Isoforms of Mammalian Cytochrome C Oxidase: Correlation with Human Cytochrome C Oxidase Deficiency

Isoforms of Mammalian Cytochrome c Oxidase: Correlation with Human Cytochrome c Oxidase Deficiency

NANCY G . KENNAWAY. ROQUE D. CARRERO -VAL ENZUELA. GA RY EWART. VIJAY K. BALAN. ROB ERT L1GHTOWLERS. YU -ZHONG ZH ANG . BERKLEY R. POW ELL RODERICK A. CAPA LDI. AN D NEILR. M. BUIST Departmen ts ofMedica! Generics [NG.K., R.D.C.- v., V.K.B.. N R.M.B.j and Pediatrics [B.R.P., N R.M. B.j, Oregon Health Sciences Univcrsitv. Portla nd. Oregon 9720i and Institute ofMolecular Biology [G.E.. R.L., Y-z.z., R.A.Cj. University ofOregon, Eugene. Oregon 97403

ABSTRACf. We have reviewed the structure, function, IV (cytochro me c oxidase) (5) have all been described. Th e and biogenesis of mammalian cytochrome c oxida se, ex severity of these disorders is very variable, and the spectrum of amined the tissue-specific expression of isoforms of cyto clinical disease is impressive, ran ging from isolated myopath y or chrome c oxidase subunits in different mammals, and at cardiomyopathy in some patients to a multisystem disease such tempted to correlate the data with our knowledge of cyto as Leigh's syndrome in others. Thi s variability presum ably results chrome c oxidase deficiency, illustrated by one particular in part from the complexity ofthe respirato ry chain proteins, the patient. Cytochrome c oxidase was isolated from bovine smallest of which (complex II) comprises four or five individual tissues, and individual subunits examined by SDS-PAGE, subunits, and the largest of which (complex I) comprises at least N-terminal peptide sequencing, and antibody binding. Iso 26 indi vidual subunits (6). Some of these subunits appear as forms of subunits VIa, VIla, and VIII were identified, tissue-specific isoforms in mammals. manifesting one pattern of expression in heart and skeletal The ident ification of patients with respiratory chai n deficien muscle, and another in liver, kidney, and brain . In rat heart cies limited to on e or a few tissues suggests that isoforms of some and liver, only one form of subunit VIla was identified. of these proteins may also exist in humans and provide an Northern analysis of bovine and rat tissues suggested that explan ation for the tissue-specific expression of the disease in the tissue-specific expression of subunits VIa and VIII is such patients. Until very recentl y, attempts to define the tissue regulated transcriptionally in liver, kidne y, and brain, and specificity of the respiratory chain have focused mai nly on beef posttranscriptionally in heart and skeletal muscle. In hu and rat. not only because of the ready avai lability of tissue mans, antibody binding documented isoforms of subunits material. but also because of the assumptio n that the pattern of VIa and VIla, with the pattern of expression in heart and expression ofthese isoforms would be conserved in all mammals, skeletal muscle differing from that in liver, kidney, and includ ing humans. There is preliminary evidence of isoforms of brain; our data suggested that both isoforms of subunit VIa some subunits of complex I (7); however, the mo st comprehen may be expressed in human heart. In a patient with cyto sive data have been obtained on cytochrome c oxidase. In this chrom e c oxidase deficiency, the clinical, morphologic, and report we will: 1) review the stru cture, function , and biogenesis biochemical manifestations were much more severe in heart of mammalian cytochro me c oxidase ; 2) discuss the tissue spec than in skeletal muscle. Antibod y binding suggested partial ificity and expression of isoforms of cytochrome c oxidase in assembly of the enzyme in heart. These and other data mammals, including humans; and 3) relate thi s briefly to the suggest considerably more variability in the tissue-specific pheno types of cytochrome c oxidase deficiency, and to on e expression of isoforms of cytochrome c oxidase subunits patient that we have studi ed recently in parti cular. than previously recognized . (Pediatr Res 28: 529-535, 1990) CYT OCHROME c OX IDASE: TE RMINA L OXIDASE IN RESPIRATI ON Abbreviations Structure.funaion, and biogenesis. Cytochrome c oxidase, the mtDNA, mitochondrial DNA terminal component of the respiratory chain, spans the mito chond rial inner membrane and tran sfers electrons from reduced cytochrome c to molecular oxygen with generatio n of a trans mem brane proton gradient. In mammals, the enzyme is a mul Over the past few years, there has been increasing recognit ion ticomponent complex containing 13 different polypeptides (8). of the importance of inherited disorders of the mit ochondrial The three largest subunits are coded on mtDNA and synthesized respiratory chain in a wide variety of human diseases (I ). Dis in the mit ochond rion. The se mitochond rially-coded subunits orders of complex I (NADH:ubiquino ne oxidoreductase) (2), appear to be the functional core of the enzyme, but the func complex II (succinate:ubiquinone oxidoredu ctase) (3), complex tion(s) of the nuclear coded subunits is presentl y uncle ar. Figure III (ubiquinol:cytochro me c oxidoreductase) (4), and complex 1 shows the resolut ion of all the subunits of beef heart cyto chrom e c oxidase on highly resolving SDS-PAGE and presents Corre spondence and reprint requests: Dr. Nan cy G, Kcn naway, Department of two nomenclatures currently in use for the different subunits; Medical Genetics/ L473. Oregon Health Sciences Un iversity. 3181 S.W. Sam Jack the nomenclature of Kadenbach et al. (9) is used in this review . son Park Rd.. Portland. OR 9710 1-3098. Supported in part by NI H G rant 22050. a gran t from the Musc ular Dystrophy The biosynthesis of cytochrome c oxidase is complicated, Association. and a grant from the Medical Research Foundation of Oregon. involving components coded on two geno mes and synthesized 529 530 KENNAWAY ET AL.

