Molecular analysis of mouse alcohol dehy drogenase: nucleotide sequence of the Adh-1 gene and genetic mapping of a related nucleotide sequence to chromosome 3

171

Gene. 59(1987)171-182 Elsevier GEN 02164

Molecular analysis of mouse alcohol dehydrogenase: nucleotide sequence of the Adh-1 gene and genetic mapping of a related nucleotide sequence to chromosome 3 (Recombinant boundary;

DNA;

restriction

glucocorticoid fragment

response

element;

phage A vector;

transcription

start point;

intron/exon

length polymorphism)

Jeffrey D. Ceci, Yao-Wu Zheng and Michael R. Felder Department of Biology, University of South Carolina, Columbia, SC 29208 (U.S.A.) Received Revised

2Y May 1987 7 August

Accepted

1987

Y August

1987

SUMMARY

The mouse has three genes (A&) encoding alcohol dehydrogenase (ADH) enzymes of different tissue specificity and catalytic properties. Identified regulatory loci are known to affect the expression of A&Z and A&-3, which are closely linked on chromosome 3. The A&z-l gene product is expressed predominantly in liver, and its mRNA product is androgen-inducible in kidney. In this study, genomic clones of A&-I were obtained from a Balb/cJ DNA library. The nucleotide sequences of all exons, intron/exon boundaries and 5’- and 3’-flanking regions were obtained. The gene spans nearly 13 kb and is divided into nine exons and eight introns. The transcription start point of this gene was determined by Sl nuclease mapping studies and presumptive regulatory regions in the 5’ -flanking regions were identified, including a TATA box and a glucocorticoid-responsive element. A restriction fragment length polymorphism in the A&-l gene was identified among inbred strains and mapped at the [A&-I, A&-3] complex on chromosome 3. An additional ‘A&-like’ sequence in the genome was also mapped to chromosome 3 approx. 9 centiMorgans from A&-l.

INTRODUCTION

Alcohol : NAD + oxidoreductase (EC 1.1.1.1; ADH) catalyzes the first step in the major pathway of ethanol metabolism in the mammalian liver. Human and horse liver ADHs exist as a heteroCorrespondence to: Dr. M.R. Felder, University

of South

Carolina,

Department

Columbia,

of Biology,

SC 29208

(U.S.A.)

Tel. (803)777-5135. Abbreviations:

Adh, gent /I-subunit

encoding

037X-I 11L)‘X7~$1)3.500

tary to RNA; CM, centiMorgan; element;

kb, kilobase

nt, nucleotide(s);

aa, amino acid(s); ADH, alcohol dehydrogenase;

of ADH;

geneous group of isozymes whose structural, catalytic, developmental and evolutionary properties have been widely investigated (Eklund et al., 1976; Wagner et al., 1983; von Bahr-Lindstrdm et al., 1986). Based upon structural and functional features, human ADH can be grouped into three

ADH;

pdh,

bp, base pair(s);

human cDNA,

gene

encoding

the

DNA complemen-

Pollk,

Klenow

RFLP,

restriction

nant inbred;

fragment

fragment phosphate,

messenger

response RNA;

1,4-piperazine-diethanesulfonic of E.

coli

DNA

length polymorphism;

SDS, sodium dodecyl

0.01 M sodium

1987 Elsevier Science Publishers B.V. (Biomedical Diwsion)

Pipes, (large)

GRE, glucocorticoid

or 1000 bp; mRNA,

sulfate;

1 mM EDTA,

acid;

polymerase

I;

RI, recombi-

SSPE, 0.15 M NaCl, pH 7.4; u, units.

172

classes (Strydom and Vallee, 1982) with three genes encoding three different class I polypeptides (Smith et al., 197 I), and one gene each encoding and Magnes,

class If (Li

1975) and class III (Pares and Vallee,

198 1) polypeptides. The mouse Adh-1, Adh-3 different properties respectively

In addition,

has three specificity

ADH

genes,

encoding and

designated

enzymes having

an Adh-1 RFLP

is reported

and found to map at the [Adh-1, Adh-31 complex on chromosome 3. Other ADH or ‘Adh-like’ sequences not found in A&-I and one sequence

and Adh-2,

tissue

sequences.

3 at a distance

are identified

in genomic

DNA,

is found to map to chromosome

from Ad&I.

with

catalytic

similar to class I, class II and class III, (Algar et al., 1983). The major

activity in mouse liver is encoded

ADH

MATERIALS

AND METHODS

by the Adh-I gene

located on chromosome 3. This gene is closely linked to Adh-3 which specifies an enzyme of divergent

(a) Materials

substrate specificity expressed in lung, stomach and reproductive tissues (Holmes et al., 198 1a). Several features of the mouse ADH system make it attractive for studying gene expression. Variations in liver ADH activity among inbred strains are due to allelic variation at a single locus designated Adh-I-t (Balak et al., 1982). This locus exerts its effect by controlling the rate of ADH synthesis in liver. Because Adh-I-t is tissue- and developmental stage-specific, it is classified as a temporal gene. In addition, a closely Iinked, &-acting, temporal locus controls the tissuespecific expression of the ADH-3 isozyme in the

