Model Selection for Solving Kinematics Problems

There has been much interest in the area of model-based reasoning within the Artificial Intelligence community, particularly in its application to diagnosis and troubleshooting. The core issue in this thesis, simply put, is, model-based reasoning is fine, but whence the model? Where do the models come from? How do we know we have the right models? What does the right model mean anyway? Our work has three major components. The first component deals with how we determine whether a piece of information is relevant to solving a problem. We have three ways of determining relevance: derivational, situational and an order-of-magnitude reasoning process. The second component deals with the defining and building of models for solving problems. We identify these models, determine what we need to know about them, and importantly, determine when they are appropriate. Currently, the system has a collection of four basic models and two hybrid models. This collection of models has been successfully tested on a set of fifteen simple kinematics problems. The third major component of our work deals with how the models are selected.

The identification and characterisation of Rab4 interacting proteins

Rab4 is a member of the Rab superfamily of small GTPases. It is localized to the early sorting endosome and plays a role in regulating the transport from this compartment to the recycling and degradative pathways. In order to further our understanding of the role Rab4 plays in endocytosis, a yeast two-hybrid screen was performed to identify putative Rab4 effectors. A constitutively active mutant of Rab4, Rab4Q67L, when used as bait to screen a HeLa cDNA library, identified a novel 80kDa protein that interacted with Rab4-GTP. This protein was called Rab Coupling Protein (RCP). RCP interacts preferentially with the GTP-bound form of Rab4. Subsequent work demonstrated that RCP also interacts with Rab11, and that this interaction is not nucleotide-depenedent. RCP is predominantly membrane-bound and localised to the perinuclear recycling compartment. Expression of a truncation mutant of RCP, that contains the Rab binding domain, in HeLa cells, results in the formation of an extensive tubular network that can be labelled with transferrin. These tubules are derived from the recycling compartment since they are inaccessible to transferrin when the ligand is internalised at 18oC. The truncation mutant-induced morphology can be rescued by overexpression of active Rab11, but not active Rab4. This suggests that RCP functions between Rab4 and Rab11 in the receptor recycling pathway, and may act as a ‘molecular bridge’ between these two sequentially acting small GTPases. Quantitative assays demonstrated that overexpression of the truncation mutant results in a dramatic inhibition in the rate of receptor recycling. Database analysis revealed that RCP belongs to a family of Rab interacting proteins, each characterised by a carboxy-terminal coiled-coil domain and an amino-terminal phospholipid-binding domain. KIAA0941, an RCP homologue, interacts with Rab11, but not with Rab4. Overexpression of its Rab binding domain also results in a tubular network, however, this tubulation cannot be rescued by active Rab11.

Characterisation of the non-muscle α-actinins

Actinins are cytoskeleton proteins that cross-link actin filaments. Evolution of the actinin family resulted in the formation of Ca++-insensitive muscle isoforms (actinin-2 and- 3) and Ca++-sensitive non-muscle isoforms (actinin-1 and -4) with regard to their actin-binding function. Despite high sequence similarity, unique properties have been ascribed to actinin-4 compared with actinin-1. Actinin-4 is the predominant isoform reported to be associated with the cancer phenotype. Actinin-4, but not actinin-1, is essential for normal glomerular function in the kidney and and is able to translocate to the nucleus to regulate transcription. To understand the molecular basis for such isoform-specific functions I have comprehensively compared these proteins in terms of localisation, migration, alternative splicing, actin-binding properties, heterodimer formation and molecular interactions for the first time. This work characterises a number of commercially available actinin antibodies and in doing so, identifies actinin-1, -2 and -4 isoform-specific antibodies that enabled studies of actinin expression and localisation. This work identifies the actinin rod domain as the predominant domain that influences actinin localisation however localisation is likely to be effected by the entire actinin protein. si-RNA- mediated knockdown of actinin-1 and -4 did not affect migration in a number of cell lines highlighting that migration may only require a fraction of total non-muscle actinin levels. This work finds that the Ca++-insensitive variant of actinin-4 is expressed only in the nervous system and thus cannot be regarded as a smooth muscle isoform, as is the case for the Ca++-insensitive variant of actinin-1. This work also identifies a previously unreported exon 19a+19b expressing variant of actinin-4 in human skeletal muscle. This work finds that alternative splice variants of actinin-1 and -4 are co-expressed in a number of tissues, in particular the brain. In contrast to healthy brain, glioblastoma cells express Ca++-sensitive variants of both actinin-1 and -4. Actin-binding properties of actinin-1 and -4 are similar and are unlikely to explain isoform-specific functions. Surprisingly, this work reveals that actinin-1/-4 heterodimers, rather than homodimers, are the most abundant form of actinin in many cancer cell lines. Taken together this data suggests that actinin-1 and -4 cannot be viewed as distinct entities from each other but rather as proteins that can exist in both homodimeric and heterodimeric forms. Finally, this work employs yeast two-hybrid and proteomic approaches to identify actinin-interacting proteins. In doing so, this work identifies a number of putative actinin-4 specific interacting partners that may help to explain some of the unique functions attributed the actinin-4. The observation of alternative splice variants of actinin-1 and -4 combined with the observed potential of these proteins to form homodimers and heterodimers suggests that homodimers and heterodimers with novel actin-binding properties and interaction networks may exist. The ability to behave in this manner may have functional implications. This may be of importance considering that these proteins are central to such processes as cell migration and adhesion. This significantly alters our view of the non-muscle actinins.

