Short sequence-paper

Cloning and characterization of the cDNA for a plastid c factor1 fromthe moss Physcomitrella patens

Keishi Hara, Mamoru Sugita, Setsuyuki Aoki *Division of Biological Informatics, Graduate School of Human Informatics, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan

Received 15 August 2000; received in revised form 2 October 2000; accepted 4 October 2000

Abstract

We isolated a cDNA PpSig1 encoding a plastid c factor from the moss Physcomitrella patens. The PpSIG1 protein is composed of theconserved subdomains for recognition of 310 and 335 promoter elements, core complex binding and DNA melting. Southern blot analysisshowed that the moss sig1 gene is likely a member of a small gene family. Transient expression assay using green fluorescent proteindemonstrated that the N-terminal region of PpSIG1 functions as a chloroplast-targeting signal peptide. These observations suggest thatmultiple nuclear-encoded c factors regulate chloroplast gene expression in P. patens. ß 2001 Elsevier Science B.V. All rights reserved.

Keywords: c factor; Transcription; Plastid; Physcomitrella patens

Plastid gene expression is controlled by various environ-mental and endogenous factors, e.g., light conditions, de-velopmental state of plastid and circadian clock [1]. Plas-tid-speci¢c RNA polymerases and their cofactors arelikely involved in the regulation by these factors [1], butthe detailed mechanisms of such regulation are mainlyunknown. In higher plants, plastid genes are transcribedby at least two types of RNA polymerases (RNAP), abacteriophage-type single subunit RNAP and a bacteria-type multisubunit RNAP [2]. The latter consists of a pu-tative core catalytic complex (KLLPLQ) and a c factor. Byanalogy with the functions of the c factors in bacteria [3],the plastid c factor is thought to confer promoter specif-icity on the core complex. Recently, genes encoding plastidc factors have been identi¢ed in red algae [4^6] and higherplants [7^13]. These plastid c factor genes appear to bemembers of a gene family in the nuclear genome of eachspecies [6^8,11,13], and show various types of regulationsuch as tissue- or organ-speci¢c expression [8,9,11,13,14],chloroplast development-dependent expression [14], light-dependent expression [5^10,12] and circadian clock-con-trolled expression [12,15]. Moreover, even under thesame light conditions, members of the c factor gene familywithin the same plant species are di¡erentially regulated

[6,8,11,13]. These facts suggest that the plastid c factorsare speci¢city factors in plastid gene expression possiblyactivating di¡erent sets of genes in response to variousenvironmental and endogenous signals. However, in vivofunctions of plastid c factors have not yet been demon-strated. A high frequency of transformation and homolo-gous recombination have been reported for the moss Phys-comitrella patens [16^19], which enable e¤cient analysis ofgene functions in vivo by targeted gene disruption [20^23].Therefore, the cloning of c factor cDNAs from Physcomi-trella is particularly useful to assess the regulatory role ofthe c factor in plastid gene expression. Based on this idea,we isolated and characterized a cDNA for plastid c factorfrom Physcomitrella.

We used a polymerase chain reaction (PCR) strategy toclone the Physcomitrella plastid c factor cDNAs. Proto-nemata of wild-type P. patens subsp. patens were grownunder continuous illumination (LL; intensity was about 40Wmol m32 s31 from white £uorescent lamps) as the stan-dard culture conditions described by Nishiyama et al. [24].Total cellular RNA was isolated from protonemata grownfor 5 days in LL using the RNeasy Mini Kit (Qiagen), and¢rst-strand cDNA was synthesized using the AMV Re-verse Transcriptase First-strand cDNA Synthesis Kit(Life Sciences). Based on the amino acid sequences con-served among six plant c factors (gene products of Arabi-dopsis thaliana SIG2 (AB004821), Sinapis alba Sig1(Y15899), Sorghum bicolor sig1 (Y14276), Zea mays sig1and sig2 (AF058708; AF058709), Oryza sativa sigA

0167-4781 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 7 - 4 7 8 1 ( 0 0 ) 0 0 2 5 0 - 5

* Corresponding author. Fax: +81-52-789-5376;E-mail : aoki@info.human.nagoya-u.ac.jp