Restriction endonucleases were purchased from New England Biolabs or Bethesda Research Laboratory and used according to the suppliers’ recommendations. DNA polymerase was from Bethesda Research Laboratory. PolIk was from New England Nuclear. The M 13 primer, deoxy- and dideoxynucleoside triphosphates were obtained from New England Nuclear or Pharmacia. The M 13 phage vectors, polynucleotide kinase, and Sl nuclease were from Pharmacia. The pGem3 cloning vector was from Promega Biotec. Three synthetic oligodeoxynucleotides were prepared at the University of South Carolina Oligonucleotide Core Facility. High-M, DNAs were obtained from Dr. Ben Taylor at the Jackson Laboratory. [ a-32P]dCTP (3000 Ci/mmol) and [~r-‘~sldATP (500 Ci/mmol) were from New England Nuclear. [ Y-~‘P]ATP was from ICN. Nitrocellulose and Nytran membranes were from Schleicher & Schuell. Inbred mice were purchased from The Jackson Laboratory and bred at the

mouse (Holmes et al., 1981b). All inbred strains expre.js this gene product in stomach, lung and kidney, but allelic variation at this Adh-3t locus results in certain strains failing to express this enzyme in reproductive tissues. No variation has been described in the expression of Adh-2; however, unlike Adh-1 and Adh-3 the protein product of this gene is widely distributed in mouse tissues (Holmes, 1978). Because the Adh genes are expressed in a tissueand developmental stage-specific manner, subject to hormonal control, and modulated by identified regulatory loci, this system should yield useful information at the molecular level concerning gene expression. A partial cDNA clone for the Adh-I gene was sequenced and used to demonstrate that the androgen induction of ADH-I activity in kidney tissue is accompanied by a corresponding increase in Adh-l mRNA (Ceci et al., 1986). The complete amino acid sequence of mouse ADH-1 protein has also been determined by sequencing a full-length cDN.4 clone (Edenberg et al., 1985). In this paper, we report the structure of the Adh-I gene, its transcription start point and presumptive regulatory

University

of South Carolina.

(b) Isolation and subcloning of genomic clones Two

bacteriophage

ACharon4A

genomic

libra-

ries, generously provided by Leroy Hood (Davis et al., 1980), were screened. The Adh-I cDNA clone pADHm9 (Ceci et al., 1986), which contains 458 nt of coding sequence and 170 nt of 3’-untranslated sequence, was used as a hybridization probe. pADHm9 was digested with WI, the insert was separated, purified (Maxam and Gilbert, 1980) and labeled with [ r-‘2P]dCTP by nick translation (Rigby et al., 1977). Screening of the library was performed (Benton and Davis, 1977) at an initial density of 50 000 plaques/ 150 cm’ plate. Hybridizations and

173

washings

were as previously

described

(Ceci et al.,

(d) Sl nuclease analysis

1986). Phage DNA from purified plaques was isolated

(Maniatis

et al.,

1982)

and

exon-containing

The

subclone

pADHgl8-1

contains

a 5.2-kb

regions were identified by digesting each lADH clone with EcoRI, HindIII, BumHI, and &I, and performing blot hybridizations (Southern, 1975)

EcoRI fragment from the 5’-end of the A&-l gene. A 96-bp &I-PvuII fragment containing a portion of

using

(Maxam

probes.

pADHm9

and pADHm16

pADHm16

in the 5’ direction as previously cDNA

extends

than pADHm9

described

as hybridization

approx.

400 bp farther

and was isolated

(Ceci et al., 1986). Since our

clone was not a full-length

thetic oligodeoxynucleotides the published full-length

clone, three syn-

were made based upon Ad/z-l cDNA sequence

(Edenberg et al., 1985). Two oligodeoxynucleotides were used as primers for sequencing and to identify exons 1 and 2 (see Fig. 1). The third encoded aa 75-80 and was used to locate exon 3. Each oligodeoxynucleotide was labeled with [ Y-~~P]ATP (Maxam and Gilbert, 1980) and used to probe EcoRI-digested /z clones. Hybridization was done overnight at 55°C in 6 x SSPE (1 x SSPE contains 0.15 M NaCl, 0.01 M sodium phosphate, 1 mM EDTA, pH 7.4) 5 x Denhardt’s medium (Denhardt, 1966) 1% SDS and 50 pg/ml denatured salmon sperm DNA. The filters were washed extensively in 6 x SSPE at room temperature and for 10 min at 55 ‘C. All of the exons were located on four EcoRI fragments of 3.8, 2.3, 1.1 and 5.2 kb. Purified AADH22 DNA was digested with EcoRI, separated by agarose gel electrophoresis, and the individual EcoRI fragments were recovered by blotting onto DEAE membranes. Seven fragments were individually ligated (Maniatis et al., 1982) with EcoRIdigested pGem3 and used to transform competent DH-1 cells (Dagert and Ehrlich, 1979). To obtain the 5’ end of the gene, a 5.2-kb EcoRI fragment from iADH18 was subcloned into pGem3. (c) Restriction

