Oxidative metabolism in neuronal and non-neuronal mitochondria' JAMES C . K .

LAr2

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 12/02/14 For personal use only.

Department sf Phamsaceuticcal Sciences, College of P h c a m q , and Center for Toxicology Research, Idaho State University* Campus BOX $334, Pscarel/o, ID 832m-OW, U-S.A. Received September 23, 1991 LAI, J. C. K. 1992. Oxidative metaboIism in neuronal and non-neuronal mitochondria. Can. J. Physiol. Pharmacol. 78: S130-S137. Methodological advances have allowed the isolation of two populations of synaptic (SM and SM2) and two populations of nonsymptic (A and B) mitochondria from rat forebrain. All four populations of brain mitochondria are metabolically active and essentially free from nonmitmhondrial contaminants. They (SM, SM2, A, and B) can oxidize a variety of substrates; the best substrate is pymvate. With pymvate as the substrate, the respiratory control ratios (i.e., state 3Istate 4) in all four populations are routinely > 6 . Results from numerous enzyme activity measurements provide strong support for the hypothesis that brain mitochondria are very heterogeneous with respect to their enzyme contents and that the enzymatic activities in a particular population s f mitochondria, be they synaptic or nonsynaptic, differ from those in another population of mitochondria derived from either the same or another brain region. The major methodological advances in brain mitochondrial isolation greatly facilitate metabolic studies. For example, we have demonstrated that the K+ stimulation of brain mitochondrial pymvate oxidation is mediated through a K+-induced elevation of the activation state of the pymvate dehydrogenase complex and the K+ stimulation of the flux through the pymvate dehydrogenase complex. Our previous and ongoing studies using primary cultures of hypothalamic neurons and astrocytes are consistent with the proposal that brain cells are heterogeneous with respect to heir capabilities in energy metabolism. I can envisage that in the not-so-distant future, one could adapt these preparations of cells as the starting material for the isolation of mitochondria of h s w n cellular origin for metabolic studies. Key words: heterogeneity of brain mitochondria, regulation of interndiary metabolism. EAH,J. C. K O 1992. Oxidative metabolism in neuronal and nsn-neuronal mitochondria. Can. J. Physioi. Pharmacol. 70 : S130-S137. Les dCveloppements mCthodologiques ont permis d'isoler deux populations de mitochondries synaptiques (MS et MS2) et deux populations de mitochondries non synaptiques (A et BB)dans le cerveau antQieur du rat. Les quatre populatisns de mitochondries c6rCbrales sont mktaboliquement actives et essentiellement libres Be contaminants non mitochondriaux, Elles (MS, MSZ, W et B) peuvent oxyder une variCtd de substrats, le plus adCquat Ctant le pynvate. En prCsence de ce substrat, les rapports du contrdle respiratoire (c.4-d. Ctat 3lCtat 4) dans les quatre populations sont wonmadement >6. Les resultats de nombreuses maesures d7activitCemymatique apportent un solide soutien a l'hypothkse que les mitochondries cdr6brales ssnt trks hCtCrogknes par rapport i leurs contenus enzymatiques et que les activitks enzymatiques d'une population particulikre de mitschondries, synaptiques ou now, diffkrent de celles d'une autre population de mitochondries, si@& dans la mtme rCgion ou dans une autre rCgion du cerveau. Les dCveloppements m6thodologiques majeurs au wivmu de l'isolation des mitochondries cCrCbrales facilitent considCrablement les Ctudes mktaboliques. Winsi, nous avsns d6rnontrC que la stimulation par le Kfde l'oxydation du pymvate mitochondrial est mCdiCe par une augmentation, induite par le K + , de l'Ctat d9activation du complexe pymvate dCshydrogCnase et par la stimulation par le K9 du flux dam le complexe pymvate d6shydrog6na8e. Nos Cmdes anterieures et actuelles sur des cultures primaires d'astrwytes et de neurones hypothalamiques sont em accord avec l'hypowse que les cellules cCrCbrales presentent des caractkristiques hCtCrog&nes au niveau du m6tabolisme $nergCtique. Je peux envisager que, dans un prsche avenir, il sera possible d'adapter ces prCparations de cellules de fagon 8 ce qu9ellessewrent d'outils de base pour 19isolationde mitochondries d'srigine cellulaire connue pour des 6tudes m6taboliques. Mots elks : hCtCrogCnCitC des mitochondries cCrCbrales, rkgulation du mktabolisme intermediaire. [Tradnit par la rCdactiom]

Intrduction From electron-microscopis studies, it is evident that brain mitochondria are very heterogeneous with respect to ( i ) their size, shape, and cristal structure, and (it) their cellular (i.e., neurond and (or) gllid) and subcellular (i.e., cell body, dendritic, nerve ending) localization (see Raine 1981 for diseussion and refs.). What is the metabolic significance of the diverse forms and distribution of brain mitochondria? We have focused our research efforts on this question for some time. 'This paper was presented at the satellite symposium of the International Brain Research Organization meeting held August 10- 14, 1991, University of Saskatchewan, Saskatosn, Sask., Canada, entitled Ions, Water, and Energy in Brain Cells, and has undergone the Journal's usual peer review. 'Address for correspondence: Department of Pharmaceutical Sciences, College of Pharmacy, Idaho State University, Campus Box $334, Pocatello, ID 83289-rn9, U.S.A. Printed in Can& i Imprime au Canada

When we first started our research program, prior measdological problems spurred US OW to seek me&odologicd advances that have led to new ways of looking at the heterogeneity of mitochondria and allowed one to address fundamental issues concerning the regulation of brain mitochondrial metabolism.

What were the methodological problems associated with brain dtoebondriaig studies? There were two major problems associated with studies using inadequate brain mitocheswdrial isolation methodology, The problems were concerned with (i) the purity of the isolated mitochondria and (it) their; metabolic integrity. For example, differential centrifugation procedures originally designed for isolating mitochondria from peripheral tissues (e.g., liver and kidney) yielded crude preparations of brain mitochondria that retained some metabolic activity but were invariably and variously contaminated with myelin fragments.

