1 3

Arch Microbiol (2014) 196:157–168DOI 10.1007/s00203-014-0953-7

ORIGINAL PAPER

Asymmetric cell division in Mycobacterium tuberculosis and its unique features

Srinivasan Vijay · Mukkayyan Nagaraja · Jees Sebastian · Parthasarathi Ajitkumar

Received: 1 August 2013 / Revised: 6 November 2013 / Accepted: 18 January 2014 / Published online: 31 January 2014 © Springer-Verlag Berlin Heidelberg 2014

tuberculosis patients’ sputum samples, which are known for the prevalence of oxidative stress conditions, also con-tained short cells at the same proportion as that in the mid-log phase population. The probable physiological signifi-cance of the generation of the short cells through unusually deviated asymmetric cell division is discussed.

Keywords Asymmetric septum · Asymmetric cell division · Mycobacterium tuberculosis · Short cell · Nucleoid · Symmetric cell division

Introduction

Symmetric binary fission is the hallmark of rod-shaped bacteria to generate two equally sized daughter cells. For instance, symmetric binary fission with the septum pre-cisely placed at the mid-cell site occurs in Escherichia coli (Yu and Margolin 1999) and Bacillus subtilis (Migocki et al. 2004). Both the pathogenic and the non-pathogenic mycobacterial species also have been found to divide through symmetric binary fission (see Hett and Rubin 2008 for a review on the various aspects of mycobacterial cell division). Transmission and scanning electron microscopy (TEM and SEM) studies performed several decades ago had shown how the equally sized daughter cells physi-cally separate at the completion of the septation at the mid-cell site in Mycobacterium leprae (Edwards 1970; Hirata 1978), Mycobacterium avium (McCarthy 1974; Rastogi and David 1981), and Mycobacterium vaccae (Takade et al. 1983). Recent closer analyses of Mycobacterium smegma-tis, Mycobacterium marinum, and Mycobacterium bovis BCG cell division, using PBP1a-mCherry (Joyce et al. 2012), FtsZ-mCherry (Singh et al. 2013), and Wag31-GFP (Santi et al. 2013) localization, combined with septum and

Abstract Recently, several reports showed that about 80 % of mid-log phase Mycobacterium smegmatis, Myco-bacterium marinum, and Mycobacterium bovis BCG cells divide symmetrically with 5–10 % deviation in the sep-tum position from the median. However, the mode of cell division of the pathogenic mycobacterial species, Myco-bacterium tuberculosis, remained unclear. Therefore, in the present study, using electron microscopy, fluorescence microscopy of septum- and nucleoid-stained live and fixed cells, and live cell time-lapse imaging, we show the occur-rence of asymmetric cell division with unusually deviated septum/constriction in 20 % of the 15 % septating M. tuber-culosis cells in the mid-log phase population. The remain-ing 80 % of the 15 % septating cells divided symmetri-cally but with 2–5 % deviation in the septum/constriction position, as reported for M. smegmatis, M. marinum, and M. bovis BCG cells. Both the long and the short portions of the asymmetrically dividing M. tuberculosis cells with unu-sually deviated septum contained nucleoids, thereby gen-erating viable short and long cells from each asymmetric division. M. tuberculosis short cells were acid fast positive and, like the long cells, further readily underwent growth and division to generate micro-colony, thereby show-ing that they were neither mini cells, spores nor dormant forms of mycobacteria. The freshly diagnosed pulmonary

Communicated by Erko Stackebrandt.

Electronic supplementary material The online version of this article (doi:10.1007/s00203-014-0953-7) contains supplementary material, which is available to authorized users.

S. Vijay · M. Nagaraja · J. Sebastian · P. Ajitkumar (*) Department of Microbiology and Cell Biology, Indian Institute of Science, Bangalore 560012, Karnataka, Indiae-mail: ajit@mcbl.iisc.ernet.in

158 Arch Microbiol (2014) 196:157–168

1 3

membrane staining (Joyce et al. 2012; Santi et al. 2013; Singh et al. 2013), showed 5–10 % deviation in the posi-tion of the septum from the mid-cell site in about 80 % of the symmetrically dividing cells, thereby generating slightly unequally sized daughter cells (Joyce et al. 2012; Santi et al. 2013; Singh et al. 2013). One of these studies showed that although septum was always placed symmetri-cally with minor deviation (with in 10 %) from the mid-cell site (Joyce et al. 2012), in M. smegmatis and M. bovis BCG cells, differential polar growth post placement of septum leads to more asymmetry in cell division. Another study showed that the division site placement in M. mari-num and M. smegmatis occurs at non-medial and medial sites at equivalent proportions. However, compensatory mechanisms operate in non-medially septated cells to gen-erate predominantly equal-sized daughter cells (Singh et al. 2013). A very recent study found about 5–10 % deviation in the placement of septum, in the majority of M. smegmatis cells, which generated differently sized daughter cells that grew at different velocities (Santi et al. 2013). An apparent asymmetry was also reported among M. smegmatis daugh-ter cells due to the deterministic difference in the elonga-tion rates of the daughter cells, post-symmetric placement of the septum, and division (Aldridge et al. 2012). Never-theless, in all these studies, the mode of cell division of the cells in the remaining lower proportion of the three myco-bacterial species in the mid-log phase population remained unknown.

