- Email: [email protected]
Isolation and purification of Mycobacterium tuberculosis from H37Rv infected guinea pig lungs
Tuberculosis 94 (2014) 525e530
Contents lists available at ScienceDirect
Tuberculosis journal homepage: http://intl.elsevierhealth.com/journals/tube
Isolation and purification of Mycobacterium tuberculosis from H37Rv infected guinea pig lungs Libin Shi, Gavin J. Ryan, Suresh Bhamidi, JoLynn Troudt, Anita Amin, Angelo Izzo, Anne J. Lenaerts, Michael R. McNeil, John T. Belisle, Dean C. Crick**, Delphi Chatterjee* Mycobacteria Research Laboratories, Department of Microbiology, Immunology and Pathology, Colorado State University, Fort Collins, CO 80523, USA
a r t i c l e i n f o
s u m m a r y
Article history: Received 15 May 2014 Accepted 23 May 2014
Evidence suggests that Mycobacterium tuberculosis grown in vivo may have a different phenotypic structure from its in vitro counterpart. In order to study the differences between in vivo and in vitro grown bacilli, it is important to establish a reliable method for isolating and purifying M. tuberculosis from infected tissue. In this study, we developed an optimal method to isolate bacilli from the lungs of infected guinea pigs, which was also shown to be applicable to the interferon-g gene knockout mouse model. Briefly, 1) the infected lungs were thoroughly homogenized; 2) a four step enzymatic digestion was utilized to reduce the bulk of the host tissue using collagenase, DNase I and pronase E; 3) residual contamination by the host tissue debris was successfully reduced using percoll density gradient centrifugation. These steps resulted in a protocol such that relatively clean, viable bacilli can be isolated from the digested host tissue homogenate in about 50% yield. These bacilli can further be used for analytical studies of the more stable cellular components such as lipid, peptidoglycan and mycolic acid. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Mycobacterium tuberculosis Bacterial isolation Guinea pig infection
1. Introduction Tuberculosis (TB) is an airborne infectious disease that is both preventable and curable; however, it continues to be a major cause of death worldwide. In 2012, there were an estimated 8.6 million new cases globally, and 1.3 million people died from TB including 320,000 deaths in HIV-infected people [1]. The WHO has set a goal to reduce the global burden of TB dramatically by 2015 [2], therefore, development of new tools for optimal diagnosis, prevention and treatment of TB is urgent [3]. However, in terms of identifying novel diagnosis and validated drug targets, most of the present knowledge of the tubercule bacillus is derived from the bacilli grown in vitro, which may not accurately represent organisms living in a host lung. Mycobacterium tuberculosis is typically acquired by inhalation of aerosolized droplets from infected individuals resulting in the infection of alveolar macrophages, which release various cytokines resulting in the formation of granulomas, a hallmark of TB disease [4]. M. tuberculosis is an obligate aerobic organism, however, it adapts to
* Corresponding author. Tel.: þ1 970 491 7495; fax: þ1 970 491 1815. ** Corresponding author. Tel.: þ1 970 491 3308; fax: þ1 970 491 1815. E-mail addresses: [email protected] (D.C. Crick), delphi.chatterjee@ colostate.edu (D. Chatterjee). http://dx.doi.org/10.1016/j.tube.2014.05.003 1472-9792/© 2014 Elsevier Ltd. All rights reserved.
oxygen limitation during the development of granulomas as they become relatively hypoxic compared to the surrounding tissue in guinea pig, rabbit, and nonhuman primate models of disease [5], and the bacilli appear to change their physical characteristics in response to stimuli encountered in the host microenviroment. Seiler et al. [6] reported that with persistence of M. tuberculosis infection in mice and human patients, the classical cell-wall component-dependent Ziehl-Neelsen (ZN) staining gradually reduced while the detection by cell-wall independent staining using a polyclonal anti-Mycobacterium bovis BCG serum continued, suggesting that the cell-wall of M. tuberculosis was either lost or reorganized. Recently, applying a dual-staining method including fluorescent acid-fast auraminerhodamine targeting the mycolic acid containing cell-wall followed by immunofluorescence targeting bacterial proteins using an antiM. tuberculosis whole cell lysate polyclonal antibody, Ryan et al. [7] observed three M. tuberculosis populations in hypoxic culture; similar subpopulations of bacilli were also found in the lungs from M. tuberculosis infected mice and guinea pigs. These three populations appear to be either exclusively acid-fast positive, exclusively immunofluorescent positive or acid-fast and immunofluorescent positive, a result suggesting that multiple populations of M. tuberculosis may exist in a single microenvironment. Using transmission electron microscopy, Cunningham et al. [8] showed that microaerobically and anaerobically cultured
526
L. Shi et al. / Tuberculosis 94 (2014) 525e530
M. tuberculosis bacilli have thicker cell-walls than aerobically cultured bacilli. The cell-wall thickening could consist of polysaccharides and lipids and may help the bacilli to survive in oxygen-deficient microenvironment in vivo. In addition to hypoxia, M. tuberculosis bacilli also encounter low pH and low nutrient availablility within the caseous tissue in the development of granulomas [9,10]. Under nutrient starvation conditions, there is a global metabolic shift and lipids are thought to become the sole source of energy for mycobacteria [11e13]. Recently, Markova et al. [14] reported L-form variants of M. tuberculosis under nutrient starvation stress, which lost their acidfast character and changed their morphology. Single-stress conditions such as hypoxia and nutrient starvation provide information regarding the physiological changes that occur during TB development, but cannot fully simulate conditions found in the granuloma. Deb et al. [15] adopted an in vitro multiple-stress condition including low oxygen (5%), high CO2 (10%), low nutrient (10% Dubos medium) and acidic pH (5.0) in which M. tuberculosis stopped replicating, lost acid-fastness, and accumulated triacylglycerol and wax ester. Collectively, these results suggested that the bacilli under stress may alter their cell-wall, and potentially their metabolic status to enter a state of nonreplicating persistence and phenotypic drug resistance [16e20]. Ideally, the changes induced in the physiology and biochemistry of the bacilli during disease progression would best be studied in bacteria isolated from the lungs of infected hosts. To date very little data has been published regarding the physiological and metabolic state of M. tuberculosis isolated from host tissue; although there are striking differences reported in the metabolic characteristics with respect to substrate response between tubercle bacilli grown in culture and their counterparts isolated from moribund infected mice [21]. In this study, we aimed at developing a method to isolate M. tuberculosis from the lungs of infected guinea pigs that surmount the significant technical difficulties involved in reaching this goal. Initially, we attempted to adopt a protocol reported for isolation and purification of Mycobacterium leprae from armadillo livers, which has been widely used for studies of M. leprae genomics, proteomics, lipidomics and cell-wall structure [22e24]. Unfortunately, this method of isolation, which was developed to provide vaccine grade material, failed to give a desirable yield of viable bacilli when applied to M. tuberculosis infected lungs. The fact that guinea pig lungs contained relatively low numbers of bacilli, ~1056 bacilli/g of lung, relative to 10910 bacilli/g of M. leprae in infected armadillo liver possibly contributed to the problems associated with the isolation. In addition, the sodium hydroxide treatment resulted in the death of more than 90% of the bacilli making CFU enumeration for bacilli tracking impossible, the protein digestion did not efficiently liberate bacilli from homogenized lung tissue and it was difficult to perform biphasic partitioning between polyethylene glycol and dextran solution in a Biosafety Level 3 (BSL3) environment. A protocol previously described by Bolch et al. [21] to isolate M. tuberculosis H37Rv from the infected mice was also investigated. This protocol resulted in incomplete release of bacilli from the tissue and contamination with sucrose and mitochondria complicated downstream analysis. The low yields of viable bacilli and severe host tissue contamination suggested that neither method was optimal for isolating M. tuberculosis from infected guinea pig lungs. Thus, a new facile and reproducible method was developed for isolating M. tuberculosis bacilli from host tissue in high yield and purity. 2. Methods and materials 2.1. Bacteria and plasmid M. tuberculosis H37Rv (Trudeau Institute) was grown from lowpassage seed lots in Proskauer-Beck liquid medium containing
