Na u e Mic obiology | Volume 10 | Ap il 2025 | 958–972 958
na u e mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Pseudomonas sy ingae subpopula ions
coope a e by coo dina ing lagella and
ype III sec e ion spa io empo al dynamics
o acili a e plan in ec ion
Nie es López-Pagán1,7, José S. Ru ián 1, Julien Luneau2,3,
Ma ía-An onia Sánchez-Rome o4, Lau en Aussel 5, Simon an Vlie 2,6,
Ja ie Ruiz-Albe 1 & Ca men R. Beuzón 1
Isogenic bac e ial popula ions can display p obabilis ic cell- o-cell a ia ion
in esponse o challenges. This pheno ypic he e ogenei y can a ec i ulence
in animals, bu i s impac on plan pa hogens is unknown. P e iously, we
showed ha exp ession o he ype III sec e ion sys em (T3SS) o he plan
pa hogen Pseudomonas sy ingae displays pheno ypic a ia ion in plan a.
He e we use low cy ome y and mic oscopy o in es iga e single-cell lagella
exp ession in ela ion o T3SS exp ession, showing ha bo h sys ems
unde go pheno ypic he e ogenei y in i o in apoplas -mimicking medium
and wi hin apoplas ic mic ocolonies h oughou coloniza ion o Phaseolus
ulga is. S ochas ic, spa ial and ime ac o s shape he dynamics o a
pheno ypically di e se pa hogen popula ion ha displays di ision o labou
du ing coloniza ion: e ec o s p oduced by T3SS-exp essing bac e ia ac as
‘common goods’ o supp ess immuni y, allowing mo ile lagella-exp essing
bac e ia o inc ease and lea e in ec ed issue be o e nec osis. These
esul s showcase he mechanisms o bac e ial specializa ion du ing plan
coloniza ion in an en i onmen ally and ag icul u ally ele an sys em.
Flagella mo ili y is an impo an ai o en i onmen al adap a ion.
Membe s o he genus Pseudomonas can colonize soil, plan o animal
issues. The Pseudomonas sy ingae species complex includes mos o
hose pa hogenic o plan s, collec i ely causing diseases in o e 500
species, many ag icul u ally ele an 1. P. sy ingae can li e on he lea
su ace, bu mos a ou endophy ic coloniza ion, using wounds and
s oma a o en e he lea . Flagella mo ili y con e s ad an ages o epi-
phy ic popula ions o P. sy ingae p . sy ingae2 and acili a es lea en y
o many P. sy ingae pa ho a s3,4. Inside he lea , P. sy ingae colonizes
he in e cellula spaces (apoplas ). Once inside, lagella mo ili y is no
equi ed o sys emic sp ead3, while lagellin, he main componen o
he pilus, igge s immuni y
5
. Fi ingly, lagella exp ession is down-
egula ed in he apoplas 6–8.
Flagella biosyn hesis is depic ed as a de e minis ic p og amme
wi h a complex and ie ed egula o y hie a chy, whe e p omo e s a e
sequen ially ac i a ed, main aining his s a e h ough ac i e g ow h.
Recei ed: 19 Ma ch 2024
Accep ed: 19 Feb ua y 2025
Published online: 2 Ap il 2025
Check o upda es
1Ins i u o de Ho o u icul u a Sub opical y Medi e ánea “La Mayo a”, Uni e sidad de Málaga-Consejo Supe io de In es igaciones Cien í icas
(IHSM-UMA-CSIC), Málaga, Spain. 2Depa men o Compu a ional Biology, Uni e si y o Lausanne, Lausanne, Swi ze land. 3Depa men o Fundamen al
Mic obiology, Uni e si y o Lausanne, Lausanne, Swi ze land. 4Depa amen o de Mic obiología y Pa asi ología, Facul ad de Fa macia, Uni e sidad de
Se illa, Se illa, Spain. 5Aix Ma seille Uni , CNRS, LCB UMR7283, IMM, Ma seille, F ance. 6Biozen um, Uni e si y o Basel, Basel, Swi ze land. 7P esen
add ess: Depa men o Chemical Biology and Gene ics, Cen e o he Region Haná o Bio echnological and Ag icul u al Resea ch, Facul y o Science,
Palacký Uni e si y, Olomouc, Czech Republic. e-mail: [email p o ec ed]
Na u e Mic obiology | Volume 10 | Ap il 2025 | 958–972 959
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
HIM medium, mos bac e ia exp ess liC::GFP3 (Flagella
ON
) wi h consi-
de able cell- o-cell a ia ion, bu a subpopula ion does no (Flagella
OFF
)
and can be isualized only by memb ane s aining, o e lapping
wi h non- luo escen bac e ia in FC analyses (Fig. 1a). Exp ession
o GFP3 om a cons i u i e p omo e is homogeneous22, uling ou
exp ession noise as he liC::GFP3 sou ce o he e ogenei y.
In keeping wi h lagella ole in acili a ing lea en y2–4, lea -su ace
bac e ia a e mos ly Flagella
ON
. None heless, images o bean (Phaseolus
ulga is) lea s sp ayed wi h liC::GFP3 bac e ia cons i u i ely exp ess-
ing eCFP show ha lagella exp ession is also he e ogeneous on lea
su aces (Ex ended Da a Fig. 2). Fou days pos p essu e inocula ion
(dpi), bac e ia eco e ed om he apoplas display lowe a e age
exp ession han media-g own popula ions (Fig. 1b), ag eeing wi h
ansc ip omic da a
6–8
. Howe e , mos apoplas -ex ac ed bac e ia
a e Flagella
ON
(Fig. 1a). Cell- o-cell di e ences in luo escence we e
mo e p onounced among apoplas -ex ac ed bac e ia han in media,
wi h a wide dis ibu ion e ealed by FC (Fig. 1a ( igh ),c). This a i-
abili y is also appa en in plan a, wi h apoplas ic mic ocolonies o
liC::GFP3 eCFP bac e ia showing he e ogeneous GFP3 luo escence
o e laid wi h homogeneous eCFP luo escence dis ibu ion (Fig. 1d).
Homogeneous eCFP exp ession was epo ed p e iously23.
LB popula ions we e so ed by luo escence-ac i a ed cell so ing
based on GFP luo escence (Fig. 1e), en iching popula ions wi h low
o high liC::GFP3 exp ession (Fig. 1e igh ) o use in swimming assays.
Since so ing causes physical lagella damage, i can educe mo ili y
di e ences be ween so ed subpopula ion pai s and inc ease a i-
a ion be ween so ing e en s. Thus, we compa ed mo ili y be ween
high/low subpopula ions sepa a ed in he same so ing e en (Fig. 1 ):
we calcula ed he a io be ween swimming halo diame e s o high
GFP3- and low GFP3-exp essing subpopula ions (pe so ed e en ).
High- o-low mo ili y a ios o so ed LB popula ions we e signi ican ly
highe han 1.0, indica ing ha liC exp ession posi i ely co ela es
wi h mo ili y (Fig. 1 ). Simila esul s we e ob ained using HIM popula-
ions (Fig. 1g,h). These esul s indica e ha lagella a e pheno ypically
he e ogeneous in P. sy ingae and show ha , despi e lagellin being an
immune elici o , pa o he apoplas ic popula ion exp esses lagella.
Flagella and T3SS show coun e - egula ion a he cellula le el
Nex , we used a s ain ca ying ansc ip ional usions liC:: dT ( o
dToma o; AAV52169 1) and h pL::GFP3 (WP_004656240.1) o e alu-
a e T3SS and lagella exp ession simul aneously. Popula ion p o iles
o liC:: dT a e simila o hose o liC::GFP3 (Ex ended Da a Fig. 3a,b).
Po en ial in e e ences be ween luo opho es ( ha is, bleed h ough
leading o alse posi i es o Fö s e esonance ene gy ans e (FRET)
leading o alse nega i es) we e uled ou (Ex ended Da a Fig. 3c,d).
