Low Dose Ae osol Fi ness a he Inna e Phase o Mu ine
In ec ion Be e P edic s Vi ulence amongs Clinical
S ains o
Mycobac e ium ube culosis
Neus Cace es
1,2
, Isaac Llopis
3
, Elena Ma zo
1,2
, Cla a P a s
3
, C is ina Vilaplana
1,2
, Da io Ga cia de
Viedma
2,4
, So ı
´a Sampe
2,5,6
, Daniel Lopez
3
, Pe e-Joan Ca dona
1,2
*
1Uni a de Tube culosi Expe imen al, Fundacio
´Ins i u pe a la In es igacio
´en Cie
`ncies de la Salu Ge mans T ias i Pujol, Uni e si a Au o
`noma de Ba celona, Badalona,
Ca alonia, Spain, 2CIBER En e medades Respi a o ias, Bunyola, Balea s, Spain, 3Depa amen de Fı
´sica i Enginye ia Nuclea , Escola Supe io d’Ag icul u a de Ba celona,
Uni e si a Poli e
`cnica de Ca alunya, Cas ellde els, Ca alonia, Spain, 4Se icio de Mic obiologı
´aClı
´nica y En e medades In ecciosas, Hospi al G ego io Ma an
˜o
´n, Mad id,
Spain, 5Ins i u o A agone
´s de Ciencias de la Salud, Za agoza, Spain, 6Hospi al Uni e si a io Miguel Se e , Za agoza, Spain
Abs ac
Backg ound:
E alua ion o a quick and easy model o de e mine he in insic abili y o clinical s ains o gene a e ac i e TB
has been se by assuming ha his is linked o he i ness o Mycobac e ium ube culosis s ain a he inna e phase o he
in ec ion. Thus, he highe he bacilla y load, he g ea e he possibili y o induc ing lique ac ion, and hus ac i e TB, once
he adap i e esponse is se .
Me hodology/P incipal Findings:
The i ulence o se en clinical Mycobac e ium ube culosis s ains isola ed in Spain was
es ed by de e mining he bacilla y concen a ion in he spleen and lung o mice a weeks 0, 1 and 2 a e in a enous (IV)
inocula ion o 10
4
CFU, and by de e mining he g ow h in i o un il he s a iona y phase had been eached. Co d
dis ibu ion au oma ed analysis showed wo clea pa e ns ela ed o he high and low i ness in he lung a e IV in ec ion.
This pa e n was no seen in he in i o i ness es s, which clea ly a o ed he e e ence s ain (H37R ). Subsequen
de e mina ion using a mo e physiological low-dose ae osol (AER) inocula ion wi h 10
2
CFU showed a hi d pa e n in which
he h ee bes alues coincided wi h he highes dissemina ion capaci y acco ding o epidemiological da a.
Conclusions/Signi icance:
The i ness ob ained a e low dose ae osol adminis a ion in he p esence o he inna e immune
esponse is he mos p edic i e ac o o de e mining he i ulence o clinical s ains. This gi es suppo o a mechanism o
he induc ion o ac i e TB de i ed om he dynamic hypo hesis o la en ube culosis in ec ion.
Ci a ion: Cace es N, Llopis I, Ma zo E, P a s C, Vilaplana C, e al. (2012) Low Dose Ae osol Fi ness a he Inna e Phase o Mu ine In ec ion Be e P edic s Vi ulence
amongs Clinical S ains o Mycobac e ium ube culosis. PLoS ONE 7(1): e29010. doi:10.1371/jou nal.pone.0029010
Edi o : Ma kus M. Heimesaa , Cha i e
´, Campus Benjamin F anklin, Ge many
Recei ed Augus 29, 2011; Accep ed No embe 18, 2011; Published Janua y 3, 2012
Copy igh : ß2012 Cace es e al. This is an open-access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s
un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed.
Funding: The esea ch leading o hese esul s has ecei ed unding om he Eu opean Communi y’s 7 h F amewo k P og amme HEALTH-2007-2.3.2-2 (FP7/
2007–2013: TOPLATENT-TB p ojec unde g an ag eemen 200999). The unde s had no ole in s udy design, da a collec ion and analysis, decision o publish, o
p epa a ion o he manusc ip .
Compe ing In e es s: The au ho s ha e decla ed ha no compe ing in e es s exis .
* E-mail: [email p o ec ed]
In oduc ion
The concep o i ulence has been in e p e ed om bo h a hos
and pa hogen poin o iew [1]. As a as M. ube culosis is
conce ned, i was p edominan ly de ined and de e mined in
e ms o su i al and his opa hology om he beginning o he
20 h cen u y [2] un il he 1990s, when i ulence begun o be
measu ed by de e mining he g ow h a e [3] acco ding o he
Da winian concep o i ulence: ‘‘likelihood o su i e and
ep oduce’’ [4]. Vi ulence in M. ube culosis is he e o e commonly
de e mined by i s maximum g ow h a e, which is exp essed
nume ically by he slope o he eg ession cu e in a semi-
loga i hmic ep esen a ion o bac e ial g ow h and is consensu-
ally called ‘‘Fi ness’’.
Al hough he i ulence o s ains in ol ed in human ou b eaks
could o iginally only be deduc ed om epidemiological e o-
spec i e s udies, expe imen al models a e now used. The i s
s udies ega ding he i ulence o M. ube culosis we e pe o med
using in i o models [5]. Howe e , al hough animal models a e
s ill used o de e mine i ulence [6], new app oaches using issue
cul u es (mac ophages, dend i ic cells and pneumocy es) [7] and
in i o liquid cul u es using he Bac ec and he MGIT sys ems
[8,9,10] ha e been designed. Vi ulence s udies a e highly ocused
on analyzing he inc ease o dec ease o he i ness o d ug-
esis an clinical and mu an s ains. These eme ging d ug-
esis an s ains a e s udied e ospec i ely om an epidemio-
logical poin o iew and hei i ulence is subsequen ly
cha ac e ized expe imen ally. These s udies ha e shown ha
he e is a s ong selec ion p essu e o d ug- esis ance-con e ing
mu a ions ha cause minimal i ness de ec s [4]. Mo eo e , W-
Beijing s ains ha e shown o be specially i ulen in animal
models o in ec ion [11]. Indeed, ce ain lineages ha e been
ela ed o an inc eased ansmissibili y and/o pa hogenici y
[12,13],andi hasbeensugges ed ha he eisaco ela ion
be ween he i ulence pa ame e s and epidemiological cha ac-
e is ics o new Beijing s ains [14].
PLoS ONE | www.plosone.o g 1 Janua y 2012 | Volume 7 | Issue 1 | e29010
P e ious li e a u e epo s encou aged us o de elop a new
quick and easy expe imen al model o es he i ulence o clinical
s ains in o de o p edic he po en ial damage ha hese
s ains could cause in a popula ion. As esul s om ou p e ious
expe imen s sugges ed ha a cons an endogenous ein ec ion
occu s du ing La en Tube culosis In ec ion (LTBI) [15], we
conside ed ha he speed o g ow h be o e he onse o he speci ic
immune esponse, oge he wi h ein ec ion in he uppe lobe,
whe e a highe oxygen p essu e p omo es as e bacilla y g ow h,
could be c ucial o he in ec ion o p og ess o ac i e disease.
