Experimental and in Silico Analyses of Glycolytic Flux Control in Bloodstream Form Trypanosoma brucei

Share Embed


Descrição do Produto

Supplemental Material can be found at: http://www.jbc.org/content/suppl/2005/06/15/M502403200.DC1.html THE JOURNAL OF BIOLOGICAL CHEMISTRY © 2005 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 280, No. 31, Issue of August 5, pp. 28306 –28315, 2005 Printed in U.S.A.

Experimental and in Silico Analyses of Glycolytic Flux Control in S Bloodstream Form Trypanosoma brucei*□ Received for publication, March 3, 2005, and in revised form, June 7, 2005 Published, JBC Papers in Press, June 14, 2005, DOI 10.1074/jbc.M502403200

Marie-Astrid Albert‡§, Jurgen R. Haanstra¶, Ve´ronique Hannaert‡, Joris Van Roy‡, Fred R. Opperdoes‡, Barbara M. Bakker ¶, and Paul A. M. Michels‡储 From the ‡Research Unit for Tropical Diseases, Christian de Duve Institute of Cellular Pathology and Laboratory of Biochemistry, Universite´ Catholique de Louvain, B-1200 Brussels, Belgium and the ¶Department of Molecular Cell Physiology, Faculty of Earth and Life Sciences, Vrije Universiteit, NL-1081 HV Amsterdam, The Netherlands

* This work was supported in part by a grant from the Belgian Fonds de la Recherche Scientifique Me´dicale and European Commission INCO-DEV Program Contract ICA4-CT-2001-10075 (to P. A. M. M.) and by a Vernieuwingsimpuls grant from the Dutch Organization for Scientific Research (to B. M. B.). □ S The on-line version of this article (available at http://www.jbc.org) contains Supplemental “Experimental Procedures,” abbreviations, references, Fig. S1, and Tables S1 and S2. § Supported by a Ph.D. scholarship from the Fonds pour la Formation a` la Recherche dans l’Industrie et dans l’Agriculture and the Universite´ Catholique de Louvain and by a COST-B9 Action travel grant from the European Commission for a short work visit to the laboratory of Dr. Christine Clayton. 储 To whom correspondence should be addressed: Research Unit for Tropical Diseases, Christian de Duve Institute of Cellular Pathology, ICP-TROP 74.39, Universite´ Catholique de Louvain, Ave. Hippocrate 74, B-1200 Brussels, Belgium. Tel.: 32-2-764-7473; Fax: 32-2-762-6853; E-mail: [email protected].

Trypanosomatid parasites (Trypanosoma and Leishmania) are responsible for serious diseases of mankind in tropical and subtropical countries worldwide. These diseases affect millions of people, and hundreds of millions are at risk of becoming infected. Unfortunately, available treatments are largely inadequate. Currently used drugs are inefficacious and toxic. There is a desperate need for new effective and safe drugs, particularly in view of the development and spreading of drug resistance (1). Glycolysis plays an important role in the energy metabolism of these protozoan organisms, notably of Trypanosoma brucei when it lives in the blood of its mammalian host, causing a disease called sleeping sickness or human African trypanosomiasis in man and nagana in cattle (2, 3). This bloodstream form of T. brucei is entirely dependent on the conversion of the blood sugar glucose into pyruvate for its ATP supply. Oxidative metabolism involving mitochondrial Krebs cycle enzymes and oxidative phosphorylation are largely repressed. Therefore, glycolysis has been perceived as a potentially good target for anti-trypanosome drugs. Moreover, the glycolytic pathway in trypanosomatids is organized in a unique manner: the majority of the glycolytic enzymes are sequestered inside peroxisomelike organelles, designated glycosomes (Fig. 1) (2–5). Our previous research has shown that this unusual organization of the pathway and the long evolutionary distance between trypanosomatids and other organisms have led to the endowment of distinct kinetic, regulatory, and structural properties to most of the trypanosomatid glycolytic enzymes. These properties are currently being exploited for the structure-based design of parasite enzyme-selective inhibitors to be used as lead drugs (6). All enzymes involved in the glycolysis of T. brucei have been purified previously (either the authentic enzymes from parasites grown in rats or enzymes expressed from recombinant genes in Escherichia coli), and their kinetic properties have been determined (7–9). These data and the information about the enzymes’ subcellular organization and specific activities in lysates of trypanosomes grown in rats have been used to develop an in silico model of glycolysis in the parasite (10). The fluxes and cytosolic metabolite concentrations as predicted by the model were in accordance with data experimentally determined in non-growing trypanosomes under both anaerobic and aerobic conditions. Following the principles of Metabolic Control Analysis (11–13), which imply that, in general, there is no single rate-limiting step in a metabolic pathway, but that control can be shared among several steps, the model was subsequently used to calculate what controls the flux of trypanosome glycolysis under physiological conditions. Moreover, the extent to which different enzymes should be inhibited to achieve a certain decrease in glycolytic flux was estimated (14, 15). Such information may be rele-

28306

This paper is available on line at http://www.jbc.org

Downloaded from www.jbc.org by guest, on February 19, 2013

A mathematical model of glycolysis in bloodstream form Trypanosoma brucei was developed previously on the basis of all available enzyme kinetic data (Bakker, B. M., Michels, P. A. M., Opperdoes, F. R., and Westerhoff, H. V. (1997) J. Biol. Chem. 272, 3207–3215). The model predicted correctly the fluxes and cellular metabolite concentrations as measured in non-growing trypanosomes and the major contribution to the flux control exerted by the plasma membrane glucose transporter. Surprisingly, a large overcapacity was predicted for hexokinase (HXK), phosphofructokinase (PFK), and pyruvate kinase (PYK). Here, we present our further analysis of the control of glycolytic flux in bloodstream form T. brucei. First, the model was optimized and extended with recent information about the kinetics of enzymes and their activities as measured in lysates of in vitro cultured growing trypanosomes. Second, the concentrations of five glycolytic enzymes (HXK, PFK, phosphoglycerate mutase, enolase, and PYK) in trypanosomes were changed by RNA interference. The effects of the knockdown of these enzymes on the growth, activities, and levels of various enzymes and glycolytic flux were studied and compared with model predictions. Data thus obtained support the conclusion from the in silico analysis that HXK, PFK, and PYK are in excess, albeit less than predicted. Interestingly, depletion of PFK and enolase had an effect on the activity (but not, or to a lesser extent, expression) of some other glycolytic enzymes. Enzymes located both in the glycosomes (the peroxisome-like organelles harboring the first seven enzymes of the glycolytic pathway of trypanosomes) and in the cytosol were affected. These data suggest the existence of novel regulatory mechanisms operating in trypanosome glycolysis.

