Non-transcribed Sequences Research Articles

Discrimination of related transcribed and non-transcribed repetitive DNA sequences from the Y chromosomes of Drosophila hydei and Drosophila eohydei.

The short arm of the Y chromosome of Drosophila hydei carries a single male fertility gene, gene Q, which forms the lampbrush loop pair Nooses. Conflicting observations have been reported concerning the identity of the repetitive DNA sequences that are transcribed in this loop pair. It has been claimed by other investigators that the loop transcripts contain repeats of two distinct, but related families of Y-specific repetitive DNA sequences, ay1 and YsI. We reinvestigated this issue, using as probes single ay1 and YsI repeats which, under stringent conditions, hybridize only to members of their own family. Under non-stringent conditions, both repeats hybridize in situ to Nooses transcripts. However, if hybridization conditions are stringent, only the ay1 probe hybridizes to loop transcripts. Hybridizations to Northern blots of testis RNA confirm these results. Further, YsI repeats are not found the closely related species D. eohydei. We conclude that the YsI repeats are not relevant for the function of fertility gene Q.

Characterization of an unusually conserved AluI highly reiterated DNA sequence family from the honeybee, Apis mellifera.

An AluI family of highly reiterated nontranscribed sequences has been found in the genome of the honeybee Apis mellifera. This repeated sequence is shown to be present at approximately 23,000 copies per haploid genome constituting about 2% of the total genomic DNA. The nucleotide sequence of 10 monomers was determined. The consensus sequences is 176 nucleotides long and has an A + T content of 58%. There are clusters of both direct and inverted repeats. Internal subrepeating units ranging from 11 to 17 nucleotides are observed, suggesting that it could have evolved from a shorter sequence. DNA sequence data reveal that this repeat class is unusually homogeneous compared to the other class of invertebrate highly reiterated DNA sequences. The average pairwise sequence divergence between the repeats is 2.5%. In spite of this unusual homogeneity, divergence has been found in the repeated sequence hybridization ladder between four different honeybee subspecies. Therefore, the AluI highly reiterated sequences provide a new probe for fingerprinting in A. m. mellifera.

Intranuclear compartmentalization of transcribed and nontranscribed c-myc sequences in Namalva-S cells.

This investigation is centered on the intranuclear localization of transcribed and nontranscribed c-myc sequences in human Namalva-S cells bearing t(8;14) translocation. Southern hybridization showed that the breakpoint in the truncated allele of c-myc lies outside the characteristic 12.8-kbp EcoRI fragment: as Northern analysis indicated, this reorganization induces a high level of c-myc transcription. Following high-salt treatment, EcoRI digestion and centrifugation, isolated nuclei from the same cells are separated into two residual fractions: a heavier P fraction including nuclear matrix structures and a light S fraction representing dehistonized chromatin fibres. Comparative hybridization experiments revealed that the above procedure separates the c-myc sequences between the two fractions. To locate the site of intranuclear c-myc transcription, we performed run-on experiments with two fractions, topologically analogous to the residual P and S fractions but maintaining the original chromatin organization. These experiments indicated that chromatin P fraction harbours actively transcribed c-myc sequences while chromatin S harbours nontranscribed ones. Further experiments have clarified that the transcribed c-myc sequences are firmly bound to the matrix by multiple attachment sites, arranged throughout the entire gene locus. It was found, moreover, that at the site of attachment the interaction between DNA and the matrix components is realized via proteins. Controls with the beta-globin gene, which is constitutively nonexpressed in Namalva-S cells but upon induction is highly expressed in murine erythroleukemia cells, completely confirmed the conclusion we had made for the intranuclear localization of c-myc. Thus the experiments presented here support the more common idea that the transcribed and nontranscribed sequences are precisely compartmentalized.

Transcription affects the rate but not the extent of repair of cyclobutane pyrimidine dimers in the human adenosine deaminase gene.

