Nucleic Acids Research

Growth of cyanobacterium

Anabaena variabilis strain M3 was grown photoautotrophically as described previously (2,13). In standard temperature shift experiments cells grown at 38°C were transferred to 22°C in the light. The temperature of the culture medium reached the new temperature within 2 min. In the experiments in Figure 2A cells that had been grown at 38°C were transferred to various temperatures ranging from 45 to 10°C. Two hours later the cells were harvested for extraction of RNA. In the experiments in Figure 2B, cells were transferred from 38 to 22°C with addition of various inhibitors or in the dark. In the experiments in Figure 3, cells that had been grown at 38°C were first transferred to 22°C to induce accumulation of rbpA1 transcript. Two hours later the culture was divided into four 100 ml portions. Rifampicin was added to two of the portions to a final concentration of 15 µg/ml and then two portions (one with rifampicin and the other without it) were transferred to 38°C, while the other two portions were kept at 22°C. For purification of DNA binding proteins, cells grown at 38°C were collected by centrifugation and stored frozen at -75°C until use.

RNA blot analysis

RNA was extracted from the cells and subjected to RNA blot analysis as described previously (2). In most experiments radiolabeled probes were used, whereas in the experiment in Figure 5A a digoxigenin (DIG)-labeled probe was used.

Construction and expression of rbpA1-lacZ fusion genes

A 12.3 kb genomic DNA fragment (clone A2a) containing the rbpA1 gene and its upstream region was obtained by screening a genomic library of A.variabilis strain M3 in [lambda]GEM11 with the coding sequence of rbpA1 as a probe. A terminal 4.7 kb EcoRI-XhoI fragment (this XhoI site was located in the cloning site of the vector) was subcloned in pBluescript SK+ and then used to construct a 3[prime] deletion series. This plasmid was named p26EXh. The 4.7 kb fragment contained a 3.5 kb sequence upstream of the rbpA1 gene, the rbpA1 gene, the rpsU gene and a short downstream sequence until nucleotide (nt) 1257 from the transcription start site (Fig. 5). The transcription start site of the rbpA1 gene was mapped in a previous study (14). The plasmids for transformation were constructed in the following way (Fig. 4). A linker fragment (HindIII, PstI, SmaI, KpnI, EcoRI, a 0.4 kb stuffer DNA fragment, BamHI) was inserted into the multi-linker (XbaI, HindIII, PstI, BamHI) located in front of the lacZ gene of pRL576 [a derivative of pRL119 (15) that contains the promoter of the psbA gene of Amaranthus hybridus in front of the npt gene, obtained from J.Elhai]. Then this plasmid (pRL576m) was digested with HindIII, blunted with the Klenow fragment of DNA polymerase I and then digested with KpnI. This was ligated with a 4.7 kb SmaI-KpnI fragment of p26EXh that contained the whole genomic insert with the linker sequences of pBluescript. The resultant plasmid (pRL576rbp) containd a 4.7 kb genomic DNA that was placed upstream of the lacZ gene through a linker sequence XhoI-KpnI-EcoRI-(stuffer)-BamHI. The XhoI and KpnI sites were cleaved and a series of deletions from the XhoI site was created by digestion with exonuclease III followed by digestion with mung bean nuclease (deletion kit from Takara). Then the BamHI site was cleaved to remove the stuffer and blunted. The blunted BamHI site was ligated to the blunt end created by the deletion reaction. Since these plasmids did not contain a cyanobacterial replication origin they were not expected to be maintained in A.variabilis cells after conjugal transfer (16). Instead, the plasmid was expected to be integrated into the genome by homologous recombination. The recombinants obtained by selecting with neomycin (30 µg/ml) contained the lacZ gene integrated in the cyanobacterial genome at the fusion site created within the rbpA1 region in the transforming plasmid. Another copy of the rbpA1 gene (and rpsU gene located just downstream) remained intact downstream of the integration site, as reported in a similar experiment (15). These copies of the rbpA1 gene were confirmed by Southern blot hybridization before and after digesting the genomic DNA with EcoRI: wild-type and recombinant copies of the rbpA1 gene were detected as DNA fragments of 6.5 and [sim]8.5 kb (depending on the extent of deletion) respectively.

Reporter shuttle plasmids were assembled from the cyanobacterial replication origin of pRL488 (15), which was originally derived from pDU1 of Nostoc sp. PCC 7924 (17), the lacZ gene and the kanamycin resistance gene (from Tn5) of pRL576, and named pRA101 (Fig. 4). The following primers were used to prepare DNA fragments of the promoter region of the rbpA1 gene. The upstream primer (UP) was 5[prime]-GAGGTACCAGAGCGAATTGCTGAGAA (nt 793-810 in GenBank accession no. D17710). The underlined sequence is a KpnI site introduced for manipulation of the PCR products. The downstream primers (DN) were: DN3, 5[prime]-ATGGATCCGTTACTGTCTAGAGATGTAAAT (1072-1051); DN1, 5[prime]-TGGGATCCAAAATTGTTACTGTCTAGAG (1081-1059); DN2, AAGGATCCTTGAGTAACTTCGTACG (1132-1115). Added BamHI sites are underlined. PCR-generated DNA fragments of the promoter region of the rbpA1 gene were inserted into the multiple linker region located upstream of the lacZ gene of pRA101. This series of plasmids was designed to exist as plasmids in A.variabilis cells after conjugal transfer, as reported (15). The results of Southern blot analysis with or without digestion with EcoRI indicated that the copy number of all three plasmids used was about four per genome. In this analysis the chromosomal copy was detected as a 6.5 kb EcoRI fragment, whereas the plasmid copy was detected as 4.3 and 9.4 kb fragments.

