Subunit specific monoclonal antibodies show different steady-state levels of various cytochrome-c oxidase subunits in chronic progressive external ophthalmoplegia

ELSEVIER

Biochimica et Biophysica Acta 1315 (1996) 199-207

Biochi~ic~a et BiophysicaA~ta

Subunit specific monoclonal antibodies show different steady-state levels of various cytochrome-c oxidase subunits in chronic progressive external ophthalmoplegia Jan-Willem Taanman a.*, Miriam D. Burton b, Michael F. Marusich c, Nancy G. Kennaway Roderick A. Capaldi ~

b

Institute of Molecular Biology, UnicersiO"qf Oregon, Eugene, OR 97403, USA b Department of Molecular and Medical Genetics, Oregon Health Sciences UnicersiO', Portland, OR 97201, USA c Institute ofNeuroscience, Unicersitv of Oregon, Eugene, OR 97403, USA

Received 20 September 1995; revised 8 December 1995: accepted 19 December 1995

Abstract Monoclonal antibodies recognizing the mitochondrially encoded subunits I and II, and the nuclear-encoded subunits IV, Va, Vb and VIc of human cytochrome-c oxidase were generated. These antibodies are highly specific and allow the assessment of subunit steady-state levels in crude cell extracts and tissue sections. In the experimental human cell line 143B206, which is devoid of mitochondrial DNA, immunovisualization with the antibodies revealed that the nuclear-encoded subunits IV and Va were present in amounts close to that of the parental cell line despite the absence of the mitochondrially encoded subunits. In contrast, the nuclear-encoded subunits Vb and VIc were severely reduced in cell line 143B206, suggesting that unassembled nuclear-encoded subunits are degraded at different rates. In skeletal muscle sections of a patient with chronic progressive external ophthalmoplegia known to harbor the 'common deletion' in a subpopulation of her mitochondrial DNA, most cytochrome-c oxidase activity negative fibers had greatly reduced levels of subunits I, II, Va, Vb and VIc of cytochrome-c oxidase. The steady-state level of subunit IV, however, was less affected. This was particularly evident in cytochrome-c oxidase activity negative fibers with accumulated mitochondria ('ragged-red' fibers) where immunodetection with anti-subunit IV resulted in intense staining. The data presented in this paper demonstrate that the battery of monoclonal antibodies can be employed for diagnostic purposes to analyze steady-state levels of mitochondrially and nuclear-encoded subunits of cytochrome-c oxidase. Keywords: Cytochrome e oxidase; Mitochondrial protein; Immunohistochemistry: Monoclonal antibody: Chronic progressive external ophthalmoplegia; (Human)

1. Introduction Mitochondria play a central role in cellular energy provision by the process of oxidative phosphorylation. In this metabolic pathway, mitochondria generate ATP driven by a transmembrane proton gradient across the mitochondrial inner membrane. The proton gradient is sustained by the respiratory chain which couples electron transfer from

Abbreviations: bp, base pairs; CPEO, chronic progressive external ophthalmoplegia; kb, kilobase pairs; PBS, phosphate-buffered saline: SDS, sodium dodecyl sulfate. Corresponding author. Present address: Department of Clinical Neurosciences, Royal Free Hospital School of Medicine, Rowland Hill Street. London NW3 2PF, UK. Fax: +44 171 431 1577; e-marl: [email protected]. 0925-4439/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved SSDI 0 9 2 5 - 4 4 3 9 ( 9 5 ) 0 0 1 2 7 - 1

N A D H and F A D H 2 to molecular oxygen with proton pumping across the membrane. The respiratory chain is comprised of four multisubunit enzyme complexes: NADH:ubiquinone oxidoreductase (EC 1.6.5.3), s u c c i n a t e : u b i q u i n o n e o x i d o r e d u c t a s e (EC 1.3.5.1), ubiquinol:ferricytochrome-c oxidoreductase (EC 1.10.2.2) and f e r r o c y t o c h r o m e - c : o x y g e n oxidoreductase (cytochrome-c oxidase, EC 1.9.3.1) [1]. All subunits of succinate:ubiquinone oxidoreductase are encoded on nuclear D N A but some subunits of the other complexes are encoded on mitochondrial D N A [2]. Cytochrome-c oxidase deficiency is emerging as one of the most common defects in mitochondrial diseases. In mammals, this enzyme is composed of three mitochondrially encoded polypeptides (I, II and III) and ten