To study the relationship between transcription and strand-specific repair of UV-induced cyclobutane pyrimidine dimers, dimer removal was analyzed in a cell line containing two alleles of an inactivated adenosine deaminase (ADA) gene. The cell line was derived from a patient suffering from severe combined immunodeficiency. The disease was caused by a deletion of the complete promoter of the gene as well as the first exon of the ADA gene. This resulted in a true null allele without any detectable transcription (Berkvens, T.M., Gerritsen, E. J. A., Oldenburg, M., Breukel, C., Wijnen, J. T. H., Van Ormondt, H., Vossen, J. M., Van der Eb, A. J., and Meera Khan, P. (1987) Nucleic Acids Res. 15, 9365-9378). Despite this lack of transcription, repair of the ADA gene in this cell line was found to be very efficient with 80% of the dimers being removed within 24 h after UV irradiation. However, the initial rapid repair which is associated with the transcribed strand in normal cells is absent. Dimer removal from two inactive loci, 754 and coagulation factor IX, was much less efficient with only 40% dimers removed after 24 h. From this data, we conclude that transcription is not required for efficient repair of a gene, but forms an additional signal for accelerated repair of the transcribed strand. Furthermore, we suggest that different levels of repair exist between non-transcribed sequences in active genes and those in repressed loci. The results are discussed in terms of the current ideas about the mechanism of preferential DNA repair in human cells.

The Mouse Intestinal Fatty Acid Binding Protein Gene: Nucleotide Sequence, Pattern of Developmental and Regional Expression, and Proposed Structure of Its Protein Product

The rat intestinal fatty acid binding protein (I-FABP) gene has been used as a model to study temporal and spatial differentiation of the gut epithelium while its protein product has been used as a model for examining the atomic details of noncovalent fatty acid-protein interactions. We have isolated the mouse I-FABP gene (Fabpi) and determined its nucleotide sequence. Comparisons of the orthologous mouse, rat, and human I-FABP genes revealed three conserved domains in their otherwise divergent 5' nontranscribed sequences. RNA blot hybridization and multilabel immunocytochemical methods were used to compare the developmental stage-specific patterns of activation of the rat and mouse genes. In addition, Fabpi expression in enterocytes was examined as a function of their differentiation along the crypto-to-villus and duodenal-to-colonic axes of the small intestine. Based on the similar temporal and geographic patterns of mouse and rat I-FABP expression described here and the results of our earlier studies of transgenic mice containing rat Fabpi/human growth hormone fusion genes, we propose that one of the conserved domains, spanning nucleotides -500 to -419 in mouse Fabpi, and/or a 14-bp element, are necessary for establishing and maintaining its region-specific expression along the duodenal-to-colonic axis of the perpetually renewing gut epithelium. Finally, predictions of the structure of mouse I-FABP using the refined 2.0 A model of rat I-FABP, suggest that a proline found at position 69 of the mouse, but not rat, protein may affect its ligand binding properties.

Replication terminator protein of Escherichia coli is a transcriptional repressor of its own synthesis.

We have investigated the regulation of synthesis of the replication terminator protein (Ter) of Escherichia coli and have discovered that the protein is a repressor of its own synthesis at the transcriptional step. Since the synthesis of Ter protein was observed to be down-regulated in vivo, these results are consistent with autoregulation as one control mechanism of Ter protein within the cell. Analysis of the tus gene that encodes the Ter protein revealed that transcription was initiated from a single promoter located within the upstream nontranscribed sequence. In vitro footprinting experiments have revealed that Ter protein prevented binding of RNA polymerase to the promoter sequence when both proteins were incubated with promoter DNA. However, once bound to the promoter, RNA polymerase could not be displaced by Ter protein. Conversely, prebound Ter protein could not be dislodged from its binding site at the promoter when challenged with RNA polymerase. Therefore, Ter protein can serve as a transcriptional repressor of its own synthesis by preventing RNA polymerase from binding to the tus promoter when both proteins are present in the cell milieu.

Developmental regulation of expression of the malate synthase gene in transgenic plants

The cucumber malate synthase (MS) gene, including 1856 bp of 5' non-transcribed sequence, has been transferred into Petunia (Mitchell) and Nicotiana plumbaginifolia plants using an Agrobacterium binary vector. The transferred gene is found in variable copy number in different transformants, and is stably transmitted in each case as a single Mendelian character. Transgene mRNA accumulates in the seedling during the first three days of germination, then declines in amount as the cotyledons emerge from the seed. The decline is more pronounced in light-grown seedlings than in dark-grown seedlings. Expression of the MS transgene is also detected at a low level in petals of transformed Petunia plants. In these respects the pattern of MS gene expression is similar in cucumber and in transformed plants, showing that the transferred DNA fragment contains a functional MS gene. A 1076 bp fragment of 5' sequence was linked to the beta-glucuronidase reporter gene and transferred into Nicotiana, where it was shown to direct temporal and spatial patterns of expression similar to that of the complete MS gene. However, histochemical localisation of beta-glucuronidase activity demonstrated that the chimaeric gene is expressed not only in cotyledons of transgenic plants, but also in endosperm and some hypocotyl cells during early germination. The relevance of these findings to the control of malate synthase gene expression is discussed.

