D(

AGU

R

zdamdcobAc9ahcaaprd

c

bdotmppeaie

Wfic

Biochemical and Biophysical Research Communications 286, 1082–1086 (2001)

doi:10.1006/bbrc.2001.5511, available online at http://www.idealibrary.com on

0CA

evelopmental Expression of Alcohol DehydrogenaseADH3) in Zebrafish (Danio rerio)

sok K. Dasmahapatra, Herman L. Doucet, Champa Bhattacharyya, and Michael J. Carvan, III1

reat Lakes Water Institute and Marine and Freshwater Biomedical Sciences Center,niversity of Wisconsin-Milwaukee, Milwaukee, Wisconsin 53204

eceived July 30, 2001

their development produced FAS-like abnormalities oftmftbip

bdsAwtlAAdtteAtAada(smIado

M

so

Alcohol dehydrogenase (ADH) is the primary en-yme responsible for metabolism of ethanol to acetal-ehyde. One class of ADH has been described in fish,nd has been found to be structurally similar to mam-alian class III ADH (glutathione-dependent formal-

ehyde dehydrogenase) but functionally similar tolass I ADH (primarily responsible for ethanol metab-lism). We have cloned a cDNA by RT-PCR from ze-rafish (Danio rerio) liver representing the zebrafishDH3 gene product, with a coding region of 1131 nu-leotides. The deduced amino acid sequences share0% identity to ADH3 from the marine fish Sparusurata, and 82 and 81% identity to the mouse anduman sequences, respectively. Using a quantitativeompetitive RT-PCR assay, ADH3 mRNA was detectedt all timepoints analyzed and was lowest between 8nd 24 h postfertilization. Thus, differential ADH3 ex-ression may be at least partly responsible for tempo-al variations in the sensitivity of zebrafish embryos toevelopmental alcohol exposure. © 2001 Academic Press

Key Words: alcohol dehydrogenase; ADH3; zebrafish;ompetitive RT-PCR; ethanol; teratogenesis.

Fetal alcohol syndrome (FAS) refers to a pattern ofirth defects seen in some children exposed to alcoholuring the prenatal period. The major manifestationsf this disease are primarily physical and mental re-ardation, craniofacial malformations, and joint abnor-alities. Although the exact mechanisms for FAS are

resently unclear, it has been shown that polymor-hisms in alcohol dehydrogenase (ADH) genes influ-nce susceptibility to FAS in humans (1). Numerousnimal models have been utilized to investigate thenfluence of genetics on susceptibility to FAS. Alcoholxposure of zebrafish (Danio rerio) embryos during

1 To whom correspondence should be addressed at Great Lakesater Institute, University of Wisconsin-Milwaukee, 600 E. Green-

eld Avenue, Milwaukee, WI 53204. Fax: (414) 382-1705. E-mail:arvanmj@uwm.edu.

1082006-291X/01 $35.00opyright © 2001 by Academic Pressll rights of reproduction in any form reserved.

he face, eye, ear, and heart similar to that seen inammals (2). Only one full-length cDNA sequence

rom a teleost fish (Sparus aurata) has been reported inhe literature (3), and nucleotide sequences in the ze-rafish EST database that are identified as ADHs arencomplete and their relationship to other ADH generoducts is unclear.The dimeric zinc-containing alcohol dehydrogenases

elong to the protein super family of medium-chainehydrogenases/reductases and consist of a complexystem with different forms and extensive multiplicity.DHs are able to catalyze the reversible oxidation of aide variety of xenobiotic and endogenous alcohols to

he corresponding aldehydes (4). ADH1 is the classicaliver enzyme responsible for ethanol metabolism.DH2 is also found in liver with a high Km for ethanol.DH3 is a glutathione-dependent formaldehyde dehy-rogenase that can oxidize ethanol at high concentra-ions. ADH3 is the ancestral ADH and has been iden-ified in all species analyzed (4, 5). ADH4 is expressedxtrahepatically and is involved in retinol metabolism.DH5 and ADH6, have been poorly investigated and

heir substrate specificity is not well understood.DH7 has been described only in the chicken embryo,nd showed activity with retinoids. ADH8 has beenescribed in amphibians and has a tissue distributionnd substrate specificity similar to mammalian ADH46). Previously, only one class of ADH has been de-cribed in fish, which are structurally similar to mam-alian ADH3 and functionally similar to ADH1 (7, 8).

n the present communication, we report the deducedmino acid sequence of zebrafish ADH3, its expressionuring development, and its relationship to ADHs fromther species.

