FEMS Microbiologyt,ctlcrs 95 (It~92)213 -218 ~' lt,~;'2Federation of [-umpcan MicrobiologicalSocieties 1137N-11197/~2/$115.1111 Published by Elsevier
Z13
FEMSLE 11511011
An overlap between operons involved in carotenoid and bacteriochlorophyll biosynthesis in Rhodobacter capsulatus D e b r a A. Y o u n g , M o n i k a B e c k c r R u d z i k
~
a n d B a r r y L. M a r t s
DuPont ('el/Ira/Rt'.~l'ltr¢'t!t/lid lh't ehJimwnt. Science alld Engineering l ahoratorw~. I: ur'rmr'ntal Station. llihmnglon. Ih'htwarc. USA
Rccci,.cd 16 April It)t12 Rcvi~,ionrccci~cd 21}May I~U,12 Accepted 21 May I~ttl2 Key words: Carotenoid: Bacterioehlorophyll; Supert)peron: R h o d o b a c t e r capsuhmts; Photosynthetic bacteria
I. S U M M A R Y
2. I N T R O D U ( q ' I O N
A new example of sul3eroperonal g e n t arr a n g e m e n t has b e e n d o c u m e n t e d in the R h o d o b a c t e r calzs'ulatus photosynthetic gene cluster. The p r o m o t e r for the opcron initiated by the b c h l gene is e m b e d d e d within an upstream opcron for carotenoid synthesis. The stop codon for the crtA gone, the only genc in the first opcron, overlaps the start codon of the downstream b c h l gene. As a consequence of this overlap, thc promotcr(s) for the bch opcron must bc located within the crtA structural gone. The b c h l gene is shown here for the first time to bc required for the conversion of p m t o p o r p h y r i n IX to subsequent intermediates in bacteriochlorophyll biosynthesis.
The genes that encode and control the fi~rmation of the photosynthetic apparatus of the members of the genus R h o d o b a c t e r have been studied to understand the regulatory mechanisms inw)lved. Most of these genes are clustered in a single region of the chromosome, and many of them have been sequenced. An interesting regulatory motif, t e r m e d the s u p e r o p e r o n , has emerged in the course of these studies. Superoperons are strings of adjacent operons in which transcription of the downstream opcrons may bc initiated from either their own promoters or from those of the upstream operons. Thus, for one reason or another, termination of transcription does not occur between the operons in a superoperon. In each of the two instances that have bccn characterized to date [I-3], the promoter for a downstream opcron is e m b e d d e d within the coding sequence of the last gene of the adjacent upstream operon, creating overlapping operons. In these two examples, the upstream promoters have been less active z:nd less strongly regulated
Corr¢.ff~omh'nce to: B.I,. Marrs. DuPont ('cntral Rot,catch and
Devdopment, Experimental station p.o. Box ;q(1173,Wilmington, DE 19881)-11173,USA. t Present address: Wcstminslcr College, Biology Dcpartmcnl. New Wilmington. PA 16172-01X)1,USA.
214 than those downstream, thus the net effect has been to maintain low level constitutive synthesis of photosynthetic apparatus transcripts under generally repressing conditions, while allowing for strong induction of the apparatus under other conditions. Armstrong et al. [4] published and analysed ;.in l1039-base pair (bp) sequence of part of the photosynthetic region encoding eight of the nine known carotenoid biosynthetic genes. They proposed, based upon their interpretation of sequences in thc vicinity of the crtA gene, a third example of overlapping operons in the photosynthesis region. Wc show here that a sequencing error misled Armstrong et al. [4] into believing that interrupting the crtA gene within about l kilobasc (kb) from its 3' end could prevent transcription of an adjacent downstream operon de-
w~ted to bacteriochlorophyll (BChl) biosynthesis. In fact, the crtA gcne is about 1 kb shorter than proposed by Armstrong et al. [4], and the remaining coding sequence is the bchl gene, the first gene in the next operon. Therefore, the data presented by Armstrong et al. [4] in support of it third superoperon do not indicate the existence of overlap between the crt,4 and hcll operons. Ironieally~ we also show here that these two operons do overlap, although not as originally proposed by Armstrong et al. [4]. The functional significance of this superoperon is unknown, in addition, we have analysed the phcnotypic differenccs between pohtr and non-polar mutations in the bchl gene. and have shown unambiguously for the first time that a functional bchl gene is required for conversion of protoporphyrin IX to subsequent metabolites.
DE346
H
SB1003
M
~,.lm
~r*
,-i
lira
~H
H
,,,
k\\\\\\\\\"4 m
crtA I
i
I
I
bchl
or*
m
DE392 crtA '
crtA '
A bchI
Fig. I. Genetic-physical map of three strains of R h o d o t m c w r cap~ululu.~ with respect to the cm,4-bchl gcn¢ region. Strain SBliI(13 is the wild-type. Strain DE34f~ has the .(2 interposon conferring spectinonzycin resistance in.,,erted into the EcoRV site of the h c h l gene. Strain DE392 has the 12 intcrposon inserted into the Sphl site of the crt.4 gene and in addilion has an in-frame deletion of the 246.bp Sinai fragment located in tile hchl gene. Bases are n u m b e r e d according to A r m s t r o n g et al. i4] befl~r¢ the correction.
