Detection of Protein-Protein Interactions Using Protein-Fragment Complementation Assays (PCA)

Protein-protein interactions play a central role in many cellular processes. Their characterisation is necessary in order to analyse these processes and for the functional identification of unknown proteins. Existing detection methods such as the yeast two-hybrid (Y2H) and tandem affinity purification (TAP) method provide a means to answer rapidly questions regarding protein-protein interactions, but have limitations which restrict their use to certain interaction networks; furthermore they provide little information regarding interaction localisation at the subcellular level. The development of protein-fragment complementation assays (PCA) employing a fluorescent reporter such as a member of the green fluorescent protein (GFP) family has led to a new method of interaction detection termed Bimolecular Fluorescent Complementation (BiFC). These assays have become important tools for understanding protein interactions and the development of whole genome interaction maps. BiFC assays have the advantages of very low background signal coupled with rapid detection of protein-protein interactions in vivo while also providing information regarding interaction compartmentalisation. Modified forms of the assay such as the use of combinations of spectral variants of GFP have allowed simultaneous visualisation of multiple competing interactions in vivo. Advantages and disadvantages of the method are discussed in the context of other fluorescence-based interaction monitoring techniques.

The IQGAP-myosin light chain interaction: what is its function?

The first members of the IQGAP family of proteins were
characterised over 15 years ago. It is now known that these molecules act
at the interface between cellular signalling pathways and the actin
cytoskeleton. They bind to a diverse range of signalling molecules –
including those involved in calcium, GTPase, kinase and growth factor
signalling. One intriguing interaction is that between mammalian
IQGAP1 and the myosin essential light chain isoform, Mlc1sa. Although
this has been demonstrated in vitro, its in vivo role is not known. Indeed,
it would be tempting to dismiss it as an experimental artefact, except for
the existence of a parallel interaction in the budding yeast,
Saccharomyces cerevisae. In this organism, the IQGAP-like protein
(Iqg1p) interacts with a myosin essential light chain (Mlc1p). This interaction is critical for the correct execution of cytokinesis. IQGAP-like
proteins also play key roles in cytokinesis in other fungi. Recent work
implicating mammalian IQGAP1 in cytokinesis may help explain the role
of the interaction in higher eukarytotes.

Ethanol-induced water stress in yeast

This review considers the effect of ethanol-induced water stress on yeast metabolism and integrity. Ethanol causes water stress by lowering water activity (a(w)) and thereby interferes with hydrogen bonding within and between hydrated cell components, ultimately disrupting enzyme and membrane strut and function. The impact of ethanol on the energetic status of water is considered in relation to cell metabolism. Even moderate ethanol concentrations (5 to 10%, w/v) cause a sufficient reduction of a(w) to have metabolic consequences. When exposed to ethanol, cells synthesize compatible solutes such as glycerol and trehalose that protect against water stress and hydrogen-bond disruption. Ethanol affects the control of gene expression by the mechanism that is normally associated with (so-called) osmotic control. Furthermore, ethanol-induced water stress has ecological implications.

Biological membranes : biophysical properties and aquaporin function

Tese de doutoramento, Farmácia (Bioquímica), Universidade de Lisboa, Faculdade de Farmácia, 2014

Spatiotemporal mechanisms for actomyosin ring assembly and contraction in budding yeast cell division

Dissertation presented to obtain the Ph.D degree in Molecular Medicine

Dissecting the function of the SpoIIIJ and YqjG membrane protein insertases during bacterial spore development

Dissertation presented to obtain the Ph.D degree in Biology, Microbial Biology

Contribution to drug discovery and development for tauopathies using yeast as a model