In his con ex , he highe he bacilla y load he highe he
p obabili y ha i will induce lique ac ion (and pos e io
ca i a ion) once he immune esponse appea s [16][17]. Because
o his o iginal app oach o he na u al his o y o ube culosis (TB),
we also conside he maximum g ow h a e be ween 0 and 2 weeks
pos -inocula ion o be he bes means o de e mining he i ulence
o a s ain.
A simple and sho ime- amed mu ine model was de eloped
and alida ed by es ing M. ube culosis clinical s ains wi h
demons a ed epidemiological impo ance in Spain in he las 20
yea s in h ee Spanish egions (G and Cana ia, Mad id and
Za agoza). The de elopmen o his i ulence model led o a need
o de e mine he quali y o he inoculum by s udying he
cha ac e is ics o i s co ding p ope ies, i.e., he M. ube culosis
endency o join in o bacilla y agg ega es wi h a pa allel
o ien a ion o cells’ axes [18,19]. This de e mina ion esul ed in
a aluable ‘‘in i o’’ es o p edic he ‘‘in i o’’ beha io . Two
di e en i ulence ankings we e ob ained in he in i o models
depending on he in ec ion ou e used. The esul s ob ained
sugges his is because he low-dose ae osol model (AER) mainly
desc ibes he i ness o isola ed bacilli, whe eas he in a enous
model (IV) gi es in o ma ion ega ding bo h isola ed and co ded
bacilli. The AER model has been ound o bes i he
dissemina ion o TB cases in he communi y. This ac ein o ces
he concep ha a highe i ness a he beginning o he in ec ion
p ocess co ela es wi h he abili y o induce TB disease, hus
suppo ing he concep de i ed om he dynamic hypo hesis o
LTBI.
Resul s
In i o i ness clea ly a o s he s anda d labo a o y
s ain
All s ains g ew in 7H9 Middleb ook supplemen ed medium as
expec ed excep o s ain 7, which, al hough i was cul u ed h ee
imes, showed no e idence o g ow h du ing he pe iod o ime in
which i was p e iously epo ed o do so (i.e. 1 mon h). Fo his
eason, no CFU da a om he S ain 7 we e used in he i ness
calcula ion.
The i ness o he e e ence s ain (H37R Pas eu ) was
conside ably highe han ha o he o he s ains (Figu e 1),
ollowed, in dec easing o de , by he clinical s ains (Table 1).
Rema kably, isola es 4 and 5 (ac ually he same s ain) showed a
simila i ness, hus gi ing some obus ness o his me hod. No
co ela ion was ound be ween i ness and lag phase (da a no
shown).
The i ness in lungs a e in a enous (IV) inocula ion
appea s o be a eliable i ulence me hod
The i s in i o model in mice in ol ed in ec ing he animals by
he in a enous ou e (IV). The bacilla y load was de e mined in
lung and spleen a day 0 and weeks 1 and 2 pos -in ec ion
(Figu e 2).
Spleen i ness esul s we e highe han he co esponding alues
o lung i ness in all cases (Figu e 1), whe eas he lag phase was 0
in all s ains (da a no shown). In e es ingly, all s ains excep s ain
4 showed a highe i ness han H37R in he lung (Table 1).
Figu e 1. Fi ness esul s. Fi ness esul s om in i o and in i o s udies. Top: i ness alues o lung ae osol and IV expe imen s; bo om: in i o
assay and spleen IV in ec ion alues. Do ed lines indica e he i ness alue o he H37R s ain (shown as ‘‘S ’’).
doi:10.1371/jou nal.pone.0029010.g001
Low Dose Ae osol Fi ness P edic ing Vi ulence
PLoS ONE | www.plosone.o g 2 Janua y 2012 | Volume 7 | Issue 1 | e29010
S ains 4 and 5 showed o ally di e en i ness, hus sugges ing a
poo obus ness o his me hod.
No ela ion could be es ablished be ween lung i ness and lag
phase (da a no shown). The H37R s ain showed a medium
posi ion in he i ness anking (Table 1), and isola es 4 and 5
showed a simila i ness, which was he lowes .
The co ding phenomenon is ela ed o i ness in he IV
model
Co ding was de ec ed in all s ains om he ozen la e-log
phase s ocks o each s ain used o inocula e each expe imen al
model. A eas om single and agg ega ed bacilli in la ge co ds
om he ozen s ocks o each clinical s ain we e analyzed by
image analysis. The g aphical ep esen a ion o he a ea
dis ibu ion o all s ains displayed an asymme ic shape
(Figu e 3A and B). Two ypes o a ea dis ibu ion pa e ns we e
dis inguished: he i s pa e n displayed a log-no mal dis ibu ion
whe eby he maximum numbe o agg ega es (any pa icle
de ec ed, om single bacilli o la ge co ds) co esponds o he
second column o he g aph, which comp ises he a ea be ween
2.5 and 6.3 mm
2
( ew bacilla y agg ega es, up o 5 bacilli). An
exponen ial dec ease is obse ed om 6.3 mm
2
o highe a eas
(Figu e 3A). The second pa e n displayed a mono onous
exponen ial dec ease in which he maximum numbe o
agg ega es co esponds o he i s column p esen ed in he g aph,
which comp ises he a ea be ween 1 and 2.5 mm
2
(single bacilli).
An exponen ial dec ease is obse ed om 2.5 mm
2
o highe a eas
(Fig. 3B). I is impo an o compa e hese da a wi h he ac ha
he size o in ec i e ae osols is a ound 7 mm
2
[20].
Once he clinical s ains had been classi ied in o he i s o
second pa e n, s a is ical analysis o hese wo g oups showed
signi ican di e ences be ween hem o he pa ame e s lung
i ness om IV in ec ion (Figu e 4A), skewness (a s a is ical
pa ame e ha indica es he asymme y o a p obabili y
dis ibu ion; Figu e 4B), co ding p opo ion and single bacilli
p opo ion (Figu e 4C). As a as he lung i ness esul s o he IV
in ec ed mice a e conce ned, s ains we e g ouped as as (highes
i ness alues: s ains 1, 2 and 3) o slow (lowes i ness alues:
s ains 4, 5, 6 and 7). The H37R s anda d was also classi ied as a
as s ain. The as s ains showed a clea ly highe co ding
p opo ion and lowe single bacilli p opo ion han slow s ains.
The AER model gi es a hi d i ness anking ha is
ela ed o he epidemiologic da a a he han he
co ding abili y
Summa izing he p e ious da a, he i ness ob ained in lung
a e IV in ec ion showed he bes p o ile in e ms o de e mining
he i ulence in clinical s ains. In ligh o his, he in o ma ion
p o ided by he co ding dis ibu ion, which co ela es wi h he
highes o lowes i ulence, ep esen s a no able achie emen as
his in i o model could be a use ul, quick and eliable ma e o
de e mining his pa ame e in la ge numbe s o samples. In his
ega d, we planned o alida e hese esul s by es ing he mos
physiological me hod, namely low-dose ae osol in ec ion (AER), o
be e mimic he human in ec ion in he lung (Figu e 2).
The da a ob ained using he AER model ga e a new i ness
anking (Table 1), wi h all alues being wi hin he ‘‘high i ulence’’
ange ob ained using he lung IV model (Figu e 1). In his model,
he H37R s anda d showed he second highes alue. In e es -
ingly, he i ness in lung showed a signi ican co ela ion wi h lag
phase, and isola es 4 and 5 showed simila esul s in e ms o
i ness (wi hin he e o ma gins o his echnique; Table 2).