215 3. M A T E R I A L S A N D M E T H O D S ,t. I. Straht construction T h e larger of two Sail fragments of the B a m H I - H fragment o f R. cap.~tdattts D N A [5,6] was cloned into the Sail site of vector pBR322 to produce plasmid pDAY93. T h e omega (t,~) spcctinomycin resistance intcrposon [7] was cloned into the unique E c o R V site of plasmid pDAY93 to create plasmid pDAY71. The .t,) interposon contains flanking transcription and translation termination signals. The 246-bp Sma ! fragment ( 11}00-1246 bp, [4]) of R. CalZ~'ulatus D N A was removed from plasmid pDAY93 to produce plasmid pDAYO4. The ,t_} cartridge was cloned into the Sphl site of R. capsulatus D N A in plasmid pDAY94 to produce plasmid pDAY99. Plasmids pDAY71 and pDAY99 were mated into the gene transfer agent ( G T A ) twerproduccr strain C B I I 2 7 [I] using the mobilizing plasmid pDPTS0 [5]. A G T A cross from the above mated isolates to recipient wild-type strain SB111113 was as previously described [8] to yield, respectively, strains DE346 and DE392 (Fig. I).
3.Z Analytical methods Fluorimetric spectral analysis of a c e t o n e : methanol ( 7 : 2 ) extracts was done as previously described [9]. Carotenoids were characterized its previously described [10] using cells grown in RCV + media [9] u n d e r high aeration [11. Both strands of D N A wcrc sequenced by the dideoxynucleotide method of Simgcr ct al. [11] using two 21-mer oligonucleotide:;: DI starting at
Table
I
Rhodohacter capsuhtlu.~ strains
Dcsignalion
Rc|cvanl markers
Sourccfrclcrcnc¢
CB1127
rif I0 crtDl21: GTA ovcrproduccr r(FlO g!::ht'hl rilZlO g2::crt.4 ht'hl hchDH)O.~' hf-IO r(f 10 crtA203
This F::pcr This paper Marrs [15] Marrs [151 Yen and Marrs 181
DE346 DE392 MBIII08 SB|1}t13 SB3
Young ct al. [11
bp 1864 ( 5 ' - ( ' G ( i A T G T G G C A A A G A T A A C G C 3') and 1)2 starting at bp 1612 ( 5 ' - G G T C A C C A A A G A ( ' C A G A A C G C - 3 ' ). For polymcrasc chain reaction ( P C R ) analysis of strains, two oligonuclcotides wcrc used which flanked the region of interest: DYI, bp 15111-1485 ( 5 ' - G C A A A T C G A C C G C G G T G - 3 ' ) and DY2. bp 8 4 1 - 8 5 7 ( 5 ' - G C ( ~ C A G T C A T A A A G C G C - 3 ' }. PCR wits pcrfl~rmcd using the G e n e A m p PCR Kit with Nat;re D N A Polymcrase (Pcrkin Elmer Cclus) ;.is per instructions.
4. R E S U L T S A N D DISCUSSION Figure 2A shows a codon preference plot of the crt,,! gene proposed by Armstrong et al. [41, Noting that i n t c r p o s o , s in the 5' region blocked carotenoid synthesis but not Bchl synthesis. whereas intcrposons in the 3' region blocked BChl synthesis, Armstrong et ill. [4,12] proposed that this region represented an ~wcrlap between two adjilccnt opcrons, creating a supcroperonal gcnc arrangement. Armstrong c t a ! . [4] could not test the CrtA phcnotypc of their mutant with an interposon in C.c 3' end of the region because the strain carricd an additional crt mutation that blocked cltrotcnoid synthesis earlier in the pathway. Since wc noticed the region of poor codon usage in the middle of the proposed crtA gcnc (arrows in Fig. 2A), wc thought it possible that there were actually t ~ o separate genes in the region labelled crt,4 in Fig. 2A, with the downstream g e n t simply the first gcnc in an operon devoted to BChl synthesis. If this ~,ere correct, the argument advanced by Armstrong et ill. [4] for a superoperon would disappear. To investigate this issue further, we construetcd a strain similar to strain SB226 of Giullano ct al. [0], in which an .Q interposon cartridgc, flanked by transcript ion and translation termination signals, wits placed into the E c o R V site (bp 1303 of Armstrong ct al. [4], see Fig. 1). Giuliano ct al. [6] noted the B c h D - phenotype of SB226 but did not specify the carotenoid status of SB226. The BChl and carotenoid content of the new construct, DE346, was examined.
