Cyanide-resistant respiration, a very frequent metabolic pathway in yeasts

Abstract

It has recently been shown that cyanide-resistant respiration (CRR) is very common in Crabtree-negative yeasts (incapable of aerobic fermentation) and in non-fermentative yeasts. It is conferred by a salicylhydroxamic acid-sensitive alternative oxidase that transfers electrons from ubiquinol to oxygen, bypassing the cytochrome chain. An interesting finding is that, in general, whenever CRR is present, complex I is also present. In this article we briefly review the occurrence of CRR, the biochemistry and molecular biology of the alternative oxidase, and summarise the putative functions that have been attributed to this ubiquitous metabolic pathway, whose usefulness for the yeast cells still remains obscure.

1 Introduction

In plants, in many microorganisms and in most yeasts there is an alternative pathway to cytochrome respiration that transfers electrons directly from the ubiquinone pool to oxygen, bypassing complex III and cytochrome c oxidase, two sites of energy conservation in the cell. This cyanide-resistant respiration (CRR) is conferred by a protein, the alternative oxidase (AOX), sensitive to salicylhydroxamic acid (SHAM) and insensitive to conventional inhibitors of cytochrome respiration. Fig. 1 summarises the flow of electrons and H⁺ in the mitochondrial membrane, taking into account the relative position of AOX in the respiratory chain.

CRR in yeasts has not received much attention in the literature, probably because it is absent in Saccharomyces cerevisiae and also because its role is still obscure. Here we will discuss the occurrence, the biochemistry and molecular biology of AOX and the putative role(s) of CRR in yeasts.

2 Historical background

The first known observation of a respiration on which cyanide had no effect was made in plants, by Genevois, in 1929. Some years later, in 1937, van Herck postulated the existence, in Sauromatum guttatum, of an alternative respiratory pathway involving a non-cytochromic protein [1], and in 1955, early techniques for the isolation of relatively crude plant mitochondria allowed James and Elliot [2] to state that CRR was located in these organelles. About 20 years later, Storey [3] established that the branch point of the two electron transport chains was at the ubiquinone level. In 1978, a cyanide-resistant quinol oxidase from a plant was solubilised, and CRR was ascribed to an enzyme that was named alternative oxidase [4]. The group of McIntosh partially purified AOX of S. guttatum, and raised monoclonal antibodies against this enzyme [5], which were used for its detection in a great variety of plants [6] and eukaryotic micro-organisms [7–[9]. In prokaryotes, this cyanide-resistant, but SHAM-sensitive, respiration was never detected [10].

For many years, scientists assumed that CRR was generally absent in yeasts. This was probably due to its absence in S. cerevisiae, an organism that has been used as ‘the yeast model’ for decades. A CRR in this yeast was referred to, but with a non-mitochondrial origin [11]. Other alternative pathways were also described in yeasts, which were not due to the activity of AOX. In Kluyveromyces lactis[12] and in Schwanniomyces castellii[13] a cyanide- and SHAM-insensitive but azide-sensitive respiration was brought up, and in Candida parapsilosis a second respiratory chain was found, insensitive to antimycin A, but inhibited by amital, SHAM, myxothiazol and cyanide. This chain functions in parallel with the main chain, upstream of complex III, but merges into it at the complex IV level [14].

The first reference to CRR in yeasts, as defined above, was made in 1966, in Rhodotorula glutinis[15]. In 1972, Nyns and Hamaide-Deplus [16] reported a CRR in Candida lipolytica (now Yarrowia lipolytica) in a study that continued for several years. In parallel, the group of Medentsev and Akimenko performed a great deal of research work on the alternative pathway in this same yeast. Unfortunately, most of their studies between 1976 and 1999 were published in Russian, and only recently access to their abstracts in English has become possible through survey sites on the Internet. Meanwhile, references to CRR in yeasts were sporadic, as in the case of Candida utilis[17] and Candida albicans[18]. During the 1980s, studies on AOX in yeasts became more frequent, the work performed with Pichia anomala (formerly Hansenula anomala) by the Minagawa, Sakajo and Yoshimoto group being particularly relevant. Nevertheless, the idea that CRR was not a frequent pathway in yeasts persisted.

