P oc. Na l. Acad. Sci. USA
Vol. 95, pp. 9687–9692, Augus 1998
Plan Biology
MsPG3,aMedicago sa i a polygalac u onase gene exp essed
du ing he al al a–Rhizobium melilo i in e ac ion
(in ec ion p ocess
y
symbiosis
y
noduline
y
pec ic gene)
JOSE A. MUN
˜OZ*, CARMEN CORONADO*, JAVIER PE
´REZ-HORMAECHE*, ADAM KONDOROSI†‡,PASCAL RATET†,
AND ANTONIO J. PALOMARES*§
*Depa amen o de Mic obiologı´a y Pa asi ologı´a. Facul ad de Fa macia, Uni e sidad de Se illa, 41012 Se illa, Spain; †Ins i u des Sciences Ve´ge´ ales, Cen e
Na ional de la Reche che Scien i ique, 91198 Gi su Y e e Cedex, F ance; and ‡Ins i u e o Gene ics, Biological Resea ch Cen e , Hunga ian Academy o
Sciences, Szeged P.O. Box 521, H-6701 Hunga y
Communica ed by Donald R. Helinski, Uni e si y o Cali o nia, San Diego, La Jolla, CA, May 28, 1998 ( ecei ed o e iew Oc obe 10, 1997)
ABSTRACT Polygalac u onase (PG) is one o he mos
impo an enzymes associa ed wi h plan cell wall deg ada-
ion. I has been p oposed o pa icipa e in he ea ly s eps o
he Rhizobium–legume in e ac ion. We ha e iden i ied wo
classes o cDNA agmen s co esponding o wo classes o PG
genes in he Medicago genome. One o his class, ep esen ed
by E
2
in M. unca ula and Pl
1
in M. sa i a, seems o be ela ed
o p e iously cha ac e ized plan PG genes exp essed in
pollen. We ha e isola ed he genomic clone con aining he
en i e gene co esponding o he second class (E
3
). We showed
ha MsPG3 is a single gene in he Medicago genome coding o
PG. By e e se ansc ip ion-PCR, MsPG3 exp ession was
de ec ed in oo s 1 day a e Rhizobium inocula ion. The ea ly
induc ion o he MsPG3, as also seen by in si u hyb idiza ion
expe imen s, suppo s i s in ol emen in he ea ly s ages o
he Rhizobium-legume in ec ion p ocess. In addi ion, by ana-
lyzing he exp ession o a MsPG3 p omo e -gus cons uc in
Vicia hi su a- ansgenic oo nodules, we showed ha MsPG3
was exp essed in all cells o nodule p imo dia and in he cells
o he in asion zone. By No he n blo , MsPG3 ansc ip s a e
no de ec ed in a ious Medicago issues, indica ing ha he
unc ion o his gene is ela ed closely o symbiosis. Thus, ou
esul s s ongly sugges he in ol emen o MsPG3 gene
du ing me is em o ma ion andyo in he in ec ion p ocess,
p obably by acili a ing cell wall ea angemen , pene a ion
o he bac e ia h ough he oo hai wall, o in ec ion h ead
o ma ion and elease o bac e ia in plan cells. MsPG3
ep esen s a class o PG genes, dis inc om he pollen-
speci ic genes, and i is he i s pec ic encoded enzyme
demons a ed o be in ol ed in Rhizobium-legume symbiosis.
Soil bac e ia known as hizobia ha e he capaci y o induce he
o ma ion o nodules on he oo o hei leguminous hos
plan s. When he bac e ia a e eleased in o he plan cells, hey
di e en ia e in o bac e oids ixing he a mosphe ic ni ogen.
The ea ly s ages o he in e ac ion a e con olled by signals
exchanged be ween he wo symbio ic pa ne s, esul ing in
spa ially and empo ally egula ed exp ession o speci ic genes
(1). The mechanisms con olling he bac e ial en y, he ini ial
o ma ion o in ec ion h eads, he di ec ional g ow h and
passage h ough oo hai and co ical cell walls o hese
s uc u es, and he elease o he bac e ia in he in ec ed cells
a e s ill poo ly unde s ood. Nu man (2) p oposed ha a
edi ec ion o oo hai ip g ow h by bac e ial ac ion esul ed
in an in agina ion o he hai wall o ming he in ec ion h ead.
Fah aeus and Ljungg en (3) p oposed ha legume oo s
p oduce he enzyme polygalac u onase induced by he homol-
ogous Rhizobium s ain. The abili y o induce polygalac u o-
nase p oduc ion by he oo may be con olled by a esiden
plasmid o he Rhizobium (4). Elec on mic oscopic s udies
ha e demons a ed ha hizobia pene a e a deg aded egion
o he plan cell wall a a e y localized si e (5). These s udies
suppo he in ol emen o cell wall deg ading enzymes in he
in ec ion p ocess allowing he pene a ion o he bac e ia. In
addi ion, he p oduc ion o celluloly ic and pec oly ic enzymes
by Rhizobium has been demons a ed (6). Howe e , conclusi e
expe imen s demons a ing he ole o hese enzymes in he
in ec ion and nodula ion phenomena ha e no been epo ed
ye .
Polygalac u onase (PG) is one o he mos impo an en-
zymes associa ed wi h cell wall deg ada ion. This enzyme, in
conjunc ion wi h o he sec e ed cell wall-deg ading enzymes
such as pec in es e ase and pec a e lyase a e in ol ed la gely
in se e al biological p ocesses. Endopolygalac u onase ac i i y
has been iden i ied in ipening ui s, and i has been impli-
ca ed as he p ima y agen o polyu onide deg ada ion and
hence implica ed as he enzyme esponsible o ui so ening
ha accompany ipening (7). B own and C ouch (8) p oposed
ha PG may unc ion by depolyme izing pec in in he cell wall
o he pis il du ing pollina ion o allow pene a ion by he
pollen ubes andyo o p o ide wall p ecu so s o he g owing
ube. This enzyme is also among hose which a e sec e ed by
in asi e plan pa hogens o deg ade he hos cell wall (9).
