Divergent Nod-Containing Bradyrhizobium Sp. DOA9 with a Megaplasmid and Its Host Range
Divergent Nod-Containing Bradyrhizobium sp. DOA9 with a Megaplasmid and its Host Range
KamonlucK TeamTisong1, PongPan songwaTTana2, RujiReK noisangiam2, PongdeT PiRomyou2, nanTaKoRn BoonKeRd2, Panlada TiTTaBuTR2, Kiwamu minamisawa3, achaRa nanTagij4, shin oKazaKi5, miKiKo aBe6, ToshiKi uchiumi6*, and neung TeaumRoong2** 1&HQWHUIRU6FLHQWL¿FDQG7HFKQRORJLFDO(TXLSPHQW6XUDQDUHH8QLYHUVLW\RI7HFKQRORJ\0XDQJ1DNKRQ5DWFKDVLPD7KDLODQG 26FKRRORI%LRWHFKQRORJ\6XUDQDUHH8QLYHUVLW\RI7HFKQRORJ\0XDQJ1DNKRQ5DWFKDVLPD7KDLODQG 3*UDGXDWH6FKRRORI/LIH 6FLHQFH7RKRNX8QLYHUVLW\6HQGDL-DSDQ 46RLO0LFURELRORJ\*URXS'LYLVLRQRI6RLO6FLHQFH'HSDUWPHQWRI$JULFXOWXUH %DQJNRN7KDLODQG 5*UDGXDWH6FKRRORI$JULFXOWXUH7RN\R8QLYHUVLW\RI$JULFXOWXUHDQG7HFKQRORJ\7RN\R-DSDQ DQG 6*UDGXDWH6FKRRORI6FLHQFHDQG(QJLQHHULQJ.DJRVKLPD8QLYHUVLW\.DJRVKLPD-DSDQ (Received May 13, 2014—Accepted August 15, 2014—Published online October 4, 2014)
%UDG\UKL]RELXP sp. DOA9, a non-photosynthetic bacterial strain originally isolated from the root nodules of the legume $HVFK\QRPHQHDPHULFDQD, is a divergent QRGFRQWDLQLQJVWUDLQ,WH[KLELWVDEURDGKRVWUDQJHEHLQJDEOHWRFRORQL]HDQGHI¿- ciently nodulate the roots of most plants from the Dalbergioid, Millettioid, and Robinioid tribes (7 species of Papilionoideae). In all cases, nodulation was determinate. The morphology and size of DOA9 bacteroids isolated from the nodules of various species of Papilionoideae were indistinguishable from the free-living form. However, they were spherical in $UDFKLVK\SRJDHD nodules. GusA-tagged DOA9 also colonized rice roots as endophytes. Since broad-host-range legume symbionts often carry multiple replicons in their genome, we analyzed the replicons for symbiosis genes by electrophoresis. DOA9 carried two replicons, a chromosome (cDOA9) and single megaplasmid (pDOA9) larger than 352 kb. The genes for nodulation (QRG$, %, C DQGQLWURJHQ¿[DWLRQ nifH) were localized on the megaplasmid. Southern blot hybridization revealed two copies of QRG$ on the megaplasmid, single copies of QRG% and C on the megaplasmid, and one copy each of nifH on the chromosome and megaplasmid. These results suggested that %UDG\UKL]RELXP sp. DOA9 may have the unusual combination of a broad host range, bacteroid differentiation, and symbiosis-mediating replicons.
Key words: %UDG\UKL]RELXP, $HVFK\QRPHQHDPHULFDQD, broad host range, megaplasmid
The genus $HVFK\QRPHQH belongs to the Dalbergioid clade of the 16S rRNA gene and housekeeping genes (GQD., UHF$, of the subfamily Papilionoideae of the family Fabaceae (19). and JOQ%). DOA9 was called a divergent QRG-containing $HVFK\QRPHQH species establish a symbiotic relationship strain based on Southern blot hybridization with the QRG gene with bacteria of the genus %UDG\UKL]RELXP (19, 27). They (QRG$, %, and C) probes. In addition, on the basis of nifH KDYHEHHQFODVVL¿HGLQWRRQHRIWKUHHFURVVLQRFXODWLRQ &, sequence similarities, DOA9 was placed in a cluster of JURXSV &, JURXS VSHFLHV VXFK DV $ DPHULFDQD and $ QRQSKRWRV\QWKHWLFEUDG\UKL]RELDOVWUDLQVWKDWDUHDEOHWR¿[ HODSKUR[\ORQ, are only nodulated on their roots by non- nitrogen in the free-living form. Moreover, it can nodulate a SKRWRV\QWKHWLFEUDG\UKL]RELD&,JURXSVSHFLHVVXFKDV$ wide range of legumes, including 0DFURSWLOLXPDWURSXUSXUHXP, DIUDVSHUD and $ QLORWLFD, are nodulated on their roots and $UDFKLVK\SRJDHD, 9LJQDUDGLDWD, and $HVFK\QRPHQHDIUDVSHUD stems by both non-photosynthetic and photosynthetic (29). Since information regarding this strain is limited, we EUDG\UKL]RELD&,JURXSVSHFLHVVXFKDV$LQGLFD and $ investigated the root colonization, infection, and nodulation VHQVLWLYD, are nodulated on their roots and stems by photosyn- HI¿FLHQF\ RI '2$ LQ VHYHUDO VSHFLHV IURP WKH OHJXPH WKHWLF EUDG\UKL]RELD &, JURXS VSHFLHV DUH QRGXODWHG E\ subfamilies Papilionoideae and Mimosoideae, as well as rice. strains BTAi1 and ORS278 in a Nod factor–independent We also examined bacteroids and nodule development. manner (27). However, $LQGLFD &,JURXS ZDVUHFHQWO\ Previous studies demonstrated that symbionts with a broad found to be nodulated by a non-photosynthetic strain isolated host range generally carried multiple replicons in their from $DPHULFDQD &,JURXS 7KLVVWUDLQFRXOGQRGXODWH genome (12, 36, 42); therefore, we examined the replicon VHYHUDO&,JURXSVRI$HVFK\QRPHQH, as well as the peanut and structure, localization, and copy number of symbiosis genes. mung bean (29). %UDG\UKL]RELXP sp. DOA9 (Alphaproteobacteria) was Materials and Methods isolated from the root nodules of $DPHULFDQD in Thailand. This strain was assigned to % \XDQPLQJHQVH based on its %DFWHULDOVWUDLQVDQGJURZWKFRQGLWLRQV phenotypic characteristics and multilocus sequences analysis %UDG\UKL]RELXP sp. DOA9, % GLD]RHI¿FLHQV USDA110, and 0HVRUKL]RELXP ORWL 0$)) ZHUH FXOWXUHG DW & LQ +0 medium supplemented with l-arabinose (8). GUS-tagged DOA9 was &RUUHVSRQGLQJDXWKRU(PDLO XWWDQ#VFLNDJRVKLPDXDFMS cultured in HM supplemented with streptomycin (200 Pg mL1) (29). Tel: 81–99–285–8164; Fax: 81–99–285–8163. &RUUHVSRQGLQJDXWKRU(PDLO QHXQJ#VXWDFWK 3ODQWJURZWKDQGLQRFXODWLRQ Tel: 66–44–223350; Fax: 66–44–216345. Peanut ($UDFKLVK\SRJDHD), mung bean (9LJQDUDGLDWD), soybean
Reproduced from Microbes Environ. 29: 370-376 (2014). Kamonluck Teamtisong: Participant of the 30th UM, 2002-2003.