Glycolytic Flux Control in T. brucei

28307

vant for choosing optimal drug targets (16). According to the computer simulation, under physiological conditions, most control is exerted by the glucose transporter, but can be shifted to a set of four enzymes, aldolase (EC 4.1.2.13), glyceraldehyde-3-phosphate dehydrogenase (GAPDH1; EC 1.2.1.12), phosphoglycerate kinase (EC 2.7.2.3), and glycerol3-phosphate dehydrogenase (EC 1.1.1.8), at high glucose concentrations and/or when a small increase occurs in the level of the transporter. Similarly, these five steps appear to have the most prominent effects on the flux if their levels are decreased. Surprisingly, trypanosomes seemed to possess a substantial overcapacity of hexokinase (HXK; EC 2.7.1.1), phosphofructokinase (PFK; EC 2.7.1.11) and pyruvate kinase (PYK; EC 2.7.1.40), the kinases that are often thought to 1 The abbreviations used are: GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HXK, hexokinase; PFK, phosphofructokinase; PYK, pyruvate kinase; RNAi, RNA interference; PGAM, phosphoglycerate mutase; ENO, enolase; UTR, untranslated region.

control glycolysis (17–19). Indeed, the results predicted for the glucose transporter (trypanosome hexose transporter) were essentially confirmed by experiments showing that 40% of the total flux control is exerted by the trypanosome hexose transporter (20). This implies that earlier claims that glucose transport is the only rate-limiting step exaggerated the role of the transporter (21–24). In this work, we present our further analysis of the control of glycolytic flux in bloodstream form T. brucei. First, the model was optimized and extended with recent information about the kinetics of enzymes (25–28). Moreover, it was modified on the basis of activities of enzymes as measured in lysates of in vitro cultured growing bloodstream form trypanosomes. Such cells were then used for experimentally determining fluxes after either overexpression of enzymes or their depletion through RNA interference (RNAi). We first set out to check whether HXK, PFK, and PYK are indeed in large excess. Subsequently, two other cytosolic enzymes, phosphoglycerate mutase (PGAM; EC 5.4.2.1) and enolase (ENO; EC 4.2.1.11), were studied. The

Downloaded from www.jbc.org by guest, on February 19, 2013

FIG. 1. Aerobic glycolysis in bloodstream form T. brucei. Step 1, hexokinase; step 2, glucose-6-phosphate isomerase; step 3, phosphofructokinase; step 4, aldolase; step 5, triose-phosphate isomerase; step 6, glyceraldehyde-3-phosphate dehydrogenase; step 7, phosphoglycerate kinase; step 8, NAD-dependent glycerol-3-phosphate dehydrogenase; step 9, glycerol kinase; step 10, phosphoglycerate mutase; step 11, enolase; step 12, pyruvate kinase; step 13, glycerol-3-phosphate oxidation (involving FAD-dependent glycerol-3-phosphate dehydrogenase, ubiquinone, and the trypanosome alternative oxidase). G-3-P, glyceraldehyde 3-phosphate; 1,3-BPGA, 1,3-bisphosphoglycerate; 3-PGA, 3-phosphoglycerate; DHAP, dihydroxyacetone phosphate; Gly-3-P, glycerol 3-phosphate; PEP, phosphoenolpyruvate. The substrate (glucose) and secreted end product (pyruvate) are boxed. Enzymes whose expression was modified by RNAi and/or overexpression in this study are indicated by boldface numbered arrows.

28308

Glycolytic Flux Control in T. brucei

results of these experiments and a comparison of the data obtained experimentally and in silico have been used to gain further insight in the control and regulation of glucose metabolism in these parasites. EXPERIMENTAL PROCEDURES

Trypanosomes, Growth Conditions, Harvesting, and Preparation of Cell Extracts Bloodstream form T. brucei strain 427 (cell line 449 (29), constitutively expressing the E. coli tetracycline (tet) repressor gene integrated in its genome) was cultured at 37 °C in water-saturated air with 4% CO2 in HMI-9 medium containing 20% heat-inactivated bovine serum (Invitrogen) and 0.2 ␮g/ml phleomycin (Cayla), the selectable marker for the Tet repressor construct. Tetracycline was used at a concentration of 1 ␮g/ml. Cultures were harvested at densities of 1–3 ⫻ 106 cells/ml (the exponential growth phase range is 0.2–5 ⫻ 106 cells/ml) by centrifugation at 1900 ⫻ g for 10 min. Cells used for enzyme activity assays were washed twice with 5 ml of phosphate-buffered saline and then lysed in 0.5 ml of phosphate-buffered saline containing 0.1% Triton X-100, 150 mM NaCl, and an EDTA-free protease inhibitor mixture (1 ␮M pepstatin, 1 ␮M leupeptin, 0.25 ␮M E-64, 10 ␮M 4-(2-aminoethyl)benzenesulfonyl fluoride, and 750 ␮M phenylmethanesulfonyl fluoride). Cellular debris was removed by centrifugation, and the supernatant was kept on ice until assays were performed.

Transfection of Trypanosomes Trypanosomes were cultured to a density of 1.6 ⫻ 106 cells/ml; centrifuged; washed with 5 ml of Cytomix supplemented with 0.5% glucose, 1 mM hypoxanthine, and 100 ␮g/ml bovine serum albumin (31, 32); and resuspended in 0.4 ml of the same solution. The trypanosome suspension was incubated with 10 ␮g of linearized DNA and subsequently subjected to a single discharge using a Genetronics BTX ECM630 electroporator at settings of 1250 V, 25 ohms, and 50 microfarads to obtain a pulse with a voltage of 1000 V and a time constant of 260 ␮s. After dilution of the cell in 12 ml of culture medium, it was split into 24 wells of a microtiter plate. An antibiotic to select positive transfectants was added after 24 h. The antibiotic used was hygromycin (Sigma) at a final concentration of 5 ␮g/ml, blasticidin (Invitrogen) at 5 ␮g/ml, or G418 (Invitrogen) at 2 ␮g/ml. Antibiotic-resistant cells were detectable after 7–10 days. If antibiotic-resistant cultures appeared in ⬎3 of 24 wells, limiting dilution was performed to ascertain the existence of cloned cell lines. Correct integration of constructs into the genome was checked by Southern blotting and hybridization and/or PCR, and expression of newly introduced genes was checked by reverse transcription-PCR.

Construction of Plasmids Constructs for Conditional Gene Knockouts—To attempt to create conditional knockout HXK and PFK strains, three constructs were made for each HXK and PFK. First, for inducible expression of a new gene, a DNA fragment containing the full-length gene flanked with HindIII and BamHI restriction sites at the 5⬘- and 3⬘-ends, respectively, was introduced into vector pHD677 (29). This places the gene under the control of a T. brucei procyclic acidic repetitive protein promoter to which the tet operator has been fused. Furthermore, the vector provides a 5⬘-untranslated region (UTR) of a procyclic acidic repetitive protein gene and a 3⬘-UTR of a variant surface glycoprotein gene for the inserted gene. The vector has also a hygromycin gene that will be constitutively expressed in trypanosomes and a sequence of the ribosomal repeat spacer with a unique NotI restriction site. Linearization of the construct with NotI permits its insertion by homologous recombi-

Biochemical Methods Enzyme assays were performed as described previously: HXK, glucose-6-phosphate isomerase (EC 5.3.1.9), PFK, triose-phosphate isomerase (EC 5.3.1.1), GAPDH, phosphoglycerate kinase, and glycerol kinase (EC 2.7.1.30) (33); PGAM (using the assay based on measurement of phosphoenolpyruvate formation at 240 nm) (28); ENO (27); and PYK (34). Protein concentrations were determined using the Bio-Rad protein assay based on Coomassie Brilliant Blue staining using bovine serum albumin as a standard. SDS-PAGE and Western blotting were performed as described previously (27). Primary polyclonal antibodies were raised in rabbits against purified proteins (either natural proteins purified from bloodstream form trypanosomes as described by Misset et al. (7) or recombinant proteins purified from E. coli) and used for the detection of pro-

