Veterinary Microbiology 109 (2005) 211–216

www.elsevier.com/locate/vetmic

Rapid differentiation of Mycobacterium bovis and Mycobacterium

tuberculosis based on a 12.7-kb fragment by a single
tube multiplex-PCR
C.S. Bakshi a, D.H. Shah a,1, Rishendra Verma b,*, R.K. Singh a, Meenakshi Malik a
a
National Biotechnology Center, Indian Veterinary Research Institute, Izatnagar, Uttar Pradesh 243122, India
b
Mycobacteria Laboratory, Division of Biological Standardization, Indian Veterinary Research Institute,
Izatnagar, Bareilly, Uttar Pradesh 243122, India
Received 18 June 2004; received in revised form 25 January 2005; accepted 20 May 2005

Abstract

The aim of this work was the design and validation of a rapid and easy single tube multiplex-PCR (m-PCR) assay for the
unequivocal differential detection of Mycobacterium bovis and Mycobacterium tuberculosis. Oligonucleotide primers were
based on the uninterrupted 229-bp sequence in the M. bovis genome and a unique 12.7-kb insertion sequence from the M.
tuberculosis genome, which is responsible for species-specific genomic polymorphism between these two closely related
pathogens. The m-PCR assay was optimized and validated using 22 M. bovis and 36 M. tuberculosis clinical strains isolated from
diverse host species and 9 other non-tuberculous mycobacterial (NTM) strains. The designed primers invariably amplified a
unique 168-bp (M. bovis-specific) and 337-bp (M. tuberculosis-specific) amplicon from M. bovis and M. tuberculosis strains,
respectively. The accuracy of the assay, in terms of specificity, was 100%, as none of the NTM strains tested revealed any
amplification product. As little as 20 pg of genomic DNA could be detected, justifying the sensitivity of the method. The m-PCR
assay is an extremely useful, simple, reliable and rapid method for routine differential identification of cultures of M. bovis and
M. tuberculosis. This m-PCR may be a valuable diagnostic tool in areas of endemicity, where bovine and human tuberculosis
coexist, and the distinction of M. bovis from M. tuberculosis is required for monitoring the spread of M. bovis to humans.
# 2005 Elsevier B.V. All rights reserved.

Keywords: Mycobacterium tuberculosis; Mycobacterium bovis; PCR

* Corresponding author. Tel.: +91 581 230 1757; 1. Introduction

fax: +91 581 230 3284.
E-mail addresses: dhshah99@yahoo.com (D.H. Shah), Tuberculosis (TB), caused by Mycobacterium
rishendra_verma@yahoo.com (R. Verma). tuberculosis, is one of the most widespread infectious
1
Present address: Department of Veterinary Internal Medicine,
Biosafety Research Institute, Teaching Veterinary Hospital, College
diseases and the leading cause of death, due to a single
of Veterinary Medicine, Chonbuk National University, Jeonju 561- infectious agent, among adults in the world. Likewise,
756, South Korea. bovine TB, caused by Mycobacterium bovis, is a

