Localization of malate dehydrogenase, adenylate kinase and glycolytic enzymes in glycosomes and the threonine pathway in the mitochondrion of cultured procyclic trypomastigotes of Trypanosoma brucei

Molecular and Biochemical Parasitology, 4 (1981) 291 - 309

291

Elsevier/North-Holland Biomedical Press

LOCALIZATION OF MALATE DEHYDROGENASE, ADENYLATE KINASE AND GLYCOLYTIC ENZYMES IN GLYCOSOMES AND THE THREONINE PATHWAY IN THE MITOCHONDRION OF CULTURED PROCYCLIC TRYPOMASTIGOTES OF

TR YPANOSOMA BR UCEI

FRED R. OPPERDOES, ANTON MARKOS* and ROLF F. STEIGER**

Research Unit for Tropical Diseases, International Institute o f Cellular and Molecular Pathology (ICP), avenue Hippocrate 74, B-1200 Brussels, Belgium (Received 12 June 1981; accepted 8 September 1981)

Procyclic culture forms of Trypanosoma brueei stock 427 have been screened for the presence of enzymes involved in glycolysis, mitochondrlal energy metabolism and threonine degradation. The enzyme activities in the procyclics were compared with those of the bloodstream forms. The specific activities of glycolytic enzymes represented 30-70% of the respective levels in the bloodstream form, except for hexokinase which was 25-fold reduced. Cell fractionation showed that the enzymes involved in the early sequence of the glycolytic pathway, i.e. from hexokinase to phosphoglycerate kinase, and the enzymes NAD+-linked glycerol-3-phosphate dehydrogenase and glycerol kinase were all present in glycosomes equil~rating at a density of 1.23 g/cm 3 in sucrose gradients. Malate dehydrogenase was 8-fold more active in procyclics than in bloodstream forms. This increase in activity was the result of the appearance of malate dehydrogenase in the glycosomes of the procyclics, in addition to mitochondrial and cell-sap activities which were present in both stages of the life cycle. Glycosomes contained part of the adenylate kinase activity, which was also associated with the mitochondrion. Suecinate dehydrogenase and sn-glycerol-3-phosphate dehydrogenase, together with oligomycinsensitive ATPase, were located in the mitochondrion which had a density in sucrose ranging from 1.16 to 1.18 g/cm 3 . This organelle also contained L-threonine 3-dehydrogenase and carnitine aeetyltransferase, two enzymes involved in threonine catabolism. The latter two enzymes had activities which were, respectively, 15-and 13-fold higher in the procyclics than in the bloodstream form. Mitochondrial sn-glycerol-3-phosphate dehydrogenase was decreased 4-fold.

* Present address: Department of Physiology, Charles University, Vini~ml 7, 12844 Prague 2, Czechoslovakia. ** Present address: Ciba-Geigy, AG 9.52, P.O. Box, CH-4002 Basel, Switzerland. Enzymes: Adenylate kinase (EC 2.7.4.3); ATPase (EC 3.6.1.3); carnitine acetyltransferase (EC 2.3.1.7); fructose-bisphosphate aldolase (EC 4.1.2.13); glucose-phosphate isomerase (EC 5.3.1.9); glyeeralde(EC 2.7.1.30); NAD +-linked sn-glycerolhyde-phosphate dehydrogenase (EC 1.2.1.12); glycerol kmase " " 3-phosphate dehydrogenase (EC 1.1.1.8); sn-glycerol-3-phosphate dehydrogenase (sn-glycerol-3-phosphate:(acceptor) oxidoreductase, EC 1.1.99.5); hexokinase (EC 2.7.1.1); L-malate dehydrogenase (EC 1.1.1.37): 6-phosphofructokinase (EC 2.7.1.11); phosphoglyeerate kinase (EC 2.7.2.3.); suecinate dehydrogenase (succinate: (acceptor) oxidoreductase, EC 1.3.99.1); L-threonine 3-dehydrogenase (EC 1.1.1.103); triose-phosphate isomerase (EC 5.3.1.1). 0166-6851/81/0000-0000/$02.75 © 1981 Elsevier/North-Holland Biomedical Press

292 Key words: Trypanosoma bruceL Procyclics, Cell fractionation, Glycolytic enzymes, Mitochondria, Glycosomes, Threonine pathway, Adenylate kinase, Malate dehydrogenase.

INTRODUCTION The life cycle of the African trypanosomes belonging to the subgenus Trypanozoon, the causative agent of Sleeping Sickness in Man and Nagana in cattle, is complex and characterized by gross morphological [1] and biochemical [2] changes. The two biochemically best studied stages in the life cycle are the long slender bloodstream trypomastigote, the mammalian parasite, and the cultured procyclic trypomastigote, representative of the insect-vector stage. Bloodstream forms of pleomorphic strains (as defined by Opperdoes et al. [3]) may differentiate to culture forms following cultivation at 27°C. Transformation from bloodstream to culture forms is accompanied by the development of the mitochondrion from a single promitochondrial tube into an extensively branched network with prominent plate-like cristae. Respiration, previously cyanide-insensitive, is now inhibited by cyanide and other classical inhibitors of the respiratory chain. Cytochromes, which were absent before, can now be detected [2]. This cultured procyclic trypomastigote is thought to reflect an adaptation to the environmental conditions as they exist in the midgut of the tsetse fly, where glucose and oxygen are limited and amino acids abundant. Indeed, these procyclics consume significant amounts of proline, glutamate and threonine, in addition to glucose [4], the sole energy substrate in the bloodstream form. Established culture forms, apart from some minor morphometric differences [5], are identical to the midgut stages in the vector and they retain the potential to accomplish a full life cycle when ingested by Glossina spp. [6]. Therefore, the metabolic capabilities of the cultured procyclic trypomastigote are believed to reflect the actual biochemical properties of the midgut forms. Cell fractionation has previously produced evidence that in the bloodstream form snglycerol-3-phosphate oxidase is located in the promitochondrion [7] and a number of glycolytic enzymes in microbodies [7, 8, 9], henceforth called glycosomes. The latter property is unique to Trypanosomatids [8, 10]. We have also described marker enzymes for the plasma membrane, the flagellar pocket, lysosomes [ 11 ] and the mitochondrion [7] of the bloodstream form. So far, however, only little is known about the localization of vital enzymes and pathways in procyclic trypomastigotes and whether their localization in the respective subcellular compartments is influenced by the transformation process. In order to get a better understanding of the biology of the organism and the mechanisms involved in transformation we have initiated cell fractionation experiments on the culture form of T. brucei derived from bloodstream T. bruceL stock 427, used previously. In this paper the enzymes present in the mitochondrion and the glycosome are examined; in the accompanying paper [12] hydrolytic enzymes are studied. A brief summary of our main conclusions has been published elsewhere ([ 13 ]; F.R. Opperdoes quoted in [14]).