291 291 (*O\FLQH PD[ cv. SJ5), siratro (0DFURSWLOLXP DWURSXUSXUHXP), &HOOSHOOHWVZHUHKDUYHVWHGE\FHQWULIXJDWLRQDWîJ for 10 min. 6HVEDQLDURVWUDWD, lupin (/XSLQXVSRO\SK\OOXV), 'HVPRGLXP sp., alfalfa 7KHFHOOVZHUHUHVXVSHQGHGLQ ZY 1D&OWR2'600 1. They (0HGLFDJR VDWLYD), 0HGLFDJR WUXQFDWXOD, and $FDFLD PDQJLXP were harvested from 1 mL of a cell suspension and washed with seeds were sterilized as described previously (37). $HVFK\QRPHQH 0 VDOWV FRQWDLQLQJ 0 1D&O DQG WKHQ ZLWK P/ DPHULFDQD (a local Thai variety), $DIUDVSHUD, $HYHQLD (provided by (w/v) Sarcosyl. The supernatant was removed immediately and the (ULF*LUDXG DQG$LQGLFD (ecotype Tottori, Japan; ecotype Tomeshi, sediment was resuspended in 50 PL of lysis buffer (1 mg mLí Japan; and a local Thai variety) were sterilized by incubating in lysozyme, 1 mg mLí RNase A [33], 0.1% (w/v) bromophenol blue concentrated sulfuric acid for 30 min. The seeds of ,QGLJRIHUD in Trisborate buffer (pH 8.2, 89 mM Tris base, 12.5 mM disodium WLQFWRULD, /RWXVMDSRQLFXV, 6W\ORVDQWKHVKDPDWD, &URWDODULDMXQFHD, ('7$DQG P0ERULFDFLG>@DQG>YY@JO\FHURO /HVSHGH]D sp., /HXFDHQDOHXFRFHSKDOD, 0LPRVDSXGLFD, 1HSWXQLD Before sample loading, electrophoresis was performed on a QDWDQV, and 6DPDQHDVDPDQ were sterilized by incubating in con- DJDURVH JHO ZKLFK OHYHOHG RII ZLWK î7%( EXIIHU & centrated sulfuric acid for 10 min. The seeds of rice (2U\]DVDWLYD XQWLO LW WRXFKHG WKH JHO 7KH ZHOOV ZHUH WKHQ ¿OOHG ZLWK PL ssp. LQGLFD) were sterilized as described previously (26). All seeds Sodium Dodecyl Sulfate (SDS; 10% [w/v]) mixed with xylene cyanol were washed with sterilized water and then soaked in sterilized (1 mg mL1). The current was run for 10–15 min at 100 V with water overnight at ambient temperature. All seeds were germinated reversed polarity until SDS was 1 cm above the wells. After that, RQVWHULOL]HG ZY ZDWHUDJDUIRUWRGDW&LQWKHGDUN 50 PL of sample mixed with lysis buffer was directly loaded The germinated seeds were transferred onto Hoagland’s agar (15). and left for 15 min before 15 PL of Proteinase K (5 mg mL1) in %UDG\UKL]RELXP VS '2$ ZDV ZDVKHG ZLWK 1D&O DQG 40% (w/v) glycerol was overlaid. The wells were sealed with melted optical density at 600 nm (OD600 ZDVDGMXVWHGWRZLWKVWHULOL]HG DJDURVHJHODQGî7%(EXIIHUZDVWKHQDGGHGWRFRYHUWKHJHO water, corresponding to approximately 109 cells mL1(DFKVHHG- After 1 h, electrophoresis was carried out in a cold chamber at ling was then inoculated with 100 PL of the bacterial culture. All &7KHFXUUHQWZDVUXQDWP$IRUKDQGDWP$IRUD SODQWVZHUHJURZQDW&XQGHUDKOLJKWKGDUNF\FOHDWDOLJKW further 10 h. The DNA in the gel was stained for 30 min in ethidium intensity of 639 P(P2 s1 for 7 to 14 d (43). These samples were bromide (0.5 Pg mL1) and washed with distilled water before being used for microscopic observations. The symbiotic abilities of DOA9 viewed under UV light. ZHUHGHWHUPLQHGLQ/HRQDUG¶VMDUVFRQWDLQLQJVWHULOL]HGYHUPLFXOLWH and inoculated with 1 mL of bacterial culture, as described above. 6RXWKHUQEORWK\EULGL]DWLRQ 1IUHH+RDJODQG¶VVROXWLRQZDVDGGHGWRHDFKMDUDVUHTXLUHG3ODQWV The megaplasmid and chromosomal DNAs separated on the were harvested after 35 d, used in an analysis of nitrogenase activity gel were used for the Southern blot hybridization of nodulation by the Acetylene Reduction Assay, and the number of nodules was (QRG DQGQLWURJHQ¿[DWLRQ nifH JHQHVDVGHVFULEHG %ULHÀ\ scored. The dry weights of plants were determined after drying at probes for QRG$ (550 bp), QRG% (530 bp), and QRG& (1 kb) were &IRUK REWDLQHG WKURXJK 3&5 DPSOL¿FDWLRQ XVLQJ WKH JHQRPLF '1$ RI % \XDQPLQJHQVH SUTN6-2, % FDQDULHQVH SUTN7-2, and % 0LFURVFRSLFREVHUYDWLRQRIURRWFRORQL]DWLRQDQGQRGXODWLRQ GLD]RHI¿FLHQV USDA110, respectively. DNA fragments of the Root colonization and nodulation were revealed by GUS staining. UHVSHFWLYHVWUDLQVZHUHDPSOL¿HGZLWKWKHSULPHUSDLUVQRG$<) Samples inoculated with GUS-tagged DOA9 were immersed in QRG$<5 QRG$), nodB26/nodB625 (QRG% DQG QRG& GUS assay solution (40 mL 20 mg mL1 X-Gluc in 11-dimethyl- QRG&, QRG&) (29). The probe for nifH was derived from % formamide, 20 mg SDS, 2 mL methanol, 0.2 mL 1 M sodium phos- \XDQPLQJHQVH using nifHF/nifHI primer pairs (18). DNA probes phate buffer, and 7.76 mL distilled water) in a vacuum for 120 min ZHUHODEHOHGRYHUQLJKWDW&E\UDQGRPSULPLQJDQGK\EULGL]HG DQGOHIWLQWKLVVROXWLRQIRUKDW&*86VWDLQLQJZDVREVHUYHG with the Digoxigenin (DIG) High Prime DNA Labeling and under a light microscope. Detection Starter Kit I (Roche, Switzerland). DNA was capillary- transferred to a Hybond-NQ\ORQPHPEUDQH $PHUVKDP&DUGLII 'LIIHUHQWLDOLQWHUIHUHQFHFRQWUDVWDQGÀXRUHVFHQFHPLFURVFRSH UK) as described previously (34). Low-stringency conditions were REVHUYDWLRQVRILVRODWHGEDFWHURLGV XVHGIRUK\EULGL]DWLRQ PHPEUDQHVZHUHK\EULGL]HGDW& QRG$ Nodules were mashed in bacteroid extraction buffer (125 mM and QRG&JHQHV RU& QRG% and nifH) for 18 h and then washed .