Plant Cell Physiol. 37(2): 153-159 (1996)JSPP © 1996

Expression of the puf Operon in an Aerobic Photosynthetic Bacterium,Roseobacter denitrificans

Kohji Nishimura, Hiroshi Shimada, Hiroyuki Ohta, Tatsuru Masuda, Yuzo Shioi andKen-ichiro TakamiyaDepartment of Biological Sciences, Faculty of Bioscience and Biotechnology, Tokyo Institute of Technology, Nagatsuta, Midori-ku,Yokohama, 226 Japan

The effects of oxygen and light on the expression ofthe puf operon were investigated in Roseobacter denitri-ficans in a comparison with those in Rhodobacter sphae-roides. In darkness, the levels of the total puf mRNA inRos. denitrificans were about 1.3 times those in Rb. sphae-roides at low concentrations of oxygen, reflecting the ac-cumulation of bacteriochlorophyll and carotenoids. The ox-ygen tension, up to 94% saturation of dissolved oxygen,did not affect the levels of the total puf transcripts in Ros. de-nitrificans, whereas those in Rb. sphaeroides were reducedto 55% of the maximum level even at 50% saturation.

Four pu/-specific transcripts were detected: a 0.5-kbtranscript was the most abundant; 1.2-kb and 1.9-kb tran-scripts accumulated at low levels; and a 3.5-kb transcript ac-cumulated at very low levels under all conditions tested.The levels of the individual transcripts were barely affectedby molecular oxygen. An S-l nuclease protection assayrevealed that the 0.5-kb transcript encoded the LHI-a andLHI-/7subunits (pufBA), the 1.2-kb transcript encoded puf-BA and part of pufL, and the 1.9-kb transcript encodedpufBAL and part of pufM. It was not clear whether the3.5-kb transcript encoded the entire pufBALM and thegene for the polypeptide moiety of cytochrome c. Thedifference in levels between the 0.5-kb transcript and theother transcripts (1.2 kb, 1.9 kb, and 3.5 kb) was presumedto be due to the presence of several stem-loop structuresat the 3' terminus of the 0.5-kb transcript which acted asterminators of transcription and, possibly, as protectionagainst nucleolytic digestion.

Light inhibited the expression of the puf operon inRos. denitrificans more effectively than that in Rb. sphae-roides. The insensitivity to oxygen, as well as the sensitivityto light, of the expression of the puf operon in Ros. denitri-ficans, which was different from that in Rb. sphaeroides,seemed to represent a mode of adaptation that allowed theformer cells to avoid photodynamic damage by light underhighly aerobic conditions.

Key words: Aerobic photosynthetic bacterium — Gene ex-pression — Light inhibition — Oxygen insensitivity — pufoperon — Roseobacter denitrificans.

Purple non-sulfur bacteria, such as Rhodobactersphaeroides and Rhodobacter capsulatus, grow as a conse-quence of respiration in darkness or light under aerobicconditions, while they grow photosynthetically under anaer-obic conditions (Cohen-Bazire et al. 1957, Drews 1978,Kiley and Kaplan 1988). In these bacteria, the levels ofbacteriochlorophyll and carotenoids in the photosyntheticmembranes are mainly regulated by external factors, suchas oxygen and light. In Rb. sphaeroiodes the pigments ac-cumulate under either anaerobic light conditions or undersemi-aerobic dark conditions but not under aerobic dark orsemi-aerobic light conditions.

The photosynthetic apparatus of Rb. sphaeroiodesand that of Rb. capsulatus are composed mainly of light-harvesting complexes (LHI, LHII) and a reaction center(RC), which contain pigment-binding polypeptides, bacte-riochlorophyll and carotenoids. The puf operon encodesthe polypeptides of LHI (pufBA), the L and M polypep-tides of RC (pufLM) and other polypeptides (pufQ, pufK,pufX; Bauer and Marrs 1988, Farchaus et al. 1992, Gonget al. 1994, Kiley et al. 1987, Williams et al. 1983, 1984).The puc operon encodes the polypeptides of LHII (puc-BACDE; Tichy et al. 1989, Youvan and Ismail 1985). Thepuh gene encodes the H polypeptide of RC (Donohue et al.1986).

An aerobic photosynthetic bacterium Roseobacter de-nitrificans, which was isolated by Shiba et al. (1979), growsas a consequence of respiration in darkness and in the lightunder aerobic conditions, but it cannot grow photosynthet-ically under anaerobic conditions. In Ros. denitrificans,bacteriochlorophyll and carotenoids accumulate in the pho-tosynthetic membranes under aerobic dark conditions butnot in the light during either aerobic or anaerobic respira-tion (Harashima et al. 1980, Iba and Takamiya 1989, Shiba1987). Thus, the effects of oxygen and light on the accu-mulation of the pigments in aerobic photosynthetic bacte-ria differ from those in purple non-sulfur bacteria.