This work aimed to contribute to drug discovery and development (DDD) for tauopathies, while expanding our knowledge on this group of neurodegenerative disorders, including Alzheimer’s disease (AD). Using yeast, a recognized model for neurodegeneration studies, useful models were produced for the study of tau interaction with beta-amyloid (Aβ), both AD hallmark proteins. The characterization of these models suggests that these proteins co-localize and that Aβ1-42, which is toxic to yeast, is involved in tau40 phosphorylation (Ser396/404) via the GSK-3β yeast orthologue, whereas tau seems to facilitate Aβ1-42 oligomerization. The mapping of tau’s interactome in yeast, achieved with a tau toxicity enhancer screen using the yeast deletion collection, provided a novel framework, composed of 31 genes, to identify new mechanisms associated with tau pathology, as well as to identify new drug targets or biomarkers. This genomic screen also allowed to select the yeast strain mir1Δ-tau40 for development of a new GPSD2TM drug discovery screening system. A library of unique 138 marine bacteria extracts, obtained from the Mid-Atlantic Ridge hydrothermal vents, was screened with mir1Δ-tau40. Three extracts were identified as suppressors of tau toxicity and constitute good starting points for DDD programs. mir1Δ strain was sensitive to tau toxicity, relating tau pathology with mitochondrial function. SLC25A3, the human homologue of MIR1, codes for the mitochondrial phosphate carrier protein (PiC). Resorting to iRNA, SLC25A3 expression was silenced in human neuroglioma cells, as a first step towards the engineering of a neural model for replicating the results obtained in yeast. This model is essential to understand the mechanisms of tau toxicity at the mitochondrial level and to validate PiC as a relevant drug target. The set of DDD tools here presented will foster the development of innovative and efficacious therapies, urgently needed to cope with tau-related disorders of high human and social-economic impact.

The protein phosphatase 7 regulates phytochrome signaling in Arabidopsis.

The psi2 mutant of Arabidopsis displays amplification of the responses controlled by the red/far red light photoreceptors phytochrome A (phyA) and phytochrome B (phyB) but no apparent defect in blue light perception. We found that loss-of-function alleles of the protein phosphatase 7 (AtPP7) are responsible for the light hypersensitivity in psi2 demonstrating that AtPP7 controls the levels of phytochrome signaling. Plants expressing reduced levels of AtPP7 mRNA display reduced blue-light induced cryptochrome signaling but no noticeable deficiency in phytochrome signaling. Our genetic analysis suggests that phytochrome signaling is enhanced in the AtPP7 loss of function alleles, including in blue light, which masks the reduced cryptochrome signaling. AtPP7 has been found to interact both in yeast and in planta assays with nucleotide-diphosphate kinase 2 (NDPK2), a positive regulator of phytochrome signals. Analysis of ndpk2-psi2 double mutants suggests that NDPK2 plays a critical role in the AtPP7 regulation of the phytochrome pathway and identifies NDPK2 as an upstream element involved in the modulation of the salicylic acid (SA)-dependent defense pathway by light. Thus, cryptochrome- and phytochrome-specific light signals synchronously control their relative contribution to the regulation of plant development. Interestingly, PP7 and NDPK are also components of animal light signaling systems.

Functional differentiation of tbf1 orthologues in fission and budding yeasts.

In Saccharomyces cerevisiae, TBF1, an essential gene, influences telomere function but also has other roles in the global regulation of transcription. We have identified a new member of the tbf1 gene family in the mammalian pathogen Pneumocystis carinii. We demonstrate by transspecies complementation that its ectopic expression can provide the essential functions of Schizosaccharomyces pombe tbf1 but that there is no rescue between fission and budding yeast orthologues. Our findings indicate that an essential function of this family of proteins has diverged in the budding and fission yeasts and suggest that effects on telomere length or structure are not the primary cause of inviability in S. pombe tbf1 null strains.

Metabolite changes during Riesling icewine fermentation when yeast micronutrients are used /

Icewine is an intensely s\veet dessert \vine fermented from the juice of naturally frozen grapes. Icewine fermentation poses many challenges such as failure to reach desired ethanol levels and production of high levels of volatile acidity in the fonn of acetic acid. This study investigated the impact of micronutrient addition (GO-FERM® and NATSTEP®) during the rehydration stage of the commercial \vine yeast Saccharomyces cerevisiae KI-VIII6 during Ice\vine fermentation. Sterile-filtered and unfiltered Riesling Ice\vine juice was inoculated \vith yeast rehydrated under four different conditions: in water only; with GO-FERM®; with NATSTEP®; or the combination of both micronutrient products in the rehydration water. Using sterile-filtered Icewine juice, yeast rehydration had a positive impact of reducing the rate of acetic acid produced as a function of sugar consumed, reducing the ratio of acetic acid/ethanol and reducing the ratio of acetic acid/glycerol. In the sterile-filtered fermentation, yeast rehydrated with micronutrients generated 9-times less acetic acid per gram of sugar in the first 48 hours compared to yeast rehydrated only \vith water and resulted in a 17% reduction in acetic acid in the final \vine \vhen normalized to sugar consumed. However, the sterile-filtered fermentations likely became stuck due to the overc1arification of the juice as evidenced from the low sugar consumption (117 gIL) that could not be completely overcome by the micronutrient treatments (144 gIL sugar consumed) to reach a target ethanol of IO%v/v. Contrary to \vhat \vas observed in the sterile-filtered treatements, using unfiltered Ice\vine juice, yeast micronutrient addition had no significant impact of reducing the rate of acetic acid produced as a function of sugar consumed, reducing the ratio of acetic acid/ethanol and reducing the ratio of acetic acid/glycerol. However, in the unfiltered fermentation, micronutrient addition during yeast rehydration caused a reduction in the acetic acid produced as a function of sugar consumed up to 150 giL sugar consumed.. In contrast to the sterile-filtered fermentations, the unfiltered fermentations did not become stuck as evidenced from the higher sugar consumption (l47-174g1L). The largest effects of micronutrient addition are evident in the first two days of both sterile and unfiltered fermentations.