&OP06RGLXPVXFFLQDWHP07(6EXIIHUS+ZLWK WZLFHLQî66& ZY 6'6DW&IRUPLQDQGLQî (w/v) BSA) (24). To remove plant cell debris, the suspension was 66& ZY 6'6DW&IRUPLQ centrifuged at 100×JDW&IRUPLQ7RSUHFLSLWDWHWKHEDFWH- DOA9 genomic DNA was digested with (FRRI, (FRRV, +LQGIII, roids, the supernatant was centrifuged at 3,000×JDW&IRUPLQ or 1RWI to evaluate the copy numbers of QRG%, QRG&, and nifH; with The precipitate was observed using Differential Interference (FRRI, +LQGIII, %JOII, or 1RWI for QRG$. Fragments were separated &RQWUDVW ',& PLFURVFRS\ 5HJDUGLQJ ÀXRUHVFHQFH PLFURVFRS\ on 1% (w/v) agarose gels before hybridization as described above. WKH EDFWHURLG IUDFWLRQ ZDV ¿UVW VWDLQHG ZLWK GLDPLGLQR phenylindole (DAPI; 50 Pg mL1) and then with propidium iodide (PI; 2 Pg mL1). Results and Discussion '1$H[WUDFWLRQDQGPHJDSODVPLGGHWHFWLRQ 1RGXODWLRQRIWKHVXEIDPLO\3DSLOLRQRLGHDH %UDG\UKL]RELXP sp. DOA9 total DNA was prepared as described We tested 19 species and 3 varieties of the ecotypes of $ previously (23). Megaplasmids were isolated by electrophoresis LQGLFD (Table 1). DOA9 was able to nodulate the roots of 15 ZLWKWKHPRGL¿FDWLRQVGHVFULEHGKHUH%DFWHULDZHUHFXOWXUHG on HM broth medium (37) with 0.05% (w/v) l-arabinose, 0.05% species. Seven of the tested species were effective in terms of (w/v) yeast extract for % GLD]RHI¿FLHQV USDA110 and 0 ORWL QLWURJHQ¿[LQJ HI¿FLHQF\ DV LQGLFDWHG E\ D VLJQL¿FDQW MAFF303099, and no l-arabinose for %UDG\UKL]RELXP strain increase in the plant dry weight when inoculated with DOA9 DOA9, to reduce the production of polysaccharides. USDA110 and (39) (Table 1 and Table S1 in the supplemental material). An '2$ ZHUHFXOWXUHGDW&IRUWRGRQDURWDU\VKDNHUDW ineffective symbiotic was detected in DOA9 with 6KDPDWD, rpm until the late-log phase was reached. A total of 1% (v/v) of these 0DWURSXUSXUHXP, /HSHGH]D sp., /MDSRQLFXV, /OHXFRFHSKDOD, pre-cultures was then inoculated into new tubes containing HM and 6 VDPDQ. This strain colonized the lateral root (Fig. broth medium. The cultures were incubated for 3 d until the expo- nential growth phase (the mid-log phase) was reached. MAFF303099 $±' DQG DOVR WKH URRW VXUIDFH )LJ ( ,Q DGGLWLRQ was incubated for 2 d, inoculated into a fresh tube containing HM nodulation was related to the colonization sites (Fig. 1F–I). medium, and then cultured under the same conditions for 24–36 h. All nodules were determinate (Fig. 1F–J). A thin section of
292 293 (*O\FLQH PD[ cv. SJ5), siratro (0DFURSWLOLXP DWURSXUSXUHXP), &HOOSHOOHWVZHUHKDUYHVWHGE\FHQWULIXJDWLRQDWîJ for 10 min. Table 1. Nodulation by %UDG\UKL]RELXP sp. DOA9 and bacteroid morphology in various legumes
6HVEDQLDURVWUDWD, lupin (/XSLQXVSRO\SK\OOXV), 'HVPRGLXP sp., alfalfa 7KHFHOOVZHUHUHVXVSHQGHGLQ ZY 1D&OWR2'600 1. They Plants Nodulationa Nodule type Bacteroidc (0HGLFDJR VDWLYD), 0HGLFDJR WUXQFDWXOD, and $FDFLD PDQJLXP were harvested from 1 mL of a cell suspension and washed with seeds were sterilized as described previously (37). $HVFK\QRPHQH 0 VDOWV FRQWDLQLQJ 0 1D&O DQG WKHQ ZLWK P/ Papilionoideae Genistoids DPHULFDQD (a local Thai variety), $DIUDVSHUD, $HYHQLD (provided by (w/v) Sarcosyl. The supernatant was removed immediately and the í /XSLQXVSRO\SK\OOXV () determinate unswollen (ULF*LUDXG DQG$LQGLFD (ecotype Tottori, Japan; ecotype Tomeshi, sediment was resuspended in 50 PL of lysis buffer (1 mg mL &URWDODULDMXQFHD determinate unswollen í Japan; and a local Thai variety) were sterilized by incubating in lysozyme, 1 mg mL RNase A [33], 0.1% (w/v) bromophenol blue Dalbergioids concentrated sulfuric acid for 30 min. The seeds of ,QGLJRIHUD in Trisborate buffer (pH 8.2, 89 mM Tris base, 12.5 mM disodium $HVFK\QRPHQHDPHULFDQD (a local Thai variety) determinate unswollen WLQFWRULD, /RWXVMDSRQLFXV, 6W\ORVDQWKHVKDPDWD, &URWDODULDMXQFHD, ('7$DQG P0ERULFDFLG>@DQG>YY@JO\FHURO $HVFK\QRPHQHLQGLFD (ecotype Tottori, Japan) /HVSHGH]D sp., /HXFDHQDOHXFRFHSKDOD, 0LPRVDSXGLFD, 1HSWXQLD Before sample loading, electrophoresis was performed on a $HVFK\QRPHQHLQGLFD (ecotype Tomeshi, Japan) QDWDQV, and 6DPDQHDVDPDQ were sterilized by incubating in con- DJDURVH JHO ZKLFK OHYHOHG RII ZLWK î7%( EXIIHU & $HVFK\QRPHQHLQGLFD (a local Thai variety) centrated sulfuric acid for 10 min. The seeds of rice (2U\]DVDWLYD XQWLO LW WRXFKHG WKH JHO 7KH ZHOOV ZHUH WKHQ ¿OOHG ZLWK PL $HVFK\QRPHQHDIUDVSHUD determinate unswollen ssp. LQGLFD) were sterilized as described previously (26). All seeds Sodium Dodecyl Sulfate (SDS; 10% [w/v]) mixed with xylene cyanol $HVFK\QRPHQHHYHQLD were washed with sterilized water and then soaked in sterilized (1 mg mL1). The current was run for 10–15 min at 100 V with $UDFKLVK\SRJDHD determinate swollen water overnight at ambient temperature. All seeds were germinated reversed polarity until SDS was 1 cm above the wells. After that, 6W\ORVDQWKHVKDPDWD () determinate swollen RQVWHULOL]HG ZY ZDWHUDJDUIRUWRGDW&LQWKHGDUN 50 PL of sample mixed with lysis buffer was directly loaded Millettioids The germinated seeds were transferred onto Hoagland’s agar (15). and left for 15 min before 15 PL of Proteinase K (5 mg mL1) in *O\FLQHPD[ (cv. SJ5) %UDG\UKL]RELXP VS '2$ ZDV ZDVKHG ZLWK 1D&O DQG 40% (w/v) glycerol was overlaid. The wells were sealed with melted 0DFURSWLOLXPDWURSXUSXUHXP () determinate unswollen 9LJQDUDGLDWD determinate