216
, ~.,-.,, ~ ,,t ..,,., i ,,r~l,, , nT .,,.=,..,.,, ......,..,.,
o
t)
1-
~
, , I r-,...,.. ,,... ,,,,,,.,t i it, ,.,, ,~[~,J ,,,,~..,.| i|.. , , . ,Jl
~
-
, •
-~
I
i
0
ni . l.
.
.
.
.
.
.
.
.
.
.
.
.
i t--'l i
1,000
1.......... ,, .
.....
i
500
.
.
.
.
.
.
.
.
........
.
.
.
"T-,.
1,500
.
.
.
.
.
.
.
.
.
.
.
.
I
.
I.
. . . . h. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
. . .
~
-.
o ..........
F 0
! .......
'
I~;I ii~i
•
i
I.H~II
i i~Ii i I..............
i 500
l
............
i 1.000 Base
Pairs
i I..................
Ull III~U ..............
i 1.500
•
•
'
217 Fluorimetric spectra of strains DE346 and MBI008 a p p e a r e d identical, with a major emission peak at 634 nm and a minor emission peak at 547 rim, characteristic of protoporphyrin IX. Thus. strain DE346 does indeed exhibit a BchD phenotype. Strains SBI003, SB3, MBI008, and DE346 were analysed for the carotenoid content of aerobic cultures (data not shown). SB3 revealed the characteristic C r t A - pigment of spheroidene, while all others revealed a wild-type pattern of carotenoids, distinguished by the predominance of spheroidenone. Thus, although both MBI008 and DE346 are indeed B C h l - , both also exhibit wild-type carotenoids. This demonstrates that the E c o R V site interrupted in DE346 does not lic in the crtA gene, in contrast to the interpretation of Armstrong et al. [4]. We sequenced the suspect area and located one error. A r m s t r o n g et al. [4] reported a stretch of four G nucleotides at bp 1736-1733 ( 5 ' - 3 ' ) , whereas we clearly saw only three G nucleotides. This correction has recently b e e n submitted to G e n B a n k by the Hearst laboratory, accession n u m b e r Z ! 1165, as part of major sequence analysis of the entire photosynthesis region. A codon preference plot of the corrected sequence clearly reveals two distinct open reading frames (Fig. 2B). T h e first open reading frame contains the sites which upon insertional mutagenesis yielded the C r t A - phenotypes for A r m s t r o n g et al. [4,12] and Giuliano et al. [6], whereas the second opcn reading frame contains the sites which yielded a distinct B C h l - phenotype in prior investigations, all involving polar mutations [4,6,12,13]. Thus, we conclude that the first open reading frame is indeed the crtA gene, while the second open reading frame might be a gene involved in an early step of BChl biosynthesis. Taylor et al. [5]
showed that two point mutations in this region of the chromosome map to separate complementation groups but give the same 'BchD" phenotypes. Zscbo and Hcarst [ 13] defined two separate genes. hchl and hchD. in this region. To convincingly show that In'hi does indeed function early in the BChl biosynthetic pathway, construction of a n o n - I ~ l a r mutation in this gene was neccssary. This is because In'hi is upstream from hchD and a polar insertion in bchl might give a phcnotype caused by lack of bchD expression. We created in vitro an in-frame, non-polar deletion in the hchl gene by deletion of the scquenccs from Smal-Smal sites shown in Fig. I., linked to a markcr insertion of the _Q cartridge in the adjacent crtA gene. A G T A - m e d i a t e d cross gave rise to a 3 : I ratio of double (crtA and bchl) mutants : single (crtA) mutants upon selection for the .(.~ cartridge with spectinomycin. Analysis of DE302, one of the double mutants, rcvcalcd a non-photosynthetic, C r t A - , BChl isolate with fluorescence emission and excitation spectra attributable to protoporphyrin IX. PCR of wild-typc SBI003 chromosomal D N A with primers DYI and DY2 yielded a 660-bp fragmcnt. PCR of DE346, containing the approximately 2.0-kb .f.~ fragment gave a 2660-bp fragment, and PCR of DE392, containing the deletion of the Sinai fragment, gave a 414-bp fragment, confirming that both constructions were indeed correct. Since both a polar and a non-polar mutation in thc bchl gone exhibit blocks very early in the BChl pathway, we conclude that this is indeed a gone involved in the conversion of protoporphvrin IX to a subsequent BChl precursor. The crtA and bchi genes a p p e a r to be in separate opcrons. The bchl gene is the first gene in an
Fig. 2. Codon preference plots generated using a pn~gram written by Gribskov el al. [16]. The program identifies efficiently translated genes ;is peaks above open reading frames (open boxes below peaks) that contain few rare c~nlons(vertical dashes below the open reading frames). This is repeated for each of the three possible open reading frames. A c~}doh frequency table used to produce the plot was obtained as previously published Ill. The horizontal dashed lines indicate the b;~ckgroundcodon preference statistic, which is generated by random sequence of the szlme base composition as this DNA sample. The windrowsize and rare codon threshold were set at 25 and I).l. respectively.A. Codon preference plot using the DNA sequence published by Armstrong et al. [4] starting at the Nn~l site (bp 2471) through bp 617. I bp p~st the stop codon attributed by Armstrong et al. [4] to the crtA gene. B. Codon preference plot of the above sequence corrected hy the elimination of one (i nucleotide between bp 1736 and 1733.