Bo h sys ems we e exp essed he e ogeneously in HIM (Fig. 2a,b (le )
and Ex ended Da a Fig. 3e). FC showed wo subpopula ions, o med
by bac e ia a ou ing exp ession o T3SS (T3SSON/FlagellaOFF) o liC
Howe e , single-cell s udies e ealed s ochas ic ac i a ion pulses
and bimodal pa e ns in he lagella sys em o Salmonella en e ica,
wi h clonal popula ions displaying pheno ypic he e ogenei y
9–13
. In
P. sy ingae, egula ion o lagella syn hesis emains mos ly uncha -
ac e ized, bu gene a angemen and p omo e conse a ion wi h
P. pu ida suppo s a simila hie a chy o ansc ip ional egula ion
14
,
wi h FleQ a he op15.
Apoplas ic P. sy ingae use a ype III sec e ion sys em (T3SS) o
in oduce e ec o s (T3Es) in o he hos -cell cy osol. T3Es supp ess
plan immuni y including lagellin- igge ed esponses, allowing
bac e ial p oli e a ion and disease16. T3SS gene exp ession equi es
H pL17. We showed ha a Δh pL mu an displays inc eased swimming
mo ili y in apoplas -mimicking H p-inducing medium (HIM), while
a mu an lacking H pV, a ep esso o h pL exp ession18, displays
educed mo ili y19, sugges ing a egula o y link be ween T3SS
and lagella exp ession. La e , we es ablished ha P. sy ingae T3SS
genes display pheno ypic he e ogenei y du ing plan coloniza ion
and bis able exp ession wi h e e sible ON and OFF s a es in HIM20.
He e we show ha lagella gene exp ession displays pheno ypic
he e ogenei y in P. sy ingae, pa icula ly wi hin he plan apoplas .
While a e age exp ession le els o lagellin-encoding liC is down-
egula ed in he apoplas ic popula ion, as p e iously desc ibed6–8, a
subpopula ion exp esses high liC le els du ing coloniza ion. Analysing
lagella and T3SS exp ession simul aneously, we ound all possible
single-cell ON/OFF combina ions, e ealing he pheno ypic complexi y
a ising amid clonal popula ions. None heless, T3SSON/FlagellaOFF and
T3SSOFF/FlagellaON subpopula ions we e gene ally mo e abundan ,
sugges ing coun e - egula ion be ween hese loci. Exp ession o hese
sys ems impac s bac e ial g ow h a he popula ion and/o cellula
le el. Analysis o he spa ial dis ibu ion o exp ession wi hin apoplas ic
mic ocolonies shows ha T3SS
ON
bac e ia a e mo e abundan ea ly
in he in ec ion and close o he hos -cell su ace, while FlagellaON
bac e ia become mo e equen la e and u he away om he
hos cell. Following bac e ial apoplas ic mul iplica ion, FlagellaON
bac e ia ac i ely exi he in ec ed issue be o e he onse o nec o ic
symp oms. Ou esul s suppo a di ision o labou model wi h T3SS
ON
sec e ed e ec o s ac ing as common goods, allowing FlagellaON
bac e ia o inc ease and exi be o e issue collapse.
Resul s
Flagella exp ession in P. sy ingae is he e ogeneous
We gene a ed a ansc ip ional usion o GFP3 (AAB18957.1) o
lagellin-encoding liC (WP_002554301.1) wi hin i s genome loca ion,
p ese ing con ex and unc ion wi hou a ec ing i ness o i ulence
(Ex ended Da a Fig. 1) in model bean pa hogen P. sy ingae p . phaseo-
licola 1448A21. Using con ocal lase scanning mic oscopy (CLSM)
and low cy ome y (FC), we moni o ed liC::GFP3 exp ession in i o
(Fig. 1a). In ich lysogeny b o h (LB) medium and apoplas -mimicking
Fig. 1 | Flagella display pheno ypic he e ogenei y in Pseudomonas sy ingae.
a, Le : CLSM images and FC analysis o bac e ia ca ying liC::GFP3 g own in LB o
HIM medium, o apoplas ex ac ed om lea apoplas s a 4 dpi. FM4-64 channel
(g eyscale) display luo escence om memb ane s aining. Scale ba s, 2 µm.
Righ : FC da a ob ained om liC::GFP3 bac e ia a e ep esen ed using a bi a y
uni s in loga i hmic scale. Non-GFP g aphs show au o luo escence le els o wild
ype as e e ence o OFF subpopula ions ( e ical lines lea e 99% non-GFP da a
o he le ). Mic oscopy and FC panels show ypical esul s (n = 14 independen
expe imen s). b, Mean GFP luo escence in ensi y o liC::GFP3-ca ying bac e ial
popula ions (n = 14) om FC da a desc ibed in a. c, Robus coe icien o a ia ion
(RCV) ob ained om FC da a desc ibed in a. b and c display indi idual sample
alues, and sample se s displaying di e en le e s a e signi ican ly di e en
as de e mined using one-way analysis o a iance (ANOVA) ollowed by Tukey’s
mul iple compa isons es (P < 0.05); all compa isons showed P < 0.0001. Lines
indica e he means o all indi idual da a. E o ba s ep esen s.e.m. d, CLSM
images o liC::GFP3 bac e ia exp essing a cons i u i e eCFP gene g owing in
mic ocolonies in plan a a 3 dpi. GFP channel shows he e ogeneous dis ibu ion
o liC exp ession. CFP channel shows homogeneous dis ibu ion o CFP (n = 3).
Red in all panels co esponds o chlo oplas au o luo escence. Scale ba s, 5 µm.
In a and d, con as and b igh ness we e adjus ed o imp o e isualiza ion
and kep cons an ac oss samples and channels. e,g, His og ams showing GFP
luo escence s cell coun om liC::GFP3 o e nigh LB-g own cul u es (e) o 24 h
HIM-g own cul u es (g) be o e (le ) and a e ( igh ) so ing. Red lines indica e
ga ing used o sepa a e liC::GFP3 bac e ia acco ding o luo escence in ensi y.
,h, Rela i e a io be ween halo diame e s in swimming pla es seeded wi h cells
displaying high GFP exp ession, di ided by he diame e s o halos ob ained om
seeding cells displaying low GFP exp ession, esul ing om he same so ing
e en (iden i ied wi h di e en shades) om cul u es g own in LB ( ) o HIM
(h) (n = 9 as shown). Lines indica e he means o all indi idual da a. E o ba s
ep esen s.e.m. As e isks indica e signi ican di e ences es ablished by an
unpai ed wo- ailed - es (P < 0.05). P alues a e indica ed.
Na u e Mic obiology | Volume 10 | Ap il 2025 | 958–972 960
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
(T3SSOFF/FlagellaON) (Fig. 2a,b (le )). Al hough hese combina ions we e
mo e abundan , a con inuous dis ibu ion wi h bac e ia exp essing
bo h (T3SS
ON
/Flagella
ON
) o none (T3SS
OFF
/Flagella
OFF
) was obse ed
(Fig. 2a,b (le ) and Ex ended Da a Fig. 3e). Apoplas -ex ac ed bac-
e ia (4 dpi) also displayed all pheno ypic combina ions, al hough
T3SS
ON
/Flagella
OFF
bac e ia we e mo e abundan (Fig. 2a,b ( igh )).
As epo ed
20
, apoplas ic popula ions did no display bimodal GFP
dis ibu ion (Fig. 2a,b and Ex ended Da a Fig. 3 ). Analysis o T3SS
e ec o gene hopAB1 ( e . 20)(WP_011282445.1) ende ed simila esul s
(Fig. 2c,d and Ex ended Da a Fig. 3e, ).
High
5 µm
High
GFP
luo escence
in ensi y
Plan
ab
LB
HIM c
d
GFP FM4-64 Me ged
GFP CFP Me ged
liC::GFP3 eCFP
ab
c
a
b
c
P = 0.0412
*
e
GFP luo escence in ensi y
Coun
HIM
Coun
Low
g
GFP luo escence in ensi y
Low
LB
Low
Cell size
liC::GFP3
Non-GFP
GFP luo escence
Cell size
Non-GFP
Cell size
Non-GFP
liC::GFP3
liC::GFP3
OFF
ONOFF
OFF
ON
ON
P = 0.0112
*
h
High Low High
GFP
luo escence
in ensi y
101102103104105
1,200
Mean
LB HIM Plan
LB HIM Plan
Fluo escence in ensi y
mean (A.U.)