unswollen optical density at 600 nm (OD600 ZDVDGMXVWHGWRZLWKVWHULOL]HG DJDURVHJHODQGî7%(EXIIHUZDVWKHQDGGHGWRFRYHUWKHJHO 9 1 'HVPRGLXP sp. determinate unswollen water, corresponding to approximately 10 cells mL (DFKVHHG- After 1 h, electrophoresis was carried out in a cold chamber at P /HVSHGH]D sp. ( ) determinate unswollen ling was then inoculated with 100 L of the bacterial culture. All &7KHFXUUHQWZDVUXQDWP$IRUKDQGDWP$IRUD ,QGLJRIHUDWLQFWRULD determinate unswollen SODQWVZHUHJURZQDW&XQGHUDKOLJKWKGDUNF\FOHDWDOLJKW further 10 h. The DNA in the gel was stained for 30 min in ethidium 2 1 1 Robinioids intensity of 639 P(P s for 7 to 14 d (43). These samples were bromide (0.5 Pg mL ) and washed with distilled water before being /RWXVMDSRQLFXV (ecotype Miyagi, Japan) () determinate unswollen used for microscopic observations. The symbiotic abilities of DOA9 viewed under UV light. 6HVEDQLDURVWUDWD ZHUHGHWHUPLQHGLQ/HRQDUG¶VMDUVFRQWDLQLQJVWHULOL]HGYHUPLFXOLWH IRLC (inverted repeat–lacking clade) and inoculated with 1 mL of bacterial culture, as described above. 6RXWKHUQEORWK\EULGL]DWLRQ 0HGLFDJRWUXQFDWXOD 1IUHH+RDJODQG¶VVROXWLRQZDVDGGHGWRHDFKMDUDVUHTXLUHG3ODQWV The megaplasmid and chromosomal DNAs separated on the 0HGLFDJRVDWLYD () determinate unswollen were harvested after 35 d, used in an analysis of nitrogenase activity gel were used for the Southern blot hybridization of nodulation 7ULIROLXPUHSHQV (white clover) by the Acetylene Reduction Assay, and the number of nodules was (QRG DQGQLWURJHQ¿[DWLRQ nifH JHQHVDVGHVFULEHG %ULHÀ\ Mimosoideae scored. The dry weights of plants were determined after drying at probes for QRG$ (550 bp), QRG% (530 bp), and QRG& (1 kb) were Mimoseae &IRUK REWDLQHG WKURXJK 3&5 DPSOL¿FDWLRQ XVLQJ WKH JHQRPLF '1$ RI 0LPRVDSXGLFD % \XDQPLQJHQVH SUTN6-2, % FDQDULHQVH SUTN7-2, and % /HXFDHQDOHXFRFHSKDOD () determinate unswollen 0LFURVFRSLFREVHUYDWLRQRIURRWFRORQL]DWLRQDQGQRGXODWLRQ GLD]RHI¿FLHQV USDA110, respectively. DNA fragments of the 1HSWXQLDQDWDQV Root colonization and nodulation were revealed by GUS staining. UHVSHFWLYHVWUDLQVZHUHDPSOL¿HGZLWKWKHSULPHUSDLUVQRG$<) Ingeae 6DPDQHDVDPDQ () determinate unswollen Samples inoculated with GUS-tagged DOA9 were immersed in QRG$<5 QRG$), nodB26/nodB625 (QRG% DQG QRG& GUS assay solution (40 mL 20 mg mL1 X-Gluc in 11-dimethyl- QRG&, QRG&) (29). The probe for nifH was derived from % a , positive with effective nodules; (), positive with ineffective nodules; , negative. formamide, 20 mg SDS, 2 mL methanol, 0.2 mL 1 M sodium phos- \XDQPLQJHQVH using nifHF/nifHI primer pairs (18). DNA probes phate buffer, and 7.76 mL distilled water) in a vacuum for 120 min ZHUHODEHOHGRYHUQLJKWDW&E\UDQGRPSULPLQJDQGK\EULGL]HG DQGOHIWLQWKLVVROXWLRQIRUKDW&*86VWDLQLQJZDVREVHUYHG with the Digoxigenin (DIG) High Prime DNA Labeling and under a light microscope. Detection Starter Kit I (Roche, Switzerland). DNA was capillary- transferred to a Hybond-NQ\ORQPHPEUDQH $PHUVKDP&DUGLII 'LIIHUHQWLDOLQWHUIHUHQFHFRQWUDVWDQGÀXRUHVFHQFHPLFURVFRSH UK) as described previously (34). Low-stringency conditions were REVHUYDWLRQVRILVRODWHGEDFWHURLGV XVHGIRUK\EULGL]DWLRQ PHPEUDQHVZHUHK\EULGL]HGDW& QRG$ Nodules were mashed in bacteroid extraction buffer (125 mM and QRG&JHQHV RU& QRG% and nifH) for 18 h and then washed .&OP06RGLXPVXFFLQDWHP07(6EXIIHUS+ZLWK WZLFHLQî66& ZY 6'6DW&IRUPLQDQGLQî (w/v) BSA) (24). To remove plant cell debris, the suspension was 66& ZY 6'6DW&IRUPLQ centrifuged at 100×JDW&IRUPLQ7RSUHFLSLWDWHWKHEDFWH- DOA9 genomic DNA was digested with (FRRI, (FRRV, +LQGIII, roids, the supernatant was centrifuged at 3,000×JDW&IRUPLQ or 1RWI to evaluate the copy numbers of QRG%, QRG&, and nifH; with The precipitate was observed using Differential Interference (FRRI, +LQGIII, %JOII, or 1RWI for QRG$. Fragments were separated &RQWUDVW ',& PLFURVFRS\ 5HJDUGLQJ ÀXRUHVFHQFH PLFURVFRS\ on 1% (w/v) agarose gels before hybridization as described above. WKH EDFWHURLG IUDFWLRQ ZDV ¿UVW VWDLQHG ZLWK GLDPLGLQR phenylindole (DAPI; 50 Pg mL1) and then with propidium iodide (PI; 2 Pg mL1). Results and Discussion '1$H[WUDFWLRQDQGPHJDSODVPLGGHWHFWLRQ 1RGXODWLRQRIWKHVXEIDPLO\3DSLOLRQRLGHDH %UDG\UKL]RELXP sp. DOA9 total DNA was prepared as described We tested 19 species and 3 varieties of the ecotypes of $ previously (23). Megaplasmids were isolated by electrophoresis LQGLFD (Table 1). DOA9 was able to nodulate the roots of 15 ZLWKWKHPRGL¿FDWLRQVGHVFULEHGKHUH%DFWHULDZHUHFXOWXUHG on HM broth medium (37) with 0.05% (w/v) l-arabinose, 0.05% species. Seven of the tested species were effective in terms of (w/v) yeast extract for % GLD]RHI¿FLHQV USDA110 and 0 ORWL QLWURJHQ¿[LQJ HI¿FLHQF\ DV LQGLFDWHG E\ D VLJQL¿FDQW MAFF303099, and no l-arabinose for %UDG\UKL]RELXP strain increase in the plant dry weight when inoculated with DOA9 Fig. 1. ([DPSOHVRI $±( URRWFRORQL]DWLRQ )±- QRGXOHPRUSKRORJ\DQG .