The regulatory VirG protein specifically binds to a cis-acting regulatory sequence involved in transcriptional activation of Agrobacterium tumefaciens virulence genes

Virulence genes of Agrobacterium tumefaciens are induced in parallel in the presence of plant phenolic compounds such as acetosyringone and the two regulatory vir genes virA and virG. In this study we identified a cis-acting regulatory sequence in the 5'-noncoding region of the virE operon that is essential for this activation. To do this, we constructed a series of deletion mutants by using exonuclease Bal 31. Western blot (immunoblot) analysis showed that the 70 base pairs upstream of the transcriptional start site were sufficient for full virE gene induction. A conserved dodecadeoxynucleotide sequence (vir box), which was previously identified in the nontranscribed sequences of all vir genes, was located at 5' end of the minimum required promoter sequence. Deletion of this vir box only completely abolished induction of the virE gene. This demonstrates that the vir box functions as an upstream regulatory sequence. To study the role of the VirG protein in the activation process, we overproduced the native-sized VirG protein in Escherichia coli by fusing the lacZ' start codon ATG with the second virG codon AAA using site-directed mutagenesis. The VirG protein was purified and renatured from E. coli and was shown to bind to a specific sequence in two vir gene promoters. Footprinting analysis of the virE and virB promoters identified the 12-base-pair vir box as the VirG-binding core sequence.

Biochemical characterization of chromatin fractions isolated from induced and uninduced friend erythroleukemia cells

Chromatin fractions from Friend erythroleukemia cells after induction of differentiation by dimethylsulfoxide (DMSO) were compared in their biochemical characteristics to fractions from uninduced cells. Fractions were prepared by extracting chromatin from nuclei after mild micrococcal nuclease treatment with increasing concentrations of NaCl according to Sanders. This procedure has been found to release chromatin containing hyperacetylated histones preferentially. The fractions obtained by this procedure were analysed in respect to the amount of chromatin released, the amount of histone H1, the degree of acetylation of histone H4, the presence of non-histone proteins and the concentration of transcribed and non-transcribed sequences. It was found that the fractions differ in the amount of histone H1 present, in several non-histone proteins and in the acetylation of histone H4, regardless whether induced or uninduced cells were analysed. The distribution of transcribed sequences versus non-transcribed sequences among the fractions was the same, demonstrating that this fractionation procedure, although leading to fractions with biochemical differences, is not able to discriminate functional states of chromatin and that the biochemical characteristics of the fractions may be common to both, active as well as inactive states of chromatin.

Demethylation enhances removal of pyrimidine dimers from the overall genome and from specific DNA sequences in Chinese hamster ovary cells.

We have examined the effects of changes in cytosine methylation on DNA repair in UV-irradiated Chinese hamster ovary (CHO) cells. A hypomethylated derivative of the CHO K1B11 line, B11aza, was established by passaging B11 cells over several months in increasing concentrations of 5-azacytidine; greater than 60% demethylation was consistently demonstrated in these conditioned cells. Following a UV dose of 10 J/m2, the amount of repair replication performed within 24 h was approximately twofold higher in B11aza cells than in control B11 cells. Removal of T4 endonuclease V-sensitive sites (ESS) from specific restriction fragments within and around the dihydrofolate reductase (DHFR) gene was then examined in B11aza cells and compared with that in B11 cells. Although demethylation had little or no effect on repair in the 5' half of the DHFR gene, within a nontranscribed sequence immediately downstream from the gene, or within an extragenic region further downstream from the DHFR gene, significant increases in repair were observed at the 3' end of the DHFR gene and within an extragenic region upstream of the DHFR gene. However, the increases in DNA repair were not accompanied by any changes in overall cellular resistance to UV when colony-forming ability was assayed. We suggest that the level of DNA methylation may play an indirect role in the regulation of DNA repair, perhaps through an effect on chromatin structure or transcriptional activity.

Demethylation enhances removal of pyrimidine dimers from the overall genome and from specific DNA sequences in Chinese hamster ovary cells.

