Vol. 3, No. 5MOLECULAR AND CELLULAR BIOLOGY, May 1983, p. 796-8020021-9193/83/050796-07$02.00/0Copyright © 1983, American Society for Microbiology

Repeated Family of Genes Controlling Maltose Fermentationin Saccharomyces carlsbergensis

RICHARD B. NEEDLEMAN'* AND CORINNE MICHELS2

Department ofBiochemistry, Wayne State University School of Medicine, Detroit, Michigan 48201,1 andDepartment ofBiology, Queens College of City University ofNew York, Flushing, New York 113672

Received 10 September 1982/Accepted 27 January 1983

Maltose fermentation in Saccharomyces spp. requires the presence of any one

of five unlinked genes: MALI, MAL2, MAL3, MAL4, or MAL6. Although thegenes are functionally equivalent, their natures and relationships to each other are

not known. At least three proteins are necessary for maltose fermentation:maltase, maltose permease, and a regulatory protein. The MAL genes may codefor one or more of these proteins. Recently a DNA fragment containing a maltasestructural gene has been cloned from a MAL6 strain, CB11, to produce plasmidpMAL9-26. We have conducted genetic and physical analyses of strain CB11. Thegenetic analysis has demonstrated the presence of two cryptic MAL genes inCB11, MALJg and MAL3g (linked to MALI and to MAL3, respectively), inaddition to the MAL6 locus. The physical analysis, which used a subclone ofplasmid pMAL9-26 as a probe, detected three HindIII genomic fragments withhomology to the probe. Each fragment was shown to be linked to one of the MALloci genetically demonstrated to be present in CB11. Our results indicate that thecloned maltase structural gene in plasmid pMAL9-26 is linked to MAL6. Since theMAL6 locus has previously been shown to contain a regulatory gene, the MAL6locus must be a complex locus containing at least two of the factors needed formaltose fermentation: the structural gene for maltase and the maltase regulatoryprotein. The absence of other fragments which hybridize to the MAL6-derivedprobe shows that either MAL2 and MAL4 are not related to MAL6, or the DNAcorresponding to these genes is absent from the MAL6 strain CB11.

Maltose fermentation by the Saccharomycesyeasts requires at least two enzymes, maltase(E.C. 3.2.1.20, cx,D-glucosidase) and maltosepermease. The synthesis of these enzymes isinduced by adding maltose to the growth medi-um. In the absence of inducer, only low basallevels of each enzyme are present. Based onthese physiological studies as well as on geneticanalyses of the maltose fermentation genes (seebelow), it has been proposed that a regulatoryprotein controls the production of both maltaseand maltose permease.The genetic characterization of the genes in-

volved in maltose utilization in the Saccharomy-ces yeast species has identified a family ofunlinked loci, MALI, MAL2, MAL3, MAL4,and MAL6 (14-16). The presence of any one ofthese loci is sufficient for maltose fermentation.Maltose-nonfermenting mutants which synthe-size only basal levels of wild-type maltase havebeen isolated in a strain carrying the MAL6locus, and their mutations have been shown tomap to the MAL6 locus, forming a single com-plementation group. These noninducible mu-tants appear to be defective in the regulatory

protein controlling the induction of maltase (12).Similar mutants have been isolated in MAL2strains, and it has been postulated that all of theMAL loci code for at least this regulatory protein(17). No mutations in the maltase structural geneor in the maltose permease gene have yet beenidentified, despite extensive mutational analy-sis. One explanation for the inability to isolatestructural gene mutations proposes the presenceof multiple genomic copies of these genes.The possibility that the MAL loci code for

more than just the regulatory protein has beenpresented (5, 6). Naumov, working with naturalisolates of Saccharomyces spp., found a com-plementing gene system involving two comple-mentation groups which he called MALp andMALg. When two maltose-nonfermentingstrains from the two complementation groupswere crossed, the diploid was able to fermentmaltose. Dissection of such diploids enabledNaumov to identify in his strains two MALgenes, of which one was allelic to the MI genedescribed by Winge (and thus allelic to MALl)and the other was allelic to Winge's M2 gene(and thus allelic to MAL3) (6, 13). We refer to