200

J.- W. Taanman et al. / Biochimica et Biophysica Acta 1315 (1996) 199-207

nuclear-encoded polypeptides (IV, Va, Vb, Via, VIb, VIc, VIIa, VIIb, VIIc and VIII). The mitochondrially encoded subunits are homologous to the three polypeptides found in bacterial cytochrome-c oxidases and constitute the catalytic core of the enzyme (for recent reviews see Refs. [3-5]). The function(s) of the nuclear-encoded subunits is enigmatic, although recent observations have clarified the role of one of these subunits. Kadenbach was the first to postulate that some of the nuclear-encoded subunits are involved in allosteric modification of cytochrome-c oxidase activity [6,7]. Support for this idea is provided by the presence of tissue and developmental specific isoforms of some of the nuclear-coded subunits (reviewed in Ref. [8]) which could optimize the enzymatic activity to the metabolic demands of different tissues. Recently, we demonstrated in yeast that subunit Via, which is expressed as isoforms in mammals, contains an ATP binding site and inhibits the electron transfer activity in response to high physiological concentrations of ATP [9]. The dual genetic origin and isoform complexity of the enzyme may explain the remarkable clinical and biochemical heterogeneity of cytochrome-c oxidase deficiencies, with different tissues being critically affected and with variable onset of symptoms (reviewed in Refs. [10-12]). With the exception of one patient with a 15-bp deletion in the gene for cytochrome-c oxidase subunit III [13], no molecular abnormality of any subunit of cytochrome-c oxidase has been shown to be associated with cytochrome-c oxidase deficiency. Partial defects of cytochrome-c oxidase activity, frequently accompanied by other respiratory chain abnormalities, are often associated with specific point mutations in tRNA genes or deletions in a subpopulation of the mitochondrial DNA (reviewed in Refs. [1416]). For instance, in patients with ocular myopathy as a key feature, as in chronic progressive external ophthalmoplegia (CPEO) and Kearns-Sayre syndrome, the focal defect of cytochrome-c oxidase in individual muscle fibers has been attributed to large-scale deletions or duplications in mitochondrial DNA. Reported deletions range in size but a deletion 'hotspot' between two 13-bp direct repeats results in the most common 5-kb deletion. This deletion includes the genes for subunit III of cytochrome-c oxidase, subunits 3, 4, 4L and 5 of the NADH:ubiquinone oxidoreductase complex and several unique tRNAs. A characteristic feature in these patients is a subsarcolemmal accumulation of morphologically abnormal mitochondria in muscle fiber sections, visualized as 'ragged-red fibers' with modified Gomori trichrome stain. Diagnosis of cytochrome-c oxidase deficiency is generally made by histochemical and biochemical analysis of the tissue involved, in concert with an analysis of mitochondrial DNA. Recently, we described the development of antibodies against subunit isoforms to study Mendelian-inherited, tissue specific deficiencies of cytocbrome-c oxidase [17]. Here, we report the development of monoclonal antibodies against other nuclear and mito-

chondrially encoded subunits of the enzyme to further dissect the various forms of cytochrome-c oxidase deficiency at the protein level.

2. Materials and methods 2.1. C y t o c h r o m e - c o x i d a s e

Human heart cytochrome-c oxidase was isolated in the laboratory of Dr. A.O. Muijsers (E.C. Slater Institute for Biochemical Research, Academic Medical Center, University of Amsterdam) according to published procedures [18]. The enzyme was dissociated in 3% ( w / v ) sodium dodecyl sulfate (SDS) and subunits were resolved on standard gelfiltration columns [19]. To remove excess of SDS, subunits I and II were precipitated by addition of 3 volumes of ethanol at room temperature. The protein pellet was washed with 80% ( v / v ) ethanol, dried, and dissolved in phosphate-buffered saline (PBS), 0.1% ( w / v ) SDS for immunization. Bovine heart cytochrome-c oxidase was purified as described [20] and stored in small aliquots at - 7 0 ° C until used for preparation of monoclonal antibodies against the nuclear-encoded subunits. To isolate single subunits, the enzyme was resolved on SDS-polyacrylamide gels [21] followed by electro-elution of subunits in 10 mM Tris-HCl (pH 8.6), 0.1% ( w / v ) SDS using an ISCO Electrophoretic Concentrator. 2.2. A n t i b o d i e s

Table 1 lists the immunogen, mouse strain and myeloma cell line used to make each hybridoma. In each case, mice were immunized intraperitoneally with immunogen (either 50 /zg holo-enzyme or 10 /xg isolated subunit) emulsified in Complete Freund's Adjuvant and then boosted intraperitoneally at monthly intervals with immunogen in Incomplete Freund's Adjuvant until high titers of specific antibody were identified with immunoblots. Final boosts of immunogen in saline were then delivered intraperitoneally at 4 and 3 days prior to splenocyte collection. Hybridoma cell lines were produced and cultured as previously described [22]. Standard HAT (hypoxanthine, aminopterin and thymidine) selection was used to select B A L B c / G N + P3X63-Ag8.653 hybridomas and B A L B c / G N + S P 2 / 0 hybridomas, whereas AAT (adenine, aminopterin and thymidine) selection [23] was used to select RBF/DnJ + FOX-NY hybridomas. Myeloma cell lines were obtained from the American Type Culture Collection. Hybridoma supernatants were screened for anti-cytochrome-c oxidase antibodies using solid phase binding assays with native holo-cytochrome-c oxidase of the appropriate species as target antigen [17]. Antibody binding specificities were further characterized by immunoblot analysis [8] of cytochrome-c oxidase or mitochondria from

201

J.-W. Taanman et al. / Biochimica et Biophysica Acta 1315 (1996) 199-207

various species. The specificities of the monoclonal antibodies against subunit Va and Vb (6E9-B 12-D5 and 16H12, respectively) were checked on immunoblots with bovine cytochrome-c oxidase separated on a polyacrylamide gel (16.5% T, 6% C) containing 6 M urea, under the conditions described by Sch~igger and Von Jagow [24]. On this gel system, subunit Va and Vb switch in position compared to their location on the gel system described by Kadenbach et al. [21]. The specificity of the monoclonal antibody against subunit VIc (3G5-F7-G3) was determined on immunoblots with isolated subunit VIb and bovine cytochrome-c oxidase extracted with Triton X-100. Treatment of the enzyme with this detergent removes subunit VIb (Y.-Z. Zhang and R.A. Capaldi, unpublished observations, [25]) which co-migrates with subunit VIc on SDSpolyacrylamide gels. Monoclonal antibodies were purified from serum-free hybridoma culture supernatants as described by Marusich et al. [26], or used as untreated cell culture supernatants. The development of monoclonal antibodies specific for bovine (and human) cytochrome-c oxidase subunit IV has previously been described [17]. Monoclonal antibodies against subunits I, II, IV and VIc have been made generally available (Molecular Probes, Inc,). 2.3. I m m u n o b l o t analysis

