Antibiotic Persistence in Mycobacterium tuberculosis · 42 Mycobacterium tuberculosis, which is the causative agent of tuberculosis, is one of the 43 most successful human pathogens

1

Reduced Permeability to Rifampicin by Capsular Thickening as a Mechanism of 1

Antibiotic Persistence in Mycobacterium tuberculosis 2

3

4

Jees Sebastian, Sharmada Swaminath, Parthasarathi Ajitkumar* 5

Department of Microbiology and Cell Biology, Indian Institute of Science, Bangalore, 6

Karnataka, India. 7

8

Running title: Rifampicin persistence in Mtb by capsular thickening 9

10

*Address correspondence to P. Ajitkumar, [email protected]. 11

12

Key words: Mycobacterium tuberculosis; Rifampicin permeability; Persistence; Thick 13

outer layer; Polysaccharides 14

15

16

.CC-BY-NC-ND 4.0 International licenseunder acertified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available

The copyright holder for this preprint (which was notthis version posted May 1, 2019. ; https://doi.org/10.1101/624569doi: bioRxiv preprint

https://doi.org/10.1101/624569

http://creativecommons.org/licenses/by-nc-nd/4.0/

2

ABSTRACT 17

Persisters constitute a subpopulation of bacteria that can tolerate lethal concentrations 18

of antibiotics. Multiple mechanisms have been suggested for bacterial persistence 19

against antibiotics. With mycobacteria being no exception to this behaviour, we had 20

reported the de novo emergence of genetically antibiotic-resistant Mycobacterium 21

tuberculosis from persister cells upon prolonged exposure to microbicidal 22

concentrations of the anti-tuberculosis drugs, rifampicin and moxifloxacin. Here, we 23

present evidence for reduced permeability to rifampicin as a mechanism for 24

persistence of Mycobacterium tuberculosis in vitro. We observed that rifampicin 25

persistent M. tuberculosis cells developed a thick outer layer (TOL) capsule. The TOL 26

restricted the entry of fluorochrome-conjugated rifampicin, 5-carboxyfluorescein-27

rifampicin (5-FAM-rifampicin), which retained only 2.5% of its original bactericidal 28

activity, but high levels of permeability, on actively growing mid-log phase cells. Gentle 29

mechanical removal of TOL significantly enhanced 5-FAM-rifampicin entry into the 30

persister cells. The level of 5-FAM-rifampicin in the persister cells was not affected by 31

the pre-incubation of the cells with verapamil, a drug efflux pump inhibitor, ruling out 32

the involvement of efflux pumps in the reduced intracellular concentration of 5-FAM-33

rifampicin. GC-MS analysis of TOL showed the presence of ~7-fold, ~5-fold and ~2-34

fold higher levels of α-D-glucopyranoside, 1,2,5-linked-mannitol, and 3,4-linked 35

mannose, respectively, among ~2-fold higher levels of derivatives of several other 36

types of sugars such as arabinose and galactose. Taken together, the present study 37

reveals that rifampicin-persistent M. tuberculosis cells develop TOL that enables the 38

bacilli to restrict entry of rifampicin and thereby remain tolerant to the antibiotic in vitro. 39

40



https://doi.org/10.1101/624569


3

INTRODUCTION 41

Mycobacterium tuberculosis, which is the causative agent of tuberculosis, is one of the 42

most successful human pathogens due to its ability to survive under diverse extreme 43

stress conditions. Like many other pathogenic bacteria, M. tuberculosis also exhibits 44

several survival strategies to overcome the lethality imposed by antibiotic exposure. 45

One of the unique features of M. tuberculosis that has enabled the bacilli to remain 46

refractory to many commonly used antibiotics is the distinct cell wall structure (1). The 47

complex cell wall of M. tuberculosis is well known for imposing limited permeability to 48

several host-derived antimicrobial biomolecules that makes the bacilli tolerant to many 49

toxic components from its normal habitats (2-5). Studies have been conducted on the 50

physiological significance of the changes that the cell wall undergo inside 51

macrophages (6), under extreme nutritional stress conditions (7) and hypoxia (8). 52

However, there is no information on whether the cell envelope has any contribution to 53

the persistence of the bacilli against antibiotics. 54

Persistence is one of the survival strategies of mycobacteria to remain tolerant 55

to lethal concentrations of antibiotics. Although several mechanisms that contribute to 56

mycobacterial persistence have been reported (9, 10), very little information is 57

available on the ultrastructural features of mycobacterial antibiotic persisters and their 58

contribution, if any, to antibiotic tolerance. Earlier studies in Escherichia coli and 59

Staphylococcus aureus have shown that exposure to a sub-lethal concentration of 60

antibiotics can cause the emergence of resisters through the generation of reactive 61

oxygen species and consequential mutagenesis (11). Subsequently, in our study, M. 62

tuberculosis persisters were found in the presence of lethal concentrations of 63

rifampicin and moxifloxacin, leading to the de novo emergence of antibiotic resisters 64

(12). These two studies raised the possibility that the bacilli might have some hitherto 65



https://doi.org/10.1101/624569


4

unknown strategy to ensure sub-lethal concentration of antibiotics inside persister cells 66

even though they remain exposed to lethal concentrations of the antibiotics. In order 67

to verify this possibility, we examined whether the permeability of rifampicin into M. 68

tuberculosis persister cells was affected by any ultrastructural changes. The present 69

study reports the ultrastructural changes of rifampicin persistent M. tuberculosis cells 70

and their contribution to the restricted permeability of rifampicin, in comparison with 71

the entry of rifampicin into actively growing mid-log phase cells. 72

73

MATERIALS AND METHODS 74

75

Rifampicin treatment of Mycobacterium tuberculosis 76

Mycobacterium tuberculosis H37Ra (National JALMA Institute of Leprosy and Other 77

Mycobacterial Diseases, Agra, India) was used in all the experiments. Bacteria were 78

grown in Middlebrook 7H9 broth (BD) supplemented with 0.2% glycerol, 0.05% 79

Tween-80 and 10% ADS (albumin, dextrose, sodium chloride) until mid-log phase 80

(MLP, 0.6 at OD600 nm) at 37ºC under shaking at 170 rpm. Cultures were then exposed 81

to 1 µg/ml (10x MBC; 12) of freshly prepared rifampicin (Sigma) under the same 82

condition. Aliquots of the culture were withdrawn once in every 24 hrs and used for 83

serial dilution and plating on sterility checked Middlebrook 7H10 agar supplemented 84

with 10% ADS in the absence of rifampicin. Plates were sealed with parafilm and 85

incubated at 37ºC in CO2 incubator (5% CO2) for determining CFU. For experiments 86

using 5-FAM-rifampicin (Merck-Millipore), MLP or persistence phase cells were 87

exposed to 1.5 µg/ml (concentration equimolar to 10x MBC rifampicin) for the required 88

period and taken for flow cytometry analysis (see below). 89

90



https://doi.org/10.1101/624569


5

Transmission electron microscopy 91

M. tuberculosis cells from different stages of rifampicin treatment were harvested and 92

processed for transmission electron microscopy as previously described (13). In brief, 93

the cells were prefixed in 1% (w/v) osmium tetroxide (solution in double-distilled water) 94

in 0.15 M cacodylate buffer (pH 7.2) for 1 hr at room temperature and washed once in 95

the same buffer. Further, cells were fixed using 2% (v/v) glutaraldehyde and 2% (w/v) 96

tannic acid in 0.15 M cacodylate buffer for 2 hrs at room temperature. Cells were 97

washed once in the same buffer and re-fixed in 1% osmium tetroxide overnight at 4ºC. 98

Samples were dehydrated by a series of ethanol washes (with 25%, 50%, 75% and 99

95% ethanol) (5 min incubation between every wash) and infiltrated with 50% LR white 100

resin in 50% ethanol for 24 hrs at 4°C. Dehydrated cells were collected by 101

centrifugation and the pellet was used for making blocks in gelatine capsules with 102