Downloaded from www.jbc.org by guest, on February 19, 2013

Molecular Biological Methods For most experiments in molecular biology, standard methodologies were used (30), or protocols were followed as provided by the suppliers of enzymes used for various forms of DNA and RNA manipulation. PCR was performed with either genomic DNA from T. brucei (1 ␮g) or plasmid DNA (100 ng), the four deoxynucleotides (1 mM each), a sense and antisense primer (10 pmol each), and Taq DNA polymerase (2.5 units; TaKaRa) in a reaction volume of 50 ␮l. Amplification involved a denaturation step at 95 °C for 3 min, followed by cycles of denaturation at 95 °C for 1 min, hybridization at a temperature specific for the primer sets chosen for 60 s, and elongation at 72 °C for 60 s. The number of cycles required was determined visually after electrophoresis and ethidium staining of the DNA. Amplification was stopped by incubation at 72 °C for 7 min. Semiquantitative PCR was used to assess changes in RNA levels resulting from RNAi. First, cDNA was prepared using the RevertAid H minus first strand cDNA synthesis kit (Fermentas). 1 ␮g of total RNA purified both from trypanosomes in which RNAi was induced and from cells in which it was not induced was treated with DNase for 30 min. Subsequently, cDNA was synthesized using 0.2 ␮g of random primers and 200 units Moloney murine leukemia virus reverse transcriptase in the presence of 20 units of RNase inhibitor at 25 °C for 10 min, followed by 42 °C for 60 min. The reaction was stopped by heating at 70 °C for 10 min. The cDNA was stored at ⫺20 °C. 0.2 ␮g of cDNA was used as a template for PCR with 0.04 mM each dATP, dTTP, and dGTP and 0.01 mM [␣-32P]dCTP (3000 Ci/mmol; Amersham Biosciences), 10 pmol of a sense and antisense primer, and 2.5 units of Taq DNA polymerase in a reaction volume of 50 ␮l. The number of cycles (denaturation at 95 °C for 2 min, hybridization at a temperature specific for the primer sets chosen for 60 s, and elongation at 72 °C for 60 s), terminated with a single 7-min incubation at 72 °C, was chosen such that the experiment remained within the linear range of amplified DNA/cycle. This was tested by migration of the amplified radioactive DNA on a polyacrylamide gel and measuring the 32P signal of the bands using a PhosphorImager.

nation into this transcriptionally silent region of the genome. Two other constructs were made for the HXK and PFK genes to disrupt both their alleles. These constructs were made with a pBluescript-based vector containing a G418 (neomycin) or blasticidin resistance cassette (pBSTubNeo and pBS-Blast, respectively). (The latter one was made from pBS-Hyg by replacing the hygromycin gene with the blasticidin gene; pBS-Tub and pBS-Hyg were kindly donated by Dr. Christine Clayton (Zentrum fuer Molekular Biologie Heidelberg, Universita¨t Heidelberg).) For HXK, which is encoded by two tandemly linked genes, a 540-bp fragment of the 5⬘-part of the coding region and a 572-bp region of the unique 5⬘-UTR were amplified and placed on either side of the cassettes. These constructs should allow the disruption of one or both genes of the two HXK alleles by homologous recombination upon transfection of trypanosomes. For PFK, which is encoded by only a single gene, a 405-bp region of the 5⬘-UTR and a 417-bp region of the 3⬘-UTR were amplified and fused to the genes conferring antibiotic resistance. Constructs for RNAi—To induce RNAi of HXK, PFK, PGAM, ENO, and PYK, constructs were prepared for the tetracycline-inducible synthesis of an intramolecular (hairpin) double-stranded RNA molecule. This was achieved by cloning, in an direct inverted arrangement, a gene-specific PCR fragment in the sense orientation and the corresponding but somewhat shorter fragment (32–50 bp) in the antisense orientation in either plasmid pHD1136 or pHD677. pHD1136 differs from pHD677 in providing a 3⬘-UTR of the actin gene to the inserted DNA fragment and having a blasticidin resistance marker gene. The following fragments were cloned: HXK, a 323-bp fragment of the 3⬘UTR of the second gene in the HXK locus in the sense orientation and a 280-bp fragment in the antisense orientation in pHD1136 (clone pRH3⬘, selected for RNAi studies); PFK, 466-bp (sense) and 418-bp (antisense) fragments of the 3⬘-UTR in pHD1136 (clone pRP3⬘); ENO, 499-bp (sense) and 434-bp (antisense) fragments of the 5⬘-part of the coding region in pHD677 (clone pHD677-ENO); PGAM, 550-bp (sense) and 500-bp (antisense) fragments of the coding region in pHD677 (clone pHD677-PGAM); and PYK, 432-bp (sense) and 400-bp (antisense) fragments of the 3⬘-end of the gene including a small part of its 3⬘-UTR in pHD677 (clone pHD677-PYK). All RNAi constructs were prepared for chromosomal integration in the ribosomal repeat spacer. To minimize off-target effects in RNAi studies, the specificity of all gene fragments selected for making RNAi constructs was checked by performing BLAST searches against the genome data base of T. brucei (available at www.geneDB.org/genedb/tryp/blast.jsp).

Glycolytic Flux Control in T. brucei teins on immunoblots at the following dilutions: HXK, 1:120,000; PFK, 1:100,000; aldolase, 1:150,000; GAPDH, 1:150,000; PGAM, 1:10,000; ENO, 1:150,000; PYK, 1:100,000; and glycerol kinase, 1:100,000. The rate of glucose consumption was determined by measuring the production of tritiated water, exchanged for “bulk” water in the aldolase-catalyzed reaction as described by Bontemps et al. (35), by incubation of 1 ⫻ 108 trypanosomes with 0.3 ␮Ci/ml D-[3-3H]glucose (Amersham Biosciences) in 2 ml of HMI-9 medium at 37 °C for 30 min. For determination of the rate of pyruvate and glycerol production, minimally 5 ⫻ 107 trypanosomes were cultured in 2 ml of HMI-9 medium at 37 °C for 30 – 45 min. Batches of cells were lysed before and after culturing for a certain period of time, respectively, and the reactions were stopped by addition of 7% (final concentration) HClO4. The extracts were incubated for 10 min on ice and centrifuged, and the supernatant was neutralized with 3 M KOH/KHCO3, followed by another centrifugation step. The supernatant thus obtained was kept on ice, and the amounts of pyruvate and glycerol present in it were measured by their quantitative conversion into other metabolites with the concomitant oxidation of NADH, followed spectrophotometrically at 340 nm. The rate of oxygen consumption was monitored polarographically at 37 °C in a thermostatted vessel with a Clark-type electrode as described (20).