TABLE1. Substrate-dependent oxygen uptake (natm oxygen min-' . tng-' protein; 1 atm = 101.325 kPa) by forebrain synaptic (SM and SM2) and nonsynaptic (A and B) mitochondria

SM2

SM

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 12/02/14 For personal use only.

[K'l (mRI)

State 3

RGW

State 3

B

A

RCW

State 3

WCR

State 3

WCR

NOTE: Values are means i SD of 3 -8 separate experiments and are compiled from the data of Lai and Clark (1979. 1989) and from unpublished observations s f J. C . K. Lni. The respiratory control ratios (WCRs) are the ratios of state 3 rate (Leo,respiratory rate measured in the presence of saturlating levels of substrateds) and ADP) to state 4 rate (i.e., basal rate, measured in the presence of saturating levels of suhstrate(s) but limiting level of ADB) (not shown). *p < 0.05 (determined by ANOVA using the BAsac program by J. @. K. Eai) versus corresponding values in the other mitochondrial populations.

synaptosornes, and rnicrosomes (see Clark and NicUas 1978, 1984; Lai and Clark 1989 for discussion and refs*).The procedures for sucrose density gradient centrifugation pioneered by De Robertis (De Wobertis and de Lores Arnaiz 1969) and Whittaker whittaker 1969, 1984) and their co-workers entailed long centrifugation times and prolonged exposure to very hypertonic sucrose. Although relatively pure, the mitochondrid preparations so obtained were metabolically inactive and poorly coupled (see Clark and NicMas 1970, 1984 Lai and Clark 1989 for discussion and refs.). Consequently, there was an obvious need for isolation procedures that yield brain mitochondria that meet both the criteria of purity and metabolic integrity. New and improved methodology for isolation of brain mitochondria Clark and Nicklas (1978) were the first to design a discontinuous Ficoll density gradient procedure whereby nonsynaptic mitochondria (fraction B) could be routinely isolated from eight rat forebrains. Subsequently, Lai and Clark (see Lai and Clark 1979, 1989 for discussion and refs.) developed procedures whereby synaptic mitochondria (i. e., mitochondria derived from lysed synaptosomes) could be routinely prepared in yields large enough for metabolic studies. One of the procedures (Lai et al. 1977; Lai and Clark 1979, 1989) allows the isolation of two populations of synaptic (fractions SM and SM2) and one population of nonsynaptic (fraction A) mitochondria derived from the same homogenate originated from four forebrains. The buoyant densities (as determined by sucrose density gradient centrihgation in a zonal rotor) of synaptic mitochondria (buoyant density of SM less than that of SM2) are significantly lower than those of nonsynaptic mitochondria (A and B) (9. C. K. Lai, unpublished data). Furthermore, rate-zonal separations reveal that the A fraction of nonsynaptic mitochondria consists of heterogeneous subpopulwtions whereas the B fraction of nonsynaptic mitochondria is comparatively homogeneous (3. C . K. Lai, unpublished observations). All four populations of mitochondria (i.e., SM, SM2, A, and B) meet both the criteria of purity and metabolic integrity (as indicated by high respiratory control ratios; Table 1) (see Eai and Clark 1979, 1989 for a detailed discussion). Additionally, in contrast with the Clark and Nicklas (1970) procedure, our method (kai and Clark 1979, 1989) has the advantages of (i) improved mitochondrial yield (e.g., using four instead of eight forebrains) and (ki) simultaneous isolation of synaptic and nonsynaptic mitochondria from the same hsmogenate.

Thus, the Lai and Clark (1979, 1989) method allows one to address metabolic issues pertaining to both synaptic and nonsynaptic mitochondria in a single fractionation experiment. More recently, the Lai and Clark (1979, 1989) method has been successfully adapted for the routine isolation of synaptic and nonsynaptic mitochondria from discrete brain regions (9. C. K. Lai, unpublished data; also see below).

Heterogeneity of forebrain mitochondria Oxygen uptake in I00 nzM K+ medium All the mitochondrial populations can oxidize a variety of substrates, including pyruvate and glutamate (Table 1; also see Lai and Clark 1979, 1989). However, pyruvate is by far the best substrate. Among the four mitochondrial populations (SM, SM2, A, and B), pyruvate-dependent oxygen uptake is lowest in the SM2 fraction of synaptic mitochondria (Table 1) (g < 0.05), in accord with the distribution of the activities of pyruvate dehydrogenase complex in these fractions (Table 2). Glutamate-supported oxygen uptake is lower in synaptic mitochondria (SM and SM2), especially SM2, compared with nonsynaptic mitochondria (A and B). These data correspond quite well to the distribution of the activities of aspartate arninotransferase (AAT; Table 2). Tricarboxylic acid cycle and relc~tedenzymes Irrespective of the mitochondrial fractions, the activities of the pyruvate dehydrogenase (PDHC) and a-ketoglutarate dehydrogenase (KGDHC) complexes are the lowest among all the tricarboxylic acid (TCA) cycle enzymes investigated (Table 2). Thus, since both complexes catalyze essentially irreversible reaction steps, their low activities strongly suggest that they are rate limiting in regulating the flux into (in the case of PDHC) and through (in the case sf KGDHC) the TCA cycle. The activities of these complexes are significantly lower ( p < 0.05) in the SM2 fraction of synaptic mitochondria compared with those in the other three populations of mitochondria (Table 2). Activities of citrate synthase (Cit Syn) are generally lower ( p < 0.05) in synaptic mitochondria than in nonsynaptic mitochondria (Table 2). NAD-linked isocitrate dehydrogenase (NAD-ICDH), hrnarase (Fum), and AAT activities are lower in the SM2 fraction of synaptic mitochondria than in the other three populations (SM, A, and B Table 2) (g < 0.05). On the other hand, the activities of NADP-linked ICDH (NADPICDH) and NAD- and NABP-linked glutamate dehydrogenases (NAD-GDH and NADP-GDH) are higher ( p < 8.85)

CAN. J. PHYSIOL. PHARMACOL. VOL. 90, 1992

TABLE 2. Enzymatic activities (mU/mg protein) in forebrain synaptic (SM and SM2) and nonsynaptic (A and 8)mitochondria

SM2

SM

Mean k SB

n

Mean f S D

B

A

n

Mean f SD

n

Mean

+ SD

n

PBHC

Cit Syn NAD-HCDH

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 12/02/14 For personal use only.