ATERIALS AND METHODS

The maintenance of zebrafish embryos, larvae and adults are theame as described by Westerfield (9). Eggs were collected within 2 hf fertilization and maintained at 28.5°C in embryo water (purified

wi

zDo(wwipAaar

cs(Abi4fifipmMw

wRTgasrTpn

smRdtcMspna

wBR(mu

a0ebbrttstwfT

sNmwfm

R

zzoaahmid1piaddGfis

c

Vol. 286, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ater with 60 mg/l Instant Ocean and 25 ppb methylene blue tonhibit fungal growth).

RNA isolation. The procedures for isolation of RNA from theebrafish liver, embryos, and larvae are similar to those reported byasmahapatra et al. (10). Briefly, approximately 10 mg of liver (fromne fish) or 10 embryos or larvae were homogenized in 1 ml TrizolLife Technology) and left at 280°C until analysis. The homogenateas extracted with chloroform (0.2 ml) and nucleic acids precipitatedith isopropyl alcohol using glycogen (10–20 mg) as a carrier. The

solated RNA was treated with DNase I (Promega, Madison) in theresence of 1 mM MnCl2, and 40 units of RNAsin at 37°C for 30 min.fter enzyme digestion, DNase was removed by phenolic extraction,nd RNA precipitated with ethanol. RNA was resuspended in nucle-se free water and used for reverse transcriptase-polymerase chaineaction (RT-PCR), or competitive RT-PCR (cRT-PCR).

RT-PCR. Reverse transcriptase-polymerase chain reaction wasarried out using the Access-RT-PCR system from Promega (Madi-on, WI) according to the manufacturer’s specifications. Forward59-CTT TCC ATC GAG GAG GTG GAG-39) and reverse (59-GAGAG GTG CTG GTG CCC ATG AAG TGG-39) primers were designedased on ADH sequences in GenBank. The reactions were conductedn a Perkin Elmer (Norwalk, CT) Model 480 DNA thermal cycler at8°C for 45 min, followed by 94°C for 2 min. RT product was ampli-ed by 33 cycles of 94°C, 30 s; 60°C, 1 min; 68°C, 2 min, followed bynal template extension at 68°C for 7 min. The PCR product was gelurified and cloned using the pGEM-T-Easy Vector system (Pro-ega, Madison, WI). The vector insert was sequenced on an ABIodel 373 automated DNA sequencer and all obtained sequencesere compared to sequences published in GenBank.

Rapid amplification of cDNA ends. The complete coding regionas obtained by Rapid amplification of cDNA ends (RACE) using 39ACE System and 59 RACE System, Version 2 (Invitrogen Lifeechnologies, Carlsbad, CA). For 39 RACE, a sequence specific oli-onucleotide was used in conjunction with the manufacturer’sdapter primer. For 59 RACE, the first strand of cDNA was synthe-ized using a sequence-specific oligonucleotide with superscript IIeverse transcriptase (Life Technology, Grand Island, NY) at 42°C.he amplification of 59 end was performed using sequence-specificrimers and the manufacturer’s anchor abridged primer (Life Tech-ology, Grand Island, NY).

Quantitative cRT-PCR. cRT-PCR reactions were carried out es-entially as described in Dasmahapatra et al. (10). A 382-bp frag-ent from the ADH3 cDNA was used for the synthesis of competitorNA. The cloned cDNA template was amplified with a specificallyesigned forward primer that produced a fragment 10% shorter thanhe original cDNA (target) sequence. The in vitro transcription wasarried out using Riboprobe in vitro transcription systems (Promega,adison, WI), according to manufacturer’s protocol. The synthesized

ingle stranded competitor RNA was isolated by phenol extraction,recipitated with isopropyl alcohol, separated from unincorporateducleotides by size-exclusion chromatography through a small Seph-dex G-50 column (5 Prime3 3prime, Boulder, CO) and precipitated