The cell line 143B206 [27], which has been completely depleted of mitochondrial DNA ( p ° cells) by long-term exposure to ethidium bromide, and the parental osteosarcoma cell line 143B - T K - were kindly provided by Dr. G. Attardi (Division of Biology, California Institute of Technology, Pasadena, CA). Cells were lysed in 5% ( w / v ) SDS. Protein concentrations were determined according to Smith et al. [28] with the bicinchoninic acid kit (Pierce) using bovine serum albumin as standard. Similar amounts of protein from the two cell lines were resolved on 5.5 M urea, 10-20% polyacrylamide gradient minigels [29], blotted onto poly(vinylidene difluoride) (PVDF) membranes (Millipore), and strips of the immunoblots were incubated with antibodies (ID6-E1-A8 and 5D4-F5, 1 /xg I g G / m l ; 12C4-F12, 0.1 /.~g I g G / m l ; 10G8-D12-C12 and 6E9-B12-

Table l Anti-cytochrome-coxidase monoclonal antibodies Subunit binding Immunogen specificity Subunit I Subunit I Subunit II Subnnit IV Subunit Va Subunit Vb Subunit VIc

human subunit I human subunit I human subunit II bovine holo-enzyme ~ bovine subunit Va bovine holo-enzyme bovine holo-enzyme

d Bovine subunit IV in final boost.

D5, 2 /zg I g G / m l ; 16H12, 50 /zl culture medium/ml; 3G5-F7-G3, 5 /zg I g G / m l ) and developed using the 5bromo-4-chloro-3-indolyl phosphate and nitro blue tetrazolium-alkaline phosphatase reaction as described [8]. 2.4. H i s t o c h e m i s t ~ and i m m u n o h i s t o c h e m i s t ~ '

Patient and age-matched control skeletal muscle specimens, obtained by open biopsy, were immediately frozen in liquid nitrogen-cooled isopentane and stored at -70°C. The patient suffered chronic progressive external ophthalmoplegia (CPEO) and was known to carry the 'common' 5-kb mitochondrial DNA deletion in a relatively high percentage of her skeletal muscle mitochondrial DNA. The samples were obtained, with permission, following protocols approved by the OHSU Human Studies Committee. Muscle sections (8 /zm thick) were mounted on poly-Llysine coated slides. Demonstration of the activity of cytochrome-c oxidase and succinate dehydrogenase were performed as described [30,31]. Fiber type was determined by myofibrillar actomyosin ATPase staining at pH 9.4 after preincubation at pH 4.5 [32]. Ragged-red fibers were stained with modified Gomori trichrome [33]. Muscle sections (8 /zm thick) used for all immunostaining studies were air-dried, followed by fixation for 5 min in 4% ( v / v ) methanol-free formaldehyde, 0.1 mM CaCl 2, 0.1 M sodium phosphate buffer (pH 7.2). Prior to immunostaining, the sections were treated with 0.6% ( v / v ) H 2 0 2 in 80% ( v / v ) methanol for 30 min at room temperature. Sections were pre-incubated in PBS supplemented with 1% ( w / v ) bovine serum albumin for 10 rain at room temperature. Sections were then incubated for 2 h in PBS, 1% ( w / v ) bovine serum albumin, containing primary antibodies. Primary antibodies were used at dilutions optimized to give consistent immunostaining in all fibers of the control muscle sections. Following subsequent incubation with biotinylated anti-mouse immunoglobulins (Vector Laboratories) for 1 h and incubation with avidin-biotinylated horseradish peroxidase complex (Vectastain ABC kit, Vector Laboratories) for 1 h, sections were reacted with 3,3'-diaminobenzidine (Sigma) in PBS. Controls included a 2-h incubation in PBS, 1% ( w / v ) bovine serum albumin, without primary antibody.

Hybridoma

Isotype

Mouse strain

Myelomacell line

l D6-El-A8 5D4-F5 !2C4-F12 10G8-D12-CI2 6E9-B12-D5 16H12 3G5-F7-G3

IgG2a~¢ IgG2aK IgG2a~c IgG2aK IgG2aK IgG2bK IgG2bK

RBF/DnJ RBF/DnJ RBF/DnJ BALBc/GN BALBc/GN BALBc/GN BALBc/GN

FOX-NY FOX-NY FOX-NY P3X63-Ag8.653 SP2/0 P3X63-Ag8.653 P3X63-Ag8.653

202

J.-W. Taanman et al. / Biochimica et Biophysica Acta 1315 (1996) 199-207

Coornassie

I-1 c

p*

I-2 c

p*

II C

IV

p*

c

Va

p*

c

Vb c p*

p*

VIc ¢ p*

~" ,,.

,,.

...........

Mr

_

18k

_

14k

-

6k

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

,.

Fig. 1. Immunoblot analysis of monoclonalantibodies against cytochrome-c oxidase subunits. Immunoblot strips with total cellular protein (25 /xg/strip) of the human control cell line 143B. TK- (c) and its derivative cell line 143B206 (p0) were exposed to the monoclonals:5D4-F5 (I-1), 1D6-E1-A8(I-2), 12C4-F12 (II), 10G8-C12-DI2(IV), 6E9-B12-D5 (Va), 16H12 (Vb) and 3G5-F7-G3 (VIc). One pair of strips was stained with Coomassie brilliant blue. Bands of interest are indicated by arrows. Migration of M r standards is indicated.