KADEN BACH CAPALDI CYTOPLASM et al1983 at al1987

II Mtll III Mtm

IV CI y

Va Cy Vb C ---- YI VIa ASA VIb AED VIc STA "- VII a ____ CYII . VII b IHQ Fig. 2. Schematic representatio n of the synthesis and assembly ofthe respirator y chain complexes in mammals. Reproduced from Takamiya :======: VII c Gym VIII 1'1al. (12), with permi ssion . CIX Fig. I. Subunit str ucture ofbeef heart cytochro me c oxidase, showing Rat . two nome nclature s cur rentl y in use (9, 10). The enzyme was purified In their original experiments , Merle and Kadenbach (14) from mit ochondrial preparatio ns essentially as described (11), sub units found an altered migration of subunit VIa when the composition separated on an 18.5% polyacrylamide gel containing 6 M urea (9), and of the heart and liver enzyme was compared by SDS-PAGE (14). proteins stained with Coo massie Brilliant Blue. The presence of isoforms of subunit VIII as well as VIa was indicated by differences in labeling of these two subunits in different tissues by sulfhydryl reagents (15). Characterization of in two compart ments, namely the mitochondrial matrix and the two cDNA, one isolated from a rat liver, the other from a rat cytoplasm. This overall process is shown schematically in Figure heart cDNA library, offers definitive evidence that there are two 2. Briefly, nuclear coded subunits, synthesized on free ribosomes, forms of subunit VIa (16). The two forms of subunit VIa encoded are released in the cytoplasm as precursor proteins, frequently by these cDNA show only 50% homology. The expression of the containing an N-terminal extension or leader peptide that targets isoforms of subunit VIa was explored by Schlerf et al. (16) using the protein to the mitochondrion and that may help direct the Northern analysis; the heart form of this subunit was found to protein to the correct location within the organelle, i.e. intracris be transcribed in heart and skeletal muscle only but, surprisingly, tal space, inner membrane, or matrix. Mitochondrially synthe the liver form was transcribed in all tissues examined (16). sized proteins are transcribed as a polycistronic message that is More recently, Kuhn- Nentwig and Kadenbach (17) have used processed to give mRNA for individual subunits. The three quantitative immunoassays to compare subunits from a variety subunits of cytochrome c oxidase are synthesized as mature of adult and fetal rat tissues. Based on these studies, they have polypeptides in mammals and may be cotran slationally inserted concluded that all of the nuclear coded subunits of cytochrome into the mitochondrial inner membrane. Steps involved in the c oxidase may occur in different tissue-specific forms in mam addition of heme and coppers to subunits I and II have not been mals, that more than three different isoforms are present for identified yet, but may occur at the inner membran e concurrent many subunits, and that there are major differences between with the assembly of the mitochondrially and nuclear coded fetal and adult tissues. However, these results are open to ques subunits into the complex. tion, because some of the observed antigenic differences may have resulted from acquired conformatio nal differences such as ISOFORMS OF MAMMALIAN CYTOCHROME c OXIDASE the extent of denat uration of the same subunit from different tissues. The occurrence of tissue-specific isoforms of mammalian cy The early data of Kadenbach et al. (13) showed no difference tochrome c oxidase was first suggested by Kadenbach et al. (13), in the migration of cytochrome c oxidase subunit VIla in rat based on slight differences in the apparent molecular weight of heart and liver by SDS-PAGE, in contrast to what was seen in some subunits in different tissues from each of several species some other mamm als (13). We have recently obtained prelimi including rat, pig, chicken, and beef. We have focused our nary data also suggesting that isoforms of subunit VIla may not attention on tissue-specific variation of subu nits of this enzyme occur in rat. Figure 3 shows the N-terminal sequence data on in beef, where extensive protein sequence data of the heart this subunit from rat heart and rat liver and compares these data enzyme was available already, and in humans, as part of our with the sequences of the two isoforms of this subunit in beef. program on mitochondrial diseases. A review of our current The rat sequences are identical at several positions in which the understanding of tissue specificity of cytochrome c oxidase in two isoforms in beef are different. This suggests an important these different animals follows. difference in the tissue-specific pattern of isoforms of cytochrome ISOFORM S OF CYT OCH ROME c OXIDASE 531