Very little information is available on the mode of cell division in the slow-growing human pathogenic species, Mycobacterium tuberculosis. An ultrastructural study con-ducted almost three decades ago showed the presence of M. tuberculosis cells with asymmetric septum, although it went unreported (see Fig. 1b in Barksdale and Kim 1977). Similarly, in a scanning electron microscopy study per-formed a decade ago, the ridges on the surface of the M. tuberculosis cells closer to the poles were speculated to be either due to the scar of the previous division or due to asymmetric septum beneath, with daughter cells break-ing off from one pole of the elongated parental cell (Dahl 2004). In a recent study, different proportions of asymmet-rically dividing tubercle bacilli were found in the exponen-tial phase cultures of drug-susceptible, multidrug-resistant (MDR), extensively drug-resistant (XDR), and extremely drug-resistant (XXDR) clinical isolates (Farnia et al. 2010). Nevertheless, the mode of division of M. tuberculosis cells still remained unclear. Therefore, in the present study, the mode of cell division of exponential phase cultures of M. tuberculosis was determined. Initially, to find out the posi-tion of the septum in the cells, a large number of exponen-tial phase M. tuberculosis cells was screened, using TEM of longitudinal ultrathin sections of fixed cells and fluores-cence microscopy of septum-stained and nucleoid-stained

live and fixed cells. Subsequently, using live cell time-lapse imaging, the cell division process of M. tuberculosis cells was monitored for 2–3 generations, to determine the posi-tion of the division constriction and the mode of division of the progeny cells. These experiments revealed that low proportion of M. tuberculosis cells in the population divide with unusually deviated asymmetric septum/constriction. On the contrary, major proportion of the cells divides with minor deviation in the septum position, as reported for the saprophytic mycobacterial species (Joyce et al. 2012; Santi et al. 2013; Singh et al. 2013). The short cells, which were generated by the unusually deviated asymmetric division, were also found in the freshly diagnosed pulmonary tuber-culosis patients’ sputum samples, which are known for the prevalence of oxidative stress conditions. The mode of divi-sion, and the ultrastructural and cellular features, of short cells were characterized and the significance of the genera-tion of short cells through such unusually deviated asym-metric division was discussed.

Materials and methods

Bacterial strains and growth conditions

Mycobacterium tuberculosis H37Ra cells (obtained from Central JALMA Institute for Leprosy and Other Myco-bacterial Diseases, Agra, India) were grown in Middle-brook 7H9 broth, with albumin-dextrose supplement, and with or without 0.05 % Tween 80, at 37 °C with shaking at 170 rpm, till OD600 nm of the culture reached 0.60 (mid-log phase). The avirulent strain, M. tuberculosis H37Ra, was used in the study, instead of the virulent H37Rv strain, for the following reasons: (1) all the cell division genes are present in the H37Ra strain, like in H37Rv strain (Zheng et al. 2008); (2) both H37Ra and H37Rv divide with same generation time (Hiriyanna and Ramakrishnan 1986); and (3) for want of biosafety facility.

Imaging with transmission electron microscopy

Mycobacterium tuberculosis cells were processed for TEM, as described (Takade et al. 1983), but with minor modi-fications (Vijay et al. 2012). In brief, mid-log phase cells were fixed in 1 % (v/v) osmium tetroxide in 0.15 M caco-dylate buffer, pH 7.2, for 1 h at room temperature. They were washed once with the same buffer and fixed in 0.15 M sodium cacodylate buffer, pH 7.2, containing 2 % (w/v) tannic acid and 2 % (v/v) glutaraldehyde for 2 h at room temperature. The cells were then washed once with the same buffer, re-fixed in 1 % (v/v) osmium tetroxide over-night at 4 °C, dehydrated in graded series (20, 30, 50, and 75 %) of 95 % ethanol, and embedded in LR white resin.

159Arch Microbiol (2014) 196:157–168

1 3

Ultrathin sections were generated with glass knife, stained with 0.5 % uranyl acetate and 0.04 % lead citrate, and observed under JEOL JEM 100 CX II electron microscope, at 80 kV. The cell length measurements were made from the TEM images.

Fluorescence microscopy of septum-, nucleoid-, and membrane-stained live and fixed cells

Septum staining of live cells was carried out using vanco-mycin-BODIPY (VBP) (Daniel and Errington 2003; Reyn-olds 1989; Thanky et al. 2007). VBP (1 μg/ml in PBS) was added to the cells in culture and incubated with shaking at 170 rpm for 24 h at 37 °C. The cells were then adhered to poly-l-lysine-coated slides. For WGA-Alexa488 (2 μg/ml in 1 × PBS) staining (Sizemore et al. 1990), the cells were fixed in 4 % para formaldehyde, adhered to poly-l-lysine-coated multiwell slides, washed with 1 × PBS for 1 min, incubated with lysozyme (2 mg/ml) for 15 min at room tem-perature, washed three times with 1 × PBS, for 1 min each, and stained for 15 min. For staining nucleoid and membrane, 2 μl of the membrane-specific dye, TMA-DPH (40 nM) (Hett et al. 2008), and 1 μl of DNA-specific dye/RNA-specific dye, RiboGreen (1 % final conc) (Weinbauer et al. 1998), were added to the live cells (200 μl of mid-log phase culture cells re-suspended in 100 μl of 1 × PBS) and incu-bated in the dark for 45 min at 37 °C. The cells were then adhered to poly-l-lysine-coated multiwell slides, incubated for further 30 min in the dark and washed once with 50 μl of PBS. The cells in all the samples were observed under oil immersion 100× objective using Zeiss Axio Imager M1 microscope, after mounting them in 90 % glycerol. The fluo-rescence and the respective DIC images were procured.

Imaging with live cell time-lapse microscopy

Live cell time-lapse imaging of M. tuberculosis cell division was performed using agarose pad method, as described (de Jong et al. 2011; Joyce et al. 2011), but with slight modifications. Middlebrook 7H9 medium containing 1.5 % low-melting agarose was used to form agarose pad on glass slide. About 10 μl of mid-log phase M. tuberculo-sis cells was placed on top of the solidified agarose pad, the agarose/cells mixture was spread evenly by tilting the slide, and covered with a cover slip. The agarose pad with the cells was kept at 37 °C for 1 h incubation to enable adhe-sion of the cells. This slide was used for live cell time-lapse imaging with Z-stacking at 37 °C. The M. tuberculosis cells were observed for 2–3 generations. DIC images were taken at every 10 min interval. The cell lengths were deter-mined from the DIC images, using Axio vision 4 software. The tracking of the live cell time-lapse imaging movies was performed using the ImajeJ version 1.43m (Rasband 2012).

