PCR diagnostics of Mycobacterium tuberculosis in historic human long bone remains from 18th century burials in Kaiserebersdorf, Austria
© Haring et al; licensee BioMed Central Ltd. 2008
Received: 12 March 2008
Accepted: 17 September 2008
Published: 17 September 2008
In the present pilot study we applied recently published protocols for detecting Mycobacterium tuberculosis in human remains. We screened long bones from an 18th century cemetery and skulls from the anatomical "Weisbach collection" (19th century). In addition, besides the study of abundance of tuberculosis in inmates of the poorhouse itself, we were interested to test whether in this particular instance tuberculosis can be identified from cortical bones, which are rarely affected by tuberculosis, but mostly better preserved than the vertebral bodies or epiphyses.
The DNA extractions from the bone samples were obtained following established ancient DNA protocols. Subsequently extracts were subjected to a series of PCR amplifications using primer pairs published previously [1, 2]. PCR products of the expected size were subsequently sequenced.
Only primers targeting the repetitive IS6110 insertion sequence yielded PCR products of appropriate size. In one sample only (skull sample WB354 of the "Weisbach collection") sequence analysis revealed an authentic M. tuberculosis sequence that matched to a reference sequence from GenBank.
With a variety of established PCR approaches we failed to detect M. tuberculosis DNA in historic human femurs from an 18th century cemetery relating to a poor house in Kaiserebersdorf, Austria. Our data may indicate that in this particular case, thoracic or lumbar vertebrae, i.e. bones that are severely affected by the disease, would be more suitable for molecular diagnostics than long bones. However, the unpredictable state of DNA preservation in bones from museum collections does not allow any general recommendation of any type of bone.
Bone tuberculosis (Spondylitis tuberculosa, "Pott's disease") and joint tuberculosis are caused by Mycobacterium tuberculosis and appear among others as late manifestations of a tuberculosis infection. Because of the inflammatory bony changes, e.g., partial or total destruction of the vertebral bodies or joint elements, manifestations of tuberculosis infections of (pre)historic human skeletal remains are very often identifiable by means of macroscopic inspections. Since the spread of M. tuberculosis depends on population density, its verification is also of high interest for the reconstruction of population dynamic processes in archaeology and anthropology . Nevertheless, post mortem destructions with substantial substance loss or other pathologies with a similar appearance such as, e.g., destruction and remodeling of joint surfaces caused by a fracture and dislocation of a joint element, or an idiopathic avascular necrosis of the femoral head, may lead occasionally to erroneous diagnoses, which may miss- or under-represent the prevalence of tuberculosis in past human populations. An unambiguous molecular identification of tuberculosis for historic human bone remains has therefore been highly appreciated by anthropologists and paleo-epidemiologists.
In recent years, tuberculosis has been diagnosed from a variety of historic human bone remains using ancient DNA methodology. Spigelman and Lemma  were the first to detect authentic DNA of M. tuberculosis in pre-European-contact human remains from Borneo by means of PCR amplification. This study has been criticized to some extent for technical issues, but the results were confirmed some ten years later . Other examples of detecting M. tuberculosis in historic samples include ancient Egyptian mummies  and twelve approximately 140 – 1,200 years old mummies excavated from the Andes Mountain region of South America . In seven samples Mycobacterium DNA could be detected out of which two proved positive for M. tuberculosis. In another study several loci of the M. tuberculosis genome were targeted in DNA extracted from naturally mummified remains from three 18th century individuals from Hungary by means of PCR and spoligotyping , and M. tuberculosis rather than Mycobacterium bovis was identified as the cause of the disease. Taylor et al.  were able to confirm the morphology based diagnosis of tuberculosis for approximately 2,200 years old human remains excavated from Tarrant Hinton, Dorset, United Kingdom, using a series of sensitive PCR amplifications. In addition, they identified a member of the "modern" type of M. tuberculosis and excluded other strains such as M. bovis as the source of infection of this particular individual. More recently, M. tuberculosis DNA could also be genotyped from five Iron Age individuals from Aymyrlyg, South Siberia . So far, the most comprehensive studies have been conducted by Zink et al. [9, 10]. Remains of 41 individuals from the ancient Egyptian population (3,000-500 BC) have been screened for M. tuberculosis, among them 21 samples without morphological osseous changes. M. tuberculosis was detected in 9 samples including 2 cases without pathological bone changes . In 2005, Zink et al.  screened numerous samples of different age and origin, and determined high frequencies of tuberculosis infections in historic populations.