1 DDBJ accession No. AB046872.

BBAEXP 91490 16-1-01 Cyaan Magenta Geel Zwart

Biochimica et Biophysica Acta 1517 (2001) 302^306

www.elsevier.com/locate/bba

(AB005290)) clustered in the phylogenetic tree [11], severalsets of primers were designed. Preferential codon usage forPhyscomitrella was applied [25,26]. Using two primers, asense primer SigUp1 and an antisense primer SigDn2, weampli¢ed a DNA fragment with expected size of 172 bpfrom the moss cDNA. SigUp1 was based on the peptidesequences NVRLV(I/M)SIA with the nucleotide sequence5P-CGTGCGGTTGGTIATI(A/T)(G/C)IAT(A/C/T)GC-3P(`I' represents an inosine). SigDn2 was based on the pep-tide sequence WWI(K/R)QGVS, which overlapped with ahighly conserved region, `rpoD box' [27], with the nucleo-tide sequence 5P-GGACACTCCCTGI(T/C)(T/G)(T/G/A)ATCCACCA-3P. The 172 bp band was cloned into apGEM-T easy vector (Promega), and the nucleotide se-quence was determined. The deduced amino acid sequencefrom this DNA fragment showed signi¢cant identity with

the c factors from cyanobacteria, red algae and plants,and thus the putative corresponding gene was namedsig1. A 3P-terminal portion of 0.8 kb in length was isolatedby the 3P-rapid ampli¢cation of cDNA ends (RACE)method using protonemal ¢rst-strand cDNA as template.This 0.8 kb PCR product was used as a probe to screen aPhyscomitrella cDNA library (gift of A. Cuming and S.Bashiardes; 8U105 pfu/Wl). From a screen of 2.0U105

plaques, 13 cDNA clones were isolated, and partial se-quence analysis indicated that all the clones were derivedfrom the same gene. The cDNA insert of the longest clonewas entirely sequenced. Since the 5P portion was still trun-cated as judged from the sequence, it was isolated by 5P-RACE using the 5P-RACE System for Rapid Ampli¢ca-tion of cDNA Ends Version 2.0 (Gibco BRL).

The full-length cDNA (named PpSig1) was 2207 bp in

Fig. 1. Alignment of the c factor gene products from various species. The alignment of C-terminal regions including subdomains 1.2^4.2 is shown.PpSIG1 represents the amino acid sequence studied in this paper. The other sequences are as follows: 7942rpoD1, gene product of rpoD1 of Synecho-coccus sp. PCC 7942 (D10973); EcoliRpoD, gene product of rpoD of E. coli (J01687); AtSigB, gene product of sigB of A. thaliana (AB004293). Theconserved subdomains are indicated above the sequences. Identical amino acids are indicated by asterisks. Dashes represent gaps introduced to maxi-mize homology. Multiple alignment was performed with the program ClustalW [28].

Fig. 2. Southern and Northern blot analyses of the PpSig1. Physcomitrella genomic DNA was digested with BamHI, HindIII, SmaI and XbaI, and hy-bridized with a PpSig1-speci¢c probe under high stringency (a) or low stringency conditions (b). The positions of DNA markers are indicated. 0.5 Wgof poly(A)�RNA from Physcomitrella protonema grown in the light (LL) was subjected to Northern blot analysis with the PpSig1-speci¢c probe (c).The positions of RNA markers are indicated.

BBAEXP 91490 16-1-01 Cyaan Magenta Geel Zwart

K. Hara et al. / Biochimica et Biophysica Acta 1517 (2001) 302^306 303

length, and encoded a putative protein of 579 amino acids(DDBJ accession No. AB046872) with an estimated mo-lecular mass of 64 kDa. A database search using theBLAST program revealed that the protein encoded bythe PpSig1 cDNA (PpSIG1 protein) had the highest ho-mology score with the c factor encoded by the rpoD1 geneof the cyanobacterium Synechococcus sp. PCC 7942.Alignment was carried out using the ClustalW program[28], and the overall amino acid identity was 35.8% be-tween the moss and Synechococcus sequences. Amonghigher plant plastid c factors, SigB protein from A. thali-ana was most similar with PpSIG1 with an overall aminoacid identity of 36.1%.

Fig. 1 shows the alignment of the amino acid sequenceof PpSIG1 protein with those of the c factors from Syn-echococcus sp. PCC 7942 (rpoD1 gene product, D10973),Escherichia coli (rpoD gene product, J01687) and A. thali-ana (sigB gene product, AB004293). Based on this align-ment, PpSIG1 clearly exhibits the subdomains supposed tobe involved in the recognition of 310 and 335 promoter

elements (subdomains 2.4 and 4.2), in core complex bind-ing (subdomain 2.1) and in DNA melting (subdomain 2.3)and a subdomain of presently unclear function (subdo-main 1.2). The sequence similarity supports the idea thatPpSIG1 is a c factor, most likely a subunit of the bacteria-type plastid RNAP.