We ha e cha ac e ized a Medicago sa i a polygalac u onase
gene (MsPG3) speci ically exp essed du ing symbiosis wi h R.
melilo i. Ou exp ession s udies sugges ha his enzyme may
play mul iple oles du ing he symbio ic in e ac ion.
MATERIALS AND METHODS
G ow h Condi ions o Plan s and Bac e ia. Al al a seeds
s e iliza ion and plan g ow h unde ae oponic condi ions
we e done acco dingly o Baue e al. (10). To gene a e oo
nodules, he oo s we e inocula ed wi h wild- ype R. melilo i
s ain 41 and he non-nodula ing mu an de i a i e ZB138 as
desc ibed (10).
Nucleic Acid Manipula ions. Genomic DNA o M. sa i a ssp.
sa i a c . Nagysze´na`si and M. unca ula c . Jemalong was
p epa ed acco ding o Dellapo a e al. (11). Res ic ion
enzyme diges ions we e pe o med unde s anda d condi ions.
To al RNA was isola ed om di e en o gans o M. sa i a ssp.
a ia o M. sa i a ssp. sa i a using he guanidinium hiocyana e
me hod and subsequen cesium chlo ide g adien cen i uga-
The publica ion cos s o his a icle we e de ayed in pa by page cha ge
paymen . This a icle mus he e o e be he eby ma ked ‘‘ad e isemen ’’ in
acco dance wi h 18 U.S.C. §1734 solely o indica e his ac .
© 1998 by The Na ional Academy o Sciences 0027-8424y98y959687-6$2.00y0
PNAS is a ailable online a www.pnas.o g.
Abb e ia ions: RT-PCR, e e se ansc ip ion—PCR; PG, polygalac-
u onase; GUS,
b
-glucu onidase.
Da a deposi ion: The sequence epo ed in his pape has been
deposi ed in he GenBank da abase (accession no. Y11118).
§To whom co espondence should be add essed. e-mail: paloma e@
cica.es.
9687
ion acco ding o Samb ook e al. (12). Poly(A
1
) RNA we e
p epa ed acco ding o Samb ook e al. (12). T ans e o he
nucleic acids o Hybond-N nylon memb anes (Ame sham) and
subsequen hyb idiza ion we e ca ied ou by using he man-
u ac u e ’s ins uc ions.
Polygalac u onase Gene Isola ion. Two oligonucleo ide
p ime s named 370: 59-GGTGATGATTGTAT(TyC)TCA-
AT-39and 389: 59-GTCTTGATCCTAACTCC-39co espond-
ing espec i ely o he conse ed amino acid sequences GD-
DCISI and GVRIKT p esen in di e en plan PG genes we e
used o PCR ampli y se e al pa ial Medicago DNA agmen s
homologous o PG. Then 100 ng o genomic DNA o one- en h
o he cDNA syn hesis eac ion mix we e used as empla es o
ampli ica ion in 100
m
lo Taq polyme ase bu e wi hou Mg
21
(10 mM T iszHCl, pH 9.0y50 mM KCly0.1% T i on X-100), 6
mM MgCl
2
, 0.12 mM dNTP, 70 pmol o 370 and 389 p ime s,
and 1.5 uni s o Taq polyme ase, (P omega). Ampli ica ion was
pe o med by using 30 cycles o 1-min dena u a ion a 94°C,
1-min p ime annealing a 48°C, and 1-min elonga ion a 72°C.
The PCR p oduc s we e size- ac iona ed on a 1.5% aga ose
gel, ans e ed o nylon il e s, and hyb idized o a adiola-
beled oma o PG cDNA clone (13). A 186-bp PCR p oduc
ha hyb idized was blun -ended by using T4 DNA polyme ase
and cloned in o he EcoRV si e o pBluesc ip IIKS (S a -
agene). To isola e he en i e al al a MsPG3 gene, he E
3
agmen (see Resul s) was used as a p obe o sc een a genomic
lib a y p epa ed om M. sa i a ssp. sa i a c . Nagysze´na`si
genomic DNA cloned in o he
l
EMBL-4 ec o by using
s anda d p ocedu es (12). Inse s o he pu i ied phages we e
subcloned in o pBluesc ip IIKS (S a agene).
Re e se T ansc ip ion–PCR (RT-PCR). The echnical as-
pec s o he RT-PCR assay and he necessa y con ols we e
desc ibed p e iously (10). Based on he nucleo ide sequence
alignmen o agmen s E
3
,E
2
and Pl
1
, speci ic p ime s we e
designed. P ime 427 (59-CCATTGTCAGTATTTG(AyT)-
GT-39) is speci ic o E
3
and p ime 428 (59-CATTGTCG-
GTTG(TyC)TGTCAG-39) o E
2
and Pl
1
. The ampli ica ion
a e was in a linea ange o all PCR p oduc s. One- en h o
he PCR p oduc s we e elec opho esed and ans e ed o
Hybond-N nylon memb ane. The memb ane was i s hyb id-
ized a high s ingency condi ions o ei he he Msc27 o Rhe2
PCR agmen s o quan i ica ion. The blo was s ipped and
ehyb idized o ei he he E
2
o E
3
PCR agmen s.
DNA Sequence Analysis. DNA sequencing analysis was
ca ied ou by he dideoxy chain- e mina ing me hod (12). Fo
agmen s longe han 0.5 kb, dele ion de i a i es we e ob-
ained by using he E ase-a-Base Sys em Ki (P omega).
Compu e analyses o sequence in o ma ion was pe o med by
using he so wa e package o he Uni e si y o Wisconsin
Gene ics Compu e G oup.