Most studies mentioned above used either dried skin  or thoracic and lumbar vertebrae as well as ribs [2, 8], i.e. those parts of the skeleton that are affected most by Pott's disease, as sources for DNA extractions. Nevertheless, to some extent long bones have been shown suitable for detection of M. tuberculosis, e.g., Fusegawa et al. , although vertebrae, spines, and ribs remain the preferred samples. Zink et al.  also screened 51 long bones for the presence of M. tuberculosis, and showed that such samples may also be suitable for investigating osseous tuberculosis, although with probably lower success rate.
In the present pilot study, we screened 10 long bones from the 18th century burials of the Kaiserebersdorf castle, Austria, for M. tuberculosis DNA. In addition to study the abundance of tuberculosis in inmates of the poorhouse itself, we were interested to test whether in this particular instance tuberculosis can be identified from cortical bones (here we used the most robust element, the femur). Such bones are rarely affected by tuberculosis but are frequently much better preserved than the vertebral bodies or epiphyses, and preservation of authentic ancient DNA in the bones' middle regions has been shown likely . According to historical sources, the cemetery was localized close to a chapel and a poor house, which was established by the Habsburg empress Maria Theresia. The skeletal remains of 34 individuals were systematically investigated, and 24 individuals were identified as elder females. Nearly all individuals exhibited severe skeletal changes or pathological alterations. Among them not only progressive degenerative joint lesions indicating long-lasting hard labor, but also features such as cribra orbitalia (anemia), chronic Vitamin C deficiency (newly build bone structures and porosities) and Vitamin D deficiency (rachitic, osteomalacia, osteoporosis) as well as severe inflammations (meningitis and sinusitis) and symptoms of infectious diseases such as syphilis and tuberculosis were observed .
All samples were taken from the collections housed at the Department of Anthropology, Natural History Museum Vienna, Austria. Ten samples were taken from the midshaft femur of ten different individuals recovered from the burial site at the castle of Kaiserebersdorf, south-east of Vienna: KE1, KE8, KE18, KE20, KE22b, KE26b, KE22, KE23, KE26c, KE29. . Tuberculosis was clearly identified in one individual (KE20), a well preserved skeleton of an end-mature female, which shows characteristic osteolytic changes among two vertebral bodies and the adjacent portion of the os sacrum. Individual KE23, a mature female, very likely represents a case of beginning joint tuberculosis. The proximal metaphysis of the left ulna shows an active osteolytic process in form of a cavity. From an anthropological point of view, the inmates of the poor house in Kaiserebersdorf represent a group of hard working, extremely insufficiently nourished people. It is considered likely that infectious diseases such as tuberculosis have spread within this community.
As positive controls, three bone samples (core samples obtained from skull vaults) from the anatomical "Weisbach collection", dated to the end of the 19th century were used. Individual data such as sex, age-of-death, and cause-of-death were recorded in handwritten protocols : WB 565, a 22 year old male, died from meningitis tuberculosa. The internal layer exhibits no signs of inflammation, but many small (1–2 mm in diameter), often confluent depressions, which have been interpreted as features of tubercles. WB 354, a 24 year old male, died from tuberculosis of the peritoneum. The internal layer shows few very tiny new bone formations within the sagittal sinus and small depressions by tubercles in the middle and occipital cranial base. The sample has been taken from a pathomorphologically completely inconspicuous region of the skull. WB 594, a 22 year old male, died from tuberculosis of the peritoneum and meningitis. The internal lamina shows slight newly build bone structures within the sinus and imprints of tubercles exclusively at the posterior cranial base. These skull remains are macroscopically well preserved because they have never been exposed to soil.
Finally, two bone samples (also from the midshaft femur) were taken from the site Hainburg (Lower Austria) that dates back to the early Bronze Age (ca. 2,000-1,500 BC): HB284 (a female, age-of-death 21–24), and HB310 (a female, age-of-death 30–35). From this epoch, bones with morphological alterations due to tuberculosis have never been reported from Austrian localities (Ninvestigated = 2000). Therefore, we considered them as negative controls.