The transcript of the chimeric lacZ gene was analyzed by RNA blot analysis with a DIG-labeled lacZ probe as described previously (2). [beta]-Galactosidase activity of intact cells was determined essentially according to the method developed for E.coli (18), except that the density of cyanobacterial cells was measured by turbidity at 750 nm (instead of 600 nm) and the reaction mixture was centrifuged at 15 000 g for 2 min to obtain clear supernatant. One unit of activity was defined as 10 × A₄₂₀/(A₇₅₀ × t × V), where A₇₅₀ is the absorbance at 750 nm due to turbidity of cyanobacterial cells, V is the volume (ml) of cyanobacterial culture added to the reaction mixture (normally 0.1 ml), t is the reaction time (min) and A₄₂₀ is the absorbance at 420 nm due to hydrolysis of o-nitrophenyl-[beta]-d-galactoside. Here we used a scaling factor of 10 instead of the 1000 used in the original method for E.coli, for simplicity.

Gel mobility shift analysis

Gel mobility shift analysis for DNA binding activity was performed essentially according to Stone et al. (19). The basal plasmid for preparation of substrate DNA (pUTR) was constructed by cloning the 5[prime]-UTR of the rbpA1 gene [from the transcription start site to nt 151; see Sato et al. (14) for the revised sequence], which had been amplified by PCR with primers LONG2F (5[prime]-TAGGATCCATAGATGCAGAGAATTGGCT, restriction site underlined) and LONG2R (5[prime]-ATGGTACCCTCCAAAATTGTTACTGTCTAGAG), into pBluescript SK+. A DIG-labeled DNA probe was prepared by PCR with DIG PCR labeling mixture (Boehringer) and primers LONG2F and LONG2R using pUTR as template. DNA probe ([sim]50 ng) and protein ([sim]5 µg, except for column fractions, which were not quantitated) as well as poly(dI-dC) (0.5 µg; Pharmacia) were mixed in a 10 µl reaction mixture containing 50 mM KCl, 20 mM HEPES-KOH, pH 7.7, 5 mM MgCl₂, 10% glycerol, 0.01% bromophenol blue. Here protein concentration was measured by the method of Lowry et al. (20) with bovine serum albumin as the standard. After standing on ice for 30 min the mixture was loaded onto a non-denaturing 5% polyacrylamide gel (80 mm height × 84 mm width) which had been pre-electrophoresed for 2 h. The electrophoresis buffer contained 6.7 mM Tris, 3.3 mM acetic acid, 1 mM EDTA, pH 7.9. After electrophoresis at 15 V/cm for [sim]1 h DNA was transferred to a nylon membrane (Hybond N⁺; Amersham). DIG-labeled DNA was detected according to the standard procedure recommended by the manufacturer (Boehringer). The specific competitor, DNA 1 (Fig. 6B), was a synthetic double-stranded DNA, shown in Figure 5B.

Purification of DNA binding proteins

All manipulations were done at 4°C or on crushed ice. Frozen cells ([sim]3 g fresh weight or 150 mg protein) of A.variabilis from 2.5 l cultures were suspended in 50 ml TEGP medium [20 mM Tris-HCl, pH 7.5, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride (PMSF), 5 mM 2-mercaptoethanol, 10% glycerol]. The cells were lysed by digestion with 1 mg/ml lysozyme for 30 min, followed by two passages through a nitrogen-driven pressure cell (Parr homogenizer) operated at [sim]100 kg/cm². The cells were further treated with a sonicator (model UR-20P; Tomy Seiko, Tokyo) five times for 30 s each. Cell debris was removed by centrifugation at 2000 g for 5 min. Ammonium sulfate was added to 30% saturation. After standing for 1 h the mixture was centrifuged at 10 000 g for 10 min. Then further ammonium sulfate was added to the supernatant to 50% saturation. After 1 h the mixture was centrifuged at 10 000 g for 10 min. The precipitate was dissolved in 25 ml TEGP buffer and dialyzed against TEGP buffer.

Figure 1. Time course of accumulation of rbpA1 transcript after a temperature shift from 38 to 22°C. Cells of A.variabilis M3 grown at 38°C were transferred to 22°C at time zero. Aliquots of 25 ml culture were taken at 10, 30 min, 1, 2, 3, 5 and 10 h and then RNA was extracted. The total RNA was electrophoresed and then transferred to a nylon membrane (Biodyne A). The membrane was probed with a radiolabeled rbpA1 probe. The intensity of each band was quantified by densitometry and plotted as a function of time. The level at 2 h was taken as 100.