3 Occurrence of CRR in yeasts

In 2000, Veiga et al. [19] reported that forced aeration of starved cells of Pichia membranifaciens could induce CRR. With this very simple methodology, a large number of yeast species were surveyed. Considering these results together with previous data in the literature revealed that, after all, CRR is a very common pathway among yeasts, but almost exclusively found in non-fermentative and Crabtree-negative yeasts. Interestingly, Cryptococcus albidus, the only non-fermentative yeast species reported in that work in which CRR had not been found, was further investigated. The new result was that CRR is present in four other strains recently studied in our laboratory (L. Lima, personal communication). Therefore, it was proposed that, in general, yeasts have developed two strategic catabolic pathways as an alternative to cytochrome respiration: either aerobic fermentation in the so-called Crabtree-positive yeasts, or CRR in non-fermentative and in Crabtree-negative yeasts (capable of fermentation but not under aerobic conditions) [19].

4 Alternative oxidase

AOX catalyses the reduction of molecular oxygen by ubiquinol, resulting in the production of H₂O, but not superoxide or H₂O₂. The electron flux through AOX is not associated with energy conservation, the change in redox potential being lost as heat [20].

Structural features of AOX were first revealed in plants. It is present in the mitochondrial inner membrane as a 65-kDa dimer, which can be found in two different states. In the oxidised state, the participating monomeric subunits are covalently linked by an intermolecular disulfide bond. In the reduced state, which is substantially more active than the oxidised one, the interaction between the sulfhydryl components is non-covalent (see [10,21] for reviews on the structure of AOX). Although limited in scope, some studies on the nature of AOX from the yeast Pichia stipitis[22], and from other micro-organisms [10], have also been performed. They have revealed that, unlike in plants, in micro-organisms AOX is always present as a monomer. However, it has been suggested that this difference in structure may not be catalytically relevant. There is evidence that the monomeric subunits of AOX from plants act as a catalytically functional unit [10]. Since iron is present in most, if not all, of the terminal oxygen-reducing oxidases, it is probable that iron is also present in AOX protein. Although no paramagnetic signal has ever been ascribed to this oxidase, evidence for the presence of iron in this enzyme was reported for Pichia anomala, in which the ferrous ion was essential for a considerable insensitivity to cyanide [23]. It was proposed that the active centre of AOX might have a bi-nuclear iron centre, similar to the one in mono-oxygenase [24].

As regards AOX activity, it is well established that, in vivo, it depends on the amount of protein, and on the substrate (ubiquinol) concentration [20]. However, several other factors have also been shown to affect the activity. In plants, it is stimulated by some α-ketoacids, pyruvate being the most effective [25]. In yeasts, as well as in other micro-organisms, stimulation by organic acids has not been observed. Comparing the N-terminal sequences of AOX from five fungi with the corresponding sequences in plants, Joseph-Horne et al. [26] explain this lack of activation by the absence, in the fungal AOX, of a cysteine residue involved in the pyruvate activation in plants. Nevertheless, other regulators of AOX activity in micro-organisms have been described. Cyanide-resistant respiratory activity was stimulated by purine nucleosides, like AMP, ADP, d-AMP or GMP, in the yeasts P. anomala[27], C. parapsilosis[28] and Y. lipolytica[29], in some moulds and in protozoa, but the mechanism of action of these nucleotides is yet to be determined [30]. Stimulation of the alternative pathway by the cytochrome chain inhibitors, cyanide and antimycin A, accompanied by a marked decrease in ATP and an increase in ADP and AMP, was reported in Y. lipolytica[29]. Since in yeasts the activity of AOX is increased by AMP, Medentsev et al. [30] interpreted the increase in oxygen consumption often observed in the presence of cyanide as due to activation of AOX mediated by an increase in AMP concentration. However, this activation occurs immediately upon cyanide addition. Veiga et al. [31], observing the same phenomenon in P. membranifaciens and Debaryomyces hansenii, suggested that this increase in oxygen consumption through the alternative pathway may correspond to an overall increase in metabolic flux, as if mitochondria were counterbalancing the rapid switch to this relatively low-yielding proton-motive force pathway.

Unlike stimulation, the inhibition of AOX is not dependent on the type of organism. SHAM is the most commonly used AOX inhibitor in cells and it probably acts by competing with ubiquinol [10], but several other inhibitors have also been described. Hoefnagel et al. [32] performed a screening of possible inhibitors of AOX and found powerful inhibitory effects in several compounds such as 2,5-dibromo-3-methyl-6-isopropyl-p-benzoquinone, disulfiram, 5-n-decyl-6-hydroxy-4,7-dioxobenzothiazole, valinomycin, n-propyl gallate (frequently used in isolated mitochondria) and different hydroxamic acids beside SHAM, such as m-iodo-benzhydroxamic acid, benzhydroxamic acid and chloroquine. More recently, Affourtit et al. [33] introduced maesaquinone as a new inhibitor of all respiratory enzymes that use ubiquinone as substrate, including AOX.

