70-3-359

Pathogens and Disease ISSN 2049-632X
RESEARCH ARTICLE
Expression of antimicrobial drug tolerance by attached communities

of Mycobacterium tuberculosis
David F. Ackart1, Laurel Hascall-Dove1, Silvia M. Caceres2, Natalie M. Kirk1, Brendan K. Podell1,
Christian Melander3, Ian M. Orme1, Jeff G. Leid4, Jerry A. Nick2 & Randall J. Basaraba1
1 Department of Microbiology, Immunology and Pathology, Mycobacterial Research Laboratories, Colorado State University, Fort Collins, CO, USA
2 Department of Medicine, National Jewish Health, Denver, CO, USA
3 Department of Chemistry, North Carolina State University, Raleigh, NC, USA
4 Medical Products Division, W.L. Gore and Associates, Flagstaff, AZ, USA
Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

The development of novel anti-TB drugs would benefit from improved in vitro assays. In this study, drug tolerance is linked
to the formation of complex mycobacterial communities that attach to the surface of the well of the extracellular matrix from
lysed leukocytes. With some modifications, the assay used in this study would allow high-throughput screening of
antimicrobials targeting drug-tolerant Mycobacterium tuberculosis.
Keywords Abstract
Mycobacterium tuberculosis; drug resistant;
drug tolerant; biofilm; microbial communities. There is an urgent need to improve methods used to screen antituberculosis
drugs. An in vitro assay was developed to test drug treatment strategies that
Correspondence specifically target drug-tolerant Mycobacterium tuberculosis. The H37Rv strain of
Randall J. Basaraba, Department of Microbi- M. tuberculosis survived antimicrobial treatment as attached microbial communi-
ology, Immunology and Pathology, Colorado ties when maintained in tissue culture media (RPMI-1640) with or without lysed
State University, 200 W. Lake, 1619 Campus human peripheral blood leukocytes. When cultured planktonically in the presence
Delivery, Fort Collins, CO 80523, USA. of Tween-80, bacilli failed to form microbial communities or reach logarithmic
Tel.: 970 491 3313 phase growth yet remained highly susceptible to antimicrobial drugs. In the
fax: 970 491 6030 absence of Tween, bacilli tolerated drug therapy by forming complex microbial
e-mail: [email protected] communities attached to untreated well surfaces or to the extracellular matrix
derived from lysed human leukocytes. Treatment of microbial communities
Received 11 December 2013; revised 15 with DNase I or Tween effectively dispersed bacilli and restored drug susceptibility.
January 2014; accepted 16 January 2014. These data demonstrate that in vitro expression of drug tolerance by M. tuber-
Final version published online 24 February culosis is linked to the establishment of attached microbial communities and that
2014. dispersion of bacilli targeting the extracellular matrix including DNA restores drug
susceptibility. Modifications of this in vitro assay may prove beneficial in a
doi:10.1111/2049-632X.12144 high-throughput platform to screen new antituberculosis drugs especially those
that target drug-tolerant bacilli.
Editor: Tom Coenye
conditions that exist during M. tuberculosis infection (Deb

Introduction
et al., 2009; Patel et al., 2011). The persistence of extra-
Current first-line tuberculosis treatment consists of a mini- cellular bacilli sequestered within necrotic lung granulomata
mum of 6–9 months of combination antimicrobial drug represents at least one population of bacilli that may
therapy, which is presumed necessary to completely contribute significantly to the overall expression of pheno-
eradicate persistent populations of drug-tolerant bacilli typic drug resistance or drug tolerance in vivo, but the
(Mitchison & Davies, 2012). In vivo drug tolerance is not mechanisms are poorly understood (Ulrichs & Kaufmann,
only expressed during active disease in humans but can be 2006; Basaraba, 2008).
modeled in a variety of laboratory animal species and M. tuberculosis is considered an obligate aerobe, and the
in vitro. The in vitro susceptibility of Mycobacterium tuber- inherent slow rate of growth is further reduced when bacilli
culosis (M. tuberculosis) to antimicrobial drugs is influenced are maintained under low-oxygen conditions. Using an
by specific nutrient depletion, decreasing environmental in vitro model, Wayne & Hayes (1996) and Wayne &
oxygen concentrations or a combination of nutrient deple- Sohaskey (2001) demonstrated that the gradual consump-
tion, low oxygen and the introduction of other stress tion of oxygen by M. tuberculosis forces bacilli into multiple
Pathogens and Disease (2014), 70, 359–369, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved 359
Drug tolerance of attached Mycobacterium tuberculosis D.F. Ackart et al.
transition phases of non- or slowly replicating persistence, with 0.05% Tween-80 (Sigma, St. Louis, MO) at 37 °C with
which coincides with decreased antimicrobial drug suscep- shaking. Cultures were aliquoted and frozen at 80 °C from
tibility. It is the slow rate of bacterial replication that is cultures grown to an optical density 600 (OD600 nm) of 0.6–
thought to limit the effectiveness of some antimicrobial 0.9. For in vitro assays, bacilli were thawed and resus-
drugs especially those that target cell wall synthetic path- pended at a concentration of 1.5 9 107 colony-forming units
ways (Erdemli et al., 2012). The original Wayne model with (CFU) mL 1 in RPMI 1640 with L-glutamine, phenol red (Life
and without modifications is still commonly used to better Technologies, Carlsbad, CA) and 2% heat-inactivated
understand the factors that influence in vitro bacterial growth bovine platelet poor plasma (Bioreclamation, Liverpool,
and M. tuberculosis drug susceptibility (Leistikow et al., NY) (complete RPMI-1640).
2010). However, the design and complexity of these assays
are not well suited for high-throughput screening of antitu-
Human leukocyte isolation
berculosis drugs or novel treatment strategies. Moreover,
these assays fail to take into account host factors that likely Peripheral blood leukocytes enriched for neutrophils were

