Identification of two variant short chain acyl coenzyme A dehydrogenase alleles, each containing a different point mutation in a patient with short chain acyl coenzyme A dehydrogenase deficiency

Identification of two variant short chain

acyl-coenzyme A dehydrogenase alleles, each

containing a different point mutation in a patient

with short chain acyl-coenzyme A

dehydrogenase deficiency.

E Naito, … , Y Indo, K Tanaka

J Clin Invest.

1990;

85(5)

:1575-1582.

https://doi.org/10.1172/JCI114607

.

Two distinct mutant alleles of the precursor (p) short chain acyl-CoA dehydrogenase

(SCAD) gene were identified in a SCAD-deficient patient (YH2065) using the polymerase

chain reaction to amplify cDNA synthesized from total RNA from her fibroblasts. Cells from

this patient had previously been shown to synthesize a labile variant SCAD in contrast to

the normal stability of variant SCADs in two other SCAD-deficient cell lines (Naito, E., Y.

Indo, and K. Tanaka. 1989. J. Clin. Invest. 84:1671-1674). In the present study, both mutant

alleles of YH2065 were found to contain a C----T transition, one at position 136 and the

other at position 319 of the coding region of pSCAD cDNA. Clones of cDNA amplified from

this region showed only one of the C----T transitions, indicating that each mutation was

derived from different pSCAD alleles. Each of these mutations altered a known restriction

endonuclease site, and restriction analysis of additional cDNA clones from amplified mutant

cDNA and Southern blotting of mutant genomic DNA confirmed the presence of two unique

mutant alleles in YH2065, indicating YH2065 is a compound heterozygote. These C----T

transitions result in the substitution of Arg-22 and Arg-83 of the mature SCAD with Trp and

Cys, respectively.

Research Article

Find the latest version:

Identification of Two Variant Short Chain Acyl-Coenzyme

A

Dehydrogenase

Alleles, Each Containing

a

Different Point Mutation

in

a

Patient with Short Chain Acyl-Coenzyme A Dehydrogenase Deficiency

Etsuo Naito, Yasuhiro Indo, and Kay Tanaka

Yale University School ofMedicine, Department ofHuman Genetics, New Haven, Connecticut 06510

Abstract

Two distinct mutant alleles of the precursor (p) short chain acyl-CoA dehydrogenase

(SCAD)

gene were

identified

in a

SCAD-deficient

patient

(YH2065) using

thepolymerase chain reaction to amplify cDNA synthesized from total RNA from her

fibroblasts.

Cells from this patient had previously been shownto

synthesize

a

labile variant

SCAD in contrast tothe normal stability of variant

SCADs

in two other SCAD-defi-cient cell lines (Naito, E., Y. Indo, and K. Tanaka. 1989. J.

Clin.

Invest.

84:1671-1674).

Inthe present study, bothmutant

alleles ofYH2065werefoundtocontainaC -- T

transition,

one atposition 136 and the otheratposition 319 of the coding region of

pSCAD

cDNA.Clones of cDNA

amplified

from this region showed only oneof the C - Ttransitions, indicating that eachmutationwasderived from

different

pSCAD alleles. Eachof these

mutations

altered a knownrestriction

endonucle-ase

site,

and restriction

analysis

of additional cDNA clones from

amplified

mutantcDNA and

Southern

blotting of

mutant

genomic DNA

confirmed

the presence oftwo

unique

mutant

alleles in

YH2065, indicating YH2065

isacompound hetero-zygote. These C-- _T

_transitions

_{result in thesubstitution}

_of

Arg-22 and

Arg-83

of the mature SCAD with

Trp

and

Cys,

respectively. (J.

Clin.

Invest. 1990.

85:1575-1582.)

short chain

acyl-CoA

dehydrogenase.

variant

alleles. point

muta-tion * sequence analysis * altered restricmuta-tion sites

Introduction

Short

chain acyl-CoA dehydrogenase

(SCAD)'

deficiency

isan

inborn error of

fatty acid metabolism

that has

recently

been

reported

in three

infants (1, 2).

The

main

clinical

symptoms in

two

of

them were an

early

episode

of

lethargy

and

metabolic

acidosis,

whereas in the

other,

progressive

skeletal

muscle weakness,developmental

delay,

and

evidence of

fatty

infiltra-tion of

muscle and

liver

were the

prominent

features

(2).

In

AddressreprintrequeststoDr. K.Tanaka, YaleUniversitySchoolof Medicine, Department ofHuman Genetics, 333 Cedar Street, P.O. Box3333,NewHaven, CT 06510.

Receivedforpublication20October1989andinrevisedform 10

January1990.

1.Abbreviations used in this paper:IVD, isovaleryl-CoA

dehydroge-nase; MCAD, medium chain acyl-CoA dehydrogenase;p, precursor;

PCR,polymerasechain reaction; SCAD,shortchain acyl-CoA

dehy-drogenase.

ourprevious studies (3, 4), we have shown that the cells from all threepatients synthesized a variant SCAD protein of nor-mal size. This suggested that the defect of SCAD in these pa-tients is caused by a

point

mutation in the SCAD gene (3). Furthermore, usingimmunoblot analysis and pulse-chase

ex-periment,

we showed that variant SCAD synthesized by the cells from oneof the patients, reported by Amendt et al. as L.N. (1) and coded by us as YH2065 (3, 4), was labile, whereas the

stability

ofthe variant SCAD

protein

in the cells

from

two otherpatientswascomparabletonormalcells, indicating

het-erogeneity

of the

point

mutations (4).

The

biochemical function

of SCAD is to dehydrogenate

butyryl-CoA, initiating

the

13-oxidation

cycle

for

this substrate

atthe laststageoffatty acid oxidation (5, 6). It is a member of the

acyl-CoA

dehydrogenase

family

(7). Other enzymes inthis genefamily arelong chain acyl-CoA dehydrogenase and me-dium

chain

acyl-CoA dehydrogenase (MCAD),

isovaleryl-CoA

dehydrogenase

(IVD), and

2-methyl-branched

chain

acyl-CoA

dehydrogenase(7).