796

MAL GENE FAMILY OF S. CARLSBERGENSIS 797

these MAL genes as MALJg and MAL3g. TheMALp strains identified by Naumov and used tocomplement these MALg strains were shown tocarry a MALp gene which also mapped to theMAL] locus and was therefore designated asMALJp. Genetic analysis of a maltose regula-tory mutant (mal6-17) isolated by ten Bergeshowed it to be complemented by MALJp butnot by MALJg, implying that MALJp codes for amaltose regulatory protein which is functionallyequivalent to the regulatory protein coded for bythe MAL6 locus (7).

Recently, a segment of DNA from a MAL6strain (CB11) has been cloned into a yeastplasmid vector to form plasmid pMAL9-26 (3).This plasmid is able to transform a maltose-nonfermenting strain into a maltose fermentor.It was also demonstrated that this segment con-tains a structural gene for maltase. In the ab-sence of any structural gene mutations, and inlight of the proposed multiple copies of thestructural gene, the possibility that this fragmentmay in fact be unlinked to MAL6 must beconsidered. Rigorous proof that any clonedDNA fragment actually represents the MAL6locus requires that it be mapped to that locus.The observed complementation by pMAL9-26may be due to a variety of causes, including agene dosage effect resulting from the use of amulticopy plasmid.We show here that the cloned segment in the

plasmid pMAL9-26 does indeed representMAL6. MAL6 is therefore a complex locusconsisting of at least two genes, the regulatorygene and the maltase structural gene. We dem-onstrate that the 7.3-kilobase (kb) HindIll frag-ment cloned from the MAL6 region containsMALg activity as well as the maltase structuralgene as shown by Federoff et al. (3). Whereasthis finding suggests that the MALg function isdependent upon the presence of a structuralgene, it is not yet possible to identify the MALgfunction as coding for the maltase structuralgene.Given the functional identity of MAL]

through MAL4 and MAL6 and the availability ofa MAL6 probe, it was also of interest to explorethe possibility that these genes are related at thenucleotide level. We demonstrate that strainCB11 contains multiple unlinked DNA segmentswith significant homology to MAL6. Two ofthese fragments segregate with active MALggenes identified in this strain by our geneticanalysis. These active MALg genes are linked tothe MAL] and MAL3 loci and therefore repre-sent MALlg and MAL3g genes. At least three ofthe genes of the MAL family, MAL], MAL3, andMAL6, are therefore related at the nucleotidelevel. No indication of the presence of MALsequences homologous to the MAL6 probe was

found at other positions in the genome. TheMAL2 and MAL4 loci either lack homology withMAL6 or are deleted for the segment corre-sponding to the probe.

MATERIALS AND METHODSStrains. The MALlp, MALlg, and malo tester

strains were constructed from homothallic strainskindly provided by G. Naumov, Institute for theGenetics of Microorganisms, Moscow. The MAL6strain, CB11 (a MAL6 adel), was obtained from thecollection of A. M. A. ten Berge and is identical to thatused by Federoff et al. (3). However, although ourCB11 strain was derived from the same diploid(NCYC74), it is not identical to the CB11 (otMAL6Iys2) strain characterized by Naumov (7). A MALIstandard strain, 4059, was provided by R. Haekel. Thestandard MAL2 (1453-3B), MAL3 (1412-4D), andMAL4 (1403-7A) strains were obtained from theBerkeley Yeast Stock Center collection, Berkeley,Calif. A large number of tester strains were construct-ed in the course of this study. Key strains will beprovided to the Berkeley Yeast Stock Center. PlasmidYIP5 was kindly supplied by Tom Donahue.

Genetic analysis. Standard techniques of mating,sporulation, dissection, and tetrad analysis were used(4). Maltose fermentation was determined in 5-mlDurham tubes of 1% yeast extract-1% peptone-1%maltose. Fermentation was indicated by the produc-tion of gas 1 to 3 days after inoculation.DNA preparation. Two procedures were used for

preparing total genomic DNA from S. carlsbergensis.One was a modification of the method of Cryer et al.(1), and the other, for preparations from small culturevolumes, was described by Sherman et al. (9).