NADP-ICDH

KGBHC Fum NAD-MDH NAB-GDH

NADP-GDH AAT NOTE:Data are compiled from Lai and Clark 41979, 1989) and unpublished observations of 9. @. K. Lai. BDHC, pymvate dehydrogenase complex; @it Syn, sitrate synthase; NAD- and NADP-ICDH, NAB- and NADP-dependent isocitrate dehydrogenase; KGDHC, a-ketoglutarate dehydrogenase complex: Fum, fumarase; NAD-MDH, NAD-dependent malate dehydrogenase; NAD- and NABP-GDH, NAD- and NADPdependent glutamate dehydrogenase; AAT, aspartate arninotransferase. *p < 00.5 (determined by ANOVA using the BAsac program by S. C. K. Lai) versus corresponding values in the other mitochondria1 populations.

TABLE3. Enzymatic activities (mU/mg protein) in synaptic (SM2) and nonsynapeic (A) mitochondria isolated from cerebral cortex, striaturn, and pons and medulla Cerebral cortex

PDHC

Cit Syn NAB-ICDH

Fum

20f3 429+83* 50f I1 134f 38*

81 + 6 1210f189

198f 36 494 f33

Striaturn

14k 1 6Wf 131 42&4 200 f48

61 +9* 1057f 141" 103f27" 409 f53"

Pons and medulla

13f4* 605+40 37i-5" 234 f28

65+8 1188f197 135f19

656 f90

NOTE:Valeies are the means i -SD sf 3 or more experiments and are compiled from the data of Lesng et al. (1984) and unpublished observations of J. @. K. Lai. For more details see Table 2.

in this fraction (SM2) than in the other mitochondrial populations (Table 2). The activities of NAD-linked malate dehydrogenase (NAD-MDH) in the various mitochondrial populations do not markedly differ (Table 2).

Synaptic ( S W ) and nsnsynaptic (A) mitochondria isolated from brain regions As alluded to earlier, the structural and biochemical heterogeneity of brain suggests that brain mitochondria, in addition to being heterogeneous at the whole brain level, may d s o be heterogeneous at the regional level. Indeed, Sokoloff and eolleagues (see Sokoloff 1983 and refs. therein) have clearly demonstrated that brain regiond glucose utilization is not uniform. We have studied the activities of more than 15 enzymes of intermediary metabolism, including glycolytic and TCA cycle enzymes, and found their brain regiond distribution to be uneven (Leong et al. 19811; Lai et al. 1982, 1984, 1985a; Sheu et al. 1984; Lai and Cooper 1986). Other workers have also reported similar findings (Butterworth and Gigukre 11984; Leong and Clark 119Ma, 1984b, 1 9 8 4 ~Malloch ; et al. 1986; Martin et al. 11987). Although the regional distributions s f the activities of various enzymes we have investigated differ somewhat in the rank orders, there is the consensus that enzymatic activities are generally higher in phylogenetically younger brain regions (such as the cerebral cortex) than in

phylogenetically older regions (such as the pons and medulla). If this generalization holds for brain regional distribution s f enzymatic activities, then one would ask if the various synaptic and nonsynaptic mitochondria isolated from discrete regions may also have enzymatic characteristics that are consistent with the generalization. Since SM2 and A consistently show more interesting metabolic differences in comparison with SM (Tables l and 21, and the yields of mitchondrid protein are higher in A and SM2 than in SM (Lai et all. 1977; Lai and Clark 1979, 1989), we have focused our investigations on SM2 and A isolated from cerebral cortex (CC), striaturn (ST), and pons and medulla (PM). Irrespective of the brain region, enzymatic activities are generally lswer in the SM2 fraction of synaptic mitochondria than in the A fraction of nonsynaptic mitochondria (Table 3). (The activities of the same enzymes in the SM2 fraction isolated from the forebrain are also lswer than the corresponding enzymatic activities in the A fraction isolated from forebrain (compare data in Tables 2 and 31.) Activities of PDHC and NAD-ICBH are lowest in the SM2 fraction from PM among the SM2 fractions from the three regions investigated. Activities of these enzymes are lowest in the A fraction from ST among the A fractions from these regions (Table 3). Activities of Cit Syn and Fum are lowest in the SM2 fraction from CC among the SM2 fractions from the three regions inves-

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 12/02/14 For personal use only.

tigated whereas these enzymatic activities are lowest in the A fraction from ST among the A fractions studied (Table 3). Our data on brain regional distribution of enzymes (Table 3; also see Lai et al. I981 ; Leong et al. 1984), taken together, strongly suggest that the enzymatic activities in a particular population of mitochondria, be they synaptic or nonsynaptic, differ from those in another population of mitochondria derived from either the same or another brain region.

Use of isdated brain mitcpshondria to study problems concerned with metabolic regulation The major methodological advances in brain mitochondrial isolation allow one to carry out a variety of metabolic studies using functionally intact organelles. To illustrate how metabolically active preparations of brain mitochondria can be usefully exploited to address issues of regulation of brain metabolism, several aspects of pyruvate metabolism in Ronsynaptic mitochondria (fraction A) isolated from the CC are discussed. hpsrtance of pyruvase oxida~iveme~b&)&ism via pyruvate dehydrogerease complex The importance is twofold (see Lai and Clark 1989 for discussion). (i) Because the adult brain is dependent on aerobic glucose oxidation for energy, the regulation of pyruvate oxidative metabolism (in mitochondria) plays a key role in the channelling of glycolytic carbon into the TCA cycle. Thus, a better understanding of the regulation of the PDWC is important in this regard. (ii) Glucose-derived pyruvate is converted to acetyl CoA, which provides carbon for TCA cycle flux and for syntheses of neurotransmitters such as acetylcholine (ACh), glutamate (Glu) , aspartate (Asp), y-aminobutyric acid (GABA) and, quite possibly, several other neurotransmitter amino acids related to the TCA cycle. (However, it is important to note that for net syntheses of TCA cycle and related intermedithe presence of functional "anaates (e.g., Glu, Asp, GABA), plerotic" enzyme(s) (such as pyruvate carboxylase and (or) malic enzyme) is required.) ?