Like

four other members of the gene

family,

SCADisa

tetrameric

mitochondrial flavoenzyme

(5,

6).

Its 4 1-kDsubunit is encoded in the nucleus and

synthe-sized

in the cytosol as a precursor (p) that contains a 3-kD

extra

peptide (leader peptide)

atthe

NH2-terminus of

the ma-tureenzyme that is

required for

the transport of

pSCAD

tothe

mitochondria.

The

subunit

precursor

is

then

imported into

mitochondria, proteolytically

processed

into

the mature

sub-unit,

andassembled into the tetrameric form (8).

Wehave recently reported the

cloning

andsequencingof a cDNA

encoding

human

pSCAD (3).

The

coding region

of humanpSCAD cDNA is

1,236-bp

long, and encodes the

412-amino acid pSCAD.

The

first

72

nucleotides

at the 5'-end encodethe 24-amino acid leader

peptide

that is

contiguous

to

the

NH2-terminus of

the

388-amino acid

mature SCAD

(3).

With

these sequence data in hand, we sought to

characterize

thenatureof

mutation(s) responsible

for

SCAD-deficiency

at

the gene level

using

the

polymerase

chain reaction

(PCR).

This

technique provides

a

convenient procedure

to

greatly

enrich theamountofa

specific

DNAsequence,evenwhen present in minute amounts,

thereby

fascilitating

the

analysis

ofDNA

sequence(9, 10). Inthis paper,we reportthe identification of twodistinctvariant SCAD allelesinYH2065 cells,each witha pointmutation in anargininecodon at differentpositions.

Methods

Radioisotopes and chemicals.

[35S]dATP

and

[32P]dCTP

were

pur-chasedfromAmershamCorp.(ArlingtonHeights,IL).Most reagents and enzymesformolecularcloningwereprocured fromPromega

Bio-tec(Madison, WI). Taq DNApolymerase wasfrom Perkin-Elmer Cetus (Norwalk, CT). Random hexamerprimers pd(N)6 werefrom Pharmacia LKB Biotechnology (Piscataway, NJ) and oligo(dT)was

from Bethesda ResearchLaboratories (Gaithersburg, MD). The ran-J.Clin. Invest.

C) TheAmerican Society for Clinical Investigation,Inc. 0021-9738/90/05/1575/08 $2.00

dom-primed labeling kitwasfrom Boehringer Mannheim

Biochemi-cals (Indianapolis,IN).Guanidiumisothiocyanate was obtained from Eastman Kodak Co. (Rochester, NY). Cell culture materials were from

Gibco Laboratories (GrandIsland, NY).

Source of cells and culture method. The sources of the three SCAD-deficient cells have been previouslyreported (3, 4). The normal

celllineswereobtained fromthe NIGMS Human Genetic Mutant Cell

Repository (Camden,NJ).Fibroblastswere grown in monolayer cul-turein Eagle's minimum essential medium supplemented with 10%

fetalcalfserumand kanamycin (GibcoLaboratories) at370Cin 5%

CO2 atmosphere. The cellsweregrown toconfluenceand harvested by

trypsinization 3-6 d afterthe lastmediumchange.

RNAisolation and cDNA synthesis. Total cellular RNA was iso-lated from SCAD-deficient (YH2065) and normal (GM05756) cell

lines, and normal human liver using guanidium isothiocyanateas

de-scribedpreviously (3). Human liver poly(A)+RNAwasseparated from

total RNA by oligo(dT)-cellulose chromatography. First-strand cDNAs weresynthesized from total RNA or poly(A)' RNA using Moloney murine leukemia virus reverse transcriptaseand a cDNA

synthesis kit (Bethesda Research Laboratories)asfollows: 10 sgoftotal RNA or5

Ag

of poly(A)+ RNA was mixed onicewith either 5 ,ug of

random hexamer primers pd(N)6when the5'-and middlefragments

were to beamplifiedsubsequently or 2.5

Ag

of oligo(dT) for subsequent

amplification of the 3'-fragment. Transcriptionwasthencarried out at

37°C for 60 min in 50

Al

of reaction mixture containing 50 mM Tris-HCl (pH 8.3),75 mM KCl, 3mM MgCl2, 10 mMDTT,0.5 mM dNTPs(dATP, dCTP, dGTP, and dTTP) and 2.5 !l ofthe above reversetranscriptase.

Synthesis ofPCRprimers, andamplificationofthe human SCAD mRNAand genomic DNA. Three pairs of 30-mer primers were selected

asdescribed under Results and synthesized usinga DNAsynthesizer (model 380A, Applied Biosystems, Inc., Foster City, CA).Ineachpair, theupstreamprimer correspondedtotheRNA sensestrandsequence,

and thedownstreamprimer correspondedtothe antisense strand se-quenceof normal human SCADmRNA(3). Theprimerswere

puri-fied usinganoligonucleotide purificationcartridge (Applied

Biosys-tems,Inc.)(1 1). PCR amplificationwascarriedoutusingan

appro-priate aliquot(typically, 8-10

il)

ofthefirst-strandsynthesismixture

as template and 1 MM of each ofa pair of the PCR primers. The reaction mixture contained 10mM Tris-HCI, pH 8.3, 3.5 mMMgCl2,

50 mMKCI, 0.01%gelatin,and

200MgM

eachof dATP, dCTP, dGTP, and dTTP in 100Ml.Thereaction mixturewasboiledfor2min, cooled

toroomtemperature,0.5

,l

ofTaqDNApolymerasewasadded,and thereaction mixturewascovered with 100

Al

mineral oil(Brand-Nu

Laboratories, Inc., Meriden, CT). 30 cycles of PCRwereperformedin

aDNAthermalcycler(Perkin-Elmer Cetus) usingthefollowing

pro-file:1 mindenaturationat94°C, followed by2 minannealingat50°C,

and 3 min extensionat72°C.