218 o p e r o n presumably including hchD as the second; t h e r e may bc additional d o w n s t r e a m g e n e s in this o p e r o n . Since the s t o p c o d o n o f the crtA g c n e (a monocistronic o p e r o n ) overlaps the start c o d o n o f the d o w n s t r e a m hchl g e n e ( b c k m g i n g to the hchlD o p e r o n ) , it a p p e a r s that this region o f D N A is also an example o f s u p c r o p c r o n a l organization o f g e n c s in IClunh~bactercapstdattts. C o d o n p r e f e r e n c e plot analysis selects A T G (bp 1671)1668) as the start m e t h i o n i n c o f the hchl gone r a t h e r than A T G (bp 1577-1575) which is 311 a m i n o acids d o w n s t r e a m . A d d i t i o n a l evidence to s u p p o r t A T G (bp 16711-1668) as the start is s e c n in thc strong a m i n o acid homology b c t w c e n the R. capsuhmls bchl g c n e p r o d u c t and the A r a bulopsis thaliana chloroplast protein cs p r e c u r s o r [14]. C o m p a r i s o n o f these two a m i n o acid seq u e n c e s reveals a striking 49.3C~ identity in a 341-amino acid overlap. T h e strong homology b e t w e e n t h e s e two p r o t e i n s starts midway between the above two A T G s . t h e r e b y strongly supporting the proposal that the first A T G (bp 1670-1668) e n c o d e s the first m c t h i o n i n e o f the Bchl protein, it is this A T G that overlaps the stop c o d o n for the crtA gone. T h e r e must be a p r o m o t e r for the bchlD o p e r o n e m b e d d e d in the crtA structural gene. b e c a u s e strains containing an .f.~ cartridge insertion at the Sphl site in crtA retain a high level o f Bchl synthetic activity. Transcripts initiating at the crtA p r o m o t e r may transcribe the d o w n s t r e a m hchiD o p e r o n , since t h e r e is no intcrcistronic space in which t e r m i n a tion s e q u e n c e s could be located. A t e r m i n a t i o n s e q u e n c e could hypothetically bc e m b e d d e d in
the bchl g e n e if an anti-termimttion system existed that was able to distinguish b e t w e e n transcription initiated at the hch p r o m o t c r c o m p a r e d to the crt p r o m o t e r . Evidence fiw a n t i - t e r m i n a lion in R. ¢'LqZ~'llhltllS r e m a i n s to bc ,liscewcrcd.
REFERENCES [I) Young, D.A.. Bauer. ('.F,.. Williams. J.('. and Marts, B.L 11989) Mol. Gcn. Gcncl. 218. 1-12.
[2] Wellington. ('.1.. and Bcally. T. (1091) J. Bactcriol. 173. 1432-1443. [3] Batncz'.C.E.. Buggy. LJ.. Yang. Z. and Marrs. Fkl,. ( lOOl) Mol. Gcn. Gcnct. 228. 433-444. [4] Armstrong. G.A.. AIbcrti. M.. l.cach. F. and IIcarst. J.E. (IL)S9) Mol. Gen. Gcnct. 211~.254-208. [5] Taylor. D.P.. Cohen, S.N.. Clark. W.G. and Marrs, 13.L.
(11)83) J. Baclcriol. 154. 581}-591). 16] Giuliano. G.. Pollock. D.. Stapp. ]I. and Scolnik. P.A. (1088) Mol. (;ell. Gencl. 213. 78-83. [7] Prcnlki. P. and Krisch. ll.M. 11984) Gone 29. 31)3-313. [8] Yen. II.C. and Marrs. B.L 11076) J. Baclcriol. 121~. 610-629. [9] Biel. A.J. and Marrs. 13.L. (lOS3)J. Bactcriol. 156. fi86694. [lf)] Scolnik. P.A.. Walker. M.A. and Marrs, B.L. 11080)J. Biol. ('hem. 255. 2427-2432. [I I] Sangcr, F.. Nicklens. S. and ('oulson. A.R. 11977) Proc. Nail. Acad. Sci. USA 74. 5463-5467. [ 12] Armstrong. G.A.. Schmidt. A.. Sandmann. G. and Hearst. J.E. (1901})J. Biol. Chem. 265. 8329-8338. [13] Zscbt~. K.M. and Hearst. J.E. 11984) (.'ell 37. 937-947. [141 Koncz. C.. Maycrhofcr. R.. Koncz-Kalman. Z.. Nawralh, C.. Rciss. B., Rcadi. G.P. and Schcll, J. (It~91)) EMBO J. 9. 1337. [15] Marrs. 13. (I081)J. Bactcriol. 146. 11)03--1(}12. [Ib] Gribskov. M.. Dcvcrcux. J. and Burgess. R.R. 11984) Nucleic Acids Rcs. 12. 539-547.