Mechanisms that regulate brain pyruv~te dehydmgemase comp&ex Two kinds of mechanisms h o w n to control peripheral tissue PDWC (Weed 1981; Randle 1981) may also regulate brain PDHC. (i) The PDHC is regulated by a pkosphorylation (inactivation) - dephosphorylation (activation) cycle, mediated by PDH, kinase and PDHb phosphatase. respectively. We were the first to design assay procedures for characterizing these two regulatory enzymes in brain (Sheu et al. 1983, 1984; Lai et al. 1985a). (ii) Brain PDHC is also regulated by product inhibition by acetyl CoA and NADW (see Lai and Sheu 1987; Lai et al. 1989 for additional discussion and refs.).

K+ stimulation of brain pyruvate metabolism It has k e n h o w n for some time that elevation of external (i.e., extracellular) Kf concentrations will lead to the stimulation of oxidative metabolism of several brain tissue preparations including slices and synaptosomes (see Hertz 1965, 1990; Lai and Sheu 1985 for discussion). One mechanism that could account, at least in part, for this K+ stimulation of aerobic oxidation of glucose by brain tissue is through the stimulation of pyruvate metabolism. If this hypothesis is tenable, one would anticipate that raising the external Kf level should result in a stimulation of pyruvate-supported oxygen uptake by mitochondria. Indeed, that is exactly what we found in all the brain

mitochondrial preparations we have studied so far, irrespective of the brain regional or subcellular eompartrnent(s) from which the mitochondria were isolated. For example, when the medium K+ level is raised from 5 to 100 mM, the state 3 rates with pymvate as the substrate are increased by 100- 150% (Table 1). Thus, the Kf stimulation appears to be a universal phenomenon. (Nevertheless, one must caution against overinterpreting the stimulation by K+ of pyruvate-supported oxygen uptake by brain mitochondria in vitro. Since the b6norma199 intracellular Kf level in brain is high, one might argue that brain mitochondrial respiration is normally stimulated and this K" stimulation might not be directly related to the K+ stimulation detected in whole cells or in slices.) The K+ stimulation of pyruvate-supported oxygen uptake by brain mitochondria could be due to at %easttwo separate mechanisms, since oxygen uptake is an indirect measure of metabolic flux. (i) The stimulation can be mediated through a direct cationic effect of the K+ upon the mitochondrial respiratory chain or selected compcments thereof. Indeed, some years ago, Clark and NicMas (Clark and NicHas 1970; Niclclas et al. 1971) did find such a direct ionic effect (i.e., the first mechanism; see above) when they investigated this aspect, using nonsynaptic mitochondria isolated from the forebrain. However, since they observed that the direct ionic effect on the respiratory chain could, at best, only account for 20% of the overall Kf stimulation of respiration, they speculated that in addition to the direct effect on the respiratory chain, K + probably exerts a major stirnulatory effect on the PDHC itself (i.e., the second mechanism: see below). (ii) K can exert a direct effect on the PDHC itself and thereby accelerate the flux through the PBHC. When we recently set out to address the second hypothesis (i.e., the direct K+ effect on brain PDHC), we had to overcome several methodological difficulties. Only two key methsdological advances are mentioned here. (i) Because the activation state of the PDHC can influence the flux through the PDHC, we developed a method to "freeze" the activation state of PDHC in situ such that its value under different metabolic states can be determined. (ii) Since pymvate-dependent oxygen uptake is not a direct indicator of flux through the PDHC, we had to devise a more direct method of assessing PDHC flux. We designed a procedure to measure I4CO2 production (which is a direct measure of PDHC from fl-'"]pyruvate flux). Once the metkodologicd difficulties had been resolved, we were ready to address two questions: Does K+ stimulate flux through the PDHC in brain mitochondria (as determined by [l-MC]pymvatedecarboxylation)? Does K+ exert a stimulatory effect on the activation state of the PDHC in brain mitochondria? In good agreement with the results concerning pymvatesupported oxygen uptake by brain mitochondria (Table 1; also see Lai and Clark 19'79, 1989), in respiratory state transition from state 4 (i .e., basal state) to state 3 (i.e., ADP stimulated), the PBHC flux in cerebrocorticd nonsynaptic mitschondria is increased by B 12% in 5 mM Kf medium and by 282% in 100 mM K+ medium (Table 4). More importantly, our data provide direct evidence that K+ does stimulate flux through PDHC. For example, raising the external K + level from 5 to BOO mM results in increases in PDHC flux of 36 and 9476, respectively, under state 4 and state 3 conditions (Table 4). In parallel with the K+ stimdation of flux through PDHC, there are corresponding elevations of the PDHC activation +

CAN. J. PHYSIOE. PHARMACBL. VOL. 76, 1992

S134

TABLE4. Pymvate dehydrogenase complex (PDHC) activation state and flux in cerebrocortical nomsynaptic mitochondrka in low (5 mM) and high (108 mM) K+ media

[Kf1 (mh.l)i

+ 2.5 mM malate 5 mM pymvate + 2.5 mM rnalate + 1 mM ADP (state 3) 5 mkl pyruvate + 2.5 mRI malate (state 4) 5 mM pymvate + 2.5 mM rnalate +

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 12/02/14 For personal use only.

