BV.I. BIBLIOTEEk o university free state \\",\\\3\43\0\00\0\04\2\49\6I4 il"'" Universiteit \/f'1staat RESPONSES OF YEASTS TO HYPO-OSMOTIC STRESS By Gerald Kayingo A thesis submitted in fulfilment of the requirements for the degree of PHILOSOPHIAE DOCTOR In the Faculty of Natural Science, Department of Microbiology and Biochemistry, University of the Orange Free State, Bloemfontein, South Africa. March 2001 Promoter: Prof B.A. Prior Co-promoters: Prof S.G. Kilian Prof S. Hohmann DECLARA nON I, the undersigned, hereby declare that the work contained in this dissertation is my own original work and has not been previously in its entirety or in part been submitted at any other university for a degree. ~e£t~~t, . GERALD KAYlNGO Date 11 AKCNOWLEDGEMENTS I wish to convey my gratitude to all the people who have contributed to making this task a success. I am especially indebted to my study leaders for their guidance and encouragement. Special thanks go to the National Research Foundation (NRF) without whose financial support this study would have perhaps never been possible. The assistance and moral support I have received from colleagues in Bloemfontein, Stellenbosch, Leuven and Goteborg cannot go unrecorded. My siblings in Uganda, thank you for enduring the difficult times that my absence may have caused. I have received much help both morally and professionally from many people. If! have in a way failed to express my appreciation, lowe you my apology. III PREFACE This thesis is presented as a compilation of manuscripts. Each chapter is introduced separately and written according to the style of the journal to which the manuscript was or will be submitted. Chapter 2: Microbial water and glycerol channels. Parts of this chapter have been published in Trends in Microbiology 2000,8: 33-38. Current Topics in Membranes 2001,51:335-370. Chapter 3: Growth, conservation, and release of osmolytes by yeasts during hypo-osmotic stress Manuscript in preparation. Chapter 4: Effect of ergosterol on survival and glycerol release from Saccharomyces cerevisiae cells after osmotic downs hock . Parts of this chapter have been submitted for publication. Chapter 5: An investigation of the possible existence of homologues of FPSl, a glycerol facilitator of Saccharomyces cerevisiae, in the osmotolerant yeasts Zygosaccharomyces rouxii and Pichia sorbitophila. Parts of this chapter have been published in Molecular Biology and Physiology of water and Solute Transport. Kluwer academicIPlenum Publishers. Pages 393-404. Chapter 6: Isolation and characterisation of the TIM10 homologue from the yeast Pichia sorbitophila: A putative component of the mitochondrial protein import system. Published in Yeast 2000, 16:589-596. Chapter 7: Characterisation of a putative glycerol facilitator in the fission yeast Schizosaccharomyces pombe Manuscript in preparation. IV CONTENTS 1. Introduction 1 The osmotic stress response in yeast 1 Physiological and morphological changes in response to osmotic stress 2 Synthesis, accumulation and transport of osmolytes 2 Signal transduction mechanisms controlling the osmotic stress response in yeast 4 Aims of the study 5 h~~ 7 2. Microbial water channels and glycerol facilitators 11 Abstract Il Introduction 12 Microbial aquaporins and glycerol facilitators 21 Classification of microbial:MIP channels into subfamilies 21 Distribution of:MIP channels in microorganisms 23 Origin of microbial:MIP channels 27 Transport properties and channel selectivity of microbial:MIP channels 27 From primary to quaternary structure in microbial:MIPs 29 A structure-function analysis of microbial:MIP channels 29 Physiological roles 31 Glycerol facilitators in the uptake of substrates 31 Microbial:MIP channels in osmoregulation 34 Microbial aquaporins 34 The yeast osmolyte system: Control of glycerol metabolism 37 The Fps1p solute exporter 39 Control of the function of microbial:MIP channels 43 Control of protein activity 44 Conclusions and future perspectives 45 References 47 v 3. Growth, Conservation and Release of Osmolytes by Yeasts during Hypo-osmotic stress 59 Abstract 59 . Introduction 60 Materials and Methods 62 Yeast strains and growth conditions 62 Assay of viability 62 Osmotic stress and efflux experiments 62 Extraction and measurements of osmolytes 63 Dry weight measurements 63 Data treatment and reproducibility 64 Results and Discussion 65 Growth / survival of yeast cells during hypo-osmotic stress 65 Release of osmolytes from yeast exposed to hypo-osmotic shock 66 Osmolyte release results from effects ofturgor stress but not water stressper se 66 Osmolyte release is unaffected by ganadolium and CCCP inhibitors 73 Is osmolyte export a passive process, membrane leaks, carrier or channel mediated? 73 Regulation of osmolyte export 74 References 76 4. Effect of Ergosterol on Survival and Glycerol release from Saccharomyces cerevisiae cells after Osmotic Downshock 80 Abstract 80 Introduction 81 Materials and Methods 82 Yeast strains and growth conditions 82 Osmotic downshock sensitivity experiments 82 Nystatin sensitivity experiments 83 Intracellular glycerol determination 83 Results and Discussion 84 Effect of ergosterol on survival of yeast cells after osmotic downshock 84 Effect of ergosterol on the release of glycerol after osmotic downshock 84 Effect of nystatin on the growth and release of glycerol fromfpsl iJ strain 85 References 94 VI 5. An investigation of the possible existence of homolognes of FPSl, a glycerol facilitator of Saccharomyces cerevisiae, in the osmotolerant yeasts Zygosaccharomyces rouxii and Pichia sorbitophila 98 Abstract 98 Introduction 99 Materials and Methods 100 Strains, growth conditions and transport experiments 100 Molecular and genetic techniques 101 Southern blot hybridisation 101 Polymerase chain reaction 102 Degenerate reverse transcriptase polymerase chain reaction 103 Complementation experiments 104 Heterologous expression oftheJPslLllungated channel in Z. rouxii 105 Results and Discussion 111 Analysis of FPSI homologues by PCR and DNA probes 111 Complementation analysis 116 The yeast DOM34 is required for osmotolerance 117 Conclusion 117 References 124 6. Isolation and Characterization of the l'IMIO Homologue from the Yeast Pichia sorbitophila: A putative component of the mitochondrial protein import system 129 Abstract 129 Introduction 130 Materials and Methods 131 Strains and growth conditions 131 Nucleic acid manipulation and analysis 131 Isolation of mitochondria and Western blotting 132 Functional complementation analysis 132 Results and Discussion 133 Isolation and characterization of the Pichia TIM 10gene 133 Vil Similarity to other proteins in the public databases; the Tim family of proteins in the mitochondrial intermembrane space 134 Pichia TimlOp cross-reacts with the S. cerevisiae TimlOp antiserum 135 Pichia TIMIO is a functional homologue of S. cerevisiae TIMIO 135 References 143 7. Characterisation of a Putative Glycerol Facilitator in the Fission Yeast Schizosaccharomyces pombe 146 Abstract 146 Introduction 147 Material and Methods 149 Molecular and genetic methods 149 Isolation and cloning of S. pombe mip I 150 Disruption of the S. pombe mip l 151 Glycerol transport assays and osmotic stress experiments 151 Results and Discussion 155 Glycerol transport in S. pombe 155 Isolation and sequence analysis of the S. pombe mip I 156 Heterologous expression of the S. pombe mip l in S. cerevisiae and functional analysis 156 The S. pombe mip I' mutant phenotypes 158 Expression of S. pombe mip l is slightly induced by osmotic stress (NaCl) 159 Concluding remarks 160 References 176 8. Summary 182 CHAPTER 1 Introd uction Yeasts are found in diverse habitats where conditions such as temperature and water availability vary considerably and often impose severe stress on growth. However, yeast cells are endowed with adaptive mechanisms that sense and respond to environmental changes in order to protect the cell and ensure cellular activity. Over the last two decades, there has been considerable interest in understanding the response of yeast cells to hyper-osmotic stress. This thesis examines the yeast osmotic stress response with special emphasis on hypo-osmotic stress, where less attention has been given. The osmotic stress response in yeast What is osmotic stress? Various terms are currently used to describe the amount of thermodynamically available water in the environment of an organism, namely water potential, osmotic potential, osmotic pressure, water activity (as), osmolarity and turgor pressure (Brown, 1978). Water activity, one of the most commonly used parameter, refers to the mole fraction of water in a solution whereas turgor pressure refers to the difference between internal and external osmotic pressure maintained by the cell envelope. Changes in solute concentration automatically affect the osmolarity of the medium and leads to concomitant flux of water in or out of the cell. This osmotic flow of water might cause a physiological burden and affect the normal functioning of the cell. Therefore, osmotic stress loosely refers to the adverse effect of increased or reduced F (B) '----- P. mu/fcoda G IpF (B) .----( E. co/iGlpF (B) ~ S. !l9xneri GIpF (B) Y. pestis GlpF (B) '-- __ -{_ S. ~phi GIpF (col1lig460) (B) S. ~phimuriumGlpF (B) S. enterioe serover typhimurium PduF (B) subfamily 2 S. Iyphimurium PduF (Bl S. IyphiPduF (contig411) (B) S. Iyphi PduF (oon6g443) (B) K. pneumoniaeG IpF (oontig848) (8) v. cho19rae GlpFc (B) K.pneumoniaeGIpF' (oontig757) (8) B. burgdal9ri GlpF (B) D. radiodurans GlpF (B) r---------l ---uL___;.----- T. bruce! GIpFb (P)S. c ... ""is"eYFL054 (Y) S. pombe NO (Yl uc;_ B. ctneree G I>FD.C (F)B. antlTacis GlpF (gba1391) (B)B. subtiBs GlpF (B) S. eureus GlpF (B) T.msritima GlpF (B~ C. perlringens G IpF (B) E./secaNs GlpF (ge!6204) (B) S.pneumoniseGIpF (sp7) (B) S.pyogems AGP (conl.g115) (B) M. gsDisepticum G IpF (B) subfamily 3 S. ocetaaor 9 ylAb (B) M. cea coum GlpF (B) C. diphthfTis G1.I'Fb(conlg423) (B) L. "'ctis Ydp1e (B) S.pyogemsGIpF (conlig104) (B) L S.pneumon"eAGP (sp42) (B) C. 8CebbutDcum AGP (B) L- _ H. in/kJenzae G IpF (B) E. teecats GlpF (gef6176) (B) 1-------------------------------------------------------------------------c--::-=-=--=-== M. genfta6um Glpl' (B) -----M.pneumonBeG IpF (B)S. c ... ""is"e Fps1c (Y) Figure 1. Phylogenetic tree of microbial MIP channels. Phylogenetic analysis of the MIP channels listed in Table 1. A, archeaobacterium; B, bacterium; F, fungus; P, protozoan; Y, yeast 23 Subfamily 2 comprises the glycerol facilitators of Escherichia coli (Sweet et al., 1990) and Pseudomonas aeruginosa (Schweizer et al., 1997) as well as the propanediol facilitator from Salmonella typhimurium (Walter et al., 1997). The transport specificities of the latter two proteins have not been determined experimentally, but the genes are respectively part of the well-characterized glp operon in Pseudomonas aeruginosa (Schweizer et al., 1997) which is required for glycerol catabolism, and the pdu operon in Salmonella typhimurium (Fig. 2) which is required for propanediol utilization (Walter et al., 1997). The third subfamily contains the Lactobacillus laetis glycerol facilitator, which has been shown to transport both water and glycerol (Froger et al., 2000). Whether this is a general feature of the third subfamily is unknown and hence conclusions about functionality can only be speculative. In order to encompass the possible roles of these proteins in transporting water and glycerol, the term aquaglyceroporin (AGP) has been suggested for them. Some microbial MIP channels do not appear to fall into any of the three subfamilies, such as the putative glycerol facilitators from Enterococcus faecalis, Mycobacterium genitalium, Mycobacterium pneumoniae and Fpslp from Saccharomyces cerevisiae, which has a number of unusual features (see further). This may reflect functional specialization as is apparent for Fps1p, which functions mainly as an export channel. B. Distribution of MIP channels in microorganisms Table 1 lists the known MIP channels that were found in a total of 23 complete microbial genome sequences by the end of March 2000, as well as those from ongoing sequencing projects. The data from the completed microbial genomes allows some conclusions to be drawn on the distribution of MIP channels in microorganisms. For example, there are apparently some organisms that lack MIP channels altogether, such as the Archaea Methanococcus jannaschii and Pyrococcus horikoshii and the Bacteria Aquifex aeolicus, Helicobacter pyroli, Mycobacterium tuberculosis, Treponema pallidum, Chlamydia trachomatis, Chlamydia pneumoniae, Rickettsia prowazekii and Campylobacter jejuni. The majority of these microbes are either animal pathogens or deep-sea dwellers. It is plausible that in such habitats microbes might not experience stressful osmolarity changes that would require MIP channel mediated-water/solute flux. Interestingly most microorganisms lacking a glycerol facilitator gene also do not possess a glycerol kinase gene suggesting that these organisms might not utilize glycerol as a carbon source or as a metabolic precursor. 24 Of course it is possible that these organisms have other currently undefined mechanisms for water or solute transport. (a)Pseudomonas aeruginosa glpFK operon glpD· (b) Salmonella typhimurium pdu/cob operon Pdu operon cob operon ... (c) Bacillus subtilis glpFK operon ORFI Figure. 2. Operon structure Operon organization of glycerol and propanediol facilitator genes in different bacteria. a) Operon organization of the glpFK-containing region of the Pseudomonas aeruginosa chromosome. The genes encoding the glycerol facilitator and glycerol kinase are indicated as glpF and glpK respectively; glpX and gipR encode regulatory proteins and glpD sn-glycerol- 3-phosphate dehydrogenase (Schweizer et al., 1997). b) Salmonella typhimurium pdu/cob operon containing the pduF gene. The pdu operon controls the degradation of propanediol whereas the cob operon controls the synthesis of cobalamin, which is required for propanediol catabolism. The region between the two operons encodes two proteins, the propanediol facilitator PduF and PocR, a regulatory protein, which mediates the induction of the pdu/cob operon by propanediol (Chen et aI., 1994; Chen et al., 1995). The letters A, B and C designate the first three genes in the pdu operon. The pduA gene encodes a hydrophobic protein with high similarity to the carboxysome-forming proteins of several photosynthetic bacteria whereas pduB and pduC encode proteins of unknown function. The arrows indicate the direction of gene transcription. c) The Bacillus subtilis glpPFKD region containing genes essential for growth on glycerol or glycerol 3-phosphate. The genes encoding a glycerol facilitator and a glycerol kinase are indicated as glpF and glpK respectively. The glpP gene encodes a regulatory protein whereas glpD encodes a glycerol 3-phosphate dehydrogenase. The four genes represent three separate transcription units (glpP, glpFK, glpD) and the activities of glpFK and glpD are controlled by glpP, the phosphoenolpyruvate:sugar phosphotransferase system (PTS) and glucose repression (Beijer et aI., 1993). Inverted repeats are indicated by hairpin symbols. 25 In general, it appears that most bacteria whose genomes have been fully sequenced possess a glycerol facilitator homolog that is part of the same operon as the gene for glycerol kinase: an indication of a role in glycerol metabolism, as we discuss below. In addition, several genomes contain a second MIP channel, which may be an aquaporin homolog, such as in Escherichia coli, Pseudomonas aeruginosa or Shigellaflexneri. The second MIP channel may also be (i) an additional glycerol facilitator as in Pseudomonas putida, (ii) a propanediol facilitator, as in Salmonella typhimurium or (iii) a glycerol facilitator from a different subfamily, as in Bacillus anthracis, Streptococcus pneumoniae and Haemophilus influenzae. Some bacteria appear to have more then two MIP channels, often one from each subfamily as found in Enterococcus faecalis and Streptococus pneumoniae or from at least two different subfamilies as in Klebsiella pneumoniae and Salmonella typhi. The presence of more than one MIP channel from the same subfamily is found only in very few cases, such as in Klebsiella pneumoniae (two putative, quite distinct water channels and two closely related subfamily 2 members), Pseudomonas putida (two closely related subfamily 2 glycerol facilitators) and Salmonella typhi (two highly similar, putative propanediol facilitators). With the possible exception of the two Klebsiella pneumoniae water channels, two proteins from the same subfamily are likely to be the result of a recent gene duplication event and may fulfill the same physiological role in that microorganism. In general, however there appears to be a tendency to maintain only one (if any) MIP channel per subfamily in a given organism. This distribution pattern supports the idea that the members of the different subfamilies may indeed exhibit different functions. This is further corroborated by the finding that in bacteria which only have a single MJ:P channel, the protein is in most cases found in subfamily 3, whose members may transport both water and glycerol (Froger et al., 2000). Hence, these proteins may fulfill functions as water channels and glycerol facilitators. Whether this is of any physiological relevance has yet to be addressed. From the four sequenced Archeal genomes, only two encode a MIP channel and in both instances it is a putative water channel. Hence, from this limited information it appears that glycerol is either not utilized by these organisms or that alternative uptake systems are required. The only eukaryotic microorganism for which a complete genome sequence is available is that of Saccharomyces cerevisiae. This genome encodes four MIP channels (André, 1995): two aquaporin homologs, 86% identical to each other, and two related glycerol facilitator homologs. 26 The functions of only one of the aquaporins and one of the glycerol facilitators have been confirmed (Luyten et al., 1995; Bonhivers et al., 1998). Strikingly, most laboratory yeast strains seem to have mutations that inactivate both aquaporin genes and even industrial strains and yeasts isolated from Nature appear to have mutated versions of the AQY2 gene (Laizé et aI., 2000). So far, only one laboratory yeast strain, l:1278, a derivative from an industrial isolate, has been found to have two complete aquaporin genes. AQYl from l:1278 differs from that of other laboratory strains by three amino acids and, due to a frame-shift mutation, the entire carboxy-terminus is different. Two of the amino acid substitutions seem to be responsible for functional alteration in most laboratory strains (Bonhivers et al., 1998). AQY2 in all laboratory strains investigated so far contains an Il bp deletion in the center of the gene leading to a premature translational stop. Various different alleles for AQY2 have also been found in laboratory and industrial strains as well as in natural isolates (Laizé et al., 2000). Although strain Ll278 has a complete open reading frame for AQY2, it has not been possible to confirm in the Xenopus oocyte system whether AQY2 actually encodes a functional aquaporin (Laizé et al., 2000). With regard to other eukaryotic microorganisms, two aquaporins have been found in the slime mold, Dictyostelium discoideum; the function of one of these, WacA, has been confirmed experimentally (Flick et aI., 1997). Similarity searches reveal putative aquaporins in the pathogenic yeast Candida albicans and the filamentous ascomycetes, Aspergillus nidulans and Neurospora crassa. Glycerol facilitators such as that found in the parasitic protozoan Trypanasoma brucei, are also well represented in eukaryotic microorganisms. In particular, Fpslp from Saccharomyces cerevisiae has been well characterized as a glycerol and polyol facilitator and will be discussed later (Luyten et al., 1995; Sutherland et al., 1997; Tamás et al., 1999). This organism has a second open reading frame encoding a putative glycerol facilitator, YFL054c. Fpslp and Yfl054p are 32% identical in their six transmembrane domain cores. Like Fps1p, the protein encoded by YFL054c has an approximately 300 residue amino-terminal extension. Unfortunately, analysis of strains deleted for YFL054 have not yet lead to the elucidation of the protein's function (Tamás, 1999). Recently, glycerol facilitators have been recognized in the fission yeast Schizosaccharomyces pombe and in the plant pathogenic ascomycete Botrytis cinerea. Strikingly, sequence comparison suggests that these are homo logs of Saccharomyces cerevisiae Yfl054p. Yfl054p and the homolog from Schizosaccharomyces pombe are 76% identical within the transmembrane core and share 30% identity even within their extensions (Tamás and Hohmann, unpublished data). 27 However, the extensions of these two proteins and that of Fpslp appear totally unrelated. Hence Yfl054p may be the founding protein of a glycerol facilitator subfamily in fungi. C. Origin of microbial MIl? channels It is generally believed that the MIP family emerged as a result of an intragenic duplication event, which probably took place 2.5 to 3 billion years ago (Wistow et al., 1991; Pao et al., 1998; Heymann and Engel, 2000). It has been postulated that a single gene arose in prokaryotes shortly before the emergence of eukaryotes and that subsequent gene duplication and divergence resulted in the various MIP family genes. However, the occurrence of both highly similar and dissimilar MIP channels in a single organism suggests that some MIP proteins did not arise via gene duplication, but rather have been acquired horizontally from other organisms (park and Saier, 1996). For instance, the G+C content of Saccharomyces cerevisiae AQY1 and FPSI is 50% and 43% respectively, whereas the overall G+C content of Saccharomyces cerevisiae is 40%. This deviation lends support to the notion that AQYl could have been acquired horizontally and that the second highly similar yeast aquaporin gene (86% identity at protein level) was the result of a subsequent duplication event. HI. Transport properties and channel selectivity of microbial MlD> channels The transport properties of MIP channels can be studied in a number of ways including the use of the heterologous Xenopus laevis oocyte system or the osmotic swelling of whole cells, spheroplasts or membrane vesicles to determine water or solute transport (Hohmann et al., 2000). Most available data from these types of measurements indicate that transport is very rapid, has low activation energy and can be sensitive to mercury compounds, such as HgCh, if a cysteine residue lines the pore. The transport properties of microbial MIP channels have been well characterized in just a few cases and in most instances functional studies have been performed in Xenopus laevis oocytes. It is reasonable to assume that in such a heterologous system, a different lipid environment will affect transport function or specificity (Truniger and Boos, 1993) and this could explain conflicting data obtained in some cases. The most informative transport coefficients such as osmotic water permeability (hydraulic conductivity), solute permeability and reflection coefficient have been determined for only a limited number of channels. 28 For example, the Escherichia coli glycerol channel, GlpF, transports polyols, glyceraldehyde, glycine and urea (Heller et al., 1980) but little or no water (Maurel et al., 1994; Calamita et al., 1995). Consistent with a pore-type mechanism, glycerol transport via GlpF has a low activation energy (Ea = 4.5 kcaVmol) and is non-saturable (Maurel et al., 1994). GlpF- mediated glycerol uptake is also sensitive to the membrane lipid composition (Truniger and Boos, 1993). Similar transport properties are expected for other microbial glycerol channels since the molecular architecture of bacterial glycerol uptake systems is apparently highly conserved. With regard to microbial aquaporins, expression of the prokaryotic aquaporin gene, Escherichia coli aqpZ, in Xenopus laevis oocytes results in a IS-fold increase in osmotic water permeability, but negligible solute transport. The observed water transport has a low activation energy (Ea = 3.8 kcaVmol) and is insensitive to HgC~ (Calamita et al., 1995). The water permeability of the putative Saccharomyces cerevisiae aquaporin encoded by AQYl, from both wild type and laboratory strains, has been evaluated in Xenopus laevis oocytes. Only oocytes expressing AQYl from the strain L1278, which is closely related to industrial isolates, exhibit an increase in water permeability (Bonhivers et al., 1998). Transport assays using yeast membranes and yeast vesicles also lead to similar observations (Coury et al., 1999). In contrast, data is not yet available for AQY2, since it could not be functionally expressed in oocytes (Laizé et al., 1999; Laizé et al., 2000). The transport' characteristics of the glycerol exporter, Fps 1p, are similar to those of Escherichia coli GlpF although Fpslp is known to be a regulated channel (Tamás et al., 1999), which we discuss later. In contrast to most other Jv1IP channels, the transport properties of Fpslp can be studied homologously in Saccharomyces cerevisiae. At least three groups have also tried to functionally express Fpsl p in oocytes, but this has been unsuccessful to date. Fps 1p transports glycerol, erythritol and xylitol and probably other polyols (Luyten et al., 1995; Sutherland et al., 1997 and Karigren et al., 2000). Sorbitol and mannitol seem to be transported only with very low efficiency if at all (Karigren and Hohmann, unpublished data) and water does not seem to be transported (Coury et al., 1999). Like GlpF (Sanders et aI., 1997), Fpslp also seems to transport antimonite (Wysocki, unpublished data), probably because the hydrated form of this ion resembles a polyol. These observations are consistent with poor substrate specificity, which is probably determined by pore size and interactions between the polyol and residues lining the channel. 29 Although the open reading frame YFL054 is predicted to encode a glycerol facilitator, it has not yet been possible to determine its substrate specificity. The mechanisms dictating the commonly observed water/solute channel selectivity described above remain poorly understood The amino acid content and length of the predicted loop region of MIP channels may playa role in determining specificity. Froger and colleagues have proposed that the molecular basis of substrate selectivity is determined by key amino acids (Froger et al., 1998) as the substitution of two amino acids can switch the selectivity of an insect aquaporin to a glycerol channel in the Xenopus oocyte system (Lagree et al., 1999). It is possible that these residues influence channel pore size, since this factor is also likely to be key to a complete understanding of selectivity. For example, it appears from structural data at 4.SÁ that the pore of human AQP1 is large enough to allow the passage of water, but too small for solutes such as glycerol (Mitsuoka et al., 1999). However, size alone cannot explain the fact that microbial glycerol facilitators such Saccharomyces cerevisiae Fps 1pand Escherichia coli GlpF transport glycerol and not water. More studies on microorganisms are thus needed to clearly elucidate the factors governing channel selectivity and determine its significance in vivo. IV. From primary to quaternary structure in microbial MIPs A. Structure- function analysis of microbial MIP channels From an analysis of their amino acid sequences, all MIP channels are predicted to have six transmembrane domains, and to share highly conserved residues. This is no less true of the microbial branch of the family. As for all MIPs, the most notable of the conserved residues are present in the presumed channel-forming loops (Heymann et al., 1998), B and E, and comprise the family's signature sequences, Ser-Gly-X-His-X-Asn-Pro-Ala-Val-Thr and Asn-Pro-Ala-Arg, respectively, the so-called 'NPA boxes' being underlined. However, striking differences can be observed between the sequences of microbial MIPs both within these signature motifs and at the termini. This is well illustrated in the case of the glycerol facilitator, Fps1p, from Saccharomyces cerevisiae. Although Fps1p is clearly related to bacterial glycerol facilitators such as GlpF from Escherichia coli (31% identity within the core of six transmembrane domains), it is - so far - unique in the MIP family for a number of reasons (Hohmann et al., 2000). 