For PCRamplificationofgenomicDNA sequencescorresponding tothemiddle1 18-bp section ofcDNA, highmolecularweight

genomic

DNAwasisolated from three SCAD-deficient and three normal cell linesusingguanidium isothiocyanateasdescribedpreviously (3). Ge-nomicDNA(1

Mg)

wasannealedto1MM each ofthepairofprimersfor the middle I18-bpfragment.Afterboilingthesamplesfor2min, 1Ml ofTaqDNApolymerasewasadded, and 30cyclesofthereactionswere carriedoutasdescribed above.

Subcloning ofPCRamplified products.ThePCR-amplified prod-uctderived from totalRNAwasfirstdigestedwithHinc II,TaqI,or

Bam HI, andpurified by electrophoresis using a 1.5% low-melting agarosegel (SeaPlaque;FMCBioProducts, Rockland, ME). Fragments

with the expected sizewereexcised,extracted withphenol/chloroform,

andprecipitatedwith ethanol.Theywerethen subcloned into

pBlue-scriptvector(Stratagene,LaJolla,CA),which had beendigestedwith Hinc II,Acc I or Bam HI. TheI18-bpPCRamplified productfrom

genomicDNAwasdigestedwith bothTaq Iand HincII, and then subcloned intobluescriptvectorwhich had beendigestedwith AccI

and Sma I. Thenewlyconstructed

plasmid

was

amplified

by

trans-formingEscherichiacolistrainXLI-Blue

(Stratagene).

DNA sequencing. DNA sequencing was performed directly using the recombinant pBluescript plasmid with its cDNA insert by the dideoxy-sequencing method ( 12) using an appropriate oligonucleotide (T7 or T3; Stratagene),

[35S]dATP

and the Sequenase kit (United States Biochemical Corporation, Cleveland, OH).

Digestion of the PCR products by restriction enzyme. After PCR, the reaction mixture was extracted with chloroform to remove the mineral oil. Two

20-MAl

aliquots were digested with either Hpa II or Pvu

II,and the resulting DNA fragments were resolved by electrophoresis ona 4%agarose gel (NuSieve; FMC BioProducts) containing ethidium bromide. The separated DNA fragments were visualized with ultravio-letlight.

Southern blot analysis of genomic DNA. 10 Mgeach of genomic DNA isolated from three SCAD-deficient and three normal cell lines were digested with Hpa II or Pvu II. The resulting fragments were separated on a 1.0% agarose gel (Bio-Rad Laboratories, Richmond, CA)andtransferred onto a sheet of Hybond-N membrane (Amersham Corp.) according to the manufacturer's protocol. The membrane was then hybridized using a 489-bp fragment (nucleotide position 23-51 1) of human SCAD cDNA as a probe. Final washes were performed at

650C

in 0.2 XSSPE/0.I% SDS. Blots were exposed to Kodak X-AR film for autoradiography.

Prediction ofthe secondary structure ofvariant SCAD using a com-puter.Acomputer-assisted prediction of the secondary structure per-turbations in variant SCAD caused by an amino acid substitution was performed bythemethod of Chou and Fasman(13)using the Univer-sity of Wisconsin Genetics Computer Group software developed by Wolf et al. (14). Hydrophilicity/hydrophobicity moments were also calculatedand overlaid on the secondary structure plots ( 15).

Results

Selection

of

primers

for

PCR

We planned to

amplify

theentire SCAD

cDNA-coding

region andsome

untranslated flanking region in

three sections, using three sets

of 30

oligonucleotide primers

that were selected and

synthesized according

tothe normal human SCAD cDNA se-quence (3) as

illustrated

in

Fig.

1. The segment between each

pair of primers

overlapped the

neighboring

one(s), such that

the

entire coding region plus

some in the

flanking regions of

the

SCAD

mRNA sequence could be

determined.

Inorder to

facilitate

the

isolation of

the

PCR

products,

primers

were se-lected in such a way that each contained an

appropriate

restric-tion site for

subcloning. Natural restriction sites

arepresent

in

the

primer pair for

the

middle

fragment and

in the

down-stream

primer for the 5'-fragment. 2,

1, and1

nucleotide

mis-matches,

respectively,

were

introduced into

the

primers for the

3'-fragment

and the upstream

primer for

the

5'-fragment

to create an

appropriate restriction site.

The three

restriction

en-zymes were selected because these enen-zymes

produced

no

re-striction

fragment length

polymorphism

in two normal and three

SCAD-deficient

cell

lines (3).

The

sizes

of

the

expected

amplification products

are

399, 638,

and

499

bp for

the

5'-,

middle, and

3'-fragments, respectively.

Amplification

of

the human

SCAD mRNA

using PCR

Asshown in

Fig.

2 A, the

size of

the

major

band in

each

of

the

5'-,

middle

and

3'-fragments amplified

from cloned

human SCAD cDNA

(HS- 1) exactly correspond

to

the

predicted

sizes

shown in

Fig.

1. In some

of

the

amplification products

from

total RNA

from

a normal and

YH2065

cell,

a

single major

band

was

detected.

In

others,

numerous

bands

were

detected,

but the

size of the

major

band in

each was

identical

to

the

corresponding band amplified from the cloned SCAD cDNA.

Taql TGA1239

l

51

3'

Taql Taql

(252-281) (860-889)

HincIl* Hinci!

(-30--l) (340-369)

bp

A 5-fragment Middle fragment 3-fragment

I

M

A

B

CIA

B

C

A

B

C

1159-1093

805

514\

468

448 '-

339-264

BamHI* BamHfo

(803-832) (1272-1301)

'

Onetotwomismatcheswereintroduced to accomodate the restrictionsite.