5 mM pymvate (state 4)

I mM ADP (state 3)

PDHC activation state (76 active)*

PBHC flux (nanol 1"0, min- I . mg- protein)

Mean f SD

41

Mean f SB

9.1 +1.8t

pf

100

78f 8 t

9

I00

54 i-

7

5

52210

5

6.7k1.4

7

5

30+ 10$

7

14.2+3.7$

7

2'7.5 f3.4-1-$

12 16

NOTE:Data are compiled from Lai and Sheu (1985) and unpublished observations of J. C. K. Lai. For other details see Table 2. "Expressed as percentage of the total PDHC activity (1 16 20 rnBJImg protein) for 8 experiments. Tps < 0.05 versus corresponding values in the 5 mM K + medium. $ p < 0.05 versus corresponding state 4 values.

+

TABLE 5. Enzymatic activities (mU/mg protein) in hypothalamic neurons, astrocytes, and mixed cultures of neurons on astroglial substratum, and in hypothalamus from 6-day-old rats

Neurons

Astrocytes

Mixed culture

Hypothalamus froen 6-day-old rats

HK LDH G6PDH GDH MAB-A MAO-B AIB AChE Na -K ATPase Mg ATPase NOTE:Values arc means f SD of 4-6 separate experiments and are unpublished data of J. C. K. Lai. T. K. C. Leung, and L. Lim. The enzymatic activities were assayed as describecl previously (Lai and Clark 1979. Lai et al. 1980b, 1984; Leung et al. 1981). The prlmary cultures of hypothalam~cneurons, astrosytes, and mixed cultures of neurons on astroglial substratum were prepared as descr~bedby Whatlcy et al. r 198 1) and were maintained for 10- 12 days in vitro before the experiments. HK, hexokinase; LBH, lactate dehydrogenase; G6PBH, glucosc-6-phosphate dehydrogenase; GDH, NAD-linked glutatnate dehydrogenase: MAO-A, and MAQ-B. type A and B monoamine oxidase: AiB, MAO-A to MAO-B activity ratio; AChE, acetyicholinesterase; Na-K ATPase, Na,K-actsvated adenosine triplnosphatase: Mg ATPase, Mg-activated adenosine triphosphatase; ND. not determined. For other details see Table 2. * p < 0.05 versus corresponding values in the other groups.

s@te in eerebrocofiical mitschsndria under both state 4 and state 3 conditions when the external Kf levels are increased from 5 to 100 mM (Table 4). Thus, our data provide, for the first time, direct evidence that raising the extramitocksndrial Kf level does result in an elevation of the activation state of PDHC in brain mitochondria. It is noteworthy that under state 4 conditions, the percent increase (36%) in PDHC flux induced by K f is similar to the percent elevation (46%) of the PDHC activation state induced by K t (Table 4). By contrast, under state 3 conditions, the percent increase (94%) in PDHC flux induced by Kf is higher than the percent elevation (59%) of the PDHC activation state induced by Kf (Table 4). Nonetheless, it is important to note that irrespective of the metabolic state (and many metabolic states have been i~vestigated)the PBHC activation state is always high enough to account for the flux through the csaaaplex. (See Lai and Sheu 1985, 1987 for additional discussion.)

One surprising finding is that whereas both the flux through PDHC and the pyruvate-dependent oxygen uptake are increased in the state 4 to state 3 respiratory state transition in cerebrocortical mitochondria, the activation state of PDHC is actually depressed in the state transition (Table 4 and Figs. 1 and 2). The ADP-induced lowering of the PDHC activation state is dependent on the respiratory substrate (Figs. 1 and 2) as well as the concentration sf the added ADP (Pig. I). For example, e the respiratory substrates, the when glutamate and d ~ t are maximum lowering of the PDHC activation state is attained with 0.1 rnM added ADP (Fig. I). By contrast, when pymvate and rnal~teare the substrates (Fig. I), the maximum lowering s f the BDHC activation state is whieved only at ADP ceneentrations above 0.5 mM (Fig. 1). Moreover, the ADP-induced lowering of the PDHC activation state is dependent, at least in part, dan the conversion sf ADP to ATP by mitochondria1 oxidative phssphorylatisn, since the rate of ADP-induced

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 12/02/14 For personal use only.

ADP (mM) FIG. 2 . BBHC activities in cerebrocortical nonsynaptic mitochondria at different ADP concentrations. Cerebrocortical nonsynaptic mitochondria were isolated as described by Lai and Clark (19'99>1989). Mitochondria were preinsubated with 5 mM pymvate plus 2.5 anM malate (c)or 10 mM glutamate plus 2.5 mM malate ( 8 )in 30 mM K + medium with oligornyein (75 pg1rnL) (Lai and Sheu 1985, 1987). At 1 min, ADP was added at the concentration indicated, and the incubation was continued for a further 2 min. Samples (i.e., mitschon$rial suspensions) were withdrawn, "sttop frozen9 (see Lai and Sheu 1984, I987 for experimental details) and assayed for PDHC activity. Unpublished data of J. C. K. Lai and K.-F. R. Skeu. ?

lowering of the PDHC activation state in uncoupled cerebrocortical mitochondria (by the addition of the uncoupHer CCCP (carbonyl cyanide m-ehlorophenylhydrazone) to the incubation medium) is significantly less than that observed in coupled cerebrocortical mitochonadria (Fig. 2). The lowering sf the PBHC activation state during the state 4 to state 3 transition in brain mitochondria contrasts sharply with the corresponding results obtained with peripherd tissue (e.g ., heart and liver) mitochondria (see Lsi and Sheu 1987 for discussion and refs,) . In peripheral tissue mitochondria, all three paradigms (i .e., py ruvate-dependent oxygen uptake, flux through the PDHC, and the activation state of the PDHC) are increased in the state 4 to state 3 transition. Another difference that is quite pronounced between brain and peripheral tissue mitochondria (but is not discussed by workers in this field) is that whereas in peripheral tissue mitochondria the activation state of the PDHC is remarkably low under state 4 conditions, the activation state of the PDHC in cerebrocortical mitochondria is very much higher under these conditions (i.e., 5075% active, depending on the substrate), irrespective sf the K+ stimulation.

