Volume 28]. number !, it:~-8tl

FEBS 09694

May 1991

1991 Fcdera|iol~ of ~UrOl~:tn Biochemical $ ~ i ~ t l ~ O014~'/9~t~JllS].~lO

A DONI,~ 0014~"/9~91004011t

Structural features of stomach aldehyde dehydrogenase distinguish dimeric aldehyde dehydrogenase as a 'variable' enzyme 'Variable' and 'constant' enzymes within the alcohol and aldehyde dehydrogenase families Shih-Jiun Yin, N i k o l a o s Vag¢lopoulos, Sung-Ling W a n g a n d H a n s J6rnvall Departntcnt of Chemistry I, Karol#tska Instituter. S.104 Ol Stockhob~L Sweden and Department af Biochcmistr),, Nathmal Defcn,r¢ Medical Center. PO Bo.~ 90048. Talp¢l. Taiwan R,O.C;

Rece~vc~d8 March :199t Stomach aldehyde dehydrosenase was structurally evaluated by analysis of pcptide fragments of the human enzyme and ¢ontparisons with corresponding parts from other characterized aldehyde dehydrogenases. The results establish a large part of the structure, confirming that the stomach enzyme Is identica; to the inducible or tumor, derived dimcric aldehyde dchydrogena~, In addition, species variations between identical sets of different aldehyde lind alcohol dchydr0genases reveal that stomach aldehyde dehydrogenas¢ exhibits a fairly rapid rate of evolutionary changes,

similar to th:tt for the likewise 'variable' classical alcohol dehydrogenase, sorbitol dehydrogenase, and cytosolic aldehyde dehydrogena~ but in contrast to the 'constant' class Ill '.alcohol dehydrogenase and mitoehondrial aldehyde dehydrogenas¢, This establishes that rates of divergence in the aldehyde and alcohol dehydrog©nascs are unrelated to subunit size or quaternary structure, highlights the unique nature of class 111 alcohol dchydrogena~, and positions the stomach aldehyde dehydrogenas¢ in a group with more ordinary i'eaturcs. D0hydrogcnase; Enzyme variability; Divergence

1, INTRODUCTION Alcohol and aldehyde dehydrogenases both exhibit considerable multiplicity with mammalian structures known for four [1] and three [2] 'classes', respectively, and with at least three additional mammalian enzymes [3-6] clearly belonging to these two families (sorbitol dehydrogenase and ~'-crystallin in the case of the alcohol dehydrogenase family; glutamic ,y-semialdehyde or l-pyrroline-5-carboxylate dehydrogenase in the case of the aldehyde dehydrogenase family). In both cases, most studies have concerned the liver enzymes, but alcohol and aldehyde dehydrogenases have wide-spread occurrence, and the two act ivities also occur in stomach, although only limited amounts of the enzymes have been available for study from this source. Recently, however, the stomach alcohol dehydrogenase was established to constitute a new structural class, class IV [1], while the stomach aldehyde dehydrogenase has been concluded [7-10] to be identical to one of the already known aldehyde dehydrogenases, the form that is inducible or tumor-derived in liver [2,11] and also occurs in bladder [12]. Because of this difference between the stomach alcohol and aldehyde dehydrogenases, it is Correspondence address: H, JOrnvall, Department of Chemistry 1, Karolinska Institutet, S-104 01 Stockholm, Sweden. Fax: (46)(8) 33 74 62

Published by Elsevier Science Publishers B. V,

essential to characterize this aldehyde dehydrogenase further. Only a brief report [10] on limited data without further details exists. Furthermore, knowledge on the structural variation of dimeric aldehyde dehydrogenase is of special interest, since other alcohol and aldehyde dehydrogenases differ widely in being either fairly 'constant' (class III alcohol dehydrogenase and mitochondrial aldehyde dehydrogenase [13-15]) or 'variable' (class I alcohol dehydrogenase, sorbitol dehydrogenase and cytosolic aldehyde dehydrogenase [13,16]). We therefore now characterized human stomach aldehyde dehydrogenase by analysis of an Asp-specific digest of the protein, giving close to half of the protein structure. This allows firm conclusions on the structural relationships, and shows stomach aldehyde dehydrogenase to be in the 'variable, of the two categories of alcohol and aldehyde dehydrogenases. 2. MATERIALS AND METHODS Human stomach aldehyde dehydrogennse was prepared as described [7,8], yielding about 3 nmol enzyme from 10 g wet weight tissue. The enzyme was reductiveiy alkylated by ['4C]carboxymethylation and submitted to cleavage with Asp-N protease at a substrate/protease ratio of 1:100 in I M urea/0.15 M ammonium bicarbonate, pH 8, for 22 h at 37°C, The digest obtained was separated by reverse phase HPLC on Ultropac TSK ODS-I20T (5 tzm; Pharmacia/LKB, Bromma, Sweden) in 0.1°7o trifluoroacetic acid with a gradient of acetonitrile [I]. All fractions obtained were analyzed for purity by solid-phase sequencer degradation in a MilliGen Prosequencer 6600