Table 2 shows a compa ison o he i ness pa ame e s o he
s ains acco ding o hei duplica ion ime in all he expe imen al
models used. This compa ison shows how he lowes doubling ime
(bes i ness) was ob ained in he spleen a e IV Inocula ion,
ollowed by in i o, lungs a e AER and lungs a e IV, wi h spleen
g ow h a e IV inocula ion and lung g ow h a e AER being he
models in which he lowes in e -s ain a ia ion was ob ained.
The in i o model showed he highes a iance mos ly because o
he low doubling ime de ec ed o he s anda d labo a o y s ain
H37R (9.55 h) and he huge amoun o ime equi ed by M3
s ain (42.78 h), us conside ing hese a iances being due o he
adap a ion o each s ain o he in i o g ow h.
The AER model was he only one o gi e a good co ela ion
wi h he epidemiological da a. A co ela ion was ound be ween
he calcula ed i ness in he in i o AER model and he numbe o
TB cases ela ed o each s ain (Figu e 5). Thus, he AER model
ga e he maximum alue o s ain 7, which appea s o be he
Table 1. Fi ness anking in he di e en models.
Fas es (le ) o Slowes ( igh )
In i o
H37R 4 5 1 2 6 3
IV spleen 2 5 1 7 3 6 H37R 4
IV lung 1 3 2 7 H37R 6 4 5
AER lung 7 H37R 5 3 4 6 1 2
Numbe s co espond o s ains de ailed in Table 3. S ains a e a anged om
as es (le ) o slowes ( igh ) in a dec easing o de acco ding o Fi ness.
doi:10.1371/jou nal.pone.0029010. 001
Figu e 2. Fi ness in he
in i o
models. Bacilla y load in he lung om ae osol-in ec ed mice and lung and spleen om IV-in ec ed mice on day 0,
week 1 and week 2. The H37R s ain is indica ed as S . Do ed lines show he inocula used o in ec each mouse.
doi:10.1371/jou nal.pone.0029010.g002
Low Dose Ae osol Fi ness P edic ing Vi ulence
PLoS ONE | www.plosone.o g 3 Janua y 2012 | Volume 7 | Issue 1 | e29010
s ain wi h he highes capaci y o dissemina e h oughou he
popula ion wi h he ewes associa ed isk ac o s (Tables 3 and 4).
None o he in i o models induced a speci ic immune
esponse
In o de o ensu e ha he CFU alues ob ained in he mice
expe imen s we e due only o he g ow h capaci y o each s ain
and we e no in luenced by he adap i e immune esponse o he
hos , a new expe imen was pe o med, in which mice we e
in ec ed IV wi h he e e ence o s ain 1 (co esponding o high
and low i ness alues, espec i ely).
Nei he ESAT-6 no PPD-speci ic IFN-c elease could be
de ec ed by ELISPOT assay a any ime poin (day 0, week 1,
week 2) in ei he spleen o hila node samples o animals in ec ed
wi h any o he s ains e alua ed (Figu e S1).
Discussion
Concep ually, he i ulence should be assessed conside ing he
hos and he pa hogen, and i is ue ha a human le el, e e y
case should be ho oughly s udied o exac ly de e mine how much
weigh he hos ci cums ances ha e in he e olu ion o he
in ec ion and he disease. Howe e , a he expe imen al basis, we
decided o ocus in he pa hogen beha io in he mouse model
( hus in a p ecise hos wi h conc e e ci cums ances) aking in o
accoun he in ec ion ou e (ae osol s. in a enous), hus
conside ing he concep o i ulence de e mined by he g ow h
a e and de ined in he In oduc ion sec ion.
This wo k aimed o es ablish an easy, quick and use ul
expe imen al model o analyzing he i ulence o a ecen ly
isola ed clinical M. ube culosis s ain. To alida e his model, se en
clinical s ains isola ed in Spain and o some epidemiological
in e es we e e alua ed oge he wi h a s anda d e e ence s ain
(H37R Pas eu ).
All s ains in ec ed he hos success ully, e en i e iden
di e ences we e encoun e ed be ween he bacilla y load ound
in he issues and he inocula used o in ec he animals, e lec ing
he di e en abili y o success ully in ec he hos o he di e en
s ains. In gene al, he esul s ob ained show ha he i ness
anking is speci ic o each model. Thus, he i ness pa e n
p o ided by he in i o model clea ly a o s he i ulen s anda d
s ain H37R , which could be due o he ac ha his s ain was
main ained subcul u ed o a long ime un il eeze-d y echnology
appea ed in he 1960s. The e e ence s ain showed a ma kedly
highe alue in he in i o assay han in he in i o s udies. This migh
be due o an agonis ic pleio opy, which explains he i ness cos s o
adap ing o a new sys em [21]. A apid inc ease o pa asi e-induced
educ ion o hos i ness is he mos gene al esul o se ial passage
Figu e 3. Agg ega e a ea dis ibu ion esul s. Co ding a ea analysis esul s o he s ains inocula ed, as ob ained by image analysis. A and B:
ep esen a ion o agg ega e a ea dis ibu ion showing he wo cha ac e is ic pa e ns ound. A: log-no mal dis ibu ion obse ed o s ain 3
(included in he as s ains g oup). B: dec easing exponen ial dis ibu ion obse ed o s ain 5 (included in he slow s ains g oup). C: pic u e om a
Ziehl-Neelsen s ained s ain 1 sample showing di e en sized co ds. O iginal magni ica ion: 2006.
doi:10.1371/jou nal.pone.0029010.g003
Figu e 4. Di ision o clinical s ains in o as and slow s ains. The s ains we e di ided in as (1, 2 and 3) and slow s ains (4, 5, 6 and 7)
acco ding o hei i ness in lung in he in i o model IV (panel A). Panel B and C show he ep esen a ion o as and slow s ains acco ding o
pa ame e s de e mined om he co ding a ea analysis: skewness (panel B), and p opo ion o single bacilli (b own ci cles) and o agg ega es (b igh
ed ci cles) (panel C). A signi ican di e ence was ound be ween as and slow s ains o all h ee g aphs.
doi:10.1371/jou nal.pone.0029010.g004
Low Dose Ae osol Fi ness P edic ing Vi ulence
PLoS ONE | www.plosone.o g 4 Janua y 2012 | Volume 7 | Issue 1 | e29010
expe imen s [22], and one passage h ough mice o guinea pig is no
su icien o es o e he i ulence [23]. The i ness alues ob ained in
i o and in i o o he e e ence s ain used he ein i wi h hese
s a emen s, as H37R Pas eu is equen ly cul u ed in i o and is
p obably be e adap ed o g ow in nu ien b o h han clinical
s ains. Fo his eason, he use o he e e ence s ain in
expe imen al in i o modeling has been also c i icized [24]. The
inabili y o s ain 7 o g ow in he 7H9 Middleb ook cul u e, and he
ac ha i achie ed he highes posi ion in he i ness anking o he
lung o AER-in ec ed mice, ein o ces he idea o his s ain being
p obably mo e used o su i ing and g owing h ough human hos
passages a he han h ough cul u e medium. These esul s clea ly
show ha he in i o i ness is no a good model o p edic ing he
i ulence o M. ube culosis s ains.