30 For example, neither of the family's signature NPA boxes is fully preserved, being Asn-Pro- Ser (NPS) and Asn-Leu-Ala (NLA), respectively. In fact, only four additional microbial MIP sequences contain motifs other than NP A, and it is apparent that NP A is preserved in loop B but not in loop E in most of these cases. In Enterococcus faecalis, the presumed glycerol facilitator (gef 6176) contains an NQA motif in loop E, in Chlorobium tepidum, the putative aquaporin (get 5) has an NPV motif (Hohmann et al., 2000) and the Botrytis cinerea putative glycerol facilitator contains an NPS. The only deviation from NPA in loop B of a microbial MIP occurs in the Salmonella typhimurium (contig 1308) which has an NLA motif. Unfortunately, these MIPs are incompletely characterized and thus their transport characteristics cannot be used to aid our understanding of the role of these atypical features. This is of particular relevance to a general understanding of MIP channels since their generic NPA motifs are believed to be integral to the formation of a continuous solute channel, the so- called 'hourglass' (Jung et al., 1994). One possible functional consequence of these atypical motifs is that they influence MIP channel transport properties, resulting in transport specialization. Recently, it has been suggested that loop B in particular may be involved in the determination of transport direction following a comparison of the glycerol transport properties ofFpslp and Escherichia coli GlpF. Physiologically, these proteins are a glycerol exporter and an uptake facilitator, respectively. Comparison of mutants where the NPA motifs were 'restored' in Fpslp with those where GlpF was made more Fpslp-like by mutating NPA to NPS and/or NLA, indicated that the NPS of loop B may be important in influencing Fps1p's export characteristics (Bill et al., 2000). In addition to a deviation from the family's signature motifs in the channel-forming loops, Fps 1p further distinguishes itself from most other microbial MIPs by having long amino- and carboxy-terminal hydrophilic extensions. This results in a protein of 669 amino acids, compared with 281 amino acids for GlpF and other typical family members. As mentioned above, the putative second Saccharomyces cerevisiae glycerol facilitator, Yfl054p, as well as the similar Schizosaccharomyces pombe protein, also have long amino-terminal extensions unrelated to that ofFps1p. Standard secondary structure predictions suggest that MIPs are rich in a-helical segments. This is yet to be confirmed experimentally since to date, low-resolution structural data (at 4.5Á) are only available for human AQP1 (Mitsuoka et al., 1999). Even though the most divergent members of the MIP family are less than 20% identical (Park and Saier, 1996), it is expected that the gross structural features apparent in AQP1 will also be present in other 31 water and glycerol channels. For example, AQP1 is functionally homotetrameric; this quaternary structure is thus anticipated for all other .J\.1IPs.A study of the crystal organization of.J\.1IPand Escherichia coli AqpZ and GlpF confirms the close overall structural relationship between .J\.1IPchann~ls (Hasler et aI., 1998; Ringler et al., 1999) although it has been proposed recently that glycerol channels could be functionally monomeric (Lagree et al., 1999). v. Physiological roles As we have already mentioned, the physiological role of microbial .MIP channels has been mostly studied in the bacteria E. coli, P. aeruginosa, T. flavus, and S. typhimurium, in the yeast S. cerevisiae, and in the protozoan D. discoideum. In E. coli, S. cerevisiae and D. discoideum, the genes for microbial .J\.1IPchannels have been deleted and the phenotype of the mutant strains compared to wild type strains. Studies on the expression of genes encoding MIP channels as well as on the location of bacterial genes in operons have provided additional information on their physiological roles. In general, the function of microbial .J\.1IP channels is in osmoregulation, metabolism - via the uptake of glycerol or related compounds as sources of carbon and energy- or disposal of metabolic end products. A. Glycerol facilitators in the uptake of substrates Although glycerol and other uncharged small molecules are able to move across microbial membranes by simple diffusion, there is currently sufficient evidence to suggest that .J\.1IP channel proteins facilitate the uptake of these solutes. In bacteria, the organization of genes in operons, which are co-expressed and hence co-regulated, usually points to their function in a common pathway. Hence, knowledge of the function and regulation of a single gene in a bacterial operon allows the function of other genes in the same operon to be predicted. Such a relationship is not observed in eukaryotic microorganisms. Genes encoding glycerol facilitators are commonly part of the glp operon in both Gram- positive and Gram-negative bacteria (Table 1, Fig. 2) The glp operon comprises glpF, encoding the facilitator, glpK encoding a glycerol kinase, glpD encoding a glycerol-3- phosphate dehydrogenase and two genes, glpX and glpR, which presumably encode regulators of the operon (Schweizer et al., 1997). 32 The glycerol kinase appears to be closely associated with the facilitator resulting in glycerol phosphorylation during uptake which prevents re-export of glycerol (Voegele et al., 1993). Exceptions to this operon organization have been found in a number of bacteria such as Lactococcus laetis (P22094), Streptococcus pneumoniae (SP42), Corynebacterium acetobutylicum (AE001437) and Haemophilus influenzae (U32782) where putative glycerol facilitator genes do not form part of the glycerol operon, suggesting that these MIP family proteins might also have functions other than glycerol uptake (park and Saier, 1996). Escherichia coli mutants lacking glpF grow poorly on low glycerol concentrations presumably due to insufficient glycerol permeating the cell (Voegele et al., 1993). Hence it has been proposed that the facilitator is required for efficient uptake of glycerol, especially at low concentrations. The glpF mutant also shows altered kinetics for glycerol phosphorylation and the fact that free glycerol is undetectable in wild type cells utilizing glycerol supports the conclusion that transport and phosphorylation of glycerol are closely coupled (Voegele et al., 1993). Substantial glycerol transport by passive diffusion through the lipid bilayer is also apparent since a glpF mutant does not show a glycerol-negative phenotype at high glycerol concentrations (Voegele et al., 1993). Recently, a gene encoding a glycerol facilitator in Pseudomonas aeruginose has been cloned and a chromosomal !::.glpFK mutant isolated (Schweizer et al., 1997). This mutant, which lacks both the facilitator and the glycerol kinase, does not grow on medium containing glycerol as the sole carbon source and does not transport glycerol. The Salmonella typhimurium pdu operon is required for the catabolism of 1,2-propanediol. The pdu operon (Fig. 2) is closely linked to the cob operon, which controls the synthesis of adenosyl-cobalamin (vitamin B12), a cofactor required for the catabolism ofpropanediol. The region between the pdu and cob operons encodes two proteins, PduF, the putative propanediol transporter, and PocR, a regulatory protein that mediates the induction of the pdu/cob operon by propanediol (Chen et al., 1994; Chen et al., 1995). The transport characteristics of PduF have not been determined experimentally but it is reasonable to assume that this protein is involved in the uptake of 1,2-propanediol, a compound closely related to glycerol. It is not yet known whether PduF can transport glycerol in addition to 1,2-propanediol or, indeed, whether other GlpFs can transport 1,2-propanediol. However, in addition to PduF, Salmonella typhimurium has a gene encoding a glycerol facilitator that forms part of a glpFK operon. This suggests, in fact, that the organism possesses two facilitators, one for catabolism of propanediol and another for glycerol uptake. 33 Whether members of the M1P channel protein family from eukaryotic microorganisms playa role in the uptake of solutes such as glycerol is less clear. Extensive analysis of the glycerol transport characteristics of yeast wild type and fpsl Li mutants has demonstrated that this protein can transport glycerol in both directions (Luyten et al., 1995; Lages and Lucas, 1997; Sutherland et al., 1997; Tamás et al., 1999). However, mutants lacking Fpslp, Yfl054p or the double mutant lacking both putative yeast glycerol facilitators, do not show a defect in glycerol utilization (Tamás et al., 1999; Tamás and Hohmann, unpublished data). Since the yeast plasma membrane seems to be relatively impermeable to glycerol (Luyten et al., 1995; Tamás et al., 1999), the existence of uptake proteins involved in glycerol catabolism has been suggested. In addition, it has been demonstrated that yeast cells have at least one system for the active uptake of glycerol (Van Zyl et al., 1990; Lages and Lucas, 1997; Lages et al., 1999). It has also been shown that yeast mutants unable to produce any glycerol themselves, and which therefore do not grow on medium containing Nael (see below), can be rescued by as little as 5rnM glycerol in the growth medium, indicative of an active uptake system mediating accumulation of glycerol against a concentration gradient (Holst et al.. 2000). This effect has been used to identify a yeast gene, GUP I, which is required for rescue of the glycerol-negative mutant by low concentrations of glycerol. Gup 1p is a membrane protein that either takes up glycerol or at least controls glycerol uptake, a process that appears to be closely coupled to glycerol phosphorylation. This phosphorylation is catalyzed by the glycerol kinase Gut 1p. Gup 1P is not a M1P channel (Holst et al, 2000). Surprisingly, deletion of the genes encoding the glycerol facilitators in Escherichia coli and Saccharomyces cerevisiae results in diminished passive diffusion of glycerol and altered cellular lipid composition (Truniger and Boos, 1993; Sutherland et al., 1997). The reason for this observation is not clear, but it is possible that glycerol uptake and phosphorylation via the facilitator/kinase system provides glycerol-3-phosphate for phospholipid metabolism. There are several pathways that lead to the synthesis of glycerophospholipids, and the balance between different precursors seems to be critical for phospholipid biosynthesis (Daum et al., 1998). In this context it is of interest that the regulation of expression of the yeast glycerol kinase gene, GUTJ, was recently reported to be not only controlled by glucose repression and glycerol induction but also by the same regulators that control genes encoding enzymes involved in phospholipid metabolism (Grauslund et al., 1999). In conclusion it appears as there may be a connection between glycerol transport and phosphorylation on the one hand and cellular phospholipid metabolism on the other. 34 VI. Microbial MIP channels in osmoregulation A. Microbial aquaporins Most available information on the physiological roles of microbial aquaporins is derived from studies on E. coli AqpZ, S. cerevisiae Aqylp, and D. discoideum WacA. AqpZ has been shown to have a direct role in the way that Escherichia coli adjusts cell turgor within the range needed for growth and survival (Calamita et al., 1998). This function has been demonstrated by comparing Escherichia coli cells carrying a null mutation in the aqpZ gene with their parental wild-type strain. Disruption of aqpZ is not lethal, but the viability of cells in which aqpZ has been knocked out is strikingly reduced when they are grown at low osmolarity. On the contrary, no significant changes in cell viability are observed when these cells are grown in high osmolarity medium. The reduced growth observed in hypo-osmotic conditions can be rescued by transforrning the knockout strain with a plasmid bearing a functional aqpZ gene. Overall, these data are consistent with the results of regulatory studies which show a marked increase of the aqpZ transcription rate when Escherichia coli is grown in low osmolarity medium and a reduced expression in high osmolarity medium (Calamita et al., 1998). Although questions on the physiological necessity of a water channel during prolonged hypo-osmotic stress remain to be answered, this finding clearly indicates involvement of AqpZ in prokaryotic osmoadaptation. In fact, AqpZ seems to be required both during the long-term osmoregulatory response triggered by hypo-osmotic stress and the short- term responses that occur suddenly after changes in the extra-cellular osmolarity. Although further investigation is required to elucidate the precise role of AqpZ, especially during the osmotic response to hypo-osmotic stress, it is likely that involvement in osmoregulation is a general feature ofrnicrobial aquaporins (Booth and Louis, 1999). Escherichia coli wild type and aqpZ mutant strains have been used in cryoelectron microscopy studies to demonstrate, in vivo, the ability of AqpZ to mediate rapid outward- and inward-directed water fluxes triggered by sudden up- and down-shifts of the extracellular osmolarity, respectively (Delamarche et al., 1999). Besides demonstrating the functional expression of AqpZ in Escherichia coli, these studies indicate that AqpZ transports water in both directions, a property that has been also reported for many mammalian aquaporins (Meinild et aI., 1998). A role for AqpZ in mediating the bulk water uptake needed for cell expansion during rapid growth is suggested both by its maximal expression at the rnid- logarithmic phase of growth and the reduced viability characterizing the Escherichia coli aqpZ mutant grown at 39°C, a temperature where the growth rate is highest. 35 However, this function apparently contrasts with the assumption that sufficient water for cell division may be absorbed by simple diffusion across the cytoplasmic membrane during its 20-30 minute generation time (Haines, 1994). Additional studies are therefore required to better elucidate the physiological relevance of AqpZ expression during the exponential growth phase of Escherichia coli. Interestingly, an AqpZ-like protein seems to be necessary for the expression of certain surface antigens (Kopecko et al., 1980), which in tum appear to be one of the requirements for pathogenic bacteria to invade epithelial cells. In fact, it has been observed that the Shigella sonnei ORF10, an open reading frame with striking sequence similarity to the aqpZ coding region, is part of a gene cluster composed of ten contiguous ORFs located in a plasmid (pHH201) encoding the form I antigen. It has been found that deletions ofORF10 and/or any of the other nine cluster ORFs eliminate form I antigen expression of Shigella sonnei (Houng and Venkatesan, 1998). An identical gene cluster including an aqpZ-like coding region (ORF10p) is also found in the pathogenic species Plesiomonas shigelloides (serotype 017) where in association with other genes it leads to the expression of a cell surface O-antigen (Chida et aI., 2000). A possible role for AqpZ in the virulence mechanisms of pathogenic bacteria is an exceedingly appealing hypothesis, which deserves investigation. We note, however, that several pathogenic organisms lack MIP channels alltogether and hence the importance of aquaporins in virulence, if any, must be restricted. Although a role in osmoregulation seems likely for the Saccharomyces cerevisiae aquaporin Aqy1p, surprisingly, the related null mutant yeast strain tolerates osmotic changes better than the wild-type strain under laboratory conditions (Bonhivers et al., 1998). In these experiments wild type and mutant cells were co-cultivated and repeatedly osmotically shocked. The wild type cells did not survive this treatment and were consequently depleted in the culture, whereas the aqyl mutant cells survived. This evidence together with the observation that the AQYI gene product appears to be non-functional in many laboratory strains (Bonhivers et al., 1998; Laizé et al., 2000) led to the suggestion that functional aquaporins might have been lost in strains maintained under laboratory growth conditions (Bonhivers et al., 1998). However, the interpretation of these observations is complicated by the fact that the AQYI gene is very poorly expressed during vegetative growth, conditions under which the above-mentioned phenotype was determined. 36 Expression of AQYl is, however, strongly stimulated when diploid yeast cells enter sporulation and hence AQYl is clearly a developmentally-regulated yeast gene, similar to Dictyostelium discoideum wacA (see below; Chu et al., 1998; Laizé and Hohmann, unpublished data). Key questions currently under study are the precise localization of Aqy 1P and whether the protein has any role during sporulation, spore maturation or spore germination. A potential role in mediating the extrusion of water during prespore cell encapsulation has been suggested for the Dictyostelium discoideum WacA aquaporin (Flick et al., 1997). Although the wacA gene is expressed only in prespore cells, it has been observed that disruption of the gene does not lead to any apparent alterations in prespore cells or their ability to germinate or respond to osmotic stresses. However, the lack of an apparent phenotype could possibly be explained by the presence of unknown alternative aquaporins or the fact that the Dictyostelium wacA mutant was not exposed to the selective challenges of its natural habitat, the soil. According to Cotter (Cotter and Raper, 1968), germinating spores of Dictyostelium take up water very rapidly unlike prespore cells, which extrude water during encapsulation. Such rapid water fluxes imply the involvement of aquaporins. A similar argument may be used to support the involvement of Aqy1p in yeast sporulation/spore germination. Further research is required to address this intriguing question. As outlined above, yeast AQY2 seems to be mutated in most yeast strains (Laizé et al., 2000). This is a rather unusual scenario and consequently one might speculate that there is some selective pressure against maintaining this gene. The AQY2 ORF is complete in strain L1278, which consequently can express both AQYl and AQY2 (Laizé et al., 2000). In this strain AQY2 is expressed during vegetative growth and expression is stimulated after a hyperosmotie shock. This observation suggests that Aqy2p might be involved in the uptake of water during the recovery of cells from osmotic upshock. However, at present the subcellular localization of Aqy2p has not been determined and hence the protein could perhaps be located in an intracellular compartment controlling water fluxes within the cell. In fact, the two yeast aquaporins are most similar to the tonoplast aquaporins from plants and hence Aqy2p could, in fact, be a vacuolar protein. So far, attempts to assign a phenotype to cells deleted for the AQY2 gene have been unsuccessful (Laizé and Hohmann, unpublished data). 37 The unresolved issues discussed in the preceding paragraphs clearly illustrate that no well- defined physiological role can yet be assigned to any microbial aquaporin. The identification and characterization of additional aquaporins in genetically tractable systems will provide further insight into the role of water transport in microbial water relations and in osmoregulation in general. Indeed, such studies certainly deserve a more widespread interest from microbiologists. B. The yeast osmolyte system: control of glycerol metabolism A common strategy in osmoadaptation is the accumulation of compatible solutes (Yancey et al., 1982). A range of quite different compounds are employed as compatible solutes by microorganisms, such as polyols (glycerol, D-arabitol, D-mannitol and meso-erythritol) in fungi (Spencer and Spencer, 1978; Yancey et aI., 1982) and potassium ions, trehalose and amino acids or their derivatives in Bacteria and Archea (Measures, 1975; da Costa et al., 1998). The yeast Saccharomyces cerevisiae employs glycerol for this purpose (Brown, 1978; Brown and Edgley, 1980; Blomberg and Adler, 1992). This is surprising, because glycerol is known to diffuse through lipid bilayers and hence one might expect that yeast cells could lose the glycerol they produce under hyperosmotie stress. However, this does not appear to be the case to any significant extent since the yeast plasma membrane is relatively impermeable to glycerol (Luyten et al., 1995; Sutherland et al., 1997; Tamás et al., 1999). In fact, there is evidence that yeast cells can actively control the permeability of their plasma membrane to glycerol under osmotic stress, perhaps by altering the lipid composition (Sutherland et al., 1997). Glycerol is produced in two steps from the glycolytic intermediate dihydroxy acetone phosphate (Fig. 3; Blomberg and Adler, 1992) catalyzed by the enzymes glycerol-3- phosphate dehydrogenase (Gpd) and glycerol-3-phosphatase (Gpp), respectively. Both enzymes have two isoforms whose expression is differentially regulated (Norbeck et al., 1996; Ansell et al., 1997). The expression of GPDl and GPP2 is strongly induced by hyperosmotie stress (Albertyn et aI., 1994; Norbeck et aI., 1996) and hence these two proteins appear to account for most of glycerol production capacity under osmotic stress. The expression of GPPI, which is more strongly expressed under normal growth conditions than GPP2, is also somewhat induced under osmotic stress (Rep et al., 2000). 38 Glycerol Glycerol I I Glc-6-PPhospholipids '- Gpp 1,2 . Glycerol . NAD t ~ ~ PI yprl~,l .. NADH ~I I GC~l"": iIGlycerol1 GIy-3-P AT? DHA •Dakl:2'p Fru-l,6-bPAlP Gutl NAD Gpdl,2 ~ PI~ NADH GIy-3-P I ~ ._ A1dl !DHAP I Tpll . 41 ~ GA-3-P ! il Pyruvate mitochondria (GlycolysiS) Figure. 3. Yeast glycerol metabolism Schematic overview of yeast glycerol metabolism. Glycerol catabolism starts with uptake, presumably through Guplp. Glycerol is phosphorylated to glycerol-3-phosphate, Gly-3-P, by a glycerol kinase, Gutlp, and oxidized by an FAD-dependent, mitochondrial glycerol-3-phosphate dehydrogenase to dihydroxyacetonephosphate, DHAP, a glycolytic intermediate. For glycerol production DHAP is converted to Gly -3-P by an NADR-dependent, cytosolic glycerol-3- phosphate dehydrogenase, Gpd1p or Gpd2p, and subsequently dephosphorylated by glycerol-3- phosphatase, Gpp1P or Gpp2p, to glycerol. Glycerol is either accumulated within the cell or exported through the osmoregulated glycerol facilitator Fps1p, a MIP channel. DRAP and glycerol can also be interconverted via dihydroxyacetone, DRA, but the relevance of this pathway in Saccharomyces cerevisiae is unclear. The actions of the FAD-dependent Gutl pand the NADR-dependent Gpd1p/Gpd2p provide a shuttle for electrons into the mitochondrial electron transport chain (adapted fromHohmann, 1997). 39 GPD2 and GPPI expression are stimulated under anaerobic conditions (Ansell et al., 1997) since in the absence of oxygen, glycerol production is essential for the reoxidation ofNADH to NAD, which is normally performed by the respiratory chain (Van Dijken and Scheffers, 1986; Hohmann, 1997). The mechanisms that control the induction of GPDl and GPP2 under osmotic stress are being studied extensively. The High Osmolarity Glycerol (HOG) response pathway plays a central, though not exclusive, role in the induction of GPDl and GPP2 (Rep et al., 1999a; Rep et al., 1999b; Rep et al., 2000). This pathway is a prototypical MAP (Mitogen Activated Protein) kinase cascade (Fig. 4), as found in all eukaryotes. Recent transcriptome analysis indicates that this pathway controls the expression of more than 100 yeast genes upon osmotic shock (Rep et al., 2000) and that it mediates its effects via different transcription factors such as Hotlp, Msn1p, Msn2p, Msn4p and Sko1p. Of these, Hotlp and Msn1p are involved in controlling the glycerol biosynthesis genes (Rep et al., 1999b). Upon osmotic shock yeast cells rapidly stimulate the production of glycerol and substantial levels of glycerol are built up in the cell within a few hours. Concentrations of up to 1M of glycerol have been reported (Blomberg and Adler, 1992). In their natural environment, yeast cells are frequently exposed to high osmolarity, especially to high sugar concentrations. Equally common is exposure to hypo-osmotic shock, for instance during rainfall. Under these conditions the cell has to rapidly dispose of accumulated glycerol in order to diminish turgor pressure. Hence, yeast has developed an efficient system to export the majority of its accumulated glycerol within a few minutes through the MIP channel, Fps1p. C. The Fpslp solute exporter The FPSI gene was originally isolated as a multicopy suppressor of a growth defect on fermentable sugars, such as glucose, of a mutant with defective feedback control of glycolysis (Van Aelst et al., 1991). Subsequently it wasshown that this growth defect could be partially corrected by overproduction of glycerol (Luyten et al., 1995). As outlined above, Fps1p is an unusual MIP channel (Fig. 5). lts 'NP A' motifs are not fully conserved, being NPS and NLA in loops Band E, respectively and its A loop being unusually long. In addition, Fps1p is 669 amino acids long due to amino- and carboxy-terminal cytosolic extensions. Apart from two other fungal proteins, only the Drosophila BIB (Big Brain) protein has such long extensions. However, Fps lp's extensions do not show any sequence similarity to other proteins. 40 Glycerol GIUëëSë GIc-6-P 1.....T.re.h.a.lo.S.e.Glycerol Glycogen®t ? protein kinaseA GIy-3-P @ Fru-l.6-bP \ GE) ?_et uncharacterisedEthanol pathway ICen surfoce ossembIV? I 1 ~ Figure 4. Signaling upon osmotic shock in yeast Schematic overview of signaling upon an osmotic shock in Saccharomyces cerevisiae. Central to the response is the High Osmolarity Glycerol (HOG) pathway. Osmotic shock is sensed by at least two putative transmembrane osmosensors; one of those, Slnlp-Ypdlp-Ssklp forms a phosphorelay system similar to bacterial two-component systems. The signal is then transmitted through a MAP kinase cascade eventually to Hoglp, which is translocated to the nucleus. Hoglp controls andlor interacts with different transcription factors to stimulate expression of more than 100 genes. Examples are the glycerol biosynthesis genes GPDl and GPPl, the gene for the sodium pump ENAl, the gene GRE2 whose product may be involved in detoxification of oxygen radicals, AR09, which encodes in enzyme in amino acid metabolism, CITl, which encodes a catalase and HSPl2, which encodes a heat shock protein of unknown function. NCE3 encodes carbonic anhydrase and an unknown signaling pathway mediates its induction by osmotic stress. The transcription factors Msn2p and Msn4p mediate a general stress response and their subcellular localization is controlled by protein kinase A. The scheme on the left hand side depicts glycerol metabolism. Grey: osmosensors; black with white text: protein kinases; gray fountain fill: protein phosphatases; shaded: enzymes; dotted fill: transcription factors. 