100% LR white resin. Blocks were solidified by incubating at 65°C for 2-3 days in a 103

dry bath. Ultrathin sections of 80 to 100 nm were prepared from the blocks using 104

ultramicrotome and collected on copper grids. Sections were stained with uranyl 105

acetate (0.5%) and lead citrate (0.04%) and observed under JEOL-100 CX II electron 106

microscope at 100 kV. 107

For ruthenium red staining, we followed the previously reported method (14). In 108

brief, the cells were harvested and washed once in Middlebrook 7H9 medium and 109

prefixed in the presence of paraformaldehyde and glutaraldehyde (2.5% w/v each), 110

CaCl2 and MgCl2 (5 mM each) and 0.05% (w/v) ruthenium red in 0.1 M cacodylate 111

buffer (pH 7.2) for 2 hrs at 4ºC. Cells were washed once in the same buffer and 112

incubated with 2.5% w/v glutaraldehyde and 0.05% (w/v) of ruthenium red in 113

cacodylate buffer. Cells were washed once and post-fixed in osmium tetroxide (OsO4) 114

and uranyl acetate (1% w/v of each) containing ruthenium red (0.05% w/v). Fixed cells 115



https://doi.org/10.1101/624569


6

were dehydrated with a series of ethanol washes (with 25%, 50%, 75% and 95% 116

ethanol) (5 min incubation between every wash, ethanol contained 0.05% w/v of 117

ruthenium red) and infiltrated with 50% LR white resin in 50% ethanol for 24 hrs at 118

4°C. Cells were collected by centrifugation and the pellet was used for making blocks 119

in gelatine capsules with 100% LR White resin. Blocks were solidified by incubating at 120

65°C for 2-3 days in dry bath. Ultrathin sections of 80-100 nm were prepared from the 121

blocks using ultra-microtome and placed on copper grids. Sections were stained with 122

uranyl acetate (0.5%) and lead citrate (0.04%) and observed under JEOL-100 CX II 123

electron microscope at 100 kV. 124

125

Bacterial hydrophobicity assay 126

A modified protocol of a previously reported method (15, 16) was used. Cells from 127

MLP and rifampicin persistence phase were harvested by centrifugation and 128

resuspended in filter sterilised PUM buffer (100 mM K2HPO4.3H2O, 50 mM KH2PO4, 129

33.3 mM urea, 1 mM MgSO4.7H2O, in double-distilled autoclaved water, pH 7.1) to 130

get an approximate density of 0.7 at OD600 nm. Phase extraction was performed in 131

siliconised borosilicate tubes using one volume of PUM buffer containing cells against 132

three volumes of n-hexadecane and vortexed for 8-10 seconds and left at room 133

temperature for 15 min. The aqueous phase was collected into siliconised 134

microcentrifuge tubes using siliconised tips, mildly sonicated and used for serial 135

dilution and plating on sterility checked Middlebrook 7H10 agar plates containing 10% 136

ADS supplement. The difference in the CFU of the aqueous phase before and after 137

phase separation was used for calculating the percentage of cells with hydrophilic 138

outer layer. 139

140



https://doi.org/10.1101/624569


7

Estimation of zeta potential of mycobacterial persister cells 141

For the determination of surface charge of M. tuberculosis cells, aliquots of the culture 142

were withdrawn on different days during rifampicin treatment and washed once with 143

Middlebrook 7H9 broth. Cells were resuspended in fresh medium and used for zeta 144

potential measurement using zeta sizer nano series (Nano-ZS90, Malvern 145

Instruments). For the estimation of the isoelectric point of the cells, the cells were 146

resuspended in solutions of varying pH from 2 to 10 in PPMS buffer (40 mM of 147

K2HPO4, 20 mM of KH2PO4, and 1.5 mM of MgSO4.7H2O per litre of Milli Q water, pH 148

adjusted using HCl or NaOH). Zeta potential values of the cells at different pH were 149

calculated and the graph was plotted, as described (17, 18). 150

151

GC-MS analysis of OL 152

Cells from MLP and persistence phase were collected and used for OL extraction as 153

previously reported (19) with minor modifications. Briefly, cells were washed once in 154

1x PBS and resuspended in 20 ml distilled water followed by incubation with 10 gm of 155

sterile 4 mm glass beads at 50 rpm for 15 min. The cell suspension was collected and 156

centrifuged at 12000 xg for 10 min. The supernatant was filtered through a 0.2 micron 157

filter and the filtrate was lyophilized. The sample was derivatised and subjected to GC-158

MS performed at Mass spectrometry glycomics facility at C-CAMP Bangalore. Data 159

analysis was also performed at the same facility. 160

161

Fluorescence microscopy 162

Cells from MLP and persistence phases were harvested and resuspended in 100 µl of 163

Middlebrook 7H9 broth. For the staining of cells, 5-FAM-rifampicin (1.5 µg/ml) or 164

calcofluor white (CFW) (1:1000 dilutions of 0.1% solution, Sigma) was added into the 165



https://doi.org/10.1101/624569


8

cell suspension and incubated for 1 hr at 37ºC incubator shaker. Propidium iodide 166

(1:1000) was added (to discriminate dead cells) and layered over poly-L-lysine treated 167

multiwell slides 20 min in the dark. Wells were washed once with phosphate buffered 168

saline (PBS) and mounted with glycerol and cover slips and observed under 100X 169

objective in Zeiss AxioVision fluorescence microscope. Bead beating was performed 170

by incubating 20 ml culture with 10 grams of glass beads (4 mm) at 37ºC for 15 min 171

at 50 rpm, prior to the addition of 5-FAM-rifampicin and processed similarly as above. 172

173

Flow cytometry analysis 174

Cell suspension (500 µl) was exposed to 1.5 µg/ml of 5-FAM-rifampicin and incubated 175

at 37ºC incubator for 1 hr in the dark. Aliquots were collected at specified intervals and 176

cells were harvested by centrifugation at 12000 x g for 5 min at 4ºC and washed once 177

with ice-cold Middlebrook 7H9 broth and used for flow cytometry analysis. For bead 178

beating, 20 ml culture was incubated with 10 grams of glass beads (4 mm) at 37ºC for 179

15 min at 50 rpm prior to the addition of 5-FAM-rifampicin and processed similarly for 180

flow cytometry with 488 nm solid state laser and 527/32 nm emission filter (for GFP) 181

at low or medium flow rate. For relative permeability estimation of rifampicin persisters 182

for 5-FAM-rifampicin, cells from MLP, persistence phase and verapamil-pretreated (50 183

µg/ml) persisters were incubated with 1.5 µg/ml of 5-FAM-rifampicin and incubated for 184

1 hr at 37ºC in the dark and processed in the same way as described earlier. 185

For CFW staining and OL analysis, 500 µl aliquots of the cells before and after 186

bead beating were incubated with 1:1000 dilutions of CFW (0.1% solution, Sigma) for 187

1 hr at 37ºC. Cells were washed once with Middlebrook 7H9 broth and used for flow 188

cytometry analysis in BD FACSVerse flow cytometer. 189



https://doi.org/10.1101/624569


9

About 10000 gated cells were analysed and used for the estimation of the median 190

fluorescence value of the 5-FAM-rifampicin stained cells while keeping the 191

autofluorescence at 2-log10 value. The photomultiplier tube (PMT) voltage settings 192

used for measuring the 5-FAM-rifampicin fluorescence of the cells were 208 (FSC), 193

333 (SSC). The instrument was calibrated using FACSuite cytometer set up and 194

tracking beads (CS&T, Becton Dickinson). Data were processed and analysed using 195

