4.9pNP5::TnJO (A12) 4.0pNP5::TnlO (A13) 3.1pNP5::TnlO (A24) 4.8

a Count(s) per minute per nanogram of total protein (30).b Unit(s) of optical density at 340 nm per minute per nanogram of total

protein (29).c NC, not characterized.d_d, no measurable enzyme activity; i.e., <0.01 cpm/ng or <0.01 unit of

optical density at 340 nm per min per ng.

FIG. 5. In vitro transcription and translation of the L. helveticusgalK and galT region. Lanes: 1, pACYC184; 2, pNP5; 3,pNP5::TnJO (isolate Al); 4, pNP5::TnlO (isolate A4); 5, pNP5::TnlO(isolate A5); 6, pNP5::TnJO (isolate A6). Numbers represent themolecular masses of the markers in kilodaltons. Protein bands a, ,

-y, and 8 are 53, 44, 41, and 38 kDa in size.

helveticus as a host system, the mode of expression in thisorganism remains unsolved.Many organisms have the genes constituting the Leloir

pathway for galactose metabolism, the galK, galT, and galEgenes (the last coding for the uridine diphosphogalactose4-epimerase), clustered or organized as one operon (1, 2, 50).L. helveticus, however, shows a clustering of only the galKand galT genes, which are followed downstream by a puta-tive galM gene. The DNA sequence of the upstream flankingregion, determined for about 1 kb more than presented inFig. 2, did not show any long ORF. Therefore, the galE genewas not found immediately next to the galK and galT genes.Furthermore, plasmid pNP5 did not complement the appro-priate E. coli galE mutant strains (data not shown). A similarsituation was found in Erwinia stewartii, in which the galEgene is also not directly linked to the galK and galT cluster(15). It was shown that it was linked to genes for thesynthesis of capsular polysaccharide (cps), in which it playsan important biosynthetic role, and that it was expressedconstitutively. The galKT operon was inducible by galac-tose. A similar observation was also made for Vibrio chol-erae (19).

DNA sequence analysis revealed that the putative galMgene of L. helveticus is located downstream of the galKToperonlike cluster. As reported for S. thermophilus byPoolman et al. (41), we also found good homology on theamino acid sequence level between the determined galMpart and the carboxy-terminal part of the GALIO geneproduct from Kluyveromyces lactis (24% amino acid iden-tity) (51). However, we did not detect the galE gene in frontof the galM gene, as in S. thermophilus or as within theapparent hybrid epimerase GALIO in K. lactis.

It is interesting to observe that the number and order ofgenes within the gal operons of different organisms vary. InE. coli and other enterobacteria, the operon is organized asgalETK (2, 20); in Streptomyces lividans, it is organized asgalTEK (1). In L. helveticus, there are only two genes withinthe proposed operon, with galK located in front of galT.These differences in gene organization suggest that theseoperons do not have a common structure as an ancestor butevolved separately from individual genes into regulatedexpression units. This assumption is consistent with theobservation that the galK and galT gene families differ intheir evolutionary relationship, as will be discussed furtherbelow.

gal gene products. As shown in Fig. 3, the L. helveticusand E. coli galactokinases show significant homology (35%amino acid identity) throughout their entire amino acidsequences. We also included the reported amino acid se-quences of the galK proteins from S. lividans (1), K. lactis(51), and Saccharomyces cerevisiae (10) in the comparativestudy. A computer-generated alignment of all these se-quences is shown in Fig. 6. The three different galactoki-

_-o

J. BACTERIOL.

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GALACTOSE UTILIZATION IN L. HELVETICUS 4471

1 18WAKE ELLKK..EA TSN....

NSLKE KTQSL.. FAN AFGP....MG EAVGE.. PSA SGSGSCTRS

MSVPIV PTDFAHAPKL ECLNEFVEC YDTS....NTKSHSEEVI VPWNSSAKE LPPLAEKP SI1KK..FlS AYD....