RAT HEART FENKVP EK Q K LF Q E DNGMP VH L KG

RATUVER F EN KVPEK Q KLF QE DNGMPVHL

BEEFHEART F KQK LF Q E KG E DN"[:JP VHL BEEFLIVER F ENKVPE KQK LF Q E DNGIPVHL KG

Fig. 3. N-ter minal seq uences of subunit Vila of cytochrome c oxidase from rat and beef tissues. After electrophoresis of pu rified enzyme (as described in Figure I), subunits were transferred to po lyvinylidene difl uoride memb ranes as described (18 ). After staining, the blots were air dried and sto red at - 20·C. Polypeptides of int erest were cut from the membran e and sequenced in a gas phase protein sequence r equ ipped with an on line phenylthioh ydantoin analyzer. c oxidase in rat compared with other mammals such as beef, LLHHB B MM described in more detail below. Beef Our studies of the beef enzym e began with a comparison of the subunits of liver cytochrome c oxidase with those of the heart enzyme, for which sequence data were available (19). Subunits IV, Va, Vb, Vlb, VIc, VIlb, and VIlc in liver all had the same N-terminal sequences as those for heart and , to date, we have. no evidence for isoforms of any of these subunits. I However, any of these subunits may have a second isoforrn, differing in only a few amino acids at the C terminus, and there may be as yet unidentified isoforms of any subunit present in low amounts in one or more tissues that are not identified by protein chemical analysis of the isolated enzyme. Three subunits, VIa, VIla, and VIII, migrated differently when II heart and liver enzyme were compared by SDS-PAGE (I 9). N III terminal sequencing of the liver forms of these three subunits showed homologies, but many amino acid differences, compared to the corresponding heart subunits. We have since cloned cDNA encoding the heart form and liver form of subunit VIII; the mature proteins show only 52% hom ology to each other and the IV leader sequences only 40% homology (20). The differences in the two isoforms reside mainly in the N-terminal hydrophili c part of the polypeptid es. Recentl y, we have examined the subunit profile of cytochr ome Va c oxidase from a number of bovine tissues. The migrations of Vb subunits VIa, VIla, and VIII in skeletal muscle are the same as in heart , and the migrations of these three subunits in kidney VIa (not shown) and brain are the same as in liver (Fig. 4). N VIb terminal sequence analysis demonstrates clearly that brain and VIc kidney isoforms of these subunits differ from their counterparts ---- in heart , and that the skeletal muscle isoform s differ from those in liver. The results also suggest that the heart and muscle isoforms are identical, as are those of liver, kidne y, and brain VII a (Fig. 5). Althou gh protein sequencing identifies the main isoform of a parti cular subunit, it does not preclude the presence of small VII b, C amounts of the alternative isoform in the same tissue. We VIII excluded this possibility for the heart isoform of subunit VIa in tissues other than heart and skeletal muscle, using a MAb specific to the beef heart isoform . As seen in Figure 6, there is no (L), detectabl e cross-reactivity with cytochrome c oxidase from beef Fig. 4. Comparison of the subunit structure of beef liver heart (H). brain (B), and skeleta l muscle (M) cytoc hrome c oxida se. showing liver, kidney , or brain, thus excluding the presence of small differences in migration amounts of the heart isoform in these tissues. of subunits Via, Vil a, and VIII. Methods as in Northern analysis using cDNA for the two isoform s of rat Figure I, except that subunits were sepa rated on a 18.75% po lyacryl amide gel co ntai ning 6 M urea. subunit VIa (kindly provided by Dr. B. Kadenbach) showed that, as in rat, the mR NA for the H form of this subunit is present in bovine heart and skeletal muscle only, but the L form is present Humans. Our knowledge of the tissue specificity of human in all tissues (data not shown). The same pattern of expression cytochro me c oxidase is rudimentary. We have found that a was found for subunit VIII: the H form is only present in heart majorit y of the polyclonal antibodies and several of the MAb and skeletal muscle tissue, whereas the L form is present not made against the bovine enzyme subunits react with human only in brain, liver, and kidney, but in heart and skeletal muscle cytochrome c oxidase. Figure 8 shows a typical immunologic as well (Fig. 7) (20). However, the subunit profiles of cytochro me experime nt in which mitochondria from different tissues were c oxidase (Fig. 4) show little or no incorporation of the L form treated with subunit-specific antibodies using the Western blot of subunit VIII into funct ional enzym e in heart or muscle, a ting technique. result confirmed by careful analysis including protein sequencing The polyclonal antibody to subunit VIa reacts well with iso studies (20). These data suggest that the expression of the H lated beef heart cytochrome c oxidase but only poorly with the isoform of cytochrome c oxidase is cont rolled transcriptionally beef liver enzyme, establishing that this antibody can distinguish but the expression of the L isoform is regulated posttranscrip between the Hand L isoforms. In humans, the antibody reacts tionall y. strongly with heart and skeletal muscle and less well but signifi- 532 KENNAWAY ET AL.