Restriction digestion and Southern analysis. Restric-tion digestion was done under conditions recommend-ed by the commercial supplier (Bethesda ResearchLaboratories, Bethesda, Md.). Fragments were sepa-rated on 0.8% horizontal agarose gels. The gels weretreated as described by Southern (10), and the nitrocel-lulose filters were hybridized to 32P-labeled probes andexposed to Kodak XAR film at -70°C with an intensi-fying screen.

Nick translation. The probe DNA was labeled bynick translation with [32P]dCTP (Amersham Corp.,Arlington Heights, Ill.) according to procedures de-scribed by Rigby et al. (8).

RESULTSOur goal was to define genetically and physi-

cally the MAL genes present in the MAL6 strainCB11. This strain was used as the DNA sourcein the construction of the clone pMAL9-26,which contains the maltase structural gene. AMAL6 strain related to CB11 was analyzedgenetically by Naumov and found to containmultiple MALg genes unlinked to MAL6 (7).Therefore, as a first step toward the completeanalysis of CB11, we decided to determinewhether any unlinked MALg genes were presentin CB11. Before proceeding with the completegenetic analysis of CB11 we wish to introduce anew notation which we have found extremely

VOL. 3, 1983

798 NEEDLEMAN AND MICHELS

useful for concisely describing the genotype ofMAL strains.

Genetic notation for the MAL loci. The largenumber of loci involved in maltose metabolismalong with the presence of unlinked MALg genesgreatly complicates the notation of genotypes.As will be made clear by the results presentedbelow, the standard MAL strains availablethrough the Yeast Genetics Stock Center, ordescribed in the literature as MAL2 or MAL6,etc., are only partially defined genetically withregard to the MAL loci. Complete genotypicanalysis of these strains requires the use of theMALJp and MALJg strains, which have onlyrecently become available. Using such strains,we demonstrated that the standard maltose-fermenting MAL strains contain, in addition tothe designated dominant MAL locus, additionalunlinked MALg genes. This is shown here (i.e.,in this paper) for a MAL6 strain and also holdstrue for other MAL strains obtained from theYeast Genetics Stock Center (C. Michels andR. B. Needleman, manuscript in preparation).The following notation has been established.

(i) The standard allele tester strains for deter-mining the presence ofMALg or MALp functionare the MALJp and MALJg alleles obtainedfrom G. Naumov. Strains complementing theMALJg strain are said to have MALp function;strains complementing the MALJp strain aresaid to have a MALg function.

(ii) Only active genes are noted. A strainwritten as MALJp indicates that no additionalMALp or MALg genes are present at any locus.

(iii) Strains which are maltose negative andwhich lack genes capable of complementingeither MALJp or MALJg are written as mal°.This convention differs from previous notationswhich have used malo to denote maltose segre-gational nonfermenters. As shown elsewhere(Michels and Needleman, in preparation), suchsegregational nonfermenters often containMALg genes.

(iv) When the genotype of a maltose-ferment-ing strain is written, e.g., as MAL6 (malo), theparentheses around (malo) are used to indicatethat there are no genes unlinked to MAL6 thatcan complement either MALJp or MALJg. Suchstrains were constructed by repeatedly back-crossing the MAL strains obtained from theYeast Genetics Stock Center to maP' strainsuntil no segregating MALg loci could be detect-ed.

Genetic analysis of a standard MAL6 strain,CBll. Strain CB11, a standard MAL6 strain,was crossed to strain 6-2A, a maltose-nonfer-menting malo strain. In nine tetrads resultingfrom this cross (cross W208), maltose fermenta-tion segregated 2:2. When the two negativesegregants in each tetrad were crossed to a

MALJp strain, 13 segregants complemented theMALJp strain to give maltose fermentation,indicating the presence ofMALg genes unlinkedto MAL6. The segregation of MALg in thesenonfermenters gave five tetrads with 2 MALg:OmaP', three tetrads with 1 MALg:1 malo and onetetrad with 0 MALg:O mar', suggesting that morethan one MALg gene was involved. This resultwas confirmed by detailed genetic analysis aspresented below.

