Human heat shock transcription factor 1 (HSF1) is responsible for stress-induced transcription of heat shock protein genes. The activity of the HSFI transcriptional activation domains is modulated by a separate regulatory domain, which confers repression at control temperature and heat inducibility. We show here that two specific proline-directed serine motifs are important for function of the regulatory domain: Mutation of these serines to alanine derepresses HSF1 activity at control temperature, and mutation to glutamic acid, mimicking a phosphorylated serine, results in normal repression at control temperature and normal heat shock inducibility. Tryptic mapping shows that these serines are the major phosphorylation sites of HSF1 at control temperature in vivo. Stimulation of the Raf/ERK pathway in vivo results in an increased level of phosphorylation of these major sites and the regulatory domain is an excellent substrate in vitro for the mitogen-activated MAPK/ERK. We conclude that phosphorylation of the regulatory domain of HSF1 decreases the activity of HSF1 at control temperature, and propose a mechanism for modification of HSF1 activity by growth control signals.

Calcium/calmodulin-dependent serine/threonine kinase type II (CaMKII) is one of the most abundant proteins in the mammalian brain, where it is thought to regulate synaptic plasticity and other processes. Activation of the multisubunit kinase by calcium is effectively cooperative and can persist long after transient calcium rises. Despite extensive biochemical characterization of CaMKII and identification of numerous in vitro kinase targets, little is known about its function in vivo. Here we report that unc-43 encodes the only Caenorhabditis elegans CaMKII. A gain-of-function unc-43 mutation reduces locomotory activity, alters excitation of three muscle types and lengthens the period of the motor output of a behavioural clock. Null unc-43 mutations cause phenotypes generally opposite to those of the gain-of-function mutation. Mutations in the unc-103 potassium channel gene suppress a gain-of-function phenotype of unc-43 in one tissue without affecting other tissues; thus, UNC-103 may be a tissue-specific target of CaMKII in vivo.

Heat shock factor (HSF) activates transcription in response to cellular stress. Human HSF1 has a central regulatory domain which can repress the activity of its activation domains at the control temperature and render them heat shock inducible. To determine whether the regulatory domain works in tandem with specific features of the HSF1 transcriptional activation domains, we first used deletion and point mutagenesis to define these activation domains. One of the activation domains can be reduced to just 20 amino acids. A GAL4 fusion protein containing the HSF 1 regulatory domain and this 20-amino-acid activation domain is repressed at the control temperature but potently activates transcription in response to heat shock. No specific amino acids in this activation domain are required for response to the regulatory domain; in particular, none of the potentially phosphorylated serine and threonine residues are required for heat induction, implying that heat-induced phosphorylation of the transcriptional activation domains is not required for induction. The regulatory domain is able to confer heat responsiveness to an otherwise completely heterologous chimeric activator that contains a portion of the VP16 activation domain, suggesting that the regulatory domain can sense heat in the absence of other portions of HSF1.

bEnteropathogenic Escherichia coli (EPEC) continues to be a leading cause of mortality and morbidity in children around the world. Two EPEC genomes have been fully sequenced: those of EPEC O127:H6 strain E2348/69 (United Kingdom, 1969) and EPEC O55:H7 strain CB9615 (Germany, 2003). The O55:H7 serotype is a recent precursor to the virulent enterohemorrhagic E. coli O157:H7. To explore the diversity of O55:H7 and better understand the clonal evolution of O157:H7, we fully sequenced EPEC O55:H7 strain RM12579 (California, 1974), which was collected 1 year before the first U.S. isolate of O157:H7 was identified in California. Phage-related sequences accounted for nearly all differences between the two O55:H7 strains. Additionally, O55:H7 and O157:H7 strains were tested for the presence and insertion sites of Shiga toxin gene (stx)-containing bacteriophages. Analysis of non-phage-associated genes supported core elements of previous O157:H7 stepwise evolutionary models, whereas phage composition and insertion analyses suggested a key refinement. Specifically, the placement and presence of lambda-like bacteriophages (including those containing stx) should not be considered stable evolutionary markers or be required in placing O55:H7 and O157:H7 strains within the stepwise evolutionary models. Additionally, we suggest that a 10.9-kb region (block 172) previously believed unique to O55:H7 strains can be used to identify early O157:H7 strains. Finally, we defined two subsets of O55:H7 strains that share an as-yet-unobserved or extinct common ancestor with O157:H7 strains. Exploration of O55:H7 diversity improved our understanding of the evolution of E. coli O157:H7 and suggested a key revision to accommodate existing and future configurations of stx-containing bacteriophages into current models.