the first exon and 5’-flanking and

radiolabeled separated

Gilbert,

1980). The

at its 5’ termini, on a nucleotide

coding strand

region

was isolated fragment

was

and the strands

were

sequencing

gel. The non-

was labeled with 32P 20 nt upstream

from the AUG start codon. The strand was annealed to 30 pg of total RNA in 25 ~1 of a buffer containing 40 mM Pipes, pH 6.8, 0.4 M NaCI, 1 mM EDTA, and 0.4 pg/pl tRNA. This mixture was heated at 90°C for 45 s and then incubated at 60°C for 12 h. After annealing, 250 ~1 of a cold buffer containing 30 mM Na. acetate, pH 4.75, 0.25 M NaCl, 1 mM ZnCl, and 5% glycerol was added. lo-100 u of Sl nuclease were added and the mixture was incubated at 30’ C for 30 min. Following S 1 nuclease digestion (Berk and Sharp, 1977) 25 ~1 of 0.6 M Tris. HCl, pH 8, 2.2% SDS, 55 mM EDTA, and 0.2 pg/pl tRNA were added. After phenol-chloroform extraction, ethanol precipitation, and a 70% ethanol wash, the samples were analyzed on an 8 y0 sequencing gel. The end-labeled, noncoding strand was sequenced (Maxam and Gilbert, 1977). RNA was isolated as previously described (Chirgwin et al., 1979). (e) Computer analysis of nucleotide sequences Nucleotide sequences were analyzed on a VAX computer of the University of Wisconsin Genetics Computer Group and compared to Genebank and EMBL sequence banks (Devereux et al., 1984). (f) Genetic mapping of DNA polymorphisms recombinant inbred (RI) mice

using

mapping

All of the subclones were singly and doubly digested with EcoRI, HindIII, PstI and BarnHI. Each subclone was used to probe AADH22 DNA cut with EcoRI, HindIII, BumHI and PstI to orient the fragments in relationship to each other. To further facilitate the mapping, a 3.8-kb Hind111 fragment was subcloned into pGem3. Single and double digestions and Southern blot hybridizations were also performed with this subclone.

A computer printout of phenotypes of known genetic markers in BXD RI lines was obtained from Dr. Ben Taylor of the Jackson Laboratory. Three markers on chromosome 3 near the [A&-I, A&-J] complex were used in the calculation of map distances. The cdm gene controls a difference in resistance to cadmium-induced testicular damage between C57BL/6J and DBA/2J mice (Taylor et al., 1973). XP-24 is a DNA polymorphism identified on chromosome 3 using a probe for endogenous xeno-

174

tropic

murine

(Wejman are variant

leukemia

virus-related

sequences

et al., 1984). The progenitor for an enzyme

electrophoretic

mobility

strains

polymorphism

of the stomach

coded by the A&-3 gene (Holmes

also

affecting enzyme

en-

et al., 1981a).

the analysis of LADH18 clones and exon-containing quenced

as indicated

revealed

that the Adh-l

arranged

as nine exons interspersed

The introns nucleotide RESULTS

The characterization

gene spans

analysis

13 kb and is by eight introns.

range in size from 62 bp to 3.55 kb. The sequences

for all exon-containing

DNA,

3’-flanking

The Adh-I gene contains 374 codons plus the initial methioninc codon, thus indicating that human,

gene

and sequence

genomic were se-

(Fig. 1). The sequence

intron/exon boundaries, and 5’- and regions were obtained (Fig. 2).

AND DISCUSSION

(a) Cloning the mouse Adz-2

and AADH22 M 13 clones

analysis

of a

cDNA (pADHm9) encoding the C-terminal 151 aa of the mouse liver ADH-1 protein was reported previously (Ceci et al., 1986). Using the insert in pADHm9 as a hybridization probe, a mouse genomic library prepared from Balb/cJ DNA was screened. Eleven clones were selected and digestion of these clones with EcoRI and subsequent analysis (Southern, 1975) revealed that all the clones were related but extended farther in either the 5 ’ or the 3 ’ direction. Orientation of the EcoRI fragments was accomplished by probing with a Sau3A-PstI subclone of pADHm9 containing about 200 bp of 3’-erd sequence (Ceci et al., 1986), pADHm9, and pADiIml6. Only the 3.8-kb fragment hybridized with the 3’-specific probe, while both the 3.8- and l.l-kb fragments hybridized with pADHm9. pADHm16 contains an additional 5’-end sequence and, when used as probe, 3.8- and l.l-kb fragments

horse, and mouse ADH contain

an equal number

amino

1986).

acids

(Duester

et al.,

The

of

initial

methionine codon is contained within the sequence GGCATGA which is similar to that found at most initiation codons (Kozak, 1981). The AATAAA polyadenylation signal (Proudfoot and Brownlee, 1976) is located 25 bp upstream from the polyadenylation site deduced from the cDNA sequence (Edenberg et al., 1985). Each of the eight introns is bounded by the consensus GTjAG sequences found in all such junctions (Breathnach and Chambon, 1981). The encoded amino acid sequence in the exons of the Adh-l gene perfectly matches the sequence predicted from cDNA sequences (Ceci et al., 1986; Edenberg et al., 1985). When the nucleotide sequences of the cDNAs are compared with the coding

plus an additional 2.3-kb fragment hybridized, suggesting an orientation of the EcoRI fragments as

sequence of the genomic clone, only a single substitution is noted in the codon for valine at aa 73. In the cDNA sequence reported by Edenberg et al. (1985) this valine is specified by GTC whereas GTT is used in the Ad/z-l gene from Balb/cJ mice. Since the