Significance of DNA base excision repair in the maintenance of mitochondrial function in Saccharomyces cerevisiae /

Mitochondria have an important role in cell metabolism, being the major site of ATP production via oxidative phosphorylation (OXPHOS). Accumulation of mtDNA mutations have been linked to the development of respiratory dysfunction, apoptosis, and aging. Base excision repair (BER) is the major and the only certain repair pathway existing in mitochondria that is in responsible for removing and repairing various base modifications as well as abasic sites (AP sites). In this research, Saccharomyces cerevisiae (S. cerevisiae) BER gene knockout strains, including 3 single DNA glycosylase gene knockout strains and Ap endonuclease (Apn 1 p) knockout strain were used to examine the importance of this DNA repair pathway to the maintenance of respiratory function. Here, I show that individual DNA glycosylases are nonessential in maintenance of normal function in yeast mitochondria, corroborating with previous research in mammalian experimental models. The yeast strain lacking Apn 1 p activity exhibits respiratory deficits, including inefficient and significantly low intracellular ATP level, which maybe due to partial uncoupling of OXPHOS. Growth of this yeast strain on respiratory medium is inhibited, but no evidence was found for increased ROS level in Apn 1 p mitochondria. This strain also shows an increased cell size, and this observation combined with an uncoupled OXPHOS may indicate a premature aging in the Apnlp knockout strain, but more evidence is needed to support this hypothesis. However, the BER is necessary for maintenance of mitochondrial function in respiring S.cerevisiae.

Aspects of the hemes and modulation of hydrogen donors in catalases from bovine liver, yeast, and escherichia coli

Catalase is the enzyme which decomposes hydrogen peroxide to water and oxygen. Escherichia coli contains two catalases. Hydroperoxidase I (HPI) is a bifunctional catalase-peroxidase. Hydroperoxidase II (HPII) is only catalytically active toward H202. Expression of the genes encoding these proteins is controlled by different regimes. HPJI is thought to be a hexamer, having one heme d cis group per enzymatic subunit. HPII wild type protein and heme containing mutant proteins were obtained from the laboratory of P. Loewen (Univ. of Manitoba). Mutants constructed by oligonucleotidedirected mutagenesis were targeted for replacement of either the His128 residue or the Asn201 residue in the vicinity of the HPII heme crevice. His128 is the residue thought to be analogous to the His74 distal axial ligand of the heme in the bovine liver enzyme, and Asn201 is believed to be a residue critical to the function of the enzyme because of its role in orienting and interacting with the substrate molecule. Investigation of the nature of the hemes via absorption spectroscopy of the unmodified catalase proteins and their derived pyridine hemochromes showed that while the bovine and Saccharomyces cerevisiae catalase enzymes are protoheme-containing, the HPII wild type protein contains heme d, and the mutant proteins contain either solely protoheme, or heme d-protoheme mixtures. Cyanide binding studies supported this, as ligand binding was monophasic for the bovine, Saccharomyces cerevisiae, and wild type HPII enzymes, but biphasic for several of the HPII mutant proteins. Several mammalian catalases, and at least two prokaryotic catalases, are known to be NADPH binding. The function of this cofactor appears to be the prevention of inactivation of the enzyme, which occurs via formation of the inactive secondary catalase peroxide compound (compound II). No physiologically plausible scheme has yet been proposed for the NADPH mediation of catalase activity. This study has shown, via fluorescence and affinity chromatography techniques, that NADPH binds to the T (Typical) and A (Atypical) catalases of Saccharomyces cerevisiae, and that wild type HPII apparently does not bind NADPH. This study has also shown that NADPH is unlike any other hydrogen donor to catalase, and addresses its features as a unique donor by proposing a mechanism whereby NADPH is oxidized and catalase is protected from inactivation via the formation of protein radical species. Migration of this radical to a position close to the NADPH is also proposed as an adjunct hypothesis, based on similar electron migrations that are known to occur within metmyoglobin and cytochrome c peroxidase when reacted with H202. Validation of these hypotheses may be obtained in appropriate future experiments.

