Control of mitochondrial L-oxidation: sensitivity of the trifunctionalprotein to [NAD�]/[NADH] and [acetyl-CoA]/[CoA]

Simon Eaton a; b;*, Bruce Middleton c, Kim Bartlett a

a Sir James Spence Institute of Child Health, Royal Victoria In¢rmary, Newcastle-upon-Tyne NE1 4LP, UKb Unit of Paediatric Surgery, Institute of Child Health, 30 Guilford Street, London WC1N 1EH, UK

c Department of Biochemistry, University of Nottingham Medical School, Queens Medical Centre, Nottingham NG7 2UH, UK

Received 27 August 1998; received in revised form 15 October 1998; accepted 15 October 1998

Abstract

Isolated human mitochondrial trifunctional protein was incubated with 2-hexadecenoyl-CoA, CoA and NAD� and theresultant CoA esters measured. Steady state with respect to the concentrations of the intermediates 3-hydroxyhexadecanoyl-CoA and 3-ketohexadecanoyl-CoA and the rate of formation of the product tetradecanoyl-CoA was reached within 4 min.Flux was greatly enhanced by the addition of Tween 20 (0.2% v/v) which stimulated 3-ketoacyl-CoA thiolase activity by over7-fold. When 3-ketoacyl-CoA thiolase was not stimulated, 3-hydroxyhexadecanoyl-CoA was the prominent CoA esteraccumulated, presumably due to inhibition of 3-hydroxyacyl-CoA dehydrogenase activity by accumulated 3-ketoacyl-CoA,analogous to the inhibition of short-chain 3-hydroxyacyl-CoA dehydrogenase by 3-ketoacyl-CoA. When [NAD�]/[NADH]was varied at a fixed total [NAD�+NADH], the overall flux was only inhibited by [NAD�]/[NADH] less than 1. In contrast,when [acetyl-CoA]/[CoA] was varied at a fixed total [CoA], much greater sensitivity was observed. ß 1998 Elsevier ScienceB.V. All rights reserved.

Keywords: Mitochondrial L-oxidation; [NAD�]/[NADH]; Acetyl-CoA/CoA; Trifunctional protein; 3-Hydroxyacyl-CoA dehydrogenase;3-Ketoacyl-CoA thiolase; 2-Enoyl-CoA hydratase

1. Introduction

In recent years, it has become apparent that thereare membrane-associated enzymes responsible for themitochondrial L-oxidation of long-chain fatty acids.The existence of a long-chain 2-enoyl-CoA hydratasewas demonstrated by Schulz [1], and that of a long-chain 3-hydroxyacyl-CoA dehydrogenase by El-Fakhri and Middleton [2]. These activities wereshown to be part of a single protein, the trifunctionalprotein, together with that of a long-chain speci¢c

3-ketoacyl-CoA thiolase [3,4]. The activities and sub-units of the trifunctional protein are indicated inFig. 1. Inborn errors of the trifunctional proteinhave been demonstrated which either involve a de¢-ciency of 3-hydroxyacyl-CoA dehydrogenase activitywith a mild reduction in 2-enoyl-CoA hydratase and3-ketoacyl-CoA thiolase activities [5] due to a com-mon point mutation [6], or severely diminished activ-ities of all three components together with completeabsence of immunodetectable trifunctional protein[7]. Also associated with the inner mitochondrialmembrane are very-long-chain acyl-CoA dehydo-genase [8], electron-transfer £avoprotein:ubiquinoneoxidoreductase [9] and complex I of the respiratory

0167-4838 / 98 / $ ^ see front matter ß 1998 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 7 - 4 8 3 8 ( 9 8 ) 0 0 2 4 6 - 5

* Corresponding author.

BBAPRO 35807 1-12-98

Biochimica et Biophysica Acta 1429 (1998) 230^238

chain so that a L-oxidation metabolon consisting ofthe long-chain enzymes of L-oxidation and their re-oxidants can be envisaged [10].