We have examined the effects of changes in cytosine methylation on DNA repair in UV-irradiated Chinese hamster ovary (CHO) cells. A hypomethylated derivative of the CHO K1B11 line, B11aza, was established by passaging B11 cells over several months in increasing concentrations of 5-azacytidine; greater than 60% demethylation was consistently demonstrated in these conditioned cells. Following a UV dose of 10 J/m2, the amount of repair replication performed within 24 h was approximately twofold higher in B11aza cells than in control B11 cells. Removal of T4 endonuclease V-sensitive sites (ESS) from specific restriction fragments within and around the dihydrofolate reductase (DHFR) gene was then examined in B11aza cells and compared with that in B11 cells. Although demethylation had little or no effect on repair in the 5' half of the DHFR gene, within a nontranscribed sequence immediately downstream from the gene, or within an extragenic region further downstream from the DHFR gene, significant increases in repair were observed at the 3' end of the DHFR gene and within an extragenic region upstream of the DHFR gene. However, the increases in DNA repair were not accompanied by any changes in overall cellular resistance to UV when colony-forming ability was assayed. We suggest that the level of DNA methylation may play an indirect role in the regulation of DNA repair, perhaps through an effect on chromatin structure or transcriptional activity.

Variability of human rRNA genes: inheritance and nonrandom chromosomal distribution of structural variants of nontranscribed spacer sequences.

Human rRNA genes contain variable regions, one of which is located in nontranscribed spacers (NTSs) closely downstream from the 3'-end of the transcribed region. This polymorphism may be detected by means of blot hybridization analysis as a set of distinct restriction fragments corresponding to this part of the rRNA genes. We have analyzed DNA of 51 individuals and found eight structural NTS variants of this region; two of these were common to all individuals analyzed, and six others were found in different combinations and with different frequencies. The copy number of each variant also differed but was not less than 10-20 copies per cell. The analysis of DNA isolated from leukocytes of the members of 11 families indicated that some of the structural variants (of the NTS region) are inherited as a single Mendelian locus. We propose that rRNA genes that belong to one particular structural variant form clusters on separate chromosomes. To test this proposition, we developed a combined method, including AgNO3-staining of chromosomes, in situ hybridization, and DNA analysis with methylation-sensitive restrictases, and used it for study of persons who had methylated rRNA genes located on AgNO3-negative nucleolar organizers. It was found that in three of four cases methylated genes really belonged to one structural variant. This approach may be used for detailed localization of separate classes of NTS structural variants of human rRNA genes.

Cis-acting elements involved in the regulated expression of a human hsp70 gene

Expression of reporter genes under the control of upstream sequences of a human hsp70 gene was examined in microinjected Xenopus oocytes, transfected monkey COS and human HeLa cells. The genes were strictly heat-regulated in all three cell systems. A 69 nucleotide long segment of hsp70 5′ non-transcribed sequence that included at least one functional heat shock regulation sequence was sufficient for heat-controlled expression in Xenopus and monkey cells but not in human HeLa cells. An additional segment of about 200 nucleotides in length was required for optimal activity. This segment contains two heat shock regulation elements, each of which appears to contribute to the overall activity in heat-treated human cells. Upstream non-transcribed sequences of the human hsp70 gene are capable of conferring heat regulation on a heterologous promoter. The potential roles in transcription regulation of a bending center in the TATA box region, a CCAAT-like sequence and some of many potential Sp1 binding sites in the hsp70 5′ non-transcribed region were investigated.

The characterization of ribosomal RNA gene chromatin from Physarum polycephalum.