Segregant W208-2B of diploid W208. StrainW208-2B, a nonfermenter, was crossed to themar' strain W106-1B (diploid W243). In sixcomplete tetrads, MALg segregated 2:2, indicat-ing the presence of one MALg gene. When theMALg segregant from cross W243 (W243-2B)was crossed to a MALl (mar') strain, W203-3B,all 31 maltose-nonfermenting segregants result-ing from this cross complemented MALJp. Thusthe MALg gene present in W208-2B is linked toMALl and is therefore MALJg.

Segregant W208-2C of diploid W208. StrainW208-2C, a maltose fermenter, was crossed tomar' strain W240-1C. Of 14 maltose-nonfer-menting segregants, all were mar'. W208-2Ctherefore lacks unlinked MALg genes and istherefore MAL6 (malr).

Segregant W208-2A of diploid W208. StrainW208-2A, a maltose nonfermenter, was crossedto the mar' strain 3-2B (W242). When dissected,it yielded eight asci which showed a 4 MALg:Omar' or a 3 MALg:1 mar' segregation pattern,clearly indicating the presence of more than oneMALg gene. Tetrad 3 of diploid W242, whichgave four MALg segregants, was subjected tofurther analysis by mating each segregant to amalo strain. The results (detailed below) indicat-ed that each segregant of tetrad 3 contained onlyone MALg gene. When W242-3B was crossed toa MALI (mar') strain, W203-7B, 23 of the 23maltose-nonfermenting spores were MALg.Thus W242-3B has the MALJg gene. A straincontaining the MALg gene derived from W242-3D was crossed with a MAL3 (mar') strain, W48-2C, and four-spored and three-spored tetradswere analyzed. Maltose fermentation segregated2:2 (in four-spored tetrads), and all 38 maltose-nonfermenting segregants were MALg. W242-3D, thus, carries the MAL3g gene. The genotypeof W208-2A is, therefore, MALJg MAL3g.

In summary the genotypes are: 208-2A,MALJg MAL3g; 208-2B, MALJg; and 208-2C,MAL6 (mar'). CB11 therefore has the genotypeMAL6 MALJg MAL3g.

Southern analysis of strain CB11. Figure 1shows the restriction map of the major HindIIIfragment of the pMAL9-26 S. carlsbergensisDNA insert. This map differs slightly from thatpreviously published (3) in the location of asmall 350-base pair EcoRI fragment. The

MOL. CELL. BIOL.

MAL GENE FAMILY OF S. CARLSBERGENSIS 799

PvPs

H3 R Hp S A HP R

I I 11 .,I

lkbl

PsB Hz Hp HP BR R A

V

D-1I

FIG. 1. Restriction endonuclease map of the major HindlIl fragment from plasmid pMAL9-26. The Dl probewas constructed by subcloning the EcoRI fragment containing Dl into pBR325 and then deleting the region fromthe BglII side of the insert to the BamHI side of the vector. A, AvaI; B, BglII; H2; HincII; H3, HindIII; Hp,HpaI; Ps, PstI; Pv, PvuII; R, EcoRI; S, Sall.

HindIII fragment was subcloned into the non-replicating vector YIP5 (5) to form plasmid Y6.Plasmid Y6 behaves as a replicating plasmid inS. carlsbergensis presumably owing to the pres-ence of a chromosomal replication origin on theHindIII fragment. When introduced into either amalo or a MALg strain, the strain remainedunable to ferment maltose. However, whentransformed into a MALJp strain, the plasmidcomplemented to give maltose fermentation, in-dicating that MALg-complementing activity waspresent on the cloned HindIll fragment.