Qualitative studies of the control of mitochondrialL-oxidation, using intact mitochondria incubatedwith hexadecanoate, hexadecanoyl-CoA or hexade-canoyl-carnitine and measurement of acyl-CoA andacyl-carnitine esters, NAD� and NADH have dem-onstrated that L-oxidation of long-chain substrates issensitive to respiratory chain inhibition [11^13]. Fur-thermore, when complex I of the respiratory chain isinhibited, the concentration of 3-hydroxyacyl-CoAesters increases [14,15]. In addition, it has beenpostulated that as 3-ketoacyl-CoA esters are potentfeedback inhibitors of the 3-hydroxyacyl-CoA dehy-drogenases, the 2-enoyl-CoA hydratases and theacyl-CoA dehydrogenases [16,17] and the 3-ketoac-yl-CoA thiolases are sensitive to inhibition by the[acetyl-CoA]/[CoA] ratio [18], overall, L-oxidation£ux may be modulated by the [acetyl-CoA]/[CoA]ratio [19]. In view of the importance of the NAD�/NADH and acetyl-CoA/CoA couples in the regula-tion of L-oxidation £ux and the involvement of thenewly discovered trifunctional protein, we examinedthe sensitivity of the isolated trifunctional protein toalterations in the [NAD�]/[NADH] and [acetyl-CoA]/[CoA] ratios. This was achieved by mass anal-

ysis of CoA esters accumulating from the incubationof the trifunctional protein with 2-hexadecenoyl-CoA, with the [NAD�]/[NADH] and [acetyl-CoA]/[CoA] ratios varied using, as bu¡ers, lactate dehydro-genase/lactate/pyruvate and carnitine acetyl transfer-ase/carnitine/acetyl-carnitine, respectively.

2. Materials and methods

2.1. Materials

2-Hexadecenoyl-CoA was synthesised enzymati-cally from hexadecanoyl-CoA (Sigma, Dorset, UK)using acyl-CoA oxidase (Boehringer-Mannheim, Sus-sex, UK). To 1.7 mM hexadecanoyl-CoA in 25 mMKH2PO4 pH 7.5 was added 1.125 U acyl-CoA oxi-dase/Wmol substrate. The reaction was allowed tocontinue for 30 min at 30³C (with constant shakingto ensure oxygenation), after which the progress waschecked by HPLC (see below). A further aliquot ofacyl-CoA oxidase was added if the reaction had notgone to completion. 2-Hexadecenoyl-CoA was puri-¢ed on a C18 Bond-Elut cartridge [13]. 3-Ketohexa-decanoyl-CoA was synthesised enzymatically from2-hexadecynoyl-CoA with crotonase [20]. Both CoAesters were greater than 95% pure as analysed by

Fig. 1. Diagram showing the activities and subunits of the trifunctional protein of L-oxidation, together with the recycling systemsused in the experimental section. LDH, L-lactate dehydrogenase; CAT, carnitine acetyl transferase.

BBAPRO 35807 1-12-98

S. Eaton et al. / Biochimica et Biophysica Acta 1429 (1998) 230^238 231

HPLC (see below). Human liver trifunctional proteinwas isolated as described previously [3]. It was pre-pared using octylthioglucoside, a detergent with ahigh critical micellar concentration, so that the de-tergent could be e¡ectively removed by dilution intothe assay or by rapid gel ¢ltration. HPLC grade sol-vents were from Rathburn (Peeblesshire, UK) andHPLC grade bu¡ers were from Merck (Dorset,UK). Pigeon liver carnitine acetyl transferase, bovineliver glutamate-dehydrogenase (type II), bis^tris-pro-pane and Tween 20 were from Sigma (Dorset, UK)and pig heart L-lactate dehydrogenase from Boehr-inger-Mannheim (Sussex, UK). Acetyl-carnitine wasfrom Aldrich (Dorset, UK).

2.2. Measurement of enzyme activity

Trifunctional enzyme was assayed in the forwarddirection using 25 WM 2-hexadecenoyl-CoA as sub-strate in 50 mM bis^tris-propane HCl pH 9.0, con-taining 1 mM NAD�, 1 mM CoA and 0.2% (v/v)Tween 20, and the activity followed by the increasein absorbance of NADH at 340 nm. The activity(long-chain 3-hydroxyacyl-CoA dehydrogenase) inthe reverse direction was assayed using 3-ketohexa-decanoyl-CoA as substrate as described by El-Fakhriand Middleton [2] following the decrease in absorb-ance at 340 nm. The activity in the forward directionwas 1^3 U/mg protein and in the reverse directionwas 11^25 U/mg protein. 2-Enoyl-CoA hydratase ac-tivity was measured following the initial rate of dis-appearance of absorbance at 263 nm, using 25 WM2-hexadecenoyl-CoA as substrate [21]. 3-Ketoacyl-CoA thiolase activity was measured at 303 nm and30³C as the rate of cleavage of 3-ketohexadecanoyl-CoA (10 WM) in 50 mM bis^tris-propane, 2 mMMgCl2, 0.2% (v/v) Tween 20 pH 9.0 in the presenceof 0.2 mg/ml CoASH. The O mM for 3-ketohexade-canoyl-CoA was found to be 16.0 at 303 nm.

