FEMS Microbiology Letters 205 (2001) 113^117
www.fems-microbiology.org
Novel use of guanidinium isothiocyanate in the
isolation of Mycobacterium tuberculosis DNA from
clinical material
Soumitesh Chakravorty, Jaya S. Tyagi *
Department of Biotechnology, All India Institute of Medical Sciences, New Delhi-110029, India
Received 29 August 2001; accepted 25 September 2001
First published online 30 October 2001
Abstract
Nucleic acid amplification technologies offer great promise for the rapid, sensitive and specific diagnosis of tuberculosis. However, the
isolation of inhibitor-free DNA from biological specimens is a bottleneck of the PCR assay. Here we describe a simple method for the
isolation of PCR-amplifiable DNA of Mycobacterium tuberculosis from all types of samples of pulmonary and extrapulmonary origin
tested. Briefly, it involves concentration of the bacilli by high-speed centrifugation, removal of PCR inhibitors by a wash solution containing
guanidinium isothiocyanate and the release of bacterial DNA by heating in the presence of detergents and Chelex-100 resin. The entire
process is accomplished within V3 h. The method has been validated on 780 samples of human, bovine and guinea pig origin including
sputum, cerebrospinal fluid, pulmonary fluids, pus, fine needle aspirate, tissue, blood and milk. ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
Keywords : Mycobacterium tuberculosis DNA isolation ; Clinical sample; Guanidinium isothiocyanate
1. Introduction
A crucial step in the diagnostic PCR assay for tuberculosis is the isolation of ampli¢able template from clinical
material. Tissue and cell constituents present in biological
material are oftentimes inhibitory to PCR leading to the
reporting of false negative results. A natural consequence
has been the lowering of test sensitivity, which has led to
the quest for improved DNA isolation protocols. Most of
the procedures have been developed for sputum since it
constitutes the most common and most infectious biological specimen. Sputum in itself presents a challenge as its
physical characteristics may vary from containing copious
amounts of mucus to being purulent, haemoptyic or tenacious. The same is true for extrapulmonary specimens like
* Corresponding author. Tel. : +91 (11) 652 4491;
Fax: +91 (11) 686 2663.
E-mail address : [email protected] (J.S. Tyagi).
Abbreviations : M., Mycobacterium ; GITC, guanidinium
isothiocyanate; AFB, acid-fast bacillus; CTAB, cetyltrimethylammonium
bromide
pus, tissue biopsy, etc. The commonly employed DNA
isolation methods involve the use of enzymes such as lysozyme and proteinase K, cetyltrimethylammonium bromide (CTAB) treatment and extraction with organic solvents [1], detergent-induced lysis in conjunction with
proteinase K and lysozyme [2] or lysis using guanidinium
isothiocyanate (GITC)-containing solutions [3,4]. Chelex100 resin has been used in DNA isolation from clinical
samples [5,6]. Though many methods have been published,
no single method is available that is universally applicable
to all types of clinical specimen. Commercial houses market kits for pathogen DNA isolation optimized for sputum. However, they are out of place in high disease incidence countries like India owing to their exorbitant cost
and the increasing prevalence of extrapulmonary TB.
Therefore there is a need for a reproducible, simple and
inexpensive protocol for mycobacterial DNA isolation
from all types of biological samples. Here, we report a
method for the isolation of Mycobacterium tuberculosis
DNA that ful¢ls these requirements. We have employed
GITC to e¤ciently wash away PCR inhibitors from clinical samples. Mycobacterial DNA is subsequently obtained upon bacterial lysis by heating with Chelex-100
and detergents.
0378-1097 / 01 / $20.00 ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 1 0 9 7 ( 0 1 ) 0 0 4 5 2 - 9
FEMSLE 10198 20-11-01
114
S. Chakravorty, J.S. Tyagi / FEMS Microbiology Letters 205 (2001) 113^117
natant. CSF, BAL, pleural e¡usion: Centrifuge at high
speed, discard supernatant.