Figure 1.Strategy for PCR amplification ofp-SCAD cDNA, subclon-ing, andsequencing of the amplifiedfragments. Thesectionsshown

withashaded rectangle andlinesonbothsides indicate the coding

andnoncoding region of humanprecursorSCAD, respectively.The

base number for the AoftheATG initiation codonandtheterminal

Aof the TGAstopcodonareindicatedasI and 1,239, respectively.

Theposition of restriction sites, which exist inthecoding region of pSCAD and usedforsubcloning of the amplified fragments,is

indi-catedonthetoprectangle.The sixPCRprimerswereusedto

am-plifythreesections of pSCAD,shown eachwithhatched rectangles, fromfirst-strandcDNAsynthesizedfrom total RNA. Eachprimer

wasselectedsothateachpair containedanappropriate cloning site

(HincII, Taq I,orBamHI).Whenanappropriate endogenous

re-striction sitewasnotpresent,one ortwomismatcheswere

intro-ducedtoaccommodatetherestrictionsite. Thesearemarkedwithan

asterisk. The location indicated inparentheseswith basenumberis accordingtothepublished pSCAD cDNAsequence(3).The

ex-pected sizesof thePCR-amplified copiesare399,638,and499 bp, respectively.

Minorbandsseen were

presumably

produced by nonspecific

priming of other RNAs.

To definitively identify the bona-fide SCAD fragment in thePCR-amplification products, thesamegelwasfurther

an-alyzed by Southern blot

hybridization

using the cloned human cDNA (HS-1)as aprobe. In each lane, only the major band

anticipatedto correspondtoSCADsequence

specifically

hy-bridized (Fig. 2 B), positively identifying it as an authentic

copyof the intendedsegmentofhumanSCAD cDNA.

Identification

ofpoint mutations in YH2065 cell line. In the determination of the nucleotide sequences ofthe amplified cDNA fragments,wecarriedouttheentireprocessincluding the RNAisolation,reversetranscription,andPCR amplifica-tion

independently

two to three times foreach of thethree

fragments of mRNAs from YH2065, normal fibroblasts, and humanliver, and each productwassubcloned inorderto

ex-cludetranscription errorsincDNAsynthesisorin amplifica-tionby the PCR. Multiple clones of each product from each roundofamplificationwere

sequenced.

Inthesystematic

no-menclature for identifying the position of the nucleotide and amino acid substitutionsinthefollowing description, the

po-sitionsinthesequence of theprecursorSCAD isused unless otherwise noted.

Inall, fivedifferent nucleotide substitutions,notmatching the

previously

reported normal human SCADcDNA

se-quence (3), were consistently detected in YH2065 in each

roundof amplification. Three of themwerein the 5'-fragment

involving the codingsequence,whereastwootherswereinthe

noncoding region in the 3'-fragment. One of the three

substi-tutions found in the 5'-fragment was aC -- T transition at

B

3 -4ra C,- B--, cC--e,'

A B C A B C A B C

Figure2.PCR amplification of the first-strandcDNAsynthesized fromtotal RNAfrom normalandSCAD-deficientfibroblasts. The

coding regionofthe humanSCADmRNAwasamplifiedinthree sections from cDNAgenerated byreversetranscriptionofthetotal cellular RNA of the fibroblasts. 20 ofthereaction mixturewas

an-alyzedon a 1.5%agarosegel containingethidiumbromide. LaneM,

themolecularweightmarkers with PstI-digested XDNA;laneAin eachsection, thePCR-amplifiedfragmentsgeneratedfrom SCAD cDNA(HS- l) (3);lane B ineachsection,thePCR-amplified

frag-mentsderivedfromanormalcellline;and lane C in each section,

thePCR-amplifiedfragments derived from YH2065 cell line.(A) Analysisonethidium bromide-stained gel. (B)Southern blotanalysis

of the PCRproducts.Theethidium bromide-stainedgelshownin

Fig.2wasdenatured, transferredtoaHybond-Nmembrane, and

hy-bridized with the insert of humanpSCADcDNA clone(HS- 1) (3).

Fragment sizeswerecalculatedbasedonthesizeof PstI-digested

XDNAfragments, whichwereelectrophoresedinthesamegel.

position 136

(counted

taking theA inthe initiation codonas1) inaCGGcodon. This point mutation alters the 46th amino

acid residuein the pSCAD, changing Arg-22 in the mature SCAD,toTrp (Fig.3A). ThisC-- Ttransitionwasdetectedin

three of the sixclones from the firstround ofamplification. Two other substitutions found in the 5'-fragment from

YH2065 both involved the CGT codon encoding the 107th residue ofpSCAD that is Arg-83 of themature SCAD. One

was aC-- Ttransitionatposition 319,thatwasfound in three

remaining clones from the first round amplification. This

transition causedasubstitutionofArg-83withCys (Fig.3B).

ThesetwoC Ttransitionswereconsistently found in the

products from the later amplification experiments with

YH2065mRNA, butnotfrom those of normalmRNA. In the

independently repeated experiments, the C -- _T

136-transi-tionwasfound in 12 of the24clones, andintheremaining 12

clones,C T-319transitionwasfound. Wedesignatedthese

mutations R46W and R 107C, respectively. The other substi-tution involving the CGTcodon forArg-83 (mature SCAD)

was a T-- _{C transition at nucleotide 321, producingCGC.}

However, this appears to be a normal

polymorphism

since

bothCGTandCGCencode

Arg,

and both of themare

repeat-edly found in the amplified products from normal mRNA.

ATG TaqiHincli

Normal

SCAD

deficient

(GM05756)

(YH2065)

T-144

A

C G

T

A

C

G

T

T

Trp-46

-_126

Thedotted lineindicates aHpall site.

SCAD deficient

(YH2065)

_s<,eD%@

3' 5'

>c

AC G T

A

C

G T

A 1P _ i

-wo G___.Ar,."now E~~i4 A G T C G A C A C C G A C a- 312 A A . Cys-107 G G c G 326 5 3

Thedotted line indicatesaPvult site.