Robus coe icien o a ia ion
High/low a io
High/low a io
1,000
800
600
400
200
150
100
50
0
0
101102103104105
101102103104105
50
40
30
20
10
0
101102103
6
4
2
0
0
1
2
3
4
High/low
LB
101102103
6
4
2
0
101102103
101102103
101102103
101102103
3,0K
400
300
200
100
0
0
0.5
1.0
1.5
300
200
100
0
2,0K
1,0K
0
HIM
High/low
5 µm5 µm
RCV
Na u e Mic obiology | Volume 10 | Ap il 2025 | 958–972 961
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Exp ession o T3SS o lagella ca ies g ow h penal ies
Exp ession o Salmonella pa hogenici y island 1 (SPI1) T3SS causes
g ow h delays, which ha e been p oposed o d i e pheno ypic
he e ogenei y24,25. We used compe i i e index (CI) assays (Fig. 3a) o
compa e g ow h o a Δh pL mu an and wild- ype bac e ia (Fig. 3b).
CIs show ha Δh pL mu an s ou g ow wild- ype bac e ia in HIM
(CI = 1.75 ± 0.15), bu no in ich medium (CI = 1.07 ± 0.08), whe e T3SS
genes a e no exp essed. G ow h a es o Δh pL bac e ia a e also highe
in HIM (Fig. 3c). In plan a, wild- ype bac e ia ou g ow a de i a i e con-
s i u i ely exp essing H pL (CI = 0.67 ± 0.06) (Fig. 3d). Cons i u i e
exp ession o H pL esul s in o e exp ession o T3SS genes
20
, suppo -
ing ha T3SS exp ession ca ies a g ow h penal y.
Flagella o en ca ies g ow h penal ies26–28. Howe e , p e ious CI
assays did no show signi ican g ow h di e ences o Δ liC bac e ia in
P. sy ingae29. Fi ingly, we did no obse e signi ican g ow h di e -
ences o Δ liC o Δ leQ (WP_011169094.1) (lacking lagella egula o
FleQ) mu an s based on CI (LB: CI = 1.08 ± 0.22 and 1.00 ± 0.13, espec-
i ely; HIM: CI = 1.21 ± 0.22 and 1.31 ± 0.13, espec i ely), al hough
a ia ion in LB Δ liC CI was la ge (Fig. 3b), o g ow h a es (Ex ended
Da a Fig. 4a). Howe e , a Δh pL Δ liC mu an ema kably ou g ows wild
ype, wi h CI alues (LB: CI = 4.195 ± 0.9975: HIM: CI = 25.79 ± 5.156)
4–15 old hose ob ained o single Δh pL mu an (Fig. 3b). This sup-
po s he no ion ha lagella exp ession ca ies a cos , a leas when
simul aneously exp essed wi h he T3SS. Cons i u i e exp ession o
leQ om pFleQ led o se e e g ow h delays (Ex ended Da a Fig. 4a),
al hough pleio opic e ec s could no be uled ou gi en he se e i y
o he pheno ype.
In plan s ea ed wi h lagellin o oid di e ences associa ed wi h
lagellin-media ed immuni y, CI assays e ealed g ow h ad an ages
o Δ liC mu an s (CI = 1.88 ± 0.34) (Fig. 3d). Howe e , we canno ule
ou ha Δ liC mu an ad an ages in lagellin- ea ed plan s may s em
om o he easons.
Since T3SS and lagella exp ession a e bo h highly he e ogene-
ous, popula ion-le el assays could be noisy and may lead o unde -
es ima ing g ow h penal ies. Thus, we analysed hese impac s a he
single-cell le el using HIM-aga pads and ime-lapse mic oscopy.
In his se ing, hopAB1::GFP3 ac i a ion akes place oo la e in he
expe imen o allow o con iden quan i ica ion (Supplemen a y
Videos 1–4). Howe e , ac i a ion o liC::GFP3 akes place ea lie
(Supplemen a y Videos 5 and 6, and Fig. 3e). Time-lapse analyses
show ha liC exp ession nega i ely co ela es wi h bac e ial
g ow h (Fig. 3 ). Spli ing he liC::GFP3 popula ion in o wo using
he median luo escence in ensi y alue, we de ec ed a 6.5% g ow h
penal y o he highe - han-median- luo escence hal compa ed
wi h he lowe hal (Fig. 3g). Cons i u i e exp ession o eGFP does
no nega i ely co ela e wi h g ow h (Ex ended Da a Fig. 4b– ).
Pheno ypic specializa ion du ing mic ocolony de elopmen
Apoplas ic FlagellaON/T3SSOFF subpopula ions a e po en ially
mo e ulne able o plan de ences since hey igge bu canno
supp ess immuni y. We used p opidium iodide which binds o
memb ane-comp omised (no iable) bac e ia o de e mine any co -
ela ion be ween pheno ypic a ian s and apoplas killing a es
30,31
.
No signi ican di e ences we e obse ed be ween dead/li e a ios
o Flagella o T3SS ON s OFF subpopula ions (Ex ended Da a
Fig. 5), suppo ing ha Flagella
ON
/T3SS
OFF
cells a e p o ec ed om plan
esponses, pe haps by T3SS
ON
bac e ia supp essing de ences in ans.
We es ablished ha T3SS mu an bac e ia may p oli e a e up
o wild- ype le els when g owing wi hin he same mic ocolony as
wild- ype bac e ia22, while wild- ype and T3SS mu an bac e ia g ow-
ing in sepa a e mic ocolonies de elop acco ding o hei espec i e
abili ies o supp ess local immuni y. The e o e, ans complemen a-
ion be ween T3SS
ON
and Flagella
ON
/T3SS
OFF
bac e ia would equi e
pheno ypic a ian s o ise du ing g ow h o he mic ocolony. Fluo-
escence dis ibu ion o liC:GFP3 in plan a shows ha o lagella, ON/
OFF a ian s appea closely loca ed wi hin he mic ocolony (Fig. 1d).
Close examina ion o liC::GFP3 dis ibu ion (Supplemen a y Video 7)
e eals a eas wi h s onge luo escence, sugges ing a esponse
o local s imuli (o siblings’ inhe i ance o an exp ession pa e n), bu
also isola ed bac e ia displaying GFP luo escence s ikingly di e en
om hose o nea by pee s, in keeping wi h s ochas ic di e ences.
We ollowed exp ession o T3SS and lagella genes du ing he
de elopmen o apoplas ic mic ocolonies, analysing lea es in il a ed
wi h hopAB1::GFP3 liC:: dT o h pL::GFP3 liC:: dT bac e ia a di e en
imepoin s (Fig. 4a,b). S ochas ic T3SS he e ogenei y was de ec ed
a
b
h pL::GFP3 liC:: dT
HIM
h pL::GFP3 liC:: dT
Plan
HIM Plan
liC:: dT
dToma o luo escence
GFP luo escence
h pL::GFP3
GFP luo escence
h pL::GFP3
liC:: dT
dToma o luo escence
T3SSOFF/FlagellaON
T3SSON/FlagellaOFF
T3SSON/FlagellaON
T3SSOFF/FlagellaON
T3SSON/FlagellaOFF
T3SSON/FlagellaON
hopAB1::GFP3 liC:: dT hopAB1::GFP3 liC:: dT
c
d
Plan
HIM Plan
liC:: dT
dToma o luo escence
GFP luo escence
hopAB1::GFP3
GFP luo escence
hopAB1::GFP3
liC:: dT
dToma o luo escence
HIM
103
102
101
103
102
101103
102
101
103
102
101
103
102
101
103
102
101103
102
101
103
102
101
Fig. 2 | Flagella and T3SS exp ession display independen he e ogeneous
exp ession a he single-cell le el. a, CLSM images o dual- epo e h pL::GFP3
liC:: dT s ain g own in HIM o ex ac ed om lea apoplas s a 4 dpi. Scale
ba s, 2 µm. Con as and b igh ness we e adjus ed o imp o e isualiza ion
and used h oughou . b, FC analysis o bac e ia desc ibed in a. G aphs show
GFP luo escence in ensi y s ha o dToma o. Resul s shown in he igu e
a e ep esen a i e o 3 o mo e independen expe imen s. c, CLSM images o
dual- epo e hopAB1::GFP3 liC:: dT s ain g own in HIM o ex ac ed om lea
apoplas s a 4 dpi. Scale ba s, 2 µm. Con as and b igh ness we e adjus ed o
imp o e isualiza ion and used h oughou . d, FC analysis o bac e ia desc ibed
in c. Resul s shown in he igu e a e ep esen a i e o a leas 3 independen
expe imen s.