Cytokinesis proteins Tum and Pav have a nuclear role in Wnt regulation.

Wg/Wnt signals specify cell fates in both invertebrate and vertebrate embryos and maintain stem-cell populations in many adult tissues. Deregulation of the Wnt pathway can transform cells to a proliferative fate, leading to cancer. We have discovered that two Drosophila proteins that are crucial for cytokinesis have a second, largely independent, role in restricting activity of the Wnt pathway. The fly homolog of RacGAP1, Tumbleweed (Tum)/RacGAP50C, and its binding partner, the kinesin-like protein Pavarotti (Pav), negatively regulate Wnt activity in fly embryos and in cultured mammalian cells. Unlike many known regulators of the Wnt pathway, these molecules do not affect stabilization of Arm/beta-catenin (betacat), the principal effector molecule in Wnt signal transduction. Rather, they appear to act downstream of betacat stabilization to control target-gene transcription. Both Tum and Pav accumulate in the nuclei of interphase cells, a location that is spatially distinct from their cleavage-furrow localization during cytokinesis. We show that this nuclear localization is essential for their role in Wnt regulation. Thus, we have identified two modulators of the Wnt pathway that have shared functions in cell division, which hints at a possible link between cytokinesis and Wnt activity during tumorigenesis.

Identification of the endophilins (SH3p4/p8/p13) as novel binding partners for the beta1-adrenergic receptor.

Several G-protein coupled receptors, such as the beta1-adrenergic receptor (beta1-AR), contain polyproline motifs within their intracellular domains. Such motifs in other proteins are known to mediate protein-protein interactions such as with Src homology (SH)3 domains. Accordingly, we used the proline-rich third intracellular loop of the beta1-AR either as a glutathione S-transferase fusion protein in biochemical "pull-down" assays or as bait in the yeast two-hybrid system to search for interacting proteins. Both approaches identified SH3p4/p8/p13 (also referred to as endophilin 1/2/3), a SH3 domain-containing protein family, as binding partners for the beta1-AR. In vitro and in human embryonic kidney (HEK) 293 cells, SH3p4 specifically binds to the third intracellular loop of the beta1-AR but not to that of the beta2-AR. Moreover, this interaction is mediated by the C-terminal SH3 domain of SH3p4. Functionally, overexpression of SH3p4 promotes agonist-induced internalization and modestly decreases the Gs coupling efficacy of beta1-ARs in HEK293 cells while having no effect on beta2-ARs. Thus, our studies demonstrate a role of the SH3p4/p8/p13 protein family in beta1-AR signaling and suggest that interaction between proline-rich motifs and SH3-containing proteins may represent a previously underappreciated aspect of G-protein coupled receptor signaling.

The Function and Regulation of Photobodies in Phytochrome Signaling

Light is a critical environmental signal that regulates every phase of the plant life cycle, from germination to floral initiation. Of the many light receptors in the model plant Arabidopsis thaliana, the red- and far-red light-sensing phytochromes (phys) are arguably the best studied, but the earliest events in the phy signaling pathway remain poorly understood. One of the earliest phy signaling events is the translocation of photoactivated phys from the cytoplasm to the nucleus, where they localize to subnuclear foci termed photobodies; in continuous light, photobody localization correlates closely with the light-dependent inhibition of embryonic stem growth. Despite a growing body of evidence supporting the biological significance of photobodies in light signaling, photobodies have also been shown to be dispensable for seedling growth inhibition in continuous light, so their physiological importance remains controversial; additionally, the molecular components that are required for phy localization to photobodies are largely unknown. The overall goal of my dissertation research was to gain insight into the early steps of phy signaling by further defining the role of photobodies in this process and identifying additional intragenic and extragenic requirements for phy localization to photobodies.