0378-1135/$ – see front matter # 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.vetmic.2005.05.015
212 C.S. Bakshi et al. / Veterinary Microbiology 109 (2005) 211–216
significant veterinary disease that can spread to
humans. M. tuberculosis is the most common cause
of human TB, but an unknown proportion of cases
occur due to M. bovis (Acha and Szyfres, 1987).
Moreover, disease caused by M. bovis in human
immunodeficiency virus (HIV)-positive individuals is
also an increasing concern (Daborn et al., 1993;
Blazquez et al., 1997). There is a direct correlation
between M. bovis infection in cattle and disease in the
human population (Cosivi et al., 1998). In industria-
lized countries, animal TB control and elimination
programs, together with milk pasteurization, have
drastically reduced the incidence of disease caused by
M. bovis in both cattle and humans. In developing
countries, however, animal TB is widely distributed,
control measures are not applied or are applied
sporadically, and pasteurization is rarely practiced.
For instance, M. bovis infection is responsible for
about 2% and 8% of new cases of human pulmonary
and extra-pulmonary TB, respectively, in Latin
America (Cosivi et al., 1998). In Asia, 94% and
>99% of the total cattle and buffalo populations,
respectively, are found in countries where bovine TB
is either partly controlled or not controlled at all. Thus,
94% of the Asian human population lives and is at a
risk, in countries where cattle and buffaloes undergo
no or only limited control for bovine TB (Cosivi et al.,
1998). M. bovis and M. tuberculosis are closely related
organisms and have almost identical genomes
(Imaeda, 1985). Despite the high degree of DNA
homology between the two organisms, M. tuberculosis
causes disease almost exclusively in humans and
rarely in other animals, whereas M. bovis can cause
TB in a wide range of animal hosts, including humans
(WHO, 1994). In areas where bovine and human TB
coexist and are endemic, the separation of M. bovis
from M. tuberculosis is important in monitoring the
spread of M. bovis to humans. In addition, such a
distinction has an impact on the treatment of the
disease, since M. bovis strains are naturally resistant to
pyrazinamide (PZA) (Konno et al., 1967); thus, human
TB caused by M. bovis cannot be treated with PZA.
Currently, differentiation of M. bovis from M.
tuberculosis is based on conventional culture and
biochemical tests. In addition to being tedious and
slow, current methods, such as biochemical typing, are
not 100% reliable due to the advent of intermediate
strains, such as niacin and T2CH variant M.
tuberculosis (Verma et al., 1987; Grange et al.,
1996) and PZA variant M. bovis (Niemann et al.,
2000). Partly due to close genetic relatedness and
variable biochemical patterns, definitive detection of
M. bovis and M. tuberculosis, up to species level, is
time consuming and difficult. Methods, such as PCR,
could be the best alternative strategy to meet this
purpose.
Recently, sequence analysis of the M. tuberculosis
and M. bovis genomes has shown that M. bovis lacks a
12.7-kb fragment present in the genome of M.
tuberculosis (Zumarraga et al., 1999). Further analysis
of the 12.7-kb fragment suggested that it represents a
deletion in M. bovis rather than an insertion in M.
tuberculosis. This deletion removes most of the mce-3
operon, one of the four closely related operons, which
may be involved in cell entry. Therefore, it was
suspected that this deletion might contribute to
differences in virulence or host range in the two
species. Interestingly, all the M. tuberculosis isolates
studied showed the presence of the 12.7-kb fragment,
while all the M. bovis strains lacked this fragment
(Zumarraga et al., 1999). Therefore, the 12.7-kb
fragment may be a useful marker to differentiate M.
bovis from M. tuberculosis. In this report, we describe
a new single tube multiplex-PCR (m-PCR) assay,
based on a 12.7-kb fragment, for the rapid and easy
differential detection of M. bovis and M. tuberculosis.
2. Materials and methods
2.1. Mycobacteria strains and DNA
M. tuberculosis strains (35) used in this study were
clinical strains, i.e. isolates from human patients with
pulmonary TB from the Medical Hospital, IVRI,
Izatnagar (UP), India (9 strains), District Tuberculosis
Hospital, Bareilly (UP), India (15 strains), bovines (9
strains), guinea pig (1 strain) and swine (1 strain). M.
bovis strains (20) were isolated from bovines (17
strains), deer (2 strains) and black buck (1 strain). The