DNA extraction, PCR, and DNA sequencing
Small samples were chopped off the bones and ground into a fine powder using a mortar under liquid nitrogen. Approximately 0.1–0.3 gram of bone powder was subsequently transfered into a 2.0 ml screw-capped centrifuge tube and DNA was extracted according to the silica based protocol with modifications [15, 16].
Primers used for amplifying M. tuberculosis target DNA
Length of product
Sequence 5' – 3'
Insertion sequence IS1081
oxidative stress regulator OxyR pseudogene
flanking sections of TbD1 region
insertion sequence IS6110
PCR amplifications were performed using the PfuTurbo DNA polymerase (Stratagene, La Jolla, CA, USA) or the DyNAzyme (Finnzymes, Espoo, Finland) DNA polymerase in 1× Reaction Buffer, 0.2 mmol/L of each dNTP, 0.04% bovine serum albumin (BSA), 1 μmol/L of each primer, and 2 μL of DNA extract (DNA concentration not quantified). Thermal cycling conditions were: 94°C for 2 min; 32 cycles of 94°C for 50 sec, 48° – 60°C for 50 sec, and 72°C for 1 min; and a final extension at 72°C for 10 min. PCR products were purified for direct sequencing using 8 μL 10× diluted exoSAP-IT solution (USB Corporation, Cleveland, OH, USA) per reaction. Cycle sequencing, using the same primers as in the PCR reaction, was performed in 10 μL reactions using 2 μL BigDye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA, USA), 2 μL 5× Sequencing Buffer, 10 pmol primer, and 3 μL purified PCR product. Sequencing products were purified with ethanol precipitation and analyzed using an ABI 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).
PCR products that could not be sequenced directly were cloned in the pCR4Blunt-TOPO plasmid vector by means of the TOPO TA Cloning Kit (Invitrogen) and individual clones were subsequently sequenced.
Inhibitory effect of DNA extracts
Ancient DNA extracts may contain PCR inhibiting components that can cause misleading results. In such instances failure to obtain PCR products cannot be assigned to the absence of the target DNA but simply to inhibition of the reaction. In order to test for co-extracted PCR inhibitors in the samples used in the current study, established test PCRs targeting a ~1000 bp and a ~300 bp fragment of the mitochondrial genome of microtine rodents were spiked with ancient DNA extracts. Spiking PCR amplifications  were performed in a final volume of 25 μL using the Taq DNA polymerase (Roche) in 1× Reaction Buffer, 0.2 mmol/L of each dNTP, 1 μmol/L of each primer, 50 ng genomic DNA of Microtus maximowiczii and 2 μL of DNA extract (DNA concentration not quantified). Thermal cycling conditions were: 94°C for 2 min; 35 cycles of 94°C for 20 sec, 56°/57°C for 30 sec, and 72°C for 60/40 sec; and a final extension at 72°C for 7 min. The primers were Pro+ (5'-ACC ATC AGC ACC CAA AGC TG-3') and Phe- (5'-AAG CAT TTT CAG TGC TTT GCT T-3') for the ~1000 bp fragment, and Pro+ and mico3- (5'-GTA AAA GAA GCA TTA ATT AAA-3') for the ~300 bp fragment . None of the ancient DNA extracts inhibited amplification of the test PCRs, as assessed by three replicates. PCR inhibition through ancient DNA extracts was therefore not considered a significant problem for the selected samples.
Precautions against contamination
The bone samples were processed independently in the DNA labs of the Natural History Museum Vienna and the Natural History Museum Oslo following standards to avoid contamination of samples: These include in particular, usage of DNA, DNAse, and RNAse-free consumables and reagents, as well as pipette tips with aerosol protection filters, and UV-treatment of tubes prior to extraction and PCR amplifications. Negative controls for extractions (without any sample) as well as PCR reactions (without any DNA) were included in the experimental setup. The samples from Hainburg (Lower Austria, Bronze Age) were considered as additional negative controls. Samples from presumably TB-containing cases were never processed before the experiments on the negative controls and samples of unknown infection status had been finished. In both labs no studies of modern mycobacterial or human DNA had been performed before, and especially no modern mycobacterial DNA was used as positive PCR controls.
