William B. Gurley

The Hsf world: classification and properties of plant heat stress transcription factors

Based on the partial or complete sequences of 14 plant heat stress transcription factors (Hsfs) from tomato, soybean, Arabidopsis and maize we propose a general nomenclature with two basic classes, i.e. classes A and B each containing two or more types of Hsfs (HsfA1, HsfA2 etc.). Despite some plant-specific peculiarities, essential functional domains and modules of these proteins are conserved among plants, yeast, Drosophila and vertebrates. A revised terminology of these parts follows recommendations agreed upon among the authors and representatives from other laboratories working in this field (see legend to Fig. 1). Similar to the situation with the small heat shock proteins (sHsps), the complexity of the hsf gene family in plants appears to be higher than in other eukaryotic organisms.

Plants contain a novel multi-member class of heat shock factors without transcriptional activator potential.

Based on phylogeny of DNA-binding domains and the organization of hydrophobic repeats, two families of heat shock transcription factors (HSFs) exist in plants. Class A HSFs are involved in the activation of the heat shock response, but the role of class B HSFs is not clear. When transcriptional activities of full-length HSFs were monitored in tobacco protoplasts, no class B HSFs from soybean or Arabidopsis showed activity under control or heat stress conditions. Additional assays confirmed the finding that the class B HSFs lacked the capacity to activate transcription. Fusion of a heterologous activation domain from human HSF1 (AD2) to the C-terminus of GmHSFB1-34 gave no evidence of synergistic enhancement of AD2 activity, which would be expected if weak activation domains were present. Furthermore, activity of AtHSFB1-4 (class B) was not rescued by coexpression with AtHSFA4-21 (class A) indicating that the class A HSF was not able to provide a missing function required for class B activity. The transcriptional activation potential of Arabidopsis AtHSFA4-21 was mapped primarily to a 39 amino acid fragment in the C-terminus enriched in bulky hydrophobic and acidic residues. Deletion mutagenesis of the C-terminal activator regions of tomato and Arabidopsis HSFs indicated that these plant HSFs lack heat-inducible regulatory regions analogous to those of mammalian HSF1. These findings suggest that heat shock regulation in plants may differ from metazoans by partitioning negative and positive functional domains onto separate HSF proteins. Class A HSFs are primarily responsible for stress-inducible activation of heat shock genes whereas some of the inert class B HSFs may be specialized for repression, or down-regulation, of the heat shock response.

Sequence organization of the soybean genome.

The total complexity of one constituent soybean (Glycine max) genome is estimated to be 1.29 . 10(9) nucleotide pairs, as determined by analysis of the reassociation kinetics of sheared (0.47 kilobase) DNA. Single copy sequences are estimated to represent from 53 to 64% of the genome by analysis of hydroxyapatite binding of repetitive DNA as a function of fragment length. From 65 to 70% of these single copy sequences have a short period interspersion with 1.11--1.36 kilobase lengths alternating with 0.3--0.4 kilobase repetitive sequence elements. The repetitive sequences of soybean DNA are interspersed both among themselves and among single copy regions of the genome.

Specific Interactions with TBP and TFIIB in Vitro Suggest That 14-3-3 Proteins May Participate in the Regulation of Transcription When Part of a DNA Binding Complex

The 14-3-3 family of multifunctional proteins is highly conserved among animals, plants, and yeast. Several studies have shown that these proteins are associated with a G-box DNA binding complex and are present in the nucleus in several plant and animal species. In this study, 14-3-3 proteins are shown to bind the TATA box binding protein (TBP), transcription factor IIB (TFIIB), and the human TBP–associated factor hTAFII32 in vitro but not hTAFII55. The interactions with TBP and TFIIB were highly specific, requiring amino acid residues in the box 1 domain of the 14-3-3 protein. These interactions do not require formation of the 14-3-3 dimer and are not dependent on known 14-3-3 recognition motifs containing phosphoserine. The 14-3-3–TFIIB interaction appears to occur within the same domain of TFIIB that binds the human herpes simplex virus transcriptional activator VP16, because VP16 and 14-3-3 were able to compete for interaction with TFIIB in vitro. In a plant transient expression system, 14-3-3 was able to activate GAL4-dependent β-glucuronidase reporter gene expression at low levels when translationally fused with the GAL4 DNA binding domain. The in vitro binding with general transcription factors TBP and TFIIB together with its nuclear location provide evidence supporting a role for 14-3-3 proteins as transcriptional activators or coactivators when part of a DNA binding complex.

