PDF hosted at the Radboud Repository of the Radboud University Nijmegen

The following full text is a publisher's version.

For additional information about this publication click this link. http://hdl.handle.net/2066/114105

Please be advised that this information was generated on 2021-10-11 and may be subject to change. MITOCHONDRIAL CREATINE KINASE some clinical, biochemical and morphological aspects

Jan A.M. Smeitink

MITOCHONDRIAL CREATINE KINASE some clinical, biochemical and morphological aspects

Jan A.M. Smeitink

MITOCHONDRIAL CREATINE KINASE SOME CLINICAL, BIOCHEMICAL AND MORPHOLOGICAL ASPECTS

EEN WETENSCHAPPELIJKE PROEVE OP HET GEBIED VAN DE MEDISCHE WETENSCHAPPEN, IN HET BIJZONDER DE GENEESKUNDE

PROEFSCHRIFT

TER VERKRIJGING VAN DE GRAAD VAN DOCTOR AAN DE KATHOLIEKE UNIVERSITEIT NIJMEGEN VOLGENS BESLUIT VAN HET COLLEGE VAN DECANEN IN HET OPENBAAR TE VERDEDIGEN OP

DINSDAG 6 OKTOBER 1992,

DES NAMIDDAGS TE 1.30 UUR PRECIES

DOOR

JOHANNES ALBERTUS MARIA SMEITINK

GEBOREN OP 21 JUNI 1956 TE ARNHEM IV

Promotores : Prof. Dr. R.C.A. Sengers Prof. Dr. J.M.F. Trijbels Co-Promotores : Dr. W. Ruitenbeek Dr. R.A. Wevers Aan mijn ouders AanWillemien en Mark CONTEN

CHAPTER 1 Introduction and aim of the study

CHAPTER 2 Mitochondrial creatine kinase: a key enzyme of aerobic energy metabolism Biochimica et Biophysica Acta (Reviews on Bioenergetics): in press

I. Introduction II. Biochemical studies of Mi-CK ΠΙ. Functional studies of Mi-CK IV. Integration of Mi-CK in cellular energy metabolism V. Perspectives

CHAPTER 3 A method for quantitative measurement of mitochondrial creatine kinase in human skeletal muscle Annals of Clinical Biochemistry Ì99229:196-201

CHAPTER 4 Maturation of mitochondrial and other isoenzymes of creatine kinase in skeletal muscle of preterm bom infants Annals of Clinical Biochemistry 199229:302-306 CHAPTER 5 Mitochondrial creatine kinase activity in patients with 123 a disturbed energy generation in muscle mitochondria Journal of Inherited Metabolic Disease: accepted

CHAPTER 6 Mitochondrial creatine kinase containing crystals, 133 creatine content and mitochondrial creatine kinase activity in chronic progressive external ophthalmoplegia Neuromuscular Disorders: in press

CHAPTER 7 Considerations and perspectives 143

Summary 150

Samenvatting 152

Woorden van dank 155

Curriculum vitae 159

List of publications 161 This study was performed at the Department of Pediatrics of the University of Nijmegen, The Netherlands. It is part of the research program "Disorders of the neuromuscular system".

Publication of this thesis was financially supported by: Ethifarma, Milupa, Fonds Bevordering Wetenschapsbeoefening Afdeling Kindergeneeskunde (Academisch Ziekenhuis, Nijmegen) en Stichting ter bevordering van Onderwijs en Speurwerk ten behoeve van de Kindergeneeskunde (Wilhelmina Kinderziekenhuis, Utrecht).

CIP-DATA KONINKLIJKE BIBLIOTHEEK, DEN HAAG

Smeitink, Jan A.M.

Mitochondrial creatine kinase : some clinical, biochemical and morphological aspects / Jan A.M. Smeitink. - [S.l. : s.n.] (Wijk bij Duurstede : Addix). - 111. Thesis Nijmegen. - With ref. - With summary in Dutch. ISBN 90-9005239-9 Subject headings: mitochondria/myopathy

No part of this book may be reproduced in any form or by any means without permission from the author. Introduction and aim of the study

Chapter 1 2 CHAPTER 1 Introduction 3

Introduction and aim of the study

Mitochondrial creatine kinase (Mi-CK; EC 2.7.3.2) is localized at the outer surface of the inner mitochondrial membrane. The enzyme catalyzes the transfer of they-phosphate group of adenosine triphosphate (ATP) to creatine (Cr), thereby producing phosphocreatine (PCr) plus adenosine diphosphate (ADP). Mi-CK plays an essential role in the shuttling of high-energy phos­ phates produced by the intramitochondrially situated energy-generating system in organs with a high and fluctuating energy demand like skeletal muscle, heart and brain. A reduction in the activity of Mi-CK may cause severe problems with respect to energy-consuming processes. Saks and co-workers showed that inhibition of Mi-CK by a specific antibody resulted in a decreased total ATP production by mitoplasts [1 ]. The physiological role of the enzyme has hardly been investigated in human pathology. It may be assumed that a deficiency of Mi-CK leads to a high intra­ mitochondrial ATP/ADP ratio and a decreased amount of PCr (Figure 1). The ATP/ADP ratio has its influence on several metabolic steps. In the first place, the activity of the pyruvate dehydrogenase complex (PDHc) is inacti­ vated by a high ATP/ADP ratio mediated by a phosphorylation of the El subunit of PDHc [2]. Furthermore, in the presence of a physiological, intact coupling state of the oxidative phosphorylation, an increased ATP/ADP ratio causes inhibition of the electron flow through the respiratory chain. The resulting increase of the NADH/NAD+ ratio also produces a feedback inhibition of the PDHc, leading to accumulation of pyruvate and lactate. As a result of the increased intrami tochondrial NADH /N AD+ ratio the cy tosolic NADH/NAD+ ratio is shifted likewise through the action of the malate- aspartate shuttle. The final result is an increased lactate/pyruvate (L/P) ratio in body fluids. Both lactic acidosis and an increased L/P ratio is frequently found in body fluids of patients with mitochondrial myopathy. A mitochondrial myopathy can be defined as a muscle disease characterized by structurally or numerically abnormal mitochondria and / or abnormal functioning of mitochondria. Such a disease may be expected in patients with muscle weakness and exercise intolerance. In most of the laboratories involved in the diagnosis of mitochondrial disorders the follow­ ing biochemical measurements are performed in skeletal muscle biopsies of such patients: substrate oxidation rates, ATP production rates and activities of PDHc and respiratory chain complexes [e.g. 3]. However, despite these intensive studies, no specific enzyme deficiency could be detected in approximately 30% of the muscle samples with a disturbed pyruvate and 4 CHAPTER 1

malate oxidation rate. Therefore, extension of the diagnostic program includ­ ing determination of Mi-CK activity is warranted. The aim of this study was to develop a method to measure Mi-CK activity in human skeletal muscle and to study its applicability with respect to mitochondrial myopathies.

Outline of the study

In this study the results of investigations of Mi-CK activity in human skeletal muscle are described. A critical study of the literature concerning Mi-CK is presented in Chapter 2. The first objective of this study was to develop a method to determine Mi-CK activity in a small amount of human skeletal muscle and to collect reference values for Mi-CK activity. Method and reference values are presented in Chapter 3. From scarce literature data and the experience from oui group it appears that the activity of enzymes involved in the energy metabolism of muscle mitochondria inaeases with age during development and reaches adult levels approximately a few months post partum [4,5]. Knowledge about the existence of age-dependency is important to the interpretation of the biochemical studies performed in the very young age group with the suspicion of a mitochondrial disorder. Chapter 4 contains data concerning the development of Mi-CK and other CK isoenzymes in preterm born infants. Reference values for this specific age group are given. Sub­ sequently Mi-CK activity in patients with a disturbed pyruvate and malate oxidation rate, in which no specific defect of the pyruvate dehydrogenase complex and the respiratory chain could be established, was studied. Results of this investigation are presented in Chapter 5. A striking structural ab­ normality regularly observed in the intermembrane space of mitochondria of patients suffering from mitochondrial myopathies is the occurrence of crystals. Recently it was shown that these crystals contain Mi-CK protein [6,7]. We studied the relationship between Mi-CK containing crystals and the muscle aeatine content. This is presented in Chapter 6. Finally, considera­ tions about the work performed in this study and recommendations for future investigation are given in Chapter 7. Introduction 5

Glucose

. LDH \ Cr^/^ PCr | Lactate f -^ *- Pyruvate · j j Cytosol

Fig. 1. Hypothetical biochemical consequences of a Mi-CK deficiency. LDH = Lactate dehydrogenase; PDHc = Pyruvate dehydrogenase complex; OM = Outer mitochondrial membrane; IM = Inner mitochondrial membrane.

= porm

= adenine nucleotide translocator

= pyruvate translocator

= long chain fatty acids are, in contrast to medium and short chain fatty

acids, transported via a carnitine dependent transport system

= respiratory chain

•• F^-ATPase 6 CHAPTER 1

References 4. Smeitink JAM, Sengers RCA, Trijbele JMF, et al. Fatal neonatal cardiomyopathy 1. Kuznetsov AV, Khuchua ZA, Saks VA. associated with cataract and mitochondrial Mitochondrial synthesis of phosphocrcatine myopathy. Eur J Pediatr 1989;148:656-659 . under physiological conditions. In: Creatine 5. Speri W, Sengers RCA, Trijbels JMF, et al. phosphate. Biochemistry.pharmacologyand Enzymeactivitiesofthemitochondrial energy clinical efficiency (Saks VA, Bobkov YG, generating system in skeletal muscle tissue Strumia E, eds) 1987:15-30, Edizioni Minerva of preterm and full term neonates. Ann Clin Medica, Torino. Biochem: in press. 2. Wieland OH. The mammalian pyruvate 6. Stadhouders A, Jap P, Wallimann T. dehydrogenase complex: structure and Biochemical nature of mitochondrial crystals. regulation. Rev Physiol Biochem Pharmacol J Neurol Sci 1990;98:304-305. 1983; 96:123-170. 7. Stadhouders A, Jap P, Winkler HP, 3. Fischer JC, Ruitenbeek W, Cabreéis FJM, et Wallimann T. Pathologic intra-organelle al. A mitochondrial encephalo-myopathy: crystallization of mitochondrial creatine the first case with an established defect at the kinase(Mi-CK)inmitochondrial myopathies. level of coenzyme Q. Eur J Pediatr J Muscle Res Cell Motility 1992;13:255A. 1986;144:441-444. Mitochondrial creatine kinase: a key enzyme of aerobic energy metabolism

Maikus Wyss1, Jan Smeitink2, Ron A. Wevers3 and Theo Wallimann1

1 Institute for Cell Biology, Swiss Federal Institute of Technology, Zurich, Switzerland. 2 University Children's Hospital, "Het Wilhelmina Kinderziekenhuis", Utrecht, The Netherlands. 3 Institute of Neurology, University Hospital Nijmegen, The Netherlands.

Chapter 2

Biochimica et Biophysica Acta (Reviews on Bioenergetics): in press 8 CHAPTER 2

Abbreviations ANT, adenine nucleotide translocator; CS, contact sites between inner and outer mitochondrial membranes; CK, creatine kinase; Mi-, M- and B-CK, mitochondrial, muscle cytosolic and brain cytosolic isoforms of CK; Cr, creatine; PCr, phosphoaeatine; GPA, ß-guanidinopropionic acid; GPAP, phosphorylated GPA; GBA, ß-guanidinobutyric acid; cCr, cyclocreatine; PcCr, phosphocyclocreatine; AK, arginine kinase; CSF, cerebrospinal fluid. Review 9

I. Introduction

The primary source of energy for many crucial processes in living cells is ATP. However, even though cellular pools of ATP are rather small, no significant decrease in [ATP] is detected during cell activation (e.g., muscle contraction, brain stimulation, phototransduction in retina or initiation of sperm motility) [for reviews see Refs. 318,417]. In all these tissues or cells with high andfluctuating energ y requirements, ATP is continuously replenished from phosphocreatine (PCr) by the action of the creatine kinase (CK, EC 2.7.3.2) system. The CK isoenzymes catalyze the transphosphorylation reac­ tion between PCr and ADP:

PCr2" + MgADP' +(χ)·Η+ « » MgATP2 + Cr

The indicated charges are approached above pH 6.5 where χ approximates 1. When the pH is lowered below 6.5, the charges considerably change and χ decreases. Five CK isoenzymes are currently known in avian and mammalian tissues. Three of them are found within the cytoplasm and two are strictly mitochondrial. The cytosolic CK isoenzymes form only dimeric molecules, namely MM-CK, MB-CK and BB-CK, composed of two types of subunits, the M or muscle type subunit and the В or brain type subunit [95,115]. MM-CK is predominantly found in mature skeletal muscle and mammalian myocardium; BB-CK in mammalian brain and neural tissues, embryonic skeletal and cardiac muscle, and avian myocardium; and MB-CK in adult mammalian heart as well as in striated muscles during the developmental transition from BB- to MM-CK [115,227,530]. Subcellular fractionation, e.g. isolation of myofibrils, sarcoplasmic reticulum, plasma membranes, etc., as well as direct in situ immunolocalization studies revealed a cellular and subcellular compartmentation of the cytoplasmic CK isoenzymes [for re­ views see Refs. 38379,491,492]. In muscle, for example, a small but significant fraction of MM-CK is specifically associated with the myofibrillar M-band where it directly rephosphorylates ADP generated by the actin-activated myosin ATPase. The two mitochondrial CK (Mi-CK) isoenzymes, due to their tissue- specificity of expression on one hand and their relative isoelectric points on the other hand, were called either ubiquitous and sarcomeric Mi-CK

[162,163,244,344] or Mia- and Mib-CK (only for chicken) [397,516], respec­ tively. Both Mi-CK isoenzymes are located within the mitochondrial 10 CHAPTER 2

intermembrane space and form, in contrast to the cytosolic CK isoenzymes, octameric as well as dimeric molecules. In recent years, the obvious importance of the CK isoenzymes for cellular energetics has attracted considerable attention, and three main functions were assigned to the CK/PCr system [491,492]: (i) In many tissues, the concentrations of Cr and PCr are much higher than those of ADP and ATP, thus enabling efficient buffering oí the ADP and ATP concentrations within these cells, (ii) The higher concentrations together with the higher diffusion coefficients of Cr and PCr relative to ADP and ATP [524,525] ensure a significantly enhanced maximal rate of delivery of "high-energy phos­ phates" to sites within the cells where energy is consumed and where ATP has to be locally regenerated (transport function of the CK system), (iii) Since the CK system is only involved in one particular reaction pathway, and since the CK isoenzymes are subcellularly compartmentalized, the CK system offers an attractive opportunity to specifically regulate local ATP / ADP ratios as well as cellular energy metabolism in general. For reviews on the physiology of the CK isoenzyme system and on the biochemistry and evolution of the cytoplasmic CK isoenzymes, the reader is referred to Refs. 37,38,209,218,230,240,261,311,379,416,488,491-494. The purpose of this review is to summarize the important findings on Mi-CK since its discovery in 1964 and to convey our current ideas on the physiological significance and on the structure-function relationships of this enzyme important for the bioenergetics of cells with high and fluctuating energy turnover. Review 11

IL Biochemical studies of Mi-CK

Д-А. Purification of Mi-CK isoenzymes

In order to obtain homogeneous starting material for biochemical and biophysical experiments as well as for the production of polyclonal and monoclonal antibodies, Mi-CK has been purified from a variety of animal and human tissues, namely bovine heart [123,166,167,280], chicken heart [63,396], dog heart [363-365], human heart [49,50,156,231,457], pig heart [355], rabbit heart [132], rat heart [81,82,387], chicken brain [516], human liver [231], pigeon pectoralis muscle [281], and sea urchin spermatozoa [464]. Most of the purification procedures described take advantage of the intracellular locali­ zation of Mi-CKwhichis attached to theoutersurfaceof the inner mitochondrial membrane [213,411 ]. In a first step, mitochondria are enriched by differential centrifugation. Then, Mi-CK is released from the mitochondrial inner mem­ brane ("extraction"), and finally, it is separated from contaminating proteins by ethanol or ammonium sulphate fractionation, chromatofocussing, affinity, ion exchange, hydrophobic interaction, or gel permeation chromatography. Some aspects of the purification procedures will now be discussed in more detail. Simply rupturing the outer mitochondrial membrane, followed by extraction of the enzyme under iso-osmotic conditions, does not give satisfac­ tory yields of Mi-CK. Instead, swelling of the mitochondria by incubation in a hypotonic medium or in an isotonic phosphate solution prior to extraction is a prerequisite for an optimal release of Mi-CK from the inner mitochondrial membrane [122,130]. Upon incubation with sodium phosphate, mitochondria swell because of a passive uptake of sodium as well as phosphate ions. Swelling with potassium phosphate is only observed in respiring mitochondria, since the uptake of potassium ions is an active process. Accordingly, respiratory inhibitors block potassium-induced swelling of the mitochondria [122,130]. Release of Mi-CK from swollen mitochondria can be achieved by a variety of conditions. In most studies, 10-100 mM phosphate was used at pH values of 6.5-9.0 [31,63,81,82,101,123,130,132,156,166,167,210,213,231, 243,296,324,363-365,374,377,396,401,409,434,457,481,482,487,498,502, 504,516]. Extraction of Mi-CK by phosphate was shown to be strongly pH- dependent, being more efficient at pH values higher than 7.0 [481]. Besides phosphate ions, adenine nucleotides [265,296,385,482], negatively charged organic mercurials [81,82,130,131,324,482,498], and 100-250 mM KCl [265,280,481,482,498] were also used to release Mi-CK from the inner mitochondrial membrane. Organic mercurials like parahydroxymercuri- 12 CHAPTER 2

benzoate and mersalyl at concentrations as low as 10 μΜ seem to be the most potent releasing agents resulting in more than 80 % solubilization of Mi-CK [482]. Since mersalyl blocks the uptake of phosphate ions into mitochondria, however, it should not be used in combination with phosphate. Organic mercurials have the additional disadvantage that they inhibit the enzymatic activity of Mi-CK, probably by reacting with the "essential" sulphydryl group of the enzyme, but enzymatic activity can easily be recovered by incubation with excess amounts of reducing agents (2-mercaptoethanol or dithiothreitol). Because KCl extracts Mi-CK, the binding of the latter to mitochondrial membranes was believed to depend on ionic interactions and therefore to be sensitive to the ionic strength of the medium [281,498]. Instead, Saks et al. [385] suggested that the decisive factor is not the ionic strength of the medium by itself but rather the ion composition. Whereas in their experiments 125 mM KCl released more than 50 % of Mi-CK, a "physiological salt solution" containing only 10 mM chloride, but with higher ionic strength and osmolarity, released only 12 %. Chloride (and other anions) therefore seem to have a special effect on the release of Mi-CK, probably by binding to the active site of the enzyme [316,494]. Three types of affinity matrices have been used up to now for Mi-CK purification, namely Cibachrome-Blue-based matrices [50,132,396,487,516], ADP-hexane-agarose [63,132], and ATP-hexane-agarose [167]. Whereas only 2-5 ml columns of the latter two matrices were used. Blue Sepharose CL-6B columns of up to 250 ml were routinely employed for the purification of the chicken Mi-CK isoenzymes [140,396,516], thus allowing the application of large amounts of protein. Nevertheless, the most powerful affinity chroma­ tography technique is probably the transition state-analogue affinity chro­ matography procedure developed by Brooks et al. [63]. In this method, the interaction of Mi-CK with the ADP-hexane-agarose matrix is selectively strengthened by the addition to the application buffer of magnesium ions, creatine and nitrate, thus inducing a transition state-analogue complex of the enzyme. The nitrate ion mimics the phosphate group to be transferred during catalysis [316], which is thought to be planar in the transition state of the reaction. A considerable problem in the purification of Mi-CK is its separation from the cytosolic CK isoenzymes, because in many tissues, Mi-CK makes up only a few percent of the total CK activity. A large part of the contaminating cytosolic CK can be eliminated by differential centrifugation. In this respect, brain is more delicate, because synaptosomes formed during homogeniza- tion of the tissue contain cytosolic CK [136,504] and co-sediment with mitochondria. In chicken and rat, however, swelling of the mitochondrial (and synaptosomal) suspension breaks up the synaptosomes, and a first Review 13 separation of the mitochondrial and cytosolic CK isoenzymes can be achieved by an additional centrifugation step [516,518]. Mi-CK can also be separated from the cytosolic CK isoenzymes by ethanol fractionation [166,503] or, because Mi-CK isoenzymes of most species have much higher isoelectric points (IEP > 8; see II-D.) than the cytosolic CK isoenzymes (IEP < 7) [185,483,526], by ion exchange chromatography. In addition, mitochondrial and cytosolic CK isoenzymes can be separated by hydrophobic interaction chromatography [9,49,50,401,499], indicating that Mi-CK is more hydropho­ bic than cytosolic CK. Finally, both a complete separation of the isoenzymes as well as an additional purification of Mi-CK itself were achieved by affinity chromatography [167,396,487,516]. A second serious problem during puri­ fication emerges from the "oligomeric heterogeneity" of Mi-CK. It is well established now that Mi-CK isoenzymes form octameric as well as dimeric molecules which are - depending on protein concentration, substrate concen­ trations, pH, etc. - readily interconvertible (see III-D.). Mi-CK octamers and dimers behave quite differently in most of the purification methods men­ tioned above so that one oligomeric form has to be strongly favoured by appropriate experimental conditions in order to obtain maximal yields (Schlegel, J., Wyss, M. and Wallimann, T., unpublished data). The purification procedures for two Mi-CK isoenzymes differ from the "general scheme" described above. The purification of human heart Mi-CK by Blum et al. [50] includes three chromatography steps, but no enrichment of mitochondria by differential centrifugation. The low specific activity of their Mi-CK preparation may be explained by the rather harsh and lengthy procedure used. A purification scheme based on differential centrifugation may turn out to be superior if one takes into account the fact that an optimal release of human sarcomeric Mi-CK from the mitochondrial membranes can only be achieved by addition of 0.05 % Triton X-100 to the extraction buffer [434]. Because differential centrifugation of spermatozoa is almost impossi­ ble, this step was also omitted from the purification scheme of sea urchin sperm Mi-CK [464]. Furthermore, phosphate failed to release significant amounts of Mi-CK from the sperm membranes, and instead, 5% of the non-ionic detergent Nonidet P-40 was used [463,464]. Additional purifica­ tion of sperm Mi-CK was achieved by selective precipitation in a buffer of low ionic strength [464]. The specific activities of the purified Mi-CK isoenzymes shall not be listed here, since they strongly depend on the reaction direction and on the experimental conditions used. The most reasonable range in the direction of ATP synthesis at 30CC and pH near neutrality seems to be 100-200 ^ol/min/mg protein [26,63,281,287]. 14 CHAPTER 2

II-B. Molecular size

Whereas it is generally accepted that cytosolic CK isoenzymes exclusively form dimeric molecules (MM-, MB- and BB-CK) [95,115,494], the naturally occurring oligomeric forms of Mi-CK were a matter of debate for quite a long time. Some research groups presented evidence that Mi-CK, like the cytosolic

CK isoenzymes, is also exclusively a dimeric molecule with a Mr of 80 500-

84 000 [23,49,50,63,156,364,365,439] and even claimed that the higher Mr aggregates reported by others represent "induced artifacts" [156,365]. Nevertheless, an increasing body of evidence, starting with the pioneering studies of Farrell et al. [122], Saks et al. [374] and Jacobs [211], demonstrated that Mi-CK, at least in vitro, forms stable dimeric and octameric molecules. Dimeric and octameric Mi-CK are readily interconvertible (see Ш-D.) and have molecular masses of 75-91 kDa and 306-380 kDa, Stokes radii of 36-38 Â and 59-65 Â, and sedimentation coefficients of 4.9-5.4 S and 11.6-13.5 S, respectively (Table 1). The most thorough analysis concerning the molecular mass was per­

formed on the Mi-CK isoenzymes from chicken, namely ubiquitous Mia-CK from brain and sarcomeric Mib-CK from heart. Gel permeation chromatography and analytical ultracentrifugation of the purified isoenzymes revealed two oligomeric forms each with M/s of 83 000-86 000 and 306 000- 364 000, respectively, with no indication for the existence of an intermediate

form [396,397,403,516]. Considering the protomeric Mr of -43 000 (Table 1), the cDNA sequences (see II-F.) and hybridization experiments [516], the

lower Mr form in all likelihood corresponds to dimeric Mi-CK molecules. In contrast, the higher Mr value, which in addition was confirmed by direct mass measurements of single Mi-CK molecules by scanning transmission electron microscopy [403,516], is strongly indicative for an octameric molecule. This view is also supported by electron microscopical examination of Mi-CK molecules (see II-C), which suggests that four ellipsoid Mi-CK dimers are arranged in parallel to each other to form an octamer.

In several studies [26,27,82,100,122,156,167,278,464,504], the higher Mr value of Mi-CK was underestimated to be in the range 180 000-250 000. The reasons for these underestimations seem to be manifold: (i) already in 1972, Farrell et al. [122] published a Stokes radius of 65 Â for beef heart Mi-CK which is fully in line with the available data for octameric Mi-CK (Table 1).

However, probably due to inappropriate Mr standards, the molecular mass was calculated by the same authors to be only 250 kDa; (ii) due to its

recommended Mr fractionation range of 5 000-250 000, Sephacryl S-200 clearly represents a bad choice for the estimation of the higher Mr value of Mi-CK [100,156,167]. In addition, dimeric Mi-CK revealed a Mr of only 65 000 Review 15

Table L Molecular masses, sedimentation coefficients and Stokes radii ofMi-CK isoenzymes

Species - Tissue Molecular masses [kDa] Sedimentation Stokes radu [A] coefficients [S]

protomer dimer oc ta mer dimer oc tamer dimer octamer Ref bovine heart 41^4 80-89 317-355 5 3-5 4 118-12 4 37 59-65 a chicken heart 42^3 78-87 321-378 50 12 8-13 5 38 3* 62.4* b chicken brain 42 85 306-352 4 9-5 3 116-12 0 36 3* 60 3» с dog heart 41^4 82-84 - - - - - d human heart 41-42 80-85 350 50 - - . e pig heart - 75 342 - (117) - - f pigeon pectoralis 44 87-91 331-380 54 12 0 374 59 3 В rabbit heart 40-45 81 332-377 - - 36-39* 56" h rat heart 42^3 82-87» 340-345* - - 37 7* 611* 1 sea urchin sperm 44-50 - 353 - 12 4 - 62 8 1

References a) 26-29,122,123,211,212,280,282,283,286, b) 63,396,397,403, с) 397,516, d) 23,347,348,363-365, e) 49,50,156,231,243, f) 242,296,483, g) 281,283, h) 132,296,353,439,482,483, ι) 81,82,266,344,382,387,396,

))463,464^17/*)Wyss,M,Schlegel,] and Wallimann,T,unpublished data^ViaLC,unpublished data.

on Sephacryl S-200, but one of 80 000 on Sephadex, indicating that the choice of the gel filtration matrix is crucial for molecular mass determination of Mi-CK [156]; (iii) using a FPLC Superose-12 column from Pharmacia, Tombes and Shapiro [464] determined a Stokes radius of only 45 Â for sperm Mi-CK of the sea urchin Strongylocentrotus purpurei us, whereas we [517] obtained a Stokes radius of 62 8 Â and a Mr of 353 000 for the same protein from the sea urchin Psammechmus miliaris (Table 1) Since Tombes and Shapiro for their ultracentrifugation experiments used a calculation method where the Mr is a function of the Stokes radius whereas the sedimentation coefficient is not, they underestimated the Mr as 235 000, but obtained a "correct" sedimenta­ tion coefficient of 12 4 S Correction for a Stokes radius of 62 8 Â gives a Mr of 328 000 instead of 235 000; (iv) the Mr of 240 000 reported for rat heart Mi-CK [82] can be explained by partial dissociation of the Mi-CK octamers during the gel filtration run. First, in the very same experiments, a Mr > 100 000 was obtamed for dimeric Mi-CK [81], second, faster gel permeation chromatography on a FPLC Superose-12 column revealed a M of 340 000 [396], and third, purified rat heart Mi-CK octamers proved to readily dissociate [398]; (v) the observations on bovine heart Mi-CK made by Belousova and co-workers [26-29,123] are an exception, since these authors are the only ones who provided experimental evidence for two different high 16 CHAPTER 2

Mr forms, namely "hexameric" (Mr 240 000) as well as "octameric" Mi-CK (Mr 340 000). In their most recent article, however, "hexameric" and "octameric" Mi-CK both were shown to have a sedimentation coefficient of 12.3-12.4 S and to display a four-fold symmetry on electron micrographs. Since a hexamer would be highly inconsistent with a four-fold symmetry, one is now tempted to suggest that the two forms represent different conforma­ tions of octameric Mi-CK [28,29]. In vitro translation of Mi-CK mRNA [162,348] as well as determination of the cDNA sequences of the two human and rat Mi-CK isoenzymes

[162,163344] revealed that Mi-CK is synthesized as a precursor protein of Mr 47 000-48 000 containing anN-terminal mitochondrial target sequence which is proteolytically removed after the import into the mitochondria to yield

mature Mi-CK subunits with a Mr of approximately 42 000. A Mr of 42 000 was also reported for Mi-CK protomers from rooster spermatozoa [489] and chicken retina [490,496]. Gel filtration experiments performed by Yasui et al. [523] revealed two oligomeric forms of Mi-CK from human skeletal muscle,

heart and stomach having Mr's of 80 000 and 370 000, but only one form of Mr 370 000 for human brain Mi-CK. Similarly, two different oligomeric forms with M/s of 80 000 and 350 000 were described for human liver Mi-CK [231]. However, since liver only contains minute amounts of CK [40,518], the assumption that this Mi-CK in fact originated from liver cells is unlikely. Instead, the extracted Mi-CK may be derived from blood vessels because vascular and intestinal smooth muscle cells have recently been shown to contain appreciable amounts of octameric Mi-CK [207]. As a last point it seems worth mentioning that already in 1968, Keto and Doherty [242] enriched a CK form with a sedimentation coefficient of 11.73 S from a particulate fraction of pig heart. Even though the authors claimed that this form was not of mitochondrial origin, they were, in hindsight, probably the first to describe octameric Mi-CK.

JJ-C. Three-dimensional structure: Electron microscopy and protein crystallography

By electron microscopy, the octameric forms of Mi-CK isoenzymes from all species investigated so far reveal a very similar structure. Bovine heart [28,29,122,286], rat heart [82], chicken heart [396,397,403-407,507], chicken brain [397,404] as well as sea urchin sperm Mi-CK [517] seem to be cube-like molecules with a side-length of 10 nm, displaying a four-fold symmetry and a central stain-filled indentation or cavity. The fact that not only the oligomeric state but also the overall three-dimensional structure of Mi-CK has been conserved throughout evolution from sea urchins to mammals points to an important physiological role of this structure. Review 17

At least two different projections have to be assumed for an octameric, cube­ like molecule. However, negative staining (Fig. 1A,B) [396,397,403,404,507] and rotary shadowing (Fig. 1C,D) [403,407,507] of single molecules of chicken sarcomeric Mib-CK revealed only one single view of the molecule, indicating that the top and bottom faces of the octamer are identical and have a distinctly higher affinity for a variety of support films used for electron microscopy than the side faces [407]. Since the cross-like surface depression of Mib-CK visualized by rotary shadowing (Fig. 1C,D) is slightly twisted in dock-wise direction relative to the side faces of the octamer [407,507], and since computer averaging of the structure of negatively stained and rotary shadowed Mib-CK octamers revealed a four-fold symmetry [507], a single view for all four side faces has to be assumed. This side view of the octamer has only recently been obtained under one set of particular conditions.

Overnight incubation of chicken Mib-CK with neutral uranyl acetate, fol­ lowed by dialysis against distilled water, resulted in the formation of linear unbranched Mi-CK filaments, in which Mi-CK octamers were stacked by their top and bottom faces on top of each other (Fig. IE) [407]. The side view of the filaments, therefore, also displays the side face of the Mi-CK octamers. One possibility for the overall three-dimensional structure of Mi-CK octamers is that a central cavity is present inside the octamer (Fig. IF) and is connected with the exterior of the molecule by two smaller channels, protruding from the midst of the top and bottom faces right into the middle of the octamer. This possibility is favoured by the facts that negative staining, visualizing a projected view of the entire volume of the molecule, reveals a central accumulation of stain with a diameter of about 2.5 nm (Fig. 1 A,B), and

that rotary shadowing, reflecting the surface structure of Mib-CK, gives rise to a small orifice only (Fig. 1C,D). Alternatively, the side view displayed by the linear Mi-CK filaments (Fig. IE), as well as the recent observation of a second projection of single Mi-CK octamers (Kaldis, P., Schnyder, T. and Wallimann, T., unpublished data), favour the notion that the banana-shaped dimers are arranged in such a way that they form funnel-like indentations in the middle of the top and bottom faces (Fig. IF). Further insight into the structural organization within the octamer was gained from experiments on

the formation of heterooctameric molecules of chicken ubiquitous Mia- and sarcomeric Mib-CK [516]. MiaMib-CK heterooctamers were formed out of a mixture of Mia- and Mib-CK homodimers and subsequently stored for three months at 4°C. Upon re-dissociation of the octamers, cellulose polyacetate

electrophoresis only revealed Mia- and Mib-CK homodimers, but no MiaMib-CK heterodimers, thereby proving that within the octamer, discrete dimers are the stable building blocks, with no subunit exchange between them. 18 CHAPTER 2

%

-r\S ¿if •· в Ща • /Л GODO E -—' і 5nm Review 19

Fig. 1. The three-dimensional structure of chicken sarcomeric Mi-CK octamers deduced from electron micrographs. A) Negative staining of Mi-CK octamers; B) contour representation of A) after circular harmonic averaging [507]; C) Mi-CK octamers rotary-shadowed with Ta/ W at low temperature and ultra-high vacuum [407]; D) contour representation of Mi-CK octamers rotary-shadowed with Pt/lr/C after circular harmonic averaging [507]; E) linear filaments of positively stained Mi-CK octamers. The inset on the right side displays an averaged stretch of the filaments. The side view of one octamer is outlined by dots [407]; F) model representations of the structure of an octamer. Above: top or bottom view; below: the two possible side views displayed as sections through the center of an octamer (see text).

Though crys tallization of cy tosolic CK isoenzymes has already been reported by several groups [69,151,185,304], and though these crystals diffracted to up to 2.0 À resolution, no three-dimensional structure has been solved up to now. One reason for this probably is the microheterogeneity of purified MM- and BB-CK preparations [185356] which is primarily due toposttranslational modifications. This problem may be overcome by crystallizing chicken sarcomeric Mib-CK [404,405,408] for which no posttranslational modifica­ tions are currently known. Precipitation of Mib-CK with polyethylene glycol 1000, either in the presence or absence of ATP, yielded two different types of tetragonal crystals with the space groups P4212 and P422, containing one octamer and one dimer per asymmetric unit, respectively. The former crystal type diffracts to at least 3 Â resolution [405]. A change in the crystal form in the presenceof ATP might be indicative for a conformational change induced by substrate binding, a phenomenon that has attracted much interest in the study of the cy tosolic CK isoenzymes [39]. As far as Mi-CK is concerned, conformational changes induced by substrates were assumed to influence the dimer to octamer ratio [280,281,296], since in a variety of studies, formation of a "transition state-analogue complex" of Mi-CK with MgADP, Cr and nitrate resulted in the complete dissociation of the octamers into dimers (see Ш-D.). Conformational differences of octameric Mi-CK were also thought to be the basis for the apparent "hexameric" and "octameric" forms of bovine heart Mi-CK described by Belousova and co-workers [28,29,123] (see Π-Β.). Thefinding that the modification of Mi-CK dimers, "hexamers" and octamers with SH group reagents is biphasic, with the first half of the subunits being more readily modified than the second half, was taken as an argument for an asymmetric association of the subunits within the dimer [26,27,123], but might be explained as well by a conformational change of discrete dimers induced by modification of only one of their subunits. Surprisingly, however, only half of the active sites within the octamer bound MgADP in the presence of Cr and nitrate [124]. 20 CHAPTER 2

II-D. Isoelectric point

With the exception of sea urchin sperm Mi-CK [464], all Mi-CK isoenzymes currently investigated have higher isoelectric points than the cytosolic CK isoenzymes of the respective species. Whereas pi's of 8.2-9.7 were reported for most Mi-CK isoenzymes (Table 2), pi's of 6.2 and 7.0 were obtained for sea urchin sperm [464] and frog heart Mi-CK [273]. The results on the human Mi-CK isoenzymes are somewhat contradictory. From the cDNA sequences of cardiac and placental Mi-CK [162,163], pi's of 7.42 and 7.49 can be calculated for the mature subunits. These values agree quite well with those determined for native human heart and brain Mi-CK (6.8-7.0) [50,502,504], but are distinctly lower than the value of 9.35 reported for human heart Mi-CK by Khuchua et al. [243]. Besides the differences between particular Mi-CK isoenzymes, distinct isoelectric points were also observed for dimeric and octameric Mi-CK. Whereas a higher pi for the octamer than for the dimer was reported for Mi-CK from rabbit heart, chicken heart and chicken brain, an inverse relationship was found for the isoenzymes from bovine heart and pigeon pectoralis (see Table 2). However, these latter results have to be questioned since in cellulose acetate electrophoresis experiments, octameric bovine heart Mi-CK migrated further towards the cathode than dimeric [100,167] (see also HI-D.), thus suggesting that the octamer is also more positively charged in bovine heart. Mi-CK can be released from the inner mitochondrial membrane by high salt concentrations (see II-A.) and is therefore thought to be bound via ionic interactions [82,369,395,498]. In fact, the basic pi of Mi-CK indicates that, within the mitochondrial intermembrane space, positively charged Mi-CK may bind to negatively charged membrane phospholipids. Accordingly, octameric Mi-CK from chicken heart, probably due to i ts higher pi, was found to interact more strongly with mitoplast and model membranes than dimeric Mi-CK [369399]. Even though ionic interactions of Mi-CK with mitochondrial membranes are important, hydrophobic interactions can not be excluded, especially in the case of sea urchin sperm Mi-CK which can only be detached from the membranes by relatively high concentrations of detergents, but not by any other treatment known to release Mi-CK isoenzymes [464]. This may be due to its acidic pi of 6.2, indicating that ionic interaction with the mitochondrial membranes is reduced, but compensated for by increased hydrophobic interaction. Because of the higher pi, Mi-CK isoenzymes can be nicely separated from the cytosolic CK isoenzymes under native conditions by electrophoresis on cellulose polyacetate strips as well as starch or agar gels, with Mi-CK migrating more cathodically than cytosolic CK [13,50,208,222,326,338, Review 21

Table II. Isoelectric points of Mi-CK

Species - Tissue dimeric octameric other forms1 refs. Mi-CK Mi-CK bovine heart 9.67 8.93 9.2 - 9.6 a chicken heart 9.3 9.4 - 9.5 8.25' b chicken brain 8.4-8.5 8.7 - 9.0 7.94» с frog heart - - 7.0 d human heart 6.94 - 6.8; 9.35; 7.42» e human brain 7.01 - 7.49» f pig heart 8.20 - - g pigeon pectoralis 9.56 8.91 10.25 h rabbit heart 8.24 8.83 - i rat heart - - 9.4; 7.88» i rat brain - - 7.73» к sea urchin sperm - - 6.2 1

'Included in this column are values determined for Mi-CK which was either monomeric or for which the oligomeric form was not determined. Values marked with a * were calculated from the cDNA sequence. References: a) 26,27,208,211,282; b) 196,399,516; c) 399,508,516; d) 273; c) 50,163,243,502; 0 162,504; g) 354; h) 282; i) 354,483; j) 266,344,387; k) 344; 1) 464.

374,390,410,441,516,518]. The only exceptions known are human and frog heart Mi-CK. Human heart Mi-CK, because its pi is very similar to that of MM-CK, may be obscured by the latter isoenzyme [173]. Frog (Rana esculenta) heart Mi-CK, though having a pi which is distinctly higher than those of its cytosolic counterparts (7.0 vs. 5.5-5.8), migrated in between the cytosolic CK isoenzymes on cellulose polyacetate strips [273], as did the isoenzyme termed CK-V of Xenopus frogs which is expressed in lung and heart [511]. Since CK-I of Xenopus is supposed to be of mitochondrial origin as well, but is expressed in eye, brain and stomach, these results may indicate that frog, like chicken, rat and man, also contains two different Mi-CK isoenzymes. Inter­ estingly, the fact that the mitochondrial isoenzymes have a higher isoelectric point than the cytosolic ones is not restricted to CK, but seems to be a rather general phenomenon [175]. Therefore, the hypothesis has been raised that the increased pi is either a prerequisite for an efficient import of the precursor proteins into the mitochondria or an adaptation to the metabolic conditions within these organelles. 22 CHAPTER 2

II-E. Kinetic constants

Since the kinetic constants of the Mi-CK isoenzymes strongly depend on the calculation method used, on the species investigated, on the purity of the enzyme preparation, as well as on buffer composition, temperature and pH, it is beyond the scope of this review to mention all of the available data [for discussion see Ref. 419]. Instead, those studies are selectively picked out which compared different purified CK isoenzymes under exactly the same conditions, in which a variety of kinetic constants were determined for a single isoenzyme, or which allow interesting conclusions. As already mentioned, the CK isoenzymes catalyze the trans- phosphorylation reaction between PCr and ADP (see I.). Most studies on the reaction mechanism were performed on the cytosolic CK isoenzymes [for a review see Ref. 240]. For Mi-CK, it was only shown that at pH 7.4 the reaction mechanism is of the "rapid-equilibrium random" type [287,375], that one cysteine [27,123] as well as two arginine residues per subunit [323,420] are "essential" for catalytic activity, and that probably Asp-335 is involved in substrate binding [223] (see also II-F.). The pH optima of the reaction catalyzed by Mi-CK from bovine heart [100,167], chicken heart [516], human heart [49,50], rabbit heart [439], guinea pig heart [338], rat heart [213], chicken brain [516], guinea-pig brain [456] and sea urchin spermatozoa [464] all range from 6.0-7.0 in the direction of ATP synthesis and from 7.5-9.0 in the direction of PCr synthesis, therefore being in about the same range, or slightly lower than the respective values of the cytosolic CK isoenzymes [338,341,439,464,494] as well as of all other known Phosphagen kinases [for a review see Ref. 322]. These findings are fully in line with pH studies on rabbit MM-CK which revealed that a single group with

a pKa near 7, probably a histidine residue, acts as an acid-base catalyst and must be unpro tonated in the direction of PCr formation and protonated in the direction of ATP formation [90].

All purified Mi-CK isoenzymes investigated so far have a lower Km value

for MgADP than for MgATP, and a lower Km value for PCr compared to Cr [23,26,50,63,100,167,231,287,397,401,410,448,464,523]. The same relationships have been described for the cytosolic CK isoenzymes [for a review see Ref.

48]. The Km values of the Mi-CK isoenzymes were reported to be: 42 μΜ-1.7 mM for MgATP, 15-150 μΜ for MgADP, 3.4-62 mM for Cr, and 0.23-4.1 mM for PCr. Human and chicken ubiquitous Mi-CK display a 2-fold and 4-fold

lower Km value for PCr than the respective sarcomeric Mi-CK isoenzymes [397,523], possibly reflecting a metabolic adaptation to the lower PCr concen­ trations in brain compared to cardiac and skeletal muscle. Furthermore, the

fact that in intact mitochondria, Mi-CK has a lower Km for ATP produced in Review 23 the matrix compartment and presented by the adenine nucleotide translocator (ANT) than for exogeneously added ATP was taken as an argument that Mi-CK has privileged access to intr amitochondrially synthesized ATP due to a direct interaction of Mi-CK with ANT [169,266,267,382,38537]. However, the differences in Km values could also be explained by restricted diffusion of adenine nucleotides across the outer mitochondrial membrane, signifying that especially at high velocities of the Mi-CK reaction, the substrate concen­ trations within the intermembrane space would not correspond to the extramitochondrial ones [146-148,257]. This controversy between "direct interaction of Mi-CK with ANT" and "restricted diffusion of adenine nucleotides" will be discussed in more detail in chapter Ш-С. The indications that the ANT has a much lower Km for ADPand a higher one for ATP than Mi-CK were thought to explain why PCr is the high-energy phosphate compound leaving the mitochondria as the net product of oxidative phosphorylation [167]. However, on the basis that the Km values of Mi-CK and ANT should not be compared since Mg ADP and MgATP are the effective substrates of Mi-CK, whereas only the uncomplexed adenine nucleo tides are transported by ANT, this interpretation has to be seriously questioned. Different cooperativities of substratebindingbetween individual subunits might be a distinguishing feature of dimeric and octameric Mi-CK and thereby represent an effective means of regulating the enzymatic activity of the Mi-CK isoenzymes. However, the Hill coefficients of sarcomeric and ubiquitous Mi-CK octamers for the binding of PCr in the presence of excess MgADP are only 1.2 and 1.0, respectively [397], meaning that almost no communication between the subunits in the binding of this substrate occurs. Cooperativity in the binding of the adenine nucleotides is more likely to occur, since ATP and ADP are not only more tightly bound to Mi-CK, but are also more potent in inducing conformational changes of CK isoenzymes [238]. Arrhenius plots of the temperature dependence of the enzymatic activity were used to calculate the activation energies of the CK reaction in the direction of ATP formation. For the human isoenzymes, significantly higher activation energies of 101-142 kj/mol were found for ubiquitous and sarcomeric Mi-CK than for MM-, MB- and BB-CK (49-76 kj/mol) [164,340,448,449,523], indicating that the substrates PCr and ADP are more strongly bound by the mi tochondrial than by the cy toplasmic CK isoenzymes [231,401,448,503]. Whether this 2-fold difference in activation energies is of physiological importance, for example reflecting an adaptation to lower PCr or ADP concentrations in the intermembrane space compared to the cytosol, remains to be elucidated. Yet, a higher activation energy by all means causes a lowered maximal reaction velocity (V ). Accordingly, lower Vmax values 24 CHAPTER 2

in the direction of ATP formation were reported for the chicken mitochondrial than for the cytosolic CK isoenzymes [518].