Liebetantz et al. (1991) determined the sequence of thepuf operon of Ros. denitrificans, which consisted of puf-BALM'but lacked pufQ, pufK and pufX, and they demon-strated that the accumulation of pigments was inhibited bydim light. To date, however, no detailed study of the tran-scriptional organization of the genes has been reported.

In the present study, we investigated the effects of ox-

153

Dow

nloaded from https://academ

ic.oup.com/pcp/article/37/2/153/1820311 by guest on 10 August 2022

154 K. Nishirnura et al.

ygen and light on the expression of the puf operon inRos. denitrificans in a comparison with that in Rb. sphae-roiodes, and we demonstrated that the expression of the pufoperon in Ros. denitrificans was much more oxygen-insen-sitive but more light-sensitive than that in Rb.sphae-roiodes.

Materials and Methods

Bacterial strains, media, and growth conditions—The aerobicphotosynthetic bacterium Ros. denitrificans was cultured in themodified PPES-II medium of Taga (Shioi 1986, Taga 1968), andthe anaerobic photosynthetic bacterium Rb.sphaeroides 2.4.1.was cultured in the organic medium of Cohen-Bazire et al. (1957).The cells were grown at 27°C in a water bath on a reciprocalshaker (Personal-11; Taitec Co., Koshigaya, Japan) in darknessor in the light. For studies of the effects of oxygen tension, 5 ml ofa suspension of cells was inoculated into a 500-ml Sakaguchi flaskthat contained 200 to 300 ml of medium. The cells were grown onthe reciprocal shaker at 110 to 130 strokes per min in darkness andthe ratio of dissolved oxygen of each culture to that of standard so-lution saturated with oxygen was monitored with an oxygen elec-trode (OM-14; Horiba Co., Kyoto, Japan) immersed in the cul-ture medium. For studies of the effects of light, we used 100-mlL-shaped tubes that contained 75 ml of medium. After inocula-tion of 5 ml of a suspension of cells, the culture was incubated onthe reciprocal shaker which was operated at 130 strokes per min.Light irradiation from 100-W tungsten bulbs was supplied perpen-dicularly to the shaking L-shaped tubes. In most cases, a pair oflight-irradiated and foil-wrapped tubes was shaken side by side.The intensity of the white light was 2.5 x 106 ergs cm"2 s~' of thesurface of the tubes. We used bacterial cells in the middle of thelogarithmic phase of growth throughout the experiments. Bacte-rial growth was monitored by measuring the absorbance of cul-tures at 660 nm.

For the preparation of hybridization probes, Escherichia coliJM109 harboring plasmid pUC118 (Yanisch-Perron et al. 1985)that included the region of the puf operon was grown at 37°C inLuria-Bertani medium (Maniatis et al. 1982).

Extraction of RNA—Thirty ml of the bacterial culture wererapidly cooled on ice, and cells were immediately harvested by cen-trifugation. The cells were suspended in 500 /A of solution A (200mM acetic acid, 10 mM EDTA, and 0.5% SDS) and added to anequal volume of phenol that had been saturated with 200 mMacetic acid and 10 mM EDTA. The suspension was incubated at60°C for 5 min and then centrifuged at 12,000 x g for 5 min to sep-arate the aqueous phase from the organic phase. The RNA in theaqueous phase was precipitated with ethanol at room tempera-ture. The pelleted precipitate was suspended in 400 fj\ of solutionA for precipitation with ethanol at room temperature. This precip-itation in ethanol was repeated three times. The precipitate wasdissolved in 500//I of buffer B [50 mM Tris-HCl (pH 7.8) and 5mM MgCl2]. The DNA was digested by incubation with RNase-free DNase I (22.5 units; Pharmacia LKB Biotechnology Inc.,Uppsala, Sweden) at 37°C for 3 h. DNase I was removed byphenol extraction, and the RNA was finally collected by ethanolprecipitation. The ratio of absorbance at 260 nm to that at 280 nmvaried from 1.8 to 2.0. Total RNA was used for Northern hybridi-zation (lO^glane"1) and S-l nuclease mapping or primer exten-sion analyses (40/ig lane""1).