As a as he in i o models a e conce ned, i is impo an o no e
ha he duplica ion ime in lung a e IV inocula ion was he
highes egis e ed. As he lung appea s o il e he la ge pa icles
a e IV inocula ion, his sugges s ha he bacilli inside co ds
equi e a longe ime o g ow h. Likewise, hese co ds seem no o
each he spleen, which he e o e mainly ecei es single bacilli, as
is he case in lung a e AER in ec ion. The pa adox is ha when
compa ing he pulmona y g ow h o he di e en s ains in he IV
model, co ding appea s o a o a highe i ness. This means ha
o he ac o s, such as he na u e o he mac ophages ha
phagocy ose bacilli, which a e less agg essi e in he spleen han
in he lung, may play a ole [25].
The lag phase can be conside ed o be ‘‘ he in e al ha elapses
be ween he ime o seeding and he ime a which maximum a e
o g ow h begins’’ [26] o as ‘‘ he in e al ha elapses be ween he
ime o seeding and he ime o de ec able g ow h’’ [9,10].
De e mina ion o he lag phase con ains an inhe en e o
esul ing om he di e en measu emen s used in expe imen al
p ocedu es, he e o e we es ima ed his pa ame e by adap ing he
me hodology o P a s e al. [27,28], which es ima es he lag phase
on he basis o expe imen al da a.
The lag phase has been used as an indica o o M. ube culosis
i ulence only occasionally. Indeed, i is mo e o en used o
de e mine he i ness o E.coli [29,30]. No co ela ion was ound
be ween he i ness and lag phase pa ame e s o IV-in ec ed mice
and he in i o assay in he p esen s udy. This lack o co ela ion
could be due o he less equen moni o ing o CFU load in ou
expe imen s in compa ison o he hou ly moni o ing o he
Table 2. Compa ison o he i ness pa ame e s acco ding
hei duplica ion ime in hou s.
Model Max Min Mean SD % CV
IV Spleen 23.88 15.54 18.27 2.68 14.70
in i o 42.78 9.55 24.60 9.85 40.06
Ae . Lung 30.50 20.14 25.87 3.21 12.41
IV Lung 50.49 20.12 35.92 11.27 31.38
doi:10.1371/jou nal.pone.0029010. 002
Figu e 5. Fi ness and Lag phase esul s. Compa ison o i ness alues om he di e en expe imen s wi h espec o he numbe o clinical cases
de ec ed (A) and he e alua ed lag phase(B–D). A: co ela ion be ween i ness alues om lung ae osol in ec ions and he numbe o clinical cases
epo ed, which esul ed signi ican ; B–D: i ness (le axes, illed ci cles) and lag phase ( igh axes, un illed ci cles) o each expe imen wi h espec o
he di e en s ains in he lung a e ae osol in ec ion (B), lung a e IV in ec ion (C) and in i o assay (D).
doi:10.1371/jou nal.pone.0029010.g005
Low Dose Ae osol Fi ness P edic ing Vi ulence
PLoS ONE | www.plosone.o g 5 Janua y 2012 | Volume 7 | Issue 1 | e29010

BACTEC MGIT sys em in he Toungousso a pape , which ound
a end in e ms o g ow h a e in ela ion o lag phase [9]. In
con as , he lag phase esul s ob ained wi h he lungs o AER-
in ec ed mice co ela ed wi h he co esponding i ness alues.
Howe e , his co ela ion was posi i e, hus indica ing ha , in
con as o wha we expec ed o ind (high i ulence has p e iously
been associa ed wi h a sho lag phase [9]), highe i ulence esul s
in a longe lag phase. This inding could be a esul o he
me hodology used o calcula e he lag phase, which does no
conside he lag phase as he ime a which he s ain s a s o g ow
bu he ime o in e sec ion be ween he ini ial cell concen a ion
and he p olonga ion o he exponen ial g ow h phase line in a
semi-loga i hmic ep esen a ion o he g ow h cu e [17,27]. This
migh e lec he ac ha , wi h no co ding, a highe lag phase may
be impo an o adap o he s ess ul milieu o he mac ophage.
The bacilla y load in e ms o CFU esul s ob ained in he
animal model sugges ed he AER expe imen o be he bes model
o es ing i ulence as i showed a co ela ion wi h he i ness
de e mined om epo ed lung and epidemiological da a,
exp essed as numbe o diagnosed cases. I mus be no ed,
howe e , ha his i ness is ob ained wi h no implica ion o he
adap i e immune esponse. This concep comes om he dynamic
hypo hesis o la en ube culosis in ec ion, which p oposes ha a
cons an ein ec ion p ocess is needed o main ain he in ec ion
and a oid he cons an d ainage o non- eplica ing bacilli. In his
con ex , i his ein ec ion akes place in he uppe lobe, whe e a
highe g ow h is po en ia ed by he high oxygen p essu e, his
highe g ow h inc eases he likelihood o lique ac ion (and hus
ac i e disease) once he adap i e esponse appea s [16,17].
The AER model is he only one ha co ela es wi h he
physiological ansmission o TB seen in humans, hus making i
he gold s anda d o in ec ion. This also i s wi h he idea ha
cons an endogenous ein ec ion main ains he LTBI acco ding o
he dynamic hypo hesis [16]. The di e ences ound be ween his
and he IV model migh pa ially e lec he ac ha he in ec ing
ae osols measu e less han 7 mm
2
, hus meaning ha hey ha bo
i e bacilli each a mos [20]. This model does no he e o e
conside he co d- o ming abili y o each s ain, which mus be
aken in o accoun in all expe imen al models in which he s ain is
inocula ed di ec ly, whe he in i o (in anasal, IV o in a acheal
challenge) o ex i o (mac ophage in ec ion). The co d- o ming
abili y o each s ain may play a ole in all hese cases.
O he au ho s ha e ob ained a e y good co ela ion be ween
he su i al ime and he abili y o dissemina e ac i e TB when
using an in a acheal model o e alua e he su i al ime o mice
a e induc ion o massi e pneumonic disease by di ec ly
inocula ing 10
6
CFU in he lung [14]. Likewise, in e e ence
om co ding induc ion has been obse ed when e alua ing he
i ness o s ains using he mac ophage model, wi h he highes
alues being obse ed o s ains 1 and 2, o example, which did
no demons a e any dissemina ion acco ding o he epidemio-
logical da a. In e es ingly, hese s ains we e also classi ied as
highly i ulen on he basis o he co d o ma ion and i ness
ob ained using ou IV model, whe eas hey p o ided he lowes
Table 3. Fea u es o he in ec ious agen .
Iden i ica ion S ain n
Beijing
Lineage Loca ion o diagnosis Pe iod MQ i ulence le el n seconda y cases Ab esis an
Mad id 1 1 YES Mad id High - NO
Mad id 2 2 YES Mad id High - INH-R
Mad id 3 3 NO Mad id 1997–2004 Low 45 NO
Mad id 4 4 YES Mad id* Low - NO
GC 1237 5 YES G an Cana ia 1993–1996 ND 75 NO
HMS 1292 6 NO Za agoza ND - NO
MTZ 7 NO Za agoza 2001–2004 ND 85 NO
Abb e ia ions: n: numbe ; ND: No Done; MQ: Mac ophage; Ab: An ibio ic.
(*): Isola es 4 and 5 a e ep esen a i e o he same s ain and co esponded o ini ial (Isola e 5, G an Cana ia) o inal (Isola e 4, Mad id) s ages o i s dissemina ion.
doi:10.1371/jou nal.pone.0029010. 003
Table 4. Epidemiological ea u es o he hos .