2. Materials and methods
2.1. Clinical specimens
Seven hundred and eighty samples were included in the
development and validation aspects of this study. These
comprised 300 sputa and 216 other pulmonary and extrapulmonary samples of human origin, namely, CSF
(n = 130), pus (n = 14), ¢ne needle aspirate (n = 14), pleural
e¡usion (n = 10), bronchoalveolar lavage (n = 11), tissue
biopsy (n = 15), blood (n = 17) and para¤n-embedded tissue (n = 5). The samples were provided by hospital consultants, namely, Drs. M. Dudeja, Sunderlal Jain Charitable
Trust Hospital ; S.K. Sharma and S. Singh, AIIMS ; New
Delhi, V.M. Katoch, JALMA, Agra and consultants at Sri
Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram. Guinea pig spleens (n = 4) were
provided by Dr. V.D. Ramanathan of the Tuberculosis
Research Centre, Chennai. Bovine samples (n = 260) were
also successfully processed by this method (Dr. H.K. Prasad, personal communication) including milk (n = 65),
blood (n = 65), rectal pinch (n = 65) and prescapular lymph
gland biopsy (n = 65).
2.2. DNA isolation method
2.2.1. Preprocessing
Sputum: Add 0.2 vol. of 2.5% N-acetyl L-cysteine
(NALC, in 68 mM phosphate bu¡er pH 7.5). Vortex
and stand for 5^10 min. Centrifuge at 25 000Ug (high
speed) at room temperature for 15 min. Discard supernatant. Tissue: Mince and homogenize tissue in 5 M GITC,
50 mM Tris^Cl, pH 7.5, 25 mM EDTA, 0.5% Sarcosyl, 0.2
M L-mercaptoethanol (inhibitor removal solution, IRS) in
a mini bead beater (Biospec, USA) using 1-mm glass beads
for 30^60 s. Centrifuge at 600Ug for 3 min. Centrifuge the
supernatant at high speed and discard supernatant. Paraf¢n-embedded tissue : Depara¤nize tissue by xylene treatment. Proceed as above. Blood: Centrifuge at high speed.
Wash pellet with 1 M NaCl. Milk : Centrifuge at high
speed. Discard supernatant. Remove milk fat by wiping
tube wall with a cotton swab. Pus: To purulent and thick
samples, add an equal volume of phosphate-bu¡ered saline and vortex. Centrifuge at high speed; discard super-
2.2.2. Inhibitor removal
Resuspend pellet from step 1 in an equal volume of IRS.
Note: For highly mucoid, tenacious or purulent samples,
complete resuspension in IRS may require incubation at
37³C for up to 60 min or an additional wash with IRS. An
additional wash with the IRS may also be necessary for
tissue samples. Centrifuge at high speed and discard supernatant. Rinse pellet with water. Dry pellet.
2.2.3. DNA extraction
To pellet, add approximately 5 vols. of Chelex-100 resin
[5] (Bio-Rad, USA), Tween 20 and Triton X-100 at a ¢nal
concentration of 10%, 0.3% and 0.03% respectively. Heat
the tubes at 90³C for 40 min. Use supernatant for PCR.
2.3. Stability and handling of IRS
The IRS was stable for at least 2 months at room temperature if stored in a dark and airtight container. The
IRS was prepared in a fume hood and GITC-containing
waste was disposed o¡ in 10 M NaOH ensuring that the
¢nal concentration did not fall below 0.3 M [3].
2.4. PCR
DNA ampli¢cation was assessed by the devR assay that
is speci¢c for organisms of the M. tuberculosis complex
[7,8]. Brie£y, a 40-Wl reaction was set up containing 0.5
WM of primers devRf (5P-GGTGAGGCGGGTTCGGTCGC-3P) and devRr (5P-CGCGGCTTGCGTCCGACGTTC-3P), 1.5 mM of MgCl2 , 0.2 mM dNTPs, 1 U of Taq
DNA polymerase and 4 or 10 Wl of specimen DNA (neat
and 1/10 dilution). In parallel, inhibitor check reactions
were set up containing 10^20 ng of M. tuberculosis DNA
and sample DNA. The thermal cycling parameters were 10
min at 94³C, 40/45 cycles each of 1 min at 94³C, 1 min and
30 s at 65³C/70³C, and a ¢nal extension of 10 min at 72³C.
23S rDNA PCR assay was performed on para¤n-embedded tissue as described [7]. PCR products were visualized by ethidium bromide staining after agarose gel electrophoresis.