Preparation and radioactive labeling of hybridization probes—The protocols for preparation of plasmid DNA and the purifica-

tion of the hybridization probes were those of Maniatis et al.(1982). Figure 1 shows the puf operons of Ros. denitrificans andRb. sphaeroides and the DNA templates used for the present analy-ses. Probes for Ros. denitrificans (probes a, b, and d in Fig. 1)were derived from the 2.7-kb EcoKI-EcoRI fragment that encod-ed pufBALM (Liebetantz et al. 1991) cloned in pUC118 (p8EpE)by digestion with appropriate restriction enzymes. The probe forRb. sphaeroides (probe e in Fig. 1) was the 1.5-kb Pstl-Kpnl frag-ment that encoded the polypeptides from the carboxyl terminus ofbchA to the amino terminus of pufL (Shimada et al. 1993) clonedin pUC118 (p8PpK). The probes for Northern hybridization werelabeled with [a-32P]dCTP (ICN Biomedicals Inc., Costa Mesa,CA, U.S.A.) with a tfcaBEST™ Labeling Kit (Takara ShuzouCo., Kyoto, Japan; a specific radioactivity, 2.0x 109dpmper/*gof probes d or e). The 5' end probe for the S-l nuclease protectionassay (probe a in Fig. 1) was labeled with [y-32P]ATP (ICN) usingMEGALABEL™ (Takara; a specific radioactivity, 2.0 x 108 dpmper//g of probe a). The 3' end probe for the S-l nuclease protec-tion assay (probe b in Fig. 1) was labeled with [a-32P]dCTP,in view of the base to be filled in by the Klenow fragment ofDNA polymerase I from E. coli (Takara; a specific radioactivity,2.0 x 108dpmperji/g of probe b; Maxam and Gilbert 1980). Thepuf oligonucleotide probe (5 -TCTGTAAGACCTGTGAAGGA-3') for primer extension analysis (probe c in Fig. 1) and the 16SrRNA probe (5-AACCACATGCTCCACCGCTTGTGCGGGCC-CCCGTCAATTCCTTTGAGTTT-3'; Dryden and Kaplan 1990)for Northern hybridization were labeled with [y-32P]ATP usingMEGALABEL™ (a specific radioactivity, 3.2 x 10'dpm per jigof probe c in Fig. 1 or the 16S rRNA probe).

Northern hybridization—The protocol for Northern hybridi-zation was that described by Maniatis et al. (1982). For the de-tection of pw/mRNAs (using probes d and e in Fig. 1), the nitrocel-lulose membrane (BA-S 85; Schleicher and Schuell, Inc., Keene,NH, U.S.A.) was washed twice with 2xSSC, 0.1% SDS at 25°Cfor 15 min and twice with 0.2xSSC, 0.1% SDS at 25°C for 15min. For the detection of rRNA, the filter was washed twice with0.1 x SSC, 0.1% SDS at 65CC for 30 min. The washed filter wasexposed to X-ray film. For quantitative assessment of mRNAs ofthe Northern hybridization, the filter was scanned with an ImageAnalyzer (Fujix BAS2000; Fuji Photo Film, Tokyo, Japan). Theamount of puf transcripts was normalized by reference to theamount of 16S rRNA.

S-l nuclease protection assay—S-l nuclease protection analy-sis was performed by the method of Zhu et al. (1986). For precisemapping of a 5' end, DNA sequencing ladders were generated bythe sequencing method of Maxam and Gilbert for an appropriateend-labeled restriction fragment (Maxam and Gilbert 1980). Thesamples were subjected to electrophoresis in a denaturing poly-acrylamide gel (6% polyacrylamide and 8 M urea). The dried gelwas exposed to X-ray film.

Primer extension analysis—Primer extension analysis was per-formed by the method of Lee et al. (1989b). For sequencing ladd-ers, 10 pmol of the same oligonucleotide probe were allowed to hy-bridize to 12 fig of single-stranded DNA template p8EpE. Thesequencing reaction was performed with a BcaBEST sequencingkit. The dried gel was exposed to X-ray film.

Other methods—Proteins were quantitated by the method ofLowry et al. (1951). Extraction and quantitation of bacteriochloro-phyll and carotenoids were performed by the methods of Iba andTakamiya (1989). S-l nuclease and avian myeloblastosis virusreverse transcriptase were purchased from Takara Shuzou Co.