FACSuite software and the statistical significance between the time points was 196

calculated using paired t-test of GraphPad Prism version 5.0. 197

198

5-FAM-rifampicin bioassay 199

Rifampicin sensitive Staphylococcus aureus (ATCC 25923) was used for the bioassay 200

of 5-FAM-rifampicin (20). In brief, rifampicin (sigma) and 5-FAM-rifampicin were 201

dissolved in DMSO to make a stock solution of 2 mg/ml. Dilutions were made from the 202

stock solution and used for the agar diffusion assay. LB agar plates were made by 203

mixing 50 µl of the Staphylococcus aureus glycerol stock with 100 ml LB agar (warm 204

to the touch). Wells were made in the agar using a stainless steel puncture with 0.5 205

cm diameter. Known concentrations of rifampicin or 5-FAM-rifampicin solutions (50 µl) 206

were added into the well and incubated overnight at 37°C. The diameter of the zone 207

of inhibition was measured using Vernier caliper and used for plotting the standard 208

graph for rifampicin, from which and the zone of inhibition of 5-FAM-rifampicin, the 209

bioactivity of 5-FAM-rifampicin was calculated. 210

211

5-FAM-rifampicin permeability assay 212

To estimate the extent of 5-FAM-rifampicin entry into MLP cells and rifampicin 213

persisters, cultures were treated with a final concentration of 1.5 µg/ml of the drug 214



https://doi.org/10.1101/624569


10

conjugate in 20 ml culture and incubated in a shaker at 37°C. Cultures for both (MLP 215

and persisters) were mildly beaten with 4 mm glass beads to remove the outer 216

capsular layer, as a control for the experiment. Aliquots were taken at every 15 min 217

intervals and quickly washed once with ice-cold Middlebrook 7H9 broth and used for 218

flow cytometry analysis using BD FACSVerse system. Samples before 5-FAM-219

rifampicin addition (0 min) were used as the control for autofluorescence and median 220

of fluorescence was kept at 102. 221

To construct a standard graph for 5-FAM-rifampicin entry, MLP cells were 222

incubated with two-fold increasing concentrations of 5-FAM-rifampicin in an incubator 223

shaker for 1 hr at 37°C. Cells were harvested and quickly washed once with ice-cold 224

Middlebrook 7H9 broth and used for flow cytometry analysis in BD FACSVerse 225

system. Autofluorescence median of the untreated cells was kept at 102 and the 226

normalised median values were used for plotting standard graph. Rifampicin 227

persisters, which were isolated from the 12th day of the treatment, were incubated with 228

5-FAM-rifampicin equivalent of 10x MBC rifampicin. These cells were also similarly 229

processed and readings were taken from 10000 cells. Normalised fluorescence values 230

from six independent persister cell samples were used for calculating the relative 231

concentration of 5-FAM-rifampicin in persister cells from the standard graph. 232

233

RESULTS AND DISCUSSION 234

Earlier studies from our laboratory had shown that actively growing MLP M. 235

tuberculosis cells exposed to 10x MBC concentration of rifampicin consistently 236

showed a triphasic response involving killing phase (0 to 10th day of exposure), 237

persistence phase (10th to 15th day of exposure), followed by regrowth phase (beyond 238

15th day of exposure) (12). As part of our earlier characterisation of the triphasic 239



https://doi.org/10.1101/624569


11

response of M. tuberculosis cells to 10x MBC rifampicin, we had found that the 240

rifampicin concentration stays at ~5x MBC even on the 15th day of the prolonged 241

exposure that spanned for 20 days (12). Therefore, for the present ultrastructural 242

studies, the persistence phase cells from the 12th day of exposure to 10x MBC 243

rifampicin were used. 244

245

Rifampicin persister cells possess thick outer layer (TOL) 246

Transmission electron micrographs of rifampicin persister M. tuberculosis cells, 247

stained with tannic acid-lead citrate combination (13), showed a significantly thick, but 248

strikingly uneven, loosely bound and deeply stained outer layer (OL) (Fig. 1A). On the 249

contrary, the MLP cells (control) showed an evenly thin outer layer, which we called 250

normal outer layer (NOL) (Fig. 1B). The TOL thickness of persister cells ranged 251

between 25-130 nm (n = 27), while the MLP cells showed a more-or-less evenly thin 252

layer of NOL with an average thickness of ~20 nm (Fig. 1C). However, the thickness 253

of electron transparent layer (ETL) of the persister and the control cells were found 254

comparable and were morphologically like the already published data (Fig. 1D) (14, 255

21, 22). Since rifampicin exposure for the experimental sample was started from the 256

MLP cells, cultures from MLP stage for an equivalent number of days, but unexposed 257

to rifampicin did not show any TOL (Fig. S1). Thus, the ultrastructural difference 258

between the rifampicin persister cells and the control cells was confined to the outer 259

layer (OL) thickness. 260

Transmission electron microscopy of persister and MLP cells stained with 261

ruthenium red was performed to specifically stain polysaccharides for detection (14). 262

It showed the presence of polysaccharides as an uneven layer on the surface of 263

rifampicin persister cells unlike a very thin layer on the MLP control cells (n = 22) (Fig. 264



https://doi.org/10.1101/624569


12

S2). The thin OL on the MLP also contained polysaccharides, as reported (19). The 265

presence of polysaccharides was further confirmed by staining with a polysaccharide 266

specific fluorophore, calcofluor white (CFW; 23, 24) (Fig. 1E, top panel). The persister 267

cells, which were gently bead beaten to remove OL, as described (19), and stained 268

with CFW, showed loss of CFW fluorescence (Fig. 1E, bottom panel). Flow cytometry 269

analysis of CFW stained native and bead beaten persister cells also showed loss of 270

CFW fluorescence. However, the MLP cells (Fig. 1F, H, J) did not show an appreciable 271

difference in CFW fluorescence between the native and the bead beaten samples, 272

compared to persister cells (Fig. 1G, I, J). Thus, by using multiple staining methods 273

with two different polysaccharide specific reagents confirmed the differential 274

polysaccharide content on the MLP and persister cells. 275

276

TOL imparts hydrophilicity to rifampicin persister cells’ surface 277

It was shown that the constituents of the OL of actively growing mycobacteria are 278

mostly polysaccharides and proteins with low lipid content (19). Therefore, due to the 279

thickening of the OL in persister cells, the possibility of a change in the hydrophilicity 280

of the surface of persister cells, as compared to that of the actively growing MLP cells, 281

was verified. For this purpose, cells from the rifampicin persistence phase and MLP 282

were subjected to phase separation between highly hydrophobic hexadecane and 283

aqueous buffer to measure cell-surface hydrophobicity, as described (15). The cells 284

retained in the aqueous phase after the phase extraction reflects the proportion of cells 285

that have a hydrophilic surface. We observed that an average 7% of the persister 286

population was hydrophilic and retained in the aqueous phase, while only 0.03% of 287

MLP cells being hydrophilic (Fig. 2A). This showed that the persister population 288

contained a relatively higher proportion of cells having hydrophilic surface. 289



https://doi.org/10.1101/624569


13

290

Rifampicin persister M. tuberculosis cells possess negative surface charge 291

The presence of anionic polysaccharides in the TOL could impart a negative surface 292

charge to the persister cells and confer hydrophilicity. The cell surface negative charge 293

can be measured by the zeta potential (ZP) of intact cells (17, 18). The ZP of M. 294

tuberculosis cells was measured for MLP cells and the cells from the killing, 295

persistence and regrowth phases of rifampicin exposure, which were described before 296

(12). M. tuberculosis MLP cells showed a ZP value of -3.91 mV, with a gradual 297

increase in the negative surface charge over the course of rifampicin exposure (Fig. 298

2B). The gradual increase in the negative ZP value upon continued exposure to 299

rifampicin indicated accumulation of negatively charged molecules on the cell surface. 300

Cells from the killing phase showed a ZP of -19.26 mV indicating a remarkable 301

increase in their negative surface charge compared to that of the rifampicin-unexposed 302

MLP cells. The persistence phase cells showed a peak ZP value of -32.76 mV, 303

showing that the persister cells have significantly high negative surface charge density 304

compared to that of the cells from any other phase of rifampicin exposure. The high 305

negative ZP of persister cells indicated the accumulation of anionic polysaccharides 306

on the cell surface. The regrowth phase cells showed a reduction in the negative ZP 307

value indicating the loss of TOL once the cells gained rifampicin resistance and came 308

out of the persistence phase. In addition, we observed multiple peaks in the ZP 309

histogram of RG phase population, suggesting the possibility of heterogeneity in terms 310

of surface charge among the rifampicin-tolerant/resistant regrowing cells (Fig. 2C). 311

In order to determine ionic properties of the mycobacterial TOL, the isoelectric 312

point (pI) of M. tuberculosis cells was measured at different pH values (2, 4, 6, 8 and 313