Individual fibers were followed through the serial cross-sections and each staining was visually rated proportional to the intensity of its punctuate appearance, rather than more diffuse background, from 0 (no staining) to 4 (most deeply stained), by two observers independently.

3. Results 3.1. I m m u n o b l o t analysis

Monoclonal antibodies against nuclear-encoded subunits of cytochrome-c oxidase were raised against the bovine enzyme. To obtain monocional antibodies against the mitochondrially encoded subunits, the isolated subunits of human cytochrome-c oxidase were used. An overview of the monoclonal antibodies employed in this study is given in Table 1. We used immunoblot strips with total cellular protein from the human control cell line 143B. T K - and its derivative cell line 143B206, which lacks mitochondrial D N A ( p ° cells) [27], to test the (cross-)reaction and specificity of the monoclonal antibodies (Fig. 1). Of the two monoclonal antibodies raised against subunit I, one

antibody (1D6-E1-A8) recognized a single band on the strip with control cell lysate, migrating slightly slower than the 43k marker, and showed no reaction on the strip with p0 cell lysate. The other antibody (5D4-F5) predominantly reacted with a band co-migrating with the only band recognized by 1D6-EI-A8 on the strip with control cell lysate. Monoclonal 5D4-F5 did not identify this band on the strip with p0 cell lysate. These results clearly indicate that both monoclonals recognize human cytochrome-c oxidase subunit I on immunoblots because both antibodies react with a band in control cells migrating slightly slower than the 43k marker (which is the position of subunit I) and neither antibody reacts with this band in cells lacking the gene for subunit I. However, the reaction of 1D6-EI-A8 is more specific than that of 5D4-F5 on immunoblots. The monoclonal antibody raised against cytochrome-c oxidase subunit II (12C4-F12) reacted strongly with a single band on the strip with control cell lysate, migrating between the 29k and 14k markers (the position of subunit II), but failed to react with any protein band on the strip with p0 cell lysate (Fig. 1). This result clearly demonstrates the specificity of monoclonal 12C4-F12 for subunit II. Previous studies have revealed that monoclonal antibody 10G8-D12-C 12, raised against bovine cytochrome-c

Fig. 2. Sections of skeletal muscle of a patient with CPEO. (a) Serial cross-sections. Panels show: histochemical staining for myofibrillar actomyosin ATPase activity for fiber typing (ATP), histochemical staining for cytochrome-c oxidase activity (COX), and immunohistochemical staining with monoclonal antibodies 5D4-F5 (I), 12C4-F12(II), 10G8-C12-D12(IV), 6E9-B12-D5 (Va), 16H12 (Vb) and 3G5-F7-G3 (VIc). Marked fibers are: type 1, normal cytochrome-coxidase activity (A); type 1, reduced cytochrome-coxidase activity (B); type 2b, reduced cytochrome-coxidase activity (C); type 2b, no cytochrome-c oxidase activity (D); type 2a, normal cytochrome-c oxidase activity (E); type 2a, reduced cytochrome-c oxidase activity (F); type 2a, no cytochrome-c oxidase activity (G). (b) Longitudinal section histochemically stained for cytochrome-c oxidase activity.

J.-W. Taanman et al./Biochimica et Biophysica Acta 1315 (1996) 199-207

oxidase subunit IV, cross-reacts with subunit IV from human and chicken [8,17]. Fig. 1 illustrates that this monoclonal specifically recognizes the human subunit in a

lal

Ibl

203

total cell lysate. Similar amounts of protein were loaded on the two gels with either control or p0 cell lysate, as indicated by the two strips stained with Coomassie brilliant

204

J.- W. Taanman et al. / Biochimica et Biophysica Acta 1315 (1996) 199-207

blue (Fig. 1). The signal of the anti-subunit IV monoclonal is only slightly less on the strips with p0 cell lysate compared to the strips with control cell lysate. Thus, despite the absence of mitochondrially encoded cytochrome-c oxidase subunits in p0 cells, subunit IV is present in almost normal amounts. The monoclonal antibodies against subunit Va (6E9B 12-D5) and subunit Vb (16H 12) predominantly recognize one protein band on the strip with control cell lysate, migrating between the 14k and 6k markers (the position of subunits Va and Vb). The specificity of these monoclonals for either subunit Va or Vb was determined on highly resolving SDS-polyacrylamide gels with bovine cytochrome-c oxidase as detailed in Section 2. Whereas incubation of a p0 cell lysate strip with the anti-subunit Va monoclonal resulted in only a slight decrease in staining, incubation with the anti-subunit Vb monoclonal resulted in a faint band compared to the predominant band on the control strip. Apparently, like subunit IV, subunit Va is present at almost normal steady-state levels in p0 cells. Steady-state levels of subunit Vb, however, are severely decreased in cells devoid of mitochondrial DNA. Monoclonal antibody 3G5-F7-G3 was raised against bovine cytochrome-c oxidase. On immunoblots with the bovine enzyme, 3G5-F7-G3 reacted with a band that contains the closely migrating subunits VIb and VIc (not shown). The specificity of this monoclonal for subunit VIc was determined with bovine cytochrome-c oxidase preparations in which subunit VIb was removed (see Section 2). Moreover, we have demonstrated that this monoclonal still recognizes a protein band with the correct M r of subunit VIc in enzyme preparations incubated with trypsin, which selectively digests the subunits Via and VIb [9]. Incubation of a human control cell lysate strip with monoclonal 3G5-F7-G3 resulted in an intensely stained protein band co-migrating with the 6k marker (position of subunit VIc) and less intensely stained higher M r protein bands (Fig. 1). The same pattern of higher M r protein bands show up on the p0 cell lysate strip but the intensely stained lower M r protein band on the control strip is absent (Fig. 1). These results strongly suggest that the intensely stained lower M r protein band represents subunit VIc and that the amount of this subunit is greatly reduced in cells lacking mitochondrially encoded subunits. 3.2. I m m u n o h i s t o c h e m i c a l a n a l y s i s