standard strains examined were M. tuberculosis
H37Rv, M. bovis AN5, M. bovis BCG (Mycobacteria
Laboratory, IVRI), M. paratuberculosis (Teps strains,
Division of Biological Products, IVRI) and the non-
tuberculous mycobacterial (NTM) strains, including
M. avium, M. intracellulare, M. fortuitum, M.
 C.S. Bakshi et al. / Veterinary Microbiology 109 (2005) 211–216
chelonae, M. gordonae, M. smegmatis, M. phlei and
M. xenopi, were obtained from the Tuberculosis
Research Centre, Chennai, India. For the extraction of
PCR amplifiable DNA, a loop-full of mycobacterial
growth was suspended in a microfuge tube containing
400 ml of 1
TE (10 mM Tris–HCl, 1 mM EDTA, pH
8.0). The suspension was than subjected to boiling for
10 min followed by a brief centrifugation and the
supernatant was directly used as a template for PCR.
2.2. Species characterization
Frozen stocks of M. bovis and M. tuberculosis
strains, maintained at the Mycobacteria Laboratory at
IVRI, were freshly grown on Lowenstein–Jensen (LJ)
medium. All M. tuberculosis and M. bovis strains were
identified, as described previously, by classical
methods (Verma and Srivastava, 2001), as well as
by PCR based on allele-specific amplification of pncA
and oxyR genes (Shah et al., 2002).
2.3. Primers and PCR conditions
The amplification primers for single tube m-PCR,
designed in this study, were based on the previously
described sequences (Zumarraga et al., 1999). The
primers targeted a 229-bp sequence in M. bovis, which
in the case of M. tuberculosis, is interrupted at position
197 by a unique 12.7-kb fragment ORF MTCY 227.28c
encoding a hypothetical protein ‘Rv1506c’ (accession
no. Z79701). Oligonucleotide sequences of the primers
used in the study were: the common forward primer,
CSB1 (50 -TTCCGAATCCCTTGTGA-30 ), and two
reverse primers, including M. bovis-specific, CSB2
(50 -GGAGAGCGCCGTTGTA-30 ), and M. tuberculo-
sis-specific, CSB3 (50 -AGTCGCGTGGCTTCTCTT-
TTA-30 ).
The m-PCR reactions were performed in a total
volume of 50 ml consisting of the following: 5 ml of
the template DNA, 25 pmol of each primer (CSB1,
CSB2 and CSB3), 200 mM of each dNTPs, 1.5 U of
Taq DNA polymerase (Bangalore Genei, Bangalore,
India), 10 mM Tris–HCl (pH 8.0), 50 mM KCl,
1.5 mM MgCl2 and 0.01% (w/v) gelatin. The cycling
parameters were: initial denaturation at 94 8C for
5 min, followed by 30 three-step cycles, including
denaturation at 94 8C for 1 min, annealing at 52.3 8C
for 1.5 min, extension at 72 8C for 1 min and a final
213
extension at 72 8C for 5 min. The amplification
products were analyzed by electrophoresis on 1.5%
(w/v) agarose gel and visualized by ethidium bromide
fluorescence. The unique amplification product of
either 168 bp (M. bovis-specific) or 337 bp (M.
tuberculosis-specific) must be visualized using DNA
from the respective species.
2.4. Specificity and sensitivity of multiplex PCR
assay
To determine the specificity of the m-PCR assay,
genomic DNAs from various non-tuberculous myco-
bacteria mentioned above, were subjected to ampli-
fication by the cited primer combinations. Serial 10-
fold dilutions of DNA from both M. bovis AN5 and M.
tuberculosis H37Rv strains were subjected to ampli-
fication by the designed primers to determine the
sensitivity of the m-PCR assay.
3. Results
3.1. Design of primers and optimization of m-PCR
conditions
A common forward primer, CSB1, was designed to
hybridize the 229-bp sequence found in both M. bovis
and M. tuberculosis, and complemented bases 50–66.
The M. bovis-specific reverse primer, CSB2, comple-
ments bases 217–202 of the 229-bp sequence and is
also expected to hybridize to the 229-bp sequence
from both organisms, but should generate a unique
168-bp PCR product in the case of M. bovis only and
not in M. tuberculosis. This is expected since the 229-
bp sequence, present in M. tuberculosis, is interrupted
at position 197 by a unique 12.7-kb fragment. In
principle, in a PCR reaction, the 12.7-kb size insertion
in M. tuberculosis is beyond the amplification limits of
Taq DNA polymerase and, hence, cannot be amplified
using primer CSB2. In contrast, M. tuberculosis-
specific reverse primer CSB3, which complements
bases 23,729–23,708 of the 12.7-kb fragment, is
designed to hybridize to the 12.7-kb fragment and is
expected to generate a unique 337-bp PCR product
specific to M. tuberculosis.
The m-PCR assay was initially tested with genomic
DNA from standard strain M. bovis AN5 and M.
214 C.S. Bakshi et al. / Veterinary Microbiology 109 (2005) 211–216