10). The pH at which cells show zero ZP value was considered as the pI of the cells. 314



https://doi.org/10.1101/624569


14

To estimate the isoelectric point (pI) of M. tuberculosis cells, at different pH values, we 315

measured the ZP of cells from various time points during rifampicin exposure. The 316

positive ZP of MLP cells at pH 2.0 and pH 4.0 were dropped to -2.58 at pH 6.0 317

indicating that the pI of MLP cells were between pH 4.0 and pH 6.0 (Fig. 2D). The 318

persistence phase cells showed a similar pI profile with a lower ZP value at high pH 319

(Fig. 2E). Although the cells in the killing phase also showed a similar pI, the extent of 320

negative ZP potential at high pH was considerably lesser than that of the persistence 321

phase cells (Fig. 2F). Thus, M. tuberculosis cells from MLP, killing and persistence 322

phases showed similar pI values with a varying surface charge at higher pH, probably 323

due to the high levels of anionic polysaccharide content on their surface. The 324

comparable pI values of the MLP cells and persisters indicated that the ionic properties 325

of the TOL might not have changed over the period of rifampicin exposure. This 326

alluded to the possibility that the nature of the molecules might be similar but their 327

relative composition might be different (see GCMS data described below), thereby 328

having similar pI. 329

330

Molecular analysis of TOL composition 331

A comparative molecular analysis using GC-MS was performed to find out the 332

polysaccharide composition of TOL on persistence phase cells and its difference from 333

that of the NOL on MLP cells. For this purpose, the NOL from MLP cells and TOL from 334

persister cells were gently extracted, as described (19) and used for mass 335

spectrometry. The percentage of relative abundance of the monosaccharides detected 336

are listed in Table 1. We observed ~5-fold increase in the levels of 1, 2, 5-mannitol 337

and ~6-7-fold increase in the α-D-glucopyranoside levels. This composition is 338

consistent with an earlier study that the NOL of actively growing M. tuberculosis (19). 339



https://doi.org/10.1101/624569


15

The thickening of the OL has caused several fold increase in the levels of the 340

constituents that were present in the NOL of MLP cells (see Table 1). For instance, 341

the 6-7-fold higher levels of α-D-glucopyranoside, as a breakdown product in the GC-342

MS analysis, alluded to the possibility of the presence of trehalose, which is α-D-343

glucopyranosyl-α-D-glucopyranoside (glucose disaccharide), known to desiccate 344

bacterial cells against severe stress conditions (25-27). Similarly, glucan, which is a 345

component of the M. tuberculosis cell surface, is a polymer of glucose that is 346

expressed in vitro and in vivo (28). The presence of high levels of α-D-glucopyranoside 347

might be an indication of the presence of glucan as well. The increased levels of 348

arabinose and mannose probably signify the presence of arabinomannan, which is 349

one of the significant components of OL (1, 29, 30). The high levels of polysaccharides 350

in the OL are known to be used as a bacterial decoy for antimicrobial peptides, for 351

respiratory tract colonisation, pathogenesis, cellular invasion, antiphagocytosis (16, 352

30-32). The presence of high levels of polysaccharides in the TOL of rifampicin 353

persister M. tuberculosis cells brings up another role for the OL polysaccharides in 354

mycobacterial physiology. Although differences in terms of relative abundance of 355

polysaccharides were observed, the molecular composition of OL between the MLP 356

and the persister cells was comparable. 357

358

Persister cells restrict rifampicin permeability 359

Reduced cell wall permeability is known to be a factor contributing to antibiotic 360

resistance in Neisseria meningitidis and Staphylococcus aureus (33-35). In 361

mycobacteria also, the role of membrane permeability in rifampicin resistance in the 362

actively growing populations of Mycobacterium intracellulare has been reported (36). 363

Therefore, presuming that the TOL might restrict the entry of rifampicin, experiments 364



https://doi.org/10.1101/624569


16

were carried out to understand the permeability characteristics of rifampicin persister 365

cells. For this purpose, 5-carboxyfluorescin (5-FAM) conjugated rifampicin (5-FAM-366

rifampicin) was used to monitor entry of the same into persister cells, in comparison 367

with its entry into MLP cells. 368

Conjugation of the small hydrophobic fluorophore, 5-FAM, to rifampicin 369

generated two possible isomers of 5-FAM-rifampicin depending on the ester bond 370

formed on the two aliphatic hydroxyl group on rifampicin molecule (Fig. 3A). 5-FAM 371

group was selected due to the nonpolar nature and smaller size of the fluorophore. 372

This was to avoid any kind of alteration in the polarity of the molecule and to minimise 373

the molecular size to maintain its entry into rifampicin-unexposed cells. Thus, it is like 374

rifampicin, which is nonpolar in nature and believed to passively diffuse through the 375

mycobacterial cell wall and accumulate inside the mycobacterial cell within 20 min of 376

exposure (37). Further, the conjugation of 5-FAM to rifampicin reduced the bactericidal 377

activity of the antibiotic to 2.5% of its original bioactivity, as calculated from the bio-378

assay (Fig. 3B; Fig. S3). An average zone of inhibition of 1.77 ± 0.035 cm was 379

obtained for 5-FAM-rifampicin at a concentration of 29.2 µg/ml, while 0.74 µg/ml of 380

rifampicin gave the same zone of inhibition, showing the drop in bioactivity of 5-FAM-381

rifampicin to 2.5% compared to that of rifampicin. Owing to these molecular properties, 382

use of 5-FAM-rifampicin avoided inflicting lethality on the cells which would have 383

otherwise affected the permeability assay. Radioactively labelled native rifampicin was 384

not used as it would have affected the viability of the cells. The concentration-385

dependent entry of 5-FAM-rifampicin into MLP cells confirmed its high level of 386

permeability (Fig. 3C, D). The extent of increase in the permeability showed a linear 387

correlation to the concentration of 5-FAM-rifampicin (Fig. 3E). This indicated that the 388

conjugation of 5-FAM to rifampicin did not affect its uptake by the actively growing 389



https://doi.org/10.1101/624569


17

MLP cells. The permeability characteristics of 5-FAM-rifampicin into MLP cells 390

validated that the uptake of 5-FAM-rifampicin by the persister cells could be 391

considered as a measure of its extent of permeability into the persister cells as well, 392

in comparison to MLP cells. 393

MLP and persister cells were incubated with 1.5 µg/ml (concentration equivalent 394

to 10x MBC rifampicin used in our earlier study; 12) of 5-FAM-rifampicin and analysed 395

using fluorescence microscopy. The low levels of fluorescence in the persister cells, 396

as compared to that in the MLP cells, indicated restricted entry of 5-FAM-rifampicin 397

into persister cells (Fig. 4A). Fluorescence microscopy of persister cells, which were 398

gently bead beaten to remove TOL and incubated with 5-FAM-rifampicin, showed 399

enhanced levels of 5-FAM fluorescence indicating increased 5-FAM-rifampicin entry 400

into the persister cells (Fig. 4B). Further, we determined the relative time-dependent 401

entry of 5-FAM-rifampicin into MLP and persister cells, with or without bead beating. 402

For this purpose, the cells were incubated over a period of one hour with 1.5 µg/ml of 403

5-FAM-rifampicin and aliquots were withdrawn every 15 min followed by flow 404

cytometry analysis. Both the native and the bead beaten MLP cells showed a time-405

dependent steady increase in the 5-FAM-rifampicin fluorescence showing that the thin 406

OL of MLP cells did not play any role in the permeability of 5-FAM-rifampicin (Fig. 4C, 407