To explore the use of the monoclonal antibodies in immunohistochemical studies, we applied the antibodies on serial skeletal muscle sections of a CPEO patient known to harbor the 'common deletion' in a subpopulation of her mitochondrial DNA. First, we followed 30 fibers on serial cross-sections of an age-matched control muscle specimen. Cytochrome-c oxidase activity was demonstrated in all control muscle fibers. The staining intensity indicated that the activity was relatively high in type 1

(slow-twich, oxidative) fibers, followed by type 2b (fasttwitch, glycolytic) and type 2a (fast-twitch, oxidative-glycolytic) fibers (not shown). Immunohistochemical detection of individual subunits of cytochrome-c oxidase in control muscle with the monoclonal antibodies listed in Table 1 showed particulate staining in all fibers that was more pronounced in type 1 than in type 2a and 2b fibers (not shown). In the patient's specimen, 72 fibers were studied in serial cross-sections. One set of serial sections in which fibers were followed is shown in Fig. 2a. The myofibrillar actomyosin ATPase profile displayed the normal checkerboard distribution of type 1 (dark), type 2b (intermediate) and type 2a (light) fibers (cf. Ref. [34]). Sections stained with modified Gomori trichrome revealed that less than 1% of the fibers was ragged-red (not shown). Approximately 30% of all fibers, including all ragged-red fibers, exhibited no cytochrome-c oxidase activity in cross-sections. In longitudinal sections, some fibers showed virtually no cytochrome-c oxidase activity but other fibers had normal activity or the enzyme defect was segmentally expressed (Fig. 2b). The immunohistochemical examination of serial crosssections with the cytochrome-c oxidase subunit specific monoclonal antibodies revealed a scattered pattern of staining (Fig. 2a). Staining intensities were visually rated from 0 (no staining) to 4 (darkest staining) in all 72 fibers studied in cross-sections. The semi-quantitative data for type 1 fibers are compiled in Fig. 3, where they are subdivided into three categories: those with relatively normal cytochrome-c oxidase activity, those with reduced cytochrome-c oxidase activity, and ragged-red fibers which had no apparent cytochrome-c oxidase activity. In type 1 fibers with relatively high cytochrome-c oxidase activity levels (Fig. 2a, fiber A; Fig. 3a), immunodetectable proteins were mostly present at relatively high levels and no defects were found for succinate dehydrogenase. Type 1 fibers with reduced cytochrome-c oxidase activity (Fig. 2a, fiber B; Fig. 3b) showed in general an associated loss of immunodetectable proteins and slightly reduced succinate dehydrogenase activity. Type 1 fibers virtually lacking cytochrome-c oxidase activity are not present in the section shown in Fig. 2a but all were ragged-red and showed high levels of succinate dehydrogenase activity (Fig. 3c). In these cytochrome-c oxidase activity negative fibers, the cytochrome-c oxidase subunits I, II, Va, Vb and VIc were undetectable with the monoclonals in some fibers but low levels were demonstrated in other fibers. Immunodetection with the monoclonal against subunit IV, however, resulted in intense staining of all ragged-red fibers. The correlation between enzyme activities and immunodetectable proteins in type 2a and type 2b fibers (Fig. 2a, fibers C - G ) followed the same trend as described for type 1 fibers. Of the 25 type 2a fibers studied, none of the 12 fibers lacking cytochrome-c oxidase activity was raggedred, whereas of the 16 type 2b fibers studied, only 1 of the

J. - W. Taanman et al. / Biochimica et Biophysica A cta 1315 (1996) 1 9 9 - 2 0 7

41nn 3 2

1

e-

0

4

9 3 c

2

.>

o

°

n-

3

B

c

2

2

COX

I

II

IV

Va

Vb

VIc

$DH

Fig. 3. Correlation between cytochrome-c oxidase activity and immunodetectable cytochrome-c oxidase subunits in type 1 muscle fibers of a CPEO patient. Fibers are classified according to cytochrome-c oxidase activity (normal activity, panel A; reduced activity, panel B; no activity, panel C). Bars indicate mean of staining intensity for: cytochrorne-c oxidase activity (COX), immunodetectable cytochrome-c oxidase subunit I (I), subunit II (II), subunit IV (IV), subunit Va (Va), subunit Vb (Vb) and subunit VIc (VIc), and succinate dehydrogenase activity (SDH). Standard deviations of mean values are indicated. All cytochrome-c oxidase activity negative fibers (i.e., all fibers in panel C) were ragged-red with Gomori trichrome stain.

5 fibers lacking cytochrome-c oxidase activity was raggedred and had relatively high succinate dehydrogenase activity. In cytochrome-c oxidase activity negative type 2a and 2b fibers, subunit II, Vb and VIc of cytochrome-c oxidase could not be demonstrated with the antibodies. The subunits I and Va, however, were still demonstrated in some of these fibers, though at reduced levels. In these cytochrome-c oxidase activity negative fibers, steady-state levels of subunit IV did not differ significantly from fibers with apparently normal cytochrome-c oxidase activity. The single type 2b ragged-red fiber lacked immunodetectable subunits I, II, Va, Vb, and VIc, but subunit IV was demonstrated to be present at a steady-state level higher than in type 2b fibers with apparently normal cytochrome-c oxidase activity.

