Trypanosoma brucei is the protozoan parasite that causes African trypanosomiasis, a neglected disease of people and animals. Co-metabolite analysis, labelling studies using [methyl-2H3]-methionine and substrate/product specificities of the cloned 24-SMT (sterol C24-methyltransferase) and 14-SDM (sterol C14demethylase) from T. brucei afforded an uncommon sterol metabolic network that proceeds from lanosterol and 31-norlanosterol to ETO [ergosta-5,7,25(27)-trien-3β-ol], 24-DTO [dimethyl ergosta-5,7,25(27)-trienol] and ergosterol [ergosta-5,7,22(23)-trienol]. To assess the possible carbon sources of ergosterol biosynthesis, specifically 13C-labelled specimens of lanosterol, acetate, leucine and glucose were administered to T. brucei and the 13C distributions found were in accord with the operation of the acetate–mevalonate pathway, with leucine as an alternative precursor, to ergostenols in either the insect or bloodstream form. In searching for metabolic signatures of procyclic cells, we observed that the 13C-labelling treatments induce fluctuations between the acetyl-CoA (mitochondrial) and sterol (cytosolic) synthetic pathways detected by the progressive increase in 13C-ergosterol production (control<[2-13C]leucine<[2-13C]acetate<[1-13C]glucose) and corresponding depletion of cholesta-5,7,24-trienol. We conclude that anabolic fluxes originating in mitochondrial metabolism constitute a flexible part of sterol synthesis that is further fluctuated in the cytosol, yielding distinct sterol profiles in relation to cell demands on growth.
- ergosterol biosynthesis
- sterol C24-methyltransferase
- sterol C14-demethylase
- Trypanosoma brucei
Eukaryotes exhibit extensive diversity in the composition of their membrane sterols, which have a fossil record that extends back billions of years. The most valuable feature giving rise to the different sterol patterns in nature are largely due to the presence of various combinations of C24-alkyl groups in the sterol side chain, which are synthesized by a family of structurally similar 24-SMT (sterol C24-methyltransferase) enzymes . Algae, fungi and protozoa typically synthesize 24β-alkyl sterols, whereas vascular plants synthesize 24α-alkyl sterols from the acetate–MVA (mevalonate) or MVA-independent pathways. In previous years, much effort has been devoted to understanding the 24-alkyl sterol metabolic networks that operate throughout Nature [2–4], in order to identify the evolutionary pathway for how these structure specificities arose [5–7] and, in the case of neglected diseases, to develop chemotherapeutic leads designed to selectively inhibit their production [8–11].
For some time our laboratories have been engaged in a study designed to establish the in vitro properties of cloned enzymes that act on sterols and in vivo sterol compositions from organisms of the protozoan order Kinetoplastida [12–17]. In the present study, a further examination of the enzyme properties of the 24-SMT and 14-SDM (sterol C14-demethylase), and the identities and amounts of sterols from one of these members responsible for sleeping sickness, Trypanosoma brucei, has been undertaken; in addition, we studied the incorporation of different 2H- and 13C-labelled carbon sources into the final biosynthesized products.
One reason why we have pursued probing the sterol biosynthesis abilities of T. brucei is to establish the basis of variability in ergosterol homoeostasis as the parasite responds to adaptations in the changing environments of human bloodstream and alimentary gut of the tsetse fly, which to date remains enigmatic. Thus the bloodstream forms of T. brucei have been reported to be blocked at some unknown stage of isoprenoid metabolism, generating ergosterol auxotrophy to live [18–21]. To satisfy the sterol requirements for the bloodstream form, it is proposed that host cholesterol is absorbed by the cell via a receptor-mediated lipoprotein uptake mechanism . As a result, it has long been held that their ability to scavenge cholesterol from their animal hosts, and their inability to undergo sterol biosynthesis in this adaptive stage, makes them resistant to antiparasitic drugs that target ergosterol or ergosterol biosynthesis [20–23].