5 Molecular biology of AOX

The genetic basis for the alternative pathway activity was first proposed in Neurospora crassa. Preliminary studies in this mould [34], later confirmed in yeasts and plants [7,35], indicated that AOX is encoded in the nucleus. Bertrand et al. [36] proposed that its activity required a structural gene called aod-1 and an additional induction component called aod-2. The plant equivalent to aod-1 has been isolated and termed AOX1[35], but the existence of a gene analogous to aod-2 has never been demonstrated in plants.

The production of monoclonal antibodies raised against AOX from S. guttatum[5] paved the way for the isolation of clones, and subsequently genes, from several other plants, protozoa and moulds [10]. Genes from the yeasts P. anomala[7] and C. albicans[37] were also isolated. When the C. albicans AOX1 gene was expressed in S. cerevisiae, the respiration of this yeast became partially resistant to cyanide [37]. In Schizosaccharomyces pombe, a cyanide-sensitive yeast, functional expression of the plant S. guttatum AOX was also achieved [38]. Day et al. [6] compared the predicted proteins from sequences of AOX1 cDNA clones from four different plant species and P. anomala. The plant sequences display high identity (at least 69%), whereas they are only approximately 35% identical to the P. anomala sequence. In this yeast, as in moulds, AOX is encoded by a single-copy nuclear gene [39]. In C. albicans, it seems that it is encoded by a family with two members [37], as happens in higher plants, in which AOX is encoded by gene families.

Besides the regulation of AOX activity by α-ketoacids and nucleotides, AOX expression may be regulated at the transcriptional level. Diverse factors were described to affect CRR expression in vivo. In plants, it is dependent on the type of tissue, but may also be stimulated by low temperatures [40], osmotic shock, wounding, pathogen attack, elevated carbohydrate status, cell culture stage, addition of ethylene, ripening, increase of salicylic acid [41], and citrate levels [42]. In some non-fermentative yeasts, such as P. membranifaciens[43] and Y. lipolytica[30], when glucose was used as substrate, CRR was observed only in the transition between exponential- and stationary-phase cultures. In some Crabtree-negative yeasts (C. parapsilosis[28], Debaryomyces occidentalis [formerly S. castellii] [44] and C. albicans[45]) it was detected under all conditions tested and in D. hansenii it was absent only in the very early exponential phase [43]. A systematic analysis of the in vivo emergence of CRR in yeasts is yet to be performed.

Different ways of inducing CRR in yeasts, as well as in other micro-organisms and plants, have been reported. The most common is to incubate the culture in the presence of cytochrome chain inhibitors, such as antimycin A or cyanide, which have been shown to induce the expression of the AOX gene in yeasts [30,37,43,46], moulds [47], and plants. Chloramphenicol, an inhibitor of mitochondrial protein synthesis, has also been described as a CRR inducer [48]. Other means for the induction of CRR, which do not involve the addition of inhibitors, have been sporadically reported. Externally added salicylic acid induced CRR in plants and algae [49]. Thermal shock promoted the appearance of an alternative pathway in N. crassa[50], and H₂O₂ induced CRR in plants [42,51]. In P. anomala it was induced by the superoxide anion [52] or by incubation of the cultures in the presence of sulfur compounds, such as cysteine, methionine, or glutathione [53]. Several stress situations such as high temperature, decreased pH of the cultures, or the presence of reactive oxygen species (ROS), e.g. H₂O₂ or superoxide (induced by menadione), were shown to induce CRR in P. membranifaciens and D. hansenii. In the latter, it was also induced by increasing the pH of the cultures and by the presence of NaCl [43]. As it is well established that stress situations may generate ROS and as these species constitute regulating signals for the expression of a wide range of genes, it has been suggested that all CRR-inducing situations would result in an increase in ROS production that would trigger the expression of the AOX gene [49]. Nevertheless, there is not yet well-structured evidence to support this suggestion.

Induction of CRR was prevented by the addition of cycloheximide to P. membranifaciens, D. hansenii[43] and P. anomala[54]. In Aspergillus niger, besides cycloheximide, other cytosolic translation inhibitors, such as emetine and puromycin, had the same effect, which was also achieved with carbonylcyanide m-chlorophenylhydrazone (uncoupler) and actinomycin D (transcription inhibitor) [55]. These results suggest that AOX protein is synthesised de novo in the cytosol and transported into mitochondria.

