Download “Understanding dissemination of Mycobacterium tuberculosis from

1
“Understanding dissemination of Mycobacterium tuberculosis from the lungs
2
during primary infection”
3
4
Running title: M. tb dissemination from the alveolus
5
6
Michelle B. Ryndaka, Dinesh Chandraa,c, Suman Laalb,a, *
7
a
8
York, 10016 United States of America
9
b
Department of Pathology, New York University School of Medicine, New York, New
Veterans Affairs New York Harbor Healthcare System, New York, New York, 10010
10
United States of America
11
c Present
12
College of Medicine, Bronx, NY, 10461 United States of America
address: Dinesh Chandra, Department of Microbiology, Albert Einstein
13
14
*Corresponding Author: Suman Laal
15
Email: [email protected]; Phone: 212-263-4164
16
17
Key words: Mycobacterium tuberculosis; in vitro model; alveolar barrier; ESAT-6;
18
HBHA; adhesins; toxins; dissemination; primary infection
19
Subject category: Pathogenicity and Virulence
20
Word count: Abstract (197 words); Manuscript (3707 words)
21
22
23
1
24
ABSTRACT
25
Understanding how inhaled M. tuberculosis achieve dramatic replication and cross the
26
alveolar barrier to establish systemic latent infection before adaptive immunity is elicited
27
in humans is limited by the small inoculum and the inability to identify time of infection.
28
M. tuberculosis is believed to disseminate via infected macrophages; however, like
29
other Gram-positive bacteria, M. tuberculosis could also cross the barrier directly using
30
adhesins and toxins. An in vitro alveolar-barrier mimicking gas-exchange regions of the
31
alveolus was devised comprising monolayers of human alveolar epithelial and
32
endothelial cells cultured on opposing sides of a basement membrane. Migration of
33
dissemination-competent strains of M. tuberculosis, and dissemination-attenuated M.
34
tuberculosis and M. bovis mutant strains lacking adhesin/toxin ESAT-6 and adhesin
35
HBHA were tested for macrophage-free migration across the barrier. Strains that
36
disseminate similarly in vivo migrated similarly across the in vitro alveolar-barrier.
37
Strains lacking ESAT-6 expression/secretion were attenuated, and absence of both
38
ESAT-6 and HBHA increased attenuation of bacterial migration across the barrier.
39
Thus, like other Gram-positive bacteria, M. tuberculosis utilizes adhesins and toxins for
40
macrophage-independent crossing of the alveolar-barrier. This in vitro model will allow
41
identification and characterization of
42
tuberculosis to establish systemic LTBI during primary infection.
molecules/mechanisms employed
by
43
44
Key words: Mycobacterium tuberculosis; in vitro model; alveolar barrier; ESAT-6;
45
HBHA; adhesins; toxins; dissemination; primary infection
46
2
M.
47
INTRODUCTION
48
Tuberculosis (TB) is initiated by a few Mycobacterium tuberculosis (M. tb) bacilli
49
entering the alveoli. In TB-endemic countries, infection mostly occurs during childhood,
50
and while ~10% of the infected individuals progress to primary TB, the remaining
51
infections lead to latent TB infection (LTBI). Approximately 15-20% of the reactivation
52
TB in immunocompetent individuals, and ~50% in HIV+ cases occurs exclusively at
53
extrapulmonary sites, indicating systemic bacterial dissemination during primary
54
infection. Studies in animal models demonstrate extensive bacterial replication (>20,000
55
fold) in the lungs and systemic dissemination during the first few days and weeks (2-3
56
weeks) post-aerosol infection and prior to the elicitation of adaptive immunity (Wolf et
57
al., 2008; Chackerian et al., 2002; Ordway et al., 2007). Earlier studies reported the
58
presence of M. tb DNA in normal lung and adipose tissue from autopsied non-
59
tuberculous subjects living in TB-endemic settings (Hernandez-Pando et al., 2000;
60
Neyrolles et al., 2006). Subsequently, LTBI with viable M. tb in macrophages,
61
endothelial cell (EC) and/or epithelial cells in autopsied tissue from lungs, liver, spleen
62
and kidney was demonstrated (Barrios-Payan et al., 2012). In many subjects, LTBI was
63
detected only in the extrapulmonary sites. Same cell-types infected with M. tb are also
64
present in apparently normal tissues in the mouse model of latency (Barrios-Payan et
65
al., 2012). Together these studies confirm the establishment of systemic LTBI during
66
primary infection (Barrios-Payan et al., 2012; Laal, 2012).
67
Systemic dissemination is fundamental to the pathogenesis of M. tb (Vidal
68
Pessolani et al., 2003); however, understanding M. tb dissemination during primary
69
infection in humans is limited by the small number of the bacteria that are inhaled, and
3
70
the inability to identify the time of infection. It is believed that dissemination of M. tb
71
occurs by the “Trojan Horse” mechanism wherein alveolar macrophages that
72
phagocytose M. tb invade the epithelium, elicit inflammatory responses, and recruit
73
macrophages from the blood stream, which in turn transport the bacteria systemically.
74
However, other bacteria also utilize repertoires of adhesins, which promote
75
adherence/invasion of the barrier cells, and toxins, which lyse these cells to increase
76
barrier permeability, in order to directly cross their relevant barriers. Studies with an in
77
vitro alveolar bilayer model (A549- apical:EAhy926-basolateral) implicated the use of
78
both the “Trojan horse” and the “direct migration” mechanisms by M. tb (Bermudez et
79
al., 2002). However, this model lacked a basement membrane (BM) between the
80
alveolar epithelial cell (AEC) and EC bilayer, and the direct bacterial migration of M. tb
81
could be ascribed to an incomplete BM.
82
M. tb expresses several adhesins, HBHA, malate synthase (Rv1837c; MS), and
83
ESAT-6, that enable bacterial attachment to the AEC (Pethe et al., 2001; Kinhikar et al.,
84
2006; Kinhikar et al., 2010). HBHA, a heparan sulphate proteoglycan receptor-binding
85
adhesion, promotes adherence to AEC but not to macrophages (Pethe et al., 2001).