41 Fps 1P has been demonstrated by direct transport assays with radio labelled glycerol to mediate transport of glycerol into and out of the yeast cell (Luyten et al., 1995; Sutherland et al., 1997; Tamás et al., 1999). The glycerol facilitator from Escherichia coli, GlpF, when expressed in yeast, can replace Fpslp's glycerol transport function lending further support to the role ofFpslp as a glycerol transporter (Luyten et al., 1995; Sutherland et al., 1997; Tamás et al., 1999). The phenotype associated with deletion ofFpslp clearly classifies this protein as a glycerol exporter, namely the inability to grow under anaerobic conditions and its sensitivity to hypo-osmotic shock (Tamás et al., 1999). As indicated above, yeast cells produce glycerol when grown in the absence of oxygen for redox balancing (Ansell et al., 1997). In cells lacking Fpsl p, glycerol accumulates inside the cell and inhibits growth, presumably because it leads to an osmotic imbalance (Tamás et al., 1999). Under these conditions glycerol can be regarded as a metabolic end or waste product and hence Fps 1p serves as a waste product exporter. When yeast cells are grown in high osmolarity medium and then shifted to low osmolarity, they dispose of 80% of their accumulated glycerol within 5 minutes (Luyten et al., 1995). In contrast, cells lacking Fpslp require 60 minutes to achieve the same low glycerol level and survive a hypo-osmotic shock in a 100-fold lower proportion than wild type cells. Those cells that survive resume growth more slowly (Luyten et al., 1995; Tamás et al., 1999) and, moreover, if the fpsl á mutation is combined with a mutation that weakens the cell wall, a hypo-osmotic shock is lethal (Tamás et al., 1999). Since cells lacking Fpslp grow in low osmolarity medium as well as the wild type, the fpsl á mutant is specifically sensitive to a hypo-osmotic shock and thus far is the only yeast mutant known to display such a phenotype (Tamás et al., 1999; Ferreira and Hohmann, unpublished observations). The role ofFpslp in osmoregulation and the control of cellular glycerol content are supported by further observations. Mutants lacking Fpslp exhibit diminished signaling through the HOG pathway, apparently because they can accumulate glycerol after a hyperosmotie shock faster than wild type cells (Tao et al., 1999; Tamás, Rep and Hohmann unpublished observations). Strikingly, fpsl LImutant cells show a defect in cell fusion during the mating process of haploid yeast cells; this defect is apparently also associated with altered osmoregulation because it can be suppressed by deletion of the GPDI gene and hence by reduction of glycerol production (philips and Herskowitz, 1997). Cell fusion requires local cell wall degradation and it appears that yeast cells have to relieve osmotic pressure at this point in order to prevent cell bursting. 42 KIlIO BGO·yCCVD .·.IlR.A'1Il1aSTy t ·B . I,S::~O(o)O -O()o0 60 1-0_ =~.~ ~";i :l ;~ 40 3:-9 40 II)'_s:: =II)'_s:: - eoe.....~~ 01-0 20 20E~ E~'-" - '-" 0 -5 0 5 io 15 20 25 30 -5 0 5 la 15 20 25 30 Time after hypo-osmotic shock (min) Time after hypo-osmotic shock (min) Figure la. Intracellular glycerol content of S. cerevisiae (0), S. pombe (e) and Z. rouxii CT) at 28°C (A) and O°C (B) when subjected to hypo-osmotic shock (from 0.86 M NaCl to O.09 M Nael in glucose- YNB). 68 100 ,.-., 100 c ..00 -~..... 80 80c Cl) oc 0 60 60 (.) 40 40 co0····0 ...0. ·'0'" ..... 20 20 o +---,--.---r----.-_-,----,_-.--' -5 0 5 10 15 20 25 30 -5 0 5 10 15 20 25 30 Time after hypo-osmotic shock (min) Time after hypo-osmotic shock (min) Figure lb. Intracellular content of the osmolytes glycerol (0), arabitol (V) and erythritol (0) by the Pichia sorbitophila at 28°C (A) and O°C (B) after hypo-osmotic shock (from 0.86 M NaCl to 0.09 M NaCl in glucose- YNB). 69 ISO -.. ..-.t:: 160 01) ei) ....... -0 ~ 140 C "0 120 ~ 8 '-'" s:: 100 0 '.0 --&- Glycerol release at 2SoC-roI-< SOCl ··0·· Glycerol release at OOC Cl) os:: o0 60 ë5 I-< Cl) o 40 ;:>-, CS 20 0 0 5 la 15 20 25 30 35 . Time after hypoosmotic shock (min) Figure 2. Extracellular glycerol from Z. rouxii after hypo-osmotic shock (0.S6 M NaCI - 0.09 NaCI in YNB-glucose) 70 100 -&- Glycerol .0····· . ·0·· Arabitol - 80.';..t_:.., 0 .Q...). .. ;>. 0 E 60 til .0 0.... cd o-: :2 ~ CJ .c..d.. 40 ..s... o 20 o 2 4 6 8 10 12 14 16 NaCl concentration of the hypo-osmotic media (%) Figure 3a. Release of osmolytes (at 280C) as a function of shock intensity (15% to 0 % NaCl) in Z. rouxii NRRL 2547. Intacellular osmolytes were extacted 5 min after hypo-osmotic shock and expressed in percentages, taking the iso-osmotic values as 100 % 71 100 ~u "v0 0..s:: ,,-., c:: Cl) c:: .~ u 0 80 ~'';::'g -I8- 0 .I.-o ...Cl) c:: Iv- 0 vI U 60 U 0 c:: "E>h. 0.. 0c-, u I-..s::~ "r3o 1V-. •.;_::1 _ t:=:c:: 40 - cd· ..... 8rco:.:_c.-. 0 '-::=;18/')~'-" 20 o o 1 2 3 4 5 6 Nael (% difference) Figure 3b. Release of accumulated glycerol as a function of shock intensity in S. cerevisiae. 72 r.zzzl Glycerol Arabitol NaCI YNB PEG Ethanol Glucose (Control) Solutes in transfer medium Figure 4. Retention of osmolytes (mean ± standard deviation of triplicate determinations) by Z. rouxii when cultivated at 0.972 60 aa "'evIlC I (J 8 a aC ~ c Glycerol aell 0..;>... (J (%pem Arabitol rI-o...c~I-..;::l 40"'3 (1).- Il!!JBi Erythritol _-c..t..:.:ccd·- ~roc.-c.....a .t::8~ .5V)'-' 20 0 YNB CCCP Gd Figure 5. Effect of the protonophore CCCP (50 ~ and the channel blocker gadolinium (50 mM) on osmolyte release by Z rouxii (A) or P. sorbitophila (B) during hypo-osmotic stress from 0.86 M NaCl (0.972 a ) to 0.09 M NaCI (0.996 a ). w w 76 REFERENCES Adler, L., Blomberg, A, and Nilsson, A (1985). Glycerol metabolism and osmoregulation in the salt tolerant yeast Debaryomyces hansenii. J Bacterial. 162,300-306. Ajouz, B., Berrier, C., Garrigues, A, Besnard, M., Ghazi, A (1998). Release of Thioredoxin via the mechanosenstive channel mscL during osmotic downs hock of Escherichia coli cells. J Bioi. Chem. 273, 26670-26674. Attfield, V. P. (1998). Physiological and molecular aspects of hyper osmotic stress tolerance in yeasts. In Recent Research Developments in Microbiology Vol. 2 Part IT, S.G. Pandala, ed. (Research Signpost, Trivandrum, India), pp.427-440. Batiza, AF., Schulz, T., and, Masson, P.H. (1996). Yeast respond to hypotonic shock with a calcium pulse. J Bioi. Chem. 271,23357-23362. Berrier, C., Coulmbe, A, Szabo, 1., Zoratti, M. and Ghazi, A (1992). Gadolinium ion inhibits loss of metabolites induced by osmotic shock and large stretch- activated channels in bacteria. Eur. J Biochem. 206, 559-565. Blomberg, A, and Adler L. (1992). Physiology of osmotolerance in fungi. Adv. Microbial. Physiol. 33, 145-212. Brown, AD. (1974). Microbial water relations: Features of the intracellular compostion of sugar-tolerant yeasts. J Bacterial. 118, 769-777. Brown, A D., and Edgley, M. (1980). Osmoregulation in yeast. In Symposium on genetic engineering of osmoregulation, D. W. Rains, R. C. Valentine and A Hollander, eds. (New York: Plenum Press), pp. 75-90. Brown, AD., and Simpson, J.R. (1972). Water relations of sugar-tolerant yeasts: the role of intracellular polyols. l.Gen. };1icrob. 72, 589-591. Chamberlin, M.E and Strange, K. (1989). Anisosmotie cell volume regulation: a comparative view. Am. J Physiol. 257, C159-C173. 77 Ganthala, B., Marshall, lB., May, W.I. (1994). Xerotolerance in fission yeasts and the role of glycerol as compatible solute. Arch. Microbiol. 162, 108-113. Glaasker, E., Konings, W.N., and Poolman, B. (1996). Glycine betaine fluxes in Lactobacillus plantarum during osmostasis and hyper- and hypo-osmotic shock. J Bioi. Chem. 271, 10060-10065. Gilles, R. (1987). Volume regulation in cells ofeuryhaline invertebrate. In Cell Volume Control: Fundamental and comparative aspects in animal Cell, R. Gilles, A. KleinzeIler and L. Bolis, eds. (New York: Academic press), pp. 205-247. Goyal, A. (1989). Intracellular glycerol in DunalielIa is depleted by intracellular metabolism in response to hypoosmotic stress by dilution. FEMS Mierob. Lett. 61, 145-148. Hallsworth, J.E. (1998). Ethanol induced water stress in yeast. J Ferment. Bioeng. 85, l25-137. Kayingo, G., Casaleggio c., Tamás, M. L, Hohmann, S., and Prior, B. A. (2000). An investigation of the possible existence of homologues of FPSl, a glycerol facilitator of Saccharomyces cerevisiae, in the osmotolerant yeast Zygosaccharomyces rouxii. In Molecular Biology and Physiology of Water and Solute Transport. (Kluwer academicIPlenum Publishers), pp.393-404 Kirk, K. (1997). Swelling-activated organic osmolyte channels. J Membr. Bioi. 158, 1-16 Lages, F., and Lucas, C. (1995). Characterization of a glycerol/H" symport in the halotolerant yeast Pichia sorbitophila. Yeast 11, 111-119. Lages, F., and Lucas, C. (1997). Contribution to the physiological characterization of glycerol active uptake in Saccharomyces cerevisiae. Biochim. Biophys. Acta 1322,8-18. Lages, F., Silva-Graca, M., and Lucas, C. (1999). Active glycerol uptake is a mechanism underlying halotolerance in yeasts: a study of 42 species. Microbiology 145, 2577-2585. 78 Levin, D.E., and Errede, B. (1995). The proliferation of MAP kinase siganalling pathways in yeast. Curro Opin. Cell. Biol. 7, 197-207. Loureiro-Dias, M C. (1987). Glucose and polyol transport systems in Candida intermedia and their regulation J Gen. Microbial. 133,2737-2742. Lucas, C., da Costa, M. and Van uden, N. (1990). Osmoregulatory active sodium- glycerol co-transportin the halotolerant yeast Debaryomyces hansenii. Yeast 6, 187-191. Luyten, K., Albertyn, l, Skibbe, W.F., Prior, B.A.l Ramos, l, Thevelein, lM., Hohmann, S. (1995). Fps1, a yeast member of the MIP family of channel proteins, is a facilitator for glycerol uptake and efflux and is inactive under osmotic stress. EMBO J 14, 1360-1371. Norkrans, B. (1966). Studies on marine occurring yeasts: Growth related to pH, NaCl concentration and temperature. Arch. Microbial. 54, 374-392. Onishi, H. (1963). Osmophilic yeasts. Adv. Food Res. 12,53-94. Perozo, E., Cortes, D.M., and Cuello, L.G. (1999). Structural rearrangements underlying K+-channel activation gating. Science 285, 73-78. Poolman, B., and Glaasker, E. (1998). Regulation of compatible solute accumulation in bacteria. Mol. Microbial. 29,397-407. Ruffert, S., Lambert, C., Peter, H., Wendish, Y.F., and Kramer, R. (1997). Efflux of compatible solutes in Corynebacterium glutamicum mediated by osmoregulated channel activity. Eur. J Biochem. 247, 572-580. Schleyer, M., Schmid, R. and Bakker, E.P. (1993) Transient, specific and extremely rapid release of osmolytes from growing cells of Escherichia coli K-12 exposed to hypoosmotic shock. Arch. Microbial. 160,424-431. 79 Tamás, M. J, Luyten, K., Sutherland, F. C. W., Hernandez, A., Albertyn, J, Valadi, H., Li, H., Prior, B. A., Kilian, S. G., Ramos, J, Gustafsson, L., Thevelein, J and M., Hohmann, S. (1999). Fps1p controls the accumulation and release of the compatible solute glycerol in yeast osmoregulation. Mol. Microbiol. 31, 1087-1104. Thomas, B. J. and Rothstein, R. J (1989). Elevated recombination rates in transcriptionally active DNA. Cell 56, 619-630. Van Eck, J. H., Prior, B. A., and Brandt, E. V. (1989). Accumulation of poly hydroxy alcohols by Pichia anomala in response to water stress. J Gen. Microbiol. 135, 3505-3513, Van Eck, JH., Prior, B.A. and Brandt, E.V. (1993). The water relations of growth and polyhydroxy alcohol production by ascomycetous yeasts. J Gen. Microbiol. 139, 1047-1054. Van Zyl, P. J., Kilian, S. G., and Prior, B. A. (1990). The role of an active transport mechanism in glycerol accumulation during osmoregulation by Zygosaccharomyces rouxii. Appl. Microbiol. Biotechnol. 34,231-235. Van Zyl, PJ., and Prior, B.A. (1990). Water relations of pol yoI accumulation by Zygosaccharomyces rouxii in continious culture. App. Microbiol. Biotechnol. 33, 12-17. Yancey, P. H., Clark, M. E., Hand, S. C., Bowlus, R. D., and Somero, G. N. (1982). Living with water stress: evolution of osmolyte systems. Science 217,1214-1222. Yang, x.C. and Sachs, F. (1989). Block of stretch-activated ion channels in Xenopus oocytes by gadolinium and calcium ions. Science 243, 1068-1071. 80 CHAPTER4 Effect of Ergosterol on Survival and Glycerol Release from Saccharomyces cerevisiae cells after Osmotic Downshoek Abstract Lipid composition influences the membrane permeability of solutes during adaptation of yeast cells to osmotic stress. In this study, we investigated the effect of ergosterol on survival and glycerol release from Saccharomyces cerevisiae cells after an osmotic downshock. The wild- type strain survived the osmotic downshock and grew to a similar degree in the presence or absence of ergosterol supplements. By contrast, S. cerevisiae cells lacking a glycerol facilitator (the jpsILlstrain), grew poorly upon an osmotic downshock, but apparently survived the shock better, and recovered more rapidly, if ergosterol was supplied. The erg-I disruption mutant, which is unable to synthesize ergosterol, survived and recovered from the osmotic shock more successfully at the higher ergosterol concentration. Transport studies showed a rapid efflux of glycerol from the. wild-type cells upon osmotic downshock. The glycerol content of wild-type cells was reduced to about 20% of its initial value within 5 min regardless of addition of exogenous ergosterol. However, the glycerol content of the jpsi LI strain was only reduced to 80% of its initial value within 5 min, and remained at this level for at least 30 min. When exogenous ergosterol was supplied, the rate and amount of glycerol release was markedly enhanced in the fpsl á mutant. The polyene antibiotic nystatin, which affects membrane permeability, caused S. cerevisiae cells to release a large amount of glycerol and equally inhibited the growth of wild-type and jpsi deletion strains in medium containing 5% (w/v) NaCl. This study demonstrated the role of ergosterol in glycerol efflux and survival in S. cerevisiae after an osmotic downs hock and provided additional evidence for the significance of membrane permeability and glycerol conservation in yeast osmoregulation. 81 INTRODUCTION The role of glycerol and membrane lipid composition in yeast stress responses has been studied extensively (Swan and Watson, 1996; Prior and Hohmann, 1997; Attfield, 1998). It has been known for decades that most yeast cells accumulate substantial amounts of glycerol in their cytosol in response to hyper-osmotic stress. The accumulated glycerol serves as an osmolyte (compatible solute) and maintains turgor without interfering with cellular functions (Brown, 1978; Yancey et al. 1982). It is also well established that yeast cells regulate their membrane lipid composition in response to osmotic stress (Tunblad-Johansson and Adler, 1987; Watanabe and Takakuwa, 1987; Hosono, 1992; Yoshikawa et al., 1995). However, little is known about the relationship between membrane lipid composition, glycerol transport and survival during hypo-osmotic stress. Studies with liposomes indicate that glycerol permeation is influenced by membrane composition (De Gier, 1993) but the key components in the yeast membrane that influence glycerol transport are not yet well defined. In Saccharomyces cerevisiae, the movement of glycerol across the cell membrane occurs via active transport, by channel-mediated diffusion (Fpsl protein), and by passive diffusion across the plasma membrane (Sutherland et al., 1997). The extent to which glycerol permeates the cell may then be influenced by the membrane lipid composition as observed in osmotolerant yeasts when grown under osmotic stress (Watanabe and Takakuwa, 1987). During hypo-osmotic stress, yeast cells rapidlyrelease their intracellular glycerol and this release appears to be controlled by a membrane channel protein Fps 1p (Luyten et al.. 1995; Tamás, et al., 1999). Cells lacking the Fpsl glycerol channel grow poorly during hypo- osmotic stress, while those expressing an unregulated channel are sensitive to hyperosmotie stress (Tamás et al., 1999). Sutherland et al. (1997) revealed that the deletion of FPSI results in diminished passive diffusion of glycerol and alters cellular lipid composition. Interestingly, the deletion of the glycerol facilitator (glpF) in Escherichia coli, a close homologue of FPSl, also reduces the permeability across the plasma membrane (Truniger and Boos, 1993). Recently, we have observed that FPSI deletion leads to an alteration in the ergosterol contents of the whole cell and the plasma membrane (Toh et al., 2000). These observations raise the question whether a link between glycerol channel proteins, membrane permeability and cellular lipid synthesis existed. Therefore, the present study was carried out to investigate the relationship between ergosterol content, glycerol release and survival of yeast cells during hypo-osmotic stress. Ergosterol is the most abundant sterol in yeast membranes (Zinser et al., 1993) and plays an important role in determining the integrity of biological membranes. 82 In turn, ergosterol apparently influences the membrane permeability to solutes (Rattray 1988; Van der Rest et al., 1995). Here, we demonstrate the role of ergosterol in glycerol efflux and further highlight its significance in osmotolerance. MATERIALS AND METHODS Yeast strains and growth conditions The Saccharomyces cerevisiae strains lacking a glycerol facilitatorjpslL1::HIS3 (Tamás et ai., 1999) and jpsll!:.::LEU2 (van Aelst et al., 1991) were derived from the haploid laboratory strain W303-1A (Thomas and Rothstein, 1989). The ergosterol (ergl-disruption) mutant and its corresponding wild-type strain (X2180-lB MAT a, SUC2, mal, mel, gal2 cuplR) were kindly supplied by Drs. L.W. Parks and J.H. Crowley of North Carolina State University. All yeast strains were routinely grown on YEPD (10 g rl yeast extract, 20 g r' each of peptone and glucose, pH 6) or on defined medium (6.7 g r' yeast nitrogen base, 20 g r' glucose and amino acids, pH 6, glucose- YNB) with supplements as indicated in the text. Osmotic downshock sensitivity experiments Yeast cells were grown on medium (0.998 aw) without ergosterol supplementation (control) and on medium with ergosterol supplementation at a final concentration of 10 ug ml-I ergosterol. For the ergl-disruption mutant and the X2180-lB isogenie wild-type strains, ergosterol (Sigma) and fatty-acid supplements were supplied as described by Swan and Watson (1998). The preparation of osmotically stressed cells is similar to that described previously (Luyten et al., 1995). At a standardised time point (OD6oo 1.0), the cells were harvested and resuspended in the same media supplemented with 50 g r' NaCI (0.972 aw) and incubated at 30°C for 3 h. At the end of that period, the cultures were harvested, quickly washed once in growth medium without salt (0.998 aw) and finally resuspended in the same medium to a predetermined optical density. Spot tests were carried out by serial decimal dilutions of the cell suspension prior to plating (Luyten et al., 1995). Five microlitres of the cell suspensions were spotted onto agar plates without NaCI addition and incubated at 30°C until the colonies were visible. Aliquots of the cell suspensions were also removed for biomass dry weight and intracellular glycerol determinations. 83 Nystatin sensitivity experiments The polyene antibiotic nystatin (Sigma, USA) was dissolved in ethanol (20 mg mr') and added to exponentially growing cells to evaluate its effect on the growth of fpsi L1mutant strain and its isogenie wild type. The final concentrations of the antibiotic in the culture media were, 0.02, 0.05, 0.1 or 0.2 ug mr' while that of ethanol was < 0.5 % (v/v). To investigate the effect of nystatin on intracellular glycerol levels, yeast cells were grown in glucose- YNB until early exponential phase, harvested and then resuspended in glucose- YNB containing 5% NaCI. After three hours in hyper-osmotic media, nystatin (0.2 ug mr') was added and samples taken for glycerol determination. Intracellular glycerol determination Aliquots of cells exposed to hypo-osmolar media or to nystatin treatment were periodically taken in a 1.5 ml Eppendorf tubes. Cells were quickly harvested, resuspended in lO ml Tris lysis buffer (van Eck et al.. 1989) and then boiled for 10 min in a water bath with regular vortex agitation. The suspensions were centrifuged at 16000 r.p.m. for 5 min to remove cell debris and the supernatant was retained for glycerol determination. Intra- and extracellular glycerol concentrations were determined spectrophotometrically at 340nm with the aid of a commercial enzymatic kit (Kit No. 148270; Boehringer-Mannheim, Germany). 84 RESUL TS AND DISCUSSION Effect of ergosterol on survival of yeast cells after osmotic downshock Osmotic downshock sensitivity experiments were carried out to test the effect of cellular ergosterol levels on survival, the rate of glycerol release and subsequent recovery of the yeast cells after an osmotic downshock. The wild-type strain recovered from osmotic downshock, and grew to a similar degree regardless of ergosterol supplementation (Fig. lB), By contrast, cells of the fpsl á strain grew very poorly on glucose- YNB, but apparently survived the osmotic downshock better, and recovered more rapidly, if ergosterol was supplied (Fig. lA). The difference between the growth of the fpsl á strain in the presence or absence of exogenous ergosterol was apparent throughout the incubation period between 12 and 30 h. The erg-l disruption mutant which is unable to synthesise ergosterol (parks et al., 1995) and X2180-lB wild-type strain were subjected to osmotic downshock, and their growth and survival were assessed. The wild-type strain survived and recovered from the osmotic downs hock equally well irrespective of the concentration of ergosterol supplied (Fig. lB). By contrast, the erg-I disruption strain survived and recovered from the osmotic downshock more successfully at the higher ergosterol concentration (Fig. lB). This result was consistently observed regardless of the incubation time after the downshock (between 12 and 30 h). Effect of ergosterol on the release of glycerol after osmotic downshock It has been proposed that theJPsl.1 mutant grows poorly upon an osmotic downshock due to its inability to rapidly dispose of the accumulated glycerol (Luyten et al.. 1995). To determine whether the improved survival upon a downshock observed above was associated with changes in solute transport, the effects of ergosterol supplementation on glycerol efflux was investigated by monitoring the intracellular glycerol concentrations prior to and immediately after an osmotic downs hock. There was a rapid efflux of glycerol from cells of the wild-type strain upon osmotic downs hock as reported previously (Luyten et al.. 1995). The glycerol content of W303-1 A wild-type cells was reduced to about 20% of its initial value within 5 min, and remained at this level for at least 30 min, regardless the addition of exogenous ergosterol (Fig. 2A). 85 On the other hand, the glycerol content of fpsi L1cells was only reduced to 80% of its initial value within 5 min of downshock, and remained at this level for at least 30 min. However, when exogenous ergosterol was supplied, the rate and amount of glycerol release was markedly enhanced (Fig. 2A). In the presence of ergosterol, fpsi L1 mutant reduced its intracellular glycerol concentration to around 60% of its initial value within 5 min. During iso-osrnotic conditions, no glycerol release was observed in both the wild-type (Fig. 2B) and the fpsl S. strain (Fig. 2C) regardless the addition of exogenous ergosterol. Glycerol release occurs after an osmotic downsock to balance up the turgor pressure difference created as cells are moved from high to low osmolarity environments. Our observations that ergosterol may playa role in glycerol efflux is in agreement with previous reports in Candida albicans where mutants inhibited in ergosterol synthesis exhibited decreased permeability for glycerol (pesti and Novak, 1984). However, the mechanisms by which ergosterol improve glycerol release and survival of yeast cells are not yet clear. It could presumably be related to the membrane stablising effects of ergosterol (Hossack and Rose 1976) and the associated improvements in membrane fluidity (Arami et al., 1997). Effect of nystatin on the growth and release of glycerol fromfpslL1 strains The susceptibility to nystatin is known to correlate with the ergosterol content in the plasma r l membrane of yeast cells (Hamilton-Miller 1973). For instance, mutants of S. cerevisiae thatcannot synthesise ergosterol are more tolerant to nystatin (Bard, 1972). In this study, we investigated the effect of nystatin on the growth of S. cerevisiae wild-type cells and fpsi L1 mutant which exhibits an alteration in the cellular ergosterol contents (Toh et al., 2000). As shown in Figure 3, nystatin inhibited the growth of both strains in YEPD media with 5% NaCl whereas cells grown in media without Nael were only slightly affected. The presence of 5% NaCI alone also caused a delay in the growth of yeast cells compared to the growth of cells without NaCI but addition of nystatin aggravated the effect. Both strains were equally sensitive to the various concentrations of nystatin tested. It has been suggested that nystatin interacts with ergosterol and forms an aqueous channel in the plasma membrane (Kobayashi and Medoff, 1977). These pores result in the leakage of intracellular osmolytes that are indispensable for growth of yeast cells in hyper-osmolar media (Hosono, 2000). 86 We further investigated the changes in intracellular glycerol levels after addition of 0.2 ug nystatin mr! to salt-stressed cells. As shown in Figure 4, nystatin caused salt-stressed cells to release large amount of glycerol. The intracellular glycerol decreased by more than 70% in 30 minutes of incubation with nystatin. Since the action of nystatin is generally thought to be due to an interaction with ergosterol in the membrane, it was anticipated that the fps l deletion strain which shows about 26% reduction in ergosterol content, could be less susceptible to nystatin compared to the wild-type. However, no significant difference was observed in susceptibility or glycerol release between the two strains. This result suggest that although a 26% difference in ergosterol content might be sufficient to alter membrane solute transport properties as we have shown in Figure 1 above, it might not be adequate to cause significant differences in susceptibility to the concentrations of nystatin used in this study. On the other hand, the ergosterol differences reported by Toh et al., (2000) were observed during steady state conditions. The ergosterol content in the plasma membrane increases drastically when yeast cells are exposed to high salinity media. For instance, the ergosterol content in the plasma membrane of Zygosaccharomyces rouxii grown in YEPD media containing 15% Nael increased 2. 9 fold higher than that of unstressed cells (Hosono, 1992). It is therefore possible that when S. cerevisiae cells were grown in media containing 5% NaCl, the ergosterol content of both the wild-type and the fpsl deletion mutant increased to proportions that could not allow detectable differences in susceptibility to polyene antibiotics. 87 A Wild-type .- fps l Lt strain + l0J.lg ... :.: - ergosterol mr! .'. JO. ., {psILt strain •- •~w .. ~.'. . ..,.~~~.. B 1 2 3 4 5 1= erg-l disruption mutarit + 5 ug ergosterol mlo! 2 = erg-l disruption mutant + 10 ug ergosterol ml" 3 = Wild-type without ergosterol 4 =Wild-type + 5 ug ergosterol mlo! 5 =Wild-type + 10 ug ergosterol ml:' Figure 1. Survival of S. cerevisiaeW303-1A euáfpsl á strains (A) as well as the erg- 1 desruption mutant and the X2180-1B wild type strains (B) grown on glucose- YNB after an osmotic downshock with and without ergosterol supplementation. 88 100 -so-::- .~ . __ ., '\l-- .. ~ 80 V'-"-"---V" -\1-"-"-"\1Q) C) os:: C) ~..--.S- .........60 .__ -""?"-----~ -------- -~ ....···············0 O+-----~----~----~----~------~----~----~~ -5 o 5 10 15 20 25 30 Time after osmotic downshock (min) Figure 2A. Glycerol release from the fpsl é: (triangles) and W303-1A wild-type strain (circles) before and after osmotic downshock with (solid symbols) and without (open symbols) ergosterol supplementation. 89 140 ,-.... c:: 0 ...r..o.......... 120c:: Cl) eu 0 cu 100.pa ·2 c.,.... 0 80 ~.._., 0... Cl) 60 u :>-. O..hro. :§ 40 Q) ·r ... uo ...... 