Results

Transmission electron microscopy shows M. tuberculosis cells with unusually deviated septum

In the present study, irrespective of the wide range in the lengths of the dividing mother cells, any division with more than 10 % deviation in the septum position from the mid-cell site was considered asymmetric with unusually devi-ated septum, generating a short cell and a long cell. This definition was necessary to distinguish the asymmetric division with unusually deviated (more than 10 %) septum of M. tuberculosis cells from the symmetric division with minor deviation (5–10 %) of the septum in the majority of the population (80 %) of M. smegmatis, M. marinum, and M. bovis BCG cells, reported recently (Joyce et al. 2012; Santi et al. 2013; Singh et al. 2013). Screening of a large number of mid-log phase M. tuberculosis cells (n = 1,000), using TEM, revealed that about 20 % (n = 150 septating cells) of the 15 % septating mid-log phase population pos-sessed septum at unusually deviated position from the mid-cell site, thereby dividing the mother cell into a short por-tion and a long portion (Fig. 1a–d). M. tuberculosis cells, at the initiation or progression of asymmetric septation, could also be seen (Fig. 1e, f, respectively). More images of the cells with asymmetric septum are given in Fig. S1 a and b. This population also included ‘V’-shaped cells, which undergo typical ‘snapping postfission’ mode of mycobac-terial division (Dahl 2004; Thanky et al. 2007) (Fig. S1b). On the contrary, majority (80 %, n = 150 septating cells) of the 15 % septating mid-log phase M. tuberculosis cells possessed symmetric septum, with minor deviation in the position from the mid-cell site, as shown by the ultrastruc-tural analyses (see Fig. 1g, h). More images of the cells with asymmetric and symmetric septum are given in Fig. S1 a & b and c & d, respectively. The presence or absence of Tween 80 in the medium did not affect the proportion of M. tuberculosis subpopulation having cells with unusually deviated asymmetric septum, indicating that culture condi-tions do not affect unusually deviated asymmetric division.

Determination of the position of the asymmetric sep-tum, with respect to the nearest pole, from the TEM images showed that the length between the septum and the near-est pole was 0.73 ± 0.10 μm (n = 26), with the average total length of the asymmetrically septating cells being 2.80 ± 0.70 μm (n = 26) (Table 1). Thus, with the mid-cell position in the asymmetrically septating cells being 1.40 ± 0.35 μm, the deviation in the position of the asym-metric septum from the mid-cell to the nearest pole was 0.67 ± 0.25 μm (i.e., 1.40 ± 0.35 μm minus 0.73 ± 0. 10 μm) (Table 1). This extent of deviation of the asymmet-ric septum amounted to a deviation of 24 ± 4 % (n = 26) of the length, from the median (Fig. 2). The length between

160 Arch Microbiol (2014) 196:157–168

1 3

the septum and either of the poles in the cells with sym-metric septum was 1.31 ± 0.16 μm (n = 24), with the average total length of the symmetrically septating cells being 2.63 ± 0.32 μm (n = 24) (Table 1). Thus, the extent of deviation of the septum in the symmetrically dividing cells was 2.30 ± 2 % (n = 24) of the cell length, from the median (Fig. 2). In spite of the high deviation in the sep-tum position in the asymmetrically septating cells, the aver-age lengths of the cells with symmetric septum with minor deviation and asymmetric septum with major deviation were comparable (Table 1).

Both the short-cell portion and the long-cell portion (n = 150 septating cells) of the asymmetrically septating cells contained nucleoids (Fig. 1). The ultrastructure of the nucleoid in the short-cell portion of the cells showed compact appearance, while that in the long cell portion appeared non-compact (Fig. 1). Similarly, the more or less equal-sized portions of the cells with symmetric septum also contained nucleoids (Fig. 1g, h; Fig. S1 c & d). The ultrastructural profile of the nucleoid in the asymmetrically dividing cells was similar to that of the nucleoids in the

symmetrically septating cells (compare images in Fig. 1, Fig. S1). Thus, TEM imaging revealed that a subpopulation of exponential phase M. tuberculosis cultures contains cells with highly deviated asymmetric septum, which compart-mentalized the mother cell into a short-cell portion and a long-cell portion, both with nucleoids. On the other hand, the majority of the dividing exponential phase population contained symmetric septum with minor deviation in the septum position.

Septum, membrane, and nucleoid staining shows M. tuberculosis cells with unusually deviated septum

M. tuberculosis cells with unusually deviated asymmetric septum could also be found in the mid-log phase live cells (n = 20 septating cells), which were stained for septum with VBP that binds only to the terminal d-Ala-d-Ala of nascent peptidoglycan (Daniel and Errington 2003; Reyn-olds 1989) before these termini get protected by process-ing either through the formation of crosslinks or through hydrolysis by carboxypeptidases (Holtje 1998) (Fig. 3a).

Fig. 1 Transmission electron microscopic images of mid-log phase M. tuberculosis cells with asymmetric and symmetric septum. a–d M. tuberculosis cells with asymmetric septum. e, f M. tuberculosis cells at the initial stages of asymmetric septation. g, h TEM images

of mid-log phase M. tuberculosis cells with symmetric septum. The wide septum, which is characteristic of M. tuberculosis cells, can be seen. n indicates nucleoid. Arrow indicates the position of the septum

161Arch Microbiol (2014) 196:157–168

1 3

Septum staining of the mid-log phase-fixed M. tuberculosis cells (n = 20 septating cells) with WGA-Alexa488, which specifically binds to N-acetylglucosamine in the peptidogly-can layer (Sizemore et al. 1990), also showed the presence of cells with asymmetric septum (Fig. 3b–d). Relatively more intense VBP and WGA-Alexa488 staining of the poles could be observed, probably indicative of growth from the poles, which is typical of mycobacteria (Dahl 2004; Thanky et al. 2007). The presence of nucleoids, which were stained with RiboGreen that binds both RNA and DNA (Sizemore et al. 1990), could be seen on both the sides of the TMA-DPH (Hett et al. 2008)-stained asymmetric and symmet-ric septum in the mid-log phase cells (n = 42) (Fig. 3 e, f, respectively). Unlike in TEM, these experiments, which did not involve cell sectioning, ruled out the possibility of oblique sectioning as a reason for asymmetric septum. Thus, these observations from the septum/nucleoid-stained cells confirmed the data from the TEM images that very low proportions of exponential phase cultures of M. tuberculosis

cells contain unusually deviated asymmetric septum, which compartmentalizes the mother cell into a short-cell and a long-cell portion, both with nucleoids.