Study of new yeast strains as novel starter cultures for Riesling icewine production

Icewine is a sweet dessert wine fermented from the juice of grapes naturally frozen on the vine. The production of Icewine faces many challenges such as sluggish fermentation, which often yields wines with low ethanol, and an accumulation of high concentration of volatile acidity, mainly in the form of acetic acid. This project investigated three new yeast strains as novel starter cultures for Icewine fermentation with particular emphasis on reducing acetic acid production: a naturally occurring strain of S. bayanus/S. pastorianus isolated from Icewine grapes, and two hybrids between S. cerevisiae and S. bayanus, AWRI 1571 and AWRI 1572. These strains were evaluated for sugar consumption patterns and metabolic production of ethanol, glycerol and acetic acid, and were compared to the performance of a standard commercial wine yeast KI-VI116. The ITS rONA region of the two A WRI crosses was also analyzed during fermentations to assess their genomic stability. Icewine fermentations were performed in sterile filtered juice, in the absence of indigenous microflora, and also in unfiltered juice in order to mirror commercial wine making practices. The hybrid A WRI 1572 was found to be a promising candidate as a novel starter culture for Icewine production. I t produced 10.3 % v/v of ethanol in sterile Riesling Icewine fermentations and 11.2 % v/v in the unfiltered ones within a reasonable fermentation time (39 days). Its acetic acid production per gram sugar consumed was approximately 30% lower in comparison with commercial wine yeast K I -V 1116 under both sterile filtered and unfiltered fermentations. The natural isolate S. bayanus/S. pastorianus and AWRI 1571 did not appear to be suitable for commercial Icewine production. They reached the target ethanol concentration of approximately 10 % v/v in 39 day fermentations and also produced less acetic acid as a function of both time and sugar consumed in sterile fermentations compared to KI-V1116. However, in unfiltered fermentations, both of them failed to produce the target concentration of ethanol and accumulated high concentration of acetic acid. Both A WRI crosses displayed higher loss of or reduced copies in ITS rDNA region from the S. bayanus parent compared to the S. cerevisiae parent; however, these genomic losses could not be related to the metabolic profile.

A mechanism for co-transcriptional recruitment of mRNA localization factor on nascent mRNAs in budding yeast

Le transport et la localisation des ARN messagers permettent de réguler l’expression spatiale et temporelle de facteurs spécifiques impliqués dans la détermination du destin cellulaire, la plasticité synaptique, la polarité cellulaire et la division asymétrique des cellules. Chez S.cerevisiæ, plus de trente transcrits sont transportés activement vers le bourgeon cellulaire. Parmi ces transcrits, l’ARNm ASH1 (asymetric synthesis of HO) est localisé à l’extrémité du bourgeon pendant l’anaphase. Ce processus va entrainer une localisation asymétrique de la protéine Ash1p, qui sera importée uniquement dans le noyau de la cellule fille, où elle entraine le changement de type sexuel. La localisation asymétrique de l’ARNm ASH1, et donc de Ash1p, implique la présence de différents facteurs de localisation. Parmi ces facteurs, les protéines She (She1p/Myo4p, She2p et She3p) et les répresseurs traductionnels (Puf6p, Loc1p et Khd1p) participent à ce mécanisme. La protéine navette She2p est capable de lier l’ARNm ASH1 et va entrainer le ciblage de cet ARNm vers l’extrémité du bourgeon en recrutant le complexe She3p-Myo4p. Des répresseurs traductionnels régulent la traduction de cet ARNm et évitent l’expression ectopique de la protéine Ash1p pendant son transport. Alors que la fonction cytoplasmique de She2p sur la localisation des ARNm est connue, sa fonction nucléaire est encore inconnue. Nous avons montré que She2p contient une séquence de localisation nucléaire non classique qui est essentielle à son import nucléaire médié par l’importine α (Srp1p). L’exclusion de She2p du noyau par mutation de son NLS empêche la liaison de Loc1p et Puf6p sur l’ARNm ASH1, entrainant un défaut de localisation de l’ARNm et de la protéine. Pour étudier plus en détail l’assemblage de la machinerie de localisation des ARNm dans le noyau, nous avons utilisé des techniques d’immunoprécipitation de chromatine afin de suivre le recrutement des facteurs de localisation et des répresseurs traductionnels sur les ARNm naissants. Nous avons montré que She2p est recruté sur le gène ASH1 pendant sa transcription, via son interaction avec l’ARNm ASH1 naissant. Puf6p est également recruté sur ASH1, mais d’une manière dépendante de la présence de She2p. De façon intéressante, nous avons détecté une interaction entre She2p et la plus grande sous-unité de l’ARN polymérase II (Rpb1p). Cette interaction est détectée avec la forme active en élongation de l’ARN polymérase II. Nous avons également démontré que She2p interagit avec le complexe d’élongation de la transcription Spt4p/Spt5p. Une délétion de SPT4 ou une mutation dans SPT5 (Ts spt5) à température restrictive empêche l’interaction entre She2p et Rpb1p, et diminue le recrutement de She2p au gène ASH1, entrainant un défaut de localisation de l’ARNm et un défaut de localisation asymétrique de la protéine Ash1p. De manière globale, nos résultats montrent que les facteurs impliqués dans la localisation cytoplasmique des ARNm et dans leur contrôle traductionnel sont recrutés de façon co-transcriptionnelle sur les ARNm naissants via leur interaction avec la machinerie de transcription, suggèrant un rôle important de la machinerie transcriptionelle dans la localisation des ARNm.