Since this cloned fragment has MALg activityand multiple MALg genes are present in CB11, itwas of interest to determine whether the frag-ment cloned was unique in the genome and, ifnot, whether it could be used as a probe todistinguish physically the multiple MALg genespresent in CB11 by using Southern analysis.Total DNA was prepared from CB11, from 6-2A(the marP parent), and from four tetrads resultingfrom the diploid W208. Southern analysis ofthese strains after digestion with HindIlI isshown in Fig. 2. The probe used, Dl, is aninternal EcoRI-BgIII subclone of the 7.3-kbHindIII fragment and is shown in Fig. 1.

Initial observation of the results showed thatCB11 had at least two HindlIl genomic frag-ments with significant homology to the probeand that the malo parent did have homology tothe probe despite its lack of all MALp- andMALg-complementing activity. No commonbands were present in both parents. CB11 had aband at approximately 10.7 kb (the A band) anda very intense band at 7.3 kb (the B band). Strain6-2A had one band at 7.0 kb (the C band). Giventhe existence of these restriction polymor-phisms, it becomes possible to correlate maltosefermentation and function with the presence ofparticular HindIll fragments. Bands A and Ccan be seen to segregate 2:2 in the selectedtetrads shown in Fig. 2. Band B represents adoublet, which we will refer to as bands B andB'.

Figure 2 also indicates which segregants weremaltose fermenters, and therefore contained theMAL6 locus. Based on these four tetrads andthree others (data not shown) we can correlatemaltose fermentation with the presence of one ofthe 7.3-kb HindlIl bands (B or B'). Segregantscontaining only band C, e.g., W208-1A andW2084D, are malo like their 6-2A parent. Figure

< - Tetrad I Tetrad 2 Tetrod 4 Tetrad 7N _I__I

m I I' Ii I~

w o A B C D A B C D A B C D A B C D

10.7

7.3 -a m- -_-__ _- I=

^

7.0 '"- '

maltose ++- - I0++t-0fermentationl 0 + ° 0 + + ° ° + + + ° + ° ° + + °

MALg genotype 0 g 0 0 9 9 9 o g 9FIG. 2. Southern analysis of segregants from the diploid W208 (CB11 x 6-2A). CB11 is a standard MAL6

strain. Strain 6-2A is a mal° strain (see text). Total DNA preparation from the two parent strains and 16segregants were digested with HindIll. The fragments were separated on an 0.8% agarose gel and analyzed bythe method of Southern (10). The nitrocellulose filters were probed by using a plasmid containing region Dl (seeFig. 1).

HzHP Pv Hs

VOL. 3, 1983

800 NEEDLEMAN AND MICHELS

2 also shows the results of MALIp comple-mentation tests done for each maltose-nonfer-menting segregant, indicating the presence orabsence of MALg genes. All maltose-nonfer-menting segregants that contained MALg genesalso contained a 7.3-kb band (B or B'), band A,or both. No segregant entirely lacked homologyto the probe. This can be explained by theobservation that bands A and C segregated inrepulsion. Therefore they are alleles or closelylinked. The segregation patterns of A and the B-B' doublet in the seven tetrads analyzed pro-vides no indication of linkage. A possible hy-pothesis, which we prove below, is that band Brepresents the MAL6 locus, and that bands Aand B' represent the unlinked MALg genes.The genotype of tetrad 2 (Fig. 2) was detailed

above. Segregant W208-2B, shown to carry onlythe MALJg locus, can be seen in Fig. 2 tocontain only band A. Southern analysis of segre-gants from the diploid W243 (W208-2B x malo)shows that band A segregated with the MALJggene in all six tetrads tested (data not shown).Thus, band A is linked to MALIg.

Segregant W208-2C carries only the MAL6locus and lacks all unlinked MALg genes. FromFig. 2, W208-2C is seen to contain a 7.3-kbfragment and the 7.0-kb fragment seen in allmalo strains. Thus, one of the 7.3-kb fragmentsis associated with the MAL6 locus.