contribute to the expression of in vivo drug tolerance in isolated from freshly collected whole blood from healthy
humans and animals. volunteers by the plasma Percoll method as previously
The Wayne model is designed to specifically mimic the described (Walker et al., 2005). These studies were con-
low-oxygen concentration measured in lung lesions of ducted in accordance with the Declaration of Helsinki,
M. tuberculosis humans and some animal models (Tsai approved by the Colorado State University Biosafety Com-
et al., 2006; Via et al., 2008; Heng et al., 2011; Harper mittee and the National Jewish Health Institutional Review
et al., 2012). Prolonged hypoxia of inflammatory cells that Board. Written informed consent was obtained from all
respond to lung infection likely contributes to individual cell clinical healthy leukocyte donors. Isolated human peripheral
death and lesion necrosis, which releases bacilli from blood leukocytes enriched for neutrophils were washed and
infected cells into the extracellular matrix (Basaraba, resuspended in complete RPMI-1640 and seeded in each
2008). A feature of necrotic lesions therefore is the coex- well of a 96-well plate at a concentration of 7.5 9 106
istence of both intracellular and extracellular populations of cells mL 1. Cells were lysed by freezing and stored at
bacilli (Lenaerts et al., 2007; Hoff et al., 2011). In vivo, 80 °C. Lysis and the loss of viability were confirmed
extracellular bacilli survive entrapped in an extracellular microscopically using the trypan blue dye exclusion tech-
matrix composed of macromolecules derived primarily from nique.
necrotic and lysed leukocytes, which we hypothesize con-
tributes to the expression of antimicrobial drug tolerance
M. tuberculosis cultures and the establishment of
(Canetti, 1955; Lenaerts et al., 2007; Basaraba, 2008). The
drug tolerance
other consequence of necrotic lesions is blood flow is limited
which not only reduces oxygen availability, but further limits M. tuberculosis H37Rv or HN878 (7.5 9 106 bacilli mL 1)
the delivery of therapeutic drug concentrations directly to the was cultured in 96-well flat-bottom plates (Becton Dickinson,
site of infection (Barclay et al., 1953; Manthei et al., 1954; Franklin Lakes, NJ) or Lab-Tek II Chamber Slides (Nunc,
Prideaux et al., 2011). Therefore, the in vivo expression of Rochester, NY) in complete RPMI-1640 with and without
phenotypic resistance by M. tuberculosis is multifactorial in Tween-80 at a final concentration of 0.05%. M. tuberculosis
which lesion necrosis is central to the pathogenesis (Lena- was also cultured in complete RPMI-1640 in the presence
erts et al., 2007; Ahmad et al., 2009). of lysed human leukocytes. Plates were incubated for
To more closely mimic the host contribution to the 7 days at 37 °C in a humidified incubator without supple-
expression of in vivo drug tolerance, we developed a simple mental carbon dioxide. To determine the growth rates from 7
assay that enriches for extracellular M. tuberculosis that to 14 days in the presence or absence of drug treatment
expresses in vitro drug tolerance in the presence of or drug carriers, attached communities of bacilli were
macromolecules derived from lysed human cells. This scraped from well surfaces, dispersed mechanically into a
model system takes into account the possible contribution single-cell suspension, and serial dilutions plated on 7H11
host-derived macromolecules play in the establishment of in agar. Data are expressed as CFU mL 1 or log10 percent
vitro and in vivo drug tolerance. Through the use of this survival.
assay, we determined that the ability of M. tuberculosis to
attach to abiotic or biotic surfaces is an important determi-
Antimicrobial drug treatment
nant of the expression of drug tolerance, which we show
here, can be reversed by dispersion of attached microbial Each of the three separate culture conditions, complete
communities. RPMI-1640 with Tween-80, complete RPMI-1640 without
Tween-80 and complete RPMI-1640 with lysed leukocytes,
was established for 7 days and then treated for an
Materials and methods
additional 7 days with isoniazid (10 lg mL 1, Sigma) or
rifampin (25 lg mL 1; Sigma) alone or in combination with
Bacterial strains
pyrazinamide (3 mg mL 1; Sigma), dissolved in water
The H37Rv and HN878 strains of M. tuberculosis were prop- (isoniazid and pyrazinamide) or dimethyl sulfoxide (DMSO)
agated in Proskauer and Beck liquid broth supplemented (rifampin). In separate experiments, molecular grade,
360 Pathogens and Disease (2014), 70, 359–369, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved
D.F. Ackart et al. Drug tolerance of attached Mycobacterium tuberculosis
protease-free DNase I (0.5 mg mL 1, Sigma) was added and analyzed with Nikon NIS Elements AR imaging software
on days 7 and 10 of culture to established drug-tolerant version 3.20.00 or VOLOCITY 6.3.0.
cultures with or without isoniazid and incubated to day 14.
Also in a separate experiment, Tween-80 (0.05% final
Statistical analysis
concentration) was added to established drug-tolerant
cultures of M. tuberculosis with and without isoniazid and SAS 9.3 software was used for statistical analysis with an
incubated for 7 days. At 24-h intervals, bacilli were scraped, alpha equal to 0.05. CFUs were normalized to day 7 starting
mechanically dispersed and plated on 7H11 agar and CFUs values and log-transformed to correct for unequal variance.
counted. The CFU data were expressed as percent survival A three-way factor analysis of variance was used to
of the original inoculum or on day 7 of culture when drug determine differences between growth conditions. A
treatment was initiated. The minimum inhibitory concentra- Tukey-adjusted least squares means method was employed
tions (MIC) of isoniazid and rifampin were determined in the to determine statistical differences between growth condi-
different culture conditions using the alamar blue (Life tions for each treatment–time point combination. Differences

Technologies) as previously described (Franzblau et al., within growth conditions were determined using a two-way
1998). Starting on day 7, cultures were treated for 5 days ANOVA and Tukey adjusted for multiple comparisons between
with twofold serial dilutions of isoniazid or rifampin. On day treatments for each time point.
5, 50 lL of a solution of a 5X alamar blue and 5% Tween–
80 solution was added to wells and incubated at 37 °C.
Results
After 48 h, the MIC was defined as the lowest drug
concentration that does not result in alamar blue reduction
Mycobacterium tuberculosis cultured in nutrient-rich
as indicated by color change.
mammalian cell culture media fails to reach log phase
growth in vitro
Laser scanning confocal microscopy
The relative rate of growth and survival of the H37Rv strain
In separate cultures prepared for fluorescence microscopy, of M. tuberculosis cultured in complete RPMI-1640 with and
media were removed from M. tuberculosis cultures grown without Tween-80 and lysed human leukocytes is shown in
on Lab-Tek II chamber slides and allowed to dry prior to Fig. 1. Despite being maintained in nutrient-rich media,
fixation with 4% paraformaldehyde for 30 min. Slides were bacilli not only failed to reach a logarithmic stage of growth,
blocked with a serum-free blocking solution (Dako, Carpin- but there was an approximate one log10 reduction in viable
teria, CA) and then stained with Rhodamine B (Sigma) bacilli by day 7 as determined by counting CFUs. Data were
solution. After decolorization with acid-alcohol, slides were expressed as CFUs per mL (Fig. 1a), which was used to
stained with biotinylated wheat germ agglutinin (Vector calculate percent survival of the original inoculum (Fig. 1b)
Laboratories, Burlingame, CA) and visualized with an and responses to drug treatment. There were no significant
avidin-conjugated AMCA (Vector Laboratories). Slides were differences in the viability of bacilli between the three
counterstained for nucleic acids with TO-PRO-3 (Life Tech- different culture conditions across all time points even out to
nologies) and mounted with Prolong Gold (Life Technolo- 14 days in culture. Even in media further enriched by a
gies). To assess the spatial distribution of viable and complex mixture of macromolecules derived from lysed
nonviable bacilli within attached communities, the LIVE/ human leukocytes, there was no net growth increase.
DEAD BacLight Bacterial Viability assay (Life Technologies) However, in the presence of lysed leukocytes, bacilli
was utilized as per the manufacturer’s instructions prior to numbers returned to that of the original inoculum by day 7
paraformaldehyde fixation. Confocal stacks were captured compared with bacilli grown in the presence or absence of
with a Zeiss Laser Scanning Axiovert Confocal Microscope Tween-80, but differences between treatment groups were
(a) (b)
Fig. 1 Mycobacterium tuberculosis cultured in nutrient-rich mammalian cell culture media fails to proliferate in vitro. The H37Rv strain of
M. tuberculosis was cultured in mammalian cell culture media (RPMI-1640 + 2% bovine plasma) with Tween-80, without Tween-80 and with lysed
human leukocytes; all 3 culture conditions show no net increase in viable bacilli numbers out to 7 days. Data are expressed as (a) mean log10
CFUs SEM or (b) mean percent survival SEM of CFUs cultured on 7H11 agar (n = 3).
not statistically significant. In contrast, the HN878 strain of 3b) was significantly higher between days 10 through 14
M. tuberculosis showed evidence of slow growth by day 14 (P ≤ 0.0001) compared with bacilli cultured in
under all three different culture conditions, yet also failed to Tween-80-supplemented media. Tolerance to monotherapy
reach log phase growth (Supporting Information, Fig. S1). was further enhanced when M. tuberculosis was cultured in
the presence of lysed human leukocytes without Tween-80
(Figs 2c and 3c), which showed less than a 0.5 log10
Mycobacterium tuberculosis expresses differences in
reduction in viable bacilli out to day 14. The rapid ca. 2.5
tolerance to isoniazid and rifampin monotherapy in
log10 reduction in viable bacilli treated with rifampin contin-
vitro
ued out to day 13, but like isoniazid treatment, there were c.
The differences in susceptibility of M. tuberculosis exposed 0.3% of bacilli that survived drug treatment and persisted
to high doses of isoniazid (10 lg mL 1) and rifampin throughout the remainder of the culture period. The survival
(50 lg mL 1) alone under different in vitro growth conditions of M. tuberculosis in the presence of lysed leukocytes and
are shown in Figs 2 and 3, respectively. Despite the lack of treated with drugs alone was significantly higher between