293 MATERIALS AND METHODS

Organism. Trypanosoma brucei was obtained as procyclic trypomastigotes from Dr. Reto Brun (Swiss Tropical Institute, Basel). It is a derivative of T. (T.)brucei, stock 427 [ 15], cultures of which had been initiated from an established midgut infection of Glossina morsitans. Procyclics were grown at 27°C in a semi-defined synthetic medium (SDM-79) in the presence of 100 U penicillin/ml, essentially as described elsewhere [16]. Cells were harvested in the late log phase of growth (cell density 2 × 10 7 /ml) and thoroughly washed by centrifugation and resuspension in isotonic homogenization medium (3 mM imidazole-HC1, 0.25 M sucrose and 1 mM EGTA, pH 7.0). Enzyme determinations. Glycolytic enzymes [8], glycerol kinase [8], mitochondrial snglycerol-3-phosphate dehydrogenase [17], malate dehydrogenase [17] and adenylate kinase [ 18] were measured according to published procedures. Total ATPase activity in the fractions was measured according to Pullman and Penefsky [19] and liberated inorganic phosphate was determined according to the method of Fiske and Subbarow as described by Sumner [20]. The difference between the total ATPase activity and the activity resistent to 10/~g oligomycin/ml represents the oligomycin-sensitive or mitochondrial ATPase [ 17]. Mitochondrial sn-glycerol-3-phosphate dehydrogenase and succinate dehydrogenase were measured in reaction mixtures containing 50 mM Tris/HC1, 0.08 mM dichlorophenol indophenol and 0.5 mM phenazine methosulfate, pH 7.0. The reaction was started by the addition of 20 mM D,L-glycerol-3-phosphate or sodium succinate. L-Threonine dehydrogenase activity was measured by following the reduction of NAD ÷ at 340 nm at 25°C. The assay mixture contained (mM): NAD ÷, 2.5; Tris-HC1 pH 8.6, 250; L-threonine, 120; KC1,250; enzyme preparation, 50/lg in a final volume of 1.0 ml. Carnitine acetyltransferase was followed spectrophotometrically at 324 nm by using Aldrithiol-4(4,4'-dipyridyl disulfide) to measure the production of the free thiol group from coenzyme A at 25°C. The assay mixture contained 0.1 M Tris-HC1 (pH 8.0), 0.1 mM acetyl-coenzyme A, 0.1 mM 4,4'-dipyridyl disulfide and 50/~g enzyme preparation in a f'mal volume of 1.0 ml. The reaction was initiated by the addition of 10 mM D,Lcarnitine. Protein was determined fluorometrically [22] after diluting and solubilizing the extracts in 0.2 M carbonate-buffered 1% deoxycholate (pH 11.3) using bovine serum albumin as standard. Cytochrome spectra. Difference spectra of the cytochromes were recorded on intact cells and submitochondrial particles with an Aminco-Chance DW-2 double-beam dual-wavelength spectrophotometer at room temperature as well as after cooling the samples with liquid nitrogen. Submitochondrial particles were prepared by treating mitochondria with

294 hypotonic morpholinopropane sulphonate buffer, essentially as described by Kusel and Storey [21 ].

Cellfractionation. Homogenates of T. bruceiculture forms were prepared and processed according to the previous protocol [ t l ] , except that density equilibration of cell extracts on sucrose gradients was performed in a Beckman VTi 50 vertical rotor in a L 5-50 preparative ultracentrifuge, equipped with a slow acceleration accessory. Loading of the polyallomer centrifuge tubes was done by the successive addition of a 3-ml cushion of sucrose with a density of 1.34 g/cm 3 and 27 ml of a linear sucrose gradient (p = 1.071.25 g/cm 3) prepared with a Beckman gradient former. Both sucrose solutions contained 25 mM Tris/HC1 and 1 mM EDTA, pH 7.2. Finally, a post large-granule extract in homogenization medium (supernatant after 10 min centrifugation at 5000 X g)containing a total of approximately 50 mg protein in 5 ml, was layered on top of the gradient. The rotor was slowly accelerated to 1400 rev./min and then rapidly to 48 000 rev./min (190 000 X gay)" After 150 min the rotor was decelerated, with the brake, to 1400 rev./min and then with the brake off, to rest. The gradient was collected by suction from the bottom into 12-13 fractions of equal volume (2.5-3.0 ml). Protein and density were measured immediately thereafter before the fractions were stored undiluted in liquid nitrogen until analysis. The distribution patterns of the differential and density-gradient centrifugation are presented as before [ 11 ]. Using post large-granule supernatants, instead of a total cellular extract (supernatant of 10 min at 1000 × g), for isopycnic centrifugation resulted in more reproducible distribution prof'des, probably because of the absence of artefactually produced 'baskets' of plasma membrane attached to subpellicular microtubules. These 'baskets' contain all subcellular constituents except the cytosol and tend to equilibrate at densities between 1.20 and 1.25 g/cm 3. Because of this procedure the amounts of particulate material and especially of the mitochondria are underestimated with respect to soluble material in the distribution profiles. Latency experiments. For the determination of the degree of latency of a number of particle-bound enzymes a washed preparation of a combined large- and small-granule fraction (material sedimentable between 1000 and 14 500 × g for 10 min) was used Free activity was always determined directly in an isotonic enzyme-assay mixture to which 0.25 M sucrose was added. Total activity was measured under identical conditions but after addition of 0.2 ~ Triton X- 100.

Materials. All chemicals were of the highest purity available. Bovine serum albumin (crystallized) was purchased from Miles Laboratories, oligomycin and phenazine methosulfate were from Sigma, enzymes, substrates and co-substrates from Boehringer (Mannheim), 2,6-dichlorophenol indophenol sodium salt from Merck, antimycin A from Serva (Heidelberg) and salicylhydroxamic acid and Aldrithiol-4 from Aldrich.

295 RESULTS

Screening for enzymes. Since cultured procyclic trypomastigotes derived from Trypanosoma brucei stock 427 have never been characterized with respect to their mitochondrial activity, we first screened this organism for the presence of cytochromes and mitochondrial enzymes. In thoroughly washed intact cells, as well as in submitochondrial particles, cytochromes were detected in reduced minus oxidized difference spectra. In intact cells at room temperature (Fig. 1C) cyt aa3 absorbed at 595 run as described for Leishmania tropica [23] and cyt b had a maximum at 563 nm with a shoulder towards shorter wavelengths due to the presence of cytochromes c + cl. The contribution of these c cytochromes becomes especially apparent in the low-temperature spectrum of submitochondrial particles (Fig. 1B), when cyt b is shifted towards shorter wavelengths, whereas cyt csss is not [21 ]. A contribution of cyt cl at 552 nm can also be seen. Haemin, 428

~

s~7 A

A Abs J-

Ts

0.01 T

563

595

8

I

400

I

I

I

500 WAVELENGTH

I

I

600 (nm)

Fig. t. Difference spectra from T. brucei procyclic trypomastigotes. A. Low-temperature (77°K) spectrum of intact, thoroughly washed, cells (20.5 mg protein/ml) reduced with Na2S~O 4 minus oxygenated reference. B. Low-temperature spectrum of sub-mitochondrial particles reduced with Na2S20 , minus oxygenated reference. C. Room-temperature spectrum of intact cells reduced with 1 mM potassium cyanide m i n u s oxidized (ferricyanide) reference.