4. Discussion

Antibodies have been used by others to investigate cytochrome-c oxidase deficiencies at the protein level. In early studies, antisera against the holo-enzyme were used (e.g., Refs. [35-39]). More recently, either subunit specific polyclonal [39-46] a n d / o r monoclonal antibodies [47-55] were employed. However, many of these 'subunit specific' antibodies, including monoclonal antibodies, recognize apparent clusters of subunits that are difficult to resolve on

205

SDS-polyacrylamide gels, such as subunits Va and Vb, or subunits Via, VIb and VIc [40-50]. Thus, accurate identification of the cytochrome-c oxidase subunits has not been possible. We have developed a battery of truly subunit specific monoclonal antibodies which permit a more detailed analysis of the mutant enzyme. When applied on immunoblots with whole cell extracts, most of the monoclonal antibodies listed in Table 1 exclusively react with their respective subunit (Fig. l), illustrating the high specificity of these monoclonals. One of the monoclonals against subunit I (5D4-F5) and the monoclonal against subunit VIc also bind to some extent to other proteins on the immunoblot strips with total cellular protein (Fig. 1) but the cytochrome-c oxidase subunit can still easily be identified. Thus, the subunit steady-state levels can be determined in crude cell and tissue extracts. In the experimental human cell line 143B206, which is devoid of mitochondrial DNA, the monoclonal antibodies reveal that the nuclear-encoded cytochrome-c oxidase subunits IV and Va are present in amounts close to those of the parental cell line, whereas the amounts of the nuclearencoded subunits Vb and Vlc are severely reduced (Fig. 1). Similar results have been obtained with a myoblast cell line from a patient with mitochondrial DNA depletion syndrome (J.-W. Taanman, J.M. Cooper and A.H.V. Schapira, unpublished observations). In the CPEO patient, most cytochrome-c oxidase activity negative fibers have greatly reduced levels of subunit I, II, Va, Vb, and VIc of cytochrome-c oxidase (Figs. 2 and 3). The steady-state levels of subunit IV, however, were less affected. Particularly in fibers with accumulated mitochondria (ragged-red), immunostaining with anti-subunit IV resulted in high signals. In yeast null mutants for subunit 1 or II of cytochrome-c oxidase, we also observed differential steadystate levels of the nuclear-encoded subunits, with 50-100% remaining of the yeast homologs for the (mammalian) subunits IV and Va, and 20-30% remaining of the yeast homolog for subunit Vb (L.F.M. Calavetta, J.-W. Taanman and R.A. Capaldi, unpublished observations). Recently, Nijtmans and colleagues [56] reported that the steady-state levels of all nuclear-encoded subunits of cytochrome-c oxidase are without exception severely reduced by increased proteolytic degradation in human Molt-4 cells depleted of mitochondrial gene products. In contrast, our results indicate that some nuclear-encoded subunits are more stable than others in the absence of their mitochondrially encoded partners. This is particularly evident in ragged-red fibers where increased amounts of subunit IV were demonstrated. These findings are corroborated by other immunohistochemical studies with subunit IV specific antibodies [51-53]. Moreover, subunit IV remained at almost normal steady-state levels during inhibition of mitochondrial protein synthesis in human HepG2 cells [57]. The function of most nuclear-encoded subunits of cytochrome-c oxidase is obscure and it cannot be excluded that some have an additional function besides their role in

206

J.-W. Taanman et al. / Biochimica et Biophysica Acta 1315 (1996) 199-207

the oxidase. Deletion analyses of nuclear genes coding for mitochondrial proteins have revealed a mosaic of cis-acting upstream and intron sequence elements required for maximal transcription [58]. For instance, a recognition sequence for the transcriptional activator NRF-1 [59] has been identified in the upstream regions of a large number of genes for mitochondrial proteins, including cytochrome-c oxidase subunit Vb [60] and VIc [61] but could not be demonstrated in the gene for subunit IV [62]. The apparent differential regulation of subunit IV gene transcription supports the idea of an additional function of subunit IV and may explain the relatively high steady-state levels of this subunit in cells lacking cytochrome-c oxidase activity. In the immunohistochemical analysis of skeletal muscle of a patient with CPEO, occasional discrepancies were observed in the serial cross-sections where decreased immunoreactivity was accompanied by normal cytochrome-c oxidase activity or the other way around. For instance, fiber 'E' in Fig. 2a shows normal cytochrome-c oxidase activity despite severly reduced levels of immunodetectable subunit I. These anomalous situations are probably due to the segmental nature of the pathology as demonstrated in a longitudinal section stained for cytochrome-c oxidase activity (Fig. 2b). Overall, the immunohistochemical analysis clearly demonstrates that the focal defect of cytochrome-c oxidase activity in muscle of CPEO patients, as observed in this study and in other reports [39,41,49,63], is based on a nearly complete loss of the mitochondrially encoded subunits I and II and some nuclear-encoded subunits of the enzyme complex. The loss of subunits I and II may be the result of a global mitochondrial translation failure, due to limiting amounts of some indispensable tRNAs whose genes are included in the 5-kb deletion. Alternatively, the loss of subunits I and II may be caused by increased proteolytic degradation because assembly does not occur in absence of subunit III whose gene is in the deleted region and may be present at restrictive copy numbers. The results presented in this paper demonstrate that our panel of monoclonal antibodies can be used for diagnostic purposes to reveal steady-state levels of mitochondrially and nuclear-encoded cytochrome-c oxidase subunits. To date, deficiency of the cytochrome-c oxidase is in general determined by analysis of the enzyme activity but this does not reveal the underlying defect. It would be of particular interest to know whether the disease is caused by a nuclear or mitochondrial mutation. We are currently investigating cell lines from other cases with cytochrome-c oxidase deficiency to classify patients by steady-state subunit patterns and try to correlate this with the genetic origin of the defect.