VI a 8< Wha t is not clear is how many of the nuclear coded subunits have different isoforms or how many isoforms th ere are of a ny HEART I I Y £ K sub unit. Moreover, it remains possible and even like ly, given the SKELE TAL MUSCLE above da ta, th at th e pattern of tissue specificity differs in different m ammals. The lim ited data on humans provide evide nc e that th ere are at least two diffe rent forms of subunits VIa and VIla LIVER G.O-SA Mern AeP with an H and an L form of these sub units, as in beef. More BRAIN SSG AH GEE G - - S A RM L LP ...... definitive I\:SUJiS wiii req uire th e isolatio n of hu man cytochrome tU DNl: Y SSCAn cr.r.-- -S II. It M ---" - -- c oxidase from d iffercnt tissues and the seq ucncing ofco m ponent polypeptides, Functionul signilicaru:« (!I" tissue-snccific isofon ns of (FIO• VII (1 chrome c oxidase. From the above review ofthe sub un it str ucture or cytochrome c oxidase and th e tissue-spec ific exp ression or ,. L N AV AL x 0 K L r 0 I: " N G fL1 P II S K • isoforms or so me subunits. two qu estion s eme rge: I): What are Mu s e u : r E N V x 0 )( I. .. 0 t: 0 N G l:J I' V II LK the fu nctions of the nuclear coded subunits of th is enzyme'! a nd 2) W ha t is the advantage ofth e tissue specificity orsome ofth ese

subunits? f f K O' L r 0 " n N v n L' HRAIN r e N K V I-' l: K 0 K I. r IJ E n N c I I' vi Seve ral possible roles (or the nuclear coded subunits of cyto chrome c ox idase have bee n suggested . These include regul ation KI ONr.Y • r r. NKVP f: K 0 )< l. r 0 r. n N r; I P vJ - -_ . - -_.._ - - of the rate of electro n tran sfer or prot on pumping function by effec tors such as nuclcotidcs or horm o nes (24). The evidence for

VJl l th is is weak at the momen t, although ATP may alter the affi ni ty •• ofthe enzyme lor cytochrome c at low co nce ntra tions of substrate (25) . It is not clea r if there is an advantage in the tissue-specific II s '" A I p • x T P T S " .-••:oj ; -.--oJP1 expression of iSO «Jl'IllSorcytoc hro me c ox idase subu nits. II'th ere SKt: U :l AL MUSCL t; 1 GLW _ IL------is, it presumably relates to the functio n of the subunit in the complex. T hus. it see ms likel y that the tissue-specific expression

1. 1VER 1 ';;-; K f' A I c L T of th ese isoforrns, and thei r apparently variable patterns or expressio n in mam mals. co uld reflect d ifferences in regulatory UJ{AIN I 1\ 5 K f' I' It t; 0 L G T M LJ I '" 1 G rncch unisrns of the enzyme in different tissues. Stu dy of patients KIONt:Y I II SK P r It t: 0 L G 1 M lJ I A ! G with cytochrome c ox idas e defi cien cy will ultimat ely help to Fig. 5. Nvtcrm inu l sequen ces or subunits Vi a. Vila. and VII I fro m clarify these issues. different beef tissues. Method s as in Figure :I. H L M K 8 ca ntly with liver. bra in . and kid ney (taking into account that there are di fferen t amounts of cytochrome c oxidase in these different preparations as judged by a ntibod y to subunit IV). It is also apparent tha t subunit Via in skeletal muscle m igrates slightly mor c slowly than th at in bra in, kidney, or liver. Intriguingly. with the an ti-Via antibody. the reacti ve band in heart is broad . and close inspec tion ofth e gel shows two species present, a heavy band co migrating with su bu nit Via of" m uscle and a light er band corn igrating with subunit Via or liver, brain, and kid ney . Al th ough it is possib le th at the lighter ba nd represents a degradation product of" the heart isoform, th is is un likely, inasm uch as it has been seen in several preparations of" heart, but not mu scle, mi toch on dria, inc luding so me iso lated in the pre sence of" protease inh ibit or s. T hese da ta im ply that th ere are at least two isoforrn s ofsubuni t Via in human tissues, II an d L form s, as in bee r. with VIa the H form pre sent in heart and skeletal m uscle and the L form found in brain. kid ney . and liver, an d possib ly also in heart. The a ntibody aga inst subunit Vila reacts well with bee f" heart enzyme but very poorly with liver. The anti-Vila antibody also reacts well with human heart a nd skeletal muscle but no t with liver. kidney. or brain. ind icat ing that in hum an s th ere arc two isoforrn s of" this subunit with ap parently the sa me tissue d istri bution as in beer (Fig . R). Our antibody aga inst bovine heart subunit VIII is one ofthe few beefantibodies that does not react with the hum an enzyme, and it was not possib le to assess tissue spec ificity o f this sub uni t im m uno logically. However, subu nit VIII of human heart cytochrome c oxidase has bee n isolated recently and seq ue nced (22). This po lypep tide is remarkably similar to bo vine liver subunit VIII and much different from the Fig. 6. Western hlot showing reaction or MAh to the II limn or beef bovine heart isofo rrn. T he protein seq ue nce obtained by va n heart subu nit Via with cytoc hro me c oxida se fro m beef hc.nt ( II). liver Kuilcnburg ('I al. (22) is ide ntical to th e predicted product of th e ( I .). skeletal m usclc (M). kidncy (1\' ). and brain (8). Each lane was loaded cDNA cloned from human liver (23). T he mig ration of sub unit with 25 Ilg or enzyme, Elec tro pho resis was performed as described in VIII from hum an heart and liver is th e sa me. also suggest ing that Figu re I. a nd pol ypeptides tra nsferre d to nitrocellulose mem bra nes ac there may be on ly one form ofsub uni t VIII in hu ma ns. cording to Towhin ('I (/1, (21 ). Bind ing or primary antibody was detected In sum mary. th ere is un equivocal evidence of isoforrn s of" by reaction with a lkaline ph osphatase co nj uga ted goat antimousc Ig( ; several subunits ofcytoc hrome c oxidase in rat , beef: a nd hu man. antibod ies. ISOFORMS OF CYTOCHROME c OXIDASE 533