20 ~ 0 -5 o 5 10 15 20 25 30 Time after osmotic downshock (min) Figure 2B. Glycerol release from S. cerevisiae W303-1A wild-type strain during iso- osmotic conditions (triangles) or after osmotic downshock (circles) with (solid symbols) and without (open symbols) ergosterol supplementation. 90 '"2 120 o '';::: . . 'il ..e.... ::;;.- 5 cCJ 100 ~--- ..o . CJ ~. \b...-... '. _.().. ---0 '2 80 \ 0-- .. - ..0--" -D-"-" \ tt. ............... __ ... 60 --e----- ..-- 40 20 O+-----.----.----,-----.----.----,-----.-~ -5 o 5 10 15 20 25 30 Time after osmotic downshock (min) Figure 2e. Glycerol release from fpsi Li strain during iso-osmotic conditions (triangles) or after osmotic downshock (circles) with (solid symbols) and without (open symbols) ergosterol supplementation. 91 8 7 A 6 .0··················0 5 0 0 \Cl Q 4 0 .,,;rt 3 »: /"" »: /"" 2 . -..rv. _- -~ ---'V" .. _,'-"--'\1-::;:::::.-'-=-- .. - .. ~ 1 0 0 2 4 6 8 10 12 14 16 18 20 Time (hours) 92 8 B 7 6 0·················.0 o 5 o \0 § 4 _. 3 _...,_...,_..., -~ _...,2 - ~ .-. -"---"- _ .. ......sv- .. -" - .. ----sv 1 o 2 4 6 8 10 12 14 16 18 20 Time (hours) ____._ YEPD + 0.21lgmr1 nystatin .. ·0··· YEPD + 5% NaCI ~ YEPD + 5%NaCI + 0.05 mgml" nystatin -'\1'" YEPD + 5% NaCl + 0.11lgmr1 nystatin Figure 3. The effect of nystatin on the growth of S. cerevisiae W303-1A wild-type strain (A) and theJPsIL1 mutant (B). The polyene antibiotic nystatin was dissolved in ethanol (20 mg ml") and added to exponentially growing cells to the final concentrations, 0.05, 0.1 or 0.2 ug mr1 while that of ethanol was < 0.5 % (v/v). 93 ,.-... c 100 .0~ .t"c. d '.". Q) uc 80 0u .ta~ :5 60 o. e:::_R, '0 Q"'") u ;>, 40 ___._ fpsl Llmutant ""5'0 "'" .. -0 .. W303-1A wild-typetd :2 '0 Q) u 20 ,. t .". d s'.". o +-------.------.------~------.------.------~----~ o 5 10 15 20 25 30 35 Time after addition of nystatin (min) Figure 4. The effect of nystatin on the intracellular glycerol content of S. cerevisiae W303-1A wild-type and thefpsILl mutant strains during osmotic stress (5% NaCl). Yeast cells were grown in glucose- YNB until early exponential phase; harvested and then resuspended in glucose-YNB containing 5% Nael. After three hours in hyper- osmotic media, nystatin (0.2 ugml") was added' and samples taken for glycerol determination. 94 REFERENCES Arami, S., Hada, M.and Tada, M. (1997). Reduction of ATPase activity accompanied by photodecomposition of ergosterol by near-UV irradiation in plasma membranes prepared from Saccharomyces cerevisiae. Microbiology 143, 2465-2471. Attfield, V. P. (1998). Physiological and molecular aspects of hyper osmotic stress tolerance in yeasts. In Recent Research Developments in Microbiology Vol. 2 Part II, S.G. Pandala, ed. (Research Signpost, Trivandrum, India), pp.427-440. Bard, M. (1972). Biochemical and genetic aspects of nystatin resistance in Saccharomyces cerevisiae. J. Bacteriol. 111,649-657. Brown, AD. (1978). Compatible solutes and extreme water stress in eukaryotic micro- organisms. Adv. Microbial. Physiol. 17, 181-242. De Gier, J. (1993). Osmotic behaviour and permeability properties of Iipo somes. Chemo Phys. Lipids 64, 187-96. Hamilton-Miller, lM.T. (1973). Chemistry and Biology of the polyene macrolide antibiotics. Bacteriol. Rev. 37, 166-196. Hosono, K (2000). Effect of nystatin on the release of glycerol from salt-stressed cells of the salt-tolerant yeast Zygosaccharomyces rouxii. Arch. Mierobtol. 173,284-287. Hosono, K (1992). Effect of salt stress on lipid composition and membrane fluidity of the salt-tolerant yeast Zygosaccharomyces rouxii. J. Gen. Mierobtol. 138,91-96. Hossack, lAand Rose, AH. (1976). Fragility of plasma membranes in Saccharomyces cerevisiae enriched with different sterols. J. Bacteriol. 127,67-75 Kobayashi, G. S., Medoff, G. (1977). Antifungal agents: Recent developments. Ann. Rev. Microbiol. 31,291-308. Luyten, K, Albertyn, L, Skibbe, W. F., Prior, B. A, Ramos, L,Thevelein, J. M., and Hohmann, S. (1995). Fps1, a yeast member of the MJP family of channel proteins, is a facilitator for glycerol uptake and efflux and it is inactive under osmotic stress. £MBO 1. 14, 1360-1371. Parks, L.W., Smith, S. J, and Crowely, J. H. (1995). Biochemical and physiological effects of sterol alteration in yeast- a review. Lipids 30, 227-230. 95 Prior, B.A, and Hohmann, S. (1997). Glycerol production and osmoregulation. In: Yeast sugar metabolism, F. K Zimmermann and KD. Entian, eds. (Technomic Puplishing company Inc., Lancaster, PA), pp. 313-335. Pesti, M., and, Novak, R. (1984). Decreased permeability of glycerol in an ergosterol- 'less mutant of Candida albicans. Acta Microbial Hung 31,81-84. Rattray, 1. B. M. (1988). Microbial Lipids. In Yeasts.Vol1, C. Ratledge and S. G.Wilkinson, eds. (Academic press, London), pp. 555-697. Sutherland, F. C. W., Lages, F., Lucas, C., Luyten, K, Albertyn, 1., Hohmann, S., Prior, B. A and Kilian, S. G. (1997). Characteristics ofFPS1-dependent and- independent glycerol transport in Saccharomyces cerevisiae. J Bacterial. 179, 7790-7795. Swan, I.M. and Watson, K (1997). Membrane fatty acid composition and membrane fluidity as parameters of stress tolerance in yeast. Can. J Microbial. 43, 70 -77. Swan, I. M. and Watson, K (1998). Stress tolerance in a yeast sterol auxotroph: role of ergosterol, heat shock proteins and trehalose. FEMS Microbial. Lett. 169, 191-197. Tamás, M. J., Luyten, K, Sutherland, F. C. W., Hernandez, A, Albertyn, J., Valadi, l H., Li, H., Prior, B. A, Kilian, S. G., Ramos, J., Gustafsson, L., Thevelein, J.M. and Hohmann, S. (1999). Fps1p controls the accumulation and release of the compatible solute glycerol in yeast osmoregulation. Mol. Microbial. 31, 1087-1104. Thomas, B. 1. and Rothstein, R. J. (1989). Elevated recombination rates in transcriptionally active DNA Cell 56, 619-630. Toh, T. H., Kayingo, G., Hallsworth, J. E., Kilian, S. G., Hohmann, S. and Prior B. A (2000). Mutations in glycerol metabolism influence membrane lipid compostion, glycerol efflux and survival during hypo-osmotic stress in Saccharomyces cerevisiae. In Preparation. 96 Truniger, V. and Boos, W. (1993). Glycerol uptake in Escherichia coli is sensitive to membrane lipid composition. Res. Microbiol. 144,56-574. Tunblad-Johansson, I, and Adler, L. (1987). Effects of sodium chloride concentration on phospholipid fatty acid composition of yeasts differing in osmotolerance. FEMS Microbiol. Lett. 43, 275-278. Van Aelst, L., Hohmann, S., Bulaya, B., de Koning, W., Sierkstra, L., Neves, M. J., Luyten, K., Alijo, R, Ramos, J., Coccetti, P., Martegani, E., de Magelháes, N. M., Brandáo, R L., Van Dijck, P., Vanhalewyn, M., Durnez, P., Jans, A W. H, and Thevelein, J. M. (1993). Molecular cloning ofa gene involved in glucose sensing. Mol. Microbiol. 8,927-943. Van Eck, J. H., Prior, B. A, and Brandt, E. V. (1989). Accumulation of polyhydroxy alcohols by Pichia anomala in response to water stress. J. Gen. Microbiol. 135, 3505-35l3. Van der Rest, M. E., Kamminga, A H., Nakano, A, Anraku, Y, Poolman, B. and Konings, W. N. (1995). The plasma membrane of Saccharomyces cerevisiae: structure, function and biogenesis. Microbiol. Rev. 59, 304-322. Yancey, P. H, Clark, M. E., Hand, S. C., Bowlus, R D., and Somero, G. N. (1982). Living with water stress: evolution of osmolyte systems. Science 217, 1214-1222. Yoshikawa, S., Mitsui, N., Chikara, K.-I., Hashimoto, H., Shimosaka, M. and Okazaki, M. (1995). Effect of salt stress on plasma membrane permeability and lipid saturation in the salt-tolerant yeast. J. Ferm. Bioeng. 80, 131-135. Watanabe, Y and Takakuwa, M. (1987). Effects of sodium chloride on lipid composition of Saccharomyces rouxii. Agric. Bioi. Chemo 48,2415-2422. 97 Zinser, E., Paltauf, F. and, Daum., G. (1993). Sterol composition of yeast organelle membranes and subcellular distribution of enzymes involved in sterol metabolism. J Bacteriol. 175, 2853 - 2858. 98 CHAPTERS An investigation of the possible existence of homolognes of FPS1, a glycerol facilitator of Saccharomyces cerevisiae, in the osmotolerant yeasts Zygosaccharomyces rouxii and Pichia sorbitophila Abstract Yeast species differ in their mechanisms of accumulating and transporting glycerol across the plasma membrane. In S. cerevisiae, glycerol accumulation is mainly controlled by a transmembrane channel protein Fpslp during osmoregulation The yeast Fpslp is a member of the MIP family of water channels and glycerol facilitators. In this study, we investigated the presence of channel-mediated glycerol transport and FPSI homologues in the osmotolerant yeasts Z. rouxii and P. sorbitophila. Transport studies revealed a rapid glycerol efflux which was characteristic of glycerol facilitators. Analysis of FPSI homologues by PCR and DNA probes resulted in weak hybridisation signals suggesting that the putative glycerol channel-encoding gene might have low sequence similarity to FPSI. S. cerevisiae mutants, in which FPSI has been deleted, survive poorly an osmotic downshock due to an inability to rapidly dispose of the accumulated glycerol. This phenotype was used to screen a Z. rouxii library for genes that might complement or suppress this growth defect. Analysis of complementing clones revealed genes encoding homologues of the S. cerevisiae CDCIO and DOM34 These genes, none of which is a MIP family member, have been implicated in the yeast cytoskeleton, cell cycle, or cell integrity pathway. Functional analysis indicated that the S. cerevisiae dom34 mutants are sensitive to osmotic stress and that the Z. rouxii DOM34 complements this growth defect suggesting that the DOM34 gene is involved in yeast osmotolerance. Physiological and genetic data point to the occurrence of glycerol facilitator protein(s) in the osmotolerant yeasts Z. rouxii and P. sorbitophila, previously assumed to permeate glycerol only via active transport and simple diffusion. However, these homologues might have low sequence similarity to FPSI as indicated by Southern blot hybridisation and PCR analysis. 99 INTRODUCTION Glycerol plays a protective role during osmotic stress and its formation is essential for redox balancing in yeast (Brown, 1990; Ansell et al., 1997). Yeast cells have consequently developed adaptive mechanisms to control glycerol flux within limits suitable for growth including; modulation of glycerol production and dissimilation, conservation and increased retention as well as regulating glycerol transport across the plasma membrane (for review see Prior and Hohmann, 1997; Attfield, 1998). However, the mechanisms of glycerol transport and its intracellular conservation differ considerably between yeasts and may also differ according to growth conditions (Sutherland et al., 1997; Lages et al., 1999). In S. cerevisiae, glycerol conservation is mainly controlled by a transmembrane channel protein Fps1 (Luyten et al., 1995; Sutherland et al., 1997; Tamás et al., 1999). Fps1p is a member of the M1P (Major Intrinsic Protein) family of transport proteins and is closely related to bacterial glycerol facilitators. However, Fps1p is unique among glycerol facilitators; the characteristic NP A motifs found in M1P proteins are not fully preserved. Secondly, Fps1p has long amino- and carboxy- terminal hydrophilic extensions, resulting in a protein of 669 amino acids, compared with the 250-300 amino acids for most M1P proteins (for recent review, Hohmann et al., 2000). The gene encoding Fps 1p was originally identified as a multicopy suppressor of the growth defect ofthefdplmutant in controlling glycolysis (van Aelst et al., 1991). Subsequent studies have shown that Fps1p is indeed a glycerol transport protein controlling both glycerol influx and efflux (Luyten et al., 1995; Sutherland et al., 1997). However, Fps1-mediated glycerol transport appears to be more important during efflux than in uptake (Tamás et al., 1999). S. cerevisiae mutants lacking FPSl cannot rapidly release glycerol and subsequently fail to cope with hypo-osmotic stress (Luyten et al., 1995). In addition, fpsl L1mutants grow poorly under anaerobic conditions and exhibit slightly diminished osmotic induction of GPDl expression (Tamás et al., 1999). Deletion of FPSl also causes cell fusion defects during mating (philips and Herskowitz, 1997) and leads to prolonged phosphorylation of the Mpk1p kinase in the PKC pathway after a hypo-osmotic stress. All these effects are consistent with a role of Fps1p in controlling the intracellular glycerol content. It appears that Fps1p opens during hypo-osmotic conditions to release glycerol and closes during hyper-osmotic conditions thereby conserving glycerol inside the cell (Luyten et al., 1995; Tamás et al., 1999). 100 Whether a similar mechanism of glycerol conservation occurs in other yeasts is still unknown and no other yeast protein has been reported to be homologous to Fps1p both in structure, function and regulation. Glycerol transport in the osmotolerant yeast Z. rouxii is so far known to occur via simple diffusion and by an osmotically active transport system (Edgley and Brown, 1978; Van Zyl and Prior 1990; Van Zyl et al., 1990). The purpose of this study was to explore the possible occurrence of a channel mediated glycerol transport and FPSi homologues in the osmotolerant yeasts Z. rouxii and P. sorbitophila, which in turn might provide additional insights into yeast stress tolerance. MA TERIALS AND ~THODS Strains, growth conditions, and transport experiments The following yeast strains were used: Pichia sorbitophila CBS 7064 (CSlR Y170), Zygosaccharomyces rouxii NRRL Y2547, Z. rouxii NRRL Y998, Z. rouxii NRRL Y225, Schizosaccharomyces pombe CBS 5682 (CSlR Y457) and Debaryomyces hansenii CBS 0767 (CSlR Y953). These wild-type yeast strains were obtained from the Microbial Resource Centre (MlRCEN), Department of Microbiology and Biochemistry, University of the Free State, South Africa. The various laboratory strains used in this study are listed in Table 1. The Saccharomyces cerevisiae strains lacking a glycerol facilitator fpsiL1::HIS3, fpsiL1::LEU2 mpkiL1::TRPi (Tamás et al, 1999) andfpsiL1::LEU2 !psiL1::TRPI (Luyten et al .. 1995) were derived from the haploid laboratory strain W303-1A (Thomas and Rothstein, 1989). The S. cerevisiae dom34 mutant was constructed and obtained from the laboratory of Dr JEngebrecht (State University of New York). A diploid strain MA121-3 of Z. rouxii (Ushio et al., 1988; 1996) was derived from the strain NRRL 2547 (MATa, prototrophic [ciro], i.e., harbouring no plasmid). All wild-type yeast strains were routinely grown on YEPD (10 g r' yeast extract, 20 g r' each of peptone and glucose, pH 6). Mutant strains were grown on defined medium (6.7 g r' yeast nitrogen base, 20 g r1 glucose and amino acids, pH 6) with supplements as indicated in the text. For the growth of the tpsl L1fpsi L1double mutant, glucose was replaced by 2% galactose whereas thefpsiL1 mpkl á double mutant was grown in presence of 1 M sorbitol. The Escherichia coli cells (TOPIOF') were routinely grown at 37°C overnight in LB media (10 g r' NaCl, 10 g r1 bacto-trypton, 5 g r' yeast extract). Osmotic stress and transport experiments were carried out as described previously in Chapter 3 and 4. 101 Molecular and genetic techniques Yeast and bacterial transformations were carried out using the lithium acetate method (Gietz and Schiestl, 1995) and the calcium chloride method (Maniatis et al., 1982) respectively. In general, nucleic acid manipulations were carried out according to Sambrook et al. (1989) unless otherwise mentioned. We routinely isolated plasmid DNA from bacteria using cetyltrimethylammonium bromide (CTAB) following the method of Del Sal et al. (1988). To isolate plasmids from yeast, an overnight culture was harvested, disrupted, and homogenised using acid washed glass beads. The homogenised cell suspension was centrifuged (13000 rpm, 5 min) and plasmid DNA was recovered from the supernatant. For sequencing experiments, plasmid DNA was purified using the Nucleobond AXI00 cartridges of Macherey-Nagel, Germany (Cat. No. 740521). Yeast genomic DNA was prepared from mid- stationary cultures as described by Sherman et al. (1986) whereas total RNA was isolated from exponentially growing yeast cells cultivated in YEPD with or without 5% NaCI. Cells were harvested and lysed with Zymolase (0.2 mg/ml) and l3-mercaptoethanol (8 ul/rnl) in extraction buffer (1 M sorbitol, 100 mM Nas-citrate, 60 mM EDTA, pH 7.0). After cell lysis (3 hours at 37°C) and homogenization, total RNA was isolated from the yeast spheroplasts using the Rneasy kit (Qiagen, Cat. No. 74103). Poly A+ rnRNA was purified from total RNA following the oligotex (Qiagen, Cat. No. 70022) purification protocols. Southern Blot Hybridisation Digested yeast genomic DNA (10 !lg) was fractionated by gel electrophoresis and transferred to a positively charged nylon membrane (Magna Graph, MSI, Cat. No. NJOHYOOO10) by capillary blotting (Sambrook et al.. 1989). The entire reading frame or a 426 bp fragment of FPSJ corresponding to the most conserved region (Fig. I), were generated by PCR from S.cerevisiae W303-1A genomic DNA, and used as probes. Hybridisation was carried out overnight at 68°C followed by stringent washes (2 X 5 min, 2 X SSC, l%SDS at room temperature, then 2 X 15 min, O.IX SSC, 0.1% SDS) under constant agitation. Less stringent hybridisation was performed at 55°C with a single wash (2 X 5 min, 2 X SSC, 1% SDS) at room temperature. Probe labelling and detection were performed using the DIG system (Boehringer Mannheim). 102 Polymerase chain reactions Oligonucleotide primers: Two pairs of specific primer were designed from S. cerevisiae sequences. Pair A corresponds to the Fpslp amino acid sequences, ISGAHL (sense) and ARDLGP (antisense) just outside the NPS and NLA boxes respectively (Fig. lA). Pair B and was designed from the N- and C-termini nucleotide sequences of the FPSI. Pair A GS S'ATCTCAGGTGCTCATTTG3' G6 5' TGGGCCCAGATCACGAGC 3' PairB FPSFO 5' GCTCTAAACGACTTTCTGTCCA 3' FPSRE 5' CCATAATGCGAATCTTCTGATG 3' Also, two pairs of degenerate primers were designed (Fig. I). Pair C was designed from the LNPSIT (sense) and NLARDL (antisense) motifs of the FPSI gene. Pair D was designed from the consensus motifs of the entire MlP family LNPAVT (sense) and NPARSF (antisense). Pair C sense Gl 5' GGGATCCYTNAAYCCNTCNATHAC 3' .......BamHI antisense G2 5' CGGAATTCRTCNCGNGCNARRTT 3' ........ EcoRI Pair D. sense ;G3SCTTCTAGAYTNAAYCCNGCNGTNAC 3' ........Xbal antisense G4.S'CGGAATTCAANSWNCKNGCNGGRTT3' EcoRl lUB (international union of Biochemistry) codes are used to represent mixed positions (degeneracy); R = AlG M = AlC W = AlT S = G/C K = GIT Y = CIT V = AlG/C D = AlG/T H = AlC/T B = G/C/T N = AlG/CIT. With 0.5 !JM specific primers, PCR amplifications were carried out (94°C, 1 min; 52°C, 50 sec; noc, 1 min; 30 cycles) by 2.5 units of Taq DNA polymerase in presence of 200 !JM dNTPs with 1.5 mM MgCh (Boehringer Mannheim). In all cases template DNA (50-100 ng) was denatured for 2 min at 95°C. 103 Degenerate Reverse Transcriptase Polymerase Chain reaction (RT-PCR) To optimize the degenerate RT-PCR procedure for homology cloning (Fig. l B), several parameters such as primers, MgCl! and template concentration were adjusted until the desired amplification was reached (Fig. 2). S. cerevisiae and E coli from which the MIP homologues are already known were also included as positive controls. The Titan™ One Tube RT-PCR system (Boehringer Mannheim Cat. No. 1888382) was used in presence of 2.5 !-JM primers, 1 mM dNTPs and 3 mM MgCh concentrations. The reaction also included Dithiothreitol (DTT) at a final concentration of 5 mM, 8 units of RNase inhibitor and the following cycling parameters. 30 min (reverse transcription) 94°C 2 min (initial denaturation) 1 min (cycle denaturation) 46°C 1 min (annealing) 3 min (elongation) for 35 cycles 68°C 7 min (prolonged elongation). Samples were analysed on 1.2% agarose gel. Bands of about 400 bp (typical of the MIP family) were obtained and immediately purified (Boehringer Mannheim purification system Cat. No. 1732676). To obtain large quantities of DNA for subsequent manipulations, purified samples were reamplified. Aliquotes from purified PCR products were digested by respective enzymes, ligated into pUC18 and transformed into E coli. A schematic scheme for the cloning strategy and the optimised conditions are shown in Figure lB. Plasmid DNA was isolated from positive clones digested and analysed for the presence of the correct insert (Fig 2B). Candidate clones were picked and prepared for sequencing (Thermo sequenase dye terminator cycle sequencing system, Amersham Kit No. 204279). Sequence data analysis and homology searches were performed using the FAST A program and the BLAST network service (National Center for Biotechnology Information). 104 Complementation experiments To investigate the occurrence of functional homologues or suppressors in Z. rouxii, fpsl S. mutants were transformed with a genomic library constructed in the autonomously replicating yeast plasmid pKU24 using partially digested DNA from Z. rouxii NRRL 2547 (Ushio et al., 1996). Transformed cells were grown on selective media (YNB-glucose without leucine) for about 5 days until a colony size of about 1-2 mm. About 300,000 transformants were replica- plated on 5% NaCl YEPD plates. After 3 days, salt stressed cells were replica-plated on plates without NaCl (osmotic downshock) and monitored for survival following an osmotic downshock. Fast recovering transformants (compared to the control cells transformed with empty plasmid) were selected and rescreened using spot assay. For these assays, 5 III of a 10- fold dilution of cell suspension (OD6oo= 1) was spotted on defined media and monitored for 3 days. Each promising isolate was compared with the wild-type and the fpsi Li strain for growth during osmotic stress (hyper- and hypo-osmotic stress), and anaerobic conditions. Isolates were also analysed for the ability to release glycerol after a hypo-osmotic shock as described in detail in chapters 3 and 4. To confirm activity, plasmid loss experiments were carried out by growing the transformants in non selective media (YEPD) followed by several rounds of replica-plating on YEPD agar plates for single colonies. Yeast cells (colonies) that could not grow on selective media (i.e those that had lost the plasmid) were selected and monitored again for survival during osmotic downshock. Plasmids containing the putative complementing or suppressing activity were isolated from yeast cells and sub cloned into the shuttle vector YEplac195. The different subelones were transformed back into the fpsl á strain to confirm and localise activity (Fig. 3). Promising subelones were isolated and the corresponding inserts were sequenced for analysis. The DNA fragments were also labelled and used as a probe in Southern blots to confirm that indeed they originate from Z. rouxii. In addition to the fpsi Li mutant, complementation experiments were also performed using the tpsl áfpsl Li and the gpdi Li gpd2Li double mutants, which are unable to grow on glucose or on high osmolarity media (such as 3 M glycerol) respectively. The tpsiLi fpsi Li double deletion mutant was grown in YEPGalactose and transformed with the Z. rouxii library as mentioned above. Complementation in the transformed strains was analysed on YNB-glucose. For the complementation of the gpdi Li gpd2Li double mutant, cells were transformed with the Z. rouxii library and grown on selective media (YNB-glucose without leucine) until colonies were visible. About 20,000 transformants were replica-plated on the same selective plates containing either 2 or 3 M glycerol or 1M xylitol and monitored for survival for about 7 days. 105 Heterologous expression of the S. cerevisiaefpsl-Al ungated channel in Z. rouxii. A diploid strain MA121-3 of Z. rouxii, was transformed with the truncated construct (jpsI- Lil) lacking the N-terminal 13-230 amino acids that are essential for the closing and opening of the Fpslp channel (Tamás et al., 1999). The YEpFPS1 construct containing the FPSI gene on a 3.8 kb Sail/HindI II fragment (Van Aelst et al, 1991) and the construction ofYEpjpsl- ál have been described previously (Tamás et al.. 1999). The YEpjps1-Li1 construct was digested by Sail and Hindiii to release the truncated FPS1. The resulting fragment was then gel purified and ligated into the E. coli - Z. rouxii shuttle vector to create the construct pKU24fPs1-Li1. This construct was subsequently transformed into the Z. rouxii MA121-3 strain and its effects on osmotolerance monitored. 106 Table 1. Laboratory yeast strains used in this study Strain Genotype Reference S. cerevisiae W303-1A MAT a leu2-3/ll2 ura3-l trp I-I his3-ll/l5 Thomas and Rothstein, 1989 ade2-l canl-100 GAL SUe2 malO S. cerevisiae YMT 2 W303-1A jpslLJ.::HIS3 Tamás et al., 1999 S. cerevisiae YSH 6.161.-20 W303-1A gpdlLJ.::TRPl gpd2LJ.::URA3 Ansell et al., 1997 S. cerevisiae YSH 302 W303-1A tpsl,1::TRPljpslLJ.::LEU2 Luyten et al., 1995 S. cerevisiae YSH 7.86.-1A W303-1A mpklLJ.::TRPl jpslLJ.::LEU2 Tamás et al., 1999 S. cerevisiae Y743 1:1278b MATa /MATa ura3/ura3 Davis and Engebrecht, 1998 S. cerevisiae Y739 Ll278b MATa / MATa ura3/ura3 Davis and Engebrecht, 1998 dom34tJ.: :LEU2/ dom34,1: :LEU2 Z. rouxii MA121-3 MATa / MATa leu2/leu2 Ushio et al., 1996 107 Sacch FPSI SGAHrNPSIT LllNLVYRGFP LKKVPIYFAG QLlGAFTGAL ILFIWYKRVL QEAYSDliIiMN ES.VAGMFCV FPK.PYLSSG RQFFSEFLCG AMLQAGTFAL TDPYTCL.SS DVFPLMMFIL IFlINASMAY QTGTAMNLAR DLGP B. cinerea SGAHrNPAIs DlLWrYRGFP LRKVPHYVLA QILGAFIAAL ISFGLYQTNI VEY. GGTDLK TSD'lMGAFIT YPRYAWINAS TSFFTEFVGT AlLAVAVLAL GODMNAPPGA GMSAFILGLV ITVLSMAFGY NTGAALNPSR DLGP YFF4_YEAST SGGHINPAVT ISMAIFRKFP WKKVPVYIVA QIIGAYFGGA MAYGYFWSSI TEFEGGPHIR TTATGACLFT DPK.SYVTWR NAFFDEFlGA SILVGCLMAL LDDSNAPPGN GMTALIIGFL VAAIGMALGY QTSFTINPAR DLGP Sch. pombe SGGavNPAVT ISLAIFRKFP WYKVPIYIFF QIWGAFFGGA LAYGYHWSSI TEFEGGKDIR TPATGGCLYT NPK.PYVTWR NAFFDEFIGT AVLVGCLFAl LDDTNSPPTQ GMTAFIVGLL I.AAIGMALGY QTSFTrNPAR DLGP C albicans SGGNLNPAVT LTLVLAQAVP PIRGLnlMVA ~lAGMAAAG AASAMTPGPI A .•...•• FT NGLGGGA .•...••••• BKA RGVFLEAFGT CIL •• CLTVL MMAVEKSRAT IiMAPFVIGIS LFLGHLICVY YTGAGrNPAR SFGP Sacch.AQYl sGGALNPAvs LSLCLARAVS PTRCVVMWVS QlVAGMAAGG AASAMTPGEV L ••••... FA NSLGLGC .....•.... SRT RGLFLEMFGT AlL •. CLTVL MTAVEKRETN FMAALPIGIS LFIAHVALTA YTGTGVNPAR BLGA Consensus SGghlNPart 1.1. .. r.fp .. kvp.y.v. Qiiga •• aa .. a •••.... i .e •••...... S ••• G ..... p ••... S •• r.fF.Ef.gt aiL •. cl.aL .d .......••• m.al.igi. .f .... a .. y .Tg ... NpaR dlGp Figure lA. Partial alignment ofFpslp and other fungal/yeast members of the MIP family The two most conserved regions (consensus motifs NPA about 400 bp apart) used for designing primers are represented in bold. Two pairs of degenerate primers were designed. Pair C was designed from the Fpslp amino acid sequences LNPSIT (sense) 5' and NLARDL (antisense). Pair D was designed from the consensus motifs of the entire MIP family (park and Saier, 1996); LNPAVT (sense) and NPARSF (antisense). The aligned proteins and their accession numbers were, S. cerevisiae Fpslp (P23900), Botrytis cinerea putative glycerol facilitator (ALI12633), YFF4_ Yeast, the S. cerevisiae putative glycerol facilitator Yfl054p (P43549), S. pombe putative glycerol facilitator (SPAC977.17), Candida albicans putative aquaporin (contig4-2389), and S. cerevisiae aquaporin (AAC69713). 107 108 Growth and induction of gene expression (e.g. osmotic shock) Extracted totail RNA mRNA 1/ One tube RT-PCR (50°C 30min, 94°C 2min, 94°C Imin, 46°C Imin , 68°C 3min, 2.5~ primers, lmM dNTPS, 3mM MgCh" 5mM DTT) l Separated an aliquot ofPCR products on an agarose gel Harvested the band(sl of expected size -400bp for MIP Purified the DNA from agarose (Qiagen protocol) l Reamplified individually the purified product (band) as with the initial PCR Cleaned the PCR product (phenol Chloroform extraction) l Digested the purified PCR product and the cloning vector with restriction enzymes. Also purified the digest and concentrated the sample to a small volume (about 101l1). For T- overhang vectors On1ligated Ligation into pUC18 or any other sequencing vector, Mixed insert DNA with vector in different ratios e.g. 3: 1 or 1:1. Concentrated ligation mixture to -IOIlL Transformed compelt 100 IIIE coli cells with the ligation mixture, included controls and plated on LB plates with Ampicillin, IPTG and X-Gal. Analysis of recombinant clones, isolate plasmid DNA and digest with appropriate restriction enzymes. Analyse the inserts on agarose gel J Sequencing of the promising inserts, sequence comparisons with known family members and homology searches in public genome data banks Figure lB. Optimised procedure used for isolating MIP family genes following the degenerate Reverse Transcriptase Polymerase chain reaction (RT-PCR). 