Regulation of the puf Operon

The LH-Ⅰ complexes are found in the photosynthetic membrane in a fixed molar excess（approximately twelvefold）over the RC complex（Chory and Kaplan 1983）.However，each of the two polypeptides within an LH-Ⅰ complex（β and α）or within the RC（L and M）are present in a 1∶1 stoichiometry.As these four genes plus C2397 are closely linked in a single operon both in R.sphaeroides and R.capsulatus，it would be of particular interest to elucidate how this stoichiometry and differential expression are achieved.Belasco et al.（1985）have reported that a 2.7 kb mRNA containing information for the LH-Ⅰ（β，α）and RC（-L，-M）proteins and a small 0.5 kb mRNA containing information for LH-Ⅰ（β，α）in R.capsulatus are derived from the same transcriptional initiation site.They came to the conclusion that the ratio of the small to large transcripts encoding the LH-I and RC proteins in R.capsulatus results predominantly from the selective degradation of a specific segment of the initial large transcript at the 3'end of the molecule，resulting in a stabilized small transcript.The final ratio of small to large transcripts（（10—20）：1）would then reflect the relative 3'processing activity.

In R.sphaeroides Zhu and Kaplan（1985）have also observed the presence of two transcripts in a similar（10—20）：1 ratio.Additionally，it has been demonstrated that the half life of the large transcript is 9 minutes and that of the small transcript is 20 minutes（Zhu et al.1986b）.The difference in the stability of the two mRNAs indicates that some type of posttranscriptional control is involved in the differential expression of pufB and A and pufL and M，but it cannot be the major event responsible for the（10—20）：1 ratio of small to large transcripts.Further，we have demonstrated that the two transcripts have different 5'terminals as measured by both S-Ⅰ nuclease protection analysis and the primer extension technique（Zhu et al.1986b）.The small transcript has a 5'end 104 bp upstream from the first ATG codon of pufB，whereas the large transcript has a 5'end 75 bp upstream from the same start codon.A model is presented as follows：When the RNA polymerase binds to the sequence 104 bp upstream from the ATG start codon of pufB，the transcription terminates between pufA and pufL；whereas when the RNA polymerase binds to the sequence 75 bp upstream from the start codon of pufB，it reads through the first termination site between pufA and pufL，and then stops at the second termination site at the end of the puf operon.The mRNA secondary structure of the intercistronic region between the pufB and A and pufL and M is compatible with a rho factor-dependent termination signal（Zhu et al.1986b）.Based on the model described it is suggested that a combination of differential transcriptional initiation coupled with transcription termination led to the（10—20）：1 ratio of small to large transcripts.These results seem to conflict with those of Belasco et al.（1985）who found that both the large and the small messengers initiate from the same initiation site about 35 bp downstream of the EcoRI site.This conclusion is based mainly on the S-Ⅰ mapping in which only one strong band was observed under low-oxygen conditions，although another band corresponding to an initiation site 70 bp downstream of the EcoRI site was noticed.Since this minor band accounted for only 1%—2%of the puf mRNA 5'ends，it was neglected（Belasco et al.1985）.By carefully maintaining the growth under strictly anaerobic photosynthetic conditions instead of low-oxygen conditions，which may cause the degradation of the labile large mRNA，we have been able to obtain similar results in R.capsulatus and in R.spharoides using the S-l mapping technique.Two bands were detected with a correct ratio of（10—20）：1 and an approximately 35 bp space between their 5'ends（Zhu and Hearst，unpublished observations）.We suggest that the transcripts of the puf operon in R.capsulatus also have two 5'ends.

Since the 5'ends determined by S-1 nuclease analysis may not be the intiation sites，the authors did not exclude the possibility that the 5'ends observed might themselves be the products of processing of a primary transcript in R.sphaeroides（Zhu et al.1986b）.It was also suggested that if this is the route by which these two transcripts are derived，the mechanism of posttranscriptional processing of the primary puf transcript of R.sphaeroides is far more complex than that envisioned for R.capsulatus，because the 5'end of the small transcript（0.5 kb）mapped upstream of that of the large transcript（2.6 kb）cannot be simply processed from the large transcript（Zhu et al.1986b）.Recently，Beatty et al.（1986）reported that the stable puf transcripts might be derived from short-lived precursors.Marrs'group（Bauer et al.1987）has described an ORF upstream of the puf operon，which is designated pufQ.Recently they have（Bauer et al.1988）hypothesized that the pufQ gene product is the carrier protein for Bchl synthesis.(https://www.daowen.com)

The consensus sequence of the promoter of the puf operon and the mechanism controlling the termination and antitermination are unknown as yet.It has in general been accepted that the regulatory sequences reside in the 5'region upstream of the structural genes.Kiley et al.（1987a）have reported that a promoter region for transcription intitiation signal is within the region between 100 bp and 200 bp proximal to the first codon of pufB in R.sphaeroides.Beatty et al.（1985）have defined a DNA sequence that seems to be necessary for oxygen-regulated transcriptional initiation of the puf operon in R.capsulatus.Marrs'group has also reported that the puf operon may have two promoters：an oxygenregulated promoter and a constitutive promoter upstream of pufQ（Bauer et al.1988）using lac fusion.Recently，we have found that this Q region includes two ORFs，both controlled by oxygen and light（Zhu et al.，unpublished observations）.These two ORFs share strong homologies with the corresponding sequences in R.sphaeroides（Kiley and Kaplan，personal communication）.The fact that the“regulatory gene or genes”are regulated by oxygen，raises a further question：what is the fundamental mechanism for oxygen regulation？

Since most of the photosynthetic genes are coordinately expressed in response to oxygen，there might be a common mechanism for their coordinate expression，although some of these genes also could be regulated individually.It has been suggested that DNA supercoiling may control the prokaryotic gene expression（Yamamoto and Droffner 1985；Kranz and Haselkorn 1986）.Recently we have shown that the expression of the photosynthetic genes in R.capsulatus in response to oxygen requires the action of the topoisomerase，gyrase（Zhu and Hearst 1988）.The supercoiling of the DNA in bacteria is determined by a dynamic balance between gyrase，which supercoils the DNA，and topoisomerase I，which relaxes the DNA.We have found that the synthesis of Bchl，LH-Ⅰand LH-Ⅱ/Bchl complexes，and the expression of most photosynthetic genes coding for LH，RC，and Bchl biosynthesis as well as for form Ⅱ RuBPCase is immediately inhibited by novobiocin and coumermycin at various concentrations（5—100 μg/ml）（Zhu and Hearst 1988）.However，the constitutively expressed puf mRNA under aerobic growth conditions，as we mentioned early，is not inhibited by novobiocin.These results imply that the DNA gyrase is necessary for the anaerobic expression of those genes.The oxygen-activated crtA gene and other crt genes whose transcription is relatively enhanced by oxygen were inhibited by the antibiotics to a far smaller extent.The expression of Q gene also requires gyrase.At the present time we believe these gyrase effects to relate to a need for DNA unwinding during heavy transcription as opposed to the regulation of the initiation of transcription.Although transcriptional regulation plays a major role，we believe that the posttranscriptional and posttranslational regulations，including the assembly of the gene products in the membrane（Zhu et al.1986a and 1986b；Zhu and Hearst 1987），are also required to elaborate the efficient and complex regulation of the photosynthetic apparatus in purple bacteria in response to environmental signals.A proposed mechanism of the regulation of the photosynthetic genes of purple nonsulfur bacteria at the transcriptional，posttranscriptional，and posttranslational levels is summarized in Figure 6.