JÍ-F. Nucleic and amino acid sequences

Compared to the cytosolic CK isoenzymes [for references see Ref. 223], Mi-CK sequence data are rather sparse. Besides several partial protein sequences [50,82,137,196,223,464,516], six cDNA sequences [162,163,196, 344,508] and two genomic sequences [162,244] were published up to now. Originally, determination of the aminoterminal sequences of chicken heart and brain Mi-CK proved that these tissues contain two different Mi-CK isoenzymes [196], as had been suggested earlier for the rat from agarose gel electrophoresis experiments [391]. In the meantime, communication of two different Mi-CK cDNA sequences each for chicken [196,508], man [162,163] and rat [344] confirmed these findings and further demonstrated that they are due to two different nuclear genes rather than differential splicing of a single gene. Since on RNA blots, specific probes for the sarcomeric and ubiquitous human Mi-CK isoenzymes only detected minute amounts of Mi-CK RNA in uterus and testis, tissues known to contain significant CK activities including Mi-CK, Payne et al. [344] assumed that a third Mi-CK isoenzyme is expressed in these tissues. This might indicate, together with the finding that in rainbow trout a CK isoenzyme termed TCK1 is expressed mainly in testis but only faintly in other tissues displaying CK activity [142], that testis has its own particular set of CK isoenzymes. Since in a few instances contradictory results were published for Mi-CK sequences, the respective references shall be shortly mentioned to avoid confusion in case the original literature is studied. In the cDNA sequence for human ubiquitous Mi-CK presented in Fig. 3 of the article written by Haas et al. [162], the 11 aminoterminal residues were erroneously replaced by the 10 corresponding residues of human sarcomeric Mi-CK, which the authors already knew at that time [162]. Furthermore, 4 of the 11 aminoterminal amino acids determined for human heart Mi-CK [50], 5 of the 120 amino adds specified for chicken ubiquitous Mi-CK [516] and 7 of the 40 aminoterminal amino acids determined for rat sarcomeric Mi-CK [82] by protein chemical methods do not agree with the respective cDNA sequences [163,344,508]. Finally, Benfield et al. [30] published a partial sequence of a rat CK gene differing from M- and B-CK which probably corresponds, in spite of a variety of differences, to rat ubiquitous Mi-CK [344]. The gene structures of human ubiquitous [162] and sarcomeric Mi-CK [244] differ considerably from those of the cytosolic CK isoenzymes. The Review 25 former genes span 5.5 and 37 kb, are located on chromosomes 15 [447] and 5 [244] and contain 9 and 11 exons, respectively. The lengths of the coding region exons as well as the locations of the exon junctions are absolutely identical between the two human Mi-CK isoenzymes, but differ considerably from those of the cytosolic CK isoenzymes, indicating that from an evolutio­ nary point of view, a first duplication even t of a common primordial CK gene resulted in ancestral mitochondrial and cytosolic CK genes. Only at a later stage, further gene duplications gave rise to the four different CK isoforms known for avian and mammalian species [244]. Very interestingly, exons 1 and 2 of the human sarcomeric Mi-CK are untranslated and followed by an 11 and 7 kb-intron, respectively [244]. Since sarcomeric contractile proteins were found to share a similar 5' organization, with an untranslatedfirst exo n followed by a relatively large intron [250], it has been hypothesized that this common feature is responsible for the coordinate transcriptional activation of sarcomeric Mi-CK and contractile proteins [244]. In contrast to M- and B-CK mRNA, Mi-CK mRNAs code, in addition to the native protein, for an aminoterminal mitochondrial target peptide which is proteolytically removed after import into the mitochondria [162,163, 196,344,347]. This target peptide is 39 amino acids long, rich in serine, threonine and basic amino acid residues, and has a characteristic tripartite structure [174]. Its amino terminus (residues 1-16) is positively charged and hydrophilic, and likely represents a mitochondrial matrix-targeting signal. Residues 17-32 constitute a stretch of uncharged amino acids which might act as a stop-transfer signal, therefore directing the protein into the intermembrane space. And finally, residues 33-39 are again positively charged and hydrophilic and are probably responsible for proper cleavage to yield the native enzyme. Sequence comparisons show that within the native protein, within the mitochondrial targeting peptide, as well as within the 3' and 5' untranslated regions, the sarcomeric Mi-CK isoenzymes on one hand and the ubiquitous Mi-CK isoenzymes on the other hand of chicken, man and rat are more closely related to each other than the sarcomeric and ubiquitous Mi-CK isoenzymes of one particular species, indicating that two different Mi-CK isoenzymes already appeared before divergent evolution of birds and mammals oc­ curred. This is in line with the likely presence of two different Mi-CK isoenzymes in frog tissues (see Π-D.). By the Needleman-Wunsch method, the native sarcomeric Mi-CK isoenzymes display 89-96 % amino acid se­ quence identity and the ubiquitous isoenzymes 91-96 %, whereas within the same species, the two Mi-CK isoenzymes have only an identity of 82-84 %. The fact that in rat and man the mitochondrial targeting peptides are much more conserved between the sarcomeric (87 %) and the ubiquitous Mi-CK isoenzymes (92 %), respectively, as compared to both Mi-CK isoenzymes of 26 CHAPTER 2

the same species (37 % in rat, 38 % in man), was taken as an argument for tissue-specific mitochondrial import receptors [344]. Furthermore, the marked isoenzyme-specific rather than species-specific conservation of the 3' untranslated regions indicates an important function of these stretches of DNA for Mi-CK gene expression. The mitochondrial and cytosolic CK isoenzymes represent a class of highly conserved proteins. When the sequences of the known isoenzymes are compared, one finds six blocks of very high homology separated by seven regions which are less conserved [196]. Among the latter are the amino- and the carboxy terminus. The highly conserved regions are likely to be involved in essential functions of the enzyme like catalytic activity and dimer forma­

tion. For example, affinity labelling of chicken sarcomeric Mib-CK with the ATP analogue T^-OJ^-chloroethyl-N-methylamino)] benzylamide ATP (C1RATP) led to the selective modification of Asp-335, a residue which is conserved in all Mi-CK and the great majority of the cytosolic CK sequences [223]. Since in C1RATP, the reactive label is attached to the γ-phospha te group of ATP, and on the basis of structural comparisons with adenylate kinase, Asp-335 was suggested to be involved in the binding of Mg2+ ions coordinated to the phosphate groups of ATP. The findings (i) of a partial loss

of enzymatic activity of chicken sarcomeric Mib-CK when Asp-335 is changed to a Thr, Asn, Ala or Lys residue by site-directed mutagenesis (Furter, R., unpublished data), and (ii) of a complete loss of enzymatic activity upon selective cleavage by proteinase К of M-CK between Ala-328 and Ala-329 [271,321] as well as of chicken sarcomeric Mi-CK between Ala-324 and Val-325 (Wyss, M., Schlegel, J. and Wallimann, T., unpublished data) further substantiate that besides the region around the "essential" Cys-278, stretches around Asp-335 are also important for substrate binding and catalysis. In contrast, the regions less conserved between the CK isoenzymes might be responsible for isoenzyme-specific properties like octamer formation, binding to cellular membranes, or binding to the myofibrillar M-band. In experiments with purified Mi-CK peptides, for example, Cheneval and Carafoli [82] found only one peptide, corresponding to the 25 amino terminal residues, which bound to cardiolipin-containing liposomes. Accordingly, the positively charged residues Arg-19, Lys-20 and His-21 have been impli­ cated to mediate the binding of Mi-CK to the negatively charged phospholipids of the mitochondrial membranes [82]. As a matter of fact, these three amino add residues are absolutely conserved among all amino termini known for the sarcomeric and ubiquitous Mi-CK isoenzymes from chicken [196,508,516], man [162,163], and rat [82,344], as well as for mouse brain Mi-CK [344], but not among the cytosolic CK isoenzymes. Review 27

III. Functional studies of Mi-CK

Ш-А. Species and tissue distribution

Highest enzymatic activities and mRNA levels of Mi-CK are present in tissues with high and fluctuating energy demands like heart [13,50,82,156,167,210,213,338,344,363,377,396,410,411, 441], skeletal muscle [31,34,173,196,210,281,293,326,344,377,410], brain [53,210,213,268,344, 397,456,504,512,516], retina [41,268,490,496] and spermatozoa [463-465,489]. Smaller amounts of Mi-CK and Mi-CK mRNA are found in smooth muscle- containing tissues like uterus [163,344], placenta [344], intestine [41,63,162, 206,207,213,239,344], vas deferens as well as aorta [207]. In contrast, no Mi-CK activity was detected in chicken gizzard in spite of its high cytosolic BB-CK activity [396]. It has been hypothesized that this difference in Mi-CK content between distinct smooth muscle types is due to the different physiological functions of these muscles [207,492]; whereas most smooth muscles respond to stimulation with very long tonic contractions, carbachol and electric field stimulation in chicken gizzard elicited only phasic con­ tractions without a tonic component. Maximal force was developed within 10-15 s, and almost complete relaxation was accomplished in less than a minute [126]. Therefore, the CK/PCr system in chicken gizzard has to "buffer" [ATP] which can easily be accomplished by the extraordinarily large amounts of BB-CK present, with no need for an additional mitochondrial isoenzyme. In contrast, "transport" of PCr and Cr between Mi-CK and cytosolic CK isoenzymes may be a prerequisite for the proper functioning of tonic smooth muscles (see also IV-B.). In contrast to earlier studies [210,213], significant amounts of Mi-CK activity and mRNA were also observed in rat kidney [1,41,137,344], where Mi-CK, together with BB-CK, was found in cortex and outer medulla, probably supporting sodium transport in the distal nephron [137]. Liver was reported to contain either no or only minute amounts of Mi-CK activity and mRNA [41,163,210, 213,344,518]. Nevertheless, Kanemitsu et al. [231] suc­ ceeded in purifying Mi-CK from human liver and provided some, albeit weak evidence indicating that this isoenzyme differs from human heart Mi-CK. Finally, Mi-CK might become increasingly important as a valuable diagnostic tool since it was also found in certain tumor cells as well as in serum (see ΠΙ-G.). These data show that Mi-CK expression is regulated ¿и α físswe- and cell type-specific fashion. In all instances, however, Mi-CK is coexpressed with at least one of the cytosolic CK isoenzymes MM-, MB- or BB-CK. As discussed in Π-R, it is now clear that at least two different Mi-CK isoenzymes exist and are products of distinct genes. Northern blot analysis 28 CHAPTER 2

revealed that mRNA for the isoenzyme termed sarcomeric Mi-CK is almost exclusively expressed in heart and skeletal muscle [163,196,244,344]. In contrast, mRNA for ubiquitous Mi-CK was mainly found in kidney, placenta, intestine and brain. Seemingly, sarcomeric Mi-CK is the counterpart of M-CK and ubiquitous Mi-CK that of B-CK. The relative proportions of Mi-CK vary considerably between different tissues and species. Mi-CK amounts to 0.5-15 % of total CK activity in brain [48,245,335,391,423] and to 10-15 % in kidney [137]. Since in fast-twitch skeletal muscle, the percentage of Mi-CK is rather low [48,173, 294,326,410,434,520], while it is considerably higher in slow-twitch skeletal muscle [210,520], the proportion of Mi-CK can be increased within a particu­ lar muscle by endurance training [9] or chronic stimulation [400]. The highest proportions of Mi-CK are found in the heart, where Mi-CK was reported to constitu te

(Mi-AK) has been described for crustacean and crab muscles [78,105,106,113] as well as for Drosophila [325], whereas most of the molluscs, insects and annelids lack Mi-AK [for a review see Ref. 113].

Ш-В. Intracellular localization of Mi-CK: Binding to mitochondrial inner and outer membranes

In most studies, Mi-CK was found to be enriched in the mitochondrial fraction [103,210,326,338, 410,441], and immunohistochemistry with anti­ bodies directed against Mi-CK also revealed an accumulation of this isoenzyme exclusively within mitochondria [396,489,490,496]. Nevertheless, CK of cathodic mobili ty resembling Mi-CK was found in the nuclear fractions of rat, human and pig heart and skeletal muscle [117,228]. Since some mitochondrial subfractionatio n experiments were performed in phosphate- [246, 350,509] or KCl-containing solutions [278,498], these conditions being known to release Mi-CK from the mitochondrial mem­ branes (see II-A.), it is not surprising that Mi-CK was suggested to be soluble in the mitochondrial intermembrane space. However, in a number of subse­ quent studies it was proven, instead, that Mi-CK is bound to the outer surface of the inner mitochondrial membrane. Rotenone, a strong inhibitor of the respiratory chain, inhibited the solubilization of Mi-CK from mitochondrial membranes by phosphate [122]. Subfractionation studies in hypotonic tris/ phosphate, hypotonic tris/hydrochloride or isotonic sucrose solutions as well as with digitonin revealed that most of the Mi-CK remains associated with the inner membrane/matrix fraction [3,64,104,208,213,409, 411]. Fur­ thermore, the facts thatMi-CK activity couldnotbe inhibited by atractyloside, a competitive inhibitor of adenine nucleotide transport across the inner membrane [213,480], and that Mi-CK can be released by phosphate only from mitoplasts, but not from intact mitochondria [130] both demonstrate that Mi-CK is located inside the outer membrane, but on the "cytoplasmic" side of the ANT. Finally, electron microscopical cytochemistry and immunohisto­ chemistry revealed an accumulation of Mi-CK protein along the inner mitochondrial membrane [13,380,396,421,496]. Recent evidence suggests that Mi-CK is not distributed at random over the inner membrane, but may be accumulated in contact sites (CS) between the inner and outer mitochondrial membrane. CS are thought to play an important role in the mitochondrial production of high-energy phosphates, because they were found to increase in extent and number during oxidative phosphorylation [40,58,68,249]. In subfractionation experiments of rat brain and kidney mitochondrial membranes, Mi-CK was predominantly found, in 30 CHAPTER 2

association with hexokinase I, in a fraction enriched for CS [1,59,256,257]. It is likely that Mi-CK in that fraction binds simultaneously to both membranes and is functionally coupled to the ANT in the inner and to porin in the outer membrane [61] (see ПІ-С). These conclusions are corroborated by the following facts:(i) 40-50 % of the total Mi-CK in rat brain mitochondria was not readily accessible to external substrates and could only be released from the membranes by rather high concentrations of digitonin [1,33,257]. (ii) After selective disruption of the ou ter membrane by digitonin, or in isolated contact site fractions, König's polyanion, known to inhibit the transport of adenine nucleotides through porin [32], inhibited Mi-CK activity by 50 % [1,257]. (iii) In inhibition experiments on intact or digitonin-treated mitochondria, 50 % of the Mi-CK was not accessible to externally added PCr and iodoacetate, the latter substance being known efficiently to inhibit purified Mi-CK. A reasonable explanation for these findings is that about half of the total Mi-CK is buried between the two mitochondrial membranes in the CS, where porin is thought to be in its cation-selective s tate, thus creating a microcompartment that is fairly inaccessible to externally added anions [257]. The localization of Mi-CK within the CS of intact mitochondria in situ was further corroborated by electron microscopical immunohistochemistry showing an accumulation of gold particles at places where inner and outer membranes approach each other [396,496], and by electron microscopical cytochemistry displaying a CK-dependent trapping of stain within CS of rat heart, brain, retina photoreceptor cells, skeletal muscle and kidney [40,41]. Dinitrophenol and amytal, known to reduce the number of CS [58], also diminished the proportion of Mi-CK-derived stain within CS [40,41]. The ability of Mi-CK to simultaneously interact with two different membranes was directly investigated in an in vitro system generated (i) by a monolayer of outer membrane lipids or spread outer membranes, and (ii) by addition to the subphase of radioactively labelled large unilamellar vesicles (LU V's) of inner membrane lipids or inner membranes, respectively. Mi-CK proved to be highly efficient in inducing close contacts between the monolayer and the vesicles, the octameric form being more potent than the dimeric one [333,370]. In contrast to Mi-CK, polylysines not only mediated contacts between LUV's and monolayer, but in addition caused membrane fusion between them, suggesting that the rather large side-length of the Mi-CK octamer is an essential prerequisite to avoid fusion between two opposing membranes. These results, however, do not imply that Mi-CK is an essential prerequisite for the formation of membrane contacts, since CS were also found in rat liver mitochondria known to be devoid of Mi-CK [68,249]. Instead, they clearly indicate that the Mi-CK octamer is a structure ideally suited for a contact site localization and possibly also for a stabilization of Review 31 such contacts. Electron micrographs of Mi-CK are fully in line with this interpretation, displaying two identical top and bottom faces for the octameric molecule and therefore suggesting that both faces have the same potential to interact with membranes (see П-С). Once it was established that Mi-CK interacts with mitochondrial mem­ branes, the biochemical nature of the "receptor" for Mi-CK binding was investigated. Because most Mi-CK isoenzymes can be released from the mitochondrial membranes by a variety of different salts in a ionic strength- dependent manner (see also П-А.), a mainly electrostatic interaction of positively charged Mi-CK with negatively charged phospholipids of the membranes was assumed [64,278]. Cardiolipin seems to be the most likely candidate, since (i) experiments on artificial liposomes composed of defined mixtures of phospholipids demonstrated that Mi-CK preferentially binds to cardiolipin [324,369], (ii) the binding of Mi-CK to cardiolipin-containing liposomes, aqueous dispersions of cardiolipin, as well as to heart and liver mitoplasts was inhibited by adriamycin [64,83,324,331], a clinically useful anticancer drug which was suggested to specifically interact with cardiolipin [80], (iii) treatment of rat heart mi tochondria with phospholipase A2, but not with phospholipase С led to the release of Mi-CK, therefore also indicating that the interaction is critically dependent on cardiolipin [395], (iv) both inner and outer mitochondrial membranes, but not other cellular membranes, contain cardiolipin [197], and (v) Mi-CK (see above) as well as cardiolipin [10] are highly accumulated in CS. In experiments with cyanogen bromide- digested rat heart Mi-CK, only one peptide representing the 25 aminoterminal residues bound to cardiolipin-containing liposomes, whereby this interac­ tion was also inhibited by adriamycin [82]. In addition, chemical modifica­ tion of arginine and lysine residues of Mi-CK drastically reduced the binding of the enzyme to the liposomes. Therefore, it was suggested that the positively charged residues Arg-19, Lys-20 and His-21 are responsible for the specific binding of Mi-CK to cardiolipin [82,344]. In fact, these three residues are absolutely conserved among all known Mi-CK sequences (see Π-F.). How­ ever, it should be kept in mind that the interaction of Mi-CK with cardiolipin might not be as specific as originally thought. First, Mi-CK proved to bind also to other negatively charged phospholipids like phosphatidyl serine or phosphatidyl inositol [81369] as well as to monolayers of spread microsomal membranes [369] known to be virtually devoid of cardiolipin [94]. Further­ more, two different classes of Mi-CK binding sites were found on mitoplasts, one with high and the other with low affinity [285,395]. The number of high- affinity binding sites approximately equalled the amount of Mi-CK in heart mitochondria [395], but in total, two to fourteen times this amount could be rebound to heart and liver mitoplasts [64,131,168,324,395]. Therefore, 32 CHAPTER 2

cardiolipin may constitute the low-affinity binding si tes. The isoelectric point of octameric and dimeric Mi-CK may be a last argument for a ionic interaction of this isoenzyme with the mitochondrial membranes [354]. Mi-CK octamers have a higher isoelectric point than the corresponding dimers (see Π-D.), and, as a matter of fact, octameric Mi-CK was found to be the exclusive or predominant oligomeric form being bound to the mitochondrial membranes (see ni-D.). Whereas Cheneval et al. [83] obtained no evidence for penetration of Mi-CK into a membrane bilayer, monolayer experiments with spread mitochondrial membranes or membrane lipids clearly showed an increase in

surface pressure upon binding of chicken sarcomeric Mib-CK, suggesting that Mi-CK penetrates into the hydrophobic domain, at least under the experimental conditions used [369,370]. This phenomenon can be ascribed to the hydrophobic nature of Mi-CK (see II-A.) and is also in agreement with experiments on the extraction of Mi-CK from chicken heart mitoplasts which were indicative for a small nonpolar effect [64]. Furthermore, detergents had to be used for an optimal release of Mi-CK from rat heart [395], human skeletal muscle [434] and sea urchin sperm mitochondria [463,464317]. As an alternative to cardiolipin, ANT may be suggested to act as a "receptor" protein for Mi-CK [376,380], since a variety of studies revealed a functional coupling between Mi-CK and ANT of the inner membrane (see Ш- C). However, incubation of Mi-CK with ANT-containing liposomes [324] as well as crosslinking experiments [406] failed so far to demonstrate such a relationship. In addition, electrostatic repulsion of the two very basic proteins was taken as an argument against a direct interaction between Mi-CK and ANT [278]. However, a specific interaction of certain domains of both proteins can not be ruled out by the mere fact that both proteins have an overall positive charge. In conclusion, Mi-CK is primarily attached to the outer surface of the inner mitochondrial membrane, but, within CS, additionally interacts with the outer membrane as well. Binding of Mi-CK molecules exclusively to the outer membrane was never observed in situ. The interaction of Mi-CK octamers with the mitochondrial membranes seems to be mostly due to an electrostatic interaction of positively charged Mi-CK with negatively charged phospholipids like cardiolipin, but hydrophobic interactions as well as the existence of a specific receptor (protein) can not yet be ruled out. The question of whether the release of Mi-CK from the membranes into the intermembrane space by phosphate ions or other effector molecules is of physiological importance remains to be elucidated. Rojo et al. [370] recently showed that Mi-CK octamers and, to a much lesser extent, Mi-CK dimers are able to mediate contact formation between two mitochondrial membranes. This Review 33 data, together with the finding of functional coupling between Mi-CK and ANT of the inner as well as porin of the outer mitochondrial membrane provoked alternative ideas about the functional involvement of Mi-CK in CS, as will be discussed in the next chapter.

Ш-С. Functional coupling

Besides the still existing controversy in assigning an exclusive "buffer" or "transport function" to the CK/PCr system (see IV-B.), the question of how PCr production is coupled to mitochondrial oxidative phosphorylation is also heavily disputed. Therefore, this topic will be discussed here in more detail. Already in 1911, Thunberg [461] realized that addition of Cr to a respiring muscle homogenate results in increased oxygen uptake. In isolated respiring mitochondria from cardiac and skeletal muscle as well as brain, Cr together with catalytic amounts of ADP or ATP also stimulates oxygen consumption [34,122,153,213,319,377,381,439,480,521]. This is due to Mi-CK continuously regenerating ADP from ATP, produced by oxidative phosphorylation, and Cr. The ADP, in turn, further stimulates oxygen consumption. The stimulatory effect is dependent on the concentration of Cr, with Cr-stimulated respiration approaching the state-3 rate [34,213]. Under the conditions described, PCr is the net product of oxidative phosphorylation [54,213,319,374,377]. Most importantly, the Mi-CK content of heart mitochondria is sufficient to transphosphorylate all of the ATP produced by oxidative phosphorylation to PCr at all rates of respiration [213,374,381,480]. A large number of subsequent studies demonstrated that on one hand, Mi-CK has "privileged access" to ATP, produced in the mitochondrial matrix by oxidative phosphorylation, over externally added (extramitochondrial) ATP, and that on the other hand, the ANT preferentially transports ADP liberated by Mi-CK over externally added ADP. In other words, Mi-CK and oxidative phosphorylation seem to be functionally coupled in some way or another. The major evidence supporting this view shall now be summarized:

(a) For PCr production, the Km value for ATP of intact mitochondria [119,120,216,380,381] ormitoplasts [384] is significantly lower when ATP is produced within the matrix by oxidative phosphorylation, than when

ATP is added externally. Due to this decreased Km value for ATP, PCr synthesis by intact, respiring mitochondria, compared to rotenone- plus oligomycin-inhibited mitochondria with externally added ATP, is less susceptible to product inhibition by PCr [216,381]. In other words, at identical ATP and Cr concen trations in the surrounding medium, the rate of PCr production is much higher when oxidative phosphorylation is 34 CHAPTER 2

activated compared to when it is inhibited [375,376,381] and therefore is also much higher than calculated from the substrate concentrations in the medium and the kinetic constants of Mi-CK [375]. (b) Similarly, ADP produced in the intermembrane space by Mi-CK proved

to be much more potent (lower Km value) in stimulating oxidative phosphorylation than ADP which was externally added or produced by hexokinase [65,218,220,320]. Accordingly, much less atractyloside/mg mitochondrial protein was needed to suppress respiratory stimulation by externally added ADP than for stimulation by ADP produced by Mi-CK [320,384]. Furthermore, the uptake of [3H]-labelled ADP or ATP into respiring mitochondria was found to be high in the absence, but much lower in the presence of Cr or PCr, respectively, thus suggesting as well that ADP or ATP produced by Mi-CK has preferential access to the ANT over externally added ADP or ATP [20,266,385]. (c) Externally added phosphoenolpyruvate (PEP) plus pyruvate kinase (PK) fully suppressed respiratory stimulation when ADP was produced in the extramitochondrial compartment by hexokinase, but only partially di­ minished the stimulatory effect of ADP produced in the intermembrane compartment by Mi-CK [145,147,148,257,384,386]. (d) After short incubations of respiring rabbit heart mitochondria with γ-[32Ρ]-ΑΤΡ and Cr, the specific radioactivity of PCr was only 36 % of that of ATP [522]. However, if mitochondrial respiration was uncoupled by carbonylcyanide m-chlorophenyl hydrazone or inhibited by atractyloside, the specific radioactivi ties of PCr and ATP were the same. These findings indicate that intra- and extramitochondrial ATP do not readily equili­ brate and that Mi-CK has preferential access to intramitochondrial ATP. The same conclusions were drawn from similar labelling experiments with γ-[32Ρ]-ΑΤΡ or [33P]-phosphate [118-120]. (e) Under identical experimental conditions, intact rat heart mitochondria produced 2.3 times the amount of PCr as rat liver mitochondria contain­ ing no Mi-CK themselves, but to which the same quantity of Mi-CK had been added externally [380]. (f) Whereas all experiments described so far deal with kinetic effects, a thermodynamic approach was chosen by DeFuria et al. [98] and Saks et al. [384,386] to shed more light on Mi-CK function. Normally, the mass action ratio of the CK substrates,r= [Mg ADP"] [PCr2"] [H+] / [Mg ATP2"] [Cr], relative to the equilibrium constant K, determines in which direction the reaction will proceed. An approximation to the equilibrium state of the reaction is reflected by a diminuition of the difference between Γ and K. Whereas partially purified Mi-CK or Mi-CK in inhibited, non-respiring mitochondria obeyed this thermodynamic law, PCr production in re­ spiring mitochondria or mitoplasts still proceeded even when Γ > К. Review 35

Although some experiments described above have been criticized and conflicting reports have been presented [4,5,54,55,279], it is now generally accepted that a functional coupling between Mi-CK and oxidative phosphorylation indeed exists. Two alternative explanations have mainly been proposed as basis for this coupling. On one hand, Mi-CK may be bound to the inner membrane in close proximity to or directly associated with the ANT, thus creating a microenvironment enabling efficient metabolic chan­ nelling of substrates between the catalytic sites of both enzymes [20,145,218, 220,266,267,320,375,376,380-382,384-387]. On the other hand, the outer mitochondrial membrane might act as a diffusion barrier for adenine nucleotides, leading to concentration gradients over the outer membrane [38,65,98,118,119,147,148,218,220]. In this latter case, substrate concentra­ tions in the surrounding medium would not reflect those experienced in the intermembrane space. Whereas most of the arguments mentioned above (a-f) are compatible with both ideas, several experiments favour a direct interaction between Mi-CK and ANT. Phosphate and high concentrations of chloride ions are known to release Mi-CK from mitoplasts (see Π-Α.) and can therefore also be assumed to detach Mi-CK from the membranes in intact mitochondria. Incubation of intact mitochondria in phosphate- or chloride-containing media significantly raised the Km values of Mi-CK for ADP and ATP [474] and drastically reduced the functional coupling between Mi-CK and oxidative phosphorylation [169,170,267,387], suggesting that the binding of Mi-CK to the membranes and possibly to ANT is mainly responsible for the kinetic coupling. The observation of functional coupling not only in intact mitochondria, but also in mitoplasts, supports this view [384,386]. Further­ more, in rat heart mitoplasts, rabbit anti-rat Mi-CK antibodies affected neither Mi-CK nor ANT activity, whereas chicken anti-rat Mi-CK antibodies inhibited Mi-CK and ANT activity as well as the rate of oxidative phosphorylation approximately to the same extent [266,382,385]. Removal of Mi-CK from the mitoplast membranes by 150 mM KCl plus 20 mM ADP completely abolished these inhibitory effects, thus suggesting an intimate functional and structural interaction of Mi-CK and ANT, with the catalytic sites of both enzymes lying side by side. In contrast, the ANT inhibitors atractyloside and carboxyatractyloside hadnoeffectonthekineticpropertiesofMi-CK-innon-respiringmitochondria, arguing that the active sites of Mi-CK and ANT do not simply overlap [146,474]. Instead, the outer mitochondrial membrane as a diffusion barrier for adenine nucleotides was suggested to be responsible for the functional coupling on the basis of the following experiments: 36 CHAPTER 2

(a) Labelling studies with pPj-phosphate demonstrated that selective da­ mage or removal of the outer mitochondrial membrane by digitonin abolishes the functional coupling between Mi-CK and oxidative

phosphorylation [119]. Furthermore, approximately equal Km values of Mi-CK for ATP were found for respiring, digitonin-treated mitochondria, for atractyloside-inhibited, digitonin-treated mitochondria, as well as for respiring intact mitochondria, whereas the corresponding value for atractyloside-inhibited, intact mitochondria was significantly higher.

(b) Under conditions where 70 % of Mi-CK was membrane-bound, the Km value of Mi-CK for ATP in respiring as well as in oligomycin-inhibited mitoplasts was indistinguishable from the corresponding value of the soluble enzyme, while it was significantly lower in respiring and much higher in oligomycin-treated intact mitochondria [65]. (c) As already mentioned above, externally added pyruvate kinase (PK) plus phosphoenolpyruvate (PEP) fully suppressed the respiratory stimulation of ADP produced by hexokinase [145,147,148,386]. However, if ADP was liberated either by membrane-bound Mi-CK or by the intermembrane space protein adenylate kinase, PK/PEP in both cases inhibited the respiratory stimulation by ADP only partially and approximately to the same extent, indicating that diffusion limitation by the outer membrane is more likely to be responsible for the functional coupling of Mi-CK and ANT than interaction of Mi-CK with the inner membrane [148].

Finally, membrane energization, alkalinization near the mitochondrial mem­ branes as well as unstirred layer effects were also taken into consideration to explain the functional coupling, but most of these possibilities have been dismissed as sole explanation for the observed phenomena, although in combination with each other they may play an important role [147,169,218,220,320,381,474]. To conclude, the experiments performed sofar do not allow a definite answer to be given on the true basis for the functional coupling between Mi-CK and oxidative phosphorylation. In fact, depending on the metabolic state, one, two or all of the mechanisms mentioned above may be operational at the same time. However, many recent experiments argue for a third explanation of the functional coupling, thatofmicrocompartmen ta tion and metabolite channel­ ling within the CS. Four new aspects have opened the door to this fascinating, alternative explanation (Fig. 2; page 38). First, porin of the outer mitochondrial membrane was recognized as a voltage-dependent channel for polar metabolites and is therefore also known as voltage-dependent anion-selec- tive channel (VDAC) [87]. Below a membrane potential of 30 mV, porin is anion-selective, whereas above 30 mV, it adopts a different state which is Review 37 characterized by a lower conductance and cation-selectivity [125,135,291,371]. In this latter state, ADP and ATP are excluded from transport by porin [32,33,59,61]. It was assumed that within the CS between the mitochondrial inner and outer membranes, the cation-selective state of porin is induced by "capacitive coupling" to the membrane potential of the inner membrane. Beyond the CS, where the distance between inner and outer membrane would be too large to allow capacitive coupling, porin is thought to be in its anion-selective state [33,61]. Second, Mi-CK was found to be enriched in mitochondrial CS where its activity is "latent" [1,59,257] (see also ΠΙ-Β.). In other words, Mi-CK within the CS is not readily accessible to externally added substrates. Furthermore, König's polyanion, known to block porin, reversibly inhibited the activity of Mi-CK within the CS by cessation of the adenine nucleotide diffusion through the outer membrane [257]. And finally, Mi-CK within the CS was neither accessible to product inhibition by the negatively charged substrate PCr nor to irreversible inhibition by negatively charged iodoacetate [257]. Third, structural analysis of Mi-CK revealed an intriguing octameric structure with two identical top and bottom faces and either central indentations in these two faces or a channel through the molecule (see П-С). And fourth, Mi-CK was found to be able to interact simultaneously with two different membranes probably due to its identical top and bottom faces [370]. In the two alternative models depicted in Fig. 2, all of these findings were taken into account. Within the CS, Mi-CK is buried between the two mitochondrial membranes where it is functionally (and structurally?) cou­ pled to ANT of the inner and porin of the outer mitochondrial membrane. It may even be hypothesized that octameric Mi-CK, tetrameric ANT [484], oligomeric (tetrameric?) porin [259,276] and possibly also tetrameric hexokinase [519] form a highly ordered multi-enzyme complex allowing efficient substrate channelling. In the CS, Mi-CK has preferential access to ATP produced by oxidative phosphorylation within the mitochondrial matrix. According to the first model [61,492] (Fig. 2A), Cr is "presented" to Mi-CK by porin in its cation-selective state. Mi-CK activity liberates ADP, which is preferentially and directly re-taken up into the matrix, as well as PCr, which as the net product of oxidative phosphorylation leaves the mitochondria beyond the CS through porin in its anion-selective state. This model has the disadvantage that newly synthesized PCr would equilibrate with the substrate pools in the remaining intermembrane space before its export out of the mitochondrion. Accordingly, it seems unlikely that the cytosolic phosphorylation potential in this case could be sensed by the mitochondria, either within or beyond the CS. This problem is circumvented by the second model (Fig. 2B). ANT, Mi-CK 38 CHAPTER 2

(A) Review 39

Fig. 2. Structure and function of Mi-CK mthin the mitochondrial intermembrane space with special reference to the contact sites. Within CS (surrounded by a dashed line), exclusively octameric Mi-CK interacts with both the inner (IM) and the outer mitochondrial membrane (OM) and is thought to be functionally coupled to ANT of the inner and porin (P) of the outer membrane. Below the contacts, Mi-CK may be dimeric or octameric, and it may be either bound to the inner membrane or free in the intermembrane space. In contrast to Mi-CK, adenylate kinase (AdK) is not firmly bound to the membranes. In both models, it is schematically indicated that two different microcompartments for CK substrates exist within and beyond the CS. Due to this microcompartmentation, PCr production by Mi-CK may still proceed even at a high cytosolic ATP/ADP ratio. Note that by the schematic representation it is not intended to suggest that Mi-CK is only active within the CS. Furthermore, it should not be inferred that below the CS, the two mitochondrial membranes are far from each other. In fact, electron microscopic analysis of insitu mitochondria revealed that the mitochondrial inner and outer membrane form a 5-layered structure with a width of only 12 nm, whereas a 7-layered structure would be expected for two separated membranes [295]. (A) Within CS, porin is in its cation-selective state and therefore only permeable to Cr. Porin beyond the CS is in i ts anion-selective state and thus allows diffusion of ATP, ADP and PCr through the outer membrane. (B) Tetrameric ANT, octameric Mi-CK and (tetrameric) porin form a highly organized multienzyme complex and thereby create a microcompartment allowing efficient substrate channelling between the three enzymes. PCr is produced within the central channel of the octameric Mi-CK molecule and is directly pulled out of the mitochondria by electrostatic repulsion through the superimposed porin molecule. For simplicity, the subunit "boundaries" were orni tted in the Mi-CK octamer localized within the CS. For further details see the text.

and porin form a contiguous channel spanning both mitochondrial mem­ branes. PCr is produced within the channel and is directly exported to the cytosol where it rapidly equilibrates with the cytosolic PCr/Cr and ATP/ADP pools through the action of the cytosolic CK isoenzymes which are thought to be in a near-equilibrium state [205,262,298,310]. PCr production even at a high cytosolic phosphorylation potential may be accomplished by "active" expulsion of PCr from the channel possibly due to electrostatic repulsion. On the other hand, the cytosolic phosphorylation potential could be sensed by the mitochondria in regions beyond the CS. Clearly, elucidation of the structural and functional involvement of octameric Mi-CK in the CS will be an exciting task for future studies and may establish the existence of a multi-enzyme complex suited for metabolic channelling [442,443]. Independently of both models, Mi-CK is not restricted to the CS, as was shown by in situ localization studies [396,496] and biochemical experiments. In rat brain mitochondria, only 50 % of the total Mi-CK activity was "latent" and not susceptible to inhibition by PCr or iodoacetate [1,257]. Furthermore, 40 CHAPTER 2

since CS are dynamic structures changing in number and extent depending on the metabolic state of the mitochondria (see Ш-В.), the proportion of Mi-CK within the CS is also supposed to correlate with the respiratory rate. Accordingly, dpher 1 in Fig. 2 indicates the dynamic equilibrium between the two Mi-CK localizations "in" and beyond" the CS. Whereas it is highly likely that exclusively octameric Mi-CK is present in the CS [257,370], reversible changes in the dimer to octamer ratio (ciphers 2 and 4) as well as dissociation and reassociation of Mi-CK from and to the inner membrane (ciphers 3 and 5) may be important regulatory parameters beyond the CS [399] (see also III-D.). Usually, dissociation of Mi-CK octamers into dimers is accompanied by the release of Mi-CK from the membranes, and vice versa [296,399]. In turn, release of Mi-CK from the inner membrane may abolish the functional coupling between Mi-CK and ANT, thus potentially diminishing the overall Mi-CK reaction rate. Finally, Mi-CK dimers, when bound to the inner membrane, reassociate to octamers, a process which is probably facilitated by lateral diffusion of the dimers on the membrane [399]. The involvement of Mi-CK in a specialized structure like the CS is highly indicative for a metabolic advantage of this location. The facts that Mi-CK within the CS is "buried", that it is protected from inhibition by PCr and iodoacetate, and that its activity is "latent" strongly suggest that CS represent a special microcompartment with substrate concentrations differing from those in the remaining intermembrane space as well as in the extramitochondrial compartment(s), an assumption that has been corrobo­ rated by experiments on hexokinase and glycerol kinase bound to mitochondrial CS [36,58,155,161; for reviews see Refs. 1,60,61]. In liver mitochondria, for example, hexokinase and adenylate kinase usually share the same adenine nucleotide pool. However, 5 μΜ Ca2+, known to increase the number of CS, resulted in metabolic separation of the two enzymes [58]. Interestingly, in intact respiring rabbit heart mitochondria containing both hexokinase, bound within the CS at the outer surface of the outer membrane to porin, as well as Mi-CK, the formation of glucose-6-phosphate was depressed by creatine at a low ATP/ADP ratio. However, glucose had no effect on PCr production, indicating that Mi-CK has preferential access over mitochondrial hexokinase to ATP produced in the matrix [55]. Accordingly, even within the CS, rapid equilibration of substrate concentrations does not occur. The assumption of a separate microcompartment for adenine nucleotides within the CS (see also IV-C.) has dramatic consequences for total cellular energy metabolism. First, it may explain why the cytosolic ATP/ADP ratio is considerably lower in liver than in muscle [179,247,440]. In liver lacking Mi-CK, the inner membrane potential seems to be the primary determinant Review 41 of the cytosolic ATP/ADP ratio. In muscle, however, the functional coupling of Mi-CK and ANT within the CS, the particular substrate concentrations experienced in this microcompar tment, and the vectorial export of PCr out of the mitochondria before equilibration with the substrate concentrations beyond the CS may allow the build-up of a higher ATP/ADP ratio and therefore also of a higher phosphorylation potential in the cytosol. Second, the localization of Mi-CK, hexokinase, glycerol kinase and nucleoside diphosphate kinase within the CS [1,2,59] seems to be especially relevant, since all of them, due to their involvement in the regulation of the energy metabolism and/or in the export of high-energy phosphates out of the mitochondria, influence the cytosolic phosphorylation potential. In contrast, adenylate kinase, thought to protect the cell from an accumulation of ADP [492], is not found within the CS and is not or only partially bound to mitochondrial membranes. And third, microcompartmentation within the CS may account for the finding that mitochondrial PCr production still proceeds even if, on the basis of total (intra- and extramitochondrial) substrate concentrations, the opposite CK reaction direction is favoured [98,384,386]. To conclude, it is now largely accepted that functional compartmen tation of Mi-CK within the mitochondrial intermembrane space is an essential prerequisite for efficient mitochondrial PCr production. However, it is not yet completely clear whether this compartmentation is due only to an interaction of Mi-CK with ANT, to diffusion limitations by the outer mitochondrial membrane, to both of these effects, or to microcompartmentation and metabolic channelling within the CS. A variety of recent studies favour the latter hypothesis. Consequently, in any description or mathematical modelling of mitochondrial PCr production in particular and of total cellular energy metabolism in general, the particular substrate concentrations expe­ rienced in the CS, together with the fact that the substrate concentrations within the CS are no t in equilibrium with those of the cy tosol, have to be taken into account. The same holds true for the interpretation of in vivo 31P-NMR measurements.

III-D. Dynamic regulation of the octamer to dinter ratio

Before anything was known about the molecular size of Mi-CK, two distinct cathodically migrating bands with Mi-CK activity were often observed using cellulose acetate, agarose and acrylamide gel electrophoresis at pH 8.0-8.8 [166,222,268,374,502]. These two bands were readily interconvertible, there­ fore indicating that they were not due to different Mi-CK isoenzymes [166]. 42 CHAPTER 2

Instead, they could be assigned to two distinct oligomeric forms of the same isoenzyme [100,167,504]. The band moving faster to the cathode corresponds

to octameric Mi-CK with a Mr of approximately 350 000, and the more slowly

migrating band represents dimeric Mi-CK with a Mr of approximately 85 000 [231,296,516]. The existence of two different, readily interconvertible oligomeric forms was subsequently corroborated by several techniques (see Π-Β.). In only two experiments, acrylamide gel electrophoresis [100] and isoelectric focussing of bovine heart Mi-CK [282], additional Mi-CK bands were observed which were interpreted as tetrameric, hexameric and/or aggregated Mi-CK. However, in the light of the uncertainties inherent in the methods employed and the overwhelming body of evidence demonstrating no intermediate forms between dimers and octamers, these findings have to be seriously questioned. Once it was realized that the two distinct oligomeric forms of Mi-CK might be of physiological importance, several groups searched for factors influencing the dimer to octamer ratio in vitro. This ratio is not influenced by the temperature (5-20oC) and the nature of the monovalent anions (chloride, nitrate or acetate) or cations (Na+ or K+) in the buffer used at ionic strengths between 0.02 and 0.25 M [280,281 ]. In contrast, it is strongly dependent on the Mi-CK concentration itself. The proportion of octamers rises as the Mi-CK concentrationisincreased[26,100,123,166,167,287,396,397,516].Thedimerto octamer ratio also depends on the pH value of the medium, with dissociation into dimers being more pronounced at alkaline pH values [280- 282,296,396,483]. Furthermore, partial or complete dissociation of octamers into dimers can be achieved by exceedingly high concentrations of 2-mercaptoethanol [100,166,167,448,502,504], freezing and thawing [281,396], low ionic strength conditions [281,282], 1 M KCl [296], p-hydroxy- mercuribenzoate (Wyss, M. and Wallimann, T., unpublished data) and 1-8 M urea [26,123,156,231,296,403]. The mode of action of 2-mercaptoethanol is unclear since several weak and strong oxidizing agents did not reverse its effect [502]. Upon incubation of Mi-CK octamers with urea, dimers are first formed and only on a much slower time-scale, or at higher urea concen­ trations, monomers also appear [26,403], indicating that the intersubunit interactions are much stronger within a dimer than between adjacent dimers within an octamer (see also II-C.). Single CK substrates and "unproductive" substrate combinations on one hand were reported not to influence the dimer to octamer ratio [280,281,296,396,483]. On the other hand, 4 mM MgADP or MgATP led to

partial dissociation of octameric chicken ubiquitous Mia-CK into dimers [398], suggesting that the stability of octamers differs between Mi-CK isoenzymes. In contrast, equilibrium substrate combinations (MgADP + Review 43

MgATP + Cr + PCr) and formation of a transition state-analogue complex of Mi-CK (Mi-CK + MgADP + Cr + nitrate) in all instances resulted in fast dissociation of the octamers [280-282,287,296,396,399,483,516]. However, even in the presence of MgADP, Cr and nitrate, the dimer to octamer ratio still depended on the Mi-CK concentration, but at any given concentration, the proportion of octamers was lower than in the absence of substrates [516]. At last, Lipskaya and co-workers [280,281 ] concluded from a limited set of data that at physiological concentrations of ATP and ADP, the dimer to octamer ratio of bovine heart and pigeon breast muscle Mi-CK correlates with the Cr/ PCr ratio. This finding may be explained by the fact that with changes in the Cr/PCr ratio, the probability that Mi-CK simultaneously binds two comple­ mentary substrates (ADP + PCr or ATP + Cr) also varies. Simultaneous binding of two complementary substrates in its turn is expected to have the same effect as formation of a transition state-analogue complex of Mi-CK, namely to dissociate octamers into dimers. As far as the kinetic aspect is concerned, the new equilibrium state between dimers and octamers after formation of a transition state-analogue complex of Mi-CK is already reached after 15 minutes [296,396,483], whereas it is approached very slowly in the absence of substrates [281,396]. Such a minute time-scale is clearly sufficient to envisage a physiological role for the dimer to octamer interconversions of Mi-CK (see ІП-С. and Fig. 2; page 38). Compared to all these findings, virtually no factors stabilizing octamers are known. p-Aminobenzamidine was reported to inhibit octamer dissociation induced by formation of a transition state-analogue complex of Mi-CK [296]. Furthermore, p-aminobenzamidine as well as benzamidine result in partial reassociation of dimers into octamers [396]. Finally, inorganic phosphate leads to an increased proportion of octamers as well, but only in the presence of CK substrates [280]. Now that in vitro some factors influencing the dimer to octamer ratio are known, the questions of whether changes in the oligomeric state of Mi-CK also occur in vivo and if they have any physiological bearing have yet to be answered. Radiation inactivation [353,483] and crosslinking experiments [284] revealed that the Mi-CK octamer very likely is the only oligomeric form bound to intact bovine and rabbit heart mitochondria. This was further corroborated with antibodies against purified pig and rabbit heart Mi-CK which allowed the discrimination between dimeric and octameric Mi-CK [355,483]. In extraction experiments, the vast majority of Mi-CK was released from the mitochondrial membranes in its octameric form, independently of the releasing agent used [284,296,399,483]. Dissociation into dimers, if it happened, only occurred in solution and on a slower time-scale than the release of Mi-CK from the membranes. Whereas exclusively octameric 44 CHAPTER 2