In Si u Hyb idiza ion. Fo p epa ing an isense and sense
RNA agmen s o he al al a MsPG3 gene, a cDNA agmen
o 833 bp, ob ained by RT-PCR om al al a nodule RNA, was
cloned in o pBluesc ip IISK ec o (S a agene) gi ing plas-
mid pCJ3. Fo his RT-PCR ampli ica ion, wo p ime s co -
esponding o genomic PG sequence we e syn hesized. The 59
p ime (p ime 420) is 59-CTTAATGATGCTTATCAA-
TGG-39and he 39p ime (p ime 1751) is 59-ACACATTT-
AGCGTTTACTGG-39. The p obes we e digoxigenin-labeled
by using he DIG RNA-labeling Ki om Boeh inge Mann-
heim and deg aded o '150 nucleo ides long be o e hyb id-
iza ion. Al al a nodules and nodule p imo dia induced by
wild- ype R. melilo i s ain 41 we e ixed, dehyd a ed, and
inally in il a ed and embedded in o pa a in. Sec ions (9
m
m
hick) we e hyb idized o digoxigenin-labeled an isense o
sense RNA p obes acco ding o a p ocedu e de i ed om
Bochenek and Hi sch (14). Sec ions we e pho og aphed wi h
a POLYVAR Reiche -Jung mic oscope by using T-160 Ek-
ach ome Kodak ilm.
V. hi su a T ans o ma ion and Analysis o T ansgenic Roo
Nodules. A agmen con aining he 2.7 kb-p omo e se-
quence, ending 3 bp ups eam o he ATG codon o MsPG3
gene, was subcloned in o he bina y ec o pPR97 (15) o
cons uc a ansc ip ional usion o gusA-in . V. hi su a ans-
o ma ion was pe o med as desc ibed (16) using Ag obac e-
ium hizogenes A4Tc24 con aining he new plasmid named
pPG97. Composi e V. hi su a plan s we e ans e ed o esh
ni ogen- ee medium and inocula ed wi h an o e nigh cul-
u e o Rhizobium leguminosa um b . iciae. His ochemical
b
-glucu onidase (GUS) s aining o plan ma e ial was pe -
o med as desc ibed (17). Semi hin sec ions (100
m
m) o esh
un ixed nodules we e cu by using a Mic o-Cu H1200 (Bio-
Rad) and pho og aphed wi h a POLYVAR Reiche -Jung
mic oscope by using T-160 Ek ach ome Kodak ilm.
RESULTS
Isola ion o PCR P oduc s om Medicago cDNA and
Genomic DNA Hyb idizing o a Toma o Polygalac u onase
cDNA Clone. To de e mine he p esence o PG genes in he
Medicago o Rhizobium genome, we i s used he e ologous
p obes. When Sou he n blo s con aining genomic DNA ob-
ained om al al a, clo e , Rhizobium leguminosa um b .
i olii, and Rhizobium melilo i we e p obed wi h a oma o PG
cDNA clone (13) unde low s ingency condi ions, hyb idizing
bands we e obse ed only wi h he al al a genomic DNA (da a
no shown). To clone a homologous gene p obe, we used a PCR
app oach.
The plan PG genes desc ibed so a con ain a conse ed
egion in he middle o he coding sequence including a
his idine esidue a posi ion 293 o he oma o amino acid
sequence, which may be pa o he ac i e si e o he enzyme
(18). Two oligonucleo ide p ime s (370 and 389) expec ed o
ampli y a cDNA agmen o '180 bp co esponding o his
conse ed domain, we e syn hesized. Using hese p ime s,
agmen s o his size we e ampli ied om maize o al DNA,
M. sa i a ssp. sa i a genomic DNA, M. sa i a ssp. sa i a pollen
cDNA as well as om M. unca ula genomic DNA. All o hese
PCR agmen s hyb idized o he oma o PG cDNA clone
(da a no shown).
Two classes o agmen s, dis inguishable by he p esence o
aHindIII es ic ion si e, we e cloned om he PCR p oduc s
ob ained om he genomic DNA o he diploid M. unca ula
108–1. One ep esen a i e o each class was analyzed. The
agmen con aining a HindIII es ic ion si e was named E
3
and he o he one E
2
. A hi d clone was ob ained om he M.
sa i a ssp. sa i a pollen cDNA, named Pl
1
. This clone did no
con ain a HindIII es ic ion si e. Compa ison o he nucleo-
ide sequences and he p edic ed amino acid sequences om
hese PCR p oduc s wi h he al eady published sequences
e ealed a high deg ee o homology wi h PGs om di e en
o ganisms (da a no shown). In e es ingly, analysis o he
nucleo ide sequences o hese pa ial pu a i e PG genes e-
ealed ha agmen Pl
1
and E
2
we e mo e closely ela ed o
each o he (96% simila i y) han o agmen E
3
(71% and 69%
simila i y espec i ely), despi e he ac ha Pl
1
and E
2
we e
cloned om di e en Medicago species. In addi ion, he
pep ides encoded by E
2
and Pl
1
exhibi ed 73.7% iden i y wi h
he pollen PG o Zea mays (19) whe eas E
3
showed only 63.9%
iden i y. These di e en deg ees o homology, oge he wi h
he ac ha he Pl
1
agmen was isola ed om pollen cDNA,
sugges ed ha a leas wo classes o PG genes a e p esen in
Medicago. One class, ep esen ed by E
2
in M. unca ula and Pl
1
in M. sa i a, seems o be close ela ed o he p e iously
cha ac e ized genes, which a e exp essed in pollen. Using
Sou he n blo expe imen s o he PCR based app oach de-
sc ibed abo e, we we e unable o de ec PG- ela ed sequences
in Rhizobium leguminosa um b . i olii and Rhizobium melilo i.
9688 Plan Biology: Mun˜oz e al. P oc. Na l. Acad. Sci. USA 95 (1998)
To es ima e he numbe o PG genes in Medicago, genomic
DNA om he diploid M. unca ula was diges ed wi h EcoRI
and HindIII and subsequen ly hyb idized o Pl
1
and E
3
clones,
using he same mode a e s ingency condi ions in bo h cases.