0.05% Tween 80 to early mid-log phase and frozen in aliquots at 80 C until needed. Cultures were diluted in sterile water prior to use. Plasmid pCherry3 [25], which was kindly provided by Dr. Tanya Parish, was electroporated into M. tuberculosis H37Rv and cultured on 7H11 supplemented (Becton, Dickinson and Company) with 10% BBL Middle Brook OADC enrichment (BD) plus hygromycin (Streptomyces sp., Calbiochem) at 50 mg/mL for 3 weeks at 37 C. 2.2. Animal infection All experimental protocols were approved by the Animal Care and Usage Committee of Colorado State University. Female outbred Hartley guinea pigs (approximately 500 g in weight) were purchased from the Charles River Laboratories (North Wilmington) and held under barrier conditions in a BSL3 animal laboratory. Guinea pigs were infected with 10e20 CFU of M. tuberculosis H37Rv via the low-dose aerosol route as previously described [26], or 100 CFU of bacilli via high-dose aerosol infection. The guinea pigs were sacrificed 30 days after infection, lungs were rapidly excised, immediately frozen in liquid nitrogen and stored at 80 C. Eight- to ten-week-old female specific pathogen-free C57BL/6Ifngtm1ts (GKO) mice were purchased (Jackson Laboratories). These mice have a disrupted interferon-g gene, which renders them highly susceptible to TB [27]. The standard infection for this model was performed as previously described [28]. Briefly, mice were exposed to a low-dose aerosol infection with M. tuberculosis using a Glas-Col inhalation exposure system (Glas-Col LLC.) [29]. One day after low-dose aerosol infection, three mice were euthanized to verify infection with 50e100 CFU per mouse. The mice were sacrificed after 18 days exposure, and lungs were excised and stored frozen as mentioned above. 2.3. Isolation and purification of M. tuberculosis from lungs 2.3.1. Homogenization of infected guinea pig lungs Frozen guinea pig lungs were thawed at room temperature (RT) and weighed. The lungs were rinsed with phosphate buffered saline (PBS, pH 7.2, without CaCl2 or MgCl2, Gibco) and half of the total lung mass from each guinea pig was placed in a 40 mL straight walled homogenization tube (Glas-Col LLC.). PBS containing 1 mg/ mL collagenase (Clostridium histolyticum, type XI, SigmaeAldrich), 30 mg/mL deoxyribonuclease I (DNase I, from bovine pancreas, type IV, SigmaeAldrich) and 5 mM CaCl2 was added at 3 mL per gram of lung tissue. Lungs were initially homogenized using a Tissue-Tearor (Kinematic Polytron) followed by further homogenization using a variable speed motor drive homogenizer and teflon pestle (Glas-Col LLC.) at 4000 rpm for 15 passes [30]. 2.3.2. Enzyme digestion of the homogenized lungs To obtain the maximal bacilli release from the tissue homogenate, multiple enzyme treatments were tested and four alternative conditions of enzyme digestion were adopted, which utilized collagenase followed by either proteinase K (recombinant, Roche) or pronase E (protease from Streptomyces griseus, type XIV, SigmaeAldrich). Thus, twelve low-dose aerosol infected guinea pig lungs were divided into four groups and were homogenized as described above. The homogenates were exposed to two rounds of collagenase digestion in all four protocols. This included an initial incubation with collagenase at 1 mg/mL at 37 C for 1 h, followed by addition of more collagenase to 2 mg/mL final concentration and incubation at RT for 2 more hours. Following the collagenase digestion, either proteinase K or pronase E was added. In one treatment, proteinase K was added to 0.5 mg/mL and the mixture was incubated at RT for 2 h. After centrifugation at 10,000 g for
L. Shi et al. / Tuberculosis 94 (2014) 525e530
15 min at 4 C, the pellets were resuspened in 0.5 mg/mL proteinase K for 1 more hour incubation at RT. In the second treatment, proteinase K was added to 0.5 mg/mL and the flask was incubated at 4 C for overnight. In the third and fourth treatments, pronase E was added in place of proteinase K and incubated as described. In each case, the digested homogenate was centrifuged at 10,000 g for 15 min at 4 C to obtain pellets containing the bacilli. 2.3.3. Density gradient centrifugation A working solution of percoll with a density of 1.07 g/mL was prepared according to the manufacturer's protocol (Amersham Biosciences). Stock isotonic percoll (SIP) was prepared by adding 9 parts of percoll to 1 part of 10 PBS (v/v, ri ¼ 1.123 g/mL), and then SIP was diluted to working solution (ri ¼ 1.07 g/mL) by adding 60.8 mL PBS to 75 mL SIP. The pellets obtained from enzymatic digests (Section 2.3.2.) were suspended in 3 mL PBS and homogenized using the GlasCol homogenizer as described above and gently layered on top of 35 mL working solution of percoll in 50 mL polycarbonate centrifuge tubes (Beckman). After centrifugation at 30,000 g for 30 min at 4 C with no deacceleration, nine fractions (numbered 1e9 from top to bottom) of 4 mLs each were carefully removed. Fraction 10 consisted of the remaining 2 mLs of the working solution. The pellet was resuspended in 1 mL of PBS and designated as fraction 11. 2.4. M. tuberculosis viability The number of viable organisms was determined by serial dilution of the samples and plating on 10% OADC-enriched 7H11 agar plates [23]. The number of colony forming units (CFU) were calculated to determine the number of viable bacilli in each step of the protocol. 2.5. Combined immunofluorescence and auramine-rhodamine staining PBS (1.3 mL) was added to a 200 mL aliquot from each fraction, which was then centrifuged at 14,000 g for 15 min at 4 C. The pellets were washed twice more in PBS as discussed above to remove the percoll. The washed pellets were suspended and fixed in 50 mL 4% paraformaldehyde (Electron Microscopy Supplies) in PBS, washed with PBS and water, and suspended in 5 mL water, which was spotted on a glass slide and allowed to dry. The samples were then treated with lysozyme (1 mg/mL, SigmaeAldrich) and achromopeptidase (30 U/mL, SigmaeAldrich) in 10 mM Tris (pH 7.5) at 37 C for 40 min. Antigen retrieval was performed using the Retriever™ 2100 at 121 C for 15 min using Target Retrieval Buffer solution (DAKO). Blocking was performed with 1% goat serum in PBS (Biomeda) for 30 min. The slides were then incubated at 4 C overnight with rabbit polyclonal anti-TB whole cell lysate minus LAM (NR-13820, BEI Resources) [7]. Subsequently, the slides were washed using PBS and probed with goat anti-rabbit IgG (Invitrogen) labeled with Alexafluor 488. The slides were washed with PBS, mounted with ProLong® Gold antifade reagent (Invitrogen) and photographed using a Nikon Ti-E equipped with a Nikon DSQi1Mc camera. The coverslips were removed and the slides were further stained using a TB Fluorescent Stain Kit T containing a combination of auramine O and rhodamine B (Becton, Dickinson and Company) per manufacturer's instructions. The slides were then remounted and photographed [7].
527
and other analyses for comparison with in vitro grown bacilli. Guinea pigs were chosen as the initial model animal since disease progression and lung pathology have been well characterized and they develop granulomas that are similar to those found in human disease [31]. Guinea pigs were infected M. tuberculosis H37Rv via the aerosol route as previously described and lungs excised at 30 days post infection were used for the bacilli isolation. Figure 1 shows the gross pathology of low-dose aerosol infected guinea pig lung. Multiple granulomatous lesions were visible on the surface of the lungs at this time point and the total lung burden is approximately 106 CFU. 3.1. Enumerating CFU to track the yield of viable M. tuberculosis organisms During the procedure development, it was important to track the yield of bacilli at each step, thus, several methods were used including CFU enumeration, qRT-PCR and immunofluorimetry. However, due to the low load of bacilli in the lungs and heavy lung tissue contamination, qRT-PCR and immunofluorimetry proved to be unreliable. Thus, although time consuming, we adopted CFU enumeration for calculating the numbers of viable bacilli at each step. 3.2. Optimization of the homogenization and enzyme digestion Due to the large amount of host tissue relative to the number of bacilli in an infected guinea pig lung, we explored a number of options to optimize homogenization and reduce host tissue contamination. Since the guinea pig lung has more connective tissue than armadillo liver or mouse lung, we ultimately incorporated two homogenization steps. The tissue was first transferred into homogenization tubes and PBS containing collagenase, DNase I and CaCl2 was added. The tissue was homogenized in two sequential procedures, as discussed above, to a very smooth and even homogenate with no observable loss of viable bacilli. DNase I was routinely used to reduce the viscosity caused by the DNA content in cell lysates after homogenization and had no significant effect on bacterial viability [32]. Since collagenase is commonly used to isolate primary lung cells for culture [33], we initially investigated using this enzyme to debulk the homogenate. After two rounds of collagenase digestion, we utilized low specificity proteases to enhance the reduction of the protein mass in the homogenate. For this purpose we utilized either proteinase K or the
3. Results and discussion This study was undertaken to establish a robust methodology to isolate M. tuberculosis bacilli from infected lungs to enable cell-wall
Figure 1. Excised lungs from M. tuberculosis H37Rv infected guinea pig at 30 days post low-dose aerosol infection.