Figure3. Sequenceanalysisof the amplified

copies ofthe5'-fragments ofthe SCAD cDNAfromnormal andYH2065 cells. In thefirstroundofamplificationof the 5'-frag-mentof YH2065 mRNA, the amplified copies weresubcloned. Sixclones were se-quenced. Thenucleotideresidues involved in themutations are circled. In A, the se-quencefrompositions 126-144foundin threeoftheclones are shown on the right

side withthecorrespondingpartofthe

am-plifiedcopyfromnormal cells at theleft. C-136 in thenormalcDNAissubstituted with TinYH2065 cDNA. An HpaIIsite

(CCGG)innormalSCADsequence,

indi-catedbyabrokenline, is abolishedbythis transition(CTGG)atnucleotide 136. In B, the sequenceofthe reverse strandfrom posi-tions312-326of pSCADcDNAfrom nor-mal cells and thatfoundin three other clonesofYH2065 are shown on theleftand

rightsides,respectively. C-319 in the normal cDNAis substituted withTin YH2065 cDNA.Thisnucleotide substitutioncreated arestrictionsiteforPvuII(CAGCTG) as shownbyabrokenline.

These results indicate that there are two variant alleles, one

carrying R46W and the other carryingR107C.

The two substitutions found in the 3'-noncoding region

were a G -* A transition at position 1,246 and a G -3 C

transversion atposition 1,260. In addition to the five substitu-tionsthatwereconstantly found,wedetected another

nucleo-tidesubstitution,anA-- Ttransversion,atposition346 in the

coding region ofasingle clone. However, this substitutionwas

notdetected in the fragments from the later rounds of amplifi-cation.Therefore, itwasassumed tobeatranscription artifact. Digestion of the PCR products with restrictionenzymes. In

ordertofurther showthat the R46W and R 107C substitutions

are intrinsic to the variant alleles, we took advantage of the

changes of restriction sites caused by these transitions. As shown in Fig. 3 A, the tetranucleotide, CCGG, including

C-136,constitutesaHpaIIsite that would be abolishedby the

C-- T-136 transition.Theanticipatedresultsofthis digestion

is illustrated in Fig. 4 B. There arethree Hpa II sites in the

5'-fragment from normal SCAD mRNA. Thus, digestion of

the 5'-fragments from normal cells withHpaIIproducesfour

fragments (200, 114, 51, and34bp,respectively). Incontrast, digestion ofthe amplified copies of the 5'-fragments from YH2065 isexpected toproduce fivefragments, four

normal-sizedfragments, plusanadditional longer fragment(314 bp)

from allele-R46W. The resultsoftheelectrophoretic analysis

ofthe Hpa II digestion products are shown in Fig. 4 A. In

YH2065, a strong 314-bp band was detected in addition to

four shorternormalfragmentsasanticipated (lane 6), whereas

thisbandwas not detected in normal cells. Furthermore, in YH2065, the intensity of the 114- and 200-bp bands were

considerably weaker than in the normal cell line, indicating

that the C-* T-136transition is indeed intrinsictoonlyoneof

thetwovariantallelesinYH2065.

Inallele-R 107C, a Pvu II site is created dueto the C

T-319 transition whereas no Pvu II site is presentin normal

humanSCAD cDNA.Thus, when the 5'-fragments from

nor-mal and R46W alleles are treated with Pvu II, they remain

undigested. Incontrast, digestion of the 5'-fragmentfrom

al-lele-R107Cwith Pvu IIisexpectedtoproducetwofragments,

onewith the size of 347bpand theother withthesize of 52bp.

When total amplified 5'-fragment copies from YH2065 were

directlydigestedwith PvuII,threefragments (399, 347,and52

bp) wereindeed detected, whereasthe size of the normal

5'-fragment (399 bp)did notchangeaftertreatmentwith PvuII

(Fig. 4 A, lanes 7and 8). The undigested 399-bp piece in the

Pvu II digest ofYH2065 5'-fragment is presumably due to

allele-R46W.

Identification ofthe C -- T-319 transition in genomic

DNA.Toidentifythepresenceof the C-- T-319 transition in

genomic DNA,ashort segment ofgenomicDNAwas

ampli-fied by PCR usingthe upstream primer for the middle frag-ment and the down-stream primer for the 5'-fragment. The size oftheproductof thisamplificationisexpectedtobe 118 bp encompassing positions252-369.

1578 E. Naito, Y. Indo, andK. Tanaka

A

3' 144 -T T T A G Arg-46 cD r G T A A 126 -c

B

01

Normal

se s

(GM05756)

A Digested with: (-) Hpa II Pvu II

I<

--- 11

1-z

C

o- - Lo) - LC - Lo

-,2L <C)

E~~'

c CDL- aC t-E >-N

Z> U) Z >- Z >- Z

>-1 2 3 4 5 6 7 8

HpaII HpaII

..,- 52

-51 -34

Hpa II

Normal

HpaII HpaII

FR46W

SCAD -51

i

314 134d

deficient

-YH2065 HpaII HpaII[1HpaII

I ~~~II T

LR107C

347 1 52

Figure 4.Restrictiondigestion ofthePCR-amplified copies of the 5-end fragment of pSCAD cDNA from

normal andSCAD-deficient (YH2065) cell lines. (A) Electrophoretic separationof the restriction

frag-mentsfromthePCR-amplified copiesonethidium bromide-stained4%NuSieveGTGagarosegel.Hpa II (lanes S and 6) or Pvu II (lanes 7 and 8) was used fordigestion.Norestrictionenzymes wereusedin lanes2-4. TheDNAsampleswerederivedfrom the amplification of SCAD cDNA (HS- 1) (lane 2), and

1399 bp cDNAsfromnormal

(GM05756) (lanes

3, 5,and

7)

3 andSCAD-deficientfibroblasts (YH2065) (lanes 4, 6, and8).The molecularweight markers shown in lane 1werethe PstI-digested ADNA. The size ofeach di-gestedfragment is indicatedontheright side. In the

399bp lowerpart

(B),

thelocationof the restriction sites and

3 thesize of theresultingfragmentsareschematically

illustrated.