tuberculosis H37Rv. After several rounds of amplifica- 4. Discussion

tion and testing different annealing temperatures (data
not shown), adequate conditions were found (see
Section 2.4) to distinguish two species in a single
reaction.
3.2. Rapid differentiation of M. bovis and
M. tuberculosis by m-PCR
The m-PCR assay was applied to DNA from 35 M.
tuberculosis and 20 M. bovis strains, isolated from
clinical cases of tuberculosis in human and animals. In
all cases, reaction mixtures with the template DNAs
from M. bovis strains, including M. bovis BCG,
invariably showed a unique amplicon of 168 bp, while
the reactions with the DNAs from M. tuberculosis
strain showed a unique amplicon of 337 bp (Fig. 1).
The m-PCR assay could not differentiate M. bovis
BCG from those of M. bovis non-BCG clinical strains,
but the amplification results were consistent with each
strain of M. bovis and M. tuberculosis, irrespective of
the source.
3.3. Sensitivity and specificity of m-PCR
As determined by serial 10-fold dilutions of the
template DNAs from both M. bovis AN5 and M.
tuberculosis H37Rv, the amplicons of 168 bp (M.
bovis-specific) and 337 bp (M. tuberculosis-specific)
could be visualized when amplifications were per-
formed with as little as 20 pg of chromosomal DNA
(data not shown). Templates from NTM strains
produced no detectable PCR products justifying the
specificity of the technique (Fig. 1).
The development of any species-specific PCR
assay requires that the target gene or DNA fragment be
present in all the isolates from the particular species of
interest and be absent from all other unrelated species.
In the past, the mtp-40 gene (M. tuberculosis-specific)
and a 500-bp DNA fragment (M. bovis-specific) were
targeted to develop a PCR assay to differentiate
between these two pathogens (Del-portillo et al.,
1991; Rodriguez et al., 1995). However, the specificity
of these PCR methods was invalidated due to the
findings that mtp-40 gene is also present in some M.
bovis strains and the so-called M. bovis-specific 500-
bp fragment is also present in some M. tuberculosis
strains (Weil et al., 1996; Shah et al., 2002).
In the present study, we designed a set of three PCR
primers targeting the 229-bp sequence polymorphism
generated due to the presence of a 12.7-kb fragment in
the M. tuberculosis genome. This provided an
improved method for definitive detection of these
two closely related species. The newly designed
primers correctly identified M. tuberculosis (337 bp)
and M. bovis (168 bp) at species level in a single tube
reaction. The results obtained in this study indicated
that the primer sequences designed for the differentia-
tion of closely related M. bovis and M. tuberculosis are
100% specific. This m-PCR based on a 12.7-kb
fragment polymorphism does not differentiate M.
bovis BCG isolate from those of M. bovis non-BCG
clinical strains. Our results, along with the findings of
Zumarraga et al. (1999), indicate that the presence of
the 12.7-kb fragment is unique to M. tuberculosis
strains, as is its absence in M. bovis strains, tested from

