Role of the H Protein in Assembly of the Photochemical Reaction Center and Intracytoplasmic Membrane in Rhodospirillum rubrum

Abstract

Rhodospirillum rubrum is a model for the study of membrane formation. Under conditions of oxygen limitation, this facultatively phototrophic bacterium forms an intracytoplasmic membrane that houses the photochemical apparatus. This apparatus consists of two pigment-protein complexes, the light-harvesting antenna (LH) and photochemical reaction center (RC). The proteins of the photochemical components are encoded by the puf operon (LHα, LHβ, RC-L, and RC-M) and by puhA (RC-H). R. rubrum puf interposon mutants do not form intracytoplasmic membranes and are phototrophically incompetent. The puh region was cloned, and DNA sequence determination identified open reading frames bchL and bchM and part of bchH; bchHLM encode enzymes of bacteriochlorophyll biosynthesis. A puhA/G115 interposon mutant was constructed and found to be incapable of phototrophic growth and impaired in intracytoplasmic membrane formation. Comparison of properties of the wild-type and the mutated and complemented strains suggests a model for membrane protein assembly. This model proposes that RC-H is required as a foundation protein for assembly of the RC and highly developed intracytoplasmic membrane. In complemented strains, expression of puh occurred under semiaerobic conditions, thus providing the basis for the development of an expression vector. The puhA gene alone was sufficient to restore phototrophic growth provided that recombination occurred.

Rhodospirillum rubrum is a facultatively phototrophic purple nonsulfur bacterium. Under reduced oxygen concentration, this organism forms an intracytoplasmic membrane (ICM) that is the site of the photosynthetic apparatus (15, 16, 21). This apparatus consists of the light-harvesting antenna (LH) and the photochemical reaction center (RC). The pigment-binding proteins, LHα, LHβ, RC-L, and RC-M, are encoded by the puf operon, while RC-H is encoded by puhA. The nucleotide sequences of puhA and the puf operon have been determined for R. rubrum (7, 9, 10) and related bacteria (20, 25, 28, 29, 40, 42, 43, 47, 48).

R. rubrum may grow phototrophically under anaerobic light conditions or by respiration under aerobic or anaerobic conditions in the dark. Because R. rubrum is capable of growth under conditions for which the photosynthetic apparatus is not required, and because the photosynthetic apparatus and the ICM may be induced by laboratory manipulation of oxygen concentration, this is an excellent organism in which to study membrane formation (15, 16).

In previous studies from this laboratory, the puf region was cloned and interposon mutations within this region were constructed (21). R. rubrum P5, in which most of the puf genes were deleted, was shown to be incapable of phototrophic growth and ICM formation. P5 was restored to phototrophic growth and ICM formation by complementation with puf in trans (21, 26). These results imply that in R. rubrum the puf gene products are required for ICM formation. These results differ from those obtained with a puf interposon mutant of Rhodobacter sphaeroides (17) which was phototrophically incompetent but still capable of ICM formation (24). In the case of R. sphaeroides, the formation of ICM in the absence of the puf products may be attributable to the presence of an accessory light-harvesting component (LHII) encoded by puc (23). This implies that R. rubrum is a simpler model for studies of membrane formation.

Because the puf-encoded proteins are required for ICM formation in R. rubrum and because the RC is assembled from puf and puhA products, it is important to evaluate the role of puhA-encoded RC-H in RC assembly and ICM formation in R. rubrum. This study describes the cloning, mutation, and complementation of the puhA region of R. rubrum and demonstrates that as in Rhodobacter capsulatus and R. sphaeroides, RC-H is required for the assembly of a functional photosynthetic apparatus. In addition, in R. rubrum RC-H is required for maximal ICM formation. On the basis of these studies, a model for the assembly of a membrane protein complex is proposed.

MATERIALS AND METHODS

Growth of bacteria.