5’-(2.3-, l.l-, and 3.8-kb)-3’. This was confirmed by nucleotide sequence analysis. Two clones, AADH 18 and iADH22, were chosen for further structural analysis since the Southern blots ofEcoR1 digests of those clones suggested that they extended farthest in the 5’ and 3’ regions, respectively, and together likely represented the complete Adh-I gene.

cDNA was prepared from DBAjZJ mouse liver mRNA, this base substitution may represent a strain-specific variation. Two cDNA clones sequenced by Edenberg et al. (1986) showed differences in the 3’-untranslated region of the corresponding mRNA; there is a G at nt 1281 and a C at nt 1334 in clones pZK 105-36 and

(b) Structure of the mouse Adz-1 gene The structure of /ZADH18 and LADH22 was determined by detailed restriction mapping and sequencing to explore the organization of the gene with respect to intromexon structure and the presence of presumptive 5’-regulatory sequences. The detailed restriction map of the Adh-I gene was revealed from

pZK7 whereas T was found at both positions in pZK6-6, corresponding to the sequence of the Balb/cJ Adh-I gene reported here. A cDNA sequence from SWR/J mouse liver RNA also has G and C at these two respective positions (Ceci et al., 1986). The possibility exists that the mouse haploid genome contains and expresses more than a single copy of Adh-1. The overall structure of the mouse Adh-l gene is

175

Ah Alul w+----+

I&mer

Sou34 Sou3A

A Hoe Ill

Saujn

t--t3a

AADH AADH

I

22

18 0

,

I

,

5 t

*

I

1

I

L

I

IO

*

I 5 ,

I

kb

EXON

-H

7

EXON

f-4 Haelll

HoeIll

Al”1

Haelll

Ali

AIUI \

sspi Fig. I. Structure

mapping

and sequencing.

(blackened

boxes) and introns

indicate the sequencing

strategy

used. Restriction

using the dideoxy chain-termination

ct al., 1983). Two synthetic The solid, thick arrow

oligodeoxynucieotides

represents

the direction

f

J

tiPOIl

(open boxes) and the amount

The bars above and below (0. I-kb scale) represent and sequenced

Alul

AlUl

sspi

The open bar above represents

fragments method

bars represent the restriction

of sequence

expanded

determined

the overlapping map and position

were subcloned

lADH18

and 22 clones used in

of translated

of exons l-9. Arrows

into phages M13mp18

et al., 1977) and [a-%]ATP

were used as primers

tipat

portions

at the 5’ and 3’ ends of the gene (hatched

views of the gene and positions (Sanger

0.1 kb

9

a61

>

f

+

of the mouse Ad&l gene. The two solid, horizontal

restriction

EXON

8

in t&so regions

(t8-mer

or M13mp19

boxes).

above and below

(Norrander

as the radiolabeled

of exons

et al., 1983)

nucleotide

(Biggin

A and 1%mer B; seen near the top).

of transcription.

remarkably similar to the human gene encoding the fi subunit of ADH @A&). The amino acid positions encoded by each exon are identical for both genes. There is a general similarity in size of the introns with intron 4 being the smallest in both genes (62 bp for A&-I, 67 bp for pA&). Introns 2,3,6 and 7 are quite simiIar differing by no more than 25 “/a in size. However, intron I in j?Adh is larger than A&-I (2.8 vs. 1.8 kb) as is intron 8 (2.8 vs. 0.4 kb). In contrast, intron 5 in A&-l is larger (3.55 vs. 2.0 kb).

(c) Structure of the 5’-end of the Adh-2 gene A transcription start point was identified by Sl nuclease mapping using a 5’-labeled DNA fragment as a probe (Fig. 3). The most abundant size of DNA fragment protected from digestion in hybrids formed with mouse liver mRNA is 53 nt in length su~esting that the G at nt + 1 (Fig. 2) is a transcription start point. There may be other start points l-5 nt downstream from this one or these may represent degrada-

176

ser TCA

se-r ‘KC

Fig. 2. Nucleotide shown

Ttr ACT

ThT ACA

my GGC

180 Tyr TAT

Gly GGC

cys

*,a

“a,

TGT

GCC

GTG

sequence

ser

*,a

TCT

GCC

“.a, GTC

Lys AA*

“al OTC

Ala GCC

Lys AAG

~AGGATGGACAGTG-3.55

Ph.= ‘ITT

Gly GGC

200 L‘=u CTC

Gly GGA

Gly

vail GTC

Gly GGT

Leu CTG

of the mouse Adh-I

along with 5’- and 3’-flanking

boundaries

are underlined

boxes are underlined

regions.

as is the AATAAA

and a GRE element

GGT

gene.

The complete

The approximate polyadenylation

is labeled.

,st-r TCT

“al GTC

Ilr ATC

sequence

Il.ATT

210 my cys GGC ‘RX

Lys Al.3 AAA

GCA

Ala GCA

c1y GGA

of all exons and the predicted

sizes of introns

190 Thr

m-0

0ly

Ala GCA

*1lJ GCC

Arg AGG

*cc cc* ffic

amino acid sequence

are given, and the GT and AG consensus

signal. Start and stop codons

The site of polyadenylation

“al CT0

kb-TTTCTGGAAATAC~

is indicated

are given. Presumptive by an asterisk.

are

intron

TATA and CAT

177

!z!zrI

P+II

8

m If * _

-60 1

_

-40 t

*I 1

+80

440

Lstort

Fig. 3. Transcription After hybridization

start point ofthe mouse Adh-2 gene. The diagram to RNA and digestion

gel (Maxam

and Gilbert,

S 1 nuclease

assays

1977) next to the products

contained

liver RNA digested

with (d) 50 and (e) 75 u of enzyme, itself was sequenced (not shown).