Iden i ica ion Male Female Age (yea s) Immig an s* Indigency P e ious TBC
P ison
backg ound PDA HIV T ansplan O he
Mad id 1 0 1 2 Ecuado - NO - - - - -
Mad id 2 1 0 51 Pe u
´- YES (childhood) - - - - DM II
Mad id 3 82.2% 17.8% - 0,04 4.4% - 26.7% 22.2% 0,4 2.2% -
Mad id 4 - - - - - NO - - - - -
GC1237 - - 35614.8 (3 o 81) 0,19 0,12 - 0,16 0,39 0,35 - -
HMS1292 1 0 o e 18 YES - - - - YES - -
MTZ 70.6% 29.4% 30
a
0,14 - - 0,05 0,07 15.3% - -
*: Coun y o o igin.
a
: (23 o 41 wi h pe cen il25–75) (8 mon hs o 68 yea s).
Abb e ia ions: TBC: Tube culosis; PDA: Pa en e al d ug abuse s; HIV: Human immunode icieny i us; DM II: Diabe es Melli us II.
doi:10.1371/jou nal.pone.0029010. 004
Low Dose Ae osol Fi ness P edic ing Vi ulence
PLoS ONE | www.plosone.o g 6 Janua y 2012 | Volume 7 | Issue 1 | e29010
alues in he AER model [31]. This ac sugges s ha co ding
abili y appea s no o be de e minan in p edic ing i ulence. O
he h ee mo e dissemina ed s ains, #3 showed a high co ding
abili y, whe eas s ains 7 and 5 had some o he lowes co ding
abili ies. In e es ingly, he labo a o y s ain H37R showed a
o ally unique pheno ype, combining a high co ding abili y wi h a
e y good i ness in he AER model. Taking in o accoun he ease
o his in i o echnique, i would be in e es ing o e alua e he
co ding abili y when ying o de e mine he i ness o clinical
s ains.
Co ding abili y was i s desc ibed by Koch in 1882, who
e e ed o hese s uc u es as ‘‘densely bunched and b aided
g oups’’ [18]. Co ding is cha ac e ized by he o ma ion o igh
bundles o bacilli in which he o ien a ion o he long axis o each
cell is pa allel o he long axis o he co d [19]. Al hough T ehalose
dimycola e (TDM) was i s desc ibed as he only esponsible o
he co ding phenomenon [32], i has been p o en ha al e a ions
in cell wall (non ela ed o mycolic acids) can lead o loss o co ding
p ope y [33][19].
Co d o ma ion has been p e iously ela ed o i ulence [34,35]
as well as o pa hogenici y [36], al hough non-pa hogenic myco-
bac e ia a e also able o o m co ds [19].
Co ding abili y has also been ela ed o bio ilm o ma ion, a
leas in M.ma inum [37]. We ollow he c i e ia es ablished by
Julian e al. [19], who conside ed co ding o occu when bacilla y
agg ega ion ollows a de ini e o de , when deciding whe he o use
he e m co d o agg ega e o desc ibe s uc u es. When using
shacked cul u es unde cons an ae a ion in o de o ge
homogeneous inoculums, he co ding phenomenon e ol es
owa ds he o ma ion o la ge coiled s uc u es simila o
cinnamon olls, a clea consequence o he o bi al mechanics ha
o ces his sel -coiling [unpublished esul s]. Bio ilm o ma ion is
induced once spinning is s opped [da a no shown].
Fo ma ion o a bio ilm, which is he adi ional means o
ob aining BCG [38] in non-shacked cul u es, has been ela ed o
he induc ion o ole ance o esis ance agains an ibio ics [39,40]
as his is hough o be how bacilli pe sis in nec o ic issue [39].
Indeed we conside i necessa y o he bacilli o g ow unde
ex acellula condi ions. Cu en ly unpublished indings om ou
lab [unpublished esul s] sugges ha he cul u e o M. ube culosis
wi h a high dose o Tween 20 (610) s ops g ow h due o a delayed
induc ion o co ding. Mo eo e , he p esen s udy ela es his
pa icula i y o a highe i ness once a high dose is inocula ed in
he lung. I could be conside ed ha his high dose gene a es a
s ong in lamma o y esponse, hus leading o a hos ile en i on-
men o he bacilli. In con as , he low dose AER is mo e
disc e e, and hus less p one o induce a global esponse. Co ding
would be he induce and he de ense agains his hos ile
en i onmen , hus making i a i ulence p ope y.
We should emphasize ha ex acellula g ow h also occu s in TB
once lique ac ion akes place du ing e olu ion o he g anuloma
owa ds he ca i y [41]. In his con ex , he bacilli migh also g ow
by o ming co ds, hus helping o gene a e mo e des uc ion in he
lesion by inc easing i s size and e ol ing o a ca i y.
Despi e he co ela ion ound be ween i ness in lungs om
AER-in ec ed mice and epidemiological da a, i should be no ed
ha he abili y o he hos o p oduce in ec ed AER could
in luence he ansmissibili y deg ee o a clinical s ain, as
ein o ced by s udies pe o med in humans [42] and guinea pigs
[43,44]. On he o he hand, a p e ious i ulence s udy in abbi s
sugges ed ha he success ul ansmission o s ain CDC1551 was
due o i s high ansmissibili y a he han i s high i ulence [45].
Mo eo e , Table 4 shows a signi ican p opo ion o cases
associa ed wi h i ulen s ains 3 and 5 as being d ug-use s,
HIV-in ec ed and o p isone backg ound, ein o cing he weigh
o he socioeconomic ac o s. All his means ha epidemiological
dissemina ion da a should no be based solely on he in insic
i ulence o he bac e ia bu also on he hos ’s condi ions as hese
go e n he abili y o he hos o in e ac . This s udy o e s new
insigh s ha should be aken in o accoun when de e mining he
i ulence o M. ube culosis s ains. Howe e , he in i o i ness o
M. ube culosis s ains should be in e p e ed cau iously due o he
po en ial i ness cos s o adap ing o a new sys em. Though non
conclusi e, as u he s udies migh be equi ed o co obo a e i s
p edic able alue, he AER in ec ion mouse model p esen ed
he ein has p o ed o be a eliable me hod o p edic ing he
i ulence o clinical s ains, wha should b ing he oppo uni y o
implemen p e en i e measu es on ime. Finally, he co ding
phenomenon has been ela ed o i ness o he i s ime as a esul
o an exhaus i e analysis o he dis ibu ion pa e n o bacilli in o
agg ega es.
Ma e ials and Me hods
Mice
6–8-week-old emale Balb/c speci ic-pa hogen- ee (sp ) mice
we e ob ained om Ha lan Labo a o ies (San Feliu de Codines,
Ca alonia, Spain). The animals we e shipped unde sui able
condi ions, wi h he co esponding ce i ica e o heal h and o igin.
Upon a i al, he mice we e kep unde con olled condi ions in a
P3 high secu i y acili y wi h s e ile ood and wa e ad libi um.
E hics
All animal p ocedu es we e app o ed and supe ised by he
Animal Ca e Commi ee o he Ge mans T ias i Pujol Uni e si y
Hospi al and by he Depa men o En i onmen o he Ca alan
Go e nmen (app o al numbe 4095). Mice we e weighed and
checked e e y week ollowing ap o ocol ha moni o ed weigh
loss, appa en good heal h (b is led hai and wounded skin) and
beha iou (signs o agg essi eness o isola ion). Mice we e
eu hanized wi h iso lu ane.