Na u e Mic obiology | Volume 10 | Ap il 2025 | 958–972 962
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
a
b c d
Wild ype
Mu an
1:1 mix
Incuba e
o/n 28 °C
No Ab Ab
Inocula e
plan lea
4 dpi
In i o CI In plan a CI
CI =
c. .u. mu an /c. .u. wild ype (ou pu )
c. .u. mu an /c. .u. wild ype (inpu )
Compe i i e index alue
No
di e ence
Mu an
inc ease
Mu an
dec ease
1
0.1
10
Pla e dilu ions on o
g
e
0 h 12 h8 h4 h
2 h
liC::GFP3
N = 1,263 cells
500
250
0
750
1,000
1,250
1,500
0
1,750
G ow h a e (µm h−1)
0.1 0.2 0.3 0.4 0.5
R = –0.17
P = 1.03 × 10–9
Fluo escence in ensi y (a.u.)
2,000
liC::GFP3
Low GFP High GFP
0.2
0.1
0
0.3
0.4
0.5
****
G ow h a e (µm h−1)
0.6
0 h 8 h8 h4 h2 h
∆h pL
s WT
∆ liC
s WT
∆ leQ
s WT
∆h pL
P = 0.0006
pH pL
s WT
∆ liC
s WT
P < 0.0001
P = 0.0161
P = 0.0054
∆h pL∆ liC
s WT
P = 0.0086
P = 0.0329
HIM LB HIM LB HIM LB HIM LB Pph
0.1
1
10
100 0.15
0.10
0.05
0
Compe i i e index alue
***
**
*
Plan Plan (+ lg22)
0.1
1
10
Compe i i e index alue
G ow h a e
****
**
*
5 µm 5 µm 5 µm5 µm 5 µm
5 µm
5 µm
5 µm
5 µm
5 µm
Na u e Mic obiology | Volume 10 | Ap il 2025 | 958–972 963
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
h oughou , while zonal dis ibu ion o lagella and T3SS luo escence
wi hin he mic ocolonies changed o e ime. One-dpi mic ocolonies
displayed he e ogeneous exp ession o T3SS, bu no lagella exp es-
sion, suppo ing he idea ha lagella is u ned o in ea ly in ec ion
s ages6–8. Two-dpi mic ocolonies displayed o ange and yellow ones
consis en wi h exp ession o bo h sys ems. By 3–4 dpi, p edomi-
nan ly ed ( liC:: dT) o g een (hopAB1::GFP3 o h pL::GFP3) zones wi h
ho oughly o e laid he e ogenei y becomes he no m (Fig. 4a,b).
Zones wi h highe exp ession o hopAB1::GFP3 gene ally display lowe
exp ession o liC:: dT and ice e sa, al hough he e a e also o ange
and yellow spo s indica i e o T3SSON/FlagellaON bac e ia (Fig. 4a,b).
Mos ema kably, his zonal dis ibu ion shows a consis en spa ial
pa e n: bac e ia closes o he lowe epide mis (Fig. 4c–e), o o a
spongy mesophyll hos cell (Fig. 4a ( igh ), ,g), a e p edominan ly g een,
ha is, T3SSON, whe eas hose u he away om he hos -cell su ace,
ha is, he inne mos o u he mos pa s o he mic ocolonies, a e p e-
dominan ly ed, ha is, FlagellaON (Fig. 4, and Supplemen a y Videos 8
and 9). Al hough he T3SSON zone in he p oximal side a ies in wid h,
i is consis en ly wide han a single bac e ial laye , suppo ing he
idea ha he signal(s) ac i a ing T3SS exp ession include di usible
molecules.
FlagellaON bac e ia ac i ely exi in ec ed issues
The inc ease o Flagella
ON
bac e ia in olde mic ocolonies sugges s a
unc ion a la e in ec ion s ages. We ook lea es a 1 dpi, 2 dpi (asymp-
oma ic), 5 dpi (onse o mac oscopic symp oms) and 7 dpi ( ully symp-
oma ic) and subme ged hem in o a s e ile MgCl
2
solu ion o 30 min.
We used CLSM o e alua e hopAB1::GFP3 and liC:: dT exp ession in
hose bac e ia ound in he solu ion. Bac e ia exi ed e en in ac 1 dpi
lea es. Sho e incuba ions also allowed bac e ial eco e y bu no
enough o eliable quan i ica ion. We compa ed T3SS o liC exp es-
sion be ween bac e ial popula ions exi ing he lea and whole-lea
popula ions ob ained by mechanical ex ac ion om eplica e lea es
wi hou incuba ion (Fig. 5 and Ex ended Da a Fig. 6). Exi ing and o al
apoplas -ex ac ed popula ions showed simila hopAB1::GFP3 exp es-
sion p o iles a 1 dpi, 2 o 7 dpi (Fig. 5a,b and Ex ended Da a Fig. 6), while
signi ican ly ewe T3SS
ON
bac e ia we e de ec ed in he exi ing ac ion
by 5 dpi (Ex ended Da a Fig. 6). In con as , bac e ia exi ing 1 dpi lea es
we e signi ican ly en iched in FlagellaON (Fig. 5a,b), e en hough o al
apoplas -ex ac ed bac e ial popula ions showed almos no FlagellaON
bac e ia (Figs. 4 and 5a, and Ex ended Da a Fig. 6). This suppo s he
no ion o Flagella
ON
bac e ia being mo e e icien in ac i ely exi ing he
lea in humid condi ions. We con i med his no ion using an expe i-
men al se ing mimicking ain o hea y dew, by sp aying he in ec ed
lea es e e y 5 min o 30 min and collec ing he d ipping solu ion o
mic oscopic examina ion (Fig. 5a bo om panels). The bias owa ds
Flagella
ON
bac e ia in he exi ing popula ions de ec ed a 2 and 5 dpi
and p og essi ely dec easing un il hey a e no longe de ec ed a 7 dpi
(Fig. 5a,b and Ex ended Da a Fig. 6), when lea in eg i y has been los ,
sugges s ha lagella mo ili y is no only impo an o bac e ial en y
in o he lea bu also o as and ac i e exi o apoplas -g own bac e ia
be o e nec osis.
Discussion
Bac e ial encoun e s wi h en i onmen al changes lead o gene ac i-
a ion o en assumed o happen in a non- luc ua ing homogeneous
manne . Howe e , s udies a e aging popula ion da a lose cell- o-cell
a ia ion, and an inc easing numbe o ai s ha e been shown o mani-
es di e en ly ac oss clonal popula ions. Such pheno ypic a ia ion
may e lec mic oen i onmen al di e ences bu may also be caused
Fig. 4 | S ochas ic and spa ially s uc u ed ime-cou se dis ibu ion o
lagella and T3SS exp ession wi hin apoplas ic mic ocolonies. a, CLSM
Z-s ack compila ion a di e en dpi showing apoplas ic mic ocolonies o
hopAB1::GFP3 liC:: dT bac e ia. All images examined (n = 5 o mo e) pe
biological eplica e (n = 2) and independen expe imen (n = 2 o 1 and 2 dpi;
n = 4 o 3 and 4 dpi). b, Images om a including o hogonal p ojec ions.
The colo ed lines wi hin he cen al images indica e he planes ha gene a e
each o hogonal sec ion: g een lines indica e he sec ion ha is displayed in
he g een-boxed images (images abo e pe ime poin ), ed lines indica e he
sec ion ha is displayed in he ed-boxed images (images on he igh pe ime
poin ). In g een- o ed-boxed images, he blue line co espondes o he sec ion
displayed in he cen al image. Scale ba s, 5 µm. c, Schema ic lea 3D s uc u e
wi h close-up showing loca ion o apoplas ic mic ocolonies. Posi ion and
o ien a ion o Z-s ack acquisi ion a e indica ed; 1–3 ep esen h ee planes o a
bac e ial mic ocolony g owing om he lowe epide mis inne su ace inwa ds
(1 closes o cell su ace and 3 u hes ) as shown in d and Supplemen a y Video 8.