Position

with S 1 nuclease,

in a separate

tRNA digested experiment

of a sequencing

*

96

nt *probe

*

53

nt

shows the fragment

the protected

fragments

protected fragment

used as a probe in S 1 protection

were analyzed

on an 8% polyacrylamide

ladder made from the sense strand

with (a) 25, (b) 50, and (c) 75 u of enzyme, with (f) 75 u of enzyme,

fragment

fragment

terminated

is indicated

experiments. sequencing

labeled at the PvuII site. The

androgen-induced

and liver RNA digested

to confirm that the 53-nt protected

of the s2P label on both the probe and protected

tion of the larger fragment. The band corresponding to nt -1 may result from incomplete digestion as it decreases in intensity with greater nuclease concentration. Conceivably, the transcription start point at nt + 1 may represent the position of an intron. However, the first consensus intron/exon boundary sequence is at nt + 8, and the cDNA sequence reported previously (Edenberg et al., 1985) is identical with this sequence up to nt + 6. In addition the sequence

codoi

kidney RNA digested

with (g) 75 u of enzyme. The probe at the G nt marked

by a blackened

+ 1 in Fig. 2

circle.

AAAATAT begins 25 bp upstream from the proposed transcription initiation site and closely matches the position and sequence of the TATA box associated with most eukaryotic promoters (Breathnach and Chambon, 198 1). In fact, surrounding this sequence there is a match of 19/23 nt with the human /iAdh sequence suggesting nucleotide sequence conservation surrounding the TATA box. The nucleotide sequence homology in the 5’ flanking region

178

A

consensusGFtE

T

humanfiAdhGFE1

A

A

TCACT-T

-

TTACAATTT

-226

1 human I3Adh GRE II -187 mouse

Adh-1

Fig. 4. GRE consensus

the consensus

in ADH

(Karin

sequences

genes. Two GRE

sequences

in the human

et al., 1984) and these are listed together

/Udh gene (Duester

with a putative

and all three genes or between Adh-I and either the consensus

gene. Two other regions of strong sequence homology exist between mouse Adh-1 and human @dh starting at nt -81 (13/15 nt) and nt -62 (15/16 nt). Interestingly, the full-length cDNA sequence reported previously (Edenberg et al., 1985) has an initiation codon near the 5’-end followed by codons for phenylalanine and arginine and then a stop codon. This sequence would begin at nt + 5 in Fig. 2, but that is where sequence homology between the cDNA sequence and this genomic sequence ends. Possibly this may represent alternate processing at the 5’-end such that an additional promoter region may occur upstream from the one suggested here. Alternatively, if more than one copy of Adh-1 exists in the haploid genome, these may differ at the 5’-end of their transcribed sequences. mapping

r

c

r

1 C

T

-171 -168

available for comparison between mouse Ad/z-l and the human /IADH (Duester et al., 1986) is 69%. The human gene also has two GRE sequences which are closely related to a consensus sequence derived from the metallothionein gene and mouse mammary tumor virus (Karin et al., 1984). There is a 14/17-nt sequence identity starting at nt -184 in mouse Adh-1 compared to the consensus GRE sequence and GREI and GREII found in the human /L4DH gene (Fig. 4). The mouse GRE maintains the region of dyad symmetry found to be shared between ten other GRE sequences (Jantzen et al., 1987). The position of this sequence in the mouse Adh-1 gene(Fig. 2) is similar to the location of GREII in the human /?Adh

(d) Genetic

T c

-184

GRE

sequences

sequence

r

of Ad/z-Z and ‘Ad/z-like’

DNA

polymorphisms

Adh-l gene structure was examined by Southern (1975) analysis of EcoRI digests of DNA from

et al., 1986) are suggested

GRE in the mouse Adh-1 gene. Matches sequence

from a between

or a GRE in the human gene are boxed.

Balb/cJ, A/J, C57BL/6J, AKR/J, SWR/J, DBA/2J and C3HeB/FeJ inbred strains. The analysis of sequences complementary to pADHm16 in genomic DNA and I.ADH18 and /IADH22 revealed the presence of 2.0- and 2.1-kb EcoRI restriction fragments in the genome not found in the Adh-Z structural gene (Fig. 5C). These additional genomic sequences were not revealed with pADHm9 as probe suggesting that only the middle portion of the coding region is complementary to other Adh or ‘Adh-like’ genes. DNA polymorphisms were observed in both the Adh-1 and ‘Adh-like’ sequences among the inbred strains examined. An RFLP exists in the 5’ end of the Adh-Z gene between C57BL/6J and DBA/2J (Fig. 5A) as revealed by analysis of EcoRI digests using the 5.2-kb EcoRI fragment obtained from i,ADH 18 as probe. This fragment is 6.8 kb in C57BL/6J and 5.2 kb in six other inbreds. The DNA polymorphism in the ‘Adh-like’ sequence was seen as a 2. I-kb EcoRI fragment in DNA of C57BL/6J mice and five other inbreds whereas a 7-kb fragment was found in DBA/2J mice (Fig. 5B) when pADHm16 was used as probe. The two RFLPs occurring between C57BL/6J and DBAjZJ mice enabled these nucleotide sequences to be mapped using B x D RI lines produced from these two progenitor strains. In producing RI lines, progenitors are crossed to produce F, and F, progeny. The progeny are then inbred to produce a number of RI lines, each containing various combinations of the progenitor genes. Linked allelic forms of genes found in progenitors will occur together more frequently in RI lines than will unlinked forms, and this is the basis for genetic mapping. All available RI strains derived from C57BL/6J

179

C x ADH 6

DBA

5.2 kb 3.8

Fig. 5. ADH subcloned or lADH18

sequences

and RFLPs

5’-end fragment and 1ADH22

in the mouse genome.