Figure 2. Effects of temperature, light and various inhibitors on accumulation of rbpA1 transcript. (A) Effects of temperature. Cells grown at 38°C were divided into 25 ml aliquots and then transferred to various temperatures as indicated. Two hours later the cells were harvested and RNA was extracted. The abundance of rbpA1 transcript was quantified by slot-blot RNA blot hybridization with a radiolabeled rbpA1 probe. A series of 2-fold dilutions was used. The relative abundance of rbpA1 transcript was determined by densitometry and extrapolation of the series of data to infinite dilution. (B) Effects of light and inhibitors. Cells grown at 38°C were divided into 25 ml aliquots. Various inhibitors were added as indicated and then the tubes were transferred to 22°C or kept at 38°C as indicated. Some of the tubes were wrapped with aluminum foil to keep the culture in the dark. Two hours after treatment the cells were harvested and RNA was extracted. The abundance of rbpA1 transcript was determined as described above. Rif, rifampicin (15 µg/ml); DCMU, 3-(3,4-dichlorophenyl)-1,1-dimethylurea (20 µM or 4.7 µg/ml); Cm, chloramphenicol (15 µg/ml); DMF, dimethylformamide (0.1%). DMF was used as a solvent for Rif, DCMU and Cm.

The dialyzate was loaded onto a column of Q-Sepharose (5 ml Hi-Trap Q; Pharmacia) that had been washed in advance with basal column buffer (50 mM Tris-HCl, pH 7.5, 0.1 mM EDTA, 10% glycerol, 1 mM PMSF), 1 M column buffer (1 M NaCl in basal column buffer) and again basal column buffer. Then the column was eluted with a gradient of 0-0.5 M NaCl in basal column buffer over 40 min at a flow rate of 1 ml/min. Fractions of 1 ml were collected. Appropriate fractions were pooled and analyzed for DNA binding activity. Pooled fractions 3-5 (Fig. 7) were dialyzed against HEGP buffer (20 mM HEPES-KOH, pH 7.0, 1 mM EDTA, 0.1 mM PMSF, 10% glycerol) and further subjected to affinity purification.

An affinity matrix was prepared by conjugating a biotin-labeled DNA probe with magnetic particles (Dynabeads M280 Streptavidin; Dynal, Oslo). The biotin-labeled DNA probe had been prepared by PCR with a T7 primer (Stratagene) and a biotin-labeled T3 primer using pUTR (see above) as template. Each of the dialyzed column fractions was mixed with the DNA-Dynabead conjugate in 50 mM Dynabead buffer (50 mM KCl in basal Dynabead buffer, 20 mM HEPES-KOH, pH 8.0, 0.1 mM PMSF, 20% glycerol). The mixture was rocked for 2 h. Then the magnetic particles were collected by centrifugation at 2000 g for 2 min followed by magnetic separation according to the manufacturer's protocol. The particles were washed twice with 50 mM Dynabead buffer. Then bound proteins were eluted three times with 200 µl each 0.3 M Dynabead buffer (0.3 M KCl in basal Dynabead buffer). Finally, tightly bound proteins were eluted three times with 200 µl each 1 M Dynabead buffer (1 M KCl in basal Dynabead buffer). The eluted fractions were analyzed by SDS-PAGE on a 12% polyacrylamide gel (21). The gel was stained with silver nitrate (Silver Stain Plus kit; Boehringer).

RESULTS

Cold-induced accumulation of the transcript of the rbpA1 gene

The level of rbpA1 transcript increased dramatically after the temperature shift from 38 to 22°C (Fig. 1). The increase was detected as early as 10 min after the temperature shift. The level of rbpA1 transcript became maximal at 3 h and then slightly decreased to reach a plateau. We have chosen 38°C as the high temperature because this is the optimal growth temperature of this organism ([sim]1.8 divisions/day) (14). We normally use 22°C as the low temperature (1,2,14) because this is the lowest temperature at which Anabaena cells grow at a reasonably high rate ([sim]1.2 divisions/day). Below this temperature cells grow but the color of the cells is significantly yellower, a symptom of excess irradiance. After the temperature shift from 38 to 22°C, growth of the cells is significantly slowed, to 0.9 divisions/day, during the first 10 h (22).

Figure 3. Stability of rbpA1 transcript at different temperatures. Cells grown at 38°C were transferred to 22°C and kept at this temperature for 2 h. Then they were divided into four tubes (numbered 1-4). Rifampicin (15 µg/ml) was added to tubes 1 and 2, and immediately tubes 1 and 3 were transferred to 22°C, while tubes 2 and 4 were kept at 38°C. Aliquots of 25 ml were withdrawn from tubes 1 and 2 at 5, 10, 25 and 60 min. The cells in tubes 3 and 4 were also analyzed at 60 min. RNA was extracted from the cells and analyzed by slot-blot RNA blot hybridization as in Figure 2 (not shown). After densitometry the relative abundance of rbpA1 transcript was obtained by extrapolating each data set to infinite dilution and plotted on a logarithmic scale against the time after addition of rifampicin.