As shown inFig. 5, the majorband in theamplified prod-uctfromhumanSCADcDNA(HS- 1)was118bpasexpected. The major band of the PCR products from genomicDNAof

YH2065 and normal cell lineswasofthesamesize,indicating

that thesegmentbetween nucleotide 252-369 is locatedin a

single exon. The presence of C -* T-319 in the amplified

copiesfrom YH2065 wasstudied bothby sequencingandby Pvu II digestion. For sequencing, the PCR-amplified copies

weredigested with HincIIand TaqI,and then subcloned into

pBluescript. Sequence analysis of the amplified fragments

confirmed thepresenceof the C-* T-319transition intwoof

thefour clones obtained. Thesequences oftwoother clones

were normal, consistent with thepresence oftwodistinct

al-lelesinYH2065.When digested with PvuII,the 118-bp

frag-mentamplified from normal genomic DNA remained intact (Fig. 5, lane 5), whereas the 118-bp fragment amplified from YH2065 DNA yieldedtwofaintshortbands(66 and 52 bp)in additiontotheundigested 118-bp fragment.

Southern

blot

analysis ofgenomic

DNA.We also testedthe

direct effects of PvuIIandHpaII digestionofgenomicDNA

from three normal cell lines, YH2065 andtwo other

SCAD-deficient cells.Amongfive cell lines other thanYH2065,three patternswereobserved with PvuIIdigestion (Fig. 6). In three

cell lines (two normal and one SCAD-deficient), three frag-ments (8.7, 4.3, and 1.35 kb) were detected. In one of the

remainingtwo,aSCAD-deficient cellline,no8.7-kb bandwas

observed,whereas in theother, anormal cell line, the4.3-kb band was undetectable. Thus, the 4.3- and 8.7-kb bandsare

allelic. These three patternsappeartobe duetonormal poly-morphism, becausewe observed these three patterns also in

genomicSouthern blot analysisin nine normal human lym-phoblastoid lines (datanotshown).InYH2065,a1.2-kb band wasdetectedinaddition tothethree bands mentioned above (Fig. 6). This 1.2-kb bandwasnotobserved in anyother cell

lines including the five other fibroblast lines presented here and the nine normal lymphoblastoid lines. The presence of thisunique 1.2-kb Pvu IIfragmentisconsistent with the

exis-tence of the C Ttransition in the variant SCADgenome

that isresponsiblefor the C-. T-319transition.

Hpa II recognizes a specific sequence of only four bases

(CCGG). Therefore, numerous Hpa II restriction sites were

expectedingenomicDNA. Infact,whentheHpa TI-digested genomic DNA from three normal and two SCAD-deficient cell lines other than YH2065 wassubjectedto Southern blot

analysis, multipleminor bandsweredetected andsome varia-tions ofthe patternwereobserved.Onlytwobands,oneat10.0 kb and the other at 3.7 kb, were commonly detected. The pattern in YH2065 was conspicuously different from those

seen inthe other five cell lines: theHpaII-digested YH2065-DNA contained only one very prominent band at 5.5 kb in bp

- 399 - ₃₄₇

" ₃₁₄

- ₂₀₀

- ₁₁₄

B

Digested

with:

_(-)

z

o) - LI)

0w

o

o Lo

< '

c

C) 0 I CD Z

>-Pvu

11

I<

cl z

C

o - U)

<

E

CM

C)

z

>-A

bp

B

-118

Normal

VM

a helix WMA 1_sheet NH2 _vs

_A/ARandom

_coil

~,~v..vv~xr

0 Hydrophilicity>=1.3 Variant SCAD Hydrophobicity>=1.3

(R46W)

NH2 26 76

if |

A_____

_

A__

A_

--

1 2 3 4 5 6

Figure 5. PvuIl-digestionof thePCR-amplified products from

geno-micDNAinnormal and SCAD-deficient (YH2065) cell lines. The

electrophoretic separation oftherestriction fragments from the

PCR-amplified5'-end copyofthe genomicDNA using Pvu II is shown.

Ethidiumbromide-stained 4% NuSieve GTG agarose gel was used. Lanes 1-3,PCR-amplified products without digestion and lanes 4-6, PCR-amplified products digested with Pvu II. The DNA samples werederivedfromcloned pSCADcDNA(HS-1)(lanes I and 4), and

fromgenomic DNA in normal (GM05756) (lanes 2 and 5) and YH2065fibroblasts(lanes 3 and 6).

N

Lf-) (.0

IC) CD

2o urto

00c

Kb

8.7-

4.3-1.35

1.2-1

2

3

4

SCAD

deficient

\N1

I _{I co)}

Co₀₎₀₎₀₎co O c

cN i-r- -- C

irMr

Figure 6. Southern blot analysis of genomic

DNA. 10pg of genomic

DNAisolated from fi-broblastsweredigested withPvuIIand

sub-jectedtoSouthern blot analysis. Genomic

DNAsfromthree nor-mal(lanes1, 2,and 6)

andthree SCAD-defi-cient celllines (lanes 3-5)wereanalyzed.

_ Celllinesareidentified withthe numbers at the topofeach lane. Ap-proximate sizesare

giveninkilobaseson 5 6 theleft side.

Figure 7.Computer-assisted prediction of the secondary structure of

the5'region ofnormal andvariantSCAD-R46W. Theamino acid

residue fromtheNH2-terminusand 75th residue of the mature

nor-mal andvariant SCAD-R46W were analyzed using the Peptidestruc-tureandPlotstructure oftheGeneticComputer Groupnucleic acid

and proteinanalysisprogram.The regions of random coil,fl-sheet,

f-turn, and a-helixareillustratedin the figure. Thehydrophobicity

andhydrophilicityvalues areincluded inthe secondary structure plot asdiamondsand ovals,respectively. The C-. Ttransition- 136 causessubstitution of Arginine-22to tryptophan in themature

SCAD.Note that thefl-turnatthatpositionin normal SCAD is

abolished in SCAD-R46W.

additionto avery strong 10-kbband and a faint 3.7-kb band

(datanotshown). However,thesignificanceof this data is not clear atpresent.