(A) and pADHm16 EcoRI-digested

the RFLP in the ‘Adh-like’ sequence

Sequences

(B,C) were analyzed.

in mouse genomic

Lanes containing

DNAs (1 pg each) are identified.

is shown in B where the arrows

indicate

found in the two strains. In C, note that the 2.1-kb and 2.0-kb EcoRI fragments clones. In all instances, and Gilbert,

DNA digests were fractionated

1984) and probed

as previously

described

on 0.7% agarose

DNA complementary

10 pg ofC57BL/6J

to the 5.2-kb EcoRI

(B6), DBA/ZJ (DBA), Balb/cJ,

The RFLP found in the Adh-Z gene is shown in A, while the position

of the 2.1-kb and 7.0-kb EcoRI

fragments

located in the genomic digests are not found in the lADH

gels, transferred

to Nytran

membranes

(Southern,

1975; Church

(Ceci et al., 1986).

and DBA/2J progenitor strains were scored for their phenotype for each of the two RFLPs (Table I). The strain distribution pattern shows a perfect correlation between the phenotype at the Adh-3 locus and the phenotype of the Adh-I RFLP. However, the strain distribution pattern reveals several recombi-

nant phenotypes (B x D lines 15, 18, 19,22,23,24) between ‘A&-like RFLP and A&-3. Using the equations of Haldane and Waddington (1931) to calculate map distance from the strain distribution pattern in RI lines (Bailey, 1971), the order on chromosome 3 is [Adh-I, Adh-3]6.7 CM - cdm - 6.7

I x0

I

TABLE

order

Strain distribution RFLP

in B

RI line

x D

pattern

of an Adh-1 RFLP and an ‘Adh-like’

recombinant

ADH-3

inbred

is also indicated

distance

since

is 22.8 CM, but the fldh-like’

Phenotype’

CDM

XP-24

B x D,’

mosome

3 (Holmes,

1981a), it is unlikely

‘Adh-like’ sequence identified

RFLP

ADH-1

‘ADH-like’

RFLP

to

XP-24 value is 18 CM. Since Adh-1 and Adh-3 are closely linked in chro-

lines.

RFLP

the cdm to XP-24

The chromosomal

location

here represents

that the Adh-3.

of Adh-2 is unknown

in

the mouse, but the ‘Adh-like’ sequence may represent I

B

B

B

B

B

this gene. If so, these results

2

D

D

D

D

B

Adh-3 and Adh-2 are all found in chromosome

5

B

B

B

B

D

6

B

B

B

B

B

8

D

D

D

D

B

9

B

B

B

B

B

are currently in the process of obtaining

II

D

D

D

D

D

I2

D

D

D

D

B

13

D

D

D

D

B

14

B

B

B

D

B

15

B

B

D

B

D

16

B

B

B

B

B

I8

D

D

B

D

B

I9

D

D

B

B

B

20

D

D

D

D

B

21

D

D

D

D

D

22

B

B

D

D

D

contain the ‘Adh-like’ sequences and the 2.0-kb EcoRI fragment. It is possible that the 2.0-kb fragment could be derived by a portion of the DNA remaining undigested at the EcoRI site in the fifth intron (1 1-kb site, Fig. 2) although ethidium bromidestained gels indicated that the DNA was digested to completion. Additional structural studies on these sequences will be necessary before their identity is known.

23

D

D

B

B

B

24

B

B

D’

B

B

25

D

D

D

D

D

27

B

B

B

B

B

2x

D

D

D

D

D

29

B

B

B

B

B

30

B

B

B

B

D

31

B

B

B

B

B

32

D

D

D

B

B

” The B x D RI lines are produced progenitor

inbred

h Phenotypes

from C57BL/bJ

(B. C57BL/6J;

D, DBA/2J)

Adh-3, cdm and XP-24 were obtained MATERIALS

AND METHODS,

‘ The phenotype

and bothgave

7-kb EcoRI fragment was present,

of the RI lines for

from Dr. Ben Taylor (see

3 and

are a linked multigene family. Alternatively, the identified sequence(s) may represent a pseudogene. We clones which

ACKNOWLEDGEMENTS

We thank Drs. Michael Dewey, Frank Berger, Aubrey Thompson, Robert Lawther, Gregg Duester, Moyra Smith, and Roger Holmes for many helpful discussions. We appreciate the help of Debra Williams and Clint Cook in preparing the manuscript. This work was supported by PHS grants AA06608 and AA055 12.

section f).

of this line was scored

DNA preparations fragment

and DBA/2J

strains.

suggest that Adh-Z,

using two independent

slightly ambiguous

results. The

was clearly visible, and although it was much less prevalent

the 2. I-kb

than m other

lines with the B phenotype.