19 I 63.KDVFF.. GRINVIEHT DYNGH PIS..LGVYG VYGRDDKKV.THTIQ.. w GRVNLIGEHT DYNDGFVLP AID..YTVI SCAPRDtKVRRGC6R.. GRENLIGEHT DYNGFP PCR..TRSRP SPGAN.GIL. RKFFI T NIGEHI DYQH SENDLMLACR LTSESENPSI.DFVAR.. I D AID..FDNLC AVKVLNEKNP

.Z4 II 98RL..YSGNVD D1....VEF DIDDTNVEKD E NF K.GRV..NAADYE NaL....DEF SLDAPIVAHE NYC.. VNV R.GC.RL ..HSADVD ADP...VEL RVAD .. LAPA SDKS. ATAYP S.GTLTNHDSNFA QRK.... FDL PLDGSLIEID PSVSD SNF KC...SI. .TLINAD PKFAQRKFDL PLDGSYVTID PSVSD NtYFKCKHVAHSF

99 142MIT.. .YLRE KYD. .GID.H GLYIEANL PSGSGLSSSA Al IILVVK...HLQL RNN..SfG.. GVDMVISGNV PQCGGLSSSA SL. AVGTVLVL.. . ALRE AGH. .ELT.. GADVHLASTV PSCGALSSSA ALE VRPLANLLVVQOFLQE KYWFKGPV.H GMEIYVKGDI PSGGGLSSSA AF . ......

LKK...LAPE RFA..SAPLA GLQVFCEGDI PTGSGLSSSA AFI AVALAV

143 187KDEFNLDVDR VSLAKNGQRT . . ENEfIGLN SGINDOFA.. CINGKKNSACOLYHLPLDG ACIALNGOEA . . ENCfVGCN CGINDOLI.. . SALGKKDHANDLYALALRG MOLARLCORA ..ENYVGAP VGINCTASA CCEAGTPSSS.......ICA VS TI ... Y ..SNVPAG SC 8

VKAUGPGYH MSKCNLMRIT VVADIMLVLT NAVUIRLP.. .LFAVRK IML

18 231IFLDCNTLK. . . YEYLPLAL GDYEIII ... ATNNPHTLA DSAYNNRVAELLIDCRSLG. .. TKAVSMPK G.VAVVl... INSNFKRTLV GSEYNTRRECTPATSPSGR. .. SPSTSPPR GCACSS....... TPGSSTPTAR AST.....ASYTLSSNAVEA TPFKFPQLKN HEISFYIANT LVVSNKFETA PTNYNLRVVE

232CGRALKKLtC. ..........

CETGARFFO. ..........

AARAARRAP. ..........

VTTAANVLAA TYGVVLPSGK

2"ETE...........

LDP.................EEE...........

DIESGIERLT KNLVLVEESL

265.... .CKLDI KALGELDNDT FDEYSYLIND..... .. QPALRDVT IEENAVAHE.... .RCUA . STRCDVPYAD LDAALERLGD

EGSSTNKlCN4 FDFNVYYAR YHNISTPM4G

.......... .......... ..........

. . . . . .. . . .. .. . .. .. . .. .. . .. . . . .. . ..............

. . . . . . .. . . .. . . . .. .. . .. .. . .. . . . .. . ..............

ANKKOFSVD DVAQSLNCSR EEFTRtDYLTT

26 IV 294L.Holy .......... I..RIARHA VSENCRTLRA TO... .AMD GD ...... LE.Coli .......... IV R.VRHl LTENARTVEA AS....ALE GD. L