Extension of the Metabolic Model

RESULTS AND DISCUSSION

Extension of the Metabolic Model—In this study, we compared the effects of knockdown of enzyme activities on glycolytic flux in bloodstream form trypanosomes with predictions made by the kinetic model of this pathway. To this end, the model was first updated with the most recent information. Previous model versions (10, 15) were based on specific enzyme activities and enzyme concentrations measured in lysates of T. brucei strain 427, purified from infected rats (7). This laboratory-adapted strain can grow in rats to very high densities (⬎109 cells/ml of blood, 103 to 104-fold higher than in their natural hosts) and cause acidosis of the blood or hypoglycemia. This may have affected the levels of glycolytic enzymes in the trypanosomes. Moreover, once taken from the animal and purified over an ion-exchange column, the trypanosomes do not grow anymore. For the experimental test of the model described in this study, we used an in vitro cultured cell line of growing trypanosomes derived from strain 427. The maximal densities to be reached are 5 ⫻ 106 cells/ml. Enzyme activities were measured again, now in lysates of different batches of these trypanosomes harvested at different cell densities. The specific activities appeared to be rather constant during the exponential phase of growth (cell densities between 1 and 3 ⫻ 106 cells/ml). The average values as measured for independently grown batches of cells and the Vmax values used in the model are given in Supplemental Table S1. At higher cell densities, when the growth rate is retarded and cells start entering the stationary phase, higher specific activities were found for several enzymes, notably triose-phosphate isomerase, HXK, and phosphoglycerate kinase (data not shown). Addition-

ally, the model was modified with improved kinetic data for glycerol kinase, extended by incorporation of PGAM and ENO, and the reactions catalyzed by aldolase, phosphoglycerate kinase, and the glucose transporter were slightly modified. This new version of the model predicts a steady-state flux of 90.1 nmol/min/mg of protein under aerobic conditions, which corresponds nicely with the glucose flux of ⬃100 nmol/min/mg of protein at 25 °C that we estimated from our measurements. The glucose consumed/glycerol produced/pyruvate produced ratio is 1:0.06:1.94 at steady state under aerobic conditions. This is in good agreement with literature data (38). The sensitivity of the anaerobic glycolytic flux to external glycerol under anaerobic conditions was also tested. In the computer model, we need 380 ␮M to inhibit the glucose consumption flux by 50%. This is closer to the experimentally measured value of 800 ␮M (39) than the previous model version, which predicted 50% inhibition already at 75 ␮M glycerol (14). The flux control distribution over the different enzymes is hardly changed in the new model compared with earlier versions: the glucose transporter still exerts most of the control. However, although earlier simulations (14) predicted that the glucose transporter shared control with aldolase, GAPDH, phosphoglycerate kinase, and glycerol-3-phosphate dehydrogenase under physiological conditions, the control of aldolase and phosphoglycerate kinase is now somewhat lower, and glycerol-3-phosphate oxidase and PGAM have gained some control. Notably, glycerol kinase has a negative flux control coefficient over the oxygen consumption and pyruvate flux. According to the new model, too, no control resides in HXK, PFK, and PYK, and they are still present in excess (see Supplemental Fig. S1 and below). Conditional Knockout of HXK and PFK—Initial attempts to experimentally estimate the control exerted over glycolytic flux by HXK and PFK involved the deletion of the endogenous genes in conjunction with the insertion of another gene copy under the control of the tetracycline-inducible promoter of the procyclic acidic repetitive protein. Introduction of a replacement copy of either the HXK or PFK gene proceeded without problem. The HXK clone selected (clone H3) showed, upon addition of tetracycline, 20% higher HXK activity than non-induced cells, but no measurable effect on the activity of other enzymes tested (PFK and ENO), pyruvate production, oxygen consumption, and growth rate (data not shown). Similarly, the additional gene copy of PFK (clone P7) led to 40% higher PFK activity compared with control cells, without an effect on ENO activity, pyruvate production, oxygen consumption, and growth rate. These results confirm that neither HXK nor PFK exerts control over glycolytic flux or growth under normal growth conditions. However, all attempts to disrupt even one allele of the endogenous HXK genes in clone H3 or the PFK gene in clone P7 failed under conditions under which control experiments with nonessential genes were successful. Similar negative results were observed previously for two other enzymes involved in aerobic glycolysis of bloodstream form T. brucei: none of the alternative oxidase gene alleles appeared dispensable, whereas for triose-phosphate isomerase, only one allele could be disrupted in the presence of a replacement copy (15). Knockdown of Expression of HXK and PFK by RNAi—We next used RNAi to assess experimentally to what levels HXK and PFK activity should decrease to assume control over glycolytic flux and growth. The cell lines used in this study contain constructs for tetracycline-inducible RNAi; the targeted mRNA was selectively degraded upon induction of RNAi. However, it appeared not possible to fine-tune the extent of mRNA depletion and enzyme activity decrease with different concentrations of tetracycline and thus to achieve different steady-state levels

Downloaded from www.jbc.org by guest, on February 19, 2013

A kinetic model of glycolysis in bloodstream form T. brucei was previously constructed (10), modified (14), and extended (15). The model was further extended for this study. (i) For better comparison between the model and experiments, Vmax values in the model were modified on the basis of measurement of specific activities of enzymes in lysates of in vitro cultured trypanosomes. (Earlier versions were based on data from trypanosomes isolated from rats, which were unable to grow ex vivo.) (ii) Reversible Michaelis-Menten rate equations with one substrate and one product for PGAM and ENO were included into the model, as kinetic data had become available. (iii) The aldolase equation was corrected; the equilibrium constant is now independent of the adenine nucleotide concentrations. (iv) Kinetic parameters of glucose transport, glycerol kinase, phosphoglycerate kinase, PGAM, and ENO were updated based on new experimental data. All numerical calculations of the T. brucei model were carried out using the program Jarnac (36, 37). Details on the extensions of the model (rate equations, differential equations, and parameter values) are provided in the Supplemental “Experimental Procedures.”