Segregant W208-2A was genetically shown tocarry both the MALJg and MAL3g loci. In Fig.2, this segregant can be seen to contain at leasttwo Hindlll fragments, band A and a 7.3-kbband, homologous to the Dl probe. SegregantW208-2A was mated to a malo strain to givediploid W242. Southern analysis of the segre-gants from tetrads 3, 5, and 7 of diploid W242 isshown in Fig. 3. Strain W242-2A can be seen tocontain only two HindIll fragments homologousto the Dl probe: band A and a 7.3-kb band.Tetrad 3 shows a nonparental ditype segregationpattern for these bands. Consistent with thegenetic results presented above, strain W242-3B(containing MALJg) has band A, and strainW242-3D (containing MAL3g) has a 7.3-kb bandand band C (the malo-associated band).

Identification of the MAL genes correspondingto the cloned 7.3-kb HindIII fragment. Both theHindIII fragments associated with MAL6 andMAL3g are 7.3 kb, a size identical to that of thecloned HindIII insert of plasmid pMAL9-26. Toidentify whether MAL6 or MAL3g is present inplasmid pMAL9-26 we compared the EcoRIdigestion patterns of both 7.3-kb (i.e., bands Band B') genomic fragments with that of plasmidpMAL9-26 (Fig. 4). Tetrad 7 of diploid W208(Fig. 2) shows the two 7.3-kb fragments in anonparental ditype segregation pattern. Segre-gant 7B ferments maltose and thus carries the

erad 3 etr ;- -e

FIG. 3. Southern analysis of segregants from thediploid W242 (W208-2A x 3-2B). W208-2A is a malt-ase-nonfermenting segregant of diploid W208 (see Fig.2). Strain 3-2B is mal0. Total DNA preparations fromthe parent strains and segregants from three repre-sentative tetrads were digested with HindIll and sub-jected to Southern analysis, as described for Fig. 2.The filters were probed with the Dl-containing plas-mid.

MAL6 locus; segregant 7A is a nonfermenterand carries the MAL3g gene. Both contain theMALJg band A. A strain containing the samefragment as plasmid pMAL9-26 should afterdigestion release a 2.55-kb fragment spanningthe Dl region used as the probe in these experi-ments. EcoRI-Hindlll digests of strains W208-7A and W208-7B, probed with Dl, gave tworegions of homology. The upper fragment, alsofound in W208-4B, clearly resulted from band A.This lower band differed in these two strains,and only W208-7B could be seen to release theexpected 2.55-kb fragment D. Therefore, thegenomic fragment cloned in plasmid pMAL9-26is, in fact, the MAL6 locus.

DISCUSSION

The maltase system is an attractive model forthe study of gene regulation in Saccharomycesspp. Maltase and maltose permease are highlyinducible, and both are subject to glucose re-pression. In addition, mutants with altered regu-lation have been isolated. In spite of thesefeatures, the genetic analysis of maltose regula-tion has encountered one fundamental obstacle:neither the location of the maltase structuralgene nor the number of maltase structural genesis known.There are at least two possible models for the

location of the maltase structural gene. In thefirst, the gene is linked to each dominant MALlocus; in the second, it is unlinked. Both modelsare consistent with the genetic data. Maltosefermentation, while requiring the presence of atleast three genes (the regulatory gene, the malt-ose structural gene, and the gene for maltosepermease), segregates as a single gene in crosses

MOL. CELL. BIOL.

MAL GENE FAMILY OF S. CARLSBERGENSIS

m < m

a) co m00 O 0C%J C%J

6.2- _-_

47 _

2.55- _

FIG. 4. Comparison of the EcoRl digestion patternof the two 7.3-kb HindlIl fragments present in CB11(the MAL6 parent strain). DNA preparations fromsegregants W2084B, W208-7A, and W208-7B (seeFig. 2) were digested with HindIll and EcoRI andsubjected to Southern analysis as described for Fig. 2).Dl was used as the probe.

