Upstream sequences required for efficient expression of a soybean heat shock gene.

A soybean gene (Gmhsp17.5-E) encoding a small heat shock protein was introduced into primary sunflower tumors via T-DNA-mediated transformation. RNA blot hybridizations and S1-nuclease hybrid protection studies indicated that the heat shock gene containing 3.25 kilobases of 5'-flanking sequences was strongly transcribed in a thermoinducible (40 degrees C) manner. Transcriptional induction also occurred to a lesser extent upon treatment of whole tumors with sodium arsenite and CdCl2. Basal (26 degrees C) transcription was not detected in soybean seedlings, but it was quite evident in transformed tumor tissue. A 5' deletion to -1,175 base pairs with respect to the CAP site had no effect on the levels of thermoinducible transcription, but it resulted in a large increase in basal transcription. Further removal of DNA sequences (including the TATA-distal heat shock consensus element) to -95 base pairs reduced thermoinducible transcription by 95% and also greatly decreased basal transcription. The termini of the Gmhsp17.5-E RNA in the tumor were generally the same as those present in soybean RNA, with the exception of several additional 3' termini.

Genes for low-molecular-weight heat shock proteins of soybeans: sequence analysis of a multigene family.

Soybeans, Glycine max, synthesize a family of low-molecular-weight heat shock (HS) proteins in response to HS. The DNA sequences of two genes encoding 17.5- and 17.6-kilodalton HS proteins were determined. Nuclease S1 mapping of the corresponding mRNA indicated multiple start termini at the 5' end and multiple stop termini at the 3' end. These two genes were compared with two other soybean HS genes of similar size. A comparison among the 5' flanking regions encompassing the presumptive HS promoter of the soybean HS-protein genes demonstrated this region to be extremely homologous. Analysis of the DNA sequences in the 5' flanking regions of the soybean genes with the corresponding regions of Drosophila melanogaster HS-protein genes revealed striking similarity between plants and animals in the presumptive promoter structure of thermoinducible genes. Sequences related to the Drosophila HS consensus regulatory element were found 57 to 62 base pairs 5' to the start of transcription in addition to secondary HS consensus elements located further upstream. Comparative analysis of the deduced amino acid sequences of four soybean HS proteins illustrated that these proteins were greater than 90% homologous. Comparison of the amino acid sequence for soybean HS proteins with other organisms showed much lower homology (less than 20%). Hydropathy profiles for Drosophila, Xenopus, Caenorhabditis elegans, and G. max HS proteins showed a similarity of major hydrophilic and hydrophobic regions, which suggests conservation of functional domains for these proteins among widely dispersed organisms.

Characterization of Gmhsp26-A, a stress gene encoding a divergent heat shock protein of soybean: heavy-metal-induced inhibition of intron processing.

We determined the DNA sequence and mapped the corresponding transcripts of a genomic clone containing the Gmhsp26-A gene of soybean. This gene is homologous to the previously characterized cDNA clone pCE54 (E. Czarnecka, L. Edelman, F. Schöffl, and J. L. Key, Plant Mol. Biol. 3:45-58, 1984) and is expressed in response to a wide variety of physiological stresses including heat shock (HS). S1 nuclease mapping of transcripts and a comparison of the cDNA sequence with the genomic sequence indicated the presence of a soybean seedlings with either CdCl2 or CuSO4. Analysis of the 5' termini of transcripts indicated the presence of one major and at least two minor start sites. In each case, initiation occurred 27 to 30 base pairs downstream from a TATA-like motif, and thus each initiation site appears to be promoted by the activity of a separate subpromoter. The three subpromoters are all associated with sequences showing low homology to the HS consensus element of Drosophila melanogaster HS genes and are differentially induced in response to various stresses. Within the carboxyl-terminal half of the protein, hydropathy analysis of the deduced amino acid sequence indicated a high degree of relatedness to the small HS proteins. A comparison of the primary amino acid sequence of hsp26-A with sequences of the small HS proteins suggested that this stress protein is highly diverged and may therefore be specialized for stress adaptation in soybean.

Localization of small heat shock proteins to the higher plant endomembrane system.