To examine the copy number of the Physcomitrella siggene, Southern blot analyses were performed on genomicDNA extracted from protonemata digested with restric-tion enzymes using a PpSig1 cDNA speci¢c probe(1573^2380; the ¢rst nucleotide of the PpSig1 cDNA isnumbered +1) (Fig. 2a,b). Probe labeling, hybridization,washing and signal detection were carried out using theAlkPhos DIRECT Gene images kit (Amersham Pharma-cia Biotech). The primary washes of the membranes werecarried out at 50³C and at 55³C for low and high strin-gency conditions, respectively. Under high stringency con-ditions, two or three bands were detected for each lane(Fig. 2a). When the same blot was probed under lowstringency conditions, three or four additional bands of

Fig. 3. Intracellular localization of the PpSIG1 transit peptide-GFP fusion protein. Physcomitrella protoplast cells transformed with the PpSIG1 N-ter-minal region (amino acids 1^83)-sGFP(S65T) fusion (a^c) or with the sGFP(S65T) alone (d^f) are shown. The DNA fragment encoding the ¢rst 83 ami-no acid residues of PpSIG1 was ampli¢ed from the PpSig1 cDNA with a set of primers SigTpUp1 (5P-GGATCCACGATGGCTGCTGTT-TCGTCGGCGTG-3P ; attached restriction site is underlined) and SigTpDn1 (5P-CCATGGCCAGCGCTCGACAATGCACAGT-3P). The ampli¢ed 256bp fragment encoding a putative transit peptide was cloned in pGEM-T easy vector (Promega). From the resulting plasmid, an approx. 260 bp BamHI-EcoRI fragment carrying the putative transit peptide region was excised and introduced into BamHI-EcoRI-cleaved pE7133-GUS (gift of Y. Ohashi[31]) to produce p7133-SigTP. From p7133-SigTP, a 2.5 kb HindIII-NcoI fragment carrying a fusion of the E76In promoter [31] and the transit peptideregion was excised and introduced into HindIII-NcoI-cleaved CaMV35S-sGFP(S65T)-nos3P (gift of Y. Niwa) to yield a GFP reporter plasmid p7133TP-sGFP. To construct a control plasmid p7133sGFP, a 2.2 kb HindIII-BamHI fragment of pE7133-GUS carrying the E76In promoter was introducedinto HindIII-BamHI-cleaved CaMV35S-sGFP(S65T)-nos3P. These reporter constructs were introduced into the moss protoplasts by polyethylene glycol-mediated transformation [17,24]. Two days after transformation, protoplasts were observed by confocal laser scanning microscopy (LSM-GB200, Olym-pus). Images were taken at 530 nm to detect £uorescence of sGFP(S65T) (b,e) and at over 665 nm to detect chlorophyll £uorescence (a,d) after illumi-nation at 488 nm for excitation. Superimposed images are in c and f. Incomplete overlap between the chlorophyll and GFP signals (c) was caused byrotation of the cell during the observation. Microscopy images were processed using Adobe Photoshop Ver. 5.0.

BBAEXP 91490 16-1-01 Cyaan Magenta Geel Zwart

K. Hara et al. / Biochimica et Biophysica Acta 1517 (2001) 302^306304

various intensities appeared in each lane (Fig. 2b). Theseresults indicate that sig1 is a member of a small genefamily in the moss genome.

To investigate the steady-state level of the PpSig1mRNA in protonemata, RNA gel blot analysis was per-formed. 0.5 Wg of poly(A)�RNA was hybridized with thePpSig1 cDNA probe. Probe labeling, hybridization, wash-ing and signal detection were carried out as above. A tran-script with a size of approx. 2.2 kb was detected (Fig. 2c),which is in very good agreement with the 2207-bp PpSig1cDNA.

The N-terminal region of PpSIG1 protein does not ex-hibit enriched Ser or Thr residues, which is a property ofmany plastid-targeting signals (transit peptides). On theother hand, a program for protein sorting TargetP (ver-sion 1.01, http://www.cbs.dtu.dk/services/TargetP/ [29])predicted that the N-terminal region functions as a transitpeptide with a signi¢cant reliability (reliability class 3 [29]).To examine whether the N-terminal region of PpSIG1 (1^83 amino acids) acts as a transit peptide, it was fused tothe green £uorescent protein (sGFP(S65T), a syntheticGFP [30]) reporter (Fig. 3). This fusion was transientlyexpressed in the moss protoplasts under the control ofthe E76In promoter, which enables high level expressionin plants [31]. The protoplasts were illuminated at 488 nmand emissions were measured at 530 nm to detectsGFP(S65T) and at over 665 nm to detect chlorophyll.Fluorescence of the PpSIG1 putative transit peptide-sGFP(S65T) fusion (Fig. 3b) was clearly overlapped withchlorophyll £uorescence (Fig. 3a,c). In the control experi-ment, the sGFP(S65T) alone was expressed and detectedin the nucleus, and faintly near the cell surface (Fig. 3e),but it did not overlap with the chlorophyll £uorescence(Fig. 3d,f). These results indicated that the N-terminalregion of PpSIG1 can function as a transit peptide respon-sible for chloroplast targeting.