Fig. 1 shows he esul s o he M. unca ula genomic DNA gel
blo analysis. Despi e he 68,6% sequence iden i y be ween he
wo p obes, dis inc hyb idizing pa e ns we e ob ained. Pl
1
p obe hyb idized s ongly o ou agmen s when he DNA
was diges ed wi h EcoRI and o h ee agmen s when he
DNA was diges ed wi h HindIII. E
3
p obe hyb idized s ongly
o one agmen when he DNA was diges ed wi h EcoRI and
o wo agmen s wi h he same in ensi y (bu lesse han o
EcoRI) when he DNA was diges ed wi h HindIII (in ag ee-
men wi h he ac ha E
3
con ains an in e nal HindIII
es ic ion si e). Simila esul s we e ob ained by using DNA
om he e aploid M. sa i a ssp. a ia c . A2 plan (da a no
shown).
Genes Co esponding o he E
3
and E
2
F agmen s A e
Di e en ially Exp essed. The E
3
and E
2
PCR agmen s we e
used o successi ely p obe he same RNA gel blo made o
poly(A)
1
RNA isola ed om a ious plan issues o M. sa i a
ssp. a ia c A2 (da a no shown). The E
2
agmen hyb idized
o an RNA ansc ip o '1.5 kb p esen a high le el in ma u e
lowe s. The ansc ip was de ec able a lowe le el in lowe
buds and was no de ec able in callus, lea es, nodules, o oo s,
con i ming he p e ious hypo hesis ha he E
2
agmen may
co espond o a Medicago PG gene exp essed in pollen. When
he same RNA gel blo was p obed wi h E
3
, no PG ansc ip
was de ec ed. These di e en ial exp ession da a a e in ag ee-
men wi h he Sou he n blo s udy, sugges ing ha he e a e a
leas wo classes o PG genes in Medicago.
Because o he possible low le el o ansien exp ession o
he PG gene co esponding o he E
3
agmen , he exp ession
pa e n was in es iga ed by using a RT-PCR app oach. This
me hod, in addi ion o being mo e sensi i e, allowed us o
examine he exp ession o he wo genes independen ly, de-
spi e hei high sequence homology, h ough he use o speci ic
oligonucleo ide p ime s. Alignmen o he nucleo ide se-
quences o he E
3
and E
2
yPl
1
PCR agmen s, as well as hei
p edic ed amino acid sequences, e ealed a egion which was
no conse ed a he 39end o he PCR p oduc . This egion
was chosen o designing new speci ic p ime s ha , in combi-
na ion wi h p ime 370, ep esen ed gene-speci ic p ime pai s
(370-427 o E
3
and 370-428 o E
2
yPl
1
). A e quan i ica ion,
cDNA was ampli ied wi h he E
3
-speci ic p ime s and he PCR
p oduc s we e hyb idized o E
3
. The cDNA also was ampli ied
wi h he E
2
-speci ic p ime s and hen he PCR p oduc s
hyb idized o E
2
. Fig. 2 Ashows ha he gene co esponding
o E
2
(PG-E
2
) was exp essed in lowe s bu no in oo s o
nodules (in ag eemen wi h he No he n s udy). The gene
co esponding o E
3
(PG-E
3
) was exp essed in nodules and in
oo s ha es ed 5 days a e inocula ion ei he wi h he wild-
ype s ain R. melilo i Rm41 o he Nod
2
de i a i e ZB138 bu
no in uninocula ed oo s. The PG-E
3
ansc ip s we e p esen
a highe le els in R. melilo i 41-inocula ed oo s han in oo s
inocula ed wi h he non-nodula ing mu an ZB138. No hyb id-
izing band appea ed in he lowe cDNA, which hyb idized
FIG. 1. Sou he n blo analysis o Medicago plan s. M. unca ula
genomic DNA was diges ed wi h EcoRI (E) and HindIII (H) and
hyb idized wi h he Pl
1
and E
3
p obes. The sizes o he molecula mass
ma ke s a e indica ed in kb a he le .
FIG. 2. RT-PCR exp ession s udies o PG genes co esponding o
he E
3
and E
2
DNA agmen s in di e en plan issues. (A) Exp es-
sion s udy using comple e oo , nodules, and lowe s. Medicago PGs
and he con ol Msc27 sequences we e coampli ied du ing 25 cycles
om RNA o a ious plan issues o M. sa i a ssp. sa i a c . Si el: H
2
O
(C), noninocula ed oo s (N.i.), oo s ha es ed 5 days a e R. melilo i
41 inocula ion (Rm), oo s ha es ed 5 days a e inocula ion wi h
s ain ZB138 (ZB) (in hese cases plan s we e inocula ed 3 days a e
ge mina ion), 21-day-old wild- ype nodules (Nod) and lowe s (Fl.).
The PCR p oduc s we e subjec ed o elec opho esis on 1% aga ose
gel, blo ed, and hyb idized o he co esponding p obe. The amoun
and quali y o loaded cDNA was checked by ampli ica ion wi h
Msc27-speci ic oligonucleo ides, ollowed by p obing wi h an Msc27
p obe. E
3
and E
2
sequences we e ampli ied wi h he oligonucleo ide
p ime s 370-427 and 370-428 pai s espec i ely and each Sou he n blo
was subsequen ly hyb idized o E
3
and E
2
p obes. (B) Exp ession s udy
using spo innocula ed oo s. Th ee-day-old al al a seedlings o M.
sa i a ssp. sa i a c . Si el we e inocula ed wi h Gibson plan medium
(G), R. melilo i 41 (Rm), o he non-nodula ing mu an ZB138 (ZB).
RNA om oo samples ha es ed 1 day (1 d.p.i.) and 5 days (5 d.p.i.)
a e inocula ion we e used o RT-PCR. The PCR p oduc s we e
subjec ed o elec opho esis on 1% aga ose gel, blo ed, and hyb idized
o he co esponding p obe. cDNAs we e coampli ied by PCR by using
Msc27 speci ic p ime s and 370-427 p ime s, and he PCR p oduc s
we e successi ely hyb idized o an Msc27 o E
3
p obe. (C) Analysis o
MsEnod12 exp ession: The same cDNA sample we e used as in B.Rhe2
was used as cons i u i e con ol ins ead o Msc27 in his expe imen .
cDNA was coampli ied by PCR using Rhe2 speci ic p ime s and
MsEnod12 p ime s and he PCR p oduc s we e successi ely hyb idized
o he Rhe2 o MsEnod12B p obe (10).