Bacterial strains and plasmids are listed in Table 1. R. rubrum strains were grown at 30°C in modified Ormerod's medium (33) as described previously (31). Aerobic cultures (500 ml) were grown in 2,800-ml Fernbach flasks with shaking at 300 rpm. The optical density at 680 nm of aerobic cultures did not exceed 0.5, thus avoiding reduction of oxygen in dense cultures. The photosynthetic apparatus was induced by incubation under semiaerobic conditions as described previously (16). Phototrophic cultures were grown at 25°C in screw-cap tubes on a rotating platform illuminated by four incandescent lamps at 100 W/m². R. rubrum R5 was grown in the presence of rifampin (15 μg/ml) to counterselect for donors in conjugations as previously described (21). Kanamycin (15 μg/ml for R. rubrum and 50 μg/ml for Escherichia coli), tetracycline (12.5 μg/ml), chloramphenicol (30 μg/ml), and spectinomycin (25 μg/ml) were added to the medium as appropriate. Due to its photolability, tetracycline was omitted from the medium when complemented strains were cultured in the light; in this case, obligate phototrophy selected for plasmid maintenance.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Relevant characteristics	Reference or source
E. coli
S17-1	Donor for interspecific conjugation	36
JM 109	Host used for preparation of libraries	45
R. rubrum
R5	Rifr mutant of R. rubrum S1, puhA+	21
P5	puf interposon mutant, Kanr	21
H15	puhA interposon mutant, Kanr	This study
Plasmids
pRK404E1	IncP, pRK404 (19) with second EcoRI site deleted, Tetr	G. Roberts
pH3.6+/−	3.6-kb HindIII puhA fragment cloned into pRK404E1; +/− refers to orientation with respect to lac; Tetr	This study
pB7.1+/−	7.1-kb BamHI puhA fragment cloned into pRK404E1; +/− refers to orientation with respect to lac; Tetr	This study
pUC19	ColE1 replicon, Ampr	32
pBU	3.7-kb puh upstream sequence cloned in pUC19	This study
pH15	pH3.6− modified by substitution of Kanr cassette for PstI fragment, Tetr Kanr	This study
pPH1JI	IncP, tra+ Specr	22
pE7.7−	puf region cloned into pRK404E1 in the orientation opposite that of lac, Tetr	21
pPUH	puhA structural gene and 359 bp upstream cloned into pRK404E1; direction of transcription is opposite that of the lac promoter of the vector	This work

To assess phototrophic competence of colonies of complemented strains, plates were incubated under aerobic conditions until colonies formed. The plates were then transferred to an anaerobic GasPak (BBL Microbiology Systems, Cockeysville, Md.) and incubated under illumination. Colonies that enlarged and formed photopigments were scored as phototrophically competent (PS⁺). Photosynthetically incompetent colonies remained pale pink.

Molecular biology and genetic techniques.

Plasmid DNA was isolated using the modified miniprep method (50) and a Qiagen kit (Qiagen Inc., Chatsworth, Calif.). Restriction digestion, electrophoresis of DNA, and Southern analysis were carried out using standard methods (35). Two partial libraries of size-fractionated BamHI- and HindIII-digested R. rubrum DNA were prepared in the broad-host-range vector pRK404E1. puhA clones were identified by colony hybridization with an 821-bp puhA PCR product obtained with primers designed on the basis of sequence of the region immediately flanking the puhA structural gene (10).

An interposon mutant was generated by the approach used previously (21). The PstI fragment of pH3.6+ extending from within G115 through the first 161 bp of puhA (Fig. 1; Table 1) was replaced by a kanamycin resistance cassette (Kan^r Genblock; Pharmacia Biotech, Milwaukee, Wis.) to generate pH15. E. coli S17-1 was transformed with pH15, and the plasmid was transferred to R. rubrum R5 by interspecific conjugation. A double crossover to replace the chromosomal puhA gene was obtained by the introduction of the IncP incompatible plasmid pPH1JI (spectinomycin resistant [Spec^r]) into pH15-containing R. rubrum and selection for Kan^r and Spec^r. The genetic structure of the mutants was confirmed by Southern blots probed with the Kan^r cassette and with the puhA PCR product. For complementation analysis, pH3.6+ and pH3.6− were delivered to the mutant via conjugation with E. coli S17-1. To construct a plasmid that could be used to deliver puhA to mutated strains, the puhA structural gene and sequence extending 359 bp upstream of the start codon were amplified by PCR and cloned into pRK404E1 to form pPUH (Fig. 1; Table 1).

FIG. 1.

