The HSP12 gene encodes one of the two major small heat shock proteins of Saccharomyces cerevisiae. Hsp12 accumulates massively in yeast cells exposed to heat shock, osmostress, oxidative stress, and high concentrations of alcohol as well as in early-stationary-phase cells. We have cloned an extended 5'-flanking region of the HSP12 gene in order to identify cis-acting elements involved in regulation of this highly expressed stress gene. A detailed analysis of the HSP12 promoter region revealed that five repeats of the stress-responsive CCCCT motif (stress-responsive element [STRE]) are essential to confer wild-type induced levels on a reporter gene upon osmostress, heat shock, and entry into stationary phase. Disruption of the HOG1 and PBS2 genes leads to a dramatic decrease of the HSP12 inducibility in osmostressed cells, whereas overproduction of Hog1 produces a fivefold increase in wild-type induced levels upon a shift to a high salt concentration. On the other hand, mutations resulting in high protein kinase A (PKA) activity reduce or abolish the accumulation of the HSP12 mRNA in stressed cells. Conversely, mutants containing defective PKA catalytic subunits exhibit high basal levels of HSP12 mRNA. Taken together, these results suggest that HSP12 is a target of the high-osmolarity glycerol (HOG) response pathway under negative control of the Ras-PKA pathway. Furthermore, they confirm earlier observations that STRE-like sequences are responsive to a broad range of stresses and that the HOG and Ras-PKA pathways have antagonistic effects upon CCCCT-driven transcription.

SummaryThere is rapidly expanding interest into the protective systems against reactive oxygen species (ROS) in the eukaryotic cell, now that the links between oxidative damage, various disease states, and ageing, are firmly established in higher organisms. Yeast molecular genetics should be able to provide powerful insight into these mechanisms; this potential is now starting to be exploited. A number of primary antioxidant activities and systems of metal-ion homeostasis or detoxification have now been demonstrated to contribute to oxidative-stress protection in yeast. Also, evidence is emerging that the oxidative-stress response of this organism is complex, involving separate transcription-factor responses to peroxide, superoxide anion and metal ions. Generation of reactive oxygen speciesYeasts, as aerobic cells, have to face the toxic side-effects of molecular oxygen, viz. the production of reactive oxygen species (ROS). ROS are generated during normal cellular metabolism (e.g., by the mitochondrial respiratory chain, and H 2 O 2 -generating reactions catalysed by oxidases). They can also originate from the presence of pro-oxidants, such as hydrogen peroxide, menadione or paraquat, in the medium; by increases in the oxygen pressure (hyperoxia or re-oxygenation of hypoxic cells); or by exposure to ionizing radiations. These ROS, notably superoxide (O 2 . 7) and hydroxyl (OH . ) radicals, hydrogen peroxide (H 2 O 2 ) and singlet oxygen ( 1 O 2 ), damage cellular components by oxidizing lipids, proteins and nucleic acids (Halliwell and Gutteridge, 1989). Under normal physiological conditions, antioxidant defence mechanisms are almost certainly adequate to maintain ROS at basal, unharmful levels and to repair cellular damages. These defences operate at different levels. Primary defences operate to neutralize ROS, while secondary defences repair or remove the products of oxidation damage to DNA, proteins and lipids. When ROS levels exceed the antioxidant capacity of the cells, an oxidative stress results. This review discusses current understanding of the effects of pro-oxidants, the oxidative-stress response and the acquisition of increased resistance to ROS in yeast. Molecular damages induced by ROSOxidative damage to DNA and nucleoproteins produces base and sugar damage, single-strand breaks, abasic sites and DNA-protein cross-links. Although the superoxide radical and H 2 O 2 do not react directly with DNA, OH . is generated from them by the metal-catalysed Fenton/Haber-Weiss reaction. This induces DNA damage in yeast cells (Frankenberg et al., 1993). One of the major products of base damage, 8-hydroxyguanine (8OHG), is detectable in yeast cells exposed to lethal concentrations of H 2 O 2 , but not the sublethal concentrations of H 2 O 2 that induce increased oxidative-stress resistance (Woodford et al., 1995). Furthermore, H 2 O 2 increases the frequency of intrachromosomal recombination in Saccharomyces cerevisiae, via the formation of hydroxyl radicals. It is therefore likely that oxidative DNA damage plays...

Cytosol Chaperone ssa2 Cytosol Chaperone ssa3 Cytosol Chaperone ssa4 Cytosol Chaperone ssbl Unknown Unknown ssb2 Unknown Unknown sscl Mitochondria Chaperone ssdl (kar2) Endoplasmic reticulum Chaperone Hsp6O Mitochondria Chaperone Hsp3O Plasma membrane Unknown Hsp26 Cytosol/nucleus Unknown Hspl2 Cytosol ? Unknown Ubiquitin Cytosol Protein degradation Enzymes Enolase Cytosol Glycolysis Glyceraldehyde 3-phosphate Cytosol Glycolysis dehydrogenase Phosphoglycerate kinase Cytosol Glycolysis Catalase Cytosol Antioxidative defense Abbreviations used: hsp, heat shock protein; GRE, glucocorticoid response element; HSG, heat shock granule; HSF, heat shock transcription factor; HSE, heat shock responsive element; URS, upstream repression site.

Stress response elements (STREs, core consensus AG4 or C4T) have been demonstrated previously to occur in the upstream region of a number of genes responsive to induction by a variety of stress signals. This stress response is mediated by the homologous transcription factors Msn2p and Msn4p, which bind specifically to STREs. Double mutants (msn2 msn4) deficient in these transcription factors have been shown to be hypersensitive to severe stress conditions. To obtain a more representative overview of the set of yeast genes controlled via this regulon, a computer search of the Saccharomyces cerevisiae genome was carried out for genes, which, similar to most known STRE‐controlled genes, exhibit at least two STREs in their upstream region. In addition to the great majority of genes previously known to be controlled via STREs, 69 open reading‐frames were detected. Expression patterns of a set of these were examined by grid filter hybridization, and 14 genes were examined by Northern analysis. Comparison of the expression patterns of these genes demonstrates that they are all STRE‐controlled although their detailed expression patterns differ considerably. © 1998 John Wiley & Sons, Ltd.

No abstract