2.3. Incubations for HPLC analysis

Incubations with isolated trifunctional protein and2-hexadecenoyl-CoA were carried out at 30³C in ashaking water bath, in 500 Wl of a medium consistingof 50 mM bis^tris-propane/1 mM NAD�/260 WMCoA pH 9.0 except where stated. Amounts of1.2 mU (forward) of trifunctional protein were pre-

incubated in the above medium for 2 min before theaddition of 90 WM 2-hexadecenoyl-CoA. Reactionswere allowed to continue for 4 min, except wherestated, and quenched by the addition of 40 Wl 5 MHCl, which brought the pH to 3^4. Internal standard(50 nmol heptadecanoyl-CoA) was added, togetherwith 0.2% (v/v) Tween 20, to those incubations whichdid not already include Tween 20 (the addition ofTween 20 prevented any precipitation of CoA esters).The samples were analysed directly by HPLC [11]using a 5W C18 Supelco Guard column to protectthe analytical column from injected protein and de-tergent. Quantitation was on the basis of peak areawith recovery corrected with the internal standard.Acetyl-CoA and CoA were analysed from the samesamples by HPLC, using a Hypersil 5W C18 250U4.6 mm column, and a gradient of 220 mMKH2PO4/water/acetonitrile: 6 min isocratic 90%KH2PO4/7% water/3% acetonitrile, followed by a lin-ear gradient to reach 82% KH2PO4/12% water/6%acetonitrile at 20 min and a linear gradient to 30%KH2PO4/49% water/21% acetonitrile at 40 min. The£ow rate was 1 ml/min and the column washed withwater between injections. Detection was based onUV absorbance using a photodiode array detectorbetween 200 and 300 nm. NAD� and NADH weremeasured directly by HPLC in phenol/chloroform/isoamyl alcohol quenched samples as described pre-viously [13].

3. Results and discussion

3.1. E¡ect of pH on the forward activity of thetrifunctional protein

The activity of the trifunctional protein (TP) in theforward direction was measured between pH 6.0 and10.0 using 50 mM bis^tris-propane as the bu¡er andwas greatest at pH 9.0 (Fig. 2). The pH sensitivity ofthe overall reaction is presumably a consequence ofthe dependence of the 3-hydroxyacyl-CoA dehydro-genase equilibrium on [H�] and possibly some de-pendence of the 3-ketoacyl-CoA thiolase on the ex-tent of enolisation of 3-ketohexadecanoyl-CoA. Allthe CoA esters (2-hexadecenoyl-, 3-hydroxyhexade-canoyl-CoA, 3-ketohexadecanoyl-CoA, tetradecano-yl-CoA) measured were stable at this pH for at least

BBAPRO 35807 1-12-98

S. Eaton et al. / Biochimica et Biophysica Acta 1429 (1998) 230^238232

15 min so all further incubations were carried out atthis pH unless stated.

3.2. E¡ect of Tween 20 on individual activities of thetrifunctional protein

In the presence of Tween 20 (0.2%, v/v), 3-ketoac-yl-CoA thiolase activity was stimulated by 726%,there was no signi¢cant e¡ect on 3-hydroxyacyl-CoA dehydrogenase activity (reverse direction) and2-enoyl-CoA hydratase activity was inhibited by36%.

3.3. Time courses of production of CoA esters from2-hexadecenoyl-CoA

In the absence of Tween 20, large amounts (greaterthan 30 nmol) of 3-hydroxyhexadecanoyl-CoA wereformed within 2 min (Fig. 3); however, the amountdid not increase any further during the course of theincubation. The rates of the production of tetradeca-noyl-CoA and 3-ketohexadecanoyl-CoA were linearwith respect to time, although the accumulation oftetradecanoyl-CoA was very low. In the presence of