85

Volume 28), member I

FEBS L E T T E R S

+ r ........

il

© vt,

M=y 1991

+iT

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

'7 f

i,

!

+

O~0

,+1 ¢C1+ O

Q0~

1 Ol

I

I!i

I Q"

d

f)¢

+. . . .

" - ;~

,

Om . . . . . . . . . . . . . . .

, z

i

+.+ .... 'I:

......

!+

oF

i+

............. ,+'+ opo

3,~+ i!:+

~

I

Is

~'

I.

,!

60

~0

VOLUME t a l l

Fill. i. Reverse phase HPLC of a divert with Asp-N pretense of ['~Clcarbo.xym~thylaled human stoma~;h aldehydedehydroi~enase. All major l'ra¢(ions were analyzed for amino acid sequence. Only those labelled D A . D J gave interpretable results, As shown, a few rra~tions contained nor© titan one peptide, but the mixturescould beclearly identified by different yields of constituent p©ptides anti by comparisons with adjacent framlions. Several peptides wills identical N.termlnal structures elute al more titan one position, presumably becattse of separate secondary modifications or different C,terntinal cleavages,

utilizing arylamine membranes for C.~ermina~ coupli.lL Results obtained were correlated wttlt known ~tldehydedehydrogenase structures

aldehyde dehydrogcnase, cytosolie class I and I l l alcohol dehydro. genases, a,td sorbitol dehydrogenase, reeenlly established from the structures of those rat/human enzymes (el, [13-19],),

[19] and with the different ~voh.donary rates for mitochondrial

"Fable I Primary structt|res of ten segme,ts of human s¢onn=ch aldehyde dchydrogenase now analyzed ( H u m D, for dinleri¢, top line in all 4.set sequences) versus structures known for dimeric rat inducible aldehyde dehydrogenase (Rat D , second line in eacll set), and the two tetrameric h u m a n enzymes

from cytosol and mitochondria ( H u m . C nnd Hum-M, bottom two lines in each set) Positions refer to those known for the entire structure of the tumor.derived rat dehydrogenase [2,11] while peptide designations DA-DJ refer Lo the Asp-cleavage products purified as sltown in Fig. I. Data for the bottom three lines f r o m [2,11,19]

1 (pos£tions 5o-60) Ilum P Rat D Hum C Hum M

I• DA--'-~ DLIIKNS:WNAY¥ DLGKNEWTSY¥ NGGKLYSNAYL DNGKPYVISYL

2 (83-94)

3

I Dm-----.q DEPVEKTPQTQQ DEPVAKTRQTQQ KZQGRTIPIDGN K Y H G K T I PIDGD

I, , ~ De-I DLY PV IX G G L P E T T E L L K E R - - P N LY L W K G G V P E T T E L L K E R - - F GVVN IVPGYGPTAGAAISSHMDI G V V N I V P G F G PTAGAAI A S H EDV

4 (IBI-218)

5 (252-273)

t ~D { DttI L Y T G S T G V G K I IM T A A A K - H LT P V T L E L G G R S P X Y V DHIMYTGSTAVGKIVMAAAAK-HLTPVTLELGGRSPCYV DKVAFTGSTEVGKLIKEAAGKSNLKRVTLZLGGKSPClV DKVAFTGSTEIGRVIQVAAGSSNLKRVTLELGGESPNII