86
ESAT-6 is both a laminin-binding adhesin that binds AEC and is a pore-forming toxin
87
that lyses macrophages and AEC in vivo and in vitro (Hsu et al., 2003; Kinhikar et al.,
88
2010; McDonough & Kress, 1995) Transcripts for ESAT-6, MS, and HBHA are
89
upregulated in M. tb in AEC cells (Ryndak et al., 2015; Delogu et al., 2006), while they
90
are unaffected in bacteria replicating in macrophages (Schnappinger et al., 2003;
91
Fontan et al., 2008). M. bovis BCG, M. tb H37Rv esat-6 and M. bovis BCG hbha are
92
attenuated for dissemination from the lungs in animal models (Lewis et al., 2003; Pethe
4
93
et al., 2001). Together, these results raise the possibility that like other bacteria, M. tb
94
also uses adhesins and toxins to achieve “direct migration” across the alveolar barrier
95
by transcellular and/or paracellular mechanisms.
96
To investigate roles for HBHA and ESAT-6 in the direct migration of M. tb across
97
the alveolar barrier, we modified the in vitro alveolar barrier to address the shortcomings
98
of the earlier model (Bermudez et al., 2002). While AECs express the components of
99
the BM in vitro, the provision of exogenous laminin is necessary for complete BM
100
formation (Furuyama & Mochitate, 2000). Our model therefore incorporates a BM
101
between the AEC and EC monolayers and uses the more relevant human pulmonary
102
artery EC line, HPAE-26.
103
Studies with our human in vitro alveolar barrier demonstrate the macrophage-
104
free migration of clinical M. tb strains. ESAT-6 mutants, M. bovis BCG and M. tb H37Rv
105
esat-6, and ESAT-6 secretion-deficient M. tb H37Rv phoP are similarly attenuated for
106
transmigration; and the M. bovis BCG hbha (lacking both ESAT-6 and HBHA),
107
demonstrates greater attenuation of transmigration compared to strains lacking ESAT-6
108
alone. These results provide evidence that like other Gram-positive bacteria, M. tb may
109
also employ the synergistic activity of adhesins and toxins to achieve macrophage-
110
independent direct transmigration across the alveolar barrier during primary infection.
111
112
MATERIALS AND METHODS
113
M. tb infection and replication in HPAE-26 cells: The earlier in vitro alveolar bilayer
114
barrier included bilayers of AEC (A549; human type II AEC), but a non-physiological EC
115
line (EAhy926) which is a hybrid of human umbilical vein EC and A549 cells) (Bermudez
5
116
et al., 2002). To confirm that human pulmonary artery EC (HPAE-26) (Coriell;
117
WC00107) get infected with and support M. tb replication (Fig. 1), 5 x 105/well HPAE-26
118
were seeded into wells of a 24 well plate (or glass coverslips in wells for microscopy)
119
pre-coated with 0.1% pig gelatin 10 min and cultured in F12K medium supplemented
120
with 10% fetal bovine serum (FBS), 0.1 mg/ml heparin and 0.03 mg/ml endothelial cell
121
growth factor and grown to confluency. Confluent monolayers were washed with F12K
122
medium and infected with M. tb H37Rv at multiplicities of infection (MOI) 1:1 and 5:1
123
bacteria to cell for 2 hr. Infected cells were washed thrice with F12K medium and
124
remaining extracellular bacteria were killed by amikacin treatment (200 ug/ml) for 1 hr.
125
Amikacin was removed by washing thrice with F12K medium and replaced with F12K
126
1% FCS. At indicated time points and in triplicate, infected cells were washed thrice with
127
F12K medium and lysed with 0.1% triton X-100 for colony forming units (CFUs) (Fig. 1a)
128
or fixed with 60% acetone 20 min for staining and light microscopy (Fig. 1b). Lysates
129
were serially diluted, plated onto 7H11 Middlebrook plates and incubated at 37C for
130
CFUs. Fixed infected cells on coverslips were acid-fast stained by the Kinyoun stain (TB
131
Stain kit K; Fisher Scientific), visualized under oil-emersion (Nikon Eclipse 80i light
132
microscope) and images captured using a mounted Nikon CoolPix 9400 digital camera.
133
Human in vitro alveolar barrier: To mimic the human in vivo alveolar barrier (Fig. 2a and
134
2b), BM preparation (Matrigel; BD Biosiences), diluted in ice-cold serum-free medium,
135
was pipetted (60 g/cm2) onto 3 m porous membranes of transwell permeable
136
supports (Costar), incubated 1 hr at 37°C followed by overnight at room temperature.
137
The polymerized BM was moistened for 1-2 hr at 37°C with serum-free F12K media,
138
surplus media removed, and inserts placed upside-down in wells of a 24-well plate. The
6
139
exposed bottom sides of the transwells were seeded with 10 5/100l HPAE-26 cells in
140
F12K containing glutamine, 10% FBS, heparin (0.05 mg/ml) and EC growth supplement
141
(ECGS) (0.02 mg/ml) and the cell-seeded inverted transwells left undisturbed for ~45
142
min to allow cellular adherence. The HPAE-26-seeded inserts were gently turned into
143
upright position into wells containing 500l F12K-glutamine/FBS/heparin/ECGS, and
144
200l F12K-glutamine/10% FBS added to upper chambers. The slower growth rate of
145
HPAE-26 cells compared to A549 cells requires that HPAE-26 cells be seeded to the
146
basal surface of the BM-coated transwells 2 days prior to the seeding of A549 cells to
147
the apical surface, to enable confluent monolayers of both cell-lines to form
148
simultaneously. At 48 hr post-HPAE-26 seeding, the upper chambers are seeded with
149
105/200l A549 cells (ATCC; CCL-185). Transwells with similarly seeded HPAE-26 or
150
A549 cells alone were included as controls. Starting one day post-seeding for each cell
151
type, the transmembrane electrical resistance (TER) of the monolayers/bilayer in
152
triplicate wells was monitored daily using a MilliCell ERS Volt/Ohm meter (Millipore) until
153
stabilization (Fig. 2c). TER of duplicate BM-infused, cell-free transwells were also
154
recorded daily to account for fluctuations in TER readings.