Live cell time-lapse imaging shows unusually deviated asymmetric division in M. tuberculosis cells

Live cell time-lapse imaging of M. tuberculosis cells (n = 50) on agarose pad showed asymmetric constriction and division to generate a short cell and a long cell. An M. tuberculosis cell (white star marked cell in Fig. 4; see the corresponding movie S1, where the dividing cells have been tracked) underwent asymmetric constriction and divi-sion with unusual deviation in the position of the constric-tion, in the typical ‘snapping postfission’ mode of myco-bacterial division (Dahl 2004; Thanky et al. 2007), to give a short daughter cell (triangle symbol, 2.65 μm) and a long daughter cell (diamond symbol, 3.69 μm) (Fig. 4), differing in length by 1.04 μm. Both the short and the long daughter

Table 1 Length measurements of M. tuberculosis cells determined using different microscopy methods

TEM transmission electron microscopy, DIC differential interference contrast, LCM live cell microscopy. The number of the cells measured is indicated in bracket. It was observed that live cells were slightly longer in length (around 1–2 μm) as compared to the fixed cells. ND not deter-mined; NA not applicable. The longer portion, from the asymmetric septum to the farthest pole, of the cells with asymmetric septum is called the longer-sized portion and the short portion, from the asymmetric septum to the nearest pole, is called the short-sized portion of the cells with asymmetric septum

Description of the cells undergoing division TEM (in μm)(no of cells measured)

DIC (in μm)(no of cells measured)

LCM (in μm)(no of cells measured)

Longer-sized portion of the cells with asymmetric septum/constriction (A)

1.47 ± 0.24(26)

ND 3.64 ± 0.96(25)

Short-sized portion of the cells with asymmetric septum/constriction (B)

0.64 ± 0.08(26)

ND 2.31 ± 0.64(25)

Difference in the length between the long portion and the short portion of the cells with asymmetric septum (A) minus (B)

0.83 ± 0.16(26)

ND 1.33 ± 0.62(25)

Cells with asymmetric septum/constriction (C) 2.80 ± 0.70(26)

ND 5.98 ± 1.5(25)

Mid-cell site position in the cells with asymmetric septum (D) = (C)/2

1.40 ± 0.35(26)

ND 2.99 ± .75(25)

Length between the septum/constriction and the nearest pole in the cells with asymmetric septum E)

0.73 ± 0.10(26)

ND 2.31 ± 0.64(25)

Deviation in the position of the asymmetric septum/constriction from the median to the nearest pole (D) Minus (E)

0.67 ± 0.30(26)

ND 0.66 ± 0.31(25)

Cells with symmetric septum/constriction (F) 2.63 ± 0.32(24)

ND 6.23 ± 1.2(25)

Mid-cell site position in the cells with symmetric septum (F)/2 1.31 ± 0.16(24)

NA 3.12 ± 1.2(25)

Individual non-septating cells 1.70 ± 0.26(26)

ND ND

Individual short cells in the 7H9 medium ND 0.73 ± 0.13(300)

ND

Individual short cells in the sputum sample NA 0.91 ± 0.10(300)

NA

Individual normal-sized cells in the sputum sample NA 2.91 ± 1.16(300)

NA

162 Arch Microbiol (2014) 196:157–168

1 3

cells showed further growth and symmetric division, with minor deviation in the constriction, to generate daughter cells (Fig. 4). The short cell (triangle symbol, 2.65 μm) gave rise to daughter cells of lengths 3.13 μm (pentagon symbol) and 2.47 μm (right angle triangle symbol), while the long cell (diamond symbol, 3.69 μm) divided sym-metrically with minor deviation to generate daughter cells of lengths 2.58 μm (circle symbol) and 2.96 μm (square symbol). These daughter cells further underwent sym-metric division with minor deviation in the constriction. Thus, the mother cell and its progeny cells went through three consecutive asymmetric divisions with major devia-tion and/or symmetric divisions with minor deviation to generate micro-colony. The lineages of these daughter cells have been traced back to the starting mother cell (Fig. S2). More live cell time-lapse images of asymmetric division of M. tuberculosis cells are given in Figs. S3, S4, and S5 (see the corresponding movies, S2, S3, and S4, where the dividing cells have been tracked). The average deviation in the position of the constriction in the live cell time-lapse images of asymmetrically dividing cells was found to be 0.66 ± 0.31 μm (n = 25) (Table 1). This amounted to a deviation of 11 ± 4 % (n = 25) of the cell length, from the median (see Fig. 2). Thus, live cell time-lapse imaging

of M. tuberculosis cells confirmed the occurrence of asym-metric division with unusually deviated constriction in the subpopulation and confirmed the observations from the TEM and septum/nucleoid-stained imaging experiments.

Post-completion of the constriction, both the short and the long cells grew at the elongation rates of 1.00 μm/8 h and 1.07 μm/8 h, respectively, and underwent further con-striction and division (see Fig. 4 and the corresponding movie S1). Major proportion (80 % of the 15 % septating cells) of M. tuberculosis cells showed symmetric constric-tion and division, with minor constriction deviation of 2.50 ± 2 % (n = 25) from the median (see Fig. 2), to gen-erate slightly unequally sized daughter cells (Fig. S6; see the corresponding movie S5, where the dividing cells have been tracked). Comparable elongation rates of 1.13 μm/8 h and 1.25 μm/8 h were observed for the two portions of the cell undergoing symmetric division (Fig. S6; see the cor-responding movie S5, where the dividing cells have been tracked). Growth and division per se of the M. tuberculosis cells were not affected on agarose pad, as in both asym-metric and symmetric division, the cells in the agarose pad grew and divided once in about 18–24 h, which is the divi-sion time reported for M. tuberculosis cells in liquid media (Hiriyanna and Ramakrishnan 1986).