S.Livid .......... V.. RLVRHV VTEDERVERV VALLESATPG AS. AS.Corey SPVRQVLKL Y.. R.AKHV YSESLRVLKA VK....LMTT ASFTADEDFF

295EKLGRLINAS HESLHYDYEVKRMGELMAES HASWRDFEIPSNSRATPAA RRLPHL....KQFGVLMNES QASCDKLYEC

FIG. 6. Amino acid sequence alignment of galactokinases. TheK. lactis sequence has not been determined completely (51) and isonly shown partly. Dots represent gaps in the alignments. Thenumbers correspond to the amino acid positions of L. helveticusgalactokinase. Regions of significant homology are shown as boxes,labeled I to V. Conserved single amino acids are indicated byasterisks.

nases of prokaryotic origin, L. helveticus, E. coli, and S.lividans, are all of similar sizes. Their amino acid sequencealignment did not produce many gaps and showed fiveregions of significant homology. The comparison with thesequences for S. cerevisiae and part of K. lactis, however,showed that the S. cerevisiae galactokinase is significantlylonger than those of prokaryotic origin, as manifested bynumerous gaps in the sequence alignment. Nevertheless, thefive homologous regions are conserved and are, with twoexceptions, similar to the homologous regions of S. cerevi-siae and E. coli, as reported by Citron and Donelson (10).We assume that these five regions are important for enzymefunction and substrate binding. In particular, the conservedregion at L. helveticus protein positions 331 to 350 containsa sequence motif found within the ATP binding sites ofnumerous protein kinases (11).As shown in Fig. 3, there was significantly less homology

between the L. helveticus and E. coli galactose-1-phosphateuridyl transferase sequences (21% amino acid identity).However, the in vitro enzyme assays, together with the E.coli in vivo complementation results, still suggest that wehave identified a galactose-1-phosphate uridyl transferase.We further analyzed the galT amino acid sequence by dotplot analysis and compared it with the corresponding se-quences of S. lividans (1), S. cerevisiae (45), K. lactis (51),and humans (42). No significant homologies were found. Theamino acid identities were 18% with S. lividans, 18% with S.cerevisiae, 21% with K. lactis, and 15% with humans. Ageneral homology search through the entire National Bio-medical Research Foundation protein data bank did notreveal homology to any other known protein either. It isinteresting to note that the eukaryotic galT sequence showedquite good homology to the corresponding E. coli protein(51% amino acid identity between S. cerevisiae and E. coli),whereas that of the gram-positive bacterium S. lividansshowed weaker homology to the other proteins (33% aminoacid identity with E. coli and 23% amino acid identity with S.cerevisiae). We can only speculate about the origin of the L.helveticus galT gene. Possibly it evolved from an ancestor incommon with that of the E. coli galT gene but developedindependently from very early on in the evolution. Theobservation that the same gene from another gram-positivebacterium, S. lividans, also differs significantly supports thishypothesis. However, the possibility that the L. helveticusgalT gene coevolved from a different origin to the sameenzymatic function cannot be excluded.

342 384FGGSAIAI. ...KKSEAEN FKKNVGKIY. .. RAVGYDA SFYDAEIVDGFGGCIVA j. ...PEELVPA VOOAVAEQY. EAKTGIKE TFYVCKPSCGFGGSAIVLV...EAAAVDA VTKAVEDAF. ..AAAGLKRP RVFEAVPRRGGCTVHL GGPNNGNIElt VKEALANEFY KVKYPKITDA ELENAI IVSK

385 388TKRIAGQCAAPQ.TVSRA ASPACTPPALGSCLYEL

ACKNOWLEDGMENTS

We thank F. Beutler for performing all the enzyme assays. Wethank H. Hottinger, P. Niederberger, and C. Gysler for criticalreading of the manuscript.

REFERENCES1. Adams, G. W., J. A. Fornwald, F. J. Schmidt, M. Rosenberg,

and M. E. Brawner. 1988. Gene organization and structure ofthe Streptomyces lividans gal operon. J. Bacteriol. 170:203-212.