155
Mycobacterial strains: M. tb H37Rv and clinical isolates CDC1551 and HN878 exhibit
156
similar dissemination from the lungs in aerosol-infected animal models (Palanisamy et
157
al., 2008). M. tb esat-6 and M. bovis BCG (attenuated for expression of ESAT-6 and
158
in vivo dissemination) (Hsu et al., 2003; Lewis et al., 2003; Guinn et al., 2004), and M.
159
tb H37Rv phoP (deficient for secretion of ESAT-6) (Frigui et al., 2008) were tested in
160
parallel with wild type M. tb H37Rv. In addition, M. bovis BCG hbha (attenuated for
7
161
expression of both ESAT-6 and HBHA) (Pethe et al., 2001) was evaluated for migration
162
across the human in vitro alveolar barrier.
163
Trans-alveolar barrier migration experiments: Mid-log phase bacteria (cultured in 7H9
164
broth supplemented with 0.2% glycerol, 10% ADC, 0.05% Tween80) in single-cell
165
suspensions (~107 CFU; 200 l) were added to upper chambers of fully-formed in vitro
166
alveolar barriers in triplicate for each strain. Inocula CFUs were verified by plating serial
167
dilutions on 7H11 Middlebrook agar. At 72 hr post-infection, media from the bottom
168
chambers were collected, diluted, and plated for CFUs. Percent migration was
169
calculated (# CFU in bottom chamber/# CFU in inoculum) x 100. Percent migrations
170
were normalized in comparison to the value obtained for the M. tb H37Rv assayed in
171
parallel and set at 1. (Absolute percent migrations of M. tb H37Rv ranged 0.15 – 1% in
172
separate experiments.) The Mann Whitney one tailed test was used to determine
173
statistical significance (GraphPad Prism 5 software); P values ≤ 0.05 were considered
174
statistically significant. (While attenuations of M. tb H37Rv esat-6 and M. bovis BCG
175
compared to M. tb H37Rv wild type and of M. bovis BCG hbha to BCG wild type were
176
verified in additional separate experiments, the migration of M. tb H37Rv phoP across
177
the in vitro alveolar barrier was performed once in comparison to M. tb H37Rv wild type
178
and M. bovis BCG.)
179
180
RESULTS
181
HPAE-26 cells are infected with and support M. tb replication: Human cell-line
182
A549 is widely used for studies of interactions of other pulmonary pathogens e.g.,
183
Influenza A, Streptococcus pneumoniae, Staphylococcus aureus etc., with human AEC.
8
184
In contrast to the extensive information on M. tb invasion and replication in A549 cells,
185
studies with human HPAE-26 EC are not reported. Our results demonstrate that at
186
MOIs of 1:1 and 5:1, intra-HPAE-26 M. tb H37Rv CFU increased by ~1.7 and ~3.6 fold
187
by day 3, and ~12.8 and ~8.5 fold by day 5, respectively (Fig. 1a). Kinyoun staining of
188
fixed M. tb-infected HPAE-26 revealed increasing numbers of acid-fast bacteria in the
189
cells over time (Fig. 1b). Thus M. tb infects EC and replicates intracellularly in these
190
cells.
191
Human in vitro alveolar barrier: A bilayer model of the alveolar barrier based on
192
monolayers of human AEC and EC on opposing sides of a BM-infused permeable
193
transwell membrane was devised (Fig. 2b). There was no effect on morphology or
194
cellular replication of either cell type during growth on BM (data not shown). The
195
developing monolayers of the two cell-types were monitored individually and in parallel
196
with the bilayer by measuring the TER between the upper and bottom chambers of
197
triplicate wells (Fig. 2c). TER measurements for EC-seeded transwell inserts increase
198
from 1 day post-seeding (~382 +/-14 ohms/cm2) to peak/plateau at 8 days post-seeding
199
(~464 +/- 64 ohms/cm2). TER values in AEC-seeded transwell inserts also increase
200
from 1 day post-seeding (~398 +/-4 ohms/cm2) to peak/plateau 6 days post-seeding
201
(~472 +/- 19 ohms/cm2). The initial TER for the bilayer (419 +/-11 ohms/cm2) at 1 day
202
post-A549 seeding increases to peak/plateau at ~493 +/- 9 ohms/cm2 on day 8 post-
203
EC-seeding (day 6 post- AEC-seeding) and is higher than the TER for either monolayer
204
alone. The resulting bilayer mimics the gas exchange regions of the human alveolar
205
barrier in that the AEC and EC share a BM with the apical AEC-seeded side
9
206
representing the alveolar lumen and the basolateral EC-seeded side representing the
207
capillary lumen (Fig. 2a and 2b).
208
Migration of M. tb H37Rv and clinical isolates across the human in vitro alveolar
209
barrier: M. tb H37Rv, CDC1551 and HN878, strains that disseminate similarly in vivo
210
(Palanisamy et al., 2008), demonstrated similar migration across the human in vitro
211
alveolar barrier (Fig. 3a). There was no statistically significant difference between the
212
percent migration of the three strains (0.99 +/- 0.33%; 1.0 +/- 0.14%; 1.0 +/-0.13%,
213
respectively) suggesting that the bacterial factors and mechanisms responsible for
214
direct migration across the alveolar barrier are conserved and similar in all 3 strains.
215
Migration of in vivo dissemination–attenuated mycobacterial strains across the
216
human in vitro alveolar barrier: The migrations of M. tb H37Rv esat-6, M. bovis
217
BCG, and M. tb H37Rv phoP were evaluated in parallel with M. tb H37Rv (Fig. 3b).
218
The three mutant strains were all similarly attenuated for migration across the human in
219
vitro alveolar barrier (~50% compared to H37Rv). There was no significant difference
220
between the percent migrations of the three mutant strains (Fig. 3b). These results
221
demonstrate a role for ESAT-6 in the direct crossing of M. tb across the alveolar barrier.
222
Disruption/deletion of hbha in M. tb or M. bovis BCG causes attenuated
223
dissemination from the lungs of intra-nasally-infected mice (Pethe et al., 2001). The
224
migration of H37Rv, BCG, and BCG hbha were also evaluated in parallel (Figure 3B).