±2 WKLQVHFWLRQVRIQRGXOHVLQRFXODWHGZLWK*86WDJJHG'2$ DOA9, to reduce the production of polysaccharides. USDA110 and (39) (Table 1 and Table S1 in the supplemental material). An in various legumes. (A, F, and K); $HVFK\QRPHQHDPHULFDQD, (B, G, and L); 6W\ORVDQWKHVKDPDWH &+DQG0 ,QGLJRIHUDWLQFWRULD, (D, I, and N) 0DFURSWLOLXPDWURSXUSXUHXPDQG (-DQG2 /RWXVMDSRQLFXV. '2$ ZHUHFXOWXUHGDW&IRUWRGRQDURWDU\VKDNHUDW ineffective symbiotic was detected in DOA9 with 6KDPDWD, rpm until the late-log phase was reached. A total of 1% (v/v) of these 0DWURSXUSXUHXP, /HSHGH]D sp., /MDSRQLFXV, /OHXFRFHSKDOD, pre-cultures was then inoculated into new tubes containing HM and 6 VDPDQ. This strain colonized the lateral root (Fig. broth medium. The cultures were incubated for 3 d until the expo- nodules in this group showed achynomenoid types (Fig. and $K\SRJDHD. Nodules in this group were clearly of the nential growth phase (the mid-log phase) was reached. MAFF303099 $±' DQG DOVR WKH URRW VXUIDFH )LJ ( ,Q DGGLWLRQ 1K–O) (38). DOA9 nodulated both genistoids, but only induced aeschynomenoid type (Fig. 1K–L), which were formed via was incubated for 2 d, inoculated into a fresh tube containing HM nodulation was related to the colonization sites (Fig. 1F–I). effective nodules on & MXQFHD. Among the Dalbergioids, the crack infection pathway (Fig. 1F and G) (38). This route medium, and then cultured under the same conditions for 24–36 h. All nodules were determinate (Fig. 1F–J). A thin section of DOA9 induced effective nodules on $DPHULFDQD, $DIUDVSHUD, of infection bypassed some of the complex processes
292 293 involved in infection via root hairs, which depend on the for the total nitrogen content and compared between inocu- Nod-factor structure (14). DOA9 induced nodules on all the lated and un-inoculated rice. Previous studies reported that Millettioids tested, except for *PD[, even though soybean- photosynthetic %UDG\UKL]RELXP strains induced N2¿[LQJ nodulating bacteria were distributed among three genera (6, nodules on the stems and roots of the genus $HVFK\QRPHQH 7). Rhizobia so far known to nodulate Millettioids are all (28), and also formed a natural endophytic association with classical rhizobia, mainly slow-growing (“brady-”) types the wild rice species 2U\]DEUHYLOLJXODWD (5), which grows in (38). Of the two Robinioids, only /MDSRQLFXV was nodulated; association with several aquatic legumes. However, no stud- however, these nodules were ineffective (see Fig. S1 and ies have yet been published on non-photosynthetic bradyrhi- Table S1). This strain could nodulate the model legume zobial strains in endophytic association with rice. Therefore, /MDSRQLFXV. /MDSRQLFXV may have permitted intercellular it was interesting that the non-photosynthetic DOA9 could infection of the cortex (22). However, DOA9 could not colonize and infect rice. In Thailand, the semi-aquatic $ nodulate 6URVWUDWD, even though 6HVEDQLD accepted infection DPHULFDQD frequently grows in association with rice. The via the crack entry mode (14). Among the Inverted Repeat– genes in DOA9 that are involved in rice infection, which may /DFNLQJ&ODGH ,5/& RQO\0VDWLYD was nodulated; however, also play a role in the early interaction between $DPHULFDQD these nodules were ineffective. These Papilionoideae tribes DQG'2$ VKRXOGEHH[DPLQHGLQPRUHGHWDLO7KH¿QGLQJV DUH JHQHUDOO\ QRGXODWHG ZLWK IDVWJURZLQJ ĮUKL]RELD of such a study may reveal a common determinant in the ($OSKDSURWHREDFWHULD) and typically show a high degree of symbiotic interaction between rhizobia and ancestor plants VSHFL¿FLW\EHWZHHQV\PELRWLFSDUWQHUV before the symbiotic relationship evolved. 1RGXODWLRQRIWKHVXEIDPLO\0LPRVRLGHDH 0LFURVFRSLFREVHUYDWLRQRILVRODWHGEDFWHURLGV We tested 4 species (Table 1). Although DOA9 nodulated Bacteroids isolated from the various species of /OHXFRFHSKDOD and 6VDPDQ, these nodules were not effec- Papilionoideae were indistinguishable from the free-living tive. This limited result may have been due to host range IRUP )LJV DQG 1R VLJQL¿FDQW GLIIHUHQFHV ZHUH VSHFL¿FLW\DVVSHFLHVRIWKHWULEH0LPRVHDHPD\EHQRGX- observed in the shape or size of bacteroids between this group lated by E-rhizobia (%HWDSURWHREDFWHULD) (3, 11). DQG WKH IUHHOLYLQJ IRUP )LJ $±( +RZHYHU EDFWHURLGV %UDG\UKL]RELXP sp. DOA9 is able to nodulate several from $K\SRJDHD were spherical (Figs. 3F, 4B, 4F, 4J). It is species of legumes (Table 1) (29). In addition, $DPHULFDQD widely accepted that the size and shape of N2¿[LQJ plants are commonly nodulated by %UDG\UKL]RELXP spp. of bacteroids vary widely and are controlled by the legume host WKHFRZSHDPLVFHOODQHRXVJURXS 7KHVH¿QGLQJVKLJK- rather than by the rhizobial genotype (31). The bacteroids lighted the broad host range characteristics of these bacteria. LVRODWHGIURPWKHQRGXOHVRIIRXURXWRI¿YH3DSLOLRQRLGHDH DOA9 also nodulated many of the Dalbergioids, Millettioids, and Robinioids. DOA9 has been shown to contain divergent QRG-genes (29), which may facilitate the broad host range nodulation ability. Moreover, the ability of DOA9 to invade roots via the cracks through which lateral roots emerge (29) may allow it to infect a wide variety of legumes; however, we found no infection threads in any sample. These results indi- cated that DOA9 showed a broad host range. 