28309

28310

Glycolytic Flux Control in T. brucei due to mutated revertants that escaped RNAi control (data not shown), a phenomenon frequently seen in trypanosomes, particularly in bloodstream forms (42, 43). As mentioned above, when PFK was targeted by RNAi in cell line RP3⬘, the activity dropped gradually over 24 h to 46.7 ⫾ 3.3% and remained approximately at this level until 48 h (Fig. 2B) with the concomitant reduction of glycolytic flux to 26% after 24 h (Table I). However, contrary to the RNAi experiments in which HXK was targeted, the activities of some other glycolytic enzymes (HXK, ENO, and PYK) decreased as well when PFK was depleted to ⬃60 –75% after 24 h, and they continued to decrease afterward to 40 –70% after 48 h (Fig. 2B). In contrast, the activities of glucose-6-phosphate isomerase and triose-phosphate isomerase were not affected (data not shown). Quantitative Western blotting showed that the PFK concentration dropped to 70% in 12 h and to 40.0 ⫾ 10.0% in 24 h compared with the control PEX5, whereas the concentrations of other enzymes remained constant (PYK, 106.1 ⫾ 10.0%; and GAPDH, 100.9 ⫾ 4.6%) (Fig. 4A). Only after 48 h, when the remaining PFK concentration was 42.3 ⫾ 3.1%, did the concentrations of some other enzymes decrease further: PYK to 74.1 ⫾ 5.9% and GAPDH to 91.0 ⫾ 12.4%. These changes in protein concentrations seemed to match those of the transcripts as measured by semiquantitative PCR, although these latter experiments showed somewhat larger variations between different experiments (data not shown). The decreases in activity observed for enzymes other than PFK are most likely not due to “off-target” RNAi effects. In data base searches, no sequence similarities could be found between the targeted area of the PFK gene and genes of the other proteins (including their flanking regions), which might cause unwanted transcript degradation. Moreover, the changes in enzyme activities did not match those measured for the corresponding protein and RNA levels. Therefore, these data suggest that the decrease in PFK concentration or activity down-regulates some, but not all, other glycolytic enzymes. This down-regulation is not related to compartmentalization, as both glycosomal (HXK) and cytosolic (ENO and PYK) glycolytic enzymes were affected. Possible mechanisms for such regulation and their relevance for the cell will be discussed in more detail below. Knockdown of Expression of PYK, PGAM, and ENO by RNAi—Similar experiments as those performed with HXK and PFK were performed with PYK, another enzyme that, according to the kinetic model, would be in large excess. Cytosolic PYK is the only trypanosomatid glycolytic enzyme whose activity seems to be highly regulated by effectors, most notably the allosteric activator fructose 2,6-bisphosphate (2, 3, 34). Furthermore, the control exerted by PGAM and ENO over glycolytic flux and growth was also analyzed. Upon induction of RNAi for PYK, its activity gradually decreased during the first 24 h to 23.8 ⫾ 6.5%, whereas the activity of all other enzymes tested (HXK, PFK, and ENO) remained essentially unchanged (Fig. 2C, left panel). A similar observation was made at the protein level by Western blotting (Fig. 2C, right panel). Also at the transcript level (data not shown), PYK decreased to ⬃25% in 24 h, with no significant changes for the trypanosome hexose transporter, PFK, GAPDH, and ENO when normalized to the tubulin mRNA concentration. An increase was observed for HXK mRNA, but its significance should be doubted in view of the large fluctuations observed for this transcript in different experiments. Glucose consumption dropped to 49.7% (Table I). During these first 24 h, the growth rate remained unchanged despite the ⬃50% decrease in glycolytic flux, but the further reduction of PYK after 24 h (to 10% after 30 h) led to growth arrest, followed by trypanosome death (Fig. 3C).

Downloaded from www.jbc.org by guest, on February 19, 2013

of protein, contrary to reporter gene expression that, in trypanosomes, could be controlled over a range of 4 orders of magnitude in response to the concentration of the inducer (40). Intermediate levels could, to some extent, be obtained by combining expression of an exogene with the RNAi-induced degradation of the mRNA of the endogenous genes. To do so, the RNAi was targeted to the 3⬘-UTR of the endogenous transcript, whereas the exogenous gene copy was made with the 3⬘-UTR of a variant surface glycoprotein gene. For example, 24 h after the mere induction of RNAi for PFK (in cell line RP3⬘), the enzyme activity had dropped to 46.7 ⫾ 3.3% (n ⫽ 3) of that in noninduced cells, whereas the glucose consumption and pyruvate production were decreased to 26.2 ⫾ 4.0 and 27.0 ⫾ 1.4%, respectively (see also below). The PFK activity remained rather stable between 24 and 48 h, but soon after 48 h, the cells were dying. In contrast, the combined RNAi and overexpression led to a gradual smaller decrease, with 60% PFK activity still remaining after 48 h (data not shown). No effect on the glucose consumption and pyruvate production was observed after 24 h, when the PFK activity was 64.7 ⫾ 2.5% (n ⫽ 3) of that in non-induced cells, and growth remained unaffected over 5 days. These results confirmed that PFK had no flux control, even if its activity was ⬃60% of the wild-type level. However, further depletion of PFK appeared to result in a reduction of glycolytic flux and growth retardation. Similar experiments with HXK suggested that it has an even larger overcapacity (data not shown), in agreement with computer simulations. To study further what would happen upon lowering enzyme activities, we followed cell growth, determined enzyme activities at different times after RNAi induction, and measured glycolytic flux (glucose consumption and pyruvate production) in samples taken after 24 h. Moreover, the effect of RNAi was also tested for other enzymes at the transcriptional level by performing semiquantitative reverse transcription-PCR using total RNA extracts prepared from cells sampled 24 h after induction and at the protein level by Western blotting. All these experiments were performed three or six times with independently grown batches of trypanosomes, all with essentially similar results. The only significant variation observed concerned the onset of the decrease in enzyme activity after tetracycline addition to the culture to induce RNAi. This varied maximally by 12 h within the period 12–24 h after induction. However, the values measured for all parameters between 24 and 48 h in each experiment of a series were very similar, and the correlation between all parameters (growth, activities of other enzymes, flux, etc.) remained constant also within the first period. Therefore, the time course given in Fig. 2 for the change in enzyme activities during the first 24 h represents a superposition of curves from different experiments, but slightly shifted along the time axis. Fig. 2A (left panel) shows that, upon induction, HXK activity in cell line RH3⬘ gradually decreased to 8% after 48 h. After 24 h, only 22.5 ⫾ 2.5% HXK activity was left, sufficient to sustain a flux (glucose consumption and pyruvate production) of 68% (Table I). Initially, the activity of the control enzymes (PFK and glycerol kinase) remained constant or even increased slightly, whereas some decrease could be observed at later stages. A major decrease in the HXK level was observed by Western blotting after 24 h compared with the control PEX5, a protein involved in glycosome biogenesis (41), whereas the levels of other glycolytic enzymes (PFK and ENO) remained unaffected (Fig. 2A, right panel). Despite the large decrease in activity and flux during the first 24 h, no significant reduction of the growth rate was seen during this period (Fig. 3A). However, the further decrease in activity was associated with retardation of growth. Nevertheless, the cell population survived, and after 4 days, growth resumed at its original rate

Glycolytic Flux Control in T. brucei

28311

Downloaded from www.jbc.org by guest, on February 19, 2013

FIG. 2. Changes in enzyme activities and enzyme levels in bloodstream form T. brucei upon induction of RNAi. The mRNAs targeted were for HXK (A), PFK (B), PYK (C), PGAM (D), and ENO (E). Left panels, time-dependent changes in enzyme activities in T. brucei lysates; right panels, protein levels as determined by Western blot analysis in samples taken 24 h after RNAi induction. Tetracycline (Tet) was added to the cultures at a concentration of 1 ␮g/ml to induce the formation of specific double-stranded RNAs and thus RNAi from t ⫽ 0 onward. Enzyme activities are expressed relative to specific activities as measured in lysates of trypanosomes without induction. GK, glycerol kinase; HK, hexokinase; PGK, phosphoglycerate kinase.