Bac e ia
M. ube culosis H37R Pas eu , Mad id1, Mad id2, Mad id3,
Mad id4, GC1237, HMS1292 and MTZ s ains we e used in his
s udy. They we e isola ed in Spain wi hin he las 20 yea s and
we e selec ed by he Mad id and Za agoza labo a o ies o
inclusion in he s udy (Tables 3 and 4) on he basis o being highly
p e alen (s ain 3 [46,47,48]) hei in ol emen in ou b eaks (5
and 7 [46,47,48]), hei being a epea ed o phan s ain h ough
ime (s ain 6), o showing high i ulence in he mac ophage
model (s ains 1 and 2). S ains 4 and 5 we e ac ually wo clinical
isola es o he same s ain. The i s caused a la ge ou b eak in
G an Cana ia (s ain 5) and could s ill be isola ed in Mad id some
yea s la e (s ain 4).
H37R Pas eu was used as he e e ence s ain in all s udies.
Bo h he names o he s ains and he co esponding numbe s
assigned o hem a e shown in Table 2.
In o de o s anda dize he p e-inocula ion condi ions, he
ozen s ains sen by he collabo a ing labo a o ies we e
in ape i oneally inocula ed in o Balb/c mice (3 mice/s ain).
The homogenized spleens om mice in ec ed wi h each s ain
we e cul u ed in Middleb ook 7H11 aga medium o 21 days.
Visible colonies we e hen subcul u ed in Middleb ook 7H9
ae a ed and s i ed b o h kep a 37uC and supplemen ed wi h
0.2% glyce ol, 0.5% albumin-dex ose ca alase (Bec on Dickinson)
and 0.05% Tween 80. Bac e ia we e le o g ow o a la e-log
phase and s o ed a 270uC in 3 mL aliquo s.
Low Dose Ae osol Fi ness P edic ing Vi ulence
PLoS ONE | www.plosone.o g 7 Janua y 2012 | Volume 7 | Issue 1 | e29010
In i o expe imen (Con ol o bacilla y g ow h in i o)
Each s ain was cul u ed once in a Middleb ook 7H9 b o h
(150 mL) supplemen ed wi h 0.2% glyce ol, 0.5% albumin-
dex ose ca alase (Bec on Dickinson) and 0.05% Tween 80 ( inal
concen a ion o 10
5
CFU/mL) and kep a 37uC wi h cons an
ae a ion and s i ing. Samples we e ex ac ed h ee imes a week
in o de o analyze he bacilla y load.
In i o expe imen s
The in ec ion solu ion was p epa ed by dilu ing he ozen
aliquo ed s ains wi h s e ile PBS o ob ain a inal concen a ion o
10
6
CFU/mL o he AER and 5610
4
CFU/mL o he IV
in ec ion solu ion. In he AER expe imen , he animals we e
in ec ed by inocula ing 20–50 bacilli in o he lungs using a
Middleb ook de ice. In he IV expe imen , a o al olume o
0.2 mL con aining 10
4
CFU was inocula ed o each mouse IV.
Bo h he in i o expe imen s we e conduc ed once, including six
mice pe ime poin and expe imen al g oup. The numbe o
iable bac e ia in lung and spleen homogena es was measu ed a
day 0, week 1 and week 2 by pla ing duplica ed se ial dilu ions o
he issue samples. Special ca e was aken no o include hila
lymph nodes du ing emo al o he lung in o de no o a i icially
inc ease he CFU alue.
Fo he adap i e immune esponse cha ac e iza ion, six mice
we e inocula ed IV wi h H37R Pas eu and s ain 1. Animals
we e eu hanized a day 0, week 1 and week 2, when 250,000
PBMCs om spleen and hila node we e s imula ed o 18 hou s
wi h PPD (S a ens Se um Ins i u e, Copenhague) and ESAT-6
(Lionex The apeu ics & Diagnos ics L d., B aunsweig, Ge many)
a a concen a ion o 20 and 10 mg/mL, espec i ely. An
ELISPOT assay was pe o med o de e mine he speci ic-IFN-c
sec e ed by using a MABTECH mu ine IFN-cELISPOT ki
(MABTECH, Sweeden) acco ding o he manu ac u e ’s ecom-
menda ions. The Spo Fo ming Uni s (SFU) we e subsequen ly
coun ed using an ELISPOT eade .
Bacilla y load
The bacilla y load o samples ob ained in bo h he in i o and in
i o expe imen s was de e mined by cul u ing he samples on
Middleb ook 7H11 aga pla es (Bec on Dickinson, Mad id, Spain)
a 37uC o 21 days. Visible colony o ming uni s (CFU) we e
coun ed and he bacilla y load exp essed as CFU/mL in each
o gan.
Vi ulence pa ame e s
The i ness pa ame e was de ined as he maximum g ow h a e
in he semi-loga i hmic ep esen a ion o bacilla y load (log CFU/
mL) e sus ime (days). This maximum slope was de e mined by
pe o ming a linea eg ession be ween he wo consecu i e ime
poin s ha amed he maximum g ow h. The same de ini ion was
used o bo h in i o and in i o i ness assessmen s.
The lag phase pa ame e o an in i o cul u e is usually
conside ed as he in e sec ion be ween he ini ial cell concen a ion
and he p olonga ion o he exponen ial g ow h phase line in a
semi-loga i hmic ep esen a ion o he g ow h cu e [27]. This
in e sec ion occu s when he ho izon al line esul ing om he
ini ial CFU concen a ion (y = b
1
) and he line esul ing om he
maximum g ow h in exponen ial phase (y = ax+b
2
) sha e he same
y alue, which is he same CFU concen a ion; b
1
=ax+b
2
whe e
b
1
is he CFU/mL a he ini ial ime poin , a he slope and b
2
he
Y-in e cep when x = 0 and x he lag phase. Then, lag
phase = (b
1
2b
2
)/a.
Thus, he lag phase o bo h in i o and in i o g ow h was
assessed as he in e sec ion be ween he p olonga ion o he
maximum slope line de ined abo e and he ini ial le el. When he
i ness calcula ion occu ed be ween ime poin 0 and he nex , he
Lag phase was conside ed o be 0.
Image analysis o bacilli agg ega es
A o al o wo 20 mL d ops om he s ains ozen a 270uCin
he la e-log g ow h phase (see 4.3.-Bac e ia) we e ixed on a glass
slide and s ained ollowing he Ziehl-Neelsen p ocedu e. A o al o
8 o 10 pho og aphs we e aken using an Eclipse 50i mic oscope
(Nikon, Japan) and a DS-Fi 1 came a (Nikon, Japan) a 1006
using he Nis Elemen s imaging so wa e (Nikon, Japan).
Agg ega es we e hen de ec ed by image analysis using MATLAB
so wa e (MATLAB, s 7.9.0.529; The Ma hWo ks
TM
). The
o iginal colo ed image was i s con e ed in o a g ay scale image,
hen a h eshold was ixed o con e he image in o a black and
whi e one, wi h isola ed black egions being conside ed o be
‘‘agg ega es’’ ( hese being any pa icle de ec ed, om single bacilli
o la ge co ds). The a ea o each agg ega e was de e mined and
he ollowing pa ame e s calcula ed au oma ically o each s ain:
1) A ea dis ibu ion: equency o agg ega es in 16 a ea
in e als, loga i hmically dis ibu ed be ween 1 and10
6
mm
2
.
The i s column in he a ea dis ibu ion g aphics co e-
sponds o he a eas be ween 1 and 2.5 mm
2
, which we
conside ed o ep esen single bacilli.