225
Migration of BCG hbha was 40% lower than BCG (P value = 0.05) and 72% lower than
226
H37Rv (P value = 0.05). Migration of BCG hbha, was diminished compared to both M.
227
tb esat-6 and M. tb phoP as well as BCG, all of which express hbha in a background
228
attenuated for esat-6 expression and/or ESAT-6 secretion. Importantly, simultaneous
10
229
deficiency of two adhesins demonstrated increased attenuation of bacterial migration
230
across the barrier. Thus, strains that exhibit attenuated dissemination from the lung in
231
vivo in animal models are also attenuated for migration across the human in vitro
232
alveolar barrier, validating the model for studies of macrophage-independent M. tb
233
dissemination across the alveolar barrier.
234
235
DISCUSSION
236
In vitro models of mucosal barriers have enhanced our understanding of the
237
mechanisms used by pathogens such as Shigella spp., enteropathogenic Yersinia, and
238
Streptococcus pneumoniae, just to name a few, to cross their relevant barriers
239
(McCormick, 2003). To investigate the M. tb proteins and the mechanisms employed by
240
M. tb to directly cross the alveolar barrier during primary infection, we devised an
241
accurate mimic of the human in vivo alveolar barrier. M. tb strains that disseminate
242
similarly in animal models migrated similarly across this human in vitro alveolar barrier
243
(Fig. 3a). Importantly, BCG and M. tb esat-6, which are attenuated for dissemination
244
from the lungs in aerosol-infected mice, are equally attenuated for transmigration across
245
the human in vitro alveolar barrier (Fig. 3b), as is M. tb phoP (deficient for ESAT-6
246
secretion) (Fig. 3b). The BCG hbha strain demonstrated enhanced attenuation for
247
macrophage-independent transmigration compared to the mutants lacking only ESAT-6
248
(Fig. 3b). Together, these results provide evidence for synergistic participation of ESAT-
249
6 and HBHA in direct migration across the alveolar barrier in humans.
250
The massive replication of M. tb in a non-migrating, non-antigen-presenting
251
compartment of the lungs in aerosol-infected mice (Wolf et al., 2008), the presence of
11
252
M. tb in the AEC in non-TB humans (Barrios-Payan et al., 2012; Hernandez-Pando et
253
al., 2000), and the enhanced infectivity of M. tb retrieved from type II AEC to cross a
254
bilayer of A549 and Eahy926 cells (Bermudez et al., 2002) indicate that M. tb exploits
255
the non-phagocytic intra-AEC environment to both replicate actively (Wolf et al., 2008;
256
Mehta et al., 1996) and to transition to a disseminative phenotype. This is supported by
257
the transcriptional profile of M. tb replicating in human type II AEC (Ryndak et al., 2015).
258
Thus, intra-AEC M. tb downregulate DosR (DevR) regulon genes and dormancy-
259
inducing stress-related genes, and upregulate genes involved in energy production,
260
protein synthesis, and cell-wall synthesis (Ryndak et al., 2015). In contrast, intra-
261
macrophage M. tb show upregulation of DosR (DevR) regulon genes and stress-related
262
genes associated with transition to dormancy (Schnappinger et al., 2003; Fontan et al.,
263
2008). While esat-6 is downregulated in M. tb in macrophages, esat-6 and 5 genes
264
encoding ESAT-6-like proteins, esxH, esxW, esxJ, esxK and esxP, are upregulated in
265
M. tb in type II AEC (Ryndak et al., 2015), as are genes encoding the laminin-binding
266
adhesin MS which promotes bacterial adherence to type II AEC (Kinhikar et al., 2006),
267
Ag 85C (Rv0129c) which is a member of the fibronectin-binding Ag85 complex (Kuo et
268
al., 2012) and Rv3922c which has sequence similarity to hemolysins of other bacteria.
269
Importantly, transcripts for these genes are unaffected during bacterial residence in
270
macrophages (Schnappinger et al., 2003; Fontan et al., 2008).
271
Bacterial dissemination requires that after crossing the alveolar barrier, M. tb
272
survive and adapt to the blood to eventually infect the cells of other organs. Our studies
273
of ex vivo adaptation of M. tb to whole blood from HIV- and HIV+ subjects demonstrated
274
that esat-6 is upregulated in M. tb replicating in both environments (Ryndak et al.,
12
275
2014); esat-6 was the most upregulated of ~150 upregulated M. tb genes in blood from
276
HIV+ patients. As in AEC, upregulation of esat-6 was accompanied by upregulation of
277
esxW, esxJ, esxK and esxP. The higher susceptibility of HIV+ patients to M. tb infection
278
as well as disseminated and extrapulmonary TB (EPTB) further emphasizes a potential
279
role for these ESAT-6-like proteins in dissemination of M. tb.
280
Other bacterial pathogens cross the relevant mucosal barriers by synergistic use
281
of repertoires of adhesins and lysins/toxins (Doran et al., 2013). Adhesins enable
282
bacterial anchoring to the cells of the barrier via binding to components of the ECM/BM
283
(laminin, proteoglycans, collagen, fibronectin, elastin etc), and toxins facilitate
284
dissemination by increasing barrier permeability via cell lysis. The upregulation of
285
transcripts for east-6 and hbha in M. tb replicating in AECs (Ryndak et al., 2015;
286
Kinhikar et al., 2010; Delogu et al., 2006), and the greater attenuation of BCG hbha
287
strain for migration across the in vitro alveolar barrier is reminiscent of the synergism of
288
adhesin(s) and toxin(s) reported for other Gram-positive bacteria for direct
289
dissemination. Thus, L. monocytogenes adhesins (FbpA, ActA, InlA) promote
290
attachment/invasion of small intestinal enterocytes (Dramsi et al., 2004; Suarez et al.,
291
2001; Bierne et al., 2007) and Listeriolysin O (LLO), a pore-forming toxin, impairs the
292
integrity of the intestinal epithelial barrier (Richter et al., 2009). S. pneumonia adhesins
293
PspC, CbpA and PsaA promote adherence to type II AEC (Nobbs et al., 2009; Chen et
294
al., 2010); while pneumolysin (Ply) is cytotoxic to AEC and EC resulting in increased
295
alveolar permeability (Marriott et al., 2008). S. aureus adhesin SpA promotes invasion
296
across airway epithelial cells via paracellular junctions, alters the epithelial barrier to
297
expose ECM, and enables the S. aureus fibronectin-binding proteins (FnbP) to bind to
13
298
EC (Soong et al., 2011; Sinha et al., 1999), and S. aureus α-toxin detaches epithelial
299
cells from the BM to increase the microvascular EC barrier permeability to promote
300
invasive disease (Phillips et al., 2006; Powers et al., 2012).