Three related gene families of low-molecular-weight (LMW) heat shock proteins (HSPs) have been characterized in plants. We describe a fourth LMW HSP family, represented by PsHSP22.7 from Pisum sativum and GmHSP22.0 from Glycine max, and demonstrate that this family of proteins is endomembrane localized. PsHSP22.7 and GmHSP22.0 are 76.7% identical at the amino acid level. Both proteins have amino-terminal signal peptides and carboxyl-terminal sequences characteristic of endoplasmic reticulum (ER) retention signals. The two proteins closely resemble class I cytoplasmic LMW HSPs, suggesting that they evolved from the cytoplasmic proteins through the addition of the signal peptide and ER retention motif. The endomembrane localization of these proteins was confirmed by cell fractionation. The polypeptide product of PsHSP22.7 mRNA was processed to a smaller-M(r) form by canine pancreatic microsomes; in vivo, GmHSP22.0 polysomal mRNA was found to be predominantly membrane bound. In vitro-processed PsHSP22.7 corresponded in mass and pI to one of two proteins detected in ER fractions from heat-stressed plants by using anti-PsHSP22.7 antibodies. Like other LMW HSPs, PsHSP22.7 was observed in higher-molecular-weight structures with apparent masses of between 80 and 240 kDa. The results reported here indicate that members of this new class of LMW HSPs are most likely resident ER proteins and may be similar in function to related LMW HSPs in the cytoplasm. Along with the HSP90 and HSP70 classes of HSPs, this is the third category of HSPs localized to the ER.

Analysis of plant RNA polymerase I transcript in chromatin and nuclei.

Chromatin isolated (pH 8.0) from soybean hypocotyl contains only RNA polymerase I activity as judged by its elution at low ionic strength (0.11 M ammonium sulfate) from DEAE-cellulose and DEAE-Sephadex, its total resistance to alpha-amanitin, and lack of preference for poly(dA-dT). The in vitro RNA product from this chromatin contains rRNA as a major component (36%) with little or no symmetry of transcription. The transcript from nuclei, where both RNA polymerases I and II are active, shows a dramatic increase in % rRNA (from 35 to 65%) when alpha-amanitin is present during synthesis. These observations suggest that plant RNA polymerase I is similar to animal RNA polymerase I in both its insensitivity to alpha-amanitin and preferential transcription of rRNA genes.

Induction and Regulation of Heat-Shock Gene Expression by an Amino Acid Analog in Soybean Seedlings.

The effect of the proline analog azetidine-2-carboxylic acid (Aze) on the induction and the regulation of heat-shock (HS) mRNA accumulation and heat-shock protein (HSP) synthesis in soybean (Glycine max) seedlings was studied. Treatment with Aze elicited an HS-like response at the normal growth temperature, 28[deg]C, with seven of nine HS cDNA clones tested. Two cDNA clones, Gm-Hsp22.5 and pFS2033, share 78% identity; however, transcripts hybridizing to GmHsp22.5 but not pFS2033 accumulated with Aze treatment at 28[deg]C. Substantial incorporation of radioactive amino acid into high molecular weight HSPs but not low molecular weight HSPs was observed in vivo during Aze treatment at 28[deg]C. Low molecular weight HSPs were detected using antibodies raised against an abundant member of low molecular weight class I HSPs, indicating that low molecular weight HSPs were synthesized at normal growth temperatures during Aze treatment despite a lack of substantial in vivo radioactive amino acid incorporation. In summary, Aze treatment induced accumulation of most but not all HS mRNAs and HSPs in soybean seedlings; the observations presented here suggest differential regulation among various HS genes at the transcriptional and posttranscriptional levels.

A soybean 101-kD heat shock protein complements a yeast HSP104 deletion mutant in acquiring thermotolerance.

A cDNA clone encoding a 101-kD heat shock protein (HSP101) of soybean was isolated and sequenced. Genomic DNA gel blot analysis indicated that the corresponding gene is a member of a multigene family. The mRNA for HSP101 was not detected in 2-day-old etiolated soybean seedlings grown at 28 degrees C but was induced by elevated temperatures. DNA sequence comparison has shown that the corresponding gene belongs to the Clp (caseinolytic protease) (or Hsp100) gene family, which is evolutionarily conserved and found in both prokaryotes and eukaryotes. On the basis of the spacer length between the two conserved ATP binding regions, this gene has been identified as a member of the ClpB subfamily. Unlike other Clp genes previously isolated from higher plants, the expression of this soybean Hsp101 gene is heat inducible, and it does not have an N-terminal signal peptide for targeting to chloroplasts. Transformation of the soybean Hsp101 gene into a yeast HSP104 deletion mutant complemented restoration of acquired thermotolerance, a process in which cells survive an otherwise lethal heat stress after they are given a permissive heat treatment.

