Arabidopsis and the heat stress transcription factor world: how many heat stress transcription factors do we need?

Abstract

Sequencing of the Arabidopsis genome revealed a unique complexity of the plant heat stress transcription factor (Hsf) family. By structural characteristics and phylogenetic comparison, the 21 representatives are assigned to 3 classes and 14 groups. Particularly striking is the finding of a new class of Hsfs (AtHsfC1) closely related to Hsf1 from rice and to Hsfs identified from frequently found expressed sequence tags of tomato, potato, barley, and soybean. Evidently, this new type of Hsf is well expressed in different plant tissues. Besides the DNA binding and oligomerization domains (HR-A/B region), we identified other functional modules of Arabidopsis Hsfs by sequence comparison with the well-characterized tomato Hsfs. These are putative motifs for nuclear import and export and transcriptional activation (AHA motifs). There is intriguing flexibility of size and sequence in certain parts of the otherwise strongly conserved N-terminal half of these Hsfs. We have speculated about possible exon-intron borders in this region in the ancient precursor gene of plant Hsfs, similar to the exon-intron structure of the present mammalian Hsf-encoding genes.

Fig 1.

Basic structure of Hsfs. The block diagrams in part A represent tomato (Lycopersicon peruvianum [Lp]) Hsfs with their conserved functional domains. For abbreviations see part B. Arrowheads at the block diagram of Lp-HsfA1 indicate the positions of putative introns in the ancient Hsf gene. (B) Essential structural details are represented by the tomato HsfA2. (1) The central part of the DBD is the helix-turn-helix motif (H2-T-H3) with a considerable number of amino acid residues invariant among different organisms (boldfaced letters). The arrow indicates the position of the intron conserved in all plant Hsfs. (2) The oligomerization domain HR-A/B is characterized by the heptad pattern of hydrophobic residues (dots, asterisks). The insertion of additional 21 amino acid residues between parts A and B are marked in green. (3) The bipartite NLS represents a cluster of basic residues (K, R) recognized by the NLS receptor. (4) Central elements of the activator region are short motifs (AHA elements) rich in aromatic (W, Y, F), hydrophobic (L, I, V), and acidic amino acid residues (D, E). (5) A leucine-rich motif at the C-terminus functions as an NES

Fig 3.

Phylogenetic relationship of Hsfs based on amino acid sequence comparison of the DBDs and HR-A/B regions (Clustal analysis). The consensus tree for all Hsfs was elaborated using the Clustalx 1–8_msw and Tree view software. For names, accession numbers, and identification of the Hsfs, see Table 1 and explanations given to this table. An extended version of this tree is available in the online version (Fig 5). It includes sequence information of many additional Hsfs from tomato, potato, barley, and soybean as derived from EST libraries

Fig 4.

Clustal analysis of the phylogenetic relationship based on the comparison of sequences of different parts of the N-terminal half of plant Hsfs. Three different conserved parts were used to create the phylogenetic trees: (1) the N-terminal part of the DBD until the position of the intron, (2) the C-terminal part of the DBD, and (3) the HR-A/B region, including 2 heptad repeats of HR-A, the insert, and HR-B (see borders defined in Table 2). Most Hsfs are found in identical groups irrespective of the sequence parts used for the analysis. However, a few Hsfs marked by boldfaced letters change their positions (see also the summary given at the bottom of the figure). We assume that, similar to the situation in the present mammalian Hsf genes, these 3 parts were separated by introns in the ancient plant Hsf precursor gene, ie, exon shuffling could have generated mosaic Hsfs. For AtHsfs A7a and A7b, the following situation is envisaged. The N-terminal part of the DBD and the HR-A/B region are derived from the putative A2/A7 precursor gene, whereas the C-terminal part of the DBD stems are from the B3/A7/A6a precursor gene

Fig 5.

Extended version of phylogenetic relationships of plant Hsfs. To emphasize the complexity of the plant Hsf family in general, an extended version of the phylogenetic tree was created by Clustal analysis based on the N-terminal parts of the DBD of those Hsfs contained in Fig 3 plus a considerable number of additional partial clones derived from the database, mostly from EST libraries of tomato (Lycopersicon esculentum), potato (Solanum tuberosum), soybean (Glycine max), and barley (Hordeum vulgare). For other abbreviations see Fig 3 and Table 1. New entries in the figure are marked by boldfaced letters. Information was derived from the following ESTs and accession numbers: L esculentum HsfA1a (AW933448, AW399336, AW223123), HsfA1b (BE354387), HsfA2a (AW034874), HsfA2b (AW930998), HsfA3 (BE433610, AI895834, AW034135, AW035854, AW035844, AW030642, AW033013; all ESTs represent incompletely spliced messenger RNAs), HsfA4 (AW038959, AW933529), HsfA5a (AW217982, AW041695, AW030725, BF096782), HsfA5b (AW034402), HsfA6b (AW036683, AW932142, AW222011, BE 434585, BE433803, AI895294, AI489721, BG132247), HsfA6c (BG351853), HsfA8a (AW738023, mosaic Hsf, whose annotation to the HsfA8 group depends on the C-terminal part of the DBD and the HR-A/B region), HsfA8b (AW931892, not included, because only C-terminal part with HR-A/B region available), HsfB1 (BF097217, BG134658, AI895934), HsfB2a (AW931781, AW220758, AW931176), HsfC1 (AW738534, AW979619, BE451302, AW649243); G max HsfA2a (Hsf21 fragment, Z46952), HsfA2b (AW164509), HsfA4 (BE611683, AW756148, BG405291, BE330669), HsfA6 (BG041837), HsfB2a (BE346810, BF067962), HsfB2b (AW703969), HsfB3 (BE019974), HsfB4b (BF597135), HsfC1 (BE347442, AW596493, BG352891); S tuberosum HsfA5 (BF459947), HsfB2a (BE473183), HsfB2b (BF052865), HsfC1 (BF459641); H vulgare HsfA2a (BE603513), HsfA4 (BE216310), HsfA2b (BF264338), HsfC1 (BF616419); Zea mays: ZmHsfA2/A8 (Hsfa fragment, S61458)

Fig 2.

Survey of Arabidopsis Hsfs. For explanations see legend to Figure 1

Fig 2.

Continued

Fig 6.

Promoter architecture of Arabidopsis Hsf genes with respect to the positioning of HSE clusters. For orientation and positioning of the putative TATA box, we used the well-defined start site of the open reading frames, data for genomic clones with experimentally defined transcription start sites, and the starting points of the indicated ESTs. Symbols are explained at the bottom of the figure. Numbers on the left refer to the corresponding numbers in Table 1

Fig 6.

Continued