Metabolic reprogramming during the Trypanosoma brucei life cycle

2.1. Multiple carbon sources for energy production

All living organisms use adenosine triphosphate (ATP) as an intracellular energy source. ATP is generated by the catabolism (breakdown) of nutrients. The most common nutrients or carbon sources are carbohydrates (such as glucose), fatty acids and amino acids.

Most organisms derive energy from the breakdown of glucose, by a process known as glycolysis, a universal and evolutionarily ancient metabolic pathway, which converts glucose (6-carbon) into pyruvate (3-carbon). Under aerobic conditions, pyruvate can undergo further breakdown to acetyl coenzyme A (acetyl-CoA) (2-carbon) and subsequently to carbon dioxide (CO₂) via the tricarboxylic acid (TCA) cycle with the concomitant production of reducing equivalence (NADH and FADH₂) and GTP. Transfer of electrons from these reduced cofactors to oxygen via an electron transport chain generates a proton electrochemical gradient across the mitochondrial inner membrane that is used to generate ATP by a membrane-bound ATP synthase, collectively a process termed oxidative phosphorylation (OXPHOS). The complete oxidation of each glucose molecule leads to the production of about 36 ATP molecules, showing how OXPHOS is a very efficient mechanism of producing energy.

In the absence of oxygen, the glycolysis product (pyruvate or phosphoenolpyruvate) can be further metabolised by fermentation into excreted end-products, such as lactate (for instance, in humans during running) and ethanol (for instance, in yeast) in the cytoplasm, leading to the net production of two ATP molecules per molecule of glucose consumed. Although the flux through fermentation can be very high, the pathway is energetically inefficient in terms of ATP production. In the absence of oxygen, some microorganisms use nitrate ions, sulfate ions, and carbon dioxide as final electron acceptors, in a process named anaerobic respiration. For example, the final product of glycolysis could be converted into acetyl-CoA, which enters the TCA cycle or is converted into acetate. The electrons are then donated to the final acceptor through the mitochondrial electron transport chain.

In many organisms, fatty acids can be catabolised via β-oxidation in the mitochondria to again generate the 2-carbon unit acetyl-CoA, which feeds into the TCA cycle and OXPHOS. Fatty acid β-oxidation of a palmitate molecule (a fatty acid with 16 carbons that is very abundant in mammalian adipocytes) can generate 106 ATP molecules. The balance between making and breaking down fatty acids is tightly regulated.

Amino acids can also contribute to total energy production by oxidation to urea and CO₂. The first reaction is the removal of the amino group by transaminases. While the amino group enters the urea cycle, the ketoacid carbon skeletons typically enter the TCA cycle and fuel OXPHOS.

The relative abundances of sugars, amino acids, and fatty acids along with the availability of sufficient oxygen to use OXPHOS influence which metabolic pathways are preferentially used to produce ATP. Thus, the metabolic profile of a cell is a consequence of the regulated expression of pathway-specific proteins and associated transporters in response to extracellular nutritional and environmental conditions¹⁰.

2.2. Metabolic adaptations in eukaryotes

Textbooks on metabolism explain that in nutrient-rich conditions, model unicellular organisms undergoing exponential growth often use fermentation¹¹. Proliferating cells in a multicellular organism also metabolise glucose primarily through glycolysis, secreting ethanol, lactate, or another organic acid such as acetate. When unicellular organisms are starved of nutrients, they switch and rely primarily on oxidative metabolism, as do terminally differentiated cells in a multicellular organism. It is no surprise that there are many exceptions to these generalised concepts, and as we will describe below (Section 4), T. brucei is a quintessential example of these exceptions.

The metabolism of cells is a highly regulated process that is influenced by numerous extracellular factors. For example, yeast uses glucose from the environment as their preferred carbon source. Even in the presence of oxygen, glucose is converted into excreted ethanol, with a low yield of ATP production. Although this process may seem wasteful, it is a highly efficient way to support exponential growth. When glucose levels are low, yeast undergoes a diauxic switch to another carbon source, ethanol, which requires an alteration of its mitochondrial metabolism. By using this alternative carbon source, cells are able to continue to grow and divide, but at a significantly reduced rate¹².

In mammals, most non-proliferating differentiated cells use glycolysis and OXPHOS to generate ATP and convert glucose to CO₂ and H₂O. However, most proliferating cancer cells convert glucose into pyruvate and lactate (3-carbon) even under aerobic conditions, a phenomenon known as the Warburg effect, named after its discoverer^11,13. Although this fermentation-like process is intrinsically energetically less efficient, these cells use far higher rates of glycolysis to meet their higher metabolite demand as they divide faster. This metabolic reprogramming allows cancer cells to rapidly produce the building blocks and increase total biomass for a faster propagation time¹⁴.