When subjecting the long bone samples from Kaiserebersdorf, Austria, as well as the positive and negative controls to PCR amplification targeting various regions in the genome of Mycobacterium tuberculosis, several PCR products were obtained. OxyR amplified for the skull WB594 using the primer pair R1/F4. The PCR targeting the IS6110 insertion sequences yielded PCR products of appropriate size for the skulls WB354, WB565, and WB594 using primer pair TbA/TbB, and for the long bones KE23 and HB284, and the skulls WB354 and WB594 using primer pair TbC/TbD.
When sequencing the obtained PCR products only WB354 amplified with the primer pair TbC/TbD yielded a sequence identical with the targeted IS6110 insertion sequences of M. tuberculosis (reference sequence: AF 181860), and was therefore considered positive. The sequence has been deposited in GenBank under the accession number FJ177511. None of the other cloned PCR products yielded sequences with similarity to M. tuberculosis but proved unspecific PCR amplification (sequence similarity to, e.g., sequences from Pseudomonas or various fungi).
With the exception of the unspecific amplification of non-target DNA of samples KE23 and HB284 when using the TbC/TbD primer pair, DNA extracts from long bones never yielded PCR products that were visible as distinct band of appropriate size on ethidium-bromide stained agarose gels. Even reamplification attempts were negative for the various primer pairs.
In the present study we aimed at diagnosing tuberculosis in historic human long bones from 18th century burials in Kaiserebersdorf, Austria. A positive result, i.e., PCR amplification of a DNA fragment with a nucleotide sequence matching the IS6110 insertion sequences of M. tuberculosis deposited in GenBank, was only obtained for the skull WB354. This was not unexpected since this sample served as positive control and tuberculosis had been diagnosed earlier in this individual by means of the spoligotyping approach . The amplification of the IS6110 insertion sequence proofs the suitability of the applied PCR approach. However, also for this sample positive PCR amplification was only obtained with the primer pair TbC/TbD that targets the repetitive insertion sequences IS6110, but not for any of the other molecular markers tested. This may indicate that only very limited amount of template DNA could be extracted from this particular sample. The failure to amplify any tuberculosis specific target DNA from the other positive skull samples may further support this interpretation. It should be considered that a M. tuberculosis strain which did not contain IS6110 has been reported earlier . Thus, the use of IS6110 as a single marker for PCR-facilitated detection M. tuberculosis might be insufficient. Nevertheless, the IS6110 insertion sequence is very likely the most sensitive marker  in the M. tuberculosis genome and yielded positive results while other markers such as, e.g., oxyR failed .
However, even the IS6110 insertion sequence could not de determined in the long bones of the samples KE20 and KE23 from the poor house in Kaiserebersdorf, Austria. Given the difficulties in amplifying target DNA of M. tuberculosis in the positive skull samples, it may not be surprising that none of the test bones from the burials in Kaiserebersdorf yielded any specific PCR product, although it seems likely from other anthropological evidence that at least samples KE20 and KE23 relate to individuals that suffered from tuberculosis.
Several reasons can account for the negative results on the long bone samples from Kaiserebersdorf: First, some individuals may not have suffered from tuberculosis. This can almost certainly be excluded for samples KE20 and KE23. Second, no DNA was preserved in the samples from Kaiserebersdorf. With the data at hand, we can currently not reject such interpretation. We refrained from amplifying other human target DNA in order to test for the presence of amplifiable DNA. The human actin gene has been suggested for this purpose . However, the samples from Kaiserebersdorf have been handled extensively by humans, and proof of authenticity of any human target DNA would have been close to impossible. Third, the presence of PCR inhibitors in the DNA extracts may have prevented successful amplifications. According to the results of our spiking experiments, this option can be excluded. No sample from Kaiserebersdorf, Austria, prevented successful amplification of high quality test DNA. A further reason for the failure of amplifying M. tuberculosis DNA from the samples from Kaiserebersdorf may relate to the bone samples themselves. Long bones such as, e.g., the femurs that are rarely affected by tuberculosis may represent material that is hardly suitable for the molecular detection of M. tuberculosis. This interpretation may hold for the particular samples under study, but it has been shown earlier that M. tuberculosis can be detected in historic femur samples  although with probably lower success rate.
In the current study, we failed to positively diagnose tuberculosis in DNA extracts from historic human long bones from inmates of a poor house in Kaiserebersdorf, Austria. Using thoracic or lumbar vertebrae for diagnosing tuberculosis, i.e., bones that are severely affected by the disease might have been more successful than long bones. This issue will be further explored in a future study. We conclude that the unpredictable state of DNA preservation in bones from museum collections does not allow any general recommendation of any type of bone.