This is the ¢rst report of the identi¢cation and charac-terization of a c factor-encoding cDNA from a nonvascu-lar plant. Eukaryotic c factor cDNA and genes have sofar been isolated from red algae, and monocotyledonousand dicotyledonous vascular plants. Thus, the nuclear-en-coded c factors seem to be widespread among distantlyrelated photosynthetic eukaryotes. This is consistent withthe widely accepted endosymbiotic theory of chloroplastevolution and the bacterial origin of nuclear-encoded pro-tein genes including the c factors.

The alignment of c factors showed that PpSIG1 pos-sesses all highly conserved subdomains (2.1^4.2) proposedto be functionally important domains in bacterial c factors(Fig. 1) [3]. Highest conservation was observed in domains2 and 4 as indicated in Fig. 1. These domains are impor-tant for promoter recognition and transcription initiationin eubacteria [3]. Interestingly, the region spanning subdo-mains 2.1^4.2 of PpSIG1 showed very high amino acididentity (67.3%) with the product of the rpoD1 genefrom the cyanobacterium Synechococcus sp. PCC 7942.

The most highly conserved subdomains between the twospecies are 2.4 and 4.2 with amino acid identities of 86.4%and 88.5%, respectively. Since subdomains 2.4 and 4.2 ofbacterial c factors are involved in the recognition of the310 and 335 conserved promoter elements, respectively[3], which are also conserved in most plastid photosyn-thetic gene promoters, it is very likely that PpSIG1 isalso a factor involved in promoter speci¢city during plas-tid gene transcription. Furthermore, the N-terminal 83amino acids of PpSIG1 can function as a chloroplast-tar-geting signal in the GFP transient expression assay (Fig. 3)supporting the notion that PpSIG1 regulates gene expres-sion in plastids. Thus, we suppose that PpSIG1 is im-ported into plastids to regulate gene expression, as pro-posed for other nuclear-encoded c factors.

E¤cient transformation and homologous recombinationof Physcomitrella will provide an e¡ective tool for thefurther analysis of in vivo functions for plastid c factors.

We thank K. Fujiwara and M. Hasebe for kindly sup-plying the Physcomitrella strains and the protocols forculture and transformation, Y. Hiwatashi, K. Fujiwara,M. Umeda, Y. Kobayashi and M. Hasebe for valuabletechnical advise over the period of this study. We alsothank A.C. Cuming, and S. Bashiardes for the Physcomi-trella cDNA library as part of the Physcomitrella ESTProgramme (PEP) at the University of Leeds (UK) andWashington University in St. Louis (USA), Y. Ohashi forpE7133-GUS, Y. Niwa for CaMV35S-sGFP(S65T)-nos3P,the Center for Gene Research (Nagoya University) for theconfocal scanning microscopy, T. Tezuka for the micro-homogenizer, and Kyowa Hakko Kogyo Co., Ltd. forDriserase. We are grateful to B. Piechulla for critical read-ing of the manuscript. This work was supported by grantsfrom the Japanese Society for the Promotion of Science(12740435 to SA), the Japanese Ministry of Education,Science, Sports and Culture (012025212 to SA; 12874107to MS) and the Sumitomo Foundation to SA.

References

[1] M. Goldschmidt-Clermont, Rev. Cytol. 177 (1998) 115^180.[2] M.W. Gray, B.F. Lang, Trends Microbiol. 6 (1998) 1^3.[3] J.D. Helmann, M.J. Chamberlin, Annu. Rev. Biochem. 57 (1988)

839^872.[4] K. Tanaka, K. Oikawa, N. Ohta, H. Kuroiwa, T. Kuroiwa, H. Ta-

kahashi, Science 272 (1996) 1932^1935.[5] B. Liu, R.F. Troxler, Proc. Natl. Acad. Sci. USA 93 (1996) 3313^

3318.[6] K. Oikawa, K. Tanaka, H. Takahashi, Gene 210 (1998) 277^285.[7] K. Tanaka, Y. Tozawa, N. Mochizuki, K. Shinozaki, A. Nagatani,

K. Wakasa, H. Takahashi, FEBS Lett. 413 (1997) 309^313.[8] K. Isono, M. Shimizu, K. Yoshimoto, Y. Niwa, K. Satoh, A. Yoko-

ta, H. Kobayashi, Proc. Natl. Acad. Sci. USA 94 (1997) 14948^14953.