Major metabolic changes also take place when immune cells are activated and initiate proliferation. When T cells are activated upon infection or inflammation, gene expression is reprogrammed, resulting in rapid growth, proliferation and the acquisition of new effector functions. Effector T cells, like cancer cells, rely upon aerobic glycolysis when proliferating¹¹. In contrast, T cells destined to become memory cells maintain an oxidative metabolism, which allows them to keep their quiescence and longevity. Regulatory T (Treg) cells predominantly use OXPHOS for development and survival^15,16, while activated B cells show increased glucose uptake and induction of glycolysis¹⁷. These examples demonstrate how malleable metabolism is, as cells react to environmental factors and signals to acquire new functions.

3.1. Trypanosomes have a single mitochondrion and glycosomes

Trypanosomes are characterised by the presence of a dense network of circularised interlocking rings of mitochondrial DNA termed the kinetoplast, located within the large, single mitochondrion of the cell. The single mitochondrion in PFs has a highly defined branched structure with discoid cristae, whereas in the B-LS forms the organelle is a less well-developed narrow tubular structure with an acristate morphology similar to that of the promitochondrion of anaerobic yeast^18,19. T. brucei also contains peroxisome-like organelles, named glycosomes, which contain the first six (PF) or seven (B-LS) glycolytic enzymes³. Since the glycosomal membrane is impermeable to ATP, no net ATP production occurs inside these organelles. Thus, net ATP production from glycolysis occurs during the cytoplasmic steps (Figure 2).

Figure 2. Multiple pathways to produce ATP in trypanosomes.

Catabolism of the most abundant carbon sources in procyclic form grown in glucose-depleted (A) or in glucose-containing (B) conditions and in bloodstream long slender form (C). Excreted end-products from glucose and proline degradation (pyruvate, acetate, succinate and alanine) are underlined. Arrows with different thicknesses tentatively represent the metabolic flux at each enzymatic step. In (B), the direction of ADP/ATP exchange between the cytosol and the mitochondrion (step 14) is unknown and is represented by double arrows. Key enzymatic steps: 1a, glycosomal phosphoglycerate kinase; 1b, cytosolic phosphoglycerate kinase; 2, pyruvate kinase; 3, phosphoenolpyruvate carboxykinase; 4, glycosomal malate dehydrogenase; 5, cytosolic fumarase (for simplification this reaction is placed in the glycosome); 6, glycosomal NADH-dependent fumarate reductase; 7, pyruvate phosphate dikinase; 8, acetate:succinate coenzyme A-transferase, or ASCT; 9, acetyl-coenzyme A thioesterase; 10, succinyl-coenzyme A synthetase; 11, trypanosome alternative oxidase; 12, respiratory chain; 13, F₀F₁-ATP synthase; 14, mitochondrial ADP/ATP exchanger. AcCoA, acetyl-coenzyme A; DHAP, dihydroxyacetone phosphate; G3P, glyceraldehyde 3-phosphate; Gly3P, glycerol 3-phosphate; MAL, malate; PEP, phosphoenolpyruvate; PYR, pyruvate; SUC, succinate.

In contrast, ADP and ATP molecules can be exchanged between the cytosolic and mitochondrial compartments through the TbMCP5 mitochondrial ADP/ATP exchanger²⁰. This exchanger is required because oxidative phosphorylation does not occur in B-LS and thus no ATP is generated inside this organelle. To maintain the mitochondrial proton electrochemical gradient across the mitochondrial membrane, the F1F0-ATPase operates in the reverse direction, hydrolysing ATP into ADP (Figure 2). This unusual way of generating a mitochondrial potential also implies that a functional phosphate/H⁺ exchanger must be present.

Significant differential expression of mitochondrial and glycosomal proteins occurs during the life cycle²¹. Indeed, during the differentiation of B-LS into PFs, degradation of glycosomes probably via autophagy is enhanced and new glycosomes with different enzymatic contents are produced, so that parasites become rapidly metabolically adapted to the new host environment²². In fact, the differential expression of glycosomal and mitochondrial proteins is a clear indicator of the difference in metabolic life styles between the two major life-cycle stages of T. brucei.

3.2. Trypanosomes have specific pathways and unique enzymes

Kinetoplastida are of great intrinsic scientific interest as they have diverged very early compared with most studied eukaryotic models (for example, yeast, plants and animals), on which the foundations of molecular, biochemical and cellular biology have been built. Cytochrome oxidase (COX) is the terminal oxidase of the mammalian electron transport chain and is responsible for the reduction of oxygen to water. However, T. brucei possesses an additional plant-like, non-energy-conserving terminal oxidase called alternative oxidase (TAO). Indeed, B-LS forms are unique in the sense that they do not use COX but rely on TAO (step 11 in Figure 2C). TAO is 100-fold more expressed in B-LS than PFs and thus is considered a potential drug target²³.