Dow

nloaded from https://academ

ic.oup.com/pcp/article/37/2/153/1820311 by guest on 10 August 2022

Expression of puf operon in Ros. denitrificans 155

Results and Discussion

Transcripts of the puf operon—We investigated thetranscription of the puf operon by Northern hybridization.Different [32P]-labeled probes (d and e in Fig. 1) were usedfor Ros. denitrificans and Rb. sphaeroides, respectively.Four puf-specihc transcripts were detected when Ros. de-nitrificans had been grown under aerobic conditions (ox-ygen tension, 80% saturation; Fig. 2, lane 1) in darknessand also when Rb. sphaeroides was grown under semi-aerobic conditions (2% saturation; Fig. 2, lane 2) in dark-ness, as previously reported (Zhu and Kaplan 1985). Themost abundant transcript was the 0.5-kb transcript, andthe 1.2-kb and 1.9-kb transcripts accumulated at lowlevels. In addition, the 3.5-kb transcript accumulated atvery low levels. The signals from the four transcripts couldnot be eliminated by extensive washing of blots (0.1 x SSC,0.1% SDS at 65°C for 30 min, twice; data not shown) andthese transcripts were also detected under a broad rangeof oxygen tensions in Ros. denitrificans (see below). Therelative levels of these puf transcripts were 100% (0.5 kb),9 ± 2 % (1.2kb), 15±5% (1.9kb) and 3±O.5% (3.5 kb).These values are the results from more than three independ-ent experiments. The largest transcript (3.5 kb) probably en-coded not only the entire length of pufBALM but also thegene for cytochrome c of RC, as reported in Rhodopseudo-monas viridis (Weissner et al. 1990). Because the complete

sequence of the gene for cytochrome c has not yet been de-termined, it is unclear whether the 3.5-kb transcript en-codes the entire polypeptide of cytochrome c. When weused a probe that covered pufBALM, the minor transcripts(1.2 kb, 1.9 kb, and 3.5 kb) accumulated similarly at verylow levels (data not shown).

Insensitivity to oxygen of both the expression of thepuf operon and the accumulation of pigment—-We exam-ined in detail the effects of oxygen tension on the expres-sion of the puf operon. Figure 3 shows the levels of allp«/-specific transcripts, as determined by Northern hybrid-ization, in both Ros. denitrificans and Rb. sphaeroidesgrown under various oxygen tensions in darkness. The levelof individual transcripts in Ros. denitrificans was almostconstant at various oxygen tensions (Fig. 3A-I), suggestingthat both the transcriptional activity and the stability ofmRNAs were oxygen-insensitive. By contrast, the levels ofpuf mRNA in Rb. sphaeroides decreased significantly inproportion to increases in oxygen tension (Fig. 3A-II).Total levels of these transcripts are shown in Figure 3B.The transcription of the entire puf operon in Ros. denitri-ficans was barely affected by oxygen tension up to 94% satu-ration, whereas that of Rb. sphaeroides was reduced to thehalf of the maximal level at 40% saturation. Although the0.5-kb transcript from Ros. denitrificans seems to be adoublet in the right lane in Figure 3A-I (near 100% satura-tion), the doublet is an experimental artifact.

bchA

Reaction center

pufL pu/M cyl

EcoR\ fle/II Hindi6 Sail

• 0.5 kb .

B

LHI

bchA pufQ pufKpufBpufA

Reaction center

pu/L pufM pu/X

Pst\ Kpn\

Fig. 1 A partial restriction map of thepu/intercistronic regions in Ros. denitrificans (A) and Rb. sphaeroides (B). The Figure alsoshows the DNA templates used in the Northern hybridization for the analysis of pu/operons, the S-l nuclease protection assay and theprimer extension analysis for the determination of the end points of the 0.5-kbpufBA-specific transcript. DNA template a (EcoRl-BgM)and oligonucleotide probe c were used for the 5' end analysis. DNA template b (BglH-Hincll) was used for the 3' end analysis. DNAtemplates d (EcoRI-Sa/1) and e (Pstl-Kpnl) were used for Northern hybridization of RNAs from Ros. denitrificans and Rb. sphaeroides,respectively. DNA templates a, b, and d were generated from p8EpE, and DNA template e was generated from p8PpK.

Dow

nloaded from https://academ

ic.oup.com/pcp/article/37/2/153/1820311 by guest on 10 August 2022

156 K. Nishimura et al.

0.5 0.5

Fig. 2 Northern hybridization of puf transcripts with totalRNAs from Ros. denitrificans (lane 1) and Rb. sphaeroides (lane2) that had been grown under aerobic conditions in darkness (ox-ygen tension, 80%) and semi-aerobic conditions in darkness (ox-ygen tension, 2%), respectively. Probe d in Fig. 1 was used inlane 1 and probe e in Fig. 1 was used in lane 2. Ten fig of totalRNA per lane were fractionated on an agarose gel.

I IIOxygen tension (%)

100

0.5

kb

-2.6-1.7-1.1

B

i

100

50

s\

Rb. sphaeroides

Ros.

Ok

denitrificans

50 100

Oxygen tension (%)

Fig. 3 The levels of puf mRNAs in cells grown in darkness atvarious oxygen tensions. Panel A, Northern hybridization ofRNA from Ros. denitrificans (I) and Rb. sphaeroides (II). PanelB, Dependence on oxygen tension of the total amount of pufmRNA. The total/>M/mRNA is expressed as the percentage of themaximum level. Each point and bar indicate the mean and thestandard deviation, respectively, of the results from more thanthree independent experiments. The data were normalized byreference to the level of rRNA. Experimental conditions were thesame as described in the legend to Fig. 2.