Tween 20, the amounts of accumulated 3-hydroxy-hexadecanoyl-CoA and 3-ketohexadecanoyl-CoAwere lower than in its absence and reached steadylevels within 4 min, and the amount of tetradecano-yl-CoA higher than in the absence of Tween 20, witha production rate linear with respect to time. Theaccumulation of 3-ketohexadecanoyl-CoA with re-spect to time in the absence of Tween 20 and theaccumulation of tetradecanoyl-CoA in the presenceare due to the stimulatory e¡ect of Tween 20 on3-ketoacyl-CoA thiolase activity. The dramatic e¡ectof Tween 20 on the levels of 3-hydroxyhexadecanoyl-CoA is probably due to inhibition of 3-hydroxhexa-decanoyl-CoA dehydrogenase activity by 3-ketoacyl-CoA esters analagous to the short-chain enzyme[16,17] so that in the absence of Tween 20, the in-creased amount of 3-ketohexadecanoyl-CoA formedinhibits the 3-hydroxyacyl-CoA dehydrogenase.Tween 20 did not signi¢cantly a¡ect 3-hydroxhexa-decanoyl-CoA dehydrogenase activity of the TP inthe reverse direction, although we did not measurethe e¡ect of Tween 20 on its activity in the forwarddirection. In a study in which pig heart trifunctionalprotein was incubated with 2-hexadecenoyl-CoA, no

Fig. 2. pH dependence of trifunctional protein activity in the physiological direction. Isolated trifunctional protein was incubated in50 mM bis^tris-propane with 2-hexadecenoyl-CoA, NAD�, CoA and 0.2% (v/v) Tween 20. The data are means of replicate initialrates of NADH production (see Section 2).

BBAPRO 35807 1-12-98

S. Eaton et al. / Biochimica et Biophysica Acta 1429 (1998) 230^238 233

accumulation of 3-ketoacyl-CoA esters was observed[22]. The di¡erence between that observation andthe present data might be due both to the speciesof origin and the presence of detergent (TritonX-100) in the enzyme preparation used in the otherstudy [22,23]. We ¢nd that Triton X-100 has a sim-ilar e¡ect to Tween 20 in stimulating the 3-ketoacyl-CoA thiolase activity of the human trifunctional pro-tein.

3.4. E¡ect of [NAD+]/[NADH] ratio on productionof CoA esters from 2-hexadecenoyl-CoA

We have shown accumulation of 2-enoyl- and 3-hydroxyacyl-CoA esters in the absence of grosschanges in the intramitochondrial [NAD�]/[NADH], measured directly in incubations of intactrat skeletal and cardiac muscle mitochondria with[U-14C]hexadecanoyl-CoA or [U-14C]hexadecanoate([11] and Eaton, Bartlett and Pourfarzam, submittedfor publication). This implies that either the trifunc-tional protein is very sensitive to small changes in the[NAD�]/[NADH] ratio, or that there is a pool of

NAD�/NADH channelled between the trifunctionalprotein and complex I and that it is the turnover ofthis pool which is responsible for accumulation of2-enoyl- and 3-hydroxyacyl-CoA esters in the ab-sence of changes in the total (bulk phase) mitochon-drial NAD�/NADH pool. In order to investigate thee¡ect of [NAD�]/[NADH] on the activity of the tri-functional protein, the lactate dehydrogenase/pyru-vate/lactate system was used as a dynamic bu¡er sothat [NAD�]/[NADH] could be varied at a ¢xed to-tal (NAD�+NADH) concentration. Incubationswere carried out for 4 min at pH 9.0 in the presenceof 2 U/ml lactate dehydrogenase, 10 WM NAD�,125 WM pyruvate and suitable concentrations of lac-tate to give appropriate [NAD�]/[NADH] ratios, us-ing:

Keq � �pyruvate��NADH��H���lactate��NAD�� � 4:2U10312

[24]. The [NAD�]/[NADH] ratios were measured byHPLC, both in the presence and absence of trifunc-tional protein/2-hexadecenoyl-CoA/CoA to verifythese [NAD�]/[NADH] ratios and to ensure that ad-

Fig. 3. Graph showing the accumulation of CoA esters with respect to time in incubations of human liver trifunctional protein with2-hexadecenoyl-CoA (as described in Section 2) in the presence and absence of 0.2% (v/v) Tween 20. 8, tetradecanoyl-CoA; F, 3-ke-tohexadecanoyl-CoA; R, 3-hydroxyhexadecanoyl-CoA. Symbols: ¢lled, presence of Tween 20; open, absence of Tween 20.Means þ S.D. for three incubations.