with nonfermenting strains. This implies thateither a maltase structural gene is linked to eachdominant MAL locus or that an active structuralgene is common to both the maltose-fermentingand the maltose-nonfermenting parents. Thislatter explanation is more than just a theoreticalpossibility; all maltose nonfermenters of theMALg genotype examined have basal levels ofauthentic maltase and must therefore have anactive maltase structural gene (Needleman andMichels, unpublished data).The possibility that more than one maltase

structural gene is present in a given MAL strainis suggested by two lines of evidence. The first isthat no mutants having mutations in the maltasestructural gene have been isolated despite exten-sive searches. The second is derived from thefollowing experiment. We crossed a single malt-ose-negative strain to both a MAL6 and a MAL2strain. Eight maltose-negative spores from eachcross were tested for maltase activity; all hadbasal levels of authentic maltase. This suggeststhat the maltase structural gene present in boththe MAL6 and the MAL2 strain is homozygousto the structural gene present in the maltose-nonfermenting strain. Clearly, either there aretwo structural genes in the maltose nonfer-menter allelic to both MAL2 and MAL6, or elseboth the MAL2 and the MAL6 strains have asingle maltase structural gene which is unlinkedto either MAL6 or MAL2 and is identical to thestructural gene present in the nonfermentingstrain.

If we wish to account for the inability toisolate mutants in the maltase structural gene bypostulating that in a MAL6 strain (like CB11)there are, in addition to the MAL6-linked malt-ose structural gene, other unlinked maltasestructural genes, then certain constraints areplaced on the nature of these genes. These genesmust be under the control of the MAL6 regula-tory protein and must be quite similar in se-quence, since several physical studies havefailed, with one exception, to distinguish be-tween the enzymes produced in strains carryinga single MAL gene (2). To investigate the loca-tion and number of the maltase structural genes,we have conducted genetic and physical analy-ses of strain CB11, a strain used extensively inthe analysis of maltose regulation (A. M. A. tenBerge, Ph.D. Thesis, Rijkuniversiteit te Utrecht,Utrecht, The Netherlands, 1974). The geneticanalysis utilizes MALp and MALg tester strainsrecently made available by G. Naumov; thephysical analysis relies on the recent isolation ofa maltase structural gene from CB11 (3).The genetic analysis has established the com-

plete genotype of CB11. CB11 has three genesinvolved in maltose metabolism, namely,MAL6, MALJg, and MAL3g. Strains containingthe MAL6 gene alone are able to ferment malt-ose, whereas those carrying the MALJg orMAL3g genes alone are unable to ferment malt-ose unless they are supplied with a complement-ing MALp gene. In the absence of a complemen-ing MALp gene, such MALg genes are cryptic.The physical analysis used a DNA fragment of

CB11 that had been cloned onto a replicating S.carlsbergensis plasmid. This fragment had beenshown to contain the maltose structural gene (3).The physical analysis had demonstrated the ex-istence of three genetically unlinked genomicHindlIl fragments from CB11 that show signifi-cant homology to the S. carlsbergensis DNAcontained on plasmid pMAL9-26 and homologyin particular to a subclone, Dl. Each of thesefragments segregates with one of the MAL genesdefined above: a 10.7-kb fragment with MALJg;a 7.3-kb fragment with MAL3g; and a 7.3-kbfragment with MAL6. Recently we have shownthat the EcoRI fragment from which probe Dlwas derived hybridizes to two maltose-induciblemRNAs. Dl therefore contains sequenceslinked to the maltase structural gene.

Since the Dl probe hybridized to three dis-tinct, unlinked fragments of CB11 DNA, thepossibility remained that the clone pMAL9-26contained DNA from a structural gene unlinkedto MAL6. The observed transformation of amaltose-negative strain into a maltose-positivestrain is by itself inadequate evidence for thecloning of MAL6. To resolve this difficulty,Southern analysis of EcoRI digests of strains

801VOL. 3, 1983

802 NEEDLEMAN AND MICHELS

containing MAL3g and MAL6 were compared;these demonstrated that the cloned fragment inplasmid pMAL9-26 is in fact the MAL6 locus.These results lead us to propose the following

model for the organization of the MAL genes.Each MAL locus is a gene complex consisting ofat least two genes, a maltase regulatory gene anda maltase structural gene. Preliminary evidencesuggests that another gene product under malt-ose control (presumably the permease) is alsopresent. We postulate that strains carrying asingle MAL gene frequently have, in addition tothe structural gene linked to the dominant MALgene, additional cryptic MAL genes. Thesecryptic MAL genes (e.g., MALg) lack a regula-tory gene but have structural genes which arehighly homologous to those present at the domi-nant MAL locus, and these genes can be turnedon by the regulatory protein linked to the domi-nant MAL allele.