H K L B M H K L B M

650 -

4 00 -

A B Fig. 7. Northern blot showi ng th e patt ern of transcript ion for the heart and liver for ms of cytochro me c oxida se subunit VIII in various beef tissues. A. 20 Jlg of total RNA from bovine heart (H), kid ney (K) , liver (L), brain (B) , and skeletal mu scle (M) was separated through a 1.5% agarose de naturing gel containing 2 M for maldehyde, and tran sferred to a nylon filter as described (20). The blot was probed with high sp act antisense RNA transcripts synthesized in vitro with T7 or SP6 RNA polymerase from pGEM-3Z(Promega Corporation, Madison, WI) vectors containing the cDNA for the heart form of bovine subun it VIII. Hybridiz ation and washing were performed as describ ed (20) and the washed membrane exposed at -70·C to Kodak XAR-5 film for 7- 10 d in the presence of a Cro nex intensifying screen. B. identical blot probed with a 120-b p Nco I generated fragme nt for the liver form of bovine subunit VIII. Relati ve sizes of the hybridizing RNA species are shown. Reprodu ced from J Bioi Chern 20, with permission.

BEEF HUMAN VARIABLE PRESENTATIONS OF PATIENTS WITH .. • .. CYTOCHROME c OXIDASE DEFICIE NCY IN RELATION TO M H L H LMK B TISSUE SPECIFICIT Y OFENZYM E From a clinical standpoint, to what extent can our understand ing of tissue specificity explain the variability in severity and tissue distribution of mitochondrial diseases caused by cyto v chrome c oxidase dysfunction? If a correlation can be made between the tissue distribution of disease and the tissue distri bution of isoforms of the enzyme, then the molecular analysis of a variety of mitochondrial diseases will be greatly facilitated. The many and varied clinical phenotypes associated with ,I cytochrome c oxidase deficiency have recently been discussed VI a (5). For the purposes of this review, they can be divided into two major categories: multisystem diseases such as myoclonus epi lepsy with ragged red fibers, Leigh's syndrome (subacute necro tizing encephalomyelopathy), or Kearn s-Sayre syndrome, and VII a diseases confined to one or a few tissues, such as progressive external ophthalmoplegia or fatal infantile mitochondrial my VI]I opathy with lactic acidosis. In Kearns-Sayre syndrome, which is almost always sporadic , and in some patients with progressive external ophthalmoplegia, Fig. 8.Western blot showing reaction of polyclonal antibodies to beef deletions of mtDNA have recently been described (26, 27). heart cytochrome c oxidase subunits IV, VIa, VIla, an d VIII with beef Myoclonus epilepsy with ragged red fibers, on the other hand, skeletal mu scle mitochon dria (M, 20 ug), heart (H , 3.0 ",g), and liver (L , frequently shows a maternal inheritance pattern (28), also sug 3.0 Jlg) cytochrome c oxidase and human heart (H, 90 Jlg) , liver (L, 200 gesting a defect of mtDNA in these patients. In either of these Jlg), skeletal mu scle (M, 100 Jlg), kidn ey (K , 160 Jlg), and brain (B, 200 disorders, differential tissue involvement could reflect the differ.. Jlg) mitochond ria. Subunits were separated on an 18.5% gel containing ent energy requirements, or the distribution of norm al and 6 M urea (9), blotted onto nitrocellulose (2 1), and bound antibodies mutant mtD NA in different tissues. Other multisystem diseases visualized by reaction with avidi n-biotinylated alkaline phosphatase com associated with cytochrom e c oxidase deficiency, such as Leigh's plex bo und to biotinylated protein A. syndrome, show an autosomal recessive pattern of inheritance, 534 KENNAWAY ET AL.