Genetic and restriction map of puhA region and constructs. ORFs G155, I2372, and I3087 were previously identified (10). In pH3.6+, puhA and flanking ORFs are in the same orientation as the lac promoter (plac) of the vector. In pH3.6− (not shown), the fragment is cloned in the opposite orientation with respect to the lac promoter. pH15 was constructed by substitution of the Kan^r cassette for the PstI fragment of pH3.6+. pB7.1+/− extends from 5.4 kb upstream of puhA to the BamHI site in I3087. pPUH includes the puhA structural gene and 359 bp of upstream sequence.

Double-stranded DNA was sequenced in both directions on an ABI model 373A automated DNA sequencer (Applied Biosystems Inc., Norwalk, Conn.). For sequencing, the 3.7-kb BamHI-HindIII fragment from pB7.1− was cloned into pUC19 to form pBU. Sequence comparison was performed using the BLAST algorithm (1).

Analytical procedures.

Samples were prepared for electron microscopy as previously described (14). Cell extracts were prepared, and membranes were recovered by centrifugation for 30 min at 90,000 rpm (353,000 × g) in a TLA100.2 rotor in an Optima TL centrifuge (30) (Beckman Instruments, Inc., Palo Alto, Calif.). Protein concentration was measured with the bicinchoninic acid protein assay reagent (Pierce Chemical Co., Rockford, Ill.). Bacteriochlorophyll (BCHL) and carotenoid (CRT) were measured as previously described (21). Electron spin resonance spectroscopy (ESR) was performed on a Varian E-4 X-band spectrometer at the National EPR Center of the Medical College of Wisconsin. Samples were frozen under saturating illumination or in the dark. Typical parameters were as follows: microwave power, 1 μW; modulation amplitude, 8 G; time constant, 0.25 s; scan width, 100 G; and modulation frequency, 100 kHz.

Nucleotide sequence accession number.

The nucleotide sequences reported in this paper have been submitted to the GenBank database under accession no. AF202319.

RESULTS

Cloning of puhA.

To analyze the role of puhA in RC assembly and ICM formation in R. rubrum, puhA was cloned from partial libraries. Clones were recognized by hybridization to a puhA probe that was prepared by PCR amplification using primers designed on the basis of available sequence information (10). Two clones, designated pH3.6+ and pH3.6−, contained the 3.6-kb HindIII fragment cloned into pRK404E1 in the same and opposite orientation, respectively, relative to the vector lac sequence (Fig. 1). Clones designated pB7.1+ and pB7.1− containing a 7.1-kb BamHI fragment with additional upstream sequence (Fig. 1) were isolated from a BamHI partial library.

Identification of upstream ORFs.

The sequences of puhA and flanking open reading frames (ORFs) G115, I2372, and I3087 have been reported (10). As a prelude to further studies of the puh region, the sequence of the 3.7 kb upstream of G115 was determined. Two ORFs and one partial ORF were identified (Fig. 1). On the basis of inferred amino acid sequence, these ORFs encode genes similar to the BCHL biosynthesis genes bchH (56% identical, 70% similar), bchL (53% identical, 63% similar), and bchM (53% identical, 66% similar) of R. capsulatus (11, 48). The gene organization is the same in R. rubrum and R. capsulatus.

Mutagenesis and complementation of puhA.

A puhA mutant of R. rubrum was generated by gene replacement. An interposon was substituted for the PstI fragment of pH3.6+ to construct pH15 (Fig. 1). The deleted fragment extends from upstream of puhA through the first 161 bases of puhA. This construction also has a deletion of 76% of the 3′ end of an upstream ORF G115. The construct pH15 was introduced by conjugation into R. rubrum R5, and recombinants that resulted from double reciprocal crossover were isolated. After confirmation by Southern analysis (not shown), one of these recombinants, R. rubrum H15, was selected for further analysis. Plasmids pH3.6+ and pH3.6− were introduced into H15 by conjugation.

Growth under phototrophic conditions.

R. rubrum R5, H15 and complemented H15 strains were incubated under conditions requiring phototrophic growth. R. rubrum H15 was incapable of phototrophic growth (Fig. 2). The ability to grow under phototrophic conditions was restored by complementation with either pH3.6+ or pH3.6−. While R5 and H15(pH3.6+) were capable of phototrophic growth regardless of the incubation conditions of the inoculum, cultures of H15(pH3.6−) inoculated with aerobically grown cells were incapable of phototrophic growth. A further difference between H15 complemented with the pH3.6 constructs and R5 is the growth rate. The shortest generation time was observed for R5 (Table 2). A longer generation time was observed for pH3.6− than for pH3.6+ (Table 2).