Sphingolipids are a class of lipids important in cell recognition and signal transmission. To date, T. brucei is the only organism known to make all three types of sphingolipids (sphingomyelin, inositolphosphoceramide and ethanolamine-phosphoceramide). These lipids are synthesised via four sphingolipid synthases (SLSs) that are encoded by genes organised in a tandem array. Sphingolipid synthesis is highly controlled during development: a dedicated inositolphosphoceramide synthase (SLS1) is highly upregulated in B-SS parasites and maintained in PFs²⁴. As a consequence of more ceramide being used for inositolphosphoceramide synthesis, the synthesis of sphingomyelin is reduced, causing an alteration in the levels of phosphatidylinositol species.

4.1. Amino acids: an abundant carbon source in the midgut of the fly

The midgut of the tsetse fly has a temperature of about 28°C and a variable pH and contains hardly any glucose, but is rich in amino acids, such as proline (about 100 μM)⁹. It is well accepted that, in a glucose-depleted environment, PFs primarily use proline for their energy production^25,26 (Figure 1), but catabolism of other amino acids, such as threonine and leucine, is also used^27,28. These latter amino acids feed fatty acid biosynthesis and/or enter into the mevalonate pathway to produce the building blocks to generate essential lipids, including isoprenoids and sterols. Proline is catabolised within the mitochondrion and excreted from the cell as the end-product alanine, with the production of several reduced cofactors, which are reoxidised in the respiratory chain for ATP production by OXPHOS (Figure 2A). However, if glucose is provided, PFs adjust their metabolism and produce most of their ATP via glucose degradation (glycolysis), even in the presence of proline (Figure 2B)^25,29. These findings highlight that these parasites, like most other eukaryotes, are extremely flexible at adapting their central metabolism to their environment.

4.2. Glucose: differences in consumption rate and efficiency of ATP production

So far, the only carbon source for ATP production described for replicative bloodstream parasites is glucose, which is converted via glycolysis (Figure 1 and Figure 2C). Unlike proliferative yeast and tumour cells, B-LS does not undergo fermentation per se. In fact, instead of being metabolised and generating ethanol or lactate, most pyruvate in B-LS is immediately excreted, and only about 1% is fermented into succinate^30,31. To oxidise the NADH produced during glycolysis back to NAD⁺, B-LS consume large amounts of oxygen that act as an electron acceptor in a reaction catalysed by the unusual TAO. This type of glucose metabolism is uncommon and does not fit textbook knowledge. Interestingly, B-LS cells also tolerate anaerobic conditions where they convert glucose to equimolar amounts of glycerol and pyruvate, with a two-fold reduction of the ATP production rate.

As mentioned above, although PFs rely upon proline in vivo, they prefer glucose to produce ATP²⁵. Interestingly, the rate of glucose degradation is about 10-fold higher in B-LS than in PFs^30,32. This considerable difference is probably due to metabolic adaptations developed by B-LS in response to a much higher ATP demand compared with PFs. Firstly, B-LS replicate faster than PFs (doubling times of about 7 and about 12 hours, respectively), which means that theoretically B-LS should show a 1.5-fold higher rate of ATP production. Secondly, the estimated number of ATP molecules produced per glucose consumed is about two-fold lower in B-LS. This difference is explained by the different strategies used by B-LS and PFs to degrade glucose into excreted end-products, which are mainly pyruvate in B-LS (Figure 2C) and acetate plus succinate in PFs (Figure 2B)³³. Indeed, at the end of glycolysis, PFs convert pyruvate into acetate and ATP by the acetate:succinate CoA-transferase (ASCT)/succinyl-CoA synthetase cycle^34,35. This pathway accounting for about 70% of the glycolytic flux in PFs is reduced to 5% in B-LS³⁰. In addition, the glycosomal succinate fermentation pathway (steps 3–6 in Figure 2B), pyruvate phosphate dikinase (step 7), and cytosolic localisation of phosphoglycerate kinase (step 1b in Figure 2B) improve the rate of ATP production within the cytosol of PFs^29,36.