2. Adhya, S. 1987. The galactose operon, p. 1503-1512. In F. C.

VOL. 173, 1991

L.HolvE.Coli

S.LividK..LactisS.Corov

L.HolvE.Coli

S.LividK.LactisS. Coro

L.HolvE.Coli

S.LividK.LactisS. Corev

L. HolvE. Coli

S.LividK.LactiaS. Coroy

L.HolvE. Coli

S.LividK.LactisS. Corev

L. HolvE. Coli

S.LividS. Corey

L.HolvE. Coli

S.LividS .Core

L.HolvE.Coli

S.LividS. Corev

L.HolvE.Coli

S.LilydS .Crev

V 341TGKELDTLAE ASK PLG A...RNIGGeTVPCIDTLVE IVKAVIDIt G .VRMTGGG.... LPRAGP CRRHGP GLR6 P. RRRMTGGGSCPEIDKICS IALSN. SYG S...RLTG

L.HolvE.Coli

S.LividS .Core

L.HelvE.Coli

S.LividS. Crev

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

4472 MOLLET AND PILLOUD

Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M.Schaechter, and H. E. Umbarger (ed.), Escherichia coli andSalmonella typhimurium: cellular and molecular biology, vol. 2.American Society for Microbiology, Washington, D.C.

3. Backman, K., M. Ptashne, and W. Gilbert. 1976. Constructionof plasmids carrying the cI gene of bacteriophage X. Proc. Natl.Acad. Sci. USA 73:4174-4178.

4. Bibb, M. J., P. R. Findlay, and M. W. Johnson. 1984. Therelationship between base composition and codon usage inbacterial genes and its use for the simple and reliable identifi-cation of protein-coding sequences. Gene 30:157-166.

5. Boizet, B., J. L. Flickinger, and B. M. Chassy. 1988. Transfec-tion of Lactobacillus bulgaricus protoplasts by bacteriophageDNA. Appl. Environ. Microbiol. 54:3014-3018.

6. Boyer, H. W., and D. Roulland-Dussoix. 1969. A complementa-tion analysis of the restriction and modification of DNA inEscherichia coli. J. Mol. Biol. 41:459-472.

7. Bradford, M. M. 1976. A rapid and sensitive method for thequantitation of microgram quantities of protein utilizing theprinciple of protein-dye binding. Anal. Biochem. 72:248-254.

8. Chang, A. C. Y., and S. N. Cohen. 1978. Construction andcharacterization of amplifiable multicopy DNA cloning vehi-cules derived from the plSA cryptic miniplasmid. J. Bacteriol.134:1141-1156.

9. Chassy, B. M. 1987. Prospects for the genetic manipulation oflactobacilli. FEMS Microbiol. Rev. 46:297-312.

10. Citron, B. A., and J. E. Donelson. 1984. Sequence of theSaccharomyces gal region and its transcription in vivo. J.Bacteriol. 158:269-278.

11. Debouck, C., A. Riccio, D. Schumperli, K. McKenney, J. Jeffers,C. Hughes, M. Rosenberg, M. Heusterspreute, F. Brunel, and J.Davison. 1985. Structure of the galactokinase gene of Esche-richia coli, the last (?) gene of the gal operon. Nucleic AcidsRes. 13:1841-1853.

12. Delley, M., B. Mollet, and H. Hottinger. 1990. DNA probe forLactobacillus delbrueckii. Appl. Environ. Microbiol. 56:1967-1970.

13. De Man, J. C., M. Rogosa, and E. Sharpe. 1960. A medium forthe cultivation of lactobacilli. J. Appl. Bacteriol. 23:130-135.

14. Devereux, J., P. Haeberli, and 0. Smithies. 1984. A comprehen-sive set of sequence analysis programs for the VAX. NucleicAcids Res. 12:387-395.

15. Dolph, P. J., D. R. Majerezak, and D. L. Coplin. 1988. Charac-terization of a gene cluster for exopolysaccharide biosynthesisand virulence in Erwinia stewartii. J. Bacteriol. 170:865-871.