Fig. 2 Position of the septum and constriction in the asymmetrically and symmetrically dividing M. tuberculosis cells. a Placement of the septum with respect to the cell length in the M. tuberculosis cells from TEM images. b Position of the constriction in the M. tuberculo-

sis cells from the live cell time-lapse images. (n = 50 septating/con-stricting cells in each case). The average values (with standard devia-tion) from a to b have been tabulated in the chart below

163Arch Microbiol (2014) 196:157–168

1 3

Presence of short cells in the mid-log phase culture and in the sputum of TB patients

Commensurate with the cells undergoing unusually devi-ated asymmetric division, DIC imaging of the mid-log phase cells showed the presence of short cells (Fig. 5a, b). The ultrastructure of short cells showed the thick electron-transparent envelope layer, like the characteristic arabi-nogalactan-mycolic acid layer of the cell envelope of the normal-sized M. tuberculosis cells (reviewed in Crick et al. 2008) (Fig. 5c, d, compare with the normal-sized cell in Fig. 5e). Like in the case of the short portion of asymmetric septum containing cells, the short cells contained nucleoid with compact appearance (compare Fig. 5c, d with Fig. 1a–d). From the DIC images, the sizes of the short cells were

found to be 0.73. ± 0.13 μm (n = 300) (Table 1). Since M. tuberculosis cells exhibited high levels of cell length het-erogeneity in the population, it was not possible to distin-guish the normal-sized cells from the long cells generated from asymmetric division, in the population.

In order to find out whether M. tuberculosis short cells exist in clinical samples also, acid fast stained sputum sam-ples from freshly diagnosed (before treatment) pulmonary tuberculosis patients were examined. The sputum sam-ples (n = 73) also showed the presence of about 3–5 % of short cells (n = 5,000) and 95–97 % of normal/long cells (n = 5,000) (Fig. 5f), as found in the mid-log phase cultures. The sizes of the short cells (0.91 ± 0.10 μm) and of the normal-sized cells (2.91 ± 1.16 μm) in the sputum samples were slightly higher than the sizes of the acid fast stained

Fig. 3 Fluorescence images of mid-log phase M. tuberculosis cells with asymmetrically or symmetrically placed septum. a VBP-stained live M. tuberculosis cell with asymmetric septum. b–d WGA-Alexa488-stained asymmetric septum in fixed M. tuberculosis cells.

e, f M. tuberculosis cells with asymmetric septum and symmetric septum, respectively, stained with RiboGreen for DNA and RNA and with TMA-DPH for the septal membrane. Arrows indicate the posi-tion of the septum

164 Arch Microbiol (2014) 196:157–168

1 3

165Arch Microbiol (2014) 196:157–168

1 3

short cells (0.73 ± 0.13 μm; n = 300) and the normal-sized cells (2.10 ± 0.60 μm; n = 300) from in vitro culture. The size of tubercle bacilli in sputum samples has been reported to be consistently 1-2 fold higher than that of the laboratory grown M. tuberculosis cells (Garton et al. 2002).

Discussion

The cell division studies in mycobacteria have so far focused on the mode of division of majority of the cells in the population, wherein about 80 % of the septating cells divide symmetrically, but with 5–10 % deviation in the septum position (Joyce et al. 2012; Santi et al. 2013; Singh et al. 2013). Corynebacterium glutamicum, which is phylogenetically closely related to mycobacteria, also show similar division pattern. About 97 % of the septat-ing C. glutamicum cells divide symmetrically but with less than 10 % deviation in the position of the septum (Joyce et al. 2012), which is similar to that reported in the major-ity cells of the mycobacterial population by others (Joyce et al. 2012; Santi et al. 2013; Singh et al. 2013) and by us in the majority population in M. tuberculosis in the present study. Yet another system showing symmetric division with minor deviation in septum position is Helicobacter pylori, wherein about 50 % of the cells in the population divide symmetrically with 5–10 % septum position deviation (Specht et al. 2013). Such deviations have been attributed to the absence of the topological regulators, MinCDE, and the nucleoid occlusion systems, noc/slmA, as E. coli and B. subtilis, which have these systems, divide with precision symmetry (Migocki et al. 2004; Yu and Margolin 1999). Although such symmetric division with minor deviation in the septum position occurs in the majority of the septat-ing population, the observations made in the present study reveal that asymmetric division with unusual septum posi-tion deviation (24 ± 4 %, n = 26) occurs consistently and reproducibly in low proportions of M. tuberculosis cells in the exponential phase population to generate short cells and

long cells, which differ distinctly in their lengths. The dif-ference in the deviations measured for the septum position (24 ± 4 %), as compared to that of the constriction position (11 ± 4 %) (see Fig. 2), could be due to the possible dif-ferential polar growth reducing the size difference between the daughter cells, during the substantial time interval that exists between cytokinesis and sibling cell separation, as reported for M. smegmatis (Santi et al. 2013).

Asymmetric division can occur either through the place-ment of the septum at asymmetric position, followed by constriction and division, or through the symmetric place-ment of the septum, followed by differential polar growth. Although our data on the asymmetric division of M. tuber-culosis cells does not distinguish between these two pos-sibilities, earlier studies in our laboratory had shown that septal constriction in mycobacteria occurs only subsequent to the formation of a complete septum, unlike in E. coli and B. subtilis, where septum formation and constriction are simultaneous events (Vijay et al. 2012). Therefore, the position of the constriction at asymmetric position in the live cell images, where further division occurred to gener-ate short and long cells, confirmed the unusual deviation in the position of the septum. Further, the comparable lengths of the cells with asymmetric septum and symmetric septum indicated that the septum positioning itself might have been asymmetric, rather than due to differential polar growth or elongation rate. This is in conformity with the asymmet-ric positioning of FtsZ in M. marinum and M. smegmatis (Singh et al. 2013).