On the other hand, testing sterol biosynthesis inhibitors that bind specifically to the cloned TbSMT (T. brucei 24-SMT) and TbSDM (T. brucei 14-SDM) enzymes has shown some success [12,13,16]. The existence of a functional 24-alkyl sterol pathway was demonstrated in vitro in both the procyclic and bloodstream forms [14,15]. It is noteworthy that several independent studies indicate that ergosterol (ergosta-5,7,22-trienol) is the major 24-alkyl sterol of T. brucei procyclic cells cultured on FGM (full-growth medium) [18,20,24], as reported in Kinetoplastida generally . However, we discovered that procyclic cells cultured on a lipid (cholesterol)-depleted medium synthesize primarily cholesta-5,7,24-trienol, a set of ergosterol isomers predominated by ergosta-5,7,25(27)-trienol and ergosta-5,7,24(25)-trienol and trace amounts of ergosterol . Despite the fact the trypanosmatids are considered to have a close evolutionary association to the algae [26–28], which might suggest a vestigial cycloartenol MVA-independent [plastid-derived (MEP) methylerthritol phosphate] route to ergosterol , as can occur in the non-photosynthetic alga Prototheca wickerhamii , these organisms are considered to synthesize lanosterol from the acetate–MVA biosynthesis pathway (Figure 1) [19,31,32]. Intriguingly, Goad and co-workers have shown that sterols can be biosynthesized both by the acetate–MVA route and by the direct incorporation of leucine in T. cruzi and Leishmania species [33–35]. Indeed, leucine catabolism in the mitochondria was found to be the major carbon source of sterols in Leishmania mexicana , whereas T. brucei cultured in vitro prefers glucose as the main carbon source for acetyl-CoA (originating in the mitochondrion) and lipid production under standard growth conditions [36,37]. In light of the variant network rigidity resulting from phylogenetic differences across kingdoms as well changes in medium that can affect the nature of sterols in trypanosomatids, a more direct reason for investigating the relationship between structure and biosynthesis was to establish the physiological route of 24-alkyl sterol groups connected to central metabolism utilized in optimal parasite growth of procyclic and bloodstream forms.
The isotopically labelled compounds used, [1-13C]sodium acetate (Cambridge Isotope Laboratories), [2-13C]leucine (ISOTEC), [1-13C]glucose (Cambridge Isotope Laboratories) and [methyl-2H3]methionine (Sigma–Aldrich), contained 99% atom enrichment. [27-13C]Lanosterol (99% 13C) was prepared as described previously . [methyl-3H3]AdoMet (S-adenosylmethionine; 10–15 Ci/mmol), diluted to 10 μCi/μmol for the activity assays, was purchased from PerkinElmer.
Cell cultures and growth studies
T. brucei strain 427 procyclic-form cells were grown in SDM-79 medium supplemented with 10% heat-inactivated FBS (fetal bovine serum; Atlanta Biologicals), hereafter referred to as FGM, at 27°C. Bloodstream-form cells were maintained in HMI-9 medium supplemented with 10% heat-inactivated FBS (Atlanta Biologicals) and 10% Serum Plus (SAFC Biosciences), which is also a FGM. For preparation of a CDM (cholesterol-depleted medium), procyclic forms in SDM-79 medium were supplemented with 10% heat-inactivated lipid-free FBS (Sigma–Aldrich). For a low-cholesterol medium, the bloodstream-form cells were grown in HMI-9 medium containing 10% lipid-free heat-inactivated FBS (Sigma–Aldrich) and 10% Serum Plus, which contains serum lipids equivalent to 2% of FBS. Cultures were harvested by centrifugation. (5000 g for 10 min at 4°C) Cell densities were determined using a Neubauer haemocytometer counter. Growth curves were performed in triplicate and the variation in growth at each data point never exceeded 10%.
Sterol compounds, referenced to the retention time of cholesterol in capillary GC at 13.8 min and HPLC at 22 or 37 min (analytical compared with semi-preparative C18-column), were quantified by integration of the detector signal [FID (flame ionization detector) in GC and UV at 210 in HPLC], from 10 to 30 min. Products were routinely identified by their retention times in GC and electron-impact spectrum with those of reference specimens. For select products, unambiguous identification was established after HPLC (Phenomenex C18-column linked to a diode array detector which provided UV spectra) fractionation followed by GC/MS (ZB-5 capillary column of 30 m coupled to a HP LS 6500 gas chromatograph interfaced to a 5973 mass spectrometer) and NMR analysis [spectra measured on deuterochloroform solutions using a Varian Unity Inova 500 MHz spectrometer operating at 500 MHz for protons and 125 MHz for 13C nuclei; chemical shifts were referenced to chloroform at δ=77.00 and reported as δ (p.p.m.)] [14,30]. For biosynthetically 13C-labelled sterols, broad-band de-coupled 13C-NMR spectra were measured as follows: 32520.3 Hz spectrum width, 90° pulse, 1.3 s acquisition time, 70000 transients, 84618 K data points, line broadening of 0.5 Hz before Fourier transformation and a relaxation delay of 6 s with a repetition time of 3.3 s.