[9] Y. Tozawa, K. Tanaka, H. Takahashi, K. Wakasa, Nucleic AcidsRes. 26 (1998) 415^419.

[10] M. Kestermann, S. Neukirchen, K. Kloppstech, G. Link, NucleicAcids Res. 26 (1998) 2747^2753.

BBAEXP 91490 16-1-01 Cyaan Magenta Geel Zwart

K. Hara et al. / Biochimica et Biophysica Acta 1517 (2001) 302^306 305

[11] S. Tan, R.F. Troxler, Proc. Natl. Acad. Sci. USA 96 (1999) 5316^5321.

[12] K. Morikawa, S. Ito, Y. Tsunoyama, Y. Nakahira, T. Shiina, Y.Toyoshima, FEBS Lett. 451 (1999) 275^278.

[13] S.D. Lahiri, J. Yao, C. McCumbers, L.A. Allison, Mol. Cell Biol.Res. Commun. 1 (1999) 14^20.

[14] S.D. Lahiri, L.A. Allison, Plant Physiol. 123 (2000) 883^894.[15] K. Kanamaru, M. Fujiwara, M. Seki, T. Katagiri, M. Nakamura, N.

Mochizuki, A. Nagatani, K. Shinozaki, K. Tanaka, H. Takahashi,Plant Cell Physiol. 40 (1999) 832^842.

[16] D. Schaefer, J.P. Zryd, C.D. Knight, D.J. Cove, Mol. Gen. Genet.226 (1991) 418^424.

[17] D. Schaefer, PhD Thesis, University of Lausanne, 1994 (http://www.unil.ch/lpc/docs/DSThesis.htm).

[18] W. Kammerer, D.J. Cove, Mol. Gen. Genet. 250 (1996) 380^382.[19] D.G. Schaefer, J.P. Zryd, Plant J. 11 (1997) 1195^1206.[20] R. Strepp, S. Scholz, S. Kruse, V. Speth, R. Reski, Proc. Natl. Acad.

Sci. USA 95 (1998) 4368^4373.[21] T. Girke, H. Schmidt, U. Zahringer, R. Reski, E. Heinz, Plant J. 15

(1998) 39^48.[22] A.H. Hofmann, A.C. Codon, C. Ivascu, V.E. Russo, C. Knight, D.

Cove, D.G. Schaefer, M. Chakhparonian, J.P. Zryd, Mol. Gen. Gen-et. 261 (1999) 92^99.

[23] P.A. Girod, H. Fu, J.P. Zryd, R.D. Vierstra, Plant Cell 11 (1999)1457^1472.

[24] T. Nishiyama, Y. Hiwatashi, I. Sakakibara, M. Kato, M. Hasebe,DNA Res. 7 (2000) 9^17.

[25] R. Reski, S. Reynolds, M. Wehe, T. Kleber-Janke, S. Kruse, Bot.Acta 111 (1998) 145^151.

[26] J. Machuka, S. Bashiardes, E. Ruben, K. Spooner, A. Cuming, C.Knight, D. Cove, Plant Cell Physiol. 40 (1999) 378^387.

[27] K. Tanaka, T. Shiina, H. Takahashi, Science 242 (1988) 1040^1042.[28] D.G. Higgins, P.M. Sharp, Gene 73 (1988) 237^244.[29] O. Emanuelsson, H. Nielsen, S. Brunak, G. von Heijne, J. Mol. Biol.

300 (2000) 1005^1016.[30] W.-I. Chiu, Y. Niwa, W. Zeng, T. Hirano, H. Kobayashi, J. Sheen,

Curr. Biol. 6 (1996) 325^330.[31] I. Mitsuhara, M. Ugaki, H. Hirochika, M. Ohshima, T. Murakami,

Y. Gotoh, Y. Katayose, S. Nakamura, R. Honkura, S. Nishimiya, K.Ueno, A. Mochizuki, H. Tanimoto, H. Tsugawa, Y. Otsuki, Y. Oha-shi, Plant Cell Physiol. 37 (1996) 49^59.

BBAEXP 91490 16-1-01 Cyaan Magenta Geel Zwart

K. Hara et al. / Biochimica et Biophysica Acta 1517 (2001) 302^306306