Asymmetric division of M. tuberculosis cells with unu-sual septum deviation showed specific differences, as com-pared to the symmetric division with minor septum devia-tion reported in M. smegmatis, M. marinum, and M. bovis BCG cells (Joyce et al. 2012; Santi et al. 2013; Singh et al. 2013). First of all, the asymmetric division with unusual septum deviation was found only in a low proportion of M. tuberculosis cells, while the cells in the majority popula-tion still divided symmetrically, but with minor deviation in the septum position, identical to the symmetric division with minor deviation reported in the majority proportion of other mycobacterial species (Joyce et al. 2012; Santi et al. 2013; Singh et al. 2013). Secondly, it was shown that the nucleoid gets transported from the long-cell portion to the anucleated short-cell portion across a fully formed septum in the asymmetrically dividing M. smegmatis and M. mari-num cells, often showing septum placed over the chromo-some (Singh et al. 2013). On the contrary, not even a single asymmetrically dividing exponential phase M. tuberculosis cell could be found without nucleoid in the short cell as well as long-cell portions, as revealed by the TEM imag-ing (n = 150 septating cells). Even in the majority popula-tion of M. tuberculosis cells, where the cells divided sym-metrically with minor septum deviation, both the portions

Fig. 4 Live cell time-lapse imaging of an M. tuberculosis cell undergoing asymmetric division and the daughter cells undergoing further divisions. Growth and asymmetric division of an M. tuber-culosis cell (white star) to generate a short daughter cell (triangle symbol) of length, 2.65 µm, and a long daughter cell (diamond sym-bol) of length, 3.69 µm, with the difference in their lengths being 1.04 µm. The short daughter cell (triangle symbol) showed further growth and symmetric division to generate cells of lengths, 3.13 µm (pentagon symbol) and 2.47 µm (right angle triangle symbol). The long daughter cell (diamond symbol) also underwent further growth and symmetric division to generate daughter cells of lengths, 2.58 µm (circle symbol) and 2.96 µm (square symbol). Further, all these cells grew and divided symmetrically. The cells were observed for three generations. See the corresponding movie S1, where the divid-ing cells have been tracked

◂

166 Arch Microbiol (2014) 196:157–168

1 3

Fig. 5 Profile of M. tuberculosis short cells. a, b M. tuberculosis short cells and normal-sized and elongating (for division) cells. The short arrow indicates short cells, medium arrow indicates normal-sized/long cells, and long arrow indicates elongated (for division) cells. c, d TEM images of M. tuberculosis short cells. e TEM image

of M. tuberculosis normal-sized cell. The arrows in c–e show the presence of the thick electron-transparent mycolic acid layer. f Tuber-cle bacilli in sputum sample from freshly diagnosed (before the start of treatment) TB patients (bright field view). The arrows indicate short cells

167Arch Microbiol (2014) 196:157–168

1 3

of the dividing cells always contained nucleoids. Thirdly, M. tuberculosis short cells from asymmetric division under-went next division within a generation time after birth, with-out any lag in the growth or division. This feature is differ-ent from the lack of division, but growth, of M. smegmatis short cells from the non-medial division, for one generation to attain the cell size similar to that of the daughter cells from the division of longer sister cell (Singh et al. 2013).

The rod-shaped morphology of the short cells, the pres-ence of nucleoid in the short cells (TEM images), and the growth and division of the short cells (live cell time-lapse imaging) showed that they are viable and culturable, unlike the viable-but-non-culturable (VBNC) ovoid/coccoid-shaped mycobacterial cells found only under extreme stress condi-tions (Shleeva et al. 2002, 2004). The existence of short cells in the actively growing cultures also show that they are nei-ther spores, unlike those reported in the aged cultures of M. smegmatis, M. marinum, M. bovis (Ghosh et al. 2009), and M. avium Subsp. Paratuberculosis (Lamont et al. 2012). In addi-tion, retention of acid fastness showed that the short cells are neither the non-replicating dormant forms of mycobacteria (Nyka 1974) nor the cell wall deficient l-forms of the bacilli (Markova et al. 2012). These properties and the consistent and reproducible maintenance of low proportions of cells undergoing asymmetric division with high division position deviation allude to the possibility that the generation of short cells may not be a ‘noise’ but an inherent, regulated, physi-ologically relevant feature of M. tuberculosis population.

Microbial phenotypic heterogeneity has correlation to tol-erance to stress conditions such as nutrient depletion (Nyka 1974; Smeulders et al. 1999) and antibiotics (Aldridge et al. 2012) in mycobacteria and in other bacterial systems (reviewed in Dhar and McKinney 2007). Stationary phase mycobacterial cells have been found to be short in size, and the proportion of short cells has been found to increase as the cultures approach stationary phase (Smeulders et al. 1999; Thanky et al. 2007). However, it is not clear whether these short cells have arisen through asymmetric division in response to nutrient stress. Nevertheless, the short cells pre-sent in the nutrient-rich condition in the mid-log phase may be different from the short cells present under stress condi-tions, as they are generated and present in the absence of any stress. Phenotypically divergent forms of M. tuberculosis and other mycobacterial cells have been reported in myco-bacterial cultures (Imaeda 1975; Korsak 1975; Markova et al. 2008, 2012; McCarthy 1971), mouse and guinea pig lung tissues (Ryan et al. 2010), and in patient samples (Far-nia et al. 2010; Khomenko 1987). Asymmetrically dividing tubercle bacilli have been found in the exponential phase cul-tures of drug-susceptible and MDR strains to the extent of 20 % (Farnia et al. 2010), which is consistent with our obser-vations in the mid-log phase population. This proportion of the asymmetrically dividing cells has been found to increase