Plane 4 cu s h ough a bac e ial mic ocolony g owing w apped a ound a spongy
mesophyll cell (SMC) (as mic ocolony and ame shown in and Z-s ack shown
in Supplemen a y Video 9). d, F ames o a Z-s ack acquisi ion aken om an
apoplas ic mic ocolony o h pL::GFP3 liC:: dT bac e ia a 4 dpi. Con as and
b igh ness we e adjus ed o imp o e isualiza ion and kep cons an h oughou .
Scale ba s, 10 µm. Lowe epide mis and mesophyll indica e ela i e posi ion o
ames wi hin he lea . e, Schema ic dis ibu ion o luo escence o liC:: dT and
T3SS::GFP3 wi hin he ype o apoplas ic mic ocolonies indica ed in c (1–3) and
shown in d: planes closes o lowe epide mis (abaxial side) show p edominan ly
g een bac e ia and hose u hes p edominan ly ed, wi h in e media e plane
displaying yellow. He e ogenei y is obse ed wi hin each plane. , Z-s ack ame
aken om an apoplas ic mic ocolony o hopAB1::GFP3 liC:: dT bac e ia (Z-s ack
in Supplemen a y Video 9) om expe imen s desc ibed in a. Scale ba s, 20 µm.
g, Schema ic luo escence dis ibu ion o liC:: dT T3SS::GFP3 bac e ia wi hin
apoplas ic mic ocolonies o he ype indica ed in c (4) and shown in : bac e ia
closes o he cell a e p edominan ly g een, u ning o yellow and ed as hey ge
u he away. He e ogenei y is displayed h oughou . In e and g, legend indica es
g een as T3SSON/ FlagellaOFF bac e ia, ed as T3SSOFF/ FlagellaON and yellow/o ange
as bac e ia exp essing bo h o di e en le els.
Fig. 3 | Flagella and T3SS exp ession impac on bac e ial g ow h. a, Le :
compe i i e index assay o e iew: inocula con aining 1:1 mixes o wild- ype and
mu an (o plasmid-ca ying) LB-g own cul u es a e used o inocula e lea es
(in plan a CI) o o s a esh cul u es (in i o CI). Resul ing mixed cul u es
and apoplas -ex ac ed bac e ia (4 dpi) a e se ially dilu ed and pla ed wi h o
wi hou an ibio ics o c. .u. de e mina ion (ou pu ). Righ : CI is calcula ed as
mu an - o-wild- ype c. .u. a io in ou pu di ided by a io in inpu . CIs di e en
om 1.0 indica e s ains g owing signi ican ly di e en ly: mu an ou g owing
wild ype (CI > 1.0) o ice e sa (<1). b, CIs o mu an s s wild- ype s ain (WT)
in HIM o LB (n = 3–9 as shown). c, HIM g ow h a es o Δh pL mu an and wild
ype. d, In plan a CIs o wild ype cons i u i ely exp essing h pL (pH pL)
o Δ liC, s WT a 4 dpi (n = 10 as shown). Plan s o Δ liC CI p e ea ed wi h lg22.
In b–d, da a show mean ± s.e.m. Indi idual da a o each biological eplica e
shown; di e en shades iden i y independen expe imen s, and esul s
wi h as e isks a e signi ican ly di e en om 1.0 (non-pa ame ic wo-sided
S uden ’s - es ). Signi ican P alues indica ed. e, Time-lapse images o
liC::GFP3 bac e ia g owing on aga pads: phase-con as channel ( op),
GFP luo escence (bo om). Con as and b igh ness adjus ed o imp o e
isualiza ion and kep cons an ac oss ames. , Co ela ion be ween g ow h
a es and luo escence in ensi y o indi idual liC::GFP3 cells. Each poin
ep esen s a single cell, wi h liC exp ession and g ow h a e a e aged o e
cell li e ime. G ey-shaded a ea shows 95% con idence in e al a ound linea
eg ession model i . Co ela ion es ed wi h Pea son’s co ela ion coe icien
using wo-sided es (R = −0.17; P = 1.03 × 10−9). g, G ow h a es o high-GFP- and
low-GFP-signal cells (de e mined by spli ing he popula ion using median
luo escence in ensi y) show signi ican ly slowe g ow h in high-GFP cells
( wo-sided Mann–Whi ney U- es , P = 3.6 × 10−8). Da a p esen ed as boxplo s wi h
i s and hi d qua iles as bo om and op bounds, median shown as ho izon al
line in he box, and 1.5 in e qua ile ange shown as whiske s. Ou lie s no shown.
G ey iolin plo shows dis ibu ion o all da a poin s summa ized in boxplo .
Whi e diamonds show mean g ow h a e o all cells conside ed.
Na u e Mic obiology | Volume 10 | Ap il 2025 | 958–972 964
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
by s ochas ic cellula changes (gene exp ession noise) and pa icula
egula o y ci cui s (enhancing noise)32.
P. sy ingae p oduce pola lagella
33,34
upon ac i a ion o FleQ
15
unde a egula o y cascade ha is mos ly uncha ac e ized. Gene
a angemen and p omo e mo i s conse a ion wi h P. pu ida14,
which also p oduces pola lagella, suppo a simila o ganiza ion.
None heless, we show ha P. sy ingae lagella egula ion sha es an
impo an ea u e wi h ha o pe i ichous Salmonella: s ochas ic
cell- o-cell a ia ion (Fig. 1 and Ex ended Da a Figs. 1–3)9–13,35,36. Bis able
ac i a ion o Salmonella lagella is es ablished by wo mechanisms
9–13
:
(1) a double-nega i e eedback loop and (2) a de elopmen al check-
poin . A dual mechanism is also in ol ed in u ning P. sy ingae
c
e
Lowe epide mis MesophyllZ-s ack
d
a
1 dpi
5
2 dpi
5
4
3
SMC
SMC
Lowe epide mis
Lowe
epide mis
T3SS::GFP3 liC:: dT
T3SS::GFP3 liC:: dT
123
4
2
1
3 dpi
5
g
10
1 dpi
10
2 dpi
10
3 dpi
10
4 dpi
b
Na u e Mic obiology | Volume 10 | Ap il 2025 | 958–972 965
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
T3SS noisy exp ession in o bimodali y20: (1) a double-nega i e18,37
and (2) a posi i e eedback loop38. Resea ch in o he egula ion o
lagella syn hesis in P. sy ingae will help iden i y elemen s in ol ed
in pheno ypic he e ogenei y.
Flagella mo ili y is o en cos ly
26–28
. Flagella exp ession also has a
cos in P. sy ingae (Fig. 3 and Ex ended Da a Fig. 4), al hough his cos is
clea e a he single-cell le el (Fig. 3e–g). In Salmonella, cos s linked o
exp ession o SPI1 T3SS ha e been p oposed as a ac o a ou ing bo h
coun e - egula ion wi h lagella and main enance o he e ogenei y
24,25
.
T3SS exp ession also ca ies a g ow h penal y in P. sy ingae (Fig. 3),
and he s ong g ow h ad an age o a Δ liC Δh pL double mu an indi-
ca es an e en g ea e i ness cos associa ed o exp ession o bo h.
Pheno ypic he e ogenei y is conside ed pa icula ly ad an ageous
when a ec ing loci p oducing immunogenic and/o ene ge ically
cos ly goods, wi h cos o he p oduce s and bene i s o he popula ion
de ining a coope a i e beha iou 39–41. The g ow h penal ies de ec ed
o lagella and T3SS exp ession in P. sy ingae p o ide selec i e ad an-
ages o pheno ypically he e ogeneous popula ions and may d i e
selec ion o gene ic and/o epigene ic mechanism(s) unde lying such
he e ogenei ies. Fo bac e ia engaging in hos in e ac ions, he e oge-
neous exp ession lowe s he o e all amoun o lagellin displayed by
he popula ion, o by a mic ocolony a a local le el, and could be also
ad an ageous by educing de ence elici a ion and acili a ing immune
supp ession.