28312

Glycolytic Flux Control in T. brucei

TABLE I Comparison of experimentally determined and modeled enzyme Vmax values at which experimentally measured glycolytic flux reductions were achieved exp mod % J is the relative flux (compared with that in non-induced cells) of either glucose consumption or pyruvate production. % Vmax and % Vmax are the % Vmax measured 24 h after induction (compared with that in non-induced cells) and the % Vmax in the model (relative to non-inhibited values) giving a flux corresponding to the experimentally measured one, respectively. The values given (mean ⫾ S.E., n ⫽ 3) for the glucose consumption and pyruvate production fluxes were determined 24 h after induction of RNAi (with tetracycline), targeted to the glycolytic enzyme indicated, and are expressed relative to the values measured in the same cell lines without induction (without tetracycline). The absolute values for glucose consumption and pyruvate production fluxes measured (at 37 °C) in wild-type cells were 51.71 ⫾ 0.97 and 80.38 ⫾ 5.47 nmol/min/108 cells, respectively. Glucose consumption Enzyme

exp Vmax

J %

HXK PFK PYK PGAM ENO

67.6 ⫾ 8.7 26 ⫾ 4.0 49.7 ⫾ 3.3 53.9 ⫾ 3.3 46 ⫾ 2.9

22 ⫾ 2.5 47 ⫾ 3.3 24 ⫾ 6.5 51 ⫾ 1.8 16 ⫾ 2.3

mod Vmax

exp mod Vmax /Vmax

%

%

4.1 ⬍5 8.6 38.8 12.6

5.37 ⫾ 0.61 2.79 ⫾ 0.76 1.31 ⫾ 0.05 1.27 ⫾ 0.18

Pyruvate production exp Vmax

J %

61 ⫾ 3.3 27 ⫾ 1.4 66 ⫾ 3.6 74.3 ⫾ 5.0 59.1 ⫾ 5.3

22 ⫾ 2.5 47 ⫾ 3.3 24 ⫾ 6.5 51 ⫾ 1.8 16 ⫾ 2.3

mod Vmax

exp mod Vmax /Vmax

%

%

3.4 ⬍5 12.5 57.8 17.3

6.47 ⫾ 0.73 1.92 ⫾ 0.52 0.88 ⫾ 0.03 0.92 ⫾ 0.13

Downloaded from www.jbc.org by guest, on February 19, 2013

FIG. 3. Effect of intracellular depletion of specific glycolytic enzymes by RNAi on the growth of bloodstream form T. brucei. The mRNAs targeted by RNAi were for HXK (A), PFK (B), PYK (C), PGAM (D), and ENO (E). Shown are the log10 values of cumulative cell numbers. The growth of cells transfected with a plasmid producing double-stranded RNA to decrease specific enzyme levels was determined by growing the cells in the absence (Œ) or presence (f) of the RNAi inducer tetracycline (1 ␮g/ml) from t ⫽ 0 onward.

RNAi experiments were also performed with PGAM and ENO as controls. Like PYK, these enzymes are present in the cytosol, but no special activity regulation mechanisms are known. Also for PGAM, the enzyme level decreased gradually to 51.3 ⫾ 1.8% during the first 24 h after RNAi induction, without affecting any other enzyme tested (HXK, PFK, phosphoglycerate kinase, ENO, and PYK) (Fig. 2D). The remaining glycolytic flux was 54% (Table I). Growth was only slightly affected during this period (Fig. 3D). Between 24 and 30 h after induction, all other enzyme activities dropped sharply, and cell growth stopped; cell density decreased during the next period. At 24 h after induction, the residual PGAM transcript level was ⬃30%, with tubulin as a control, whereas the PFK transcript level remained unchanged (data not shown). In all experiments in which ENO was targeted by RNAi, its activity dropped quickly to 45% after 10 h of induction and to 16.0 ⫾ 2.3% after 24 h (Fig. 2E, left panel). The resulting

glycolytic flux was 46% at 24 h (Table I), with a slight growth retardation (Fig. 3E). Growth stopped 48 h after induction, and cells started dying. As in the experiments with PFK, the cellular activity of several other enzymes assayed (HXK, PFK, PYK, and glycerol kinase) changed in a gradual manner, albeit with a delay of ⬎10 h compared with the decrease in ENO activity, and to a lesser extent (⬃0 –25% after 24 h and ⬃40 – 60% after 48 h). However, no significant decrease in the activities of glucose-6-phosphate isomerase and triose-phosphate isomerase was observed (data not shown). At the protein level, enolase decreased to 56, 43, and 29% after 12, 24, and 48 h, respectively, using PEX5 as the stable control (Fig. 4B). Other enzymes (PYK and GAPDH) showed 10 –30% variations during the 48-h induction. With regard to the transcripts, the ENO mRNA had already reached a level of only ⬃15% at 12 h after induction, where it remained at 24 and 48 h. In contrast, the levels of the PFK, PYK, and GAPDH transcripts remained at

Glycolytic Flux Control in T. brucei

28313

FIG. 4. Time-dependent changes in the levels of various enzymes in bloodstream form T. brucei upon targeting depletion of PFK (A) and ENO (B) by RNAi. Protein levels were determined by quantitative Western blotting and normalized to the control PEX5 and are expressed relative to levels in lysates without induction of RNAi (at t ⫽ 0).

would be associated with a rapid decrease in the flux. In contrast, the experiment showed that reduction of PYK activity to 24% led to a halved glucose consumption rate, suggesting that the model overestimates the overcapacity of PYK. Similar observations were made for HXK and PFK, but in these cases, the exp mod % Vmax /% Vmax ratios were even larger. Particularly for PFK, the discrepancy was large. The experiments suggested that the enzyme acquired control much more readily than predicted (observed, 26% flux at 47% activity; and predicted, 100% flux at activities ⬎15%), but the interpretation is complicated by the changes observed also in the activities of some other enzymes, possibly due to regulatory mechanisms (discussed further below). Contrary to the situation for PFK, the experiments with HXK confirmed the presence of a large activity excess (see above), although slightly less than predicted (predicted, 100% flux at more than ⬃15% HXK activity; and observed; 67% flux at 22% activity). In this study, we confirmed the presence of excess HXK, PFK, and PYK activity, predicted by the kinetic model. However, the excess of HXK and PYK appeared to be less than predicted. This discrepancy may be attributed to several factors. First, the excess may be due either to an overestimation of the activities of these enzymes in crude cell lysates compared with those in intact cells or to an underestimation of other enzymes. Indeed, enzymes in crude lysates may be inhibited by compounds used in lysis buffers or (macro)molecules (such as nucleic acids) coming from other cell compartments. Where possible, such effects have been taken into account (Supplemental “Experimental Procedures”) (10). Second, the activity of some enzymes may be regulated by mechanisms that have not yet been identified and therefore were not included in the model. When cells and cell compartments are broken and/or lysates are diluted in a cuvette, such physiological regulatory mechanisms may get lost, consequently causing the discrepancy observed. Research so far has shown a surprising apparent lack of mechanisms that regulate the activity of glycolytic enzymes in trypanosomatids. Mechanisms operating in other organisms that stimulate or inhibit the activities of HXK and PFK are virtually absent in trypanosomes. For example, only AMP has been identified so far as an allosteric activator of T. brucei and Leishmania donovani PFK (44 – 46). This paucity of regulation has been attributed to the compartmentalization of these enzymes inside glycosomes (2, 3, 47). Important activity regulation has been observed only for cytosolic PYK. However, as argued elsewhere (3), it is likely that regulation mechanisms have been overlooked. The existence of such mechanisms, possibly involving covalent modification of enzymes, is currently under study. The results will also be used for further improvement of the model.