Computer-assisted

prediction

of

the

secondary

structure. Computer-assisted analysis of the region (residue 1-80)

sur-rounding

the

Arg

-

Trp

substitution _atthe 22nd

position

of

thematureSCAD(R46W) predictsdrasticchangesinthe

sec-ondary structure of the NH2-terminal region. These changes include the loss oftwo f-turns anda random coil, with an

increased

tendencytoformana-helical conformation (Fig. 7). Nosuch drasticchangewaspredictedinthesecondary struc-ture of the variant SCAD (R107C) due to the Arg -. Cys

substitution.

Discussion

UsingPCRamplificationof SCADcDNAs,wedemonstrated

that YH2065cells containtwovariant SCAD

alleles,

one

pro-ducing

the SCADmRNAwith C-. Ttransition- 136 and the other

producing

theSCADmRNAwith C Ttransition 319. These mutations have been

unambiguously

confirmed

through

repeated

sequencing

ofclonesisolated from

indepen-dent RNA

isolations,

reverse

transcription

experiments,

and

PCR

amplification. Additionally, analysis

ofthe PCR

prod-uctsand

genomic

DNA

using

tworestriction enzymes,

Hpa

II

and Pvu

II,

has shown thatthe naturalrestrictionsite for

Hpa

II was abolished as

expected

by

C -. Ttransition- 136 anda newsite forPvu IIwascreated

by

C-- Ttransition-319. All

the

5'-fragment

clones had eitherof thetwoC -* T

transitions,

and were detected with

equal

frequency, indicating

that this

patientis acompound mutantfor these twovariantalleles. In

previous studies

using

immunoblot

analysis

and

pulse/

chase

experiments,

we have shown that unlike the variant

1580 E.Naito, Y Indo, and K. Tanaka

SCAD in two other SCAD-deficient patients, the variant SCAD-YH2065 in whole cells wasextremely labile (4). How-ever, in the previous biochemical experiments, we did not

observe clear evidencefor twodistincttypesofvariantSCAD with differing stability. Therefore, it appears at present that both SCAD-R46Wand-RI07Carelabile. It islikelythatthe

mature variant SCAD withcysteine-83 isunstable, since this

cysteine is unpairedandmaycausedimerization, leadingtoits

instability. The substitution ofArginine-22 with tryptophan

causes a drastic changein the secondarystructure of the

re-gion, possiblyleading to achange of the tertiarystructure. It maybe worthwhiletopointoutthatArginine-22inthe mature SCAD is anevolutionarilywell-conserved residue. Werecently reported cloningandsequencesof several acyl-CoA dehydro-genasesincludinghuman and rat SCAD(3, 7),and rat MCAD

(16), long chain acyl-CoA dehydrogenase, and IVD(7). Com-parison of the sequences ofthese five acyl-CoA dehydroge-nases plus human MCAD(17) revealed ahigh degree of

se-quence

homology,

with 57 invariant and 94 near-invariant residues among them. Arg-22 is one of the near-invariant resi-dues, conservedinfiveacyl-CoAdehydrogenaseswith the

sin-gle

exception of

ratIVD

(7).

It

is possible

that more accuratemeasurementoftherateof

the

loss

of

35S-labeled SCAD

inthe whole

cell-labeling

experi-ment, or the

direct

measurement

of

the

stability

of SCAD-R46W and

-RI07C,

that are

synthesized

in

vitro

using the

isolated variant SCAD

cDNA, may reveala

difference

in their

stability. Still

another

possibility is

that

only

one

of

the variant SCAD

subunit is unstable,

but that most

of

the variant

SCAD-YH2065

tetramers become

unstable

due to the

pres-ence

of

one or more

molecules

of the unstable variant subunit

participating

in thetetramer.

At the

nucleotide

level, the sequences

of

the two Arg codonsinvolved in the C-- Ttransitions_at

positions

136 and

316 both

participate in

the

formation of

a

tetranucleotide,

GCCG,

with both normal C residues

contributing

the second C.

This tetranucleotide contains CpG,

the

major

site of

meth-ylation

in

human

DNA.It

is interesting

to notein

this regard

that

an

unusually

high frequency

of

restriction fragment length

polymorphism has been observed

at various loci in human DNA

with the

use

of

restriction

enzymes whose

recognition

sequences

contain

a

CpG dinucleotide.

Ithas been

speculated

that

CpG

dinucleotides are

probably

mutation hot spots

be-causethe

methylation of

cytosine, followed by deamination of

methylcytosine

to

thymidine,

wouldresult inaC T transi-tion

(18).

Point mutations duetoC Ttransition in the

CpG

dinucleotides

have been

identified

in Factor VIII gene and

ornithine transcarbamylase

gene

(19-21). Thus, the

identifi-cation of

thetwo

point mutations described

in

this study

are

further examples of mutations

caused

by C

- T

transitions

involving CpG dimers.

Acknowledgments

Synthesisand purification of oligonucleotides werecarried out by

Monica Talmor,DepartmentofPathology, YaleUniversitySchoolof

Medicine. Cultured skin fibroblasts from threepatients with SCAD deficiencywere kindly provided byDrs. Carol Green (University of

Colorado,Denver,CO), Vivian E. Shih (Massachusetts General

Hos-pital, Boston, MA), and Susan Winter (Valley Children's Hospital,

Fresno,CA).We thank Mrs. ConnieWoznick forpreparing this

man-uscript.

Thiswork was supported by a grant from the National Institutes of Health (DK-38154).

References

1.Amendt,B. A.,C. Greene, L.Sweetman,J.Cloherty,V.Shih,A.