Fig. 1. Ethidium bromide-stained 1.5% (w/v) agarose gel showing PCR products amplified from representative strains of mycobacteria by m-
PCR assay. Lanes: (M) DNA molecular mass marker (100 bp ladder); (1) M. bovis AN5; (2) M. bovis BCG; (3:) M. bovis 2/86 clinical isolate; (4)
M. tuberculosis H37Rv; (5) M. tuberculosis 199/94 clinical isolate; (6) M. paratuberculosis; (7) M. avium; (8) M. smegmatis; (9) M. chelonae;
(10) M. fortuitum; (11) M. phlei; (12) M. intracellulare; (13) M. xenopi; (14) M. gordonae.
 C.S. Bakshi et al. / Veterinary Microbiology 109 (2005) 211–216
different geographical areas and host species. The
cross-reactivity of the designed primers with other
non-tuberculous mycobacteria was also excluded, as
none of the NTM strains tested showed presence of
PCR amplicons. The m-PCR assay was sensitive, as
the amplicons of 168 bp and 337 bp could be
visualized when the PCR was performed with as
little as 20 pg of genomic DNA from M. bovis and M.
tuberculosis strains, respectively.
Recently, two allele-specific PCR methods, based
on allelic polymorphism in pncA and oxyR genes,
have also been described (De los Monteros et al.,
1998). Although these methods can differentiate M.
bovis from M. tuberculosis cultures, both assays
require two differential amplifications to be performed
in separate PCR tubes for identification of a single
isolate up to species level. A PCR assay was described
by Zumarraga et al. (1999), but it also required
different sets of primers and separate reactions for
definitive identification of these two pathogens. Our
m-PCR assay, however, had a similar sensitivity
(20 pg) as that of the allele-specific PCR methods (De
los Monteros et al., 1998), but with an advantage of
being simpler (requiring only a single tube reaction)
and less expensive (saving reagents required for an
additional PCR reaction).
Although our data show 100% specificity, it is
possible that analysis of larger number of strains will
identify organisms for which this polymorphism of
12.7 kb fragment may break down. However, no
reasonable number of strains can permit us to rule
out this possibility. Our analysis of polymorphism of
the 12.7-kb fragment in 55 strains studied here,
along with the 20 strains studied by Zumarraga et al.
(1999), suggests that the organisms for which this
association breaks down will be relatively rare or
confined to restricted geographical localities or host
species. Moreover, from an evolutionary point of
view, it has been hypothesized that an ancestor strain
having the 12.7-kb fragment diverged toward M.
tuberculosis and M. bovis, and then M. bovis, lost
this fragment (Zumarraga et al., 1999). Therefore, it
can be expected that the 12.7-kb fragment poly-
morphism is likely to be the most stable genetic
difference between these two pathogens. The results
obtained using the m-PCR assay designed in this
study are, therefore, conclusive and indicate that the
use of m-PCR based on polymorphism in the 12.7-kb
215
fragment reliably differentiates M. bovis from M.
tuberculosis. It appears that an m-PCR based on the
12.7-kb region offers an excellent test for rapid and
easy definitive identification M. bovis or M.
tuberculosis.
In conclusion, the m-PCR protocol described here
is highly species-specific and decreased the time
needed for speciation of M. bovis and M. tuberculosis.
This m-PCR assay can be easily used as a routine
monitoring tool in veterinary and medical microbiol-
ogy laboratories, especially in areas of endemicity,
where bovine and human TB coexist, and the
separation of M. bovis from M. tuberculosis is
required for monitoring the spread of M. bovis to
humans.
Acknowledgement
We are grateful to the Director, Indian Veterinary