528
L. Shi et al. / Tuberculosis 94 (2014) 525e530
nonspecific pronase E. Proteinase K, a serine protease that cleaves peptide bonds primarily after the carboxyl group of N-substituted, hydrophobic aliphatic and aromatic amino acids, has been widely used in mitochondria isolation and nucleic acid purification [34,35]. The pronase E used in this study, a mixture of five serine proteases, two zinc endopeptidases, two zinc leucine aminopeptidases and one zinc carboxypeptidase, has no specificity and is useful for reducing proteins to single amino acids. Pronase was previously shown to be much more effective in digestion of casein than trypsin, chymotrypsin and several other proteases [36,37]. In order to optimize the enzyme digestion process for the maximal bacilli release from the tissue homogenate, multiple combinations of enzyme concentrations, digestion times and temperatures were tested. Effectiveness of digestion was based on visual observation of the residual pellet size after centrifugation and recovery of CFU. Based on these experiments, it was clear that after collagenase digestion, the two-step pronase E digestion at a final concentration of 0.5 mg/mL provided the best reduction in the size of the pellet after centrifugation and acceptable recovery of CFU (data not shown). Thus, this treatment was adopted.
3.3. Purification of M. tuberculosis by percoll gradient centrifugation Although the enzymatic treatment dramatically reduced the amount of contaminating host tissue, large amounts of tissue debris, relative to the amount of bacilli, remained. Therefore, we used a percoll gradient to further purify the bacilli. Percoll spontaneously forms a density gradient when centrifuged and has been widely used to separate cells, subcellular particles and viruses under gentle, physiological conditions [38e40]. Initially, we compared the separation of bacilli from infected lungs in a percoll gradient with M. tuberculosis H37Rv/pCherry3 grown in vitro. M. tuberculosis/pCherry3 was used as a control because the expressed fluorescent protein (a far-red reporter, mCherry) was readily visualized [22]. After centrifugation, the M. tuberculosis/pCherry3 bacilli were observed in a tight band in fraction 9, which contained 82% of the bacilli based on viability tracking as shown in Figure 2. For the separation of bacilli from infected lungs, the debris from the treated host tissue remained at top of the percoll gradient in fraction 1 containing only 4% bacilli, and 82% the bacilli were located in fraction 9.
Figure 3. The optimal procedure for the isolation and purification of M. tuberculosis in vivo from infected animal lungs.
In order to check the purity of the purified bacilli, bicinchoninic acid (BCA) protein assays were used to compare the protein content between the homogenate and fraction 9. The results showed that there was more than 500 fold purification achieved based on the protein content of the guinea pig lung homogenate versus fraction 9. 3.4. Final protocol The final protocol for isolation of bacilli involves three steps as indicated in Figure 3. First, the tissue is thoroughly homogenized in two sequential procedures using a Tissue-Tearor and followed by a Glas-Col motor driver homogenizer (Figure 4(A)). Second, a four step enzymatic digestion is utilized to reduce the bulk of host tissue (Figure 4(B)). Third, contamination by the host tissue debris is reduced by density gradient centrifugation to yield isolated bacilli (Figure 4(C and D)). This procedure takes place over a two-day period in BSL3 containment facilities. Homogenization and enzyme digestion are performed on the first day. The centrifuged pellets are stored at 4 C overnight and the process is completed on the second day. 3.5. Isolation and purification of M. tuberculosis bacilli from infected lungs Using this method, we isolated and purifed M. tuberculosis bacilli from three high-dose H37Rv infected guinea pig lungs and
Figure 2. Density gradient centrifugation of viable M. tuberculosis grown in culture and isolated from guinea pig lungs. M. tuberculosis/pCherry3 (in vitro) was cultured on 7H11 supplemented with 10% OADC plus hygromycin at 50 mg/mL for 3 weeks at 37 C. The in vivo bacilli were the mass from the infected lungs after homogenization and enzymatic digestion.
Figure 4. Appearance of lung tissue homogenate in the three steps in the bacilli isolation protocol. (A) The two half lung's is thoroughly homogenized. (B) A four step enzymatic digestion is utilized to reduce the bulk of host tissue. (C) Contamination by the host tissue debris is reduced by density gradient centrifugation to yield isolated bacilli. (D) After centrifugation at 30,000 g for 30 min at 4 C.
L. Shi et al. / Tuberculosis 94 (2014) 525e530
529
Table 1 The number of viable organisms after homogenization and enzyme digestion using three H37Rv infected guinea pigs lungs and GKO mouse lungs. Viable bacilli
Guinea pig lung 1
In homogenate (CFU) After enzyme digestion (CFU) Total recovery (%)
GKO mouse lung
2 7
1.4 10 8.6 106 61
Ave ± SEM
3 6
6.4 10 3.6 106 56
6
6.0 10 4.8 106 80
6
1 6
8.8 10 ± 2.6 10 5.6 106 ± 1.6 106 66 ± 7
Figure 5. Comparison of the percentage of viable M. tuberculosis organisms in different fractions when using the optimal procedure to isolate and purify M. tuberculosis bacilli in vivo from M. tuberculosis H37Rv infected guinea pig lungs and GKO mouse lungs.
2 6
1.1 10 5.3 105 48
Ave ± SEM
3 5
7.0 10 3.6 105 51
5
6.1 10 2.8 105 46
8.0 105 ± 1.5 105 3.9 105 ± 0.7 105 48 ± 1
three GKO mice lungs respectively. As shown in Table 1, after homogenization and enzyme digestion, 66% of viable organisms were recovered from the guinea pig lungs and 48% from the GKO mouse lungs. After gradient centrifugation, most of the purified bacilli were found in fractions 9 and 10 (Figure 5). For guinea pig lungs, 53% of the CFU were found in fraction 9, and 33% in fraction 10. Consistently, for GKO mouse lungs, 58% of the CFU were found in fraction 9 and 20% in fraction 10. Fraction 1 was grossly contaminated with host tissue, while fractions 9 and 10 gave much cleaner bacilli (Figure 6). These results suggested the established method for isolation and purification of M. tuberculosis from infected lungs is not only optimal for guinea pig model, but is applicable to the GKO mouse model and potentially other models of infection. In addition to using CFU enumeration to track the bacterial distribution in different fractions, a dual-staining method developed by Ryan et al. was adopted [7], which included fluorescent acid-fast auramine-rhodamine and immunofluorescence. As shown in Figure 6, there were three subpopulations of bacilli in vivo that were either labeled exclusively by immunofluorescence (green), exclusively by auramine-rhodamine (red) or by both techniques concurrently (yellow). These results are consistent with a previous study [7], in which the infected lungs were fixed, embedded in paraffin and then dewaxed for staining without any treatment described in our protocol. All fractions contained all three subpopulations of bacilli (Figure 6) previously described [7]. Thus, it is clear that the methodology presented here does not separate the subpopulations of the bacilli indicating that their densities are similar. As the protocol described here generated enough bacilli for analysis (106) structural analysis of mycobacterial cell-wall from bacilli purified from the host is ongoing. These studies involve applying microanalytical techniques [41] towards deciphering the cell-wall structure (mycolic acid, arabinogalactan and peptidoglycan) relative to that of M. tuberculosis H37Rv grown in vitro. This method provides an initial step in characterizing M. tuberculosis living in an infected host. A more ambitious goal will be to separate bacilli population revealed by multiple staining technique in order to understand the basis of the differential staining. Acknowledgment Acknowlegement: Bill and Melinda Gates Foundation, Grant# OPP 1034945. Funding: The funding source was Biil and Melinda Gates Foundation (BMGF). Competing interest: Ethical approval:
Figure 6. Micrograph of M. tuberculosis detected by immunofluorescence and auramine-rhodamine dual-staining for the samples in vivo from fractions 1 and 9. Extensive tissue contamination is visible, as diffuse autofluorescence, in fraction 1 and greatly reduced in fraction 9. All fractions containing bacilli contained all three populations of bacilli identified by differential staining and immunohistochemistry as indicated by the arrows.
None declared. Not required.
References [1] Williams C. Global tuberculosis control: WHO report 2011. Aust Nz J Publ Heal 2012;36:497e8. [2] Raviglione MC. The new stop TB strategy and the global plan to stop TB, 20062015. Bull World Health Organ 2007;85(5):327.