Moon,L.Teel, andW. J.Rhead. 1987.Short-chain acyl-coenzymeA

dehydrogenasedeficiency.Clinical and biochemical studies intwo

pa-tients.J.Clin. Invest. 79:1303-1309.

2. Coates,P. M., D. E. Hale,G.Finocchiaro,K.Tanaka, andS. C. Winter. 1988. Genetic deficiency of short-chainacyl-coenzymeA

de-hydrogenasein culturedfibroblasts fromapatient withmuscle

carni-tinedeficiency andsevere skeletal muscleweakness. J. Clin.Invest.

81:171-175.

3. Naito,E.,H. Ozasa, Y. Ikeda, and K. Tanaka. 1989. Molecular

cloning and nucleotidesequenceof complementaryDNAsencoding human short chainacyl-coenzyme Adehydrogenaseand thestudyof the molecular basis ofhuman shortchain acyl-coenzymeA dehydroge-nasedeficiency. J. Clin.Invest.83:1605-1613.

4.Naito,E., Y. Indo, and K. Tanaka. 1989. Short chain acyl-coen-zyme Adehydrogenase (SCAD)deficiency: Immunochemical

demon-stration of molecular heterogeneitydue tovariant SCAD withdiffering stability. J. Clin.Invest.84:1671-1674.

5. Ikeda, Y., K.Okamura-Ikeda,and K. Tanaka. 1985. Purifica-tion and characterization of short-chain, medium-chain, and long-chain acyl-CoA dehydrogenases from ratliver mitochondria.J.Biol Chem. 260:1311-1325.

6.Finocchiaro, G.,M.Ito, andK.Tanaka. 1987.Purification and properties of shortchain acyl-CoA, medium chain acyl-CoA,and

iso-valeryl-CoA dehydrogenases from human liver. J. Biol. Chem. 262:7982-7989.

7. Matsubara, Y., Y. Indo,E. Naito, H.Ozasa, R.Glassberg, J. Vockley, Y.Ikeda, J. Kraus, and K. Tanaka. 1989. Molecular cloning andnucleotidesequenceof cDNAs encoding theprecursorsofratlong chain acyl-coenzymeA, shortchain acyl-coenzymeAand

isovaleryl-coenzymeAdehydrogenases:sequencehomology of fourenzymesof the Acyl-CoA Dehydrogenase Family. J. Biol. Chem. 264:16321-16331.

8. Ikeda, Y., S. M. Keese,W. A.Fenton, andK. Tanaka. 1987. Biosynthesis of fourratliver mitochondrial acyl-CoA dehydrogenases:

Invitro synthesis,import intomitochondria, and processing of their precursors in acell-free system and in cultured cells.Arch. Biochem. Biophys. 252:662-674.

9.Saiki,R.K., D. H.Gelfand,S.Stoffel,S.J.Scharf,R.Higuchi,

G. T. Horn, K. B.Mullis, and H. A. Erlich. 1988. Primer-directed enzymaticamplification ofDNAwithathermostableDNA

polymer-ase.Science(Wash.DC). 239:487-491.

10. Gibbs,R.A.,P.-N.Nguyen,L. J.McBride, S.M.Koepf,and

C.T.Caskey. 1989. Identification ofmutations leadingtothe Lesch-Nyhan syndrome by automated direct DNA sequencing ofin vitro

amplifiedcDNA.Proc.Nati.Acad. Sci. USA. 86:1919-1923.

11.McBride, L. J., C. McCollum, S. Davidson,J. W.S.Efcavitch, A.Andrus, and S.J.Lombardi. 1988.Anew,reliable cartridgefor the rapid purification ofsynthetic DNA.Biotechniques.6:362-367.

12. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA

se-quencing withchain-terminating inhibitors. Proc. Natl. Acad. Sci. USA. 74:5463-5467.

13.Chou,P.Y., and G.D.Fasman.1978.Empirical predictions of protein conformation.Annu.Rev.Biochem. 47:251-276.

14.Wolf,H., S.Modrow,M.Motz,B.Jameson,B.Hermann, and

B.Fotsch. 1988.Anintegrated family of amino acidsequenceanalysis

15. Hopp, T. P., and K. R. Woods. 1981. Prediction of protein antigenic determinants from amino acid molecular weights. Proc. Nail. Acad. Sci. USA. 78:3824-3828.

16. Matsubara, Y., J. P. Kraus, H. Ozasa, R. Glassberg, G. Finoc-chairo, Y. Ikeda, J. Mole, L. E. Rosenberg, and K. Tanaka. 1987. Molecular cloning and nucleotide sequence of cDNA encoding the entire precursor of rat liver medium chain acyl coenzymeA dehydro-genase. J. Bio. Chem. 262:10104-10108.

17. Kelly, D. P., J. J. Kim, J. J. Billadello, B. E. Hainline, T. W. Chu, and A. W. Strauss. 1987. Nucleotide sequence of medium-chain acyl-CoA dehydrogenase mRNA and its expression in enzyme-defi-cient human tissue. Proc. Nail. Acad. Sci. USA. 84:4068-4072.

18. Barker, D., M. Schafer, and R. White. 1984. Restriction sites

containing CpG showahigherfrequency ofpolymorphismin human DNA.Cell. 36:131-138.

19. Gitschier, J., W. I. Wood,M. A. Shuman,and R. M. Lawn.

1986.Identification ofamissense mutation in the factorVIIIgene ofa

mildhemophiliac.Science(Wash. DC).232:1415-1416.

20. Maddalena, A., J. E. Spence, W.E.O'Brian,andR.L. Nuss-baum. 1988. Characterization ofpointmutations in thesamearginine

codon in three unrelatedpatientswith ornithinetranscarbamylase de-ficiency.J.Clin.Invest.82:1353-1358.

21. Hata, A.,C. Setoyama,K.Shimada,E.Takeda,Y.K. I. Aka-boshi, andI. Matsuda. 1989.Ornithinetranscarbamylase deficiency

resultingfromaC-to-T substitution inexon5of the ornithine