Research Institute, Izatnagar 243122, India, for
providing necessary facilities to carry out this work.
References
Acha, P.N., Szyfres, B., 1987. Zoonotic tuberculosis. In:
Zoonoses and Communicable Diseases Common to Man and
Animals, second ed. Pan American Health Organization/World
Health Organization, Washington, Scientific Publication No.
503.
Blazquez, B., de Los Monteros, L.E.E., Samper, S., Martin, C.,
Guerrero, A., Cobo, J., van Embden, J., Baquero, F., Gomez-
Mampaso, E., 1997. Genetic characterization of multidrug-
resistant M. bovis strains from a hospital outbreak involving
human immunodeficiency virus positive patients. J. Clin. Micro-
biol. 35, 1390–1393.
Cosivi, O., Grange, J.M., Daborn, C.J., Raviglione, M.C., Fujikura,
T., Cousins, D., Robinson, S.R.A., Huchzermeyer, H.F.A.K., de
Kantor, I., Meslin, F.X., 1998. Zoonotic tuberculosis due to M.
bovis in developing countries. Emerg. Infect. Dis. 4, 59–70.
Daborn, C.J., Grange, J.M., Kazwala, R.R., 1993. The bovine
tuberculosis cycle: an African perspective. J. Appl. Bacteriol.:
Symp. Suppl. 8, 127s–132s.
De los Monteros, L.E.E., Galan, J.C., Gutierrez, M., Samper, S.,
Garcia Marin, J.F., Martin, C., Dominguez, L., de Rafael, L.,
Baquero, F., Gomez-Manpaso, E., Blazquez, J., 1998. Allele-
specific PCR method based on pncA and oxyR sequences for
distinguishing M. bovis from M. tuberculosis: intraspecific M.
bovis pncA sequence polymorphism. J. Clin. Microbiol. 36,
239–242.
216 C.S. Bakshi et al. / Veterinary Microbiology 109 (2005) 211–216
Del-portillo, P., Murillo, L.A., Patarrogo, M.E., 1991. Amplification
of a species-specific DNA fragment of M. tuberculosis and its
possible use in diagnosis. J. Clin. Microbiol. 29, 2613–2668.
Grange, J.M., Yates, M.D., de Kantor, I., 1996. Guidelines for
Speciation Within the M. tuberculosis Complex, second ed.
World Health Organization, Geneva, unpublished document
WHO/EC/Zoo/96.4.
Imaeda, I., 1985. Deoxyribonucleic acid relatedness among the
selected strains of M. tuberculosis, M. bovis, M. bovis BCG,
M. microti and M. africanum. Int. J. Syst. Bacteriol. 35, 147–
150.
Konno, K., Feldman, F.M., McDermott, W., 1967. Pyrazinamide
susceptibility and amidase activity of tubercle bacilli. Am. Rev.
Respir. Dis. 95, 461–469.
Niemann, S., Harmsen, D., RuschGerdes, S., Richter, E., 2000.
Differentiation of clinical M. tuberculosis complex isolates by
gyrB DNA sequence polymorphism analysis. J. Clin. Microbiol.
38, 3231–3234.
Rodriguez, J.G., Mejia, G.A., Portillo, P.D., Patarroyo, M.E., Mur-
illo, L.A., 1995. Species-specific identification of M. bovis by
PCR. Microbiology 141, 2131–2138.
Shah, D.H., Verma, R., Bakshi, C.S., Singh, R.K., 2002. A multiplex
PCR for the differentiation of M. bovis and M. tuberculosis.
FEMS Microbiol. Lett. 214, 39–43.
Verma, R., Sharma, A.K., Vanamaya, P.R., Khanna, P.N., Siddique,
I.H., Gupta, B.R., 1987. Prevalence of tuberculosis at an
organized farm and its public health implications..In: Abstr.
Ann. Conf. Ind. Assoc. Vet. Immunol. Inf. Dis., HAU, Hisar, p.
68.
Verma, R., Srivastava, S.K., 2001. Mycobacteria isolated from
man and animals: twelve year record. Ind. J. Anim. Sci. 71,
129–132.
Weil, A., Bonny, B., Plikaytis, W., Butler, R., Woodlay, C., Shinnick,
T., 1996. The mtp40 gene is not present in all strains of M.
tuberculosis. J. Clin. Microbiol. 34, 2309–2311.
World Health Organization (WHO), 1994. Report of WHO Working
Group on Zoonotic Tuberculosis (Mycobacterium bovis), with
the Participation of FAO, June 14, Mainz, Germany. WHO,
Geneva, unpublished document WHO/CDS/VPH/94.137.
Zumarraga, M., Bigi, F., Alito, A., Romano, M.I., Cataldi, A., 1999.
A 12.7-kb fragment of the M. tuberculosis genome is not present
in M. bovis. Microbiology 145, 893–897.