E, G). On the contrary, incubation of persister cells with 5-FAM-rifampicin did not show 408

any noticeable fluorescence inside the cells (Fig. 4D, F upper panels, H). This 409

indicated restricted permeability of the rifampicin conjugate into the cells. Whereas, 410

the bead beaten rifampicin persisters showed a significant time-dependent increase 411

in the fluorescence for at least upto first 30 min of incubation, followed by a level of 412

saturation, suggesting the increased permeability of 5-FAM-rifampicin (Fig. 4D, F 413

lower panels, H). Thus, the removal of the TOL by bead beating allowed permeability 414



https://doi.org/10.1101/624569


18

of 5-FAM-rifampicin into the persister cells. These experiments confirmed the 415

substantial role of TOL to function as a ‘barrier’ to restrict permeability of rifampicin 416

into persister cells. 417

418

Verapamil did not affect the permeability of persisters to 5-FAM-rifampicin 419

Multidrug efflux pumps are known to contribute to antibiotic tolerance in M. 420

tuberculosis (38-40). It was reported that after incubation of rifampicin-exposed M. 421

tuberculosis cells in in vitro cultures and in infected macrophages with the efflux pump 422

inhibitor, verapamil, rifampicin levels inside the cells increased thereby enhancing 423

susceptibility (41, 42). This study showed that verapamil-sensitive efflux pump was 424

involved in the removal of rifampicin from the cells. With this background information, 425

it was of interest to find out whether efflux pumps were involved in the tolerance of M. 426

tuberculosis persister cells to rifampicin in vitro. 427

For this purpose, we exposed rifampicin persister cells to 50 µg/ml of verapamil 428

and used for 5-FAM-rifampicin permeability assay. We did not observe any difference 429

in terms of the 5-FAM-rifampicin fluorescence intensity of verapamil-treated and 430

untreated rifampicin persister cell samples (Fig. 5 A, B, C). It ruled out the possibility 431

of rifampicin efflux as a possible mechanism for the rifampicin-tolerance in persister 432

cells in vitro and further confirmed that the TOL plays a substantial role in the reduced 433

rifampicin permeability into M. tuberculosis persister cells. On the contrary, verapamil 434

was found to be an efflux inhibitor in M. tuberculosis infected mice, zebra fish and 435

macrophages (40, 43). This apparent contradiction with these works is probably due 436

to the difference in the physiological status of the cells exposed to verapamil. While 437

the cells were in the persistence phase in our study, the cells were in the infected 438

animal model and macrophages in the other studies (40, 43). 439



https://doi.org/10.1101/624569


19

440

Discussion 441

M. tuberculosis is well known for its drug tolerance and natural resistance to many 442

antibiotics. The unique and complex cell wall features of mycobacteria imposing 443

limited permeability have been attributed to be the reason for its natural antibiotic 444

tolerance. Owing to the peripheral location, OL plays a significant role in macrophage 445

interaction, cell adhesion and pathogenicity of M. tuberculosis (44-46). Further, 446

metabolic and detoxifying enzymes, including penicillinase, urease, phosphatases and 447

superoxide dismutase, are present in the OL (30). The presence of immunomodulatory 448

factors in the OL of pathogenic M. tuberculosis plays a major role in the initial host 449

immune response (47). Previous reports have shown that the OL reduces bacterial 450

interaction with the macrophage in the absence of serum opsonins thereby possess 451

anti-phagocytic activity in M. tuberculosis (16) and in Gram-negative bacteria (31, 48). 452

Also, it has been reported that the capsular polysaccharides prevent the entry of 453

potentially harmful host-derived macromolecules like bactericidal peptides in Gram-454

negative bacteria (49-51). Thus, in addition to the contribution of OL to various 455

physiological aspects of mycobacteria (16, 29; 44-47), the present findings show yet 456

another role for OL in the rifampicin tolerance by persisters. 457

The strategy seemed to be to increase the levels of polysaccharide components 458

of OL significantly and thereby confer higher negative charge to restrict entry of 459

rifampicin. The negative surface charge density was found varying, probably 460

depending upon the extent of casing of the cells by the OL. The gradual increase in 461

the surface charge density of the persister cells might be due to the accumulation of 462

OL components over the cell surface during rifampicin exposure. The decrease in the 463

negative surface charge density during regrowth phase indicated the specific role of 464



https://doi.org/10.1101/624569


20

OL thickening during persistence phase as a survival strategy during antibiotic 465

exposure which is no more present on (required for) the bacteria as they have gained 466

resistance and entered regrowth phase. Moreover, the multiple peaks of ZP values in 467

the persister and regrowth phase cells denote the presence of heterogeneous sub-468

populations with different surface charge density depending on the extent of OL 469

thickening of the cells. 470

It is possible that the persisters are getting benefited by the ‘barrier’ effect of TOL 471

in limiting the entry of antibiotics. Restricted antibiotic uptake through outer membrane 472

modification is a known mechanism of rifampicin resistance in the actively growing M. 473

intracellulare and N. meningitidis (33, 36). Previous study has reported that the use of 474

Tween 80 could improve the permeability of rifampicin in M. intracellulare in the growth 475

medium as it is known to reduce permeability barriers in mycobacteria (36) However, 476

in all our experiments we used 0.05% of Tween-80 containing 7H9 broth. Though it is 477

likely that the degradation and assimilation of Tween-80 by the mycobacterial cells 478

over the course of growth and drug exposure (52) could affect the effective 479

concentration of the Tween-80 and thereby altered permeability, the differential OL 480

thickness between the persisters and 12 day old rifampicin unexposed control cells 481

exclude this possibility (see Fig. S1). Since the surface charge of persister cells was 482

negative, the polar and charged nature of the OL need to be considered for its effect 483

on the permeability of a nonpolar antibiotic such as rifampicin. The higher negative 484

charge (polar nature) may be expected to reduce the permeability of a more non-polar 485

type of molecule, such as 5-FAM-rifampicin. However, an integrated study using 486

Wayne’s in vitro hypoxia model at pH 5.8 has shown that while many lipophilic drugs 487

(rifampicin, rifapentine, bedaquiline, clofazimine, nitrazoxamide) could reduce cfu of 488

hypoxic cells by ≥2-log10, many hydrophilic drugs (metronidazole, moxifloxacin, 489



https://doi.org/10.1101/624569


21

pyrazinamide, ethambutol, isoniazid, meropenem) could not effectively reduce the cfu 490

of hypoxic cells (53). Therefore, the polar nature of the TOL cannot be taken as the 491

sole reason for the restricted entry of non-polar (lipophilic) rifampicin. A combination 492

of increased physical thickness and negative charge may be contributing to the 493

restricted entry of rifampicin into M. tuberculosis persister cells, thereby helping 494

survival and subsequent acquisition of drug resistance. 495

496

Acknowledgements 497

PA dedicates this work as a tribute to Prof. T. Ramakrishnan (late), who led the 498

pioneering and foundation-laying work on the biochemistry and molecular biology of 499

Mycobacterium tuberculosis at Indian Institute of Science, Bangalore. 500

501

Funding 502

The work was supported by funds from the DBT-IISc partnership programme and IISc. 503

Authors acknowledge DBT-supported FACS facility in the Biological Sciences 504

Division, and the infrastructure facilities supported by DST-FIST, UGC-CAS, ICMR-505

CAS, and IISc, in the MCB Dep't. JS and SS received SRF from DBT and CSIR, 506

respectively. 507

508

Conflicts of interest statement 509

None declared. 510

511

Author Contributions 512

PA, JS conceived/designed expts; JS, SS performed expts; PA, JS, SS analysed data; 513

PA contributed reagents, materials, and analysis tools; PA, JS wrote the manuscript. 514



https://doi.org/10.1101/624569


22

515

References 516

1. Jankute M, Cox JA, Harrison J, Besra GS. 2015. Assembly of the mycobacterial 517

cell wall. Annu Rev Microbiol 69:405-423. 518

2. Lederer E, Adam A, Ciorbaru R, Petit JF, Wietzerbin J. 1975. Cell walls of 519

mycobacteria and related organisms; chemistry and immunostimulant 520

properties. Mol Cell Biochem 7:87-104. 521

3. Jarlier V, Nikaido H. 1994. Mycobacterial cell wall: structure and role in natural 522

resistance to antibiotics. FEMS Microbiol Lett 123:11-18. 523

4. Cook GM, Berney M, Gebhard S, Heinemann M, Cox RA, Danilchanka O, 524

Niederweis M. 2009. Physiology of mycobacteria. Adv Microb Physiol 55:81-525

182. 526

5. Nguyen L. 2016. Antibiotic resistance mechanisms in M. tuberculosis: an 527

update. Arch Toxicol 90:1585-1604. 528

6. Larrouy-Maumus G, Marino LB, Madduri AV, Ragan TJ, Hunt DM, Bassano L, 529

Gutierrez MG, Moody DB, Pavan FR, de Carvalho LP. 2016. Cell-envelope 530

remodeling as a determinant of phenotypic antibacterial tolerance in 531

Mycobacterium tuberculosis. ACS Infect Dis 2:352-360. 532

7. Bacon J, Alderwick LJ, Allnutt JA, Gabasova E, Watson R, Hatch KA, Clark SO, 533

Jeeves RE, Marriott A, Rayner E, Tolley H, Pearson G, Hall G, Besra GS, 534

Wernisch L, Williams A, Marsh PD. 2014. Non-replicating Mycobacterium 535



https://doi.org/10.1101/624569


23

tuberculosis elicits a reduced infectivity profile with corresponding modifications 536