Even though the domain structure of phys has been extensively studied, not all of the intramolecular requirements for phy localization to photobodies are known. Previous studies have shown that the entire C-terminus of phys is both necessary and sufficient for their localization to photobodies. However, the importance of the individual subdomains of the C-terminus is still unclear. For example a truncation lacking part of the most C-terminal domain, the histidine kinase-related domain (HKRD), can still localize to small photobodies in the light and behaves like a weak allele. However, a point mutation within the HKRD renders the entire molecule completely inactive. To resolve this discrepancy, I explored the hypothesis that this point mutation might impair the dimerization of the HKRD; dimerization has been shown to occur via the C-terminus of phy and is required for more efficient signaling. I show that this point mutation impairs nuclear localization of phy as well as its subnuclear localization to photobodies. Additionally, yeast-two-hybrid analysis shows that the wild-type HKRD can homodimerize but that the HKRD containing the point mutation fails to dimerize with both itself and with wild-type HKRD. These results demonstrate that dimerization of the HKRD is required for both nuclear and photobody localization of phy.

Studies of seedlings grown in diurnal conditions show that photoactivated phy can persist into darkness to repress seedling growth; a seedling's growth rate is therefore fastest at the end of the night. To test the idea that photobodies could be involved in regulating seedling growth in the dark, I compared the growth of two transgenic Arabidopsis lines, one in which phy can localize to photobodies (PBG), and one in which it cannot (NGB). Despite these differences in photobody morphology, both lines are capable of transducing light signals and inhibiting seedling growth in continuous light. After the transition from red light to darkness, the PBG line was able to repress seedling growth, as well as the accumulation of the growth-promoting, light-labile transcription factor PHYTOCHROME INTERACTING FACTOR 3 (PIF3), for eighteen hours, and this correlated perfectly with the presence of photobodies. Reducing the amount of active phy by either reducing the light intensity or adding a phy-inactivating far-red pulse prior to darkness led to faster accumulation of PIF3 and earlier seedling growth. In contrast, the NGB line accumulated PIF3 even in the light, and seedling growth was only repressed for six hours; this behavior was similar in NGB regardless of the light treatment. These results suggest that photobodies are required for the degradation of PIF3 and for the prolonged stabilization of active phy in darkness. They also support the hypothesis that photobody localization of phys could serve as an instructive cue during the light-to-dark transition, thereby fine-tuning light-dependent responses in darkness.

In addition to determining an intragenic requirement for photobody localization and further exploring the significance of photobodies in phy signaling, I wanted to identify extragenic regulators of photobody localization. A recent study identified one such factor, HEMERA (HMR); hmr mutants do not form large photobodies, and they are tall and albino in the light. To identify other components in the HMR-mediated branch of the phy signaling pathway, I performed a forward genetic screen for suppressors of a weak hmr allele. Surprisingly, the first three mutants isolated from the screen were alleles of the same novel gene, SON OF HEMERA (SOH). The soh mutations rescue all of the phenotypes associated with the weak hmr allele, and they do so in an allele-specific manner, suggesting a direct interaction between SOH and HMR. Null soh alleles, which were isolated in an independent, tall, albino screen, are defective in photobody localization, demonstrating that SOH is an extragenic regulator of phy localization to photobodies that works in the same genetic pathway as HMR.

In this work, I show that dimerization of the HKRD is required for both the nuclear and photobody localization of phy. I also demonstrate a tight correlation between photobody localization and PIF3 degradation, further establishing the significance of photobodies in phy signaling. Finally, I identify a novel gene, SON OF HEMERA, whose product is necessary for phy localization to photobodies in the light, thereby isolating a new extragenic determinant of photobody localization. These results are among the first to focus exclusively on one of the earliest cellular responses to light - photobody localization of phys - and they promise to open up new avenues into the study of a poorly understood facet of the phy signaling pathway.

Cyclin-dependent kinase 11(p58) interacts with HBO1 and enhances its histone acetyltransferase activity.