109 B A 1 2 3 4 5 6 400bp Figure 2. The RT- PCR and cloning of the amplified products. Reactions were performed as described in materials and methods with different concentrations of primer pair D (panel A); Lanes: 1,0.5 J.1M;2, 1.5 J.1M; 3, 2.5 J.1M;4, 5.0· J.1M with Z. rouxii RNA and lanes 5, 2.5 J.1M (S. cerevisiae RNA) 6, 2.5 J.1M (E. coli RNA). Cloned PCR products (35$)-450 bp)' were analysed by restriction digestion and gel electrophoresis (panel B;' lanes 1-8). Insert sizes were compared with the FPSl control (lane 9) and EcoRl +HindIII lambda markers (M). 110 + + + + Figure 3. Restriction map, subcloning and functional analysis of a 4.6 kb insert in plasmid pZrcomp9. Plasmid DNA was digested with different restriction enzymes; HindIII (H), BamHI (B),XbaI (X), SphI (S). The digested DNA fragments were purified, cloned into the yeast vector YEplacI95 and transformed into the fpsl á strain. Subelones that could still complement the downs hock senstivity are indicated with a plus sign (+) while those that lost activity are indicated with a minus sign (-). Not drawn according to scale. 111 RESULTS AND DISCUSSION Analysis of FPSI homologues by peR and DNA probes Although glycerol accumulation in Z. rouxii occurs by an osmotically active transport system (Van Zyl et ai., 1990), the pattern and kinetics of glycerol release upon an osmotic downshock (Fig. 4) suggest the presence of a channel-mediated transport system similar to that of S. cerevisiae Fps1p (Chapter 3). However, when Z. rouxii was transformed with the construct pKU24.fPsl-Lil encoding constitutively open glycerol channel from S. cerevisiae, the ability to grow on high salt concentration was only slightly affected unlike S. cerevisiae where the same construct severely affected growth on media containing 1 M NaCI (data not shown). We therefore explored the occurrence of FPSI homologues in osmotolerant yeasts using DNA and PCR probes. Digested genornic DNA from various yeasts was separated by agarose gel electrophoresis, blotted on nylon membranes and probed with either the entire FPSI reading frame or a 426 bp fragment corresponding to the most conserved regions in the MIJ> family (Fig. lA). Under stringent conditions, hybridisation was only observed in S. cerevisiae. However, with less stringent conditions, hybridisation also occurred in the three Z. rouxii strains investigated (Fig. SA). Two bands were observed in BamHJJHindfIl digests of Z. rouxii strains, the strongest of which was approximately 6.1 kb. More bands could be observed under non-stringent conditions when the entire FPSI reading frame was used as a probe. Some degree of DNA to DNA cross-hybridisation has been observed among bacterial MIJ> family genes (Calamita et aI., 1995). Lack of strong hybridisation to the FPSI probes under stringent conditions might suggest a low nucleotide sequence similarity with genes from other yeasts. This is consistent with the observation that MIP family genes generally share low nucleotide sequence similarity and some members display less than 20% identity. The S. cerevisiae Fps 1p appears to be the most divergent .MIJ> family protein so far known (Hohmann et ai., 2000). For example, neither of the family's signature NPA boxes is fully preserved, being NPS and NLA, respectively (Fig. lA). Fps1p also differs from other MIJ> family members by having long amino- and carboxy-terminal hydrophilic extensions. The protein has 669 amino acids, compared with 250-300 amino acids for most of the family members (Van Aelst et ai., 1991). 112 The occurrence of FPSI homologues in various yeasts was further investigated by peR using specific primers (pair B) directed against the end terminals and conserved portion of FPSI (pair A). Apart from S. cerevisiae, no amplification was observed from other strains. Amplification was, however, achieved with degenerate primers corresponding to FPSI amino acid sequences LNPSIT (sense) and NLARDL (antisense). Since several fragments ranging from 300 to 600 bp were obtained, Southern blots were prepared and probed with the FPSI gene (426 bp fragment) to find out which of these peR products was homologous to FPSI. Under stringent conditions, only a single band of about 400 bp from S. cerevisiae hybridised (not shown). However, with less stringent conditions, strong hybridisation occurred with S. cerevisiae and with P. sorbitophila (Fig. 5B) and all hybridising fragments were between 400-500 bp. Weak signals were also observed for Z. rouxii (data not shown). Physiological and genetic analysis pointed to a possible existence of a glycerol facilitator in Z. rouxii that might be homologous to MIP channel proteins. These observations promoted us to clone the corresponding genees) using RT-peR (Fig. lB) and degenerate primers from the conserved portions of the MIP family (Fig. lA). Successful peR amplification was observed in S. cerevisiae and E. coli as well as in Z. rouxii (Fig. 2). Products within the expected size range of 350-450 bp were cloned for further analysis. Sequence analysis of the cloned fragments from S. cerevisiae and E. coli yielded sequences corresponding to previously reported members of the MIP family (park and Saier, 1996). However, analysis of 32 inserts from Z. rouxii revealed sequences that were neither related to FPSI nor to others members of the MIP family. 113 100 _._--~ .....- • .(!l)..._ •-.-::;:•:.-r:><:---»- .-.------ - -_-g--O- ----. -0 O+---.---.---.---.---.---~~~~ -5 o 5 10 15 20 25 30 35 Time after hypo-osmotic shock (min) _._ Hypo-osmotic shock at 280e ---0 -- Hypo-osmotic shock at oOe -e- -- Isotonic conditions at 280e --0- -- Isotonic conditions at osc Figure 4. Intracellular glycerol content after a hypo-osmotic shock (0.86 M Nael to 0.09 M Nael) in Z. rouxii NRRL Y2547 grown in glucose- YNB. 114 Ml 2 345678 kb 21.2 ~ 5.0 5.0 2.0 .' 0.5 Figure 5A. Southern blot analysis: The FPSl (426 bp) probe hybridised to restriction digest (BamHl /Hindl1l) of genomic DNA from (1) Z rouxii NRRL Y998, (2) P. sorbitophila CBS7064, (3) Z rouxii NRRL Y2547, (4) S. pombe CBS 5682, (5) Z rouxii NRRL Y225, (6) D. hansenii 0767, (8) S. cerevisiae W303-lA under medium stringent conditions: Hybridisation at 55°C overnight, followed by 2 X 10min wash in 2 x SSC , 0.1% SDS at RT. The probes were generated by PCR using primers; oligoG5 5' ATCTCAGGTGCTCATTTG 3' and oligoG6 5' TGGGCCCAGATCACGAGC 3' which corresponds to the Fpslp amino acid sequences, ISGAHL (sense) and ARDLGP (antisense) respectively. 115 400bp Figure SB. Degenerate PCR products were probed with a probe from the most conserved fragment of FPSI. Strong hybridisation occurred on S. cerevisiae (1) as well as on P. sorbitophila (2). fpsl á fpsl á with pZrcomp9 Wild type Figure 6. Complementation analysis:_fpsIA mutants were transformed with plasmids containing Z rouxii genes (genomic library) and monitored for survival following a hypo- osmotic shock. Promising colonies were further re-screened using a spot assay. 116 Complementation analysis Since no FPSI homologues were found using thePCR method, an alternative approach. Was tried. S. cerevisiae mutants lacking Fps1p do not show a growth defect on glycerol as a sole carbon source but show defects related to glycerol export and glycolysis that could be utilised in investigating the occurrence of functional homologues or suppressors by complementation analysis. Transforming the tpsl A fpsl jj double mutant with a multcopy plasmid carrying the FPSI gene or the E. coli glycerol facilitator glpF, has been shown to restore growth on fermentable carbon sources (Luyten et al., 1995). Unfortunately, when this double mutant was transformed with the Z. rouxii library, nearly 25% of the transformed cells grew on glucose including a control without insert. The use of the mpkl áfpsl jj mutant, which can not grow in media without osmotic stabilisers (Tamás et al., 1999), was also unsuccessful. The growth of the osmosenstive gpdl jj gpdë á double mutant is known to be improved by glycerol transporters during hyper-osmotic stress (Karlgren et al., 2000) and was thus used for complementation analysis. Out of the 20,000 transformants screened, 14 colonies survived at 2 M glycerol but they grew much slower compared to the control experiments even after 10 days (data not shown) and were therefore not considered for further analysis. The most clear-cut_wsljj phenotype so far observed is that mutants survive poorly an osmotic downs hock (hypo-osmotic shock) due to inability to dispose of accumulated glycerol. This defect is partially suppressed by over-expression of plant and bacterial glycerol facilitators (Luyten et al., 1995; Weig et al., 2000). From about 300,000_wsljj colonies transformed with a Z. rouxii library, 9 transformants grew almost like the wild-type in 24 hours of hypo- osmotic stress; the rest of the transformants grew much slower. Promising transformants were re-screened using a spot assay (Fig. 6). Complementing plasmids were isolated and the inserts analysed by restriction digestion mapping (Fig. 3) and sequencing. Sequence analysis of the most strongly complementing fragments revealed genes that play a role in yeast cytoskeleton, cell cycle or the cell integrity pathway, that are unrelated to the MIP family. Of the most promising plasmids, pZrcomp9 contained an insert encoding a gene homologous to the S. cerevisiae CDCIO gene encoding a septin (Fig. 7A). Septins belong to a family of conserved proteins that has been implicated in a variety of cellular functions and in changes in the cell shape. The biochemistry and localisation of septins suggest that they form a novel cytoskeletal system or that they function as scaffolds for the assembly of signalling complexes (for review Flescher et al., 1993; Longtine et al., 1996; Field and Kellogg, 1999). 117 Septin mutations affect the yeast cytoskeleton, budding, morphogenesis and consequently cell integrity (Cid et al., 1998). The mechanism by which the Z. rouxii CDCIO improved growth of the fpsI ~ mutant during hypo-osmotic stress is still unclear as it does not appear to improve glycerol release (Fig. 8). However, there appears to be a link between cell cycle, osmoregulation, cell integrity and the external environment (Chowdhury et a/., 1992; Brewster and Gustin, 1994; Shiozaki and Russell, 1995; Degols, et a/., 1996; Schoch et al., 1997; Fillinger, et al., 2000). The yeast DOM34 is required for osmotolerance While screening for genes from the Z. rouxii library that could alleviate the growth defects of the S. cerevisiae fpsl á mutant in hypo-osmotic media, we isolated plasmid pZrcomp9A that contained an insert with a full ORF homologous to DOM34 (Fig. 7B). The S. cerevisiae DOM34 gene (Duplication Of Multilocus region) is similar to genes found in diverse eukaryotes and archaebacteria. Analysis of dom34 mutants showed that they are defective in multiple development pathways such as the failure to undergo sporulation, exhibit a G 1 delay and fail to correctly execute pseudohyphal development (Davis and Engebrecht, 1998). We investigated whether the yeast DOM34 gene could also be relevant in osmotolerance. Preliminary experiments indicated that the S. cerevisiae dom34 mutants are sensitive to osmotic stress and that the Z. rouxii DOM34 complements this growth defect (Fig.9). CONCLUSION InS. cerevisiae, glycerol efflux is regulated and mediated by the MIP family glycerol channel Fps1p. Osmotolerant yeasts such as Z. rouxii appear to have a similar protein-mediated glycerol release mechanism under hypo-osmotic stress. Whether the protein(s) mediating the rapid glycerol efflux in Z. rouxii is homologous to Fps1 p in structure and function required further investigation. Recent studies have indicated that most of the genes involved in glycerol metabolism are very well conserved among yeasts. For instance, the Z. rouxii GPDI (AJ251481), DAKI (AJ294719) and HOG I (AB012146) are highly homologous to those from S. cerevisiae and S. pombe. Furthermore, DNA to DNA cross-hybridisation among glycerol channel encoding genes has been observed in bacteria so we theoretically anticipated cross 118 hybridisation between FPSi and its homologues from other yeasts especially Z. rouxii whose G+C content is closely related to that of S. cerevisiae (Kreger- van Rij, 1984). However, lack of strong hybridisation signals with FPSi probes suggests that the putative glycerol channel-encoding gene may be divergent from S. cerevisiae FPSI. It is also possible that Z. rouxii releases osmolytes via other mechanisms as observed in bacteria where osmolyte export after hypo-osmotic shock is mediated by mechanosensitive channels. Thus, cloning and identification of the putative glycerol facilitating proteins was attempted to obtain additional information on the molecular mechanisms of glycerol transport in yeast. A versatile homology -based cloning procedure was developed but efforts to isolate glycerol facilitators from Z. rouxii and P. sorbitophila were not successful most likely due to structural diversity amongst fungal glycerol channel encoding genes. Possibly, the use of short oligonucleotide probes or antibodies, although laborious, might provide alternative strategies in identifying glycerol facilitators in other yeasts. A large number of eukaryotic secondary transport systems have been cloned by complementation of yeast mutants defective in the corresponding transporter (Kranz and Holm, 1990; Anderson et al., 1992; Sentenac et al., 1992; Fromer et al., 1993). The use of complementation strategies to isolate functional homologues yielded genes that are unrelated to FPSi and the mechanism by which these genes (homologous to cell division control genes) improved growth of the fpsi LI strain during hypo-osmotic stress is still unclear. Overall, the study signifies the complexity of yeast osmoadapation and suggests that FPSi might be involved in other physiological process in addition to glycerol export. 119 A S. cerevisiae COCIO ALKRLTEIANVIPVIGKSDTLTLDERTEFRELIQNEFEKYNFKIYPYDSEELTDEELELN Z. rouxii Put COCIO ALKKLTEIANVIPVIAKADTLTLEERAQFREIIQQEFKKHKFRIYPYDTDELTEEELELN C. a.lbicans COCIO ALKKLSEIANVVPIIAKSDSLTLDERSEFKKLLQSEFMKYNFNIYPYDSEDLYEEERQLN S. pombe SPN2 VLKRLTEVVNVVPVIAKSDSLTLEERAAFKQQIREEFVKHDINLYPYDSDDADEEEINLN **:*:*: .**:*:*.*:*:***:**: *:: ::.** *:.:. :****::: :** :** RSVRSIIPFAVVGSENEIEINGETFRGRKTRWSAINVEDIN-QCDFVYLREFLIRTHLQD ESIRSIVPFAVVGSEREIEVNGETFRGRKTRWGAVNVEDIN-QCEFVYLREFLIRTHLED EDIKSLIPFAIAGSETEIEINGEMVRGRKTKWGAINIEDVS-QCEFVFLRDFLTRTHLQD AAVRNLIPFAVVGSEKAIIVDGRPIRGRQNRWGVVNVDDEKPLRVCFFFVTFLMRTHLQD ::.::***:.*** * ::* .. ***:.:* .. :*::* ** ****:* LIETTSYIHYEGFRARQLIALKENANS--RSS--- __AHMSSN_AIQR _ LIETTSYIHYEGFRARQLIALKENASS--RSS-----AGPANGGAYQPLIRLEWIGTLAL LIETTALTHYETFRSKQLIALKENASNPNRQSQLQKDQGQTSQQSNQDLKNASGVPNAPM LIETTSYYHYEKFRFKQLSSLKEQSSLATRMG-----SPAPVYPSEPHLHTATAQ _ *****: *** ** :** :***::. * VLF-------- FQSTTGTAAAR B S.cerevisiae Dom34 AWYGEKEVVKAAEYGAISYLLLTDKVLHSDNIAQREEYLKLMDSVESNGGKALVLSTLHS Z.rouxii putDom34 AWYGEAEVMKAVDLGAVNTLLITDTLMRSDDIDQRKRFLELAQQVERLGGKVAVF _ Dros~hila Pelota AFYGKKHVLQAAESQAIETLLISDNLFRCQDVSLRKEYVNLVESIRDAGGEVKIFSSMHI *:**: .*: :*.: *:. **:: *. : : : . : : : *: . : : : * **:. :: Figure. 7. Partial sequence alignment of the putative Z. rouxii CDC 10 with its closest fungal homologues; S. cerevisiae CDCIO (P25342), Candida albicans CDCIO (P39827), S. pombe sept in homologue SPN2 (Q09116) (Panel A). Partial alignment of the putative Z. rouxii DOM34 with the S. cerevisiae DOM34 (P33309) and the Drosophila melanogaster PELOTA protein (P48612) (panel B). 120 140 ,----------------, 120 ()- - - - -d'l 100 '.-: - - - - - - - - - - -..(l- - - - - - - - - - - - - - --0 80 60 40 20 o +---~--~----~--.---~--~--~~ -5 o 5 10 15 20 25 30 Time after hypo-osmotic shock (min) Figure 8. Glycerol release from the S. cerevisiae fpsl L1mutant transformed with pKU24 only (open cycles) or with pZrcomp9 (filled cycles) when the yeast cells were subjected to hypo-osmotic stress from 0.86 M NaCl (0.972 8w) to 0.09 M NaCl (0.996 8w). 121 12 10 - ~ ....- ~ 8 - ,ti0 ~0\0 6 - / Jl'"§ / 4 - 2 - _-A~_.-/.-- .' 0 I I I I 0 5 10 15 20 Time (hours) Figure 9A. Growth of the S. cerevisiae wild-type at 300e in YEPD media without osmoticum (circles) or with 1M sorbitol (squares). 122 8 oo 6 1.0 § 4 2 o 5 10 15 20 Time (hours) ____ YEPD ___ 1.0 M sorbitol ___._ 1.5 M sorbitol ~ 0.86M NaCI o Figure 9B. Growth of the S. cerevisiae dom34 mutant at 30 C in YEPD media with different osmotica. Control without osmoticum (circles), 1M sorbitol (squares), 1.5 M sorbitol (triangles) and 0.86 M NaCI (inverted triangles). 123 5 4 - 0 3 - 0 Y• \0 0 I) 0 2 - 1 1 - .. --~L~~.... • - I' I I I I 0 5 10 15 20 25 30 Time (hours) --e- Wild-type + YEplac 195 ____ dom34 + Zr DOM34 -A- dom34+ YEplac195 Figure 9C. Z. rouxii DOM34 complements the osmosenstive phenotype of the S. cerevisiae dom34 mutant. The S. cerevisiae wild type and the dom34 mutant strains were transformed with either the YEplac 195 containing Zr DOM34 or with the empty vector. Growth of the transformed strains was then monitored in YNB-glucose containing 1M sorbitol. 124 REFERENCES Anderson, J. A., Huprikar, S. S., Kochian, L. v., Lucas, W. L, and Gaber, R F. (1992). Functional expression of a probable Arabidopsis thaliana potassium channel in Saccharomyces cerevisiae Proc. Natl. Acad Sci USA. 89, 3736- 3740. Ansell, R, Granath, K., Hohmann, S., Thevelein, J., and Adler, L. (1997). The two isoenzymes for yeast NAD-dependent glycerol3-phosphate dehydrogenase encoded by GPDJ and GPD2, have distinct roles in osmoadaption and redox regulation. EMBO J 16,2179-2187. Attfield, V. P. (1998). Physiological and molecular aspects of hyper osmotic stress tolerance in yeasts. In Recent Research Developments in Microbiology Vol. 2 Part Il, S.G. Pandala, ed. (Research Signpost, Trivandrum, India), pp.427-440. Brewster, J.L., and Gustin, M.C. (1994). Positioning of cell growth and division after osmotic stress requires a MAP kinase pathway. Yeast 10, 425-439. Brown, A. D. (1990). Microbial water stress physiology. Chichester: John Wiley and Sons. Calamita, G., Bishai, W.R., Preson, G.M., Guggino, W.B., and Agre, P. (1995). Molecular cloning and characterization of AqpZ, a water channel from Escherichia coli. J Bioi. Chern. 270, 29063-29066. Chowdhury, S., Smith, K.W., and Gustin, C.M. (1992). Osmotic stress and the yeast cytoskeleton. Phenotype-specific suppression of an actin mutation. J Cell Biol. 118, 561-571. Cid, r.v., Adamiková, L., Cenamor, R, Molina, M., Sánchez, M., Nombela, C. (1998). Cell integrity and morphogenesis in budding yeast septin mutant. Microbiology 144, 3463-3474. Davis, L., and Engebrecht, J. (1998). Yeast dom34 mutants are defective in multiple development pathways and exhibit decreased levels of polyribosomes. Genetics 149, 45-46. 125 Degols, G., Shiozaki, K., and Russel, P. (1996). Activation and regulation of the spc1 stress-activated ptrotein kinase in Schizosaccharomyes pombe. Mol. Cell. Bioi.' 16, 2870-2877. Del Sal, G., Manfioletti, G., and Schneider.C. (1988). A one-tube plasmid DNA minipreparation suitable for sequencing. Nucl. Acids Res. 16,9878. Edgley, M. and Brown, A.D. (1978). Response of the xerotolerant and non-tolerant yeasts to water stress. J. Gen. Microbiol. 104,343-345. Field, M.e., and Kellogg, D. (1999). Septins: cytoskeletal polymers or signalling GTPases. Trends Cell Bioi. 9, 387-394. Fillinger, S., Ruijter, G., Tamas, J., M., Visser, J., Thevelein, M., J., and d'Enfert, C. (2000). Inactivation of the gfdA gene encoding a NAD-dependent glycerol 3- phosophate dehydrogenase reveals a role for glycerol 3-phosphate in hyphal morphogenesis in the filamentous fungus Aspergillus nidulans. Submitted. Flescher, G.E., Madden, K., and Snyder, M. (1993). Components rquired for cytokinesis are important for bud site selection in yeast. J. Cell. BioI. 122, 373-386 Frommer, W.B., Hummel, S., and Riesmeier, J. W. (1993). Expression cloning in yeast of a cDNA encoding a broad specificity amino acid permease from Arabidopsis thaliana. Proc. Natl., Sci. USA. 90, 5944-5948. Hohmann, S., Bill, R., Kayingo, G., and Prior, B. A. (2000). Microbial MIP channels. Trends Microbiol. 8,33-38. Karigren, S., Bill, R., and Hohmann, S. (2000). A test system to identify solutes transported by MIP channels. Proceedings of the 3rd International Conference on the Molecular Biology and Physiology of Water and Solute Transport. P104, 144. Kranz, E. J., and Holm, C. (1990). Cloning by function: An alternative approach for identifying yeast homo logs of genes from other organisms. Proc. Natl. Acad Sci. USA. 87,6629-6633. 126 Kreger-van Rij, N. 1. W. (1984). Principles of classification: The species. In The Yeasts, A taxonomic Study, N.1.W. Kreger-van Rij,. Ed. (Elsevier Science Pub. Amsterdam), pp. 15-16. Lages, F., Silva-Graca, M., and Lucas, C. (1999). Active glycerol uptake is a mechanism underlying halotolerance in yeasts: a study of 42 species. Microbiology 145, 2577-2585. Longtine, M.S. (1996). The septins: role in cytokinesis and other processes. Curro Opin. Cell. BioI. 8,106-119. Luyten, K, Albertyn, 1., Skibbe, W. F., Prior, B. A, Ramos, 1., Thevelein, 1. M., and Hohmann, S. (1995). Fps1, a yeast member of the MIP family of channel proteins, is a facilitator for glycerol uptake and efflux and it is inactive under osmotic stress. EMBO J. 14, 1360-137l. Maniatis, T., Fritsch, E.F., and Sambrook, 1. (1982). Molecular cloning: A laboratory manual. Cold Spring, Harbor Laboratory, Cold Spring Harbor, New York. Park, 1.H., and Saier, M.H.Jr. (1996). Phylogenetic characterization of the MIP family of transmembrane channel proteins. J. Membrane. Bioi. 153,171-180. Philips, 1., and Herskowitz, I. (1997). Osmotic balance regulates cell fusion during mating in Saccharomyces cerevisiae J. Cell. Bioi. 138,961-974. Prior, B.A, and Hohmann, S. (1997). Glycerol production and osmoregulation. In: Yeast sugar metabolism, F. K Zimmermann and KD. Entian, eds. (Technomic Puplishing company Inc., Lancaster, PA), pp 313-335. Sambrook, 1., Fritisch, E.F. and Maniatis, T. (1989). Molecular cloning. A laboratory manual. Second edition. Cold spring Harbor Laboratory Press, Cold Spring Harbor, New York. Schoch, C. L., Bruning, A. R, E., Entian, K.-D., Pretorius, G. H, 1., and Prior, B. A. (1997). A Saccharomyces cerevisiae mutant defective in the kinesin-like protein Kar3 is sensitive to NaCI-stress. Curro Genet. 32,315-322. 127 Sentenac, H., Bonneaud, N., Minet, M., lacroute, F., Salmon, J. M., Gaymard, F., and Grignon, C. (1992). Cloning and expression in yeast of a plant potassium ion transport system. Science 256, 663-665. Sherman, F., Fink, G.R., and Hicks, lB. (1986). Laboratory course manual for methods in yeast genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Shiozaki, K, and Russell, P. (1996). Conjugation, meiosis and osmotic stress response are regulated by Spel kinase through Atfl transcription factor in fission yeast. Genes and Development 10, 2276-2288. Sutherland, F. C. W., Lages, F., Lucas, C., Luyten, K,Albertyn, l, Hohmann, S., Prior, B. A., and Kilian, S. G. (1997). Characteristics ofFPS1-dependent and -independent glycerol transport in Saccharomyces cerevisiae. J Bacteriol. 179, 7790-7795. Tamás, M. r,Luyten, K, Sutherland, F. C. W., Hernandez, A., Albertyn, r.Valadi, H., Li, H., Prior, B. A., Kilian, S. G., Ramos, r, Gustafsson, L., Thevelein, J. M., and Hohmann, S. (1999). Fps1p controls the accumulation and release of the compatible solute glycerol in yeast osmoregulation. Mol. Microbiol. 31, 1087-1104. Thomas, B. J. and Rothstein, R. J. (1989). Elevated recombination rates in transcriptionally active DNA. Cel! 56, 619-630. Ushio, K, Otsuka, H., Hoshikawa, S., Taguchi, G., Shimosaka, M., Mitsui, N., and Okazaki, M. (1996):Cloning of the SAT] gene concerned with salt tolerance of the yeast Zygosaccharomyces rouxii. J Ferment. Bioeng. 82, 16-21. Ushio, K, Tatsumi, H., Araki, H., Toh, A., and Oshima, Y. (1988). Construction of a host-vector system in the osmophilic haploid yeast Zygosaccharomyces rouxii. J Ferment. Techno!. 66,481-488. 128 Van Aelst, L., Hohmann, S., Zimmermann, F.K., Jans, A.W., Thevelein, JM. (1991). A yeast homologue of the bovine lens fibre MIP gene family complements the growth defect of a Saccharomyces cerevisiae mutant on fermentable sugars but not its defect in glucose-induced RAS-mediated cAMP signalling. EMBO 1. 10,2095-2104. Van Zyl, P. l, Kilian, S. G., and Prior, B. A. (1990). The role of an active transport mechanism in glycerol accumulation during osmoregulation by Zygosaccharomyces rouxii. Appl. Microbiol. Biotechnol. 34,231-235. Van Zyl, PJ., and Prior, B.A. (1990). Water relations of polyoI accumulation by Zygosaccharomyces rouxii in continous culture. Appl. Microbiol. Biotechnol. 33, 12-17. Weig, A. R., and Jakob, C. (2000). Functional identification of the glycerol permease activity of Arabidopsis thaliana NLMI and NLM2 proteins by heterologous expression in Saccharomyces cerevisiae. FEBS Lett. 481, 293-298. Yancey, P. H., Clark, M. E., Hand, S. C., Bowlus, R. D., and Somero, G. N. (1982). Living with water stress: evolution of osmolyte systems. Science 217, 1214-1222. 