We gratefully acknowledge the funding by the SYNTHESYS (Synthesis of Systematic Resources) program to LB in 2006 (AT-TAF-2968) for working at the Natural History Museum, Vienna, Austria.
- Konomi N, Lebwohl E, Mowbray K, Tattersall I, Zhang D: Detection of Mycobacterial DNA in Andean Mummies. Journal of Clinical Microbiology. 2002, 40: 4738-4740. 10.1128/JCM.40.12.4738-4740.2002.PubMed CentralView ArticlePubMedGoogle Scholar
- Taylor GM, Young DB, Mays SA: Genotypic analysis of the earliest known prehistoric case of tuberculosis in Britain. Journal of Clinical Microbiology. 2005, 43: 2236-2240. 10.1128/JCM.43.5.2236-2240.2005.PubMed CentralView ArticlePubMedGoogle Scholar
- University Press of Florida, Roberts CA, Buikstra JE: The Bioarchaeology of Tuberculosis, A Global View on a Reemerging Disease. 2003, University Press of FloridaGoogle Scholar
- Spigelman M, Lemma E: The use of the polymerase chain reaction to detect Mycobacterium tuberculosis in ancient skeletons. International Journal of Osteoarchaeology. 1993, 3: 137-143. 10.1002/oa.1390030211.View ArticleGoogle Scholar
- Spigelman M, Matheson C, Lev G, Greenblatt C, Donoghue HD: Confirmation of the presence of Mycobacterium tuberculosis complex-specific DNA in three archaeological specimens. International Journal of Osteoarchaeology. 2002, 12: 393-401. 10.1002/oa.638.View ArticleGoogle Scholar
- Nerlich AG, Haas CJ, Zink A, Szeimies U, Hagedorn HG: Molecular evidence for tuberculosis in an ancient Egyptian mummy. The Lancet. 1997, 350: 1404-10.1016/S0140-6736(05)65185-9.View ArticleGoogle Scholar
- Fletcher HA, Donoghue HD, Taylor GM, Zanden van der AGM, Spigelman M: Molecular analysis of Mycobacterium tuberculosis DNA from a family of 18th century Hungarians. Microbiology. 2003, 149: 143-151. 10.1099/mic.0.25961-0.View ArticlePubMedGoogle Scholar
- Taylor GM, Murphy E, Hopkins R, Rutland P, Christov Y: First report of Mycobacterium bovis DNA in human remains from the Iron Age. Microbiology. 2007, 153: 1243-1249. 10.1099/mic.0.2006/002154-0.View ArticlePubMedGoogle Scholar
- Zink AR, Haas CJ, Reischl U, Szeimies U, Nerlich AG: Molecular analysis of skeletal tuberculosis in an ancient Egyptian population. J Med Microbiol. 2001, 50 (4): 355-366.View ArticlePubMedGoogle Scholar
- Zink AR, Grabner W, Nerlich AG: Molecular identification of human tuberculosis in recent and historic bone tissue samples: the role of molecular techniques for the study of historic tuberculosis. American Journal of Physical Anthropology. 2005, 126: 32-47. 10.1002/ajpa.10409.View ArticlePubMedGoogle Scholar
- Fusegawa H, Wang BH, Sakurai K, Nagasawa K, Okauchi M, Nagakura K: Outbreak of tuberculosis in a 2000-year-old Chinese population. Kansenshogaku Zasshi. 2003, 77: 146-149.View ArticlePubMedGoogle Scholar
- Kaiser C, Bachmeier B, Conrad C, Nerlich A, Bratzke H, Eisenmenger W, Peschel O: Molecular study of time dependent changes in DNA stability in soil buried skeletal residues. Forensic Science International. 2008, 177: 32-36. 10.1016/j.forsciint.2007.10.005.View ArticlePubMedGoogle Scholar
- Teschler-Nicola M, Schamall D, Müller M: Pathological finds in a poor house cemetery of the 18th Century (Wien-Kaiserebersdorf) – preliminary results. Paleopathology Association, 34th Annual Meeting. 2007, 31-AbstractvolumeGoogle Scholar