16. Gottesman, M. E., and M. B. Yarmolinsky. 1968. Integration-negative mutants of bacteriophage lambda. J. Mol. Biol. 31:487-505.

17. Hashiba, H., R. Takiguchi, S. Ishii, and K. Aoyama. 1990.Transformation of Lactobacillus helveticus subsp. jugurti withplasmid pLHR by electroporation. Agric. Biol. Chem. 54:1537-1541.

18. Hickey, M. W., A. J. Hillier, and G. R. Jago. 1986. Transportand metabolism of lactose, glucose, and galactose in homofer-mentative lactobacilli. Appl. Environ. Microbiol. 51:825431.

19. Houng, H.-S. H., and T. M. Cook. 1986. Cloning of the galactoseutilization genes of Vibrio cholerae. First colloquium in biolog-ical sciences. Ann. N.Y. Acad. Sci. 435:601-603.

20. Houng, H.-S. H., D. J. Kopecko, and L. S. Baron. 1990.Molecular cloning and physical and functional characterizationof the Salmonella typhimurium and Salmonella typhi galactoseutilization operons. J. Bacteriol. 172:4392-4398.

21. Hutkins, R. W., and C. Ponne. 1991. Lactose uptake driven bygalactose efflux in Streptococcus thermophilus: evidence for agalactose-lactose antiporter. Appl. Environ. Microbiol. 57:941-944.

22. Kandler, O., and N. Weiss. 1986. Regular, nonsporing gram-positive rods, p. 1208-1260. In P. H. A. Sneath, N. S. Mair,M. E. Sharpe, and J. G. Holt (ed.), Bergey's manual ofsystematic bacteriology, vol. 2. The Williams & Wilkins Co.,Baltimore.

23. Kleckner, N. 1979. DNA sequence analysis of TnlO insertions:origin and role of 9 bp flanking repetitions during TnWO translo-

cation. Cell 16:711-720.24. Laemmli, U. K. 1970. Cleavage of structural proteins during the

assembly of the head of bacteriophage T4. Nature (London)227:680-685.

25. Langeila, P., and A. Chopin. 1989. Conjugal transfer of plasmidpIP501 from Lactococcus lactis to Lactobacillus delbrueckiisubsp. bulgaricus and Lactobacillus helveticus. FEMS Micro-biol. Lett. 60:149-152.

26. Lemaire, H. G., and B. MuHler-HIll. 1986. Nucleotide sequencesof the galE gene and the galT gene of Escherichia coli. NucleicAcids Res. 14:7705-7711.

27. Lennox, E. S. 1955. Transduction of linked genetic characters ofthe host by bacteriophage P1. Virology 1:190-206.

28. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

29. Maxwell, E. S., K. Kurahashi, and H. M. Kalckar. 1962.Enzymes of the Leloir pathway. Methods Enzymol. 5:174-189.

30. McKenney, K., H. Shimatake, D. Court, U. Schmeissner, C.Brady, and M. Rosenberg. 1981. A system to study promoterand terminator signals recognized by Escherichia coli RNApolymerase, p. 383-415. In J. G. Chirikjian and T. S. Papas(ed.), Gene amplification and analysis, vol. II. Analysis ofnucleic acids by enzymatic methods. Elsevier-North Holland,Amsterdam.

31. Messing, J., R. Crea, and P. H. Seeburg. 1981. A system forshotgun DNA sequencing. Nucleic Acids Res. 9:309-321.

32. Mizuuchi, K., and T. Fukasawa. 1969. Chromosome mobiliza-tion in rec-merodiploids of Escherichia coli K12 followinginfection with bacteriophage lambda. Virology 39:467-481.

33. Mollet, B., and M. Delley. 1990. Spontaneous deletion formationwithin the P-galactosidase gene of Lactobacillus bulgaricus. J.Bacteriol. 172:5670-5676.