The accumulation of bacteriochlorophyll and carote-noids showed a similar dependence on oxygen to that ofpuf transcripts in the respective bacteria (Fig. 4). Theamounts of bacteriochlorophyll and carotenoids in Ros. de-nitrificans were 1.2 times those in Rb. sphaeroides underidentical conditions (near 0% saturation, in darkness).This result is in good agreement with the result that thetotal level of puf mRNA in Ros. denitrificans was 1.3 timesthat in Rb. sphaeroides.

Determination of 5' and 3' ends of the 0.5-kb tran-script—We determined the 5' and the 3' termini of the 0.5-kb transcript, which appeared to accumulate most abun-dantly, by the S-l nuclease protection assay (Fig. 5).Protected fragments of DNA of 260 and 289 nucleotideswere detected with respect to the 5' end of the puf tran-scripts (Fig. 5A, lane 3). The former corresponds to the 5'end at nucleotide (nt) position 74, and the latter to nt 103upstream of the first base of the initiation codon of pufB.The amount of the DNA fragment of 260 nt was onlyabout 10% of that of the fragment of 289 nt. Two bandswere also detected by primer extension analysis and the in-tensity of the band that corresponded to the 5' end at nt 74was also about 15% of that corresponding to the 5' end atnt 103 (data not shown). The DNA fragment of 286 nucleo-tides was also detected (Fig. 5-A) but, since this band wasnot detected by primer extension analysis (data not shown),we decided that it was an experimental artifact. However,

50 100Oxygen tension (%)

Fig. 4 The levels of bacteriochlorophyll and carotenoids in cellsgrown at various oxygen tensions in darkness. The levels areexpressed as percentages of maximum values. Each point and barindicate the mean and the standard deviation, respectively, of theresults from more than three independent experiments. (-•-), bac-teriochlorophyll (Ros. denitrificans); (-•-), carotenoids (Ros. de-nitrificans); (-O-), bacteriochlorophyll (Rb. sphaeroides); (-D-),carotenoids (Rb. sphaeroides). A value of 100% corresponds to1.27 and 1.31 nmol bacteriochlorophyll per ml of culture ofRos. denitrificans and of Rb. sphaeroides, respectively, and to1.09 and 0.82 nmol carotenoids per ml of culture of Ros. denitri-ficans and of Rb. sphaeroides, respectively.

Dow

nloaded from https://academ

ic.oup.com/pcp/article/37/2/153/1820311 by guest on 10 August 2022

Expression of p«/operon in Ros. denitrificans 15?

2 3

242 -

238 -

Fig. 5 S-l nuclease protection analysis of the 5' and 3' ends of thepuf transcripts. Results of S-l nuclease protection analysis of 5'ends with probe a shown in Fig. 1 (A) and of 3' ends with probe b in Fig. 1 (B). Lane 1, DNA template without any treatment; lane 2,DNA template treated with S-l nuclease; lane 3, DNA template incubated with 40 ng of total RNA from Ros. denitrificans that had beengrown aerobically in darkness, followed by S-l nuclease treatment. The Maxam-Gilbert sequencing ladders for A + G are shown next tolane 1. The DNA sequence of the coding strand is shown on the left or the right of panels A and B. The termination sites are indicated byarrows.

protected fragments of DNA of 238 and 242 nt were detect-ed at almost equal levels with respect to the 3' end of thetranscripts (Fig. 5-B, lane 3), and they corresponded to the3' ends at nt 419 and nt 423 downstream of the first base ofthe initiation codon of pufB, respectively. Thus, the 0.5-

kb transcript appeared to encode pufBA, as previously re-ported in Rb.sphaeroides (Lee et al. 1989a). Althoughthe transcript with the 5' end at nt —74 was detected at lowlevels, it was unclear whether the minor bands (1.2-kb, 1.9-kb, and 3.5-kb transcripts) started at nt - 7 4 or nt —103.

AC (25°C)= -6.0 kcal/mo! AG (25"C)= -4.6 kcal/moli LS.KJ \^J *^J — ~*t.U IM»«Uf UUJI I

GAATTCTCG'ACCTGTGTGCGCCTGTCGAGTCl£AGAG|ACTGGAACCCTGGGAACAGGTTTT|£AQA2lCCCGGAGG-103 -74

B

,u Au

AC (25°C)= -24.02 kcal/mol

/ , c — G T T

UAAUAUGGCG//CUAUCAAUG CGAACUCGUAACUUAAUUUA+321 +419 +423

Fig. 6 Nucleotide sequences near the sites of initiation and termination of the mRNAs transcribed from thepw/operon in Ros. denitri-ficans. The 5' and 3' ends of the transcripts of the pi//operon are indicated by vertical arrows. (A) 5' ends of the transcripts of the pufoperon. Opposing arrows above the sequence indicate putative stem-loop structures. Two boxed sequences, CAGAG, are located justupstream of the 5' ends. (B) 3' ends of the transcripts of the puf operon. The termination codon (UAA) is underlined.