Z13
FEMSLE 11511011
An overlap between operons involved in carotenoid and bacteriochlorophyll biosynthesis in Rhodobacter capsulatus D e b r a A. Y o u n g , M o n i k a B e c k c r R u d z i k
~
a n d B a r r y L. M a r t s
DuPont ('el/Ira/Rt'.~l'ltr¢'t!t/lid lh't ehJimwnt. Science alld Engineering l ahoratorw~. I: ur'rmr'ntal Station. llihmnglon. Ih'htwarc. USA
Rccci,.cd 16 April It)t12 Rcvi~,ionrccci~cd 21}May I~U,12 Accepted 21 May I~ttl2 Key words: Carotenoid: Bacterioehlorophyll; Supert)peron: R h o d o b a c t e r capsuhmts; Photosynthetic bacteria
I. S U M M A R Y
2. I N T R O D U ( q ' I O N
A new example of sul3eroperonal g e n t arr a n g e m e n t has b e e n d o c u m e n t e d in the R h o d o b a c t e r calzs'ulatus photosynthetic gene cluster. The p r o m o t e r for the opcron initiated by the b c h l gene is e m b e d d e d within an upstream opcron for carotenoid synthesis. The stop codon for the crtA gone, the only genc in the first opcron, overlaps the start codon of the downstream b c h l gene. As a consequence of this overlap, thc promotcr(s) for the bch opcron must bc located within the crtA structural gone. The b c h l gene is shown here for the first time to bc required for the conversion of p m t o p o r p h y r i n IX to subsequent intermediates in bacteriochlorophyll biosynthesis.
The genes that encode and control the fi~rmation of the photosynthetic apparatus of the members of the genus R h o d o b a c t e r have been studied to understand the regulatory mechanisms inw)lved. Most of these genes are clustered in a single region of the chromosome, and many of them have been sequenced. An interesting regulatory motif, t e r m e d the s u p e r o p e r o n , has emerged in the course of these studies. Superoperons are strings of adjacent operons in which transcription of the downstream opcrons may bc initiated from either their own promoters or from those of the upstream operons. Thus, for one reason or another, termination of transcription does not occur between the operons in a superoperon. In each of the two instances that have bccn characterized to date [I-3], the promoter for a downstream opcron is e m b e d d e d within the coding sequence of the last gene of the adjacent upstream operon, creating overlapping operons. In these two examples, the upstream promoters have been less active z:nd less strongly regulated
Corr¢.ff~omh'nce to: B.I,. Marrs. DuPont ('cntral Rot,catch and
Devdopment, Experimental station p.o. Box ;q(1173,Wilmington, DE 19881)-11173,USA. t Present address: Wcstminslcr College, Biology Dcpartmcnl. New Wilmington. PA 16172-01X)1,USA.
214 than those downstream, thus the net effect has been to maintain low level constitutive synthesis of photosynthetic apparatus transcripts under generally repressing conditions, while allowing for strong induction of the apparatus under other conditions. Armstrong et al. [4] published and analysed ;.in l1039-base pair (bp) sequence of part of the photosynthetic region encoding eight of the nine known carotenoid biosynthetic genes. They proposed, based upon their interpretation of sequences in thc vicinity of the crtA gene, a third example of overlapping operons in the photosynthesis region. Wc show here that a sequencing error misled Armstrong et al. [4] into believing that interrupting the crtA gene within about l kilobasc (kb) from its 3' end could prevent transcription of an adjacent downstream operon de-
w~ted to bacteriochlorophyll (BChl) biosynthesis. In fact, the crtA gcne is about 1 kb shorter than proposed by Armstrong et al. [4], and the remaining coding sequence is the bchl gene, the first gene in the next operon. Therefore, the data presented by Armstrong et al. [4] in support of it third superoperon do not indicate the existence of overlap between the crt,4 and hcll operons. Ironieally~ we also show here that these two operons do overlap, although not as originally proposed by Armstrong et al. [4]. The functional significance of this superoperon is unknown, in addition, we have analysed the phcnotypic differenccs between pohtr and non-polar mutations in the bchl gene. and have shown unambiguously for the first time that a functional bchl gene is required for conversion of protoporphyrin IX to subsequent metabolites.
DE346
H
SB1003
M
~,.lm
~r*
,-i
lira
~H
H
,,,
k\\\\\\\\\"4 m
crtA I
i
I
I
bchl
or*
m
DE392 crtA '
crtA '
A bchI
Fig. I. Genetic-physical map of three strains of R h o d o t m c w r cap~ululu.~ with respect to the cm,4-bchl gcn¢ region. Strain SBliI(13 is the wild-type. Strain DE34f~ has the .(2 interposon conferring spectinonzycin resistance in.,,erted into the EcoRV site of the h c h l gene. Strain DE392 has the 12 intcrposon inserted into the Sphl site of the crt.4 gene and in addilion has an in-frame deletion of the 246.bp Sinai fragment located in tile hchl gene. Bases are n u m b e r e d according to A r m s t r o n g et al. i4] befl~r¢ the correction.