2) Skewness: a s a is ical pa ame e ha measu es he asym-
me y o he equency dis ibu ion and is posi i e when he
ail on he igh side is longe han on he le side and he
bulk o he alues lie o he le o he mean (and ice e sa).
3) Co ding p opo ion: a io be ween he numbe o co ds and
he numbe o o al agg ega es, wi h a co d being de ined as
he agg ega es in he second column o hose wi h an a ea
g ea e han 2.5 mm
2
.
4) Single bacilli p opo ion: a io be ween he numbe o single
bacilli and he numbe o o al agg ega es, wi h single bacilli
being de ined as he agg ega es in he le -hand column o
hose smalle han 2.5 mm
2
. This pa ame e can also be
de ined as: (1 – co ding p opo ion).
5) Mean a ea: a e age a ea o all he agg ega es de ec ed.
6) Maximum a ea: size o he la ges a ea de ec ed.
Co ding s ain
Slides con aining wo 5 mL samples om exponen ial phase
H37R cul u es we e s ained using he BacLigh
TM
Bac e ial
Viabili y Ki (In i ogen, Ca lsbad, CA) acco ding o he
manu ac u e ’s ins uc ions and obse ed using a Zeiss Axioskop
epi luo escence mic oscope and he Axio ision Rel.4.8 so wa e
(Ca l Zeiss, Mad id, Spain).
S a is ical analysis
Reg ession and S uden ’s analyses we e pe o med using he
G aphPad P ism So wa e ( 4.03.354; La Jolla, Cali o nia, USA).
Di e ences we e conside ed o be s a is ically signi ican a
p,0.05.
Suppo ing In o ma ion
Figu e S1 IFN-c eleasing PBMCs in he mice i ulence
s udy. ELISPOT IFN-c esul s om he spleen and hila node o
in ec ed mice. The g aphs on he le co espond o H37R -in ec ed
Low Dose Ae osol Fi ness P edic ing Vi ulence
PLoS ONE | www.plosone.o g 8 Janua y 2012 | Volume 7 | Issue 1 | e29010
mice and hose on he igh co espond o mice in ec ed wi h s ain 1.
Do ed lines indica e he h eshold o nega i e IFN-cassay,
de e mined as he maximum alue obse ed on day 0. No signi ican
di e ences be ween day 0 and weeks 1 o 2 we e de ec ed.
(TIF)
Au ho Con ibu ions
Concei ed and designed he expe imen s: NC IL CP CV DL PJC.
Pe o med he expe imen s: NC IL EM CV. Analyzed he da a: NC IL CP
CV DL PJC. Con ibu ed eagen s/ma e ials/analysis ools: DGdV SS.
W o e he pape : NC IL CP CV DL PJC. Re iewed he manusc ip :
DGdV SS.
Re e ences
1. Casade all A, Pi o ski LA (1999) Hos -pa hogen in e ac ions: edi ining he basic
concep s o i ulence and pa hogenici y. In ec Immun 67: 3703–3713.
2. Middleb ook G (1950) The expe imen al analysis o i ulence o ube cle bacilli.
Bulle in o he New Yo k Academy o Medicine 26: 498–506.
3. O dway DJ, Sonnenbe g MG, Donahue SA, Belisle JT, O me IM (1995) D ug-
esis an s ains o Mycobac e ium ube culosis exhibi a ange o i ulence o mice.
In ec ion and Immuni y 63: 741–743.
4. Bo ell S, Gagneux S (2009) In ec iousness, ep oduc i e i ness and e olu ion o
d ug- esis an Mycobac e ium ube culosis. In e na ional Jou nal o Tube culosis and
Lung Disease 13: 1456–1466.
5. Ba ne M, Bushby SRM, Mi chison DA (1953) Tube cle bacilli esis an o
isoniazid - Vi ulence and esponse o ea men wi h isoniazid in guinea-pigs and
mice. B i ish Jou nal o Expe imen al Pa hology 34: 568–581.
6. Palanisamy GS, Smi h EE, Shanley CA, O dway DJ, O me IM, e al. (2008)
Dissemina ed disease se e i y as a measu e o i ulence o Mycobac e ium
ube culosis in he guinea pig model. Tube culosis 88: 295–306.
7. Smi h I (2003) Mycobac e ium ube culosis pa hogenesis and molecula de e minan s
o i ulence. Clinical Mic obiology Re iews 16: 463–+.
8. Billing on OJ, McHugh TD, Gillespie SH (1999) Physiological cos o i ampin
esis ance induced in i o in Mycobac e ium ube culosis. An imic obial Agen s and
Chemo he apy 43: 1866–1869.
9. Toungousso a OS, Caugan DA, Sand en P, Ma iandyshe AO, Bjune G
(2004) Impac o d ug esis ance on i ness o Mycobac e ium ube culosis s ains o
he W-Beijing geno ype. Fems Immunology and Medical Mic obiology 42:
281–290.
10. on G oll A, Ma in A, S eh M, Singh M, Po aels F, e al. (2010) Fi ness o
Mycobac e ium ube culosis S ains o he W-Beijing and Non-W-Beijing Geno ype.
Plos One 5.
11. Ma hema B, Ku epina NE, Bi ani PJ, K eiswi h BN (2006) Molecula
epidemiology o ube culosis: Cu en insigh s. Clinical Mic obiology Re iews
19: 658–+.
12. Hanekom M, an de Spuy GD, S eiche E, Ndabambi SL, McE oy CRE,
e al. (2007) A ecen ly e ol ed sublineage o he Mycobac e ium ube culosis Beijing
s ain amily is associa ed wi h an inc eased abili y o sp ead and cause disease.
Jou nal o Clinical Mic obiology 45: 1483–1490.
13. Middelkoop K, Bekke L-G, Ma hema B, Shashkina E, Ku epina N, e al. (2009)
Molecula Epidemiology o Mycobac e ium ube culosis in a Sou h A ican
Communi y wi h High HIV P e alence. Jou nal o In ec ious Diseases 200:
1207–1211.
14. Aguila D, Hanekom M, Ma a D, Gey an Pi ius NC, an Helden PD, e al.
(2010) Mycobac e ium ube culosis s ains wi h he Beijing geno ype demons a e
a iabili y in i ulence associa ed wi h ansmission. Tube culosis (Edinb) 90:
319–325.
15. Cace es N, Tapia G, Ojangu en I, Al a e F, Gil O, e al. (2009) E olu ion o
oamy mac ophages in he pulmona y g anulomas o expe imen al ube culosis
models. Tube culosis (Edinb) 89: 175–182.
16. Ca dona PJ (2009) A dynamic ein ec ion hypo hesis o la en ube culosis
in ec ion. In ec ion 37: 80–86.
17. Ca dona P-J (2011) A Spo ligh on Lique ac ion: E idence om Clinical Se ings
and Expe imen al Models in Tube culosis. Clinical & De elopmen al Immu-
nology.
18. Gao Q, K ipke K, A inc Z, Voskuil M, Small P (2004) Compa a i e exp ession
s udies o a complex pheno ype: co d o ma ion in Mycobac e ium ube culosis.
Tube culosis 84: 188–196.
19. Julian E, Roldan M, Sanchez-Cha di A, As ola O, Agus i G, e al. (2010)
Mic oscopic Co ds, a Vi ulence-Rela ed Cha ac e is ic o Mycobac e ium
ube culosis, A e Also P esen in Nonpa hogenic Mycobac e ia. Jou nal o
Bac e iology 192: 1751–1760.