6 Physiological role of CRR

Among all the functions that have been proposed for CRR in plants, only one has been confirmed: in the thermogenic floral tissue of some plants, uncoupled respiration is responsible for heat production, volatilising foul-smelling compounds that attract insect pollinators [1]. However, this role does not apply to non-thermogenic tissues or to micro-organisms. Lambers [56] suggested that the alternative pathway acts as an energy overflow that allows the tricarboxylic acid cycle to continue, independently of ATP synthesis, permitting the turnover of carbon skeletons. Following the same reasoning, Vanlerberghe and McIntosh [49] proposed that the most general function of AOX could be to balance carbon metabolism and electron transport. Rapid changes in either of these two coupled processes could be counteracted by rapid and fine metabolic control of AOX activity. Likewise, longer-term changes in carbon metabolism or electron transport resulting in imbalances could be offset by coarse control of the amount of AOX protein present.

In plants and fungi, besides the Lambers hypothesis, it has been suggested that the alternative pathway could replace, in part, the phosphorylating cytochrome pathway when it is restricted in situations of biotic or abiotic stress. Approximately 800 plant species are known to produce cyanide upon wounding or when attacked by pathogens, and it has been found that the fungus Stemphylium loti, a pathogen of a cyanogenic plant, has a CRR that is thought to play a role in the energy supply for the production of a detoxifying enzyme [49]. In P. stipitis it has been proposed that AOX would act as a redox sink, preventing xylitol production during oxygen-limited fermentation of xylose [57].

The hypothesis that the alternative pathway plays a role in the anti-oxygen defence in plants was first proposed by Purvis and Shewfelt [58], and found support in subsequent experiments [59–[61]. Linked to electron transport is the formation of ROS, in particular under stress situations. Most ROS are formed at the complex III level, and bypassing electrons directly from ubiquinone to AOX would prevent their formation. A simple mechanism of action was envisaged: since most ROS act on cytochrome c oxidase, and the decrease in activity of this enzyme should, by itself, raise the level of reduction of the quinone pool, increasing the formation of the quinone radical UQ^●−, AOX could act as a quenching mechanism of radical formation. In yeasts, however, in rather preliminary experiments, the activity of CRR did not affect the level of ROS, suggesting no links between these two phenomena [43]. A new perspective was introduced by Gomes et al. [62] who proposed that AOXs are evolutionarily related to other di-iron proteins. Because all the di-iron proteins can activate dioxygen, they suggest that this function was the ancestral reminiscent feature of this family of proteins, which has remained conserved through evolution and is still present in AOXs. In a recent paper, Moore et al. [63] discussed this hypothesis: although it is probable that one of the earliest significant functions of AOX was to scavenge dioxygen in order to reduce the generation of ROS, it is important to recognise that the later development, in evolutionary terms, of the much more efficient haem–copper proteins would have reduced the necessity of this function still found in such a ubiquitous way.

In plants, it is generally assumed that complex I of the respiratory chain is present, acting as the only one, out of the three phosphorylating sites, available for energy production derived from CRR activity. In yeasts it is not evident that energy can be obtained from the activity of CRR since in many yeasts complex I is known to be absent. Although several studies have been carried out on this subject, only recently a systematic approach of the question of energy conversion during the activity of CRR in yeasts was reported, in the framework of studies performed with P. membranifaciens and D. hansenii[31]. In both yeasts, evidence for the involvement of CRR in ATP formation, as well as in the building up of a mitochondrial membrane potential, was found, and the presence of complex I was confirmed. Sorting out in the literature the simultaneous occurrence of CRR and of complex I among yeasts, complex I was present in 13 out of 14 yeasts in which CRR was also present, whereas in six out of seven yeasts in which CRR was absent, complex I was also absent. Since three phosphorylating sites are active in the main respiratory chain of yeasts, and only one in CRR, a role in the mechanism of fine adjustment of energy available to the cell was proposed for this pathway [31]. A similar function was recently suggested by Moore et al. [63] in plants. They proposed that the plant AOX performs an active role to maintain a constant energy charge at all times aiding the stabilisation of the phosphorylation potential. Our suggestion is that Crabtree-positive yeasts use aerobic fermentation for the same purpose, while non-fermentative and Crabtree-negative yeasts use the alternative oxidative pathway for the fine adjustment of the phosphorylation potential.

7 Outlook

So far, CRR, an energy-yielding, very frequent pathway, has not been assigned a clear function in yeasts. The use of molecular tools may bring new clues to answer this open question. The physiological characterisation of mutants in the AOX gene of yeasts exhibiting a high activity of CRR, such as D. hansenii, is a promising way. The heterologous expression of AOX in a strain of S. cerevisiae appropriate for a careful kinetic and energetic characterisation is another promising approach. Results derived from the studies with yeasts may lead to a better understanding of the role of CRR in living organisms in general.

Acknowledgements

A.V. received a grant from the Fundação para a Ciência e a Tecnologia SFRH/BPD/8548/2002.

References