DPSIQNQIVEK-LKKSLKEFYGE DPSIQNQIVEK-LKKSLKDFYGE EESIYDEFVRRSVERAKKYILGN QEDIYDEFVERSVARAKSRVVGN

6 (280-308) I

9

}

i0

(389-418)

,DI

DVIVHITLHSL-PFGGVGNBGMGBYHGKKSF

DVIVHZTV~n-PFGGVGNS~MGA~MGKKSp WVNCYGWSAQCPFGGFKMSGNGRELGEYGF WVNCYDVFGAQSPFGGyKMSGSGRELGEYGL

---I

DE

7 (3o9-321) DF

DYGRIISARHFQRVMGLI ..... EGQK~/A-YGGTG D Y G R I I N D R H F Q R V K G L I ..... D N Q K V A - H G G T W TQGPQIDKEQYDKILDLIESGKKEGAKLECGGGPW EQGPQVDETQFKKILGYINTGKQEGAKLLCGGGIA

86

(160-180)

~---------DG

B (349-3G9) {

I

--DH

DAATRYIAPTILT DQSSRYIAPTILV GNKGYFVQPTVFS

EAIQFINQREKPLALYMFSSV EAIQFINQREKPLALYVFSNN DVIKRANNTFYGLSAGVFT--

ADRGYFIQPTVFG

EVVGRANNSTYGLAAAVFT--

(4~5-452) DJ---

EEGLKVRYPPSPAKMTQH EEAHKARYPPSPAKMPRH NS* NS*

I Hum D Rat D Hum C Hum M

Votllmer l]~t, number I

F~ItS L~TTEKS

3, RESULTS Human stomach aldellyde dehydrol~onase fratlm=nts obtained by digestion of the carboxymethylated enzyme with Asp.N pretense were fractionated by reverse phase HPLC. Of t7 fractions obtained (FIB. IL t0 8aveinterpretable results upon solid.phase sequencer dellradalions, thus establishing 215 residues o f the proteinehaln of human stomaclt aldehyde dehydrogenase, as shown in Table [ in relation to corresponding structures for the three types of aldehyde dehydrogenase known before. Several o f the cleavages generatintl the peptides reflect cleavages at Glu residues, which are often found when the Asp.N pretense is used in the presence o f urea (which was necessary here to solubilize the protein before cleavage). A comparison o f the aldehyde dehydrogenase strut, turcs show that the human stomach enzyme is identical in subunit organization at the 215 positions examined to the inducible, or tumor-derived, liver aldehyde de. hydrogenase. Thus, both that enzyme and the present one (top two lines in each 4.line set o f sequences; Table I) show identical gap positions versus the two types of characterized aldehyde dehydrogenase, the tetrameric eyt 0solic and mitrochondrial liver enzymes (bottom two lines; Table I). The stomach enzyme and the other dimeric aldehyde dehydrogenase exhibit zero gap differences (Table II), while both have 13 gap differences towards the tetrameric aldehyde dehydr0genases. Similarly, the stomach aldehyde dehydrogenase, now analyzed from human, shows only 38 residue differences towards the tumor-derived aldehyde dehydrogenase previously known from rat [2,111, and this difference (18°'/0 in 215 positions) falls exactly in one o f the two groups typical of species differences of rat/human dehydrogenases (Table II). In summary, stomach aldehyde dehydrogenase appears identical to the tumor-derived enzyme and is clear, ly distinguishable from the c o m m o n types of normal liver aldehyde dehydrogenase. This result is in agreement with previous expectations [7-10] and considerably extends previous structural knowledge [10]. Parenthetically, it also ascribes a residue for the human stomach enzyme at three positions previously differently ascribed, thus now Glu-298, Thr-321, and Ser-298 (Table I). Finally, the results allow reasonable estimates of the species divergence for the dimeric aldehyde

May 1991

dehydroilenase in relation to divergences for identical species of five other enzymes within the aldehyde and alcohol dehydrogenase families, revealing that the stomach aldehyde dehydrollenase fails into the 'variable' Igroup of alcohol and aldehyde dehydrosenases (Table III)