CDK11(p58), a 58kDa protein of the PITSLRE kinase family, plays an important role in cell cycle progression, and is closely related to cell apoptosis. To gain further insight into the function of CDK11(p58), we screened a human fetal liver cDNA library for its interacting proteins using the yeast two-hybrid system. Here we report that histone acetyltransferase (HAT) HBO1, a MYST family protein, interacts with CDK11(p58) in vitro and in vivo. CDK11(p58) and HBO1 colocalize in the cell nucleus. Recombinant CDK11(p58) enhances the HAT activity of HBO1 significantly in vitro. Meanwhile, overexpression of CDK11(p58) in mammalian cells leads to the enhanced HAT activity of HBO1 towards free histones. Thus, we conclude that CDK11(p58) is a new interacting protein and a novel regulator of HBO1. Both of the proteins may be involved in the regulation of eukaryotic transcription.

The C-terminal kinase domain of the p34cdc2-related PITSLRE protein kinase (p110C) associates with p21-activated kinase 1 and inhibits its activity during anoikis.

The PITSLRE protein kinases are parts of the large family of p34cdc2-related kinases. During apoptosis induced by some stimuli, specific PITSLRE isoforms are cleaved by caspase to produce a protein that contains the C-terminal kinase domain of the PITSLRE proteins (p110C). The p110C induces apoptosis when it is ectopically expressed in Chinese hamster ovary cells. In our study, similar induction of this p110C was observed during anoikis in NIH3T3 cells. To investigate the molecular mechanism of apoptosis mediated by p110C, we used the yeast two-hybrid system to screen a human fetal liver cDNA library and identified p21-activated kinase 1 (PAK1) as an interacting partner of p110C. The association of p110C with PAK1 was further confirmed by in vitro binding assay, in vivo coimmunoprecipitation, and confocal microscope analysis. The interaction of p110C with PAK1 occurred within the residues 210-332 of PAK1. Neither association between p58PITSLRE or p110PITSLRE and PAK1 nor association between p110C and PAK2 or PAK3 was observed. Anoikis was increased and PAK1 activity was inhibited when NIH3T3 cells were transfected with p110C. Furthermore, the binding of p110C with PAK1 and inhibition of PAK1 activity were also observed during anoikis. Taken together, these data suggested that PAK1 might participate in the apoptotic pathway mediated by p110C.

Hybridization of a class of "second order" models of traffic flow

Despite the simultaneous progress of traffic modelling both on the macroscopic and microscopic front, recent works [E. Bourrel, J.B. Lessort, Mixing micro and macro representation of traffic flow: a hybrid model based on the LWR theory, Transport. Res. Rec. 1852 (2003) 193–200; D. Helbing, M. Treiber, Critical discussion of “synchronized flow”, Coop. Transport. Dyn. 1 (2002) 2.1–2.24; A. Hennecke, M. Treiber, D. Helbing, Macroscopic simulations of open systems and micro–macro link, in: D. Helbing, H.J. Herrmann, M. Schreckenberg, D.E. Wolf (Eds.), Traffic and Granular Flow ’99, Springer, Berlin, 2000, pp. 383–388] highlighted that one of the most promising way to simulate efficiently traffic flow on large road networks is a clever combination of both traffic representations: the hybrid modelling. Our focus in this paper is to propose two hybrid models for which the macroscopic (resp. mesoscopic) part is based on a class of second order model [A. Aw, M. Rascle, Resurection of second order models of traffic flow?, SIAM J. Appl. Math. 60 (2000) 916–938] whereas the microscopic part is a Follow-the Leader type model [D.C. Gazis, R. Herman, R.W. Rothery, Nonlinear follow-the-leader models of traffic flow, Oper. Res. 9 (1961) 545–567; R. Herman, I. Prigogine, Kinetic Theory of Vehicular Traffic, American Elsevier, New York, 1971]. For the first hybrid model, we define precisely the translation of boundary conditions at interfaces and for the second one we explain the synchronization processes. Furthermore, through some numerical simulations we show that the waves propagation is not disturbed and the mass is accurately conserved when passing from one traffic representation to another.