Mi-CK could be rebound to pig and rabbit heart mitoplasts [296,483], both dimers and octamers rebound to chicken heart mitoplasts [399]. The rebinding of both oligomeric forms was strongly pH-dependent, and sharply decreased between pH 7.5 and 8.1 for dimeric Mi-CK, but only above pH 8.1 for octameric Mi-CK. Therefore, Mi-CK octamers rebound preferentially over dimers to mitoplasts at intermediate pH values around 8.0. Since octamers have a higher isoelectric point than dimers (see Π-D.) and since the interaction of Mi-CK with membranes seems to be mostly ionic in nature (see Ш-В.), octameric rather than dimeric Mi-CK is indeed expected to bind more strongly to negatively charged groups of the membranes [354,399]. Finally, Mi-CK dimers, once bound to the mitochondrial membranes, partially reassociate into octamers [399], thus emphasizing that octameric Mi-CK may be favoured even more so under in vivo conditions. The above findings were further extended by differential digitonin extraction of Mi-CK from rat brain mitochondria [257]. Mostly octameric Mi-CK was present in intact mitochondria, and dimeric Mi-CK could be released from the membranes more easily than octameric. Most importantly, exclusively octameric Mi-CK seems to be enriched in CS between mitochondrial inner and outer membranes where it is thought to play an essential role in mitochondrial PCr production (see ПІ-С). Octamers, in contrast to dimers, appear to be ideally suited for a CS localization. Due to their identical top and bottom faces, they can simultaneously interact with two opposing membranes, as was directly shown in in vitro experiments where octameric proved to be much more potent than dimeric Mi-CK in inducing close contacts between a spread membrane monolayer and large unilamellar membrane vesicles [370]. To conclude, the octameric form of Mi-CK seems to be an indispensable prerequisite for the efficient functional coupling between Mi-CK and oxidative phosphorylation (see also Ш-С.). The existence and physiological importance of dimeric Mi-CK in vivo may be questioned by the fact that the Mi-CK concentration in the intermembrane space was estimated to be in the range of 3.5-17.2 mg/ml [280,396]. At these same concentrations in vitro, Mi-CK is predominantly octameric, even under conditions most strongly favouring dissociation into dimers, namely formation of a transition state-analogue complex [516]. Thus, one is inevitably led to the question of whether dimeric Mi-CK represents only a test-tube "artifact" or not. The possibility of its involvement in the regulation of mitochondrial energy metabolism should, however, not yet be dismissed since (i) interaction of Mi-CK octamers and dimers with mitochondrial membranes as well as the octamer to dimer ratio are sensitive to physiologically important parameters like pH and the concentration of Mi-CK substrates; (ii) a small, but significant fraction of Mi-CK is always released in its dimeric form from intact mitochondria; and (iii) Mi-CK Review 45 isoenzymes from different species and tissues form octamers varying in stability. E. g., pig heart and chicken ubiquitous Mi-CK dissociate much more readily into dimers than rabbit heart and chicken sarcomeric Mi-CK, respec­ tively [296,397,483,516]. Alternatively, different conformations of Mi-CK rather than two oligomeric forms may be envisaged as a decisive factor for the regulation of mitochondrial energy metabolism. Changes in the octamer to dimer ratio may simply reflect changes in conformation, with octamers being more stable in one conformation rather than the other(s) [281,296,398]. This view is supported by a variety of experiments (fluorescence and EPR spectroscopy, inhibition and crystallization experiments, etc.) suggesting thatCK isoenzymes undergo conformational changes upon substrate binding [39,238,405,506]. Furthermore, conformational changes might also explain why the effects of some agents on the dimer to octamer ratio and on the release of Mi-CK from the mitochondrial membranes parallel each other. In fact, MgADP, formation of a transition state-analogue complex, high chloride concentrations as well as p-hydroxymercuribenzoate all efficiently release Mi-CK from the inner mitochondrial membrane and, in addition, dissociate octamers into dimers [281,296; Wyss, M., Schlegel, J. and Wallimann, T., unpublished data] (see also II-A.). However, this parallelism may not be that strict since in the case of rabbit heart Mi-CK, p-aminobenzamidine only inhibited the dissociation of Mi-CK octamers into dimers, bul not its release from the mitochondrial membranes [296]. In dimeric bovine heart Mi-CK, the "essential" sulphydryl groups of both subunits react readily and with the same rate with alkylating agents [123]. Additionally, both subunits of rabbit MM-CK bind substrates [124]. Conversely, only half of the subunits of octameric bovine heart Mi-CK bind substrates and are readily accessible to sulphydryl group reagents [123,124]. Clearly, also in this case, conformational changes may be the underlying basis for the observed phenomena. Finally, bovine heart Mi-CK octamers were found to have a more than two-fold higher specific enzymatic activity than dimers [123], whereas others observed no significant difference in this parameter for dimeric and octameric chicken and bovine heart Mi-CK [100,396]. To conclude, the question of whether changes in the dimer to octamer ratio or simply in the conformation of Mi-CK are essential for the regulation of mitochondrial oxidative phosphorylation and PCr production deserves full attention in the future. Recombinant techniques are likely to be a powerful tool for the direct comparison of properties between native (octameric) Mi-CK and dimeric cytosolic CK, to which by genetic engineering a mitochondrial target peptide is attached, thus allowing its import into the interpiembrane space. 46 CHAPTER 2

Ш-Е. Developmental changes

Studies on the developmental changes of the CK isoenzyme system were conducted mainly with heart and skeletal muscle as well as with brain and retina of a variety of species including man [18,107,108,115,165, 199,200,292,335,435,455,496,500,530]. In all cases, these studies revealed a complex pattern of isoenzymes appearing and disappearing during develop­ mental maturation. Total CK activity considerably increases during the last stages of fetal development of cardiac and skeletal muscle of mouse, rat, rabbit, sheep and man [199,200,202,435]. Whereas total CK activity continues to rise postnatally in the mammalian heart [18,107,108,199,335,473], the specific CK activity was either found to increase [107,108,473] or to remain constant [192]. During postnatal brain development, total CK activity in­ creases markedly at a time when a greater coordination of complex nervous activity is becoming apparent. Furthermore, the similar developmental patterns of CK and hexokinase suggest that CK is involved in the overall coordination of energy metabolism and neuro transmission in the fully active adult brain [53,335]. The CK isoenzyme distribution is both tissue-dependent and develop­ mental stage-specific. In mammalian brain as well as in heart and brain of birds, BB-CK is the major CK isoenzyme at all stages of development. In contrast, in skeletal muscle of birds and mammals as well as in mammalian myocardium, a developmental transition from B-CK to M-CK mRNA and therefore also from BB-CK over MB-CK to MM-CK protein dimers takes place prenatally [115,292,466]. In mouse heart, the transition from BB-CK to MM-CK happens during the last trimester of fetal development, whereas after birth, comparatively small changes in the proportion of the cy tosolic CK isoenzymes are observed [165,199]. In human skeletal muscle, the BB- to MM-CK transition takes place around week 8 of fetal development, in contrast to human heart where MM-CK predominates from the earliest stage examined onward, i.e. 4 1/2 weeks of embryonic development [500]. Similarly, the developmental transition from B-CK to M-CK expression occurs earlier in heart compared to skeletal muscle of the rat [466]. Finally, stage-dependent regional differences in the expression of cytosolic CK isoenzymes were observed in the prenatal de velopmen t of the rat heart [176]. For example, MM-CK was first observed in the outflow tract and the trabeculae of the right ventricle at embryonic days 12-14 and only at later developmental stages in other parts of the heart. As far as Mi-CK is concerned, developmental studies stress its impor­ tance for energy metabolism. In all tissues and species examined, the Mi-CK isoenzymes are accumulated at later stages than M- and B-CK, indicating that Review 47

су tosolic and Mi-CK isoenzymes are subject to different regulatory programs [195,202,345]. In contrast, M- and Mi-CK mRNA were coordinately induced during differentiation of mouse muscle cells in culture [163]. In myocardial tissue of altricious animals like mouse, rat, rabbit and chicken, very low amounts of Mi-CK and Mi-CK mRNA are found before birth [165,192,195,199,335,418]. Cardiac Mi-CK activity rises sharply be­ tween 6 and 25 days of neonatal life in the mouse [165,199], up to about 3-9 weeks of age in the rat [107,108,199,202,418],andbetweenaboutland20 days in the rabbit [192,510], thereby reaching adult values. However, during these periods of time, no changes in a variety of mitochondrial parameters (e.g. mitochondrial ATPase activity) were observed [107,108]. In contrast to these altricious animals, Mi-CK is already present in fetal heart and/or skeletal muscle of precocious animals such as sheep and guinea pig, but astonishingly also in man [199,200,435, 479], and accumulation of Mi-CK to the adult concentrations occurs mostly before birth. In quadriceps muscle of preterm born infants, for example, Mi-CK activity as well as protein content increased significantly with gestational age [435]. Adult concentrations of Mi-CK were reached soon after birth in both sheep and man. Most interestingly, Mi-CK in cardiac muscle appears at a time when, during postnatal development, incorporation of MM-CK into the myofibrillar M-band of the mammalian heart muscles begins [71,72,192]. During the same period of time, a general maturation of the heart muscle towards its full contractile potential takes place [18,192,194]. The coordinate postnatal ap­ pearance of Mi-CK on the PCr production side and of M-line-bound MM-CK on the PCr consumption side of the PCr "circuit" emphasizes the need for a functional coupling of the two systems for optimal muscle function [492]. In addition, the developmental accumulation of "total Cr" (Cr + PCr) in the mouse, rat and possibly also sheep heart occurs in two consecutive steps [199]. First, total Cr concentration increases prenatally, almost in parallel with the accumulation of MM-CK, thereby reaching a plateau which is maintained for a certain period of time. Then, a further severalfold increase in total Cr pool size parallels the accumulation of Mi-CK, indicating that the two processes are linked in some way. This assumption is fully in line with experiments on cell cultures derived from neonatal rat hearts showing that addition of 20 mM Cr to the culture medium stimulates the synthesis of Mi-CK [418,473]. Besides, the relative proportion of Mi-CK increases with age in culture and with age of animal from which the culture is derived. In chicken leg muscle, but not in chicken heart, amounts of Mi-CK mRNA similar to those in adult heart were already found at embryonic day 19 [195]. In rat brain, Mi-CK is undetectable at birth and increases also postnatally to the adult proportion of 15 % of total CK activity [335]. In chicken retina, 48 CHAPTER 2

BB-CK content was high in late stages of embryonic development, decreased slightly around hatching and remained high during adulthood [496]. Mi-CK content, however, was low during development in ovo, rose just before hatching, at a time when visual functions have to become operational in autophagous birds to enable them to find food, and remained high through­ out the following developmental periods. Mi-CK was accumulated pre­ dominantly within the ellipsoid portion of the inner photoreceptor cell segments. 31P-NMR measurements were performed to test the functional conse­ quences of the development of the CK system. In neonatal rat brain lacking Mi-CK, ischemia led to an almost parallel decrease in the concentrations of ATP and PCr [335]. In the adult brain, however, first PCr and, after a delay, ATP concentration also decreased, indicating that Mi-CK might be essential for efficient rephosphorylation of ADP. This conclusion is corroborated by the finding that the capacity of the rodent brain to modulate the rates of glycolysis and tissue respiration in response to sudden changes in energy demand increases in the narrow time-window between days 12 and 15 of postnatal development when also the CK-catalysed reaction rate increases [193]. To monitor the developmental changes of fluxes through the CK reaction, 3,P-NMR saturation transfer experiments were performed with hearts of 3-18 day-old neonatal rabbits at different levels of cardiac perfor­ mance [300,346]. Parallel biochemical experiments demonstrated that during this developmental period, total CK activity and adenine nucleotide pool size in the heart remained constant, the proportions of MM-CK and Mi-CK as well as the Cr pool size increased [see also Ref. 199], and the proportions of MB-CK and BB-CK decreased [346]. The 31P-NMR measurements revealed that the CK reaction flux in the direction of ATP production was positively correlated with the relative proportion of Mi-CK, with the Cr pool size, and with cardiac performance [300,346]. Furthermore, the fluxes in both direc­ tions of the CK reaction were identical under all conditions tested [300]. In conclusion, the expression of the four known CK subunit isoforms is differentially regulated during development on the transcriptional and possibly also on the translational level. The reasons for these developmental changes and the functions of the different isoenzymes at the various stages of development are not fully understood. However, the transition in muscle from B- to M-CK has to do with the isoenzyme-specific subcellular localiza­ tion of CK isoenzymes. In sharp contrast to MB- or BB-CK, homodimeric MM-CK is capable of binding to the myofibrillar M-band where it fulfils its specialized function as an intramyofibrillar ATP regenerator and as a structural component of the M-band [488]. The level of expression of M- as well as Mi-CK strongly depends on the physiological requirements and is Review 49 therefore developmental stage-specific. In addition, maturation of the CK system differs qui te dramatically between altricious and precocious animals. Since Mi-CK appears only after birth in a variety of small animals, it seems not to be essential for life per se or for basic muscle function at a low work-load. Strikingly, however, the flux through the CK reaction, the mechanical performance of the heart, the fraction of M-line-bound M-CK as well as the proportion of Mi-CK increase concomitantly, in spite of a constant total CK activity [192,346]. These latter results strongly suggest that Mi-CK is crucial for energy supply at high work-loads where ADP diffusion may become a limiting factor [525].

III-F. Adaptive changes

Two main primary functions were ascribed to the CK /PCr system (see IV-B.). First, it is involved in buffering of [ATP] and especially [ADP] during abrupt changes in workload, with this function certainly being most crucial for fast-twitch skeletal muscles. Second, the CK/PCr system is involved in the transport of high-energy phosphates from sites of ATP production (mitochondria, glycolysis) to sites of ATP consumption. This function is more pronounced in "endurance" tissues like myocardium and slow-twitch skeletal muscles, and, in contrast to the "buffer" function, is thought to be facilitated by the presence of Mi-CK. In accordance with these ideas, training for long-distance running and even a marathon race itself increased the proportion of Mi-CK relative to total CK activity in human gastrocnemius muscle, with the increase being larger in female than in male runners [9]. Total CK activity, however, remained unchanged. In addition, activity and relative proportion of Mi-CK were found to increase almost linearly with the duration of chronic stimulation of fast-twitch rabbit muscle (type II, white fibers) [400], whereas they decreased in human quadriceps femoris muscle during 6 weeks of leg immobilization after knee surgery [224]. In contrast to these results, endurance training was reported not to influence total activity and relative proportion of Mi-CK in rat gastrocnemius muscle [342]. In most instances, the changes in Mi-CK activity were either directly paralleled or even surpassed by changes in the mitochondrial protein content or in the activities of mitochondrial marker enzymes like citrate synthase [224,400]. Incidentally, chronic stimulation of fast-twitch rabbit muscle was found to increase the activities of citric acid cycle enzymes as well as of total mitochondrial volume [361]. Furthermore, several mitochondrial enzymes were shown to increase after exercise [358]. Finally, Mi-CK as well as MB-CK 50 CHAPTER 2

activity were found to be positively correlated with the oxidative capacity of a muscle [457,459]. All these findings strongly indicate that the changes in Mi-CK activity induced by training or chronic stimulation are not specific events, but are due to general changes in mitochondrial content, thus reflecting a transition from a more glycolytic to a more oxidative, fatigue- resistant energy metabolism. In vivo and in vitro ischemia affects a variety of functional as well as structural properties of heart and skeletal muscles, some being reversible and others not [see for instance Ref. 388]. In rabbit heart, for example, total ischemia resulted in a progressive loss of Mi-CK activity, which was closely paralleled by a decline in left ventricular pressure [43,215,462]. Within 60 minutes of ischemia, the ratio of Mi-CK to mitochondrial malate dehydrogenase decreased by more than 70 %, indicating that the loss is not due to a general decrease in mitochondrial content, but to a specific release or selective inactivation of Mi-CK. In addition, the functional coupling between Mi-CK and ANT was depressed [462], and Mi-CK was detected in blood after ischemia (see III-G.). Inorganic phosphate was suggested to cause a selective release of Mi-CK from the inner mitochondrial membrane [462]. During prolonged, severe ischemia, intracellular [Pj] may approach 50 mM [260], a concentration that is in the range used for in vitro solubilization of Mi-CK from the mitochondrial inner membrane [169]. However, since the loss of Mi-CK appeared to be irreversible, and since Mi-CK could not be detected in the post-ischemic supernatant, Bittl et al. [43] concluded that Mi-CK is not just released from the mitochondrial inner membrane, but is irreversibly inhibited during ischemia. Again in contrast to these results, Saks et al. [388] in ischemic rat hearts found only a transient, reversible decrease in Cr-stimulated respiration and thus in Mi-CK activity. In addition, even though ventricular performance and metabolite contents of isolated perfused rat hearts were permanently depressed, a variety of respiratory parameters proved to be highly tolerant to ischemia, thus suggesting that, at least under the experimental conditions used, mitochondrial injury is not a major component of ischemic damage [388]. Clearly, more experiments are needed to unravel the apparent discrepancies. Compared to heart muscle, more pronounced mitochondrial changes were noticed in ischemic skeletal muscles [172]. For example, ischemia resulted in the appearance of giant mitochondria containing paracrystalline inclusions. These intramitochondrial inclusions had the appearance either of accumulations of finely granular material distending the intracristae space, or of plate-like structures sandwiched between the outer and the inner mitochondrial membranes or between two leaflets of the inner membrane. The inclusions were suggested to derive from aggregation of enzymes Review 51 present in the intermembrane space of muscle mitochondria such as Mi-CK (see below). To evaluate whether or not Cr is essential for normal muscle function and structure, animal models have been developed to test the effect of Cr depletion. An almost complete Cr depletion can be achieved by feeding animals with Cr analogues like cyclocreatine (cCr; l-carboxymethyl-2- iminoimidazolodine) [513], ß-guanidinobutyric acid (GBA) [269] and ß-guanidinopropionic acid (GPA) [127], which act as competitive inhibitors of Cr uptake into the cell and are in general poor substrates for the CK isoenzymes [6,129]. Instead of PCr, large amounts of phosphorylated cCr (PcCr) and GPA (GPAP) are accumulated inside the cells. In contrast, GBA is not phosphorylated by the CK isoenzymes [531]. As a result of Cr depletion, important biochemical, functional as well as morphological alterations occur, as will be discussed now in more detail. The effects of cCr on brain, heart and skeletal muscle metabolism were mainly investigated in relation to ischemia [7,366,469,470,513]. Most impor­ tantly, and also very surprisingly, ATP levels during total ischemia were found to be sustained substantially longer in several tissues of cCr-fed

animals as compared to controls [366,469,470], even though the Vmax value of CK with PcCr as substrate is about 160-fold lower than that wi th PCr [6]. Upon cCr-feeding, delayed ATP depletion was observed for mouse brain [513] as well as for chicken breast muscle [470], chicken heart [469], and rat heart [366], these muscles displaying a relatively homogeneous fiber population. On the other hand, mixed-fiber leg muscles of cCr-fed mice [7] or chicken [470] for unknown reasons did not exhibit a directly measurable ATP-sustaining activity during ischemia. The capacity of dietary cCr to sustain [ATP] may be attributed to the unique thermodynamic properties of the accumulated PcCr. Because its free energy of hydrolysis is roughly 2 kcal /mol lower than that of PCr [6], PcCr may continue to buffer the adenine nucleotide concentrations and to transport high-energy phosphates throughout the muscle fibers even at cytosolic pH values and phosphorylation potentials well below the range where the CK/PCr system can function effectively. In contrast to cCr, homocyclocreatine and GPA feeding did not delay ATP depletion [366,469]. GPA-fed animals also seem a suitable model to study the consequences of Cr depletion, although the possibility that GPA or GPAP are toxic for muscle has to be considered [349]. GPA feeding results in a variety of adaptive changes similar to those observed in transitions from a more glycolytic to a more oxidative energy metabolism, as occurs for example during endurance training (see above). In rat skeletal muscle, GPA feeding decreased the concentrations of total Cr, PCr and ATP by up to 75, 90 and 50%,respectively [127,154,313,422,427,428], whereas [Pj] remained unchanged 52 CHAPTER 2

[428]. The glycogen content as well as the activities of aerobic enzymes such as citrate synthase, 2-oxoglutarate dehydrogenase and 2-hydroxyacyl-CoA dehydrogenase were found to be increased in all fast-twitch (plantaris and gastrocnemius) muscle regions except the superficial gastrocnemius, but not in the slow-twitch soleus muscle [427]. In contrast, the activities of CK, phosphofructokinase, and glycogen Phosphorylase decreased in all skeletal muscle regions except the deep gastrocnemius [427]. As far as contractile characteristics are concerned, plantaris muscles of GPA-fed animals exhibited no abnormalities, except for a slight decrease in initial strength [349]. Surprisingly, the endurance of soleus muscle was prolonged. In addition, the isometric twitch characteristics in this latter muscle were altered, and the maximum velocity of shortening was decreased [346]. 31P-NMR experiments revealed that in resting gastrocnemius muscle of GPA-fed rats, the flux through the CK reaction is reduced in the direction of ATP synthesis [424], and that there is no measurable phosphate exchange between GPAP and ATP [313]. In addition, the rate of GPAP hydrolysis in stimulated GPAP-loaded muscle was much less than that of PCr in control muscles [313]. Intracellular pH decreased more rapidly during stimulation and recovered more rapidly afterwards in GPAP-loaded muscles compared with controls. However, despite buffering by PCr hydrolysis, the pH ul timately decreased more in control muscles. This finding is very likely due to the two-fold greater lactate accumulation in stimulated control gastrocnemius muscles as compared to GPAP-loaded muscles. These results were taken as an argument that in skeletal muscle, PCr is not essential for steady-state energy production,but that phosphate releaseby PCr hydrolysis is essential for maximum activation of glycogenolysis and/or glycolysis [313]. However, recent evidence suggests that GPAP in workingheartmuscle may be used quite efficiently as a CK substrate [88]. Heart muscle of GPA- and GBA-fed rats exhibited contractile failure, as evidenced by cardiac hypertrophy [305] or by a rise in the left ventricular diastolic pressure (LVDP) [235,236,531]. The latter, probably by impairing left ventricular filling, may be responsible for the diminished cardiac per­ formance observed in GPA- and GBA-fcd animals [235,531]. The extent of contractile failure was found to depend on the functional load and on the degree of Cr depletion. However, 3,P-NMR saturation transfer experiments revealed that a 80-90 % depletion of PCr results in only a two- to four-fold reduced flux through the CK reaction [235,531 ], thus rendering final conclu­ sions about the involvement of the CK/PCr system in contractile failure difficult. Finally, the expression of myosin isoenzymes in the left ventride of

the heart was recently shown to change from the fast form V1 to the slower forms V2 and V3 during GPA feeding for several weeks [305], thus reflecting Review 53 changes in contractile properties similar to those observed in skeletal muscles (see above). Taken together, the available data strongly suggest that, at least in the heart, the CK/PCr system is essential for proper muscle function. In addition to the biochemical and functional alterations discussed so far, Cr depletion in muscle also caused several morphological changes, thus suggesting that Cr metabolism is important for sustaining normal muscular structure. GPA feeding caused an increase in the relative proportion of type I muscle fibers (red, slow-twitch), in soleus muscle, for example, from 81 % in control rats to 100 % in GPA-fed rats [427]. In addition, GPA-feeding in general decreased type II fibres (white, fast-twitch) in size and weight, while type I fibres were unaffected [349,422,427]. Accordingly, the largest change in relative muscle size was observed for the gastrocnemius muscle displaying the greatest proportion of type IIb fibres (62 %) [427]. Selected fibres of the pectoralis and gastrocnemius muscles of chicken fed GBA exhibited loss of thick and thin filaments, disruption of the Z-band, dilated mitochondria, as well as dilated and displaced sarcoplasmic reticulum [269]. These ultrastructural changes are attributable to an abnormality of Cr metabolism, since GBA by itself seems not to be toxic. Accordingly, when chicken were given extra dietary Cr in addition to GBA, muscle [PCr] was found to be normal, and no significant ultrastructural alterations occurred. As a result of Cr depletion, abnormal mitochondria were observed in slow-twitch skeletal muscles of rats [154,339]. These mitochondria often were enlarged and contained crystal-like inclusions like those frequently observed in human mitochondrial myopathies [121,429,444], ischemia [172,180], as well as in muscles exposed to mitochondrial poisons [307,373] (see also ΠΙ­ Ο. When adult rat cardiomyocy tes were cultured in a medium devoid of Cr or in a medium supplemented with GPA, two populations of mitochondria could be distinguished [116]. Giant, cylindrically-shaped mitochondria were randomly distributed over the cell and contained inclusions highly enriched in Mi-CK, as shown by immuno-gold labelling. In contrast, small, "normal"- sized mitochondria without inclusions were localized between the myofibrils and contained much lower amounts of Mi-CK. Addition of Cr to the culture medium caused the disappearance of the giant mitochondria as well as of the crystal-like inclusions, accompanied by an increase in the intracellular concentration of total Cr. It is not readily understood why only part of the mitochondria are affected by Cr depletion and form inclusions. One possi­ bility is that subsarcolemmal mitochondria are more susceptible to metabolic alterations and react to a Cr defici t firs t by fusion to form giant mitochondria and second by a compensatory accumulation of Mi-CK, the latter resulting in the formation of Mi-CK containing crystal-like inclusions [116]. Taken together, these results corroborate the sugges tion that changes in Cr metabo- 54 CHAPTER 2

lism in mitochondrial myopathies play an important role in the formation of abnormal mitochondria as well as mitochondrial inclusions. However, in a preliminary study among six patients, no correlation was found between the occurrence of abnormal mitochondria and total [Cr] [436]. Phosphate depletion in rats produced by dietary phosphorus restriction resulted in a decreased concentration of inorganic phosphate in skeletal muscle, in an elevated phosphorylation potential, and in reduced oxygen uptake [57]. Furthermore, the specific activities of Mi-CK and of myofibrillar MM-CK were reduced. Accordingly, addition of Cr to state 4-respiring mitochondria did not increase the rate of oxygen consumption. The mecha­ nism by which phosphate depletion may induce the observed alterations is unknown. Since Mi-CK is of prime importance in regulation of cellular energy production and transport, and since these steps are impaired in skeletal muscle during phosphate depletion, the reduction in the activity of Mi-CK may be the key biochemical disturbance in the myopathy of phos­ phate depletion [57].

Ш-G. Mi-CK in pathology

III-G1. Neuromuscular diseases In recent years, a variety of studies focussed on the possible involvement of Mi-CK in the pathology of several neuromuscular disorders such as muscular dystrophies and mitochondrial myopathies [31,293]. Muscular dystrophy is a heterogeneous group of disorders of which the Duchenne muscular dystro­ phy is most frequently observed in humans. However, to our best knowledge, no extensive studies on Mi-CK were performed in human muscular dystrophy. In skeletal muscle of dystrophic chicken, compared to normal age-matched controls, Mi-CK activity progressively decreased during the course of the disease, with the pectoralis muscle being more affected than the gastrocnemius [31,293]. Furthermore, Cr-stimulated mitochondrial respiration in dystrophic chicken breast muscle was found to be decreased [31]. Consequently, a decrease in Mi-CK activity may ultimately cause the functional loss in breast muscle fibers by decreasing the efficiency of trapping available mitochondrial ATP as PCr. An inborn error of metabolism compris­ ing Mi-CK deficiency would dearly be very helpful in testing this hypothesis. A mitochondrial myopathy can be defined as a muscle disease characte­ rized by structurally or numerically abnormal mitochondria and/or abnor­ mally functioning mitochondria. In about 30 % of the patients with an in vitro observed disturbed pyruvate oxidation rate, no single enzyme deficiency of the mitochondrial respiratory chain is found in the skeletal muscle Review 55 mitochondria. As an alternative, Mi-CK activity may be affected in these patients. However, quantitative measurement of Mi-CK activity by a newly developed method [434] revealed no deficiency of Mi-CK in skeletal muscle specimens of 11 patients with disturbances in the pyruvate oxidation rate in whom no defect in the pyruvate dehydrogenase complex or in complexes of the respiratory chain could be established [437]. In two patients with an established single enzyme deficiency of the mitochondrial respiratory chain, the specific activity of Mi-CK was clearly enhanced. A possible explanation is that this increase in the specific activity of Mi-CK reflects some kind of adaptation. A similar compensatory increase was also observed for cytochrome с oxidase and citrate synthase activities in patients with single enzyme deficiencies of the respiratory chain [372]. Electron microscopical inspection of muscle samples revealed that ab­ normal mitochondria and intramitochondrial inclusions are a typical feature of neuromuscular diseases in general and of mitochondrial myopathies in particular [121]. These crystal-like mitochondrial inclusions were frequently found in the "ragged red" muscle fibers of patients suffering from ocular myopathies with clinical manifestation of progressive involvement of the external ocular muscles (chronic progressive external ophthalmoplegia; CPEO). In patients with mitochondrial myopathies, two distinct types of crystals are observed, which can be distinguished by shape, size, pattern, unit-cell dimension, specific localization in the intermembrane space, and their occurrence in different muscle fiber types (Fig. 3; page 56 and 57). So- called type I crystals (Fig. ЗА) are usually present in the intracristae space, between two folds of the mitochondrial inner membrane, whereas type Π crystals (Fig. 3B) are preferentially located in the intermembrane space between outer and inner mitochondrial membranes. Type I crystals occur only in type I muscle fibers (slow twitch type with high oxidative capacity) and type II crystals in type II muscle fibers (weak or intermediate staining for mitochondrial enzymes) [121,444]. Only recently, it was shown that these crystals are labelled by antibodies directed against Mi-CK (Fig. 3C; see also UI-F.) [445,446]. The Mi-CK immunolabelling of these crystals was uniform and irrespective of the orientation of the crystals to the plane of sectioning. However, type Π crystals were always more heavily labelled than type I crystals [445,446]. In a preliminary study among six CPEO patients, no relationship was found so far between the concentration of total Cr, free Cr or PCr and the occurrence of mitochondrial crystals in the muscle [436]. This holds true especially for one CPEO patient with an extremely low free Cr content in muscle, in whom despite thorough electron microscopical inspec­ tion, no crystals were observed. In two patients with CPEO in which Mi-CK containing crystals were found in the quadriceps muscle, Mi-CK activity was 56 CHAPTER 2

С.Йг ^; i в Review 57

Fig. 3. Accumulation ofMi-CK in intramitochondrial inclusions. Mitochondrial crystals in human skeletal muscle biopsies (m. quadriceps). (A) Transversely sectioned Type I crystals in mitochondria of "ragged red" fibers from a patient suffering from chronic progressive external ophthalmoplegia (CPEO). Note the presence of the crystals in the intracristae spaces. (B) Type II crystals in muscle fibers of a patient suffering from unclassified mitochondrial myopathy. (C) Strong anti-Mi-CK immunogold labelling of longitudinally sectioned Type I crystals from a patient with CPEO. Magnification: (A) 32 400x; (B) 46 OOOx; (C) 35 200x. For further details see the text.

significantly enhanced despite a normal pyruvate oxidation rate [436]. Recently, it was realized that long-term zidovudine therapy, used for the treatment of patients with the acquired immuno-deficiency syndrome (AIDS), can cause a toxic mitochondrial myopathy with depletion of muscle mitochondrial DNA [11,93]. Besides inflammatory changes, crystal-like mitochondrial inclusions were observed in muscle biopsies of zidovudine- treated patients. It would be worthwhile to further study these inclusions in relation to the CK/PCr system. In addition, further studies are necessary to elucidate the mechanism of crystal formation and to clarify if crystal forma­ tion is causative to or only a consequence of mitochondrial myopathies. 58 CHAPTER 2

III-G2. Cardiomyopathies A cardiomyopathy may be defined as a dysfunction of the myocardium caused by a primary disorder within the myocardium or by secondary disorders, like for example hypertension. In experiments on rat hearts, where arterial hypertension was induced by suprarenal aortic banding, total CK activity in the left ventricle rose within 4 days by about 70 % [133]. In this model of a short-term cardiomyopathy, the expression of M-, B- and Mi-CK was con­ comitantly increased, and therefore no significant change in the relative proportion of the different CK isoenzymes occurred. Seemingly, the in­ creased energy requirements in acute pressure overload are met by a generalized induction of expression and synthesis of all CK isoenzymes. In contrast, long-term cardiomyopathies (left ventricular hypertrophy due to aortic stenosis, volume overload, or hypertension; coronary artery disease; hereditary and diabetic cardiomyopathies) are characterized by an unchanged or decreased total CK activity, by a decreased in vivo flux through the CK reaction, by an increase in the relative proportions of MB- and BB-CK, and by a decrease in the total Cr content [202-205,243,390]. With the single exception of Bio 14.6 Syrian hamsters [12], cardiac Mi-CK activity was considerably decreased in all sorts of long-term cardiomyopathies in both animals and humans [12,45,202,205,243,390,393]. Accordingly, Cr-stimu- lated mitochondrial respiration was also found to be decreased [237,390,476,477]. Most interestingly, total CK and Mi-CK activity as well as the flux through the CK reaction were normal in spontaneously hypertensive rats during the first 12 months of life, this period being characterized by a stable compensated hypertrophy of the myocardium [45,202,205]. However, be­ tween the 12th and 18th month of life, a transition from compensated hypertrophy to failure occurred, whereby in parallel with the functional capacity of the heart muscle, also total CK and especially Mi-CK activity drastically decreased. Finally, in various hereditary and experimental cardiomyopathies (induced by autoimmunization or by treatment of rats with adriamycin, norepinephrine, GPA, or streptozotocin), [ATP + PCr] was decreased [237]. With the exception of GPA-treated animals, cardiac output at standard load conditions was also substantially lowered, probably due to mild bradycardia, elevated left ventricular (LV) diastolic pressure, and stiffness that limited cardiac contractile adaptation to volume or resistance overloads. The LV diastolic stiffness at maximal functional load was in­ versely correlated with the high-energy phosphate content. Its increase in cardiomyopathic hearts may be explained by the increased myofibrillar sensitivity to Ca2+ and by the loss of functional coupling of Mi-CK to oxidative phosphorylation. Since in another study, the myofibrillar MM-CK Review 59 activity of the cardiomyopathic heart was found to be normal [477], these results clearly indicate that loss of Mi-CK activity may be of prime importance for the development of cardiac failure.

III-G3. Tumor tissues Mi-CK was detected in several types of human [101,190,233,340,368,467] and animal tumors [275], in the human carcinoma cell line HeLa [473], as well as in murine Ehrlich ascites tumor cells [25]. In these tissues or cells, Mi-CK is frequently coexpressed with BB-CK, indicating a de-differentiation of trans­ formed cells, which is reflected in the appearance of an embryonic isoenzyme pattern. In Ehrlich ascites tumor cells, however, CK activity was found to be exclusively associated with mitochondria. The authors therefore suggested that the "transport" function of the CK/PCr system (see IV-B.) is not crucial for tumor cells [25]. To our knowledge, there are no reports on the Mi-CK content of brain tumors. No differences in the molecular mass, the electrophoretic mobility, or the kinetic characteristics were observed between Mi-CK in tumors of the digestive tract and Mi-CK in adjacent normal tissue [340]. Accordingly,

Mi-CK of malignant and normal liver displayed the same Mr of 320 000- 350 000, similar heat stability, and a similar behaviour in 2 M urea, whereas the electrophoretic mobility, for unknown reasons, differed dearly [233]. Interestingly, tumors of the digestive tract were shown to contain signifi­ cantly higher amounts of Mi-CK than surrounding normal tissue [340]. This may reflect that a higher expression of Mi-CK is required to meet the increased energy demands of the tumor cells. However, the facts that (i) stomach adenocarcinomas displayed lower Mi-CK activity than the sur­ rounding normal tissue [190], and that (ii) tumors depend more on anaerobic rather than aerobic energy metabolism, make this latter interpretation un­ likely.

III-G4. Body fluids Normally, no Mi-CK is detectable in normal human serum [234,288,352] or cerebrospinal fluid (CSF). This holds true for most species, although sheep may be an exception [24]. Under certain pathological conditions, however, Mi-CK is released into the blood or CSF in man. This may cause diagnostic confusion, as Mi-CK interferes in many of the methods that are routinely used in clinical chemistry for the determination of MB-CK as an indicator of myocardial damage. Serum Mi-CK is generally thought to originate from the mitochondrial compartment. Therefore, the finding of Mi-CK in a patient's serum forms an index of mitochondrial damage. It should be kept in mind, however, that the cell nucleus of heart and skeletal muscle has been suggested 60 CHAPTER 2

to contain Mi-CK forms as well [228]. Clinically, the presence of Mi-CK in serum should not be ignored, for it may be of help in finding the proper diagnosis.

Serum CK isoforms with a Mr of >80 000 are called macro CK's [527]. According to the nomenclature of Stein and co-workers [52,448], macro CK type 1 represents an autoantibody complex of BB-CK. In contrast, macro CK type 2 is generally assumed to be Mi-CK [449] because of similarities in electrophoretic mobility [52,232,449], molecular mass [232,449], activation energy [448,449], enzyme kinetics [232,448] and antibody studies [232]. The

Mr of the principal form of macro CK type 2 was estimated as 287 000-350 000, which is similar to the Mr of Mi-CK isolated from tissues [231,449]. In addition, smaller quantities of a 80 kD form [231,449] and surprisingly also of a >750 kD form of CK were found in serum [449], whereby for the latter, nothing is known about the number and stoichiometry of its constituents. To avoid confusing terminology in the following section, macro CK type 2 is referred to as Mi-CK whenever possible. The occurrence in serum of a CK isoenzyme of mitochondrial origin was first described for one out of two patients examined with Reye's syndrome [367]. The occurrence of Mi-CK in serum was later confirmed [22,222] and considered an ominous sign, since 10 out of 14 positive cases died shortly after Mi-CK was detected in the serum [222; see also Refs . 234,352]. In a prospective study among 2954 consecutive patients in a hospital for internal diseases, the prevalence of Mi-CK in serum was found to be 3.7 % among hospitalized patients and 1.1 % among outpatients [450]. In this study, Mi-CK was found predominantly in severely ill patients of all ages, mainly with malignancies (41 %) and liver diseases (25 %). In a study of 5000 random patient sera, malignancies were found in 25 of the 26 adult patients that were positive for Mi-CK in serum [514,515]. Of course, the prevalence figure depends both on the sensitivity of the test procedure for Mi-CK and on the patient group that is screened. The occurrence of Mi-CK in serum has been studied extensively in neoplastic diseases [183,317]. Serum of patients with malignancies may contain Mi-CK, sometimes in combination with mitochondrial aspartate aminotransferase, thus indicating mitochondrial damage in the tumor. Serum Mi-CK has been found in patients with primary tumors in liver [73,158,232-234,515],pancreas[515],lung[158,234,258,274303,452,515],breast [73,158,234,258,288,352,450,453,515], gastro-intestinal tract [158,190,234,258, 289,303,308,317,340,352,368,452,515], prostate [158,234,452,453,515], gallbladder [225,234,452], ovaries, and uterine cervix [234]. However, Mi-CK was not found in serum of 120 leukemia or lymphoma patients [73]. In addition, Mi-CK has never been described in patients with renal tumors. In Review 61 some cases, the tumor and its metastases were shown to contain the same macromolecular Mi-CK as the patient's serum [232,450], thus suggesting that tumor tissue itself can release Mi-CK and sometimes also BB-CK into the serum. Histological typing of tumors probably releasing Mi-CK into the blood has been performed by Kanemitsu et al. [234]. However, a patient can have a tumor expressing Mi-CK without displaying measurable quantities of Mi-CK in the serum. This was evidenced by two patients with liver metastases, in which Mi-CK was released in measurable quantities into the serum only after embolization of the hepatic artery [432]. Although hypoxia may play a role, the exact mechanism for the pathological release of Mi-CK from tumors is unknown. Xenograf ting tumor lines into a thymic mice may be a rewarding model for studying Mi-CK release from tumors into the blood [101]. Several authors have suggested the use of Mi-CK in serum as a tumor marker [234,275,288,303,308,368,450,468], or more specifically as a marker for gastrointestinal cancer [258,308,368], metastatic prostatic carcinoma [453] and adenocarcinoma [340,352]. Serum Mi-CK activity roughly seemed to reflect the tumor burden [232,275,450,452,453], so that in individual cases, serum Mi-CK may be used to monitor the initial response to therapy [303]. However, the appearance of Mi-CK in serum alone is not a specific signal for neoplastic disease (see below). Accordingly, the diagnostic sensitivity of serum Mi-CK activity for neoplastic disease in general seems rather poor [73]. The presence of Mi-CK in seru m was shown to be related to the clinical stage of neoplastic disease for some tumors [308], while it was not so for others [73]. Apart from a few patients with various chronic and acute diseases [367,450], Mi-CK was frequently found in serum of patients with liver diseases, more in particular liver cirrhosis [73,450]. In these patients, serum Mi-CK apparently originated directly from liver cells, which in this special case were found to contain Mi-CK [232]. Normal liver cells, in contrast, do not contain measurable amounts of Mi-CK (see III-A.). Myocardial damage was early recognized as an additional potential cause of Mi-CK efflux into the blood [22,222]. Mi-CK was found in sera of children with myocarditis [362,515], congestive heart failure, and cardiomyopathy after aortic valve surgery [515]. Mi-CK appeared to be present in serum of patients who have experienced periods of poor tissue perfusion [22], myocardial ischemia [158], and cardiorespiratory arrest [92]. These findings are in line with the observation that in rats during hypoxia, Mi-CK activity in the blood increased in parallel with a decrease in the heart [178]. Furthermore, Mi-CK was found in serum of individual cases after acute myocardial infarction (AMI) [158,222,289,448,458]. Peak values for serum Mi-CK were observed 24 hours after AMI [458]. The activity of Mi-CK approximated 20 % of that of MB-CK, the traditional and established 62 CHAPTER 2

indicator of myocardial damage [458]. In larger series, however, serum Mi-CK could not be detected at all after AMI or was only found in sporadic cases [73,122,352,362,367,441,450,503]. Therefore, release of Mi-CK into the blood circulation after AMI seems to be the exception rather than the rule. Besides, it remains principally unclear why Mi-CK is released into the blood in some patients with AMI and not in others. Interestingly, the drug theophylline was suggested to induce Mi-CK release into the serum in vivo [99]. As this drug is often given to patients suffering from cardiac diseases, it can not be decided yet whether in fact theophylline or simply myocardial damage is the actual cause resulting in Mi-CK release into the serum. In CSF, Mi-CK was found under several pathological conditions [75,76,503,505] : hypoxic-ischemic brain damage [76]; after surgery in relation with various central nervous system tumors [503]; apoplexia caused by a hypophysis tumor [505]; and meningitis [503]. Due to its presence in human brain [75,277,504], Mi-CK in CSF seems to derive directly from the central nervous system. Rather surprisingly, Mi-CK in CSF is not necessarily accom­ panied by BB-CK [503]. As of yet the diagnostic potential of determining Mi-CK in CSF remains an open question and may be a topic of further research. As far as clinical chemistry is concerned, electrophoresis was often used for the detection and quantitation of Mi-CK. However, interpretation of the serum CK zymograms is complicated for several reasons: (i) The human Mi-CK isoenzymes either migrate cathodally to or comigrate with MM-CK [52,289,502]. Incubation of Mi-CK in normal human serum results in modi­ fication of the most cathodal human heart Mi-CK band [231,502], with the modified Mi-CK comigrating with MM-CK [502]. In contrast, human liver Mi-CK is not influenced by serum incubation [231]. (ii) Various authors have observed one to three Mi-CK bands in serum upon electrophoresis [52,183,231,289, 329,432,449,453,502] or even more in isoelectric focussing experiments [433,449]. This multiplicity is poorly understood. Blockingofall M-CK activity in a zymogram with inhibiting antibodies generally is very informative to discriminate between MM-CK and Mi-CK [52] (see below). (iii) Comigration of adenylate kinase isoforms [241] and in single cases of a macro CK type 1 complex [52,528] with Mi-CK may further complicate the zymograms. There are several reports in the literature where the authors relied exclusively on the electrophoretic mobility to classify a cathodally migrating form as Mi-CK. However, in any publication on Mi-CK, it should adequately be shown that there is no interference with alternative enzymatic activity at stake. Since the mere presence of the inhibitors AMP and diadenosine pentaphosphate does not always guarantee full inhibition of adenylate Review 63 kinase [241 ], a control zymogram of the electrophoresed samples without PCi in the reaction mixture is required to exclude the presence of adenylate kinase. Proper discrimination between macro CK type 1 and Mi-CK, on the other hand, can be achieved by published methods [52,448]. Mi-CK in serum was often found just by chance, because (i), it interferes with most methods that are commonly used for the detection of MB-CK, and (ii), most patients with Mi-CK in serum (81-88 %) have a normal total serum CK activity [73,234,450]. As the Mi-CK activity in serum often is below 10 IU/1 [450], the sensitivity of the assay used is a crucial point. Improvement of sensitivity may be achieved by using bioluminescence in combination with immuno-inhibition of MM-CK [303,505]. Unfortunately, anti-Mi-CK anti­ bodies for a direct determination of this isoenzyme in serum have been raised on a limited scale and have been available for research purposes only. Since Mi-CK is not inhibited by antibodies against M-CK, a clinical chemist may be alerted for the possible presence of Mi-CK in serum by the often abnormally high ratio (residual CK activity after immuno-inhibition of M-CK)/(total CK activity) [450]. Accordingly, many researchers have used immuno-inhibition of M-CK, thus measuring all the non-M-CK activity [303,450,505,514], as a reliable first step in establishing the presence in serum or CSF of abnormal CK isoenzymes in general and of Mi-CK in particular. A good methodological alternative is the commercially available reagent kit that combines immuno-inhibition with precipitation of the immune-com­ plexes by a second antibody [514]. This set-up allows direct discrimination between MB-CK on one hand and macro CK forms on the other. In all techniques mentioned, however, further tests are necessary to definitely confirm the presence of Mi-CK. Among these, determination of the apparent activation energy [448,449], the molecular mass [431,448,449], the isoelectric focussing pattern [433,449] and the electrophoretic behaviour [52] would seem to be most convincing. 64 CHAPTER 2

IV. Integration of Mi-CK in cellular energy metabolism

ÍV-A. Advantages of the CK/PCr system

Some of the aspects mentioned below have been discussed in the recent review of Wallimann et al. [492], but are reinforced here in order to be able to fully understand the new arguments and to get a comprehensive picture of the physiological importance of Mi-CK. First, the potential advantages of the CK/PCr system shall be elucidated. Clearly, the most evident advantage of this system is that PCr and Cr allow a much higher flux of "high-energy phosphates" from sites of ATP production to sites of ATP utilization than ADP and ATP, since (i) within tissues with high and fluctuating energy demands, Cr and PCr are accumulated to much higher concentrations than the adenine nucleotides, and (ii) Cr and PCr are smaller-sized and less negatively charged than ADP and ATP. Accordingly, in model solutions as well as in frog muscle, the diffusion coefficients of PCr and Cr were found to be 1.3-2.3 times higher than those of ATP and ADP [219,328,524,525]. Diffusion of ATP is unlikely to be hindered by binding to subcellular structures, since in the cytosol, its diffusion was restricted to the same extent as that of other small molecules. The diffusion coefficients of ATP and PCr were both about 60 % lower in frog muscle than in model solutions [525] which is in perfect agreement with the observations that the diffusion

coefficients of several molecules with Mr 170-24 000 are 2-5-fold lower in the cytoplasm of mammalian cells than in water [297] and that living cells have a fluid phase viscosity 3-4 times greater than water [290]. In contrast to ATP, the diffusion of ADP in the cytosol seems to be severely hindered [389,390]. As much as 97 % of the ADP may be tightly bound and non-diffusable in skeletal muscle and heart [see Refs. 74,219], a finding that may also explain the apparent discrepancy between biochemically measured [ADP] (-100-500 μΜ) and effective in vivo [ADP] of 1-50 μΜ calculated from 31P-NMR spectra [42,141,143,312]. Assuming appropriate diffusion coefficients, substrate concentrations and concentration gradients of 5 %, Jacobus [219] calculated the maximal flux rates of the respective substrates to be (in μπιοΐ/ιηίη/π^ cardiac tissue): 35 for MgATP, 0.112 for MgADP, 57 for P,, 103 for Cr and 123 for PCr. Evidently, MgADP is the most diffusion-restricted of all substrates, a fact that is also reflected by 31P-pulsed gradient NMR experiments yielding mean- square lengths of diffusion of 1.8 μιη for ADP, 22 μιη for ATP, 57 μπα for PCr and 37 μτη for Cr [525]. However, the values obtained for ADP are still in the same range as the maximum measured rate of ATP utilization in the heart (0.135 mmol/min/mg tissue) [219] and the diameter of a single myofibril of approximately 1 mm [525]. Review 65