We have isolated ribosomal RNA gene (rDNA) chromatin from Physarum polycephalum using a nucleolar isolation procedure that minimizes protein loss from chromatin and, subsequently, either agarose gel electrophoresis or metrizamide gradient centrifugation to purify this chromatin fraction (Amero, S. A., Ogle, R. C., Keating, J. L., Montoya, V. L., Murdoch, W. L., and Grainger, R. M. (1988) J. Biol. Chem. 263, 10725-10733). Metrizamide-purified rDNA chromatin obtained from nucleoli isolated according to the new procedure has a core histone/DNA ratio of 0.77:1. The major core histone classes comigrate electrophoretically with their nuclear counterparts on Triton-acid-urea/sodium dodecyl sulfate two-dimensional gels, although they may not possess the extent of secondary modification evident with the nuclear histones. This purified rDNA chromatin also possesses RNA polymerase I activity, and many other nonhistone proteins, including two very abundant proteins (26 and 38 kDa) that may be either ribonucleoproteins or nucleolar matrix proteins. Micrococcal nuclease digestion of the metrizamide-purified rDNA chromatin produces particles containing 145-base pair DNA fragments identical in length to those in total chromatin and which contain both transcribed and nontranscribed rDNA sequences. Some smaller fragments (30, 70, and 110 base pairs) are also seen, but their sequence content is not known. These particles sediment uniformly at 11 S in sucrose gradients containing 15 mM NaCl, and at 4-11 S in gradients containing 0.35 M NaCl. Particles enriched in gene or nontranscribed spacer sequences are not resolved in these sucrose gradients or in metrizamide gradients. Our findings suggest that the rDNA chromatin fraction we have identified contains transcriptionally active genes and that an organized, particle-containing structure exists in active rDNA chromatin.

Reproducible compartmentalization of individual chromosome domains in human CNS cells revealed by in situ hybridization and three-dimensional reconstruction.

Specific chromosome domains in interphase nuclei of neurons and glia were studied by three-dimensional (3-D) reconstruction of serial optical sections from in situ hybridized human CNS tissue. Overall patterns of centromere organization, delineated with alphoid repeats, were comparable to those seen in mouse, and are clearly conserved in mammalian evolution. Cloned probes from other individual chromosome domains were used to define interphase organization more precisely. Homologous chromosomes were spatially separated in nuclei. In large neurons, probes specific for 9q12, or 1q12 showed that at least one homolog was always compartmentalized together with centromeres on the nucleolus, while the second signal either abutted the nucleolus or was on the nuclear membrane. A telomeric Yq12 sequence also localized together with perinucleolar centromeres in a completely non-Rabl orientation. In astrocytes, these three chromosome regions were on the membrane and not necessarily associated with nucleoli. Therefore there are different patterns of interphase chromosome organization in functionally distinct cell types. In contrast to the above domains, a 1p36.3 telomeric sequence embedded in a large Alu-rich and early replicating chromosome region, was always found in an interior euchromatic nuclear compartment in both neurons and glial cells. In double hybridizations with 1q12 and 1p36.3 probes, 1p arms were clearly separated in all cells, and arms projected radially into the interior nucleoplasm with non-Rabl orientations. There was no absolute or rigid position for each 1p arm with respect to each other or to the major dendrite, indicating that individual chromosome arms may be dynamically positioned even in highly differentiated cell types. We suggest that centromeric and other highly repeated non-transcribed sequence domains may act as general organizing centers for cell type specific interphase patterns that are conserved in mammalian evolution. Such centers would allow selected groups of chromosome arms to extend into (and contract from) an interior, presumably transcriptionally active, nuclear compartment.

Human repair gene restores normal pattern of preferential DNA repair in repair defective CHO cells.

The pattern of preferential DNA repair of UV-induced pyrimidine dimers was studied in repair-deficient Chinese hamster ovary (CHO) cells transfected with the human excision repair gene, ERCC-1. Repair efficiency was measured in the active dihydrofolate reductase (DHFR) gene and in its flanking, non-transcribed sequences in three cell lines: Wild type CHO cells, a UV-sensitive excision deficient CHO mutant, and the transfected line of the mutant carrying the expressed ERCC-1 gene. The CHO cells transformed with the human ERCC-1 gene repaired the active DHFR gene much more efficiently than the non-transcribed sequences, a pattern similar to that seen in wild type CHO cells. This pattern differs from that previously reported in CHO cells transfected with the denV gene of bacteriophage T4, in which both active and non-transcribed DNA sequences were efficiently repaired (Bohr and Hanawalt, Carcinogenesis 8: 1333-1336, 1987). The ERCC-1 gene product may specifically substitute for the repair enzyme present in normal hamster cells while the denV product, T4 endonuclease V, does not be appear to be constrained in its access to inactive chromatin.

Transcriptional regulation of ribosomal proteins during a nutritional upshift in Saccharomyces cerevisiae.