Cellular distribution QE the activities of several enzymes of intermediary metabolism: studies with primary cultures Metabolic heterogeneity with respect to cell tjpes To investigate the possibility that neural cells may exhibit metabolic characteristics that are cell-type specific (see Edmsnd et al. 1987; Sch~ueboeet al. 1988; Hertz et id, 1988 for additional discussion and refs.), we began (see Lai et al1985b) some time ago to determine the activities of several enzymes of intermediary metabolism in primary cultures of hypothalamic neurons, astrscytes, and neurons grown on an

PIG. 2. PDHC activities in cerebrocortical nonsynaptic rnitmhondria oxidizing succinate in the presence or absence of an uncoupler (CCCP). Cerebrocortical nonsynaptic mitochondria were isolated by the Lai and Clark (1979, 1989) procedure. Mitochondria were incubated in 38 mM K+ medium (Lai and Sheu 1985, 1987) with 10 glpM succinate and rotenone (10 pg/rnL) in the presence (u,m) or absence (3,e ) of CCCP (carbony1 cyanide m-chlorophenylhydrozone) ( 5 pg1mL). After 5 min, 0.5 mM ADP was added ( a , m). At the times indicated, samples (i.e., rnitmhondrid suspensions) were withdrawn, "stop frozen," and assayed for PDHC activity (see Lai and Sheu 1985, I987 for experimental details). Unpublished data of J. C. K. Lai and M.-F. W. Sheu.

astroglial substratum, and in hypothalamic hornogenates from age-matched animals (i.e., 6 day old rats). (We focused on the hypothalamus because we were interested in the role of sex steroids in the regulation of intermediary metabolism (see Lai et aH. 1980a for discussion).)

Ce8lukar distribution of the activities of enzymes of intermediary metakP~Eisrn Hexokinase (HK) activity is higher in astrocytes than in neurons or in the mixed cultures of neurons on astrocytes (Table 5). However, HK activity in hypothalamic homogenate from 6-day-old rats is higher than the activity in any of the cultured cells (Table 5). Lactate dehydrsgenase (LDH) activity is highest in astrocytes, intermediate in the mixed cultures, and lowest in neurons (Table 5). LBH activity in hypothalamic homogenate from 6-day-old rats is almsst the same as that in neurons (Table 5). Glucose-6-phosphate dehydrogenase (G6PDH) activity is highest in astrocytes, intermediate in the mixed cultures, but lowest in neurons (Table 5). Moreover, this enzymatic activity is quite low in hypothalamic homogenate from 6-day-old rats, even lower than that in neurons (Table 5). Recently Rust et al. (1991) reported the activities of 16 enzymes of intermediary metabolism (including some of the enzymes (e.g., HK, LDH, and G5PDH) of the glycolytic and hexose rnonophssphate shunt pathways) in cultured cerebrocortical astrocytes, oligodendrocytes, Schwam cells, and neurons. (They (Rust et al. 1931) also investigated these enzymes in cultured Schwann cells and neurons from rat superior cervical ganglion.) However, it is difficult to compare their data with ours because Rust et al. (199 1) expressed the enzymatic activities in the cultured cells as percentages of the corresponding enzymatic activities in mouse brain homogenates. Nevertheless, their data indicate that the G6PDH activities in

CAN. J. BHYSIOL.

S136

PHARMACOL. VOL.

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 12/02/14 For personal use only.

cultured rat cerebroco~icalastrocytes are significantly higher than those in mouse brain homogenates @ 14). (This high A/B ratio in neurons has been confirmed in studies using selective MA8-A and MAO-B inhibitors (T. K. C. Leung, 9. C . K. Lai, and L. Lim, in preparation).) Thus, one could reasonably conclude that the M A 8 activity in neurons may be exclusively type A. Indeed, our results obtained with hypothalamic cells are in good agreement with our other observations (Lai et al. 1985b; T. K. C . Leung, J. C. K. Lai, and L. Lim, in preparation), as well as those of Yu and Hertz (1982), obtained with cerebrocortical cells in primary culture. Consistent with it being a putative cholinergic marker, the activity of acetylchslinesterase (AChE) is highest in neurons, intermediate in the hypothalamus from 6-day-old rats, and lowest in the mixed cultures and in astrocytes (Table 5). Activity of sodium, potassium-activated adenosine triphosphatase (Na -K ATPase) is higher in neurons than in the mixed cultures or in the hypothalamus from 6-day-old rats (Table 5 ) . By contrast, activity of magnesium-activated adenosine triphosphatase (Mg ATPase) is highest in astrocytes, intermediate in neurons, but lowest in the mixed culture and in the hypothalamus from 6-day-old rats (Table 5 ) . Although they do not prove it, the data in Table 5 are at least consistent with the notion that co-culturing neurons with astrocytes appears to result in a lowering of the astroglial enzymatic activities. Moreover, these observations raise the distinct possibility that metabolic characteristics differ markedly, dependent on the cell type. Obviously, additional studies are required to test these hypotheses further.

ConcHuding remarks (i) I discussed how the prior problems s f brain mitochondrial isolation led us to develop new methodology. ( i i ) One of our methods allows the isolation of two populations of synaptic (namely, fractions SM and SM2) and one population of nonsy naptic (namely, fraction A) mitochondria from the rat forebrain. Another method, the Clark and Nicklas (1970) procedure, allows the isolation of a different population of nonsynaptic mitochondria (namely, fraction B) . All four preparations are metabolically active and relatively pure and are thus very suitable for metabolic and other studies. All four mitochondria1 populations are different to various degrees in their enzyme contents, and these enzymatic differences are also reflected, to some extent, in their ability to oxidize various substrates. (tii) The heterogeneity of brain mitochondria

70, 1992

observed in the forebrain also exists at the regional level. iso(iv) To illustrate how the brain mitochondrial lated by our procedures can be usefully exploited for metabolic studies, I focused on the example of the K" stimulation of mitochondria1 pyruvate metabolism. Our studies revealed that the underlying mechanisms involve a direct stimulation by K + of the flux through PDHC and a K+-induced elevation of the PDHC activation state. (at) The results from our previous and ongoing studies using primary cultures of neurons and astrocytes are consistent with the notion that brain cells are heterogeneous with respect to their capabilities in energy and intermediary metabolism. The recent advances in the neural cell culture techniques have made it possible to prepare primary cultures of neurons and astrocytes of high purity ( > 98%)and in good yield. Consequently, E can envisage that in the not-so-distant future, one could adapt these preparations as the starting material for the isolation of mitochondria of known cellular origin for metabolic studies. I look forward very much to these exciting possibilities.