FIG. 1. Deduced amino acid sequence of zebrafish ADH3. The uompetitive RT-PCR reaction. The mRNA sequences is available fro

1083

ith ethanol. The concentration of the RNA was determined witheckman DU 64 spectrophotometer at 260 nm, and the purity of theNA was checked by gel electrophoresis. Once the concentration

copy number) of the synthesized single-stranded RNA was deter-ined, the sample was aliquoted (1–2 ml) and stored at 280°C until

se.Competitor RNA was added to multiple RT-PCR reactions, gener-

lly 6–8, such that the copy number added ranged from 0.5 3 106 to.025 3 106. PCR products were then separated by agarose gellectrophoresis. DNA content in both the target and competitorands was estimated by densitometry of ethidium bromide stainedands with Kodak EDAS (Rochester, NY). Second-degree polynomialegression analysis was used to calculate the relationship betweenhe ratio of the intensity of the product and competitor bands (Y) andhe natural logarithm of the competitor copy number (X). The regres-ion equation was expressed as: Y 5 b0 1 b1 (ln X) 1 b2 (ln X)2. Fromhe estimated regression curve, the theoretical equivalence point, X,as calculated by solving the quadratic equation. The R2 ranged

rom 0.800 to 0.999, indicating the goodness of fit for these curves.he mRNA copy numbers are normalized to mg of total RNA.

Statistical analysis and sequence alignment. All data compari-ons were analyzed by one-way Anova followed by Student–ewman–Keuls test (P , 0.05). The results were expressed asean 6 SEM of at least four independent samples. The dendrogramas constructed from multiple alignments of amino acid sequences

rom GenBank using the Clustal W algorithm in VectorNTI suiteolecular biology software (Informax, North Bethesda, MD).

ESULTS

Isolation and sequencing of cDNA clones encodingebrafish ADH. A 382-bp fragment was isolated fromebrafish liver RNA using RT-PCR with primers basedn the high sequence identity in the mouse, human,nd fish ADH sequences in the region between aminocid residues 22 and 149. This fragment demonstratedigh sequence similarity with ADH3 cDNA from thearine fish Sparus aurata. The remainder of the cod-

ng region was acquired using RACE and primers wereesigned for amplification of the entire coding region of131 bp. Zebrafish ADH3 codes for a 376 amino acidrotein of approximately 40 kDa (Fig. 1) that is 90%dentical to the sequence from Sparus autata and 82nd 81% with human and mouse ADH3, respectively. Aendrogram representing the multiple alignment ofeduced vertebrate ADH amino acid sequences fromenBank is shown in Fig. 2. All fish sequences identi-ed thus far segregate with the mammalian ADH3equences. Within the ADH3 sequences there are three

erlined sequences indicate the position of the primers used for theenBank Accession No. AF399909.

ndm G

dpttsAsf

dmfietblte(pdsteb

mRNA in adult liver is, however, 50% that of larvae of4

D

dwtahtcefanrgtnwdBA1sitImn

vwncpoASt

uDt7A(P(A

Vol. 286, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

istinct branches, the first of which includes am-hioxus and hagfish, the second branch is comprised ofhe teleost fishes, and the third branch contains mul-iple mouse and human forms. This provides furtherupport for the idea that ADH3 is the ancestral form ofDH (4, 5), however it is likely that additional ADHequences will be found as fish ESTs and genomes areurther analyzed.

Expression of ADH mRNA in zebrafish liver anduring early development. The ontogeny of ADHRNA expression in zebrafish was determined at dif-

erent developmental stages and adult liver by compet-tive RT-PCR (Fig. 3). ADH3 mRNA was quantified inmbryos as early as 4 h of development—the earliestime point examined. ADH3 mRNA in the early em-ryos is likely of maternal origin, and ADH3 mRNAevels generally decrease after 4 hpf and remain lowerhrough 24 hpf. From 30 hpf onwards the ADH mRNAxpression increased and reached a peak by 96 hpfthreefold increase compared to 20-h-old embryo). Ex-ression levels at 8, 16, 20, and 24 hpf are significantlyifferent from those at 96 and 120 hpf (P , 0.05)uggesting a greater physiological need for ADH3 ac-ivity in later stages of development, which may leavembryos vulnerable to xenobiotic chemicals that maye metabolized by ADH3. The expression of ADH3