296 which might contaminate intact cells does not interfere with the spectra as can be seen from a comparison of Figs. 1A and B. Endogenous respiration of the cells (18 nmol 02 • mg protein -~ • min -~) could not be stimulated by any of the following: glucose, glycerol, succinate or proline (5 mM each). Respiration was 6 0 ~ inhibited by 5 mM KCN or 0.i mM salicylhydroxamic acid (SHAM). Total inhibition was obtained in the presence of both inhibitors. Antimycin (43 #g/mg protein) inhibited respiration by 84%. This value could be enhanced by the addition o f SHAM but not by KCN. These results suggest the presence of a functional electron-transport chain with two terminal oxidases: cytochrome c oxidase, which is sensitive to cyanide, and an alternative oxidase, sensitive to SHAM. These results are similar to those described for other stocks of culture forms of

T. brucei [24, 25]. Succinate dehydrogenase, the FAD-containing sn-glycerol-3-phosphate dehydrogenase (EC 1.1.99.5) and the oligomycin-sensitive ATPase, henceforward referred to as mitochondrial ATPase, are constituents of the mitochondrial respiratory chain of eukaryotes and have been localized as such in Trypanosomatids [3, 7, 25]. Therefore, these enzymes were selected as markers for the mitochondrial inner membrane. These activities are shown in Table I. TABLE I Specific activities of enzymes in Trypanosoma brucei bloodstream and procyclic trypomastigotes. Enzyme

Hexokinase Glucose-P-isomerase Phosphofructokinase Triose-P isomerase Glycerol-3-P dehydrogenase [NAD÷] Glycerol kinase Phosphoglycerate kinase Malate dehydrogenase Adenylate kinase ATPase (total) ATPase (oligomycin-sensitive) Succinate dehydrogenase Mitochondrial glycerol-3-P dehydrogenase Glycerol- 3-P oxidase Threonine dehydrogenase Carnitine acetyltransferase a b c d

From Opperdoes and Borst [8]. From Opperdoes, Borst and Spits [17]. Steiger and Opperdoes (unpublished data). n.d. not detectable [171.

Specific activity (#mol • min-1 • mg-1 )

Ratio procyclic

Bloodstream

Procyclic

bloodstream

0.79 a 0.54 a 1.63a 1.18a 0.73 a 2.60 a 0.43 a 0.20 b 0.17c 0.07 b 0.02 b n.d. d 0.05 c 0.09 a 0.004 c 0.06 c

0.03 0.18 0.68 0.60 0.54 1.51 0.25 1.62 0.19 0.10 0.04 0.01 0.01 0.06 0.75

0.04 0.33 0.42 0.51 0.74 0.58 0.58 8.1 1.1 1.4 2.0 0.2 15 13

297 Monoamine oxidase, a mitochondrial outer-membrane marker, could not be found with a detection limit of 8 pmol • min -1 . mg protein _1. We have already reported that this enzyme is also absent from the bloodstream form [17]. Based on preliminary studies the enzymes adenylate kinase and malate dehydrogenase were selected as putative markers of the inter-membrane space and the matrix space, respectively. The latter enzyme was more active in culture than in bloodstream forms. All the glycolytic enzymes, including glycerol kinase and the NAD+-linked sn-glycerol3-phosphate dehydrogenase (EC 1.1.1.8) previously found in glycosomes in the bloodstream form, were detected in homogenates of procyclic trypomastigotes (Tables I and III). Their activities were only slightly reduced with respect to the corresponding bloodstream form activities, with one exception: hexokinase was decreased by 96 % Glyceraldehyde-phosphate dehydrogenase and fructose bisphosphate aldolase lost their activity after freezing and thawing. Therefore, these two enzymes were not included in the cellfractionation experiments. Threonine dehydrogenase and carnitine acetyltransferase are two enzymes of the threonine catabolic pathway. Linstead et al. [26] showed that the activity of this pathway was much higher in the culture than in the bloodstream form. Table 1 confirms these f'mdings. As in the bloodstream form of T. bruceL catalase could not be detected and the glucose-6-phosphate dehydrogenase activity was low (26 nmol • min -~ • mg -1 protein). This was an unexpected result, since another insect Trypanosomatid Crithidia luciliae contains large amounts of both enzymes [14, 27].

Cell fractionation by differential centrifugation. As can be seen in the histograms (Fig. 2) the enzymes hexokinase, triosephosphate isomerase and glycerol kinase behaved in the procyclics as if they were entirely associated with sedimentable material. The other glycolytic enzymes showed similar distributions over the different fractions (Table II), with the exceptions that glucose-phosphate isomerase and phosphofructokinase had somewhat higher activities in the soluble fraction and phosphoglycerate kinase appeared to be almost entirely soluble. Malate dehydrogenase, the activity of which was 8-fold increased as compared to the bloodstream form (Table I), showed a distribution strikingly similar to that of a number of the glycolytic enzymes (Fig. 2, Table II). The mitochondrial inner-membrane markers succinate dehydrogenase and sn-glycerol3-phosphate dehydrogenase were almost completely recovered in the particulate fractions with some 40% in the nuclear fraction. This is in agreement with the fact that We are dealing with one single mitochondrial network, which probably cannot easily be released from the cell ghosts during homogenization. Adenylate kinase also behaved as a particlebound entity. However, in contrast with the other mitochondrial enzymes, a significant part was associated with membrane vesicles (microsomal fraction). Threonine dehydrogenase and carnitine acetyltransferase showed identical bimodal distributions: they were partly soluble and partly particle-bound.

298 Hexokmase

Threonine

Succmate

4EG,.o,k,n_ Carn,,, acetyltransferase

o

LL Tri°seoh°sphate Glycerol-3phosohate Phospho(/~~(' 42 f W ~ dehydrogenase glycerata kinase

[a.__. IJ u.I rr 4 I- Malatedehydrogenase

Adenylate kinase

o I

I

0

50

I

100

I

I

0

50

PROTEIN

I

I

lOO o (~)

l

I

5o

lOO

Fig. 2. Distribution of enzymes from procyclics of T. brucei in fractions obtained by differential centrffugation. The fractions are plotted in the order of their isolation; from left to right: nuclear (N), large-granule (LG), small-granule (SG), mierosomal (M) and final supernatant (S). For the number of experiments and the averaged recoveries, see Table II.