Sterol analysis, metabolic labelling and evaluation of 13C/2H-enrichment in sterol
Two sets of experiments were performed to determine the sterol composition of T. brucei cells. For the first set of experiments to isolate sterols for chemical characterization, procyclic cells were cultured on FGM to yield approximately 1–2×109 cells. T. brucei cell pellets were saponified directly in 10% KOH in 80% aqueous methanol at the reflux temperature for 1 h, which yielded total sterol (free plus esterified sterols). The neutral lipids obtained by dilution with water and extraction with hexane (Fisher) after the saponification were analysed by GC (3% SE-30 pack column operated isothermally at 245°C). Total sterols in the non-saponifiable lipid fraction were purified on an Agilent 1100 series HPLC system coupled to a diode array detector. The total sterols were loaded on to a C18 reversed-phase analytical column (Phenomenex, 4 μm) eluted isocratically with methanol at a flow rate of 1 ml/min. The standard for HPLC data was cholesterol, and the rates of movement are given relative to cholesterol (αc). HPLC fractions in the sterol region of the chromatogram (αc=0.5–2.0) were collected and analysed by GC/MS with a Hewlett-Packard LS 6500 gas chromatograph at 70 eV. GC was performed using an Agilent ZB-5 column (30m×25 μm in diameter) and in some cases by 1H-NMR (500 MHz) recorded in deuterochloroform solutions using a Varian INOVA 500 spectrometer. Cholesterol was the standard for determination of the RRTc (retention time relative to the retention time of cholesterol). Quantification of the amounts of unlabelled sterols was accomplished by GC with a standard curve for cholesterol. To several cell pellets undergoing the saponification process, 5α-cholestane was added as an internal standard. Using GC, cholesterol at 1 ng, or with HPLC, ergosterol at 1 ng, can be detected in the sample preparation .
For the metabolic labelling experiments, cells cultured in FGM containing physiological concentrations of the 13C- or 2H-labelled compound in place of the corresponding unlabelled compound was added to the growth medium and the cultures left at 27°C for 72 h. The total sterols in the organic extract after saponification of the cells were analysed as described above. For determination of the stable IEF (isotope enrichment factors), the extent of 2H and 13C incorporation were determined by GC/MS according to the calculations of Masse et al. . The 13C-NMR spectra of the mixture of [13C]ergostenols obtained from an incorporation experiment with [1-13C]glucose and a reference specimen of yeast ergosterol were recorded at 500 MHz in deuterochloroform under identical conditions using 16C in ergostenol as a reference. Relative 13C-abundance of the individual carbon atoms was then calculated by comparison of 13C signal integrals between 13C-labelled and unlabelled material.
Incorporation of [1-13C]glucose
Glucose-free SDM-79 medium (300 ml) containing D-[1-13C]Glucose (Cambridge Isotope Laboratories) (1.0 g/l) were inoculated with T. brucei 427 procyclic-form cells (3×106/ml) and were allowed to grow in the presence or absence of lanosterol (38 μM) at 27°C in a 1 litre conical flask with constant stirring. Cells were harvested after 72 h, washed twice with phosphate-buffered saline (50 ml each) and stored at −70°C.
Incorporation of L-[2-13C]leucine
SDM-79 medium was prepared using leucine-free MEM (minimal essential medium; Sigma–Aldrich) and supplemented with L-[2-13C]leucine (ISOTEC) (0.052 g/l). T. brucei 427 procyclic cells were grown in this medium in the presence or absence of lanosterol (38 μM) for 72 h and harvested as described above.
Incorporation of [1-13C]sodium acetate
The SDM-79 medium (300 ml) supplemented with [1-13C]sodium acetate (2 mM) was inoculated with 3×106 cells/ml and allowed to grow in the presence or absence of lanosterol (38 μM) at 27°C in a 1 litre conical flask with constant stirring. Cells were harvested as described above.
Incorporation of [methyl-2H3]methionine
The methionine-free SDM-79 medium (300 ml) supplemented with [methyl-2H3]methionine (0.07 g/l) was inoculated with 3×106 cells/ml and allowed to grow in the presence or absence of lanosterol (38 μM) at 27°C in a 1 litre conical flask with constant stirring. Cells were harvested as described above.
RESULTS AND DISCUSSION
Sterol composition of procyclic T. brucei
To elucidate the full complement of metabolites in the sterol metabolome from T. brucei, we conducted a thorough analysis of the sterol composition of procyclic cells cultured on FGM or CDM for 6 days (Supplementary Figure S1 at http://www.BiochemJ.org/bj/443/bj4430267add.htm). To accomplish the present study, a combination of chemical approaches was employed including: HPLC (to separate and isolate sterols originating in the non-saponifiable lipid fraction) coupled to a UV detector to gain information about the position and number of double bonds in the sterol molecule; GC profiling of the total sterol fraction relative to the chromatographic behaviour of authentic standards in our steroid collection was used to provide a tentative identification of these compounds. Proof of the sterol identity detected in GC was obtained by MS. In this spectral analysis, the molecular ion and fragmentation pattern of the compound provide relevant structural information; and finally 1H-NMR can furnish a complete stereochemical description of protozoan sterols generated in pure form by HPLC.