log phase growth of bacilli out to 14 days carrier controls days 10 and 14 compared with bacilli cultured in media
treated with water for isoniazid and DMSO for rifampin, supplemented with Tween-80 (P ≤ 0.0001). In a separate
M. tuberculosis cultured in complete RPMI-1640 supple- set of experiments, Tween-80 and isoniazid were added to
mented with Tween-80 (Figs 2a and 3a) were significantly drug-tolerant cultures of M. tuberculosis H37Rv that had
more susceptible to individual drugs alone. The rapid two been established for 7 days. The dispersion of established
log10 reduction in CFUs, which occurred in the first 4 days of communities of bacilli with Tween-80 restored drug suscep-
treatment, reached a plateau phase of survival by day 11 tibility, which was significantly different from controls by
leaving c. 1% of the original number of bacilli viable, which days 13 and 14 both in the absence (Fig. S1a) and presence
persisted throughout the remainder of the culture period. In of lysed human leukocytes (Fig. S1b). Similar to H37Rv, the
contrast, the survival of M. tuberculosis cultured in the HN878 strain of M. tuberculosis also expressed tolerance to
absence of Tween-80 and lysed leukocytes (Figs 2b and isoniazid in the absence of Tween-80 and in the presence of
(a) (b) (c)
Fig. 2 Mycobacterium tuberculosis cultured in nutrient-rich mammalian cell culture media express differences in tolerance to isoniazid in vitro. The
H37Rv strain of M. tuberculosis was cultured in mammalian cell culture media (RPMI-1640 + 2% bovine plasma) (a) in Tween-80 (0.05%), (b) in the
absence of Tween-80 (0.05%) and (c) in the presence of lysed human leukocytes which were treated for 7 days with either isoniazid (10 lg mL 1) or
equal volumes of the carrier, water. Data are expressed as the mean percent survival of viable CFUs SEM on 7H11 agar. n = 6 separate
experiments. **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 compared to carrier controls.
(a) (b) (c)
Fig. 3 Mycobacterium tuberculosis cultured in nutrient-rich mammalian cell culture media express differences in tolerance to rifampin in vitro. The
H37Rv strain of M. tuberculosis was cultured for 7 days in mammalian cell culture media (RPMI-1640 + 2% bovine plasma) containing (a) Tween-80
(0.05%), (b) in the absence of Tween-80 (0.05%) and (c) in the presence of lysed human leukocytes. Cultures were treated with rifampin (50 lg mL 1)
or equal volumes of the carrier DMSO (< 0.01%). Data are expressed as the mean percent survival of viable CFUs SEM on 7H11 agar. n = 3
separate experiments. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 compared to carrier controls.
lysed human leukocytes. Similar to the laboratory-adapted shown in Table 1. In the absence of Tween-80, the MIC for
strain, HN878 was also susceptible to isoniazid in media isoniazid increased up to sevenfold compared to media
supplemented with Tween-80, but not in the absence of containing Tween-80. In the absence of Tween-80 and
Tween-80 or the presence of leukocyte lysate (Fig. S2). presence of lysed leukocytes, isoniazid MIC increased up to
62-fold compared with M. tuberculosis cultured in the
presence of Tween-80. For rifampin, the increase in MIC
Mycobacterium tuberculosis expresses differences in
was less marked and increased up to fourfold in media in the
tolerance to combination isoniazid, rifampin and
absence of Tween-80 and an 8–15-fold increase in media
pyrazinamide (RHZ) in vitro
without Tween-80 but in the presence of lysed leukocytes.
The differences in susceptibility of M. tuberculosis to com-
bination isoniazid (10 lg mL 1), rifampin (50 lg mL 1) and
Mycobacterium tuberculosis forms attached microbial
pyrazinamide (50 mg mL 1) drug therapy under different in
communities in the absence of Tween-80 and the
vitro growth conditions are shown in Fig. 4. Similar to what

presence of lysed leukocytes
was seen in response to isoniazid and rifampin monother-
apy, treatment of M. tuberculosis cultured in complete The relative difference in colony morphology of attached
RPMI-1640 supplemented with Tween-80 (Fig. 4a) was communities of M. tuberculosis observed by fluorescence
highly susceptible to RHZ combination therapy compared confocal microscopy is shown Fig. 5. In the presence of
with the DMSO carrier control. The rapid ca. 2.5 log10 Tween-80 (Fig. 5a), bacilli were attached to tissue culture
reduction in viability continued out to day 11, but like well surfaces as individual or small clusters of bacilli and
isoniazid and rifampin monotherapy, c. 0.1% of bacilli failed to form complex microbial communities (maximal Z
survived drug treatment and persisted throughout the stack height 6.4 lm). The majority of bacilli cultured in the
culture period. Compared to bacilli cultured in media presence of Tween-80 stained poorly by the fluorescent
supplemented with Tween-80, the survival of M. tuberculo- acid-fast technique (Rhodamine B) and stain only with the
sis cultured in the absence of Tween-80 (Fig. 4b) and the DNA stain TO-PRO3. Individual or small clusters of bacilli
presence of lysed leukocytes (Fig. 4c) was significantly were either acid-fast positive and stain only with Rhodamine
higher on days 8 and 9 (P ≤ 0.0001), 10 (P ≤ 0.001), 11 and or were positive for both stains (yellow). In contrast, the vast
12 (P ≤ 0.0001), 13 (P ≤ 0.001) and 14 (P ≤ 0.0001). Also majority of bacilli cultured in the absence of Tween-80
like isoniazid and rifampin monotherapy, M. tuberculosis (Fig. 5b) were acid-fast positive with Rhodamine and form
cultured in complete RPMI-1640 in the absence of raised complex microbial communities (maximal Z stack
Tween-80 was significantly higher with only one log10 net height 20.6 lm). In addition, there were dual-stained bacilli
reduction in CFUs out to day 14. The expression of that are stratified near the attachment surface. Acid-fast
tolerance of M. tuberculosis to RHZ when cultured in the bacilli cultured in the absence of Tween-80 and the
presence of lysed leukocytes was similar to that of bacilli presence of lysed leukocytes (Fig. 5c) formed complex
cultured in the absence of Tween-80. communities of M. tuberculosis (maximum Z stack height
29.8 lm) attached to an extracellular matrix composed of
leukocyte-derived macromolecules including extracelluar
The MIC of isoniazid and rifampin against
DNA (eDNA). In addition to host-derived DNA and acid-fast
M. tuberculosis differs under different growth
bacilli, microbial communities of M. tuberculosis in the
conditions
presence of lysed leukocytes contained non-acid-fast bacilli
The differences in MIC of isoniazid and rifampin against that stain with TO-PRO-3 only as well as complex carbo-
M. tuberculosis under different in vitro growth conditions are hydrates (Fig. S3).
(a) (b) (c)
Fig. 4 Mycobacterium tuberculosis cultured in nutrient-rich mammalian cell culture media express differences in tolerance to combination
antimicrobial drug treatment in vitro. The H37Rv strain of M. tuberculosis was cultured for 7 days in mammalian cell culture media (RPMI-1640 + 2%
bovine plasma) containing (a) Tween-80 (0.05%), (b) in the absence of Tween-80 (0.05%) and (c) in the presence of lysed human leukocytes. Cultures
were treated for 7 days with of isoniazid (10 lg mL 1), rifampin (50 lg mL 1) and pyrazinamide (3 mg mL 1) combined (RHZ) or equal volumes of
the carrier DMSO (< 0.01%). Data are expressed as the mean percent survival of viable CFUs SEM on 7H11 agar. n = 2 separate experiments.
*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 compared to carrier controls.
Table 1 The MIC of isoniazid and rifampin in increased when Myco-