The relative rates of synthesis of Saccharomyces cerevisiae ribosomal proteins increase coordinately during a nutritional upshift. We constructed a gene fusion which contained 528 base pairs of sequence upstream from and including the TATA box of ribosomal protein gene rp55-1 (S16A-1) fused to a CYC1-lacZ fusion. This fusion was integrated in single copy at the rp55-1 locus in the yeast genome. During a nutritional upshift, in which glucose was added to cells growing in an ethanol-based medium, we found that the increase in the relative rate of synthesis of the beta-galactosidase protein product followed the same kinetics as the change in relative rates of synthesis of several ribosomal proteins measured in the same experiment. This demonstrates that the nontranscribed sequences upstream from the rp55-1 gene, which are present in the fusion, are sufficient to mediate the change in rates of synthesis characteristic of ribosomal proteins under these conditions. The results also suggest that a change in transcription rates is mainly responsible for the increase in relative rates of synthesis of ribosomal proteins during a nutritional upshift in S. cerevisiae.

Transcriptional regulation of ribosomal proteins during a nutritional upshift in Saccharomyces cerevisiae.

The relative rates of synthesis of Saccharomyces cerevisiae ribosomal proteins increase coordinately during a nutritional upshift. We constructed a gene fusion which contained 528 base pairs of sequence upstream from and including the TATA box of ribosomal protein gene rp55-1 (S16A-1) fused to a CYC1-lacZ fusion. This fusion was integrated in single copy at the rp55-1 locus in the yeast genome. During a nutritional upshift, in which glucose was added to cells growing in an ethanol-based medium, we found that the increase in the relative rate of synthesis of the beta-galactosidase protein product followed the same kinetics as the change in relative rates of synthesis of several ribosomal proteins measured in the same experiment. This demonstrates that the nontranscribed sequences upstream from the rp55-1 gene, which are present in the fusion, are sufficient to mediate the change in rates of synthesis characteristic of ribosomal proteins under these conditions. The results also suggest that a change in transcription rates is mainly responsible for the increase in relative rates of synthesis of ribosomal proteins during a nutritional upshift in S. cerevisiae.

Identification of a sequence element in the promoter of the Drosophila melanogaster hsp23 gene that is required for its heat activation.

The expression of Drosophila melanogaster hsp23-Escherichia coli beta-galactosidase hybrid genes containing different segments of the 5' non-transcribed sequence of the hsp23 gene has been examined at the RNA and protein levels in Xenopus oocytes. Transcription of the hybrid genes is initiated correctly. Mutant genes with hsp23 gene promoter segments of at least 140 bp in length are strongly heat-activated while genes with shorter promoter segments are expressed constitutively and at low levels. This maps an element required for the heat-controlled expression of the D. melanogaster hsp23 gene to a region, approximately 140 bp upstream from the start of the transcription site, which contains a sequence (CGAGAAGTT-TCGTGT) that is closely related to the one responsible for the heat regulation of the hsp70 gene. These findings demonstrate the importance of this regulatory sequence for a second hsp gene and support the notion that hsp genes are heat-regulated by a common mechanism. The functional element in the hsp23 gene promoter is located greater than 80 bp further upstream from the TATA box than the relevant element in the hsp70 gene promoter. Even though other related sequences are present further upstream and downstream from the functional element, they play at most an auxiliary role in the regulation of hsp23 gene expression.

The heat shock consensus sequence is not sufficient for hsp70 gene expression in Drosophila melanogaster.

A hybrid gene in which the expression of an Escherichia coli beta-galactosidase gene was placed under the control of a Drosophila melanogaster 70,000-dalton heat shock protein (hsp70) gene promoter was constructed. Mutant derivatives of this hybrid gene which contained promoter sequences of different lengths were prepared, and their heat-induced expression was examined in D. melanogaster and COS-1 (African green monkey kidney) cells. Mutants with 5' nontranscribed sequences of at least 90 and up to 1,140 base pairs were expressed strongly in both cell types. Mutants with shorter 5' extensions (of at least 63 base pairs) were transcribed and translated efficiently in COS-1 but not at all in D. melanogaster cells. Thus, in contrast to the situation in COS-1 cells, the previously defined heat shock consensus sequence which is located between nucleotides 62 and 48 of the hsp70 gene 5' nontranscribed DNA segment is not sufficient for the expression of the D. melanogaster gene in homologous cells. A second consensus-like element 69 to 85 nucleotides upstream from the cap site is postulated to be also involved in the heat-induced expression of the hsp70 gene in D. melanogaster cells.