Dow

nloaded from https://academ

ic.oup.com/pcp/article/37/2/153/1820311 by guest on 10 August 2022

158 K. Nishimura et al.

As judged from the length of the transcripts, it seems thatthe 1.2-kb and the 1.9-kb transcripts might not encode theentire length of pufBAL and pufBALM, respectively. The3.5-kb transcript probably encoded the gene for cyto-chrome c (pufBALMC), if the length of the gene for cyto-chrome c in Ros. denitrificans is similar to that in Rhodo-pseudomonas viridis (Weissner et al. 1990).

Several factors might explain the presence of hetero-geneous transcripts in purple non-sulfur bacteria, for exam-ple, differences in transcriptional activity and/or in thestability of mRNAs (Lee et al. 1989a). A study of the sec-ondary structures of the puf transcripts revealed severalstem-loop structures upstream of the 5' termini and the3' termini of the transcripts. Calculations of free energy(Tinoco et al. 1973) suggest that stable secondary structuresare probably formed near the 5' termini ( -6 .0 kcal mol"1

and - 4 . 6 kcal moP1) and the 3' termini (-24.02 kcalmol"1) of the 0.5-kb transcript (Fig. 6). The two 5' terminiare preceded by a consensus sequence CAGAG, as well as asequence with dyad rotational symmetry, as also reportedin Rps. viridis (Weissner et al. 1990). Because no consensussequence was found in the regulatory region upstream ofthe two 5' ends (at positions —103 and -74) of the pufoperon, the regulation of transcription of the puf operonin Ros. denitrificans is probably different from that inRb. sphaeroides. (Bauer et al. 1993, Liebetantz et al. 1991).The secondary structure at the 3' terminus (-24.02 kcalmol"1) is probably a terminator of transcription for pufBAmRNA. It is also possible that the structure is an end pointthat protects against exo-nucleolytic digestion of the longertranscripts, as indicated by other investigators (Belasco andChen 1988, Klug 1993). Although there is no Q gene norany consensus promoter region in the puf operon, Ros. de-nitrificans resembles Rb. sphaeroides with respect to the ex-istence of two 5' ends. These results support the hypothesisthat the two species might have a common origin, as previ-ously reported (Liebetantz et al. 1991).

Light-regulated levels of puf transcripts—We inves-tigated the effects of light on the expression of the pufoperon in Ros. denitrificans grown under aerobic condi-tions (oxygen tension, 78%), in comparison with that inRb. sphaeroides grown under semi-aerobic conditions (ox-ygen tension, 2%). Figure 7 shows the profiles of the puftranscripts in these bacteria when they were shifted fromdarkness to the light. The levels of the individual puf tran-scripts in Ros. denitrificans grown in darkness decreasedafter illumination more significantly than those in Rb. sphae-roides (Fig. 7A). Total levels of these transcripts are shownin Figure 7B. The duration of illumination required for a50% reduction in the level of expression of the puf operonin Ros. denitrificans was approximately 45 min, but that inRb. sphaeroides was approximately 100 min (Fig. 7B).Thus, light more effectively inhibited the expression of thepuf operon in Ros. denitrificans than in Rb. sphaeroides.

I IIDuration of illumination (min)

kb

3.5 -

1.9 -1.2 -

04590180

t

I0 45 90 180

B

kb

2.61.71.1

— • — Ros. denitrificans— O — Rb. sphaeroides

50 100 150 200

Duration of illumination (min)

Fig. 7 Time courses of changes in levels of transcripts of the pufoperon after illumination. Panel A, Northern hybridizationof RNA from Ros. denitrificans (I) and Rb. sphaeroides (II). Cellswere grown under aerobic conditions (oxygen tension, 18%) indarkness and under semi-aerobic conditions (oxygen tension, 2%)in darkness for 1 day, respectively, and were then shifted fromdarkness to the light (2.5 x 106ergs cm~2s~"'). Zero time repre-sents the time just before illumination. Experimental conditionswere the same as described in the legend to Fig. 2. Panel B showstime courses of changes in the levels of total puf mRNA. Eachpoint and bar indicate the mean and the standard deviation, re-spectively, of the results from more than three independent experi-ments.