Figure 4. Construction of plasmids for analysis of activity of the rbpA1 promoter. Two types of plasmids were constructed. One type (plasmid for integration) bears a piece of the 5[prime]-UTR of rbpA1 fused to the lacZ gene, as well as the npt (neomycin phosphotransferase) gene and the replication of origin of pBR322. The other type (shuttle plasmid) contained, in addition, a cyanobacterial replication origin (the whole pDU1 plasmid from Nostoc sp. PCC7524). In construction of both types of plasmids, pRL576 [a derivative of pRL119 (15)] was used as the starting material. A short piece of DNA fragment having a multi-linker sequence was inserted in the original multi-linker of pRL576 to give pRL576m. To construct the integration plasmids a 4.7 kb genomic DNA fragment containing the rbpA1 gene was cut from p26EXh and then inserted into the multi-linker of pRL576m, to obtain pRL576rbp. Then pRL576rbp was cut at the KpnI site (3[prime]-protruding) and the XhoI site (5[prime]-protruding) and the DNA subjected to deletion from the XhoI site. Variously deleted DNA was further cut at the BamHI site. Both ends were blunted and then ligated to yield plasmids that bear a 3[prime]-deleted rbpA1 gene sequence which was transcriptionally fused to the lacZ gene. To construct the shuttle plasmids pRL576m and pRL488 were digested with PstI. The large fragment of pRL576m (lacZ, pBR ori and the downstream half of npt) was ligated with the large fragment of pRL488 (pDU1 and the upstream half of npt), to give pRA101. By this ligation the npt gene was regenerated. Various DNA fragments of the 5[prime]-UTR region were prepared by PCR and inserted into the multi-linker of pRA101. Note that the filler DNA of the multi-linker was removed during this step.

Figure 5. Deletion analysis of the rbpA1 gene. (A) Activity of the rbpA1 promoter with the 3[prime]-deletion series. (Upper) (integration) A 3[prime]-deletion series was generated from a 4.7 kb genomic DNA containing the rbpA1 gene and was transcriptionally fused to the lacZ gene (Fig. 4). These chimeric genes were introduced into A.variabilis cells with a suicide plasmid vector that does not replicate in the cyanobacterial cell and were allowed to integrate into the genome. (Lower) (plasmid) Short DNA fragments prepared by PCR were inserted upstream of the lacZ gene in a shuttle vector (pRA101) and the resultant plasmids introduced into A.variabilis cells. In both cases the transformant cells were selected by resistance to kanamycin (30 µg/ml). After confirmation by Southern hybridization and PCR, successful transformants were grown at 38°C and then transferred to 22°C. For the analysis of RNA, cells were sampled just before the temperature shift (0 h) and at 2 h after the temperature shift. For measurement of [beta]-galactosidase activity, cells were sampled at zero time and at 24 h after the temperature shift. The chimeric lacZ transcript was detected by Northern blot hybridization, while the activity of [beta]-galactosidase was measured with o-nitrophenyl-[beta]-d-galactoside as substrate. The label `increase' represents the ratio of activities at 24 and 0 h. Each number in the diagram of constructs indicates position from the transcription start site, which is 932 in GenBank database entry D17710. (B) Putitive cis-acting elements in the 5[prime]-UTR of the rbpA1 gene. The 5[prime]-UTRs of rbp genes known to be regulated by low temperature are aligned to emphasize conservation of an inverted repeat (arrows). The initiation codon ATG is shown in special characters. The 3[prime]-ends of lacZ fusions used in the experiments in Figures 4 and 5 are shown above with residue number from the transcription start site. DNA 1 is a synthetic DNA that was used in the competition experiment in Figure 6. GenBank accession nos: rbpA1, D17710; rbpA2, L20890 and AB003690; rbpA3, X92980; rbpB, D50459; rbpC, D49424. Note that rbpD (D49425), which is not regulated by temperature, was not included in this alignment, since the 5[prime]-UTR of rbpD is totally different from the 5[prime]-UTR of the cold-regulated rbp genes shown here.

Figure 6. Gel mobility shift analysis of DNA binding proteins that have affinity with the 5[prime]-UTR. (A) Detection of DNA binding activity in the extract of A.variabilis cells grown at 38 and 22°C. The soluble protein extracts were fractionated with ammonium sulfate and the fractions at 30-50% saturation and 50-80% saturation were used in the analysis. Protein extract (5 µg) was mixed with a 151 bp DIG-labeled DNA probe (50 ng) as well as a non-specific competitor (0.5 µg dI-dC) and analyzed by PAGE. DIG was detected by immunoblotting. (B) Efficient competition by a 38 bp DNA (DNA 1). The protein extract (30-50% ammonium sulfate fraction) from the 38°C grown cells was mixed with the 151 bp DIG-labeled DNA probe, a non-specific competitor (dI-dC) and varying concentrations of either DNA 1 (2.5- to 250-fold molar excess) or yeast tRNA (0.01-1 µg).

Figure 7. Separation of DNA binding activities by anion exchange column chromatography. The 30-50% ammonium sulfate fraction was subjected to anion exchange column chromatography (A). After washing with basal column buffer, proteins were eluted with a gradient of NaCl from 0 to 0.5 M. Fractions were pooled according to the peaks detected by absorbance at 280 nm. After dialysis DNA binding activity of each pooled fraction was analyzed by gel mobility shift analysis (B). AS fraction, the 30-50% ammonium sulfate fraction.

Effect of temperature was studied at 2 h after the temperature shift (Fig. 2A). The results of slot-blot hybridization were quantified by densitometry and expressed in relative values. When cells that had been grown at 38°C were transferred to various temperatures the level of rbpA1 transcript was highest at 16°C. Accumulation of transcript remained low between 41 and 33°C. At 45°C the transcript seemed to accumulate to only a certain level (lane 1). This might be either a result of a general stress response or an artefact, since A.variabilis cells do not grow above 42°C (14).