Density equilibration. The averaged distribution proCdes of protein and a number of enzymes present in post large-granule extracts are shown in Fig. 3. Averaged modal densities are presented in Table III. Hexokinase, glucosephosphate isomerase, triosephosphate isomerase, NAD*-linked glycerol-3-phosphate dehydrogenase and glycerol kinase all formed sharp, almost symmetrical peaks, with a modal density of 1.23 g/cm 3 . In addition, some glucosephosphate isomerase activity was soluble and remained at the top of the gradient. Phosphofructokinase also peaked at a density of 1.23 g/cm 3 , but significant amounts of soluble and membrane-bound activity were found at lower densities. The distributions presented here for the glycolytic enzymes are essentially identical to those reported for the glycosomal enzymes of bloodstream form trypomastigotes [8]. It is therefore concluded that glycosomes also exist in cultured procyclics. By contrast, the bulk of the phosphoglycerate kinase was soluble and had an apparent 'median density' of 1.088 g/cm a (not shown). A behaviour remarkably similar to that of the glycosomal enzymes is exhibited by malate dehydrogenase with a prominent peak at p = 1.23; soluble and mitochondrial activities are also present.

19 18 22 17 26 17 32 5 31 35 26 15 47 37 25 22

11 20 17 22 22 18 14 2 21 14 23 28 27 24 27 28

LG 8 32 26 17 29 28 20 0 19 14 17 33 14 13 13 15

SG

Distribution in the fractions (%) N 19 28 21 29 21 26 31 2 18 31 14 23 8 14 2 4

M 43 2 14 15 2 1 3 91 11 6 20 1 4 12 33 31

S 105 78 93 97 103 93 97 115 91 144 98 62 116 125 99 92

Recovery (%)

3 3 2 3 3 3 2 1 3 3 2 2 2 3 2 3

No. o f experiments

Distributions are expressed as percentages of the sum of 5 fractions. Recoveries give the sum of the activities in the fractions compared to the value of the cytoplasmic extract + nuclear fraction, as a percentage. The values represent the means of multiple experiments. N, nuclear; LG, large granular; SG, small granular; M, microsomal; S, final supernatant.

Protein Hexokinase Glucose-P isomerase Phosphofructokinase Triose-P isomerase Glycerol-3-P dehydrogenase [NAD ÷1 Glycerol kinase Phosphoglycerate kinase Malate dehydrogenase Adenylate kinase ATPase (total) ATPase (oligomy cin-sensitive) Succinate dehydrogenase Mitochondrial glycerol- 3-P dehydrogenase Threonine dehydrogenase Carnitine acetyltransferase

Enzyme

Fractionation of T. brucei procyclic trypomastigotes by differential centrifugation,a

TABLE II

tO

300

ldehydrog~nase (3)

25 _ Hexokinase [

Succinate

Threonine dehydrogenase (3)

i

20

10

15

5

10 AT Pase ~oligomycin t sensitive ~ (2)

5 0 25 .

0 15

0

Carnitine acetyl tranferase (3)

15 10

Glucosephosphate isomerase 3)

j

>- 20 0 z ILl 15 ~3 0 LLI rr 10 5

15

5

Glycerol- 3- phosphate dehydrogenase ( 3 )

Malate dehydrogenase ( 3 )

0 -

15 10

J•

5

k Phosphofructokinase _ ~ 13)

Adenylate kinase (3)

0

Protein ( 3 )

15 10

10

5

5 i

0 1.05

1.15

L

1.25

0 1.05

1.15

DENSITY

1.25

1.05

1.15

1.25

( g/cm 3 )

Fig. 3. Distribution profiles of post large-granule extracts of homogenates from T. brucei procyclics after isopycnic centrffugation on linear sucrose gradients. Histograms represent means of multiple experiments + standard deviations. The number of determinations of each enzyme is given in parentheses. Averaged recoveries of post large-granule material from the gradient ranged between 74 and 130% for protein and all enzymes.

The mitochondrial enzymes succinate dehydrogenase and oligomycin-sensitive ATPase showed identical distributions in the gradients with modal densities at p = 1.18, typical o f the T. brucei mitochondrion [7]. The profile o f total ATPase (not shown, cf. Table III) was almost identical to that o f the mitochondrial ATPase, except for a minor contribution o f an oligomycin-insensitive activity at densities around 1.1 1 - 1 . 1 5 g/cm 3. This activity could be attributed to phosphatase, possibly associated with membranes derived from the trypanosomes digestive apparatus (cf. accompanying paper [ 12]). The FAD-containing sn-glycerol-3-phosphate dehydrogenase, the other mitochondrial

301 TABLE III Densities of particulate enzymes, a Enzyme

Modal density (g/cm 3)

Hexol~inase Glucose-P isomerase Phosphofructokinase Triose-P isomerase Glycerol- 3-P dehydrogenase [NAD+I Glycerol kinase Malate dehydrogenase ATPase (total) ATPase (oligomy cin-sensitive) Succinate dehydrogenase Mitochondrial glycerol-3-P dehydrogenase Mitochondrial adenylate kinase Glycosomal adenylate kinase Threonine dehydrogenase Carnitine acetyltransferase

1.228 -+0.004 1.228 _+0.004 1.229 _+0.003 1.224 _+0.006 1.228 _+0.004 1.236 1.228 +_0:010 1.170 _+0.010 1.177 _+0.001 1.177 ± 0.007 1.159 +_0.010 1.166 +-0.007 1.230 -+0.007 1.180 _+0.003 1.181 +_0.004

(3) (3) (3) (3) (3) (1) (3) (2) (2) (3) (3) (3) (3) (3) (3)

a Values listed are means +_standard deviation. The number of experiments is given in brackets. enzyme, was slightly, but significantly, shifted towards a lower density when compared to succinate dehydrogenase and mitochondrial ATPase (peak at p = 1.16); possible explanations for this discrepancy will be discussed later. The distribution o f adenylate kinase is complex and activity was found in three regions in the gradient: at the position of soluble protein and at mitochondrial and glycosomal densities. Threonine dehydrogenase and carnitine acetyltransferase again showed identical patterns. Both enzymes have a soluble and a particulate activity, of which the latter is equilibrating at mitochondrial density.