129 CHAPTER6 Isolation and Characterisation of the TIMIO Homologue from the Yeast Pichia sorbitophila: A putative component of the mitochondrial protein import system. Abstract The Saccharomyces cerevisiae TIMiO gene encodes one of the few essential mitochondrial proteins that are required for the import of nuclear-encoded precursor proteins from the cytosol and their subsequent sorting into the different mitochondrial compartments. While searching for FPSI homologues, we isolated and characterised a putative homologue of TIMiO from the halotolerant yeast Pichia sorbitophila. The Pichia TIMiO gene encodes a protein of 90 amino acids with 66% identity to S. cerevisiae Tim lOp. It was capable of suppressing the temperature sensitivity of timlil-I mutant in S. cerevisiae suggesting that Pichia TIMIO is both a functional and structural homologue of S. cerevisiae TIMIO. The putative Pichia TIMIO gene product contains all the four conserved cysteine residues and the two CX3C motifs typical of the Tim family proteins in the mitochondrial intermembrane space. Using anti-TimlOp serum, Western blots detected a protein of about lO-kDa and suggested that the Pichia Tim lOp is a mitochondrial protein. The results suggest that mitochondrial import and sorting systems might be also strongly conserved in other fungi. The coding sequence of the P. sorbitophila TIMIO has been deposited in the EMBL Nucleotide Sequence Database under Accession Number AJ243940 130 INTRODUCTION The uptake of nuclear-encoded precursor proteins into the mitochondrion and their subsequent sorting into the different mitochondrial compartments are mediated by translocases; many of which have recently been identified. These include !ranslocases in the .Quter !!!embrane-TOM system and corresponding proteins in the inner membrane-the TIM system (for review, Rassow et al., 1999). Translocase components in the intermembrane space include Tim9p, TimlOp and Tim12p which are essential for mitochondrial biogenesis and viability of yeast cells (Jarosch et al., 1996, 1997; Sirrenberg et al., 1998; Koehler et al., 1998; Adam et al., 1999). Precursor proteins carrying a cleavable amino-terminal targeting sequence are imported into the mitochondria by a general import pathway composed of cytosolic chaperons, outer membrane receptors and a general import channel (Hachiya et al., 1995; Hill et al., 1998). After crossing the outer membrane, these proteins are translocated into the matrix by the TIM17-23 complex in co-operation with Tim44p and the mitochondrial Hsp70p (Rassow et aI., 1999). However, many integral inner membrane proteins such as the metabolite carriers, are synthesised without the amino- terminal targeting signal and therefore translocated by a separate import machinery. This import machinery has recently been identified and shown to consist of the intermembrane space proteins Tim9p, Tim 1OpTim 12p, and the inner membrane proteins Tim22p and Tim54p (Koehler et al., 1998). Upon entering the intermembrane space, multispanning carrier proteins first bind to Tim91l 0 complex, are then handed over to Tim12p, and finally enter the inner membrane via Tim22p (Rassow et al., 1999). Despite their high sequence similarity, Tim9p, TimlOp and Tim12p are not functionally equivalent and neither can one substitute for the other suggesting a co-operative mode of action in mitochondrial preprotrein import. TimlOp and Tim12p were first discovered as multicopy suppressors of mitochondrial RNA splicing 'defects and were initially termed Mrs11p and Mrs5p (Waldherr et al., 1993; Jarosch et al., 1996, 1997). Tim9p has only been recently identified from a screen of other import. components that interact with Tim 1Op (Adam et al., 1999). Here, we report on the molecular cloning and functional characterisation of a gene from the halotolerant yeast Pichia sorbitophila that shares high sequence similarity to the Tim9-Tim10-Tim12 family of mitochondrial preprotein transporters. The cloned gene exhibits 66% amino acid identity to TimlOp and was capable of restoring growth of the TIMiO temperature-sensitive mutants of S. cerevisiae at 37°C. The putative Pichia TIMiO 131 gene product contains all the four typical cysteine residues and also shows similarity to zinc- finger proteins. MATERIALS AND METHODS Strains and growth conditions The following yeast strains were used: Pichia sorbitophila CBS 7064 (CSIR. Y170), Zygosaccharomyces rouxii NRRL Y2547, Schizosaccharomyces pombe CBS 5682 (CSIR. Y457) and Debaryomyces hansenii CBS 0767 (CSIR. Y953) were obtained from the Industrial Biotechnology Microbial Resource Centre, Department of Microbiology, University of Orange Free State (South Africa). The temperature-sensitive (timiO-i) mutant of S. cerevisiae (Koehler et al., 1998) and its isogenie wild type strain GA74-6A, (MATa leu2, ura3, his3, trpi, ade8) were kindly provided by Dr e.M. Koehler. All yeast strains were grown in the YEPD medium (1% yeast extract, 2% peptone, 2% glucose) or defined medium (2% glucose, 0.67% yeast nitrogen base) supplemented with the required amino acids. Escherichia coli cells (TOPIOF' InVitrogen) were routinely grown at 37° C overnight in LB media (1% NaCl, 1% bacto-tryptone, 0.5% yeast extract). Nucleic acid manipulation and analysis Standard procedures were carried out according to Sambrook et al. (1989) unless otherwise mentioned. Yeast genomic DNA was isolated as described by Sherman et al. (1986). The polymerase chain reaction was used to screen P. sorbitophila for the occurrence of membrane channel proteins homologous to the S. cerevisiae Fps1p. The following primers were .consequently designed from S. cerevisiae FPSI sequences. sense 5' ATCTCAGGTGCTCATTTG 3' antisense 5' TGGGCCCAGATCACGAGC 3' PCR amplifications were carried out (92°C, 2 min.; 52°C, 50 sec; 72°C, 1 min; 30 cycles) with 0.5~ specific primers, 2.5 units of Taq DNA polymerase, 200 ~ dNTPs, 1.5 mM MgClz (Boehringer Mannheim) with P. sorbitophila genomic DNA as a template. A PCR product of about 400 base pairs was purified (High pure TM PCR purification kit; Boehringer Mannheim) and then used to probe the P. sorbitophila (CBS 7064) genomic library in 132 YEplac352 (Hill et al., 1986). The insert DNA of one hybridising clone was digested with several restriction endonucleases. Relevant restriction fragments were purified using the Qiagen kits and then subcloned into the sequencing vector pUC18. Both strands of each subclone were sequenced with an automated sequencer using the thermo sequenase dye teminator cycle sequencing kit (Amersham). Sequence data analysis and homology searches were performed using the FASTA program and the BLAST network service (National Center for Biotechnology Information). The 1.5 kb EcoRiiSalI fragment containing the entire putative TIMiO homologue was considered for further investigation. Southern blot hybridisation was carried out overnight at 68°C followed by stringent washes; 2 xlSmin, 0.1 x SSC, 0.1% w/v SDS at 68°C (room temperature for medium stringent washes) under constant agitation. Probe labelling and detection were performed using the digoxigenin (DIG) system (Boehringer Mannheim). Isolation of mitochondria and western blotting For preparation of mitochondria, P. sorbitophila and S. cerevisiae were grown to stationary phase (OD6oo= 3) in the medium containing per litre: 25 ml of 80% lactic acid, 3.0 g yeast extract, 0.5 g glucose, 0.5 g CaCI2. 2H20, 0.6 g MgCh.2H20, 1.0 g KH2P04, 1.0 g NfuCI, and 8.0 g NaOR. The final pH was adjusted to 5.5 with KOH (Daum et al., 1982). Homogeneous preparations of yeast mitochondria were isolated from spheroplasts by differential centrifugation as described (Glick and Pon, 1995). Protein concentrations were determined using the Bradford assay (Read and Northcote, 1981). Mitochondrial proteins were separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-P AGE) consisting of a 4% stacking gel and a 15% separating gel (Laemmli, 1970). Separated proteins were either stained by Coomassie Blue or transferred to the polyvinylidine difluoride .membrane by semidry electroblotting. Membranes were probed with a polyclonal antibody raised against S. cerevisiae TimlOp (kindly provided by Dr C. M. Koehler). Other procedures for immunoblotting and immunodetection were performed with the chemiluminescence western blotting kit (Boehringer Mannheim). Functional complementation analysis The timiO-i temperature sensitive mutant of S. cerevisiae was transformed with both single and multicopy plasmids harbouring the putative TIMiO homologue from P. sorbitophila. As a negative control, temperature-sensitive mutants and the wild-type strain GA74-6A were transformed with the empty vectors YEplac195, Yéplac181 and YCplac111 (Gietz and 133 Sugino, 1988). Yeast transformations were carried out using lithium acetate (Gietz and Schiestl, 1995). Functional complementation was analysed by comparing growth of transformants at 25, 30 and 37°C. For plate assays, 5 ).11 of a 10-fold dilution of cell suspension (OD6oo= 1) was spotted on defined media and monitored for 3 days. To compare growth rates in liquid media, transformants were cultivated in 50ml cultures and monitored spectrophotometrically for 3 days. All experiments were done in triplicate and the results of a representative experiment are shown RESULTS AND DISCUSSION Isolation and characterisation of the Pichia TIMlOgene During the analysis of membrane channel proteins homologous to the yeast glycerol facilitator Fpslp, a 4.6 kb DNA fragment was isolated from P. sorbitophila genomic library and digested into smaller fragments for subeloning and sequencing. The sequence of the 1.5 kb EcoRIlSaD fragment revealed an open reading-frame with homology to the Tim9-TimlO- Tim12 family of mitochondrial intermembrane space proteins. As shown in Figure 1, there is an uninterrupted open reading frame that codes for a protein of 90 amino acids with a predicted molecular weight of 9,935 kDa and a pI of 5.54. The gene has a GC content of 44% and a codon adaptation index (CAI) of 0.186. Similarly, the homologous genes TIMIO, TIMI2 and TIM9 from S. cerevisiae exhibit CAI of 0.24, 0.07 and 0.22 respectively which implies that codon usage is little biased and that these genes are poorly expressed (Bennetzen and Hall, 1981, Sharp et al., 1986). Southern blotting confirmed that, indeed, the cloned TIMiO gene is from P. sorbitophila by revealing strong signals of about 4 kb and 5.1 kb in a EcoRV and Hindill digestions respectively (Figure 2). Cross hybridisation was also observed .in other yeasts (S. cerevisiae, Z. rouxii, Sz. pombe and D. hansenii) under medium stringency hybridisation conditions (data not shown). Since phylogenetic analysis points to the Pichia homologue being most closely related to TimlOp (Figure 3) based on sequence identity and size, the gene has been tentatively designated as Pichia TIMiO. 134 Similarity to other proteins in the public databases; the Tim family of proteins in the mitochondrial intermembrane space Sequence comparisons and similarity searches revealed other proteins homologous to the Tim 1O/Tim12/Tim9 proteins from various eukaryotic organisms including a putative homologue (85% identity) from the preliminary sequence data of Candida albicans (Stanford 5476/ con4-3104). Although other mitochondrial proteins share sequence similarity with bacterial proteins as seen between Tim17p/ Tim22p/ Tim23p and the prokaryotic amino acid permease LivH, (Rossow et al., 1999), no bacterial homologue ofTimlOp was detected. Figure 3A shows the amino-acid sequence alignment of Pichia TimlOp with other Tim proteins detected from the public databases. The putative Pichia TIMiO gene product is closely related to TimlO proteins from S. cerevisiae (66% identity), Emericella nidulans (53%) and Ciona intestinalis (44%). It was closer to Tim12 than to Tim8, Tim9 or Tim13 proteins. Over all, Tim proteins are more divergent in their C and N terminals and may be as small as 80 to 120 amino acid residues in length. Alignment of amino acids indicated that all highly conserved regions In Tim proteins including the twin CX3C motif (Koehler et al., 1999) are present in the Pichia TimlOp (Figure 3A). Sirrenberg et al. (1998) reported that zinc binding is essential for the function of TimlOp and Tim12p in protein import. The four cysteine amino acid residues that are known to constitute a zinc-binding site in TimlOp, Tim12p and Tim9p (Adam et al., 1999) are also present in the Pichia homologue. This suggests that the putative Pichia Tim 1Opmay also be a zinc finger protein. Besides the highly conserved cysteine residues, other residues consistently conserved were recognized (Figure 3A). Whether these residues might have a specialised physiological role remains to be established . .Interestingly, these residues distinguish Timl O proteins from other Tim family proteins. For instance, a consensus signature sequence EXlONXsCX2KCx9LXJEX2CLDRCVXK. was observed in Tim 10 proteins but was not conserved in Tim9 proteins. Figure 3B shows a phylogenetic tree indicating the relationship between Pichia TimlOp and other selected homologues. It is apparent that Pichia TimlOp is more closely related to other fungal/yeast hoinologues than to Tim proteins from other taxa. Four subgroups were evident and the division presented in the tree concurs with the taxonomic relationships generally accepted. 135 Pichia 'fimlOp cross-reacts with S. cerevisiae TimlOp antiserum To investigate whether Pichia TimlOp might be a mitochondrial protein, Western blots were performed on mitochondrial extracts and probed with antiserum raised against the S. cerevisiae TimlOp. A peptide of approximately 10 kDa was detected from both S. cerevisiae and P. sorbitophila (Figure 4). This was consistent with the predicted molecular mass of TimlOp (9935 Da) and suggested that the Pichia TimlOp is located in the mitochondrion. Additional localisation studies are necessary to confirm this suggestion. In addition, the antibody also reacted with proteins of much higher molecular weight (70 kDa) albeit with a weaker signal in both organisms. Pichia TIM10 is a functional homologue of S. cerevisiae TIM10 Temperature-sensitive mutants (timiO-i) of S. cerevisiae cannot grow on any carbon source at 37°C (Koehler et aI., 1998). To investigate whether the gene product of P. sorbitophila TIMiO was a functional homologue of S. cerevisiae TimlOp, the timlo-I mutant strain was transformed with plasmids harbouring the Pichia TIMiO gene. As shown in Figure 5, the temperature-sensitive mutant transformed with YEp/PichiaTimiO grew like the wild type at the restrictive temperature (37°C) while those transformed with the empty vector could not grow. However, at the permissive temperature (25°C) both mutants with or without the Pichia gene grew at wild-type rates. These results demonstrated that the Pichia TIMiO functionally complements the S. cerevisiae timlit-I mutant enabling it to grow at 37°C. The ability of Pichia TimiO to restore growth at the restrictive temperature suggests a role in the import and sorting of mitochondrial carrier proteins which are defective in timiO-i strains. Despite their ability to synthesise some proteins, mitochondria have to import most of their protein precursor from the cytosol. The intermembrane space proteins Tim9, Timl0 and Tim12 are some of the few essential proteins mediating this import through interaction with other Tim and Tom proteins. Defects in the import machinery may affect the viability of yeast cells and cause many mitochondria-related diseases in humans (Baker and Schatz, 1991~ Jarosch et al., 1997~Koehler et al., 1999, Larsson and Clayton, 1995). For instance, depletion of Tim 1Op results in accumulation of Hsp60 precursors, loss of cytochromes and changes in mitochondrial morphology as growth stops. On the other hand, overproduction of TimlOp restores respiration ability in yeast with mitochondrial RNA splicing defects (Jarosch et al., 1997). 136 The role of Pichia TimlOp in suppressing mitochondrial RNA splicing defects was not tested. However, its structural and functional homology to S. cerevisiae TimlOp tempts one to speculate that it might also suppress mitochondrial RNA splicing defects. To conclude, we have shown that the yeast P. sorbitophila possesses a structural homologue of the TimlOp that complements the timlO-l defective gene in S. cerevisiae enabling the strain to grow at 37°C. Immunoblotting with the anti-Tim 1Op serum shows that the homologous protein is located in the mitochondria. Sequence analysis from public databases and Southern blotting further revealed the occurrence of Tim 1Op homologues in other organisms. Taken together, the results suggest that mitochondrial preprotein import and sorting systems might be strongly conserved throughout eukaryotic organisms. It is anticipated that a clear understanding of yeast homologues might pave the way in elucidating the mechanisms for mitochondrial preprotein import in higher eukaryotes. 137 1 TTATGCGCACAACTACGACGCCAAGAAAGAGGATTTCGGTGGTCCAGGCGATGAGGCCAA 61 TGAAGGCTCTCCCATTTTCGAAGCGGAGCTTCGTGCCTGCCACCACACAGCATTGTACTA 121 CAAGAGCGAGGAGCGCAACTACCTCTTTTTCATGGGTGGATTCTCAAACAGCTACCTCAG 181 GCTATTTGACCCCGAGCCCTACGTGTCAGACAAGCTTGACGTGTCGAAGTACGCCAAGTT 241 CAACCTCGCCCGCAGCAACAACCAGTCCAGGGTGTTGGCACTCAACCTCCGCACGCAGAA 301 GTGGTCGTTTTTCAGATACTTCCATGACTGCAGCGAAGCCGTATCAAAGAGCTTCACTCG 361 CAGGTTGCCGATCGACCGTCCGCTAGAGGACGTCGAATTTTGCAACTACGGCGGTGCCAT 421 ATCGCTCGACGGGAAGTGCATTAACATGTGCCACGGCCTCGCCTGCCCTGTTCCGGTTAA 481 TGCAGAGGAGTACGAAGCCCTCAAAACAGAGACTTCGGAATTGAGTTTCCTATGGGGCAT 541 TGTGACCCATTTCACCTTCCCTGGCCTATAATAGCCTATAATAGCCTACGATGGCCTGTG 601 ATGCTCGTGGTGGCCCACAACGGCCCATAAATTGCTGCATTTCGACACATTCTACCAGTT 661 CGCTCCTACTATACATATTACTTCTTAGATAGACTGGACAAACTGTTTCATAATTTTACG 721 TTACTTTTAATGGTACATACTACTATATAGACTTATCCAACTATCCCTCCGCTAGTGCTC 781 GCTTTCCAACACCTTTTTTTTCAGATTTCGACTTTGTGTTGCTTTGCTAAAATCGACAAC 841 TTCAAGAAAAAAATTGGCGAATCTTTCAAACTACAGTTACAAGAAGTAGAAGTATTTATA 901 AGATTGAAAATGTTTGGATTAGGAGGAGCACCACAGATATCGTCAGAGCAGAAATTGCAA M F G L G GAP Q ISS E Q K L Q 961 GCCGCTGAGGCGGAGTTGGATATGGTTACCGGAATGTTCAACCAGTTGGTTGATCAGTGC A A E A E L DMV T G M F N Q L V D Q C 1021 CACTCTAAGTGTATTAACAAGAGCTACGGTGATTCCGACATTACCAAGCAAGAAGCACTC H S K C INK S Y G D SDI T K Q E A L 1081 TGTTTGGACAGATGTGTTGCCAAGTATTTCGATACCAATGTTCAAGTTGGAGAACACATG C L D R C V A KYF D T N V Q v G E H M 1141 CAAAAGTTGGGACAATCTGGTCAGTTTATGGGGAGAAAATAAGCGTACATAGAATCTTCA Q K L G Q S G Q F M G R K 1201 GAGTAGCCGGAGACAGAAATCATATTTAAGGATGTATGATATGTAATCTATATTTCGAAA 1261 ATGCACACTTAGCGTGATAGAATACATGTATTTACAAGAGGATGGGCCAGGTGGCGTCTG 1321 TCGATGGGAGGCCGGGGGCAGGCATCAAGCAAGACCGACGCTGCGCATCAAGCGGAGCTT 1381 CACGCCCTCCCAGGAGTTGAGCTTCGGTCTTGCGTCTATGAAAAAGCTGGCGCCCTATCC 1441 GTTGCCCGTGTAGAAGTCGATGACGTACTCGACCTGTGTGCCGCAGCGGTCGAT Figure 1. Nucleotide sequence of the 1.5 kb EcoRlISall subclone and the deduced amino acid sequences of the P. sorbitophila TIMiO in single-letter code. 138 MKERHDEH kb 21.2 5.1. 4.2 3.5 2.0 l.4 0.8 0.6 Figure 2. Southern blot analysis. Genomic DNA extracted from P. sorbitophila was digested by EcoRV (ER), HindIII (HD) and EcoRV /HindIII (EH) separated on a 0.7 % agaraose gel electrophoresis and probed by the entire coding region of the Pichia TIMIO gene under stringent conditions. L}:;f P. sorbitophila TimlOp ........ MF GLGGA.PQIS SEQKLQAAEA ELDMVTGMFN QLVDQCHSKCINKS.YGDSD ITKQEALCLD RCVAKYFDTN VQVGEHMQKL .GQSGQFMGR K . 90 C. albicans TimlOp ........ MF GLGGTTPQIS SQQKLQAAEA ELDMVTGMFN ALVSQCHTKCINKS.YNEAD ISKQESLCLD RCVAKYFETN VQVGENMQKL .GQSGQFMGR R . 91 S. cerevisiae TimlOp MSFL GFGGGQPQLS SQQKIQAAEA ELDLVTDMFN KLVNNCYKKCINTS.YSEGE LNKNESSCLD RCVAKYFETN VQVGENMQKM .GQSFNAAGK F 93 E. nidulans TimlOp ........MS FLFGGAPKMS SEQKlAAAET EVEMITDMFN RLSESCSKKCIPND.YREGD LNKGESVCLD RCVGKFFEVN IKVSEKMQGV AGQQQGGAGL SL . 93 C. intestinalis TimlOp .................. MDP QEAQKLAAEL EVEMMADMYN RMTSSCHKKCISTR.YDTGD LEKGEAVCID RCVAKYLDIH EQIGKKLTEM SQTDEEAMSK MSQKPGYSCK 92 R. norvegicus TimlOp .......MDP LRAQQLAAEL EVEMMADMYN RMTSACHRKCVPPH. YKEAE LSKGESVCLD RCVSKYLDIH ERMGKKLTEL SMQDEELMKR VQQSSGPA .. 90 H. sapiens TimlOp .......MDP LRAQQLAAEL EVEMMADMnl RMTSACHRKCVPPH. YKEAE LSKGESVCLD RCVSKYLDIH ERMGKKLTEL SMQDEELMKR VQQSSGPA .. 90 D. melanogaster TimlOp MALPQISTAD QAKLQLMQEM EIEMMSDLYN RMTNACHKKCIPPR. YSESE LGKGEMVCID RCVAKYLDIH EKIGKKLTAM FMQDEELMKK MSS . 92 C. elegans TimlOp .......MAT DAQMAQVAEL EVEMMSDMYR RMTNSCQAKCIATA. FRESE LTKGEAvCLD RCVAKYLDVH EKLGKRLTSM SQGDEAALQK IAQQ . 86 S. cerevisiae Tim12p ......MSFF LNSLRGNQEV SQEKLDVAGV QFDAMCSTFN NILSTCLEKCIPHEGFGEPD LTKGEQCCID RCVAKMHYSN RLIGGFVQTR GFGPENQLRH YSRFVAKEIA 104 L. esculentum TimlOp .MAGVPSNLE REQIFSMAEK EMEYRVEMFN KLTHTCFKKCVENK.YKDSE LNMGENSCID RCVSKYWQVT NLVGTLLGNT RPM . 81 A. thaliana TimlOp MASPIPVGVT KEQAFSMAQT EMEYRVELFN KLAQTCFNKCVDKR. YKEAE LNMGENSCID RCVSKYWQVN GMVGQLLSAG KPPV...... . . 83 S. cerevisiae Tim9p .MDALNSKEQ QEFQKVVEQK QMKDFMRLYS NLVERCFTDCVND ..FTTSK LTNKEQTCIM KCSEKFLKHS ERVGQRFQEQ NAAL GQG LGR . 87 E. nidulans Tim9p .MDGLNAAEQ RELANRMERK QMKEFMTMYS KLVQRCFDDCVND ..FTTKS LISREEGCVM RCVDKFMKGS QRLNERFQEQ NAAMMQSGQL PGR . 90 C. elegans Tim9Ap ............ MTSEQNIQ TFRDFLTQYN LVAEQCFNSCVNE ..FGSRT VSGKEESCAN NCLDKFLKMT QRVSQRFQEH QLLNAQANGA AIKVENGGKI 86 H. sapiens Tim9Ap ........MA AQIPESDQIK QFKEFLGTYN KLTETCFLDCVKD ..FTTRE VKPEETTCSE HCLQKYLKMT QRISMRFQEY HIQQNEALAA KAGLLGQPR. 89 C. elegans Tim9Bp ..MNTIQNIQ QLREFLTVYN TLSERCFNACARD ..YTTST LTKDEGSCVS QCIDKQMLVN RRFMLVFAEQ APKALFKQGE QSPTEAIKSA 86 D. melanogaster Tim9p ....MDSNLR NLKDFFTLYN KVTELCFSRCVDN ..LSQRD LGGHEDLCVD RCVTKFARFN QNMMKVYVDV QTTINAKRME EMEENARKAE 84 H. sapiens Tim9Bp ME RQQQQQQQLR NLRDFLLVYN RMTELCFQRCVPS ..LHHRA LDAEEEACLH SCAGKLIHSN HRLMAAYVQL MPALVQRRIA DYEA ASA 87 N. crassa Tim8p .MDIPQADLD LLNEKDKNE .. LRGFISNET QRQRVQGQTH ALTDSCWKKCVTSPIKT.NQ LDKTEAVCMA DCVERFLDVN LTlMAHVQKI TRGGSK 92 S. pombe Tim8p MADATKNPIA DLSESEQLE .. LSKFIESEQ QKVKLQQAIH QFTSTCWPKCIGNI ..G.NK LDKSEEQCLQ NCVERFLDCN FHIIKRYA .. LEKFGFLFCW LGFSC 98 H. sapiens Tim8Bp •••••••• MA ELGEADEAE .. LQRLVAAEQ QKAQFTAQVH HFMELCWOKCVEKP ..G.NR LDSRTENCLS SCVDRFIDTT LAITSRFAQI VQKGGQ .... 83 D. melanogaster Tim8p ......MSDF ENLSGNDKE .. LQEFLLIEK QKAQVNAQIH EFNEICWEKCIGKP ..S.TK LDHATETCLS NCVDRFIDTS LLITQRFAQM LQKRGGGDL. 88 A. thaliana Tim8p ......... M DPSMANNPE .. LLQFLAQEK ERAMVNEMVS KMTSVCWOKCITSA.PG.SK FSSSESSCLT HCAQRYMDMS MIIMKRFNSQ . 77 S. pombe Tim13p .MGIFGGNSG NAPSSEDKKS IFMKQIRQEL AVAQAGELIS KINENCFDKCIPEP ..G.ST FDPNEKSCVS KCMERYMDAW NIVSRTYISR MQREQKNLN. 95 H. sapiens Tim13p MEGGFGSDFG GSGSGKLDPG LIMEQVKVQI AVANAQELLQ RMTDKCFRKCIGKP ..G.GS LDNSEQKCIA MCMDRYMDAW NTVSRAYNSR LQRERANM 95 C. elegans Tim13p MDQLLDVETL KKLSPEQQEQ VI.SGVKQQA ALANAQNLVT DISEKCTNKCITAP ..G.SS LASGEKQCLQ RCMDRFMESW NLVSQTLQKR LQEEMASSGG MGGGFGQGPS 106 A. thaliana Tim13p .MDSYSSPPM GGSGSSVSPE VMMESVKTQL AQAYXEELIE TLRTKCFDKCVTXP ..G.SS LGGSESSCIS RCVERYMEAT AIISRSLFTQ L 87 Consensus E N C ..KC! L E ..C .. RCV.K . 139 140 Figure 3. Members of the Tim family in the mitochondrial intermembrane space. (A) Clustal amino acid sequence alignment of Pichia Timl Op with other homologues detected from the public databases using BLAST and FASTA searches. The four-conserved cysteine residues that constitute the twin CX3C motif are indicated in bold. Protein sequences analysed included: Saccharomyces cerevisiae TimIO (Z80875), TimI2* (Z35960), Tim9 (AF093244) (Z72966), Emericella nidulans TimlO (AAD40003), Tim9 (AAD400I6), Ciona intestinalis TimlO* (AAD4000I), Rattus norvegicus TimIO (AAD3997), Homo sapiens TimIO (AAD39995), Tim9A (AAD4006), Tim9B* (AAD4006), Tim8B (AAD39994), Tim13 (AAD3995 I), Drosophila melanogaster TimIO (AAD39998), Tim9* (AAD400IO), Tim8 (AAD39 I62), Caenorhabditis elegans TimIO (AAD40000), Tim9A* (AAD400I4), Tim9B* (AAD400I5), Tim13* (AAD39955), Lycopersicon esculentum TimIO (AAD40002), Arabidopsis thaliana TimIO (AAD39999), Tim8 (AAD39990), Neurospora crassa Tim8 (AAD39I6I), Schizosaccharomyces pombe Tim8 (AAD40476), Tim13 (AAD40477). , * partial sequences shown in the alignment. (B) Phylogenetic tree showing the relation ship between TimIO proteins from various eukaryotic organisms. Amino acid sequences wete retrieved from public databases and used for generating the tree by the neighbour-joining method. E. elegans C. intestinalis ,...... JR. norvegleus l.....- H. sapiens r-- D. melanogaster Po sorbitophlla C. albicans s. cerevisiae E. nidulans L. escutenta AI. thaliana Figure 3R Phylogenetic tree for Tim 10 proteins The tree was constructed using the Neighbor-Joining method after a clusterW aligned amino acid sequences. 141 A B PS SC MK kD 75'" 30... IOkD 15 ... Figure 4. Western blot analysis. Mitochondrial proteins were isolated from spheroplasts of S. cerevisiae (SC) and P. sorbitophila (PS). Extracts were analysed on SDS-PAGE (A) and immunoblotting (B) with a rabbit anti TimlOp. Immunodetection was performed with a chemiluminescence kit (Boehringer Mannheim). 142 Figure 5. Functional complementation. Growth of S. cerevisiae wild type strain GA7-6A (1), S. cerevisiae temperature sensitive mutant timlO-l transformed with the Pichia TM10 (2) or with the empty plasmid (3) was monitored at permissive and restrictive temperatures. 143 REFERENCES Adam, A., Endres, M., Sirrenberg, C., Lottspeich, F., Neupert, W. and Brunner, M. (1999). Tim9, a new component of the TIM22-54 translocase in mitochondria. Flv1BO J. 18,313-319. Baker, KP., and Schatz, G (1991). Mitochondrial proteins essential for viability mediate protein import into yeast mitochondria. Nature 349, 205-208. Bennetzen, lL.and Hall, B.D. (1981). Codon selection in yeast. 1. Bioi. Chern. 257, 3026-3031. Daum, G., Bëhni, P.e. and Schatz, G. (1982). Import of proteins into the mitochondriaZ Biol.Chem. 257, 13028-13033. Gietz, R.D., and Schiestl, R.H. (1995). Transforming yeast with DNA Meth. Mol. Cell. Bio/. 5,255-269. Gietz, R.D., and Sugino, A (1988). New yeast-Escherichia coli shuttle vectors constructed with in vitro mutagenized yeast genes lacking six-base pair restriction sites. Gene 74, 527-534. Glick, B.S., and Pon, L. (1995). Isolation of highly purified mitochondria from Saccharomyces cerevisiae. Methods Enzymol. 260,213-233. Hachiya, N., Mihara, K, Suda, K, Horst, M., Schatz, G. and Lithgow, T(1995). Reconstitution of the initial steps of mitochondrial protein import. Nature 376, 705-709 Hill, lE., Myers, AM., Koerner, Tl, Tzagoloff, A(1986).Yeast/E.coli shuttle vectors with multiple unique restriction sites. Yeast 2, 163-167. Hill, K.,Model, K, Ryan, M.T, Dietmeier, K, Martin, F., Wagner, R., and Pfanner, N. (1998). Tom40 forms the hydrophilic channel of the mitochondrial import pore for preproteins. Nature 395, 516-521. 144 Jarosch, E., Rodel, G., and Schweyen, R.l (1997). A soluble 12-kDa protein of the mitochondrial intermembrane space, Mrs Il p, is essential for mitochondrial biogenesis and viability of yeast cells. Mol. Gen. Genet. 255, 157-165. Jarosch, E., Tuller, G., Daun, G., Waldherr, M., Voscova, A, Schweyen, RI. (1996). Mrs5p, an essential protein of the mitochondrial intermembrane space, affects protein import into yeast mitochondria. J Biol. Chern. 271, 17219-17225. Koehler, e.M., Jarosch, E, Tokatlidis, K, Schmid, K, Schweyen, R.l and Schartz, G. (1998). Import of mitochondrial proteins of the intermembrane space. Science 279, 369-373. Koehler, CM, Leuenberger, D., Merchant, S., Renold, A, Junne, T. and Schartz, G. (1999). Human deafness dystonia syndrome is a mitochondrial disease. Proc. Natl. Acad Sci. USA 96,2141-2146. Larsson, N-G, and Clayton, D. A (1995). Molecular genetic aspects of human mitochondrial disorders. Ann. Rev. Genet. 29, 151-178. Laemmli, U.K (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227, 680-685. Rassow, J., Dekker, P.I.T., Wilpe, S., Meijer, M. and SolI, J. (1999). The preprotein translocase of the mitochondrial inner membrane: Function and evolution. J Mol. BioI. 286, 105-120. Read, S.M., and Northcote, D.H. (1981). Minimizing of variation in the response to different proteins of the Coomassie Blue G dye-binding assay for proteins. Anal. Biochem. 116, 53-64. Sambrook, 1., Fritish, E.F. and Maniatis, T. (1989). Molecular cloning. A laboratory manual, Second edition. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Sharp, P.M., Tuohy, T.M.F. and Mosurski, KR. (1986). Codon usage in yeast. Cluster analysis clearly differentiates highly and lowly expressed genes. Nucl. Acids. Res. 14,5125-5143. 145 Sherman, F., Fink, G.R and Hicks, lB. (1986). Laboratory course manualfor methods in yeast genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Sirrenberg, C., Endres, M., Folsch, H., Stuart, RA., Neupert, W. and Brunner, M. (1998) Carrier protein import into mitochondria mediated by the intermembrane proteins Timl0lMrsii and Tim121Mrs5. Nature 391, 912-915. Waldherr, M., Ragnini, A., Jank, B., Teply, R, Wiesenberger, G., Schweyen, Rl (1993). A multitude of suppressors of group II intron-splicing defects in yeast. CurroGenet. 24,301-306. 146 CHAPTER 7 Characterisation of a Putative Glycerol!Facilitator in the Fission Yeast Schizosaccharomyces pombe Abstract The fission yeast Schizosaccharomyces pombe can utilise glycerol as a sole carbon source or as an osmolyte during osmoregulation. Transport studies indicated that S. pombe might control glycerol flux using facilitated diffusion or a channel-mediated mechanism but the proteins involved in this transport are not known. Comparative Blast searches of the Schizosaccharomyces pombe databases revealed three putative glycerol transport proteins two of which show considerable structural similarities to known MIP family glycerol facilitators. One complete open reading frame designated as S. pombe mip I encodes a putative transmembrane protein of 598 amino acids with 52% identity to the Saccharomyces cerevisiae putative glycerol facilitator Yfl054p including the unique N- and C - terminal extensions. The gene was subsequently isolated by PCR from the S. pombe CBS 5682 (CSIR Y457) genomic DNA, verified and cloned into yeast expression vectors for functional analysis. Expression of the S. pombe mipi into the Saccharomyces cerevisiae wild-type or the fpsl L1and gpdi L1gpd2L1 strains, had no effect on osmotic sensitivity or glycerol release during hypo-osmotic stress. However, cells expressing S. pombe mip l had a lower intracellular /extracellular glycerol ratios as compared to control cells during osmotic stress. The osmosensitive phenotype of fpsi L1and gpdi L1gpdi á mutants that is suppressed by the overexpression of a glycerol facilitator in the plasma membrane, were not suppressed by the overespression of S. pombe mipi. Northern blotting experiments revealed that S. pombe mip l expression is induced during osmotic stress suggesting a role in osmoregulation. However, deletion of the gene did not have any observable effect on the growth of S. pombe under osmotic stress or on glycerol as a sole carbon source. Therefore, the physiological role of this protein as well as the actual transporter (s) controlling glycerol flux in S. pombe remains to be elucidated. 