Figure 2B shows the effects of darkness and drugs on accumulation of rbpA1 transcript after the temperature shift from 38 to 22°C. Although incubation at 38°C in the dark reduced the level of rbpA1 transcript to nearly zero, darkness had no effect on the level of rbpA1 transcript at 22°C. Rifampicin completely abolished accumulation of rbpA1 transcript, indicating that accumulation of rbpA1 transcript depends on continued transcription. Dimethylformamide (0.1%), which was used as solvent for the three drugs used in the present study, promoted accumulation of rbpA1 transcript by a factor of 1.6. This might be the result of a general stress response to this organic solvent. We also tested the effect of dimethylformamide on the level of rbpA1 transcript in cells growing at 38°C, but no significant difference in the level of transcript was detected (data not shown). The enhancement of transcript accumulation induced by dimethylformamide seems, therefore, limited to temperature shifted cells. 3-(3,4-Dichlorophenyl)-1,1-dimethylurea (DCMU), an inhibitor of photosystem II, had no further effect on the level of rbpA1 transcript. Chloramphenicol (15 µg/ml) only partially reduced the level of rbpA1 transcript at a concentration that inhibits protein synthesis in this organism by 94% (23). This result suggests that protein synthesis is essentially not required for induction of accumulation of rbpA1 transcript. The partial reduction in transcript level might be explained by general inhibition of cellular activity during the 2 h incubation period.

Increased stability of the rbpA1 transcript at low temperature

As a first step in analyzing the mechanism of accumulation of rbpA1 transcript at low temperature we analyzed the stability of rbpA1 transcript at two different temperatures (Fig. 3). In this experiment cells that had been grown at 38°C were transferred to 22°C and grown at 22°C for 2 h, to induce accumulation of rbpA1 transcript. Then the culture was divided into aliquots and each aliquot was incubated at 22 or 38°C with or without rifampicin. At 38°C the level of the rbpA1 transcript decreased rapidly, with a half-life of [sim]5 min. After 60 min only 1-2% of transcript remained, both with or without rifampicin. In contrast, >30% of transcript remained in the presence of rifampicin at 22°C. In the absence of rifampicin the level of transcript remained unchanged after 1 h. The same result was obtained by Northern blot analysis, in which no cross-hybridizing bands were detected (data not shown). The half-life of rbpA1 transcript at 22°C was estimated to be 18.5 min. Thus the stability of rbpA1 transcript in vivo was markedly dependent on the growth temperature. We should also note that the level of the transcript rapidly became very low at 38°C both with or without rifampicin. This might indicate that transcription of this gene is slow after the shift of temperature to 38°C.

Involvement of the 5[prime]-UTR in cold regulation of the rbpA1 gene

To study possible transcriptional regulation we constructed a series of 3[prime] deletions of the rbpA1 gene that were transcriptionally fused to the lacZ reporter (Fig. 4) and then these constructs were introduced into cells of A.variabilis (Fig. 5A). It should be mentioned here that the rbpA1 gene is co-transcribed with the downstream gene rpsU. Therefore, the longest chimeric gene includes the promoter and the untranslated region, as well as the rbpA1 gene and the rpsU gene. The lacZ gene was obtained from plasmid pRL576 and contained many termination codons in the three frames within the leader sequence before the initiation codon. Therefore, all the constructs in the present experiments are transcriptional fusions. In the first experiment the chimeric genes were introduced into A.variabilis cells on a suicide vector, i.e. a vector that does not replicate in cyanobacterial cells, such that the lacZ gene would be integrated into the expected sites within the rbpA1 gene in the chromosome. In this type of transformation the plasmids were expected to be integrated into the genome by a single homologous recombination event and, therefore, an intact copy of the rbpA1 gene is present downstream of the integration site. This was confirmed by Southern blot analysis (data not shown). When the lacZ gene was placed downstream of position +1257 from the transcription start site of the rbpA1 gene, i.e. after the transcriptional terminator (14), no expression of the lacZ gene was observed. When the fusion point was located at nt +832, +680 or +421 lacZ transcript accumulated after the temperature shift from 38 to 22°C. In addition, [beta]-galactosidase activity increased [sim]4-fold after the temperature shift. The increase in enzymatic activity was not as rapid as the increase in the level of rbpA1 transcript. At 2 h after the temperature shift the activity of [beta]-galactosidase had increased by only 30% (not shown). Therefore, the activity at 24 h after the temperature shift is presented in Figure 5A. When the lacZ gene was placed downstream of position +134 the level of lacZ transcript was no longer dependent on growth temperature. The level of transcript was somewhat low in this case. The activity of [beta]-galactosidase, on the other hand, remained at a constant high level both before and after the temperature shift. When the fusion point was located upstream of the transcriptional start site (position -45), expression of the lacZ gene was quite low, as expected.