Latency of enzymes. In subcellular fractions derived from the bloodstream form the glycosomal enzymes exhibit a high degree of latency ( 7 0 - 9 0 % ) [ 2 8 ] . Therefore, we have checked whether in washed particle preparations of cultured procyclics latency could be detected. Latencies ranging from 76 to 95% were found for all the glycosomal enzymes tested, including phosphoglycerate kinase (Table IV). However, the specific activity of the latter enzyme, as compared with most of the others, was very low. These results indicate that, although the majority of phosphoglycerate kinase is present in a soluble non-latent form outside the glycosomes, these organelles still contain some kinase activity which, analogously to the other enzymes tabulated in Table IV, is bounded by the glycosomal membrane. In similar preparations malate dehydrogenase was also latent (95%, Fig.4). This together with an equilibrium density of 1.23 g/cm a in sucrose suggests a localization inside microbodies. On the basis of its activation by Triton X- 100 or by freezing and thawing,

302

TABLE IV Latency of glycosomal enzymes as measured in a washed large-granule fraction, a Enzyme

Latency (%)

Total specific activity (#mol • min -1 •mg protein -1 )

Hexokinase 6- Phosphofructokinase Fructose-bisP aldolase Triose-P isomerase Glycerol-3-P dehydrogenase [NAD*I Glycerol kinase Glyceraldehyde-P dehydrogenase P-glycerate kinase

87 76 83 77 93 94 95 82

0.08 0.32 0.03 0.43 1.49 0.76 0.09 0.02

Free activity and total activity were measured in a fresh preparation in the absence and presence of 0.2% Triton X-100, respectively. Latency is expressed as (total activity-free activity)/(total activity) × 100.

malate dehydrogenase compares with s n - g l y c e r o l - 3 - p h o s p h a t e dehydrogenase, a n o t h e r NAD÷-linked e n z y m e k n o w n to be present in glycosomes. This finding is likely to reflect a glycosomal l o c a t i o n o f malate dehydrogenase in procyclics. DISCUSSION

Bloodstream versus procycfic trypomastigote. The b l o o d s t r e a m form o f T. brucei is entirely d e p e n d e n t on c a r b o h y d r a t e m e t a b o l i s m for energy. These t r y p a n o s o m e s break 100

100

/ 50

0

0

/./4

o/"

50

/O

"6 / 0

i 0

0.05 Triton X-IO0 (%)

/ • /0

0.10

/c ~ .._..._.0 ~ 0

~

1

2

3

4

Cycles

Fig. 4. Comparison of the activation of NAD+-linked sn-glycerol-3-phosphate dehydrogenase and malate dehydrogenase by different treatments. A. The effect of increasing concentrations of Triton X-100. B. The effect of successive cycles of freezing and thawing. Total activity was measured in both cases after preincubation with 0.2% Triton X-100 for 15 min. o, NAD+-linked sn-glycerol-3-phosphate dehydrogenase; e, malate dehydrogenase.

303 down glucose to pyruvate, which is not degraded further, and oxidize their NADH via a cyanide-insensitive mitochondrial sn-glycerol-3-phosphate oxidase not coupled to oxidative phosphorylation [29]. The energy requirements are met by synthesis of glycolytic ATP only. A number of the enzymes involved in the Embden-Meyerhof pathway of glycolysis - from hexokinase to phosphoglycerate kinase - are located in glycosomes [8]. This contrasts sharply with the situation in other eukaryotes where all the glycolytic enzymes are present in the cytosol. Although the bloodstream form possesses a mitochondrion, the latter is metabolically almost inactive since oxidative phosphorylation, cytochromes and a functional Krebs' cycle are absent [30]. During transformation a metabolic switch occurs, which is paralleled by mitochondrial development and activity. Cytochromes are synthesized, respiration becomes sensitive to the classical inhibitors of the mitochondrial respiratory chain and Krebs' cycle enzymes are acquired. The established culture forms rely on keto- and amino-acid oxidation linked to the phosphorylation of ADP at the level of the mitochondrial ATPase. Our cultured procyclics of T. brucei show all the biochemical characteristics of a typical culture form as described above. They have an active respiratory chain containing the cytochromes aaa, b and c+cl and two terminal oxidases: cytochrome c oxidase (cyt aa3) which is inhibited by cyanide and a SHAM-sensitive oxidase. The cyanide- and SHAM-sensitive pathways contribute equally to the total electron flow since both inhibitors blocked endogenous respiration by the same amount (60%). The fact that endogenous respiration could not be stimulated by external substrates indicates that, in contrast to the bloodstream form, the procyclic contains large amounts of endogenous substrates. Succinate dehydrogenase, which cannot be detected in the bloodstream trypomastigote, is present in the procyclic form and the mitochondrial ATPase has doubled its specific activity. The relative unimportance of glycolysis in the procyclics is reflected by a drastic decrease in hexokinase activity. We return to this point below.

The enzymes involved in glycolysis. The cell-fractionation experiments presented in this paper clearly show that procyclics possess microbody-like organelles which contain a number of glycolytic enzymes (i.e. hexokinase, glucose-phosphate isomerase, phosphofructokinase, fructose-bisphophate aldolase, glyceraldehyde-phosphate dehydrogenase, triose-phosphate isomerase and phosphoglycerate kinase) together with two enzymes involved in glycerol metabolism: glycerol kinase and NAD÷-linked sn-glycerol-3-phosphate dehydrogenase. We have designated these organelles 'glycosomes'. Except for phosphoglycerate kinase the above enzymes are almost exclusively present in" the glycosomes. However, it is difficult to assess whether soluble activity of the others represents a true cell-sap enzyme or results from broken organdies. To illustrate this, glucosephosphate isomerase is a glycosomal matrix enzyme, whereas the other enzymes are firmly particle-bound [31, 32]. This might explain why the isomerase shows a rather high percentage in the supematant fraction. On the other hand, the majority of glucosephosphate isomerase in the insect trypanosomatid Crithidia luciliae was found soluble [14], which hints at a separate cytosolic isoenzyme.