5RRWFRORQL]DWLRQDQGLQIHFWLRQRIULFH DOA9 colonized the roots and infected the tissues of rice. The expression of GUS was indicated by blue staining (Fig. 2). At 1 d after inoculation, the histochemical staining of E-galactosidase activity revealed the strong colonization of the root cap (Fig. 2A). At 5 d, this strain colonized the entire URRW )LJ % DQG LQWHUFHOOXODU FHOOV )LJ & DQG ' DV Fig. 3. ([DPSOHVRI'LIIHUHQWLDO,QWHUIHUHQFH&RQWUDVW ',& YLHZVRI shown in blue. However, we did not measure ARA in rice and %UDG\UKL]RELXP sp. DOA9 bacteroids. (A) Free-living. (B–F) Bacteroids isolated from (B) ,QGLJRIHUDWLQFWRULD & 0DFURSWLOLXPDWURSXUSXUHXP, LWV HI¿FLHQF\ FRXOG QRW EH GHFLGHG EDVHG RQ SODQW JURZWK (D) $HVFK\QRPHQH DPHULFDQD ( $HVFK\QRPHQH DIUDVSHUD, and (F) 7KLVVKRXOGEHIXUWKHUDQDO\]HGXVLQJWKH.MHOGDKOPHWKRG $UDFKLVK\SRJDHD.
Fig. 2. Light microscopic observations of 2U\]DVDWLYD ssp. LQGLFD seedlings inoculated with GUS-tagged %UDG\UKL]RELXP sp. DOA9. (A) 1 d after LQRFXODWLRQ % GDIWHULQRFXODWLRQ & &ORVHXSYLHZRIURRWKDLUDQG ' FORVHXSYLHZRIURRWWLVVXHGD\VDIWHULQRFXODWLRQ
294 295 involved in infection via root hairs, which depend on the for the total nitrogen content and compared between inocu- Nod-factor structure (14). DOA9 induced nodules on all the lated and un-inoculated rice. Previous studies reported that Millettioids tested, except for *PD[, even though soybean- photosynthetic %UDG\UKL]RELXP strains induced N2¿[LQJ nodulating bacteria were distributed among three genera (6, nodules on the stems and roots of the genus $HVFK\QRPHQH 7). Rhizobia so far known to nodulate Millettioids are all (28), and also formed a natural endophytic association with classical rhizobia, mainly slow-growing (“brady-”) types the wild rice species 2U\]DEUHYLOLJXODWD (5), which grows in (38). Of the two Robinioids, only /MDSRQLFXV was nodulated; association with several aquatic legumes. However, no stud- however, these nodules were ineffective (see Fig. S1 and ies have yet been published on non-photosynthetic bradyrhi- Table S1). This strain could nodulate the model legume zobial strains in endophytic association with rice. Therefore, /MDSRQLFXV. /MDSRQLFXV may have permitted intercellular it was interesting that the non-photosynthetic DOA9 could infection of the cortex (22). However, DOA9 could not colonize and infect rice. In Thailand, the semi-aquatic $ nodulate 6URVWUDWD, even though 6HVEDQLD accepted infection DPHULFDQD frequently grows in association with rice. The via the crack entry mode (14). Among the Inverted Repeat– genes in DOA9 that are involved in rice infection, which may /DFNLQJ&ODGH ,5/& RQO\0VDWLYD was nodulated; however, also play a role in the early interaction between $DPHULFDQD these nodules were ineffective. These Papilionoideae tribes DQG'2$ VKRXOGEHH[DPLQHGLQPRUHGHWDLO7KH¿QGLQJV DUH JHQHUDOO\ QRGXODWHG ZLWK IDVWJURZLQJ ĮUKL]RELD of such a study may reveal a common determinant in the ($OSKDSURWHREDFWHULD) and typically show a high degree of symbiotic interaction between rhizobia and ancestor plants VSHFL¿FLW\EHWZHHQV\PELRWLFSDUWQHUV before the symbiotic relationship evolved. 1RGXODWLRQRIWKHVXEIDPLO\0LPRVRLGHDH 0LFURVFRSLFREVHUYDWLRQRILVRODWHGEDFWHURLGV Fig. 4. ([DPSOHV RI VKDSHV RI OHIWKDQG FROXPQ IUHHOLYLQJ '2$ EDFWHULD DQG RWKHU FROXPQV UHVLGHQW EDFWHURLGV LVRODWHG IURP % ) - We tested 4 species (Table 1). Although DOA9 nodulated Bacteroids isolated from the various species of $UDFKLVK\SRJDHD &*. $HVFK\QRPHQHDIUDVSHUD, and (D, H, L) /HVSHGH]DVSQRGXOHV',&GLIIHUHQWLDOLQWHUIHUHQFHFRQWUDVWPLFURVFRS\ /OHXFRFHSKDOD and 6VDPDQ, these nodules were not effec- Papilionoideae were indistinguishable from the free-living '$3,ÀXRUHVFHQFHPLFURVFRS\ZLWK¶GLDPLGLQRSKHQ\OLQGROHVWDLQ 3,ÀXRUHVFHQFHPLFURVFRS\ZLWKSURSLGLXPLRGLGHVWDLQ tive. This limited result may have been due to host range IRUP )LJV DQG 1R VLJQL¿FDQW GLIIHUHQFHV ZHUH VSHFL¿FLW\DVVSHFLHVRIWKHWULEH0LPRVHDHPD\EHQRGX- observed in the shape or size of bacteroids between this group lated by E-rhizobia (%HWDSURWHREDFWHULD) (3, 11). DQG WKH IUHHOLYLQJ IRUP )LJ $±( +RZHYHU EDFWHURLGV %UDG\UKL]RELXP sp. DOA9 is able to nodulate several from $K\SRJDHD were spherical (Figs. 3F, 4B, 4F, 4J). It is species of legumes (Table 1) (29). In addition, $DPHULFDQD widely accepted that the size and shape of N2¿[LQJ plants are commonly nodulated by %UDG\UKL]RELXP spp. of bacteroids vary widely and are controlled by the legume host WKHFRZSHDPLVFHOODQHRXVJURXS 7KHVH¿QGLQJVKLJK- rather than by the rhizobial genotype (31). The bacteroids lighted the broad host range characteristics of these bacteria. LVRODWHGIURPWKHQRGXOHVRIIRXURXWRI¿YH3DSLOLRQRLGHDH DOA9 also nodulated many of the Dalbergioids, Millettioids, and Robinioids. DOA9 has been shown to contain divergent QRG-genes (29), which may facilitate the broad host range nodulation ability. Moreover, the ability of DOA9 to invade roots via the cracks through which lateral roots emerge (29) may allow it to infect a wide variety of legumes; however, we found no infection threads in any sample. These results indi- cated that DOA9 showed a broad host range.