FIG. 2. Dendrogram of ADH proteins. The tree was constructedsing Clustral W algorithm (Nucleic Acid Res. 22, 4673–4680, 1994).educed amino acid sequences of ADH used in this analysis other

han zebrafish were obtained from GenBank Accession Nos. AAF3255 (amphioxus), S51187 (hagfish), P79896 (Sparus), P81600 (codDH3H), P81601 (cod ADH3L), P28474 (mouse ADH3), XP017924.1

human ADH3), Q64437 (mouse ADH4), P40394 (human ADH4),00329 (mouse ADH1), NM000669.2 (human ADH1C), NM000668.2

human ADH1B), NM000667.2 (human ADH1A), XP011115 (humanDH5), AJ 245750.1 (mouse ADH2), and AAA51595 (human ADH2).

1084

–6 days of age.

ISCUSSION

The cytosolic zinc-dependent medium-chain alcoholehydrogenases are recognized to form several classesith different activities and corresponding gene mul-

iplicity in humans and most vertebrates. Most char-cterized ADHs are from mammalian origin, primarilyuman. Biochemical properties and evolutionary rela-ionships have been best characterized for threelasses, the classical liver enzyme with considerablethanol activity (class I), the glutathione-dependentormaldehyde dehydrogenase (class III) and the stom-ch expressed class IV enzyme with the highest etha-ol activity. After much evaluation of the evolutionaryelationships between the various vertebrate ADHenes, it has been proposed that the class structure ofhe mammalian enzymes are reflected in their geneames (e.g., class I 5 ADH1) and that additional genesould be classified according to the similarity of theireduced amino acid sequences (Duester et al., 1999).etween species, ADH1 structure is fairly variable,DH3 is more constant and ADH4 is intermediate (5,1). All these classes presumably originated from aingle gene through a series of gene duplications dur-ng early vertebrate evolution and ADH3 appears to behe ancestral type with subsequent emergence of classactivity. This hypothesis is based on estimates ofolecular evolution, the apparent absence of ADH1 in

onvertebrate eukaryotes, and the presence of ADH

FIG. 3. ADH3 mRNA expression in zebrafish embryos and lar-ae. Total RNA from whole embryos/larvae or adult hepatopancreasas used for competitive RT-PCR to determine the ADH mRNA copyumber. The lowest values, marked with an asterisk (*) are signifi-antly different (P , 0.05) than the highest values, marked with theound symbol (#). The data are presented as the mean of at least fourbservations, error bars represent the standard error of the mean.ll comparisons were analyzed by one-way Anova followed bytudent–Newman–Keuls test. Underlined sequences in Fig. 1 denotehe regions used to generate primers for competitive RT-PCR.

enzymes in nonmammalian vertebrates that exhibitsmbaes

AihasAqczeAl

dshloevo1hddtdzrpue

tgctdettatwbimdem

ternal ADH mRNA. We do not know when zebrafishetsaHHio(otehg

mlAtsgtbs

A

M(iNsWaC

R

Vol. 286, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ixed properties being functionally similar to ADH1ut structurally similar to ADH3 in fish (12). With theim to investigate ethanol metabolism and teratogen-sis in zebrafish, we found a cDNA with structuralimilarities to the ADH3 reported in fish.The liver ADH of zebrafish is most similar to theDH3 reported from Sparus aurata (3). Two ADH3

sozymes have been characterized from cod where theform (high activity toward ethanol) and l form (low

ctivity) differ such that the l form is more similartructurally and functionally to other vertebrateDH3s (12). Zebrafish ADH3 deduced amino acid se-uence is 87 and 83% identical to the l and h forms ofod ADH3, respectively. Thus, we can expect that theebrafish ADH3 will metabolize ethanol, but to a lesserxtent than the mammalian ADH1. Interestingly,HD2 in birds is also functionally similar to mamma-