Characterization and sequence analysis of a developmentally regulated putative cell wall protein gene isolated from soybean.

A cDNA clone, pTU04, which hybridizes to two different sizes of mRNA on Northern blots was isolated from soybean suspension culture cell poly(A) RNA. Northern analysis reveals that meristematic tissue produces a 1050-nucleotide mRNA while quiescent mature cells produce primarily a 1220-nucleotide mRNA homologous to pTU04. The cDNA and its corresponding genomic clone have been partially characterized. The nucleotide sequence of the gene predicts a proline-rich protein, designated SbPRP1, which contains a signal peptide sequence and 43 repeats of a sequence consisting primarily of Pro-Pro-Val-Tyr-Lys (CCA-CCA-GTT-TAC-AAG). From nuclease S1 and hybrid-select translation analyses, the cDNA clone pTU04 appears to represent the mRNA for the mature tissue 1220-nucleotide RNA observed on Northern blots. Although there is no direct proof that the encoded protein is a cell wall protein, it has the properties similar to previously isolated cell wall proteins: 1) it is very basic with a high content of Pro, Tyr, and Lys; 2) it has similar hydropathic properties; and 3) its repeating unit shares sequence homology with that of more highly characterized cell wall proteins, generally termed extensin (Chen, J., and Varner, J. E. (1985) EMBO J. 4, 2145-2151; Smith, J. J., Muldoon, E. P., Willard, J. J., and Lamport, D. T. A. (1986) Phytochemistry 25, 1021-1030.

Developmentally regulated expression of soybean proline-rich cell wall protein genes.

Previously, we reported the characterization of a developmentally regulated proline-rich cell wall protein (SbPRP1) gene of soybean; the encoded protein is represented by a consensus amino acid repeat structure of Pro-Pro-Val-Tyr-Lys [Hong, J.C., Nagao, R.T., and Key, J.L. (1987). J. Biol. Chem. 262, 8367-8376]. Two other closely related members of this family of proline-rich protein (PRP) genes (SbPRP2 and SbPRP3), which differ from the extensin family of cell wall proteins, have been characterized (J.C. Hong, R.T. Nagao, and J.L. Key, unpublished results). Here we report studies on the regulation of expression of this gene family during soybean development by analyzing various plant organs, including leaves, stems, and roots of etiolated seedlings and light-grown plants, as well as young and mature pods, seed coats, and cotyledons. These organs were tested at different stages of development (young and fully mature tissue). Although a high level of sequence homology is observed at the nucleotide and amino acid sequence level among these three PRP genes/proteins, there are marked differences in the patterns of expression of each gene in different plant organs and organ regions. SbPRP1 is highly expressed and is the predominant form of PRP transcript in the mature hypocotyl, root, and immature seed coat. SbPRP2 is the major form of PRP in the apical hypocotyl and young suspension culture cells. SbPRP3 is the major form of PRP gene expressed in aerial parts; it is highly expressed in leaves, although no expression is detected in the roots.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of a proline-rich cell wall protein gene family of soybean. A comparative analysis.

Further characterization of a proline-rich cell wall protein gene family from soybean (Glycine max (L.) Merr) has been accomplished by the isolation and sequence analysis of two additional genes, SbPRP2 and SbPRP3, which encode mRNAs of 1050 and 650 nucleotides in length, respectively. Like the proline-rich protein gene, SbPRP1, which was previously reported (Hong, J. C., Nagao, R. T., and Key, J. L. (1987) J. Biol. Chem. 262, 8367-8376), these two SbPRP genes encode proteins having a signal peptide sequence and repeats of Pro-Pro-Val-Tyr-Lys. The SbPRP2 gene encodes a protein of 26 kDa which contains a perfect alternating repeat of Pro-Pro-Val-Tyr-Lys and Pro-Pro-Val-Glu-Lys. The SbPRP3 encodes a 10-kDa protein which also contains Pro-Pro-Val-Tyr-Lys as a major amino acid repeat, but the overall amino acid sequence of this protein is more variable than that of SbPRP1 and SbPRP2. RNA blot analyses have demonstrated that there are marked differences in the pattern of expression of each SbPRP in various soybean tissues. In contrast, sequence analysis reveals that the SbPRP genes contain a high degree of sequence conservation. Nucleotide sequence homology extends 90 to 100 base pairs upstream of the transcription initiation site and includes typical CAT and TATA sequences. Approximately 80 base pairs of the 3'-noncoding sequence around the polyadenylation signal is also highly conserved. Therefore, the DNA sequence upstream of the 5'-conserved region is presumed to contain cis-elements accounting for the developmental and tissue specificity of gene expression. While the pentameric repeat structures occur in all SbPRP genes, the encoded proteins are predicted to be different in several features including basicity, substitutions of tyrosine and glutamic acid in the repeat, and the size of the mature protein.