Plan Biology: Mun˜oz e al. P oc. Na l. Acad. Sci. USA 95 (1998) 9689
wi h he E
3
agmen , indica ing ha he designed p ime s
we e each highly speci ic o only one PG clone. The coam-
pli ica ion o he cons i u i ely exp essed al al a gene Msc27
was used as in e nal con ol in he RT-PCR eac ions.
The exp ession o MsPG3 in all oo s inocula ed wi h he
s ain ZB138 could be caused by a speci ic esponse o he
Rhizobium o al e na i ely o a nonspeci ic esponse o he
oo s o he bac e ia. Pe e o e al. (20) also ha e shown ha
PG can be exp essed du ing g ow h o la e al oo p imo dium.
The e o e, o be able o s udy he speci ic esponse o he Nod
ac o ea men , he exp ession o he MsPG3 gene a he si e
o Rhizobium in ec ion was s udied by using he spo inocula-
ion echnique (10). Th ee-day-old al al a seedlings we e spo -
inocula ed wi h he wild- ype s ain 41 and i s nod mu an
de i a i e ZB138. Roo sec ions co esponding o he inocu-
la ed egions we e ha es ed 1 and 5 days a e inocula ion,
and MsPG3 exp ession was s udied by RT-PCR by using he
combina ion o speci ic p ime s 370 and 427. The highes le el
o he MsPG3 ansc ip was de ec ed in oo s collec ed 1 day
a e ea men wi h he wild- ype s ain R. melilo i 41 ( he
ea lies ime poin es ed; Fig. 2B), which ag ee wi h he esul s
o he p e ious expe imen (Fig. 2A). A 5 days a e innocu-
la ion, he amoun o ansc ip was educed in he Rm41
ea ed oo s. No ansc ip s we e de ec able in he noninocu-
la ed o in he ZB138 inocula ed oo s. Simul aneously, he
RNA samples used we e checked by compa ing he MsPG3
exp ession pa e n wi h ha o he MsEnod12 gene (Fig. 2C),
one o he bes cha ac e ized ea ly nodulin genes associa ed
wi h he ea ly s ages o nodule de elopmen . MsEnod12A and
MsEnod12B we e ampli ied by using speci ic oligonucleo ide
p ime s gi ing wo bands co esponding o he expec ed 239-
and 299-bp sizes. Acco ding o Baue e al. (10), he exp ession
o MsEnod12A was de ec able 5 days a e inocula ion wi h R.
melilo i 41, when nodules became isible and MsEnod12B
ansc ip s we e ound in bo h oo samples. These esul s
indica ed ha he PG-E3 gene beha es as an ea ly nodulin by
exp essing a e y ea ly s ages o nodules de elopmen .
Isola ion and S uc u e o he Medicago PG Gene Co e-
sponding o he E
3
F agmen . To u he cha ac e ize he
al al a PG gene co esponding o he E
3
agmen , we isola ed
he en i e gene. Fo his, a M. sa i a ssp. sa i a c . Nagysze´na`si
EMBL-4 genomic lib a y was sc eened by using he E
3
ag-
men as a p obe. Two ypes o hyb idizing phage clones we e
isola ed. One g oup hyb idized s ongly o E
3
, and wo hy-
b idizing bands appea ed when he DNA was HindIII-
diges ed. The o he g oup hyb idized weakly o E
3
and only
one HindIII-hyb idizing band was p esen , sugges ing ha
hese phages may con ain pollen- ela ed PG genes. A
l
10
clone belonging o he i s g oup was chosen o cha ac e ize
he en i e al al a PG gene exp essed in symbiosis. This gene
was designa ed MsPG3.
The nucleo ide and he deduced amino acid sequences o he
MsPG3 gene a e deposi ed in Eu opean Molecula Biology
Labo a o y (EMBL da a bank, accession no. Y11118). Anal-
ysis o he PG sequence using he TEST CODE p og am (Ge-
ne ics Compu e G oup, Madison, WI) indica ed he p esence
o ou egions wi h high coding p obabili y and h ee o he s
wi h low coding p obabili y co esponding espec i ely o he
conse ed and nonconse ed egions. This esul sugges ed he
p esence o in ons in he MsPG3 genomic sequence. The exac
posi ion o he p edic ed in ons was de e mined by compa -
ison wi h he consensus-splicing si es o dico yledoneous
plan s (21). This analysis e ealed ha he h ee in ons
ollowed he gene al exonyGT...in on...AGyexon ule.
They a e 278-, 427-, and 94-bp long, beginning a posi ions 127,
717, and 1663, espec i ely. The second and hi d in on-
splicing si es we e la e con i med by sequencing he PCR
p oduc ob ained by ampli ica ion om nodule cDNA using
oligonucleo ide p ime s 420 and 1751. In he egion o MsPG3
co esponding o he E
3
PCR agmen , he e was no in on as
expec ed. Analysis o he deduced coding sequence e ealed
he p esence o an ORF coding o a pu a i e p o ein o 395
aa wi h a p edic ed molecula mass o 42,673 Da and an
isoelec ic poin o 8.15. The sequence a ound he p oposed
s a codon (AACATGGA) was e y simila o he consensus
sequence o he ansla ion s a si e o he dico yledoneous
plan genes Aa(AyC)ATGGC) (22). The homology o o he
PG genes sugges ed ha his ATG may be he s a codon used
in i o. The p o ein encoded by he MsPG3 gene con ains h ee
po en ial N-linked glycosila ion si es acco ding o he consen-
sus sequence Asp-X-Se yTh a amino acid posi ions 138, 178,
and 263. I also con ains 14 cys eine esidues, and 12 o hese
esidues a e conse ed among plan PGs. By using he hy-
d opha y plo , i was shown ha he p edic ed p o ein is
ela i ely hyd ophilic. The N- e minal egion is hyd ophobic
displaying he p ope ies o a signal pep ide (23) and s ongly
sugges pos - ansla ional clea age eleasing a 22-aa signal
pep ide and a ma u e p oduc .