to the cell wall and extracellular matrix. PLoS One 6:e87329. 537

8. Cunningham AF, Spreadbury CL. 1998. Mycobacterial stationary phase 538

induced by low oxygen tension: cell wall thickening and localization of the 16-539

kilodalton alpha-crystallin homolog. J Bacteriol 180:801-808. 540

9. Höner zu Bentrup K, Russell DG. 2001. Mycobacterial persistence: adaptation 541

to a changing environment. Trends Microbiol 9:597-605. 542

10. Ehrt S, Schnappinger D, Rhee KY. 2018. Metabolic principles of persistence 543

and pathogenicity in Mycobacterium tuberculosis. Nat Rev Microbiol 16:496-544

507. 545

11. Kohanski MA, DePristo MA, Collins JJ. 2010. Sublethal antibiotic treatment 546

leads to multidrug resistance via radical-induced mutagenesis. Mol Cell 37:311-547

320. 548

12. Sebastian J, Swaminath S, Nair RR, Jakkala K, Pradhan A, Ajitkumar P. 2017. 549

De Novo emergence of genetically resistant mutants of Mycobacterium 550

tuberculosis from the persistence phase cells formed against antituberculosis 551

drugs in vitro. Antimicrob Agents Chemother 61:e01343-16. 552

13. Takade A, Takeya K, Taniguchi H, Mizuguchi Y. 1983. Electron microscopic 553

observations of cell division in Mycobacterium vaccae V1. J Gen Microbiol 554

129:2315-2320. 555



https://doi.org/10.1101/624569


24

14. Rastogi N, Frehel C, David HL. 1986. Triple-layered structure of mycobacterial 556

cell wall: evidence for the existence of a polysaccharide-rich outer layer in 18 557

mycobacterial species. Curr Microbiol 13:237-242. 558

15. Rosenberg M, Gutnick D, Rosenberg E. 1980. Adherence of bacteria to 559

hydrocarbons: a simple method for measuring cell-surface hydrophobicity. 560

FEMS Microbiol Lett 9:29-33. 561

16. Stokes RW, Norris-Jones R, Brooks DE, Beveridge TJ, Doxsee D, Thorson LM. 562

2004. The glycan-rich outer layer of the cell wall of Mycobacterium tuberculosis 563

acts as an antiphagocytic capsule limiting the association of the bacterium with 564

macrophages. Infect Immun 72:5676-5686. 565

17. Wilson WW, Wade MM, Holman SC, Champlin FR. 2001. Status of methods 566

for assessing bacterial cell surface charge properties based on zeta potential 567

measurements. J Microbiol Methods 43:153-164. 568

18. Ayala-Torres C, Hernández N, Galeano A, Novoa-Aponte L, Soto CY. 2014. 569

Zeta potential as a measure of the surface charge of mycobacterial cells. Ann 570

Microbiol 64:1189-1195. 571

19. Ortalo-Magne A, Dupont MA, Lemassu A, Andersen AB, Gounon P, Daffé M. 572

1995. Molecular composition of the outermost capsular material of the tubercle 573

bacillus. Microbiology 141:1609-1620. 574

20. Dickinson JM, Aber VR, Allen BW, Ellard GA, Mitchison DA. 1974. Assay of 575

rifampicin in serum. J Clin Pathol 27:457-62. 576



https://doi.org/10.1101/624569


25

21. Takade A, Umeda A, Matsuoka M, Yoshida S, Nakamura M, Amako K. 2003. 577

Comparative studies of the cell structures of Mycobacterium leprae and M. 578

tuberculosis using the electron microscopy freeze-substitution technique. 579

Microbiol Immunol 47:265-270. 580

22. Vijay S, Anand D, Ajitkumar P. 2012. Unveiling unique features of formation of 581

septal partition and constriction in mycobacteria—an ultrastructural study. J 582

Bacteriol 194:702-707. 583

23. Wood PJ. 1980. Specificity in the interaction of direct dyes with 584

polysaccharides. Carb Res 85:271-287. 585

24. Rasconi S, Jobard M, Jouve L, Sime-Ngando T. 2009. Use of calcofluor white 586

for detection, identification, and quantification of phytoplanktonic fungal 587

parasites. Appl Environ Microbiol 75:2545-2553. 588

25. Argüelles JC. 2000. Physiological roles of trehalose in bacteria and yeasts: a 589

comparative analysis. Arch Microbiol 174: 217-224. 590

26. Elbein AD, Pan YT, Pastuszak I, Carroll D. 2003. New insights on trehalose: a 591

multifunctional molecule. Glycobiology 13:17R-27R. 592

27. Chen Q, Haddad GG. 2004. Role of trehalose phosphate synthase and 593

trehalose during hypoxia: from flies to mammals. J Exp Biol 207:3125-3129. 594

28. Schwebach JR, Glatman-Freedman A, Gunther-Cummins L, Dai Z, Robbins 595

JB, Schneerson R, Casadevall A. 2002. Glucan is a component of the 596

Mycobacterium tuberculosis surface that is expressed in vitro and in vivo. Infect 597

Immun 70: 2566-2575. 598



https://doi.org/10.1101/624569


26

29. Lemassu, A, Daffé M. 1994. Structural features of the exocellular 599

polysaccharides of Mycobacterium tuberculosis. Biochem J 297:351-357. 600

30. Daffé M, Etienne G. 1999. The capsule of Mycobacterium tuberculosis and its 601

implications for pathogenicity. Tuber Lung Dis 79:153-169. 602

31. Llobet E, Tomás JM, Bengoechea JA. 2008. Capsule polysaccharide is a 603

bacterial decoy for antimicrobial peptides. Microbiology 154:3877-3886. 604

32. Moxon ER, Kroll JS. 1990. The role of bacterial polysaccharide capsules as 605

virulence factors. Curr Top Microbiol Immunol 150:65-85. 606

33. Abadi FJ, Carter PE, Cash P, Pennington TH. 1996. Rifampin resistance in 607

Neisseria meningitidis due to alterations in membrane permeability. Antimicrob 608

Agents Chemother 40:646-651. 609

34. Cui L, Ma X, Sato K, Okuma K, Tenover FC, Mamizuka EM, Gemmell CG, Kim 610

MN, Ploy MC, El-Solh N, Ferraz V, Hiramatsu K. 2003. Cell wall thickening is a 611

common feature of vancomycin resistance in Staphylococcus aureus. J Clin 612

Microbiol 41:5-14. 613

35. Cui L, Iwamoto A, Lian JQ, Neoh HM, Maruyama T, Horikawa Y, Hiramatsu K. 614

2006. Novel mechanism of antibiotic resistance originating in vancomycin-615

intermediate Staphylococcus aureus. Antimicrob Agents Chemother 50:428-616

438. 617

36. Hui J, Gordon N, Kajioka R. 1977. Permeability barrier to rifampin in 618

mycobacteria. Antimicrob Agents Chemother 11:773-779. 619



https://doi.org/10.1101/624569


27

37. Piddock LJ, Williams KJ, Ricci V. 2000. Accumulation of rifampicin by 620

Mycobacterium aurum, Mycobacterium smegmatis and Mycobacterium 621

tuberculosis. J Antimicrob Chemother 45:159-165. 622

38. Choudhuri BS, Bhakta S, Barik R, Basu J, Kundu M, Chakrabarti P. 2002. 623

Overexpression and functional characterisation of an ABC (ATP-binding 624

cassette) transporter encoded by the genes drrA and drrB of Mycobacterium 625

tuberculosis. Biochem J 367:279-285. 626

39. Pasca MR, Guglierame P, Arcesi F, Bellinzoni M, De Rossi E, Riccardi G. 2004. 627

Rv2686c-Rv2687c-Rv2688c, an ABC fluoroquinolone efflux pump in 628

Mycobacterium tuberculosis. Antimicrob Agents Chemother 48:3175-3178. 629

40. Gupta S, Tyagi S, Almeida DV, Maiga MC, Ammerman NC, Bishai WR. 2013. 630

Acceleration of tuberculosis treatment by adjunctive therapy with verapamil as 631