BBAPRO 35807 1-12-98

S. Eaton et al. / Biochimica et Biophysica Acta 1429 (1998) 230^238234

equate bu¡ering capacity was present. The resultswere quanti¢ed as the ratio:

�3-hydroxyhexadecanoyl-CoA� � �2-hexadecenoyl-CoA��tetradecanoyl-CoA� � �3-ketohexadecanoyl-CoA�

as a measure of the inhibition of the 3-hydroxyacyl-CoA dehydrogenase activity of the trifunctional pro-tein. This measure was used because the equilibriumof 2-enoyl-CoA hydratase means that inhibition ofthe 3-hydroxyacyl-CoA dehydrogenase is also re-£ected in accumulation of 2-enoyl-CoA; we haveused this measure previously as an indication of3-hydroxyacyl-CoA dehydrogenase activity in intactmitochondria [13,25]. In the absence of Tween 20,the ratio was relatively insensitive to [NAD�]/[NADH] in that inhibition was only seen when[NAD�]/[NADH] was W1 or less (Fig. 4). In thepresence of Tween 20, the ratio was more sensitiveto [NAD�]/[NADH], but only rose at [NAD�]/[NADH] of 2.5 or less. The ¢nding that the trifunc-tional protein is relatively insensitive to [NAD�]/[NADH] when total (NAD�+NADH) concentration

was ¢xed, if extended to the intact mitochondrion,would imply that our observation of 2-enoyl- and3-hydroxyacyl-CoA esters is not due to simple inhib-ition by low amounts of NADH, but is due to chan-nelling of NAD�/NADH between the 3-hydroxyacyl-CoA dehydrogenase and complex I of the respiratorychain (see [11,12]).

3.5. Investigation of glutamate dehydrogenase/2-oxoglutarate/ammonium acetate as analternative NAD+/NADH acceptor system

NAD-dependent dehydrogenases are of two types,A and B, such that NAD�/NADH can be channelledbetween the active sites of two dehydrogenases ifthey are of opposite speci¢city [26]. Pig heart L-lac-tate dehydrogenase is A-type and short-chain hy-droxyacyl-CoA dehydrogenase is B-type [27], butthe chiral speci¢city of human liver trifunctional pro-tein is unknown. We wished to study the e¡ects ofglutamate dehydrogenase as a B-type acceptor sys-tem [27] on the intermediates derived from 2-hexade-cenoyl-CoA. Incubations were carried out as de-

Fig. 4. E¡ect of varying [NAD�]/[NADH] on the activity of the trifunctional protein. The trifunctional protein was incubated with2-hexadecenoyl-CoA, lactate dehydrogenase and various [NAD�]/[NADH] ratios obtained by using varying ratios of lactate and pyru-vate. ([3-Hydroxyhexadecanoyl-CoA]+[2-hexadecenoyl-CoA])/([tetradecanoyl-CoA]+[3-ketohexadecanoyl-CoA]) was used as a measureof 3-hydroxyacyl-CoA dehydrogenase inhibition (see Section 3). R, Presence of Tween 20; F, absence of 0.2% (v/v) Tween 20. Resultsare typical for a single titration.

BBAPRO 35807 1-12-98

S. Eaton et al. / Biochimica et Biophysica Acta 1429 (1998) 230^238 235

scribed in Section 2, but with the addition of 10 mM2-oxoglutarate, 10 mM ADP, 125 WM pyruvate,0.1 M ammonium acetate and either 2 U/ml L-lactatedehydrogenase or 2 U/ml glutamate dehydrogenase,in both the presence and absence of Tween 20. Therewas no signi¢cant di¡erence between the two dehy-drogenase acceptors in terms of the intermediatesgenerated in the either the presence or absence ofTween 20 (results not shown). Flux was also identicalwith either acceptor. This indicates that under theconditions used, provision of an A- or B-type dehy-drogenase has no e¡ect on the forward activity of thetrifunctional protein, but it does not allow us tomake any assumption on the nature (A- or B-type)of the 3-hydroxyacyl-CoA dehydrogenase activity ofthe trifunctional protein.