Coope a i e i ulence can also p e en he ise o non-p oducing
mu an a ian s a he expense o he p oduce s
42–45
. Pheno ypically
OFF bac e ia can ou compe e less equen mu an OFF a ian s and
sa ely e e o ON, hus p e en ing loss o unc ion and s abilizing
he coope a i e beha io 43. Chea e mu an a ian s a ec ed in ype
III sec e ion ha e been desc ibed in P. sy ingae popula ions colonizing
A abidopsis, al hough a equencies lowe han expec ed conside ing
he po en ial o exploi a ion o T3SS e ec o s as public goods
22,46
.
Sp ay
Ex ac ion Na u al
0
50
100 liC:: dT
<0.0001
****
Me ged dToma oGFP
To al30-min incuba ion
hopAB1::GFP3 liC:: dT
b
a
1 dpi 7 dpi
Ex ac ion Na u al
0
50
100 liC:: dT
0.237
Ex ac ion Na u al
0
50
100
150
200
250
Fluo escence in ensi y (a.u.)
hopAB1::GFP3
0.349
Ex ac ion Na u al
0
50
100
150
Fluo escence in ensi y (a.u.)
hopAB1::GFP3
0.361
Fig. 5 | Ac i e exi om in ec ed issues is ca ied ou by FlagellaON bac e ia.
a, Selec ed images o apoplas -ex ac ed bac e ia (using nega i e p essu e; op
panels), bac e ia exi ing he issue on hei own (du ing 30 min lea incuba ion
wi hin an MgCl2 solu ion; middle panels) o bac e ia eco e ed om excess
solu ion d ipping om lea es sp ayed e e y 5 min o mimic ain (bo om panels).
Con as and b igh ness we e adjus ed o imp o e isualiza ion bu we e kep
cons an ac oss he di e en condi ions. b, Fluo escence quan i ica ion o he
images ob ained in a (le images) a 1 dpi (le g aphs) and a 7 dpi ( igh g aphs)
o he same ime-cou se expe imen , using Fiji so wa e. G aph shows a bi a y
uni s o GFP luo escence co esponding o he exp ession o he T3SS gene
usion hopAB1::GFP3, o dToma o luo escence co esponding o he exp ession
o liC:: dT. Each do co esponds o an indi idual bac e ium as analysed om
he image (n = 44). Da a a e p esen ed as mean ± s.e.m. G aphs show esul s
ep esen a i e o 5 independen expe imen s (n = 2 o s ain ca ying h pL::GFP3
liC:: dT and n = 3 o s ain ca ying hopAB1::GFP3 liC:: dT). Compa isons
be ween apoplas ex ac ion and na u al exi we e ca ied ou pe sample using
an unpai ed - es (****P < 0.0001) and esul s shown o each pai .
Na u e Mic obiology | Volume 10 | Ap il 2025 | 958–972 966
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
This ole o T3SS e ec o s as public goods has been epo ed using
mu an s
32
and me apopula ions, whe e mixed isogenic P. sy ingae
de i a i es lacking e ec o s, and each o e exp essing a single e ec o ,
coope a i ely es o e i ulence46.
Pheno ypic he e ogenei y can also allow bac e ial popula ions o
cope wi h apid en i onmen al changes, since p eexis ing subpopula-
ions p eadap ed o incoming s esses can o e come hese changes
as e 39,40. This is known as ‘be -hedging’41. Pheno ypic he e ogenei y
can also lead o ‘di ision o labou ’ when a popula ion di e si ies, allow
-
ing he dis ibu ion o asks among pheno ypically di e en a ian s.
While in be -hedging, one subpopula ion is i e han he o he in each
en i onmen , di ision o labou inc eases i ness o he en i e popula-
ion
41
. We ound no e idence o any T3SS/ lagella pheno ypic a ian
bene i ing du ing plan coloniza ion, as expec ed in a be -hedging
scena io. Howe e , di e en de ence elici a ion con ex s ( o example,
esis an plan s) migh di e en ially bene i ce ain pheno ypes, as
shown o genome- eo ganized a ian s
47,48
. Ou esul s suppo a
‘di ision o labou ’ scena io, wi h T3SS
ON
bac e ia complemen ing
T3SSOFF p obably h ough supp essing immuni y by ‘common goods’
T3SS e ec o s22 (Fig. 4 and Ex ended Da a Fig. 5). The pe cen age
o T3SS
ON
bac e ia equi ed o e ec i e ans complemen a ion o
T3SSOFF, since a subpopula ion is exp essing immunogenic lagellin,
is likely o be balanced by g ow h ade-o s.
As euka yo ic cell ypes wo k wi hin issues, bac e ial coope a ion
eme ges wi hin spa ially s uc u ed popula ions49. Howe e , s uc u ed
coope a ion has been a ely shown wi hin he con ex o disease
50
.
Indeed, al hough pheno ypic he e ogenei y is well documen ed in
plank onic cul u es, how hese p ocesses occu in spa ially s uc u ed
communi ies is less unde s ood51. The egula o y loops in ol ed in
es ablishing bis able exp ession o he T3SS in plank onic HIM20 cul-
u es a e in place du ing plan coloniza ion
19,52
. Howe e , apoplas ic
bac e ia encoun e di e en s imuli speci ic o hei mic oen i on-
men , which changes o e ime, po en ially leading o cell- o-cell
a ia ion o hogonal o ha gene a ed h ough T3SS and lagella
s ochas ic he e ogenei y. Time-cou se analysis o T3SS and lagella
exp ession du ing plan coloniza ion shows he e ogenei y h ough-
ou ime and loca ions, bu also an o e lapping s uc u ed pa e n
(Fig. 4). On he lea su ace, mos bac e ia a e T3SS
OFF
ega dless o
he single-cell s a us o lagella exp ession, changing o a majo i y
o T3SSON/ FlagellaOFF du ing ea ly g ow h in he apoplas (Fig. 4a,b,
1 dpi). F om 2 dpi onwa ds, he e ogenei y is he no m o bo h loci, bu
p edominan ly T3SSON o FlagellaON a eas appea . These a eas show a
h ee-dimensional (3D) pa e n: hose closely associa ed o hos -cell
su aces a e en iched in T3SS
ON
bac e ia, while dis al a eas, o he inne -
mos pa s o he mic ocolony, a e p edominan ly FlagellaON (Figs. 4 and
Ex ended Da a Fig. 7, and Supplemen a y Videos 4 and 5). This is con-
sis en wi h he no ion o T3SS
ON
bac e ia p edominan ly loca ed whe e
T3SS ac i i y is ele an o supp ess cellula immuni y and explains he
lack o selec i e killing o T3SS
OFF
bac e ia
26
(Ex ended Da a Fig. 5). Such
pheno ypic di e en ia ion in o spa ially dis ibu ed subpopula ions
coope a ing in a complex en i onmen ( ha is, in plan a s in i o)
has a ely been in es iga ed, pa icula ly o mo e han a ai
53
, bu
i is easonable o expec i o ake place in he con ex o hos coloni-
za ion p ocesses. Spa ially s uc u ed coope a ion in he con ex o
disease has been demons a ed o chole a oxin and oxin-co egula ed
pilus in Vib io chole ae in mice in es inal mic ocolonies, and o ni ic
oxide (NO)-de oxi ying hmp gene in Ye sinia pseudo ube culosis in
spleen mic ocolonies49,53. Spa ial he e ogenei y has also been desc ibed
h ough single-cell omics wi hin in i o bio ilms o human pa hogen
P. ae uginosa, including a ia ion in lagella exp ession54. Hos -cell
signals, cell- o-cell in e ac ions o lineage his o y55 could be in ol ed
in T3SS ac i a ion in hos -p oximal a eas. How en i onmen al signals
combine wi h s ochas ic p ocesses o gene a e pheno ypic a ia ion
is a cu en opic o in e es
55
, one pa icula ly challenging o s udy
beyond in i o sys ems i posi ional in o ma ion is o be ob ained55.