Procyclic cell growth was significantly inhibited in CDM. After day 4, cell number was decreased and upon longer incubation cells died. However, growth of the bloodstream form was minimally affected in CDM, indicating that the presence of 2% serum lipid from the Serum Plus is sufficient for cell growth. Bloodstream cells did not grow in the absence of Serum Plus. The major sterols (<95%) of the non-saponifiable lipids, identified by GC/MS, were qualitatively similar except the proportion of cholesterol in the sterol samples varied with culture medium; FGM and CDM yielded approximately 55% and 35% cholesterol respectively (Figure 2). The major endogenously formed sterols (several of which are shown in Figure 1) were cholesta-5,7,24-trienol (Figure 1, compound 7) with lesser amounts of ETO [ergosta-5,7,25(27)-trien-3β-ol] (Figure 1, compound 12) and ergosta-5,7,24(25)-trienol (not shown). Ergosterol (ergosta-5,7,22E-trienol) (Figure 1, compound 13) was routinely detected only as a trace compound co-migrating in GC with cholesta-5,7,24-trienol. The two sterols are readily differentiated by their mass spectra. Thus the molecular ion for ergosterol appears at M+ 396 a.m.u. (atomic mass unit), whereas for cholesta-5,7,24-trienol it appears at M+ 382 a.m.u.
A description of the distributions of 15 HPLC-purified sterols from procyclic cells cultured on CDM has been published previously ; key intermediates detected in the sterol composition were lanosterol (Figure 1, compound 1), 31-norlanosterol (Figure 1, compound 4) and zymosterol (Figure 1, compound 8) present at less than 2% of the total sterols. During the course of the present study, one additional new compound was isolated from the HPLC of the total sterol fraction from procyclic cells grown on FGM, 24-DTO [dimethyl ergosta-5,7,25(27)-trienol] which exhibited chromatographic behaviour (RRTc, 1.36), UV (λmax=282 nm) and mass spectra (M+ 410 a.m.u.) identical with an authentic sample generated by TbSMT catalysis of cholesta-5,7,24-trienol . We also isolated ETO from the HPLC to prove the stereochemistry of the C24-methyl group in the sterol side chain. The 1H-NMR spectrum of ETO afforded signals at δ H 18, 0.621 (s), H 19, 0.951 (s), H 21, 0.942 (d), H 26, 1.648 (br. s), H 27, 4.672 (br. s), H 28, 1.01 (d)- C24 methyl group is β-oriented, H 6, 5.577 (dd), H 7, 5.380 (m) p.p.m., consistent with its proposed structure [40,41].
An important purpose of repeating the sterol analyses of fresh cells cultured at Meharry Medical College (we previously studied procyclic cells grown at Rockefeller University ) was to show clearly that T. brucei normally synthesizes three major uncommon sterols, cholesta-5,7,24-trienol, ergosta-5,7,25(27)-trienol and ergosta-5,7,24-trienol, and that the proportion of them relative to cholesterol can vary dramatically with the availability of lipid (cholesterol) in the growth medium. Equally important to us was to confirm that cells can produce ergosterol, albeit at trace levels and that the usual practice of culturing T. brucei on FGM or CDM fails to affect the low abundance of this compound in the sterol mixture. On the basis of these results, we surmise that changes in sterol homoeostasis associated with the procyclic growth conditions, which are marked by high and low concentrations of cholesta-5,7,24-trienol (Figure 2, peak 2), may be the result of a heretofore unrecognized regulation mechanism of post-squalene enzyme activities.
Sterol biosynthesis in procyclic cells
In the present study, GC/MS was used to determine the number of deuterium atoms from [methyl-2H3]AdoMet (synthesized in vivo from [methyl-2H3]methionine) incorporated into the sterol side chain at carbon-28. Thus procyclic cells were cultured in the presence of [methyl-2H3]methionine and neither the growth response nor total sterol content of the cells changed compared with the control. GC/MS analysis of total sterols from [methyl-2H3]methionine-treated cells revealed significant incorporation of deuterium into the C24-methyl sterols ETO and ergosta-5,7-24-trienol showing the incorporation of none and three deuterium atoms (M+ 396 and 399 a.m.u.) and in the 24-dimethyl sterol DTO showing the addition of six deuterium atoms in the mass spectrum (M+ at 410 shifted to M+ 416 a.m.u.). (Figure 3); as expected, no deuterium was detected in cholesterol (Table 1). These results provide convincing evidence for the intermediacy of ergosta-8,25(27)-dienol and ergosta-8,24(25)-dienol in the biosynthesis of ETO and DTO respectively, and rule out ergosta-8,24(28)-dienol (fecosterol) as an intermediate to the major 24-alkyl sterols that accumulate in these cells (Figure 1). Ergosterol, detected as a trace compound in the sterol mixture, was not labelled from these incubations.