4, DISCIJSSION The results confirm that stomach aldehyde dehydrollenase is identical to other dimeric aldehyde dehydrogenases known, tile inducible or tumor-derived enzyme, obtained from liver, bladder and other sources [2,7-111, This is concluded from analysis o f close to half the entire structure, and this type of aldehyde dehydrogenase is therefore now possible to compare for large parts of the structures from two different species, rat and human, revealing a species divergence at 18% of all positions, This divergence can be directly compared to that for other dehydrogenases of the same metabolic pathways, mitochondrial aldehyde dehydrogcnase, cytosolic aldehyde dehydrogenase classical alcohol dehydrogenase, class Ill alcohol dehydrogenase, and polyol dehydrogenase, all also known from human and rat (cf. [13-161). Unexpectedly, the rate of divergence of these dehydrogenases fails into two distinct ranges (Table III). One is fairly variable, containing classical alcohol dehydrogenase, sorbitol dehydrogenase and cytosolic aldehyde dehydrogenase from before, the other is fairly constant containing class [1I alcohol dehydrogenase and mitochondrial aldehyde dehydrogenase. The two types differ by a factor of 3 or more. We now find that stomach aldehyde dehydrogenase falls into exactly one of these two types of divergence, i.e. the variant one (Table III), surprisingly with identical values in all cases. This suggests that the evolutionary rate is independent of substrate type (alcohols" aldehydes) quaternary structure (dimers-tetramers), and total size of the subunit ( - 3 5 0 to - 5 0 0 residues). The latter point is further illustrated by inclusion of glucose-6phosphate dehydrogenase, also known from rat and human (cf. [201), which, like aldehyde dehydrogenase has a large subunit in contrast to the alcohol and polyol dehydrogenases. The lack of correlation with size for the five aldehyde and alcohol dehydrogenases in Table III and their distinction into two classes of variation, is

Table II Differences among aldehyde dehydrogenases (AIdDH) emphasizing the presently analyzed stomach form versus the three types previously characterized 12 II 17] Data from Table I Human stomach AIdDH versus Rat tumor.derived AIdDH Human liver cytosolic AIdDH Human liver mitochondria] AIdDH

Differences

Total positions compared

Gaps

Residues

Sum

0 13 13

38 137 142

38 150 155

215 210 210

87

V o l u m e 283, number I

FEBS L E T T E R S

M a y 1991

Table III

Diffcr~n¢~ between Id¢~tlcal ~p¢~i¢~~Is (human/rat) of tfimerl¢ aldehyde dehydrogenate, v~r~u~oilwr aldd~yd¢ tld~ydrot~¢na~ and ihre¢ dlf, fercnl rept~,~eltlal[vcl oi' llte al~ohol dd~ydro~cm~s¢family (classical ¢la~s I cnxym¢, ~ I ~ III ~ntymc, and ~otbhol ddlytlfol~¢rla~e} AIdDH, aldehyd~ dd~ydtog~nas¢; At)H, al¢oholdehydrot~ca't~¢. Fat comparison, the idcntleztl ,tel of l~ht~,'os¢.6.pho~phaledehydrogenaw, with suhunils similarly si~ed as Ihos,z for alddtydedehydrollenase, is ttl,to shown, Dala for dtmeri¢ AIdDH from Table 11. for previously knt~wn tittle. tgre~ from 12,11;I;~-20) Ral,~'h),lmancn~')'me pair:

........... Dhneric AIdDH Mitochondrlal AIdDH Cylo~oli¢ AidDH Glass 1 ADH Clas~ ill ADH Sorbitol dehydrollenase G icose,6.pho~phate dehydrogenas¢

Position~_W!l!)_ d!/feregc~Lor 10,,.!.pO).td0n.~ con!pared . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 'C0.~laI~t~ ent);mes . . . . . . . . . . . . . . . . . 'Variab!=' en~ymo 3a of 215 (I~%) 18 or 500 t4%1 86 of 500(17%} 66 of 333 (18%} 21 or 374 (6%} 33 of 513 (6%}