As apoplas ic mic ocolonies de elop, FlagellaON bac e ia inc ease
in numbe s. Mo ile bac e ia a e expec ed o exi he lea mo e e i-
cien ly (Ex ended Da a Fig. 7), and ou esul s expe imen ally suppo
his no ion by showing ha mos ly FlagellaON bac e ia exi colonized
issue be o e nec osis (Fig. 5). The inal s ages o he in ec ion and hos
exi is an unde s udied domain, e en o model pa hogen P. sy ingae, so
whe he his ac i e ea ly exi is mo e biologically ele an han passi e
exi om nec o ic issue emains o be explo ed, bu i appea s a sa e
s a egy since nec osis leads o he elease o oxic compounds and o
he ac i a ion o addi ional de ence esponses56,57.
Only a hand ul o s udies ha e analysed exp ession o wo bis able
he e ogeneous loci simul aneously49,53,58. Ou s s udies wo loci ele an
o bac e ial–hos in e ac ion (Supplemen a y Video 10) and does i
in he con ex o disease wi hin a spa ially s uc u ed dynamic mic o-
colony. I p o ides an example o di ision o labou and coope a i e
i ulence in a plan hos , alida ing he impo ance o pheno ypic
he e ogenei y in na u e and shi ing he emphasis om i s mechanis ic
unde s anding o i s ecological and biological ele ance. Finally, he
s udy es ablishes pheno ypic he e ogenei y and coope a i e i ulence
as a conse ed s a egy by which bac e ial pa hogens cope wi h he
luc ua ing challenges encoun e ed in plan and animal hos s.
Me hods
Bac e ial s ains and g ow h condi ions
Bac e ial s ains used and gene a ed in his wo k a e de ailed in Supple
-
men a y Table 1. E. coli and P. sy ingae s ains we e g own wi h ae a ion
in LB medium59 a 37 °C o E. coli o 28 °C o P. sy ingae. An ibio ics
we e used, when necessa y, a he ollowing concen a ions: ampicil-
lin, 100 µg ml
−1
o E. coli and 500 µg ml
−1
o P. sy ingae; kanamycin,
50 µg ml−1 o E. coli and 15 µg ml−1 o P. sy ingae de i a i e s ains;
gen amycin, 10 µg ml−1; ni o u an oin, 40 µg ml−1; and cycloheximide,
2 µg ml−1.
To induce he exp ession o he h p/h c genes, bac e ia we e ini-
ially cul u ed o e nigh in LB a 28 °C, supplemen ed wi h he app o-
p ia e an ibio ic, hen washed wice in 10 mM MgCl2 be o e being
cul u ed in h p-inducing minimal medium (HIM), con aining 10 mM
uc ose60. Fo his s udy, he pH o HIM was adjus ed o 7.0 wi h 10 N
NaOH. The ini ial op ical densi y (OD) was adjus ed o 0.13 and cul u es
we e incuba ed a 28 °C wi h agi a ion.
Fluo escence labelling o bac e ial s ains
Bac e ial s ains ca ying a ch omosome-loca ed ansc ip ional
usion o liC gene o a p omo e less dToma o gene we e gene a ed
using an adap a ion o he me hod p e iously desc ibed in e . 61.
The plasmids used and gene a ed o his pu pose a e de ailed in
Supplemen a y Table 2. The p ime s used a e desc ibed in Supple-
men a y Table 3. Fo he gene a ion o he allelic exchange plasmid,
wo agmen s o ~500 pb we e ampli ied om Pph 1448A genomic
DNA using Q5 High-Fideli y DNA Polyme ase (New England Biolabs);
one o hese agmen s (A) encompasses he 3′ end o he open ead-
ing ame (ORF), including he STOP codon, while he o he agmen
(B) co e s he sequence immedia ely downs eam o he STOP codon.
All p ime s used a e desc ibed in Supplemen a y In o ma ion. Each
eac ion was ca ied ou a 98 °C o 1 min o he ini ial dena u a ion
s ep, ollowed by 30 cycles a 98 °C o 30 s, annealing a 62 °C o 30 s
and ex ension a 72 °C o 30 s, ollowed by 5 min a 72 °C o he inal
ex ension s ep. The eac ion mix u e o each PCR included 0.64 mM
deoxynucleo ide iphospha e (dNTP) mix, 0.4 ng o each p ime ,
1 ng o genomic DNA, he app op ia e enzyme bu e and comme cial
ul apu e wa e (Nalgene). A olume o 2 µl o each gel-pu i ied PCR
p oduc was employed as empla e o he subsequen usion PCR,
employing p ime s A1 and B2 in a PCR eac ion conduc ed unde he
condi ions desc ibed, wi h an ex ended elonga ion ime o 1 min. The
esul ing bands, comp ising he end o each ORF and i s downs eam
sequence sepa a ed by an EcoRV es ic ion si e, we e A/T cloned in o
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Ex ended Da a Fig. 1 | See nex page o cap ion.
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Ex ended Da a Fig. 1 | T ansc ip ional usions o dToma o do no a ec gene
unc ion. (a) A s ain ca ying a liC:: dT ansc ip ional usion displays wild
ype mo ili y in he pho og aphs (a) and quan i ica ions (b) o swimming pla es
(n = 4). No s a is ical di e ences we e obse ed as indica ed by an unpai ed
wo- ailed es (P < 0.05). Sca e do g aph shown mean +/− SEM. The liC:: dT
cons uc is in eg a ed in he na i e ch omosome loca ion o he liC gene as
es ablished by Sou he n blo analysis (using he np II gene con e ing esis ance
o Km as a p obe) o a Alw21I (The mo Fishe Scien i ic, USA) diges ion o
genomic DNA om he wild ype (Pph) and liC:: dT de i a i e s ains (c) showing
he expec ed band sizes (d). (e) Rep esen a i e images o he symp oms caused
by he wild ype s ains and hei ela i e hopAB1::GFP3 liC:: dT, h pL::GFP3
liC:: dT and h cU::GFP3 liC:: dT s ains a 5 and 7 dpi. ( ) Rep esen a i e images
o he colonies ob ained a e pla ing se ial dilu ions o he indica ed s ains
p e iously adjus ed o 0.1 op ical densi y (Abs600). (g) Quan i ica ion o he da a
ob ained in . All dilu ions each simila alues indica ing no di e ences in CFU/
ml occasioned by exp ession o luo escence epo e genes. and g illus a e
one expe imen (>3 echnical eplica es), da a om 3 addi ional independen
expe imen s ca ied ou in he same manne a e showed in h. (h) Mean numbe o
c u/ml ob ained a e pla ing a dilu ion o he cul u e p e iously adjus ed o 0.1
op ical densi y (Abs600). Da a ob ained o he di e en s ains (n = 3 biological
eplica es; echnical eplica es no used o s a is ical analysis) did no show
signi ica i e di e ences acco ding o a one-way ANOVA ollowed by Tukey’s
mul iple compa ison es (P < 0.05). G aph shows indi idual alues (sca e plo )
and mean +/− SEM is also indica ed.
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
dToma oeCFP
B igh ield
hopAB1::GFP3 liC:: dT eCFP
Ex ended Da a Fig. 2 | See nex page o cap ion.
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Ex ended Da a Fig. 2 | Flagella exp ession is he e ogeneous on plan su ace.
Selec ed images o he hopAB1::GFP3 liC:: dT eCFP s ain on he plan su ace
a 6 h pos inocula ion (hpi) by dipping he lea in o a 5 × 107 CFU/ml bac e ial
suspension. dToma o panel shows he luo escence o dToma o associa ed o
liC exp ession, and eCFP panel shows he luo escence o eCFP as cons i u i e
exp ession epo e , using g ey scale in bo h cases o imp o e con as . The GFP
panel, co esponding o hopAB1 exp ession, is no shown since no luo escence
was de ec ed. Scale ba s co espond o 20 µm. Con as and b igh ness we e
adjus ed o imp o e isualiza ion bu we e kep cons an ac oss panels. Ci cles
highligh bac e ia de ec ed in he eCFP panel wi hou exp ession in he dToma o
channel (FlagellaOFF). Righ panels co espond o he magni ica ion o he a ea
selec ed in ec angles. Images a e ep esen a i e o 3 independen expe imen s.