301
That relatively low numbers of bacteria (≤ 1%) cross the in vitro alveolar barrier is
302
likely a consequence of the limited time for which the in vitro barrier can be maintained
303
(3-4 days). The extensive replication of M. tb in AEC leads to necrosis and lysis (Dobos
304
et al., 2000; McDonough & Kress, 1995) and subsequent infection of the adjacent
305
AECs (Castro-Garza et al., 2002). Thus, in vivo, M. tb would replicate in the infected
306
AECs, transform to a disseminative phenotype, lyse the cells, spread to adjacent cells,
307
and continue to replicate/lyse, simultaneously disseminating systemically to achieve
308
widespread infection before adaptive immune responses that induce latency are elicited
309
(Wolf et al., 2008).
310
The lack of ESAT-6 and HBHA does not completely abolish dissemination from
311
the mouse lungs (Pethe et al., 2001; Lewis et al., 2003) or migration across the human
312
in vitro alveolar barrier (Fig. 3b), indicating participation of additional bacterial factors.
313
While the functional activities of EsxH, EsxW, EsxJ, EsxK and EsxP are not known, like
314
ESAT-6, these ESAT-6-like proteins are small (~100 amino acids) secreted proteins that
315
lack a secretion signal and share a known or predicted helix-turn-helix structure and
316
central WXG motif (Pallen, 2002). The conservation of these structural features with
317
ESAT-6 despite lack of significant sequence similarity suggests similarity in function
318
thus implicating a potential role for them as adhesins and/or toxins. The human in vitro
319
alveolar barrier described herein will enable exploration of their role in bacterial
320
dissemination from the lungs.
14
321
Strain/lineage-specific differences in virulence phenotype and tropism for
322
extrapulmonary sites are increasingly reported for M. tb (Gagneux & Small, 2007;
323
Palanisamy et al., 2009; Click et al., 2012; Caws et al., 2008). Thus, M. tb strains from
324
CSF of TB meningitis patients disseminate more efficiently than strains from pulmonary
325
TB (PTB) patients (Hernandez-Pando et al., 2010). W-Beijing strains of the East Asian
326
lineage are three times more likely to cause EPTB compared to non-Beijing strains
327
(Kong et al., 2007). The highly transmissible Beijing and F15/LAM4/KZN M. tb families
328
adhere and invade type II AEC with greater efficiency than genetically different clinical
329
strains and M. tb H37Rv (Ashiru et al., 2010). The human in vitro alveolar barrier could
330
also enable understanding of how M. tb strains/lineages differ in their dissemination
331
ability.
332
There are limitations to our human in vitro alveolar barrier model. First, the in vivo
333
alveolar barrier is comprised of both type I and type II AEC; the model includes only the
334
latter cell-type. There is no true type I AEC line available, and the current type I model
335
cell-line, WI-26, which is lysed by ESAT-6 (Kinhikar et al., 2010) and CFP21 (Vir et al.,
336
2014), is a lung fibroblast line with epithelial-like morphology. Second, our model mimics
337
only the gas-exchange region of the alveoli where there is no interstitial ECM. Other
338
adhesins and toxins may interact with different ECM components in the interstitium.
339
Finally, the contribution of inflammatory responses that damage the alveolar barrier in
340
vivo would not be captured in our model. Nevertheless, the current model will contribute
341
to the identification and characterization of additional M. tb adhesins and lysins that
342
participate in direct bacterial dissemination from the lungs, allow exploration of the
343
pathogen-alveolar interactions of M. tb strains with tropism for extrapulmonary sites of
15
344
infection, and promote understanding of the mechanisms employed by M. tb to establish
345
systemic latent infection.
346
347
ACKNOWLEDGEMENTS
348
We would like to acknowledge the sources of the mycobacterial strains used in this
349
study: Dr. William Jacobs Jr. (Albert Einstein College of Medicine, Bronx, NY) for M. tb
350
H37Rv and M. tb H37Rv esat-6; Dr. Michael Brennan (Aeras, Rockville, MD) for M.
351
bovis BCG and M. bovis hbha BCG; Dr. Issar Smith for M. tb H37Rv phoP and Dr.
352
Gilla Kaplan for clinical isolates CDC1551 and HN878 (both of New Jersey Medical
353
School, Rutgers, State University of New Jersey). This work was supported by research
354
funds from the Department of Veterans Affairs, Veterans Health Administration, Office
355
of Research and Development, and with salary support to Research Microbiologist and
356
Research Career Scientist, SL.
357
358
REFERENCES
359
360
ASHIRU, O. T., PILLAY, M. & STURM, A. W. (2010) Adhesion to and invasion of
361
pulmonary epithelial cells by the F15/LAM4/KZN and Beijing strains of
362
Mycobacterium tuberculosis. J Med Microbiol, 59, 528-33.
363
BARRIOS-PAYAN, J., SAQUI-SALCES, M., JEYANATHAN, M., ALCANTARA-
364
VAZQUEZ, A., CASTANON-ARREOLA, M., ROOK, G. & HERNANDEZ-PANDO,
365
R. (2012) Extrapulmonary locations of Mycobacterium tuberculosis DNA during
366
latent infection. J Infect Dis, 206, 1194-205.
16
367
BERMUDEZ, L. E., SANGARI, F. J., KOLONOSKI, P., PETROFSKY, M. & GOODMAN,
368
J. (2002) The efficiency of the translocation of Mycobacterium tuberculosis
369
across a bilayer of epithelial and endothelial cells as a model of the alveolar wall
370
is a consequence of transport within mononuclear phagocytes and invasion of
371
alveolar epithelial cells. Infect Immun, 70, 140-6.