304 We have not studied the distribution of glyceraldehyde-phosphate dehydrogenase, because of its labile nature. However, in the bloodstream form 50% of its activity is present in the cytosol and 50% in the glycosomes [8, 11, 28]. Some of the phosphofructokinase activity seems to be associated with mitochondria. Such a localization has also been reported in the case of another protozoan: Tetrahymena pyriformis [33]. The function of such a mitochondrial phosphofruc~okinase remains to be elucidated. The distribution of phosphoglycerate kinase is very different from those of the other glycosomal enzymes. It behaves as a typical cell-sap enzyme, whereas in the bloodstream form 75% or more is associated with glycosomes. Only 2% of the total activity was detectable in the combined procyclic LG and SG fractions but was highly latent, suggesting a glycosomal activity. Apparently bloodform-procyclic transformation is accompanied by a drastic shift of this kinase from the glycosome towards the cell-sap. In this respect it is of interest to mention that an intermediate situation, with equal amounts in both compartments, was encountered in Crithidia luciliae [ 14]. The role of such a cytosolic enzyme in Trypanosomatids is not yet clear. Together with cytosolic glyceraldehyde-phosphate dehydrogenase it may regulate the interrelated redox state and phosphate potential of the cell [34]. The fact that the glycolytic enzymes in the subcellular fractions from procyclics still exhibit high latencies, even though the specific activity of some of the enzymes in the glycosomes are drastically reduced, indicates that the membrane in itself is impermeable to the metabolites and co-factors of glycolysis (cf. [31 ])and that latency does not result from a limited rate of diffusion of substrates into the organelle, a possibility not ruled out by previous experiments [28]. We have not tested for the presence of the glycolytic enzymes involved in the ultimate steps of the pathway (i.e. phosphoglycerate mutase, enolase and pyruvate kinase), but the fact that procyclics of T. brucei consume glucose [4] and incorporate radioactivity from [U -14 C]glucose into alanine, malate, fumarate and succinate is suggestive of the presence of a complete functional glycolytic pathway. It is generally accepted that the insect stages of the trypanosomes consume glucose only at low rates. This holds also true of the cultured procyclic of T. bruceL Since there is no evidence for the presence of important regulatory mechanisms in the glycolytic pathway [35, 36] one of the enzymes must catalyse a rate-limiting step. From Table I it can be inferred that hexokinase is the most likely candidate, all the more so since it is the first enzyme in the pathway. Yet, it follows from Table IV that aldolase and the glycosomal phosphoglycerate kinase could also act in a similar way. At present it is not possible to pin-point the rate-limiting step as the kinetic properties of the glycolytic enzymes, as well as the intraglycosomal metabolite and co-factor concentrations, remain to be established. Glycosomes have now been found in bloodstream [8] and procyclic trypomastigotes, as well as in Crithidia luciliae [7, 14, 37]. Taylor et al. [10] found evidence for the presence of glycosomes in Chrithidia fasciculata and in all developmental stages of Trypanosoma cruzi; more recently the presence of glycosomes in Leishmania mexicana has been re-

305 ported [38]. Therefore, we conclude that glycosomes are a general feature of all Trypanosomatids. The mitochondrial markers. The fully developed mitochondrial network of the procyclics is a complicated structure which consists of a kinetoplast region associated with the basal body of the flagellum and an elaborated network of tubes protruding from the kinetoplast into all parts of the cytoplasm. Therefore, homogenization of the cells by grinding with an abrasive may result in fragmentation of a single mitochondrion into a number of mitochondrial particles. The fact that most of the activity of the inner-membrane enzymes, succinate dehydrogenase and sn-glycerol-3-phosphate dehydrogenase, is recovered in the nuclear fraction (Fig. 2) can be explained in two ways: the mitochondria that sediment at low speed have been released from the cell intact and are therefore, very big; or alternatively, the highly sedimentable mitochondrial material represents mainly membranes, still attached to the cell ghosts via the basal body of the flagellum. The peripheral branches of the mitochondrion that have been pinched off sediment at higher g values. Such damaged mitochondria could partly be devoid of matrix material. Succinate dehydrogenase, sn-glycerol-3-phosphate dehydrogenase and the mitochondrial ATPase were almost completely sedimentable and equilibrated at a density typical of mitochondria (p = 1.16-1.18). Although succinate dehydrogenase and sn-glycerol-3phosphate dehydrogenase exhibited identical sedimentation properties, the latter dehydrogenase banded at a significantly lower density than succinate dehydrogenase and mitochondrial ATPase. Such difference in density might indicate the presence of two parallel, spatially separated, respiratory chains rather than a branched chain as proposed by Njogu et al. [25]. One chain would then contain succinate dehydrogenase, ATPase and cytochromes, be capable of oxidative phosphorylation, and be typical of, and fully expressed only, in the well-established mig-gut stage. The other chain would be the sn-glycerol-3phosphate oxidase, containing a dehydrogenase and the SHAM-sensitive oxidase. This chain would be involved rather in the reoxidation of glycolytically reduced equivalents and be of the bloodstream type. Three possibilities can be put forward to explain the presence of such parallel chains. (A) An additional location of sn-glycerol-3-phosphate dehydrogenase could exist in membranes not belonging to the mitochondrion. These membranes would equilibrate at lower density. (B) In the late-log phase of growth the cell population is not homogenous. Some cells rely still on amino acid oxidation and, therefore, on oxidative phosphorylation, whereas others have already increased their level of glucose consumption. This could result in two populations with slightly different mitochondria: one enriched in ATPase and succinate dehydrogenase, equilibrating at p = 1.17, and the other enriched in sn-glycerol-3-phosphate dehydrogenase and SHAMsensitive oxidase, equilibrating at p = 1.16. (C) The distribution of enzymes in the mitochondrial inner-membrane could, in itself, be heterogeneous as, from a morphological point of view, is the case for the kinetoplast DNA, which is concentrated in one specialized region of the mitochondrion. This could result in one part of the mitochondrion specialized in the reoxidation of cytosolically-produced reducing equivalents, and another

306 part involved in oxidative phosphorylation. Fragmentation of such a mitochondrion could lead to the production of different classes of mitochondrial particles. At present is is not possible to distinguish between these three alternatives.

Malate dehydrogenase. In bloodstream forms, malate dehydrogenase is present in the cytosol (90%) and in the mitochondrion (10 %). No evidence was found for its presence in glycosomes [7]. This paper shows that, although cytosolic and mitochondrial components of malate dehydrogenase exist in the procyclics, the majority is now found in glycosomes. We have calculated that the 8-fold increase of malate dehydrogenase upon transformation is entirely due to an induction of the glycosomal activity. This is in perfect agreement with a recent observation by Kilgour [39] that culture forms of T. brucei reveal a malate dehydrogenase isoenzyme in addition to the one already present in the bloodstream form. However, her interpretation was that the appearance of this isoenzyme reflected mitochondrial development and the acquisition of the full set of Krebs' cycle enzymes, rather than the induction of a glycosomal congener. In our opinion this entity would be a suitable biochemical marker in transformation studies. The glycosomal isoenzyme could play a role in the reoxidation of glycolytically produced NADH, by replacing, in part, the dihydroxyacetone phosphate glycerol-3-phosphate shuttle. The mitochondrial sn-glycerol3-phosphate dehydrogenase activity in the procyclics has dropped by 5-fold to 0.01 /amol.min -~ .rag -~ . This is even below the activity of hexokinase, the rate-limiting step of glycolysis (Table I). This suggests that the mitochondrial respiratory chain might not be able to cope with all the reducing equivalents produced in the glycosome. A reoxidation of surplus of NADH via the glycosomal malate dehydrogenase might, therefore, solve this problem. The malate so produced is then further reduced to succinate. Trypanosomes, indeed, excrete succinate when they are grown in the presence of glucose [2, 4]. Adenylate kinase. In mammalian cells adenylate kinase is located between the inner and outer membranes of the mitochondrion [40] and it fulf'dls a role in regulating the cytosolic adenine nucleotide levels. It was, therefore, not unexpected to find mitochondrial adenylate kinase activity. The variable amounts of soluble activity found could have been released from the organelle upon rupture of the outer membrane. A significant amount of activity equilibrates at 1.23 g/cm a . This suggests an additional glycosomal location of adenylate kinase. Since glycosomes contain a number of phosphotransferases, the glycosomal adenylate kinase may prevent complete depletion of glycosomal ATP and hence inhibition of glycolysis. This regulation would be of vital importance, especially to the bloodstream form. Therefore, we have also checked for the presence of glycosomal adenylate kinase in the bloodstream form and found indeed, an identical subcellular distribution of this enzyme (unpublished results). This confirms a recent finding of adenylate kinase in the glycosome of Trypanosoma rhodesiense [41]. The threonine pathway. T. brucei consumes L-threonine [4] and converts this into equimolar amounts of acetate and glycine [26]. This pathway requires NAD+-linked