Fig. 1. Standardization of DNA isolation by IRS method. A: PCR ampli¢cation products of M. tuberculosis DNA extracted from a +4 grade AFB
positive sputum as described [5]. Lanes: 1 and 2, neat and 1/10 dilution of DNA, respectively ; 3, inhibitor check (neat sample DNA spiked with M. tuberculosis DNA) ; 4, DNA positive control. B: E¡ect of sample dilution on PCR ampli¢cation. Lanes 1, 2 and 3, neat, 1/50 diluted sputum and 1/100
diluted sputum respectively. C: E¡ect of detergents used in lysis on PCR ampli¢cation. Lanes: 1 and 2, 0.03% Triton X-100 and 0.3% Tween 20 mixture with neat and 1/10 dilution of DNA, respectively; 3 and 4, 1% NP-40, 0.1% SDS and 1% Tween 20 mixture [5] with neat and 1/10 dilution of
DNA, respectively; 5 and 6, 0.1% SDS, neat and 1/10 dilution of DNA, respectively.
FEMSLE 10198 20-11-01
S. Chakravorty, J.S. Tyagi / FEMS Microbiology Letters 205 (2001) 113^117
Fig. 2. DNA isolation by IRS method from sputa of variable volume
and bacillary load. A: PCR pro¢le of DNA extracted from varying volumes of a +4 grade AFB positive sputum. Neat and 1/10 dilution of
DNA was used for PCR (left and right lanes respectively). B: PCR pro¢le of DNA extracted from sputa varying bacillary load (+2 grade AFB
positive sputum, +1 grade, and AFB negative sputa). The AFB negative
sputa were from patients with clinical symptoms of tuberculosis who responded to antitubercular treatment.
115
grade AFB positive sputa of various physical characteristics containing V107 bacilli/ml were used during standardization. The samples were divided into two equal aliquots and one set was treated with IRS prior to DNA
isolation. Good ampli¢cation was obtained from all the
IRS-treated samples and no inhibition was noticeable.
However, PCR inhibition was noticeable with seven out
of 10 samples not treated with IRS. The IRS wash e¤ciently removed caseous material, mucin or tissue debris as
indicated by a substantial decrease in post-wash pellet size.
A representative gel picture is shown (Fig. 1D). The IRS
treatment e¡ectively removed inhibitory material and also
did not compromise with DNA yield from sputum of
varying volume (5^0.5 ml, Fig. 2A) and of varying bacterial load (AFB +2 grade to smear negative, Fig. 2B). The
process was validated on s 300 sputa ranging from +4
grade to smear-negative specimens.
3.2. Sensitivity of the IRS method
3. Results
3.1. Standardization of DNA isolation procedure
M. tuberculosis DNA isolation was attempted from several acid-fast bacillus (AFB) positive sputa as described
[5]. A substantial degree of PCR inhibition was noted
with more than half the samples (particularly purulent
and/large volume samples, Fig. 1A). Very faint ampli¢cation was observed when the sputum samples were diluted
50^100-fold with 10 mM Tris^1 mM EDTA, pH 8.0, likely
due to dilution of inhibitors (Fig. 1B). In order to improve
DNA recovery and to e¡ectively remove inhibitors, two
approaches were attempted. First, the e¤cacy of various
detergents including sodium dodecyl sulfate (SDS), Triton
X-100, Tween 20 and Nonidet P-40 (NP-40), in lysis of
tubercle bacilli, cultured in 7H9 medium containing albumin dextrose and Tween 80, was evaluated. The detergents
were used in conjunction with 10% Chelex-100 [5]. The
inclusion of SDS at the recommended concentration of
0.1% [5] inhibited DNA ampli¢cation, suggesting PCR inhibition by SDS. The inclusion of 0.03% Triton X-100 and
0.3% Tween 20 either alone (not shown) or together, consistently resulted in the best bacterial lysis in comparison
to a mixture of NP-40, SDS and Tween 20 [5] (Fig. 1C).
Second, an inhibitor removal step was introduced. Ten +4
Sensitivity was evaluated on four +3 grade sputa.
Brie£y, NALC-treated sputa were divided into aliquots
ranging from 1 ml to 3.5 Wl, the volumes were made up
to 1 ml by the addition of Tris^EDTA, DNA was isolated
as described and PCR analysis was performed. DNA ampli¢cation was detectable with sputa volumes as low as 7^
15 Wl which was equivalent to DNA from V50 to 100
bacilli. A typical ampli¢cation pro¢le is shown in Fig. 3A.