Acknowledgements 1 would like to acknowledge the contributions of my former and current collaborators: Drs. J. B. Clark, S. F. Leong, T. K. C. Leung, L. Eim, and K.-F. R. Sheu. The studies were supported, in part, by the Medical Research Council (U.K.), the Worshipful Company of Pewterers (U.K.). the Brain Research Trust (U.K.), and the National Institutes s f Health (U.S.A.). 1 thank Dr. D. L. Diedrich and Dean A. A. Nelson, Jr., for their interest and continued support. Butterworth, R. F.. and Gigukre, J.-E. 1984. Pyruvate dehydrogenase activity in regions of the rat brain during postnatal development. J. Neurochem. 43: 280-282. Clark, J. B . , and NicMas, W. J. 1970. The metabolism of brain mitochondria. J. Biol. Chem. 245: 4724 -473 1. Clark, J. B., and Nicklas, W. J. 1984. Brain mitochondria. In Handbook of neuroehemistry. Vol. 7. 2nd ed. Edited by A. Lajtka. Plenum Press, New York. pp. 139- 159. De Robertis, E.. and de Lores Arnaiz, G . R. 1969. Structural components of the synaptic region. 1r1 Handbook of neuroehemistry. Vo1. 2. 1st ed. Edited by A. Lajtha. Plenum Press, New York. pp. 365 - 392. Edmond, J . , Wobbins, R. A., Bergstrsm, S. D., Cole, W. A., and de Vellis, 3. 1987. Capacity for substrate utilization in oxidative metabolism by neurons, astrocytes, and oligsdendrocytes from developing brain in primary culture. J. Neurosci. Res. 18: 551 561. Hertz, L. 196%.Possible role of neuroglia: a potassium-mediated neuronal - neuroglial -neuronal impulse transmission system. Nature (London), 206: 1091- 1094. Hertz. L. 1990. Regulation of potassium homeostasis by glial cells. Ia Differentiation and functions of glial cells. Alan W. Liss, New York. pp. 225-234. Hertz: L., and Schousbse, A. 1987. Role of astrocytes in compartmentation of amino acid and energy metabolism. In Astrocytes. Vol. 2. Edited by S. Fedoroff and A. Vernadakis, Academic Press, Neav York. pp. 179-208. Hertz, L., Murthy, Ch. R. K., and Schousboe, A. 1988. Metabolism of glutamate and related amino acids. In The biochemical patho%ogy of astrocytes. Edited by M . D. Norenberg, L. Hertz, and A. Schousboe. Alan W. Liss, New York. pp. 395-406. Lai, J. C. K., and Clark, J. B. 1979. Preparation of synaptic and nonsynaptic mitochondria from mammalian brain. In Methods in enzymology. Vol. 55. Bart F. Edited by S. Fleischer and L. Packer. Academic Press, New York. pp. 51 -60.

Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 12/02/14 For personal use only.

Lai, J. C. K.. and Clark, J. B. 1989. Isolation and characterization of synaptic and nonsynaptlc mitochondria from mammalian brain. In Neuromethods. Vol. 11. Edited by A. A. Boulton, G. B. Baker, and R. F. Butterworth. Humana Press, Clifton, N.J. pp. 43-98. Lai, J. C. K., and Cooper, A. J. L. 1986. Brain a-ketoglutarate dehydrogenase complex: kinetic properties, regional distribution, and effects of inhibitors. J. Neurochern. 47: 1376- 1386. Lai, S . C. K., and Sheu, KO-F.R. 1985. Relationship between activation state of pymvate dehydrogenase complex and rate of pyruvate oxidation in isolated cerebro-cortical mitochondria: effects of potassium ions and adenine nucleotides. 3. Neurochem. 45: 1861 1868. Lai, J. C. K., and Sheu, K.-F. R. 1987. The effect of 2-oxoglutarate or 3-hydroxybutyrate on pyruvate dehydrogenase complex in isolated cerebro-cortical mitochondria. Neurochem. Res. 12: 715 722. Lai, J. C. K., Walsh, S. M., Dennis. S. C., and Clark, J. B. 1977. Synaptic and non-synaptic mitochondria from rat brain: isolation and characterization. J. Neurochem. 28: 625 - 63 1. Lai. 9. C. K., Leung, T. K. C . , Marr, W., and Lim, L. 1980a. The activity of glucose-6-phosphate dehydrogenase in liver and hypothalamus of thc female rat: effects of administration of ethinyl oestradiol and the progestogens northisterone acetate and D-norgestrel. Biochern. Soc. Trans. 8: 614. La,, J. C. K.. Guest, J. F., Leung, T. K. C., Lim, L., and Davison, A. N. 1980h. The effects of cadmium, manganese and aluminum on sodium-ptassium-activated and magnesium-activated adenosine triphosphatase activity and choline uptake in rat brain synaptosomes. Biochem. Pharmacol. 29: 141 - 146. Lai, J. C. K., h u n g , T. K. C., and Eim, L. 1981. Monoamine oxidase in synaptic and non-synaptic mitochoi~dria from brain regions. Abstracts of the International Society for Neurochemistry, Nottingham, U.K. p. 291. Lai, J. C. K., Leung, T. K. C., and Lim, L. 1982. Activities of the mitochondria1 NAD-linked isocitric dehydrogenase in different regions ofthe rat brain. Changes in ageing and the cffect of chronic manganese chloride administration. Gerontology, 28: 8 1 - $5. Lai, J. C . K., Leung, T. K. $1.. and Lam, L. 1984. Differences in the neurotoxic effects of manganese during development and aging: some observations on brain regional neurotransmitter and non-neurotransmitter metabolism in a developmental rat model of chronic manganese encephalopathy . Neurotoxicology , 5: 37 -48. Lai, J. C. K., Sheu, K.-F. R., and Carlson, K. C., Jr. 19850. Differences in some of the metabolic properties of mitochondria isolated from cerebral cortex and olfactory bulb of the rat. Brain Res. 343: 52-59. Lai, 9. C. K., Leung, T. K. C., and Lim, L. 198%. Effects of metal ions on neurotransmitter hnction and metabolism. Bsa Metal ions in neurology and psychiatry. Edited by S. Gabay, J. Harris, and B. T. Ho. Alan R. Liss. New York. pp. 177 - 197. Lai, J. C. K., Wimpel-Lamhaouar, K., and Cooper, A. J. L. 1989. Selective inhibition of mitochondrial dehydrogenases by ammonia and fatty acyl coenzyme A derivatives. b~zHepatic encephalopathy. Edited b y R. F . Butterworth and G . P. Layrargues. Humana Press, Clifton, N.J. pp. 91 -98. k o n g , S. F., and Clark, J . B. 1984a. Regional development of glutamate dehydrogenase in the rat brain. J. Neurochem. 43: 106- 111. Leong, %. F., and Clark, J. B. 1984b. Regiowal enzyme development in sat brain. Enzymes associated with glucose utilization. Biochem. J. 218: 131 - 138. Leong, S. F., and Clark, J. B. 1 9 8 4 ~Regional . enzyme development in rat brain. Enzymes of energy metabolism. Biochem. J. 218: 139- 145. Leong, S. F.. Lai. J. C. K., Lim, L., and Clark, J. B. 1981. Energymetabolizing enzymes in brain regions of adult and aging rats. 9 . Neurochem. 37: 1548 - 1554.