215 3. M A T E R I A L S A N D M E T H O D S ,t. I. Straht construction T h e larger of two Sail fragments of the B a m H I - H fragment o f R. cap.~tdattts D N A [5,6] was cloned into the Sail site of vector pBR322 to produce plasmid pDAY93. T h e omega (t,~) spcctinomycin resistance intcrposon [7] was cloned into the unique E c o R V site of plasmid pDAY93 to create plasmid pDAY71. The .t,) interposon contains flanking transcription and translation termination signals. The 246-bp Sma ! fragment ( 11}00-1246 bp, [4]) of R. CalZ~'ulatus D N A was removed from plasmid pDAY93 to produce plasmid pDAYO4. The ,t_} cartridge was cloned into the Sphl site of R. capsulatus D N A in plasmid pDAY94 to produce plasmid pDAY99. Plasmids pDAY71 and pDAY99 were mated into the gene transfer agent ( G T A ) twerproduccr strain C B I I 2 7 [I] using the mobilizing plasmid pDPTS0 [5]. A G T A cross from the above mated isolates to recipient wild-type strain SB111113 was as previously described [8] to yield, respectively, strains DE346 and DE392 (Fig. I).
3.Z Analytical methods Fluorimetric spectral analysis of a c e t o n e : methanol ( 7 : 2 ) extracts was done as previously described [9]. Carotenoids were characterized its previously described [10] using cells grown in RCV + media [9] u n d e r high aeration [11. Both strands of D N A wcrc sequenced by the dideoxynucleotide method of Simgcr ct al. [11] using two 21-mer oligonucleotide:;: DI starting at
Table
I
Rhodohacter capsuhtlu.~ strains
Dcsignalion
Rc|cvanl markers
Sourccfrclcrcnc¢
CB1127
rif I0 crtDl21: GTA ovcrproduccr r(FlO g!::ht'hl rilZlO g2::crt.4 ht'hl hchDH)O.~' hf-IO r(f 10 crtA203
This F::pcr This paper Marrs [15] Marrs [151 Yen and Marrs 181
DE346 DE392 MBIII08 SB|1}t13 SB3
Young ct al. [11
bp 1864 ( 5 ' - ( ' G ( i A T G T G G C A A A G A T A A C G C 3') and 1)2 starting at bp 1612 ( 5 ' - G G T C A C C A A A G A ( ' C A G A A C G C - 3 ' ). For polymcrasc chain reaction ( P C R ) analysis of strains, two oligonuclcotides wcrc used which flanked the region of interest: DYI, bp 15111-1485 ( 5 ' - G C A A A T C G A C C G C G G T G - 3 ' ) and DY2. bp 8 4 1 - 8 5 7 ( 5 ' - G C ( ~ C A G T C A T A A A G C G C - 3 ' }. PCR wits pcrfl~rmcd using the G e n e A m p PCR Kit with Nat;re D N A Polymcrase (Pcrkin Elmer Cclus) ;.is per instructions.
4. R E S U L T S A N D DISCUSSION Figure 2A shows a codon preference plot of the crt,,! gene proposed by Armstrong et al. [41, Noting that i n t c r p o s o , s in the 5' region blocked carotenoid synthesis but not Bchl synthesis. whereas intcrposons in the 3' region blocked BChl synthesis, Armstrong et ill. [4,12] proposed that this region represented an ~wcrlap between two adjilccnt opcrons, creating a supcroperonal gcnc arrangement. Armstrong c t a ! . [4] could not test the CrtA phcnotypc of their mutant with an interposon in C.c 3' end of the region because the strain carricd an additional crt mutation that blocked cltrotcnoid synthesis earlier in the pathway. Since wc noticed the region of poor codon usage in the middle of the proposed crtA gcnc (arrows in Fig. 2A), wc thought it possible that there were actually t ~ o separate genes in the region labelled crt,4 in Fig. 2A, with the downstream g e n t simply the first gcnc in an operon devoted to BChl synthesis. If this ~,ere correct, the argument advanced by Armstrong et ill. [4] for a superoperon would disappear. To investigate this issue further, we construetcd a strain similar to strain SB226 of Giullano ct al. [0], in which an .Q interposon cartridgc, flanked by transcript ion and translation termination signals, wits placed into the E c o R V site (bp 1303 of Armstrong ct al. [4], see Fig. 1). Giuliano ct al. [6] noted the B c h D - phenotype of SB226 but did not specify the carotenoid status of SB226. The BChl and carotenoid content of the new construct, DE346, was examined.
216
, ~.,-.,, ~ ,,t ..,,., i ,,r~l,, , nT .,,.=,..,.,, ......,..,.,
o
t)
1-
~
, , I r-,...,.. ,,... ,,,,,,.,t i it, ,.,, ,~[~,J ,,,,~..,.| i|.. , , . ,Jl
~
-
, •
-~
I
i
0
ni . l.