The applicability of the IRS method to tissues was also
examined. Guinea pigs infected subcutaneously with M.
tuberculosis H37Rv (four animals) were sacri¢ced 47
days post infection. Spleens were homogenized, plated as
neat and serial dilutions on Lowenstein Jensen slants,
DNA isolated as described and subjected to M. tuberculosis PCR analysis in parallel. The devR gene was ampli¢ed from all dilutions of spleen homogenate tested. The
highest dilution of spleen homogenate contained V30 genome equivalents of DNA (based on recovered cfu). A
representative result is shown in Fig. 3B.
3.3. Comparison of IRS method with proteinase
K-SDS-CTAB method
Duplicate aliquots of ¢ve +3 grade sputa (125^15 Wl
volume) were subjected to DNA extraction by the IRS
Fig. 3. Limit of detection of the IRS method. A: Ampli¢cation products obtained from decreasing volumes of sputum (1 ml^3 Wl). For each sputum
volume, neat and 1/10 dilution of DNA was used for PCR (left and right lanes respectively). B: Ampli¢cation products obtained from increasing dilutions of M. tuberculosis infected guinea pig spleen homogenates. N, neat homogenate; N1, 1/10 dilution; N2, 1/100 dilution ; N3, 1/1000 dilution; N4,
1/10 000 dilution.
FEMSLE 10198 20-11-01
116
S. Chakravorty, J.S. Tyagi / FEMS Microbiology Letters 205 (2001) 113^117
method and by another method used in our laboratory
employing proteinase K-SDS-CTAB [1]. A typical result
is shown in Fig. 4. The lower limit of detection for the IRS
method was 15 Wl compared to 31 Wl for the proteinase KSDS-CTAB method. The improved performance of the
IRS method over the latter method with particular reference to inhibitor removal was veri¢ed on 80 sputa of varying bacterial load. Sixteen out of 40 DNA samples isolated
by proteinase K-SDS-CTAB method showed PCR inhibition and required further puri¢cation. In contrast, not a
single sample out of the 40 DNA samples isolated by the
IRS method showed PCR inhibition (not shown).
Fig. 5. DNA ampli¢cation from pulmonary and extrapulmonary samples. A: pleural e¡usion ; B: bovine milk from cattle infected with Mycobacterium bovis; C: endometrial tissue biopsy; D: lymph node ¢ne
needle aspirate; E: CSF sample ; F: pus; G: bronchoalveolar lavage;
H: para¤n-embedded tissue from patient suspected to have atypical mycobacterial infection. Panels A^G show M. tuberculosis devR ampli¢cation products and H shows mycobacterial 23S rDNA ampli¢cation
product.
3.4. Application to extrapulmonary specimens
The IRS method was successfully applied to the isolation of inhibitor-free DNA from all types of extrapulmonary samples tested, including £uids and tissue samples of
human and bovine origin (see Section 2). Representative
PCR assays with M. tuberculosis DNA isolated from CSF,
pleural e¡usion, bronchoalveolar lavage, pus, FNAC,
blood, bovine milk, tissue biopsy and para¤n-embedded
tissue are shown in Fig. 5.
4. Discussion
We have developed a rapid and simple method for the
extraction of PCR-ampli¢able M. tuberculosis DNA from
clinical samples. For sample volumes less than 1.3 ml, the
DNA isolation process can be performed in a 1.5 ml microcentrifuge tube. Our method exploits the tough cell
wall characteristics of mycobacteria that confer resistance
to 5 M GITC. In a departure from the conventional use of
this chaotropic agent in e¡ecting lysis of cells for nucleic
acid isolation [3,4,9,10], our method is based on a wash
with a GITC-containing bu¡er. The introduction of a
`cleaning-before-lysis' step results in e¡ective removal of
biological constituents (and potential PCR inhibitors)
present in the specimen and creates a better environment
for e¤cient lysis of mycobacteria rendering further puri¢cation of DNA unnecessary. Residual inhibitors, if present
at all, are adsorbed by Chelex-100 resin during the DNA
isolation step.