up to 40–45 % in the XDR and XXDR isolates (Farnia et al. 2010). The presence of phenotypically divergent forms of tubercle bacilli, such as the l-forms (Imaeda 1975; Korsak 1975; Markova et al. 2008, 2012) have been implicated in the survival under nutrient starvation conditions (Markova et al. 2008; Markova et al. 2012). Similarly, tuberculosis patients’ sputum, where the presence of short cells could be detected, is known for the prevalence of oxidative stress conditions (Palanisamy et al. 2011; Phillips et al. 2007; Rintiswati et al. 2011; Sadowska et al. 2005). The presence of short cells in the sputum samples implied that the tubercle bacilli in the sputum also may divide asymmetrically with unusually devi-ated septum and that the cells therein may be tolerant to oxi-dative stress conditions. It has also been found that the par-ticle shape, not size, plays a dominant role in phagocytosis (Champion and Mitragotri 2006), which in turn may have some relevance in the case of the rod shape of the short cells. All these studies strengthen the possibility that the short cells may be involved in stress response—a survival strategy nec-essary for a pathogen such as M. tuberculosis.

Acknowledgments This work was supported by a part-Grant from the DBT supported Centre of Excellence in Tuberculosis Research to P.A. Authors acknowledge infrastructure support from DST-FIST, UGC-CAS, and IISc to the MCB Dep’t. Authors acknowledge Dr. S. S. Indi and Mr. P. V. Balasubramaniam of the TEM facility in the MCB Dep’t, for technical help. S.V. and J.S. acknowledge CSIR and DBT for SRFs.

References

Aldridge BB, Fernandez-Suarez M, Heller D, Ambravaneswaran V, Irimia D, Toner M, Fortune SM (2012) Asymmetry and aging of mycobacterial cells lead to variable growth and antibiotic suscep-tibility. Science 335:100–104

Barksdale L, Kim K-S (1977) Mycobacterium. Bacteriol Rev 41:217–372

Champion JA, Mitragotri S (2006) Role of target geometry in phago-cytosis. Proc Natl Acad Sci USA 103:4930–4934

Crick DC, Quadri L, Brennan PJ (2008) Biochemistry of the cell envelope of Mycobacterium tuberculosis. In: Kaufmann SHE, Rubin E (eds) Handbook of tuberculosis. Molecular biology and biochemistry. Wiley-VCH Verlag GmbH & Co. KGaA, Wein-heim, Germany, pp 1–19

Dahl JL (2004) Electron microscopy analysis of Mycobacterium tuberculosis cell division. FEMS Microbiol Lett 240:15–20

Daniel RA, Errington J (2003) Control of cell morphogenesis in bacteria: two distinct ways to make a rod-shaped cell. Cell 113:767–776

de Jong IG, Beilharz K, Kuipers OP, Veening J-W (2011) Live cell imaging of Bacillus subtilis and Streptococcus pneumoniae using automated time-lapse microscopy. J Vis Exp 53:e3145

Dhar N, McKinney JD (2007) Microbial phenotypic heterogeneity and antibiotic tolerance. Curr Opin Microbiol 10:30–38

Edwards RP (1970) Electron-microscope illustrations of division in Mycobacterium leprae. J Med Microbiol 3:493–499

Farnia P, Masjedi MR, Merza MA, Tabarsi P, Zhavnerko GK et al (2010) Growth and cell-division in extensive (XDR) and extremely drug resistant (XXDR) tuberculosis strains:

168 Arch Microbiol (2014) 196:157–168

1 3

transmission and atomic force observation. Int J Clin Exp Med 3:308–314

Garton NJ, Christensen H, Minnikin DE, Adegbola RA, Barer MR (2002) Intracellular lipophilic inclusions of mycobacteria in vitro and in sputum. Microbiology 148:2951–2958

Ghosh J, Larsson P, Singh B, Pettersson BMF, Islam NM, Sarkar SN, Dasgupta S, Kirsebom LA (2009) Sporulation in mycobacteria. Proc Natl Acad Sci USA 106:10781–10786

Hett EC, Rubin EJ (2008) Bacterial growth and cell division: a myco-bacterial perspective. Microbiol Mol Biol Rev 72:126–156

Hett EC, Chao MC, Deng LL, Rubin EJ (2008) A Mycobacterial enzyme essential for cell division synergises with resuscitation-promoting factor. PLoS Pathog 4:e1000001

Hirata T (1978) Electron microscopic observations of cell division in Mycobacterium leprae by means of serial ultrathin sectioning. Int J Lepr Mycobact Dis 46:160–166

Hiriyanna KT, Ramakrishnan T (1986) Deoxyribonucleic acid repli-cation time in Mycobacterium tuberculosis H37Rv. Arch Micro-biol 144:105–109

Holtje J-V (1998) Growth of the stress-bearing and shape-maintaining murein sacculus of Escherichia coli. Microbiol Mol Biol Rev 62:181–203

Imaeda T (1975) Ultrastructure of L-phase variants isolated from a culture of Mycobacterium phlei. J Med Microbiol 8:389–395

Joyce G, Robertson BD, Williams KJ (2011) A modified agar pad method for mycobacterial live-cell imaging. BMC Res Notes 4:73

Joyce G, Williams KJ, Robb M, Noens E, Tizzano B, Shahrezaei V, Robertson BD (2012) Cell division site placement and asymmet-ric growth in mycobacteria. PLoS ONE 7:e44582

Khomenko AG (1987) The variability of Mycobacterium tuberculosis in patients with cavitary pulmonary tuberculosis in the course of chemotherapy. Tubercle 66:243–253

Korsak T (1975) Occurrence of L-forms in a case of generalized mycobacteriosis due to Mycobacterium scrofulaceum. Acta Tuberc Pneumol Belg 66:445–469

Lamont EA, Bannantine JP, Armien A, Ariyakumar DS, Sreevatsan S (2012) Identification and characterisation of a spore-like morpho-type in chronically starved Mycobacterium avium Subsp. Paratu-berculosis cultures. PLoS ONE 7:e30648