- Prodinger W, Gerold F, Parson W, Teschler-Nicola M: Möglichkeiten einer Subtypisierung von Mycobacterium tuberculosis bei tuberkulösen Läsionen in anthropologischem Material: explorartive Studie an Calvarien der anthropologischen Sammlung Weisbach (19. JH AD). Proceedings 3. Kongress der Gesellschaft für Anthropologie (GfA), Göttingen. Edited by: Schultz M, Christiansen K, Greil H, Henke W, Kemkes-Grottenthaler A, Niemitz C, Rothe H, Schiefenhövel W, Schmidt HD, Schröder I, Schutkowski H, Teschler-Nicola M, Wittwer-Backofen U. 2000, Schnittstelle Mensch – Umwelt in Vergangenheit, Gegenwart und Zukunft, 354-356.Google Scholar
- Höss M, Pääbo S: DNA extraction from Pleistocene bones by a silica-based purification method. Nucleic Acids Research. 1993, 21: 3913-3914. 10.1093/nar/21.16.3913.PubMed CentralView ArticlePubMedGoogle Scholar
- Borge T, Bachmann L, Bjørnstad G, Wiig Ø: Genetic variation in Holocene bowhead whales from Svalbard. Molecular Ecology. 2007, 16: 2223-2235. 10.1111/j.1365-294X.2007.03287.x.View ArticlePubMedGoogle Scholar
- Eisenach KD, Cave MD, Bates JH, Crawford JT: Polymerase chain reaction amplification of a repetitive DNA sequence specific for Mycobacterium tuberculosis. Journal of Infectious Diseases. 1990, 161: 977-981.View ArticlePubMedGoogle Scholar
- Cave MD, Eisenach KD, McDermott PF, Bates JH, Crawford JT: IS6110: conservation of sequence in the Mycobacterium tuberculosis complex and its utilization in DNA fingerprinting. Molecular Cell Probes. 1991, 5: 73-80. 10.1016/0890-8508(91)90040-Q.View ArticleGoogle Scholar
- Donoghue HD, Spigelman M, Greenblatt CL, Lev-Maor G, Kahila Bar-Gal G, Matheson C, Vernon K, Nerlich AG, Zink AR: Tuberculosis: from prehistory to Robert Koch, as revealed by ancient DNA. The Lancet Infectious Diseases. 2004, 4: 584-592. 10.1016/S1473-3099(04)01133-8.View ArticlePubMedGoogle Scholar
- Pusch CM, Bachmann L: Spiking of contemporary human template DNA with ancient DNA extracts induces mutations under PCR and generates non-authentic mitochondrial sequences. Molecular Biology and Evolution. 2004, 21: 957-964. 10.1093/molbev/msh107.View ArticlePubMedGoogle Scholar
- Haring E, Herzig-Straschil B, Spitzenberger F: Phylogenetic analysis of Alpine voles of the Microtus multiplex complex using the mitochondrial control region. Journal of Zoological Systematics and Evolutionary Research. 2000, 38: 231-238. 10.1046/j.1439-0469.2000.384139.x.View ArticleGoogle Scholar
- van Soolingen D, de Haas PE, Hermans PW, Groenen PM, van Embden JD: Comparison of various repetitive DNA elements as genetic markers for strain differentiation and epidemiology of Mycobacterium tuberculosis. Journal of Clinical Microbiology. 1993, 31: 1987-1995.PubMed CentralPubMedGoogle Scholar
- Mays S, Fysh E, Taylor GM: Investigation of the link between visceral surface rib lesions and tuberculosis in a medieval skeletal series from England using ancient DNA. American Journal of Physical Anthropology. 2002, 119: 27-36. 10.1002/ajpa.10099.View ArticlePubMedGoogle Scholar
- Ghossein RA, Ross DG, Salomon RN, Rabson AR: A search for mycobacterial DNA in sarcoidosis using the polymerase chain reaction. American Journal of Clinical Pathology. 1994, 101: 733-737.PubMedGoogle Scholar
- Brosch R, Gordon V, Marmiesse M, Brodin P, Buchrieser C, Eiglmeier K, Garnier T, Gutierrez C, Hewinson G, Kremer K, Parsons LM, Pym AS, Samper S, van Soolingen D, Cole ST: A new evolutionary scenario for the Mycobacterium tuberculosis complex. Proceedings of the National Academy of Sciences USA. 2002, 99: 3684-3689. 10.1073/pnas.052548299.View ArticleGoogle Scholar