ian ADH1 in its ability to metabolize ethanol (13).Our results suggest that ADH3 mRNA expression is

evelopmentally regulated in zebrafish as demon-trated by the significant levels of ADH3 mRNA at 4pf, which is likely of maternal origin, the reduced

evels of ADH3 through 24 hpf, and the rapid increasef ADH3 levels between 24 and 36 hpf. Expression wasnhanced approximately three-fold in 4–6 day-old lar-ae compared to embryos 16 hpf. Similar results werebserved in Sparus aurata, with an initial decrease to2 h of development and then significant increase afteratching (3). The reduced levels of ADH3 expressionuring early development may, in part, explain theevelopmental stage-dependent sensitivity to ethanoleratogenesis described by Blader and Strahle (2). Weo not know whether ADH3 expression in developingebrafish embryo occur ubiquitously in all tissues or isestricted to particular cell types as observed in am-hioxus (14). In situ hybridization experiments arenderway to localize ADH3 expression in the zebrafishmbryo during development.ADH is responsible for the metabolism of alcohol to

he corresponding aldehyde. Ethanol, a known terato-en to zebrafish embryos, induces malformations in-luding split axis, notochord and spinal cord duplica-ion, necrosis, and cyclopia in a time and dose-ependent manner (2, 15, 16). The exact mechanism ofthanol toxicity is still unknown. It is likely that theiming of ethanol exposure, the development stage ofhe embryo, and the level of expression of ADH mRNAnd protein play a significant role in the induction ofhese teratogenic effects. Treatment of the embryosith 2.4% ethanol at dome-30% epiboly stages (latelastula-early gastrula, about 5 hpf) induced cyclopian almost 75% of the embryos. However, ethanol treat-

ent at ring stage (late gastrula, almost 6 hpf) pro-uced cyclopia in less than 1% (2). This variation inthanol-induced teratogenesis in zebrafish embryoay be related to a time of rapid degradation of ma-

1085

mbryo initiates expression of ADH but in amphioxushe zygotic expression of the ADH gene was first ob-erved during the neurula stage (late gastrula) (14),bout 10 hpf in the zebrafish. Ethanol also up regulatesSP47 gene expression in zebrafish embryos (17).SP47 is a stress-induced heat-shock protein predom-

nantly expressed in precartilagenous cells, notochord,tic vesicles, and fins in developing zebrafish embryos17). As a result, the maternal contribution and timingf embryonal HSP47 expression may also play a role inhe abnormalities typically induced by developmentalthanol exposure. The relationship between ADH3,eat shock proteins and alcohol needs to be investi-ated further.Taken together, our results demonstrate that ADH3RNA is expressed in whole zebrafish embryos and

arvae, and in adult liver, and that the expression ofDH3 mRNA is developmentally regulated during on-

ogenesis. This study is the first step toward under-tanding the mechanism of alcohol toxicity and terato-enesis in zebrafish at the molecular level. Utilizinghe powerful molecular genetics capabilities of the ze-rafish system will advance our mechanistic under-tanding of alcohol teratogenesis in other vertebrates.

CKNOWLEDGMENTS

This research was supported by the University of Wisconsin-ilwaukee, the National Institute for Alcohol Abuse and Alcoholism

NIH R03 AA12408), and the UWM Marine and Freshwater Biomed-cal Science Center (NIH P30 ES04184). H.L.D. was supported byational Environmental Health Science Training Program Fellow-

hip provided by the Medical College of Wisconsin, Milwaukee, WI.e would also like to thank our colleagues for valuable discussions

nd careful reading of this manuscript. Contribution Number---,enter for Great Lakes Studies, University of Wisconsin-Milwaukee.

EFERENCES

1. McCarver, D. G., Thomasson, H. R., Martier, S. S., Sokol, R. J.,and Li, T.-K. (1997) Alcohol dehydrogenase-2*3 allele protectsagainst alcohol-related birth defects among african americans.J. Pharmacol. Exp. Ther. 283, 1095–1101.

2. Blader, P., and Strahle, U. (1998) Ethanol impairs migration ofthe prechordal plate in the zebrafish embryo. Dev. Biol. 201,185–201.