The deduced ORF displayed 52.1, 48.2, 47.0, 36.8, 36.0, and
31.9% iden i y and 71.9, 69.2, 67.0, 59.0, 57.4, and 54.7%
simila i y a he amino acid le el o he p edic ed PG p o eins
o Oeno he a o ganensis (8), obacco (24), B assica napus (25),
oma o abscission-speci ic (26), maize (19), and oma o (27),
espec i ely. The highes conse a ion was ound be ween
amino acids 191 and 277. In his egion, he isola ed MsPG3-
genomic clone showed 91.3% iden i y a he nucleo ide le el o
he E
3
agmen and only 76.7% iden i y o he Pl
1
. These da a
s ongly sugges ha he MsPG3 gene is he gene exp essed in
nodules and Rhizobium-inocula ed oo s in al al a. To u he
suppo his conclusion, we sequenced he PCR p oduc
ob ained by he ampli ica ion o phage
l
3 by using he
unspeci ic p ime s (370–389) and belonging o he g oup
gi ing only one hyb idizing band a e HindIII diges ion. This
agmen showed 98.9% iden i y o he pollen- ela ed Pl1
agmen and only 76.2% iden i y o he co esponding egion
in he MsPG3-genomic clone s udied he e. A M. sa i a pollen
cDNA coding o PG has been ecen ly included in EMBL da a
bank (EMBL accession no. U20431; X. Qiu and L. E ickson,
pe sonal communica ion). Amino acid sequence compa ison
wi h he p edic ed MsPG3 gene p oduc ga e 79.0% simila i y
and 64.7% iden i y bu 96.7% simila i y and 93.4% iden i y
when compa ed wi h he PCR Pl
1
agmen .
Spa ial Exp ession o MsPG3 Du ing Roo Nodule De el-
opmen . To u he analyze he pa e n o exp ession o he
MsPG3 gene du ing nodule de elopmen , he spa ial localiza-
ion o MsPG3 ansc ip s in wild- ype nodules was s udied by
in si u hyb idiza ion. As shown in Fig. 3, se ial sec ions o
4-day-old nodule p imo dia, 7-day-old nodules, and 20 day-old
nodules om al al a plan s we e hyb idized agains an 833-bp
an isense MsPG3 RNA p obe. T ans e sal sec ions o oo s
ha es ed 4 days a e inocula ion and con aining nodule
p imo dia also we e hyb idized o MsPG3 an isense RNA. The
MsPG3 mRNA was p esen in all cells o nodule p imo dia
(Fig. 3C). Sense RNA p obes did no yield hyb idiza ion
signals abo e backg ound le els (Fig. 3 Aand B). In young
nodules, PG ansc ip s we e p esen in he cells o he
me is em and o he in asion zone o zone II (Fig. 3F–H)
di ec ly adjacen o he me is em whe e in ec ion h ead
g ow h and elease o bac e ia occu and in he in e zone II-III
con aining amyloplas s, whe e bac e ial p oli e a ion and en-
la gemen o in ec ed cells ake place. Once di e en ia ed in o
cells o he symbio ic zone (zone III), he MsPG3 ansc ip s
we e no longe p esen (Fig. 3 F–H). The PG ansc ip s we e
de ec able in he in ec ed cells bu no in he highly acuola ed
nonin ec ed cells o young nodules (Fig. 3 Dand E).
To u he con i m he spa ial exp ession pa e n o MsPG3,
we used ano he legume plan , V. hi su a, which also o ms
inde e mina e- ype oo nodules and o which ansgenics
oo s can apidly be ob ained (16). A e ans o ma ion wi h
plasmid pPG97 con aining a usion o a 2.7-kb agmen o he
9690 Plan Biology: Mun˜oz e al. P oc. Na l. Acad. Sci. USA 95 (1998)
MsPG3 gene ups eam o he ATG codon o he gusA-in gene,
composi e V. hi su a plan s ha bo ing ‘‘hai y oo s’’ induced by
Ag obac e ium hizogenes ca ying pPG97 we e ob ained. The
hai y oo s we e inocula ed wi h Rhizobium leguminosa um b .
iciae, and he spa ial dis ibu ion o he GUS ac i i y was
s udied in he nodules o med, as shown in Fig. 4. GUS ac i i y
was de ec ed in nodule p imo dia and in young ansgenic
nodules (Fig. 4A). In ma u e nodules, GUS ac i i y was
isualized in he cells o he in asion zone, whe eas i was ne e
obse ed in cells o he symbio ic zone (Fig. 4Band C). GUS
ac i i y also was de ec ed in ascula issues and a he junc ion
be ween nodules and oo s. Non ansgenic oo nodules did
no show any GUS s aining (da a no shown). The GUS
ac i i y pa e n obse ed in hese ansgenic nodules e ealed
ha he 2.7-kb MsPG3 ups eam sequence was su icien o
di ec he spa ial and empo al exp ession o he epo e gene
ha coincided wi h he exp ession pa e n obse ed in he in
si u expe imen on M. sa i a.
DISCUSSION
We ha e isola ed and cha ac e ized a genomic clone in M.
sa i a (MsPG3) encoding polygalac u onase (PG). We showed
ha his single gene ep esen s a class o PG genes ha is
speci ically exp essed in oo s and nodules induced by R.
melilo i and ha based on sequence and exp ession da a, is
dis inc om he pollen speci ic PG genes.