3.6. E¡ect of [acetyl-CoA]/[CoA] ratio onproduction of CoA esters from2-hexadecenoyl-CoA

[Acetyl-CoA]/[CoA] was varied similarly to[NAD�]/[NADH] by the use of a dynamic bu¡ersystem, carnitine acetyl transferase/carnitine/acetyl-

carnitine. The ratios of [acetyl-CoA]/[CoA] were cal-culated using:

Keq � �acetyl-CoA��carnitine��CoA��acetylÿ carnitine�W0:6

[28]. Carnitine acetyl transferase is markedly pH sen-sitive such that its activity at pH 9.0 is one-sixth ofthat at pH 7.5 [29]. Accordingly, it was necessary touse an excess of carnitine acetyl transferase (1.2 U/incubation). The [acetyl-CoA]/[CoA] ratios at di¡er-ent [acetyl-carnitine]/[carnitine] ratios were veri¢edby HPLC as described in Section 2, both in the ab-sence and presence of trifunctional protein/2-hexade-cenoyl-CoA/NAD� to ensure that adequate bu¡eringcapacity was present. The results are shown graphi-cally in Fig. 5. In the presence of Tween 20 (i.e.3-ketoacyl-CoA thiolase activated), the overall ratewas essentially independent of [CoA]/[acetyl-CoA]at ratios above 0.5 at pH 7.5. At pH 9.0, however,the rate increased with increasing [CoA]/[acetyl-CoA], even between ratios of 5.0 and 15.0. In theabsence of Tween 20 the rate was very much lowerwith no dependence on [CoA]/[acetyl-CoA] above5.0. Hence under physiological conditions (pH 7.5)

Fig. 5. E¡ect of varying [acetyl-CoA]/[CoA] on the activity of the trifunctional protein. The trifunctional protein was incubated with2-hexadecenoyl-CoA and carnitine acetyl transferase and various [acetyl-CoA]/[CoA] ratios obtained using di¡erent concentrations ofcarnitine and acetyl-carnitine. R, presence of Tween 20; F, absence of Tween 20. Filled symbols, pH 9.0; open symbols, pH 7.5.

BBAPRO 35807 1-12-98

S. Eaton et al. / Biochimica et Biophysica Acta 1429 (1998) 230^238236

the 3-ketoacyl-CoA thiolase activity of the trifunc-tional protein is insensitive to [CoA]/[acetyl-CoA].Between [CoA]/[acetyl-CoA] in the range 0.1^0.5;however, there was a clear diminution in £ux, whichwas associated with a signi¢cant (P6 0.01) increasein the amount of 3-ketohexadecanoyl-CoA generated(from 1.9 þ 0.2 to 2.5 þ 0.3 nmol (mean þ S.D., n = 5).This result is similar to the ¢ndings previously re-ported, except that in the previous work the totalconcentration of acetyl-CoA plus CoA was varied[22]. The data reported here are similar, but morephysiologically relevant as the total [acetyl-CoA+CoA] was ¢xed. However, there is a discrepancywith the intact mitochondrion in that in that whenmitochondria from a variety of tissues have beenincubated with hexadecanoate, hexadecanoyl-CoAor hexadecanoyl-carnitine, we have never observedthe accumulation of 3-ketoacyl-CoA esters, evenunder conditions of maximal £ux or when acetyl-CoA has accumulated to a great extent [11^13].The reasons for this discrepancy are unknown, butit is possible that other CoA-dependent steps, such ascarnitine palmitoyl transferase II, are more sensitiveto a lowering of intramitochondrial CoA than thetrifunctional protein. This would ensure that underconditions in which acetyl-CoA cannot be disposedof, entry of further acyl groups to the mitochondrionis prevented.

4. Conclusions

The trifunctional protein of mitochondrial L-oxi-dation provides an ideal system to manipulate theCoA and NAD-dependent steps in a de¢ned way.The accumulation of the intermediates is readily de-tectable and quanti¢able. The detergent Tween 20stimulated 3-ketoacyl-CoA thiolase activity, but in-hibited 2-enoyl-CoA hydratase activity, suggestingthat the membrane environment of the trifunctionalprotein may be very important. Sensitivity of theprotein to [NAD�]/[NADH] was noted, but only at[NAD�]/[NADH] much lower than observed in iso-lated mitochondria oxidising hexadecanoate, suggest-ing further that our observation of 2-enoyl- and3-hydroxyacyl-CoA esters during pulses of L-oxida-tion in mitochondria from a variety of tissues wasnot due to simple inhibition at the level of gross

[NAD�]/[NADH]. The isolated trifunctional proteinwas also sensitive to [acetyl-CoA]/[CoA], which ledto the accumulation of 3-ketohexadecanoyl-CoA.However, as we have never observed 3-ketoacyl-CoA esters in incubations of isolated mitochondriaincubated with hexadecanoate, hexadecanoyl-CoA orhexadecanoyl-carnitine, there may be another intra-mitochondrial mechanism preventing accumulationof 3-ketoacyl-CoA esters.