372
BIERNE, H., SABET, C., PERSONNIC, N. & COSSART, P. (2007) Internalins: a
373
complex
family
of
leucine-rich
repeat-containing
374
monocytogenes. Microbes Infect, 9, 1156-66.
proteins
in
Listeria
375
CASTRO-GARZA, J., KING, C. H., SWORDS, W. E. & QUINN, F. D. (2002)
376
Demonstration of spread by Mycobacterium tuberculosis bacilli in A549 epithelial
377
cell monolayers. FEMS Microbiol Lett, 212, 145-9.
378
CAWS, M., THWAITES, G., DUNSTAN, S., HAWN, T. R., LAN, N. T., THUONG, N. T.,
379
STEPNIEWSKA, K., HUYEN, M. N., BANG, N. D., LOC, T. H., GAGNEUX, S.,
380
VAN SOOLINGEN, D., KREMER, K., VAN DER SANDE, M., SMALL, P., ANH,
381
P. T., CHINH, N. T., QUY, H. T., DUYEN, N. T., THO, D. Q., HIEU, N. T.,
382
TOROK, E., HIEN, T. T., DUNG, N. H., NHU, N. T., DUY, P. M., VAN VINH
383
CHAU, N. & FARRAR, J. (2008) The influence of host and bacterial genotype on
384
the development of disseminated disease with Mycobacterium tuberculosis.
385
PLoS Pathog, 4, e1000034.
386
CHACKERIAN, A. A., ALT, J. M., PERERA, T. V., DASCHER, C. C. & BEHAR, S. M.
387
(2002) Dissemination of Mycobacterium tuberculosis is influenced by host factors
388
and precedes the initiation of T-cell immunity. Infect Immun, 70, 4501-9.
17
389
CHEN, S. M., TSAI, Y. S., WU, C. M., LIAO, S. K., WU, L. C., CHANG, C. S., LIU, Y. H.
390
& TSAI, P. J. (2010) Streptococcal collagen-like surface protein 1 promotes
391
adhesion to the respiratory epithelial cell. BMC Microbiol, 10, 320.
392
CLICK, E. S., MOONAN, P. K., WINSTON, C. A., COWAN, L. S. & OELTMANN, J. E.
393
(2012) Relationship Between Mycobacterium tuberculosis Phylogenetic Lineage
394
and Clinical Site of Tuberculosis. Clinical Infectious Diseases, 54, 211-219.
395
DELOGU, G., SANGUINETTI, M., POSTERARO, B., ROCCA, S., ZANETTI, S. &
396
FADDA, G. (2006) The hbhA gene of Mycobacterium tuberculosis is specifically
397
upregulated in the lungs but not in the spleens of aerogenically infected mice.
398
Infect Immun, 74, 3006-11.
399
DOBOS, K. M., SPOTTS, E. A., QUINN, F. D. & KING, C. H. (2000) Necrosis of lung
400
epithelial cells during infection with Mycobacterium tuberculosis is preceded by
401
cell permeation. Infect Immun, 68, 6300-10.
402
403
DORAN, K. S., BANERJEE, A., DISSON, O. & LECUIT, M. (2013 ) Concepts and
mechanisms: crossing host barriers. Cold Spring Harb Perspect Med, 3.
404
DRAMSI, S., BOURDICHON, F., CABANES, D., LECUIT, M., FSIHI, H. & COSSART,
405
P. (2004) FbpA, a novel multifunctional Listeria monocytogenes virulence factor.
406
Mol Microbiol, 53, 639-49.
407
FONTAN, P., ARIS, V., GHANNY, S., SOTEROPOULOS, P. & SMITH, I. (2008) Global
408
transcriptional profile of Mycobacterium tuberculosis during THP-1 human
409
macrophage infection. Infect Immun, 76, 717-25.
410
FRIGUI, W., BOTTAI, D., MAJLESSI, L., MONOT, M., JOSSELIN, E., BRODIN, P.,
411
GARNIER, T., GICQUEL, B., MARTIN, C., LECLERC, C., COLE, S. T. &
18
412
BROSCH, R. (2008) Control of M. tuberculosis ESAT-6 secretion and specific T
413
cell recognition by PhoP. PLoS Pathog, 4, e33.
414
FURUYAMA, A. & MOCHITATE, K. (2000) Assembly of the exogenous extracellular
415
matrix during basement membrane formation by alveolar epithelial cells in vitro. J
416
Cell Sci, 113 ( Pt 5), 859-68.
417
GAGNEUX, S. & SMALL, P. M. (2007) Global phylogeography of Mycobacterium
418
tuberculosis and implications for tuberculosis product development. Lancet Infect
419
Dis, 7, 328-37.
420
GUINN, K. M., HICKEY, M. J., MATHUR, S. K., ZAKEL, K. L., GROTZKE, J. E.,
421
LEWINSOHN, D. M., SMITH, S. & SHERMAN, D. R. (2004) Individual RD1-
422
region genes are required for export of ESAT-6/CFP-10 and for virulence of
423
Mycobacterium tuberculosis. Mol Microbiol, 51, 359-70.
424
HERNANDEZ-PANDO, R., AGUILAR, D., COHEN, I., GUERRERO, M., RIBON, W.,
425
ACOSTA, P., OROZCO, H., MARQUINA, B., SALINAS, C., REMBAO, D. &
426
ESPITIA, C. (2010) Specific bacterial genotypes of Mycobacterium tuberculosis
427
cause extensive dissemination and brain infection in an experimental model.
428
Tuberculosis (Edinb), 90, 268-77.
429
HERNANDEZ-PANDO, R., JEYANATHAN, M., MENGISTU, G., AGUILAR, D.,
430
OROZCO, H., HARBOE, M., ROOK, G. A. W. & BYUNE, G. (2000) Persistence
431
of DNA from Mycobaterium tuberculosis in superficially normal lung tissue during
432
latent infection. The Lancet, 356, 2133-2138.
433
HSU, T., HINGLEY-WILSON, S. M., CHEN, B., CHEN, M., DAI, A. Z., MORIN, P. M.,
434
MARKS, C. B., PADIYAR, J., GOULDING, C., GINGERY, M., EISENBERG, D.,
19
435
RUSSELL, R. G., DERRICK, S. C., COLLINS, F. M., MORRIS, S. L., KING, C. H.