.
.
.
.
.
.
.
.
.
.
.
.
i t--'l i
1,000
1.......... ,, .
.....
i
500
.
.
.
.
.
.
.
.
........
.
.
.
"T-,.
1,500
.
.
.
.
.
.
.
.
.
.
.
.
I
.
I.
. . . . h. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
. . .
~
-.
o ..........
F 0
! .......
'
I~;I ii~i
•
i
I.H~II
i i~Ii i I..............
i 500
l
............
i 1.000 Base
Pairs
i I..................
Ull III~U ..............
i 1.500
•
•
'
217 Fluorimetric spectra of strains DE346 and MBI008 a p p e a r e d identical, with a major emission peak at 634 nm and a minor emission peak at 547 rim, characteristic of protoporphyrin IX. Thus. strain DE346 does indeed exhibit a BchD phenotype. Strains SBI003, SB3, MBI008, and DE346 were analysed for the carotenoid content of aerobic cultures (data not shown). SB3 revealed the characteristic C r t A - pigment of spheroidene, while all others revealed a wild-type pattern of carotenoids, distinguished by the predominance of spheroidenone. Thus, although both MBI008 and DE346 are indeed B C h l - , both also exhibit wild-type carotenoids. This demonstrates that the E c o R V site interrupted in DE346 does not lic in the crtA gene, in contrast to the interpretation of Armstrong et al. [4]. We sequenced the suspect area and located one error. A r m s t r o n g et al. [4] reported a stretch of four G nucleotides at bp 1736-1733 ( 5 ' - 3 ' ) , whereas we clearly saw only three G nucleotides. This correction has recently b e e n submitted to G e n B a n k by the Hearst laboratory, accession n u m b e r Z ! 1165, as part of major sequence analysis of the entire photosynthesis region. A codon preference plot of the corrected sequence clearly reveals two distinct open reading frames (Fig. 2B). T h e first open reading frame contains the sites which upon insertional mutagenesis yielded the C r t A - phenotypes for A r m s t r o n g et al. [4,12] and Giuliano et al. [6], whereas the second opcn reading frame contains the sites which yielded a distinct B C h l - phenotype in prior investigations, all involving polar mutations [4,6,12,13]. Thus, we conclude that the first open reading frame is indeed the crtA gene, while the second open reading frame might be a gene involved in an early step of BChl biosynthesis. Taylor et al. [5]
showed that two point mutations in this region of the chromosome map to separate complementation groups but give the same 'BchD" phenotypes. Zscbo and Hcarst [ 13] defined two separate genes. hchl and hchD. in this region. To convincingly show that In'hi does indeed function early in the BChl biosynthetic pathway, construction of a n o n - I ~ l a r mutation in this gene was neccssary. This is because In'hi is upstream from hchD and a polar insertion in bchl might give a phcnotype caused by lack of bchD expression. We created in vitro an in-frame, non-polar deletion in the hchl gene by deletion of the scquenccs from Smal-Smal sites shown in Fig. I., linked to a markcr insertion of the _Q cartridge in the adjacent crtA gene. A G T A - m e d i a t e d cross gave rise to a 3 : I ratio of double (crtA and bchl) mutants : single (crtA) mutants upon selection for the .(.~ cartridge with spectinomycin. Analysis of DE302, one of the double mutants, rcvcalcd a non-photosynthetic, C r t A - , BChl isolate with fluorescence emission and excitation spectra attributable to protoporphyrin IX. PCR of wild-typc SBI003 chromosomal D N A with primers DYI and DY2 yielded a 660-bp fragmcnt. PCR of DE346, containing the approximately 2.0-kb .f.~ fragment gave a 2660-bp fragment, and PCR of DE392, containing the deletion of the Sinai fragment, gave a 414-bp fragment, confirming that both constructions were indeed correct. Since both a polar and a non-polar mutation in thc bchl gone exhibit blocks very early in the BChl pathway, we conclude that this is indeed a gone involved in the conversion of protoporphvrin IX to a subsequent BChl precursor. The crtA and bchi genes a p p e a r to be in separate opcrons. The bchl gene is the first gene in an
Fig. 2. Codon preference plots generated using a pn~gram written by Gribskov el al. [16]. The program identifies efficiently translated genes ;is peaks above open reading frames (open boxes below peaks) that contain few rare c~nlons(vertical dashes below the open reading frames). This is repeated for each of the three possible open reading frames. A c~}doh frequency table used to produce the plot was obtained as previously published Ill. The horizontal dashed lines indicate the b;~ckgroundcodon preference statistic, which is generated by random sequence of the szlme base composition as this DNA sample. The windrowsize and rare codon threshold were set at 25 and I).l. respectively.A. Codon preference plot using the DNA sequence published by Armstrong et al. [4] starting at the Nn~l site (bp 2471) through bp 617. I bp p~st the stop codon attributed by Armstrong et al. [4] to the crtA gene. B. Codon preference plot of the above sequence corrected hy the elimination of one (i nucleotide between bp 1736 and 1733.