Table I. Respiratory chain activities in mitochondria* Succinate: NADH:0 2 NADI-I :Q, cytochrome Succinate Cytochrome oxidoreductase oxidoreductase c reductase dehydrogenase c oxidase Skeletal musclebiopsy Patient 147 109 381 539 398 Controls (n = 9) 255 ± 50 173 ± 36 6 18 ± 172 606 ± 93 771 ± 250 (196- 298) (107-220) (228- 822) (46 1- 72 1) (356- 1085) Heart autopsy Patient 42 98 143 353 29 Controls (n = 9) 236 ± 67 147 ± 104 213 ± 55 407 ± 94 440 ± 131 (153-353) (55-322) (119- 316) (258- 554) (303-707) * Activities are expressed as nmol/ rng protein/min . Controls are mean ± SD with range in parentheses. Mitochondria were isolated from tissues by the method of Bookelman et al. (30) after mild trypsin digestion in the case of skeletal muscle. Activities of rotenone-sensitive NADH:oxygen oxidoreductase (31), NADH: ubiquinone oxidoreductase (31), succinate.cytochrome c reductase (32), succinate dehydrogenase (32), and cytochrome c oxidase (33) were measured according to published procedures. which could be explained by a mutation of a nuclear encoded 1 2 3 4 5 6 I 8 subunit that is not expressed in a tissue-specific manner. Diseases in which cytoch rome c oxidase deficiency is lim ited to one or a few tissues, on the other hand, are the ones most likely to result from mutati on of a tissue-specific subunit. In cluded in th is gro up are patients with severe infa ntile mitochon II drial myop athy and lactic acidosis. In some cases, the disease is --'._---,-- spo nta neously reversible, suggesting that develop me ntally regu lated isoforms of cyto chrome c oxidase may also exist, although - ..--flli....._....- .. IV no evidence for such isoform s has been obta ined in mammalian tissues (29). In other cases, the disease is fata l; it may be restri cted to skeletal muscle or it may be associated with renal dysfunctio n V a or card iomyopathy. A patient th at we have studied recently, in which cardiac ma nifest ations were predomina nt , is describ ed V l b briefly below. Fatal cardiomyopathy associated with cytochrome c oxidase Vic deficiency: case report. N.J . presented soon afte r birth with hypotonia, seizures, and lactic acidosis, after a pregnancy in which decreased fetal movement was noted. An echocardiogram, normal on d 6, showed massive bive ntricular hypert rophy by d VIf a 22. She died 2 d later of cardiac failure. Muscle biopsies at 8 and 21 d were normal by light and electron Fig. 9. Antibody bindingto subunits of cytochrome c oxidase in heart microscopy. Liver histology was also normal. At autopsy, elec mitochondria from patient N.J. Lane 1, beef, 25 ug: lanes 2,4, 5, 6, 7, tron microscopy ofcardiac muscle showed displacement of myo and 8, human controls, 120 ug: lane 3, patient, 120 ",g. Conditions were fibrils by greatly increased nu mbers of mitoc hondria, many of as described in Figure 8. which had ab normal shapes and abnorma lly arranged cristae. Cytochrome c oxidase activity was severely reduced in heart and was in the low-normal range in mu scle (Table I). It was also 1 2 3 4 dim inished in liver and kid ney «25% of controls, data not shown). The low acti vity of NADH:02 oxidored uctase in heart l' presumabl y reflects the cytoc hrome c oxidase deficiency. Cyto chrome spectra revealed co mplete absence of cytochrome aa, in IV heart mitocho ndria, whereas it was 37% of norm al in skeletal muscle. Western blot ana lysis (Fig. 9) showed marked deficien cy of cytochrome c oxidase subunits II,Vlb, Vic, and VIla, as well as " III and VIa (not shown), with relative spa ring ofsubunits IV and Va in the patient's heart, whereas in skeletal m uscle (Fig. 10), VI a, b only a m ild generalized deficiency of subunits was seen. This pat ient was unusual in several respects. Her clin ical VI c presentati on was distin ct from previously reported patients with ,I cytochro me c oxidase deficiency, with cardiac involvem ent pre dominating and relative spari ng of skeletal muscle. Altho ugh patients with fatal infantile m itochondrial myopathy may have VII a associated cardiomy opathy, the deficiency in skeletal m uscle is usually as severe or mo re severe th an in heart (34, 35). Expression of th e defect in liver as well as kid ney is also unusual, except in Fig. 10. Antibody binding to subunits of cytochrome c oxidase in Leigh's syndrome (35), and these patie nts do not normally skeletal muscle mitochondria from patient N.J. Lane 1. patient, 100 ",g: present at such an early age. lanes 2, 3, and 4, human controls, 100 ug. Conditions wereas described Th e findings in this patient, an d reports of variability of tissue