Sequence analysis o he comple e MsPG3 gene e ealed a
deduced ORF ha displayed 64.7, 52.1, 48.2, 47.0, 36.8, 36.0,
and 31.9% iden i y and 79.0, 71.9, 69.2, 67.0, 59.0, 57.4, and
54.7% simila i y a he amino acid le el o he p edic ed PG
p o eins o M. sa i a,Oeno he a o ganensis, obacco, B assica
napus, oma o abscission-speci ic, maize, and oma o ui ,
espec i ely. The highes conse a ion was ound wi h a M.
sa i a pollen cDNA coding o PG (U20431). The p edic ed
p o ein shows, in addi ion o 12 cys eines highly conse ed
among plan PG enzymes, ou domains conse ed in he
bac e ial, ungal, and plan polygalac u onase genes (24). The
highes homology is p esen be ween amino acids 191 and 277.
This sequence con ains h ee o ou conse ed domains and
includes a his idine esidue, which has been hypo he ized o be
pa o he ca aly ic si e. The encoded pep ide also con ains a
N- e minal-hyd ophobic domain displaying he p ope ies o a
signal pep ide, p esen in all epo ed PGs, and in ol ed in he
anspo o he p o ein ac oss memb anes (23).
Sou he n blo expe imen s showed ha he E
2
and E
3
agmen s displayed a di e en hyb idizing pa e ns agains M.
unca ula DNA as well as M. sa i a spp. sa i a genomic DNA,
FIG. 3. Localiza ion o MsPG3 exp ession in oo nodule p imo dia and nodules. In si u hyb idiza ion was pe o med on ans e sal sec ions
o 4-day-old nodule p imo dia (Aand C), ans e sal sec ions o 7-day-old nodules (B,D, and E), and longi udinal sec ions o 7-day-old (F) and
20-day-old nodules (Gand H). Digoxigenin-labeled MsPG3 sense (Aand B) and an isense (C–H) p obes we e used. (E) A highe magni ica ion
o he bo om pa o D.(H) The same sec ion in Gpho og aphed unde combina ion o ligh and luo escence mic oscopy. As e isks, nodule
p imo dia; s a s, me is ema ic zone; I, in asion zone; S, symbio ic zone; C, in ec ed cells; a ows, in e zone II-III; a ow heads, s a ch g ains in
nonin ec ed cells.
FIG. 4. GUS ac i i y obse a ion in ansgenic V. hi su a oo s and nodules. The ac i i y o he GUS gene di ec ed by he MsPG3-p omo e
egion was isualized in comple e young nodules and nodule p imo dia (A) and longi udinal cu s o ma u e nodules (Band C). Thick a ow, young
nodule; hin a ows, nodule p imo dium; as e isks, me is ema ic zone; a ow heads, in ec ion zone; SZ, symbio ic zone.
Plan Biology: Mun˜oz e al. P oc. Na l. Acad. Sci. USA 95 (1998) 9691

suppo ing he idea ha he e a e a leas wo di e en PG
genes in he Medicago genome. Exp ession o MsPG3 was no
de ec ed in a ious Medicago issues by No he n blo , nei he
in nodules no in Rhizobium inocula ed oo s, indica ing ha
his gene may be exp essed a a e y low le el, i any, o
exp essed only ansien ly. This is consis en wi h he hypo h-
esis ha , i his gene is in ol ed in symbiosis, he p oduc ion
o PG should be inely egula ed o a oid a de ense eac ion
induced by he enzyme p oduc ion and subsequen abo ion o
he in ec ion p ocess (9, 28).
By using RT-PCR analysis, we showed ha he E
2
- ela ed
gene was exp essed in lowe s bu no in oo s o nodules, in
ag eemen wi h ou ea ly No he n esul s, whe eas MsPG3
was exp essed in nodules and oo s inocula ed wi h Rhizobium
bu no in uninocula ed oo s no in non unc ional spon ane-
ous nodules, s ongly sugges ing i s ole in Rhizobium-legume
symbiosis. The exp ession s udies es ic ed o he spo -
inocula ed zone o he al al a oo s showed ha he MsPG3
gene was exp essed as ea ly as 1 day a e inocula ion wi h he
wild- ype R. melilo i s ain, bu no exp ession was obse ed in
uninocula ed oo s o oo s inocula ed wi h he Nod
2
mu an
s ain ZB138, in con as o he expe imen pe o med wi h
he comple e oo sys em. Because he s ain ZB138 is no able
o p oduce Nod ac o s as well as some supe icial compo-
nen s, o he molecules like polysaccha ides o lipopolysaccha-
ides migh be in ol ed in inducing MsPG3 exp ession in
addi ion o he Nod ac o . These da a oge he con i m ou
hipo hesys by which MsPG3 is p oduced by he hos plan and
induced by he mic osymbion , al hough he bac e ial signal
in ol ed in his induc ion had no been de e mined ye .
By using in si u hyb idiza ion, we showed ha MsPG3 is
exp essed e y ea ly du ing he in ec ion p ocess coinciding
wi h he exp ession pa e n o o he ea ly nodulin genes like
Enod12 and Enod5 in inde e mina e nodules (10, 29). This
exp ession pa e n is consis en wi h a ole o he pec ic
enzymes in me is em main enance, in ec ion h ead di ec-
ional g ow h, and cell wall disolu ion, o bac e ial elease
du ing he in ec ion p ocess, and cons i u e he i s molecula
e idence suppo ing he hypo hesis o Fah aeus and Ljungg en
(3) and an Sp onsen e al. (30) who sugges ed a ole o plan
cell wall deg ading enzymes in he in ec ion p ocess, allowing
he pene a ion o he bac e ia. The ac ion o hese enzymes is
p obably equi ed in conjunc ion wi h he Nod ac o s o
conduc an e icien nodule de elopmen .
The analysis o GUS ac i i y di ec ed by he MsPG3-
p omo e egion o ansgenic oo nodules induced by Rhi-
zobium on o composi e V. hi su a plan s con i med he spa ial
and empo al pa e n o exp ession ob ained by he in si u
expe imen . In addi ion, uninocula ed ansgenic oo s o he
composi e plan s showed low le el o epo e gene ac i i y in
la e al oo p imo dia and eme ging la e al oo s, in ag eemen
wi h he PG ac i i y associa ed o oo g ow h de ec ed in
Allium (20). This la e phenomenon could accoun o he
basal exp ession le el obse ed wi h RNA isola ed om he
en i e oo o al al a. These da a demons a ed ha he
p omo e egion o MsPG3 gene used o gusA-in gene is
su icien o co ec ly conduc GUS ac i i y in he he e olo-
gous sys em. The e o e, he epo e usion in combina ion
wi h immunological s udies, will ep esen a use ul ool o
s udy he ole o MsPG3 du ing symbiosis in M. unca ula
ansgenic plan s, as well as o un a el he bac e ial ac o s
in ol ed in he signal ansduc ion pa hway leading o i s
exp ession.