When [27-13C]lanosterol was added (5–100 μM) to the culture medium, growth was inhibited with an IC50 value of 38 μM. In spite of significant accumulation of the exogenous lanosterol into the cells (approximately 1:2 ratio of cholesterol/lanosterol in the total sterol mixture), the sterol profile remained similar to the control; ETO and cholesta5,7,24-trienol were labelled to a minor extent and ergosterol not at all (Table 1). We repeated this experiment by incubating [2-13C]acetate plus unlabelled lanosterol with the hope to trap biosynthetically labelled [13C]lanosterol and establish the initial biosynthetic sterol in the pathway. Quite unexpectedly, the sterol composition of the treated cells was modified to contain several distinguishing factors. First, it contains a large abundance of a GC peak that corresponds to cholesta-5,7,24-trienol (Figure 3, peak 2) and a new GC peak that elutes in the tail of peak 2. However, inspection of the mass spectra of these compounds revealed them to be different from sterols accumulating in the control; peak 2 contained [13C]ergosterol and peak 5 contained [13C]ergosta-5,7,-dienol (Figure 4 and Supplementary Figure S2 at http://www.BiochemJ.org/bj/443/bj4430267add.htm). Secondly, the labelling results of this experiment showed less than 5% labelling of lanosterol, although ergosterol and ETO (peak 3) were heavily labelled. Incubation of [2-13C]acetate yielded an unprecedented increase in ergosterol synthesis by at least 100-fold over the control. Lanosterol accumulated to levels similar to that of cholesterol by these cells is apparently not actively entering into the sterol synthetic pathway.
Sterol biosynthesis in the bloodstream form
Given the controversy and the conflicting data, a simple experiment is required to determine whether bloodstream-form cells can synthesize 24-alkyl sterol de novo. One such experiment involves cultivating bloodstream cells in the presence of [1-13C]glucose followed using a combination of HPLC and GC/MS to isolate and identify the biosynthetically formed sterol products. For this work, bloodstream-form cells were cultured on FGM or CDM supplemented medium which afforded similar growth responses and the cells yielded similar sterol profiles. Thus in these cells cholesterol was routinely the major compound (<99% total sterol) in the chromatogram at RRTc 1.0, M+ 386, followed by trace amounts of campesterol at RRTc 1.14, M+ 400 and sitosterol at RRTc 1.29, M+ 414 (Figure 5). Although radioactive acetate has been shown not to be incorporated into bloodstream-form sterol, there nonetheless is a suggestion that radioactive glucose might be actively incorporated into bloodstream-form lipids generally . For this reason, we administered [1-13C]glucose to the bloodstream form. As expected, in the initial GC/MS analysis of the bloodstream sterols none of the sterols were labelled, which indicates that they are derived from the host or the animal diet chow. However, after the total sterol fraction was injected into HPLC equipped with a UV detector a 13C-labelled compound could be detected by GC/MS. The fraction eluting just before cholesterol at αc 0.9–1 (campesterol and sitosterol elute after cholesterol in HPLC) contained a single [13C]sterol that possessed a UV spectrum at λmax=282 nm, RRTc 1.14 and mass spectrum M+ 398 (Figure 4), consistent with the structure ergosta-5,7,-dienol identified from the previous incubation of [2-13C]acetate plus lanosterol to procyclic cells. The amount of [13C]ergosta-5,7,-dienol in the cell occurs at approximately 0.5 fg/cell, compared with the total sterol content of approximately 70 fg/cell, suggesting that the endogenously formed ergostenol functions as a ‘metabolic’ or signal molecule, whereas the cholesterol from the host functions as the membrane insert. The specificity of labelling of [13C]ergosta-5,7,-dienol is supported by three previous observations: (i) two minor sterols, tentatively identified as cholesta-5,7,24-trienol and ergosta-5,7,24-trienol, were detected in bloodstream forms; (ii) zymosterol accumulates in the bloodstream form after incubation with the 24-SMT inhibitor 25-azalanoterol ; and (iii) the 24-SMT gene detected in bloodstream form by Northern blot analysis is expressed 3-fold less than in procyclic cells .