bacterium tuberculosis is cultured as attached microbial communities
Treatment of drug-tolerant M. tuberculosis with DNase I
partially restores in vitro susceptibility to isoniazid
MIC The influence that dispersion of attached communities of
Isoniazid Rifampin M. tuberculosis with DNase I had on the in vitro suscep-
(lg mL 1) (lg mL 1) tibility to isoniazid grown under different culture conditions
RPMI w/Tween-80 < 0.04 0.06–0.12 is shown in Fig. 7. The split treatment of M. tuberculosis
RPMI w/o Tween-80 0.04–0.31 0.12–0.23 cultures with DNase I on days 7 and 10 in combination
Leukocyte Lysate 1.25–2.5 0.47–0.94 with isoniazid increased the susceptibility of drug-tolerant
bacilli when cultured in media supplemented with
In the absence of Tween-80, the MIC for isoniazid increased one- to
Tween-80 (Fig. 7a) resulting in an additional one log10
sevenfold compared to media containing Tween-80. In the absence of
reduction of viability by day 14. The increased suscepti-
Tween-80 and presence of lysed leukocytes, the MIC of isoniazid
bility to isoniazid in the presence of DNase I was

increased ranged from 31 to 62-fold compared with M. tuberculosis
cultured in the presence of Tween-80. For rifampin, the increase in MIC
statistically significant on days 13 and 14, but no effect
was less marked and ranged from a two- to fourfold increase in media in of DNase I treatment alone was seen in the presence of
the absence of Tween-80 and an 8–15-fold increase in media without the water carrier control (Fig. 7a). The killing of M. tuber-
Tween-80 but in the presence of lysed leukocytes. Data are expressed culosis by isoniazid was also potentiated by DNase I when
as the mean range of MIC from three separate experiments. cultured in media without Tween-80 (Fig. 7b). Treatment
with DNase I further enhanced the bactericidal capacity of
isoniazid with an additional 0.5 log10 reduction by day 14.
Viable and nonviable bacilli are spatially stratified when Compared to isoniazid alone treated cultures, the percent
cultured as attached communities of M. tuberculosis survival was significantly different from day 10 through 14
when DNase I was combined with isoniazid. In contrast to
The spatial distribution of viable and nonviable bacilli M. tuberculosis cultured in the presence of Tween-80
cultured as attached microbial communities for 7 days in (Fig. 7a), DNase I treatment alone also significantly
media supplemented with Tween-80, and in the absence of reduced the survival from days 12 to 14 compared with
Tween-80 is shown in Fig. 6. Out to 7 days in culture, water-treated controls. When cultured in the presence of
M. tuberculosis failed to form complex microbial communi- lysed leukocytes, (Fig. 7c) treatment with isoniazid com-
ties when cultured in media supplemented with Tween-80 bined with DNase I had a transient effect, reducing
(Fig. 6a), and viable and nonviable bacilli were admixed and survival from days 8 to 12, but there were no significant
evenly dispersed as individual bacilli or small clusters. In statistical differences on days 13 and 14 compared with
contrast, in the absence of Tween-80 (Fig. 6b), viable bacilli cultures treated with isoniazid alone. Similar to M. tuber-
aggregated to form raised, complex microbial communities, culosis cultured in the presence of Tween-80, treatment of
which were attached and spatially stratified from nonviable M. tuberculosis attached to lysed leukocytes with DNase I
bacilli at the attachment surface of tissue culture wells. The had no effect on viability compared with the water control
orthogonal view demonstrates that the DNA of nonviable cultures. DNase I digestion of extracellular DNA was
bacilli stained with propidium iodide was concentrated at the effective as evidenced by the lack of extracellular DNA
well surface and served as an attachment matrix for viable staining with the DNA stain TO-PRO3 in the different
bacilli in the upper layers, which comprise the majority of the culture conditions as determined by fluorescence confocal
microbial community. microscopy (Fig. S4).
(a) (b) (c)
Fig. 5 The expression of in vitro antimicrobial drug tolerance by Mycobacterium tuberculosis is associated with attached communities of bacilli with
different staining characteristics. The H37Rv strain of M. tuberculosis was cultured for 14 days in mammalian cell culture media (RPMI-1640 + 2%
bovine plasma) containing (a) Tween-80 (0.05%), (b) in the absence of Tween-80 (0.05%) and (c) in the presence of lysed human leukocytes,
paraformaldehyde fixed and stained with rhodamine (red) and the DNA stain TO-PRO-3 and viewed by confocal microscopy. (a) In the presence of
Tween-80, bacilli stain poorly with rhodamine and are attached as individual or small clusters of bacilli (Z stack height, 6.4 lm). (b) In the absence of
Tween-80 and lysed leukocytes, the majority of bacilli stain with rhodamine and are firmly attached to well surfaces to form complex colonies of bacilli
(Z stack height, 20.6 lm). (c) In the presence of lysed human leukocytes, rhodamine-positive bacilli form complex raised colonies (Z stack height,
29.8 lm) attached to host-derived extracellular DNA. Grid = 45 lm square.
(a) (b)

Fig. 6 Viable and nonviable bacilli are spatially stratified when cultured as attached communities of M. tuberculosis. The H37Rv strain of
M. tuberculosis was cultured in mammalian cell culture media (RPMI-1640 + 2% bovine plasma) containing (a) Tween-80 (0.05%), (b) in the absence
of Tween-80 (0.05%). The H37Rv strain of M. tuberculosis was cultured for 7 days in mammalian cell culture media (RPMI-1640 + 2% bovine plasma)
containing (a) Tween-80 (0.05%), (b) in the absence of Tween-80 (0.05%) and stained with the live-dead (Syto 9 and propidium iodide, Bac-Light, Life
Technologies) prior to fixation with paraformaldehyde and viewed by confocal microscopy.
(a) (b) (c)
Fig. 7 Treatment of drug-tolerant Mycobacterium tuberculosis with DNase I partially restores in vitro susceptibility to isoniazid. The H37Rv strain of
M. tuberculosis was cultured in mammalian cell culture media (RPMI-1640 + 2% bovine plasma) containing (a) Tween-80 (0.05%), (b) in the absence
of Tween-80 (0.05%) and (c) in the presence of lysed human leukocytes. Cultures were treated for 7 days with isoniazid (10 lg mL 1) or equal
volumes of the carrier water with or without DNase I (0.5 mg mL 1) added on day 7 and 10. Data are expressed as the mean percent survival of viable
CFUs SEM on 7H11 agar. n = 6 separate experiments. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 compared to carrier controls.
the formation of attached communities and the expression