an efflux inhibitor. Am J Respir Crit Care Med 188:600-607. 632

41. Demitto Fde O, do Amaral RC, Maltempe FG, Siqueira VL, Scodro RB, Lopes 633

MA, Caleffi-Ferracioli KR, Canezin PH, Cardoso RF. 2015. In vitro activity of 634

rifampicin and verapamil combination in multidrug-resistant Mycobacterium 635

tuberculosis. PLoS One 10:e0116545. 636

42. Chen C, Gardete S, Jansen RS, Shetty A, Dick T, Rhee KY, Dartois V. 2018. 637

Verapamil targets membrane energetics in Mycobacterium tuberculosis. 638

Antimicrob Agents Chemother 62:e02107-e02117. 639

43. Adams KN, Takaki K, Connolly LE, Wiedenhoft H, Winglee K, Humbert O, 640

Edelstein PH, Cosma CL, Ramakrishnan L. 2011. Drug tolerance in replicating 641



https://doi.org/10.1101/624569


28

mycobacteria mediated by a macrophage-induced efflux mechanism. Cell 642

145:39-53. 643

44. Armstrong JA, Hart PD. 1971. Response of cultured macrophages to 644

Mycobacterium tuberculosis, with observations on fusion of lysosomes with 645

phagosomes. J Exp Med 134:713-740. 646

45. Saito R, Tanaka A, Sugiyama K, Azuma I, Yamamura Y. 1976. Adjuvant effect 647

of cord factor, a mycobacterial lipid. Infect Immun 13:776-781. 648

46. Ragas A, Roussel L, Puzo G, Rivière M. 2007. The Mycobacterium tuberculosis 649

cell-surface glycoprotein apa as a potential adhesin to colonize target cells via 650

the innate immune system pulmonary C-type lectin surfactant protein A. J Biol 651

Chem 282:5133-5142. 652

47. Geurtsen J, Chedammi S, Mesters J, Cot M, Driessen NN, Sambou T, Kakutani 653

R, Ummels R, Maaskant J, Takata H, Baba O, Terashima T, Bovin N, 654

Vandenbroucke-Grauls CM, Nigou J, Puzo G, Lemassu A, Daffé M, Appelmelk 655

BJ. 2009. Identification of mycobacterial alpha-glucan as a novel ligand for DC-656

SIGN: involvement of mycobacterial capsular polysaccharides in host immune 657

modulation. J Immunol 183:5221-5231. 658

48. Domenico P, Salo RJ, Cross AS, Cunha BA. 1994. Polysaccharide capsule-659

mediated resistance to opsonophagocytosis in Klebsiella pneumoniae. Infect 660

Immun 62:4495-4499. 661

49. Costerton JW, Cheng KJ. 1975. The role of the bacterial cell envelope in 662

antibiotic resistance. J Antimicrob Chemother 1:363-377. 663



https://doi.org/10.1101/624569


29

50. Slack MP, Nichols WW. 1982. Antibiotic penetration through bacterial capsules 664

and exopolysaccharides. J Antimicrob Chemother 10:368-372. 665

51. Campos MA, Vargas MA, Regueiro V, Llompart CM, Alberti S, Bengoechea JA. 666

2004. Capsule polysaccharide mediates bacterial resistance to antimicrobial 667

peptides. Infect Immun 72:7107-7114. 668

52. Schaefer WB, Lewis CW Jr. 1965. Effect of oleic acid on growth and cell 669

structure of mycobacteria. J Bacteriol 90:1438-1447. 670

53. Piccaro G, Poce G, Biava M, Giannoni F, Fattorini L. 2015. Activity of lipophilic 671

and hydrophilic drugs against dormant and replicating Mycobacterium 672

tuberculosis. J Antibiot (Tokyo) 68:711-714. 673

674

Figure Legends 675

FIG 1. Presence of TOL in rifampicin persister cells. Transmission electron micrograph 676

of M. tuberculosis cells showing differential thickening of OL in persister cell (A) and 677

in control MLP cells (B). Magnified images are shown right side, the yellow arrows 678

indicate OL. Thickness (in nm) of OL (C) and ETL (D) between persister and MLP 679

cells. (E) CFW staining of OL in persister cells before and after bead beating (BB). 680

Scale, 2 µm. Density plots of the flow cytometry analysed CFW stained MLP cells (F) 681

and persister cells (G) before and after BB. AF, autofluorescence. Histogram overlay 682

of the flow cytometry analysed CFW stained MLP cells (H) and persister cells (I) before 683

and after BB. (J) Bar graph showing normalised median fluorescence intensity of CFW 684

in MLP and persister cells before and after BB. One asterisk (*) indicates P value less 685

than or equal to 0.05 (P ≤ 0.05). Three asterisks (***) indicate P value lesser than 686



https://doi.org/10.1101/624569


30

0.001 (P < 0.001). The statistical significance was calculated using two-tailed paired 687

t-test. 688

689

FIG 2. Physicochemical properties of rifampicin-exposed M. tuberculosis persister 690

cells. (A) Hexadecane assay showing the proportions of hydrophilic rifampicin 691

persister cells and MLP cells. (B) ZP of cells from different phases of rifampicin 692

exposed of M. tuberculosis cells and its corresponding (C) histogram overlay. ZP of 693

TOL as a function of pH showing the isoelectric point of the TOL between pH 4.0 and 694

pH 5.0 for: (D) MLP cells, (E) persister cells, and (F) cells from killing phase. One 695

asterisk (*) indicates P value less than or equal to 0.05 (P ≤ 0.05). Two asterisks (**) 696

indicate P value less than or equal to 0.01 (P ≤ 0.01). The statistical significance was 697

calculated using two-tailed paired t-test. 698

699

FIG. 3. Characterisation of 5-FAM-rifampicin permeability into MLP and rifampicin 700

persister cells. (A) The possible structural isomers of 5-FAM-rifampicin conjugates. 701

The carboxyl group of the 5-FAM can form an ester bond with any one of the hydroxyl 702

groups in the aliphatic ring of rifampicin, generating two possible isomers. Ester bond 703

is shown in red circle. (B) Bar graph showing the relative bactericidal activity of 704

rifampicin and 5-FAM-rifampicin using Staphylococcus aureus by agar diffusion assay. 705

(C) Density plots and (E) histogram overlay of MLP cells incubated with increasing 706

concentrations of 5-FAM-rifampicin by flow cytometry. (E) Standard graph showing the 707

5-FAM-rifampicin permeability into MLP cells at two-fold increasing concentrations. 708

709

FIG 4. Permeability of M. tuberculosis to 5-FAM-rifampicin. Fluorescence microscopy 710

images of M. tuberculosis rifampicin persister cells showing differential entry of 5-FAM-711



https://doi.org/10.1101/624569


31

rifampicn in persisters (A) and MLP cells (B) before and after the removal of OL by 712

bead beating (BB). (Scale, 2 µm). Line graph from flow cytometry analysis for the time-713

dependent entry of 5-FAM-rifampicin into MLP (C) and rifampicin persisters (D) with 714

or without bead beating (BB) and its corresponding density plots (E) and histogram 715

overlay (F). (AF, Autofluorescence). 716

717

FIG 5. Estimation of the relative permeability of 5-FAM-rifampicin into rifampicin 718

persisters in presence of efflux pump inhibitor verapamil. (A) Density plots of 5-FAM-719

rifampicin entry in M. tuberculosis cells with or without verapamil exposure. (B) Bar 720

graph showing relative quantitation of the 5-FAM-rifampicin entry between MLP cells 721

and rifampicin persisters with and without verapamil (VM) treatment. (C) Histogram 722

overlay of the flow cytometry profile. 723



https://doi.org/10.1101/624569


Table 1 Relative abundance of TOLa compounds in rifampicin persister cells with

respect to MLPb cells as detected by GC-MS

___________________________________________________________________

Compound Fold increase in Persister cells

(w.r.t. MLP cells)