3. Funkenstein, B., and Jakowlew, S. B. (1996) Molecular cloning offish alcohol dehydrogenase cDNA. Gene 174, 159–164.

4. Duester, G., Farres, J., Felder, M. R., Holmes, R. S., Hoog, J. O.,Pares, X., Plapp, B. V., Yin, S. J., and Jornvall, H. (1999) Rec-ommended nomenclature for the vertebrate alcohol dehydroge-nase gene family. Biochem. Pharmacol. 58, 389–395.

5. Danielsson, O., and Jornvall, H. (1992) “Enzymogenesis”: Clas-sical liver alcohol dehydrogenase origin from the glutathione-dependent formaldehyde dehydrogenase line. Proc. Natl. Acad.Sci. USA 89, 9247–9251.

6. Crosas, B., Cederlund, E., Torres, D., Jornvall, H., Farres, J., andPares, X. (2001) A vertebrate aldo-keto reductase active withretinoids and ethanol. J. Biol. Chem. 276, 19132–19140.

7. Danielsson, O., Eklund, H., and Jornvall, H. (1992). The major

piscine liver alcohol dehydrogenase has class-mixed properties in

1

1

1

patterns in a class III alcohol dehydrogenase. Effects on the

1

1

1

1

1

Vol. 286, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

relation to mammalian alcohol dehydrogenase of classes I andIII. Biochemistry 31, 3751–3759.

8. Ramaswamy, S., el Ahmad, M., Danielsson, O., Jornvall, H., andEklund, H. (1996) Crystal structure of cod liver class I alcoholdehydrogenase: Substrate pocket and structurally variable seg-ments. Protein Sci. 5, 663–671.

9. Westerfield, M. (1995) The Zebrafish Book: A Guide for Labora-tory Use of Zebrafish (Danio rerio). University of Oregon Press.

0. Dasmahapatra, A. K., Wimpee, B. A. B., Trewin, A. L., Wimpee,C. R., Ghorai, J. K., and Hutz, R. J. (2000) Demonstration ofTCDD attenuated P450 steroidogenic enzyme mRNA transcrip-tion in rat granulosa cell in vitro using competitive reversetranscriptase-polymerase chain reaction assay. Mol. Cell. Endo-crinol. 164, 5–18.

1. Farres, J., Moreno, A., Crosas, B., Peralba, J. M., Allali-Hassani,A., Hjelmqvist, L., Jornvall, H., and Pares, X. (1994) Alcoholdehydrogenase of class IV (sigma sigma-ADH) from humanstomach. cDNA sequence and structure/function relationships.Eur. J. Biochem. 224, 549–557.

2. Danielsson, O., Shafqat, J., Estonius, M., El-Ahmad, M., andJornvall, H. (1996) Isozyme multiplicity with anomalous dimer

1086

activity and quaternary structure of residue exchanges at “non-functional” sites in a native protein. Biochemistry 35, 14561–14568.

3. Hjelmqvist, L., Estonius, M., and Jornvall, H. (1995) The verte-brate alcohol dehydrogenase system: Variable class II type formelucidates separate stages of enzymogenesis. Proc. Natl. Acad.Sci. USA 92, 10904–10908.

4. Canestro, C., Hjelmqvist, L., Albalat, R., Garcia-Fernandez, J.,Gonzalez-Duarte, R., and Jornvall, H. (2000) Amphioxus alcoholdehydrogenase is a class 3 from a single type and of structuralconservation but with unique developmental expression. Eur.J. Biochem. 267, 6511–6518.

5. Laale, H. W. (1971) Ethanol induced notochord and spinal cordduplications in the embryo of the zebrafish, Brachydanio rerio. J.Exp. Zool. 177, 51–64.

6. Sander, K., and Baumann, M. (1983) Ausleosung von embryo-nalen Fehlbildungen beim Zebrabaerbling. Biol. Zeit 13, 87–94.

7. Lele, Z., and Krone, P. H. (1997) Expression of genes encodingthe collagen-binding heat shock protein (Hsp47) and type IIcollagen in developing zebrafish embryos. Mech. Dev. 61, 89–98.