in Figure 8. ISOFORMS OF CY TOCHROME c OXIDAS E 535 involvement in other patients with mitochondrial myopathy due and structural simi larit ies betwee n the polypeptides of yeast and beef heart to cytochrome c oxidase deficiency, suggest an important feature cytochrome c oxidase. FEBS Lett 207:1 1-1 7 11. Capa ldi RA. Hayashi H 1972 The polypeptide com position of cytochrome oftissue specificity not apparent from structural studies described oxidase from beef hea rt mitochondria. FEBS Lett 26:261-263 in previous sections:there may be considerable plasticity of tissue 12. Takamiya S. Yanamura W, Capaldi RA, Kennaway NG, Bart R, Senger s RCS. specificity of this enzyme in humans. Indeed, one or more tissues T rijbels JMF. Ruitenbeek W 1986 Mitochondrial myopathi es involving the may show variability, and even co-expression of the cytochrome respiratory chain: a biochem ical analysis. Ann NY Acad Sci 488:33- 43 13. Kadenbach B, Buge U, Jarausch J. Merle P 1981 Composition and kinetic c oxidase isoforms that are present. Moreover, we speculate that properties of cytochrome c oxidase from higher eucaryotes. In: Pal mieri F. the relative amounts of the various isoforms may fluctuate with Quagliercllo E, Siliprandi N. Slater C (cds) Vectorial Reactions in Electr on developm ental stage and other cellular signals, and these and Ion Transport in Mitochondria and Bacteria. Elsevier. Am sterdam, pp amounts may accommodate to some degree to compensate for 11- 23 14. Merle P. Kaden bach B 1980 On the heterogeneity of vertebrate cytochrome c defects in any particular subunit isoform. oxidase polypept ide chain composition. Hoppe-Seylers Z Ph ysiol Che m 36 1:1257- 1259 SUMMARY AN D CONCLUSIONS 15. Stroh A. Kadenbach B 1986 Tissue-specific a nd species-specific distribut ion of -SH groups in cytoc hro me c oxidase subunits. Eur J Biochem 156:199- 204 16. Schlerf A, Droste M, Wint er M. Kadenbach B 1988 Characterization of two In this review, we have summarized the structure, function, different genes (cDNA) for cytochrome c oxidase subunit VIa from heart and biogenesis of mammalian cytochrome c oxidase, presented and liver of the rat. EMBO J 7:2387 - 2391 data supporting the existence of tissue-specific isoforms of several 17. Kuhn-Nentwig L. Kad en bach B 1985 Isolation and propert ies of cytochrome subunits, reviewed possible mechanisms underlying the abnor c oxidase from rat liver and quantificat ion of immunological differences between isozymes from various rat tissues with sub unit-specific antisera. Eur mal phenotypes of patients with cytochrom e c oxidase deficiency, J Biochem 149:147-1 58 and discussed the insights that such patients provide about 18. Matsuda ira P 1987 Sequ ence from picom ole quantities of proteins electroblot regulation ofthe synthesis of different forms of this enzyme. ted onto polyvinylidene difluoride mem bran es. J Bioi Chern 262:10035 It is clear that tissue-specific isoforms of several subunits of 10038 19. Yanamu ra W. Zh ang Y-Z. Taka miya S. Capaldi RA 1988 Tissue-specific cytochrome c oxidase do exist in mammals, including hum ans. differences between heart and liver cytoch rome c oxidase . Biochemi stry However, the lack of conservation of the patterns of tissue 27:4909- 49 14 specificity between species leaves open the question of the func 20. Lightowlers R. Ewart G, Aggeler R, Zhang Y-Z . Calavetta L, Capaldi RA 1989 tional importance of these patterns. Isolation and chara cteri zation of the cDN As encoding two isofo rms of sub unit CIX of bovine cytochrome c oxidase . J BioI Chern 265:2677- 268 1 It is likely that a mutation of one ofthe tissue-specific isoforms 2 1. Towbin H. Staehelin T. G ordo n J 1979 Electrophoretic transfer of proteins of cytochrome c oxidase underlies the tissue-specific expression from polyacrylam ide gels to nitrocellulose sheets: procedure and some ap of this enzyme deficiency in some patients. However, in most plicat ions. Pro c Natl Acad Sci USA 76:4350- 4354 patients studied to date, the phenotype does not appear to 22. van Kui lenburg ABP , Muij sers AO. Dernol H. Dekker HL. Van Beeumen JJ 1988 Human heart cytochrome c oxid ase subunit VIlI : purifi cat ion and correlate with isoform distribution as we understand it, and much determination of the complete am ino acid sequence. FEBS l ett 240:127 more needs to be learned about the expression of the human 132 enzyme before such correlations can be made. In particular, it is 23. Rizzutt o R. Nakase H. Da rras B. Fabri zi GM . Mengel T . Walsh F. Kadenbach important to understand the regulation of synthesis of