Discussion
of drug tolerance; and (4) dispersion of attached
The goal of these studies was to develop a simple assay communities of M. tuberculosis enzymatically with DNase
that better mimics the expression of phenotypic drug or nonspecifically with Tween-80 restores in vitro drug
resistance or tolerance by M. tuberculosis in humans and susceptibility.
animal models. The assay was adapted from that of studies Among the most significant findings were that maintaining
by Walker and colleagues who demonstrated that the planktonic bacilli in media supplemented with Tween-80
inclusion of human neutrophils promotes in vitro biofilm prevented the attachment and establishment of drug-
formation and the expression of antimicrobial drug tolerance tolerant communities of M. tuberculosis. The use of myco-
by Pseudomonas aeruginosa (Walker et al., 2005; Parks bacterial media containing nonionic detergents is widely
et al., 2009). The key findings in the current study are that practiced and has the desired effect of preventing the
(1) even in nutrient-rich media, M. tuberculosis failed to inherent tendency of mycobacteria to clump or attach to
reach log phase growth yet remained susceptible to mono- plastic or glass surfaces in vitro (Franzblau et al., 2012). In
or combination therapy with first antituberculosis drugs; (2) the early development of artificial mycobacterial media , it
the formation of attached microbial communities and the was determined that maintaining bacilli as a single-cell
expression of in vitro drug tolerance were prevented when suspension (planktonic) by either agitation or the inclusion of
M. tuberculosis was maintained planktonically in media detergents was necessary to achieve the shortest time to
containing Tween-80; (3) inclusion of a complex mixture of logarithmic phase growth (Dubos, 1946; Dubos & Davis,
macromolecules derived from lysed leukocytes promotes 1946). Prior to the routine use of commercially available
detergents, glycerinated bovine bile functioned as a surfac- period when combined with pyrazinamide. The ability of
tant to prevent clumping and to promote in vitro growth M. tuberculosis to establish attached microbial communities
especially in the development of the Bacillus Calmette– even in Tween containing media may be related to the
Gue rin (BCG) vaccine. However, numerous studies have production and secretion of cutinase-like phospholipases
demonstrated that the presence of Tween-80 alters in vitro that deplete Tween in late-stage cultures (Parker et al.,
drug susceptibility, but the mechanisms are unknown 2007). This fact would also explain the ability of M. tuber-
(Dubos & Davis, 1946; Sattler & Youmans, 1948). The culosis to eventually express drug tolerance in the current
inclusion of Tween-80 has been shown to serve as an studies and to ultimately form a biofilm-like pellicle in media
alternative carbon source to promote growth and to alter containing Tween-80 (Ojha et al., 2008). This possibility is
M. tuberculosis cell wall structure by stripping the outer cell further supported by our confocal data showing that
wall envelope potentially influencing drug penetration (Sani attached, but immature communities of M. tuberculosis are
et al., 2010; Stoops et al., 2010). Conflicting studies have present by day 14 in culture even in media supplemented
demonstrated that the inclusion of Tween limits the accu- with Tween-80.

mulation of some drugs in M. tuberculosis (Piddock et al., Fluorescent confocal microscopic images of M. tubercu-
2000; Sarathy et al., 2013). The in vivo relevance of the losis maintained as abiotic cultures showed that after
effect of surfactants on M. tuberculosis drug tolerance is 7 days, viable and nonviable bacilli were spatially stratified
supported by the observations that poloxamer surfactants within attached microbial communities. Nonviable bacilli or
potentiate the in vivo antimicrobial activity against M. tuber- M. tuberculosis extracellular DNA (eDNA) stained with
culosis in vivo (Stoops et al., 2010). propidium iodide served as an attachment matrix for viable
Results of the current studies suggest that Tween-80 bacilli. The presence of eDNA derived from nonviable bacilli
prevents bacilli attachment and the establishment of mature was further supported by the susceptibility to dispersion with
microbial communities of drug-tolerant bacilli. Attachment to DNase I, which partially restored drug susceptibility. The
abiotic or biotic surfaces, establishment of complex micro- detection of M. tuberculosis DNA or DNA fragments from
bial communities, production of an extracellular polymeric human tuberculosis patients especially in culture negative
substance (EPS) and the expression of drug tolerance are samples is presumed to originate from nonviable bacilli;
all features of bacterial biofilms, which remains controversial however, the possibility of secreted DNA cannot be dis-
in regard to M. tuberculosis (Ojha et al., 2008; Pang et al., counted (van Staden et al., 1998; Fenhalls et al., 2002;
2012; Sambandan et al., 2013). Attachment is an early and Green et al., 2009). Pathogen-derived eDNA secreted by
necessary step in the establishment of bacterial biofilms, some pathogenic and nonpathogenic bacteria is known to
which has been suggested recently to be due to expression function as an attachment matrix in the early stages of
of pili by pellicle grown M. tuberculosis (Ramsugit et al., in vitro biofilm formation (Whitchurch et al., 2002). Host-
2013). The growth of M. tuberculosis as a pellicle results in derived eDNA especially is abundant in vivo associated with
the expression of drug tolerance, which has been attributed necrotic tuberculosis granulomata (Lenaerts et al., 2007;
to an EPS composed of free mycolic acids and phenolic Basaraba, 2008; Ordway et al., 2010; Hoff et al., 2011).
glycolipids (Ojha et al., 2008; Pang et al., 2012; Samban- The combined treatment of drug-tolerant M. tuberculosis
dan et al., 2013). with DNase I partially restored susceptibility to isoniazid
Similar to our data, previous studies have demonstrated under all culture conditions. Of interest and unexpected was
that M. tuberculosis fails to replicate in nutrient-rich that DNase I alone was as effective as isoniazid alone in
RPMI-1640 even in the presence of macrophage lysate M. tuberculosis grown under abiotic conditions in the
(Armitige et al., 2000). In contrast, another study proposed absence of Tween-80. This response is not likely due to a
the use the mammalian cell culture media (F12) to enhance direct toxic effect of DNase, because DNase I alone had no
M. tuberculosis growth as an improved method over tradi- effect on the viability of bacilli maintained under planktonic
tional mycobacterial growth media to rapidly screen antimi- conditions. Because the attachment and formation of
crobial drugs in vitro (Nozawa & Yokota, 1983). The microbial communities provides a significant survival advan-
conflicting data between these studies are likely due to tage, we interpreted the loss of M. tuberculosis in the
differences in the basic media formulations or related to the presence of DNase and absence of drugs as a loss of in
media supplements which included 5% heat-inactivated vitro fitness of bacilli detached by enzymatic activity
fetal bovine serum and 0.2% Tween-80 in the former (Ramasubbu et al., 2005; Tetz et al., 2008; Kaplan, 2009).
compared with 2% bovine plasma and 0.05% Tween-80 in In the presence of lysed leukocytes, the increased
the current study. Others have also shown that RPMI-1640 susceptibility to isoniazid conferred by DNase I was
supplemented with 10% fetal bovine serum also failed to transient suggesting that host- or pathogen-derived macro-
support M. tuberculosis growth in vitro (Zhang et al., 1998). molecules other than eDNA support drug-tolerant commu-
Our data show that similar to traditional mycobacterial nities of M. tuberculosis. Despite the short half-life, multiple
growth media, small populations of drug-tolerant bacilli DNase treatments were effective at digesting host- and
survive in vitro antimicrobial drug treatment even despite the pathogen-derived eDNA as evidenced by the lack of DNA
presence of Tween-80 (Gruppo et al., 2006). By 11 days in staining viewed by confocal microscopy. The potential
culture, 1% of the original population express in vitro involvement of host- or bacilli-derived macromolecules other
tolerance to isoniazid, 0.3% for rifampin monotherapy and than eDNA in the establishment of drug-tolerant communi-
0.1% persisted throughout the remainder of the culture ties of M. tuberculosis is supported by the pellicle growth
model system which demonstrates the presence of bacilli-- cultures under controlled low-oxygen conditions or to intro-
derived free mycolic acids (Ojha et al., 2008; Sambandan duce other stress factors to induce in vitro drug tolerance.
et al., 2013) and by our data showing the presence of Recent studies have demonstrated that M. tuberculosis
complex carbohydrates associated with drug-tolerant micro- cultured in the presence of viable human peripheral blood
bial communities of M. tuberculosis. mononuclear cells also express in vivo drug tolerance
It is unclear whether M. tuberculosis like other pathogenic (Kapoor et al., 2013). The poorly defined extracellular matrix
bacteria actively secretes eDNA (Yang et al., 2007; Mann derived from viable or lysed leukocytes is a limitation to the
et al., 2009); however, these studies suggest that DNA use of current assay design as a high-throughput
derived from nonviable bacilli or from the host may function drug-screening platform. Our data suggest that eDNA is an
to promote the establishment of drug-tolerant microbial important extracellular matrix in vitro and in vivo, and we are
communities in vitro and in vivo (Wilson & Schwabacher, currently exploring whether purified DNA alone is sufficient to
1935). Our previous studies have shown that M. tuberculo- promote the in vitro expression of M. tuberculosis drug
sis-derived eDNA is a major component of the extracellular tolerance. Adapting this assay using a well-defined attach-