Acknowledgements

The British Heart Foundation is gratefullythanked for a project grant to S.E. and K.B. and afellowship to S.E. supporting this work.

References

[1] H. Schulz, J. Biol. Chem. 249 (1974) 2704^2709.[2] M. El-Fakhri, B. Middleton, Biochim. Biophys. Acta 713

(1982) 270^279.[3] K. Carpenter, R.J. Pollitt, B. Middleton, Biochem. Biophys.

Res. Commun. 183 (1992) 443^448.[4] Y. Uchida, K. Izai, T. Orii, T. Hashimoto, J. Biol. Chem.

267 (1992) 1034^1041.[5] S. Jackson, K. Bartlett, J. Land, E.R. Moxon, R.J. Pollitt,

J.V. Leonard, D.M. Turnbull, Pediatr. Res. 29 (1991) 406^411.

[6] L. Ijlst, J. Ruiter, J. Hoovers, M. Jacobs, R.J.A. Wanders,J. Clin. Invest. 98 (1996) 1028^1033.

[7] S. Jackson, R.S. Kler, K. Bartlett, H. Briggs, L.A. Bindo¡,M. Pourfarzam, D. Gardner-Medwin, D.M. Turnbull,J. Clin. Invest. 90 (1992) 1219^1225.

[8] K. Izai, Y. Uchida, T. Orii, S. Yamamoto, T. Hashimoto,J. Biol. Chem. 267 (1992) 1027^1033.

[9] F.J. Ruzicka, H. Beinert, J. Biol. Chem. 252 (1977) 8440^8445.

[10] S. Eaton, K. Bartlett, M. Pourfarzam, Biochem. J. 320(1996) 345^357.

[11] S. Eaton, A.K.M.J. Bhuiyan, R.S. Kler, D.M. Turnbull, K.Bartlett, Biochem. J. 289 (1993) 161^168.

[12] S. Eaton, D.M. Turnbull, K. Bartlett, Eur. J. Biochem. 220(1994) 671^681.

[13] S. Eaton, M. Pourfarzam, K. Bartlett, Biochem. J. 319(1996) 633^640.

[14] S. Eaton, D.M. Turnbull, K. Bartlett, Biochem. Soc. Trans.22 (1994) 119S.

[15] N.J. Watmough, D.M. Turnbull, H.S.A. Sherratt, K. Bart-lett, Biochem. J. 262 (1989) 261^269.

[16] J. Schi¡erdecker, H. Schulz, Life Sci. 14 (1974) 1487^1492.

BBAPRO 35807 1-12-98

S. Eaton et al. / Biochimica et Biophysica Acta 1429 (1998) 230^238 237

[17] X. He, S. Yang, H. Schulz, Anal. Biochem. 180 (1989) 105^109.

[18] Y. Olowe, H. Schulz, Eur. J. Biochem. 109 (1980) 425^429.[19] H.Y. Wang, C.F. Baxter Jr., H. Schulz, Arch. Biochem.

Biophys. 289 (1991) 274^280.[20] C. Thorpe, Anal. Biochem. 155 (1986) 391^394.[21] J.C. Fong, H. Schulz, Methods Enzymol. 71 (1981) 390^398.[22] K. Yao, H. Schulz, J. Biol. Chem. 271 (1996) 17816^17820.[23] M. Luo, X. He, H. Sprecher, H. Schulz, Arch. Biochem.

Biophys. 304 (1993) 266^271.

[24] R. Lane, E. Dekker, Biochemistry 8 (1969) 2958^2966.[25] S. Eaton, A.M. Zaitoun, C.O. Record, K. Bartlett, Clin. Sci.

90 (1996) 307^313.[26] D. Srivastava, S. Bernhard, Science 234 (1986) 1081^1086.[27] K. You, Methods Enzymol. 87 (1982) 101^126.[28] I. Fritz, S. Schultz, P. Srere, J. Biol. Chem. 238 (1963) 2509^

2517.[29] J.F.A. Chase, Biochem. J. 104 (1967) 503^509.

BBAPRO 35807 1-12-98

S. Eaton et al. / Biochimica et Biophysica Acta 1429 (1998) 230^238238