530
L. Shi et al. / Tuberculosis 94 (2014) 525e530
[3] Lienhardt C, Glaziou P, Uplekar M, Lonnroth K, Getahun H, Raviglione M. Global tuberculosis control: lessons learnt and future prospects. Nat Rev Microbiol 2012;10(6):407e16. [4] Silva Miranda M, Breiman A, Allain S, Deknuydt F, Altare F. The tuberculous granuloma: an unsuccessful host defence mechanism providing a safety shelter for the bacteria? Clin Dev Immunol 2012;2012:139127. [5] Via LE, Lin PL, Ray SM, Carrillo J, Allen SS, Eum SY, Taylor K, Klein E, Manjunatha U, Gonzales J, Lee EG, Park SK, Raleigh JA, Cho SN, McMurray DN, Flynn JL, Barry 3rd CE. Tuberculous granulomas are hypoxic in guinea pigs, rabbits, and nonhuman primates. Infect Immun 2008;76(6): 2333e40. € rg S, Krenn V, Morawietz L, [6] Seiler P, Ulrichs T, Bandermann S, Pradl L, Jo Kaufmann SH, Aichele P. Cell-wall alterations as an attribute of Mycobacterium tuberculosis in latent infection. J Infect Dis 2003;188:1326e31. [7] Ryan GJ, Hoff DR, Driver ER, Voskuil MI, Gonzalez-Juarrero M, Basaraba RJ, Crick DC, Spencer JS, Lenaerts AJ. Multiple M. tuberculosis phenotypes in mouse and guinea pig lung tissue revealed by a dual-staining approach. PLoS One 2010;5(6):e11108. [8] Cunningham AF, Spreadbury CL. Mycobacterial stationary phase induced by low oxygen tension: cell wall thickening and localization of the 16-kilodalton alpha-crystallin homolog. J Bacteriol 1998;180:801e8. [9] Fenton MJ, Vermeulen MW. Immunopathology of tuberculosis: roles of macrophages and monocytes. Infect Immun 1996;64(3):683e90. [10] Ulrichs T, Kaufmann SH. New insights into the function of granulomas in human tuberculosis. J Pathol 2006;208(2):261e9. [11] Betts JC, Lukey PT, Robb LC, McAdam RA, Duncan K. Evaluation of a nutrient starvation model of Mycobacterium tuberculosis persistence by gene and protein expression profiling. Mol Microbiol 2002;43(3):717e31. [12] Murphy DJ, Brown JR. Identification of gene targets against dormant phase Mycobacterium tuberculosis infections. BMC Infect Dis 2007;7:84. [13] Archuleta RJ, Yvonne Hoppes P, Primm TP. Mycobacterium avium enters a state of metabolic dormancy in response to starvation. Tuberculosis (Edinb) 2005;85(3):147e58. [14] Markova N, Slavchev G, Michailova L. Unique biological properties of Mycobacterium tuberculosis L-form variants: impact for survival under stress. Int Microbiol 2012;15(2):61e8. [15] Deb C, Lee CM, Dubey VS, Daniel J, Abomoelak B, Sirakova TD, Pawar S, Rogers L, Kolattukudy PE. A novel in vitro multiple-stress dormancy model for Mycobacterium tuberculosis generates a lipid-loaded, drug-tolerant, dormant pathogen. PLoS One 2009;4(6):e6077. [16] Wayne LG, Hayes LG. An in vitro model for sequential study of shiftdown of Mycobacterium tuberculosis through two stages of nonreplicating persistence. Infect Immun 1996;64(6):2062e9. [17] Wayne LG, Sohaskey CD. Nonreplicating persistence of Mycobacterium tuberculosis. Annu Rev Microbiol 2001;55:139e63. [18] Fang X, Wallqvist A, Reifman J. Modeling phenotypic metabolic adaptations of Mycobacterium tuberculosis H37Rv under hypoxia. PLoS Comput Biol 2012;8(9):e1002688. [19] Bhat SA, Singh N, Trivedi A, Kansal P, Gupta P, Kumar A. The mechanism of redox sensing in Mycobacterium tuberculosis. Free Radic Biol Med 2012;53(8): 1625e41. [20] Heng YX, Seah PG, Siew JY, Tay HC, Singhal A, Mathys V, Kiass M, Bifani P, M. Mycobacterium tuberculosis infection induces hypoxic lung Dartois V, Herve lesions in the rat. Tuberculosis (Edinb) 2011;91(4):339e41. [21] Bloch H, Segal W. Biochemical differentiation of Mycobacterium tuberculosis grown in vivo and in vitro. J Bacteriol 1956;72(2):132e41.
M, Draper P. The envelope layers of mycobacteria with reference to their [22] Daffe pathogenicity. Adv Microb Physiol 1998;39:131e203. [23] Draper P. Purification of M. leprae. World Health Organization document TDR/ IMMLEP-SWG (5)/80.3. In: Annex 4 of report of the fifth meeting of the Scientific Working Group on the Immunology of Leprosy (IMMLEP), Geneva, 24 to 26 June 1980. Geneva, Switzerland: World Health Organization; 1980. [24] Draper P, Kandler O, Darbre A. Peptidoglycan and arabinogalactan of Mycobacterium leprae. J Gen Microbiol 1987;133(5):1187e94. [25] Carroll P, Schreuder LJ, Muwanguzi-Karugaba J, Wiles S, Robertson BD, Ripoll J, Ward TH, Bancroft GJ, Schaible UE, Parish T. Sensitive detection of gene expression in mycobacteria under replicating and non-replicating conditions using optimized far-red reporters. PLoS One 2010;5(3):e9823. [26] Keyser A, Troudt JM, Taylor JL, Izzo AA. BCG sub-strains induce variable protection against virulent pulmonary Mycobacterium tuberculosis infection, with the capacity to drive Th2 immunity. Vaccine 2011;29(50):9308e15. [27] Cooper AM, Dalton DK, Stewart TA, Griffin JP, Russell DG, Orme IM. Disseminated tuberculosis in interferon-g gene-disrupted mice. J Exp Med 1993;178(6):2243e7. [28] Lenaerts AJ, Gruppo V, Brooks JV, Orme IM. Rapid in vivo screening of experimental drugs for tuberculosis using gamma interferon gene-disrupted mice. Antimicrob Agents Chemother 2003;47(2):783e5. [29] Kelly BP, Furney SK, Jessen MT, Orme IM. Low-dose aerosol infection model for testing drugs for efficacy against Mycobacterium tuberculosis. Antimicrob Agents Chemother 1996;40(12):2809e12. [30] Ordway DJ, Orme IM. Animal models of mycobacteria infection. Curr Protoc Immunol 2011 [Chapter 19:Unit19.5]. [31] McMurray DN. Guinea pig model of tuberculosis. Washington, DC. In: Bloom BR, editor. Tuberculosis: pathogenesis, protection and control; 1994. pp. 135e47. [32] Ordway DJ, Henao-Tamayo M, Shanley C, Smith EE, Palanisamy G, Wang B, Basaraba RJ, Orme IM. Influence of Mycobacterium bovis BCG vaccination on cellular immune response of guinea pigs challenged with Mycobacterium tuberculosis. Clin Vaccine Immunol 2008;15(8):1248e58. [33] Summer R, Kotton DN, Sun X, Ma B, Fitzsimmons K, Fine A. Side population cells and Bcrp1 expression in lung. Am J Physiol Lung Cell Mol Physiol 2003;285:L97e104. [34] Herrmann BG, Frischauf AM. Isolation of genomic DNA. Methods Enzymol 1987;152:180e3. [35] Ebeling W, Hennrich N, Klockow M, Metz H, Orth HD, Lang H. Proteinase K from Tritirachium album Limber. Eur J Biochem 1974;47:91e7. [36] Nomoto M, Narahashi Y, Murakami M. A proteolytic enzyme of Streptomyces griseus: VI. hydrolysis of protein by Streptomyces griseus protease. J Biochem 1960;48:593e602. [37] Nomoto M, Narahashi Y, Murakami M. A proteolytic enzyme of Streptomyces griseus: VII. substrate specificity of Streptomyces griseus protease. J Biochem 1960;48:906e18. [38] Lindley ER, Pisoni RL. Demonstration of adenosine deaminase activity in human fibroblast lysosomes. Biochem J 1993;290(Pt 2):457e62. [39] Kominami E, Ezaki J, Muno D, Ishido K, Ueno T, Wolfe LS. Specific storage of subunit C of mitochondrial ATP synthase in lysosomes of neuronal ceroid lipofuscinosis (Batten's disease). J Biochem 1992;111(2):278e82. n R, Olofsson S, Svantesson G, [40] Svennerholm B, Vahlne A, Jeansson S, Lunde Lycke E. Separation of herpes-simplex virus virions and nucleocapsids on percoll gradients. J Virol Methods 1980;1(6):303e9. [41] Bhamidi S, Shi L, Chatterjee D, Belisle JT, Crick DC, McNeil MR. A bioanalytical method to determine the cell wall composition of Mycobacterium tuberculosis grown in vivo. Anal Biochem 2012;421(1):240e9.