20. Lou eau C, Desc oix D, Ga nie L, Delamanche I, Cha a o P, e al. (2005) A
nose-only appa a us o ai bo ne deli e y o Mycobac e ium ube culosis o mice:
calib a ion o biological pa ame e s. Mic obes In ec 7: 457–466.
21. Gillespie SH (2001) An ibio ic esis ance in he absence o selec i e p essu e.
In e na ional Jou nal o An imic obial Agen s 17: 171–176.
22. Ebe D (1998) E olu ion - Expe imen al e olu ion o pa asi es. Science 282:
1432–1435.
23. Con e se PJ, Eisenach KD, Theus SA, Nue mbe ge EL, Tyagi S, e al. (2010)
The Impac o Mouse Passaging o Mycobac e ium ube culosis S ains p io o
Vi ulence Tes ing in he Mouse and Guinea Pig Ae osol Models. Plos One 5.
24. Mendelson M, Wal e s S, Smi h I, Kaplan G (2005) S ain-speci ic
mycobac e ial lipids and he s imula ion o p o ec i e immuni y o ube culosis.
Tube culosis (Edinb) 85: 407–413.
25. Leemans JC, Ju e mans NP, Flo quin S, an Rooijen N, Ve oo deldonk MJ,
e al. (2001) Deple ion o al eola mac ophages exe s p o ec i e e ec s in
pulmona y ube culosis in mice. J Immunol 166: 4604–4611.
26. Chesney AM (1916) The la en pe iod in he g ow h o bac e ia. Jou nal o
Expe imen al Medicine 24: 387–418.
27. P a s C, Gi o A, Fe e J, Lopez D, Vi es-Rego J (2008) Analysis and IbM
simula ion o he s ages in bac e ial lag phase: basis o an upda ed de ini ion.
J Theo Biol 252: 56–68.
28. Lodge RM, Hinshelwood CN (1943) Physicochemical aspec s o bac e ial
g ow h Pa IX The lag phase o bac lac is ae ogenes. Jou nal o he Chemical
Socie y. pp 213–219.
29. Smi h MA, Bidochka MJ (1998) Bac e ial i ness and plasmid loss: he
impo ance o cul u e condi ions and plasmid size. Canadian Jou nal o
Mic obiology 44: 351–355.
30. Sleigh SC, Lenski RE (2007) E olu iona y adap a ion o eeze- haw-g ow h
cycles in Esche ichia coli. Physiological and Biochemical Zoology 80: 370–385.
31. Alonso M, Alonso Rod iguez N, Ga zelli C, Ma ı
´nez Li ola M, He anz M,
e al. (2010) Cha ac e iza ion o Mycobac e ium ube culosis Beijing isola es om he
Medi e anean a ea. BMC Mic obiol 10: 151.
32. Bloch H, So kin E, E lenmeye H (1953) A oxic lipid componen o he ube cle
bacillus (co d ac o ). I. Isola ion om pe oleum e he ex ac s o young
bac e ial cul u es. Am Re Tube c 67: 629–643.
33. Glickman MS, Cox JS, Jacobs WR (2000) A no el mycolic acid cyclop opane
syn he ase is equi ed o co ding, pe sis ence, and i ulence o Mycobac e ium
ube culosis. Mol Cell 5: 717–727.
34. Middleb ook G, Dubos RJ, Pie ce C (1947) Vi ulence and mo phological
cha ac e is ics o mammalian ube cle bacilli. Jou nal o Expe imen al Medicine
86: 175–&.
35. Howa d ST, Rhoades E, Rech J, Pang XH, Alsup A, e al. (2006) Spon aneous
e e sion o Mycobac e ium abscessus om a smoo h o a ough mo pho ype is
associa ed wi h educed exp ession o glycopep idolipid and eacquisi ion o an
in asi e pheno ype. Mic obiology-Sgm 152: 1581–1590.
36. Da zins E, Fah G (1956) Co d- o ming p ope y, le hali y and pa hogenici y o
mycobac e ia. Diseases o he Ches 30: 642–648.
37. Hall-S oodley L, B un OS, Polshyna G, Ba ke LP (2006) Mycobac e ium
ma inum bio ilm o ma ion e eals co ding mo phology. FEMS Mic obiol Le
257: 43–49.
38. Die ich G, Mollenkop HJ, Webe H, Knapp B, Diehl KD, e al. (2002)
Cul i a ion o Mycobac e ium bo is BCG in bio eac o s. J Bio echnol 96:
259–270.
39. Lenae s AJ, Ho D, Aly S, Ehle s S, And ies K, e al. (2007) Loca ion o
pe sis ing mycobac e ia in a Guinea pig model o ube culosis e ealed by
207910. An imic ob Agen s Chemo he 51: 3338–3345.
40. Ojha AK, Baughn AD, Sambandan D, Hsu T, T i elli X, e al. (2008) G ow h
o Mycobac e ium ube culosis bio ilms con aining ee mycolic acids and ha bou ing
d ug- ole an bac e ia. Mol Mic obiol 69: 164–174.
41. G osse J (2003) Mycobac e ium ube culosis in he ex acellula compa men : an
unde es ima ed ad e sa y. An imic ob Agen s Chemo he 47: 833–836.
42. Fennelly KP, Ma yny JW, Ful on KE, O me IM, Ca e DM, e al. (2004)
Cough-gene a ed ae osols o Mycobac e ium ube culosis - A new me hod o s udy
in ec iousness. Ame ican Jou nal o Respi a o y and C i ical Ca e Medicine
169: 604–609.
43. Riley RL, Shi pu i DN, Wi s ad F, Og ady F, Sul an LU, e al. (1962)
In ec iousness o ai om a ube culosis wa d - Ul a iole i adia ion o in ec ed
ai - Compa a i e in ec iousness o di e en pa ien s. Ame ican Re iew o
Respi a o y Disease 85: 511–&.
44. Escombe AR, Oese C, Gilman RH, Na incopa M, Ticona E, e al. (2007) The
de ec ion o ai bo ne ansmission o ube culosis om HIV-in ec ed pa ien s,
using an in i o ai sampling model. Clinical In ec ious Diseases 44: 1349–1357.
45. Bishai WR, Dannenbe g AM, Pa ish N, Ruiz R, Chen P, e al. (1999) Vi ulence
o Mycobac e ium ube culosis CDC1551 and H37R in abbi s e alua ed by Lu ie’s
pulmona y ube cle coun me hod. In ec ion and Immuni y 67: 4931–4934.
46. Ma in A, Cha es F, Inigo J, Alonso M, Sola C, e al. (2007) Molecula ,
epidemiological and in ec i i y cha ac e isa ion o a Mycobac e ium ube culosis
s ain p e alen in Mad id. Clin Mic obiol In ec 13: 1210–1213.
47. Lopez-Calleja AI, Ga in P, Lezcano MA, Vi o ia MA, Iglesias MJ, e al. (2009)
Unsuspec ed and ex ensi e ansmission o a d ug-suscep ible Mycobac e ium
ube culosis s ain. BMC Pulm Med 9: 3.
48. Camine o JA, Pena MJ, Campos-He e o MI, Rod iguez JC, Ga cia I, e al.
(2001) Epidemiological e idence o he sp ead o a Mycobac e ium ube culosis s ain
o he Beijing geno ype on G an Cana ia Island. Am J Respi C i Ca e Med
164: 1165–1170.
Low Dose Ae osol Fi ness P edic ing Vi ulence
PLoS ONE | www.plosone.o g 9 Janua y 2012 | Volume 7 | Issue 1 | e29010