a d e v i a t i o n f r o m the g e n e r a l t r e n d o f a c o r r e l a t i o n bet w e e n size a n d e v o l u t i o n a r y rate [ 2 1 1 - T h i s d e v i a t i o n e s t a b l i s h e s that s t r u c t u r a l a n d f u n c t i o n a l r e q u i r e m e n t s i n f l u e n c e even closely r e l a t e d a n d m e t a b o l i c a l l y s i m i l a r e n z y m e s to d i f f e r e n t e x t e n t s . I n p a r t i c u l a r , the class III a l c o h o l d e h y d r o g e n a s e is especially c o n s t a n t w i t h i n t h e g r o u p ( T a b l e I l l ) in r e l a t i o n to its s m a l l s i z e , s u g g e s t i n g the p o s s i b i l i t y t h a t it is i n f l u e n c e d b y strict r e q u i r e m e n t s o n i m p o r t a n t p r o p e r t i e s , w h e r e a s the s t o m a c h a l d e h y d e d e h y d r o g e n a s e is the m o s t v a r i a b l e o f those with large subunits perhaps suggesting more variable functions w i t h i n this m e t a b o l i c patl~way.

Acknowledgements: "Tiffs work was supported by grants from the Swedish Medical Research Council (Project 03X-3532), the Knut and Alice Wallenberg Foundation, the National Science Council of the Republic of China, and by a visiting scientist fdlowship to S,-J.Y. from Karolinska Instituter. REFERENCES [I] ParEs, x.. Moreno, A,, Cederlund, E., H~Sg, J,.O, and J6rn. vail. H. (1990)FEBS Lett, 277, 115-t 18, [2] Hempel, J,, Harper, K. and Lirtdahl, R, (1989) Biochemistry 28, 1160-i 167. /3] J6rnvall, H,, van Bahr-Lindstrfm H. and Jeffery, J, (1984) Eur, J Biochem. 140, 17-23. [4] Bart.'is, T., Persson, B. and J6rnvall, H, (1989) Biochemistry 28, 6133-6139. [5] Krzywicki, K.A. and Brandriss, M.C. (1984) Mol, Cell. Biol, 4, 2837-2842.

88

(6] Forte-McRobbie, C.M. and Pietrus•ko. R. (1986) J. Biol. Chem. 261, 2154-2163 [7] Y[n, S.-J,, Liao, C..S,, Wa,lg, S..L.; then, Y..J and Wu, C,W, (1989) Bioehem, Genel. 27, 321-331. [8l Wang, S . L . , Wu, C..W., Chcag, T..C. and Yin, S...1. (1990) Bi¢,-chem. ha. 22, 199-204. [9] Koivusalo, M,. Aarnio, M. and Rautoma, P, (1989) in: Enzymology and Molecular Biology ot" Carbonyl Metabolism (Weiner, H, and Flynn, T.G. eds) pp. 19-33, Liss, New York. [10] Eckey, R,, Agarwal, D.P., Hempel, J. Timmann, R, and Gocdde, H,W, 0990) Abstr, Enzymol. Mol. Biol. Carbonyl Mctab. [11] Jones Jr,. D.E,, Brennan, M.D., Hempel, J. and Lindahl, R. 0988) Proc, Nail, Acad, Sci, USA 85, 1782-1786. [12] Lindahl, R. (1986) Cancer Res. 46, 2502-2506, [13] Jail,5, P,, Par~s, X and J6rnvall, H, (1988) Eur, J, Biochem, 172, 73-83. [14] Kaiser, R,. Hohnquist, B,, Vallee, B.L. and J6rnvall, H, (1989) Biochemistry 28, 8432-8438, [15] Farr6s, J,, Guan. K.-L. and Weiner, H, (1989) Eur, J, Biochem, 180, 67-74 [16] Karlsson, C,, JOrnvall. H and H66g, ,I,.O, (1991) Eur, J, Biochem in press, [17] Johansson, J .. yon Bahr-Lindstr6m, H., Jeck, R., Woenckhaus, C. and J6rnvall, H. (1988) Eur. J. Biochem. 172,528-533, [18] Dunn, T,.I., Ko]cske, A.J., Lindahl, R. and Pitof, H,C; (1989) J, Biol. Chem. 264. 13057-13065. [19] J6rnval], H., Persson, B., Krook, M. and Hempel, J. (1991) m: The Molecular Pathology of Alcoholism (Pahner, N, ed,) Oxford Univ. Press pp. 130-156. [20] Jeffery, J,, Barros-SSderling, J,, Murray, L., Wood, [., Nansen, R., Szepesi B. and J6rnva], H, (1989) Eur. J Biochem, 186, 551-556, [21] Dickerson, R.E, 0971) J, Mol Evol, 1, 26-45,