Patterns of soybean proline-rich protein gene expression.

The expression patterns of three members of a gene family that encodes proline-rich proteins in soybean (SbPRPs) were examined using in situ hybridization experiments. In most instances, the expression of SbPRP genes was intense in a limited number of cell types of a particular organ. SbPRP1 RNA was localized in several cell types of soybean hypocotyls, including cells within the phloem and xylem. SbPRP1 expression increased within epidermal cells in the elongating and mature regions of the hypocotyl; expression was detected also in lignified cells surrounding the hilum of mature seeds. SbPRP2 RNA was present in cortical cells and in the vascular tissue of the hypocotyl, especially cells of the phloem. This gene was expressed also in the inner integuments of the mature seed coat. SbPRP3 RNA was localized specifically to the endodermoid layer of cells surrounding the stele in the elongating region of the hypocotyl, as well as in the epidermal cells of leaves and cotyledons. These data show that members of this gene family exhibit cell-specific expression. The members of the SbPRP gene family are expressed in different types of cells and in some cell types that also express the glycine-rich protein or hydroxyproline-rich glycoprotein classes of genes.

The DNA sequence analysis of soybean heat-shock genes and identification of possible regulatory promoter elements.

The soybean possesses a gene family encoding the major low mol. wt. heat-shock proteins of 15-18 kd. We have determined the primary DNA sequences of two of the genes, both located on the same subgenomic DNA fragment. The protein coding regions are characterized by long uninterrupted open reading frames and by sequence homology of 92% and 100% with a heat-shock specific cDNA. One protein sequence deduced from the completely cloned gene hs6871 is composed of 153 amino acids with a total mol. wt. of 17.3 kd; the other protein is a truncated polypeptide containing 73 amino acids at the carboxy-terminal end of an incompletely cloned heat-shock gene designated hs6834. Investigations of the hydrophilic/hydrophobic characteristics of the polypeptides revealed a conservation of structural features between heat-shock proteins from soybean, Caenorhabditis and Drosophila and mammalian lens alpha-crystallin. The 5' end of the soybean heat-shock gene hs6871 was mapped by S1 nuclease at a position which is 100 nucleotides upstream from the translation start codon and 25 nucleotides downstream from a TATA-box sequence. Six other potential promoter elements which are homologous to the Drosophila heat-shock consensus sequence CT-GAA-TTC-AG-, are present within 150 nucleotides upstream from the TATA-box. The possible functions of these promoter elements in transcriptional regulation of expression of soybean heat-shock gene are discussed.

Identification of protein-binding DNA sequences in an auxin-regulated gene of soybean.

The promoter region of a soybean auxin-responsive gene, GmAux28, was analyzed to identify protein-binding DNA sequences that may be involved in regulation of expression. Using DNase I footprinting and gel mobility shift assays, multiple regions of interaction, including eight major protein-binding sites, were observed in the GmAux28 gene. Two sequence motifs, TGACGACA and TCCACGTGTC, related to as-1/Hex and G-box elements, respectively, found in several plant promoters, were identified. Four distinct A/T-rich domains were identified; such A/T-rich domains appear to modulate, but not to specify, the expression of many genes. Two new sequence motifs, delta-1 (D1) and delta-4 (D4) were also identified. D1 and D4 share a very similar core sequence, TAGTxxCTGT and TAGTxCTGT, respectively. In gel mobility shift analyses, D1 and D4 elements exhibit a complex interaction of binding proteins. The GmAux22 promoter also contains D1-related elements which compete with the GmAux28 elements. Sequence comparisons have identified D1/D4-like sequences in several other auxin-responsive genes suggesting the possible importance of D1/D4 and the respective binding proteins in the regulation of expression of these genes.

Multigene family of actin-related sequences isolated from a soybean genomic library.