Acknowledgements Supported by the National Institutes of Health Grant HL 22050 (to R.A.C.) and grants from the Muscular Dystro-

phy Association and the Medical Research Foundation of Oregon (to N.G.K.). Drs. Anton Muijsers and Robin Hall are acknowledged for the human oxidase preparations. Beth Prescott is thanked for expert technical assistance in selecting and culturing hybridomas, and Cordelia C. Dickinson and George Cole (Department of Pathology, OHSU) for excellent technical assistance in the (immuno)histochemical studies. Dr. A. D' Agostino (Department of Pathology, OHSU) is acknowledged for help in obtaining and diagnosing human tissues.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11] [12] [13]

[14] [15] [16] [17]

[18]

Hatefi, Y. (1985) Annu. Rev. Biochem. 54, 1015-1069. Attardi, G. and Schatz, G. (1988) Annu. Rev. Cell Biol. 4, 289-333. Capaldi, R.A. (1990) Annu. Rev. Biochem. 59, 569-596. Cooper, C.E., Nicholls, P. and Freedman, J.A. (1991) Biochem. Cell Biol. 69, 589-607. Calhoun, M.W., Thomas, J.W. and Gennis, R.B. (1994) Trends Biochem. 19, 325-330. Kadenbach, B. and Merle, P. (1981) FEBS Lett. 135, 1-1 I. Kadenbach, B. (1986)J. Bioenerg. Biomembr. 18, 39-54. Capaldi, R.A., Marusich, M.F. and Taanman, J.-W. (1995) Methods Enzymol. 260, 117-132. Taanman, J.-W., Turina, P. and Capaldi, R.A. (1994) Biochemistry 33, 11833-11841. DiMauro, S., Lombes, A., Nakase, H., Mita, S., Fabrizi, G.M., Tritschler, H.-J., Bonilla, E., Miranda, A.F., DeVivo, D.C. and Schon, E.A. (1990) Pediatr. Res. 28, 536-541. Robinson, B. (1993) Biochim. Biophys. Acta 1182, 231-244. Shanske, S. and DiMauro, S. (1994) Curr. Topics Bioenerg. 17, 21-58. Kennaway, N.G., Burton, M.D., Hall, R.E., Johnston, W.S.W., Keightley, J.A., Salas, V.M. and Tindle, N.A. (1994) Neurology 44 (Suppl. 2), A335. Shoffner, J.M. and Wallace, D.C. (1990) Adv. Hum. Genet. 19, 267-330. Schon, E.A., Hirano, M. and DiMauro, S. (1994) J. Bioenerg. Biomembr. 26, 291-299. Ballinger, S.W., Shoffner, J.M. and Wallace, D.C. (1994) Curr. Topics Bioenerg. 17, 59-98. Taanman, J.-W., Hall, R.E., Tang, C., Marusich, M.F., Kennaway, N.G. and Capaldi, R.A. (1993) Biochim. Biophys. Acta 1225, 95-100. Van Kuilenburg, A.B.P., Gorren, A.C.F., Dekker, H.L., Nieboek, P., Van Gelder, B.F. and Muijsers, A.O. (1992) Eur. J. Biochem. 205, 1145-1154.

[19] Steffens, G. and Buse, G. (1976) Hoppe-Seyler's Z. Physiol. Chem. 357, 1125-1137. [20] Capaldi, R.A. and Hayashi, H. (1972) FEBS Lett. 26, 261-263. [21] Kadenbach, B., Jarausch, J., Hartmann, R. and Merle, P. (1983) Anal. Biochem. 129, 517-521. [22] Marusich, M.F. (1988) J. Immunol. Methods 114, 155-159. [23] Taggart, R.T. and Samloff, I.M. (1983) Science 219, 1228-1230. [24] Sch~igger, H. and Von Jagow, G. (1987) Anal. Biochem. 1166, 368-379. [25] Planques, Y., Capitanio, N., Capitanio, G., De Nitto, E., Villani, G. and Papa, S. (1989) FEBS Lett. 258, 285-288. [26] Marusich, M.F., Furneaux, H.M., Henion, P.D. and Weston, J.A. (1994) J. Neurobiol. 25, 143-155. [27] King, M.P. and Attardi, G. (1989) Science 246, 500-503. [28] Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K., Gartner, F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J. and Klenk, D.C. (1985) Anal. Biochem. 150, 76-85.