the various B, DiMauro S, Franke U, Schon E 1989 A gene specifying sub un it VII of hum an cytochrome c oxidase is localized to chromosome I I an d is expressed isoforms. in both mu scle and non-mu scle tissues. J Bioi Chem 264:1059 5- 10600 Finally, description of the molecular defect in more than a few 24. Kaden bach B. Kuhn-Nent wig L. Buge U 1987 Evolu tion of a regulatory patients with cytochrome c oxidase deficiency will likely require enzyme: cytochrorne-c oxidase (complex IV). CUff Top Bioen erg 15:113 a substantial collaborative effort between clinicians, biochemists, 16 1 25. Bisson R. Schiavo G , Montecucco C 1987 ATP indu ces conformational and molecular biologists. Exchange of tissues, antibodies, and changes in mit ochondrial cytoc hrome c oxidase: effect on the cytochrome c cDNA between different laboratories will ensure the best use of binding site. J Bioi Chern 262:5992-5998 available resources, and accelerate progress toward understand 26. Holt IJ, Harding AE,Morgan- H ughes JA 1988 De letions of mu scle mitochon ing the molecular mechanisms that give rise to disorders of the drial DNA in patients with mitoch ondrial myopathies. Nature 331:717-7 19 27. Moraes CT , DiMauro S, Zeviani M. Lombes A, Shanske S. Miranda AF. respiratory chain in hum ans. Nakase H. Boni lla E, Werneck LC. Servidei S 1989 Mito chondrial DN A deletio ns in progre ssive external ophthalmoplegi a and Kearns-Sayre syn REFERENCES drome. N Eng! JMed 320:1293-1299 28. Rosing HS. Ho pkins LC. Wallace DC. Epstein CM . Weid enh eim K 1985 I. DiMauro S. Bon illa E. Zevian i M. Nak agawa M, DeVivo DC 1985 Mitochon Matern ally inherited mitocho ndrial myopath y and myoclon ic epilepsy. Ann drial myopath ies. Ann Neuro I17:52 1-538 Neurol 17:228-237 2. Morgan- Hughes JA. Schapira AHV, Coo per JM , Clark JB 1988 Molecular 29. Takam iya S, Lindorfer MA, Capaldi RA 1987 Purification of all thi rteen defects of NADH-ubiquinone oxidoredu cta se (co mplex I) in mit ochondrial polypept ides of bovine heart cytochrome c oxidase from one aliquot of diseases. J Bioen erg Biomembr 20:365-382 enzyme: characterization of bovine fetal heart cytochrome oxidase. FEBS 3. Rivner MH. Shamsnia M, Swift TR , Trefz J. Roesel RA, Carter AL, Yanamura Lett 218:277- 282 w, Hommes FA 1989 Kearns -Sayre syndrome and com plex II deficiency. 30. Bookelma n H. Trijbels JMF, Sengers RCA, Janssen AJM . Veerkarn p JH , Neurology 39:693- 696 Stadhouders AM 1978 Pyruvate oxidation in rat and human skeletal muscle 4. Kennaway NG 1988 Defects in the cytochrome be, complex in mit ochon drial mitocho ndria. Biochem Med 20:395 -403 diseases. J Bioenerg Biomembr 20:325- 352 31. Fischer JC, Ruit enbeek W, T rijbels JMF, Veerka mp JH , Stadh ouders AM , 5. DiMauro S. Zeviani M. Rizzut o R, Lombes A. Nakase H. Bonilla E. Mira nda Sengers RCA, Janssen AJM 1986 Estimation of NADH oxidat ion in human A. Schon E 1988 Molecular defects in cytochrom e oxidase in mitochondrial skeletal muscle mi tochondria. Clin Chim Acta 155:263-274 diseases. J Bioenerg Biom embr 20:353-364 32. Fischer J'C, Ruitenbeek W, Berden JA. Trijbels JM F, Veerk amp JH , Stad 6. Cap aldi RA, G on zalez-Halphen D. Zhang Y-Z, Yanamura W 1988 Complexity houd ers AM, Sengers RCA. Ja nssen AJ M 1985 Differenti al investigatio n of and tissue specificity of the mitochondrial respiratory cha in . JBioene rg the capacity of succinate oxidation in human skeletal muscle. Clin Chim Biomembr 20:291-31 I Acta 153:23-36 7. Clay VJ. Ragan CI 1988 Evidencefor the existence of tissue specific isoenzymes 33. Wharton DC, Tza goloff A 1967 Cytochrom e oxidase from beef heart mito of mit och ondrial NAD H dehydrogenase. Biochcm Bioph ys Res Co mmun cho nd ria. Methods Enzymol 10:245- 250 157:1423-1428 34. Zeviani M. Van Dvke D H, Servidei S. Bau serman Sc. Bonilla E. Beaumon t 8. Capaldi RA, Takarniya S. Zhang Y-Z. Go nzalez-Halphen D. Yan amura W ET, Sharda J, Vanderlaan K. DiMauro S 1986 Myopath y a nd fatal cardio 1987 Structure of cytochrorne-c oxidase. Curr Top Bioenerg 15:9 1-1 12 myopathy due to cytochrome c oxidase deficiency. Arch Ne urol 43:1198 9.Kadenbach B.Jarausch J, Hartmann R, Mer le P 1983 Separation of mam 1202 malia n cytochrome c oxid ase into 13 polypeptides by a sodium dodecyl 35. DiMauro S. Servidei S, Zeviani M. DiRocco M. DeVi vo DC. DiDonato S. sulfate-gel electrophoretic pro ced ure. Anal Biochem 129:517-521 Uziel G. Berry K, Hogan son G, Johnsen SD, Jo hnson PC 1987 Cytochrome 10. Capaldi RA, Go nzalez-Halphen D, Takamiya S 1986 Sequence homologies c oxidase deficiency in l eigh syndrome. An n Neurol 22:498-506