Catalytic properties of TbSDM and TbSMT
From our previous work, the favoured TbSDM substrate was shown to be 14α,31α-dimethyl ergosta-8,24(28)-dienol (obtusifoliol) which presumably made this sterol catalyst plant-like in its substrate recognition . However, key compounds now known to be native to T. brucei, 31-norlanosterol and 14α-methyl zymosterol, were not evaluated in our previous studies. To establish the optimal substrate of TbSDM, we made further measurements (kcat/Km) on these sterols and as a control repeated the incubations of substrates tested before using the TbSDM. Thus 31-norlanosterol was found to be the optimal substrate showing steady-state kinetics of approximately 4 min−1/9 μM yielding a catalytic competence of kcat/Km of 0.43; obtusifoliol and 14-methyl zymosterol yielded catalytic competences of 0.37 and 0.13 respectively, and lanosterol was not productively bound to the enzyme. When TbSDM was incubated with 31-norlanosterol, the enzyme-generated product was identified by GC/MS (M+ 396) and HPLC-UV (250 nm) analysis to be 31-nor cholesta-8,14,24-trienol (Figure 6A, peak 3). A second product (Figure 6A, peak 2) was detected from this incubation eluting before the major metabolite and its identity is under investigation.
The cloned TbSMT has been evaluated with the same set of substrates tested with the TbSDM, in addition to zymosterol and cholesta-5,7,24-trienol which naturally occur in the T. brucei sterol composition . The production of hexane-soluble products by the TbSMT as a function of substrate concentration in the range of 1–100 μM exhibited saturation kinetics, consistent with our previous findings . Computer-assisted data analysis for zymosterol afforded a Km value of 47 μM, which is similar (+/− 5 μM) to that of the other substrates tested. However, the catalytic-centre activity for zymosterol is optimal at 0.6 min−1, and for 14-methylzymosterol at 0.3 min−1 and cholesta-5,7,24-trienol at 0.2 min−1. Conversion of 14-methyl zymosterol by TbSMT led to the same type and proportion of multiple products (Supplementary Figure S3 at http://www.BiochemJ.org/bj/443/bj4430267add.htm) that we previously reported from TbSMT catalysis of zymosterol or cholesta-5,7,24-trienol [14,40]
Origin of carbon atoms in ergostenol biosynthesis
The introduction of 13C-labelled substrates has had an enormous impact on the study of sterol biosynthesis, in part because the investigator studies compounds containing stable isotopes, not radioactive compounds, that can be identified by MS and NMR techniques. The ability to detect incorporation of the 13C-label by MS and to locate sites of labelling by 13C-NMR directly has led to the elucidation of relevant precursor–product relationships. Thus incubation of [1-13C]glucose, [2-13C]acetate or [2-13C]leucine to procyclic cells yielded marked changes in the capacity for sterol biosynthesis compared with a control incubated with the corresponding non-labelled carbon source. Thus, in the GLC chromatogram of the total sterols from [1-13C]glucose incubation, a new GC peak appeared (5% of the mixture) at RRTc 1.14 with M+ 398 after the peak corresponding to cholesta-5,7,24-trienol at RRTc 1.11. The chromatographic behaviour and mass spectrum of this compound agrees with that of an authentic specimen of ergosta-5,7-dienol available from our sterol collection. In these studies, we also observed that the GC peak corresponding to cholesta-5,7,24-trienol broadened and progressively increased in height from incubation of labelled leucine, acetate and glucose respectively. Careful GC/MS analyses of this GC peak, revealed that the change in peak shape was due to increases in ergosterol (M+ 396) which co-migrates with cholesta-5,7,24-trienol (M+ 382). The ratio of cholesta-5,7,24-trienol to ergosterol in the control and 13C-incorporation experiments varied from: (i) trace to 99% from control; to (ii) 1 to 99 from [13C]leucine; to (iii) 1 to 1 from [13C]acetate; and (iv) 1 to 10 from [13C]glucose. The 13C-label was incorporated variably into ergosterol, ETO and cholesta-5,7,24-trienol showing isotope enrichment factors for each labelled compound that varied from 1.30 to 4.27% (Table 1 and Supplementary Figure S4 at http://www.BiochemJ.org/bj/443/bj4430267add.htm). In the [13C]leucine experiment, ETO, cholesta-5,7,24-trienol and ergosterol were weakly labelled (<5%). The sterol products from labelled glucose and acetate possessed the same isotopomeric composition, suggesting a common origin of their biosynthesis (Supplementary Figure S3). In contrast, cholesterol was not labelled from these incubations. These sterol-dependent differences from incubation of different 13C-labelled carbon sources are probably explained by mitochondrial and cytosolic involvement in metabolite processing affected by two pools of acetyl-CoA exchanging rapidly, i.e. they are in equilibrium. Taken together, these results signify that glucose is a better carbon source than acetate and that leucine is not a relevant physiological carbon source in sterol biosynthesis of procyclic cells.