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Ex ended Da a Fig. 3 | See nex page o cap ion.
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Ex ended Da a Fig. 3 | Flagella display he e ogenous exp ession in
Pseudomonas sy ingae. (a) CSLM o a s ain ca ying liC:: dT g own O/N in
LB (uppe ), o 24 h in HIM (cen al), o ex ac ed om bean lea apoplas s 4 dpi
(bo om). dToma o panels show dToma o luo escence as epo e o liC
exp ession and BF panel co esponds o b igh ield. Scale ba s co espond o
2 µm. Images a e ep esen a i e o 3 independen expe imen s (b) FC analysis
o liC:: dT s ain in condi ions desc ibed in a. Do plo s show cell size e sus
dToma o luo escence in ensi y, ep esen ed as a bi a y uni s in loga i hmic
scale. Da a collec ed o 100,000 e en s pe sample. The non- dT g aph show
au o luo escence le els displayed by wild ype no ca ying any luo escen
gene. Ve ical lines lea e 99% o da a o non- dT s ain o he le and is used as
e e ence o di e en ia e o OFF cells. a and b show ep esen a i e esul s o
a leas h ee independen eplica es. (c) FC analysis shows GFP3 and TdToma o
luo escence de ec ion is no bias by bleed h ough. Le g aphs co espond o
liC::GFP3 s ain and igh o he liC:: dT s ain. Uppe g aphs show dToma o
luo escence e sus cell coun and bo om g aphs show GFP luo escence e sus
cell coun . In all cases he non- luo escen s ain is included as nega i e con ol
and da a we e ea ed as in b. (d) Quan i ica ion o GFP li e ime (τ) in he p esence
o absence o dToma o. Bac e ia exp essing simila le els o h pL::GFP3
(de e mined as GFP luo escence in ensi y) and di e ing in he exp essing o
liC:: dT exp ession (as shown by di e en dToma o luo escence in ensi y) we e
chosen o he analysis (uppe panel). GFP li e ime was ob ained o each pixel
o he image (lowe -le ) and mean was calcula ed o each single bac e ium.
A e age o ou bac e ia in each condi ion is ep esen ed (lowe - igh ) and
s a is ical signi icance de e mined wi h a - es . (e and ) Do plo g aphs display
luo escence in ensi y o GFP o dToma o e sus cell size in non- luo escen
bac e ia (wild ype), o s ains ca ying h pL::GFP3 liC:: dT o hopAB1::GFP3
liC:: dT co esponding o da a shown in Fig. 4 as GFP luo escence e sus ha
o dToma o. Figu e show ep esen a i e esul s o a leas h ee independen
expe imen s. In (e), bac e ia g own 24 h in HIM a e dilu ing an O/N LB cul u e,
display ep oducible bis able exp ession o bo h h pL::GFP3 and hopAB1::GFP3
and he e ogeneous (occasionally bis able) exp ession o liC:: dT. In he case
o hopAB1::GFP3 liC:: dT s ain in HIM, an addi ional e ical line (dashed) was
added o sepa a e he subpopula ions di e ing in hopAB1 exp ession, which
is highe due o he high basal le els o his gene. In ( ), bac e ia ex ac ed
om bean lea apoplas s 4 dpi display ypical he e ogeneous (ne e bis able)
exp ession o h pL::GFP3, hopAB1::GFP3 and liC:: dT.
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Ex ended Da a Fig. 4 | See nex page o cap ion.
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Ex ended Da a Fig. 4 | T3SS and lagella exp ession impac on bac e ial g ow h
and g ow h cos associa ed wi h single cell le els o lagella p oduc ion is no
due o GFP. (a) G ow h a es o 1448 A wild ype, a de i a i e ca ying a plasmid
ha de e mines cons i u i e exp ession o h pL (pH pL), o o leQ (pFleQ), o
a leQ dele ion (Δ leQ), in HIM. Da a co espond o 4 biological eplica es. Da a
a e p esen ed as mean alues +/− SEM. Signi ica i e di e ences a e obse ed
be ween WT and pFleQ s ains acco ding o a one-way ANOVA ollowed by
Tukey’s mul iple compa ison es (P < 0.05). P alue is indica ed in he g aph.
(b) Selec ed imelapse images o wild ype cons i u i ely exp essing GFP (eGFP)
du ing he mic ocolony de elopmen on aga pads in phase con as ( op) and
GFP luo escence (bo om) channels. Con as and b igh ness we e adjus ed
and main ained h oughou . (c) Compa isons o mean g ow h a es o eGFP
s ain and liC::GFP3 s ain. No e ha GFP in he eGFP s ain has a much highe
luo escence in ensi y compa ed o liC::GFP3. eGFP shows a signi ican ly lowe
a e age g ow h a e compa ed o liC::GFP (Mann-Whi ney U es , P = 7.7 ×
10−15). (d) Compa ison o g ow h a e o cells displaying high o low GFP signal
(de e mined by spli ing he popula ion in o wo g oups using he median
luo escence in ensi y alue) o eGFP s ain shows no signi ican di e ences
be ween g ow h a es o cells exp essing high o low GFP le els (Mann-Whi ney
U es , P < 0.05). (e) Co ela ion be ween g ow h a es and luo escence in ensi y
o indi idual liC::GFP3 cells indica es ha highe liC exp ession is associa ed
wi h slowe g ow h. The shaded a ea shows he 95% con idence in e al.
( ) Co ela ion be ween g ow h a es and luo escence in ensi y o indi idual
eGFP cells indica es ha highe GFP exp ession is no associa ed wi h slowe
g ow h. The shaded a ea shows he 95% con idence in e al.
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Ex ended Da a Fig. 5 | Dead-li e s aining shows nei he bias owa ds FlagellaON
cells no o T3SSOFF cells du ing plan g ow h. G aph shows he dead/li e
a io o bac e ia ei he exp essing o no liC::GFP3, h pL::GFP3, hopAB1:GFP3
o h cU::GFP3 du ing plan g ow h. Apoplas -ex ac ed bac e ia a 4 dpi in bean
lea es we e s ained wi h a solu ion o P opidium iodide (PI). Da a we e ob ained
by low cy ome y analysis (100,000 e en s/ pe eplica e) and GFP le els
we e used o di e en ia e be ween ON and OFF cells using he non- luo escen
wild ype s ain as e e ence, as indica ed be o e. Fluo escence o PI was used
o di e en ia e dead and ali e bac e ia compa ing o he non- luo escen wild
ype. Each do ep esen s biological eplica e in wo di e en expe imen s
(Rep 1 and Rep 2).
Na u e Mic obiology
A icle h ps://doi.o g/10.1038/s41564-025-01966-0
Ex ended Da a Fig. 6 | Ac i e exi om in ec ed issues is p e e en ially
ca ied ou by FlagellaON bac e ia. (a) Selec ed images o apoplas -ex ac ed
bac e ia (using nega i e p essu e; uppe panels) o bac e ia exi ing he issue
on hei own (du ing lea incuba ion wi hin a MgCl2 solu ion; na u al exi , lowe
panels) a 1 dpi and 2 dpi (asymp oma ic issue), o 5 dpi (jus be o e isible
symp oms appea ) and 7 dpi ( ully symp oma ic nec o ic issue). Con as and
b igh ness we e adjus ed o imp o e isualiza ion bu we e kep cons an ac oss
he di e en condi ions. (b and c) Fluo escence quan i ica ion o he images
ob ained in a o 2 dpi and 5 dpi using Fiji so wa e. G aph shows a bi a y uni s
o GFP luo escence, co esponding o he exp ession o he T3SS gene usion
hopAB1::GFP3 in he samples, o dToma o luo escence co esponding o he
exp ession o liC:: dT. Each do co esponds o an indi idual bac e ium (n = 44)
as analyzed om he image. Compa isons be ween apoplas ex ac ion and
na u al exi we e ca ied ou pe each sample using an unpai ed wo- ailed - es .
Da a a e p esen ed as mean alues +/− SEM. G aphs show esul s ep esen a i e
o 3 independen expe imen s.