Downloaded from www.jbc.org by guest, on February 19, 2013

100% during the entire 48-h period or showed only minor changes (⬍15%) depending on the experiment (data not shown). The data obtained with ENO are reminiscent of those obtained for PFK and suggest a regulation of ENO activity as well as a down-regulation of the activity of other enzymes, including glycosomal ones, as a result of the decrease in the level (or activity) of this cytosolic enzyme. Effect of Enzyme Activities on Glycolytic Flux and Comparison of Experimental Data and Model Predictions—It was previously established experimentally that glucose uptake into the cell exerts considerable control over glycolytic flux in bloodstream form T. brucei, in agreement with the predictions by the kinetic model (14, 20). The experimental data presented above confirm the model predictions that HXK and PFK do not have any control under normal growth conditions. To further check the correlation between predicted and experimentally determined data, the activities of five enzymes studied in this work were separately decreased from their 100% Vmax values, as introduced in the updated model (Supplemental Table S1), to 0%, and the fluxes of glucose consumption and pyruvate production were calculated over the whole range of enzyme activities. The results are plotted as relative fluxes (% J) as a function of the relative enzyme activities (% Vmax) in Fig. 5. The experimentally determined relationship between decreased enzyme activities and flux 24 h after induction of RNAi is also indicated in these plots. Moreover, the enzyme activities (% mod Vmax ) required, according to the model, to give the measured flux (% J) are listed in Table I next to the experimentally exp determined activities (% Vmax ). The predicted and experimentally determined data for both PGAM and ENO correspond exp mod very nicely; the % Vmax /% Vmax ratios (for the glucose consumption flux) are 1.31 and 1.27, respectively. The meaning of the good fit between the model and experiment for PGAM should not be overemphasized because of an underestimation of the specific activity of this enzyme in crude lysates (28). Therefore, when constructing the model, the Vmax value for PGAM was fit such that it was in agreement with the flux data for wild-type cells (see Supplemental “Experimental Procedures”). For ENO, however, the good fit between the model and experiment was not forced and may be interpreted as an indication that the ENO kinetics determined in vitro give a good impression of its in vivo role. exp mod For PYK, a % Vmax /% Vmax ratio of 2.79 was obtained (Table I), suggesting that the true in vivo Vmax is ⬎2-fold lower than the Vmax determined in vitro that was implemented in the model. Fig. 5C clearly shows that, according to the model, the PYK activity has a considerable overcapacity. The model predicts that a decrease in activity to 20% of the Vmax value would not result in flux reduction, but that a further activity decrease

28314

Glycolytic Flux Control in T. brucei

Our data show that decreased activity or expression of PFK caused a decrease in several other glycolytic enzymes (HXK, ENO, and PYK), but not all (glucose-6-phosphate isomerase and triose-phosphate isomerase). This effect seems to be exerted at the level of the activities of these enzymes rather than at the protein expression level. The fact that it was seen only for some enzymes (glycosomal and cytosolic) suggests that it is a specific regulatory effect rather than a nonspecific one due to the stress exerted on the cells. No information is available as yet about the molecular mechanism by which such down-regulation occurs. As a result of these effects on other enzymes, no conclusions could be drawn as to whether the decrease in glycolytic flux results directly from PFK acquiring control or whether the PFK effect on flux and growth is mediated by other enzymes that are down-regulated as a result of the decreased PFK activity. The observations made upon depletion of ENO by RNAi suggest that this enzyme also participates in regulatory mechanisms. As was observed for PFK, ENO also seems to downregulate enzyme activities rather than enzyme levels. This

regulation is specific for only certain enzymes in both the glycosome (HXK, PFK, and glycerol kinase) and the cytosol (PYK), but not for others (glucose-6-phosphate isomerase and triose-phosphate isomerase). Also in this case, no information is available as yet regarding how regulation occurs. It may involve either a key metabolite (for example, phosphoenolpyruvate, the cytosolic concentration of which is expected to decrease when ENO becomes rate-limiting) and/or mechanisms of signaling to the glycosome (for example, coupled to covalent modification of organellar HXK, PFK, and glycerol kinase). No significant effects on the activities or concentrations of other proteins were seen when HXK and PYK were targeted by RNAi, at least not during the initial period. A limited activity decrease was noticed for other enzymes in the HXK RNAi experiments, but only at the later stages. This same effect was much more dramatic when PGAM was depleted. During the first 24 h, when both PGAM activity and glycolytic flux were halved, all enzymes tested, whether glycosomal or cytosolic, remained at 100%, but they decreased considerably afterward (as measured at 30 and 48 h). It seems likely that these later

Downloaded from www.jbc.org by guest, on February 19, 2013

FIG. 5. Relative rates of glucose consumption and pyruvate production as a function of enzyme activities in bloodstream form T. brucei. The glucose and pyruvate fluxes (J) and enzyme activities (Vmax) are expressed relative to values found in non-induced trypanosomes (% J and % Vmax). A–E present the results from titrating the HXK, PFK, PYK, PGAM, and ENO activities, respectively. Curves represent results determined in silico; circles are in vivo results obtained by RNAi and, for PFK, by RNAi and overexpression (as determined 24 h after induction by tetracycline addition to the culture). Gray curves/circles are Jglucose, and black curves/circles are Jpyruvate. Error bars represent S.E.

Glycolytic Flux Control in T. brucei

Acknowledgments—We are grateful to Dr. Christine Clayton for providing T. brucei cell line BF449 and plasmids pHD677, pHD1336, pBS-TubNEO, and pBS-Hyg; to Dr. Christine Clayton and coworkers for advice and assistance in setting up the methodologies for performing reverse genetics in T. brucei in our Brussels laboratory; to Dr. Fre´de´ric Bringaud (Universite´ Victor Segalen, Bordeaux, France) for making available an RNAi PYK construct and for advice about creating RNAi cell lines and various aspects of their analyses; and to Prof. Louis Hue (Universite´ Catholique de Louvain) and Prof. Hans Westerhoff and Arjen van Tuijl (Vrije Universiteit Amsterdam) for useful discussions during the course of this work. REFERENCES 1. Barrett, M. P., Burchmore, R. J., Stich, A., Lazzari, J. O., Frasch, A. C., Cazzulo, J. J., and Krishna, S. (2003) Lancet 362, 1469 –1480 2. Michels, P. A. M., Hannaert, V., and Bringaud, F. (2000) Parasitol. Today 16, 482– 489 3. Hannaert, V., Bringaud, F., Opperdoes, F. R., and Michels, P. A. M. (2003) Kinetoplastid Biology and Disease http:/ /www.kinetoplastids.com/content/ 2/1/11 4. Opperdoes, F. R., and Borst, P. (1977) FEBS Lett. 80, 360 –364 5. Oduro, K. K., Flynn, I. W., and Bowman, I. B. (1980) Exp. Parasitol. 50, 123–135 6. Verlinde, C. L. M. J., Hannaert, V., Blonski, C., Willson, M., Pe´rie´, J. J., Fothergill-Gilmore, L. A., Opperdoes, F. R., Gelb, M. H., Hol, W. G. J., and Michels, P. A. M. (2001) Drug Resist. Update 4, 50 – 65 7. Misset, O., Bos, O. J. M., and Opperdoes, F. R. (1986) Eur. J. Biochem. 157,