Previous work has shown the 13C-label in [1-13C]glucose is found at C2 of acetate . Subsequently, Δ3-IPP (isopentyl diphosphate), which is formed via the acetate–MVA pathway, becomes labelled in positions at C2, C4 and C5, ultimately giving rise to ergosterol with 15 carbons labelled in the sterol frame at the specific locations shown in Figure 1 for ergosterol 13a. However, for [1-13C]glucose conversion via the plastid-derived MVA-independent route, the pathway starts with pyruvate and glyceraldehyde 3-phosphate from glycolysis and leads to Δ3-IPP labelled at positions C1 and C5. The resulting ergosterol molecule will possess 10 labelled carbon atoms (Figure 1, ergosterol 13c). 13C incorporation into ergosterol from incubation with [2-13C]leucine can also generate six or 11 labelled carbon atoms (Figure 1, ergosterol 13b), depending on the route of leucine catabolism. In the present study, labelled ergosterol and ETO produced by procyclic cells was purified as an approximate 50:50 mixture by HPLC. GC/MS analysis of the labelled compounds showed similar levels of 13C incorporation (Figure 7). 13C-NMR of the sterol mixture showed site-specific labelling of 15 13C-enriched carbon atoms (carbon number, normalized peak area to C16) at: δ 38.39 (C1, 7.25), 70.45 (C3, 5.88), 139.79 (C5, 1.75), 116.31 (C7, 4.00), 46.27 (C9, 4.75), 42.84 (C13 and C24, 4.75), 23.00 (C15, 4.5), 55.75 (C17, 4.5), 12.06 (C18, 4.00), 16.29 (C19, 6.25), 21.12 (C21, 5.50), 135.7 (C22, 2.75), 33.10 (C25, 1.38), 19.66 (C26, 5.25), 19.96 (C27, 5.00)/1.09.37 (C27, 1.27) p.p.m. The enriched signal corresponding to C27 for ergosterol resonated at 19.96, and for ETO it resonated at 1.09.37, whereas there is no enhanced signal associated with C28 in these 24-alkyl sterols for biosynthetic reasons .
Sterol flux analysis for procyclic and bloodstream forms
The results of the present study from labelling data of sink metabolites known to be synthesized in different compartments demonstrate a sterol metabolic network that consists of three distinct, but coordinated, sets of pathways starting with glucose in the central metabolic pathway, proceeding through the isoprenoid channel consisting of acetyl-CoA, MVA and IPP which feed ergosterol synthesis from lanosterol, summarized in Figure 8. These pathways contribute to the production of energy, intermediate metabolites and membrane inserts necessary for cellular growth. In our experiments, the procyclic cells consistently used glucose more efficiently than acetate or leucine as carbon sources. Although acetate may not be a suitable precursor for sterols in the bloodstream form , as shown in the present study glucose can replace acetate as an effective carbon source in 24-alkyl sterol biosynthesis. In addition, relating the measured flux changes to sterol compositions in the different cell types revealed the natural metabolic stabilities controlling sterol homeostasis correlate with an active 24-alkyl sterol synthesis pathway. However, the output of 24-alkyl sterol production can vary markedly. Thus flux from separate compartments affording ergosterol as the major product will be determined by the respective substrate availability and culture conditions to grow the organism.
The combination of chemical and kinetic analyses of the sterol biosynthesis pathway in T. brucei suggest that 24-alkyl sterols are formed by a conical route of ten enzymatic reactions (Figure 8). In this model, the sterol C14-demethylation reaction is part of the core sterol biosynthesis pathway which operates in cholesterol synthesis generally , and the sterol C24-methylation reaction provides the critical slow step that can yield phyla-specific 24-alkyl sterol patterning unique to T. brucei. In addition, the 24-SMT activity can serve to control the ratio of 24-alkyl sterol to cholestenols synthesized in the cells, as shown through genetic modulation of the Arabidopsis sitosterol pathway engineered to produce cholesterol . To complete ergosterol biosynthesis from ETO, a sterol C22-desaturase (Figure 8, reaction 13→14) and sterol C25(27)-reductase enzyme (the corresponding gene of this enzyme is not evident in GenBank®; Figure 8, reaction 12→13) are required; notably, neither the Δ25(27)-reductase enzyme nor multiple product TbSMT has been reported in the T. cruzi sterol biosynthesis pathway. The formation of DTO from ergosta-5,7,24-trienol (Figure 8, reaction 15→16) is significant because it has not been reported from any other protozoan. The absence of 22-dehydro DTO from cells is most likely due to the extra C24α methyl group in the sterol side chain of the 24-dimethyl sterol, which is known to sterically inhibit the introduction of the Δ22 bond in yeast ergosterol biosynthesis .