Contents lists available at ScienceDirect
Tuberculosis journal homepage: http://intl.elsevierhealth.com/journals/tube
Isolation and purification of Mycobacterium tuberculosis from H37Rv infected guinea pig lungs Libin Shi, Gavin J. Ryan, Suresh Bhamidi, JoLynn Troudt, Anita Amin, Angelo Izzo, Anne J. Lenaerts, Michael R. McNeil, John T. Belisle, Dean C. Crick**, Delphi Chatterjee* Mycobacteria Research Laboratories, Department of Microbiology, Immunology and Pathology, Colorado State University, Fort Collins, CO 80523, USA
a r t i c l e i n f o
s u m m a r y
Article history: Received 15 May 2014 Accepted 23 May 2014
Evidence suggests that Mycobacterium tuberculosis grown in vivo may have a different phenotypic structure from its in vitro counterpart. In order to study the differences between in vivo and in vitro grown bacilli, it is important to establish a reliable method for isolating and purifying M. tuberculosis from infected tissue. In this study, we developed an optimal method to isolate bacilli from the lungs of infected guinea pigs, which was also shown to be applicable to the interferon-g gene knockout mouse model. Briefly, 1) the infected lungs were thoroughly homogenized; 2) a four step enzymatic digestion was utilized to reduce the bulk of the host tissue using collagenase, DNase I and pronase E; 3) residual contamination by the host tissue debris was successfully reduced using percoll density gradient centrifugation. These steps resulted in a protocol such that relatively clean, viable bacilli can be isolated from the digested host tissue homogenate in about 50% yield. These bacilli can further be used for analytical studies of the more stable cellular components such as lipid, peptidoglycan and mycolic acid. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Mycobacterium tuberculosis Bacterial isolation Guinea pig infection
1. Introduction Tuberculosis (TB) is an airborne infectious disease that is both preventable and curable; however, it continues to be a major cause of death worldwide. In 2012, there were an estimated 8.6 million new cases globally, and 1.3 million people died from TB including 320,000 deaths in HIV-infected people [1]. The WHO has set a goal to reduce the global burden of TB dramatically by 2015 [2], therefore, development of new tools for optimal diagnosis, prevention and treatment of TB is urgent [3]. However, in terms of identifying novel diagnosis and validated drug targets, most of the present knowledge of the tubercule bacillus is derived from the bacilli grown in vitro, which may not accurately represent organisms living in a host lung. Mycobacterium tuberculosis is typically acquired by inhalation of aerosolized droplets from infected individuals resulting in the infection of alveolar macrophages, which release various cytokines resulting in the formation of granulomas, a hallmark of TB disease [4]. M. tuberculosis is an obligate aerobic organism, however, it adapts to
* Corresponding author. Tel.: þ1 970 491 7495; fax: þ1 970 491 1815. ** Corresponding author. Tel.: þ1 970 491 3308; fax: þ1 970 491 1815. E-mail addresses: [email protected] (D.C. Crick), delphi.chatterjee@ colostate.edu (D. Chatterjee). http://dx.doi.org/10.1016/j.tube.2014.05.003 1472-9792/© 2014 Elsevier Ltd. All rights reserved.
oxygen limitation during the development of granulomas as they become relatively hypoxic compared to the surrounding tissue in guinea pig, rabbit, and nonhuman primate models of disease [5], and the bacilli appear to change their physical characteristics in response to stimuli encountered in the host microenviroment. Seiler et al. [6] reported that with persistence of M. tuberculosis infection in mice and human patients, the classical cell-wall component-dependent Ziehl-Neelsen (ZN) staining gradually reduced while the detection by cell-wall independent staining using a polyclonal anti-Mycobacterium bovis BCG serum continued, suggesting that the cell-wall of M. tuberculosis was either lost or reorganized. Recently, applying a dual-staining method including fluorescent acid-fast auraminerhodamine targeting the mycolic acid containing cell-wall followed by immunofluorescence targeting bacterial proteins using an antiM. tuberculosis whole cell lysate polyclonal antibody, Ryan et al. [7] observed three M. tuberculosis populations in hypoxic culture; similar subpopulations of bacilli were also found in the lungs from M. tuberculosis infected mice and guinea pigs. These three populations appear to be either exclusively acid-fast positive, exclusively immunofluorescent positive or acid-fast and immunofluorescent positive, a result suggesting that multiple populations of M. tuberculosis may exist in a single microenvironment. Using transmission electron microscopy, Cunningham et al. [8] showed that microaerobically and anaerobically cultured
526
L. Shi et al. / Tuberculosis 94 (2014) 525e530
M. tuberculosis bacilli have thicker cell-walls than aerobically cultured bacilli. The cell-wall thickening could consist of polysaccharides and lipids and may help the bacilli to survive in oxygen-deficient microenvironment in vivo. In addition to hypoxia, M. tuberculosis bacilli also encounter low pH and low nutrient availablility within the caseous tissue in the development of granulomas [9,10]. Under nutrient starvation conditions, there is a global metabolic shift and lipids are thought to become the sole source of energy for mycobacteria [11e13]. Recently, Markova et al. [14] reported L-form variants of M. tuberculosis under nutrient starvation stress, which lost their acidfast character and changed their morphology. Single-stress conditions such as hypoxia and nutrient starvation provide information regarding the physiological changes that occur during TB development, but cannot fully simulate conditions found in the granuloma. Deb et al. [15] adopted an in vitro multiple-stress condition including low oxygen (5%), high CO2 (10%), low nutrient (10% Dubos medium) and acidic pH (5.0) in which M. tuberculosis stopped replicating, lost acid-fastness, and accumulated triacylglycerol and wax ester. Collectively, these results suggested that the bacilli under stress may alter their cell-wall, and potentially their metabolic status to enter a state of nonreplicating persistence and phenotypic drug resistance [16e20]. Ideally, the changes induced in the physiology and biochemistry of the bacilli during disease progression would best be studied in bacteria isolated from the lungs of infected hosts. To date very little data has been published regarding the physiological and metabolic state of M. tuberculosis isolated from host tissue; although there are striking differences reported in the metabolic characteristics with respect to substrate response between tubercle bacilli grown in culture and their counterparts isolated from moribund infected mice [21]. In this study, we aimed at developing a method to isolate M. tuberculosis from the lungs of infected guinea pigs that surmount the significant technical difficulties involved in reaching this goal. Initially, we attempted to adopt a protocol reported for isolation and purification of Mycobacterium leprae from armadillo livers, which has been widely used for studies of M. leprae genomics, proteomics, lipidomics and cell-wall structure [22e24]. Unfortunately, this method of isolation, which was developed to provide vaccine grade material, failed to give a desirable yield of viable bacilli when applied to M. tuberculosis infected lungs. The fact that guinea pig lungs contained relatively low numbers of bacilli, ~1056 bacilli/g of lung, relative to 10910 bacilli/g of M. leprae in infected armadillo liver possibly contributed to the problems associated with the isolation. In addition, the sodium hydroxide treatment resulted in the death of more than 90% of the bacilli making CFU enumeration for bacilli tracking impossible, the protein digestion did not efficiently liberate bacilli from homogenized lung tissue and it was difficult to perform biphasic partitioning between polyethylene glycol and dextran solution in a Biosafety Level 3 (BSL3) environment. A protocol previously described by Bolch et al. [21] to isolate M. tuberculosis H37Rv from the infected mice was also investigated. This protocol resulted in incomplete release of bacilli from the tissue and contamination with sucrose and mitochondria complicated downstream analysis. The low yields of viable bacilli and severe host tissue contamination suggested that neither method was optimal for isolating M. tuberculosis from infected guinea pig lungs. Thus, a new facile and reproducible method was developed for isolating M. tuberculosis bacilli from host tissue in high yield and purity. 2. Methods and materials 2.1. Bacteria and plasmid M. tuberculosis H37Rv (Trudeau Institute) was grown from lowpassage seed lots in Proskauer-Beck liquid medium containing