We have investigated the actin-related sequences in soybean using heterologous actin DNA probes from Dictyostelium, Drosophila, and yeast. Southern blot analysis of restriction digests of soybean DNA indicates that actin is encoded in a small multigene family. In order to isolate individual members of this gene family, we have constructed a soybean genomic library in the lambda vehicle Charon 4A. A partial characterization of this library shows it to be nearly complete. We have isolated from this library a number of recombinant clones that hybridize to actin-coding sequences from all three heterologous probes. We have identified the fragments containing the actin-related sequences on the physical maps of two of these clones lambda SAc1 and lambda SAc3. These fragments were subcloned in the plasmid vehicle pBR322. Using electron microscope heteroduplex mapping we show that the subclones, pSAc1 and pSAc3, share homology with the entire actin-coding sequence (1.1 kb) of Drosophila and Dictyostelium. Furthermore, pSAc1 and pSAc3 have additional homology of approximately 0.22 kb at the 5' ends of their coding sequences. No homology is detected in the 3' flanking regions of these clones. The actin sequence in pSAc1 contains an interruption of approximately 0.30 kb located 0.39 kb from the 5' end of the actin polypeptide coding region.

DNA sequence and transcript mapping of a soybean gene encoding a small heat shock protein.

The DNA sequence of a gene (Gmhsp17.5-E) encoding a small heat shock protein of soybean, Glycine max, has been determined. Nuclease S1 mapping of the 5' terminus of the corresponding RNA indicates that the start site for transcription is located 82 bases upstream from the coding region and 24 bases downstream from a "TATA"-like region (-T-T-T-A-A-A-T-A-). The 5' flanking region of Gmhsp17.5-E contains two imperfect dyads that closely resemble regulatory elements present in the promoters of heat-inducible genes of Drosophila. One, positioned 18 bases upstream from the TATA-like region, shows 90% homology to the Drosophila heat shock consensus sequence. The other overlaps an upstream TATA sequence and is located at position -213. Analysis of the derived amino acid sequence indicates that the protein encoded by Gmhsp17.5-E is related structurally to the four small heat shock proteins of Drosophila. This relationship is most evident by comparison of hydropathy profiles; they show conservation of several major hydrophilic and hydrophobic regions, which suggests that these proteins have common structural features.

Molecular characterization of cDNAs encoding low-molecular-weight heat shock proteins of soybean.

Three cDNA clones (GmHSP23.9, GmHSP22.3, and GmHSP22.5) representing three different members of the low-molecular-weight (LMW) heat shock protein (HSP) gene superfamily were isolated and characterized. A fourth cDNA clone, pFS2033, was partially characterized previously as a full-length genomic clone GmHSP22.0. The deduced amino acid sequences of all four cDNA clones have the conserved carboxyl-terminal LMW HSP domain. Sequence and hydropathy analyses of GmHSP22, GmHSP22.3, and GmHSP22.5, representing HSPs in the 20 to 24 kDa range, indicate they contain amino-terminal signal peptides. The mRNAs from GmHSP22, GmHSP22.3, and GmHSP22.5 were preferentially associated in vivo with endoplasmic reticulum (ER)-bound polysomes. GmHSP22 and GmHSP22.5 encode strikingly similar proteins; they are 78% identical and 90% conserved at the amino acid sequence level, and both possess the C-terminal tetrapeptide KDEL which is similar to the consensus ER retention motif KDEL; the encoded polypeptides can be clearly resolved from each other by two-dimensional gel analysis of their hybrid-arrest translation products. GmHSP22.3 is less closely related to GmHSP22 (48% identical and 70% conserved) and GmHSP22.5 (47% identical and 65% conserved). The fourth cDNA clone, GmHSP23.9, encodes a HSP of ca. 24 kDa with an amino terminus that has characteristics of some mitochondrial transit sequences, and in contrast to GmHSP22, GmHSP22.3, and GmHSP22.5, the corresponding mRNA is preferentially associated in vivo with free polysomes. It is proposed that the LMW HSP gene superfamily be expanded to at least six classes to include a mitochondrial class and an additional endomembrane class of LMW HSPs.

Sequence and characterization of two auxin-regulated genes from soybean.