HSP101: A Key Component for the Acquisition of Thermotolerance in Plants

A sudden elevation in temperature triggers a stress response found in all organisms that brings about a global transition in gene expression. Typically, the expression of most genes is either shut down or greatly attenuated, and a specific group of genes, called heat shock (HS) genes, is rapidly

Plant class B HSFs inhibit transcription and exhibit affinity for TFIIB and TBP

Plant heat shock transcription factors (HSFs) are capable of transcriptional activation (class A HSFs) or both, activation and repression (class B HSFs). However, the details of mechanism still remain unclear. It is likely, that the regulation occurs through interactions of HSFs with general transcription factors (GTFs), as has been described for numerous other transcription factors. Here, we show that class A HSFs may activate transcription through direct contacts with TATA-binding protein (TBP). Class A HSFs can also interact weakly with TFIIB. Conversely, class B HSFs inhibit promoter activity through an active mechanism of repression that involves the C-terminal regulatory region (CTR) of class B HSFs. Deletion analysis revealed two sites in the CTR of soybean GmHSFB1 potentially involved in protein–protein interactions with GTFs: one is the repressor domain (RD) located in the N-terminal half of the CTR, and the other is a TFIIB binding domain (BD) that shows affinity for TFIIB and is located C-terminally from the RD. A Gal4 DNA binding domain-RD fusion repressed activity of LexA-activators, while Gal4-BD proteins synergistically activated strong and weak transcriptional activators. In vitrobinding studies were consistent with this pattern of activity since the BD region alone interacted strongly with TFIIB, and the presence of RD had an inhibitory effect on TFIIB binding and transcriptional activation.

AT-rich promoter elements of soybean heat shock gene Gmhsp 17.5E bind two distinct sets of nuclear proteins in vitro

A 33 bp double-stranded oligonucleotide homologous to two AT-rich sequences located upstream (−907 to −889 and −843 to −826) to the start of transcription of heat shock gene Gmhsp17.5E of soybean stimulated transcription when placed 5′ to a truncated (−140) maize Adh1 promoter. The chimeric promoter was assayed in vivo utilizing anaerobically stressed sunflower tumors transformed by a pTi-based vector of Agrobacterium tumefaciens.Nuclear proteins extracted from soybean plumules were shown to bind double-stranded oligonucleotides homologous to AT-rich sequences in the 5′ flanking regions of soybean β-conglycinin, lectin, leghemoglobin and heat shock genes. These proteins were also shown to bind AT-rich probes homologous to homeobox protein binding sites from the Antennapedia and engrailed/fushi tarazu genes of Drosophila. Binding activity specific for AT-rich sequences showed a wide distribution among various plant organs and species. Preliminary characterization indicated that two sets of nuclear proteins from soybean bind AT-rich DNA sequences: a diverse high-molecular-weight (ca. 46–69 kDa) group, and a low-molecular-weight (23 and 32 kDa) group of proteins. The latter meets the operational criteria for high-mobility group proteins (HMGs).

Regulatory domains of the Gmhsp17.5-E heat shock promoter of soybean.

Promoter domains required for in vivo transcriptional expression of soybean heat shock gene Gmhsp17.5-E were identified by insertion-deletion mutagenesis with transgenic expression monitored in Agrobacterium tumefaciens-incited tumors of sunflower. Removal of the TATA-distal domain from position -1175 to position -259 had little effect on overall activity. The four regions contributing to promoter activity identified by this study all map within 244 base pairs from the start of transcription. The most distal cis-acting element of major significance was located from -244 to -179 and contains a conserved TATA-dyad motif centered at -220. Sequences from -179 to -40 comprise the TATA-proximal domain and include an AT-rich region and two sites containing heat shock consensus elements (HSEs). Deletion of the HSE centered at -93 (site 2) severely reduced transcriptional activity. Heat-inducible expression was also eliminated by internal deletion of either the TATA motif or the overlapping HSEs at site 1, indicating that each of these regions is also a major determinant of promoter activity.

Mutational analysis of a plant heat shock element

A total of 32 mutations were generated within the TATA-proximal site 1 (−72 to−47) of soybean heat shock gene Gmhsp17.5E in order to functionally define the optimal configuration of sequences within the heat shock element (HSE). Mutants were tested in vivo utilizing sunflower tumors transformed by a T-DNA based vector. Promoter activity was determined by S1 nuclease hybrid protection analysis of tumor transcripts. A total of five repeats (5′-nGAAn-3′ or 5′-nTTCn-3′) which comprise the HSE at site 1 were required for full transcription induction by heat stress. Analysis of non-conserved bases flanking the central trinucleotide block indicated that 5′-aGAAg′-3′ is the optimum sequence for the 5 bp repeat.

Isolation and characterization of six heat shock transcription factor cDNA clones from soybean

Thermal stress in soybean seedlings causes the activation of pre-existing heat shock transcription factor proteins (HSFs). Activation results in the induction of DNA binding activity which leads to the transcription of heat shock genes. From a soybean cDNA library we have isolated cDNA clones corresponding to six HSF genes. Two HSF genes are expressed constitutively at the transcriptional level, and the remaining four are heat-inducible. Two of the heat inducible genes are also responsive to cadmium stress. Comparative analysis of HSF sequences indicated higher conservation of the DNA binding domain among plant HSFs than those from yeast or other higher eukaryotes. The putative plant HSF oligomerization domain contains hydrophobic heptapeptide repeats characteristic of coiled coils and seems to exist in two structural variants. The carboxy-terminal domains are reduced in size and the C-terminal heptad repeat is degenerate.