FIG. 2.

Growth curves of R. rubrum R5, H15, and complemented H15 strains incubated under obligate phototrophic conditions. Points are means for four cultures. ○, cultures inoculated with aerobically grown cells; ●, cultures incubated with cells incubated for 18 h under semiaerobic (inducing) conditions.

TABLE 2.

Generation times of phototrophic cultures^a

Strain	Inoculum	Generation time (h)
R5	Aerobic	4.7 ± 0.3
	Semiaerobic	4.5 ± 0.2
H15	Aerobic	No growth
	Semiaerobic	No growth
H15(pH3.6+)	Aerobic	13.7 ± 0.7
	Semiaerobic	8.1 ± 0.6
H15(pH3.6−)	Aerobic	No growth
	Semiaerobic	24.9 ± 0.8
P5(pE7.7−)	Aerobic	5.7 ± 0.2

^aCorresponding growth curves are shown in Fig. 2. 

Restoration of phototrophic growth by the puhA clones was due to complementation in trans rather than to recombination. When complemented strains were grown aerobically without antibiotic selection to allow spontaneous curing of the plasmid, all cells that remained Tet^r also remained PS⁺; however, most of the cells became Tet^s and PS⁻ (Table 3). While approximately 1% of the cells remained Tet^r, 12 to 17% were PS⁺. A similar quantitative discrepancy between tetracycline resistance and phototrophic competence was observed when a pRK404E1-based construct was used to complement puf mutant R. rubrum (21). These results imply that a single copy of puf or puh is required for complementation, while a greater gene dosage is required to confer resistance to tetracycline (21). Further evidence that complementation occurs in trans is provided by the displacement of pH3.6− from H15 by the introduction of an incompatible plasmid resulting in the loss of phototrophic competence (Y. S. Cheng and M. L. P. Collins, unpublished data).

TABLE 3.

Plasmid maintenance and phototrophic competence of H15(pH3.6+) and H15(pH3.6−)

Strain	−Tetracycline		+Tetracycline
	CFU/ml (106)	% PS+	CFU/ml (104)	% PS+
H15(pH3.6+)	9.1	12	3.4	100
H15(pH3.6−)	6.2	17	8.1	100

To determine if the restoration of phototrophic growth required only puhA, pPUH (Fig. 1), which contained only the puhA structural gene and a portion of the upstream ORF, was introduced into H15, and phototrophic growth was restored. However, this occurred in only some cultures after a long lag period (>10 days), implying that recombination had occurred and recombinants were selected by obligate phototrophic growth. Southern analysis of three of these phototrophic cultures confirmed that the plasmid had integrated into the chromosome by single crossover downstream of the Kan^r cassette (not shown). The difference in the lag between these cultures in which the plasmid crossed into the chromosome and H15(pH3.6+/−) (Fig. 2) provides additional evidence that in the latter, phototrophy was restored by complementation in trans.

Photopigment content and spectral analysis.

Incubation of R. rubrum under semiaerobic conditions results in gratuitous induction of the photosynthetic apparatus (15, 16, 21). This may be observed by detection of photopigments, spectral components, and ICM (21). The BCHL content of R. rubrum R5 increased 12-fold during incubation under semiaerobic conditions (Table 4). Induction resulted in an 11-fold increase in the BCHL content of H15, but the BCHL level of H15 under either aerobic or inducing conditions was only 28 to 30% of the level for R5. Similarly, CRT levels increased 4.9- and 3.8-fold in R5 and H15, respectively, but the levels in induced H15 were 34% of those of R5 (Table 4). Introduction of pH3.6+ or pH3.6− into H15 partially restored BCHL and CRT levels (Table 4). H15 strains complemented with either pB7.1+ or pB7.1− had pigment levels comparable to those of strains complemented with pH3.6+ or pH3.6− (not shown).

TABLE 4.