The auxin-regulated expression of two poly(A)+ mRNAs in soybean hypocotyl was demonstrated by cloning of the cDNAs and Northern blot hybridization analyses (Walker, J.C., and Key, J.L. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 7185-7189). The corresponding genes, designated Aux28 and Aux22, have been isolated, and the cDNAs and genes have been sequenced. The Aux28 and Aux22 genes are present at one to two copies per haploid genome, contain four and two introns, and encode hydrophilic proteins of 26.8 and 21.5 kDa, respectively. Although the cDNAs were isolated independently and do not cross-hybridize under stringent hybridization conditions, the protein coding sequences of the two cDNAs have several colinear regions of high homology at the nucleic acid (77-80%) and the amino acid (80-100%) levels; together these regions constitute approximately a third of the protein coding sequences of the cDNAs. These data, together with genomic Southern blot hybridization analysis and hybrid-select translations of mRNAs homologous to the cDNAs, show that these genes belong to two related multigene families. We have identified two sequences, TGATAAAAG and GGCAGCATGCA, that occur at similar distances upstream of the transcription start site in each gene, and the spacing between these two elements is essentially identical in the two genes. The possible significance of these sequences is under evaluation.

Regulation of the heat shock response in soybean seedlings.

The transcriptional response of soybean (Glycine max) seedlings during heat shock (HS) was investigated under two different treatment regimes. During prolonged heat treatment at 40 degrees C, active transcription of the HS genes (as measured by "runoff" transcription assays) occurs only during the first few hours. Nonetheless, mRNAs for these genes are present at relatively high abundance even after 9 hours of exposure to 40 degrees C. Because HS mRNAs have a fairly short half-life (less than 3 hours) at 28 degrees C, these results indicate that HS mRNAs are inherently more stable at 40 degrees C. During a second type of heat treatment regime-short pulses of high (45 degrees C) heat followed by 1 to 2 hours at 28 degrees C-transcription of HS genes is comparable to that achieved at 40 degrees C for the first few hours, even though the tissue is maintained at non-HS temperatures. The transcriptional responses to these two different heat treatments indicate that regulatory controls for the transcription of the HS genes must involve more than a simple sensing of ambient temperature, since transcription of these genes can be turned off at 40 degrees C (in the case of prolonged exposure) and can continue at 28 degrees C (following a short, severe heat treatment). Additional results demonstrate that the response of soybean seedlings to a particular HS depends on their prior exposure to heat; seedlings given a preheat treatment (that is known to induce thermotolerance) respond more moderately to a short heat pulse at 45 degrees C. Overall, this research indicates that plants have mechanisms for both monitoring the severity of changes in temperature and for measuring the magnitude and duration of the stress. Such information is then used to regulate the plant's response to heat both transcriptionally and posttranscriptionally.

Sequence and Expression of a HSP83 from Arabidopsis thaliana.

A full-length cDNA encoding a heat shock protein (hsp) belonging to the 83 to 90 kilodalton hsp family of Arabidopsis thaliana has been isolated and sequenced. Truncated cDNA clones were isolated by nucleic acid hybridization to a truncated soybean HSP83 cDNA probe and a fragment generated from a Drosophila HSP83 gene. A single strand DNA vector/primer based extension procedure was employed to obtain the full-length cDNA. The level of transcripts homologous to this cDNA (AtHS83) is low in 2-week-old Arabidopsis plants but is rapidly enhanced by elevated temperatures. DNA sequence comparison between this cDNA and hsp83-90 sequences from human, yeast and Drosophila reveal amino acid identities of 63 to 69%, typical identities for interspecies comparisons between hsp83 to 90 kilodalton proteins. Genomic DNA blot analysis performed with probes derived from AtHS83 indicate the presence of a HSP83 gene family estimated to be comprised of at least three genes.

A heat shock protein localized to chloroplasts is a member of a eukaryotic superfamily of heat shock proteins.

We have isolated cDNA clones from soybean and pea that specify nuclear-encoded heat shock proteins (HSPs) which localize to chloroplasts. The mRNAs for these HSPs are undetectable at control temperatures, but increase approximately 150-fold during a 2-h heat shock. Hybridization-selection followed by in vitro translation demonstrates that these HSPs are synthesized as precursor proteins which are processed by the removal of 5-6.5 kd during import into isolated chloroplasts. The nucleotide sequence of the cDNAs shows the derived amino acid sequences of the mature pea and soybean proteins are 79% identical. While the predicted transit peptide encoded by the pea cDNA has some characteristics typical of transit sequences, including high Ser content, multiple basic residues and no acidic residues, it lacks two domains proposed to be important for import and maturation of other chloroplast proteins. The carboxy-terminal region of the chloroplast HSP has significant homology to cytoplasmic HSPs from soybean and other eukaryotes. We hypothesize that the chloroplast HSP shares a common structural and functional domain with low mol. wt HSPs which localize to other parts of the cell, and may have evolved from a nuclear gene.