5RRWFRORQL]DWLRQDQGLQIHFWLRQRIULFH Fig. 5. Determination of %UDG\UKL]RELXP VS '2$ UHSOLFRQV DQG V\PELRVLV JHQHV $ 0HJDSODVPLG SUR¿OHV ODQH 0HVRUKL]RELXP ORWL 0$)) DQGNE ODQH'2$ %±( 6RXWKHUQEORWK\EULGL]DWLRQVLJQDOVRIQRGXODWLRQDQG1¿[DWLRQJHQHVRQWKHPHJDSODVPLG DOA9 colonized the roots and infected the tissues of rice. and chromosome of DOA9 under low stringency conditions: (B) QRG$ & QRG%; (D) QRG& ( nifH. The expression of GUS was indicated by blue staining (Fig. 2). At 1 d after inoculation, the histochemical staining of E-galactosidase activity revealed the strong colonization of species were not swollen and were the same size (2–4 Pm) as the root cap (Fig. 2A). At 5 d, this strain colonized the entire WKHIUHHOLYLQJIRUP )LJ$±( 0HJDSODVPLGGHWHFWLRQDQGK\EULGL]DWLRQ URRW )LJ % DQG LQWHUFHOOXODU FHOOV )LJ & DQG ' DV Fig. 3. ([DPSOHVRI'LIIHUHQWLDO,QWHUIHUHQFH&RQWUDVW ',& YLHZVRI Furthermore, PI only stained the bacteroids from $ We successfully extracted both chromosomal and shown in blue. However, we did not measure ARA in rice and %UDG\UKL]RELXP sp. DOA9 bacteroids. (A) Free-living. (B–F) Bacteroids isolated from (B) ,QGLJRIHUDWLQFWRULD & 0DFURSWLOLXPDWURSXUSXUHXP, K\SRJDHD (Fig. 4J). In addition, no colony appeared when megaplasmid DNAs. The megaplasmid size was estimated LWV HI¿FLHQF\ FRXOG QRW EH GHFLGHG EDVHG RQ SODQW JURZWK (D) $HVFK\QRPHQH DPHULFDQD ( $HVFK\QRPHQH DIUDVSHUD, and (F) the bacteroid fraction prepared from the nodules of $ IURP WKH SODVPLG SUR¿OH RI 0 ORWL MAFF303099, which 7KLVVKRXOGEHIXUWKHUDQDO\]HGXVLQJWKH.MHOGDKOPHWKRG $UDFKLVK\SRJDHD. K\SRJDHD was plated on agar media. Although further quanti- contained megaplasmids of 208 and 352 kb in size (16) (Fig. tative analyses are needed, this result suggests that bacteroids 5A lane 1). DOA9 revealed one megaplasmid, which was in the nodules of $K\SRJDHD lost their productivity. These larger than 352 kb (Fig. 5A lane 2). The megaplasmid hybrid- characteristics, the absence of PI staining, and the ability to ized with all the probes for QRG$, QRG%, QRG&, and nifH (Fig. form colonies on agar plates place DOA9 in the group that %±( 7KHnifH probe also hybridized with the chromosomal continues to reproduce after leaving the nodules (25). However, '1$V )LJ( the bacteroids isolated from $K\SRJDHD were spherical and Broad-host-range symbionts generally carry multiple repl- non-reproductive after leaving the nodules. Since the mor- icons on their genome. 6LQRUKL]RELXP sp. NGR234, which phology of bacteroids (which is determined by the host legume) has an extremely broad host range, has three replicons: a has been linked to reproductive viability (25), the host species symbiosis plasmid, megaplasmid, and chromosome (12);6 Fig. 2. Light microscopic observations of 2U\]DVDWLYD ssp. LQGLFD seedlings inoculated with GUS-tagged %UDG\UKL]RELXP sp. DOA9. (A) 1 d after can have implications for rhizobial evolution (32). IUHGLL USDA257 has a chromosome and plasmid (36); 6 LQRFXODWLRQ % GDIWHULQRFXODWLRQ & &ORVHXSYLHZRIURRWKDLUDQG ' FORVHXSYLHZRIURRWWLVVXHGD\VDIWHULQRFXODWLRQ IUHGLL++KDVRQHFKURPRVRPHDQG¿YHSODVPLGV
294 295 Fig. 6. Determination of copy numbers of (A) QRG$, (B) QRG% & QRG&, and (D) nifH by Southern blot hybridization. %UDG\UKL]RELXP sp. DOA9 genomic DNA was digested with the restriction enzymes shown, and the blot was hybridized with probes for QRG$ (from %\XDQPLQJHQVH SUTN6-2), QRG% (from %FDQDULHQVH SUTN7-2), QRG& (from %MDSRQLFXP USDA110), and nifH (from %\XDQPLQJHQVH SUTN6-2).