Investigating the role of the isomerase Rrd1/PTPA : from yeast to human

Chez Saccharomyces cerevisiae, les souches mutantes pour Rrd1, une protéine qui possède une activité de peptidyl prolyl cis/trans isomérase, montrent une résistance marquée à la rapamycine et sont sensibles au 4-nitroquinoline 1-oxide, un agent causant des dommages à l’ADN. PTPA, l’homologue de Rrd1 chez les mammifères, est reconnu en tant qu’activateur de protéine phosphatase 2A. Notre laboratoire a précédemment démontré que la surexpression de PTPA mène à l’apoptose de façon indépendante des protéines phosphatase 2A. La fonction moléculaire de Rrd1/PTPA était encore largement inconnue au départ de mon projet de doctorat. Mes recherches ont d’abord montré que Rrd1 est associé à la chromatine ainsi qu’à l’ARN polymérase II. L’analyse in vitro et in vivo par dichroïsme circulaire a révélé que Rrd1 est responsable de changements au niveau de la structure du domaine C-terminal de la grande sous-unité de l’ARN polymérase II, Rpb1, en réponse à la rapamycine et au 4-nitroquinoline 1-oxide. Nous avons également démontré que Rrd1 est requis pour modifier l’occupation de l’ARN polymérase II sur des gènes répondant à un traitement à la rapamycine. Finalement, nous avons montré que suite à un traitement avec la rapamycine, Rrd1 médie la dégradation de l’ARN polymérase II et que ce mécanisme est indépendant de l’ubiquitine. La dernière partie de mon projet était d’acquérir une meilleure connaissance de la fonction de PTPA, l’homologue de Rrd1 chez les mammifères. Nos résultats montrent que le «knockdown» de PTPA n’affecte pas la sensibilité des cellules à différentes drogues telles que la rapamycine, le 4-nitroquinoline 1-oxide ou le peroxyde d’hydrogène (H2O2). Nous avons également tenté d’identifier des partenaires protéiques pour PTPA grâce à la méthode TAP, mais nous ne sommes pas parvenus à identifier de partenaires stables. Nous avons démontré que la surexpression de la protéine PTPA catalytiquement inactive n’induisait pas l’apoptose indiquant que l’activité de PTPA est requise pour produire cet effet. Finalement, nous avons tenté d’étudier PTPA dans un modèle de souris. Dans un premier lieu, nous avons déterminé que PTPA était exprimé surtout au niveau des tissus suivants : la moelle osseuse, le thymus et le cerveau. Nous avons également généré avec succès plusieurs souris chimères dans le but de créer une souris «knockout» pour PTPA, mais l’allèle mutante ne s’est pas transférée au niveau des cellules germinales. Mes résultats ainsi que ceux obtenus par mon laboratoire sur la levure suggèrent un rôle général pour Rrd1 au niveau de la régulation des gènes. La question demeure toujours toutefois à savoir si PTPA peut effectuer un rôle similaire chez les mammifères et une vision différente pour déterminer la fonction de cette protéine sera requise pour adresser adéquatement cette question dans le futur.