34. Moliet, B., and M. Delley. 1991. A P-galactosidase deletionmutant of Lactobacillus bulgaricus reverts to an active enzymeby internal DNA sequence duplication. Mol. Gen. Genet. 227:17-27.

35. Needleman, S. B., and C. D. Wunsch. 1970. A general methodapplicable to the search for similarities in the amino acidsequence of two proteins. J. Mol. Biol. 48:443-453.

36. Novotny, J. 1982. Matrix program to analyse primary structurehomology. Nucleic Acids Res. 10:127-131.

37. Ogawa, H., K. Shimada, and J. Tomizawa. 1968. Studies onradiation-sensitive mutants of E. coli. I. Mutants defective inthe repair synthesis. Mol. Gen. Genet. 101:227-244.

38. Pilloud, N., and B. MoUet. 1990. DNA probes for the detectionofLactobacillus helveticus. Syst. Appl. Microbiol. 656:250-255.

39. Poolman, B. 1990. Precursor/product antiport in bacteria. Mol.Microbiol. 4:1629-1636.

39a.Poolman, B. Personal communication.40. Poolman, B., T. J. Royer, S. E. Mainzer, and B. F. Schmidt.

1989. Lactose transport system of Streptococcus thermophilus:a hybrid protein with homology to the melibiose carrier andenzyme III of phosphoenolpyruvate-dependent phosphotrans-ferase systems. J. Bacteriol. 171:244-253.

41. Poolman, B., T. J. Royer, S. E. Mainzer, and B. F. Schmidt.1990. Carbohydrate utilization in Streptococcus thermophilus:characterization of the genes aldose 1-epimerase (mutarotase)and UDPglucose 4-epimerase. J. Bacteriol. 172:4037-4047.

42. Reichardt, J. K. V., and P. Berg. 1988. Cloning and character-ization of a cDNA encoding human galactose-1-phosphateuridyl transferase. Mol. Biol. Med. 5:107-122.

43. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.

44. Southern, E. 1975. Detection of specific sequences among DNAfragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517.

45. Tajima, M., Y. Nogi, and T. Fukasawa. 1985. Primary structureof the Saccharomyces cerevisiae GAL7 gene. Yeast 1:67-77.

46. Thompson, J. K., and M. A. Collins. 1988. Evidence for theconjugal transfer of the broad host range plasmid pIP501 intostrains of Lactobacillus helveticus. J. Appl. Bacteriol. 65:309-319.

J. BACTERIOL.

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

GALACTOSE UTILIZATION IN L. HELVETICUS 4473

47. Turner, K. W., and F. G. Martley. 1983. Galactose fermentationand classification of thermophilic lactobacilli. Appl. Environ.Microbiol. 45:1932-1934.

48. Way, J. C., M. A. Davis, D. Morisato, D. E. Roberts, and N.Kleckner. 1984. New TnJO derivatives for transposon mutagen-esis and for construction of lacZ operon fusions by transposi-tion. Gene 32:369-379.

49. Way, J. C., and N. Kleckner. 1984. Essential sites at transposonTnlO termini. Proc. Natl. Acad. Sci. USA 81:3452-3456.

50. Webster, T. D., and R. C. Dickson. 1988. The organization andtranscription of the galactose gene cluster of Kluyveromyces

lactis. Nucleic Acids Res. 16:8011-8028.51. Webster, T. D., and R. C. Dickson. 1988. Nucleotide sequence

of the galactose gene cluster of Kluyveromyces lactis. NucleicAcids Res. 16:8192-8194.

52. Wood, W. B. 1966. Host specificity of DNA produced byEscherichia coli: bacterial mutations affecting the restrictionand modification of DNA. J. Mol. Biol. 16:118-133.

53. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. ImprovedM13 phage cloning vectors and host strains: nucleotide se-quences of the M13mpl8 and pUC19 vectors. Gene 33:103-119.

VOL. 173, 1991

on January 22, 2020 by guesthttp://jb.asm

.org/D

ownloaded from