___________________________________________________________________

α-D-Glucopyranoside 6.67

1, 2, 5-linked-Mannitol 4.97

3, 4-linked Mannose 2.22

Hexa-acetyl-Mannitol 1.84

1, 2, 3-Propanetriol 1.83

Methyl 1, 2, 3, 4-Tetrahydronaphthalene-2-Carboxylate 1.76

1, 3-Di-Iso-Propylnaphthalene 1.73

1, 2, 4-linked arabinitol 1.68

D- (1, 2-linked mannitol) 1.58


5-linked Galactonitrile 1.30

Galactose pyranoside 1.11


α- D-Mannopyranoside 1.08

___________________________________________________________________

aThick Outer Layer; bMid-log phase



https://doi.org/10.1101/624569


FIG 1. Presence of TOL in rifampicin persister cells. Transmission electron micrograph of M.

tuberculosis cells showing differential thickening of OL in persister cell (A) and in control MLP cell

(B). Magnified images are shown right side, the yellow arrows indicate OL. Thickness (in nm) of

OL (C) and ETL (D) between persister and MLP cells. (E) CFW staining of OL in persister cells

before and after bead beating (BB). Scale, 2 µm. Density plots of the flow cytometry analysed

CFW stained MLP cells (F) and persister cells (G) before and after BB. AF, autofluorescence.

Histogram overlay of the flow cytometry analysed CFW stained MLP cells (H) and persister cells

(I) before and after BB. (J) Bar graph showing normalised median fluorescence intensity of CFW

in MLP and persister cells before and after BB. One asterisk (*) indicates P value less than or

equal to 0.05 (P ≤ 0.05). Three asterisks (***) indicate P value lesser than 0.001 (P < 0.001). The

statistical significance was calculated using two-tailed paired t-test.

0

50

100

150C

*

OL

Th

ickn

ess (

nm

)

Persisters Control

0

50

100

150

D

ET

L T

hic

kn

ess (

nm

)

Persisters Control

MLP AF MLP + CW

No

BB

BB

FPer AF Per + CW

G

H No BB BB

ML

P

Pers

iste

r

I No BB BB

Pers

iste

rC

on

tro

lA

B

JE



https://doi.org/10.1101/624569


MLP

Killing

Per

RG

CBA

2 4 6 8 10

-30

-20

-10

0

10

pH

Zeta

Pote

nti

al

(mV

)

ED

2 4 6 8 10

-3

-2

-1

0

1

pH

Zeta

Pote

nti

al

(mV

)

F

2 4 6 8 10

-15

-10

-5

0

5

pH

Zeta

Po

ten

tial (m

V)

FIG 2. Physicochemical properties of rifampicin-exposed M. tuberculosis persister cells. (A)

Hexadecane assay showing the proportions of hydrophilic rifampicin persister cells and

MLP cells. (B) ZP of cells from different phases of rifampicin exposed of M. tuberculosis

cells and its corresponding histogram overlay (C). ZP of TOL as a function of pH showing

the isoelectric point of the TOL between pH 4.0 and pH 5.0 for: (D) MLP cells, (E) persister

cells, and (F) cells from killing phase. One asterisk (*) indicates P value less than or equal

to 0.05 (P ≤ 0.05). Two asterisks (**) indicate P value less than or equal to 0.01 (P ≤ 0.01).

The statistical significance was calculated using two-tailed paired t-test.



https://doi.org/10.1101/624569


A

O

O

O

N

O

O

OH

O

O

OH OHCH3

OH

NN

N

O

O

O

CH3

NH

OCH3

CH3CH3

OH

OCH3

O

O

CH3

CH3

H

CH3

CH3

O

5-FAM-RIF Isomer 1

OH OHCH3

OH

NN

N

O

O

O

CH3

NH

OCH3

CH3CH3

O

OCH3

O

O

CH3

CH3

H

CH3

CH3

OH

O

O

O N

O

O

OH

O

O

5-FAM-RIF Isomer 2

0.0 0.5 1.0 1.5 2.0 2.50

500

1000

1500

2000

2500

Conc. of 5-FAM-RIF (g/ml)

5-F

AM

-RIF

(N

or.

A.U

)

E

B

D

Rifampicin 5-FAM-rifampicin0

5

10

60

80

100

% B

acteric

ida

l a

ctiv

ity

0.0156 µg/ml 0.03125 µg/ml 0.0625 µg/ml 0.125 µg/ml

0.25 µg/ml 0.5 µg/ml 1.0 µg/ml 2.0 µg/ml

Mid-log

C

FIG 3. Characterisation of 5-FAM-rifampicin permeability into MLP and rifampicin persister

cells. (A) The possible structural isomers of 5-FAM-rifampicin conjugates. The carboxyl group

of the 5-FAM can form an ester bond with any one of the hydroxyl groups in the aliphatic ring

of rifampicin, generating two possible isomers. Ester bond is shown in red circle. (B) Bar

graph showing the relative bactericidal activity of rifampicin and 5-FAM-rifampicin using

Staphylococcus aureus by agar diffusion assay. (C) Density plots and (E) histogram overlay of

MLP cells incubated with increasing concentrations of 5-FAM-rifampicin by flow cytometry. (E)

Standard graph showing the 5-FAM-rifampicin permeability into MLP cells at two-fold

increasing concentrations.



https://doi.org/10.1101/624569


C

0 15 30 45 600

100

200

300 MLP

MLP-BB

Time (min)

5-F

AM

Med

ian

Flu

o.

(Nor.

) D

15 30 45 60

-100

0

100

200

300Per

Per-BB

Time (min)

5-F

AM

Med

ian

Flu

o. (N

or.)

MLP BBMLP

0 min

15 min

30 min

45 min

60 min

▬▬

▬

▬

▬

0 min

15 min

30 min

45 min

60 min

▬▬

▬

▬

▬

G Per Per BB

0 min

15 min

30 min

45 min

60 min

▬▬

▬

▬

▬

0 min

15 min

30 min

45 min

60 min

▬

▬

▬

▬

▬

H

Per AF Per+ 5-FAM-RIFFMLP AF MLP+ 5-FAM-RIF

No

BB

BB

E

DIC 5-FAM-RIF MERGEDP

ersi

ster

No B

B

ADIC 5-FAM-RIF MERGED

B

DIC 5-FAM-RIF MERGEDDIC 5-FAM-RIF MERGED

Per

sist

er

BB

ML

P

No B

B

ML

P

BB

FIG 4. Permeability of M. tuberculosis to 5-FAM-rifampicin. Fluorescence microscopy images

of M. tuberculosis rifampicin persister cells showing differential entry of 5-FAM-rifampicn in

persisters (A) and MLP cells (B) before and after the removal of OL by bead beating (BB).

(Scale, 2 µm). Line graph from flow cytometry analysis for the time-dependent entry of 5-

FAM-rifampicin into MLP (C) and rifampicin persisters (D) with or without bead beating (BB)

and its corresponding density plots (E) and histogram overlay (F). (AF, Autofluorescence).



https://doi.org/10.1101/624569


5-FAM-RIF

Pers

iste

r

Control

ML

PP

ers

iste

r+V

era

pam

ilA B

MLP Per Per+ VM0.0

0.5

1.0

1.5

Rel.

5-F

AM

-Con

c. (

g/m

l)

C

FIG 5. Estimation of the relative permeability of 5-FAM-rifampicin into rifampicin persisters in

presence of efflux pump inhibitor verapamil. (A) Density plots of 5-FAM-rifampicin entry in M.

tuberculosis cells with or without verapamil exposure. (B) Bar graph showing relative

quantitation of the 5-FAM-rifampicin entry between MLP cells and rifampicin persisters with

and without verapamil (VM) treatment. (C) Histogram overlay of the flow cytometry profile.



https://doi.org/10.1101/624569


Documents

Antibiotic Persistence in Mycobacterium tuberculosis · 42 Mycobacterium tuberculosis, which is the causative agent of tuberculosis, is one of the 43 most successful human pathogens