218 o p e r o n presumably including hchD as the second; t h e r e may bc additional d o w n s t r e a m g e n e s in this o p e r o n . Since the s t o p c o d o n o f the crtA g c n e (a monocistronic o p e r o n ) overlaps the start c o d o n o f the d o w n s t r e a m hchl g e n e ( b c k m g i n g to the hchlD o p e r o n ) , it a p p e a r s that this region o f D N A is also an example o f s u p c r o p c r o n a l organization o f g e n c s in IClunh~bactercapstdattts. C o d o n p r e f e r e n c e plot analysis selects A T G (bp 1671)1668) as the start m e t h i o n i n c o f the hchl gone r a t h e r than A T G (bp 1577-1575) which is 311 a m i n o acids d o w n s t r e a m . A d d i t i o n a l evidence to s u p p o r t A T G (bp 16711-1668) as the start is s e c n in thc strong a m i n o acid homology b c t w c e n the R. capsuhmls bchl g c n e p r o d u c t and the A r a bulopsis thaliana chloroplast protein cs p r e c u r s o r [14]. C o m p a r i s o n o f these two a m i n o acid seq u e n c e s reveals a striking 49.3C~ identity in a 341-amino acid overlap. T h e strong homology b e t w e e n t h e s e two p r o t e i n s starts midway between the above two A T G s . t h e r e b y strongly supporting the proposal that the first A T G (bp 1670-1668) e n c o d e s the first m c t h i o n i n e o f the Bchl protein, it is this A T G that overlaps the stop c o d o n for the crtA gone. T h e r e must be a p r o m o t e r for the bchlD o p e r o n e m b e d d e d in the crtA structural gene. b e c a u s e strains containing an .f.~ cartridge insertion at the Sphl site in crtA retain a high level o f Bchl synthetic activity. Transcripts initiating at the crtA p r o m o t e r may transcribe the d o w n s t r e a m hchiD o p e r o n , since t h e r e is no intcrcistronic space in which t e r m i n a tion s e q u e n c e s could be located. A t e r m i n a t i o n s e q u e n c e could hypothetically bc e m b e d d e d in
the bchl g e n e if an anti-termimttion system existed that was able to distinguish b e t w e e n transcription initiated at the hch p r o m o t c r c o m p a r e d to the crt p r o m o t e r . Evidence fiw a n t i - t e r m i n a lion in R. ¢'LqZ~'llhltllS r e m a i n s to bc ,liscewcrcd.
REFERENCES [I) Young, D.A.. Bauer. ('.F,.. Williams. J.('. and Marts, B.L 11989) Mol. Gcn. Gcncl. 218. 1-12.
[2] Wellington. ('.1.. and Bcally. T. (1091) J. Bactcriol. 173. 1432-1443. [3] Batncz'.C.E.. Buggy. LJ.. Yang. Z. and Marrs. Fkl,. ( lOOl) Mol. Gcn. Gcnct. 228. 433-444. [4] Armstrong. G.A.. AIbcrti. M.. l.cach. F. and IIcarst. J.E. (IL)S9) Mol. Gen. Gcnct. 211~.254-208. [5] Taylor. D.P.. Cohen, S.N.. Clark. W.G. and Marrs, 13.L.
(11)83) J. Baclcriol. 154. 581}-591). 16] Giuliano. G.. Pollock. D.. Stapp. ]I. and Scolnik. P.A. (1088) Mol. (;ell. Gencl. 213. 78-83. [7] Prcnlki. P. and Krisch. ll.M. 11984) Gone 29. 31)3-313. [8] Yen. II.C. and Marrs. B.L 11076) J. Baclcriol. 121~. 610-629. [9] Biel. A.J. and Marrs. 13.L. (lOS3)J. Bactcriol. 156. fi86694. [lf)] Scolnik. P.A.. Walker. M.A. and Marrs, B.L. 11080)J. Biol. ('hem. 255. 2427-2432. [I I] Sangcr, F.. Nicklens. S. and ('oulson. A.R. 11977) Proc. Nail. Acad. Sci. USA 74. 5463-5467. [ 12] Armstrong. G.A.. Schmidt. A.. Sandmann. G. and Hearst. J.E. (1901})J. Biol. Chem. 265. 8329-8338. [13] Zscbt~. K.M. and Hearst. J.E. 11984) (.'ell 37. 937-947. [141 Koncz. C.. Maycrhofcr. R.. Koncz-Kalman. Z.. Nawralh, C.. Rciss. B., Rcadi. G.P. and Schcll, J. (It~91)) EMBO J. 9. 1337. [15] Marrs. 13. (I081)J. Bactcriol. 146. 11)03--1(}12. [Ib] Gribskov. M.. Dcvcrcux. J. and Burgess. R.R. 11984) Nucleic Acids Rcs. 12. 539-547.