matrix that supports the expression of drug-tolerant ment matrix in a high-throughput platform has the potential to
bacilli grown under low-oxygen conditions in vitro (Ryan fill a critical unmet need to identify urgently needed new
et al., 2010). Recently, is has been demonstrated that drugs or novel therapeutic strategies specifically targeting
M. tuberculosis grown as a suspension secretes stable drug-tolerant M. tuberculosis. Modifications of this assay
double-stranded RNA fragments into the culture media system have been recently used to identify novel antituber-
(Obregon-Henao et al., 2012). The in vivo relevance of culosis drugs that specifically target copper metabolism
bacteria interacting with nucleic acids specifically eDNA is a (Speer et al., 2013). Currently, we are using this assay
major contributor to the viscosity of inflammatory exudate in system to discover small molecules that disperse bacilli from
patients with cystic fibrosis (CF) and other chronic lung lysed leukocytes as a strategy to restore drug susceptibility
diseases (Fahy et al., 1993; Puchelle et al., 1996). Recom- of drug-tolerant mycobacterium in vivo.
binant human DNase I (Pulmzymeâ, Dornase alfa, Genen-
tech USA, Inc.) is commonly used therapeutically as an
adjunct to antimicrobial drug therapy to reduce airway Acknowledegments
secretion viscosity associated with eDNA in CF patients This work was supported by NIH NIAID Grants AI083856
(Pressler, 2008; Kaplan, 2009; Bakker et al., 2011). (R.J.B.), 1RO1AI106733 (R.J.B. and C.M.), AI070456
The use of leukocytes enriched for neutrophils as the (I.M.O.), 1R01HL090991 (J.A.N.), Science Foundation
source of leukocyte lysate in these studies also has in vivo Arizona (J.G.L.), and Rebecca Runyon Bryan Chair for
relevance. Neutrophils in vivo are short-lived and are among CF, Ira, and Libbie Pink Charitable Fund (J.A.N.).
the first immune cells to encounter bacilli following myco-
bacterial infections (Canetti, 1955; Abadie et al., 2005).
Moreover, M. tuberculosis has been shown to survive References
extracellular killing by neutrophil extracellular nets (NETs) Abadie V, Badell E, Douillard P et al. (2005) Neutrophils rapidly
in vitro (Ramos-Kichik et al., 2009; Wong & Jacobs, 2013). migrate via lymphatics after Mycobacterium bovis BCG intrader-
The extrusion of eDNA from neutrophils and other cells is mal vaccination and shuttle live bacilli to the draining lymph
thought to be an alternative defense mechanism against nodes. Blood 106: 1843–1850.
extracellular bacteria but can also be damaging by contrib- Ahmad Z, Klinkenberg LG, Pinn ML et al. (2009) Biphasic kill curve
uting to the pathogenesis of infectious and noninfectious of isoniazid reveals the presence of drug-tolerant, not drug-resis-
inflammatory diseases (Brinkmann et al., 2004; Wartha tant, Mycobacterium tuberculosis in the guinea pig. J Infect Dis
et al., 2007; Ma & Kubes, 2008). It is also of interest that 200: 1136–1143.
Armitige LY, Jagannath C, Wanger AR & Norris SJ (2000)
M. tuberculosis and other mycobacteria express secreted or
Disruption of the genes encoding antigen 85A and antigen 85B
surface exposed DNA-binding proteins that may facilitate
of Mycobacterium tuberculosis H37Rv: effect on growth in culture
the interaction with host- or pathogen-derived eDNA and in macrophages. Infect Immun 68: 767–778.
(Prabhakar et al., 2004; Yeruva et al., 2006). Antibodies Bakker EM, Volpi S, Salonini E et al. (2011) Improved treatment
directed against DNABII family members of bacterial response to dornase alfa in cystic fibrosis patients using
DNA-binding proteins have been shown to be effective at controlled inhalation. Eur Respir J 38: 1328–1335.
disrupting bacterial biofilms and to render drug-tolerant Barclay WR, Ebert RH, Manthei RW & Roth LJ (1953) Distribution of
pathogens again susceptible to antimicrobial drugs C14 labeled isoniazid in sensitive and resistant tubercle bacilli
(Goodman et al., 2011). The targeting of DNA-binding and in infected and uninfected tissues in tuberculous patients.
proteins of M. tuberculosis may be beneficial in preventing Trans Annu Meet Natl Tuberc Assoc 49: 192–195.
Basaraba RJ (2008) Experimental tuberculosis: the role of
the establishment of drug-tolerant communities of bacilli and
comparative pathology in the discovery of improved tuberculo-
an attractive target in the development of new tuberculosis
sis treatment strategies. Tuberculosis (Edinb) 88(suppl. 1):
vaccines (Chen et al., 2008; Kaplan, 2009; Gordon et al., S35–S47.
2010; Liu & Gordon, 2012). Brinkmann V, Reichard U, Goosmann C et al. (2004) Neutrophil
We describe here a simple in vitro assay that favors the extracellular traps kill bacteria. Science 303: 1532–1535.
establishment of drug-tolerant M. tuberculosis, which does Canetti G (1955) The Tubercle Bacillus in the Pulmonary Lesion of
not require specific nutrient depletion, the need to maintain Man; Histobacteriology and its Bearing on the Therapy of
Pulmonary Tuberculosis. Springer Publishing Company Inc., New Leistikow RL, Morton RA, Bartek IL, Frimpong I, Wagner K &
York, NY. Voskuil MI (2010) The Mycobacterium tuberculosis DosR regulon
Chen JM, Ren H, Shaw JE et al. (2008) Lsr2 of Mycobacterium assists in metabolic homeostasis and enables rapid recovery
tuberculosis is a DNA-bridging protein. Nucleic Acids Res 36: from nonrespiring dormancy. J Bacteriol 192: 1662–1670.
2123–2135. Lenaerts AJ, Hoff D, Aly S et al. (2007) Location of persisting
Deb C, Lee CM, Dubey VS et al. (2009) A novel in vitro mycobacteria in a Guinea pig model of tuberculosis revealed by
multiple-stress dormancy model for Mycobacterium tuberculosis r207910. Antimicrob Agents Chemother 51: 3338–3345.
generates a lipid-loaded, drug-tolerant, dormant pathogen. PLoS Liu J & Gordon BR (2012) Targeting the global regulator Lsr2 as a
One 4: e6077. novel approach for anti-tuberculosis drug development. Expert
Dubos RJ (1946) Effect of long chain fatty acids on bacterial growth. Rev Anti Infect Ther 10: 1049–1053.
Proc Soc Exp Biol Med 63: 56–58. Ma AC & Kubes P (2008) Platelets, neutrophils, and neutrophil
Dubos RJ & Davis BD (1946) Factors affecting the growth of extracellular traps (NETs) in sepsis. J Thromb Haemost 6: 415–
tubercle bacilli in liquid media. J Exp Med 83: 409–423. 420.
Erdemli SB, Gupta R, Bishai WR, Lamichhane G, Amzel LM & Mann EE, Rice KC, Boles BR et al. (2009) Modulation of eDNA