REFERENCES Algar.

CM - ‘A&-like’ - 18 cM - XP-24. This order is indicated since the calculated distance from A&-Z or A&-S to cdm is 6.7 CM and from ‘Ad/z-like’ RFLP to cdm is also 6.7 CM, whereas the distance from either Adh-I or Adh-3 to the ‘Ad/z-like’ RFLP is 8.8 CM. The discrepancy in the predicted sum of these distances is not unexpected due to the relatively small number of lines available for analysis. This

E.M., Seeley, T.-L. and Holmes,

molecular

properties

of mouse

zymes. Eur. J. Biochem.

R.S.: Purification

alcohol

dehydrogenase

of mouse

1I

strains. Transplantation

(1971) 325-327. Balak. K.J., Keith, R.H. and Felder, M.R.: Genetic regulation

iso-

137 (1983) 139-147.

Bailey, D.W.: Recombination-inbred

mental

and

liver alcohol

and develop-

dehydrogenase.

J.

Biol. Chem. 257 (1982) 15000-15007. Benton,

W.D.

and

Davis,

clones by hybridization (1977) 1x0-182.

R.W.:

Screening

to single plaques

Igt

recombinant

in situ. Science

196

181

Berk, A.J. and Sharp, virus mRNAs digested

P.A.: Sizing and mapping

by gel electrophoresis

hybrids.

(198lb)

T.S. and Hong, G.F.: Buffer gradient

and ‘?!I label as an aid to rapid DNA sequence tion. Proc. Nat]. Acad.

Biochem.

P.: Organization

split genes

coding

O’Malley, alcohol

Annu.

tyrosine

Rev.

G., Hatfield,

M.P. and Felder,

probe

M.R.: Androgen

in mouse

confirmed

G.W., Smith, M.,

kidney.

by nucleotide

induction

Studies

of

with

sequence

a

analysis.

Chirgwin,

J.M., Przybyla,

W.J.: Isolation sources

A.E., MacDonald,

of biologically

enriched

Acad. Dagert,

active

in ribonuclease.

R.J. and Rutter,

ribonucleic

acid from

Biochemistry

18 (1979)

sequencing.

Proc. Nat].

incubation

in calcium

W.: Genomic

Sci. USA 81 (1984) 1991-1995. M. and Ehrlich,

chloride

improves

the competence

of ~s~herichja coli cells.

Weissman,

K., Early, P.W., Livant,

1.L. and Hood,

D.L., Joho, R.,

L.: An immunoglobulin

heavy-

chain gene is formed by at least two recombinational

events.

Nature

D.: A membrane

complementary

filter technique

DNA. Biochem.

for the detection

of

Res. Commun.

23

Biophys.

J., Haeberli,

of sequence

G.. Smith.

Molecular

programs

V. and

of the human

class

Hatfield,

I alcohol

sequence

G.W.:

dehydro-

of the gene en-

H.J., Zhang,

Ke., Fong,

and sequencing

K., Bosron,

of cDNA

W.F. and Li,

encoding

dehydrogenase.

the com-

Proc. Nat]. Acad.

Sci. USA 82 (1985) 2262-2266. Eklund,

H., Nordstrom.

Ohisson,

B., Zeppezauer,

E., Siiderland,

G.,

B.-O., Tapia, O., Brand&n,

A.: Three-dimensional

liver alcohol dehydrogenase

structure

at 2.4 A resolution.

of horse

J. Mol. Biol.

J.B.S. and Waddington,

Genetics

oxidase

C.H.: Inbreeding

and linkage.

oxidase,

analysis ofalcohol

sorbital

from mouse

tissues.

dehydrogenase,

dehydrogenase Comp.

and

Biochem.

xanthine

Physiol.

61B

R.S., Albanese,

R., Whitehead,

F.D. and Duley, J.A.:

Mouse alcohol dehydrogenase

isozymes:

localized

exhibiting

properties. Holmes,

duplicated

genes

J. Exp. Zool. 215 (1981a)

R.S., Andrews,

lation of alcohol

RI., Krauter, of DNA

and glucocorticoid

hor-

gene. Nature

308

codon

nucleotides

by eukaryotic

in recognition

ribosomes.

Nucl.

of a distinctive

mole-

Acids Res. 9 (1981) 5233-5252. Li, T.-K. and Magnes, cular

L.J.: Identification

form of alcohol

dehydrogenase

Biochem.

Biophys.

in human

livers with

Res. Commun.

63 (3975)

202-208. T., Fritsch,

A Laboratory Spring

E.F. and Sambrook,

Manual,

Harbor,

Cloning.

Laboratory,

Cold

NY, 1982.

A.M. and Gilbert,

Maxam.

J.: Molecular

Cold Spring Harbor W.: A new method

DNA. Proc. Natl. Acad.

for sequencing

Sci. USA 74 (1977) 560-564.

A.M. and Gilbert,

with base-specific Mount,

W.: Sequencing

chemical

cleavages.

end-labeled

Methods

DNA

Enzymol.

65

S.M.: A catalogue

S.J. and Beechey,

dehydrogenase

of splice junction

products

of closely

divergent

kinetic

J., Kempe,

T. and Messing, using

C.V.: Genetic

C, in the mouse:

regu-

develop-

Nucl.