Since the free energy of PCr hydrolysis (AGobs= -45 kj/mol) [270] is consist­ ently higher than that of ATP hydrolysis (AG^ = -30.5 kj/mol) [157], the CK/ PCr system efficiently "buffers" [ADP] and [ATP] and therefore also the ATP/ADP ratio as well as the phosphorylation potential in the cytosol [230,454]. This is especially important for tissues with abrupt changes in energy demand like cardiac and skeletal muscle as well as brain. During work or anoxia, first [PCr] decreases at relatively constant levels of ATP and ADP, and only when a large part of the PCr is depleted, [ATP] decreases as well [141,159,226,330,381,417]. Since ATP and ADP are key regulators of many fundamental metabolic pathways, whereas Cr and PCr are likely not to be involved in allosteric regulation of intermediary metabolism [128], the CK/PCr system, by damping fluctuations of [ATP] and [ADP] upon abrupt changes of energy demand, allows a better fine-tuning of whole cellular metabolism and therefore protects the cell from energy dissipation (see also IV-B.). By keeping [ADP] low, theCK/PCr system further protects the cells from a net loss of adenine nucleotides [144,209,226,492]. Accumulation of ADP activates adenylate kinase (myokinase) which catalyzes the transfer of a phosphate group between two molecules of ADP to give ATP and AMP. Especially in white and red fast-twitch muscles, AMP is degraded into inosine monophosphate (IMP) and ammonia by AMP deaminase [309] which is bound to the myofibrils at both ends of the A-band [91]. Cytosolic or sarcolemma-bound 5'-nucleotidase dephosphorylates both AMP and IMP into adenosine and inosine, respectively. These latter substances ultimately leave the cell since the sarcolemma is permeable to the latter two compounds [see Ref. 226]. As can be directly seen from the chemical equation of the CK reaction,

PCr2'+ MgADP'+(x).H+ « » MgATP2"+Cr

the CK/PCr system also avoids acidification of the cytosol during periods of high workload [110]. As long as PCr is present in significant amounts, [ATP] remains almost constant and thus PCr2'-» Cr + Ρ,*· is the net reaction supporting work. Since P, at a pH around 7.0 has a mean charge between -1 and -2, PCr hydrolysis may at least in part be responsible for the tissue alkalinization observed during the first stages of muscular work [89,191,264310]. Only when almost all PCr is exhausted, lactate production by glycolysis as well as net ATP hydrolysis lead to a considerable acidifica­ tion of the cytosol. Acidification has three main consequences: (i) it decreases the maximal force of a muscle, either by itself or in combination with 66 CHAPTER 2

diprotonated inorganic phosphate, H2P04', which itself is favoured over 2 HP04 " at low pH values [336]; (ii) it reduces the glycolytic flux by inhibiting phosphofructokinase [see Ref. 77], thereby also avoiding further acidifica­ tion, exhaustion of high-energy phosphates, and thus irreversible damage of the cell; and (iii) it shifts the CK equilibrium towards ATP synthesis, as can also be seen directly from the chemical equation. Thus, the CK/PCr system ensures an almost constant ATP/ ADP ratio over quite a wide rangeof energy demands and pH values which is essential for the proper functioning of all cellular ATPases. In tissues with high and fluctuating energy demands, at least two potential sites of regulation are introduced by the CK/PCr system. Whereas the Cr + PCr pool size is likely to be only important for long-term regulation and adaptation, the cytosolic and Mi-CK isoenzymes are attractive candi­ dates for short-term regulation of the overall flux through the CK reaction. Since the cytosolic CK activity can cope easily with the maximal rates of ATP production or ATP consumption and thus the cytosolic CK system is likely to be in a near-equilibrium state [205,262,298,310], regulation of cytosolic CK activity was suggested to have no influence on energy metabolism. In contrast to the cytosolic CK isoenzymes, Mi-CK is thought to be involved in metabolic channelling of high-energy phosphates from the mitochondrial matrix to the cytosol and is therefore likely to be displaced from thermodynamic equilibrium [262]. Accordingly, regulation of Mi-CK activity would directly reduce the export of PCr out of the mitochondria. The potential implications of regulation of CK isoenzymes can only fully be appreciated if one considers that it allows very specific and efficient regula­ tion of whole cellular energy metabolism. Though there have been no convincing reports up to now proving regulation of CK activity in vivo, the

recent findings that phosphorylation of BB-CK reduces the Km for PCr by about a factor of two [84,356], that BB-CK is a possible substrate of protein kinase С [85] and that a variety of CK isoenzymes are subject to autophosphorylation [184] suggest that CK regulation may be of physiologi­ cal relevance. As a last point, PCr shall be compared with other Phosphagens. In all vertebrate and some invertebrate species, PCr is the sole phosphagen. In contrast, a variety of different Phosphagens like phosphoarginine (PAr), phospholombricine (PL), phosphotaurocyamine (PTc), phospho- hypotaurocyamine (PHTc) and phosphoglycocyamine (PGc) were found in lower phyla, either alone or in combination with each other or with PCr [112,322,493]. Interestingly, exclusively PCr is found in spermatozoa of a large number of "lower" species having other Phosphagens in other tissues. Determination of the apparent equilibrium constants of the phosphagen Review 67 kinase reactions by biochemical and 3,P-NMR methods revealed that at pH 7.25,theapparentequilibriumconstantofCK(K'CK=[Cr][ATP]/[PCr][ADP]) is 3-8 times higher than the respective K' values for arginine kinase (AK), glycocyamine kinase (GK), taurocyamine kinase (TK) and lombridne kinase (LK) [112]. In other words, the free energies of hydrolysis of PAr, PGc, PTc and PL are 2.9-5.2 kj/mol lower than that of PCr. This property can be explained by the methyl group attached to the guanidine moiety of PCr which eliminates almost all resonance states and thus decreases the thermodynamic stability of PCr [112,114]. Due to the higher AGobs value of PCr hydrolysis, the ATP/ADP ratio can be buffered at a higher value which seems especially relevant in the light of experiments on vertebrate skeletal muscle, showing that the reciprocal of the relaxation rate constant is directly proportional to the cytosolic phosphorylation potential and thus also to the ATP/ADP ratio [97,112]. In addition, maintaining a high affinity (free energy change) for ATP hydrolysis has been shown to be essential for a variety of cellular ATPases [229] (see also IV-B.). On the other hand, when the cellular pH is lowered, Phosphagen kinase reactions with a lower K' value will show a smaller degree of net hydrolysis of the respective phosphagen [112]. Consequently, a pool of the highly labile phosphagen PCr would be rapidly dissipated under conditions of intracellular acidosis which especially in molluscs is a commonly observed phenomenon. In contrast, AK/PAr may act as an effective buffer system under these circumstances. These latter reflections shed some light on the functions and evolution­ ary relationships of the different phosphagens. It has been suggested that PAr is the most primitive of the phosphagens, representing an evolutionary precursor. PCr, present mostly in vertebrates, was thought to represent a functional improvement over PAr, because PAr/Ar interfere with amino add metabolism, whereas Cr represents an end-product of a distinct meta­ bolic pathway not interfering with amino acid metabolism [495]. No satisfac­ tory explanations for the occurrence of a variety of different phosphagens in lower phyla have been presented so far. However, the differential distribu­ tion in the animal kingdom of PCr on one hand and PAr, PL, PTc, etc. on the other hand may also be explained as follows. PCr is predominantly found in vertebrates, which almost perfectly maintain intracellular homeostasis. In contrast, intracellular homeostasis is less pronounced in lower phyla, causing larger fluctuations of pH, temperature, substrate concentrations, etc. upon changes in the actual environment. Changes in intracellular conditions might lead to hydrolysis of the highly labile PCr pool and thus to energy dissipation. Consequently, under conditions of poor intracellular homeostasis, more stable phosphagens might be better suited than PCr. In conclusion, the CK/PCr system has the following major advantages 68 CHAPTER 2

over a system based exclusively on ATP/ADP diffusion: (i) it allows a larger flux ofhigh-energyphosphates between sites of ATP production (mitochondria and glycolysis) and ATP utilization (all sorts of ATPases), (ii) it allows the maintenance of a higher ATP/ADP ratio throughout the cell, (iii) it avoids a net loss of adenine nucleotides, (iv) it keeps the pH almost constant during the first stages of cellular work, and (v) provides two additional potential sites for the very specific regulation of energy metabolism. Most of the models of CK function discussed in the next chapter are based primarily on only one or just a few of these advantages. Consideration of all advantages, of various models and of changes in metabolic demand specific for different develop­ mental stages and metabolic adaptations will lead to a more thorough understanding of CK function in whole cellular energy metabolism.

IV-B. Models of CK function

Since the discovery of PCr in 1927 [111], the ideas and models about the involvement of the CK/PCr system in energy metabolism have changed several times. The fact that it is not possible to explain all physiological findings with one single of these models may explain why confusion about the "real" function of the CK/PCr system still exists in the literature and why in most textbooks, energy metabolism is simplified by omitting the CK/PCr system and by assuming exclusive diffusion of ATP and ADP between sites of ATP production and ATP consumption (Fig. 4A; page 70). This model may be correct for cells and tissues lacking CK like liver, but it is clearly incomplete for tissues with high and fluctuating energy demands like heart and skeletal muscle, brain, spermatozoa, retina, kidney, etc. [see Ref. 492]. For a historical overview on the development of alternative models, the reader is referred to Refs. 35,37,214. In this chapter, the various models will be discussed only for ATP production by oxidative phosphorylation within mitochondria. How­ ever, no differences in the qualitative aspects of the models and in conclusions result from a replacement of mitochondrial oxidative phosphorylation by glycogenolysis or glycolysis, because functional coupling, like that of Mi-CK to oxidative phosphorylation (see III-C), has also been observed between cytosolic CK isoenzymes and glycolysis [for reviews see Refs. 381,492]. Since during work, [PCr] decreases whereas [ATP] remains relatively constant, it was long believed that PCr is the direct source of energy for muscular contraction, with ATP being responsible for the regeneration of PCr. Thirty years ago, however, rather specific inhibition of CK in frog skeletal muscle by l-fluoro-2,4-dinitrobenzene caused a contraction- dependent decrease in [ATP] at constant [PCr] and therefore proved that ATP Review 69 hydrolysis directly supports muscular contraction [70]. Because CK inhibi­ tion lowered the number of normal contractions of a muscle fiber from >30 to approximately 3, it was hypothesized that the CK/PCr system represents a back-up system for very efficient "buffering" of [ATP] and especially [ ADP] [70,302,485] (Fig. 4B; page 70). However, a large body of evidence challenged the validity of the "buffer model" [149,159,160,414] so that instead an (exclusive) "transport" function was proposed for the CK system [35,37,38,213,301,328, 379,415,416,463,486,488]. According to this model, which was also termed "PCr shuttle hypothesis", Mi-CK bound to the outer face of the inner mitochondrial membrane catalyzes the transfer of the γ-phosphate group of ATP, synthesized by mitochondrial oxidative phosphorylation, to Cr. PCr then diffuses out of the mitochondria to sites within the cell where energy is consumed and PCr continuously regenerates ATP. Diffusion of Cr back to the mitochondria closes the cycle (Fig. 4C; page 70). In other words, the CK/PCr/Cr system is shunted in between sites of ATP production and ATP consumption. Many experiments were interpreted as favouring one of the two models and dismissing the other. However, within a cell, both "buffer" and "trans­ port" functions of the CK/PCr system may be operational at the same time, with the relative contributions of these two aspects depending on the metabolic demands of a tissue. For example, the buffer function is likely to be more pronounced in fast-twitch muscles where bursts of ATP breakdown have to be buffered immediately and very efficiently to allow short periods of maximal work. These muscles fatigue rapidly because regeneration of high-energy phosphates is achieved on a much slower time-scale mostly by glycolysis. In slow-twitch muscles and especially in heart, however, the transport function seems more important, since in these muscles, high rates of ATP consumption and therefore also of ATP production and transport have to be ensured for longer periods of time. For each workload, a steady- state is attained where ATP production and consumption are efficiently regulated and balanced [17]. The "buffer" function of the CK/PCr system is supported by the follow­ ing facts: (i) Non-excitable cells and organs like liver with a relatively high, but continuous flux of high-energy phosphates contain only small amounts of CK and PCr or even none at all [41,518]. This might indicate that the CK/ PCr system is not essential for maintaining high flux rates of high-energy phosphates, but rather for buffering of sudden changes in energy demand (see below), (ii) For a transport function, comparable amounts of Mi-CK and cytosolic CK activities are expected within a cell. However, the proportion of Mi-CK was found to be only 0-2 % of total CK ас ti vity in some skeletal muscles and brain [8,9,48,245,294,434]. In addition, no mitochondrial isoenzymes of CHAPTER 2

В

Cr -* • Cr ATP 'Г( -PCr- 1) £/U >• PCr ADP Review 71

Flg. 4. Models of CK function. (A) Classical "textbook" model of energy transportbetween sites of ATP production (mitochondria, glycolysis) and ATP consumption (all sorts of ATPases). This model may roughly reflect the situation in tissues devoid of CK, PCr and Cr like liver, but is clearly inappropriate to describe the situation in CK-containing tissues like skeletal or cardiac muscle, brain, retina and spermatozoa. (B) "Buffer" (storage) function of the CK/PCr system. A large pool of PCr is available for immediate regeneration of ATP hydrolyzed during short periods of intense work. Due to the high cytosolicCKacti vi ties, theCK reaction remains in a near-equilibrium stateand thus keeps [ADP] and [ATP] almost constant. In other words, the CK/PCr system efficiently "buffers" theconcentrationsof ATPand especially of ADP. (C) "Transport" function of the CK/PCr system. The CK/PCr system fulfils the function of a "transport device" shunted in between sites of ATP production and ATP consumption. Note that for the "buffer" function no Mi-CK isoenzyme is needed, whereas for the "transport" function, Mi-CK may be an essential prerequisite, especially if there were diffusion limitations for adenine nucleotides across the outer mitochondrial membrane [3233,147,148]. In addition, it should be kept in mind that the two models (B) and (C) represent extremes, with the actual situation in a cell or tissue being somewhere in between. Accordingly, the physiological requirements of a tissue determine the relative importance of the "buffer" and "transport" function of theCK/PCrsystemand therefore also the relative proportion of Mi-CK. For further details see the text. Φ, sites of ATP hydrolysis, e. g. myosin ATPase or ion pumps; Η, Mi-CK; •, cytosolic CK.

arginine kinase have so far been found in a variety of arthropod flight and squidmantle muscles [113,402,451], suggesting that in all these tissues, the CK/PCr and AK/PAr systems are not important for energy "transport". However, though insectflight muscle s are capable of extremely high aerobic energy fluxes without the participation of a mitochondrial arginine kinase isoenzyme, these results do not exclude a transport function of phosphagen kinase systems, since first, rows of densely packed mitochondria are lined up in dose apposition to individual myofibrils of insectflight musde s in such a way that diffusion distances from mitochondria to myofibrils are minimized [438], and second, a mitochondrial isoenzyme is less important for a transport function of an AK/PAr than for a CK/PCr system. Theoretical considerations have shown that due to the lower equilibrium constant of AK compared to CK, the proportion of high-energy phosphate flux carried by Ρ Ar at the same ATP/ADP ratio is higher than that carried by PCr [311]. (iii) PCr content, total CK and soluble MM-CK activity were found to correlate with the glycolytic potential of a muscle [332; for a review see Ref. 492]. Consequently, they were highest in fast-twitch musdes with high glycolytic, but low oxidative potentials where ATP breakdown due to muscular work occurs on a much faster time- scale than regeneration by glycolysis [264], indicating that the transport 72 CHAPTER 2

function of the CK/PCr system is nearly irrelevant for fast-twitch glycolytic muscles, (iv) Biochemical and 31P-NMR experiments demonstrated that in brain, heart and skeletal muscle, the cy tosolic CK activity as well as the overall flux through the CK reaction are much higher than the maximal rates of ATP synthesis and ATP consumption, suggesting that the CK system in the cytosol is in a near-equilibrium state [42,88,205,262,299,310,346,381,423,475]. Under near-equilibrium conditions, [ATP] and [ADP] are efficiently buffered, as is corroborated by physiological experiments revealing that during muscular and nervous work, [ATP] remains almost constant, whereas [PCr] decreases [381], for a review see Ref. 417]. (ν) No functional coupling was detected between Mi-AK and mitochondrial oxidative phosphorylation in the heart of the horseshoe crab Limulus polyphemus, meaning that this isoenzyme is in free equilibrium with the cytosolic substrate concentrations [106]. Thus, the function of this Mi-AK isoenzyme is likely to ensure near-equilibrium conditions of the phosphagen kinase reaction in the intermembrane space rather than to participate in a "shuttle mechanism" for high-energy phos­ phates. A "transport" function for the CK/PCr system is favoured by the following arguments: (i) The presence of Mi-CK as well as of cytosolic CK isoenzymes within the same cells suggests that they have different functions. Most attractively, Mi-CK is responsible for PCr synthesis, and significant fractions of cytosolic CK for ATP regeneration (Fig. 4C; page 70). (ii) The proportion of Mi-CK increases wi th theoxidative potential and therefore also with the expected relevance of the transport function for the proper function­ ing of a tissue (see also III-E. and III-F.). It is higher in slow-twitch than in fast- twitch muscles and highest in heart (up to 50 %; see Ш-А.). In brain, depending mostly on glycolysis, the relative proportion of Mi-CK is low [48391,423]. Furthermore, during development, the proportion of Mi-CK and muscle performance rose in parallel (see III-E.). Finally, chronic stimu­ lation [400] and endurance training [9] increased the relative proportion of Mi-CK, while Mi-CK activity decreased in the immobilized human leg after surgery [224]. (iii) In the neonatal rabbit heart, the flux through the CK reaction directly correlated with the relative proportion of Mi-CK [346]. Furthermore, model calculations suggested that upon heart stimulation, the flux through Mi-CK increases severalfold [251,529]. (iv) Whereas the overall flux through the CK reaction is probably independent of the metabolic state in skeletal muscle [529], it increases with workload in the heart [42,44,46,47, 262,300,346,386,529] suggesting that the CK/PCr system in working heart is no longer in a near-equilibrium state. This conclusion was also corroborated for several tissues suffering or recovering from hypoxia or ischemia [19,21,141,159,160,206,334] indicating that merely buffering the concentra- Review 73 tions of adenine nucleotides is not the sole function of the CK/PCr system. (v) Jacobus [218] calculated that the total ATP pool would only suffice for 10 s of cardiac work at normal rates of energy utilization. Even with PCr present, the "resting" heart turns over the total high-energy phosphate pool between two and four times per minute. Similarly, ATP plus PAr would only suffice for 1.5 s of flight of Locusta migratoria L. [402]. Assuming concentration gradients of 5 %, maximal flux rates (in μιηοίεβ/πυη/π^ heart) of 35 for MgATP, 57 for P;, 123 for PCr and 103 for Cr were calculated. However, the flux of ADP (0.112 цтоіез/тіп/пг^), primarily due to its very low concen­ tration, is in the same range as the maximum measured rate of ATP utilization (0.135 цто1е8/тіп/т§), indicating that especially at higher workloads, ATP-ADP flux alone might be insufficient to maintain appropriate ATP/ADP ratios throughout the cell [219]. (vi) In an elegant series of experiments it was shown that in sea urchin spermatozoa, very specific inhibition of the mitochondrial and flagellar CK isoenzymes by l-fluoro-2,4- dinitrobenzene (FDNB) attenuated flagellar movement in the two distal thirds of the sperm tail, indicating that ATP and ADP diffusion alone are only sufficient to ensure dynein ATPase activity in the proximal third of the tail. Indeed, when ATP was added externally to permeabilized, FDNB-treated sperms, they were able to swim again normally [463], indicating that the CK/PCr system is essential for energy supply in these highly polar cells. Evolutionary studies support this conclusion since high CK activities were found in sperms of the primitive type, whereas 10-100-fold lower CK activities were observed in sperms with a modified morphology [465]. "Primitive" sperms have long flagellae, depend on aerobic energy metabo­ lism, are typical for external fertilization, and the mitochondria are localized exclusively within the head. "Modified" sperms are typical for internal fertilization, depend also on glycolytic energy flux, and the mitochondria are localized in head and tail. That the CK/PCr system is essential for highly polarized cells is also supported by the finding of high CK activities in chicken and frog photoreceptor cells [465,496] as well as in ciliated cells from rabbit oviduct epithelium [465]. (vii) The finding of intracellular compartmentation of adenine nucleotides and Cr (see IV-C.) has the direct consequences that communication between the various compartments (mitochondria, myofibrils, etc.) and thus transport of CK substrates must happen, (viii) If only the buffer function were important for the CK/PCr system, a homogeneous dis tribution of CK isoenzymes would be expected in the cytosol. However, different proportions of cytosolic CK isoenzymes were found to be tightly bound to the myofibrillar M-band, the plasma membrane and the sarcoplasmic reticulum membrane where CK is thought to locally regenerate ATP as well as to be functionally coupled to myosin ATPase, 74 CHAPTER 2

+ + 2+ Na /K -ATPase and Ca -ATPase; respectively [51,120,301,357,383,386,455; for reviews see Refs. 38,379,488,492]. Co-localization of CK with ATP- requiring enzymes together with the likely potential that CK itself is a regulated enzyme might be an effective means of high-energy phosphate channelling, (ix) Theoretical considerations have shown that even under near-equilibrium conditions, most of the high-energy phosphate will be transported in the form of PCr at physiological ATP/ADP ratios [311]. Accordingly, Meyer et al. [311 ] hypothesize that the binding of CK near sites of ATP production or utilization "serves to raise the local enzyme activity where the flux is greatest, thus ensuring overall near-equilibrium with less total enzyme activity than would be necessary with uniformly distributed enzyme". Clearly, there were also some arguments raised against a transport function of the CK/PCr or AK/PAr systems. First, the AK activities in flight muscles of four insect species as well as in squid mantle muscle were found to be considerably smaller than the maximal ATP turnover rate [332,451]. However, comparison with mammalian tissues is difficult due to profound morphological differences (see above). And second, feeding of rats with the creatine analogue ß-guanidinopropionic acid (GPA) for 6-10 weeks resulted in the heart in a nine-fold decrease in [PCr] as well as in a four-fold decrease in the flux through the CK reaction, while GPA and GPAP were accumulated and [P,], [ATP], intracellular pH, oxygen consumption and cardiac perfor­ mance remained unchanged [424,426]. Since in GPA-fed rats, the measured rates of ATP turnover were 1.5-3 times greater than the flux through the CK reaction, it was concluded that PCr cannot be an obligate intermediate of energy transduction in the heart [426]. However, it has to be stressed that during the rather long feeding periods, compensatory metabolic adaptations take place [427] and that GPAP can serve as a CK substrate to buffer [ATP] quite efficiently during transitions between work states [88]. Furthermore, the left ventricular developed pressure was significantly lower in rats fed GPA or ß-guanidinobutyric acid compared to controls [531]. Many studies in the past aimed to proof that the CK/PCr system in some way is essential for energy metabolism. However, the GPA experiments suggest that vital mammalian tissues (except spermatozoa and retina?) still function reasonably well at greatly reduced concentrations of Cr and PCr (see Ш-F.), this fact simply reflecting the overcapacity of the CK/PCr system under normal conditions. Nevertheless, the CK/PCr system in all likelihood has both a transport and buffer function and thereby increases the thermo­ dynamic efficiency of energy metabolism, as will be discussed below. A second misleading aim was the attempt to demonstrate that a "sufficient" flux of high-energy phosphates can not be attained by ADP and ATP alone. Review 75

Evidently, with sufficiently steep concentration gradients, adenine nucleotide turnover and ADP-ATP flux might be balanced as well [219]. Therefore, the function of the CK/PCr system may not simply be to guarantee a sufficient flux of high-energy phosphates from sites of ATP production to sites of ATP consumption, but rather to ensure a sufficient flux of high-energy phosphates to maintain appropriate ATP/ADP ratios throughout the cell in order to ensure a proper functioning of all ATPases in tissues with rapidly changing energy demands [96,97,143,157,229,230]. High ATP/ADP ratios point to high phosphorylation potentials and thus to high affinities for ATP hydrolysis. In resting skeletal muscle as well as in heart at a basal metabolic rate, the affinity for ATP hydrolysis was found to be in the range of 55-64 kj/mol [143,229; for a review see Ref. 230], while in other tissues, it amounts to only 45-51 kj/mol [230,440]. The facts that a variety of cellular ATPases require energies of 41-44 kj/mol [96,97,229], that the affinity for ATP hydrolysis seems tobe very efficiently buffered between 45 and 50 kj/mol [229] and that the mechanical performanceof the heart drastically decreases below 48 kj/mol [229] strongly corroborate that the ATP/ADP ratio is of prime importance for energy metabolism. Most important in this respect is the finding that the Ca2+-ATPase of the sarcoplasmic reticulum depends on a very high affinity for ATP hydrolysis to be able to reduce the cytoplasmic [Ca2+] to 100 nM and thus to ensure muscle relaxation [229]. Cessation of proper Ca2+ sequestration mayevenbe thebiochemical basis for muscular fatigue [97,157]. Kammermeier [230] further hypothesized that the affinity for ATP hydrolysis provides an explanation why CK is presen t in heart and skeletal muscle as well as in brain, but not in liver. In tissues with affinities of 55-64 kj/mol, [ADP] would have to be kept very low. Under these conditions, the diffusion gradients of ADP and ATP required to maintain the desired high-energy phosphate flux would cause an affinity gradient of 2-3 kj/mol^m. In tissues with affinities below 51 kj/mol, [ADP] would be higher and the affinity gradients much smaller. Thus, the CK/PCr system may have the additional function of flattening affinity gradients for ATP hydrolysis and thus avoiding energy dissipation [311]. However, this interpretation raises some additional questions: Why do brain, heart and skeletal muscle depend on a higher affinity for ATP hydrolysis? What are the processes or chemical reactions requiring affinities of 55-64 kj / mol? And how is the affinity for ATP hydrolysis regulated wi thin a cell? Answering these questions may provide some deeper insight into the "real" functions of the CK/PCr system. Doubts about the real functions of the CK/PCr system might also be raised by the finding of cytosolic as well as Mi-CK isoenzymes in smooth muscle tissues [207] displaying much lower maximal performances than striated muscles. However, in analogy to striated muscles, "fast" and "slow" 76 CHAPTER 2

smooth muscles may be discriminated. Most visceral and vascular smooth muscles are characterized by long-lasting tonic contractions, contain cy tosolic and Mi-CK isoenzymes and consequently can be compared with slow-twitch skeletal muscles. In contrast, chicken gizzard displays only phasic contrac­ tions without a tonic component, where maximal force is developed after 10-15 s of stimulation [126]. Since chicken gizzard, in addition, has no Mi-CK, but considerable amounts of BB-CK [207], it may be compared with fast- twitch skeletal muscles. The obvious difference between slow- and fast- twitch skeletal muscles and "slow" and "fast" smooth muscles is the time- scale of metabolic changes. An attractive explanation to investigate is that diffusion of CK substrates is much more hindered in smooth than in striated muscles, but that the CK/PCr system in both muscle types has the same functions. The last function of the CK/PCr system to be discussed is to "accelerate" and "smooth" transitions between different work states. If ADP liberated by muscular contraction or by all sorts of cellular ATPases were the signal for ATP production by mitochondrial oxidative phosphorylation (see IV-D.), and if the system under these conditions were non-linear, then oscillations around the new steady-state level would be expected following changes in workload. These oscillations would be paralleled by fluctuations in [ADP], [ATP] and in the ATP/ADP ratio. Since ATP and ADP are key regulators of many of the fundamental metabolic pathways, the net result might be a destabilization of whole cellular metabolism. Because near-equilibrium conditions only allow for a linear system, where oscillations do not occur, the CK/PCr system may dampen the oscillations mentioned above and conse­ quently stabilize whole cellular metabolism. It has to be stressed that the oscillations discussed here are completely different from those described in several practical and theoretical studies [21,134,139]. The former are due to an approach to a new steady-state level, whereas the latter result from regularly fluctuating energy demands characteristic of cardiac muscle, or

from alternating perfusion with 02 and N2. By accelerating "communication" between sites of ATP production and ATP consumption, theCK/PCr system may, in addition, reduce the transient times for reaching a new steady-state. Transient times can be shortened in two different ways, either by elevating enzyme activities in such a way as to guarantee near-equilibrium conditions [109,182] or by metabolite channel­ ling [343]. Both of these mechanisms are probably operational in the CK/PCr system, thus indicating as well that reducing transient times is important for energy metabolism in tissues with rapidly changing energy demands. The cytosolic CK isoenzymes are likely to be in a near-equilibrium state (see above), and Mi-CK is apparently involved in metabolic channelling of high- Review 77 energy phosphates out of the mitochondria (see ПІ-С). All of these latter interpretations are strongly favoured by the fact that the CK/PCr system is predominantly found in tissues with high and fluctuating energy demands like brain, cardiac and skeletal muscle, but not in tissues with high but more or less constant energy demand like liver. In conclusion, the CK/PCr system has two main primary functions: (i) it buffers the concentrations of ADP and ATP, and (ii) accelerates the transport of high-energy phosphates between sites of ATP production and ATP con­ sumption, thereby accelerating also communication and feedback regulation between the two complementary parts of the CK isoenzyme system. The remaining functions of the CK/PCr system are direct consequences of the "buffer" and "transport" functions. Besides buffering [H+] and preventing loss of adenine nucleotides (see IV-A.), (iii) the CK/PCr system, due to keeping [ADP] low, maintains a high affinity for ATP hydrolysis which seems to be crucial for various ATPases. (iv) Furthermore, transient times between different workloads are shortened by the CK/PCr system due to metabolite channelling by Mi-CK and near-equilibrium conditions of the cytosolic CK isoenzymes, (v) Finally, due to the CK reaction being in a near- equilibrium state in the cytosol, oscillations in the concentrations of high- energy phosphates may be avoided upon abrupt changes in workload.

ÍV-C. Subcellular compartmentatton of CK substrates

The function of the CK/PCr system to accelerate the "communication" between sites of ATP production and ATP consumption, as it was proposed in the preceding chapter, would be even more crucial if discrete subcellular pools of CK substrates (ADP, ATP, PCr, Cr, H+) occurred. In fact, much evidence for microcompartmentation of CK substrates has been accumulated over the last thirty years [for reviews see Refs. 38,492]. Incubation of striated muscle of the frog with tritium-labelled adenine or Cr, followed by fixation of the tissue (in the absence or presence of a lanthanum salt to "precipitate" the high-energy phosphates) and autoradiography of thin sections, revealed an accumulation of adenine nucleotides and PCr at discrete locations within the myofibrils [186-189]. Depending on the conditions of fixation, adenine nucleotides were found in a narrow disk either in the I-band or the A-band, but in both instances close to the A-I boundary [187,188]. PCr was concen­ trated in a narrow disk within the I-band, approximately half way between the A-I boundary and the Z-line [189]. For this latter region, the local [PCr] was calculated to be 166 mmol/kg muscle. MM-CK in chicken pectoralis or rat muscles was also found to be loosely bound to the I-band, together with 78 CHAPTER 2

glycolytic enzymes, in addition to its localization within the M-band of the sarcomeres [491,497]. Even though serious reservations have to be made about the method employed by Hill [186-189], the colocalization of MM-CK and its substrates nevertheless points to an important role of microcompartmentation of the CK/PCr system within the I-band of the myofibrils. After incubation or perfusion of rabbit or rat heart with 14C-labelled Cr, the specific radioactivity of PCr (SAp^), surprisingly, was significantly higher

than the 5Αα(5Αρα/5Αα 1.24-1.87) [272,392]. Subsequent washoutof excess Cr or anoxia even increased the SAp^/SA,^ ratio to 5-11, which is a dear indication for microcompartmentation. Savabi [392] from her data concluded that 55 % of the total Cr (Cr + PCr) in spontaneously beating rat heart atria is PCr, 9 % constitute theCr-pool 1 which is readily accessible to phosphorylation by (Mi-) CK, and 36 % constitute the Cr-pool 2. Cr in this latter pool is rather inaccessible to phosphorylation and may be bound to subcellular structures. Incubation of the atria with 14C-Cr causes a selective uptake of radioactive Cr into pool 1. PCr hydrolysis due to anoxia also leads to an accumulation of Cr in pool 1, suggesting that only a small proportion of the total Cr is metabolically "active". If these findings were true, the validity of the conclusions from the GPA experiments would have to be seriously questioned. For example, only about 80 % of PCr and 60 % of total Cr were depleted by feeding animals with GPA for 8 weeks [531]. If GPA-feeding selectively diminished only the Cr- pool 2, pool 1 and therefore also energy metabolism in general might remain almost unaffected. Besides sequestration of ATP and ADP in membrane-enclosed dense granules of blood platelets [471] and accumulation of ATP in the nucleus of frog oocytes [315], as much as -30 % of the intracellular ATP was found to be "trapped" in cardiac and liver mitochondria [144; for reviews see Refs. 177,247,430]. Furthermore, subcellular fractionation of rat heart and liver in non-aqueous media revealed the mitochondrial matrix [ATP] / [ADP] ratio to be much lower than the cytosolic [ATP] / [free ADP] ratio [177,430]. Together wi th the findings of separate mitochondrial and cytosolic CK isoenzymes and of diffusion limitations of adenine nucleotides across the outer mitochondrial membrane [147,148], these results clearly point to distinct adenine nucleotide pools within the mitochondria and in the cytosol. The terms "functional coupling" and "metabolite channelling" auto­ matically imply microcompartmentation. Since much evidence has been accumulated for functional coupling between Mi-CK, ANT and porin as well as between cytosolic CK isoenzymes and myosin-ATPase of the myofibrils, Ca2+-ATPase of the sarcoplasmic reticulum and Na+/K+-ATPase of the sarcolemma, and since the different CK isoenzymes are in part bound to these Review 79 subcellular structures in an isoenzyme-specif ic manner (see Ш-С. and IV-B.), it seems very likely that not only between mitochondrial matrix, intermembrane space and cytosol, but also in the cytosol itself, different microcompartments for CK substrates exist, at least at high work-loads. The latter suggestion is strengthened by experiments on hypoxic or ischemic heart and smooth muscle, indicating that in spite of the high cytosolic CK activities, theCK substrates are not in a near-equilibrium sta te [21,159,160,206]. The most likely explanation is microcompartmentation of adenine nucleotides at the myofibrils [160,301,379,478,488] where ADP was supposed tobe tightly bound to actin [144,501]. 31P-NMR data were also interpreted in favour of subcellular compart- mentation of CK substrates. In conventional saturation transfer experiments under steady state conditions, the apparent high-energy phosphate flux in the forward direction of the CK reaction (PCr synthesis) was in most cases considerably higher than that in the reverse direction (ATP synthesis) [42,44,48,141,252,299,310,337,423]. Two alternative explanations were given for this difference: (i) microcompartmentation of CK substrates [251,252, 337,529]. Two models accounting for microcompartmentation were de­ veloped. According to the model of Koretsky et al. [251], differences in flux rates between the forward and reverse direction of the CK reaction are observed in conventional saturation transfer, but not in 2D or inversion transfer NMR experiments, with these differences being a direct measure of microcompartmentation. In fact, the forward and reverse fluxes through the CK reaction were found to be equal in 2D-NMR experiments [15]. On the other hand, the model of Zahler et al. [529] is based on the assumption that some pools of cellular adenine nucleotides are NMR-invisible and conse­ quently not saturable in conventional saturation transfer experiments. This model correctly predicts that the flux PCr -> ATP increases with workload in the heart [42,44,46,47,262,300,346, 386], but not in skeletal muscle [67,141,360,425]. Furthermore, this model was shown to be consistent with a NMR-invisible ATP pool at the mitochondria, but not with localization of NMR-invisible ATP exclusively within the myofibrils. The notion that mitochondrial adenine nucleotides, probably due to the extremely high viscosity of the mitochondrial matrix [394], are restricted in rotational diffusion and are therefore NMR-invisible, is in support of the second model [529]. (ii) Alternatively, the participation of ATP in various side-reactions has been proposed as basis for the unequal forward and reverse fluxes through the CK reaction [141,299]. This interpretation is favoured by the finding that under steady state conditions, the difference between both fluxes vanished in multiple saturation transfer experiments, when for the determination of the ATP -» PCr flux, the P, as well as PCr resonances were concomitantly 80 CHAPTER 2

saturated [472]. Nevertheless, more work is needed to define clearly which of the two interpretations or which of the two models of micro- compartmentation is correct. Finally, 31P-NMR experiments on intact cells and tissues as well as anoxic muscle revealed different pH environments for P, [412; for a review see Ref. 201] which may be explained by pH gradients of up to 0.5-1.0 pH units between the cytoplasmic and mitochondrial compartments. In conclusion, telling arguments for distinct mitochondrial and cytosolic CK substrate pools have been accumulated over the years. In contrast, only indirect evidence is currently available for microcompartmentation of CK substrates in the cytosol itself. Even though it is difficult to achieve, unequivocal corroboration of this latter type of microcompartmentation will be an important task for future research, since it critically determines the functionsoftheCK/PCr system and likely alsoa variety of cellular processes.

ÍV-D. Regulation of mitochondrial oxidative phosphorylation

Since the CK/PCr system in some tissues seems to be an essential part of the energy metabolism, its possible involvement in the regulation of mi tochondrial oxidative phosphorylation shall briefly be discussed. Up to now, mitochondrial respiration was proposed to correlate with (and therefore to be regulated by) [ADP], the [ATP] / [ADP] ratio, the phosphorylation poten­ tial, the "adenylatereaction pressure", the [NADH]/[NAD+] ratio, [Ca2+] or

02 supply [152,217,263; for reviews see Refs. 17,56,102,150,171,177,181,218, 253,294,306]. Whereas experiments on isolated mitochondria as well as on liver, skeletal muscle and newborn sheep hearts revealed a good correlation between the respiration rate and the concentration of ATP hydrolysis pro­ ducts [17,181,351], these findings were challenged by experiments on in vivo or perfused adult heart, brain and kidney, which indicated a poor correlation between these parameters [16,17,74,138,181 ]. Ins lead, especially for the heart, the [NADH]/[NAD+] ratio and [Ca2+] were favoured as primary regulators of oxidative phosphorylation, since Ca2+ at physiological concentrations was found to stimulate efficiently a variety of mitochondrial dehydrogenases [17,181]. At this point, a serious problem emerges. If reduction of transient times is an important function of the CK/PCr system (see IV-B.), then the ATP hydrolysis products must be primary determinants of the respiration rate in adult heart and brain as well. How can the apparent discrepancy be ex­ plained? Let us consider the possible consequences of introducing the CK/PCr system in a cell. If, for example, [ADP] (or [ATP]/[ADP] or the Review 81 phosphorylation potential) determined the mitochondrial respiration rate [217,218], and if the diffusion of ADP were restricted, ADP concentration gradients would exist between sites of ATP consumption and ATP produc­ tion, with the steepness of the gradient increasing with workload. The larger the diffusion restrictions for ADP, the higher the apparent Km of oxidative phosphorylation for ADP will be. Introduction of the CK/PCr system would drastically reduce the concentration gradient of ADP and therefore decrease the apparent Km of oxidative phosphorylation for ADP. This latter effect would be even more pronounced in the case of diffusion restrictions for adenine nucleotides across the outer mitochondrial membrane [32,33,147,148]. In other words, when [ADP] is plotted against the respiration rate or the rate- pressure product, a much flatter line will be obtained in the presence of the CK/PCr system than in its absence, with this flatter dependency probably being barely detectable by 31P-NMR techniques [74]. These considerations may explain (i) why a clear correlation between the respiration rate and the concentration of ATP hydrolysis products was found in liver and isolated mitochondria, but not in adult heart exhibiting a fully developed CK system; (ii) why the regulation of oxidative phosphorylation by ATP hydrolysis products is apparently lost during postnatal development of the sheep heart [351 ], at about the same time when the CK system develops to its full maturity [199] and (iii) why training or chronic stimulation, known to increase the proportion of Mi-CK (see III-F.), also resulted in an apparent loss of respiratory control by the phosphorylation potential [86]. These reflections suggest that the ATP hydrolysis products ADP, P, and H+ are more important for respiratory control in the heart than currently believed, and that [Cr], [PCr] or the [Cr]/[PCr] ratio represent some sort of intermediate feedback signal for oxidative phosphorylation [42,221,314, 376,379,414,415]. Since Mi-CK is likely to be displaced from equilibrium, a prerequisite for metabolic control, it may even be envisaged that oxidative phosphorylation is rate-limited by the Mi-CK reaction [294]. Nevertheless, the ATP hydrolysis products are clearly not the sole determinants of the respiratory rate [17,177,181]. They may be regarded as primary regulators of oxidative phosphorylation, with the [NADH]/[NAD+] ratio or [Ca2+] play­ ing a more modulatory role, or vice versa. For further investigation of respiratory control in the heart, it will be essential to determine the diffusion limitations for ADP, be it in the cytosol or across the outer mitochondrial membrane. 82 CHAPTER 2

V. Perspectives

The aim of this review was to summarize the current knowledge about the biochemistry, physiology and pathology of mitochondrial creatine kinase and to present working hypotheses for future research. In addition to the "buffer" and "transport" function of the CK/PCr system, a third main function is proposed here, namely to reduce the transient times of the system to reach a new steady state upon abrupt changes in workload (see IV-B.). Transient times can be reduced in two different ways: (i) by increasing the enzymatic activities in such a way as to guarantee near-equilibrium condi­ tions. The cytosolicCK activity in heart and skeletal muscle was shown tobe severalfold higher than the maximal rates of ATP production or ATP consumption so that the cytosolic CK system is in a near-equilibrium state; (ii) by metabolic channelling of substrates. Many recent findings support the notion that octameric Mi-CK within mitochondrial contact sites (CS) is involved in metabolic channelling of high-energy phosphates across both mitochondrial membranes (see III-C). First, kinetic and thermodynamic exfjeriments revealed microcompartmentation of CK substrates within the mitochondrial intermembrane space which was either explained by enzyme- enzyme proximity of Mi-CK and ANT or by diffusion limitations for adenine nucleotides across the outer mitochondrial membrane. Second, it was rea­ lized that CS may play an important role in the export of high-energy phosphates out of the mitochondria. The ex tent of CS is variable and increases with mitochondrial stimulation. Third, Mi-CK was found to be enriched in mitochondrial CS. And finally, the highly symmetrical 3D structure of the Mi-CK octamer, with two identical top and bottom faces, seems to be ideally suited for a CS localization, since the top and bottom faces are likely to have the same potency to interact with membranes (see II-C). Together with the apparent channel through the octameric molecule, it may be proposed that octameric Mi-CK within CS simultaneously binds to both the inner and outer mitochondrial membrane and is functionally (and physically?) coupled to ANT of the inner and porin of the outer membrane (see Fig. 2.B; page 38). This highly ordered multienzy me complex could be an effective means of displac­ ing the CK reaction in the intermembrane space from equilibrium and thus of allowing PCr synthesis even at high cytosolic ATP/ADP ratios. In addi­ tion, functional coupling was also demonstrated for cytosolic CK with myosin ATPase of the myofibrils, Ca2+-ATPase of the sarcoplasmic reticulum and Na+/K+-ATPase of the plasma membrane. A further function of the CK/PCr system may be to dampen oscillations of [ATP] and [ADP] upon abrupt changes in workload (see IV-B.). Inherent Review 83 to the latter two ideas is the assumption that Cr acts as some sort of "signal transducer" for the feedback regulation of mitochondrial oxidative phosphorylation by ATP hydrolysis products (see IV-D.). Whereas the "transport" and "transient time reduction" functions of the CK/PCr system are thought to predominate in tissues with a high proportion of Mi-CK like heart and slow-twitch skeletal muscles, the "buffer" function is dearly more important in fast-twitch skeletal muscles with low amounts of Mi-CK. To dampen oscillations of [ATP] and [ADP] may be a crucial function of the CK/ PCr system in all CK-containing tissues. In order to scrutinize these ideas and to get a deeper insight into the "real" functions of the CK/PCr system, it will be indispensable to define dearly (i) the diffusion limitations for ADP, ATP, PCr and Cr in the cytosol of heart, brain, skeletal and smooth muscle, retina and spermatozoa in comparison to liver; (ii) the permeability properties of the outer mitochondrial membrane for all CK substrates and P^ (iii) the functional and/or physical coupling of Mi-CK to ANT and porin as well as the stoichiometries of the three proteins within and beyond the CS; (iv) the three-dimensional structure of the Mi-CK octamer at atomic resolution in order to see if a channel through the molecule really exists and if the active sites of the subunits are directed towards this channel, and (v) whether changes in the dimer to octamer ratio of Mi-CK as well as dissociation and reassociation of dimeric and octameric Mi-CK from and to mitochondrial membranes (see Ш-D.) also occur under in vivo conditions and how they influence mitochondrial PCr synthesis or feedback regulation of oxidative phosphorylation. These experiments may then serve as basis for mathematical modelling of the CK/PCr system which is an essential prerequisite for asking further questions. From the point of view of comparative biochemistry and physiology, it will be intriguing to investigate at which stage of evolution two distinct Mi-CK isoenzymes in different tissues of the same species appeared, and how widespread the octameric structure of the Mi-CK isoenzymes is. In this respect, the human Mi-CK isoenzymes are of particular interest, since up to now, no detailed characterization of the higher Mr forms of the purified isoenzymes has been published. In addition, ubiquitous and sarcomeric Mi-CK may serve as valuable tools to clarify whether heart and brain have different mitochondrial import machineries for precursor proteins. A variety of recent studies suggest that the significance of the CK system in human pathology is currently underestimated. Culturing adult rat cardiomyocy tes in a medium devoid of Cr or supplemented with GPA led to the formation of large cylindrical mitochondria with crystal-like indusions enriched in Mi-CK (see III-F.). Similar abnormal mitochondria containing crystal-like inclusions enriched in Mi-CK were also observed in human 84 CHAPTER 2

mitochondrial myopathies (see Ш-G. and Fig. 4; page 70). In this case, however, formation of mitochondrial inclusions did not seem to correlate with [Cr] in the tissue. Therefore, it will be an important task for future studies to determine the conditions causing the formation of abnormal mitochondria in myopathic tissues. Furthermore, animals fed with GPA or GBA may serve as models for the investigation of mitochondrial myopathies. Since the CK/ PCr system appears to be very important for spermatozoa, the relevance of defects within this system for male infertility will also be a promising subject for further research [198]. Finally, studies of patients with decreased levels of Mi-CK or with decreased tissue [PCr] may provide new views on the "real" functions of the CK/PCr system. In addition to all these challenges, the combination of established biochemical, biophysical and physiological methods with new, investigative techniques like 31P-NMR, overexpression of cloned CK genes in E. coiior yeast [14,62,79,140,254], site-directed mutagenesis, or generation of transgenic animals [66,255] will open the doors to new playing-fields for CK research.