Bianchet MA (2012) Targeting the cell wall of Mycobacterium release and degradation affects Staphylococcus aureus biofilm
tuberculosis: structure and mechanism of L, D-transpeptidase 2. maturation. PLoS One 4: e5822.
Structure 20: 2103–2115. Manthei RW, Roth LJ, Barclay WR & Ebert RH (1954) The
Fahy JV, Steiger DJ, Liu J, Basbaum CB, Finkbeiner WE & Boushey distribution of C14 labeled isoniazid in normal and infected
HA (1993) Markers of mucus secretion and DNA levels in induced guinea pigs. Arch Int Pharmacodyn Ther 98: 183–192.
sputum from asthmatic and from healthy subjects. Am Rev Respir Mitchison D & Davies G (2012) The chemotherapy of tuberculosis:
Dis 147: 1132–1137. past, present and future. Int J Tuberc Lung Dis 16: 724–732.
Fenhalls G, Stevens-Muller L, Warren R, Carroll N, Bezuidenhout J, Nozawa RT & Yokota T (1983) Rapid drug susceptibility testing of
Van Helden P & Bardin P (2002) Localisation of mycobacterial mycobacteria in tissue culture medium. Antimicrob Agents Che-
DNA and mRNA in human tuberculous granulomas. J Microbiol mother 24: 268–272.
Methods 51: 197–208. Obregon-Henao A, Duque-Correa MA, Rojas M, Garcia LF, Bren-
Franzblau SG, Witzig RS, McLaughlin JC et al. (1998) Rapid, nan PJ, Ortiz BL & Belisle JT (2012) Stable extracellular RNA
low-technology MIC determination with clinical Mycobacterium fragments of Mycobacterium tuberculosis induce early apoptosis
tuberculosis isolates by using the microplate Alamar Blue assay. in human monocytes via a caspase-8 dependent mechanism.
J Clin Microbiol 36: 362–366. PLoS One 7: e29970.
Franzblau SG, DeGroote MA, Cho SH et al. (2012) Comprehensive Ojha AK, Baughn AD, Sambandan D et al. (2008) Growth of
analysis of methods used for the evaluation of compounds Mycobacterium tuberculosis biofilms containing free mycolic acids
against Mycobacterium tuberculosis. Tuberculosis (Edinb) 92: and harbouring drug-tolerant bacteria. Mol Microbiol 69: 164–174.
453–488. Ordway DJ, Shanley CA, Caraway ML et al. (2010) Evaluation of
Goodman SD, Obergfell KP, Jurcisek JA et al. (2011) Biofilms can standard chemotherapy in the guinea pig model of tuberculosis.
be dispersed by focusing the immune system on a common family Antimicrob Agents Chemother 54: 1820–1833.
of bacterial nucleoid-associated proteins. Mucosal Immunol 4: Pang JM, Layre E, Sweet L, Sherrid A, Moody DB, Ojha A & Sherman
625–637. DR (2012) The polyketide Pks1 contributes to biofilm formation in
Gordon BR, Li Y, Wang L et al. (2010) Lsr2 is a nucleoid-associated Mycobacterium tuberculosis. J Bacteriol 194: 715–721.
protein that targets AT-rich sequences and virulence genes in Parker SK, Curtin KM & Vasil ML (2007) Purification and charac-
Mycobacterium tuberculosis. P Natl Acad Sci USA 107: 5154–5159. terization of mycobacterial phospholipase A: an activity associ-
Green C, Huggett JF, Talbot E, Mwaba P, Reither K & Zumla AI ated with mycobacterial cutinase. J Bacteriol 189: 4153–4160.
(2009) Rapid diagnosis of tuberculosis through the detection of Parks QM, Young RL, Poch KR, Malcolm KC, Vasil ML & Nick JA
mycobacterial DNA in urine by nucleic acid amplification methods. (2009) Neutrophil enhancement of Pseudomonas aeruginosa
Lancet Infect Dis 9: 505–511. biofilm development: human F-actin and DNA as targets for
Gruppo V, Johnson CM, Marietta KS et al. (2006) Rapid microb- therapy. J Med Microbiol 58: 492–502.
iologic and pharmacologic evaluation of experimental compounds Patel K, Jhamb SS & Singh PP (2011) Models of latent tuberculosis:
against Mycobacterium tuberculosis. Antimicrob Agents Chemo- their salient features, limitations, and development. J Lab Phy-
ther 50: 1245–1250. sicians 3: 75–79.
Harper J, Skerry C, Davis SL et al. (2012) Mouse model of necrotic Piddock LJ, Williams KJ & Ricci V (2000) Accumulation of
tuberculosis granulomas develops hypoxic lesions. J Infect Dis rifampicin by Mycobacterium aurum, Mycobacterium smegmatis
205: 595–602. and Mycobacterium tuberculosis. J Antimicrob Chemother 45:
Heng Y, Seah PG, Siew JY et al. (2011) Mycobacterium tubercu- 159–165.
losis infection induces hypoxic lung lesions in the rat. Tubercu- Prabhakar S, Mishra A, Singhal A, Katoch VM, Thakral SS, Tyagi
losis (Edinb) 91: 339–341. JS & Prasad HK (2004) Use of the hupB gene encoding a
Hoff DR, Ryan GJ, Driver ER, Ssemakulu CC, De Groote MA, histone-like protein of Mycobacterium tuberculosis as a target for
Basaraba RJ & Lenaerts AJ (2011) Location of intra- and detection and differentiation of M. tuberculosis and M. bovis.
extracellular M. tuberculosis populations in lungs of mice and J Clin Microbiol 42: 2724–2732.
guinea pigs during disease progression and after drug treatment. Pressler T (2008) Review of recombinant human deoxyribonuclease
PLoS One 6: e17550. (rhDNase) in the management of patients with cystic fibrosis.
Kaplan JB (2009) Therapeutic potential of biofilm-dispersing Biologics 2: 611–617.
enzymes. Int J Artif Organs 32: 545–554. Prideaux B, Dartois V, Staab D et al. (2011) High-sensitivity
Kapoor N, Pawar S, Sirakova TD, Deb C, Warren WL & Kolattukudy MALDI-MRM-MS imaging of moxifloxacin distribution in
PE (2013) Human granuloma in vitro model, for TB dormancy and tuberculosis-infected rabbit lungs and granulomatous lesions.
resuscitation. PLoS One 8: e53657. Anal Chem 83: 2112–2118.
Puchelle E, Zahm JM, de Bentzmann S, Grosskopf C, Shak S, Walker TS, Tomlin KL, Worthen GS et al. (2005) Enhanced
Mougel D & Polu JM (1996) Effects of rhDNase on purulent Pseudomonas aeruginosa biofilm development mediated by
airway secretions in chronic bronchitis. Eur Respir J 9: 765–769. human neutrophils. Infect Immun 73: 3693–3701.
Ramasubbu N, Thomas LM, Ragunath C & Kaplan JB (2005) Wartha F, Beiter K, Normark S & Henriques-Normark B (2007)
Structural analysis of dispersin B, a biofilm-releasing glycoside Neutrophil extracellular traps: casting the NET over pathogene-
hydrolase from the periodontopathogen Actinobacillus actinomy- sis. Curr Opin Microbiol 10: 52–56.
cetemcomitans. J Mol Biol 349: 475–486. Wayne LG & Hayes LG (1996) An in vitro model for sequential study
Ramos-Kichik V, Mondragon-Flores R, Mondragon-Castelan M of shiftdown of Mycobacterium tuberculosis through two stages of
et al. (2009) Neutrophil extracellular traps are induced by Myco- nonreplicating persistence. Infect Immun 64: 2062–2069.
bacterium tuberculosis. Tuberculosis (Edinb) 89: 29–37. Wayne LG & Sohaskey CD (2001) Nonreplicating persistence of
Ramsugit S, Guma S, Pillay B, Jain P, Larsen MH, Danaviah S & Mycobacterium tuberculosis. Annu Rev Microbiol 55: 139–163.
Pillay M (2013) Pili contribute to biofilm formation in vitro in Whitchurch CB, Tolker-Nielsen T, Ragas PC & Mattick JS (2002)
Mycobacterium tuberculosis. Antonie Van Leeuwenhoek 104: Extracellular DNA required for bacterial biofilm formation. Science
725–735. 295: 1487.