147 INTRODUCTION Most yeasts can utilise glycerol both as a sole carbon source or as an osmolyte in order to maintain osmotic homeostasis. Consequently, yeasts have developed elaborate mechanisms to regulate its flux across cell membranes. In the fission yeast Schizosaccharomyces pombe, glycerol appears to be the main osmolyte with which cells maintain osmotic homeostasis (Ohmiya et al., 1995). Glycerol is synthesised from the glycolytic intermediate dihydroxyacetone phosphate in two steps that are catalysed by an NADH-dependent glycerol- 3-phosphate dehydrogenase and a phosphatase (Gancedo et al., 1968). Under osmotic stress conditions, the expression of genes involved in glycerol synthesis such as gpdl are upregulated by the Sty 1- signal transduction pathway leading to the intracellular accumulation of glycerol (Aiba et al., 1995). In absence of osmotic stress, the synthesised glycerol is either utilised or released rapidly out of the cell. Glycerol was originally thought to permeate across yeast membranes only by simple diffusion. However, recent investigations have presented evidence that suggests the involvement of active transport systems and glycerol facilitating proteins (for review, Kayingo et al., 2001). Indeed, transport systems that allow yeasts to maintain a steep glycerol concentration gradient across the membrane, have been already reported (Lages et al., 1999). Little is known about the transport proteins and the genes encoding these proteins except in the yeast S. cerevisiae. Recently the gene GUP 1, encoding a putative active glycerol uptake protein in S. cerevisiae has been identified and partially characterised (Holst et al., 2000). Furthermore, a gene FPSl, which encodes a MIP channel protein with similarity to bacterial glycerol facilitators, was described in S. cerevisiae (Van Aelst et al., 1991). The Fps1p has been shown to control the movement of glycerol across the membrane and to play an important role in yeast osmoregulation (Luyten et al., 1995; Sutherland et al., 1997; Tamás, et al., 1999). Most glycerol facilitators so far identified are related to aquaporins such as AQPO, the bovine major intrinsic 12rotein (Gorin et al., 1984) from which the name MIP family was derived (Baker and Saier, 1990). These proteins exhibit a typical structure with six transmembrane spanning domains, an internal sequence repeat and highly conserved motifs (for recent review see Kayingo et al., 2001). Functional analysis using heterologous systems and deletion mutants suggest that MIP channels playa critical role in osmoregulation by regulating water and solute transport across cell membranes (for reviews, Agre et al., 1998; Hohmann et al., 148 2000). MJP channel proteins may play additional roles in a wide range of physiological processes especially where a regulated water and solute transport is critical (Ludevid et al., 1992; Kaldenhoff et al., 1995). To date, a significant number of MJP channels have been characterized in higher organisms and their role in mediating water and solute flux established. As far as microorganisms are concerned, MJP channels have only been well studied in a few organisms. In Escherichia coli, there are two MJP channel proteins, namely the water channel AqpZ (Calamita et al., 1995) and the glycerol facilitator GlpF (Sweet et al., 1990). The S. cerevisiae genome revealed four genes encoding MIP channels (André, 1995) of which FPSJ has been shown to be a glycerol facilitator whereas AQYI and AQY2 are water channels (Bonhivers et al., 1998; Laizé et al., 1999). The function of the fourth MIP gene, YFL054, is not yet known. Although its expression is significantly induced after a diauxic shift from fermentation to respiration (DeRisi et al., 1997) and during osmotic stress (Tamás, 1999), there is no evidence yet for its involvement in osmoregulation and/or cellular metabolism. Recently, and mainly from genome sequencing projects, several MIP channels have been identified in many other unicellular organisms but little is known about their physiological roles (Kayingo et al., 2001). The genome projects focusing on the fission yeast S. pombe has so far revealed two fragments encoding MIP channel proteins (http://www.sanger.ac.uk). The l fragment SPAC977 on chromosome 1, appears to encode a complete open reading frame thatwe have called S. pombe mipl. To our knowledge, the functions of these genes have not been examined. Therefore, this study aimed at characterising the S. pombe mipl with respect to its structure-functional properties, patterns of gene expression as well as its phylogenetic relationship with other known MIP channels. 149 MATERIALS AND METHODS Strains and growth conditions The strains used in this study are shown in Table 1. S. cerevisiae and S. pombe mutants were derived from the haploid laboratory strains W303- lA and L972 H respectively. All S. cerevisiae wild type strains were routinely grown on a rotary shaker at 30°C in YEPD broth (lOg r1 yeast extract, 20 g r' each of peptone and glucose) or on defined medium (6.7 g r1 yeast nitrogen base, 20 g r1 glucose and amino acids, pH 6) as discussed below. In some experiments, the glucose concentration was increased to 100 g r' or replaced by glycerol (3% v/v) as a sole carbon source. For transformations, S. pombe was grown in Edinburgh minimal medium (Moreno et al., 1991) whereas YEPD was used for normal cultivation and routine maintenance. Anaerobic growth experiments were carried out in the anaerobic incubator (Forma Scientific USA) at 30°C for 2-3 days. Molecular and genetic methods In general, nucleic acid manipulations were carried out according to Sambraak et al. (1989) unless otherwise mentioned. Large amounts of pure plasmid DNA was isolated from E. coli using the Nucleobond AXI00 cartridges ofMacherey-Nagel Germany (Cat. No 740521). For isolation of small quantities of plasmid DNA, cetyltrimethylammonium bromide (CTAB) was used following the method of Del Sal et al. (1988). Total RNA was isolated from exponentially growing yeast cells cultivated in YEPD with or without 5% NaCI. Cells were harvested and lysed with 0.2 mg/ml Zymolase and 8 Ill/ml p-mercaptoethanol in extraction buffer (1 M sorbitol, 100 mM Nar-citrate, 60 mM EDTA, pH 7.0). After cell lysis (3 hours at 370C) and homogenization, total RNA was isolated from the yeast spheroplasts using the Rneasy kit (Qiagen, Cat. No. 74103) for small quantities or using the TRI™ reagent (Sigma, Cat. No. T9424) when large quantities of RNA were desired. Total genomic DNA was isolated from S. pombe and S. cerevisiae as described by Moreno et al. (1991) and Sherman et al. (1986) respectively. For blotting experiments, nucleic acids were fractionated by gel electrophoresis and transferred to a positively charged nylon membrane (Magna Graph, MSI, Cat. No. NJOHY00010) by capillary blotting (Sambrook et al., 1989). 150 Hybridization, probe labelling and detection were performed using the digoxigenin (DIG) system (Roche Diagnostics). Isolation and cloning of S. pombe mipl The polymerase chain reaction was used to amplify the entire reading frame encoding the S. pombe MIJ> channel. The following primers (Integrated DNA Technologies) were used: Pombe up 5' gggatcc taactaaatgagcgtcc 3' BamHl Pombe do 5' ggtcgac gagttcaattattctc 3' Sail PCR conditions: 94°C, lrnin, 94°C, 30sec~ 52°C, 30sec~noc, l min; 25 cycles. Amplification was carried out with 2.5 units of Taq DNA polymerase (Roche Diagnostics) in presence of 200 !JM dNTPS, 1.5 mM MgCh and O.5 ~ primers. The PCR product (l.8 kb) was purified (High Pure™ PCR purification kit; Roche Diagnostics, Cat. No. 1732676), cloned into pGMT -easy vector (promega) and verified by restriction digestion analysis and sequencing. The insert was subsequently subeloned into yeast vectors YEplac195 and pKU24 using the Sail and BamHl linkers introduced into the primers. For heterologous expression of the S. pombe mipl, a yeast expression vector with a strong constitutive promoter and terminator sequences of the S. cerevisiae PGKl gene encoding phosphoglycerate kinase, was used. The l.8 kb HindIII fragment containing the PGK promoter and terminator sequences were subeloned into the YEplac 181 and the entire S. pombe mipl reading frame was then inserted into the BglU and Xhol isoschizomers respectively to create the construct pPGK-Spmipl. The pPGK-Spmipl plasmid (Fig. 1) was transformed into the S. cerevisiae W303-1A and fps lI:!.strains and its effect on growth under hyper- and hypo-osmotic conditions as well its effects on the glycerol conservation were analyzed. Yeast transformations were performed using the lithium acetate method of Gietz and Schiestl (1995). 151 Disruption of the S. pombe mipl The S. pombe MIP gene was disrupted in the one step procedure of Rothstein (1983) in the strain ura4-D18 (Grimm and Kohli, 1988). The pGMT-Spmipl plasmid was digested with HindIII which cuts inside S. pombe mip l to release a fragment of about 500bp. The l.8kb HindIII fragment encoding the S. pombe ura4 gene was cut from the pCG 1 plasmid (Grimm et al .. 1988) and then inserted into S. pombe mip l to create a disruption (Fig. 2). The linear fragment carrying the disrupted S. pombe mipl was gel purified (Qiagen) and transformed into the S. pombe ura4-D18 strain. Transformants were selected on minimal media without uracil and then replica-plated several rounds on YEPD plates and back to selective plates. Surviving colonies were examined for Southern blot and PCR analyses. Glycerol transport assays and osmotic stress experiments Yeast cells were grown in 250 ml defined media (glucose- YNB) with amino acids (pH 6) until OD600= l. Cells were harvested (5000 rpm, 5 min) resuspended in 100 ml of glucose- YNB with 5% NaCI and grown at 30°C for 3 hours. Samples of 1.5 ml were taken from which . cells were harvested (5000 rpm, 2 min) and resuspended in O.lml of similar media. Hypo- osmotic shock was performed by diluting (lOX) that cell suspension with media without NaCL After a given time interval, cells were immediately sedimented, washed and re- suspended in 1ml of water. Intracellular glycerol was extracted by boiling the cell suspension for l O min followed by centrifugation (13000 rpm,S min) to remove the debris. Glycerol content was analyzed using chromatographic (Dionex , MAl column) and enzymatic methods (Boehringer Manheimn Cat. No. 148270). For osmotic sensitivity experiments, cells were exposed to hyper-osmotic stress by increasing the NaCl concentration from 0 to 5% or hypo- osmotic stress by decreasing the NaCI concentration from 5 to 0 % as described previously (Tamás et al., 1999). Spot tests were also carried out by serial dilutions of the cell suspension that were subjected to osmotic stress prior to plating. Five microlitres of the cell suspensions were spotted onto agar plates with or without NaCI addition and incubated at 30°C until the colonies were visible. 152 Table 1. Yeast strains used in this study ram Genotype Reference cerevisiae W303-1A MAT a leu2-3/112 ura3-1 trpl-l his3-11/15 Thomas and Rothstein, 1989 ade2-1 canl-lOO GAL SUe2 malO cerevisiae W303-1Ayfl054iJ::TRP13 Tamás et al., 1999 cerevisiae W303-1A fpsliJ::HIS3 Tamás et al., 1999 cerevisiae W303-1A gpdliJ::TRPl gpd2iJ::URA3 Ansell et al., 1997 . pombe CBS 5682 Wild type Chapter 3 '. pombe 972h- Zeul-32 ura4-D18 Grimm et al., 1988 This study 153 S. pombe mipl ORF r " / BamHl ", / Sail , I , I , I , I , I ~ "/", , I , I , I , I , I , ", , I"I , I , I '. I PGK promoter B£lll ~ IV Xhol PGK 3' region "'~~~~_ ~indlll Hindlil ~ pPGK-Spmipl 9.3 kb g ~ YEplac181 LEU2 ~. PGK sequences ~""""~ S. pombe mipl ORF Figure 1. Construction of the yeast expression vector and insertion of the S. pombe mipl gene. The 1.8 kb HindlIl fragment containing the S. cerevisiae PGK promoter and terminator sequences were subeloned into the multicopy plasmid YEplac181. A 1.8 kb BamHl/Sall fragment encoding the entire S. pombe mipl open reading frame was inserted into the Bgll1 and Xho 1 isoschizomers respectvely to create the construct pPGK-Spmipl. 154 H H peGl 1.8 kb Hindi II fragment encoding ura4 from peG 1 R H V ~~~~~ S. pombe mip 1 V R H H .~~~~~~~~~~~~~~~~~~~~~~~~~ S. pombe mtp liura-t' ~""""~ S. pombe mip 1 sequences ~.c<:'.c<:'.c<:'.K.( S. pombe ura4 sequences Figure 2. Partial restriction map and strategy for disruption of S. pombe mipl. Partial restriction map indicating the relevant sites of the S. pombe mipl ::ura4 disruption construct. H, Hindiil; R,EcoRI; V, EcoRY.Not drawn on scale. 155 RESULTS AND DISCUSSION Glycerol transport in S. pombe. The patterns and kinetics of glycerol export from S. pombe upon a hypo-osmotic shock (osmotic downshock) are shown in Figure 3. There was a rapid glycerol efflux upon an osmotic downshock and cells lost more than 75 % of their intracellular glycerol in 5 minutes regardless of the downshock incubation temperature. The pattern of glycerol release suggests that a specific protein mediated glycerol transport system may exist in S. pombe with which cells might regulate glycerol flux across the membrane. These observations prompted a search for transporters that might be involved. Analysis of the S. pombe genome database (http://www.sanger.ac.uk) for sequences homologous to known glycerol transporters revealed three putative glycerol transport proteins. The one encoded by SPAC24H6. OJc (ORF of 231 amino acid residues) is homologous (49% identity) to the N-terminal part of Gupl, a membrane protein involved in active glycerol uptake in S. cerevisiae (Holst et al., 2000). The other two putative glycerol transporters encoded by SPAC977.J7 (598 amino acids) and by SPACJ86 (incomplete) are homologous to glycerol facilitators of the MIP family (Hohmann et al., 2000; Kayingo et al., 2001). In particular, the former is most closely related (52% identical) to the to the yeast putative glycerol facilitator YFL054p (Fig. 4A), and more diverse to MIP channels from other fungi (Fig. 4B). This gene designated as S. pombe mip I codes for a putative protein of 598 amino acids, with a predicted molecular weight of 65815.73 and a pI of6.12. 156 Isolation and sequence analysis of S. pombe mipl The S. pombe :MIP homologue was isolated by PCR using primers designed from the sequence SPAC977.17 (WWW.sanger.ac.k/projects/s_pombe). A PCR product of about 1.8- kb was obtained from both S. pombe 972h- and S. pombe CBS 5682 (Fig. SA). The PCR products were subsequently cloned into pGMT -easy vector and verified by restriction digestion analysis and sequencing. Southern and Northern blots probed with the labeled PCR product, confirmed that indeed the cloned gene originated from S. pombe (Fig. 5). No hybridization of this probe to other yeasts such as Saccharomyces cerevisiae, Zygosaccharomyces rouxii, Debaryomyces hansenii, Pichia sorbitophila and Candida tropicalis was observed when the Southern blotting was conducted even under medium stringent conditions (not shown). Sequence analysis of the S. pombe mipl showed all typical features of the MIP family (park and Saier, 1996) such as the canonical, tandemly arranged NPA motifs and other conserved residues (Fig. 4A and 6). A hydrophobicity plot (Kyte and Doolittle, 1982) indicated that the putative protein has six transmembrane domains like other MIP proteins (Fig. 4A, 6 and 7). The S. pombe mipl has an extended N-terminal extension of approximately 300 amino acids leading to a protein of about 598 amino acids unlike most :MIP proteins which consist of about 250-300 amino acids. The S. pombe mipl exhibits the amino acid residues typical for glycerol facilitators (Fig. 6) suggesting that it might be involved in the transport of glycerol and lor other related solutes (Froger et al., 1998; Kayingo et al., 2001). However, unlike most glycerol facilitators where positions 4 and 5 (Fig. 6) are usually a proline followed by a nonaromatic residue, the S. pombe mip 1 and its closest homologue Yfl054p possess an alanine in P4 and a tryptophan in PS. Heterologous etpréssion of S. pombe mipl in S. cerevisiae and functional analysis PCR analysis arid N~rthern blotting confirmed that S. pombe mipi was properly expressed in S. cerevisiae strains. A transcript was detected in cells transformed with pPGK-Spmipl as opposed to the control cells (Fig. 8). In this study, the effects of S. pombe mipi on growth and glycerol conservation during osmotic stress conditions were analyzed both in the fpsl Ll strain and its isogenie wild-type S. cerevisiae W303-1A. 157 The growth of S. cerevisiae strains expressing the S. pombe mip l gene commenced later and grew slightly slower than the strains containing only the vector in glucose- YNB with or without 5% Nael (Fig. 9) suggesting that expression of the gene affects growth properties. The intracellular glycerol content of cells transformed with the S. pombe mipl was significantIly lower than that of cells without the gene (Table 2) suggesting that the expression of S. pombe mip l in S. cerevisiae might interfere with the retention of glycerol in both wild-type and in a strain lacking the glycerol facilitator Fpslp. However, this gene does not appear to alleviate the inability of the fpsl d strain to release glycerol upon hypo-osmotic shock (downshock). The pattern of glycerol release was identical in the /psidstrain expressing S. pombe mipi to the/psi d strain without S. pombe mip l (Fig. lOA). Furthermore, the expression of S. pombe mipi in the fpsl d strain was unable to restore the hypo-osmotic stress sensitivity of the fpsl d strain to the wild-type (Fig. lOB). These experiments suggest that the S. pombe mipl affects glycerol retention in S. cerevisiae but does not appear to act as a glycerol export channel similar to Fpslp found in S. cerevisiae (Luyten, et al., 1995; Tamás et al., 1999). Previous studies have indicated that expression of bacterial or eukaryotic glycerol facilitators alleviate thefpsi d growth and transport defects (Luyten et al., 1995, Sutherland, et. aI., 1997; Weig and Jakob, 2000; Prudent et al., 2000). Whether the S. pombe mipl transports glycerol in ungated way albeit with a low turnover insufficient to alleviate the downshock sensitivity of fpsl d strains, remains to be established. We tested further for growth phenotypes under conditions where glycerol export is required. Ansell et al. (1997) showed that during anaerobic growth, glycerol production is essential for redox balancing. Under these conditions, the fpsl dmutants accumulate large amounts of glycerol inside the cell and grow much slower than the wild type (Tamás et al., 1999). Expression of S. pombe mip l did not improve growth ofthefpsiL1 mutants during anaerobic conditions (Fig. II). To investigate whether the S. pombe mipi is involved in the uptake of glycerol and or other polyols such as xylitol, it was expressed in the S. cerevisiae gpdi L1gpd2L1 strain. This strain is unable to produce glycerol (Ansell et al., 1997) and is thus sensitive to hyper-osmotic stress caused by 2 M glycerol, 1 M erythritol or 1 M xylitol. However, its growth can be improved when transformed with a solute transporter that is capable of equilibrating the solute in and out of the cell (Karlgren et al., 2000). Expression of S. pombe mip l did not improve growth of the gpdl á gpd2L1mutants either on 2 M glycerol or 1M xylitol (data not shown). 158 These results suggest that under the conditions tested, the S. pombe mipl was neither involved in the uptake of glycerol nor xylitol across the plasma membrane of S. cerevisiae. Whether the S. pombe mipl may require to be regulated by components not present in S. cerevisiae or whether it is involved in other functions unrelated to solute transport, is not yet clear. The S. pombe mipl' mutant phenotypes To further investigate the function of S. pombe mip l, a disruption mutant was constructed (Fig. 2). The disruption was confirmed by PCR analysis and Southern blot using ura4 as a probe (Fig. 12). The S. pombe mip li.ura-t" mutant phenotypes were then tested under different physiological conditions where MIP channels have been shown to playa role. Lack of a glycerol facilitator has been reported to affect growth of some microorganisms on medium containing glycerol as a sole carbon source (Voegele et al., 1993; Schweizer et al., 1997). Disruption of the S. pombe mipl did not cause any observable growth defect on glucose or glycerol as a sole carbon source (Fig. 13). It is therefore likely that glycerol uptake for metabolic utilization does not occur via this channel. In addition, S. pombe does not present activity of mediated high or low affinity glycerol uptake (Lages et al. 1999) unlike S. cerevisiae where an active glycerol transport protein Gup! (a non MIP protein), mediates glycerol uptake for the metabolic needs of the organism (Holst et al., 2000). Therefore other transport systems may be involved. In addition to glycerol uptake, microbial MIP channels have been shown to play a role in osmoregulation (for reviews, Booth and Louis, 1999; Hohmann et ai., 2000). Therefore, the growth of S. pombe mipli.ura-i" mutant under osmotic conditions was monitored. Surprisingly, the S. pombe mip l.turas" mutant survived both the hyper- and hypo-osmotic shocks in a similar way as the wild-type cells (Fig. 14) suggesting that the S. pombe mipl might not be involved in short-term adaptation to osmotic stress. I However, as pointed out previously (Chapter 2; Kayingo etal., 2001), phenotypes associated with deletions of genes encoding MIP channels are not always clear and are difficult to interpret. They can be strain, concentration or growth phase dependent. For instance, deletion of E. coli glpF causes a growth defect only at low glycerol concentrations (Truniger et al., 1992, Voegele et al., 1993). Furthermore, deletion of the aqpZ gene causes a visible sensitivity to low osmolarity only after co-cultivation in competition with wild type (Calamita et al., 1998). Hence, to detect the phenotype of a MIP channel may require very careful inspection as well as knowledge of the conditions the, organism experiences in its natural environments. 159 Expression of S. pombe mipl is slightly induced by osmotic stress (NaCI) A number of MlP genes appear to be affected by osmotic stress and their expression patterns may vary according to the growth phase of the organism (Flick et al., 1997). In this study, the expression of S. pombe mip l was monitored by Northern blotting using the S. pombe ura4 as a control. A decrease in the S. pombe mipl mRNA level was observed as cells entered stationary phase (Fig. 13). However, growing S. pombe cells in media containing 5% NaCI resulted in a 2-3 fold induction of S. pombe mipl (Fig. 15). This is in contrast to other osmoregulated genes such as gpdl (Ohmiya, et al., 1995~Aiba et al., 1998) whose expression can be induced up to 20-fold by an increase in external osmolarity. Fission yeast cells respond to increased osmolarity via a mitogen-activated protein (MAP) kinase cascade, involving the Wis1 MAP kinase kinase and the Styl MAP kinase (for review, Banuett, 1998). The Styl MAP kinase transduces a signal to a target gene via a basic leucine zipper (bZIP) transcription factor, Atfl. Several stress responsive genes such as fop 1, eft1, and gpdl have upstream activating sequences (UAS) on which Atfl binds directly and then trigger gene expression (Wilkinson et al., 1996). The signaling pathway controlling the observed induction in the S. pombe mip l is unknown. Analysis of the upstream sequences of S. pombe mipl revealed three potential Atfl binding sites (CRE motif) at nucleotide -859 to - 852 (CAACGTTC), -448 to -441 (AAACGTTC), and -180 to -187 (ACACGTTG) relative to the translation initiation codon (ATG). These sequences suggest that S. pombe mipl could be controlled in Atfl-Wis- Styl dependent way. However, this observation requires experimental confirmation. MIP channels can be regulated both at gene and protein level. Phosphorylation sites controlling the activity of a significant number of eukaryotic MIP channels have been observed by many workers (Weaver et al., 1994~ Johnsson et al., 1996~Maurel et al., 1997). To date, little is known about MIP channel phosphorylation in microorganism but putative phosphorylation sites occur in the S. pombe mipl. Recently, a domain required to regulate the closure of the yeast Fpslp glycerol facilitator was identified (Tarnás et al., 1999). The Fps1p regulatory domain is located in the N-terminal extension, which is an unusual feature among the MIP family proteins. Interestingly, S. pombe mipl also exhibits an extended N-terminal extension but unlike its closest homologue Yfl054p, it has no sequence that shows some similarity to the regulatory domain ofFpslp (Fig. 4A). It is thus likely that S. pombe mipl is regulated in a way different from that ofFpslp. 160 Concluding remarks Although in si/ico analysis suggests that the S. pombe mipl encodes a glycerol facilitator, in vivo experiments did not indicate its involvement in glycerol transport across the plasma membrane. It is still unknown whether the protein is involved in transport of yet unidentified solute. However, the observation that S. cerevisiae cells overexpressing S. pombe mipl leak glycerol during osmotic stress tempts one to believe that the gene may encode for ungated glycerol channel. Possibly the protein permeates glycerol with a low tum over insufficient to make cells sensitive to hyper-osmotic stress or to alleviate the downshock sensitivity offpsl.t1 mutants. While the expression studies point to a role in osmoregulation, the mutant does not show phenotypes related to osmoregulation. Similarly, the closest homologue from S. cerevisiae YFL054 is induced by osmotic stress in a Hog-dependent way suggesting a role in osmoregulation but the mutants also show no osmotic stress related phenotypes (Tamás, 1999). Furthermore, YFL054 shows a 5.6 fold induction during a diauxic shift from fermentation to respiration (DeRisi et al., 1997) but the significance of this induction is still obscure as there is no evidence for its involvement in metabolism. Therefore, the physiological role as well as the transport properties of these MIP channels are still intriguing questions. Localization studies on S. cerevisiae MIP proteins indicate that Yfl054p might be a vacuole protein (Tamás, 1999b) whereas aqy2-1 p is abundant on microsomal vesicles (Meyrial and Tacnet, 2000). Localization studies in homologous systems and heterologous expression in other systems such as Xenopus oocytes might provide clues on the physiological roles of S. pombe mipl as well as its transport properties. 161 ,,s-:.:,.-. 1000.. ~""~os:: 80 0 . (.).-..r..o...'.2... 60 ~0~-8~ (.) ~ 40 00 -~--;:::l8 20ro .~..... o 10 20 30 Time after hypo-osmotic shock (min) Figure 3. Glycerol release from S. pombe CBS 5682 (CSIR Y457) during hypo- osmotic stress. Cells were grown at 30°C YEPD broth until mid-exponential growth phase, resuspended in YEPD + 5% NaCl and cultured for 3 hours, harvested and then resuspended in YEPD without NaCI (osmotic downshock) held at O°C(0) or at 28°C (e). 162 spombeMIP MSVPLRFSTPSS----SPSASDN--------ESVHDDGPTTELDT--------------- 33 YFLOS4C M..S..YESGRSSSSSESTRPPTLKEEPNGKIAWEESVKKSRENNENDSTLLRRKLGETRKAIE 60'*:. ** . '*: .:: ***: .. '* *: spombeMIP --------FNTTDVPRRVNTTKARQMRPKN-TLKVAFSSPNLKGLDNTADSDSQPWLGGY 84 YFLOS4C TGGSSRNKLSALTPLKKVVDERKDSVQPQVPSMGFTYSLPNLKTLNSFSDAEQARlMQDY 120 : : '* . : : *: :: .::* **** *:. :*:: . . ." spombeMIP LAGRLEDISGQSRRNYVDPYYEELN--AGRRPNKPVWSLNGPLPHVLGNSVVEKlSQKNQ 142 YFLOS4C L..S---RGVNQGNSNNYVDPLYRQLNPTMGSSRNRPVWSLNQPLPHVLDRGLAAKMIQKNM 177: ***** *.:** .. *:****** ****** ..... *: *** spombeMIP EARSRANSRVNSRANSRANSSVSLAGMDGSPNWKRKMKSAVFGSRVKLNDEEAQLPRNKS 202 YFLOS4C DARSRASSRRGSTDISRGGSTTSVK------DWKRLLRGAAPGK--KLGDlEAQTQRDN- 228 ** .. *: . *: :*** ::. *. * **.