Acknowledgements We are especially grateful to E. Furter-Graves for valuable comments on the manuscript. F. Trijbels, R. Sengers, W. Ruitenbeek, and H. M. Eppenberger are gratefully acknowledged for continuous support and critical reading of the manuscript. We are also obliged to E. Gnaiger, J. A. Hoerter, R. Ventura-Clapier, J. Clark and V. A. Saks for the stimulating discussions at the 2nd MERGEshop in Innsbruck, 15.-19. of December, 1991. In addition, R. Furter, W. Hemmer, T. Schnyder, A. M. Stadhouders, J. A. Hoerter, R. Ventura-Clapier, С Vial, T. Yu. Lipskaya, D. Brdiczka, K. Nicolay, P. Kaldis and T. Wirz are gratefully acknowledged for helpful discussion and for providing unpublished results. This work was supported by a graduate student training grant from the ΕΤΗ Zürich (to M.W.) and by grants from the Swiss National Science Foundation (No. 31-26384.89), the Swiss Foundation for Muscle Diseases, and the Helmut Horten Foundation (to T.W.). Review 85

References 21. Barbour, R.L, Sotak, C.H., Levy, G.C. and Chan, S.H.P (1984) Biochemistry 23, 6053- 6062. 1. Adams, V., Bosch, W., Schlegel, J., 22. Bark, C.J. (1980) J. Am. Med. Assoc. 243, Wallimann, T. and Brdiczka, D. (1989) 2058-2060. Biochim. Biophys. Acta 981,213-225. 23. Basson, CT., Grace, A.M. and Roberts, R. 2. Adams, V., Griffin, L., Towbin, ]., Gelb, В., (1985) Mol. Cell. Biochem. 67,151-159. Worley, K. and McCabe, E.R.B. (1991) 24. Beatty, E.M. and Doxey, D.L. (1983) Res. Biochem. Med. Metab. Biol. 45,271-291. Vet. Sci. 35,325-330. 3. Addink, A.D J., Boer, P., Wakabayashi, T. 25. Becker, S. and Schneider, F. (1989) Biol. and Green, D.E. (1972) Eur. J. Biochem. 29, Chem. Hoppe-Seyler 370,357-364. 47-59. 26. Belousova, L.V., Lipskaya, T.Yu., Temple, 4. Altschuld, R.A. and Brierley, G.P. (1977) J. V.D. and Rostovtsev, A.P. (1983) Adv. Mol. Cell. Cardiol. 9,875-896. Myocardiol. 3,585-595 (Chazov, E., Smimov, 5. Altschuld,R.(1980)inHeartCreatineKmase V. and Dhalla, N.S., eds). Plenum medical (Jacobus, W.E. and Ingwall, J.S., eds), pp. book company. New York. 127-132, Williams & Wilkins, Baltimore. 27. Belousova, L. V., Fedosov, S.N., Rostovtsev, A.P.,Zaitseva,N.N.andMyatlev,V.D.(1986) 6. Annesley, T.M. and Walker, J.B. (1977) Biochemistry USSR 51,405-420. Biochem. Biophys. Res. Commun. 74,185- 190. 28. Belousova, L.V., Fedosov, S.N., 7. Annesley, T.M. and Walker, J.B. (1980) J. Stelmaschuk, V.J. and Orlova, E.V. (1990) Biol. Chem. 255,3924-3930. Muscle Motil. 2, 31-36, Intercept Ltd, Andover, UK. 8. Apple, F.S., Rogers, Μ. Α., Sherman, W.M., Costili, D.L., Hagerman, F.C. and Ivy, J.L. 29. Belousova,L.V.,Fedosov,S.N.,Orlova,E.V. (1984) Clin. Chem. 30,413^16. and Stel'mashchuk, V.Ya. (1991) Biochem. 9. Apple,F.S.andRogers,MA.(1986)J.Appl. Int. 24,51-58. Physiol. 61,482^185. 30. Benfield, P.A., Graf, D., Korolkoff, P.N., 10. Ardali, D., Privat, J.-P., Egret-Charlier, M., Hobson, G. and Pearson, M.L. (1988) Gene Levrat, C, Lerme, F. and Louisot, P. (1990) 63,227-243. J. Biol. Chem. 265,18797-18802. 31. Bennett, V.D., Hall, N., DeLuca, M. and 11. Amaudo, E., Dalakas, M., Shanske, S., Suelter, C.H. (1985) Arch. Biochem. Biophys. Moraes, CT., DiMauro, S. and Schon, E.A. 240,380-391. (1991) Lancet 337,508-510. 32. Benz, R., Wojtczak, L., Bosch, W. and 12. Awaji, Y., Hashimoto, H., Matsui, Y., Brdiczka, D. (1988) FEBS Lett. 231,75-80. Kawaguchi, К., Okumura, К., Ito, Т. and 33. Benz, R., Kottke, M. and Brdiczka, D. (1990) Satake, T. (1990) Cardiovasc. Res. 24, 547- Biochim. Biophys. Acta 1022,311-318. 34. Bessman S.P.andFonyo,A.(1966)Biochem. 554. / 13. Baba, N., Ют, S. and Farrell, E.C. (1976) J. Biophys. Res. Commun. 22,597-602. Mol. Cell. Cardiol. 8,599-617. 35. Bessman, S.P. (1980) in Heart Creatine 14. Babbitt, P.C., West, B.L., Buechter, D.D., Kinase (Jacobus, W.E. and Ingwall, J.S., eds), Kuntz, I.D. and Kenyon, G.L. (1990) pp. 75-79, Williams & Wilkins, Baltimore. Biotechnology 8,945-949. 36. Bessman, S.P. and Geiger, P.J. (1980) Coir. 15. Balaban, R.S., Kantor, H.L. and Ferretti, Top. Cell. Regul. 16, 55-86. J.A. (1983) J. Biol. Chem. 258,12787-12789. 37. Bessman, S.P. and Geiger, P.J. (1981 )Science 16. Balaban, R.S., Kantor, H.L., Katz, L.A. and 211,448-452. Briggs, R.W. (1986) Science 232,1121-1123. 38. Bessman, S.P. and Carpenter, CL. (1985) 17. Balaban, R.S. (1990) Am. J. Physiol. 258, Annu. Rev. Biochem. 54,831-862. C377-C389. 39. Bickerstaff, G.F. and Price, N.C. (1978) Int. J. Biochem. 9,1-8. 18. Baldwin, K.M., Cooke, D.A. and Cheadle, W.G. (1977) J. Mol. Cell. Cardiol. 9,651-660. 40. Biermans, W., Bemaert, I., De Bie, M., Nijs, B. and Jacob, W. (1989) Biochim. Biophys. 19. Baneijee, Α., Grosso, M.A., Brown, J.M., Acta 974,74-80. Rogers, K.B. and Whitman, G.J.R. (1991) Am. J. Physiol. 261, H590-H597. 41. Biermans, W., Bakker, A. and Jacob, W. 20. Barbour, R.L., Ribaudo, J. and Chan, S.H.P. (1990) Biochim. Biophys. Acta 1018, 225- (1984) J. Biol. Chem. 259,8246-8251. 228. 86 CHAPTER 2

42. Bittl, J.A. and Ingwall, J.S. (1985) J. Biol. Biochem Biophys. 253,122-132. Chem. 260,3512-3517. 65. Brooks, S.P.J. and Suelter,C.H. (1987) Arch. 43. Bittl,J.A., Weisfeldt, M.L. and Jacobiîs, W.E. Biochem. Biophys. 257,144-153. (1985) J. Biol. Chem. 260,208-214. 66. Brosnan, M.J., Chen, L, Wheeler, CE., Van 44. Bittl, J.A. and Ingwall, J.S. (1986) Circ. Res. Dyke, T.A and Koretsky, A.P. (1991) Am. J. 58,378-383. Physiol. 260, C1191-C1200. 45. BitÜJ.A.andIngwallJ.S.(1987)Circulation 67. Brown, T.R. (1982) Fed. Proc. 41,174-175. 95,196-1101. 68. Bücheier, К., Adams, V. and Brdiczka, D. 46. Bittl, J.A., Balschi, J.A. and Ingwall, J.S. (1991) Biochim. Biophys. Acta 1056, 233- (1987) J. Clin. Invest. 79,1852-1859 242. 47. Bittl, J.A., Balschi, J.A. and Ingwall, J.S. 69. Burgess, A.N., Liddell, J.M., Cook, W., (1987) Ciro Res. 60,871-878. Tweedhe, R.M. and Swan, I.D.A. (1978) J. 48. BittI,J.A.,DeLayre,J.andIngwalIJ.S.(1987) Mol. Biol. 123,691-695. Biochemistry 26,6083-6090. 70. Cain, DJ. and Davies, R.E. (1962) Biochem. 49. Blum, H.E., Weber, В., Deus, В. and Gerok, Biophys. Res. Commun. 8,361-366. W. (1981) in Creatine Kinase Isoenzymes 71. Carlsson, E., Kjorell, U. and Thomell, L.-E. (Lang, H., ed), pp. 19-30, Spnnger-Verlag, (1982) Eur. J. Cell. Biol. 27,62-73. Heidelberg. 72. Carlsson, E., Grove, B.K., Walhmann, T., 50. Blum, H.E., Deus, B. and Gerok, W (1983) Eppenberger,H.M.andThomell,L.-E.(1990) J. Biochem. 94,1247-1257. Histochemistry 95,27-35. 51. Blum, H., Balschi, J. A. and Johnson, R.G. 73. Castaldo, G., Salvatore, F. and Sacchetti, L. (1991) J. Biol Chem 266,10254-10259. (1990) Clin Biochem. 23, 523-527. 52. Bohner,J.,Stein,W.,Steinhart,R.,Wurzburg, 74. Chance, В., Leigh, J.S., Kent, J., McCully, К., U. and Eggstein, M. (1982) Clin. Chem 28, Nioka, S., Clark, B.J., Mans, J.M. and 618-623. Graham, T. (1986) Proc. Natl. Acad. Sci. 53. Booth, R.F.G. and Clark, J.B. (1978) Biochem. U.S.A. 83,9458-9462. J. 170,145-151. 75. Chandler, W.L., Clayson, K.J., Longstreth, 54. Borrebaek, B. (1980) Arch Biochem. W Τ and Fine, J.S. (1984) Clin. Chem. 30, Biophys 203,827-829. 1804-1806. 55. Borrebaek, B. and Haviken, J T. (1985) 76. Chandler, W.L., Clayson, К J., Longstreth, Biochem. Med. 33,170-179. W.T. and Fine,J.S. (1986) Am. J. Clin. Pathol. 56. Brand, M.D. and Murphy, M P. (1987) Biol. 86,533-537. Rev. 62,141-193. 77. Cheetham, M.E., Boobis, L.H., Brooks, S. 57. Brautbar, Ν., Carpenter, С, Baczynski, R., and Williams, С (1986) J. Appi. Physiol. 61, Kohan, R. and Massry, S.G. (1983) Kidney 54-60. Int. 24,53-57 78. Chen, C.-H. and Lehninger, A.L. (1973) Arch. 58. Brdiczka,D.,KnolI,G.,Riesinger,I.,Weiler, Biochem Biophys. 154,449-459. U., Klug, G., Benz, R. and Krause, J. (1986) in 79. Chen, L.H., Babbitt, P.C., Vásquez, J.R., Advances in Experimental Medicine and West, B.L. and Kenyon, G.L. (1991) J. Biol. Biology, vol. 194 "Myocardial and Skeletal Chem 266,12053-12057. Muscle Bioenergetics" (Brautbar, Ν., ed), 80. Cheneval, D., Muller, M., Tom, R., Ruetz, S. pp. 55-69, Plenum Press, New York. and Carafoh, E. (1985) J. Biol. Chem. 260, 59. Brdiczka, D., Adams, V., Kottke, M. and 13003-13007. Benz, R. (1989) in Anion Carriers of 81. Cheneval, D. (1987) Dissertation No. 8377, Mitochondrial Membranes ( Azzi, Α., Nalecz, ΕΤΗ Zurich, Switzerland. Κ. Α., Nalecz, M. J. and Wojtczak, L, eds), 82. Cheneval, D. and Carafoh, E. (1988) Eur. J. pp. 361-372, Springer Verlag, Berlin. Biochem 171,1-9.

60. Brdiczka, D. (1990) Expenentia46/161-167. 83. Cheneval, D., Carafoh, E., Powell, G.L. and 61. Brdiczka, D. (1991) Biochim. Biophys Acta Marsh, D. (1989) Eur. J. Biochem. 186,415- 1071,291-312. 419 62. Brindle, K., Braddock, P. and Fulton, S. 84. Chida, K., Tsunenaga, M., Kasahara, K., (1990) Biochemistry 29, 3295-3302. Kohno, Y. and Kuroki, T. (1990) Biochem. 63. Brooks, S.P.J., Bennett, V D. and Suelter, Biophys. Res. Commun. 173, 346-350. C.H. (1987) Anal. Biochem. 164,190-198. 85. Chida, K., Kasahara, K., Tsunenaga, M., 64. Brooks, S.P.J. and Suelter, C.H. (1987) Arch. Kohno, Y., Yamada,S.,Ohmi,S. and Kuroki, Review 87

T. (1990) Biochem. Biophys. Res. Commun. Comp. Physiol. 160B, 459-468. 173,351-357. 107. Dowell, R.T. (1986) Biochem. Biophys. Res. 86. Clark, B.J., Acker, M.A., McCully, K., Commun. 141,319-325. Subramanian, H.V., Hammond, R.L., 108. Dowell, R.T. (1987) Biochem. Med. Metab. Salmons, S., Chance, В. and Stephenson, Biol. 37,374-384. L.W. (1988) Am. J. Physiol. C258-C266. 109. Easterby, J.S. (1991) J. Theor. Biol. 152,47- 87. Colombini, M. (1979) Nature 279,643-645. 48. 88. Conley, K. and Kushmerick, M.J. (1990) 110. Edström, L., Hultman, E., Sahlin, K. and Proc. Annu. Meet. 9th Int. Soc. Magn. Reson. Sjöholm, H. (1982) J. Physiol. 332,47-58. Med. 2,902. 111. Eggleton, P. and Eggleton, G.P. (1927) 89. Connett, R.J. (1987) J. Appi. Physiol. 63, Biochem. J. 21,190-195. 2360-2365. 112. Ellington, W.R. (1989) J. Exp. Biol. 143,177- 90. Cook,P.F.,Kenyon,G.L.andCleland,W.W. 194. (1981) Biochemistry 20,1204-1210. 113. Ellington, W.R. and Hines, A.C. (1991 ) Biol. 91. Cooper, J. and Trinick, J. (1984) J. Mol. Biol. Bull. 180,505-507. 177,137-152. 114. Ennor, A.H. and Morrison, J.F. (1958) 92. Csako, G., Papadopoulos, N.M., Jett, G.K. Physiol. Rev. 38,631-674. and Mcintosh, C.L. (1982) Clin. Chem. 28, 115. Eppenberger, H.M., Eppenberger, M., 2170-2172. Richtcrich, R. and Aebi, H. (1964) Dev. Biol. 93. Dalakas, M.C., Ilia, I., Pezeshkpour, G.H., 10,1-16. Laukaitis, J.P., Cohen, B. and Griffin, J.L. 116. Eppenberger-Eberhardt, M., Riesinger, I., (1990) N. Engl. J. Med. 322,1098-1105. Messerli, M., Schwarb, P., Müller, M., 94. Daum, G. (1985) Biochim. Biophys. Acta Eppenberger, H.M. and Wallimann,T. (1991) 822, Ы2. J. Cell Biol. 113,289-302. 95. Dawson, D.M., Eppenberger, H.M. and 117. Erashova, N.S., Saks, V.A., Sharov, V.G. Kaplan, N.O. (1967) J. Biol. Chem. 242,210- and Lyzlova,S.N. (1978) Biochem. Biopyhs. 217. Res. Commun. 82,1217-1222. 96. Dawson, M.J., Gadian, D.G. and Wilkie, 118. Erickson-Viitanen, S., Viitanen, P., Geiger, D.R. (1980) J. Physiol. 299,465-484. P.J., Yang, W.C.T. and Bessman, S.P. (1982) 97. Dawson, M.J., Gadian, D.G. and Wilkie, J. Biol. Chem. 257,14395-14404. D.R. (1980) Philos. Trans. R. Soc. Lond. 119. Erickson-Viitanen,S.,Geiger,P.J.,Viitanen, 289B, 445-155. P. and Bessman, S.P. (1982) J. Biol. Chem. 98. DeFuria, RA., Ingwall, J.S., Fossel, E.T. and 257,14405-14411. Dygert,M.K.(1980)inHeartCreatine Kinase 120. Erickson-Viitanen, S., Geiger, P., Yang, (Jacobus, W.E. and Ingwall, J.S., eds), pp. W.C.T. and Bessman, S.P. (1982) in Advances 135-139, Williams & Wilkins, Baltimore. in Experimental Medicine and Biology, vol. 99. Delahunty,T.J. (1983) Clin. Chem. 29,1484- 151 "Regulation of Phosphate and Mineral 1487. Metabolism" (Massry, S.G., Letteri, J.M. and 100. DeLuca, M. and Hall, N. (1980) in Heart Ritz, E.,eds), pp. 115-125, Plenum Publishing Creatine Kinase (Jacobus, W.E. and Ingwall, Corporation. J.S., eds), pp. 18-27, Williams & Wilkins, 121. Farrants, G.W., Hovmöller, S. and Baltimore. Stadhouders, A.M. (1988) MuscleNerve 11, 101. DeLuca, M., Hall, N., Rice, R. and Kaplan, 45-55. N.O. (1981) Biochem. Biophys. Res. 122. Farrell, E.G., Baba, Ν., Brierley, G.P. and Commun. 99,189-195. Grümer, H.-D. (1972) Lab. Invest. 27, 209- 102. Denton, R.M. and McCormack, J.G. (1990) 213. Annu. Rev. Physiol. 52,451-466. 123. Fedosov, S.N. and Belousova, L.V. (1988) 103. Desjardins, P.R. and Pesclovitch, R. (1983) Biochemistry USSR 53,478-491. Clin. Chim. Acta 135, 35-40. 124. Fedosov, S.N. and Belousova, L.V. (1989) 104. Dmitrenko, N.P. and Bukhanevich, A.M. Biochemistry USSR 54,39-50. (1973) Doklady Akademii Nauk SSSR 213, 125. Fiek, C., Benz, R., Roos, N. and Brdiczka, D. 963-965. (1982) Biochim. Biophys. Acta 688,429-440. 105. Doumen, С and Ellington, W.R. (1990) J. 126. Fischer, W. and Pfitzer, G. (1989) FEBS Lett. Comp. Physiol. 160B, 449-157. 258, 59-62. 106. Doumen, С and Ellington, W.R. (1990) J. 127. Fitch,C.D.,Jellinek,M.,Fitts,R.H.,Baldwin, 88 CHAPTER 2

K.M. and Holloszy,J.O. (1975) Am. J. Physiol. Acta 890,117-126. 228,1123-1125. 148. Gellerich, F.N., Bohnensack, R. and Kunz, 128. Fitch, CD., Chevli, R. and Jellinek, M. (1979) W.(1989)in Anion Carriersof Mitochondrial J. Biol. Chem. 254,11357-11359. Membranes (Azzi, Α., Nalecz, Κ. Α., Nalecz, 129. Fitch, CD. and Chevli, R. (1980) Metabolism M.J. and Wojtczak, L., eds), pp. 349-359, 29,686-690. Springer-Verlag, Berlin. 130. Font, В., Vial, C, Goldschmidt, D., 149. Gereken, G. and Schiette, U. (1968) Eichenberger, D. and Gautheron, D.C (1981 ) Experientia 24,17-19. Arch. Biochem. Biophys. 212,195-203. 150. Gibbs, С (1985) J. Mol. Cell. Cardiol. 17, 131. Font, В., Vial, C, Goldschmidt, D., 727-731. Eichenberger,D.andGautheron,D.C.(1983) 151. Gilliland, G.L., Sjölin, L. and Olsson, G. Arch. Biochem. Biophys. 220,541-548. (1983) J. Mol. Biol. 170,791-793. 132. Font, В., Eichenberger, D., Goldschmidt, D. 152. Gnaiger, E. and Jacobus, W.E. (1989) and Vial, С (1987) Mol. Cell. Biochem. 78, Biophys. J. 55,568a. 131-140. 153. Godinot, C, Vial,C., Font, B. and Gautheron, 133. Fontanel, H.L., Trask, R.V., Haas, R.C., D. (1969) Eur. J. Biochem. 8,385-394. Strauss, A.W., Abendschein, D.R. and 154. Gori, Z., De Tata, V., Pollera, M. and Billadello, J.J. (1991) Circ. Res. 68,1007-1012. Bergamini, E. (1988) Br. J. Exp. Pathol. 69, 134. Fossel, E.T., Morgan, H.E. and Ingwall, J.S. 639-650. (1980) Proc. Natl. Acad. Sci. U.S.A. 77,3654- 155. Gots, R.E., Gorin, F.A. and Bessman, S.P. 3658. (1972) Biochem. Biophys. Res. Commun. 49, 135. Freitag, H., Neupert, W. and Benz, R. (1982) 1249-1255. Eur. J. Biochem. 123,629-636. 156. Grace, A.M., Perryman, M.B. and Roberts, 136. Friedhoff, A.J. and Lerner, M.H. (1977) Life R. (1983) J. Biol. Chem. 258,15346-15354. Sci. 20,867-872. 157. Griese, M., Perlitz, V., Jüngling, E. and 137. Friedman, D.L. and Perryman, M.B. (1991 ) Karranermeier,H. (1988) J. Mol. Cell. Cardiol. J. Biol. Chem. 266, 22404-22410. 20,1189-1201. 138. From,A.H.L.,Zimmer,S.D.,Michurski,S.P., 158. Grobbel, M.A., Lawson, N.S. and Calam, Mohanakrishnan, P., Ulstad, V.K., Thoma, R.R. (1982) Clin. Chem. 28,1995-1996. W.J. and Ugurbil, K. (1990) Biochemistry 29, 159. Gudbjarnason, S., Mathes, P. and Ravens, 3731-3743. K.G. (1970) J. Mol. Cell. Cardiol. 1,325-339. 139. Funk, C, Clark, A. and Connett, R.J. (1989) 160. Gudbjarnason, S. (1971/72) Cardiol. 56, Adv. Exp. Med. Biol. 248, 687-692. 232-244. 140. Furter, R., Kaldis, P., Furter-Graves, E.M., 161. Gumaa, K.A. and McLean, P. (1969) Schnyder, T., Eppenberger, H.M. and Biochem. Biophys. Res. Commun. 36,771- Wallimann, T. (1992) submitted. 779. 141. Gadian, D.G., Radda, G.K., Brown, T.R., 162. Haas, R.C., Korenfeld, C, Zhang, Z., Chance, E.M., Dawson, M.J. and Wilkie, Perryman, В., Roman, D. and Strauss, A.W. D.R. (1981) Biochem. J. 194,215-228. (1989) J. Biol. Chem. 264, 2890-2897 and 142. Garber, A.T., Winkfein, R.J. and Dixon, 16332 (correction). G.H. (1990) Biochim. Biophys. Acta 1087, 163. Haas, R.C. and Strauss, A.W. (1990) J. Biol. 256-258. Chem. 265, 6921-6927. 143. Gard,J.K.,Kichura,G.M.,Ackerman,J.J.H., 164. Hagelauer, U. and Faust, U. (1982) J. Clin. Eisenberg, J.D., Billadello, J.J., Sobel, B.E. Chem. Clin. Biochem. 20,633-638. and Gross, R.W. (1985) Biophys. J. 48, SOS- 165.. Hall, N. and DeLuca, M. (1975) Biochem. SIS. Biophys. Res. Commun. 66,988-994. 144. Geisbuhler, T., Altschuld, R.A., Trewyn, 166. Hall, N., Addis, P. and DeLuca, M. (1977) R.W., Ansel, A.Z., Lamka, K. and Brierley, Biochem. Biophys. Res. Commun. 76,950- G.P. (1984) Circ. Res. 54,536-546. 956. 145. Gellerich, F. and Saks, V.A. (1982) Biochem. 167. Hall, N., Addis, P. and DeLuca, M. (1979) Biophys. Res. Commun. 105,1473-1481. Biochemistry 18,1745-1751. 146. Gellerich, F.N., Schlame, M. and Saks, V.A. 168. Hall, N. and DeLuca, M. (1980) Arch. (1983) Biomed. Biochim. Acta 10,1335-1337. Biochem. Biophys. 201,674-677. 147. Gellerich, F.N., Schlame, M., Bohnensack, 169. Hall, N. and DeLuca, M. (1984) Arch. R. and Kunz, W. (1987) Biochim. Biophys. Biochem. Biophys. 229,477-482. Review 89

170. ΗβΙΙ,Ν.βηαΟβΙ,ικβ,Μ (1986) in Advances 195. Hossle, J.P. (1987) Dissertation No. 8362, inExpenmen tal Medicine and Biology, Vol ΕΤΗ Zunch, Switzerland 194 "Myocardial and Skeletal Muscle 196. Hossle,J.P.,Schlegel,J ,Wegmann,G ,Wyss, Bioenergebcs"(Brautbar,N ,ed),pp 71-82, M , Bohlen, Ρ , Eppenberger, Η Μ , Plenum Press, New York Wallimann, Τ and Pemard, J -C (1988) 171. Hansford, R.G. (1985) Rev Physiol Biochem Biophys Res Commun 151,408- Biochem Pharmacol 102,1-72 416 172. Hanzlikova, V. and Schiaffino, S (1977) J 197. Hovius, R., Lambrechts, H, Nicolay, К and Ultrastruct Res 60,121-133 de Kruijff, В (1990) Biochim Biophys Acta 173. Harm, K., Musolf, К -M, SiragEldin, E and 1021, 217-226 Voigt, К D (1987) Çrztl Lab 33,259-266 198. Huszar,G ,Vigne,L andCorrales,M (1990) 174. Hart!, F.-U., Pfanner, Ν , Nicholson, D W J Androl 11,40-46 and Neupert, W (1989) Biochim Biophys 199 Ingwall, J.S., Kramer, M F and Friedman, Acta 988,1-15 W F (1980) in Heart Creatine Kinase 175. Hartmann, С, Christen, Ρ and Jaussi, R (Jacobus, WE and Ingwall, J S,eds),pp 9- (1991) Nature 352,762-763 16, Williams & Wilkins, Baltimore 176. Hasselbaink, H.DJ., Labruyère, WT, 200. Ingwall, J.S., Kramer, M F, Woodman, D Moorman, A F M and Lamers, W H (1990) and Friedman, W F (1981)Pediatr Res 15, Anat Embryol 182,195-203 1128-1133 177. Hassmen, 1.(1986) Biochim Biophys Acta 201. Ingwall, J.S. (1982) Am J Physiol 242, 853,135-151 H729-H744 178. Hayashi, T. and Tanaka, Τ (1985) Clin 202. Ingwall, J.S. and Fossel, E T. (1983) Chem 31,533-536 Perspectives in Cardiovascular Research 7, 179. Hebisch,S.,Sies,H andSoboll,S (1986)Pfl 601-617 gers Arch 406,20-24 203. Ingwall, J.S. (1984) Eur Heart J 5(Suppl F), 180. Heine, H. and Shaeg, G (1979) Acta Anat 129-139 103,1-10 204. Ingwall, J S., Kramer, M F, Fifer, M A, 181. Heineman, F.W. and Balaban, R S (1990) Lorell, В H , Shemin, R, Grossman, W and Annu Rev Physiol 52,523-542 Allen,PD (1985)N Engl J Med 313,1050- 182. Heinrich,R andSchuster,S (1991)J Theor 1054 Biol 152,57-61 205. Ingwall, J.S., Atkinson, D E , Clarki, К and 183. Heinz, J.W., O'Donnell, N J and Lott, ] A Fetters,J К (1990) Eur Heart J 11 (Suppl (1980) Clin Chem 26,1908-1911 В), 108-115 184. Hemmer, W., Glaser, S J, Hartmann, G R, 206. Ishida, Y. and Paul, RJ (1989) in Progr. Eppenberger,HM andWallimann,T (1991) Clin Biol Res, vol 315 Muscle Energetics NATO ASI Ser H56,143-147 (Paul,RJ ,Elzmga,G andYamada^^s), 185. Hershenson, S., Helmers, Ν , Desmueles, pp 417-128, Alan R Liss, Ine, New York Ρ and Stroud, R (1986) J Biol Chem 261, 207. Ishida, Y., Wyss, M, Hemmer, W and 3732-3736 Wallimann,T (1991) FEBS Lett 283,37-13 186. Hill,D.K.(1959)J Physiol 145,132-174 208 Iyengar, M R. and Iyengar, С L (1980) 187. Hill,D.K.(1960)J Physiol 150,347 373 Biochemistry 19,2176-2182 188. Hill, D.K. (1960) J Physiol 153,433 446 209. Iyengar, M.R. (1984) J Muscle Res Cell 189. Hill,D.K.(1962)J Physiol 164,31 50 Motil 5,527-534 190. Hirata, R.D.C, Hirata, M H , Strufaldi, В, 210. Jacobs, H., Heldt, H W and Klingenberg, Possik, R A andAsai,M (1989) Clin Chem M (1964) Biochem Biophys Res Commun 35,1385-1389 16,516-521 191. Hochachka,P W.andMommsen,T Ρ (1983) 211. Jacobs, Η К. (1974) Fed Proc, Fed. Am Science 219,1391-1397 Soc Exp Biol 33,1534 192. Hoerter, JA., Kuznetsov, A and Ventura- 212. Jacobs, Η К. and Graham, M (1978) Fed Clapier, R (1991) Cire Res 69,665-676 Proc , Fed Am Soc Exp Biol 37,1574 193. Holtzman,D.,McFarland,E W ,Jacobs,D, 213. Jacobus,W.E.andLchmnger,AL (1973)] Offutt,MC andNeuringer,LJ (1991)Dev Biol Chem 248,4803-4810 Brain Res 58,181-188 214. Jacobus, W.E. (1980) in Heart Creatine 194. Hopkins, S.F , McCutcheon, E Ρ and Kinase (Jacobus, W E and Ingwall, J S,eds), Wekstein,DR (1973) Cire Res 32,685-691 pp 1-5, Williams ácWilkins, Baltimore 90 CHAPTER 2

215. Jacobus,W.E.,Bittl,J.A.andWeisfeldt,M L. Ν Α., Lakomkm, V.L., Steinschneider, Α.Y., (1980) in Heart Creatine Kinase (Jacobus, Sevenna, M.Y., Veksler, V.l. and Saks, V.A. W.E. and Ingwall, J.S., eds), pp. 155-175, (1988) J. Mol. Cell. Cardiol. 20,465^179. Williams & Wilkins, Baltimore. 236. Kapelko, V.l., Saks, V.A., Novikova, N.A., 216. Jacobus, W.E. and Saks, V.A. (1982) Arch. Gohkov, M.A., Kupnyanov, V.V. and Biochem. Biophys. 219,167-178. Popovich, M.I. (1989) J. Mol. Cell. Cardiol. 217. Jacobus, W.E., Moreadith, R.W. and 21,79-83. Vandegaer, K.M. (1982) J. Biol. Chem. 257, 237. Kapelko, V.l., Veksler, V.l., Popovich, M.I. 2397-2402. and Ventura-Clapier, R.(1991) Am.J. Physiol. 218. Jacobus, W.E. (1985) Annu. Rev. Physiol. Suppl. 261,39-44. 47,707-725. 238. Keighren, M.A. and Price, N.C. (1978) 219. Jacobus, W.E. (1985) Biochem. Biophys. Res. Biochem. J. 171,269-272. Commun. 133,1035-1041. 239. Keller, T.C.S. and Gordon, P.V. (1991) Cell 220. Jacobus, W.E., Vandegaer, K.M. and МоЫ. Cytoskel. 19,169-179. Moreadith, R.W. (1986) in Advances in 240. Kenyon, G.L. and Reed, G.H. (1983) in Experimental Mediane and Biology, Vol. Advances in Enzymology (Meister, Α., ed), 194 "Myocardial and Skeletal Muscle pp. 367-426, J. Wiley & Sons, New York. Bioenergeücs" (Brautbar, N., ed), pp. 169- 241. Keshgegian, AA. and Marchant, B.L. (1983) 191, Plenum Press, New York Clm. Chem. 29,1727-1730. 221. Jacobus, W.E. and Diffley, D.M. (1986) J. 242. Keto,A.I.and Doherty, M.D. (1968) Biochim. Biol. Chem. 261,16579-16583. Biophys. Acta 151,721-724 222. James, G.P. and Harrison, R.L (1979) Clin. 243. Khuchua, Z.A., Ventura-Clapier, R., Chem. 25,943-947. Kuznetsov, A.V., Gnshin, M.N. and Saks, 223. James, P., Wyss, M., Lutsenko, S., V.A. (1989) Biochem. Biophys. Res. Wallimann, T. and Carafoli, E. (1990) FEBS Commun. 165, 748-757. Lett. 273,139-143. 244. Klein, S.C., Haas, R.C., Ferryman, M.B., 224. Janssen, E., Sylvén, С, Arvidsson, I. and Billadello, J.J and Strauss, A.W. (1991) J. Eriksson,E. (1988) Acta Physiol. Scand 132, Biol. Chem. 266,18058-18065. 515-517. 245. Kleine, Т.О. (1965) Nature 207,1393-1394. 225. Jaynes, P. and Feld, R.D. (1981) Clin. Chem. 246. Klingenberg, M. and Pfaff, E. (1966) in 27,1316-1317. Regulation of Metabolic Processes in 226. Jennings, R.B. and Steenbergen, С. (1985) Mitochondria (Tager, J.M., Papa, S., Annu. Rev. Physiol. 47,727-749. Quaghanello, E. and Slater, E.C., eds), pp. 227. Jockers-Wretou, E., Giebel, W. and 180-201,Elsevier Publishing, Co.,New York. Pfleiderer, G. (1977) Histochemistry 54,83- 247. Klingenberg, M. and Heldt, H.W. (1982) in 95. Metabolic Compartmentation (Sies, H., ed), 228. Jockers-Wretou, E. (1984) Clin. Chem. 30, pp 101-122, Academic Press, London. 1268-1269. 248. Klingenberg, M. (1985) in The Enzymes of 229. KammermeierjHvSchmid^P.andJunghng, Biological Membranes (Martonosi, A.N., ed), E. (1982) J. Mol. Cell. Cardiol. 14,267-277. vol 4, pp. 511 -553, Plenum Press, New York. 230. Kammermeier, H. (1987) J. Mol. Cell. 249. Knoll, G. and Brdiczka, D. (1983) Biochim. Cardiol 19,115-118. Biophys. Acta 733,102-110. 231. Kanemitsu, F., Kawamshi, I. and 250. Konieczny, S.F. and Emerson, C.P. (1987) Mizushima, J. (1982) Clin. Chun. Acta 119, Mol. Cell. Biol. 7,3065-3075. 307-317. 251. Koretsky,A.P.,Basus,V.J.,James,T.L.,Klein, 232. Kanemitsu, F., Kawamshi, I. and M P. and Werner, M.W. (1985) Magn. Res. Mizushima, J. (1982) Clin. Chun. Acta 122, Med 2,586-594. 377-383. 252. Koretsky,A.P.,Wang,S.,Klein,M.P.,James, 233. Kanemitsu, F., Kawamshi, I and Τ L. and Weiner, M.W. (1986) Biochemistry Mizushima, J. (1983) Clin. Chim. Acta 128, 25,77-84 233-240. 253. Koretsky, A.P., Katz, L. A. and Balaban, R.S. 234. Kanemitsu, F., Kawamshi, I., Mizushima,J. (1989)J Mol.Cell.Cardiol.21 (Suppl.I),59- and Okigaki, T. (1984) Clin. Chim. Acta 138, 66. 175-183. 254. Koretsky, A.P. and Traxler, B.A. (1989) FEBS 235. Kapelko,V.I.,Kupriyanov,V.V.,Novikova, Lett. 243,8-12. \ Review 91

255. Koretsky, A.P., Brosnan, MJ., Chen, L., Stricklcr, J.E. and Wilson, K.J. (1986) Chen, J. and Van Dyke, T. (1990) Proc. Natl. Biochem J. 233,51-56. Acad. Sci. U.S.A. 87,3112-3116. 272. Lee, Y.C.P. and Visscher, M.B. (1961) Proc. 256. Kottke, M., Adams, V., Riesinger, I, Bremm, Natl Acad. Sci. U.S.A. 47,1510-1515. G., Bosch, W., Brdiczka, D., Sandn, G and 273. Legssyer, A. and Arno-Dupont, M. (1988) Panfili, E. (1988) Biochim. Biophys Acta Comp. Biochem Physiol. 89B, 251-255. 935,87-102. 274. Lentjes, E.G.W.M. and Backer, E.T. (1987) 257. Kottke, M., Adams, V., Wallimann, T., Clin. Chim. Acta 168, 75-79. Kumar Nalam, V. and Brdiczka, D. (1991) 275. Lin, L.M. and Chen, Y.K. (1991) J. Oral Biochim. Biophys. Acta 1061,215-225. Pathol Med. 20,479-485. 258. Koven, IЛ., Freedman, M., Miller, D., Reece, 276. Linden, M. and Gellerfors,P. (1983) Biochim. S., Maitland, Α., Sigurdson, E. and Biophys. Acta 736,125-129. Blackstein, M.E. (1983) Surgery 94,631-635. 277. Lindsey, G.G. and Diamond, E.M. (1978) 259. Krause, J., Hay, R., Kowolhk, С and Biochim Biophys. Acta 524, 78-84. Brdiczka, D. (1986) Biochim. Biophys Acta 278. Lipskaya, T.Yu., Temple, V.D., Belousova, 860,690-698. L V.,Molokova,E.V andRybina,L.V.(1980) 260. Kubier, W. and Katz, A.M. (1977) Am J. Biokhimiya 45,1155-1166. Cardiol. 40,467-471. 279. Lipskaya, T.Yu., Temple, V J., Belousova, 261. Kuby, S.A. and Noltmann, E A. (1962) in L.V and Molokova, E.V. (1980) Biokhimiya The Enzymes, 2nd ed. (Boyer, P.D, Lardy, 45,1347-1351. H. and Myrbèck, К., eds), pp. 515-603, 280. Lipskaya, T.Yu., Kedishvih, N.Yu. and Academic Press, New York. Kalenova, M.E. (1985) Biochemistry USSR 262. Kupriyanov, V.V., Steinschneider, A.Y., 50,1339-1348. Ruuge, E.K., Kapel'ko, V.l., Zueva, M.Y., 281. Lipskaya, T.Yu. and Rybina, I.V. (1987) Lakomkin, V.L., Smirnov, V.N and Saks, Biochemistry USSR 52,594-603. V.A. (1984) Biochim. Biophys Acta 805,319- 282. Lipskaya, T.Yu., Bonsova, Τ.Α., Trofimova, 331. Μ E. and Kedishvih, N.Yu. (1987) 263. Kupriyanov, V.V., Lakomkin, V.L., Biochemistry USSR 52,1308-1318. Korchazhkina, O.V, Steinschneider, A.Y., 283. Lipskaya, T.Yu. (1989) Synopsis of Doctor Kapelko, V.l. and Saks, V A. (1991) Am J. of Science Thesis, Moscow University Physiol. Suppl. 261,45-53. Publishers, Moscow. 264. Kushmerick, M.J. (1986) in Advances in 284. Lipskaya, T.Yu. and Trofimova, M.E. (1989) Experimental Medicine and Biology, vol. Biochem. Int. 18,1029-1039. 194 "Myocardial and Skeletal Muscle 285. Lipskaya, T.Yu. and Trofimova, M.E. (1989) Bioenergehcs" (Brautbar, Ν., ed), pp 647- Biochem Int. 18,1149-1159. 663, Plenum Press, New York. 286. Lipskaya, T.Yu., Moiseeva, N.S. and 265. Kuznetsov, A.V. and Saks, V A. (1986) Trofimova, ME. (1989) Biochem Int. 18, Biochem. Biophys. Res Commun. 134,359- 1161-1171 366. 287. Lipskaya, T.Yu., Trofimova, M.E and 266. Kuznetsov, A.V., Khuchua, Ζ A and Saks, Moiseeva, N.S. (1989) Biochem. Int. 19,603- V.A. (1987) in Creatine Phosphate 613 Biochemistry, Pharmacology and Clinical 288. Liu,T.Z.,Shen,J T.,Lee,Y.-TN.andShohet, Efficiency (Saks, V.A., Bobkov, Y.G and S В. (1980) Clin. Chem. 26,1765. Strumia,E.,eds),pp 15-30,Edizioni Minerva 289. Loshon, CA., Rittenhouse, S E, Bowers, Medica, Tonno, Italy. G N and McComb, R B. (1986) Clin. Chem. 267. Kuznetsov,A.V.,Khuchua,Z.A.,Vassireva, 32,207-210. E.V.,Medvcd'eva,N.V.andSaks,V.A (1989) 290. Luby-Phelps, K., Castle, P.E., Taylor, D.L. Arch. Biochem. Biophys. 268,176-190. and Lang, F (1987) Proc. Natl. Acad. Sci. 268. Lapin, E., Maker, H.S and Lehrer, G M. USA 84,4910-4913 (1974) J. Neurochem. 22,11-14. 291. Ludwig, О., Krause, J , Hay, R. and Benz, R. 269. Laskowski, M.B., Che vh, R and Fitch, С D. (1988) Eur Biophys J 15, 269-276. (1981) Metabolism 30,1080-1085 292. Lyons, G.E., Muhlebach, S,, Moser, Α., 270. Lawson, J.W.R. and Veech, R.L. (1979) J. Masood, R., Paterson, B.M., Buckingham, Biol. Chem. 254,6528-6537. M.E andPernard,J.-C (1991) Development 271. Lebherz, H.G., Burke, T., Shackelford, J.E., 113,1017-1029. 92 CHAPTER 2

293. Mahler, M. (1979) Blochem. Biophys. Res. 316. Milner-White, E.J. and Watts, D.C. (1971) Commun. 88,895-906. Biochem. J. 122,727-740. 294. Mahler, M. (1985) J. Gen. Physiol. 86,135- 317. Miyake, S., Taketa, K., Izumi, M., 165. Nagashima, H., Nishina, Y., Kawanishi, K., 295. Malhotia, SJC. (1966) J. Ultrastruct. Res. 15, Ofuji, T. and Shimono, K. (1980) Clin. Chim. 14-37. Acta 108,323-328. 296. Marcillat, O., Goldschmidt, D., 318. Mommaerts, WJЛ.М. (1969) Physiol. Rev. Eichenberger, D. and Vial, С (1987) Biochim. 49,427-508. Biophys. Acta 890,233-241. 319. Moore, CL., Strasberg, P.M. and Kovac, С 297. Mastro, A.M., Babich, M.A., Taylor, W.D. (1973) Texas Rep. Biol. Med. 31,367-384. and Keith, A.D. (1984) Proc. Natl. Acad. Sci. 320. Moreadith, R.W. and Jacobus, W.E. (1982) J. U.S.A. 81,3414-3418. Biol. Chem. 257,899-905. 298. Matthews, P.M., Bland, J.L., Gadian, D.G. 321. Morris, G.E. (1989) Biochem.J. 257,461-469. and Radda, G.K. (1981) Biochem. Biophys. 322. Morrison, J.F. (1973) in The Enzymes, 3rd Res. Commun. 103,1052-1059. ed., vol. 8 (Boyer, P.D., ed), pp. 457-486, 299. Matthews, P.M., Bland, J.L., Gadian, D.G. Academic Press, New York. and Radda, G.K. (1982) Biochim. Biophys. 323. Moskvitina, E.L. and Belousova, L.V. (1985) Acta 721,312-320. Doklady Akademii Nauk SSSR 281, 209- 300. McAuliffé,J.J., Perry, S.B., Brooks, E.E. and 213. Ingwall, J.S. (1991) Biochemistry 30, 2585- 324. Müller, M., Moser, R., Cheneval, D. and 2593. Carafoli, E. (1985) J. Biol. Chem. 260,3839- 301. McClellan, G.,Weisberg,A. and Winegrad, 3843. S. (1983) Am. J. Physiol. 245, C423-C427. 325. Munneke, L.R. and Collier, G.E. (1988) 302. McGilvery, R.W. and Murray, T.W. (1974) Biochem. Genet. 26,131-141. J. Biol. Chem. 249,5845-5850. 326. Murone, Land Ogata,K. (1973) J. Biochem. 303. McGing, P.G., Teeling, M., McCann, Α., 74,41-48. Купе, F. and Camey, D.N. (1990) Clin. Chim. 327. Murphy, M.P., Hohl, С, Brierley, G.P. and Acta 187,309-316. Altschuld, R.A. (1982) Cire. Res. 51,560-568. 304. McPherson, A. (1973) J. Mol. Biol. 81,79-86. 328. NSgle, S. (1970) Klin. Wochenschr. 48,332- 305. Mekhfi, H., Hoertcr, J., Lauer, С, 341. Wisnewsky, С, Schwartz, К. and Ventura- 329. Nakagawa,H.,Kida,N.,Maeda,M.,Wakuta, Clapier,R. (1990) Am. J. Physiol. 258, Hll 51- Y. and Ohtaki, S. (1982) Clin. Chem. 28,723- H1158. 725. 306. Mela-Riker, LM. and Bukoski, R.D. (1985) 330. Neurohr, K.J., Gollin, G., Barrett, E.J. and Annu. Rev. Physiol. 47,645-663. Shulman, R.G. (1983) FEBS Lett. 159, 207- 307. Melmed, C, Karpati, G. and Carpenter, S. 210. (1975) J. Neurol. Sci. 26,305-318. 331. Newman, RA., Hacker, M.P. and Fagan, 308. Mercer, D.W. and Talamo, T.S. (1985) Clin. M.A. (1982) Biochem. Pharmacol. 31,109- Chem. 31,1824-1828. 111. 309. Meyer, R.A. and Terjung, R.L. (1980) Am. J. 332. Newsholme, E.A., Beis, I., Leech, A.R. and Physiol. 239, C32-C38. Zammit, V.A. (1978) Biochem. J. 172, 533- 310. Meyer, RA., Kushmerick, M.J. and Brown, 537. T.R. (1982) Am. J. Physiol. 242, CI-СП. 333. Nicolay, K., Rojo, M., Wallimann, T., Demel, 311. Meyer, R.A., Sweeney, H.L. and R. and Hovius, R. (1990) Biochim. Biophys. Kushmerick, M.J. (1984) Am. J. Physiol. 246, Acta 1018,229-233. C365-C377. 334. Norwood, W.I., Norwood, CR., Ingwall, 312. Meyer, RA., Brown, T.R. and Kushmerick, J.S. and Fossel, E.T. (1979) Biophys. J. 25, M.J. (1985) Am. J. Physiol. 248, C279-C287. 275a. 313. Meyer, RA., Brown, T.R., Krilowicz, B.L. 335. Norwood,W.I.,Ingwall,J.S.,Norwood,C.R. and Kushmerick, M.J. (1986) Am. J. Physiol. and Fossel, E.T. (1983) Am J. Physiol. 244, 250, C264-C274. C205-C210. 314. Meyer, RA. (1988) Am. J. Physiol. 254, 336. Nosek, T.M., Fender, К.Y. and Godt, R.E. C548-C553. (1987) Science 236,191-193. 315. Miller, D.S. and Horowitz, S.B. (1986) J. 337. Nunnally, R.L. and Hollis, D.P. (1979) Biol. Chem. 261,13911-13915. Biochemistry 18,3642-3646. Review 93