Interferon-induced transmembrane 3 binds osteopontin in vitro: expressed in vivo IFITM3 reduced OPN expression

Osteopontin is a secreted, integrin-binding and phosphorylated acidic glycoprotein, which has an important role in tumour progression. We have shown that Wnt, Ets, AP-1, c-jun and beta-catenin/Lef-1/Tcf-1 stimulates OPN transcription in rat mammary carcinoma cells by binding to a specific promoter sequence. However, co-repressors of OPN have not been identified. In this study, we have used the bacterial two-hybrid system to isolate cDNA-encoding proteins that bind to OPN and modulate its role in malignant transformation. Using this approach we isolated interferon-induced transmembrane protein 3 gene (IFITM3) as a potential protein partner. We show that IFITM3 and OPN interact in vitro and in vivo and that IFITM3 reduces osteopontin (OPN) mRNA expression, possibly by affecting OPN mRNA stability. Stable transfection of IFITM3 inhibits OPN, which mediates anchorage-independent growth, cell adhesion and cell invasion. Northern blot analysis revealed an inverse mRNA expression pattern of IFITM3 and OPN in human mammary cell lines. Inhibition of IFITM3 by antisense RNA promoted OPN protein expression, enhanced cell invasion by parental benign non-invasive Rama 37 cells, indicating that the two proteins interact functionally as well. We also identified an IFITM3 DNA-binding domain, which interacts with OPN, deletion of which abolished its inhibitive effect on OPN. This work has shown for the first time that IFITM3 physically interacts with OPN and reduces OPN mRNA expression, which mediates cell adhesion, cell invasion, colony formation in soft agar and metastasis in a rat model system. Oncogene (2010) 29, 752-762; doi: 10.1038/onc.2009.379; published online 9 November 2009

Heat shock protein 70-mediated sensitization of cells to apoptosis by Carboxyl-Terminal Modulator Protein

Background: The serine/threonine protein kinase B (PKB/Akt) is involved in insulin signaling, cellular survival, and transformation. Carboxyl-terminal modulator protein (CTMP) has been identified as a novel PKB binding partner in a yeast two-hybrid screen, and appears to be a negative PKB regulator with tumor suppressor-like properties. In the present study we investigate novel mechanisms by which CTMP plays a role in apoptosis process.

Regulation of cyclin D1 by the BRCA1-BARD1 complex

Background BRCA1 and cyclin D₁ are both essential for normal breast development and mutation or aberration of their expression is associated with breast cancer [1,2]. Cyclin D₁ is best known as a G₁ cyclin where it regulates the G₁ to S phase transition by acting as a rate-limiting subunit of CDK4/6 kinase activity. More recently, however, Stacey has demonstrated that cyclin D₁ levels in G₂/M determine whether a cell continues to proliferate or exits the cell cycle [3]. The majority of BRCA1 in the cell is bound to BARD1 through their N-terminal RING domains. Heterodimerization is essential for the stability and correct localization of the complex and confers ubiquitin ligase activity to BRCA1. The importance of the ligase activity of BRCA1 to breast cancer development is inferred from the fact that N-terminal diseaseassociated mutations are proposed to reduce ligase activity [4]. Methods Protein–protein interactions were demonstrated using yeast-two-hybrid and coimmunoprecipitation. Protein levels were altered through overexpression, siRNA and antisense technology. The effect of proteasome inhibitors and cycloheximide treatment was also examined. Results We initially identified cyclin D₁ as a binding partner of BARD1 in a yeast-two-hybrid screen and defined the minimal binding region as the N-terminus of BARD1. This interaction was confirmed in vivo by coimmunoprecipitation. The N-terminus of BARD1 also binds BRCA1 and imparts ubiquitin ligase activity to the complex. Covalent modification of proteins with ubiquitin is a common regulatory mechanism in eukaryotic cells. Traditionally polyubiquitin chains linked through lysine 48 target proteins for degradation by the 26 S proteasome. We have demonstrated that cyclin D₁ protein levels are inversely related to BRCA1 and BARD1 levels in several model systems. Furthermore, regulation of cyclin D1 levels occurs through a post-transcriptional mechanism and requires the ligase activity of BRCA1. Interestingly, this phenomenon is cell-cycle regulated, occurring in G₂/M. Conclusion We propose that cyclin D₁ is a potential substrate for BRCA1 ubiquitination and that this targets cyclin D₁ for proteasomal-mediated degradation. Future work will focus on ascertaining the functional consequence of cyclin D₁ regulation by the BRCA1–BARD1 complex; in particular, the impact of BRCA1, mediated through regulation of cyclin D₁, on the proliferation versus differentiation decision.