D. G ie son is acknowledged o p o iding he oma o PG p obe. E.
Kondo osi is g a e ully acknowledged o p o iding he M. sa i a
genomic lib a y. We hank M. C espi and T. Coba o hei help wi h
he in si u hyb idiza ion expe imen s and P. Baue o help wi h he
MsEnod12 exp ession expe imen . Resea ch suppo was p o ided by
Comision In e minis e ial de Ciencia y Tecnologia G an s BIO93-
0427 and PB95-1268. J.A.M., C.C., and J.P.H. we e suppo ed by he
Spanish Minis y o Educa ion. J.P.H. also was he ecipien o a Sho -
Te m Eu opean Molecula Biology O ganiza ion ellowship.
1. Nap, J. P. & Bisseling, T. (1990) Science 250, 948–950.
2. Nu man, P. S. (1956) Biol. Re . Biol. P oc. Camb idge Philos. Soc.
31, 109–151.
3. Fah aeus, G. & Ljungg en, H. (1959) Physiol. Plan . 12, 145–154.
4. Paloma es, A., Mon oya, E. & Oli a es, J. (1978) Mic obios 21,
33–39.
5. Callaham, D. A. & To ey, J. G. (1981) Can. J. Bo . 59, 1647–
1664.
6. Ma eos, P. F., Jime´nez-Zu do, J. I., Chen, J., Squa ini, A. S.,
Haack, S. K., Ma ı´nez-Molina, E., Hubbell, D. H. & Dazzo, F. B.
(1992) Appl. En i on. Mic obiol. 58, 1816–1822.
7. Fische , R. L. & Benne , A. B. (1991) Plan Mol. Biol. 42,
675–703.
8. B own, S. M. & C ouch, M. L. (1990) Plan Cell 2, 263–274.
9. Collme , A. & Keen, N. T. (1986) Annu. Re . Phy opa hol. 24,
383–409.
10. Baue , P., C espi, M. D., Sze´csi, J., Allison, L. A., Schul ze, M.,
Ra e , P., Kondo osi, E. & Kondo osi, A. (1994) Plan Physiol.
105, 585–592.
11. Dellapo a, S. L., Wood, J. & Hicks, J. B. (1983) Plan Mol. Biol.
Rep. 4, 19–21.
12. Samb ook, K. J., F i sch, E. F. & Mania is, T. (1989) Molecula
Cloning: A Labo a o y Manual (Cold Sp ing Ha bo Lab. P ess,
Plain iew, NY), 2nd Ed.
13. Sla e , A., Maunde s, M. J., Edwa ds, K., Schuch, W. & G ie son,
D. (1985) Plan Mol. Biol. 5, 137–147.
14. Bochenek, B. & Hi sch, A. M. (1990) Plan Mol. Biol. Rep. 8,
237–248.
15. Szabados, L., Cha ie , B., Kondo osi, A., B uijn, F. J. & Ra e ,
P. (1995) Mol. B eeding 1, 419–423.
16. Quand , H.-J., Pu¨hle , A. & B oe , Y. (1993) Mol. Plan –Mic ob.
In e ac . 6, 699–706.
17. Je e son, R. A, Ka anagh, T. A & Be an, M. W (1987) EMBO
J. 13, 3901–3907.
18. Cap a i, C., Ma ei, B., Basile, M. L., Sal i, G., C escenzi, V., De
Lo enzo, G. & Ce one, F. (1996) Mol. Plan –Mic ob. In e ac . 9,
617–624.
19. Niog e , M. F., Dubald, M., Manda on, P. & Mache, R. (1991)
Plan Mol. Biol. 17, 1155–1164.
20. Pe e o, R., Fa a on, F., Be ini, V., de Lo enzo, G., Ma ini, S.,
Alghisi, P., Ce one, F. & Bon an e, P. (1992) Plan a 188,
164–172.
21. B own, J. (1986) Nucleic Acids Res. 14, 9549–9559.
22. Ca e e , D. R. & Ray, S. C. (1991) Nucleic Acids Res. 19,
3185–3192.
23. Ve ne , K. & Scha z, G. (1988) Science 241, 1307–1312.
24. Tebbu , S. J., Roge s, H. J. & Lonsdale, D. L. (1994) Plan Mol.
Biol. 25, 283–297.
25. Robe , L. S., Alla d, S., Ge s e , J. L., Cass, L. & Simmonds, J.
(1993) Plan Mol. Biol. 23, 1273–1278.
26. Kalai zis, P., Koehle , S. M. & Tucke , M. L. (1995) Plan Mol.
Biol. 28, 647–656.
27. Bi d, C. R., Smi h, C. J. S., Ray, J. A., Mo eau, P., Be an, M. W.,
Bi d, A. S., Hughes, S., Mo is, P. C., G ie son, D. & Schuch, W.
(1988) Plan Mol. Biol. 11, 651–662.
28. Dazzo, F. & Hubbell, D. (1982) in Ni ogen Fixa ion, ed. B ough-
on, W. (Cla endon, Ox o d).
29. Pichon, M., Jou ne , E. P., Dedieu, A., de Billy, F., T uche , G.
& Ba ke , D. G. (1992) Plan Cell 4, 1199–1211.
30. Van Sp onsen, P. C., Bakhuizen, R., an B ussel, A. A. N. &
Kijne, J. W. (1994) Eu . J. Cell Biol. 64, 88–94.
9692 Plan Biology: Mun˜oz e al. P oc. Na l. Acad. Sci. USA 95 (1998)