ACKNOWLEDGMENTS

This work was supported by National Science Foundationgrant PCM7911780 to Richard Needleman and by PublicHealth Service grant GM28216 from the National Institutes ofHealth to Corinne Michels.We thank Jim Broach for careful reading of the manuscript

and for several helpful suggestions, and Robert Dubin forexpert technical assistance.

LITERATURE CITED

1. Creyer, D. R., R. Eccleshall, and J. Marmur. 1975.Isolation of yeast DNA. Methods Cell Biol. 7:39-44.

2. Eaton, N. R., and F. K. Zimmermann. 1976. Thermalinactivation of maltase and its application to temperature-sensitive mutants of yeast. Mol. Gen. Genet. 148:199-204.

3. Federoff, H., J. D. Cohen, T. R. Eccleshall, R. B. Needle-man, B. A. Buchferer, J. Giacalone, and J. Marmur. 1982.Isolation of a maltose structural gene from Saccharomy-ces carlsbergensis. J. Bacteriol. 149:1064-1070.

MOL. CELL. BIOL.

4. Mortimer, R. K., and D. C. Hawthorne. 1975. Geneticmapping in yeast. Methods Cell Biol. 6:221-233.

5. Naumov, G. I. 1969. Comparative genetics of yeast. I.Complementation of maltose utilizing genes in maltose-negative species of Saccharomyces. Genetika 5:142-149.(In Russian)

6. Naumov, G. I. 1970. Comparative genetics of yeast. IV.Identification of maltose complementing factors in Sac-charomyces. Genetika 6:121-124. (In Russian)

7. Naumov, G. I. 1976. Comparative genetics of yeast. XVI.Maltose fermenting genes in Saccharomyces carlsbergen-sis strain NCYC74. Genetika 12:87-100. (In Russian)

8. Rigby, P. W. J., M. Dieckmann, C. Rhodes, and P. Berg.1977. Labelling deoxyribonucleic acid to high specificactivity in vitro by nick translation with DNA polymeraseI. J. Mol. Biol. 113:237-251.

9. Sherman, F., G. Fink, and J. Hicks. 1981. Methods inyeast genetics. Cold Spring Harbor Laboratory, ColdSpring Harbor, N.Y.

10. Southern, E. M. 1975. Detection of specific sequencesamong DNA fragments separated by gel electrophoresis.J. Mol. Biol. 98:503-517.

11. Struhl, K., D. T. Stinchcomb, S. Scherer, and R. W. Davis.1979. High-frequency transformation of yeast: autono-mous replication of hybrid DNA molecules. Proc. Natl.Acad. Sci. U.S.A. 76:1035-1039.

12. ten Berge, A. M. A., G. Zoutewelle, and K. W. van de Poll.1973. Regulation of maltose fermentation in Saccharomy-ces carlsbergensis. I. The function of the gene MAL6 asrecognized by mal6 mutants. Mol. Gen. Genet. 123:233-246.

13. Winge, O., and C. Roberts. 1948. Inheritance of enzymat-ic characters in yeasts, and the phenomenon of long-termadaptation. C. R. Trav. Lab. Carlsberg 24:263-315.

14. Winge, O., and C. Roberts. 1950. The polymeric genes formaltose fermentation in yeasts and their mutability. C. R.Trav. Lab. Carlsberg 25:36-83.

15. Winge, O., and C. Roberts. 1950. Identification of thegene for mraltose fermentation in Saccharomyces italicus.Nature (London) 166:1114.

16. Winge, O., and C. Roberts. 1955. Identification of themaltase genes in some American haploid and Europeandiploid yeasts. C. R. Trav. Lab. Carlsberg 25:331-340.

17. Zimmerman, F. K., and N. R. Eaton. 1974. Genetics ofinduction and catabolite repression of maltase synthesis inSaccharomyces cerevisiae. Mol. Gen. Genet. 134:261-272.