0.05% Tween 80 to early mid-log phase and frozen in aliquots at 80 C until needed. Cultures were diluted in sterile water prior to use. Plasmid pCherry3 [25], which was kindly provided by Dr. Tanya Parish, was electroporated into M. tuberculosis H37Rv and cultured on 7H11 supplemented (Becton, Dickinson and Company) with 10% BBL Middle Brook OADC enrichment (BD) plus hygromycin (Streptomyces sp., Calbiochem) at 50 mg/mL for 3 weeks at 37 C. 2.2. Animal infection All experimental protocols were approved by the Animal Care and Usage Committee of Colorado State University. Female outbred Hartley guinea pigs (approximately 500 g in weight) were purchased from the Charles River Laboratories (North Wilmington) and held under barrier conditions in a BSL3 animal laboratory. Guinea pigs were infected with 10e20 CFU of M. tuberculosis H37Rv via the low-dose aerosol route as previously described [26], or 100 CFU of bacilli via high-dose aerosol infection. The guinea pigs were sacrificed 30 days after infection, lungs were rapidly excised, immediately frozen in liquid nitrogen and stored at 80 C. Eight- to ten-week-old female specific pathogen-free C57BL/6Ifngtm1ts (GKO) mice were purchased (Jackson Laboratories). These mice have a disrupted interferon-g gene, which renders them highly susceptible to TB [27]. The standard infection for this model was performed as previously described [28]. Briefly, mice were exposed to a low-dose aerosol infection with M. tuberculosis using a Glas-Col inhalation exposure system (Glas-Col LLC.) [29]. One day after low-dose aerosol infection, three mice were euthanized to verify infection with 50e100 CFU per mouse. The mice were sacrificed after 18 days exposure, and lungs were excised and stored frozen as mentioned above. 2.3. Isolation and purification of M. tuberculosis from lungs 2.3.1. Homogenization of infected guinea pig lungs Frozen guinea pig lungs were thawed at room temperature (RT) and weighed. The lungs were rinsed with phosphate buffered saline (PBS, pH 7.2, without CaCl2 or MgCl2, Gibco) and half of the total lung mass from each guinea pig was placed in a 40 mL straight walled homogenization tube (Glas-Col LLC.). PBS containing 1 mg/ mL collagenase (Clostridium histolyticum, type XI, SigmaeAldrich), 30 mg/mL deoxyribonuclease I (DNase I, from bovine pancreas, type IV, SigmaeAldrich) and 5 mM CaCl2 was added at 3 mL per gram of lung tissue. Lungs were initially homogenized using a Tissue-Tearor (Kinematic Polytron) followed by further homogenization using a variable speed motor drive homogenizer and teflon pestle (Glas-Col LLC.) at 4000 rpm for 15 passes [30]. 2.3.2. Enzyme digestion of the homogenized lungs To obtain the maximal bacilli release from the tissue homogenate, multiple enzyme treatments were tested and four alternative conditions of enzyme digestion were adopted, which utilized collagenase followed by either proteinase K (recombinant, Roche) or pronase E (protease from Streptomyces griseus, type XIV, SigmaeAldrich). Thus, twelve low-dose aerosol infected guinea pig lungs were divided into four groups and were homogenized as described above. The homogenates were exposed to two rounds of collagenase digestion in all four protocols. This included an initial incubation with collagenase at 1 mg/mL at 37 C for 1 h, followed by addition of more collagenase to 2 mg/mL final concentration and incubation at RT for 2 more hours. Following the collagenase digestion, either proteinase K or pronase E was added. In one treatment, proteinase K was added to 0.5 mg/mL and the mixture was incubated at RT for 2 h. After centrifugation at 10,000 g for
L. Shi et al. / Tuberculosis 94 (2014) 525e530
15 min at 4 C, the pellets were resuspened in 0.5 mg/mL proteinase K for 1 more hour incubation at RT. In the second treatment, proteinase K was added to 0.5 mg/mL and the flask was incubated at 4 C for overnight. In the third and fourth treatments, pronase E was added in place of proteinase K and incubated as described. In each case, the digested homogenate was centrifuged at 10,000 g for 15 min at 4 C to obtain pellets containing the bacilli. 2.3.3. Density gradient centrifugation A working solution of percoll with a density of 1.07 g/mL was prepared according to the manufacturer's protocol (Amersham Biosciences). Stock isotonic percoll (SIP) was prepared by adding 9 parts of percoll to 1 part of 10 PBS (v/v, ri ¼ 1.123 g/mL), and then SIP was diluted to working solution (ri ¼ 1.07 g/mL) by adding 60.8 mL PBS to 75 mL SIP. The pellets obtained from enzymatic digests (Section 2.3.2.) were suspended in 3 mL PBS and homogenized using the GlasCol homogenizer as described above and gently layered on top of 35 mL working solution of percoll in 50 mL polycarbonate centrifuge tubes (Beckman). After centrifugation at 30,000 g for 30 min at 4 C with no deacceleration, nine fractions (numbered 1e9 from top to bottom) of 4 mLs each were carefully removed. Fraction 10 consisted of the remaining 2 mLs of the working solution. The pellet was resuspended in 1 mL of PBS and designated as fraction 11. 2.4. M. tuberculosis viability The number of viable organisms was determined by serial dilution of the samples and plating on 10% OADC-enriched 7H11 agar plates [23]. The number of colony forming units (CFU) were calculated to determine the number of viable bacilli in each step of the protocol. 2.5. Combined immunofluorescence and auramine-rhodamine staining PBS (1.3 mL) was added to a 200 mL aliquot from each fraction, which was then centrifuged at 14,000 g for 15 min at 4 C. The pellets were washed twice more in PBS as discussed above to remove the percoll. The washed pellets were suspended and fixed in 50 mL 4% paraformaldehyde (Electron Microscopy Supplies) in PBS, washed with PBS and water, and suspended in 5 mL water, which was spotted on a glass slide and allowed to dry. The samples were then treated with lysozyme (1 mg/mL, SigmaeAldrich) and achromopeptidase (30 U/mL, SigmaeAldrich) in 10 mM Tris (pH 7.5) at 37 C for 40 min. Antigen retrieval was performed using the Retriever™ 2100 at 121 C for 15 min using Target Retrieval Buffer solution (DAKO). Blocking was performed with 1% goat serum in PBS (Biomeda) for 30 min. The slides were then incubated at 4 C overnight with rabbit polyclonal anti-TB whole cell lysate minus LAM (NR-13820, BEI Resources) [7]. Subsequently, the slides were washed using PBS and probed with goat anti-rabbit IgG (Invitrogen) labeled with Alexafluor 488. The slides were washed with PBS, mounted with ProLong® Gold antifade reagent (Invitrogen) and photographed using a Nikon Ti-E equipped with a Nikon DSQi1Mc camera. The coverslips were removed and the slides were further stained using a TB Fluorescent Stain Kit T containing a combination of auramine O and rhodamine B (Becton, Dickinson and Company) per manufacturer's instructions. The slides were then remounted and photographed [7].
527
and other analyses for comparison with in vitro grown bacilli. Guinea pigs were chosen as the initial model animal since disease progression and lung pathology have been well characterized and they develop granulomas that are similar to those found in human disease [31]. Guinea pigs were infected M. tuberculosis H37Rv via the aerosol route as previously described and lungs excised at 30 days post infection were used for the bacilli isolation. Figure 1 shows the gross pathology of low-dose aerosol infected guinea pig lung. Multiple granulomatous lesions were visible on the surface of the lungs at this time point and the total lung burden is approximately 106 CFU. 3.1. Enumerating CFU to track the yield of viable M. tuberculosis organisms During the procedure development, it was important to track the yield of bacilli at each step, thus, several methods were used including CFU enumeration, qRT-PCR and immunofluorimetry. However, due to the low load of bacilli in the lungs and heavy lung tissue contamination, qRT-PCR and immunofluorimetry proved to be unreliable. Thus, although time consuming, we adopted CFU enumeration for calculating the numbers of viable bacilli at each step. 3.2. Optimization of the homogenization and enzyme digestion Due to the large amount of host tissue relative to the number of bacilli in an infected guinea pig lung, we explored a number of options to optimize homogenization and reduce host tissue contamination. Since the guinea pig lung has more connective tissue than armadillo liver or mouse lung, we ultimately incorporated two homogenization steps. The tissue was first transferred into homogenization tubes and PBS containing collagenase, DNase I and CaCl2 was added. The tissue was homogenized in two sequential procedures, as discussed above, to a very smooth and even homogenate with no observable loss of viable bacilli. DNase I was routinely used to reduce the viscosity caused by the DNA content in cell lysates after homogenization and had no significant effect on bacterial viability [32]. Since collagenase is commonly used to isolate primary lung cells for culture [33], we initially investigated using this enzyme to debulk the homogenate. After two rounds of collagenase digestion, we utilized low specificity proteases to enhance the reduction of the protein mass in the homogenate. For this purpose we utilized either proteinase K or the
3. Results and discussion This study was undertaken to establish a robust methodology to isolate M. tuberculosis bacilli from infected lungs to enable cell-wall
Figure 1. Excised lungs from M. tuberculosis H37Rv infected guinea pig at 30 days post low-dose aerosol infection.