J.-W. Taanman et al. / Biochimica et Biophysica Acre 1315 (1996) 199-207

[29] Laemmli, U.K. (1970) Nature, 227, 680-685, [30] Seligman, A.M., Karnovsky, M.J., Wasserkrug, H.L. and Hanker, J.S. (1968) J. Cell Biol. 38,1-14. [31] Nachlas, M.M,, Tsou, K.-C., De Souza, E., Cheng, C.-S. and Seligmam A.M. (1957)J. Histochem. Cytochem. 5,420-436. [32] Dubowitz, V. (1985) Muscle Biopsy: A Practical Approach, 2nd ed., pp. 56-73, Baitli~re Tindall, London. [33] Engel. W.K. and Cunningham, G.G. (1963) Neurology 13, 919-923. [34] Cunning, W.J.K., Fulthorpe, J,, Hudgson, P. and Mahon, M. (1994) Color Atlas of Muscle Pathology, pp. 15-44, Mosby-Wolfe, London. [35] Bresolin, N., Zeviani, M., Bonilla, E., Miller, R.H., Leech, R.W., Shanske. S., Nakagawa, M. and DiMauro, S. (1985) Neurology 35, 802-812. [36] MUller-HiScker, J., St~nkel, S., Pongratz, D. and Hiibner, G. (1985) J. Neurol. Sci. 69, 27-36. [37] Tanaka, M., Nishikimi, M., Suzuki, H., Tada, M., Ozawa, T., Okini, E. and Takahashi, H. (1986) Biochem. Biophys. Res. Commun. 137, 911-916. [38] Romero, N.B.. Lestienne, P., Marsac, C., Paturneau-Jouas, M., Nelson, 1., Francois, D., Eymard, B. and Fardeau, M. (1989) J. Neurol. Sci. 93, 297-309. [39] Collins, S., Dennett, X., Byrne, E. and Marzuki, S. (1991) Acta Neuropathol. 82, 185-192. [40] Miiller-H~cker, J., Johannes, A., Droste, M., Kadenbach, B., Pogratz, D. and Hfibner, G. 11986) Vichows Arch. B 52, 353-367. [41] Johnson, M.A., Kadenbach, B,, Droste, M., Old, S.L. and Turnbull, D.M. (1988) J. Neurol. Sci. 87, 75-90. [42] Miiller-H~cker, J., Droste, M., Kadenbach, B., Pongratz, D., and Hiibner, G. (1989) Human Pathol. 20, 666-672. [43] Mtiller-H~cker, J., Ibel, H., Paetzke, I., Deufel, T., Endres, W., Kadenbach, B., Gokel, J.M. and Hiibner, G. (1991) Vichows Arch. A 419, 355-362. [44] Mtiller-H~cker, J., Schneiderbanger, K., Stefani, F.H. and Kadenbach, B. (1992) Mutation Res. 275, 115-124, [45] Tritschler, H.-J., Andreetta, F., Moraes, C.T., Bonilla, E., Arnaudo, E., Danon, M.J., Glass, S., Zelaya, B.M., Vamos, E., Telerman-Toppet, N., Shanske, S. Kadenbach, B., DiMauro, S. and Schon, E.A. (1992) Neurology 42, 209-217. [46] M~ller-H~cker, J., H~bner, G., Bise, K., F~rster, C., Hauck, S., Paetzke, l., Pongratz, D. and Kadenbach, B. (1993) Arch. Pathol. Lab. Med. 117, 202-210.

207

[47] Tritschler, H.-J., Bonilla, E., Lombes, A., Andreetta, F., Servidei, S., Schneyder, B., Miranda, A.F., Schon, E.A., Kadenbach, B. and DiMauro, S. (1991 ) Neurology 41,300-305. [48] Lombes, A., Nakase, H., Tritschler, H.-J., Kadenbach, B., Bonilla, E., deVivo. D.C., Schon, E.A. and DiMauro, S. 11991) Neurology 41,491-498. [49] Oldfors, A.. Larsson, N.G., Holme, E., Tulinius, M., Kadenbach, B. and Droste. M. (1992)J. Neurol. Sci. 110, 169-177. [50] Possekel, S., Lombes, A., Ogier de Baulny, H., Cheval, M.-A., Fardeau, M., Kadenbach, B. and Romero, N.M. (1995) Histochemistry 103, 59-68. [51] Mira, S., Schmidt, B., Schon, E.A., DiMauro, S. and Bonilla, E. (1989) Proc. Natl. Acad. Sci. USA 86, 9509-9513. [52] Moraes, C.T., Shanske, S., Tritschler, H.-J., Aprille, J.R., Andreetta, F., Bonilla, E., Schon, E.A. and DiMauro, S. (1991) Am. J. Hum. Genet. 48, 492-501. [53] Moreas, C.T, Ciacci, F., Bonilla, E., Jansen, C., Hirano, M., Rao, N., Lovelace, R.E., Rowland, L.P., Schon. E.A. and DiMauro, S. (1993) J. Clin. Invest. 92, 2906-2915. [54] King, M.P.~ Koga, Y., Davidson, M. and Schon, E.A. (1992) Mol. Cell. Biol. 12, 480-490. [55] Hao, H., Bonilla, E., Manfredi, G., DiMauro, S. and Moraes, C.T. (1995) Am. J. Hum. Genet. 56, 1017-1025. [56] Nijtmans, L.G.J., Spelbrink, J.N., Van Galen, M.J.M., Zwaal, M., Klement, P. and Van den Bogert, C. (1995) Biochim. Biophys. Acta 1265, 1174126. [57] Chrzanowska-Lightowlers, Z., Turnbull, D.M., Bindoff, L.A. and Lightowlers, R.N. (1993) Biochem. Biophys. Res. Commun. 196, 328-335. [58] Nagley, P. 11991) Trends Genet. 7, 1-4. [59] Evans, M.J. and Scarpulla, R.C. 11989) J. Biol. Chem. 264, 1436114368. [60] Hoshinaga, H., Amuro, N., Goto, Y. and Okazaki, T. 11994) J. Biochem. 115, 194-201. [61] Evans, A.J. and Scarpulla, R.C. (1990) Genes Dev. 4, 1023-1034. [62] Virbasius, J.V. and Scarpulla, R.C. (1990) Nucleic Acids Res. 18, 6581-6586. [63] Johnson, M.A., Turnbull, D.M., Dick, D.J. and Sherratt, H.S.A. (1983) J. Neurol. Sci. 60, 31-53.