J.: Construction

forms with unique

Biophys.

Res. Commun.

Proudfoot,

kinetic

in eukaryotic

muta-

liver alcohol

dehydro-

characteristics.

Biochem.

98 (1981) 122-130.

N.J. and Brownlee,

quences

of im-

oligonucleotide-directed

X. and Vallee, B.L.: New human

genase

G.G.: 3’ Non-coding

messenger

region se-

RNA. Nature

263 (1976)

21 I-214. Rigby,

P.W.J.,

Labeling

Dieckmann,

M.,

deoxyribonucleic

Rhodes,

C. and

acid to high specific with DNA polymerase

Berg,

P.:

activity

in

I. J. Mol. Biol.

113 (1977) 237-251. Sanger,

F., Nicklen,

S. and Coulson,

chain-terminating

inhibitors.

A.R.: DNA sequencing

Proc. Natl. Acad.

with

Sci. USA 74

(1977) 5463-5467. M., Hopkinson,

changes Southern,

D.A. and

and polymorphism

Harris,

H.: Developmental

in human alcohol dehydrogenase.

34 (1971) 251-271.

E.: Detection

ments separated

of specific sequences

by gel electrophoresis.

among DNA frag-

J. Mol. Biol. 98 (1975)

503-517. D.J. and

alcohol liquid

Vallee,

dehydrogenase chromatographic

B.L.: Characterization isoenzymes peptide

of human

by high-performance

mapping.

Anal.

Biochem.

123 (1982) 422-429. Taylor,

151-157.

sequences.

Acids Res, 10 (1982) 459-472.

Strydom,

(1978) 339-346. Holmes,

role of Banking

Ann. Hum. Genet.

16 (1931) 357-374.

Holmes, KS.: Electrophoretic nldehyde

M.: Possible

Smith,

102 (1976) 27-59. Haldane.

Kozak,

vitro by nick translation

I., Boiwe, T., SSderberg,

C.-I. and Okeson,

of the

(1984) 513-519.

Pares,

plete mouse liver alcohol

of

metallothionein-II,

Gene 26 (1983) 101-106.

J. Biol. Chem. 261 (1986) 2027-2033.

T.-K.: Cloning

induce human

genesis

the /I subunit.

Edenberg,

which cadmium

for the VAX. Nucl. Acids Res.

M., Bilanchone.

analysis

W.,

H., Richards,

M 13 vectors

genase gene family and nucleotide coding

mones

F., Schmid,

G.: Cooperativity

H.M. and Beato, M.: Characterization

proved

0.: A comprehensive

12 (1984) 387-395. Duester.

2

gene. Cell 49 (1987) 29-38.

set

P. and Smithies,

analysis

and

Genet.

located far upstream

A., Holtgreve,

through

Norrander,

(1966) 641-645. Devereux,

(ffdh-3t)

(1980) 499-560.

283 (1980) 733-739.

Denhardt,

sequences

Maxam,

Gene 6 (1979) 23-28.

elements

amino transferase

Karin, M., Haslinger,

Maniatis,

S.D.: Prolonged

Davis, M.M., Calarme.

response

high activity.

5294-5299. Church, G.M. and Gilbert,

R. and Schiitz,

of the AUG initiator

Gene 41 (1986) 217-224.

locus

3. Develop.

U., Gloss. B., Stewart,

M., Mikicek,

P., Westphal,

R., Duester,

dehydrogenase

H.-M., Strahle,

glucocorticoid

and expression

for proteins.

Jantzen,

of the temporal

of Adh-3 on chromosome

89-98.

Boshart,

50 (1981) 349-383.

Ccci, J.D., Lawther,

cDNA

gels

determina-

Sci. USA 80 (1983) 3963-3965.

B. and Chambon,

of eukaryotic

consequences

positioning

Cell 12 (1977) 721-732.

Biggin, M.D., Gibson,

Breathnach,

mental

of early adeno-

of Si endonuclease-

B.A., Heiniger,

resistance

H.J. and Meier, H.: Genetic

to cadmium-induced

testicular

damage

Proc. Sot. Exp. Biol. Med. 143 (1973) 629-633.

analysis

of

in mice.

182

von Bahr-Lindstrdm,

H., Hii@, J.-O., Heden,

Fleetwood,

L., Larsson,

Holmgreen,

A., Hempel,

K.,

Wagner,

dehydrogenase.

by class I isoenzymes.

Biochemistry

B.A., Jenkins,

B.,

Endogenous

H.:

quences

map to chromosome

of human liver

phocyte

antigens.

xenotropic

oxidation

of alcohols

22 (1983) 1857-1863.

Communicated

murine

N.A. and Copeland, leukemia

regions

virus-related

encoding

J. Virol. 50 (1984) 237-247.

25 (1986) 2465-2470.

F.W., Burger, A.R. and Vallee, B.L.: Kinetic properties

of human liver alcohol dehydrogenase:

J.C., Taylor,

R.,

Holmquist,

for the a subunit

Biochemistry

Wejman,

L.-O., Kaiser,

M.,

J., Vallee, B.L. and Jdrnvall,

cDNA and protein structure alcohol

Lake,

by S.T. Case.

N.G.: se-

mouse lym-