To analyze expression in more detail we constructed a series of plasmids based on shuttle vector pRA101, each of which contained a short DNA fragment in the 5[prime]-untranslated region (5[prime]-UTR) of the rbpA1 gene fused to the lacZ reporter (Fig. 4). In this case the plasmids were designed to be maintained as episomes in the transformant A.variabilis cells. The copy number as determined by Southern blot hybridization was about four in the three constructs used (data not shown). The short DNA fragments, one from -139 to +201 and another from -139 to +149, both conferred cold regulation on the lacZ gene. This was evident in both transcript level and [beta]-galactosidase activity (Fig. 5A, lower). Regulation as monitored by enzyme activity was only about two. In contrast, the DNA fragment from -139 to +140 conferred high constitutive expression on the lacZ reporter at both temperatures. These results suggest that the promoter of the rbpA1 gene is constitutive but the 5[prime]-UTR sequence down to +149 is necessary to reduce transcription of the rbpA1 gene at high temperature. Alternatively, the 5[prime]-UTR, when transcribed, might increase degradation of rbpA1 transcript at high temperature. Another possibility is that the 5[prime]-UTR is involved in temperature-dependent transcription attenuation.

Identification of proteins that bind to the 5[prime]-UTR of the rbpA1 gene

We tried to identify proteins that might bind to the 5[prime]-UTR of the rbpA1 gene. During the initial stages of the experiment we tried to find RNA binding proteins that can bind to the 5[prime]-UTR RNA, but this attempt was unsuccessful. We also know that expression of the rbpA1 gene disrupted with a kanamycin resistance cassette is induced at low temperature, which indicates that the RbpA1 protein itself is not involved in temperature-dependent expression of the rbpA1 gene (8). We therefore attempted to detect DNA binding proteins. The 5[prime]-UTR sequence from +1 to +151 was used as probe in a gel mobility shift analysis. First, we analyzed DNA binding proteins in various ammonium sulfate fractions from 30 up to 80% saturation. The results in Figure 6A indicate that the fraction from 30 to 50% saturation contained major DNA binding activities. Two major DNA-protein complexes were detected, named Complex 1 and Complex 2 respectively. Other complexes were also eventually detected, but they were not reproducible. Interestingly, the protein fraction from cells grown at 38°C contained significant DNA binding activity, whereas very low activity was detected in the corresponding fraction from cells grown at 22°C. Densitometric quantitation indicated that the protein extract from 22°C cells contained [sim]10% of the DNA binding activity (both Complex 1 and Complex 2) present in the extract from 38°C cells. This result may suggest that several DNA binding proteins bind to the 5[prime]-UTR of the rbpA1 gene at 38°C to repress transcription. The binding specificity was tested by competition experiments (Fig. 6B). DNA 1 (Fig. 5B), which corresponds to the 3[prime]-end of the 5[prime]-UTR, strongly competed with formation of Complexes 1 and 2. Under comparable conditions yeast tRNA did not compete at all. Note that these experiments were done in the presence of a high concentration (50 µg/ml) of dI-dC as non-specific competitor. These results suggest that there are DNA binding proteins that bind specifically to the 3[prime]-terminal part of the 5[prime]-UTR of the rbpA1 gene in cells grown at 38°C.

We then tried to separate the two activities that formed Complexes 1 and 2. The 30-50% fraction obtained by ammonium sulfate fractionation was subjected to anion exchange column chromatography (Fig. 7). An activity that formed Complex 1 was eluted at [sim]22 min, whereas another activity that formed Complex 2 was eluted at 11-19 min. This result indicates that Complexes 1 and 2 are formed by distinct proteins.

The proteins that bind to the 5[prime]-UTR were finally purified by an affinity technique using magnetic particles (Dynabeads) conjugated with the 5[prime]-UTR (Fig. 8). Column fractions 3-5 were individually subjected to affinity purification. After extensive washing DNA-Dynabead conjugates were treated with 0.3 M buffer to elute DNA binding proteins. In column fractions 3 and 4, a 72 kDa polypeptide was detected in the 0.3 M eluate. The DNA-Dynabeads were further treated with 1 M buffer to elute tightly bound proteins. In column fractions 4 and 5 a, 75 and a 32 kDa polypeptide were detected in the 1 M eluate. A part of the 75 kDa polypeptide was also eluted with 0.3 M buffer. These results suggest that the 75 and 32 kDa polypeptides are involved in formation of Complex 1, whereas the 72 kDa polypeptide is involved in formation of Complex 2. Unfortunately we were unable to obtain a larger amount of these polypeptides sufficient for protein sequencing, even when the starting material was increased to 12 g fresh weight (10 l cultures).

Figure 8. Affinity purification of DNA binding proteins. Fractions 3-5 from the anion exchange column chromatography were individually subjected to affinity purification with DNA-Dynabeads. Bound proteins were eluted from the affinity matrix with 0.3 and then with 1 M KCl. These eluates were analyzed by SDS-PAGE. The gel was stained with silver nitrate. A 72 kDa polypeptide was detected in the 0.3 M eluate from column fraction 3, whereas polypeptides of 75 and 32 kDa were detected in the 1 M eluate from column fraction 5. All of these polypeptides were found in the eluates from column fraction 4. Because of the high sensitivity of the staining, bands of unidentified contaminants that often appear in the silver staining were found in the gel as marked.