To determine whether the strong inhibition of the ac-cumulation of puf mRNAs after illumination in Ros. de-nitrificans was due to differences in transcriptional activi-ties or to those in stability between the bacteria, it may benecessary to determine the half-lives of puf mRNAs inRos. denitrificans in a comparison with those in Rb. sphae-roides (Dehoff et al. 1988).

The sensitivity to light of the expression of the pufoperon in Ros. denitrificans might help cells to avoid photo-dynamic damage due to bacteriochlorophyll under highlyaerobic conditions, with insensitivity to oxygen and sensitiv-ity to light compensating for one another. The compensa-tion might be a strategy that allows this bacterium to adaptto highly aerobic conditions.

Dow

nloaded from https://academ

ic.oup.com/pcp/article/37/2/153/1820311 by guest on 10 August 2022

Expression of puf operon in Ros. denitriflcans 159

This work was supported financially by a Grant for ScientificResearch (no. 05454015) and a Grant-in-Aid for Scientific Re-search on Priority Areas (no. 04273101) from the Ministry ofEducation, Sports, Science and Culture, Japan.

References

Bauer, C , Buggy, J. and Mosley, C. (1993) Control of photosystem genes• in Rhodobacter capsulatus. Trends Genet. 9: 56-60.

Bauer, C.E. and Marrs, B.L. (1988) Rhodobacter capsulatus pufoperon en-codes a regulatory protein (PufQ) for bacteriochlorophyll biosynthesis.Proc. Natl. Acad. Sci. USA 85: 7074-7078.

Belasco, J.G. and Chen, C.A. (1988) Mechanism of pu/mRNA degrada-tion: the role of an intercistronic stem-loop structure. Gene 72: 109-117.

Cohen-Bazire, G., Sistrom, W.R. and Stanier, R.Y. (1957) Kinetic studiesof pigment synthesis by non-sulfur purple bacteria. / . Cell. Comp.Physiol. 49: 25-68.

Dehoff, B.S., Lee, J.K., Donohue, T.J., Gumport, R.I. and Kaplan, S.(1988) In vivo analysis of puf operon expression in Rhodobacter sphae-roides after deletion of a putative intercistronic transcription terminator.J. Bacterial. 170: 4681-4692.

Donohue, T.J., Hoger, J.H. and Kaplan, S. (1986) Cloning and expressionof the Rhodobacter sphaeroides reaction center H gene. J. Bacteriol.168: 953-961.

Drews, G. (1978) Structure and development of the membrane system ofphotosynthetic bacteria. In Current Topics in Bioenergetics. Edited bySanadi, D.R. and Vernon, L.P. Vol. 8. Photosynthesis: Part B. pp. 161-207. Academic Press, Inc., New York.

Dryden, S.C. and Kaplan, S. (1990) Localization and structural analysis ofthe ribosomal RNA operons of Rhodobacter sphaeroides. Nucl. AcidsRes. 18: 7267-7277.

Farchaus, J.W., Barz, W.P., Grunberg, H. and Oesterhelt, D. (1992)Studies on the expression of the pufX polypeptide and its requirementfor photoheterotrophic growth in Rhodobacter sphaeroides. EMBO J.11: 2779-2788.

Gong, L., Lee, J.K. and Kaplan, S. (1994) The Q gene of Rhodobactersphaeroides: Its role in puf operon expression and spectral complexassembly. J. Bacteriol. 176: 2946-2961.

Harashima, K., Hayasaki, J., Ikari, T. and Shiba, T. (1980) O2-stimulatedsynthesis of bacteriochlorophyll and carotenoids in marine bacteria.Plant Cell Physiol. 21: 1283-1294.

Iba, K. and Takamiya, K. (1989) Action spectra for inhibition by light ofaccumulation of bacteriochlorophyll and carotenoid during aerobicgrowth of photosynthetic bacteria. Plant Cell Physiol. 30: 471-477.

Kiley, P.J., Donohue, T.J., Havelka, W.A. and Kaplan, S. (1987) DNA se-quence and in vitro expression of the B875 light-harvesting polypeptidesof Rhodobacter sphaeroides. J. Bacteriol. 169: 742-750.

Kiley, P.J. and Kaplan, S. (1988) Molecular genetics of photosyntheticmembrane biosynthesis in Rhodobacter sphaeroides. Microbiol. Rev.52: 50-69.

Klug, G. (1993) The role of mRNA degradation in the regulated expressionof bacterial photosynthesis genes. Mol. Microbiol. 9: 1-7.

Lee, J.K., DeHoff, B.S., Donohue, T.J., Gumport, R.I. and Kaplan, S.(1989a) Transcriptional analysis of puf operon expression in Rhodo-bacter sphaeroides 2.4.1. and an intercistronic transcription terminatormutant. / . Biol. Chem. 264: 19354-19365.