DOA9 has a chromosome (cDOA9) and megaplasmid %UDG\UKL]RELXP sp. located on a megaplasmid. A detailed (pDOA9) larger than 352 kb. A large number of the natural investigation of both replicons of DOA9 is warranted. isolates of 5KL]RELXP spp. have been shown to carry various Previous studies on the host range of rhizobia have focused large plasmids (21). Genes for a few functions have been on 6LQRUKL]RELXP sp. strain NGR234, 6 IUHGLL USDA257, detected on these large plasmids, particularly the QRG and nif and 6IUHGLL HH103. Since the genome sequences of these genes of fast-growing rhizobia. In contrast, the published three strains share a high degree of synteny (35, 36, 42), the genomes of %UDG\UKL]RELXP spp. are composed of only a host range of rhizobia may be related to the number of spe- chromosome, except for %UDG\UKL]RELXP sp. BTAi1, which cialized protein secretion systems they carry (35). Therefore, also harbors a plasmid of 228 kb (9). There have not yet been a future study is needed to determine and compare the protein any studies published on symbiosis genes localized on the secretion systems of DOA9. plasmids of %UDG\UKL]RELXP (21). The symbiotic islands of %GLD]RHI¿FLHQV USDA110 have been located on a chromo- Conclusion some, while the symbiosis genes of 6LQRUKL]RELXP sp. were located on plasmids (13). This characteristic may contribute Since many of the photosynthetic bradyrhizobial (PB) to the broad host range of these rhizobia. It will be interesting strains have been reported to be naturally endophytic with WR¿QGRXWZKHWKHUV\PELRVLVJHQHVRQSODVPLGKDYHH[WHQGHG rice (5), we concluded that rice may have evolved approxi- their host range. mately 120 million years before legumes (1, 20). Thus, PB strains may also be an ancestor of non-photosynthetic &RS\QXPEHU bradyrhizobia (41). In the present study, DOA9 exhibited the At least two fragments were detected using the probe for characteristic of a broad host range in legume and rice infec- QRG$ (Fig. 6A). This result suggested that QRG$ on the mega- tions. Therefore, DOA9 may represent the interval of evolu- plasmid was a single or two copies. Regarding QRG% and QRG&, tion in the nod factor-independent symbiotic system, but may a single fragment was detected on each lane, suggesting have later lost photosynthetic activity. Moreover, DOA9 is a that these two genes may be a single copy on the mega- divergent QRG-containing strain rendering the incomplete SODVPLG )LJ % DQG & $W OHDVW WZR IUDJPHQWV ZHUH production of nod factors. This may also support DOA9 detectable when probed by nifH. The result supported the two associating with host plants in a nod-independent manner nifH genes being separately located on the megaplasmid and since and during its evolution. To more clearly understand the FKURPRVRPH )LJV(DQG' evolution of this group, whole genome assessments, symbi- DOA9 has at least two copies of nifH, one on the megaplas- otic genes comparisons and disruptions, as well as protein mid and one on the chromosome. Two copies of nifH were secretion systems and lipopolysaccharide structures should also shown to be present in photosynthetic %UDG\UKL]RELXP be explored in more detail. ORS278 and BTAi1 (30), and one copy in %GLD]RHI¿FLHQV USDA110 (17), in all cases on the chromosome. We already Acknowledgements sequenced QRG$ of DOA9 (Accession No. DF 820426.1), This work was supported by Suranaree University of Technology; which was the corresponding fragment of the QRG$ probe E\WKH+LJKHU(GXFDWLRQ5HVHDUFK3URPRWLRQDQG1DWLRQDO5HVHDUFK used in this study, and no restriction sites of (FRRI, +LQGIII, 8QLYHUVLW\ 3URMHFW RI 7KDLODQG 2I¿FH RI WKH +LJKHU (GXFDWLRQ %JOII, or 1RW, ZHUH LGHQWL¿HG WKHUHIRUH WKH K\EULGL]DWLRQ &RPPLVVLRQ DQGE\D-6365RQSDNXVFKRODUVKLS7KDQNVDUHH[WHQGHG data probed by QRG$ suggested that DOA9 may have at least to Dr. Issra Pramoolsook for advice and comments on the manuscript. two copies of QRG$ on the megaplasmid. An increase in the References copy number of the symbiotic region has been indicated to promote the plant phenotype. The inoculation of alfalfa with 1. Allen, K.D. 2002. Assaying gene content in Arabidopsis. Proc. Natl. a moderate increase in the copy number of the symbiotic Acad. Sci. U.S.A. 99:9568–9572. $UJDQGRxD 0 ) 0DUWtQH] &KHFD , /ODPDV ( 4XHVDGD DQG $ region of 6PHOLORWL resulted in enhanced plant growth (4). Moral. 2003. Megaplasmids in Gram negative, moderately halophilic 7KHUHIRUHWKLVLVWKH¿UVWVWXG\RQWKHV\PELRVLVJHQHVRI EDFWHULD)(060LFURELRO/HWW ±
296 297 %DUUHWW&)DQG0$3DUNHU&RH[LVWHQFHRI%XUNKROGHULD, 24. McRae, D., R. Miller, W. Berndt, and K. Joy. 1989. Transport of &XSULDYLGXV, and 5KL]RELXP sp. nodule bacteria on two 0LPRVD spp. &GLFDUER[\ODWHVDQGDPLQRDFLGVE\5KL]RELXPPHOLORWL bacteroids. LQ&RVWD5LFD$SSO(QYLURQ0LFURELRO ± Mol. Plant-Microbe Interact. 2:273–278. &DVWLOOR 0 0 )ORUHV 3 0DYLQJXL ( 0DUWtQH]5RPHUR 5 0HUJDHUW 3 7 8FKLXPL % $OXQQL * (YDQQR $ &KHURQ 2 Palacios, and G. Hernández. 1999. Increase in alfalfa nodulation, &DWULFH $( 0DXVVHW ) %DUOR\+XEOHU ) *DOLEHUW DQG $ QLWURJHQ¿[DWLRQDQGSODQWJURZWKE\VSHFL¿F'1$DPSOL¿FDWLRQLQ .RQGRURVL(XNDU\RWLFFRQWURORQEDFWHULDOFHOOF\FOHDQGGLIIHU- 6LQRUKL]RELXPPHOLORWL$SSO(QYLURQ0LFURELRO ± entiation in the Rhizobium–legume symbiosis. Proc. Natl. Acad. Sci. &KDLQWUHXLO&(*LUDXG<3ULQ-/RUTXLQ$%k0*LOOLV3GH U.S.A. 103:5230–5235. /DMXGLH DQG % 'UH\IXV 3KRWRV\QWKHWLF EUDG\UKL]RELD DUH 0LFKp/DQG-%DODQGUHDX(IIHFWVRIULFHVHHGVXUIDFHVWHULO- natural endophytes of the African wild rice 2U\]DEUHYLOLJXODWD. Appl. ization with hypochlorite on inoculated %XUNKROGHULDYLHWQDPLHQVLV. (QYLURQ0LFURELRO ± $SSO(QYLURQ0LFURELRO ± &KHQ:;*+ 296 297