338. Ogunio, ЕЛ., Peters, T.J. and Hearse, D J. 464. (1977)Cardiovasc. Res. 11,250-259. 357. Quest, A.F.G. and Shapiro, B.M. (1991) J. 339. Ohira,Y.,Kanzaki,M.andChen,C.-S (1988) Biol. Chem. 266,19803-19811. Jpn. J. Physiol. 38,159-166. 358. Raimondi, G.A., Puy, R.J.M., Raimondi, 340. Okano, K., Yamamoto, К., Ohba, Y., А С ,Schwarz,E.R.andRosenberg,M.(1975) Matsumura, К. and Miyaji, T. (1987) Clin. Biomedicine 22,496-501. Chim. Acta 169,159-164. 359. Ratto, Α., Shapiro, B.M. and Christen, R. 341. Orth, H.D. (1981) in Creatine Kinase (1989) Eur. J Biochem. 186,195-203. Isoenzymes (Lang, H , ed), pp. 10-18, 360. Rees, D., Smith, M.B., Harley, J. and Radda, Springer-Verlag, Berlin G К (1989) Magn Res. Med. 9,39-52. 342. Osca^L.B.andHolloszyJ.CUWDJ Biol. 361. Reichmann, H., Hoppeler, H., Mathieu- Chem. 246,6968-6972 Costello, O., von Bergen, F. and Fette, D. 343. Ovádi,J.(1991)J.Theor Biol. 152,1-22. (1985) Pfl gers Arch. Eur. J. Physiol. 404,1- 344. Payne, R.M., Haas, R C. and Strauss, A.W. 9. (1991) Biochim. Biophys. Acta 1089, 352- 362. Rizzotti, P., Cocco, С, Burlina, Α., Marcer, 361. V, Plebani, M. and Burlina, Α. (1985) Clin. 345. Perriard,J.-C,Eppenberger,HM.,Hossle, Biochem 18,239-241. J.P. and Schafer, В (1987) in Isozymes Curr. 363. Roberts, R. and Grace, A.M. (1980) J. Biol. Top Biol. Med Res, vol. 14 "Mol Cell. Chem. 255,2870-2877. Biol.", pp. 83-101, Alan R. Liss, Inc., New 364. Roberts, R. (1980) Expenentia 36,632-634. York. 365. Roberts, R. (1980) in Heart Creatine Kinase 346. Perry, S.B., McAuliffe, J., Baischi, J.A., (Jacobus,W E andIngwall,J.S.,eds),pp.31- Hickey, P.R. and Ingwall, J.S. (1988) 45, Williams & Wilkins, Baltimore. Biochemistry 27,2165-2172. 366. Roberts, J.J. and Walker, J.B. (1982) Am. J. 347. Penyman, M.B., Strauss, A.W , Olson, J. Physiol. 243, H911-H916. and Roberts, R. (1983) Biochem Biophys. 367. Rock, R.C., Dreskin, R, Kickler, T. and Res Commun. 110,967-972 Wimsatt, T. (1975) Clin. Chim Acta 62,159- 348. Peiryman, M.B., Strauss, A W., Buettner, 162. T.L. and Roberts, R. (1983) Biochim Biophys. 368. Rogalsky, V.Y., Koven, I.H, Miller, D.R. Acta 747,284-290. and Pollard, A. (1985) Clin. Biochem. 18, 349. Petrofsky,J.S.andFitch,C D (1980) Pflugers 338-341 Arch. 384,123-129. 369. Roj o, M., Hovius, R., Demel, R., Walhmann, 350. Fette, D. (1966) in Regulation of Metabolic T.,Eppenberger,H M andNicolay,K.(1991) ProcessesinMitochondna(Tager,J M ,Papa, FEBS Lett. 281,123-129. S., Quagliancllo, E. and Slater, E.C, eds), 370. Roj o, M., Hovius, R., Demel, R.Α., Nicolay, pp. 28-50, Elsevier Publishing, Co., New К and Walhmann, T. (1991) J. Biol. Chem. York. 266,20290-20295. 351. Portman, M.A., Heineman, F W and 371. Roos, Ν., Benz, R. and Brdiczka, D. (1982) Balaban, R.S (1989) J. Clin. Invest. 83,456- Biochim Biophys. Acta 686,204-214. 464. 372. Ruitenbeek, W., Trijbels, J M.F, Fischer, 352. Pratt, R., Vallis, L M , Lim, С W. and J C, Sengers, R.C.A., Janssen, A J M. and Chisnall, W.N. (1987) Pathology 19, 162- Kerkhof, C.M.C. (1989) J. Inher. Metab. Dis. 165. 12 (Suppl 2), 352-354 353. Quemeneur, E., Eichenberger, D., 373. Sahgal, V., Subramam, V., Hughes, R.,Shah, Goldschmidt, D., Vial, С, Beauregard, G. A and Singh, H. (1979) Acta Neuropathol. and Pober, M. (1988) Biochem. Biophys. 46,177-183. Res. Commun. 152,1310-1314. 374. Saks, V.A., Chemousova, G.B., Voronkov, 354. Quemeneur,E.,Marcillat,0.,Eichenberger, lu I., Smirnov, V N and Chazov, EI. (1974) D. and Vial, С. (1989) Biochem. Int. 18,365- Cire. Res. 34/35, Suppl. III, 138-148. 371. 375. Saks, V.A., Chemousova, G В., Gukovsky, 355. Quemeneur, E., Eichenberger, D. and Vial, D E., Smirnov, V.N. and Chazov, E I. (1975) С. (1990) FEBS Lett 262,275-278 Eur. J. Biochem 57,273-290. 356. Quest, A.F.G., Soldati, T., Hemmer, W., 376. Saks,V.A.,Lipina,N.V.,Smimov,V.N.and Perriard, J.-C, Eppenberger, H M and Chazov, E.I (1976) Arch. Biochem. Biophys. Walhmann, T. (1990) FEBS Lett 269, 457- 173, 34-41. 94 CHAPTER 2

377. Saks, V.A., Seppet, E.K. and Lyulina, N.V. 392. Savabi,F.(1988)Proc.Natl.Acad.Sd.US.A. (1977) Biokhimiya 42,579-588. 85,7476-7480. 378. Saks, V.A., Lipina, N.V., Sharov, V.G., 393. Savabi, F. (1988) Biochem. Biophys. Res. Smimov, V.N., Chazov, E. and Grosse, R. Commun. 154,469-475. (1977) Biochim. Biophys. Acta 465,550-558. 394. Scalettar, B.A., Abney, J.R. and 379. Saks, V.A., Rosenshtraukh, L.V., Smimov, Hackenbrock, CR. (1991) Proc. Natl. Acad. V.N. and Chazov, E.I. (1978) Can. J. Physiol. Sci. U.S.A. 88,8057-8061. Pharmacol. 56, 691-706. 395. Schlame, M. and Augustin, W. (1985) 380. Saks, V.A., Kupriyanov, V.V., Elizarova, Biomed. Biochim. Acta 44,1083-1088. G.V. and Jacobus, W.E. (1980) J. Biol. Chem. 396. Schlegel, J., Zurbriggen, В., Wegmann, G., 255,755-763. Wyss, M., Eppenberger, H.M. and 381. Saks, V.A. (1980) in Heart Creatine Kinase Wallimann, T. (1988) J. Biol. Chem. 263, (Jacobus, W.E. and Ingwall, J.S., eds), pp. 16942-16953. 109-124, Williams & Wilkins, Baltimore. 397. Schlegel, J., Wyss, M., Schiirch, U., 382. Saks,V.A.,Chemousov,G.B., Lyulina,N.V., Schnyder, T., Quest, Α., Wegmann, G., Khuchua, Z.A., Preobrazenskiy, A.N. and Eppenberger,H.M.andWallimann,T.(1988) Ventura-Clapier, R.N. (1984) in J. Biol. Chem. 263,16963-16969. Abhandlungen der Akademie der 398. Schlegel, J. (1989) Dissertation No. 8766, Wissenschaften der DDR, Abteilung ΕΤΗ Zürich, Switzerland. Mathematik - Naturwissenschaften - 399. Schlegel, J., Wyss, M., Eppenberger, H.M. Technik, pp.41-48, Akademie- Verlag, Berlin. and Wallimann, T. (1990) J. Biol. Chem. 265, 383. Saks, V.A., Ventura-Clapier, R., Khuchua, 9221-9227. Z.A., Preobrazhensky, A.N. and Emelin, 400. Schmitt, T. and Fette, D. (1985) FEBS Lett. I.V. (1984) Biochim. Biophys. Acta 803,254- 188,341-344. 264. 401. Schneider, C, Stull, G.A. and Apple, F.S. 384. Saks, V.A., Kuznetsov, A.V., Kupriyanov, (1988) Enzyme 39,220-226. V.V., Miceli, M.V. and Jacobus, W.E. (1985) 402. Schneider, Α., Wiesner, R.J. and Grieshaber, J. Bio!. Chem. 260,7757-7764. M.K. (1989) Insect Biochem. 19,471-480. 385. Saks, V.Α., Khuchua, Z.A., Kuznetsov, A.V., 403. Schnyder, T., Engel, Α., Lustig, A. and Veksler, V.l. and Sharov, V.G. (1986) Wallimann, T. (1988) J. Biol. Chem. 263, Biochem. Biophys. Res. Commun. 139,1262- 16954-16962. 1271. 404. Schnyder, T., Engel, Α., Gross, Η., 386. Saks, V.A., Kuznetsov, A.V., Huchua, Z.A. Eppenberger, H.M. and Wallimann,T. (1989) and Kupriyanov, V.V. (1986)in Advances in in Cytoskeletal and Extracellular Proteins Experimental Medicine and Biology, Vol. ( Aebi, U. and Engel, J., eds), Springer Series 194 "Myocardial and Skeletal Muscle in Biophysics 3, 39-41, Springer-Verlag, Bioenergetics" (Brautbar, Ν., ed.), pp. 103- Berlin. 116, Plenum Press, New York. 405. Schnyder,T., Sargent, D.F.,Richmond, T.J., 387. Saks, V.A., Khuchua, Z.A. and Kuznetsov, Eppenberger, H.M. and Wallimaim,T. (1990) A.V. (1987) Biochim. Biophys. Acta 891,138- J. Mol. Biol. 216,809-812. 144. 406. Schnyder, T. (1990) Dissertation No. 9250, 388. Saks, V.A. Kapelko,V.l., Kupriyanov,V.V., ΕΤΗ Zurich, Switzerland. Kuznetsov, A.V., Lakomkin, V.L., Veksler, 407. Schnyder, T., Gross, H., Winkler, H., V.l., Sharov, V.G., Javadov, S.A., Seppet, Eppenberger, H.M. and Wallimann, T. (1991) E.K. and Kairane, С (1989) J. Mol. Cell. J. Cell Biol. 112,95-101. Cardiol. 21 (Suppl. I), 67-78. 408. Schnyder, T., Winkler, H., Gross, H., 389. Saks, V.A., Belikova, Y.O. and Kuznetsov, Eppenberger, H.M. and Wallimann,T. (1991) A.V. (1991) Biochim. Biophys. Acta 1074, J. Biol. Chem. 266,5318-5322. 302-311. 409. Scholte,H.R.,Weijers,P.J.andWit-Peeters, 390. Saks, VA.,Belikova, Y.O., Kuznetsov, A.V., E.M. (1973) Biochim. Biophys. Acta 291, Khuchua, Z.A., Branishte, Т.Н., 764-773. Semenovsky, M.L. and Naumov, V.G. (1991 ) 410. Schölte,H.R. (1973) Biochim.Biophys. Acta Am. J. Physiol. Suppl. 261,30-38. 305,413-427. 391. Sanders, J.L., Joung, J.I. and Rochman, H. 411. Schölte, H.R. (1973) Biochim.Biophys. Acta (1976) Biochim. Biophys. Acta 438,407-411. 330,283-293. Review 95

412. Seeley, P.J., Busby, SJ.W., Gadian, D.G., 432. SiragEldm, E. and Klapdor, R. (1985) Klin. Radda, G.K. and Richards, R.E. (1976) Wochenschr. 63, 257-261. Biochem. Soc. Trans. 4,62-64. 433. SiragEldin, E., Gereken, G. and Harm, K. 413. Seraydarian, M.W., Sato, E., Savageau, M. (1986) J. Clin. Chem. Clin. Biochem. 24,847- and Harary, I. (1969) Biochim. Biophys. Acta 860. 180,264-270. 434. Smeitink, J., Wevers, R., Hulshof, J., 414. Seraydarian, M.W., Artaza, L. and Abbott, Ruitenbeek, W., v. Lith, T., Sengers, R., B.C. (1974) J. Mol. Cell. Cardiol. 6,405-413. Trijbels, F., Korenke, С. and Wallimann, T. 415. Seraydarian, M.W. and Artaza, L. (1976) J. (1992) Ann. Clin. Biochem. 29,196-201. Mol. Cell. Cardiol. 8,669-678. 435. Smeitink, J., Ruitenbeek, W., v. Lith, T., 416. Seraydarian, M.W. and Abbott, B.C. (1976) Sengers, R., Trijbels, F., Wevers, R., Speri, J. Mol. Cell. Cardiol. 8,741-746. W. and de Graaf, R. (1992) Ann. Clin. 417. Seraydarian,M.W.(1980)inHeartCreatine Biochem. 29,302-306. Kinase (Jacobus, W.E. and IngwallJ.S., eds), 436. Smeitink, J., Stadhouders, Α., Sengers, R., pp. 82-91, Williams & Wilkins, Baltimore. Rui tenbeek, W., Wevers, R., ter Laak, Η. and 418. Seraydarian, M.W. and Yamada, T. (1986) Trijbels, F. (1992) Neuromuscular Disorders in Advances in Experimental Medicine and (in press). Biology, vol. 194 "Myocardial and Skeletal 437. Smeitink, J., Ruitenbeek, W., Sengers, R., Muscle Bioenergetics" (Brautbar, Ν., ed), Wevers, R., v. Lith, T. and Trijbels, F. (1992) pp. 41-53, Plenum Press, New York. J. Inher. Metab. Dis. accepted. 419. Seraydarian, M.W. and Vial, С (1987) in 438. Smith, D.S. (1966) Progr. Biophys. Mol. The Heart Cell in Culture, vol. II (Pinson, Α., Biol. 16,107-142. ed), pp. 41-61, CRC Press, Boca Raton. 439. Sobel, B.E., Shell, W.E. and Klein, M.S. 420. Severin, S.E., Belousova, L.V. and (1972) J. Mol. Cell. Cardiol. 4,367-380. Moskvitina, E.L. (1983) Biochem. Int. 6,149- 440. Soboll,S.,Scholz,R.andHeldt,H.W.(1978) 156. Eur. J. Biochem. 87,377-390. 421. Sharov, V.G., Saks, V.A., Smimov, V.N. 441. Somer, H., Dotila, Α., Konttínen, A. and and Chazov, E.I. (1977) Biochim. Biophys. Saris, N.-E. (1974) Clin. Chim. Acta 53,369- Acta 468,495-501. 372. 422. Shields, R.P., Whitehair, CK., Carrow, R.E., 442. Srere, P.A. (1987) Annu. Rev. Biochem. 56, Heusner, W.W. and van Huss, W.D. (1975) 89-124. Lab. Invest. 33,151-158. 443. Srivastava, D.K. and Bernhard, S.A. (1987) 423. Shoubridge, E.A., Briggs, R.W. and Radda, Annu. Rev. Biophys. Biophys. Chem. 16, G.K. (1982) FEBS Lett. 140,288-292. 175-204. 424. Shoubridge, E.A. and Radda, G.K. (1984) 444. Stadhouders, A.M. and Sengers, R.C.A. Biochim. Biophys. Acta 805, 79-88. (1987) J. Inher. Metab. Dis. 10 (Suppl. 1), 62- 425. Shoubridge, E.A., Bland, J.L. and Radda, 80. G.K. (1984) Biochim. Biophys. Acta 805,72- 445. Stadhouders, A.Jap, P. and Wallimann, T. 78. (1990) J. Neurol. Sci. 98 (Suppl.), 304-305. 426. Shoubridge, E.A., Jeffry, F.M.H., Keogh, 446. Stadhouders, Α.,Jap, P., Winkler, H. P. and J.M., Radda, G.K. and Seymour, A.-M.L. Wallimann, T. (1992) J. Muscle Res. Cell (1985) Biochim. Biophys. Acta 847,25-32. Motil. 13,255A. 427. Shoubridge, E.A., Challiss, R.A.J., Hayes, 447. Stallings, R.L., Olson, E., Strauss, A.W., D.J. and Radda, G.K. (1985) Biochem. J. 232, Thompson, L.H., Bachinski, L. and Siciliano, 125-131. M.J. (1988) Am. J. Hum. Genet. 43,144-151. 428. Shoubridge, E.A. and Radda, G.K. (1987) 448. Stein, W., Bohner, J., Steinhart, R. and Am. J. Physiol. 252, C532-C542. Eggstein, M. (1982) Clin. Chem. 28,19-24. 429. Shy, G.M.and Gonatas,N.K. (1964) Science 449. Stein,W.,Bohner,J.andBah]inger,M.(1985) 145,493-495. Clin. Chem. 31,1952-1958. 430. Siess,E.A.,Brocks,D.G.andWieland,O.H. 450. Stein, W., Bohner, J., Renn, W. and (1982)inMetabolicCompartmentation(Sies, Maulbetsch, R. (1985) Clin. Chem. 31,1959- H., ed), pp. 235-257, Academic Press, 1964. London. 451. Storey,K.B. (1977) Arch. Biochem. Biophys. 431. Sion, J.-P.,Laureys, M., Gerlo, E. and Gorus, 179,518-526. F. (1989) J. Chromat. 496,91-100. 452. Ström, S. and Bendz, R. (1983) Acta Med. 96 CHAPTER 2

Scand. 213,289-294. 476. Veksler,V.I.,Ventura-Clapier,R.,Lechêne, 453. Ström, S. and Bend ζ, R. (1986) Clin. Chim. P. and Vassort,G. (1988) J. Mol. Cell.CardioI. Acta 159,219-228. 20,329-342. 454. StuckiJ.W.UgSOJEur.J.Biochem.lO^S/- 477. Veksler,V.I.,Murat,I.andVentura-Clapier, 267. R. (1991) Can. J. Physiol. Pharmacol. 69,852- 455. Sum!, T. and Kishino, Y. (1983) Cell. Mol. 858. Biol. 29,175-180. 478. Ventura-Clapier,R.,Mekhfi,H.andVassort, 456. Swanson, P.D. (1967) J. Neurochem. 14, G. (1987) J. Gen. Physiol. 89,815-837. 343-356. 479. Ventura-Clapier, R., Hoerter, J.Α., 457. Sylvén, С, Jansson, E., Kallner, A. and Kuznetsov, Α., Khuchua, Z. and Clark, J. Book, К. (1984) Scand. J. Clin. Lab. Invest. (1992) in Guanidino compounds in Biology 44,611-615. and Medicine (De Deyn, P.P., Marescau, B. 458. Sylvén, С, Kallner, Α., Henze, Α., Larson, and Stalon, V., eds), John Libbey & Сотр., F., Liska, J. and Mogensen, L. (1985) Clin. in press. Chim. Acta 151,111-119. 480. Vial, C, Godinot, C. and Gautheron, D. 459. Sylvén, С, Jansson, E. Szamosi, A. and (1972) Biochimie 54,843-852. Book, К. (1989) Scand. J. Thor. Cardiovasc. 481. Vial, C, Font, В., Goldschmidt, D. and Surg. 23,63-67. Gautheron, D.C. (1979) Biochem. Biophys. 460. SylvÄn, С, Lin, L., Kallner, Α., Sotonyi, P., Res. Commun. 88,1352-1359. Somogyi, E. and Jansson, E. (1991) Eur. J. 482. Vial, C, Marcillat, O., Goldschmidt, D., Clin. Invest. 21,350-354. Font, B. and Eichenberger, D. (1986) Arch. 461. Thunberg,T.(1911)Z.Physiol.25,915-916. Biochem. Biophys. 251,558-566. 462. Toleikis, A.I., Kal'venas, A.A., Dzheya, P.P., 483. Vial, C, Eichenberger, D. and Quemeneur, Prashkyavichyus, A.K. and Yasaitis, A.A. E. (1988)inSarcomericand Non-Sarcomeric (1988) Biokhimiya 53,649-654. Muscles Basic and Applied Research 463. Tombes, R.M. and Shapiro, B.M. (1985)Cell Prospects for the 90's (Carraro, U, ed), pp. 41,325-334. 613-618, Unipress, Padova, Italy. 464. Tombes, R.M. and Shapiro, B.M. (1987) J. 484. Vignais, P.V., Brandolin, G., Boulay, F., Biol. Chem. 262,16011-16019. Dalbon,P.,Block,M.R.andGauche,I.(1989) 465. Tombes, R.M. and Shapiro, B.M. (1989) J. in Anion Carriers of Mitochondrial Exp. Zool. 251,82-90. Membranes (Azzi, Α., Nalecz, K.A., Nalecz, 466. Trask, R.V. and Billadello, J.J. (1990) M.J. and Wojtczak, L, eds), pp. 133-146, Biochim. Biophys. Acta 1049,182-188. Springer-Verlag, Berlin. 467. Tsung, S.H. (1983) Clin. Chem. 29, 2040- 485. Vincent,A.andBlair,J.M.(1970)FEBSLett. 2043. 7,239-244. 468. Tsung, J.S.H. (1986) CRC Crit. Rev. Clin. 486. Wallimann, T. (1975) Dissertation No. 5437, Lab. Sci. 23,65-75. ΕΤΗ Zürich, Switzerland. 469. Turnier, D.M. and Walker, J.B. (1985) Arch. 487. Wallimann, T., Zurbriggen, В. and Biochem. Biophys. 238, 642-651. Eppenbergcr, H.M. (1985) Enzyme 33,226- 470. Turner, D.M. and Walker, J.B. (1987) J. Biol. 231. Chem. 262,6605-6609. 488. Wallimann, T. and Eppenberger, H.M. 471. Ugurbil, K., Holmsen, H. and Shulman, (1985) in Cell and Muscle Motility (Shay, R.G. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, J.W., ed). Vol. 6, pp. 239-285, Plenum 2227-2231. Publishing Corp., New York. 472. Ugurbil, K., Petein, M., Maidan, R., 489. Wallimann, T., Moser, H., Zurbriggen, В., Michurski, S. and From, A.H.L. (1986) Wegmann,G.and Eppenberger, H.M. (1986) Biochemistry 25,100-107. J. Muscle Res. Cell Motil. 7,25-34. 473. van Brussel, E., Yang, J.J. and Seraydarian, 490. Wallimann, T., Wegmann, G., Moser, H., M.W. (1983) J. Cell. Physiol. 116,221-226. Huber, R. and Eppenberger, H.M. (1986) 474. Vandegaer, K.M. and Jacobus, W.E. (1982) Proc. Natl. Acad. Sci. USA 83,3816-3819. Biochem. Biophys. Res. Commun. 109,442- 491. Wallimann, T., Schnyder, T., Schlegel, J., 448. Wyss, M., Wegmann, G., Rossi, A.-M., 475. Veech, R.L., Lawson, J.W.R., Cornali, N.W. Hemmer, W., Eppenberger, H.M. and Quest, and Krebs, H.A. (1979) J. Biol. Chem. 254, A.F.G. (1989) in Progr. Clin. Biol. Res., vol. 6538- 6547. 315 "Muscle Energetics" (Paul,R.J.,Elzinga, Review 97

G. and Yamada, К., eds), pp. 159-176, Alan 511. Wolff,J.andKobel,H.R.(1985)J.Exp.Zool. R. Liss, Inc., New York. 234,471-480. 492. Walllmann, T., Wyss, M., Brdiczka, D., 512. Wood, T. and Swanson, P.D. (1964) J. Nicolay, K. and Eppenberger, H.M. (1992) Neurochem. 11,301-307. Biochem. J. 281,21-40. 513. Woznicki, D.T. and Walker, J.B. (1980) J. 493. Watts, D.C. (1971) in Biochemical Evolution Neurochem. 34,1247-1253. and the Origin of Life (Schoffeniels, E., ed), 514. Wu, A.H.B. and Bowers, G.N. (1982) Clin. pp. 150-173, North-Holland Pubi., Chem. 28,2017-2021. Amsterdam. 515. Wu, А Л .В., Hcrson, V.C. and Bowers, G.N. 494. Watts, D.C. (1973) in The Enzymes, 3rd ed., (1983) Clin. Chem. 29,201-204. vol. 8 (Boyer, P.D., ed), pp. 383-455, 516. Wyss, M., Schlegel, J., James, P., Academic Press Inc., New York. Eppenberger,H.M.andWalIimann,T.(1990) 495. Watts, D.C. (1975) Symp. Zool. Soc. Lond. J. Biol. Chem. 265,15900-15908. 36,105-127. 517. Wyss, M. (1992) Dissertation No. 9777, ΕΤΗ 496. Wegmann, G., Huber, R., Zanella, E., Zürich, Switzerland. Eppenberger, RM. and Wallimann, T.(1991) 518. Wyss, M., Wallimann, T. and Köhrie, J. Differentiation 46,77-87. (1992) submitted. 497. Wegmann, G., Zanolla, E., Eppenberger, 519. Xie, G. and Wilson, J.E. (1990) Arch. H.M. and Wallimann, T. (1992) J. Muscle Biochem. Biophys. 276, 285-293. Res. Cell Motil, (in press). 520. Yamashita, K. and Yoshioka, T. (1991) J. 498. Wenger,W.C, Murphy, M.P., Brierley, G.P. Muscle Res. Cell Motil. 12,37-44. and Altschuld, R.A. (1985) J. Bioenerg. 521. Yang, W.C.T. and Dubick, M. (1977) Life Biomembr. 17,295-303. Sci. 21,1171-1178. 499. Weselake, R.J. and Jacobs, H.K. (1983) Clin. 522. Yang, W.C.T., Geiger, P.J., Bessman, S.P. Chim. Acta 134,357-361. and Borrebaek, B. (1977) Biochem. Biophys. 500. Wessels, Α., Vermeulen,].L.M.,Virágh,S.Z., Res. Commun. 76,882-887. Kálmán, F., Morris, G .E., Man, N.T., Lamers, 523. Yasui, T., Uzawa, R., Ishizawa, S., Takagi, W.H. and Moorman, A.F.M. (1990) Anat. Y., Hayama, T., Comi, К. and Ishii, T. (1984) Ree. 228,163-176. Electrophoresis'83 (Hirai, H., ed), pp. 417- 501. West, J.J., Nagy, В. and Gergely, J. (1967) J. 422, Walter de Gruyter St Co., Berlin. Biol. Chem. 242,1140-1145. 524. Yoshizaki, K., Nishikawa, H. and Watari, 502. Wevers, R.A., Mul-Steinbusch, M.W.F.J. and H. (1987) Jap. J. Physiol. 37,923-928. Soons, J.B.J. (1980) Clin. Chim. Acta 101, 525. Yoshizaki, K., Watari, H. and Radda, G.K. 103-111. (1990) Biochim. Biophys. Acta 1051, 144- 503. Wevers, R.A. (1981) Dissertation, State 150. University Utrecht, The Netherlands. 526. Yue, R.H., Jacobs, H.K., Okabe, K., Keutel, 504. Wevers, R.A., Reutelingsperger, C.P.M., H.J. and Kuby, S.A. (1968) Biochemistry 7, Dam, B. and Soons, J.B.J. (1982) Clin. Chim. 4291-4298. Acta 119,209-223. 527. Yuu, H., Takagi, Y., Senju, O., Hosoya, J.-I., 505. Wevers, R.Α., Jansen, P.H.P., vanWoerkom, Gomi, K. and Ishii, T. (1978) Clin. Chem. 24, L.M.J., Doesburg, W.H. and Hommes, O.R. 2054-2057. (1984) Clin. Chim. Acta 143,193-201. 528. Yuu, H., Ishizawa, S., Takagi, Y., Gomi, K., 506. Williamson, J., Greene, J., Chérif, S. and Senju, O. and Ishii, T. (1980) Clin. Chem. 26, Milner-White, E.J. (1977) Biochem. J. 167, 1816-1820. 731-737. 529. Zahler, R., Bittl, J. A. and Ingwall, J.S. (1987) 507. Winkler, H., Gross, H., Schnyder, T. and Biophys. J. 51,883-893. Kunath, W. (1991) J. Electron Microscopy 530. Ziter, F.A. (1974) Exp. Neurol. 43,539-546. Technique 18,135-141. 531. Zweier,J.L.,Jacobus, W.E., Korecky, B. and 508. Wirz, T. (1991) Dissertation No. 9409, ΕΤΗ Brandejs-Barry, Y. (1991 ) J. Biol. Chem. 266, Zürich, Switzerland. 20296-20304. 509. Wit-Pee ters, E.M., Schölte, H.R., van den Akker, F. and De Nie, I. (1971) Biochim. Biophys. Acta 231, 23-31. 510. Wolf, W.J., Rex, Κ.Α., Geshi, E. and Sordahl, L.A. (1991) Am. J. Physiol. 261, H1-H8. 98 A method for quantitative measurement of mitochondrial creatine kinase in human skeletal muscle

J. Smeitmk1, R. Wevers2, J. Hulshof1, W. Ruitenbeek1, T. v. Lith1, R. Sengeis1, F. Trijbels1, C. Korenke1 and T. Wallimann3

1 Institute of Paediatrics, University Hospital Nijmegen, The Netherlands. 2 Institute of Neurology, University Hospital Nijmegen, The Netherlands. 3 Institute for Cell Biology, Swiss Federal Institute of Technology, Zürich, Switzerland.

Chapter 3

Annals of Clinical Biochemistry 1992;29:196-201 100 CHAPTER 3

Summary

Defects in the mitochondrial energy generating system in patients with a mitochondrial myopathy are known to be localized in various enzyme complexes involved in energy production. Such a defect may exist at the level of mitochondrial creatine kinase (Mi-CK). On that account we have deve­ loped a method for measurement of the enzyme activity in human skeletal muscle biopsy material (>10 mg). Interfering creatine kinase isoenzymes are removed by anion exchange and affinity chromatography. The activity of Mi-CK in reference skeletal muscle homogenates amounts to 240 ± 88 mU/ mg protein (30 ± 8.0 mU/mg wet weight). Measurement of mitochondrial creatine kinase 101

Introduction

Creatine kinase (CK, EC 2.7.3.2) occurs as five isoenzymes in human skeletal muscle. Three of them, MM-, MB- and BB-CK, are found within the cytoplasm and two are strictly mitochondrial. MM-CK is the predominant cytoplasmic form in muscle. The existence of mitochondrial CK (Mi-CK), was first described in 1964 by Jacobs et al. [1]. It is located within the mitochondrial intermembrane space in tissues with a high energy demand such as muscle, brain, photoreceptor cells and spermatozoa. Mi-CK is restricted to mitochondria and seems to be well adapted to generate phosphocreatine (PCr) from ATP produced within the mitochondrial matrix by forming a functionally coupled micro compartment with the ATP/ADP-translocator [2]. PCr is then made available by a PCr-shuttle or by facilitated diffusion to those fractions of CK that are localized at specific intracellular sites of high energy demand [3]. Recently, based on mRNA studies, evidence was ob­ tained for more than one gene encoding for Mi-CK[4]. Since the discovery in 1964 of Mi-CK in isolated mitochondria from rat brain, skeletal and heart muscle [1], there have been only a few studies on human skeletal muscle Mi-CK. Quantitative data were presented by Apple and Rogers who showed an obvious training effect on the activity of the enzyme in gastrocnemius muscle from long-distance runners [5]. As yet there are no clinical studies on the relevance of Mi-CK in neuro-muscular diseases. Determination of Mi-CK in skeletal muscle is complicated by the pre­ sence of large quantities of cytoplasmic CK isoenzymes, mainly MM-CK. The similar isoelectric points of both isoenzymes in humans do not enable a quantitative assay based solely on differences in charge. Various mitochondrial myopathies are caused by defects in complexes of the mitochondrial energy generating system, but the cause can not be pinpointed to a single enzyme defect in all patients [6]. This may also occur at the level of Mi-CK. For that reason we have developed a method for quantitative measurement of Mi-CK activity in small amounts of human skeletal muscle and this is now reported.

Materials and methods

Preparation of muscle homogenates Muscle tissue samples (m. quadriceps) were obtained from patients (age range 22 months-59 years, 10 males and 9 females) clinically suspect for a neuromuscular disease. Normal results were found for muscle pyruvate and 102 CHAPTER 3

malate oxidation rates and for single enzymes of the respiratory chain [7], thus excluding a defect in the energy-generating system except for fatty add oxidation. Muscle tissue was freed from fat and connective tissue and homogenized (10% w/v) in a 10 mmol/L Tris/HCl buffer pH 7.4, containing 250 mmol/L sucrose, 2 mmol/L EDTA, 50 U/mL heparin (SETH-buffer) using a teflon-glass Potter-Elvehjem homogenizer and three different teflon pestles [7].

DEAE-Cellulose The anion exchanger DEAE-Cellulose (Whatman DE52) was prepared ac­ cording to Nealon and Henderson [8] and stored at 4 0C as a 50% (v/v)

suspension in a 30 mmol/L Na2HP04 / МаН2Ю4 buffer pH 7.4 containing 0.2 mmol/L phenylmethanesulfonyl fluoride and 0.2 mmol/L dithiothreithol (buffer A) with addition of 0.01% NaNj.

Sepharose immobilized anti-M-CK antibodies CNBr-activated Sepharose 4B (Pharmacia, Uppsala, Sweden, No. 17-0430- 01) was used for the coupling of anti-M-CK antibodies (polyclonal goat anti- M-CK antibody, Boehringer, Mannheim, Germany, No. 418 234). The protein concentration of the antibody sample after overnight dialysis against the coupling buffer recommended by Pharmacia was 2.5 g/L. The coupling was carried out according to the instructions of the manufacturer. Approximately 5 mg protein was bound per mL of activated Sepharose. Sepharose coupled anti M-CK was stored in buffer A (+ 0.01% NaNj) at 4 0C. After use, regeneration of the immobilized antibody was achieved with a 1 h incubation

in 2 mol/L NH4SCN (ammonium thiocyanide), followed by extensive washing of the Sepharose with buffer A.

The assay of mitochondrial CK Figure 1 gives a schematic representation of the assay procedure. The first step was the extraction of Mi-CK in a hypotonic phosphate buffer. Fifty microlitres muscle homogenate was diluted five-fold with buffer A with 0.05% Triton X-100 (v/v). After incubation under gentle mixing (1 h; room temperature), and centrifugation at 40 000 χ g for 30 min at 4 0C, the supernatant were used for measurements. One hundred and fifty microlitres of the supernatant was added to 50 μL DEAE-Cellulose slurry binding MB-CK and BB-CK. After incubation (10 min; room temperature) and centrifugation (1 min; 8300 χ g), 100 [iL supernatant and 150 \LL Sepharose bound anti M-CK were mixed and incubated (10 min at room temperature). The Sepharose immobilized antibodies against the M-CK chain bind most (>99%, see Results) of the MM-isoenzyme. After centrifugation (2100 χ g; 1 min) in an Ultrafree-MC Filter Unit (Millipore, Bedford, MA 01730 USA) to Measurement of mitochondrial creatine kinase 103 avoid contamination with Sepharose bound MM-CK, CK activity was measured in the presence and in the absence of inhibiting antibodies against the M-CK chain (polyclonal goa t antì-М-СК antibody, Boehringer, Mannheim, Germany, No. 418234). At this stage Mi-CK can be accompanied by a minimal amount of MM-CK, which is easily inhibited by the antibody. CK activity was measured (CBR-CK NAC-activated kit, Boehringer, Mannheim, Germany, No. 475742) at 30 0C on a COBAS-Mira analyzer (Hoffman La Roche, Basel, Switzerland). Cytochrome с oxidase, citrate synthase and protein were determined as described before [9,10,11].

Internal standard Approximately 2 mg cytochrome с (horse-heart, Boehringer, Mannheim, Germany, No. 103889) were added to 200 μι muscle extract as an internal standard to calculate the dilution factor during the chromatographic met­ hods. This factor was calculated from the absorbance at 550 run in oxidized and reduced form (2 mg of K3Fe(CN)6 and 5 mg of Na2S204 respectively) before and after the chromatographic steps (Fig. 1).

Electrophoresis Cellulose acetate electrophoresis was performed at 20 0C for 35 min at 300 V (Boskamp: Kammer F21 Microphor, power supply Pherostat 273 and MTB Folien No. 49206-04 / sample volume ΙμΙ with total CK activity >500 U/L) using a 73 mmol/L Tris-barbi tal-buffer pH 8.6 containing 21.7 mmol/L

Mi-CK extraction D 40 000 χ g supernatant + cyt с

DEAE Cellulose MB+BB-CK (batch wise) Cytochrome с D determination Sepharose bound MM-CK A 550 nm anti-M-CK (batch wise)

V Mi-CK activity measurement

Fig. 1. Schematic representation of the assay procedure. 104 CHAPTER 3

5,5-diethylbutyric acid and 1 mmol/L dithiothreitol. An agar-overlay nitro blue tetrazolium technique was used for specific CK staining [12].

Behaviour of purified CK isoenzymes in the Mi-CK assay CK isoenzymes were purified from human autopsy material (BB-CK from brain, MB-CK and Mi-CK from heart, MM-CK from skeletal muscle) to establish their behaviour during the Mi-CK assay. The various isoenzymes were separated from contaminating CK-forms using anion exchange chromatography on mini-columns [8]. The purity was checked electrophoretically. These purified CK isoenzymes were submitted to the Mi-CK assay.

Results

Mitochondrial CK extraction Tissue obtained by muscle biopsy was homogenized in an isotonic sucrose containing Tris buffer. Under these conditions Mi-CK remains bound to the mitochondrial inner membrane. We have studied the release of the enzyme after five-fold dilution of the homogenate in hypotonic phosphate buffers with varying molarity (20-200 mM). Maximum extraction was obtained with an 1 h incubation at room temperature in buffer A with addition of 0.05% Triton X-100 (v/v). The buffer had the additional advantage that the 40 000 χ g supernatant can be used directly in the subsequent DE AE-Cellulose step. With a higher phosphate molarity in the buffer MB-CK and BB-CK would not bind to the anion exchanger. The extract contained all CK isoenzymes, the amount of BB-CK being very low (Fig. 2, lane 1). For optimal extraction of Mi-CK the addition of a detergent showed to be obligatory. We tested Triton X-100 (Boehringer, Mannheim, Germany), n-dodecyl-ß-D-maltoside (Boehringer, Mannheim, Germany), n-octyl- glucoside (Boehringer, Mannheim, Germany), sodium dodecyl sulphate (Merck, Darmstadt, Germany), deoxycholic acid sodium salt (Fluka, Buchs, Switzerland), Brij 35 solution (Sigma, St. Louis, USA), Tween 20 (Sigma, St. Louis, USA) and Zwittergent detergent 3-14 (Calbiochem, San Diego, USA) in different concentrations. Triton X-100 0.05% produced the most efficient release of Mi-CK from the mitochondrial inner membrane. Sonification or freeze-thawing yielded 30% extra Mi-CK activity in the absence of Triton Χ­ Ι 00 but there was no additional effect in its presence. Extraction at pH 8.0 and 8.5 resulted in 25% less Mi-CK than extraction at pH 7.4. Addition of Triton X-100 at 0.05% showed no interference with the different steps of the procedure or enzyme assay. Measurement of mitochondrial creatine kinase 105

new *

MB

M f

1 2 3 4 5 6

Fig. 2. Electrophoretic pattern on cellulose polyacetate strips of CK-isoforms in the various steps of the assay procedure. (1) Supernatant of Mi-CK extraction; (2) and (4) CK-isotrol (Sigma No. c0153); (3) after DEAE-anion exchanger; (5) after Sepharose bound anti-M-CK; (6) ibid after addition of inhibiting M-CK antibodies. -» = origin; + = anode; - = cathode.

Anion exchange and affinity chromatography Fig. 2 (lane 3) shows that by batch-wise DEAE-anion exchange-chromato- graphy MB-CK and BB-CK were quantitatively removed from the sample. MM-CK and Mi-CK did not bind to the anion exchanger. The majority of the MM-CK was removed from the resulting supernatant by affinity chromato­ graphy with Sepharose bound anti M-CK (Fig. 2, lane 5). Apart from Mi-CK, which remained unbound, the supernatant contained a limited amount of MM-CK, that was inhibited by the addition of anti-M-CK antibodies (Fig. 2, lane 6). To calculate the correct dilution factor during the various steps in the analysis cytochrome с was used as an internal standard. It does not bind to the DEAE-Cellulose or to the Sepharose bound anti-M-CK. Furthermore it was shown to have no effect on the activity of Mi-CK and does not interfere with the assay system for CK.

Linearity and reproducibility The linearity of the assay was checked by dilution of the 40 000 χ g supernatant in 30 mmol/L phosphate buffer. Mi-CK activity in the sample with 10% of the original activity could still be measured reliably. The coefficient of corre- 106 CHAPTER 3

lation was 0.9997. Reproducibility was measured with five samples of the same muscle homogenate undergoing the complete assay procedure. The within-run coefficient of variation was 4%, the day to day variation was 14%.

Specificity The specificity of the assay was tested with purified preparations of MM-CK, MB-CK, BB-CK and Mi-CK in buffer A. Total CK activity of these prepara­ tions was 70 000,4200,2820 and 8420 U/ L, respectively. Mi-CK recovery after the various steps of the assay procedure was good (Table 1). The table furthermore illustrates that MB-CK and BB-CK in this order of magnitude were efficiently removed by the anion exchanger while a major percentage of MM-CK was removed by the Sepharose bound anti-M-CK. The small percentage of MM-CK that failed to bind in this affinity step was inhibited by anti-M added in the final CK activity measurement. Care must be taken to ensure that the binding capacity in the two chromatographic steps is not surpassed, since this would lead to falsely high values for Mi-CK. Therefore, total CK activity in the 40 000 χ g supernatant after Mi-CK extraction should not exceed 100 000 U/L. Furthermore it is recommended that the MM-CK binding efficiency of the Sepharose is checked by also measuring total CK activity (without addition of inhibiting anti-M-CK antibodies) in the supernatant of the Sepharose. The assay will produce reliable results if the inhibiting antibodies in the final measurement have enough capacity to inhibit the enzymatic activity of all MM-CK present in the supernatant of the Sepharose.

Reference values Reference values for Mi-CK were obtained from skeletal muscle biopsies of 19 patients (m. quadriceps; 10 male and 9 female; Table 2). Mi-CK was not only expressed per mg protein and per mg wet weight, but also on two mitochondrial marker enzymes e.g. per cytochrome с oxidase (an inner membrane bound one) and per citrate synthase (a mitochondrial matrix enzyme). We have experience that the mitochondrial content of human skeletal muscle can vary to a considerable extent, especially in pathological muscle tissue. In case of a decreased Mi-CK activity it is possibly to distin­ guish between a lowered mitochondrial content and a real deficiency, by expressing the Mi-CK per unit activity of mitochondrial reference enzymes. Total CK activity was used as a skeletal muscle enzyme marker. Table 3 shows the means with the standard deviations of these 19 patients. Although not statistically significant male patients showed a higher specific activity of total CK and Mi-CK compared to the females. Mi-CK as percentage of total CK showed no differences between the sexes and amounted to 0.9%. Mean Measurement of mitochondrial creatine kinase 107

Table 1. Behaviour of various CK isoenzymes in the different steps for the determination of Mi-CK

Partially purified CK isoenzymes

MM MB BB Mi

Extract 100% 100% 100% 100%

Supernatant after 98% <1.5% <0.5% nd DEAE Cellulose

Supernatant after <1% 0% 0% 113% anti-M Sepharose

After anH-M 0% 0% 0% 113% immunoinhibition nd= not determined.

Mi-CK activity of all patients, males and females, was 240 mU/mg protein or 30 mU/mg wet weight.

Discussion

Various authors have used electrophoretic techniques to demonstrate the presence of Mi-CK in heart tissue. In an earlier study we have used agarose electrophoresis to show the presence of Mi-CK in human skeletal muscle. This technique can only be used semiquanti ta lively and is hampered by the similar electrophoretic behaviour of MM-CK and Mi-CK. Weselake and Jacobs were the first to describe a technique for the separation of heart Mi-CK from cytoplasmic CK forms by hydrophobic interaction chromatography [13]. This technique has been used by Apple and Rogers [5] for quantitative determinations. Wallimann et al. used affinity chromatography with Ciba- chrome-blue Sepharose for the separation of chicken heart Mi-CK from MM-CK and from the quantità lively most important cytoplasmic CK-form in chicken heart BB-CK [14]. A complication in human skeletal muscle is the considerably higher amount of MM-CK present in comparison with heart tissue. This led us to develop a technique that effectively and quantitatively removes BB-CK and MB-CK from the homogenate in a first step (DEAE anion exchanger) followed by a second step that removes almost all MM-CK activity (anti-M-CK immobilized on Sepharose). The small percentage of MM-CK that is not bound by this immunoabsorben t is inhibited in the final CK activity 108 CHAPTER 3

Table 2. Reference values for Mi-CK in human m. quadriceps (n=19). Specific activity and Mi-CK expressed per unit of mitochondrial reference enzymes

Mean SD Range

(mU/mg protein) 28 037 7524 13 707-42468 (mU/mg wet weight) 3568 773 1871 - 4751

(mU/mg protein) 240 88 131-490 (mU/mg wet weight) 30 8.0 19-52 (U/U total CK) 0.0086 0.0015 0.0061-0.012 (U/U СОХ) 1.7 0.36 1.1-2.4 (U/U CS) 2.7 0.80 1.4-4.1

СОХ = Cytochrome с oxidase. CS = Citrate synthase.

Table 3. Comparison of total CK and Mi-CK activity values of 10 male and 9 female patients clinically suspect for a neuromuscular disease.