Markova N, Michailova L, Jourdanova M, Kussovski V, Valcheva V, Mokrousov I, Radoucheva T (2008) Exhibition of persistent and drug tolerant L-form habit of Mycobacterium tuberculosis during infection in rats. Cent Eur J Biol 3:407–416

Markova N, Slavchev G, Michailova L (2012) Unique biologi-cal properties of Mycobacterium tuberculosis L-form variants: impact for survival under stress. Int Microbiol 15:61–68

McCarthy C (1971) Electronic counting in growth studies of Myco-bacterium avium. Appl Microbiol 22:546–551

McCarthy C (1974) Effect of palmitic acid on cell division in Myco-bacterium avium. Infect Immun 9:363–372

Migocki MD, Lewis PJ, Wake RG, Harry EJ (2004) The midcell rep-lication factory in Bacillus subtilis is highly mobile: implications for coordinating chromosome replication with other cell cycle events. Mol Microbiol 54:452–463

Nyka W (1974) Studies on the effect of starvation on mycobacteria. Infect Immun 9:843–850

Palanisamy GS, Kirk NM, Ackart DF, Shanley CA, Orme IM, Basaraba RJ (2011) Evidence for oxidative stress and defective antioxidant response in guinea pigs with tuberculosis. PLoS ONE 6:e26254

Phillips M, Cataneo RN, Condos R, Erickson GAR, Greenberg J, Bombardi VL, Munawar MI, Tietje O (2007) Volatile biomarkers of pulmonary tuberculosis in the breath. Tuberculosis 87:44–52

Rasband WS (2012) ImageJ, U. S. National Institute of Health, Bethesda. http://imagej.nih.gov/ij/

Rastogi N, David HL (1981) Growth and cell division of Mycobacte-rium avium. J Gen Microbiol 126:77–84

Reynolds PE (1989) Structure, biochemistry and mechanism of action of glycopeptide antibiotics. Eur J Clin Microbiol Infect Dis 8:943–950

Rintiswati N, Wibawa T, Asmara W, Soebono H (2011) Effect of oxi-dative stress on AhpC activity and virulence in katG Ser315Thr Mycobacterium tuberculosis mutant. Indones J Biotechnol 16:100–110

Ryan GJ, Hoff DR, Driver ER, Voskuil MI, Gonzalez-Juarrero M et al (2010) Multiple M. tuberculosis phenotypes in mouse and guinea pig lung tissue revealed by a dual-staining approach. PLoS ONE 5:e11108

Sadowska AM, van Overvelda FJ, Goreckab D, Zdralc A, Filews-kac M, Demkowc UA, Luytena C, Saenena E, Zielinskib J, De Backer WA (2005) The interrelationship between markers of inflammation and oxidative stress in chronic obstructive pulmo-nary disease: modulation by inhaled steroids and antioxidant. Respir Med 99:241–249

Santi I, Dar N, Bousbaine D, Wakamoto Y, McKinney JD (2013) Sin-gle-cell dynamics of the chromosome replication and cell divi-sion cycles in mycobacteria. Nat Commun 4:2470. doi:10.1038/ncomms3470

Shleeva MO, Bagramyan K, Telkov MV, Mukamolova GV, Young M et al (2002) Formation and resuscitation of “nonculturable” cells of Rhodococcus rhodochrous and Mycobacterium tuberculosis in prolonged stationary phase. Microbiology 148:1581–1591

Shleeva MO, Mukamolova GV, Young M, Williams HD, Kaprelyants AS (2004) Formation of “nonculturable” cells of Mycobacterium smegmatis in stationary phase in response to growth under subop-timal conditions and their Rpf-mediated resuscitation. Microbiol-ogy 150:1687–1697

Singh B, Nitharwal RG, Ramesh M, Pettersson BMF, Kirsebom LA, Dasgupta S (2013) Asymmetric growth and division in Mycobac-teria spp.: compensatory mechanisms for non-medial septa. Mol Microbiol 88:64–76

Sizemore RK, Caldwell JJ, Kendrick AS (1990) Alternate gram-stain-ing technique using a fluorescent lectin. Appl Environ Microbiol 56:2245–2247

Smeulders MJ, Keer J, Speight RA, Williams HD (1999) Adaptation of Mycobacterium smegmatis to stationary phase. J Bacteriol 181:270–283

Specht M, Dempwolff F, Schatzle S, Thomann R, Waldner B (2013) Localisation of FtsZ in Helicobacter pylori and consequences for cell division. J Bacteriol 195:1411–1420

Takade A, Takeya K, Taniguchi H, Mizuguchi Y (1983) Electron Microscopic Observations of Cell Division in Mycobacterium vaccae V1. J Gen Microbiol 129:2315–2320

Thanky NR, Young DB, Robertson BD (2007) Unusual features of the cell cycle in mycobacteria: polar-restricted growth and the snap-ping-model of cell division. Tuberculosis 87:231–236

Vijay S, Anand D, Ajitkumar P (2012) Unveiling unique features of formation of septal partition and constriction in mycobacteria—an ultrastructural study. J Bacteriol 194:702–707

Weinbauer MG, Beckmann C, Hofle MG (1998) Utility of green fluo-rescent nucleic acid dyes and aluminum oxide membrane filters for rapid epifluorescence enumeration of soil and sediment bacte-ria. Appl Environ Microbiol 64:5000–5003

Yu X-C, Margolin W (1999) FtsZ ring clusters in min and partition mutants: role of both the min system and the nucleoid in regulat-ing FtsZ ring localisation. Mol Microbiol 32:315–326

Zheng H, Lu L, Wang B, Pu S, Zhang X, Zhu G, Shi W, Zhang L, Wang H, Wang S, Zhao G, Zhang Y (2008) Genetic basis of viru-lence attenuation revealed by comparative genomic analysis of Mycobacterium tuberculosis strain H37Ra versus H37Rv. PLoS ONE 3:e2375. doi:10.1371/journal.pone.0002375