Thirdly and probably the most important reason for a higher rate of glucose degradation in B-LS is that some biological processes require more ATP in B-LS compared with PFs. This is the case of endocytosis, which is at least about 10-fold upregulated in B-LS compared with PFs and other trypanosomatids^37,38. The high endocytic activity observed in B-LS is required for rapid recycling of cell-surface glycosylphosphatidylinositol (GPI)-anchored variant surface glycoprotein (VSG) for internalisation and removal of bound antibodies, facilitating escape from the host immune defences, but also for nutrient scavenging from the mammalian host. Knockdown of actin resulted in a significant decrease (>70%) in endocytic activity and clearance of anti-VSG antibodies by B-LS forms, but did not significantly affect cellular ATP levels^38,39. Surprisingly, measurement of the rates of pyruvate production and oxygen consumption, under conditions identical to those employed for the ATP and transferrin uptake assays, revealed a decrease of about four-fold in both rates after a knockdown of 15 hours (D. P. Nolan, unpublished data). Although the consumption of glucose was not measured, these data suggest that membrane trafficking in the B-LS may represent a significant additional ATP demand compared with the PFs and even more surprisingly that the rate of ATP utilisation may also influence its rate of production via glycolysis. However, knockdown of actin also led to a rapid cessation in cell division and eventual cell death, so the implications of these preliminary metabolic investigations may not be so straightforward.

Interestingly, although B-SS live in a glucose-rich environment, they undergo morphological and gene expression alterations that are consistent with a preparation to survive within the tsetse midgut environment^7,24 (Figure 1). These adaptions also include increased sensitivity to specific environmental cues that signal entry to the tsetse fly vector, as well as resistance to extracellular acidic and proteolytic stress^40,41. Given that B-SS are non-proliferative and existent only in pleomorphic strains, it is more difficult to obtain large and pure quantities of this life-cycle stage in vitro. As a result, its metabolism has been less characterised. Nevertheless, we know that B-SS consume glucose and produce pyruvate and intermediate levels of acetate⁴², suggesting that metabolism is being pre-adapted to the conditions in which procyclic forms will live within the tsetse midgut. Transcriptomic studies have confirmed the downregulation of several genes that encode for components of the glycosomes and are involved in glucose uptake and breakdown²⁴. Genes upregulated in B-SS include TAO, fructose-2,6- biphosphatase, specific membrane proteins, and specific lipid biosynthesic genes, including TbSLS1 involved in inositolphosphoceramide synthesis. Further biochemical studies will be necessary to characterise and allow a better understanding of B-SS metabolism.

4.3. Lipids: responding to a great demand for myristate

In the bloodstream, trypanosomes are able to survive extracellularly in the mammalian host as they are coated by a dense homogenous layer of GPI-anchored VSGs. VSG coats are periodically exchanged by a mechanism of antigenic variation, protecting parasites against the host’s innate and adaptive immune responses⁴³. In B-LS forms, GPI anchors exclusively contain two myristate molecules (14 carbon fatty acid); however, myristate is present at very low levels within the mammalian bloodstream, which could not sustain the B-SL demand⁴⁴. Thus, it was initially thought that de novo synthesis of myristate occurred via a type II prokaryotic-like synthase^45,46, but this synthesis is not sufficient for the GPI requirement⁴⁷. Hence, it has been discovered that T. brucei uses four microsomal elongases, with different specificities, to synthesise fatty acids in a stepwise manner⁴⁷. These elongases are responsible for the majority of de novo fatty acid synthesis in T. brucei. In bloodstream forms, downregulation of the elongase 3 in the pathway (which converts C14 to C18) explains the high production of myristate required for GPI anchors.

Trindade et al. have shown that at the RNA level, parasites in adipose tissue and blood are quite different and that many genes associated with metabolism are differentially expressed⁶. Using pulse-chase experiments with stable isotope-labelled myristate, the authors showed that the first three enzymatic steps of fatty acid β-oxidation were active in parasites occupying the adipose tissue. In this pathway, a fatty acid molecule is converted into acetyl-CoA and a shorter fatty acid. The fate of each of these molecules remains unknown. The most ‘classic’ scenario would be that fatty acids undergo several cycles of β-oxidation, releasing acetyl-CoA, which could enter the TCA cycle, leading to the production of NADH and FADH₂, which could result in the production of ATP by OXPHOS. This hypothesis, however, implies that complexes III and IV of the respiratory chain are active, and this has not been observed so far in T. brucei mammalian forms. An alternative scenario is that the fatty acid chains (normally, 16 carbon palmitate) released by adipocytes are shortened through β-oxidation to enter an anabolic process in order to produce complex lipids. The resulting acetyl-CoA could be converted into acetate and lead to the production of one ATP molecule by the action of ASCT enzyme.

Trypanosoma cruzi also infects adipocytes (specialised cells of adipose tissue) and is capable of consuming stored lipids⁴⁸. It is unclear how T. brucei, an extracellular parasite, accesses lipids that are stored inside adipocytes, which constitute the largest storage of lipids in a mammalian host. Inside adipocytes, the stored triglycerides can be converted via lipolysis into fatty acids and glycerol, which are eventually secreted. During a T. brucei infection, animals typically lose weight and serum shows hyperlipidaemia^6,49. We speculate that during a T. brucei infection, lipolysis is increased, leading to the secretion of fatty acids, which could be readily taken up and used by parasites.