Ryan GJ, Hoff DR, Driver ER et al. (2010) Multiple M. tuberculosis Wilson GS & Schwabacher H (1935) The relationship between
phenotypes in mouse and guinea pig lung tissue revealed by a moist weight and numbers of total and viable organisms in
dual-staining approach. PLoS One 5: e11108. cultures of tubercle bacilli. Tubercle 17: 161.
Sambandan D, Dao DN, Weinrick BC et al. (2013) Keto-mycolic Wong KW & Jacobs WR Jr (2013) Mycobacterium tuberculosis
acid-dependent pellicle formation confers tolerance to drug-sen- exploits human interferon gamma to stimulate macrophage
sitive Mycobacterium tuberculosis. MBio 4: e00222–00213. extracellular trap formation and necrosis. J Infect Dis 208: 109–
Sani M, Houben EN, Geurtsen J et al. (2010) Direct visualization by 119.
cryo-EM of the mycobacterial capsular layer: a labile structure Yang L, Barken KB, Skindersoe ME, Christensen AB, Givskov M &
containing ESX-1-secreted proteins. PLoS Pathog 6: e1000794. Tolker-Nielsen T (2007) Effects of iron on DNA release and
Sarathy J, Dartois V, Dick T & Gengenbacher M (2013) Reduced biofilm development by Pseudomonas aeruginosa. Microbiology
drug uptake in phenotypically resistant nutrient-starved nonrepli- 153: 1318–1328.
cating Mycobacterium tuberculosis. Antimicrob Agents Chemo- Yeruva VC, Duggirala S, Lakshmi V, Kolarich D, Altmann F & Sritharan
ther 57: 1648–1653. M (2006) Identification and characterization of a major cell
Sattler TH & Youmans GP (1948) The effect of “Tween 80”, bovine wall-associated iron-regulated envelope protein (Irep-28) in Myco-
albumin, glycerol, and glucose on the growth of Mycobacterium bacterium tuberculosis. Clin Vaccine Immunol 13: 1137–1142.
tuberculosis var. hominis (H37Rv). J Bacteriol 56: 235–243. Zhang M, Gong J, Lin Y & Barnes PF (1998) Growth of virulent and
Speer A, Shrestha TB, Bossmann SH et al. (2013) Copper-boosting avirulent Mycobacterium tuberculosis strains in human macro-
compounds: a novel concept for antimycobacterial drug discov- phages. Infect Immun 66: 794–799.
ery. Antimicrob Agents Chemother 57: 1089–1091.
Stoops JK, Arora R, Armitage L et al. (2010) Certain surfactants
show promise in the therapy of pulmonary tuberculosis. In Vivo Supporting Information
24: 687–694.
Tetz GV, Artemenko NK & Tetz VV (2008) Effect of DNase and Additional Supporting Information may be found in the online
antibiotics on biofilm characteristics. Antimicrob Agents Chemo- version of this article:
ther 53: 1204–1209.
Tsai MC, Chakravarty S, Zhu G et al. (2006) Characterization of the Fig. S1. Treatment of drug-tolerant Mycobacterium tuber-
tuberculous granuloma in murine and human lungs: cellular culosis with Tween-80 restores in vitro susceptibility to
composition and relative tissue oxygen tension. Cell Microbiol 8: isoniazid.
218–232. Fig. S2. The HN878 strain of Mycobacterium tuberculosis
Ulrichs T & Kaufmann SH (2006) New insights into the function of cultured in nutrient-rich mammalian cell culture media
granulomas in human tuberculosis. J Pathol 208: 261–269. expresses differences in tolerance to isoniazid in vitro.
van Staden M, van der Ryst E, Attwood EM, Hendricks ML, Joubert Fig. S3. Mycobacterium tuberculosis form complex
G & Weich DJ (1998) Detection of Mycobacterium tuberculosis in
microbial communities attached to the complex extracellular
serum samples using the polymerase chain reaction. J Infect 36:
matrix derived from lysed human leukocytes.
273–277.
Via LE, Lin PL, Ray SM et al. (2008) Tuberculous granulomas are Fig. S4. Treatment of attached communities of Mycobac-
hypoxic in guinea pigs, rabbits, and nonhuman primates. Infect terium tuberculosis with DNase I is effective at depleting
Immun 76: 2333–2340. extracellular DNA.

70-3-359

Uploaded by

Copyright:

Available Formats

70-3-359

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

70-3-359

Uploaded by

Copyright:

Available Formats

Pathogens and Disease ISSN 2049-632X

Expression of antimicrobial drug tolerance by attached communities

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

conditions that exist during M. tuberculosis infection (Deb

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

(a) (b) (c)

(a) (b) (c)

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

(a) (b) (c)

Table 1 The MIC of isoniazid and rifampin in increased when Myco-

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

(a) (b) (c)

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

(a) (b) (c)

the formation of attached communities and the expression

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

Downloaded from https://academic.oup.com/femspd/article/70/3/359/568302 by guest on 03 December 2024

You might also like