Photopigment content of membranes

Strain	Condition	Photopigment content (nmol/mg of protein)a
		BCHL	CRT
R5	Aerobic	2.88 ± 0.66	4.21 ± 0.73
	Semiaerobic	34.1 ± 4.73	20.8 ± 2.73
H15	Aerobic	0.87 ± 0.35	1.85 ± 0.08
	Semiaerobic	9.52 ± 1.87	6.97 ± 0.41
H15(pH3.6+)	Semiaerobic	27.4 ± 1.97	15.4 ± 2.03
H15(pH3.6−)	Semiaerobic	22.8 ± 2.43	13.6 ± 1.42

^aMean ± standard deviation for two to five cultures. 

Spectral peaks at 800 and 880 nm are characteristic of RC and LH, respectively. Spectral analysis (Fig. 3) showed that mutant strain H15 had a reduced LH content and undetectable RC. Complementation by pH3.6+ or pH3.6− restored the RC and increased the LH content but not to the wild-type level. Comparison of H15(pH3.6−) and H15(pH3.6+) showed consistently in three independent experiments a higher level of LH in the latter.

FIG. 3.

Absorbance spectra of membranes (125 μg of protein/ml) prepared from R5, H15, H15(pH3.6+), and H15(pH3.6−) from cells incubated under semiaerobic conditions.

ESR provides a means to detect and quantitate the photooxidized BCHL dimer of the RC, which has a characteristic signal at g = 2.0026 (27). ESR of R5 cells reveals a characteristic light-dependent signal at g = 2.0030 ± 0.0008 (Fig. 4). ESR analysis of H15 reveals that this signal is present and detectable but at 7 to 11% of the wild-type level.

FIG. 4.

ESR spectra. Equivalent amounts of R5 and H15 cells cultured under semiaerobic conditions were frozen in the dark (gray) or under saturating illumination (black).

ICM formation.

Ultrastructural analysis was used to evaluate ICM formation in wild-type, mutant, and complemented strains incubated under semiaerobic conditions (Fig. 5). As previously demonstrated (15, 16), wild-type R. rubrum forms abundant ICM under these inducing conditions. In contrast, H15 was impaired in ICM formation. Most cells observed in thin section contained no ICM. When ICM was present, it was usually observed as a single vesicle. H15 complemented with either pH3.6+ or pH3.6− formed ICM at a level intermediate between the wild-type and H15 levels.

FIG. 5.

Electron micrographs of R5, H15, and complemented H15 strains. Bar = 0.5 μm. ICM is indicated by arrows.

DISCUSSION

Requirement for RC-H for phototrophic growth and role in ICM formation.

The inability of the puhA deletion mutant H15 to grow under phototrophic conditions and the restoration of phototrophic competence by complementation with pH3.6+/− and by integration of pPUH indicate that RC-H is required for phototrophic growth, confirming previous studies of other species (12, 37, 44). The failure of pPUH to restore phototrophy by complementation is probably attributable to the absence of 3′ sequences that could form an RNA stem-loop. This proposed stem-loop has been suggested to function as a transcription terminator (8); alternatively or additionally, it may provide transcript stability. Integration of pPUH into the chromosome restores a complete puh copy including the putative stem-loop sequence. A construct which included the puh structural gene, 359 bp of flanking upstream sequence, and 214 bp of flanking downstream sequence complemented H15 to phototrophy (Cheng and Collins, unpublished data).

The requirement of RC-H for phototrophic growth of R. rubrum is consistent with the characteristics of puhA mutants of R. sphaeroides (12, 37) and R. capsulatus (44). However, in contrast to the results obtained with puhA mutant R. sphaeroides (37), R. rubrum H15 is impaired in ICM formation. The residual level of ICM in H15 may be attributable to the presence of LH, which is absent from the R. rubrum puf mutant P5 (21). ICM formation is restored by complementation with pH3.6+ or pH3.6−. The ICM present in complemented H15 may be due to expression of puhA, resulting in assembly of the RC. Alternatively or additionally, this may be due to the increased LH in the complemented strains. In either case, these results suggest that ICM proliferation requires the assembly of the major ICM proteins, consistent with observations for the puf mutant P5 (21).

Gene organization and expression.