Leong, S. F., Lai, J. C. K., Lirn, L . , and Clark. J. B. 1984. The activities of some energy-metabolizing enzymes in nonsynaptic (free) and synaptic mitochondria derived from selected brain regions. 9. Neurochem. 42: 1306 - 131%. Leasng. T. K. C., Lai, J. C. K., and Lim, L. 1981. The regional distribution of monoamine oxidase activities towards different substrates: effects in rat brain of chronic administration of manganese chloride and of ageing. J. Neurochem. 36: 2037 -2043. Malloch, G. D. A., Munday, L. A., Olson. M. S., and Clark, 9. B. 1986. Comparative development of the pymvate dehydrogenase complex and cltrate synthase in rat brain mitochondria. Biochem. 9. 238: 729-736. Martin. R. J., Bird, M. I., Saggerson, E. D., Munday, L. A.. and Clark, J. B. 1987. Enzyme activities in regions of the hypsthalamus. J. Neurochem. 48: 738-740. Nicklas, W. J., Clark, J. B., and Williamson, J. 8 . 1971. Metabolism of rat brain mitochondria. Studies on the potassium ionstimulated oxidation of pymvate. Biochem. 9. 123: 83 - 95. Raine, C. S. 1981. Neurocellular anatomy. ha Basic neurochemistry . 3rd ed. Edited by G. J. Siegel, R. W. Albers, 8 . W. Agranoff, and W. Katzman. Little, Brown, Boston. pp. 21 -47. Randle, P. 3. 1981. Phosphorylation -dephosphorylation cycles and the regulation of fuel selection in mammals. In Current topics in cellular regulation. Vol. 18. Edited by R. W. Eshbrook and P. Srere. Academic Press, New York. pp. 107-128. Reed, L. J. 198 1. Regulation of mammalian pymvate dehydrogenase complex by a phosphol-ylation-dephcpsphoqlation cycle. I n Current topics in cellular replation. Vo1. 18. Edited by R. W. Estabrook and P. Srere. Academic Press, New York. gp. 95 - 104. Rust, R. S., Jr., Carter. J. G., Martin, D., Nerbonne, J. M., Lampe, P. A., Pusateri, M . E., and Lowry. O. H. 1991. Enzyme levels in cultured astrocytes, oligodendrocytes and Schwann cells. and neurons from the cerebral cortex and superior cervical ganglia of the rat. Neurochem. Res. 16: 991 -999. Schousboe. A., Earsson, 0 . M., Krogsgaard-Larsen, P., Brejer, J.. and Hertz, L. 1988. Uptake and release processes for neurotransmitter amino acids in astrocytes. In The biochemical pathology of astrocytes. Edited b y M. D. Norenberg, L. Hertz, and A. Schousboe. Alan R. Liss, New York. pp. 381 -394. Sheu. K.-F. R., Lai, J . C. K., and Blass, J. P. 1983. Pyruvate dehydrogenase phosphate (PBH,) phosphatase in brain: activity, properties, and subcellular localization. J. Neurochem. 40: 13661372. Sheu, K.-H. R.,Lai, 9 . C. K., and Blass, J. P. 1984. Properties and regional distribution of pymvate dehydrogenase kinase in rat brain. J . Neurochem. 42: 230-236. Sokoloff, L. 1983. Measurement of local glucose utilization in the central nervous system and its relationship to local functional activity. Handbook of neurochemistry . Vol. 3.2nd ed. Edited by A. Lajtha. Plenum Press, New York. pp. 225-257. Whatley, S. A., Hall, C., and Lirn, L. 1981. Hypothalamic neurons in dissociated cell culture: the mechanism of increased survival times in the presence of non-neuronal cells. 9. Neurochem. 36: 2052 -2056. Whittaker, V. P. 1969. The synaptosome. Bra Handbook of neurochemistry. Vol. 2. 1st ed. Edited lay A. Lajtha. Plenum Press, New York. pp. 327-364. &%ittaker, V . P. 1984. The synaptosome. Im Handbook of neurochemistry- Vol. 7. 2nd ed. Edited b y A. Lajtha. Plenum Press, New York. pp. 1-39. Yu,P. H., and Hertz, L. 1982. Differential expression of type A and type B monoamine oxidase of mouse astrocytes in primary culture. J. Neurochem. 39: 1492 - 1495.