528
L. Shi et al. / Tuberculosis 94 (2014) 525e530
nonspecific pronase E. Proteinase K, a serine protease that cleaves peptide bonds primarily after the carboxyl group of N-substituted, hydrophobic aliphatic and aromatic amino acids, has been widely used in mitochondria isolation and nucleic acid purification [34,35]. The pronase E used in this study, a mixture of five serine proteases, two zinc endopeptidases, two zinc leucine aminopeptidases and one zinc carboxypeptidase, has no specificity and is useful for reducing proteins to single amino acids. Pronase was previously shown to be much more effective in digestion of casein than trypsin, chymotrypsin and several other proteases [36,37]. In order to optimize the enzyme digestion process for the maximal bacilli release from the tissue homogenate, multiple combinations of enzyme concentrations, digestion times and temperatures were tested. Effectiveness of digestion was based on visual observation of the residual pellet size after centrifugation and recovery of CFU. Based on these experiments, it was clear that after collagenase digestion, the two-step pronase E digestion at a final concentration of 0.5 mg/mL provided the best reduction in the size of the pellet after centrifugation and acceptable recovery of CFU (data not shown). Thus, this treatment was adopted.
3.3. Purification of M. tuberculosis by percoll gradient centrifugation Although the enzymatic treatment dramatically reduced the amount of contaminating host tissue, large amounts of tissue debris, relative to the amount of bacilli, remained. Therefore, we used a percoll gradient to further purify the bacilli. Percoll spontaneously forms a density gradient when centrifuged and has been widely used to separate cells, subcellular particles and viruses under gentle, physiological conditions [38e40]. Initially, we compared the separation of bacilli from infected lungs in a percoll gradient with M. tuberculosis H37Rv/pCherry3 grown in vitro. M. tuberculosis/pCherry3 was used as a control because the expressed fluorescent protein (a far-red reporter, mCherry) was readily visualized [22]. After centrifugation, the M. tuberculosis/pCherry3 bacilli were observed in a tight band in fraction 9, which contained 82% of the bacilli based on viability tracking as shown in Figure 2. For the separation of bacilli from infected lungs, the debris from the treated host tissue remained at top of the percoll gradient in fraction 1 containing only 4% bacilli, and 82% the bacilli were located in fraction 9.
Figure 3. The optimal procedure for the isolation and purification of M. tuberculosis in vivo from infected animal lungs.
In order to check the purity of the purified bacilli, bicinchoninic acid (BCA) protein assays were used to compare the protein content between the homogenate and fraction 9. The results showed that there was more than 500 fold purification achieved based on the protein content of the guinea pig lung homogenate versus fraction 9. 3.4. Final protocol The final protocol for isolation of bacilli involves three steps as indicated in Figure 3. First, the tissue is thoroughly homogenized in two sequential procedures using a Tissue-Tearor and followed by a Glas-Col motor driver homogenizer (Figure 4(A)). Second, a four step enzymatic digestion is utilized to reduce the bulk of host tissue (Figure 4(B)). Third, contamination by the host tissue debris is reduced by density gradient centrifugation to yield isolated bacilli (Figure 4(C and D)). This procedure takes place over a two-day period in BSL3 containment facilities. Homogenization and enzyme digestion are performed on the first day. The centrifuged pellets are stored at 4 C overnight and the process is completed on the second day. 3.5. Isolation and purification of M. tuberculosis bacilli from infected lungs Using this method, we isolated and purifed M. tuberculosis bacilli from three high-dose H37Rv infected guinea pig lungs and
Figure 2. Density gradient centrifugation of viable M. tuberculosis grown in culture and isolated from guinea pig lungs. M. tuberculosis/pCherry3 (in vitro) was cultured on 7H11 supplemented with 10% OADC plus hygromycin at 50 mg/mL for 3 weeks at 37 C. The in vivo bacilli were the mass from the infected lungs after homogenization and enzymatic digestion.
Figure 4. Appearance of lung tissue homogenate in the three steps in the bacilli isolation protocol. (A) The two half lung's is thoroughly homogenized. (B) A four step enzymatic digestion is utilized to reduce the bulk of host tissue. (C) Contamination by the host tissue debris is reduced by density gradient centrifugation to yield isolated bacilli. (D) After centrifugation at 30,000 g for 30 min at 4 C.
L. Shi et al. / Tuberculosis 94 (2014) 525e530
529
Table 1 The number of viable organisms after homogenization and enzyme digestion using three H37Rv infected guinea pigs lungs and GKO mouse lungs. Viable bacilli
Guinea pig lung 1
In homogenate (CFU) After enzyme digestion (CFU) Total recovery (%)
GKO mouse lung
2 7
1.4 10 8.6 106 61
Ave ± SEM
3 6
6.4 10 3.6 106 56
6
6.0 10 4.8 106 80
6
1 6
8.8 10 ± 2.6 10 5.6 106 ± 1.6 106 66 ± 7
Figure 5. Comparison of the percentage of viable M. tuberculosis organisms in different fractions when using the optimal procedure to isolate and purify M. tuberculosis bacilli in vivo from M. tuberculosis H37Rv infected guinea pig lungs and GKO mouse lungs.
2 6
1.1 10 5.3 105 48
Ave ± SEM
3 5
7.0 10 3.6 105 51
5
6.1 10 2.8 105 46
8.0 105 ± 1.5 105 3.9 105 ± 0.7 105 48 ± 1
three GKO mice lungs respectively. As shown in Table 1, after homogenization and enzyme digestion, 66% of viable organisms were recovered from the guinea pig lungs and 48% from the GKO mouse lungs. After gradient centrifugation, most of the purified bacilli were found in fractions 9 and 10 (Figure 5). For guinea pig lungs, 53% of the CFU were found in fraction 9, and 33% in fraction 10. Consistently, for GKO mouse lungs, 58% of the CFU were found in fraction 9 and 20% in fraction 10. Fraction 1 was grossly contaminated with host tissue, while fractions 9 and 10 gave much cleaner bacilli (Figure 6). These results suggested the established method for isolation and purification of M. tuberculosis from infected lungs is not only optimal for guinea pig model, but is applicable to the GKO mouse model and potentially other models of infection. In addition to using CFU enumeration to track the bacterial distribution in different fractions, a dual-staining method developed by Ryan et al. was adopted [7], which included fluorescent acid-fast auramine-rhodamine and immunofluorescence. As shown in Figure 6, there were three subpopulations of bacilli in vivo that were either labeled exclusively by immunofluorescence (green), exclusively by auramine-rhodamine (red) or by both techniques concurrently (yellow). These results are consistent with a previous study [7], in which the infected lungs were fixed, embedded in paraffin and then dewaxed for staining without any treatment described in our protocol. All fractions contained all three subpopulations of bacilli (Figure 6) previously described [7]. Thus, it is clear that the methodology presented here does not separate the subpopulations of the bacilli indicating that their densities are similar. As the protocol described here generated enough bacilli for analysis (106) structural analysis of mycobacterial cell-wall from bacilli purified from the host is ongoing. These studies involve applying microanalytical techniques [41] towards deciphering the cell-wall structure (mycolic acid, arabinogalactan and peptidoglycan) relative to that of M. tuberculosis H37Rv grown in vitro. This method provides an initial step in characterizing M. tuberculosis living in an infected host. A more ambitious goal will be to separate bacilli population revealed by multiple staining technique in order to understand the basis of the differential staining. Acknowledgment Acknowlegement: Bill and Melinda Gates Foundation, Grant# OPP 1034945. Funding: The funding source was Biil and Melinda Gates Foundation (BMGF). Competing interest: Ethical approval:
Figure 6. Micrograph of M. tuberculosis detected by immunofluorescence and auramine-rhodamine dual-staining for the samples in vivo from fractions 1 and 9. Extensive tissue contamination is visible, as diffuse autofluorescence, in fraction 1 and greatly reduced in fraction 9. All fractions containing bacilli contained all three populations of bacilli identified by differential staining and immunohistochemistry as indicated by the arrows.
None declared. Not required.
References [1] Williams C. Global tuberculosis control: WHO report 2011. Aust Nz J Publ Heal 2012;36:497e8. [2] Raviglione MC. The new stop TB strategy and the global plan to stop TB, 20062015. Bull World Health Organ 2007;85(5):327.