The genes encoding pigment-binding proteins and enzymes involved in pigment biosynthesis are organized in a cluster that is conserved among photosynthetic bacteria (4, 6). The cloning and sequencing of bchL, bchM, and the partial bchH in this study provide further evidence for the conservation of the structure of this region. These genes have been suggested to be organized in transcriptional units termed superoperons (reviewed in references 3 and 41). As a result of this transcriptional organization, the puf and puhA operons are expressed from strong oxygen-repressed proximal promoters embedded in adjacent upstream genes and by transcriptional readthrough from distal promoters that are less tightly regulated by oxygen. The results of this study support the organization of R. rubrum puhA in a superoperon. The phototrophic growth rate of the wild-type R5 was greater and the lag was shorter than for either complemented strain (Fig. 2; Table 2), consistent with the suggestion that expression from a distal upstream promoter facilitates transition to phototrophic growth (5). As we did not observe a difference between H15 complemented with pH3.6+/− or pB7.1+/−, we conclude that the upstream promoter is probably not within the 5.4 kb upstream of puhA contained on the latter plasmid. This would be consistent with the detection of an 11-kb puhA transcript in R. capsulatus (5).

Because both pH3.6+ and pH3.6− can complement H15, it may be concluded that sequences sufficient for puhA expression are contained within the 3.6-kb HindIII fragment. Based on similarity to R. capsulatus puhA and puf, R. viridis puf, and R. sphaeroides puf, a promoter sequence was proposed for R. rubrum puhA (2). This sequence is also similar to the proposed proximal R. rubrum puf promoter (26). This conserved sequence is located −281 to −301 bp upstream from the puhA ATG and is contained within the cloned fragment in pH3.6+/−. Despite the similarity between the putative R. rubrum puhA and puf promoters, this study shows that the former is expressed under semiaerobic conditions whereas the latter is not (21, 26). Thus, while the R. rubrum puhA promoter is regulated by oxygen, it is more tolerant of oxygen than is the puf promoter. These findings have provided the basis for the construction of an expression vector for R. rubrum based on the puh proximal promoter, thus providing for oxygen-regulated transcription of the cloned gene (Cheng and Collins, unpublished data).

The higher growth rate (Table 2) of H15(pH3.6+) in comparison to H15(pH3.6−) suggests that there is increased expression of puhA when the insert is in the same orientation with respect to the lac promoter of the vector. This implies that the lac promoter is functional in R. rubrum.

Effect of puhA region on LH.

The level of LH is lower in H15 than in the wild-type R5 (Fig. 3). This may be attributable to any of the following: (i) partial deletion of the upstream ORF G115, (ii) an effect on expression of the downstream ORFs I2372 and/or I3087, or (iii) deletion of puhA. This is consistent with studies of related bacteria. Mutation of ORF 1696, which is similar to ORF G115, reduces LHI in R. capsulatus (5, 46, 49). Recent studies with R. capsulatus suggest that ORFs similar to I2372 and I3087 are involved in the assembly of photochemical components (44). It has also been suggested that sequences downstream of puhA in R. sphaeroides are required for optimal phototrophic growth (12). The reduction of LHI in a nonpolar puhA deletion mutant of R. capsulatus suggests that puhA has an effect on LH synthesis or assembly (44). The construct pH3.6+ is more effective in restoration of LH (Fig. 3) than pH3.6−, but neither restores the wild-type levels. The partial restoration of LH in semiaerobic cultures of H15(pH3.6−) may be attributable to expression of puhA and/or the downstream ORFs I2372 and I3087 from the proximal puhA promoter. The greater increase in LH with pH3.6+ in comparison to pH3.6− may be due to expression of G115 from the lac promoter in the vector in pH3.6+. These effects on LH in H15 are likely to be due to a posttranscriptional event. The presence of residual LH, as well as evidence for RC-L and RC-M (see below), indicates that the puf operon is transcribed. Moreover, on the basis of Northern analysis and expression of lac fusions, respectively, puf expression was reported to be unaffected in puhA mutants of R. sphaeroides and R. capsulatus (37, 44).

Roles of RC-H.

RC-H is not absolutely essential for primary photochemical activity because a low amount of the photooxidized BCHL dimer was detected in H15 by ESR (Fig. 4). This is consistent with studies of R. sphaeroides in which this activity was partially retained when RC-H was modified by mutagenesis or removed from purified RCs in vitro (18, 38). The failure of H15 to grow phototrophically may be attributable to the need for RC-H for electron transfer between the quinones (18, 38) or proton transfer to a bound quinone molecule (34). The phototrophic incompetence of H15 may also be due to a structural requirement for RC-H for assembly of the RC.