* *** *:: spombeMIP SVSlAEQAASRPKVSFSLQSSRQPSlAEEQPQTQRKSSAlTVEHAENAEPETPRNNVSFS 262 YFLOS4C --Tv@ADVKPTKLEPENPQKPSN~lENVSRKKKRTSHNVNFSL.GDESYASSlADAESRK 286*.. .. * : . :.: *. * ............... .. spombeMIP RKPSlAEQDSSQDlTMPPNEllAEES-----LDS---GSDTET-----LYLNYWCKlRHF 309 YFL054C LK..NMQTLDGSTPVYTKLPEELlEEENKSTSALDGNElGASEDEDADlMTFPNFWAKlRYH 346: :. *: * *:*:* ** .... *:. : : *:*.***: . spombeMlP FREGFAEFLGTLVLVVFGVGSNLQATVTNGAGGSFESLSFAWGFGCMLGVYlAGGlSGGH 369 YFLOS4C MREPFAEFLGTLVLVIFGVGGNLQATVTKGSGGSYESLSFAWGFGCMLGVYVAGGlSGGH 406 TMD1 TMD2 :** ***********:****.*******:*:***:****************:******** spombeMlP VNPAVTlSLAlFRKFPWYKVPlYlFFQlWGAFFGGALAYGYHWSSlTEFEGGKDlRTPAT 429 YFL054C lNPAVTlSMAlFRKFPWKKVPVYlVAQllGAYFGGAMAYGYFWSSlTEFEGGPHlRTTAT 466 TMD3 :*******:******** ***:** ** **:****:****.********** ***.** spombeMlP GGCLYTNPKPYVTWRNAFFDEFlGTAVLVGCLFAlLDDTNSPPTQGMTAFlVGLLlAAlG 489 YFL054C GACLFTDPKSYVTWRNAFFDEFlGASILVGCLMALLDDSNAPPGNGMTALllGFLVAAlG 526 TMD4 TMDS *.**:*:**.**************:: :*****:*:***:*:** :****:*:*:*:**** spombeMlP MALGYQTSFTLNPARDLGPRMFAWWlGYGPHSFHLYHWWWTWGAWGGTlGGGlAGGLlYD 549 YFL054C MALGYQTSFTlNPARDLGPRlFASMlGYGPHAFHLTHWWWTWGAWGGPlAGGlAGALlYD 586 TMD6 **********:*********:** ******:*** *********** * *****.**** spombeMlP LVlFTGPESPLNYPDNGFlDKKV--------HQlTAKFEKEEEVENLEKTDS--PlENN- 598 YFLOS4C lFlFTGCESPVNYPDNGYlENRVGKLLHAEFHQNDGTVSDESGVNSNSNTGSKKSVPTSS 646 :.**** ***:******:*:::* ** ........ *. *: ... :*.* Figure 4A. Sequence alignment of S. pombe mipl and its closest homologue Yfl054p from S. cerevisiae. The highly conserved NPA motifs are indicated in bold and the amino acid residues in Yfl054p that shows homology to the regulatory domain of Fpslp are shown in a box. The six transmembrane domains (TMD) as predicted using the TopPred2 programme (http://www.biokemi.su.se) are underlined. Stars and dots indicate where there is perfect or imperfect match respectively. 163 S. cerevisiae FPSl B. cinerea PgF S. cerevisiae YFL054 S. pombe mipl ~ Calbicans AQP S. cerevisiae AQYl Figure 4B. Phylogenetic tree of fungal MIP channels. The tree was constructed using the Neighbor-Joining (NI) method after a clustalW aligned amino acid sequences; S. cerevisiae glycerol facilitator FPSl (P23900), YFL054 (P43549), . aquaporinAQYl (AAC69713), Candida albicans putative aquaporinAQP (contig4- 2389), Botrytis cinerea putative glycerol facilitator (ALl12633) and S. pombe mipl (SPAC977.17). 164 A B C 1 2 1 2 1 2 ._ -6kb gure 5. Isolation and verification of the S. pombe mipl . 1.8 kb S. pombe mipl PCR product was isolated from S. pombe strains (Panel A), labeled and d in Southern blots (not shown) and Northern blot hybridization (Panel B) to confirm the gin and expression of the gene from S. pombe CBS 5682 wild-type strain (1) and S. pombe Zh laboratory strain (2). The two S. pombe strains were also verified by Southern blotting g ura4 as a probe on yeast genomic DNA digested with EcoRl (panel C). 165 Repeat 1 Extracellular 1 Intracellular eOOH • NPA boxes Conserved between repeats • Conserved throughout the family * Residues influencing specificity Figure 6. Schematic representation ofthe membrane topology for S. pombe mipl. The extended NH2-terminus (about 310 amino acids) is indicated by a blue arrow. The putative membrane channel protein is predicted to have six transmembrane domains (1-6) and five connecting loops (A-E). The highly conserved residues common to most MIP family members as well as those residues that influence channel activity are highlighted (Froger et al., 1998; Hohmann et al., 2000). 166 100 200 300 400 500 600 s. 4pombe mipl 3 2 1 0 -1 -2 100 200 300 400 500 600 Amino acids figure 7. Hydropathy profiles of the S. pombe mip 1 channel protein calculated according to yte and Doolittle (1982). 167 1 2 Figure 8. Analysis of expression of the S. pombe mip l in S. cerevisiae by Dot blotting Total RNA from the JPslL1 strain + pPGK vector only (1) and JPslL1 + pPGK-Spmipl (2) was diluted, spotted on a nylon membrane and probed by the digoxigenin labeled fragment of S. pombe mip 1 168 3 g §\O 2 1 o 10 20 30 40 50 Time (hours) --+- WT + vector only - -0- - WT + S. pombe mip l __, - fpsl á + vector only --z:;;:r- fpsl á + S. pombe mip l ---II- WT + vector only with 5% NaCI -0- WT + S. pombe mip l with 5% NaCI ......._ fpsl Lt + vector only with 5% NaCI -<>- fpsl Lt + S.pombe mip I with 5% NaCI Figure 9. Growth of S. cerevisiae W303-1A and/psILt strains transformed with the pPGK vector with (open circles) or without (closed symbols) the S. pombe mipl gene in glucose YNB containing 0% (circles, triangles) or 5% Nael (squares, diamonds) 169 Table 2. Distribution of intra-and extracellular glycerol content in S. cerevisiae W303-1A and fpsl ~ transformed with pPGK vector with or without S. pombe mip l when cultivated in glucose -YNB with 5% Nael till mid-exponential growth phase (OD6oo= 1). Values represent mean ± standard deviations of triplicate determinations. Sample Intracellular glycerol gil Extracellular glycerol gil % Intracellular glycerol fpsl á + vector only 0.195 ± 0.011 0.222 ± 0.009 46.76 ± 0.22 fpsl + S. pombe mip I 0.168 ± 0.006 0.274 ± 0.029 38.01 ± 1.65á WT + vector only 0.241 ± 0.018 0.39 ± 0.001 38.19 ± 1.70 WT + S. pombe mipl 0.114 ± 0.005 0.261 ± 0.004 30.44 ± 0.75 A 170 100 ~ .... ..:..9.... 0····0·········0 - --y--Y----v --r> •• -'\A •• V· • sv ''7-'.-. O-L--...------.---_-.- r-___' o 10 20 30 Time after hypo-osmotic shock (min) ___._ S. cerevisiae W303-IA (WT) .. -0 .. S. cerevisiae fpsl Lt ~ fpsl Lt + S. pombe mip l ---'\1... S. pombe CBS 5682 B (psi Lt +pPGK vector only WT +pPGK vector only (psi Lt +pPGK-Spmipi WT +pPGK- Spmip l Figure 10. Effect of the expression of of S. pombe mip I on glycerol release (A) and survival (B) in S. cerevisiae W303-1A and/psiLt strains during hypo-osmotic stress. Cultures were grown in glucose- YNB containing 5% NaCI for 3 hours followed by a downshock in glucose- YNB lacking NaCl. 171 WT +pPGK vector only WT +pPGK-Spmipi {psi Li +pPGK vector only fpsl á +pPGK-Spmipi Figure 11. Growth of S. cerevisiae W303-1A andjpsi,1 strains with or without S. pombe mipi cultivated on glucose-YNB under anaerobic conditions. 172 M 1 2 Figure 12. Southern blotting analysis of the deletion mutant. Genomic DNA from the S. pombe 972h (lane 1) and YGK200 (lane 2) was digested with Xbal, transferred to a nylon membrane and probed with the ura4 Dig labeled fragment as indicated in the materials and methods 173 S.pombe S.pombe mipl::ura4 972h- S. cerevisiae S. cerevisiae yfl054A W303-1A A B Figure 13. Growth of the S. pombe and S. cerevisiae strains 00 YNB with 2% glucose (A) or with 3 % glycerol (B). 174 S. pombeYGK2000 S. pombe 972h- Figure 14. Growth of the S. pombe mip l: ura4 (YGK2000) mutant and its isogenetic wild-type S. pombe 972h- on YEPD agar plates containing 3% Nael. 175 + Growth phase ME ME ST NaCI + S. pombe mipl S. pombe ura4 A B Figure 15. Northern Blot analysis (A). S. pombe CBS 5682 was grown at 30°C in YEPD medium with or without 5% NaCI until Atioo reached approximately 0.5 (mid- exponential phase ME) or 3.0 (stationary phase ST). Total RNA was isolated and subjected to Northern hybridization analysis using S. pombe mipl and ura4 as probes. An ethidium blomide stained gel (B) was used to check the quality and quantity of RNA prior to Northern blotting. 176 REFERENCES Agre, P., Brown, D., and Nielsen, S. (1995). Aquaporin water channels: unanswered questions and unresolved controversies. Curro Opin. Cell. Bio/. 7,472-483. Agre, P., Bonhivers, M., and Borgnia, M. 1. (1998). The aquaporins, blueprints for cellular plumbing systems. J. Biol. Chern. 273, 14659-62. Aiba, H., Yamada, H, Ohmiya, R, and Mizuno, T. (1995). The osmo-induciblegpdl gene is a target of the signaling pathway involving Wis 1 MAP-kinase kinase in fission yeast. FEBS Lett. 376, 199-201. Aiba, H., Kawaura, R, Yamamoto, E., Yamada, H., Takegawa, K., and Mizuno, T. (1998). Isolation and characterization of the high-osmolarity senstive mutants of fission yeast. J. Bacterial. 180, 5038-5043 André, B. (1995). An overview of membrane transport proteins in Saccharomyces cerevisiae. Yeast 11, 1575-1611. Ansell, R, Granath, K., Hohmann, S. Thevelein, 1. M., and Adler, L. (1997). The two isozymes for yeast NAD-dependent glycerol 3-phosphate dehydrogenase encoded by GPDl and GPD2. have distinct roles in osmoadaption and redox regulation. £MBOJ. 16,2179-2187. Baker, M.E., and Saier, M.H, Jr. (1990). Common ancestor for bovine lens fibre major intrinsic protein, soy bean nodulin-26 protein, and E. coli glycerol facilitator. Cell 60, 185-186. Banuett, F. (1998). Signalling in the yeasts: an informational cascade with links to the filamentous fungi. Microbial. Mol. Bioi. Rev. 62, 249-274. Blomberg, A., and Adler L. (1992). Physiology of osmotolerance in fungi. Adv. Microbial. Physiol. 33, 145-212. 177 Bonhivers, M., Carbrey, J. M., Gould, S. J, and Agre, P. (1998). Aquaporins in Saccharomyces. Genetic and functional distinctions between laboratory and wild-type strains. J. Bioi. Chern. 273, 27565-27572. Booth,1. R, and Louis, P. (1999). Managing hypo osmotic stress: aquaporins and mechanosensitive channels in Escherichia coli. Curro Opin. Microbiol. 2, 166-169. Calamita, G., Bishai, W.R, Preson, G.M., Guggino, W.B., and Agre, P. (1995). Molecular cloning and characterization of AqpZ, a water channel from Escherichia coli. J. Bioi. Chern. 270,29063-29066. Calarnita, G., Kempf, B., Bonhivers, M., Bishai, W.R, Bremer, E., and Agre, P. (1998). Regulation of the Escherichia coli water channel gene aqp'Z. Proc. Natl. Acad. Sci. USA. 95,3627-3631. Del Sal, G., Manfioletti, G., and Schneider, C. (1988). A one-tube plasmid DNA minipreparation suitable for sequencing. Nuc/. Acids Res. 16, 9878. DeRisi, J.L., Vishwanath, RI., and Brown, P.O. (1997). Exploring the metabolic and genetic control of gene expression on a genomic scale. Science 278, 680-686. Flick, K.M., Shaulsky, G., and Loomis, W. F. (1997). The wacA gene of Dictyostelium discoideum is a developmentally regulated member of the MIP family. Gene 195, 127-30. Froger, A, Tallur, B., Thomas, D., and Delamarche, C. (1998). Prediction of functional residues in water channels and related proteins. Protein Sci. 7, 1458-1468. Gancedo, c., Gancedo, lM., and Sols, A (1968). Glycerol metabolism in yeasts. Pathways of utilisation and production. Eur. J. Biochem. 5, 165-172. Gietz, RD. and Schiestl, RH. (1995). Transforming yeast with DNA Methods Mol. Cell. Bioi .. 5, 255-269. 178 Gorin, M.B., Yancey, S.B., Cline, L, Revel., lP., and Horwitz, J. (1984). The major intrinsic protein (MIP) of the bovine lens fibre membrane: characterization and structure based on cDNA cloning. Ce1l39, 49-59. Grimm, C and Kohli, J. (1988). Observations on intergrative transformation in Schizosaccharomyces pombe. Mol. Gen. Genet. 215, 87-93. Grimm, c., Kohli, J and. Maundrell, K. (1988). Genetic engineering of Schizosacchromyces pombe: a system for gene disruption and replacement using the ura4 gene as a selectable marker. Mol. Gen. Genet. 215, 81-86. Hohmann, S., Bill, R, Kayingo, G., and Prior, B. A (2000). Microbial MIP channels. Trends Microbiol. 8, 33-38. Johansson, I.,Karlsson, M., Shukla, V. K., Chrispeels, M. L, Larsson, c., and Kjellbom, P. (1998). Water transport activity of the plasma membrane aquaporin PM28A is regulated by phosphorylation. Plant Cell 10, 451-459. Kaldenhoff, R., Kolling, A, Meyers, L, Karmann, u., Ruppel, G., Richter, G. (1995). The blue light-responsive AthE2 gene of Arabidopsis thaliana is primarily expressed in expanding as well as in differentiating cells and encodes a putative channel protein of the plasmalemma. Plant Physiol. 91,6269-73. Karlgren, S., Bill, R, and Hohmann, S. (2000). A test system to identify solutes transported by MIP channels. Proceedings of the 3rd International Conference on the Molecular Biology and Physiology of Water and Solute transport. Poster 104, page 144. Kayingo, G., Bill, R, Calamita, G., Hohmann, S., and Prior, B. A (2001). Microbial water and glycerol channels. Current Topics in Membranes, 51, 335-370. Kyte, L, and Doolittle, RF. (1982). A simple method for displaying the hydropathic character ofa protein. J. Mol. BioI. 157, 105-132. 179 Lages, F., Silva-Graca, M., and Lucas, C. (1999). Active glycerol uptake is a mechanism underlying halotolerance in yeasts: a study of 42 species. Microbiology 145, 2577-85. Laizé, V, Gobin, R., Rousselet, G., Badier, c., Hohmann, S., Ripoche, P., and Tacnet, F. (1999). Molecular and functional study of AQY1 from Saccharomyces cerevisiae. Role of the C-terminal domain. Biochem. Biophys. Res. Commun. 257, 139-144. Ludevid, D., Hëfte, H., Himelblau, E., Chrispeels, MJ. (1992). The expression pattern of the tonoplast intrinsic protein y-TIP in Arabidopsis tha/iana is correlated with cell enlargement. Plant Physiol. 100, 1633-9. Luyten, K., Albertyn, J, Skibbe, W.F., Prior, B.A, Ramos, J, Thevelein, JM., and Hohmann., S. (1995). Fps1, a yeast member of the MIP family of channel proteins, is a facilitator for glycerol uptake and efflux and is inactive under osmotic stress. £MBO J. 14, 1360-137l. Meyrial, v., and Tacnet, F., (2000). Yeast aquaporins: expression, localization, function. Proceedings of the 3rd International Conference on the Molecular Biology and Physiology of Water and Solute Transport. P100, page 140. Moreno, S., Klar, A, Nurse, P. (1991). Molecular genetic analysis of fission yeast Schizosaccharomyces pombe. Methods Enzymol. 194, 795-823. Ohmiya, R., Yamada, H., Nakashima, K., Aiba, H., and Mizuno, T. (1995). Osmoregulation of the fission yeast: Cloning of two distinct genes encoding glycerol-3-phosphate dehydrogenase, one of which is responsible for osmotolerance for growth. Mol. Microbiol. 18, 963-973 Park, JH., and Saier, M.H.Jr. (1996). Phylogenetic characterization of the MIP family of transmembrane channel proteins. J. Membrane Biol. 153, 171-180. 180 Pidoux AL., Fawell, E.H., and Armstrong, J. (1990). Glycerol-3-phosphate dehydrogenase homologue from Schizosaccharomyces pombe. Nucleic Acids Res. 18, 7145. Prudent, S., Charbonnier, M, and Marty, F. (2000). Functional complementation of the yeast Saccharomyces cerevisiae mutant SP1-L1fpsl by the plant aquaporin BobTIP26 gene. Proceedings of the 3rd International Conference on the Molecular Biology and Physiology of Water and Solute Transport. Poster 114, page 154. Rothstein, Rl (1983). One step gene disruption in yeast. Methods Enzymol. 101,202-21l. Schweizer, H. P., Jump, R, and Po, C. (1997). Structure and gene-polypeptide relationships of the region encoding glycerol diffusion facilitator (glpF) and glycerol kinase (glpK) of Pseudomonas aeruginosa. Microbiology 143,287-1297. Sherman, F., Fink, G.R. and Hicks, lB. (1986). Laboratory course manual for methods in yeast genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Sutherland, F. C. W., Lages, F., Lucas, c., Luyten, K., Albertyn, l, Hohmann, S., Prior, B. A and Kilian, S. G. (1997). Characteristics ofFPS l-dependent and - independent glycerol transport in Saccharomyces cerevisiae. J Bacteriol. 179, 7790- 7795. 'Sweet, G., Gandor, C., Voegele, R, Wittekindt, N., Beuerly, r., Truniger, v., Lin, C. E., and Boos, W. (1990). Glycerol facilitator of Escherichia coli: Cloning and of glpF and identificatio of the glpF product. J. Bacteriol. 172,424-430. Tamás, MJ., Luyten, K., Sutherland, W.C.F., Hernandez, A, Albertyn, l, Valadi, H, Hong Li, Prior, B.A, Killian, S.G., Ramos, l, Gustafsson, L., Thevelein, lM., and Hohmann, S. (1999). Fps1p Controls the accumulation and release of the compatible solute glycerol in yeast osmoregulation. Mol. Microbiol. 31, 1087-1104. 181 Tamás, MJ. (1999). The Saccharomyces cerevisiae glycerol channel protein Fps1p: Analysis of its function and regulation during osmotic stress. PhD thesis, Catholic University ofLeuven Thomas, B. land Rothstein, R. J. (1989). Elevated recombination rates in transcriptionally active DNA Cell 56, 619-630. Truniger, V., Boos, W., and Sweet, G. (1992). Molecular analysis oftheglpFKX regions of Escherichia coli and Shigellaflexneri. J Bacterial. 174,6981-6991. Van Aelst, L.,Hohmann, S., Zimmermann, F.K., Jans, AW., Thevelein, lM. (1991). A yeast homologue of the bovine lens fibre MIP gene family complements the growth defect of a Saccharomyces cerevisiae mutant on fermentable sugars but not its defect in glucose-induced RAS-mediated cAMP ·signalling. EMBO J. 10,2095-2104. Voegele, R. T., Sweet, G. D., and Boos, W. (1993). Glycerol kinase of Escherichia coli is activated by interaction-with the glycerol facilitator. J Bacterial. 1713, 1087-1094. Weig, R. A, and Jakob, C. (2000). Functional identification of the glycerol permease activity of Arabidapsis thaliana NLM1 and NLM2 proteins by heterologous expression in Saccharomyces cerevisiae. FEBS Le~. 481, 293-,298. Wilkinson M.G., Samuels, M. Takeda, T., Toone, W.M., Shieh, r.c, Toda, T., Millar, IRA, and Jones, N. (1996). The Atfl transcription factor is a target for the Styl stress-activated MAP kinase pathway in fission yeast. Genes Div.10, 2289-23Q1:' 182 CHAPTERS SUMMARY This study examined the responses of yeasts to hypo-osmotic stress with special emphasis to osmolyte export and facilitating proteins. All yeast strains studied (Z rouxii, P. sorbitophila, S. cerevisiae and S. pombe) rapidly release their intracellular osmolytes upon a decrease in external osmolarity (osmotic downshoek or hypo-osmotic shock). Osmolyte release is very rapid, specific and is not affected at reduced temperatures neither inhibited by the channel blocker gadolinium or the protonophore CCCP. The export process is well controlled and the amount of osmolyte released is proportional to the shock intensity. Osmolyte release occurs with minimal cell lysis and thus the survival as well as the subsequent growth of yeast cells is largely unaffected after hypo-osmotic shock. The patterns and export kinetics suggested the involvement of channel proteins similar to that of Fps1 p previously reported in S. cerevisiae. However, search for FPSl homologues from other yeasts using PCR and DNA probes resulted in weak hybridization signals suggesting that the putative glycerol channel encoding genes might have low sequence similarity to FPSl. It appears that although the mechanism of osmolyte release is conserved among yeasts, the proteins involved in this release might be divergent. This finding was in contrast to the general view that most of the genes involved in glycerol metabolism and stress responses such as GPDl (NAD+-dependent glycerol-3- phosphate dehydrogenase), DAKl (dihydroxyacetone kinase) and HOGl (MAP kinase of the HOG pathway) are well conserved in all yeasts. Isolation and cloning of the corresponding genees) involved in osmolyte export will shed more light on the molecular nature and physiological roles of these exporters. In yeast, osmolyte transport across the cell membrane occurs via active transport, mediated by channel proteins and by passive diffusion. The extent to which osmolytes permeates the cell membrane may then be influenced by the membrane lipid composition. In this study, the role of ergosterol (the most abundant sterol in yeast membranes) in osmolyte release and survival of yeast cells during hypo-osmotic stress was investigated. Cells lacking a glycerol facilitator (the JPsL:1 strain) grow very poorly upon an osmotic downs hock, but apparently survived the shock better and recovered more rapidly if ergosterol was supplied. Furthermore, the rate and amount of glycerol release was markedly enhanced in the JPsL1 mutant when exogenous ergosterol was supplied. The erg-l disruption mutant which is unable to synthesise ergosterol, survived and recovered from the osmotic shock more successfully at the higher ergosterol 183 concentration. Although the mechanism by which ergosterol improves glycerol release and survival of yeast cells is not well understood, it could presumably be related to the membrane stabilizing effects of ergosterol and the associated improvements in membrane fluidity. The polyene antibiotic nystatin, which affects membrane permeability in an ergosterol dependent way, caused S. cerevisiae cells to release a large amount of glycerol and equally inhibited the growth of wild-type and_hJsldeletion strains in medium containing 5% (w/v) Nael. This study demonstrated the role of ergosterol in glycerol efflux and survival in S. cerevisiae after an osmotic downshock and provided additional evidence for the significance of membrane permeability and glycerol con servation in yeast osmoregulation. The ability to regulate water and solute flux across cell membranes is critical in ensuring a constant turgor pressure as well as the proper functioning of biochemical processes. In most organisms, this process appears to be mediated by the MIP family transmembrane channel proteins, most of which have been characterized in higher animals. An in silicio phylogenetic analysis of microbial MIP channels revealed two major groups, the glycerol facilitators and the water channels (aquaporins), but further divided the glycerol facilitators into two subfamilies. Water channels seem to be important for growth after drastic changes in medium osmolarity, especially to lower osmolarity. Glycerol facilitators appear to exist in all microbial groups where they function in the uptake of glycerol and related compounds for their catabolism. The S. cerevisiae glycerol facilitator has been shown to be involved in osmoregulation by controlling the accumulation and release of glycerol. The occurrence of glycerol facilitators in other yeasts and their role in osmoregulation were investigated in this study. Blast searches in the S. pombe data bases revealed three putative glycerol transport proteins one of which shows considerably structural similarities to known MIP family glycerol facilitators. However, heterologous expression and subsequent functional analysis of this S. pombe mip l did not indicate its involvement in glycerol transport across the plasma membrane. It is still unknown whether the protein is involved in glycerol transport across other organelle membranes or whether it is involved in transport of a yet unidentified solute. The expression of S. pombe mip l is induced by osmotic stress suggesting a role in osmoregulation. However, deletion of S. pombe mip I does not cause any observable effects on growth of S. pombe cells during osmotic stress. Therefore, the physiological role of the S. pombe mipl as well as the actual transporter(s) controlling glycerol flux in S. pombe remains to be elucidated. 184 OPSOMMING Hierdie studie het die respons van giste op hipo-osmotiese spanning ondersoek, veral osmoliet transport en die fasiliterende proteïene. Die studie het getoon dat al die giste- bestudeer (Z.rouxii, P. sorbitophila, S. cerevisiae and S. pombe) intrasellulêre osmoliete vinnig vrystel na dat die eksterne osmolariteit verlaagis (afwaardse osmotiese skok of hipo-osmotiese skok). Die vrystelling van osmoliete vind vinnig en spesifiek plaas en word nie deur verlaagde temperatuur beinvloed. Verder word die proses nie deur die kanaalblokkeerder ganadolium of die protonofoor eeep geinhibeer nie. Die vrystelling van osmoliete word gereguleer sodat die hoeveelheid osmoliet wat vrygestel word korreleer met die verandering in eksterne osmolariteit. Omdat osmoliet vrysteling met minimale lise van selle plaasvind, word die oorlewing en groei van selle nie deur hipo-osmotiese skok geafekteer nie. Die kinetiese parameters van osmoliet vrystelling was soortgelyk aan die van die FPSl proteien van S. cerevisiae. Hibridisasie eksperimente (Pf'R en DNA peilers) het egter getoon dat daar 'n laë ooreenkoms tussen DNS volgordes van FPSl homoloë en 'n moontlike gliserol-kanaal-geen was. Dit blyk dus dat alhoewel die metodes om osmoliete vry te stel in giste gekonserveerd is dat die proteiene betrokke by die vrystelling verskil. Hierdie bevinding verskil van die algemeen aanvaarde opvatting dat die gene, betrokke by gliserol metabolisme en skok respons (bv. GPDl; NAD+-dependent glycerol-3-phosphate dehydrogenase, DAKl; dihydroxyacetone kinase and HOGl; MAP kinase of the HOG pathway), gekonserveerd is in giste. Isolasie en klonering van die geen wat wel betrokke by vrystelling van osmoliete is, sal meer lig werp op die molekulêre aard en fisiologiese rol van transport-proteiene. In giste word osmoliete opgeneem deur aktiewe transport deaur kanaal proteiene, asook passiewe diffusie oor die selmembraan. Die mate waar toe die selmembraan deurlaatbaar is, vir osmoliete, m~g bepaal word deur die lipied samestelling van die selmembraan. In hierdie studie is die moontlike rol van ergosterol (die mees algemene lipied in gis selmembrane) in die vrysteling van osmoliete tydens hipo-osmotiese spanning ondersoek. Selle sonder 'n gliserol fasiliteerder @sl.1 stain) het swak gegroei na 'n afwaardse osmotiese skok, maar indien ergosterol in die medium ingesluit is het sulke selle beter oorleef en vinniger herstel. Verder het die tempo waarteen gliserol deur diefpsl .1 stain vrygestel is merkbaar verhoog in die teenwoordigheid van eksterne ergosterol. Die erg-l disrupsie mutant, wat nie ergosterol kan sintetiseer nie, het ook beter oorleef en herstel in die teen woordigheid van eksterne ergosterol. Alhoewel die meganisme van bogenoemde effek van ergosterol op gliserol vrystelling onbekend IS. is dit waarskynlik a.g.v. die stabiliseering en verbetering is 185 selmembraan vloeibaarheid. Die antibiotikum nystatin, wat membraan deurlaatbaarheid op 'n ergosterol afhanklike wyse beinvloed, het veroorsaak dat S.cerevisiae selle groot hoeveelhede giserol vrystel en het dus die groei van wilde tipe ea fpsl á delesie mutante in 5% (w/v) Nael geinhibeer. Hierdie studie het getoon dat ergosterol 'n rol speel in die vrylating van gliserol uit selle asook in die oorlewing van S. cerevisiae na osmotiese skok. Dit verskaf ook verdere bewys vir die belangrikheid van membraam deurlaatbaarheid en gliserol bewaring in osmoregulering van giste. Dit was duidelik dat die vermoë van selle om die beweging van water en ander oplosmiddels oor selmembrane te beheer belangrik is in die handhawing van korrekte turgordruk en biochemiese funksionaliteit. In meeste organismes blyk dit dat die proses deur transmembraanproteiene van die MIP familie beheer word. 'n "in silico" filogenetiese analise van mikrobiese kanaalproteiene het getoon dat daar twee hoofgroepe IS nl. gliserolfasiliteerders en waterkanale (aquaporins). Die gliserol fasiliteerders was onderverdeel in twee groepe. Die waterkanale blyk belangrik te wees vir groei na verandering in medium osmolaritiet, veral verandering na laë osmolariteit. Gliserol fasiliteerdes kom waarskynlik in alle mikrobe organismes voor en speel 'n rol in die opname van gliserol en verbindings betrokke by gliserol katabolisme. Die gliserol fasiliteerder van S. cerevisiae beheer osmoregulering deurdat dit die opname of vrystel van gliserol kontroleer. Die voorkoms van gliserolfasiliteerders in ander giste en hulle rol in osmoregulering is in hierdie studie ondersoek. 'CSLAST" ondesoeke van DNS data van S. pombe het drie moontlike gliserol fasiliteerders van die MIP familie getoon. Heteroloë uitdruking en funksionele analises van S. pombe mip 1 kon egter geen bewys lewer dat dit betrokke is in gliserol transport oor die plasma membraan. Dus is dit nog onbekend of die proteïen betrokke is by gliserol transport oor ander organel membrane en of dit betrokke is by die transport van ander oplosmiddels. Uitdruking van S. pombe mipI word deur osmotiese spanning geinduseer, wat 'n rol daaarvoor tydens osmoregulering aandui. Delesie van S. pombe mip l het egter nie die groei van S. pombe tydens osmotiese spanning geaffekteer nie. Die fisiologiese rol van S. pombe mipl en die werklike transporter(s) wat gliserol opname/vrystelling beheer in S. pombe moet dus nog ontrafel word. D. ·.V•• IaHor