530
L. Shi et al. / Tuberculosis 94 (2014) 525e530
[3] Lienhardt C, Glaziou P, Uplekar M, Lonnroth K, Getahun H, Raviglione M. Global tuberculosis control: lessons learnt and future prospects. Nat Rev Microbiol 2012;10(6):407e16. [4] Silva Miranda M, Breiman A, Allain S, Deknuydt F, Altare F. The tuberculous granuloma: an unsuccessful host defence mechanism providing a safety shelter for the bacteria? Clin Dev Immunol 2012;2012:139127. [5] Via LE, Lin PL, Ray SM, Carrillo J, Allen SS, Eum SY, Taylor K, Klein E, Manjunatha U, Gonzales J, Lee EG, Park SK, Raleigh JA, Cho SN, McMurray DN, Flynn JL, Barry 3rd CE. Tuberculous granulomas are hypoxic in guinea pigs, rabbits, and nonhuman primates. Infect Immun 2008;76(6): 2333e40. € rg S, Krenn V, Morawietz L, [6] Seiler P, Ulrichs T, Bandermann S, Pradl L, Jo Kaufmann SH, Aichele P. Cell-wall alterations as an attribute of Mycobacterium tuberculosis in latent infection. J Infect Dis 2003;188:1326e31. [7] Ryan GJ, Hoff DR, Driver ER, Voskuil MI, Gonzalez-Juarrero M, Basaraba RJ, Crick DC, Spencer JS, Lenaerts AJ. Multiple M. tuberculosis phenotypes in mouse and guinea pig lung tissue revealed by a dual-staining approach. PLoS One 2010;5(6):e11108. [8] Cunningham AF, Spreadbury CL. Mycobacterial stationary phase induced by low oxygen tension: cell wall thickening and localization of the 16-kilodalton alpha-crystallin homolog. J Bacteriol 1998;180:801e8. [9] Fenton MJ, Vermeulen MW. Immunopathology of tuberculosis: roles of macrophages and monocytes. Infect Immun 1996;64(3):683e90. [10] Ulrichs T, Kaufmann SH. New insights into the function of granulomas in human tuberculosis. J Pathol 2006;208(2):261e9. [11] Betts JC, Lukey PT, Robb LC, McAdam RA, Duncan K. Evaluation of a nutrient starvation model of Mycobacterium tuberculosis persistence by gene and protein expression profiling. Mol Microbiol 2002;43(3):717e31. [12] Murphy DJ, Brown JR. Identification of gene targets against dormant phase Mycobacterium tuberculosis infections. BMC Infect Dis 2007;7:84. [13] Archuleta RJ, Yvonne Hoppes P, Primm TP. Mycobacterium avium enters a state of metabolic dormancy in response to starvation. Tuberculosis (Edinb) 2005;85(3):147e58. [14] Markova N, Slavchev G, Michailova L. Unique biological properties of Mycobacterium tuberculosis L-form variants: impact for survival under stress. Int Microbiol 2012;15(2):61e8. [15] Deb C, Lee CM, Dubey VS, Daniel J, Abomoelak B, Sirakova TD, Pawar S, Rogers L, Kolattukudy PE. A novel in vitro multiple-stress dormancy model for Mycobacterium tuberculosis generates a lipid-loaded, drug-tolerant, dormant pathogen. PLoS One 2009;4(6):e6077. [16] Wayne LG, Hayes LG. An in vitro model for sequential study of shiftdown of Mycobacterium tuberculosis through two stages of nonreplicating persistence. Infect Immun 1996;64(6):2062e9. [17] Wayne LG, Sohaskey CD. Nonreplicating persistence of Mycobacterium tuberculosis. Annu Rev Microbiol 2001;55:139e63. [18] Fang X, Wallqvist A, Reifman J. Modeling phenotypic metabolic adaptations of Mycobacterium tuberculosis H37Rv under hypoxia. PLoS Comput Biol 2012;8(9):e1002688. [19] Bhat SA, Singh N, Trivedi A, Kansal P, Gupta P, Kumar A. The mechanism of redox sensing in Mycobacterium tuberculosis. Free Radic Biol Med 2012;53(8): 1625e41. [20] Heng YX, Seah PG, Siew JY, Tay HC, Singhal A, Mathys V, Kiass M, Bifani P, M. Mycobacterium tuberculosis infection induces hypoxic lung Dartois V, Herve lesions in the rat. Tuberculosis (Edinb) 2011;91(4):339e41. [21] Bloch H, Segal W. Biochemical differentiation of Mycobacterium tuberculosis grown in vivo and in vitro. J Bacteriol 1956;72(2):132e41.
M, Draper P. The envelope layers of mycobacteria with reference to their [22] Daffe pathogenicity. Adv Microb Physiol 1998;39:131e203. [23] Draper P. Purification of M. leprae. World Health Organization document TDR/ IMMLEP-SWG (5)/80.3. In: Annex 4 of report of the fifth meeting of the Scientific Working Group on the Immunology of Leprosy (IMMLEP), Geneva, 24 to 26 June 1980. Geneva, Switzerland: World Health Organization; 1980. [24] Draper P, Kandler O, Darbre A. Peptidoglycan and arabinogalactan of Mycobacterium leprae. J Gen Microbiol 1987;133(5):1187e94. [25] Carroll P, Schreuder LJ, Muwanguzi-Karugaba J, Wiles S, Robertson BD, Ripoll J, Ward TH, Bancroft GJ, Schaible UE, Parish T. Sensitive detection of gene expression in mycobacteria under replicating and non-replicating conditions using optimized far-red reporters. PLoS One 2010;5(3):e9823. [26] Keyser A, Troudt JM, Taylor JL, Izzo AA. BCG sub-strains induce variable protection against virulent pulmonary Mycobacterium tuberculosis infection, with the capacity to drive Th2 immunity. Vaccine 2011;29(50):9308e15. [27] Cooper AM, Dalton DK, Stewart TA, Griffin JP, Russell DG, Orme IM. Disseminated tuberculosis in interferon-g gene-disrupted mice. J Exp Med 1993;178(6):2243e7. [28] Lenaerts AJ, Gruppo V, Brooks JV, Orme IM. Rapid in vivo screening of experimental drugs for tuberculosis using gamma interferon gene-disrupted mice. Antimicrob Agents Chemother 2003;47(2):783e5. [29] Kelly BP, Furney SK, Jessen MT, Orme IM. Low-dose aerosol infection model for testing drugs for efficacy against Mycobacterium tuberculosis. Antimicrob Agents Chemother 1996;40(12):2809e12. [30] Ordway DJ, Orme IM. Animal models of mycobacteria infection. Curr Protoc Immunol 2011 [Chapter 19:Unit19.5]. [31] McMurray DN. Guinea pig model of tuberculosis. Washington, DC. In: Bloom BR, editor. Tuberculosis: pathogenesis, protection and control; 1994. pp. 135e47. [32] Ordway DJ, Henao-Tamayo M, Shanley C, Smith EE, Palanisamy G, Wang B, Basaraba RJ, Orme IM. Influence of Mycobacterium bovis BCG vaccination on cellular immune response of guinea pigs challenged with Mycobacterium tuberculosis. Clin Vaccine Immunol 2008;15(8):1248e58. [33] Summer R, Kotton DN, Sun X, Ma B, Fitzsimmons K, Fine A. Side population cells and Bcrp1 expression in lung. Am J Physiol Lung Cell Mol Physiol 2003;285:L97e104. [34] Herrmann BG, Frischauf AM. Isolation of genomic DNA. Methods Enzymol 1987;152:180e3. [35] Ebeling W, Hennrich N, Klockow M, Metz H, Orth HD, Lang H. Proteinase K from Tritirachium album Limber. Eur J Biochem 1974;47:91e7. [36] Nomoto M, Narahashi Y, Murakami M. A proteolytic enzyme of Streptomyces griseus: VI. hydrolysis of protein by Streptomyces griseus protease. J Biochem 1960;48:593e602. [37] Nomoto M, Narahashi Y, Murakami M. A proteolytic enzyme of Streptomyces griseus: VII. substrate specificity of Streptomyces griseus protease. J Biochem 1960;48:906e18. [38] Lindley ER, Pisoni RL. Demonstration of adenosine deaminase activity in human fibroblast lysosomes. Biochem J 1993;290(Pt 2):457e62. [39] Kominami E, Ezaki J, Muno D, Ishido K, Ueno T, Wolfe LS. Specific storage of subunit C of mitochondrial ATP synthase in lysosomes of neuronal ceroid lipofuscinosis (Batten's disease). J Biochem 1992;111(2):278e82. n R, Olofsson S, Svantesson G, [40] Svennerholm B, Vahlne A, Jeansson S, Lunde Lycke E. Separation of herpes-simplex virus virions and nucleocapsids on percoll gradients. J Virol Methods 1980;1(6):303e9. [41] Bhamidi S, Shi L, Chatterjee D, Belisle JT, Crick DC, McNeil MR. A bioanalytical method to determine the cell wall composition of Mycobacterium tuberculosis grown in vivo. Anal Biochem 2012;421(1):240e9.