Male Female

Mean SD Mean SD

Total CK (mU/mg protein) 31798 7015 23 859 5889 Mi-CK (mU/mg protein) 273 103 203 48 Mi-CK (% of total CK) 0.9 0.2 0.9 0.2

measurement by M-CK inhibiting antibodies. The resulting enzymatic activ­ ity was shown to be caused solely by Mi-CK (Table 1; page 107). This was also confirmed electrophoretically after addition of anti-M-CK to the sample (Fig. 2, lane 6; page 105). The assay for Mi-CK can be applied on amounts of muscle tissue as small as 10 mg and can thus be carried out on needle biopsy material. Because of the similar physico-chemical properties of Mi-CK in human heart, skeletal muscle and brain tissue the assay described in this paper is expected to be useful also for the quantitation of Mi-CK in human heart and brain tissue [15,16]. Total CK activity per mg wet weight in our patient reference group (3568 U/g) is comparable with the values found by Apple and Rogers for male marathon runners (2985U/g)buthigher than in female runners (pretraining: 1750 U/g) [5]. Mi-CK activity per g wet weight (30 U/ g) is in a similar range of activity Measurement of mitochondrial creatine kinase 109 as given by Apple and Rogers for females: 27.2 U/g, but lower than they reported for males: 104 U/g [5]. Expressed per unit total CK, Mi-CK activity in m. quadriceps of our patients (0.9%) is probably in conformity with the non-running control muscles investigated by Apple and Rogers although they did not present exact data. Our mean Mi-CK percentage is somewhat lower than those found in marathon runners (female 1.5%, male 3.5%). This finding can be explained satisfactorily by the effect of training shown by Apple and Rogers [5]. Furthermore differences, like the homogenization buffer, between both studies in CK and Mi-CK extraction from muscle may play a role. Apart from the study by Apple and Rogers there are to our knowledge no other reports in literature presenting quantitative data on human skeletal muscle Mi-CK. Recently, two Mi-CK mRNAs have been described [4]. However, as yet nothing is known about the relationship between the different Mi-CK genes and the relative contribution of the putative isoenzymes to the total Mi-CK activity in muscle. We present reference values based on measurements in patients sus­ pected clinically to suffer from a neuromuscular disease. It would be interesting to compare our data with data on (untrained) healthy volunteers. Care should be taken in judging Mi-CK activity levels found in patients with extreme muscular disuse, in whom the opposite of the effect of training described by Apple and Rogers [5] may be anticipated. We have found no statistically firm indications for a sex dependent activity as described pre­ viously [5]. Further study will concentrate on a influence of age on the reference values of the enzyme. The youngest of our patients was 22 months old. Reference values for neonates and young children remain to be established. After extensive biochemical and histological examinations of muscle tissue a group of patients remains in whom there is evidence for a mitochondrial myopathy while the exact localization of the defect at enzyme level remains unresolved. The clinical spectrum of such patients is very heterogeneous and the age of onset, and the severeness of symptoms, may vary considerably. Frequently encountered symptoms in neonates are muscle weakness, hypotonia and failure to thrive most often accompanied by respiratory insufficiency and central nervous system involvement. In children and adult patients either myopathic symptoms alone, i.e. excercise intolerance and muscle weakness, or a multisystemic disorder occurs, the latter mostly being a combination of myopathic, encephalopathic and or ocular involvement. A defect at the level of Mi-CK may explain some of these cases. The determina­ tion of Mi-CK may prove to be a useful biochemical diagnostic tool in patients with a mitochondrial myopathy. Studies are underway to screen patients for defects in Mi-CK using the methodology described in this paper. 110 CHAPTER 3

References creatine kinase isoenzymes in serum by ion- exchange column chromatography. Clin 1. Jacobe H, Heldt HW, Klingenberg M. High Chem 1975; 21:392-7. actìvity of creatine kinase in mitochondria 9. Coopers teln SJ, Lázarow A. A micro- from muscle and brain and evidence for a spectrophotometric method for the separa te mitochondrial isoenzyme of creatine determination of cytochrome oxidase. J Biol kinase. Biochem Biophys Res Commun 1964; Chem 1951; 189: 665-70. 16:516-21. 10. Srere PA. Ci tra te synthase, EC4.1.3.7 citrate 2. Schlegel^ZurbriggenB^egmannCWyss oxaloacetate-lyase (CoA-acetylating). In: M, Eppenberger HM, Wallimann T. Native Lowenstein JM (ed) MethodsinEnzymology, mitochondrial creatine kinase forms Academic Press, London, 1969; 13: 3-11. octameric structures (1). J Biol Chem 1988; 11. LowiyOH,RosebroughNJ,FaiTAL,RandaIl 263:16942-53. RJ. Protein measurement with the Polin 3. WallimannT,EppenbergerHM.Localization phenol reagent.JBiolCheml951;193:265-75. and function of MM-CK, M-band model and 12. Harm K, Musolf K-M, SiragEldin E, Voigt CP-shu ttle. In: Cell and muscle motili ty (Shay KD. Verteilungsmuster cytosolischer und JW, editor) New York: Plenum Publishing mitochondrialer Kreatinkinase-Isoenzyme Corp., 1985; 6:239-85. im menschlicher Oberschenkel- und Augen- 4. HaasRC,StraussAW.Separatenucleargenes musculatur. Ärtzl Lab 1987; 33:259-66. encode sarcomere-specific and ubiquitous 13. Weselake RJ, Jacobs HK. Separation of human mitochondrial creatine kinase cytoplasmic and mitochondrial isoenzymes isoenzymes. J Biol Chem 1990; 265:6921-27. of crea tine kinase by hydrophobic interaction 5. AppleFS,RogersMA.Mitochondrialcreatine chromatography. Clin Chim Acta 1983; 134: kinase activity alterations in skeletal muscle 357-61. during long-distance running. J Appi Physiol 14. Wallimann T, Zurbriggen B, Eppenberger 1986;61:482-5. HM. Separation of mitochondrial creatine 6. SengeraRCA^tadhoudersAHTrijbelsJMF. kinase (MiMi-CK) from cytosolic creatine Mitochondrial myopathies: clinical, morpho­ kinase isoenzymes by cibachrome blue logical and biochemical aspects. Eur J Pediatr affinity chromatography. Enzyme 1985; 33: 1984; 141:192-207. 226-31. 7. Fischer JC, Ruitenbeek W, Cabreéis FJM, 15. Wevers RA, Mul-Steinbush MWFJ, Soons Janssen AJM, Renier WO, Sengers RCA, JBJ. Mitochondrial CK (EC 2.73.2) in the Stadhouders AM, Ter Laak HJ, Trijbels JMF, human heart. Clin Chim Acta 1980; 101:103- VeerkampJHA. A mitochondrial encephalo- 11. myopathy: the first case with an established 16. Wevers RA, Reutelingsperger CPM, Dam B, defect at the level of coenzyme Q. Eur J Ped Soons JBJ, Mitochondrial creatine kinase (EC 1986; 144:441-4. 2.7.3.2) in the brain. Clin Chim Acta 1981; 8. Nealon DA, Henderson AR. Separation of 119:209-23. Maturation of mitochondrial and other isoenzymes of creatine kinase in skeletal muscle of preterm bom infants

J. Smeitink1, W. Ruitenbeek1, T. v. Lith1, R. Sengers1, F. Trijbels1, R. Wevers1, W. Speri3 and R. de Graaf4

1 Institute of Paediatrics, University Hospital Nijmegen, The Netherlands. 2 Institute of Neurology, University Hospital Nijmegen, The Netherlands. 3 Children's Hospital, University of Innsbruck, Austria. 4 Department of Medical Statistics, University of Nijmegen, The Netherlands.

Chapter 4

Aimais of Clinical Biochemistry 1992;29:302-306 112 CHAPTER 4

Summary

We studied pre- and postnatal changes in total creatine kinase (CK) activity, mitochondrial creatine kinase (Mi-CK) activity and immunochemical reac­ tivity with anti-Mi-CK antibodies in skeletal musde specimens from 12 infants, 10 of them preterm born, after a pregnancy varying between 28 and 40 weeks. Our results demonstrate that Mi-CK is present in fetal human quadriceps musde and that the spedfic activity of Mi-CK increases during prenatal development from week 28 to 40 by a factor of about two. Generally, adult levels have not been reached at birth, indicating a further postnatal increase of the activity of the enzyme. The Mi-CK protein content also increases during prenatal development. These results suggest that in human skeletal muscle the expression and accumulation of Mi-CK starts at mid- gestation, later than is known to occur for cytosolic CK. Maturation of mitochondrial creatine kinase 113

Introduction

CK (E.C. 2.7.3.2) catalyses the reversible transphosphorylatíon reaction between phosphocreatine (PCr) and ATP:

PCr2" + MgADP"+(x).H+ « » MgATP2'+Cr

Five different isoenzymes of CK are currently known; MM-CK, MB-CK and BB-CK are three non-mitochondrial cytosolic isoenzymes, which are dimers comprising a combination of two different enzyme subunits, the M or muscle type and the В or brain type subunit [1]. Most of these cytosolic isoenzymes are soluble, but a small percentage may be compartmentalized subcellularly at sites of high ATP-tumover. In muscle, for example, MM-CK has been found both free and in association with the myofibrillar M-line [2]. In mature human skeletal muscle the relative percentages of the various cytoplasmic isoenzymes are: 96-100% MM-CK, 0-3% MB-CK and 0-1% BB-CK [3]. Kuby and co-workers [4] showed that skeletal muscle of an 11-12 week old human female fetus already contained BB-, MB- and MM-CK. During skeletal muscle development an isoenzyme transition from BB-CK to MM-CK takes place [5]. Immunohistochemically it has been demonstrated that in human skeletal muscle a switch from BB-CK to MM-CK takes place around week 8 of development [6]. To our knowledge no data is available about the prenatal development in skeletal muscle of the other type of CK; that is mitochondrial creatine kinase (Mi-CK) which in adult human quadriceps muscle amounts to 0.9% of total CK [Τ]. We studied muscle biopsies of preterm and term bom infants to examine if there are changes of Mi-CK expression and accumulation during pre- and postnatal development as has been shown for other enzymes involved in skeletal muscle energy metabolism [8].

Clinical data

Muscle tissue (m. quadriceps) was removed within 1 h post mortem from ten preterm bom infants (four females; six males) and two term bom infants (two males) with the informed consent of the parents. The gestational age varied between 28 and 36 weeks in the preterm bom infants; their weights at birth were 460-3350 g. The infants died within the first 6 days of life. 114 CHAPTER 4

One infant died from septic shock, all others died from cardiorespiratory failure. Six infants suffered from a congenital disorder, i.e., triploidy, hemia of diaphragm (two infants), congenital heart disease (aorta ascendens hypoplasia), hydrops fetalis, and single arteria umbilicalis. Anatomical examination of the brain showed signs of intracerebral haemorrhage in four infants. Three infants had normal anatomy. The patient with the triploidy showed an aplasia of the left n. olfactorius. No examination of the brain was performed in the other five patients. None of the infants showed clinical signs of a neuromuscular disorder.

Materials and methods

Preparation of muscle homogenates Muscle samples were either immediately frozen in liquid nitrogen and stored at -70 0C or immediately homogenized and stored at -70 0C. Muscle homogenates (10% w/v) were prepared according to Fischer et al. [9]. Cytochrome с oxidase activity was measured according to Cooperstein and Lazarow [10] and citrate synthase activity according to Srere [11]. Protein was determined by the method of Lowry et al. [12].

Electrophoresis of CK-isoenzymes Electrophoresis and detection of the CK isoenzymes, using the agargel overlay technique, was performed, with slight modifications, as described by Harm et al. [13]. The homogenate was five times diluted with a 30 mmol/L sodium phosphate buffer (pH 7.4) containing 0.2 mmol/L phenylmethane- sulfonyl fluoride, 0.2 mmol/L dithiothreitol and 0.05% Triton X-100 (v/v) to extract Mi-CK. After a 1 h incubation (room temperature) and centrifugation at40 000xg(30 min;40C) the supernatant (sup. 1) was used forelectrophoresis. Six microliters of sup. 1 was applied to a cellulose acetate strip for electrophoresis. Detection of the CK isoenzymes with nitro blue tetrazolium was performed in the presence of a final concentration of 10 μιηοΙ/L diadenosine pentaphosphate to inhibit adenylate kinase activity.

Measurement of mitochondrial CK Mi-CK activity was measured according to Smeitink et al. [7]. In brief, 50 \iL muscle homogenate was diluted five times with the above mentioned buffer. The 40 000g supernatant was used for measurement of total CK activity and further treatments. Total CK specific activity was measured (CBR-CK NAC- activated kit, Boehringer Mannheim, Germany, No. 475742) at 30 "С. Fifty microlitres DEAE-Cellulose slurry (Whatman DE 52) [14] added to 150 pL of Maturation of mitochondrial creatine kinase 115 the 40 000g supernatant, were used to bind BB-CK and MB-CK. After incubation (10 min; room temperature) and centrifugation at 8300xg (Imin), 100 μΐ, DEAE supernatant and 150 μι Sepharose (CNBr-activated Sepharose 4B, Pharmacia, Uppsala, Sweden) with anti-M-CK (polyclonal goat anti-M- CK antibody, Boehringer Mannheim) bound to it were mixed and incubated (10 min; room temperature). After centrifugation for 1 min at 2100xg in an Ultrafree-MC Filter Unit (Millipore, Bedford) to avoid contamination with Sepharose bound MM-CK, CK activity was measured in the presence and absence of inhibiting antibodies against the M-CK-chain (CBR-CK NAC- activated kit, Boehringer Mannheim) at 30 0C. The hexokinase/glucose-6-P dehydrogenase coupled production of NADPH was measured at 340 nm using a COBAS-Mira analyzer (Hoffmann-La Roche, Basel, Switzerland).

Analysis of the amounts of Mi-CK protein Sup. 1 of the same muscles as used for electrophoresis was analysed by immunochemical methods. Sodium dodecyl sulphate (SDS)-polyacrylamide gel electrophoresis was performed on a 10% gel in the Mini-protein two system (Biorad, Richmond, USA) using the buffer system as described by Laemmli [15]. The separated proteins (12.5 μg protein) were transferred to nitrocellulose membranes essentially as described by Towbin et al. [16]. Proteins were stained with Ponceau S (Sigma Chemical Co., St Louis, Missouri, USA). The nitrocellulose membranes were washed and incubated with a specific polyclonal rabbit anti-chicken Mi-CK antibody [17]. Antigen-antibody complexes were detected using alkaline phosphatase labelled porcine anti-rabbit serum as second ligand and nitroblue tetrazolium as substrate. Finally, the immunoblots were scanned with a laser densitometer (Ultrascan XL, Pharmada/LKB Uppsala, Sweden).

Statistics A multiple and a simple linear regression analysis were performed to examine for total CK and Mi-CK the dependency of gestational age and postnatal age. 116 CHAPTER 4

Results

Table 1 contains the various enzyme activities measured in thefirst si x days after birth compared with those measured in an older reference group of 19 persons [7]. Using a multiple linear regression analysis with gestational age and postnatal age as explanatory variables no dependency of the latter age could be established (P>0.10). Therefore we proceeded with simple linear regres­ sion based on gestational age. The specific activity of total CK showed a dear developmental increase with duration of pregnancy (P=0.001;Fig. 1). Electrophoresis of the CK isoenzymes showed that with increasing ges tational age the content of BB-CK decreased and that of MM-CK increased (Fig. 2; page 118). Mi-CK activity was not detected with this technique before week 29 of gestation. It should be noted that only one Mi-CKband could be detected by cellulose polyacetate electrophoresis (Fig. 2; page 118). The specific activity of Mi-CK also showed a developmental increase with increasing gestational age (P=0.002;Fig.l). Neither Mi-CK expressed relative to citrate synthase, a mitochondrial matrix enzyme, nor Mi-CK expressed relative to cytochrome с oxidase, an inner membrane bound mi tochondrial enzyme, showed a developmental increase

Table 1. Ranges of total CK, Mi-CK, q/lochrome с oxidase (COX) and citrate synthase (CS) enzyme activities in m. quadriceps of preterm bom infants and of a reference group

Infants Reference group (0-6 days) (2 months-60 years) (n = 12) (n = 19)

Total CK (U/mg protein) 10.3 - 33.0* 13.7 . 42.5 (U/mg wet weight) 0.65 - 4.03' 1.87 - 4.75

Mi-CK (mU/mg protein) 60.0 - 191 131 . 490 (mU/mg wet weight) 5.2 - 20.3 19.2 - 51.9 (mU/U total CK) 4.2 - 10.2 6.2 - 12.4 (U/U COX) 0.61 - 2.87 1.09 - 2.35 (U/U CS) 0.69 - 3.23 1.36 - 4.10

COX (mU/mg protein) 29 150 65 . 247 CS (mU/mg protein) 30 - 125 45 - 143

'n=ll Maturation of mi tochondrial creatine kinase 117

with the gestational age (P=0.15 and 0.34, respectively). Immunoblot experiments appeared to show an increase in Mi-CK pro­ tein with gestational progress (Fig.3). Despite the low number of measure­ ments linear regression analysis of the densitometric results approached statistical significance (P=0.08).

Total CK (π\U/m g protein) **KJ~

о о^. 30- о

20- О ^^Ό

-"'о О О 10- 1 28 ЗО 32 34 36 38 40 Gestational age (weeks)

Mi-CK (mU/mg protein) 200

150-

100

Gestational age (weeks)

Fig. 1. LinearregressUmamlysìsofthespeàficacHvityoftotalCKfupperpaneDandMi-CKdower panel) as a function of gestational age (weeks) 118 CHAPTER 4

BB

MB

.^^ммши-ІК ЛЕЧ^^ШЛлтЁт. •JH^fl^^^AA MM

Fig. 2. Electrophoresis pattern of CK isoenzymes stained for CK activity. Supematants of 10% m. quadriceps homogenates were applied after extraction with a hypotonic phosphate buffer in the presence of 0.05% Triton X-100 to release the CK isoenzymes. Gestational ages (weeks) in lanes 1-5 are: 26,28,355/7,37, and 40, respectively. Lane 6: control serum (CBR-NAC-activated kit). -> origin; +anode; - cathode.

8

Fig.3. ImmunochemicaldetectionofMi-CKproteinafterelectrophoresisofsupermtantlonSDS- PAGE followed by immunoblotting on nitrocellulose in 6 preterm and two term bom infants. Gestational ages (weeks) in lanes 1-8 are as follows: 26, 28, 285/7, 294/7, 355/7,37, 384/7 and 40. Markers of molecular weight (kD) are shown on the left and right side of the figure. On the left side from top to bottom:180,116,84,58,48.5 and 36.5. On the right side from top to bottom: 205,116,97.4,66,45 and 29. Maturation of mitochondrial creatine kinase 119

Discussion

The purpose of this investigation was to study the developmental expression and activity of total CK and Mi-CK during pre- and postnatal development in human skeletal muscle. A prenatal increase in both activities has been reported for sheep heart [18]. Data about the developmental changes of Mi- CK activity in skeletal muscle are of importance both for understanding the physiological role of this enzyme in energy metabolism and also for the interpretation of results from infants suspected to suffer from a mitochondrial myopathy. Mi-CK deficiency may explain the disturbed pyruvate oxidation rate found in a substantial percentage of infants and adults, for the decreased substrate oxidation rate can not be ascribed to a defect in the pyruvate dehydrogenase complex, tricarboxylic acid cycle or respiratory chain [8]. No data on prenatal developmental changes of Mi-CK in human skeletal muscle are available at this moment, due in part to difficulties in obtaining human muscle samples. The relatively lower contribution of Mi-CK to the total CK activity in skeletal muscle (mean ± 0.9%) in comparison with heart muscle (mean ± 9%) forms another problem in collecting adequate reference values for skeletal muscle. Recently, we developed a technique for measure­ ment of the Mi-CK activity after removal and inhibition of the surplus of other CK isoenzyme activity [7]. We were unable to evaluate Mi-CK and total CK specific activity related to age in days after birth because the postnatal lifespan hardly varied in our group of infants. To reach average adult levels (mean ± SD = 240 ± 88) the Mi-CK activity in mU per mg protein (118 ± 44) has to increase by a factor 2 after birth (Table 1.). A comparable increase of Mi-CK was found between 28th and 40th week of pregnancy (Fig. 1; page 117). Thus, on average, Mi-CK increases in m. quadriceps by a factor of about four from the 28th week of pregnancy to adulthood. Immunoblot studies of Mi-CK with a polyclonal rabbit anti-chicken Mi-CK antibody showed a prenatal increase in the Mi-CK protein content (Fig. 3). Thus, the increase in Mi-CK activity during gestation appears to be paralleled by a rise in Mi-CK protein. Cytochrome с oxidase, a membrane bound enzyme like Mi-CK, shows a similar prenatal development to Mi-CK. The prenatal increase in total CK activity can not be explained by the observed increase in Mi-CK activity since the proportion of Mi-CK relative to the total CK activity is small. The major contribution to this increase is due to a developmental increase of the MM-CK isoenzyme, the most abundant CK isoenzyme in skeletal muscle, as could be shown with the agargel overlay 120 CHAPTER 4

technique after cellulose polyacetate electrophoresis (Fig. 2.; page 118). In conclusion, these data provide an insight into the maturation of Mi-CK and so extend our knowledge of human skeletal muscle energy metabolism in neonates. They also provide a baseline for the correct interpretation of results from patients suspected to suffer from a mitochondrial myopathy.

Acknowledgements. The authors would like to thank their colleagues from the neonatal intensive care unit for obtaining the muscle biopsies, Ing. J. Mulder from the Department of Medical Statistics for processing the data. Miss С Kerkhof for technical assistance and Dr Theo Wallimann (ΕΤΗ, Zürich, Switzerland) for the kind gift of the polyclonal rabbit anti-chicken Mi-CK antibodies and for his critical review of this manuscript. Maturation of mitochondrial creatine kinase 121

References 10. Cooperstein SJ, Lazarow A. A micro- spectrophotometric method for the 1. Eppenberger HM, Eppenberger M, determination of cytochrome oxidase. J Biol Richterich R, Aebi H. The ontogeny of creatine Chem 1951; 189: 665-70. kinase isozymes. Dev Biol 1964; 10:1-16. 11. Srere PA. Citrate synthase, EC 4.1.3.7 citrate 2. Wa]limaimT,EppenbergerHM.Localization oxaloacetate-lyase (CoA-acetylating); in and function of M-line-bound creatine LöwensteinJM(ed):MethodsinEnzymology, kinase. M-band model and crea tine phosphate Academic Press, London, 1969; 13: 3-11. shuttle. In Shay JW (ed): Cell and muscle 12. LowTyOH,RosebroughNJ,FarrAL,Randall motility. Plenum Publishing Corp 1985; 6: RJ. Protein measurement with the Folin 239-85. phenol reagent. J Biol Chem 1951; 193:265-75 3. Jockers-Wretou E, Pfleiderer G. Quantitation 13. Harm K, Musolf K-M, SiragEldin E, Voigt of creatine kinase isoenzymes in human KD. Verteilungsmuster cytosolischer und tissueand sera by an immunological method. mi tochondrialer Kreatinkinase-isoenzyme in Clin Chim Acta 1975; 58:223-32. menschlicher Oberschenkel- und Augen­ 4. Kuby SA,KeutelHJ,OkabeK,etal. Isolation muskulatur. Ärztl Lab 1987; 33:259-66. of the human ATP creatine trans- 14. Nealon DA, Henderson AR. Separation of phosphorylases (creatine phosphokinases) creatine kinase isoenzymes in serum by ion- from tissues of patients with Duchenne cxhange column chromatography (Mercer's muscular dystrophy. J Biol Chem 1977; 252: method, modified to increase sensitivity). 8382-90. Clin Chem 1975; 21:392-7. 5. Goto I, Nagamine M, Katsuki S. Creatine 15. Laemmli UK. Clcavagcof structural proteins Phosphokinase isoenzymes in muscles. Arch during the assembly of the head Neurol 1969:20:422-29. bacteriophage T4. Nature 1970; 227: 680-85. 6. Wessels A, Vermeulen JLM, Virágh SZ, et al. 16. Towbin H, Staehelin T, Gordon J. Spatial distribution of "tissue-specific" Electrophoretic transfer of proteins from antigens in the developing human heart and Polyacrylamide gels to nitrocellulose sheets: skeletal muscle. 1. An immunohistochemical procedure and some applications. Proc Natl analysis of creatine kinase isoenzyme Acad Sci USA 1970; 76:43504. expression patterns. 17. Schlegel J, Zurbriggen B, Wegmann G, Wyss Anat Ree 1990; 228:163-76. M, Eppenberger HM, Wallimann T. Native 7. Smeitink J, Wevers R, Hulshof J, et al. A mitochondrial creatine kinase forms method for quantitative measurement of octameric structures. 1. Isolation of two mitochondrialcreatine kinase in human interconvertible mitochondrial creatine skeletal muscle. Ann Clin Biochem 1992; 29: kinase forms, dimeric and octameric 196-201. mitochondrial creatine kinase: charcter- 8. Smeitink JAM, Sengers RCA, Trijbels JMF, ization, localization and structure-function et al. Fatal neonatal cardiomyopathy relationships. Biol Chem 1988; 263:16942-53. associated with cataract and mitochondrial 18. Ingwall JS, Kramer MF, Woodman D, myopathy. Eur J Pediatr 1989; 148: 656-9. Friedman WF. Maturation of energy 9. Fischer JC, Ruitenbeek W, Stadhouders AM, metabolism in the lamb: changes in myosin et al. Investigation of mitochondrial ATPase and creatine kinase activities. Pedia tr metabolism in small human skeletal muscle Res 1981; 15:1128-33. biopsy specimens. Improvement of pre­ paration procedure. Clin Chim Acta 1985; 145:89-100. 122 Mitochondrial creatine kinase activity in patients with a disturbed energy generation in muscle mitochondria

J. Smeitink1, W. Ruitenbeek1, R. Sengers1, R. Wevers2, T. v. Lith1 and F. TrijbeU1

1 Institute of Paediatrics, University Hospital Nijmegen, The Netherlands. 2 Institute of Neurology, University Hospital Nijmegen, The Netherlands.

Chapter 5

Journal of Inherited Metabolic Disease: accepted 124 CHAPTERS

Summary

Eleven patients with an established disturbance in muscle mitochondria energy generation, in whom no defect in the pyruvate dehydrogenase complex or in the complexes of the respiratory chain could be detected, were investigated for a possible deficiency in the mitochondrial creatine kinase (Mi-CK) (EC 2.7.3.2). Moreover, four patients with a defect in one of the complexes of the respiratory chain were also investigated as to the Mi-CK activity. In none of the investigated patients a Mi-CK deficiency was found. Surprisingly enough two of the four patients with a defect in one of the respiratory chain complexes showed an enhanced activity of Mi-CK. It can be concluded that a Mi-CK deficiency is not frequently found as the primary defect in patients with a disturbance in the mitochondrial energy generation but more patients should be examined to allow a definite conclusion. Mitochondrial creatine kinase and mitochondrial myopathies 125

Introduction

A mitochondrial myopathy can be defined as a muscle disease characterized by structurally or numerically abnormal mitochondria and/or abnormally functioning mitochondria [1 ]. Clinically such disorders should be considered in all floppyinfant s and in those babies with perinatal problems like failure to thrive, seizures and cardiomyopathy [2]. The most important symptoms in older infants and adults are muscle weakness and exercise intolerance [2]. The biochemical approach for diagnostics of mitochondrial myopathies, as applied in our department, consists of a metabolic screening in body fluids (lactate, amino acids and organic acids), measurement of mitochondrial substrate oxidation rates and ATP plus PCr (phosphocreatine) production rates by muscle preparations. In this way a defect in the mitochondrial energy metabolism can be established [2]. A more exact localization of a defect is based on enzymatic procedures [e.g. 3], including measurement of the activity of the pyruvate dehydrogenase complex (PDHc) and of the com­ plexes of the respiratory chain. Applying the afore mentioned diagnostic approach regularly patients have been detected with a defect in the mitochondrial energy production in whom no specific enzyme defect could be established in the PDHc or in the respiratory chain. Clinical symptoms like fatigue, exercise intolerance and muscle weak­ ness with lactic acidosis and/or an increased lactate/pyruvate ratio, found in most patients suffering from a mitochondrial myopathy, can theoretically be caused by a Mi-CK deficiency. Therefore the diagnostic program was extended by measurement of the Mi-CK activity as a possible cause of the disturbance in the energy generation. Here the results of these investigations are presented.

Materials and methods

Substrate oxidation rates, ATP plus PCr production rates, PDHc activity, single enzyme activities of the respiratory chain and Mi-CK activity were measured in skeletal muscle samples (m. quadriceps) of 15 patients (11 males/4 females;age range: 1 month - 35 years) who were biopsied after informed consent. In all patients a muscle biopsy was performed because of the suspicion, based on clinical examination and laboratory investigations, that they suffered from a mitochondrial myopathy. The oxidation rates of the various radiolabelled substrates and production 126 CHAPTERS

rates of ATP plus PCr were measured in supernatants of fresh musde homogenates as described previously [3,4]. Cytochrome с oxidase (COX) and dtrate synthase (CS) activities were determined according to Cooperstein

and Lazarow [5] and Srere [6], respectively. NADH:02 oxidoreductase, NADHrQ, oxidoreductase and succinatexytochrome с oxidoreductase ac­

tivities were measured according to Fischer et al. [3]. In the NADH:02 oxidoreductase assay Mg2* was not added because omission results in higher activities in most supernatants. Measurement of total PDHc activity in muscle homogenate was per­ formed as described earlier [7]. Mitochondrial and total creatine kinase activity was determined according to Smeitink et al. [8]. Protein was deter­ mined by the method of Lowry et al. [9].

Results

Most important dinical data and the highest measured blood ladate concen­ trations are summarized in Table 1. All patients investigated showed a disturbance in the mitochondrial energy generation. The [l-14C]pyruvate oxidation rate in the presence of malate ranged in our patients from 0.27-2.7 nmol. (h. mU CS)"1 (control range: 3.6 - 7.5; n= 14) (Table 2). The ATP and PCr production rates with pyruvate plus malate as substrates ranged from 0-30 nmol.(h. mU CS)"1 (control range: 42 - 81; n= 16) (Table 2). Four patients exhibited a decreased pyruvate oxidation rate caused by a defidency in one of the complexes of the respira­ tory chain (Table 3; page 128). Patients 1, 3 and 4 showed a deficiency of complex I, measured as NADHiQj oxidoreductase, being 24,22 and 25 mU/ U CS, respectively (control range: 53 - 220; n=l 1). Patient 2 exhibited mainly a complex IV defidency, COX activity being 81 mU/U CS (control range: 300 - 2770; n=ll) (Table 3; page 128), but also the other parts of the respiratory

chain were reduced. The NADH:02 oxidoreductase activity, reflecting the overall capadty of the respiratory chain, was diminished in all 4 patients. In the other 11 patients no spedfic defed could be established. Measurement of the adivity of Mi-CK expressed on basis of protein, dtrate synthase and cytochrome с oxidase, the latter being mitochondrial reference enzymes, revealed no dearly decreased activity in 12 patients with a decreased pyruvate oxidation rate, although the values in patient 5,6,12 and 14 were only borderline normal (Table 4; page 129). In 2 of the 4 patients with an established deficiency in the respiratory chain the spedfic activity of Mi-CK was enhanced (patients 2 and 3, Table 4). Mi-CK expressed per unit total CK adivity was markedly high in the two latter patients. Mitochondrial creatine kinase and mitochondnal myopathies 127

Table 1. Some clinical data and highest measured blood lactate concentration (mmol/l) of the imestigated patients. patient 1 2 3 4 5 6 7 8 9 10 Π 12 13 14 15 muscle weakness + + . _ + + + + . . _ + + + + hypotonia + + - - + + + - + - - + + + + exercise intolerance - + + - + - + + - - - - + - + motor retardation + + ± - + + + + + - + + - + + failure to thrive + + - - + - - + - + + - + - + mental retardation + + - + + + + + - • + + • + + convulsions + - - + + + + - - - + - - - - ophthalmoplegia + + - - - + + - - - - - + - - cardiomyopathy ------+ - - - lactate 3.1 3.2 1.7 4.7 3.4 nm 5.0 5.0 1.4 4.4 7.1 4.8 1.9 3.0 3.8

+ = present; ± = mild; - = absent; nm = not measured

Table 2. Oxidation rates and production rates of ATP plus PCr by intact skeletal muscle mitochondria, measured in 600 g supernatant. patient [1-14C] pyruvate + malate oxidation* ATP+PCr production from pyruvate+malate*

1 1.59 14.2 2 0.27 0.97 3 0.82 6.2 4 1.61 14.3 5 1.52 17.1 6 2.67 11.2 7 2.65 26.2 8 2.50 24.0 9 1.87 12.6 10 1.62 n.d. 11 1.60 21.6 12 2.27 29.7 13 0.75 11.2 14 1.17 12.5 15 1.56 20.1

Controls η = 14 n=16 mean 5.52 63.9 SD 0.99 10.1 range 3.6-7.5 42.1 - 81.2

»Values are expressed in: nmol. (h. mU CS)"1; n.d. = not detectable 128 CHAPTERS

Table 3. Enzyme activities of skeletal muscle supernatant or homogenate (PDHc)

patient NADH:02 NADttQ, Succinate: Cytochrome с Citrate PDHc oxidoreductase oxidoreductase cytochrome с oxidase Synthase oxidoreductase

1 76 24 175 2189 146 nm 2 12 22 79 81 203 nm 3 42 22 433 2143 293 nm 4 98 25 431 1096 166 59 5 278 127 444 3628 63 34 6 347 193 400 1911 45 53 7 303 131 344 1688 32 88 8 412 184 488 1581 43 nm 9 503 205 385 2385 39 108 10 248 85 383 1353 34 73 11 394 124 275 1196 51 32 12 341 144 707 2390 41 59 13 run 207 253 2839 87 72 14 ran 46 218 1500 28 65 15 204 154 322 996 67 51

Controls n = 9 n = ll η = 35 n = ll η = 32 n = 25 mean 470 140 450 1490 91 81 SD 140 45 150 680 32 24 range 280-620 53 - 220 180-800 300 - 2770 48-162 34-122

Values are expressed in mU (U CS)"1, except for citrate synthase which is given in mU (mg protein)"1 nm = not measured

Discussion

Mi-CK, located at the outer surface of the inner mitochondrial membrane as well as free in the innermembrane space, phosphorylates creatine by dephosphorylation of ATP, thereby producing PCr plus ADP [10,11,12]. Recently a quantitative method for the measurement of Mi-CK activity in small amounts of human skeletal muscle has been reported [8]. This assay has now been added to the conventional diagnostic program of patients suspected to suffer from a disturbance in the mitochondrial energy genera­ tion because the latter could also be caused by a Mi-CK deficiency. A reduced activity of Mi-CK probably leads to a diminished PCr concentration and to an elevated intramitochondrial ATP/ ADP ratio which in turn inactivates PDHc mediated by a phosphorylation of the El subunit. Furthermore, in the presence of a physiological, intact coupling state of the oxidative phosphorylation, an ADP lack causes inhibition of the electron Mitochondrial creatine kinase and mitochondrial myopathies 129

Tab le 4. Mitochondrial creatine kinase activities patient Mi-CK (mU/mg protein) Mi-CK (U/U COX) Mi-CK(U/UCS) Mi-CK(mU/UCK)

1 292 1.32 1.65 12.1 2 857 34 2.53 53.3 3 1341 3.26 5.55 50.4 4 302 ??9 2.63 12.4 5 189 1.07 1.29 9.4 6 101 0.89 1.17 3.9 7 139 2.24 1.39 6.8 8 173 1.93 7 7? 7.0 9 378 2.22 3.97 6.8 10 138 1.39 1.89 8.4 11 184 1.56 1.62 9.7 12 102 1.04 1.23 6.0 13 366 2.83 4.42 17.0 14 95 0.95 1.39 5.2 15 150 2.72 1.76 6.3

Controls η = 19 η = 19 η = 19 η = 19 mean 240 1.66 2.70 8.6 SD 88 0.36 0.80 1.5 range 131-490 1.09-2.35 1.36-4.10 6.2 -12.4

flow through the respiratory chain. The resulting increase of the NADH/ NAD+ ratio also produces a feedback inhibition of the PDHc, leading to accumulation of pyruvate and lactate. As a result of the increased intramitochondrial NADH/NAD+ ratio the cytosolic NADH/NAD+ ratio is shifted likewise through the action of the malate-aspartate shuttle. The main clinical symptoms of patients with a Mi-CK deficiency are probably identical to those which has been observed in patients with mitochondrial myopathies, e.g. fatiguability, exercise intolerance and mus­ cle weakness. In all investigated patients a disordered pyruvate oxidation rate and ATP plus PCr production rate was found pointing to a defect in energy generation. Besides, most of them showed lactic acidemia and/or clinical symptoms as observed in patients with mitochondrial myopathies. In four of them the localization of the enzyme deficiency in the respiratory chain was known. The eleven patients with an unexplained decrease in activity of the oxidative phosphorylation showed a normal Mi-CK activity. Surprisingly, two of the four patients with a localized defect showed a clearly increased activity of Mi-CK. The contribution of Mi-CK to total CK activity was 130 CHAPTERS

enhanced in these patients, too (Table 4; page 129). Such a phenomenon has been found earlier for other mitochondrial enzymes, like dtrate synthase in complex I or complex IV deficiencies [e.g. 13,14]. This enhancement may be caused by a mechanism to compensate for a diminished ATP production by the mitochondria. From these results it can be concluded that Mi-CK deficiency is not frequently the primary defect in patients with a disturbance in the mitochondrial energy production but more patients should be investigated to allow a definite conclusion in this respect. Mitochondrial creatine kinase and mitochondrial myopathies 131

References mitochondrial creatine kinase in human skeletal muscle. Ann Clin Biochem 29:196- 1. SengeMRCA,StadhoudersAM,TrijbelsJMF 201. (1984). Mitochondrial myopathies: clinical, 9. LowryOH,RosebroughNJ,FarrAL,Randall morphological and biochemical aspects. Eur RJ (1951).Protein measurement with the Folin J Pediatri«: 192-207. phenol reagent. J Biol Chem 193: 265-275. 2. Trijbels JMF, Sengers RCA, Ruitenbeek W, 10. Schölte HR, Weyers PJ, Wit-Peeters EM Fischer JC, Bakkeren JAJM, Janssen AJM (1973). The localization of mitochondrial (1988). Disorders of the mitochondrial creatine kinase, and its use for the deter­ respiratory chain: clinical manifesta tionsand mination of the sideness of submitochondrial diagnostic approach. Eur J Ped 148: 92-97. particles. Biochim Biophys Acta 291:764-773. 3. Fischer JC, Ruitenbeek W, Cabreéis FJM, et 11. Wallimann T, Wyss M, Brdiczka D, Nicolay al. (1986). A mitochondrial encephalo- K, Eppenberger HM (1992). Intracellular myopathy: the first case with an established compartmentation, structure and function of defect at the level of coenzyme Q. Eur J creatine kinase isoenzymes in tissues with Pediatr 144:441-444. high and fluctuating energy demands: the 4. Bookelman H, Trijbels JMF, Sengers RCA, "phospho-creatìne circuit" for cellular energy Janssen AJM, Veerkamp JH, Stadhouders homeostasis. Biochem J 281:21-40. AM (1978). Pyruvate oxidation in rat and 12. Wyss M, Smeitink J, Wevers RA, Wallimann human skeletal muscle mitochondria. Τ (1992). Mitochondrial creatine kinase: a Biochem Med 20:395-403. key enzyme of aerobic energy metabolism. 5. Coopeistein SJ and Lazarow A (1951). A Biochim Biophys Acta (Reviews on microspectrophotometric method for the Bioenergetics): in press. determination of cytochrome oxidase. J Biol 13. Ruitenbeek W, Trijbels JMF, Fischer JC, Chem 189: 665-70. Sengers RCA, Janssen AJM, Kerkhof CMC 6. Srere PA (1969). Citrate synthase, EC 4.1.3.7 (1989). Mitochondrial myopathies: multiple citrate oxaloacetate lyase (CoA-acetylating). enzyme defects in the respiratory chain. J In: Löwenstein J.M. (ed) Methods in Inher Metab Dis 12 Suppl. 2: 352-354. Enzymology. Academic Press, London, 13: 14. Korenke G.-C, Bentlage HACM, Ruitenbeek 3-11. W., et al. (1990). Isolated and combined 7. Speri W, Ruitenbeek W, Kerkhof CMC, et al. deficiencies of NADH dehydrogenase (1990). Deficiency of the alpha and beta (complex I) in muscle tissue of children with subunits of pyruvate dehydrogenase (Ej) in mitochondrial myopathies. Eur J Ped 150: a patient with lactic acidosis and unexpected 104-108. sudden death. Eur J Pediatr 149:487-492. 8. Smeitink J, Wevers R, Hulshof J, et al. (1992). A method for quantitative measurement of 132 Mitochondrial creatine kinase containing crystals, creatine content and mitochondrial creatine kinase activity in chronic progressive external ophthalmoplegia

J. Smeitink1, A. Stadhouders2, R. Sengers1, W. Ruitenbeek1, R. Wevere3, H. ter Laak3 and F. Trijbels1

1 Institute of Paediatrics, University Hospital Nijmegen, The Netherlands. 2 Department of Cell Biology and Histology, University of Nijmegen, The Netherlands. 3 Institute of Neurology, University Hospital Nijmegen, The Netherlands.

Chapter 6

Neuromuscular Disorders: in press 134 CHAPTER 6

Summary

Mitochondrial crystals containing mitochondrial creatine kinase (Mi-CK) protein were described recently. From in vitro studies it has been suggested that alterations in creatine concentration are connected to the occurrence of these crystals. In the present study, free, phosphorylated and total creatine concen­ trations as well as Mi-CK activity were determined in muscle samples of six patients with chronic progressive external ophthalmoplegia (CPEO). Two of them showed Mi-CK containing mitochondrial crystals. The activity of Mi-CK was found to be dearly enhanced in those muscle samples in which mitochondrial crystals were present. No relationship was found between the concentration of total, free or phosphorylated creatine and the occurrence of mitochondrial crystals. An up to now unknown mechanism seems to cause the formation of Mi-CK contai­ ning crystals in human muscle mitochondria. Crystals and mitochondrial creatine kinase 135

Introduction

Mitochondrial disorders can be defined as diseases characterized by struc­ turally or numerically abnormal mitochondria and/or dysfunctioning mitochondria. These disorders are clinically heterogeneous in severity and organ involvement, ranging from distinct myopathic syndromes to complex multisystem disorders. Chronic progressive external ophthalmoplegia (CPEO), in which a slowly progressive paresis of the muscles that rotate the eyes in the orbit and elevate the eyelids is the predominant sign, is an appropiate example of the clinical heterogeneity observed in mitochondrial disorders [1]. For instance neurological, cardiac and renal impairment can occur in this syndrome [1]. Morphological and biochemical studies in CPEO have been performed mostly on skeletal muscle mainly based on practical considerations. How­ ever, similar changes as can be found in skeletal muscle have occasionally been observed in other tissues [2]. A structural skeletal muscle alteration in CPEO, also occurring in other mitochondrial disorders, is the occurrence of so-called "ragged red" fibres with increased succinate dehydrogenase activity. These fibres possess irregularly shaped peripheral aggregates of mitochondria [1]. Electron microscopy of these fibres reveals that the numerically increased mito­ chondria are often markedly enlarged and possess aberrant configurations of cristae. The mitochondrial matrix sometimes contains electrondense in­ clusions or shows vacuolization. The occurrence of highly ordered inclusions in the intermembrane space is probably the most striking structural abnor­ mality in patients with mitochondrial disorders and is regularly observed. These inclusions are real crystals, composed of protein material [3,4]. By now little is known about the underlying mechanism of mito­ chondrial crystal formation in human mitochondrial myopathies. In particu­ lar, the question whether the appearance of the mitochondrial crystals is causative to or a consequence of the disease cannot be answered yet. Recently it was observed that intramitochondrial crystals arise, looking similar as those in human mitochondrial myopathies, in rats fed with the creatine analogue and competitor ß-guanidinopropionic add (ß-GPA) as well as in creatine depleted cultured adult rat cardiomyocytes [5,6]. These crystals contained mitochondrial creatine kinase (Mi-CK) [6]. In this study we investigated the possible relation between the creatine content and mitochondrial crystal formation as was shown in cultured adult rat cardiomyocytes [6]. Furthermore, we studied whether in patients with such crystals, Mi-CK activity was altered compared with those patients in which no Mi-CK containing crystals were present. 136 CHAPTER 6

Materials and methods

Subjects From six patients (two females; four males) fulfilling the clinical criteria of CPEO [1] a muscle biopsy (m. quadriceps) was obtained, after informed consent, for diagnostic purposes: light and electron microscopic investiga­ tions and biochemical studies were also performed. The ages at which the biopsy was undertaken in patients 1 to 6 (Table 1) were, respectively: 25,15,8,29,20 and 27 yr. Patients 1,2,3 and 6 showed besides an ophthalmoplegia also exercise intolerance. Patient 1 furthermore showed some pigmentary degeneration of the retina and sensoneural hearing loss. The knee jerk and Achilles jerk in patient 3 were absent. Patient 6 also presented muscle weakness of the proximal shoulder and pelvis. Patien ts 4 and 5 only showed ophthalmoplegia.

Microscopic studies The muscle specimens were submitted for conventional histology, enzyme histochemistry and electron microscopy.

Immuno-electron microscopic studies For immuno-electron microscopy the specimens were fixed in 0.1 M phos­ phate-buffered 2% paraformaldehyde/0.1% glutaraldehyde solution (pH 7.3; 40C) for 1 h and after brief immersion periods in 50% and 70% ethanol, subsequently transferred via a mixture of LR Whi te (London Resin Company Ltd, Basingstoke, Hampshire, U.K.) - 70% ethanol to pure LR White as the embedding medium. Polymerization occurred at 50° С for 24-48 h. Thin LR White sections on copper grids were blocked for 30 min in 1% BSA in 0.1 M phosphate buffer (pH 7.3) and after three rinses with the same buffer incubated with the polyclonal rabbit anti-chicken Mi-CK antibody (dilution 1:200). After thorough rinsing the sections were incubated for 30 min with Protein A -10 nm gold complex (Janssens Pharmaceutica, Beerse, Belgium). Thereafter the sections were rinsed three times in the above buffer, followed by three rinses in distilled water and air dried. Further treatment was as for conventional electron microscopy.

Biochemical investigations Biochemical studies of muscle samples were performed as described by Fischer et al. [7]. The muscle specimen was cooled in ice-cold 10% SETH- medium [7] and transported directly to the laboratory. A 10% (w/v) homogenate was prepared in SETH-medium. For determination of free and phosphorylated creatine concentrations in skeletal muscle a 300 μΐ portion of Crystals and mitochondrial creatine kinase 137

Table 1. Concentration of free and phosphorylated creatine, specific activity of Mi-CK and Mi-CK expressed per milliunit citrate synthase in muscle of six patients with chronic progressive external ophthalmoplegia.

Patient Free Cr PCr Mi-CK Mi-CK (nmol/g ww) (μιηοΐ/g ww) (mU/mg prot.) (mU/mU CS)

1 13.1 6.2 680 5.0 2 12.0 8.8 539 5.1 3 23.3 3.7 285 3.0 4 13.8 7.0 267 3.1 5 18.0 7.2 177 2.8 6 0.5 13.4 246 5.1

Controls n=16 n=16 n=19 n=19 mean 14.9 8.5 240 2.7 SD 3.4 2.6 88 0.8 range 9.2-21.2 2.3 -13.1 131-490 1.4-4.1

Crystals were present in muscle mitochondria of patients 1 and 2. ww = wet weight; prot. = protein; CS = citrate synthase

the total homogenate was denaturated by 3 M HC104 and the supernatant

neutralized with 1 M KHC03. The assay was performed essentially as described in Bergmeyer using pyruvate kinase and hexokinase-mediated NAD(P)H coupled reactions [8]. Reference values were determined from muscle samples of patients suspect to suffer from a mitochondrial disorder in whom, however, despite thorough biochemical and microscopic examinations no abnormalities were found. Protein was determined by the method of Lowry et al. [9]. Mitochondrial creatine kinase activity in muscle was determined according to Smeitink et al. [10].

Results

Light microscopic studies revealed normal data except for the existence of ragged red fibres in two muscle samples (patients 1 and 3). In some fibres of patient 2 the sarcolemmal activity was accentuated but no real ragged red fibres were seen. Histochemistry of cytochrome с oxidase was performed in patients 1 and 2 (normal results) and patient 6, the latter showing many cytochrome с oxidase negative fibres. 138 CHAPTER 6

Electron microscopy of a muscle sample of patient 3 showed, especially in the ragged red fibres, giant mitochondria, up to 5 mm, with abberant configura­ tions of the cristae (tubular or angular appearance). Mitochondrial crystals (type I) were regularly observed by electron microscopy in patient 1. In patient 2 initial stages of crystal formation (type I) were frequently observed. In both muscle samples the crystals could be heavily labelled with the anti- Mi-CK antibody. Figures 1A and В show the mitochondrial crystals and immunolabelling of the crystals with Mi-CK antibodies of patient 1 and the initial stages of crystal formation and immunolabelling as observed in patient 2, respectively. Biochemical studies, i.e. measurement of the oxidation rates of pyruvate and malate, and the adenosine triphosphate plus phosphocreatine produc­ tion by intact mitochondria, revealed normal activities of the mitochondrial energy metabolism in all six investigated patients. Table 1 (page 137) summarizes the concentrations of free (non-phospho- rylated) and phosphorylated creatine, the activity of Mi-CK is expressed per Crystals and mitochondrial creatine kinase 139

Fig. 1. (A)Biopsyofm.quadricepsofpatient 1.Mitochondrialtypelcrystalspredominantly present in the outer mitochondrial compartment show distinct immunogold labelling with Mi-CK antibodies. Magnification; 64 OOOx.