307 threonine dehydrogenase [42], coenzyme A-requiring aminoacetone synthase and acetylCoA hydrolase activity [43], which probably results from the concerted action of carnitine acetyltransferase and acetylcarnitine hydrolase. The flux through this catabolic pathway is higher in the proeyclic than in the bloodstream form and is associated with a manifold increase of the threonine dehydrogenase activity. In the cultured procyclics the acetyl-CoA derived from threonine serves as the preferred source of 2-carbon units in the synthesis of lipids, even in the presence of excess amounts of exogenous acetate [42]. Since this suggests some kind of compartmentation of the acetyl-CoA derived from threonine within the cell, we have localized two of the enzymes involved: threonine dehydrogenase and carnitine acetyltransferase. Our studies show that the majority of both enzymes (70%), as can be inferred from their equilibrium density of 1.18 g/cm 3, are mitochondrial in origin. However, there is a clear discrepancy in the differential centrifugation experiments between the distributions of the membrane markers (succinate dehydrogenase and sn-glycerol-3-phosphate dehydrogenase) on the one hand and the enzymes involved in the threonine pathway on the other. This discrepancy can be explained as follows. The enzymes of the threonine pathway are easily solubilized [26]. They, therefore, probably are enzymes of the mitochondrial matrix. If part of the membrane markers in the N fraction represents ruptured mitochondria attached to cell ghosts, which have lost their matrix content, then the enzymes of the matrix would be recovered in the final supernatant S. It is tentatively concluded that threonine dehydrogenase and carnitine acetyltransferase are exclusively present in the mitochondrial matrix. An exclusive mitochondrial localization also fits with the observation that only one isoenzyme band of threonine dehydrogenase is detectable in culture forms [39]. With the threonine pathway assigned to the mitochondrial matrix, it also becomes obvious why the acetyl-CoA derived from threonine does not readily mix with acetate entering the cell from the exogenous medium [42]. ACKNOWLEDGEMENTS We thank Professors Christian de Duve and Mikl6s M/iller for constant interest and valuable discussions, Dr. Reto Brun for providing us with the procyclic forms of T. brucei, Professor Andr6 Goffeau and Dr. Michel Briguet for their help with the cytochrome spectra, and Dr. Roger Klein for communicating details on the carnitine acetyltransferase assay. The excellent technical assistance of Miss Dominique Cottem and Mr. Joris Van Roy is gratefully acknowledged. A.M. was the recipient of an exchange fellowship of the Belgian Ministry of French Culture. This investigation received also financial support from the World Health Organization. REFERENCES 1 Vickerman, K. (1971) Morphological and physiological considerations of extracellular blood

308

protozoa. In: Ecology and Physiology of Parasites, (Fallis, A.M., ed.), pp. 5 8 - 9 1 . University of Toronto Press, Toronto. 2 Bowman, I.B.R. (1974) Intermediary metabolism of pathogenic flagellates. In: Trypanosomiases and Leishmaniases with Special Reference to Chagas' Disease. Ciba Foundation Symposium 20 (New Series), pp. 255-271. Elsevier/North-Holland Biomedical Press, Amsterdam. 3 Opperdoes, F.R., Borst, P. and De Rijke, D. (1976) Oligomycin sensitivity of the rnitochondrial ATPase as a marker for fly transmissability and the presence of functional kinetoplast DNA in African trypanosomes. Comp. Biochem. Physiol. 55B, 2 5 - 3 0 . 4 Cross, G.A.M., Klein, R.A. and Lindstead,D.J. (1975)Utilization of amino acids byTrypanosoma brucei in culture: L-threonine as a precursor for acetate. Parasitology 7 1 , 3 1 1 - 3 2 6 . 5 Hecker, H. (1980) Application of morphornetry to pathogenic trypanosomes (protozoa, mastigophora). Pathol. Res. Pract. 166, 2 0 3 - 2 1 7 . 6 Srivastava, H.K. and Bowman, I.B.R. (1972) Metabolic transformation of Trypanosoma rhodesiense in culture. Nat. New Biol. 2 3 5 , 1 5 2 - 1 5 3 . 7 Opperdoes, F.R., Borst, P., Bakker, S. and Leene, W. (1977) Localization of glycerol-3-phosphate oxidase in the mitochondrion and NAD+-linked glycerol-3-phosphate dehydrogenase in the microbodies of the bloodstream form of Trypanosoma brucei. Eur. J. Biochern. 76, 2 9 - 3 9 . 8 Opperdoes, F.R. and Borst, P. (1977) Localization of nine glycolytic enzymes in a microbodylike organelle in Trypanosoma brucei: the glycosome. FEBS Lett. 8 0 , 3 6 0 - 3 6 4 . 9 Oduro, K.K., Flynn, I.W. and Bowman, I.B.R. (1980) Trypanosoma bruce," activities and subcellular distribution of glycolytic enzymes from differently disrupted cells. Exp. Parasitol. 50, 123 135. 10 Taylor, M.B., Berghausen, H., Heyworth, P., Messenger, N., Rees, L.J. and Gutteridge, W.E. (1980) Subcellular localization of some glycolytic enzymes in parasitic flagellated protozoa. Int. J. Biochem. 1 1 , 1 1 7 - 1 2 0 . 11 Steiger, R.F., Opperdoes, F.R. and Bontemps, J. (1980) Subcellular fractionation of Trypanosoma brucei bloodstream forms with special reference to hydrolases. Eur. J. Biochem. 1 0 5 , 1 6 3 - 1 7 5 . 12 Opperdoes, F.R. and Steiger, R.F. (1981) Localization of hydrolases in culturated procyclic trypomastigotes Trypanosoma brucei. Mol. Biochem. Parasitol. 4,311 - 323 13 Markos, A., Van Roy, J. and Opperdoes, F.R. (1980) Subcellular fractionation of procyclic forms of Trypanosorna brucei with special reference to mitochondrial enzymes. J. Protozool. 26, 68A. 14 Klein, R.A. and Miller, P.G.G. (1981) Alternate metabolic pathways in protozoan energy metabolism. Parasitology 82, 1 - 3 0 . 15 Brun, R., Jenni, L., Tanner, M., Sch6nenberger, M. and Schell, K.F. (1979) Cultivation of vertebrate infective forms derived from metacyclic forms of pleomorphic Trypanosoma brucei stocks. Acta Trop. 3 6 , 3 8 7 - 3 9 0 . 16 Brun, R. and Sch6nenberger, M. (1979) Cultivation and in vitro cloning of procyclic culture forms of Trypanosoma brucei in a semi-defined medium. Acta Trop. 3 6 , 2 8 9 - 2 9 2 . 17 Opperdoes, F.R., Borst, P. and Spits, H. (1977) Particle-bound enzymes in the bloodstream form of Trypanosoma brucei. Eur. J. Biochem. 76, 2 1 - 2 8 . 18 Bergmeyer, H.U. (1974) Methods of Enzymatic Analysis, 2nd Edn., Academic Press, New York. 19 Pullman, M.E. and Penefsky, H.S. (1963) Preparation and assay of phosphorylating submitochondrial systems: mechanically ruptured mitochondria. In: Methods in Enzymology, (Colowick, S.P. and Kaplan, N.D., eds.), Vol. 6, pp 2 7 7 - 2 8 4 , Academic Press, New York. 20 Sumner, J.B. (1944) A method for the colorimetric determination of phosphorous. Science 100, 413-414. 21 Kusel, J.P. and Storey, B.T. (1973) Low-temperature spectral properties of the respiratory chain cytochromes of mitochondria from Crithidia fasciculata. Biochim. Biophys. Acta 3 0 5 , 5 7 0 - 5 8 0 . 22 Stein, S., Bohlen, P., Stone, J., Dairman, W. and Udenfriend, S. (1973) Amino acid analysis with fluorescamine at the picomole level. Arch. Biochem. Biophys. 1 5 5 , 2 0 3 - 2 1 2 .