BRD7, a subunit of SWI/SNF complexes, binds directly to BRCA1 and regulates BRCA1-dependent transcription

We carried out a yeast two-hybrid screen using a BRCA1 bait composed of amino acids 1 to 1142 and identified BRD7 as a novel binding partner of BRCA1. This interaction was confirmed by coimmunoprecipitation of endogenous BRCA1 and BRD7 in T47D and HEK-293 cells. BRD7 is a bromodomain containing protein, which is a subunit of PBAF-specific Swi/Snf chromatin remodeling complexes. To determine the functional consequences of the BRCA1-BRD7 interaction, we investigated the role of BRD7 in BRCA1-dependent transcription using microarray-based expression profiling. We found that a variety of targets were coordinately regulated by BRCA1 and BRD7, such as estrogen receptor alpha (ERalpha). Depletion of BRD7 or BRCA1 in either T47D or MCF7 cells resulted in loss of expression of ERalpha at both the mRNA and protein level, and this loss of ERalpha was reflected in resistance to the antiestrogen drug fulvestrant. We show that BRD7 is present, along with BRCA1 and Oct-1, on the ESR1 promoter (the gene which encodes ERalpha). Depletion of BRD7 prevented the recruitment of BRCA1 and Oct-1 to the ESR1 promoter; however, it had no effect on the recruitment of the other Swi/Snf subunits BRG1, BAF155, and BAF57 or on RNA polymerase II recruitment. These results support a model whereby the regulation of ERalpha transcription by BRD7 is mediated by its recruitment of BRCA1 and Oct-1 to the ESR1 promoter.

BcsK(C) Is an Essential Protein for the Type VI Secretion System Activity in Burkholderia cenocepacia That Forms an Outer Membrane Complex with BcsL(B)

The type VI secretion system (T6SS) contributes to the virulence of Burkholderia cenocepacia, an opportunistic pathogen causing serious chronic infections in patients with cystic fibrosis. BcsK(C) is a highly conserved protein among the T6SSs in Gram-negative bacteria. Here, we show that BcsK(C) is required for Hcp secretion and cytoskeletal redistribution in macrophages upon bacterial infection. These two phenotypes are associated with a functional T6SS in B. cenocepacia. Experiments employing a bacterial two-hybrid system and pulldown assays demonstrated that BcsK(C) interacts with BcsL(B), another conserved T6SS component. Internal deletions within BcsK(C) revealed that its N-terminal domain is necessary and sufficient for interaction with BcsL(B). Fractionation experiments showed that BcsK(C) can be in the cytosol or tightly associated with the outer membrane and that BcsK(C) and BcsL(B) form a high molecular weight complex anchored to the outer membrane that requires BcsF(H) (a ClpV homolog) to be assembled. Together, our data show that BcsK(C)/BcsL(B) interaction is essential for the T6SS activity in B. cenocepacia.

The GalF protein of Escherichia coli is not a UDP-glucose pyrophosphorylase but interacts with the GalU protein possibly to regulate cellular levels of UDP-glucose

We report the functional characterization of the galF gene of strain VW187 (Escherichia coli O7:K1), which encodes a polypeptide displaying structural features common to bacterial UDP-glucose pyrophosphorylases, including the E. coli GalU protein. These enzymes catalyse a reversible reaction converting UTP and glucose-1-phosphate into UDP-glucose and PPi. We show that, although the GalF protein is expressed in vivo, GalF-expressing plasmids cannot complement the phenotype of a galU mutant and extracts from this mutant which only produces GalF are enzymatically inactive. In contrast, the presence of GalU and GalF proteins in the same cell-free extract caused a significant reduction in the rate of pyrophosphorolysis (conversion of UDP-glucose into glucose-1-phosphate) but no significant effect on the kinetics of synthesis of UDP-glucose. The presence of GalF also increased the thermal stability of the enzyme in vitro. The effect of GalF in the biochemical properties of the UDP-glucose pyrophosphorylase required the co-synthesis of GalF and GalU, suggesting that they could interact as components of the oligomeric enzyme. The physical interaction of GalU and GalF was demonstrated in vivo by the co-expression of both proteins as fusion products using a yeast two-hybrid system. Furthermore, using a pair of galF-/galU+ and galF/galU+ isogenic strains, we demonstrated that the presence of GalF is associated with an increased concentration of intracellular UDP-glucose as well as with an enhancement of the thermal stability of the UDP-glucose pyrophosphorylase in vivo. We propose that GalF is a non-catalytic subunit of the UDP-glucose pyrophosphorylase modulating the enzyme activity to increase the formation of UDP-glucose, and this function is important for bacterial adaptation to conditions of stress.