The methods currently employed for the isolation of
mycobacterial DNA from clinical samples su¡er from
one or many drawbacks that include long processing
times, use of organic solvents and enzymes and multiple
steps resulting in either DNA loss or in the ine¤cient
removal of PCR inhibitors. Multistep methods increase
false positive results due to the risk of cross-contamination
and also augment false negative reporting due to losses in
DNA recovery. These drawbacks have been overcome in
the IRS method. The processing time is reduced to V3 h,
resulting in the reporting of test ¢ndings on the same day
as the receipt of the sample. The process involves the use
of reagents that are readily available and inexpensive.
Most importantly, the process can e¤ciently remove all
kinds of inhibitors present in clinical samples. The method
has been validated on 780 pulmonary and extrapulmonary
samples. E¤cient removal of the PCR inhibitors permitted
the addition of more DNA equivalents to the PCR reaction, thereby increasing the sensitivity of the assay procedure. Enhanced sensitivity is invaluable when handling
low volume and/or paucibacillary specimens.
Acknowledgements
Fig. 4. Comparison of IRS method with proteinase K-SDS-CTAB method. Ampli¢cation products obtained with DNA isolated from serial dilutions of a sputum sample (0.125^0.015 ml). For each sputum volume,
neat and 1/10 dilution of DNA was used for PCR (left and right lanes
respectively).
This work was supported by a grant from the Department of Biotechnology, Government of India and by
AIIMS to J.S.T. S.C. is thankful to the Department of
Biotechnology for ¢nancial assistance.
FEMSLE 10198 20-11-01
S. Chakravorty, J.S. Tyagi / FEMS Microbiology Letters 205 (2001) 113^117
References
[1] Singh, K.K., Muralidhar, M., Kumar, A., Chattopadhyaya, T.K.,
Kapila, K., Singh, M.K., Sharma, S.K., Jain, N.K. and Tyagi, J.S.
(2000) Comparison of in house polymerase chain reaction with conventional techniques for the detection of Mycobacterium tuberculosis
DNA in granulomatous lymphadenopathy. J. Clin. Pathol. 53, 355^
361.
[2] Perera, J., Chandrasekharan, N.V. and Karunanayake, E.H. (1994)
PCR based detection of Mycobacterium tuberculosis : e¡ect of sample
preparation. Southeast Asian J. Trop. Med. Public Health 25, 693^
697.
[3] Boom, R., Sol, C.J.A., Salimans, M.M., Jansen, C.L., Wertheim-van
Dillen, P.M.E. and van der Noordaa, J. (1990) Rapid and simple
method for puri¢cation of nucleic acids. J. Clin. Microbiol. 28,
495^503.
[4] Noordhoek, G.T., Kaan, J.A., Mulder, S., Wilke, H. and Kolk,
A.H.J. (1995) Routine application of the polymerase chain reaction
for detection of Mycobacterium tuberculosis in clinical samples.
J. Clin. Pathol. 48, 810^814.
117
[5] de Lamballerie, X., Zandotti, C., Vignoli, C., Bollet, C. and de Micco, P. (1992) A one-step microbial DNA extraction method using
`Chelex-100' suitable for gene ampli¢cation. Res. Microbiol. 143,
785^790.
[6] Walsh, P.S., Metzger, D.A. and Higuchi, R. (1991) Chelex-100 as a
medium for simple extraction of DNA for PCR-based typing from
forensic material. BioTechniques 10, 505^513.
[7] Dasgupta, N., Kapur, V., Singh, K.K., Das, T.K., Sachdeva, S.,
Jyothisri, K. and Tyagi, J.S. (2000) Characterization of a two-component system, devR-devS, of Mycobacterium tuberculosis. Tubercle
Lung Dis. 80, 141^159.
[8] Singh, K.K., Nair, M.D., Radhakrishnan, K. and Tyagi, J.S. (1999)
Utility of PCR assay in diagnosis of En-Plaque tuberculoma of the
brain. J. Clin. Microbiol. 37, 467^470.
[9] Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter, W.J.
(1979) Isolation of biologically active ribonucleic acid from sources
enriched in ribonuclease. Biochemistry 18, 5294^5299.
[10] Chomczynski, P. and Sacchi, N. (1987) Single-step method of RNA
isolation by acid guanidinium isothiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156^159.
FEMSLE 10198 20-11-01