309

23

24

25 26 27

28 29 30

31

32

33 34 35 36 37 38

39 40 41 42 43

Martin, E. and Mukkada, A.J. (1979) Identification of the terminal respiratory chain in kinetoplast mitochondrial complexes ofLeishmania tropica promastigotes. J. Biol. Chem. 254, 1219212198. Hill, G.C. (1976) Characterization of electrontransport systems present during the life cycle of African trypanosomes. In: Biochemistry of Parasites and Host-Parasite Relationships, (Van den Bossche, H., ed.), pp 31-50, Elsevier/North-Holland Biomedical Press, Amsterdam. Njogu, R.M., Whittaker, C.J. and Hill, G.C. (1980) Evidence for a branched electron transport chain in Trypanosoma brucei. Mol. Biochem. Parasitol. 1, 13-29. Linstead, D.J., Klein, R.A. and Cross, G.A.M. (1977) Threonine catabolism in Trypanosoma brucei. J. Gen. Microbiol. 101,243-251. Eeckhout, Y. (1972) Studies on acid hydrolases and on catalase of the Trypanosomatid Crithidja lucilliae. In: Comparative Biochemistry of Parasites, (Van den Bossche, H., ed.), pp 297-315, Academic Press, New York. Visser, N. and Opperdoes, F.R. (1980) Glycolysis in Trypanosoma brucei. Eur. J. Biochem. 103, 623-632. Grant, P.T. and Sargent, J.R. (1960) Properties of L-a-glycerophosphate oxidase and its role in the respiration of Trypanosoma rhodesiense. Biochem. J. 76,229-237. Bowman, I.B.R. and Flynn, I.W. (1976) Oxidative metabolism of trypanosomes. In: Biology of the Kinetoplastida, (Lumsden, W.H.R. and Evans, D.A., eds.), Vol. 1, pp. 435-476, Academic Press, New York. Opperdoes, F.R. and Nwagwu, M. (1980) Suborganellar localization of glycolytic enzymes in the glycosome of Trypanosoma brucei In: The Host-Invader Interplay, (Van den Bossehe, H., ed.), pp 683-686, Elsevier/North-Holland Biomedical Press, Amsterdam. Oduro, K.K., Bowman, I.B.R. and Flynn, I.W. (1980) Trypanosoma bruce," preparation and some properties of a multienzyme complex catalysing part of the glycolytic pathway. Exp. Parasitol. 50,240-250. Eldau, M. and Blum, J.J. (1973) Localization of phosphofructokinase on the mitochondrla of Tetrahymena pyriformis. J. Biol. Chem. 248, 7445-7448. Ericifiska, M. and Wilson, D.F. (1978) Homeostatic regulation of cellular energy metabolism. Trends in Biochem. Sci. 3,219-223. Berens, R.L. and Mart, J.J. (1977) Phosphofructokinase of Leishmania donovani and Leishmania brasiliensis and its role in glycolysis. J. Protozool. 24,340-344. Nwagwu, M. and Opperdoes, F.R. (1981) Regulation of glycolysis in Trypanosoma brucei: hexokinase and phosphofructokinase activity. Acta Trop., in press. Opperdoes, F.R. (1978) The glycosome: a new type of organdie in trypanosomes. Abstr. 3421, 12th Meeting of the Fed. Eur. Biochem. Soc., July 1978, Dresden. Hart, D.T. and Coombs, G.H. (1980) The importance and subcellular localisation of #-oxidation, glycolysis and amino acid oxidation in Leishmania mexicana mexicana amastigotes and promastigotes. Abstr. 16 Proc. 3rd Eur. Multicoll. Parasitol. 7 - 1 3 September 1980, Cambridge. Kilgour, V. (1980) The electrophoretic mobilities and activities of eleven enzymes of bloodstream and culture forms of Trypanosorna brucei compared. Mol. Biochem. Parasitol. 2, 51-62. Janski, A.M. and CorneU, N.W. (1980) Subcellular distribution of enzymes determined by rapid digitonin fractionation of isolated hepatocytes. Biochem. J. 186,423-429. McLaughlin, J. (1981) Association of adenylate kinase with the glycosome of Trypanosoma rhodesiense, Biochem. Intern. 2, 345-353. Klein, R.A. and Linstead, D.J. (1976) Threonine as a preferred source of 2-carbon units for lipid synthesis in Trypanosoma brucei. Biochem. Soc. Trans. 4, 48-50. Klein, R.A., Miller, P.G.G. and Linstead, D.J. (1976) The enzymic hydrolysis of acetyl-coenzyme A by Trypanosomatid flagellates. Biochem. Soc. Trans. 4, 285-287.