Model: RC-H is a foundation protein.

On the basis of this work, we propose a speculative model suggesting that RC-H serves as a foundation protein on which the other RC components are assembled (Fig. 6). The detection of a low level of the photooxidized BCHL dimer in H15 suggests that without RC-H, low levels of RC-L and RC-M may be present in the membrane. Similarly, weak primary photochemical activity was reported for a puhA deletion mutant of R. sphaeroides (37). The presence of low levels of RC-L and RC-M in R. rubrum H15 is consistent with sodium dodecyl sulfate-polyacrylamide gel electrophoresis detection of RC-L and RC-M in puhA mutant R. capsulatus (44) and immunoblot detection of RC-M in puhA mutant R. sphaeroides (37). While it is capable of being photooxidized, the incomplete RC formed in H15 is not fully functional because it cannot support phototrophic growth (Fig. 2). This finding implies that while RC-L and RC-M can assemble in the absence of RC-H, this assembly is unstable or inefficient and not fully functional. This would be consistent with a structural role for RC-H.

FIG. 6.

Model for RC assembly. The diagram depicts RC-H, -L, and -M in membranes of wild-type (R5), mutant, and complemented strains cultured under different conditions and the ability of aerobic or semiaerobic inocula of these strains to initiate a phototrophic culture. RC-L and RC-M (represented by dark and light gray ovals, respectively) are not present in aerobic cells because the puf operon is not expressed. RC-H (black oval) is absent from H15 and from H15(pH3.6−) grown under aerobic conditions. It is present in aerobically grown R5 and P5(pE7.7−) due to expression from a distal upstream superoperonic promoter. RC-H is suggested to be expressed from the vector lac promoter in pH3.6+. Under semiaerobic conditions, puhA is proposed to be expressed from an oxygen-regulated promoter contained within the cloned fragment in pH3.6−. While H15 lacks RC-H, the presence of a weak light-dependent ESR signal characteristic of the photooxidized BCHL dimer suggest that RC-L and RC-M are present in the membrane albeit in a low amount because in the absence of H the RC-L–RC-M complex is unstable or inefficiently assembled and functionally impaired.

Low levels of RC-H are detectable by immunoblot analysis of aerobically grown wild-type R. rubrum (Cheng and Collins, unpublished data) and have also been found in aerobically grown R. sphaeroides (13). The presence of RC-H in aerobically grown R. sphaeroides led Kaplan and collaborators to postulate that RC-H may be localized at discrete sites within the cytoplasmic membrane of aerobic cells that serve as insertion sites for BCHL and BCHL-binding proteins upon transition to inducing conditions (13). These investigators presented evidence that RC-M is unstable and present in reduced amounts in the membrane of puhA R. sphaeroides grown under dark anaerobic conditions (37, 39).

This study extends these findings by providing evidence for the assembly pathway. The effects of inoculum on phototrophic competence of the wild-type and complemented strains (Fig. 2) suggest that RC-H must be preexisting in the membrane for functional RC assembly. R5 and H15(pH3.6+) cultures inoculated with aerobic cells grew phototrophically, whereas H15(pH3.6−) did not. In the case of R5, RC-H is probably expressed from the upstream superoperonal promoter that is weakly expressed under aerobic conditions. RC-H is probably expressed in H15(pH3.6+) from the lac promoter under aerobic conditions. In H15(pH3.6−) because RC-H could not be expressed from the proximal puhA promoter under aerobic conditions, only cells induced by semiaerobic conditions had the RC-H foundation and could be used to initiate a phototrophic culture. Together these observations suggest a model in which RC-H must be preexisting in the membrane to serve as a foundation for functional RC assembly. In contrast to the requirement for preexisting RC-H, it is not necessary for the puf-encoded RC-L and RC-M proteins to be present prior to transition to phototrophic growth. This is indicated by the ability of aerobically grown cultures of the complemented puf mutant P5(pE7.7−) to initiate phototrophic growth (Table 2) despite the failure of the puf genes in this construct to be expressed under aerobic or even semiaerobic conditions (21). We propose that RC-L and RC-M are assembled upon the RC-H foundation.