Lee, J.K., Kiley, P.J. and Kaplan, S. (1989b) Posttranscriptional controlof puc operon expression of B8OO-B85O light-harvesting complex forma-tion in Rhodobacter sphaeroides. J. Bacteriol. 171: 3391-3405.

Liebetantz, R., Hornberger, U. and Drews, G. (1991) Organization of thegenes coding for the reaction-centre L and M subunits and B870 antennapolypeptides a and fi from the aerobic photosynthetic bacterium Erythro-bacter species OCH114. Mol. Microbiol. 5: 1459-1468.

Lowry, O.H., Rosenbrough, N.J., Farr, A.L. and Randall, R.J. (1951)Protein measurement with the Folin phenol reagent. / . Biol. Chem. 193:265-275.

Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular Cloning: aLaboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Har-bor, New York.

Maxam, A.M. and Gilbert, W. (1980) Sequencing end-labeled DNA withbase-specific chemical cleavages. Methods Enzymol. 65: 499-560.

Shiba, T. (1987) O2 regulation of bacteriochlorophyll synthesis in theaerobic bacterium Erythrobacter. Plant Cell Physiol. 28: 1313-1320.

Shiba, T., Shimizu, U. and Taga, N. (1979) Another aerobic bacteriumwhich contains bacteriochlorophyll a. Bull. Jap. Soc. Sci. Fish. 45: 801.

Shimada, H., Ohta, H., Masuda, T., Shioi, Y. and Takamiya, K. (1993) Aputative transcription factor binding to the upstream region of the pufoperon in Rhodobacter sphaeroides. FEBS Lett. 328: 41-44.

Shioi, Y. (1986) Growth characteristics and substrate specificity of aerobicphotosynthetic bacterium, Erythrobacter sp. (OCh 114). Plant CellPhysiol. 27: 567-572.

Taga, N. (1968) Some ecological aspects of marine bacteria in the Kuroshiocurrent. Bull. Misaki Mar. Biol. Inst. Kyoto Univ. 12: 56-76.

Tichy, H.V., Oberle, B., Stiehle, H., Schiltz, E. and Drews, G. (1989)Genes downstream from pucB and pucA are essential for formation ofthe B800-850 complex of Rhodobacter capsulatus. J. Bacteriol. 171:4914-4922.

Tinoco, I., Borer, P.N., Dengler, B., Levine, M.D., Uhlenbeck, O.C.,Crothers, D.M. and Gralla, J. (1973) Improved estimation of secondarystructure in ribonucleic acids. Nature (London) New Biol. 246: 40-41.

Wiessner, C , Dunger, I. and Michel, H. (1990) Structure and transcriptionof the genes encoding the B1015 light-harvesting complex P and a sub-units and the photosynthetic reaction center L, M, and cytochrome c sub-units from Rhodopseudomonas viridis. J. Bacteriol. 172: 2877-2887.

Williams, J.C., Steiner, L.A., Feher, G. and Simon, M.I. (1984) Primarystructure of the L subunit of the reaction center of Rhodopseudomonassphaeroides. Proc. Natl. Acad. Sci. USA 81: 7303-7307.

Williams, J.C., Steiner, L.A., Ogden, R.C., Simon, M.I. and Feher, G.(1983) Primary structure of the M subunit of the reaction center fromRhodopseudomonas sphaeroides. Proc. Natl. Acad. Sci. USA 80: 6505-6509.

Yanisch-Perron, C , Vieira, J. and Messing, J. (1985) Improved M13phage cloning vectors and host strains: nucleotide sequences of the M13mpl8 and pUC19 vectors. Gene 33: 103-119.

Youvan, D.C. and Ismail, S. (1985) Light-harvesting II (B8OO-85O com-plex) structural genes from Rhodopseudomonas capsulata. Proc. Natl.Acad. Sci. USA 82: 58-62.

Zhu, Y.S. and Kaplan, S. (1985) Effects of light, oxygen, and substrates onsteady-state levels of mRNA coding for ribulose-l,5-bisphosphate carb-oxylase and light-harvesting and reaction center polypeptides in Rhodo-pseudomonas sphaeroides. J. Bacteriol. 162: 925-932.

Zhu, Y.S., Kiley, P.J., Donohue, T.J. and Kaplan, S. (1986) Origin of themRNA stoichiometry of the puf operon in Rhodobacter sphaeroides. J.Biol. Chem. 261: 10366-10374.

(Received September 28, 1995; Accepted December 22, 1995)

Dow

nloaded from https://academ

ic.oup.com/pcp/article/37/2/153/1820311 by guest on 10 August 2022

Dow

nloaded from https://academ

ic.oup.com/pcp/article/37/2/153/1820311 by guest on 10 August 2022