436
& JACOBS, W. R., JR. (2003) The primary mechanism of attenuation of bacillus
437
Calmette-Guerin is a loss of secreted lytic function required for invasion of lung
438
interstitial tissue. Proc Natl Acad Sci U S A, 100, 12420-5.
439
KINHIKAR, A., VARGAS, D., LI, H., MAHAFFEY, S. B., HINDS, L., BELISLE, J. T. &
440
LAAL, S. (2006) Mycobacterium tuberculosis malate synthase is a laminin
441
binding adhesin. Mol . Microbiol., 60, 999-1013.
442
KINHIKAR, A., VERMA, I., CHANDRA, D., SINGH, K. K., WELDINGH, K., ANDERSEN,
443
P., HSU, T., JACOBS, W. R., JR. & LAAL, S. (2010) Potential Role for ESAT-6 in
444
Dissemination of M. tuberculosis via Human Lung Epithelial Cells. Mol Microbiol,
445
75, 92-106.
446
KONG, Y., CAVE, M. D., ZHANG, L., FOXMAN, B., MARRS, C. F., BATES, J. H. &
447
YANG, Z. H. (2007) Association between Mycobacterium tuberculosis Beijing/W
448
lineage
449
epidemiologic and clinical characterization of the three principal genetic groups of
450
M. tuberculosis clinical isolates. J Clin Microbiol, 45, 409-14.
strain
infection
and
extrathoracic
tuberculosis:
Insights
from
451
KUO, C. J., BELL, H., HSIEH, C. L., PTAK, C. P. & CHANG, Y. F. (2012) Novel
452
mycobacteria antigen 85 complex binding motif on fibronectin. J Biol Chem, 287,
453
1892-902.
454
455
LAAL, S. (2012) How does Mycobacterium tuberculosis establish infection? J Infect Dis,
206, 1157-9.
20
456
LEWIS, K. N., LIAO, R., GUINN, K. M., HICKEY, M. J., SMITH, S., BEHR, M. A. &
457
SHERMAN, D. R. (2003) Deletion of RD1 from Mycobacterium tuberculosis
458
mimics bacille Calmette-Guerin attenuation. J Infect Dis, 187, 117-23.
459
MARRIOTT, H. M., MITCHELL, T. J. & DOCKRELL, D. H. (2008) Pneumolysin: a
460
double-edged sword during the host-pathogen interaction. Curr Mol Med, 8, 497-
461
509.
462
463
McCORMICK, B. A. (2003) The use of transepithelial models to examine host-pathogen
interactions. Curr Opin Microbiol, 6, 77-81.
464
MCDONOUGH, K. & KRESS, Y. (1995) Cytotoxicity for lung epithelial cells is a
465
virulence-associated phenotype of Mycobacterium tuberculosis. Infect. Immun.,
466
63, 4802-4811.
467
MEHTA, P. K., KING, C. H., WHITE, E. H., MURTAGH, J. J., JR. & QUINN, F. D.
468
(1996) Comparison of in vitro models for the study of Mycobacterium tuberculosis
469
invasion and intracellular replication. Infect Immun, 64, 2673-9.
470
NEYROLLES, O., HERNANDEZ-PANDO, R., PIETRI-ROUXEL, F., FORNES, P.,
471
TAILLEUX, L., BARRIOS PAYAN, J. A., PIVERT, E., BORDAT, Y., AGUILAR,
472
D., PREVOST, M. C., PETIT, C. & GICQUEL, B. (2006) Is adipose tissue a place
473
for Mycobacterium tuberculosis persistence? PLoS One, 1, e43.
474
475
NOBBS, A. H., LAMONT, R. J. & JENKINSON, H. F. (2009) Streptococcus adherence
and colonization. Microbiol Mol Biol Rev, 73, 407-50, Table of Contents.
476
ORDWAY, D., PALANISAMY, G., HENAO-TAMAYO, M., SMITH, E. E., SHANLEY, C.,
477
ORME, I. M. & BASARABA, R. J. (2007) The cellular immune response to
478
Mycobacterium tuberculosis infection in the guinea pig. J Immunol, 179, 2532-41.
21
479
PALANISAMY, G. S., DUTEAU, N., EISENACH, K. D., CAVE, D. M., THEUS, S. A.,
480
KREISWIRTH, B. N., BASARABA, R. J. & ORME, I. M. (2009) Clinical strains of
481
Mycobacterium tuberculosis display a wide range of virulence in guinea pigs.
482
Tuberculosis (Edinb), 89, 203-9.
483
PALANISAMY, G. S., SMITH, E. E., SHANLEY, C. A., ORDWAY, D. J., ORME, I. M. &
484
BASARABA, R. J. (2008) Disseminated disease severity as a measure of
485
virulence of Mycobacterium tuberculosis in the guinea pig model. Tuberculosis
486
(Edinb), 88, 295-306.
487
488
PALLEN, M. J. (2002) The ESAT-6/WXG100 superfamily -- and a new Gram-positive
secretion system? Trends Microbiol, 10, 209-12.
489
PETHE, K., ALONSO, S., BIET, F., DELOGU, G., BRENNAN, M. J., LOCHT, C. &
490
MENOZZI, F. D. (2001) The heparin-binding haemagglutinin of M. tuberculosis is
491
required for extrapulmonary dissemination. Nature, 412, 190-4.
492
PHILLIPS, J. R., TRIPP, T. J., REGELMANN, W. E., SCHLIEVERT, P. M. &
493
WANGENSTEEN, O. D. (2006) Staphylococcal alpha-toxin causes increased
494
tracheal epithelial permeability. Pediatr Pulmonol, 41, 1146-52.
495
POWERS, M. E., KIM, H. K., WANG, Y. & BUBECK WARDENBURG, J. (2012)
496
ADAM10 Mediates Vascular Injury Induced by Staphylococcus aureus alpha-
497
Hemolysin. J Infect Dis.