DISCUSSION

Increased stability of the rbpA1 transcript at low temperature

Available data suggest that cold-dependent expression of the rbpA1 gene of A.variabilis is regulated at both the transcriptional and post-transcriptional levels. We first discuss post-transcriptional regulation. The results in Figure 3 show that stability of the rbpA1 transcript was markedly increased at low temperature. Sakamoto and Bryant (9) reported that the stability of the transcript of a cold-regulated desaturase gene (desA) is markedly increased in vivo at low temperatures in a cyanobacterium, Synechococcus sp. PCC 7002. They also showed that the stability of the transcript of a constitutively expressed desaturase gene (desC) was unaffected by growth temperature. Although one might tend to assume that degradation of RNA in vivo is generally slowed down at low growth temperatures, the results with desC indicate that this is not so. We found that extract from cells grown at low temperatures contains greater RNA degrading activity in vitro (unpublished results). One might be able to argue, therefore, that the stability of transcripts is normally kept fairly independent of growth temperature by increasing the activity of general RNA degradation at low temperatures. In this light, there must be a mechanism for the marked increase in stability of the rbpA1 transcript at low temperature. The cspA gene of E.coli is also known to be regulated by increased stability of the transcript at low temperature (10,11). These authors also showed that the coding sequence was necessary for increased stability of the transcript.

Transcriptional regulation

Despite clear difference in the stability of rbpA1 transcript at different growth temperatures, rapid degradation of the transcript upon transfer to high temperature (Fig. 3) suggests that transcription of the rbpA1 gene stops after the temperature shift, as reported previously (1). The results of expression of lacZ fusion genes suggest that there is a transcriptional control. The lacZ fusion transcript was clearly regulated by low temperature if the fusion point was located downstream of nt 149 from the transcription start site (Fig. 5A), but was constitutive if the fusion point was located at nt 134 or 140. These results clearly indicate that activity of the promoter of the rbpA1 gene is not dependent on temperature without the 5[prime]-UTR down to nt 149. The 3[prime]-end of the regulatory region was, therefore, estimated to be located between nt 140 and 149, whereas the 5[prime]-end of the regulatory sequence should be determined by different experiments.

DNA binding proteins

We showed the importance of the 5[prime]-UTR in temperature-dependent regulation by identifying temperature-regulated putative trans-acting factors. The presence of DNA binding proteins that have affinity for the 5[prime]-UTR of the rbpA1 gene in the extract of 38°C cells supports the idea of transcriptional regulation. We detected two types of DNA-protein complexes that were formed with the 5[prime]-UTR. These complexes showed different electrophoretic mobilities on non-denaturing gel electrophoresis (Fig. 6). The proteins involved in formation of these two complexes were separated by anion exchange column chromatography (Fig. 7). Based on the electrophoretic mobility, Complex 1 seemed to be fairly large. Affinity purification of DNA binding proteins was efficient and successful. The polypeptides of 75 and 32 kDa were co-purified and these polypeptides exhibited a strong affinity for the DNA probe used (151 bp 5[prime]-UTR), since they were eluted from the DNA-Dynabeads conjugates with 1 M KCl. These polypeptides were significantly less abundant than the 72 kDa polypeptide, which was found to form Complex 2. In contrast, the band of Complex 1 detected in the gel mobility shift analysis was more intense than the band of Complex 2. These findings suggest that the 75 and/or 32 kDa polypeptides bind to the 5[prime]-UTR more strongly than does the 72 kDa polypeptide. Whether these proteins really regulate expression of the rbpA1 gene will be answered if the genes encoding these proteins are identified and disrupted in A.variabilis cells.

The experiments with the lacZ fusion genes defined the 3[prime]-end of the regulatory sequence, which was located between nt 140 and 149. This region was found to form an inverted repeat (Fig. 5B). A synthetic DNA probe, DNA 1, efficiently competed with binding of the 151 bp 5[prime]-UTR with the DNA binding proteins (Fig. 6B). These findings suggest that the DNA sequence in the inverted repeat region is a target site for the DNA binding proteins and might serve as a cis-acting element in cold-regulation. In this respect we are obliged to use the expression `repression at high temperature' or `cold-derepression' rather than `cold-regulation' or `cold-induction'. Interestingly, the inverted repeat is conserved in other cold-regulated rbp genes of A.variabilis (Fig. 5B). The only constitutive rbp gene, rbpD, does not contain a similar sequence nor an inverted repeat (not shown).

The only other examples of cold-regulated genes in cyanobacteria are the genes for fatty acid desaturases. We tried to find a similar inverted repeat in the 5[prime]-region of these genes, but we found that the desaturase genes are quite divergent from the rbp genes in the 5[prime]-region. In fact, regulation of rbp genes is different in various aspects from that of des genes. We showed that cold-derepression of the rbpA1 gene was not severely inhibited by the protein synthesis inhibitor chloramphenicol (Fig. 2B) at a concentration (15 µg/ml) that inhibits protein synthesis by 94% (23). This is in contrast to cold-induction of the desA and desB genes of Synechococcus sp. PCC7002, which was totally abolished by addition of chloramphenicol (9).

Based on the results of the present study we propose a simple model of transcriptional regulation of the rbpA1 gene in A.variabilis by growth temperature. This gene has a constitutive promoter, but is repressed by one or more protein factors which bind to the 5[prime]-UTR of the rbpA1 gene at high temperature. After a shift to a low temperature the protein factors rapidly lose binding activity; this loss of activity does not require de novo protein synthesis. We do not yet know how the DNA binding activity is regulated by temperature, but one plausible hypothesis is regulation by protein modification. We need to identify the DNA binding proteins found by affinity purification to analyze the mechanism of regulation of DNA binding activity by temperature in more detail.