During the course of the present study, we observed for the first time metabolite-mediated modifications that steer the branch point intermediate zymosterol to favour C24 methylation to form ergosta-8,25(27)-dienol en route to forming ergosterol through differentially expressed genes and enhanced enzyme activities in T. brucei. This cytosolic diversion of carbon flux, which appears to be initially mediated by 24-SMT action, was independently observed in the varying growth experiments of procyclic cells. In cells cultured on FGM, TbSMT is active such that zymosterol undergoes C24 methylation, thereby significantly increasing the production of ergosta-5,7,25(27)-trienol and ergosta-5,7,24-trienol in the sterol mixture. In contrast, cells cultured on CDM, where TbSMT activity appears less efficient, but the core sterol enzymes continue to be highly active, the major portion of the cholesta-8,24-dienol (zymosterol) converts into the final product cholesta-5,7,24-trienol, which lacks a C24 alkyl group. Despite dramatic changes in the nature of accumulated end products in procyclic cells, the stability of the central metabolism to continue to generate carbon for 24-alkyl sterol biosynthesis in the adaptive bloodstream form was confirmed. Interestingly, the up-regulation of ergosterol production and remodelling of the membrane architecture in procyclic cells did not correspond to any obvious changes in growth consistent with the broad sterol requirements of T. brucei. However, retention of 24-alkyl sterol biosynthesis in the bloodstream form, shown from incubation of [1-13C]glucose, suggests a structure-based selectivity whereby the 24-alkyl sterol can function differently from cholesterol, perhaps as a ‘metabolic’ or signal molecule to spark cell proliferation, as shown in yeast [45,46].
The current study documents a novel 24-alkyl sterol biosynthesis pathway in which all but one enzymes [sterol C25(27)-reductase] have been detected by protein–protein BLAST searches of the catalogue of predicted T. brucei sterol biosynthetic enzymes (results not shown). In both the procyclic and bloodstream form, the energy-expensive C24-methylation pathway continues to play an essential role in sterol biosynthesis so long as glucose metabolism is not limiting. The reason why T. brucei synthesizes multiple 24-alkyl sterol products is not clear; if some of them are found to be functionally redundant, such as fecosterol and DTO, then we have a glimpse of sterol evolution at play in which separate routes are available for sterol biosynthesis but only one of them yields physiologically relevant end products. We surmise that the major factors affecting changes in the proportion of ergostenols to cholestenols in procyclic and bloodstream forms may operate at the level of compartmented metabolic fluxes linked to gene expression regulation through a sterol-sensing mechanism of the sort utilized by opportunistic pathogens . More detailed studies on the factors that modulate biosynthesis of the unusual ergostenols in trypanosomatids, and the selective pressures that drive this process, are warranted.
Fernando Villata, Michael Waterman, Galina Lepesheva, Minu Chaudhuri and W. David Nes designed the experiments and analysed data; Craigen Nes, Ujjal Singha, Jialin Liu, Kulothungan Ganapathy, Galina Lepesheva and Minu Chaudhuri performed the experiments; and W. David Ness wrote the paper.
This work was supported, in whole or in part, by the National Science Foundation [grant number MCB-0929212 (to W.D.N.)], and the National Institutes of Health [grant numbers GM 067871 (to M.R.W. and G.I.L.), AI 080580 (to F.V.) and GM 081146 (to M.C.)].
Abbreviations: AdoMet, S-adenosylmethionine; a.m.u., atomic mass unit; CDM, cholesterol-depleted medium; 24-DTO, dimethyl ergosta-5,7,25(27)-trienol; ETO, ergosta-5,7,25(27)-trien-3β-ol; FBS, fetal bovine serum; FGM, full-growth medium; IPP, isopentyl diphosphate; MVA, mevalonate; RRTc, retention time relative to the retention time of cholesterol; 14-SDM, sterol C14-demethylase; 24-SMT, sterol C24-methyltransferase; TbSDM, T. brucei 14-SDM; TbSMT, T. brucei 24-SMT
- © The Authors Journal compilation © 2012 Biochemical Society