498
RICHTER, J. F., GITTER, A. H., GUNZEL, D., WEISS, S., MOHAMED, W.,
499
CHAKRABORTY, T., FROMM, M. & SCHULZKE, J. D. (2009) Listeriolysin O
500
affects barrier function and induces chloride secretion in HT-29/B6 colon
501
epithelial cells. Am J Physiol Gastrointest Liver Physiol, 296, G1350-9.
22
502
RYNDAK, M. B., SINGH, K. K., PENG, Z. & LAAL, S. (2015) Transcriptional Profile of
503
Mycobacterium tuberculosis Replicating in Type II Alveolar Epithelial Cells. PLoS
504
One, 10, e0123745.
505
RYNDAK, M. B., SINGH, K. K., PENG, Z., ZOLLA-PAZNER, S., LI, H., MENG, L. &
506
LAAL, S. (2014) Transcriptional profiling of Mycobacterium tuberculosis
507
replicating ex vivo in blood from HIV- and HIV+ subjects. PLoS One, 9, e94939.
508
SCHNAPPINGER, D., EHRT, S., VOSKUIL, M. I., LIU, Y., MANGAN, J. A., MONAHAN,
509
I. M., DOLGANOV, G., EFRON, B., BUTCHER, P. D., NATHAN, C. &
510
SCHOOLNIK, G. K. (2003) Transcriptional Adaptation of Mycobacterium
511
tuberculosis within Macrophages: Insights into the Phagosomal Environment. J
512
Exp Med, 198, 693-704.
513
SINHA, B., FRANCOIS, P. P., NUSSE, O., FOTI, M., HARTFORD, O. M., VAUDAUX,
514
P., FOSTER, T. J., LEW, D. P., HERRMANN, M. & KRAUSE, K. H. (1999)
515
Fibronectin-binding protein acts as Staphylococcus aureus invasin via fibronectin
516
bridging to integrin alpha5beta1. Cell Microbiol, 1, 101-17.
517
SOONG, G., MARTIN, F. J., CHUN, J., COHEN, T. S., AHN, D. S. & PRINCE, A. (2011)
518
Staphylococcus aureus protein A mediates invasion across airway epithelial cells
519
through activation of RhoA GTPase signaling and proteolytic activity. J Biol
520
Chem, 286, 35891-8.
521
SUAREZ, M., GONZALEZ-ZORN, B., VEGA, Y., CHICO-CALERO, I. & VAZQUEZ-
522
BOLAND, J. A. (2001) A role for ActA in epithelial cell invasion by Listeria
523
monocytogenes. Cell Microbiol, 3, 853-64.
23
524
VIDAL PESSOLANI, M. C., MARQUES, M. A., REDDY, V. M., LOCHT, C. & MENOZZI,
525
F. D. (2003) Systemic dissemination in tuberculosis and leprosy: do
526
mycobacterial adhesins play a role? Microbes Infect, 5, 677-84.
527
VIR, P., GUPTA, D., AGARWAL, R. & VERMA, I. (2014) Interaction of alveolar epithelial
528
cells with CFP21, a mycobacterial cutinase-like enzyme. Mol Cell Biochem, 396,
529
187-99.
530
WOLF, A. J., DESVIGNES, L., LINAS, B., BANAIEE, N., TAMURA, T., TAKATSU, K. &
531
ERNST, J. D. (2008) Initiation of the adaptive immune response to
532
Mycobacterium tuberculosis depends on antigen production in the local lymph
533
node, not the lungs. J Exp Med, 205, 105-15.
534
535
FIGURE LEGENDS
536
Fig. 1. M. tb H37Rv invades and replicates in HPAE-26 cells. (a) Intracellular M. tb
537
CFU recovered from HPAE-26 infected with M. tb at MOI (bacteria:cells) 1:1 and 5:1 on
538
Days 0 (2 hr), 3 and 5 post-infection. (b) Representative light microscopy images of
539
HPAE-26 (stained with Brilliant Green) infected with M. tb (Kinyoun stain -pink) MOI 1:1
540
under the same conditions as described in panel A.
541
Fig. 2. Creating the human in vitro alveolar barrier. (a) Schematic illustration of the
542
in vivo alveolar barrier. The alveolar lumen is lined with a monolayer of type I (light grey)
543
and type II (blue) alveolar epithelial cells (AEC) which are separated from the alveolar
544
capillary endothelial cells (EC) (dark grey) by interstitial extracellular matrix (ECM) or by
545
a fused basement membrane (BM). To cross this barrier directly, M. tb (fuschia) must
546
invade and replicate in the AEC, exit these cells, cross the ECM/BM/EC and exit to the
24
547
blood circulation. (b) The in vitro alveolar barrier was constructed by growing
548
monolayers of A549 (human type II AEC line) and HPAE-26 (human pulmonary artery
549
EC line) cells on opposite sides of a BM-infused transwell permeable membrane. The
550
upper chamber (A549 side) represents the alveolar lumen while the lower chamber
551
(HPAE-26 side) represents entrance into the blood circulation. (c) Daily transmembrane
552
electrical resistance (TER) (ohms/cm2) measurements monitored in parallel and in
553
triplicate; BM-infused transwells seeded with HPAE-26 alone (red diamonds), A549
554
alone (green squares), or both HPAE-26 and A549 (blue triangles) starting one day post
555
seeding. TER measurements for BM-infused transwells with no cells (purple circles)
556
were also taken in duplicate and in parallel to control for fluctuations in meter reading.
557
Measurements were taken out to 9 days post-HPAE-26 seeding and 7 days post-A549
558
seeding.
559
Fig. 3. Percent migrations of different mycobacterial strains across the human in
560
vitro alveolar barrier. (a) Percent migrations of laboratory M. tb H37Rv and clinical
561
isolates CDC1551 and HN878. (b) Percent migrations of M. tb H37Rv, M. tb H37Rv
562
esat-6, M. tb H37Rv phoP, M. bovis BCG, and M. bovis BCG hbha. All migrations
563
were normalized to M. tb H37Rv = 1. Error bars represent standard errors of triplicate
564
barrier infections. Asterisks (*) indicates statistically significant difference (P value ≤
565
0.05 by Mann-Whitney one tail test); “ns” indicates not statistically different.
566
567
25