ACKNOWLEDGEMENTS

This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture, Japan (nos 09874167, 08454254 and 09251206). This work was started when the authors were at Tokyo Gakugei University. We thank Drs C.P.Wolk and J.Elhai for providing us with pRL576 and pRL488 as well as various tools for conjugal transfer of DNA into cyanobacterial cells.

REFERENCES

1. Sato,N. (1994) Plant Mol. Biol., 24, 819-823. MEDLINE Abstract

2. Sato,N. (1995) Nucleic Acids Res., 23, 2161-2167. MEDLINE Abstract

3. Sato,N. and Maruyama,K. (1997) Plant Cell Physiol., 38, 81-86. MEDLINE Abstract

4. Sato,N., Maruyama,K., Nishiyama,Y. and Murata,N. (1997) J. Plant Res., 110, 405-410.

5. Sugita,M. and Sugiura,M. (1994) Nucleic Acids Res., 22, 25-31. MEDLINE Abstract

6. Mulligan,M.E., Jackman,D.M. and Murphy,S.T. (1994) J. Mol. Biol., 235, 1162-1170. MEDLINE Abstract

7. Birney,E., Kumar,S. and Krainer,A.R. (1993) Nucleic Acids Res., 21, 5803-5816. MEDLINE Abstract

8. Sato,N. and Wada,A. (1996) Plant Cell Physiol., 37, 1150-1160. MEDLINE Abstract

9. Sakamoto,T. and Bryant,D.A. (1997) Mol. Microbiol., 23, 1281-1292. MEDLINE Abstract

10. Brandi,A., Pietroni,P., Gualerzi,C.O. and Pon,C.L. (1996) Mol. Microbiol., 19, 231-240. MEDLINE Abstract

11. Goldenberg,D., Azar,I. and Oppenheim,A.B. (1996) Mol. Microbiol., 19, 241-248. MEDLINE Abstract

12. Jiang,W., Fang,L. and Inouye,M. (1996) J. Bacteriol., 178, 4919-4925. MEDLINE Abstract

13. Kratz,W.A. and Myers,J. (1955) Am. J. Bot., 42, 282-287.

14. Sato,N., Tachikawa,T., Wada,A. and Tanaka,A. (1997) J. Bacteriol., 179, 7063-7071. MEDLINE Abstract

15. Elhai,J. and Wolk,C.P. (1990) EMBO J., 9, 3379-3388. MEDLINE Abstract

16. Elhai,J. and Wolk,C.P. (1988) Methods Enzymol., 167, 747-754. MEDLINE Abstract

17. Reaston,J., van den Hondel,C.A.M.J.J., van Arkel,G.A. and Stewart,W.D.P. (1982) Plasmid, 7, 101-104. MEDLINE Abstract

18. Miller,J.H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

19. Stone,S.R. Hughes,M.J. and Jost,J.-P. (1991) In Jost,J.-P. and Saluz,H.P. (eds), A Laboratory Guide to In Vitro Studies of Protein-DNA Interactions. Birkhauser Verlag, Basel, Switzerland, pp. 163-194.

20. Lowry,O.H., Rosebrough,N.J., Farr,A.L. and Randall,R.J. (1951) J. Biol. Chem., 193, 265-275.

21. Laemmli,U.K. (1970) Nature, 227, 680-685. MEDLINE Abstract

22. Sato,N. and Murata,N. (1980) Biochim. Biophys. Acta, 619, 353-366. MEDLINE Abstract

23. Sato,N. and Murata,N. (1981) Plant Cell Physiol., 22, 1043-1050.

---

*To whom correspondence should be addressed. Tel: +81 48 858 3623; Fax: +81 48 858 3384; Email: naokisat@molbiol.saitama-u.ac.jp

---

This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 17 Apr 1998
Copyright©Oxford University Press, 1998.
CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				S. Ehira, M. Ohmori, and N. Sato Identification of Low-temperature-regulated ORFs in the Cyanobacterium Anabaena sp. Strain PCC 7120: Distinguishing the Effects of Low Temperature from the Effects of Photosystem II Excitation Pressure Plant Cell Physiol., August 1, 2005; 46(8): 1237 - 1245. [Abstract] [Full Text] [PDF]

				S. Jager, E. Evguenieva-Hackenberg, and G. Klug Temperature-dependent processing of the cspA mRNA in Rhodobacter capsulatus Microbiology, March 1, 2004; 150(3): 687 - 695. [Abstract] [Full Text] [PDF]

				D. Chamot and G. W. Owttrim Regulation of Cold Shock-Induced RNA Helicase Gene Expression in the Cyanobacterium Anabaena sp. Strain PCC 7120 J. Bacteriol., March 1, 2000; 182(5): 1251 - 1256. [Abstract] [Full Text]

This Article Abstract Print PDF (439K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (10) Request Permissions Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Sato, N. Articles by Nakamura, A. Search for Related Content PubMed PubMed Citation Articles by Sato, N. Articles by Nakamura, A. Social Bookmarking          What's this?