441– 453 8. Bakker, B. M., Westerhoff, H. V., and Michels, P. A. M. (1995) J. Bioenerg. Biomembr. 27, 513–525 9. Opperdoes, F. R., and Michels, P. A. M. (2001) Int. J. Parasitol. 31, 482– 490 10. Bakker, B. M., Michels, P. A. M., Opperdoes, F. R., and Westerhoff, H. V. (1997) J. Biol. Chem. 272, 3207–3215 11. Kacser, H., and Burns, J. A. (1973) Symp. Soc. Exp. Biol. 27, 65–104 12. Heinrich, R., and Rapoport, T. A. (1974) Eur. J. Biochem. 42, 89 –95 13. Fell, D. (1997) Frontiers in Metabolism: Understanding the Control of Metabolism, Vol. 2, Portland Press Ltd., London 14. Bakker, B. M., Michels, P. A. M., Opperdoes, F. R., and Westerhoff, H. V. (1999) J. Biol. Chem. 274, 14551–14559 15. Helfert, S., Estevez, A. M., Bakker, B., Michels, P., and Clayton, C. (2001) Biochem. J. 357, 117–125 16. Bakker, B. M., Westerhoff, H. V., Opperdoes, F. R., and Michels, P. A. M. (2000) Mol. Biochem. Parasitol. 106, 1–10 17. Boiteux, A., and Hess, B. (1981) Philos. Trans. R. Soc. Lond. B Biol. Sci. 293, 5–22 18. Rapoport, T. A., Heinrich, R., Jacobasch, G., and Rapoport, S. (1974) Eur. J. Biochem. 42, 107–120 19. Stryer, L. (1995) Biochemistry, 4th Ed., W. H. Freeman & Co., New York 20. Bakker, B. M., Walsh, M. C., Ter Kuile, B. H., Mensonides, F. I., Michels, P. A. M., Opperdoes, F. R., and Westerhoff, H. V. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 10098 –10103 21. Gru¨nberg, J., Sharma, P. R., and Deshusses, J. (1978) Eur. J. Biochem. 89, 461– 469 22. Eisenthal, R., Game, S., and Holman, G. D. (1989) Biochim. Biophys. Acta 985, 81– 89 23. Ter Kuile, B. H., and Opperdoes, F. R. (1991) J. Biol. Chem. 266, 857– 862 24. Seyfang, A., and Duszenko, M. (1991) Eur. J. Biochem. 202, 191–196 25. Chevalier, N., Rigden, D. J., Van Roy, J., Opperdoes, F. R., and Michels, P. A. M. (2000) Eur. J. Biochem. 267, 1464 –1472 26. Kralova, I., Rigden, D. J., Opperdoes, F. R., and Michels, P. A. M. (2000) Eur. J. Biochem. 267, 2323–2333 27. Hannaert, V., Albert, M. A., Rigden, D. J., da Silva Giotto, M. T., Thiemann, O., Garratt, R. C., Van Roy, J., Opperdoes, F. R., and Michels, P. A. M. (2003) Eur. J. Biochem. 270, 3205–3213 28. Guerra, D. G., Vertommen, D., Fothergill-Gilmore, L. A., Opperdoes, F. R., and Michels, P. A. M. (2004) Eur. J. Biochem. 271, 1798 –1810 29. Biebinger, S., Wirtz, L. E., Lorenz, P., and Clayton, C. (1997) Mol. Biochem. Parasitol. 85, 99 –112 30. Sambrook, J., and Russell, D. W. (2001) Molecular Cloning: A Laboratory Manual, 3rd Ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 31. Van den Hoff, M. J., Moorman, A. F., and Lamers, W. H. (1992) Nucleic Acids Res. 20, 2902 32. McCulloch, R., Vassella, E., Burton, P., Boshart, M., and Barry, J. D. (2004) Methods Mol. Biol. 262, 53– 86 33. Misset, O., and Opperdoes, F. R. (1984) Eur. J. Biochem. 144, 475– 483 34. Callens, M., Kuntz, D. A., and Opperdoes, F. R. (1991) Mol. Biochem. Parasitol. 47, 19 –29 35. Bontemps, F., Hue, L., and Hers, H. G. (1978) Biochem. J. 174, 603– 611 36. Sauro, H. M. (2000) Proceedings of the 9th International Meeting on BioThermoKinetics: Jarnac, a System for Interactive Metabolic Analysis, in Animating the Cellular Map (Hofmeyr, J. H. S., Rohwer, J. M., and Snoep, J. L., eds) Stellenbosch University Press, Stellenbosch, South Africa 37. Sauro, H. M., Hucka, M., Finney, A., Wellock, C., Bolouri, H., Doyle, J., and Kitano, H. (2003) OMICS 7, 355–372 38. Eisenthal, R., and Panes, A. (1985) FEBS Lett. 181, 23–27 39. Fairlamb, A. H., Opperdoes, F. R., and Borst, P. (1977) Nature 265, 270 –271 40. Wirtz, E., and Clayton, C. (1995) Science 268, 1179 –1183 41. De Walque, S., Kiel, J. A. K. W., Veenhuis, M., Opperdoes, F. R., and Michels, P. A. M. (1999) Mol. Biochem. Parasitol. 104, 106 –119 42. Chen, Y., Hung, C. H., Burderer, T., and Lee, G. S. (2003) Mol. Biochem. Parasitol. 126, 275–279 43. Durand-Dubief, M., Kohl, L., and Bastin, P. (2003) Mol. Biochem. Parasitol. 129, 11–21 44. Nwagwu, M., and Opperdoes, F. R. (1982) Acta Trop. 39, 61–72 45. Cronin, C. N., and Tipton, K. F. (1985) Biochem. J. 227, 113–124 46. Lo´pez, C., Chevalier, N., Hannaert, V., Rigden, D. J., Michels, P. A. M, and Ramirez, J. L. (2002) Eur. J. Biochem. 269, 3978 –3989 47. Bakker, B. M., Mensonides, F. I., Teusink, B., Van Hoek, P., Michels, P. A. M., and Westerhoff, H. V. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 2087–2092

Downloaded from www.jbc.org by guest, on February 19, 2013

effects are due to a general stress experienced by the cells, as reflected by the decrease in the density of the cell population after 24 h, rather than a specific form of regulation. All results obtained in the RNAi experiments with HXK, PFK, PYK, PGAM, and ENO indicate the existence of an ⬃50% glycolytic flux threshold for sustaining normal growth (at least for periods of ⬃24 h). At fluxes ⬍50%, the levels of several proteins were affected, and cell numbers declined. Obviously, no reliable conclusions may be drawn about flux control and metabolic regulation in such dying cells. The RNAi experiments provide useful information only during a relatively short period, during the period that depletion of the level of the targeted enzyme has not resulted in glycolytic flux below ⬃50% and cell growth has not yet ceased. The finding that trypanosome growth was affected when the glycolytic flux was decreased below an ⬃50% threshold is in agreement with earlier indications that trypanosomes cannot survive for longer periods of time (⬎12 h) by completely anaerobic metabolism with a yield of one (instead of two) ATP molecule/glucose molecule (15). These findings have interesting implications for drug targeting. The corollary is that only partial inhibition of enzymes may be sufficient to kill trypanosomes. The fact that HXK and PYK were shown to be in less excess than predicted by the simulations made with the kinetic model renders it plausible that also the activity of these two enzymes may be decreased to such levels (by using appropriate high affinity inhibitors at low intracellular concentrations) that trypanosomes will be killed. The same conclusion will most likely also hold true for PFK, but more information should be obtained about the mechanism by which glycolytic flux is decreased and about the cause of trypanosome death upon knockdown of this enzyme.

28315

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.