(B) Biopsy of m. quadriceps of patient 2. Subsarcolemmal accumulation of mitochondria showing irregularly arrangement of cristae. Immunogold particles can be seen to be present above dilated intracristal compartment with moderate electron density. Arrowhead indicating initial stages of crystal formation. Magnification: 44 OOOx.

milligram protein as well as expressed per milliunit of citrate synthase. The free creatine content was lowered only in patient 6. Phosphorylated creatine was normal in all subjects. The specific activity of Mi-CK as well as the activity of Mi-CK expressed per milliunit of citrate synthase, a mitochondrial reference enzyme, was enhanced in patients 1 and 2. Patient 6 also showed an enhancement of the latter ratio but this was due to a low normal specific activity of citrate synthase (48; control range: 48-162 mU/mg protein)^ 140 CHAPTER б

Discussion

Gori et al. [5] described the occurrence of abnormal mitochondria in type 1 (slow twitch) skeletal muscles of rats after feeding them the creatine analogue and competitor ß-guanidinopropionic acid (ß-GPA), a drug known to cause total creatine depletion (free Cr + PCr) in muscle tissue. The compound inhibits creatine entry into muscle cells from blood plasma [11-13]. These abnormal mitochondria often contain, according to the authors, crystal-like inclusions which are very similar to those induced by ischemia or poisons and to those often observed in patients with a mitochondrial myopathy. They proposed that a decrease in the ATP buffering capacity, caused by a PCr depletion, could play a role in the pathogenesis of human myopathies. Eppenberger-Eberhardt et al. [6] observed that crystal-like mitochon­ drial inclusions were also formed in adult rat cardiomyocytes, when they were cultured in a creatine free medium or in a medium which contained ß-GPA. Immunolabelling of the crystal-like inclusions with antibodies against Mi-CK revealed a strong specific decoration of these mitochondrial inclusions [6]. The latter authors speculated that "accumulation of Mi-CK at the mitochondrial inclusions could be a compensatory phenomenon to a metabolic stress situation of the cultivated cardiomyocytes caused by low intracellular total creatine levels". Abnormal mitochondria and the occurrence of intramitochondrial crys­ tals are also found in human myopathies [4]. There are two distinct types of crystals, which can be distinguished by shape, size, pattern, unit-cell dimen­ sion, specific localization in the intermembrane space and their occurrence in different muscle fibre types. So-called type I crystals are usually present in the intracristal space, between two folds of the inner mitochondrial mem­ brane, whereas type II crystals are preferentially localized in the intermembrane space between outer and inner mitochondrial membrane [4]. Recently it has been shown by immunohistochemical studies, using rabbit anti-chicken Mi-CK antibodies, that both types of crystals in the mito­ chondria of human myopathic muscle consist mainly of Mi-CK [14,15]. The type Π crystals were always more heavily labelled than the type I crystals. Image processing of the type Π crystals revealed arrays of regularly ordered square-shaped particles with dimensions of 10 χ 10 nm and with a central cavity. Thus, the building blocks of the crystals appeared to be very reminis­ cent of isolated Mi-CK octamers [16]. From these observations it was sug­ gested that the mitochondrial inclusions seen under pathological conditions consist mainly of Mi-CK octamers [15]. On the analogy of the hypothesis for cultured adult cardiomyocytes [6] the question arose as to whether a disturbance in creatine metabolism could Crystals and mitochondrial creatine kinase 141 account for the formation of mitochondrial crystals containing Mi-CK. In our patients no relationship was found between either the free, phosphorylated or total creatine concentrations of the investigated muscle samples and the presence of mitochondrial crystals. Due to in vitro hydrolysis the in situ phosphorylated creatine concentration may be underestimated. Comparing the total creatine concentration in patients 1 and 2 with that in patients 3,4 and 5 indicated that crystal formation is not related to total creatine concentration because the total creatine concentration is normal in both groups of patients. Whether hydrolysis has caused the lowered phosphorylated creatine concen­ tration in patient 3 is not clear. The ATP concentration in this sample was quite normal. In patient 6 which showed, for unknown reasons, a clearly dimi­ nished free creatine content in the muscle, no crystals were found despite thorough electron microscopic study. So, there is no indication that a low free creatine concentration induces Mi-CK containing crystals. In two CPEO patients mitochondrial crystals (type I), showing distinct labelling with Mi-CK antibodies, were present. Our study shows that, at least in patients with CPEO, the free, phosphorylated and total muscle creatine concentrations likely do not cause crystal formation. Apparently, the condi­ tions resulting in the formation of Mi-CK containing crystals are different from those reported by Eppenberger-Eberhardt et al. [6] who performed their experimental work on adult rat cardiomyocytes in vitro. No data are available about the occurrence of mitochondrial crystals in the human heart muscle. In both patients the specific activities of Mi-CK as well as Mi-CK activities expressed per milliunit citrate synthase were dearly enhanced. This high activity of Mi-CK in both patients with electron microscopically observed Mi-CK labelled crystals is remarkable. Although the number of patients is low one can hypothesize about this phenomenon. It may reflect the presence of more than the normal number of Mi-CK molecules located at the outer surface of the mitochondrial irmermembrane or free in the intermembrane space, but can also be a consequence of solubilization of the crystals during the homogenization procedure and/ or the subsequent re-activation of the Mi-CK protein. In conclusion no relationship was found between human skeletal muscle creatine content and mitochondrial crystallization in CPEO patients. Mi-CK activity was dearly enhanced in those CPEO pa tients who were found to have mitochondrial crystals containing Mi-CK. Further studies are necessary to eluddate the nature and the mechanism of formation of these crystals.

Acknowledgements Theauthors would liketothankDr Theo Wallimann (ETH,Zürich,Switzerland)for his kind gift of the Mi-CK antibodies and Mr Theo van Lith for technical support. 142 CHAPTER б

References Folin phenol reagent. J Biol Chem 1951; 193: 265-275. 1. Hurko О, Johns D R, Rutledge S L, et al. 10. Smeitink J, Wevers R, Hulshof J, et al. A Heteroplasmy in chronic external ophthal­ method for quantitative measurement of moplegia: clinical and molecular obser­ mitochondrial creatine kinase in human vations. Pediatr Res 1990; 28:542-548. skeletal muscle. Ann Clin Biochem 1992; 29: 2. Schneck L, Adachi M, Briet Ρ, Wolintz A, 196-201. Volk В W. Ophthalmoplegia plus with U. Fitch CD, Jellinek M, Mueller E J. morphological and chemical studies of Experimental depletion of creatine and cerebellar and muscle tissue. J Neurol Sci phosphocreatine from skeletal muscle. J Biol 1973; 79:633-641. Chem 1974; 249:1060-1063. 3. Stadhouders A M, Sengers RCA. Morpho­ 12. Shields R P, Whitehair С К. Muscle creatine: logical observations in skeletal muscle from in vivodepletion by feeding beta-guanldino- patients with a mitochondrial myopathy. J propionic acid.Can J Biochem 1973;51:1046- Inherited MetabDisl987;10(Suppll):62-80. 1049. 4. Fanants G W, Hovmöller S, Stadhouders A 13. Shoubridge E A, Jeffry F M H, Keogh J M, M. Two types of mitochondrial crystals in Radda G К, Seymour A-M L. Creatine kinase diseased human skeletal muscle fibers. kinetics, ATP turnover and cardiac perfor­ Muscle Nerve 1988; 11:45-55. mance in hearts depleted of creatine with the 5. Gori Z, Tata V de, Pollera M, Bergamini E. substrateanaloguebeta-guanidinopropionic Mitochondrial myopathy in rats fed with a acid. Biochim Biophys Acta 1985; 847:25-32. diet containing beta-guanidine propionic 14. Stadhouders A, Jap P, Wallimann T. add, an inhibitor of creatine entry in muscle Biochemical natu re of mitochondrial crystals. cells. Br J Exp Path 1988; 69:639-650. J Neurol Sci 1990; 98: 304-305. 6. Eppenberger-Eberhardt M, Riesinger I, 15. Stadhouders A, Jap P, Winkler Η Ρ, Messerli M, et al. Adult rat cardiomyocytes Wallimann T. Pathologic intra-organelle cultured in crea tine-deficient medium display crystallization of mitochondrial creatine large mitochondria with paracrystalline kinase (Mi-CK)inmitochondrialmyopathies. inclusions, enriched for creatine kinase.JCell J Muscle Res Cell Motility 1992; 13:255A. Biol 1991; 113:289-302. 16. Schnyder T, Engel A, Lustig A, Wallimann 7. Fischer J C, Ruitenbeek W, Cabreéis F J M, et T. Native mitochondrial creatine kinase forms al. A mitochondrial encephalomyopathy: the octameric structures. 2. Characterization of first case with an established defect at the dimers and octamers by ul tracentri fugation, level of coenzyme Q. Eur J Pediatr 1986; 144: direct mass measurements by scanning 441-444. transmission electron microscopy, and image 8. Bergmeyer HU. Methods of enzymatic analysis of single mitochondrial creatine analyses (in German). New York: Weinheim, kinase octamers. J Biol Chem 1988; 263:16954- 1974. 16962. 9. Lowiy Ο H, Rosebrough N J, Fair A L, Randall R J. Protein measurement with the Considerations and perspectives

Chapter 7 144 CHAPTER? Considerations and perspectives 145

Mitochondrial creatine kinase and mitochondrial myopathies

In this thesis clinical, biochemical and morphological aspects of mitochon­ drial creatine kinase (Mi-CK) are described. Thus far this enzyme has hardly been studied in relation to human pathology and has not been investigated in relation to mitochondrial myopathies. The study was initiated as a consequence of the observation that, in a substantial number of patients with a mitochondrial myopathy, no specific enzyme deficiency could be detected despite an observed disturbance in the pyruvate oxidation rate. Since a deficiency of Mi-CK might also be respon­ sible for a disturbance in the pyruvate oxidation rate, a method to determine Mi-CK activity in human skeletal muscle samples was developed (chapter 3). So far, in 80 muscle samples, sent for diagnostic purposes, Mi-CK activity has been estimated and has been found normal. Twenty-nine of these muscle samples showed decreased substrate oxidation rates e causa ignota. Among the remaining specimens were localized enzyme defects of the respiratory chain as well as specimens with normal oxidation rates. To allow for a definite conclusion as to the occurrence of Mi-CK deficiencies in patients with a mitochondrial myopathy, a much larger number of Mi-CK activity determinations is required. For patients and their relatives a correct diagnosis of the biochemical defect responsible for the patient's complaints is important for genetic counselling and a therapeutic approach. Moreover, the clinical and bioche­ mical features of patients with a Mi-CK deficiency can extend our knowledge about the putative role of Mi-CK in cellular metabolism. An interesting feature of Mi-CK is that its activity in skeletal muscle of preterm born infants increases during pregnancy (chapter 4). The physio­ logical basis for this increase, which is also observed in animals, is not dear at present and needs further study. Regulation by hormones, as suggested for ubiquitous Mi-CK mRNA in rat uterus and placenta [1], can serve as a possible explanation of the observed developmental increase of Mi-CK in skeletal muscle. Another important result of the study of Mi-CK in relation to mitochondrial myopathies is the observation that crystals, observed in mitochondria in skeletal muscle fibers of some patients suffering from chronic progressive external ophthalmoplegia or from other mitochondrial encephalomyopathies, are highly enriched in Mi-CK [2,3; chapter 6]. At present the mechanism of this crystal formation is not clear. In cultured adult rat cardiomyocytes a relationship between creatine concentration and the occurrence of such crystals was observed [4]. Such a relationship could not be confirmed in the investigated patients with chronic progressive external ophthalmoplegia. 146 CHAPTER?

Disturbed pyruvate and malate oxidation rates e causa ignota: recommendations for further investigation

An important step in the diagnostic procedure for the detection of a mitochondrial myopathy is the morphological and biochemical investigation of a muscle biopsy. Before a muscle biopsy is taken, patients have to meet certain clinical criteria. Furthermore, analysis of body fluids, whether after a provocative test or not, must be indicative of a disturbance in the pyruvate oxidation. Despite such precautions, the cause of an in vitro observed disturbed pyruvate oxidation rate remains unknown in a substantial part of the investigated muscle samples of carefully selected patients. Probably the diagnostic program is still incomplete. Enzymes like Mi-CK, the adenine nucleotide translocator and the FQF,-ATPase are not generally included in the diagnostic program. Moreover, deficiencies of co-factors involved in energy generation as well as transport disturbances across the mitochondrial inner- membrane could account for the disturbed oxidation rates. Also, the search for factors with a secondary influence on the in situ oxidation of pyruvate and malate like: (i) the various causes of ischemia and anoxia, (ii) malnutrition, (iii) defects at the level of hormones and neurotransmitters, (iv) viral and bacterial infections, (v) poisoning by exogenous components and (vi) extra­ mitochondrial or extracellular inborn or acquired errors of metabolism has to be performed [5]. Determination of Mi-CK activity in a muscle sample of patients with an unexplained lactic acidosis is urged.

Molecular biology of mitochondrial creatine kinase

Separate nuclear genes encode two closely related, tissue-specific isoenzymes of Mi-CK [6; chap ter 2]. The method described in chapter 3 measures the total Mi-CK activity in skeletal muscle. The two Mi-CK isoenzymes (ubiquitous and sarcomeric) can be differentiated by using specific antibodies. However, thus far no such antibodies are available. A possibility to differentiate between ubiquitous and sarcomeric Mi-CK is the use of specific cDNA or genomic probes which recently became available [6]. With the use of a human ubiquitous Mi-CK cDNA probe the DNA of nine patients with a pyruvate oxidation rate e causa ignota was examined using the restriction fragment length polymorphism technique. No major rearrangements within the ubi­ quitous Mi-CK gene were observed using Southern blot analysis in these patients. Studies with a sarcomeric human Mi-CK cDNA probe are in progress. A reduced Mi-CK activity has very recently been found in patients with Considerations and perspectives 147 cardiomyopathy [7]. Up till now no structural rearrangements in the ubiqui­ tous Mi-CK gene were detected by us in six patients with a dilated cardio­ myopathy, three patients with a hypertrophic obstructive cardiomyopathy, four patients with a hypertrophic non-obstructive cardiomyopathy and two patients with a syndrome consisting of hyper-trophic cardiomyopathy, lactic acidosis, mitochondrial myopathy and congenital cataract (No. 21235 in the McKusick register, 1983; [8,9]). This study will be extended using a sarco- meric human Mi-CK cDNA probe. A recently developed genetic method i.e. gene targeting by homolo­ gous recombination in embryonic stem cells enables study of the conse­ quences of Mi-CK deficiency in an animal model. Such studies are in progress now (K. Steeghs, personal communication). Extrapolation of the results obtained by such experiments could be of great importance to the under­ standing of Mi-CK deficiencies and related human pathology.

Mi-CK and pathology

A review of the literature concerning Mi-CK and pathology is included in chapter 2. The existence of Mi-CK was already reported in 1964. The possible involvement of the enzyme with respect to human pathology has hardly been investigated. Based on the existence of separate Mi-CK genes, two possible clinical presentations of Mi-CK deficiency may be anticipated. Disturbances in the sarcomeric gene will mainly involve heart and/or skeletal muscle, whereas disturbances in the ubiquitous Mi-CK gene consequently will lead to a multisystem disorder. However, if a disturbance occurs at the mito­ chondrial membrane level or during protein assemblage, only a single tissue or organ may be affected. Up till now reduced Mi-CK activities in man are only found in cardio- myopathic patients. Further investigation into the involvement of a reduced Mi-CK activity in mitochondrial myopathies and other, up to now unex­ plained, pathological conditions in man is necessary. Most probably the enzyme is also present in other, as yet unexamined human cells or tissues like the retina as found in chicken [10]. A complete study of the tissuedistributio n of Mi-CK in man certainly will be helpful with respect to the understanding of physiological and patho-physiological processes.

In condusion, the work described in this thesis clarifies some aspects of Mi-CK in relation to mitochondrial myopathies and hopefully gives a push for further investigation of this enzyme and its putative role in human pathology. 148 CHAPTER?

References human mitochondrial creatine kinase isoenzymes. J. Biol. Chem. 265:6921-6927, 1. Payne RM, Haas RC, Strauss AW. Structural 1990. characterization and tissue-specific expres­ 7. Saks VA, Belikova YO, Kuznetsov AV, sion of the mRNAs encoding isoenzymes Khuchua ZA, Branishte TH, Semenovsky from two rat mitochondrial creatine kinase ML, Naumov VG. Phosphocreatine pathway genes. Biochim.Biophys. Acta 1089:352-361, for energy transport: ADP diffusion and 1991. cardiomyopathy. Am. J. Physiol. Suppl. 2. Stadhouders A, Jap P, Wallimann T. 261:30-38,1991. Biochemical natureof mi tochondrialcrystals. 8. Sengers RCA, Ter Haar BGA, Trijbels JMF, J. Neurol. Sci. 98:304-305,1990. Willems JH, Daniels O, Stadhouders AM. 3. Stadhouders A, Jap P, Winkler HP, Congenital cataract and mitochondrial Wallimann T. Pathologic intra-organelle myopathy of skeletal and heart muscle crystallization of mitochondrial creatine associated with lactic acidosis after exercise. kinase(Mi-CK) inmitochondrial myopathies. J. Pediatr. 86:873-880,1975. J. Muscle Res. Cell Motil. 13:255A, 1992. 9. Smeitink JAM, Sengers RCA, Trijbels JMF, 4. Eppenberger-Eberhardt M, Riesinger I, Ruitenbeek W, Daniels O, Stadhouders AM, Messerli M, Schwarb Ρ, Müller M, Kock-Jansen MJH. Fatal neonatal cardio­ Eppenberger HM, Wallimann T. Adult rat myopathy associated with cataract and cardiomyocytesculturedincreatine-defident mitochondrial myopathy. Eur. J. Pediatr. medium display large mitochondria with 148:656-659,1989. paracrystalline inclusions, enriched for 10. Wegmann G, Huber R, Zanolla E, creatine kinase. J. Cell. Biology 113:289-302, Eppenberger HM, Wallimann T. Differential 1991. expression and localization of brain-type and 5. Schölte HR. The biochemical basis of mito­ mitochondrial creatine kinase isoenzymes chondrial diseases. J. Bioenerg. Biomembr. during development of the chicken retina: 20:161-191,1988. Mi-CK as a marker for differentiation of 6. HaasRQStraussAW.Separatenudeargenes photoreceptor cells. Differentia tion46:77-87, encode sarcomere-specific and ubiquitous 1991. 149 Summary

Mitochondrial creatine kinase (Mi-CK; EC 2.7.3.2) plays an important role in the processing of high energy phosphates generated by the mitochondrial oxidative phosphorylation. The enzyme is localized at the outer surface of the inner mitochondrial membrane. Highest enzymatic activities of Mi-CK are present in organs with high and fluctuating energy demands like skeletal muscle, heart and brain. A deficiency of this enzyme can have important consequences as to the function of cellular metabolism. Up till now data on investigations of Mi-CK in man are rather limited. Investigations of Mi-CK in man are of importance to extend our knowledge about normal cellular processes as well as for the understanding of patho­ logical processes concerning energy metabolism. For example, investigation of Mi-CK in patients with mitochondrial myopathies can possibly contribute to the elucidation of a decreased mitochondrial pyruvate oxidation rate e causa ignota. A deficiency of Mi-CK may induce a dysfunctioning of the respiratory chain and thus a reduced pyruvate oxidation rate occurs. In the present study we investigated the activity of Mi-CK in normal and diseased human skeletal muscle.

In chapter 1 a short introduction to the topic and the subsequent chapters is given. In a subs tantial part of patients, suspected to suffer from a mitochondrial myopathy, detailed biochemical examination of skeletal muscle specimen shows a decreased pyruvate oxidation rate. Such a decrease may be caused by a deficiency of one of the enzymes or enzyme complexes of the pyruvate dehydrogenase complex, the dtric acid cycle or the respiratory chain. However, in approximately 30% of the patients with a decreased pyruvate oxidation rate this decrement can not be explained by a deficiency of one or more enzymes of the mitochondrial energy-generating system. This observa­ tion makes it necessary to extend the diagnostic program. A deficiency of Mi-CK could account for such a disturbance in the pyruvate oxidation. The aim of this study was to develop a method to determine Mi-CK activity in human skeletal muscle, to investigate the developmental aspects of the enzyme and to study Mi-CK with respect to mitochondrial myopathies.

Chapter 2 comprises the state of the art of Mi-CK. The aim of this chapter is to give a comprehensive overview on what is known about biochemical, physiological and pathological aspects of Mi-CK. Summary 151

In chapter 3 a method is described for the measurement of Mi-CK activity in human skeletal muscle. The method is based on the removal and specific inhibition of interfering CK isoenzymes. Reference values for Mi-CK activity are included in this chapter.

Chapter 4 contains the results of the study concerning the maturation of mitochondrial and other isoenzymes of CK in skeletal muscles of preterm born infants. This study was performed because earlier observations have shown an age-related development of various other enzymes involved in mitochondrial energy metabolism. As some of the patients with mito­ chondrial myopathies already show symptoms immediately after birth, knowledge about the existence of an age-dependency is necessary for correct interpretation of biochemical investigations in skeletal muscles of this age group. Evidence has been obtained for a prenatal increase in total CK and Mi-CK activity.

The results of the measurements of Mi-CK activities in patients with un­ known and known causes of defects of the pyruvate oxidation are given in chapter 5. Normal results were obtained with regard to the Mi-CK activity in the patients investigated so far with an unexplained disturbance of the pyruvate oxidation rate. In patients with a localized defect in the pyruvate oxidation pathway, more specifically patients with an isolated enzyme defect of the respiratory chain, no combination of a defect co-existing with Mi-CK was noticed.

In chapter 6 the results are presented of the study of mitochondrially localized crystals in patients with a chronic progressive external ophthalmoplegia. In this study we investigated the possible relationship between creatine content, Mi-CK activity and the presence of mitochondrial crystals. Recently it was shown that these crystals contain Mi-CK protein. Furthermore, it was shown that in these cases the Mi-CK activity was enhanced. No relationship was observed between the occurrence of such crystals and the creatine content of the skeletal muscle. This in contrast to animal and cell culture studies.

Finally, in chapter 7, considerations about the work performed in this thesis and recommendations for future investigation are given. In the twenty-nine investigated patients with a decreased substrate oxidation rate no Mi-CK deficiency was found. Molecular biological investigations concerning Mi-CK are in progress. Both, molecular biological and enzymological investigations of Mi-CK will extend our insight in the Mi-CK functioning in normal and pathological conditions. 152 Samenvatting

Samenvatting

Mitochondriaal creatine kinase (Mi-CK; EC 2.7.3.2) speelt een belangrijke rol bij deverwerkingvan het door demitochondriale oxydatievephosphorylering gevormde ATP. Het enzym bevindt zich aan de buitenzijde van de mito­ chondriale binnenmembraan. De hoogste Mi-CK activiteit is aanwezig in organen met hoge en fluctuerende energie-behoeften zoals skeletspier, hart en hersenen. Een deficiëntie van het Mi-CK kan belangrijke gevolgen hebben voor het functioneren van het celmetabolisme. Tot nu toe is er bij de mens nauwelijks onderzoek verricht naar Mi-CK. Onderzoek naar de diverse aspecten van dit enzym is echter van belang voor het vergroten van onze kennis omtrent het normale verloop van cellulaire processen en de consequenties van stoornissen hierin. Resultaten van onderzoek betreffende Mi-CK zouden een verklaring kunnen geven voor de bij patiënten met een mitochondriale myopathie veelvuldig gevonden, onverklaarde, verlaagde mitochondriale pyruvaat oxydatiesnelheid. Een deficiëntie van het Mi-CK kan namelijk het functioneren van de ademhalingsketen nadelig beïnvloeden waardoor een dergelijke verlaagde pyruvaat oxydatiesnelheid ontstaat. In dit onderzoek bes tudeerden wij de activiteit van Mi-CK in normale en pathologische humane skeletspier.

In hoofdstuk 1 worden de achtergronden van het onderwerp en het doel van de studie weergegeven. Een belangrijk aantal van de onderzochte patiënten van wie wordt aan­ genomen dat ze een mitochondriale myopathie hebben, laat bij biochemisch onderzoek van de skeletspier in vitro een verlaagde pyruvaat oxydatiesnel­ heid zien. Een dergelijke verlaging doe t een deficiëntie vermoeden in één van de enzymen of enzymcomplexen van het pyruvaat dehydrogenase complex, de citroenzuurcyclus of de ademhalingsketen. Bij ongeveer 30% van de patiënten met een verlaagde pyruvaat oxydatiesnelheid is deze afwijking echter niet te verklaren op basis van een defect in een of meer enzymen van het mitochondriale energie-producerende systeem. Deze bevinding maakt het noodzakelijk het diagnostische pakket uit te breiden. Op theoretische gronden kan een Mi-CK deficiëntie een verlaging in de oxydatiesnelheid van substraten tot gevolg hebben. Het doel van dit onderzoek is het ontwikkelen van een methode voor de bepaling van de activiteit van het Mi-CK in de humane skeletspier, studie verrichten naar de rijping van dit enzym en het bepalen van de Mi-CK Samenvatting 153

activiteit bij patiënten van wie aangenomen wordt dat ze een mitochondriale myopathie hebben.

Hoofdstuk 2 geeft een overzicht van de huidige kennis over met name de biochemische, fysiologische en pathologische aspecten van Mi-CK.

In hoofdstuk 3 wordt de door ons ontwikkelde methode voor het bepalen van de activiteit van Mi-CK in humaan spierweefsel beschreven. De methode is gebaseerd op verwijdering en specifieke remming van de storende CK isoenzymen. Referentiewaarden worden gegeven.

Hoofdstuk 4 geeft de resultaten weer van het onderzoek naar de rijping van Mi-CK en de andere CK isoenzymen in de skeletspier van prematuur geborenen. Deze studie werd verricht naar aanleiding van een eerdere observatie waaruit bleek dat enkele daartoe onderzochte enzymen, betrokken bij het mitochondriale energie-metabolisme, een aan leeftijd gerelateerde ontwikkeling doormaakten. Kennis over de ontwikkeling van deze enzymen is in de diagnostiek van patiënten met een mitochondriale aandoening van belang, omdat bij patiënten met dergelijke ziekten de symptomen zich al direct post partum kunnen manifesteren. Zowel de totale CK activiteit als de Mi-CK activiteit nemen toe met de duur van de zwangerschap.

Hoofdstuk 5 geeft de resultaten weer van de bepaling van de Mi-CK activiteit in patiënten met een te lage pyruvaat oxydatiesnelheid, veroorzaakt door een onbekend of een bekend defect in het energie-metabolisme. In de tot nu toe onderzochte patiënten met een onverklaarde verlaging in de substraat oxydatiesnelheid werd een normale activiteit geconstateerd voor Mi-CK. Er werd geen combinatie gevonden van een gelocaliseerd defect in de ademha­ lingsketen met een deficiëntie van de Mi-CK activiteit.

In hoofdstuk 6 worden de resultaten gepresenteerd van onderzoek dat verricht werd bij patiënten met een chronische progressieve externe Ophthalmoplegie. Bij deze patiënten werd gezocht naar een relatie tussen de Mi-CK activiteit, de creatine concentratie van de skeletspier en het al dan niet aanwezig zijn van mitochondriaal gelocaliseerde kristallen. Onlangs werd aangetoond dat dergelijke kristallen Mi-CK bevatten. Wij toonden aan dat de Mi-CK activiteit verhoogd was in de skeletspier van deze patiënten zonder dat er sprake was van een gestoorde substraat oxydatie. Er werd geen relatie gevonden tussen de aanwezigheid van dergelijke kristallen en de creatine concentratie van de skeletspier. Dit laatste in tegenstelling tot bevindingen bij dier en celkweek studies. 154 Samenvatting

In hoofdstuk 7 wordt een beschouwing gegeven over de uitgevoerde studie en worden aanbevelingen gedaan voor verder onderzoek. Bij deonderzochte patiënten met een verlaagde pyruvaat oxydatiesnelheid is tot nu toe geen deficiëntie van het Mi-CK gevonden. Er is een start gemaakt met moleculair biologisch onderzoek van Mi-CK. Zowel moleculair biologisch als enzymo- logisch onderzoek van Mi-CK zal onze kennis over het functioneren van dit enzym onder normale en pathologische condities doen toenemen. 155

Woorden van dank

Op deze plaats wil ik iedereen bedanken die op enigerlei wijze heeft bijgedragen aan de totstandkoming van dit proefschrift. Allereerst gaat mijn dank uit naar alle patiënten en hun ouders zonder wiens medewerking dit proefschrift nooit geschreven had kunnen worden. Vervolgens wil ik mijn beide promotores bedanken. Prof. Dr. R.C.A. Sengers. Rob, jij hebt mij de mogelijkheid gegeven dit onderzoek op te zetten en uit te voeren. Het idee onderzoek te doen naar het mitochondriaal creatine kinase in de humane skeletspier was van jou afkom­ stig. Bedankt dat je ondanks je drukke werkzaamheden de tijd vond om nadrukkelijk aanwezig te zijn bij de discussies. Prof. Dr. J.M.F. Trijbels. Frans, gedurende het gehele onderzoek heb ik me steeds van jouw steun verzekerd kunnen voelen. Door je rustige en deskundige manier van optreden was je een grote steun voor mij. Dr. W. Ruitenbeek. Wim, de vele discussies die ik met je mocht voeren over dit onderzoek en over mitochondriale myopathieën, zijn voor mij van zeer groot belang geweest. Ik dank je voor de vele uren die je hiervoor hebt vrijgemaakt en voor het feit dat ik je op de meest onmogelijke uren van de dag om advies mocht vragen. Jouw prettige en deskundige medewerking hebben mij de spirit gegeven om door te gaan. Dr. RA. Wevers. Ron, jij hebt een buitengewoon belangrijke bijdrage geleverd aan de ontwikkeling van de methode zoals beschreven in hoofdstuk 3. Dankzij jouw steun en medewerking hierbij heeft het onderzoek nieuwe impulsen gekregen. Bedankt voor je aanwezigheid en inbreng in de discussies over hoofdstuk 2 tijdens het verblijf in Zwitserland. Drs. J. Hulshof. Jos, bedankt voor de tijd die je tijdens jouw stage klinische chemie hebt gestoken in de vervolmaking van de methode om het mitochondriaal creatine kinase te bepalen. Τ. ν. Lith. Theo, jij was verantwoordelijk voor het grootste deel van de experimenten die in dit proefschrift zijn beschreven. Vanaf het begin van dit onderzoek heb ik op een bijzonder prettige manier met je samengewerkt. De dank je voor het enthousiasme en de accuratesse waarmee je gewerkt hebt. Dr. W. Speri (Universitäts Kinderklinik, Innsbruck, Oostenrijk). Wolfgang, bedankt voor al je bijdra gen aan het onderzoek o ver de leef tijdsafhankelijkheid van enzymen betrokken bij het mitochondriale energie-metabolisme, voor de prettige samenwerking en de belangstelling voor dit onderzoek. A.J.M. Janssen, C. Kerkhof, M. Brückwilder, Α. Friebel, Drs. H. Bentlage en alle andere medewerkers van het laboratorium metabole ziekten van het AZN (Academisch Ziekenhuis Nijmegen). Anton, Christine, Marloes, Ans en Herman, ik dank jullie voor de bijdragen aan dit onderzoek. De prettige samenwerking en de niet aflatende belangstelling voor het onderzoek waren voor mij van groot belang. Dr. B. van Oost. Bernard (Laboratorium antropogenetica, AZN), bedankt voor je aanvullend onderwijs in de moleculaire biologie dat ik van je mocht krijgen in de trein naar München. Bedankt, dat ik op jouw laboratorium een deel van de experimenten mocht doen. De medewerkers van het laboratorium Antropogenetica van het AZN dank ik voor de prettige werksfeer, de belangstelling en de vele nuttige discussies. Drs. K. Steeghs (Afdeling celbiologie; hoofd: Prof. Dr. B. Wieringa). Karin, bedankt voor de ontwikkeling van de humane Mi-CK ubiquitous cDNA probe en de discussies hieromtrent. Prof. Dr. A.M. Stadhouders. Bedankt voor de discussies rond het onderzoek van de mitochondriale kristallen en de bijzonder fraaie electronen- microscopische opnamen die dit heeft opgeleverd. Dr. H. ter Laak en H. Kuppen (Laboratorium experimentele morfologie, Instituut voor Neurologie, AZN). Henk en Hennie, bedankt voor al het morfologisch spieronderzoek wat jullie in het kader van dit onderzoek hebben verricht. Dr. R. de Graaf en Ing. J. Mulder (Mathematisch-statistische advies­ afdeling). Ruurd en Jan, bedankt voor de verwerking van de resultaten. De wil jullie bij deze nogmaals dank zeggen voor de snelheid waarmee jullie in een bepalende fase van het onderzoek hebben gewerkt. Dr. R. Lippens (Afdeling kinderoncologie, AZN). Rob, als mentor tijdens mijn opleiding tot kinderarts, heb je door je nuchtere kijk op dingen en je adviezen met name in de beginfase van dit onderzoek, een bijzondere bijdrage geleverd aan de totstandkoming van dit proefschrift. Prof. Dr. B.T. Poll-The (Hoofd klinische afdeling metabole ziekten, Wilhelmina Kinderziekenhuis, Utrecht). Bwee-Tien, bedankt voor de tijd die ik, ondanks alle drukke klinische werkzaamheden, in het onderzoek mocht steken. Prof. Dr. R. Berger, Dr. M. Duran, Dr. L. Dorland (Laboratorium metabole ziekten, Wilhelmina Kinderziekenhuis, Utrecht). Ruud, Ries en Bert, bedankt voor jullie interesse in mijn onderzoek. Ruud bedankt voor je adviezen in de laatste fase van het onderzoek.

Dr. T. Wallimann, Institute for Cell Biology, Swiss Federal Institute of Technology, Zürich, Switzerland (Head: Prof. Dr. H. Eppenberger). Dear Theo, I am very grateful for your intensive participation during the prepara­ tion of several of the manuscripts described in this thesis. I am still impressed by the rapidity with which you corrected the manuscripts. Your knowledge concerning mitochondrial creatine kinase and your enthousiasm in this field 157 of investigation gave me the necessary stimuli to continue. Thank you very much for your gift of the Mi-CK antibodies. I would like to thank you, and your colleagues of the institute, very much for the high standard of work you perform in the field of mitochondrial creatine kinase. I hope that our cooperation may continue.

Dr. M. Wyss (Institute for Cell Biology, Swiss Federal Institute of Technology, Zürich, Switzerland). Dear Markus, I would like to thank you very much for your cooperation and for the important contribution which you made to the preparation of the manuscript as described in chapter 2. I enjoyed our concerted action and the enthousiasm with which it was fulfilled enor­ mously. I enjoyed all the discussions we had about Mi-CK by letter, fax, phone as well during the stay at your institute. Hopefully we can continueour cooperation.

Woorden van dank gaan ook uit naar degenen die het mij mogelijk maakten de opleiding kindergeneeskunde te volgen. Prof. Dr. G.B.A. Stoelinga, Prof. Dr. R.C.A. Sengers enProf. Dr. J.W. Stoop. Mijn collegae arts-assistenten en kinderartsen voor de belangstelling die jullie toonden tijdens het onderzoek.

Mijn ouders dank ik voor de kansen die zij mij hebben gegeven. Pa en ma, zonder jullie was dit alles nooit zover gekomen. De hoop dat ik vanaf nu wat meer tijd voor jullie kan reserveren.

Willemien, zonder jou had ik dit nooit volbracht. Ik wist mij steeds van jouw steun verzekerd. Je hebt me steeds de vrijheid gelaten die nodig was voor het schrijven en je voelde perfect aan wanneer het tijd werd om een pauze in te lassen. Keer op keer heb je geluisterd naar de inhoudelijke en politieke zaken rondom dit onderzoek. Dit alles is vast niet altijd eenvoudig geweest. Het is duidelijk dat je een grote invloed hebt gehad op de totstandkoming van dit proefschrift. Mark, over een aantal jaren zal ik je uitleggen waarom mama in jouw eerste levensmaanden zoveel meer aandacht aan je heeft gegeven dan ik. We halen dit zeker in.

Rest mij nog waardering uit te spreken voor het enthousiasme, de belang­ stelling en de hulp van diegenen die ik vergeten ben met name te noemen. 158 159

Curriculum Vitae

De auteur van dit proefschrift werd geboren te Arnhem op 21 juni 1956. Vanaf 1973 volgde hij de opleiding tot klinisch chemisch analist aan de school voor laboratorium personeel te Eindhoven, waar hij in 1976 het diploma HBO-A klinische chemie behaalde. Na het vervullen van zijn militaire dienstplicht begon hij in 1978 zijn studie in de Geneeskunde aan de Katholieke Universiteit te Nijmegen. In 1983 slaagde hij voor zijn doctoraal examen, in mei 1985 werd het artsexamen met succes afgelegd. Van augustus 1985 tot augustus 1986 was hij werkzaam als agnio op de Afdeling Kindergeneeskunde van het St. Radboudziekenhuis te Nijmegen, waar hij vanaf augustus 1986 tot april 1991 in opleiding was tot kinderarts (Opleider: Prof. Dr. G.B.A. Stoelinga). Op 1 mei 1991 werd hij als kinderarts ingeschreven in het specialisten-register. Van januari 1988 tot augustus 1991 was hij voorzitter van de Junior Afdeling van de Nederlandse Vereniging voor Kindergeneeskunde. Uit hoofde van deze functie was hij van januari 1988 tot november 1990 lid van het Bestuur, het Concilium Paediatricum en de Plenaire Visitatie Commissie van de Nederlandse Vereniging voor Kindergeneeskunde. Vanaf april 1991 tot heden is hij verbonden aan de klinische afdeling metabole ziekten (Hoofd: Prof. Dr. B.T. Poll-The) van het Wilhelmina Kinderziekenhuis te Utrecht. De auteur is gehuwd met Willemien E.M. Berkers en samen hebben zij een fantastische zoon. Mark. 160 161

List of publications

1. J. Smeitink, J. de Vries, R. v. Empelen. Motorisch onderzoek bij zes patiënten met het Prune Belly syndroom. Ned Τ Fysiother (1985) 95:118-120. 2. J. Smeitink, В. Hamel, R. ν. Empelen, J. de Vries, L Monnens. Het Prune Belly syndroom: ervaringen bij 9 patiënten. Ned Tijdschr Geneeskd (1987) 12:489-493. 3. J. Smeitink, M. Verreusel, C. Schroder, R. Lippens. Nephrotoxicity associated with ifosfamide. Eur J Pediatr (1988) 148:164-166. 4. J. v. Proosdij, J. Smeitink. Sarcoidose. Tijdschr Kindergeneesk (1989) 57:49-53. 5. J. Hoekx, J. Smeitink, H. Brunner, L. Monnens. Het syndroom van De Barsy. Tijdschr Kindergeneesk (1989) 57:53-58. 6. J. Smeitink, R. Sengers, F. Trijbels, W. Ruitenbeek, O.Daniëls, A.Stadhouders, M Kock- Jansen. Fatal neonatal cardiomyopathy associated with cataract and mitochondrial myopathy. Eur J Pediatr (1989) 148: 656-659. 7. J. Smeitink, R. Wevers, J. Hulshof, R. Sengers, W. Ruitenbeek, С. Korenke, F. Trijbels. A quantitative method for measurement of mitochondrial creatine kinase in human skeletal muscle. J Neurol Sci (1990) 98 (Suppl.): 250. 8. J. Smeitink, R. Wevers, J. Hulshof, W. Ruitenbeek, T. v. Lith, R. Sengers, F. Trijbels, C. Korenke, T. Wallimann. A method for quantitative measurement of mitochondrial creatine kinase in human skeletal muscle. Ann Clin Biochem (1992) 29:196-201. 9. J. Smeitink, F. Beemer, M. Espeel, R. Donckerwolcke, C.Jacobs, R. Wanders, R. Schutgens, R. Roels, M. Duran, L. Dorland, R. Berger, B.T. Poll-The. Bone dysplasia associated with phytanic acid accumulation and deficient plasmalogen synthesis: a peroxisomal entity amenable to plasmapheresis. J Inher Metab Dis (1992):in press. 10. J. Smeitink, W. Ruitenbeek,T. v. Lith,R. Sengers, F.Trijbels, R.Wevers, W.Speri,R. de Graaf. Maturation of mitochondrial and other isoenzymes of creatine kinase in skeletal muscle of preterm bom infants. Ann Clin Biochem (1992) 29: 302-306. 11. W. Speri, R. Sengers, F. Trijbels, W. Ruitenbeek, W. Doesburg, J. Smeitink, L. Kollée, H. Boon. Enzyme activities of the mitochondrial energy generating system in skeletal muscle tissue of preterm and full term neonates. Ann Clin Biochem (1992): in press. 12. G. v. Eekeren, Α. Stadhouders J. Smei tink, R. Sengers. A retrospective study of patients with the hereditary syndrome of congenital cataract, mitochondrial myopathy of heart and skeletal muscle and lactic acidosis. Eur J Pediatr (1992): in press. 13. J. Smeitink, A. Stadhouders, R. Sengers, W. Ruitenbeek, R. Wevers, H. ter Laak, F. Trijbels. Mitochondrial creatine kinase containing crystals, creatine content and mitochondrial creatine kinase activity in chronic progressi ve external ophthalmoplegia. Neuromuscular Disorders (1992): in press. 14. J. Smeitink, W. Ruitenbeek, R. Sengers, R. Wevers, T. v. Lith, F. Trijbels. Mitochondrial creatine kinase activity in patients with a disturbed energy generation in muscle mitochondria. J Inher Met Dis (1992): accepted. 15. M. Wyss, J. Smeitink, R. Wevers, T. Wallimann. Mitochondrial creatine kinase: a key enzyme of aerobicenergy metabolism. BiochimBiophys Acta (Reviews on Bioenergetics) (1992): in press. Dit proefschrift werd vormgegeven met behulp van Aldus PageMaker 4.0, CorelDRAW2.0 en Microsoft Windows 3.1. Deze programmatuur draaide op een personal computer met een 80486 microprocessor. Het document werd geprint op een Postscript laserprinter en cameraready aan de drukker afgeleverd. Deze heeft een verkleiningsfactor van 80% toegepast. De gebruikte lettertypen zijn Palatino, Palatino Bold en Palatino Italie. Voor de standaard tekst werd een 12 punts letter, voor tabellen een 10 punts letter en voor koppen een 18 of 14 punts letter gebruikt. De interlinie bedroeg 135% van de puntgrootfë. / Vormgeving: Eric Tetteroo о»лк Figuren hoofdstuk 1 en 2: Markus Wyss Ontwerp omslag: Willemien Smeitink-Berkers 163

Stellingen

Behorend bij het proefschrift

MITOCHONDRIAL CREATINE KINASE some clinical, biochemical and morphological aspects

Nijmegen, 6 oktober 1992 Jan A.M. Smeitink I Bij de interpretatie van de resultaten van onderzoek naar de activiteit van het mitochondriaal creatine kinase in de humane skeletspier dient rekening gehouden te worden met de leef tijdsafhankelijkeheid van de activiteit van dit enzym. Dit proefschrift.

II De aanwezigheid van intramitochondriaal gelocaliseerde kristallen in de Ьхяпапе skeletspier bij patiënten met een chronische progressieve externe Ophthalmoplegie lijkt niet gerelateerd te zijn aan de creatine concentratie van de skeletspier. Dit proefschrift.

III Omdat mitochondriaal creatine kinase gecodeerd wordt door twee verschillende genen zal een deficiëntie van dit enzym tot een wisselende klinische expressie aanleiding geven. Dit proefschrift.

IV Bij een patiënt met spierklachten in combinatie met een sterk verhoogd plasma CK dient voor er invasieve diagnostiek plaatsvindt de schildklierfunctie onderzocht te worden. Soomers, persoonlijke mededeling.

V Me tabool onderzoek gericht op stoornissen in de mitochondriale ß-oxidatie van vetzuren bij kinderen die onverwacht overlijden is een diagnostische uitdaging met vele valkuilen. Duran et al. Pediatrics (1986) 78:1052-1057.

VI Een stoornis in de assemblage van peroxisomen, leidend tot het Zellweger syndroom, kan meer oorzaken hebben. Shimozawa et al. Science (1992) 255:1132-1134. Gärtner et al. Nature genetics (1992) 1:16-23.

VII De vorming van peptide-bindingen bij de eiwitsynthese wordt met grote waarschijnlijkheid gekatalyseerd door RNA-moleculen in het ribosoom. Nolier et al. Science (1992) 256:1416-1419. VIH De door velen gehanteerde ondergrens vannormoglycaemie dient naar boven bijgesteld te worden.

IX Specifiek laboratorium onderzoek ten behoeve van de diagnostiek van erfelijke stofwisselingsziekten dient een artikel 18 voorziening te worden.

X Zowel tijdens als na behandeling met ifosfamide dient men bedacht te zijn op het ontstaan van een glomerulo-tubulaire stoornis. Smeitink et al. Eur. J. Pediatr. (1988) 148:164-166.

XI De recentelijke waarneming van een nieuw syndroom bestaande uit hydrops f oetalis op basis van congenitale anemie met afwezige duimen, neu tropenie en immuundef iciëntie bevestigt het concept van een ontogenetisch veld dat zowel de radiale straal als het hematopoëtisch en immunologisch systeem omvat. Semmekrot et al. Am. J. Med. Gen. (1992) 42:736-740.

XII De kwaliteit van de kindergeneeskundige zorg dient onafhankelijk te zijn van de assertiviteit van de ouders.

XIII De voorgeschreven afstand van 7 cm tussen de spijlen van een kinderbed is, zeker voor zuigelingen, te groot.

XIV Milieubelangen zijn niet altijd tegenstrijdig aan economische belangen getuige de snel groeiende markt voor milieu-adviesbureau's.

XV De wens om zelf te bouwen is voor velen een ideaal. Als het ideaal verwezenlijkt kan worden ontbreekt het velen de moed om durf te tonen in de vormgeving van dit ideaal.

XVI De door de resultaten van aerodynamisch onderzoek gewijzigde vormgeving van moderne automobielen heeft het adequaat anticiperen in het verkeer een stuk moeilijker gemaakt.

• i-^

• • • • I

,J^"

.J- ^

[ Wnyg \л*ш!!£ *22r ЩЕ£*2£\ 1 ítía ^Ф1 IBS .:

^^H ^^1 ¿imf% I ^jpjra^ ^^Н , :ЭЭБ| ^^^H ¡Ж *ЗП 1 [\ 7#^ИС? 1 ---'" I ; ^^H ^^F' ^pf м^І І^н В*ДЗ 1

1 ^И Sy^?* 1 ^^^^1

^^^^i • | ^^^^^^и ^^^H 1 [ [ ^^н

^^^Bf^v^^^u в [ 1 • ^^^^н ^у Jjjl 1 ι B^^L I 1 ' i 1 ^н^нн^н ^н | • ^^^^^^H [ ^^H ^^^H ^^^Н

» ^^^H I ^^^^H \ \ '^S^tf'

ί,

I •

ί • ^^^^B r

• •

• ^ж^в ^^^H ^^^H ^^^^н*^9^ ^яК ^Н ¡fa'

[ • • •" · ¿i»qj .ifaflJ ^^^H g ДЦЯ ^В ' ^^H ^^^^H ^Cn^^T ^^^B ^^^Н l МИШ ДУІ^ДЮДИКДДИК jgM

;ί'; Ж i^S 9Я

1 ^^^B • ^^^^^^^| ' ^І

• . ^^^Н ^^^Н 3^ί^Φ ДіР щш ^^^^Н ^^^^^| ^ч^Р ^^^H ______^_ _^^__^__ ^^ __ •Ш ШшШ ЯІШЯЯШ ш ІШ я ш тштшшш1 Ш ^н