Molecular analysis of Mycobacterium tuberculosis DNA from a family of 18th century Hungarians

Helen A. Fletcher1,{dagger}, Helen D. Donoghue1, G. Michael Taylor2, Adri G. M. van der Zanden3 and Mark Spigelman1

1 Department of Medical Microbiology, University College London, London W1T 4JF, UK
2 Imperial College London, Centre for Microbiology and Infection, Flowers Building, Armstrong Rd, London SW7 2AZ, UK
3 Medical Microbiology and Infectious Diseases, Location Lukas, Gelre hospitals, Albert Schweitzerlaan 31, 7300 DS Apeldoorn, The Netherlands

Correspondence
Helen D. Donoghue
h.donoghue{at}ucl.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The naturally mummified remains of a mother and two daughters found in an 18th century Hungarian crypt were analysed, using multiple molecular genetic techniques to examine the epidemiology and evolution of tuberculosis. DNA was amplified from a number of targets on the Mycobacterium tuberculosis genome, including DNA from IS6110, gyrA, katG codon 463, oxyR, dnaA–dnaN, mtp40, plcD and the direct repeat (DR) region. The strains present in the mummified remains were identified as M. tuberculosis and not Mycobacterium bovis, from katG and gyrA genotyping, PCR from the oxyR and mtp40 loci, and spoligotyping. Spoligotyping divided the samples into two strain types, and screening for a deletion in the MT1801–plcD region initially divided the strains into three types. Further investigation showed, however, that an apparent deletion was due to poor DNA preservation. By comparing the effect of PCR target size on the yield of amplicon, a clear difference was shown between 18th century and modern M. tuberculosis DNA. A two-centre system was used to confirm the findings of this study, which clearly demonstrate the value of using molecular genetic techniques to study historical cases of tuberculosis and the care required in drawing conclusions. The genotyping and spoligotyping results are consistent with the most recent theory of the evolution and spread of the modern tuberculosis epidemic.

Abbreviations: MTB, Mycobacterium tuberculosis

{dagger}Present address: Molecular Immunology (Hill Group), Nuffield Department of Clinical Medicine, John Radcliffe Hospital, Oxford OX3 9DU, UK.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Building works carried out in 1994 at the Dominican church in Vác, Hungary, led to the chance discovery of two crypts, which had been sealed since 1838. One of these was found to contain a total of 263 human bodies within wooden coffins; about two-thirds of the bodies were naturally mummified. The human remains proved to be those of prominent local families and clerics accorded interment between the years 1731 and 1838. Information from the coffin plates and local archival records, combined with the appearances of some of the bodies, led to the suggestion that many of the individuals were suffering from tuberculosis at the time of death. Certainly, tuberculosis was one of the most prevalent infections in Europe during this period, accounting for one in four deaths from the 16th to 18th centuries (Dubos & Dubos, 1952). The presence of tuberculosis in the Vác samples was confirmed by computerized tomographic (CT) examination of one of the bodies at the Natural History Museum in Budapest, histopathology of lung tissue, and a pilot study using PCR to amplify the IS6110 element specific to the Mycobacterium tuberculosis (MTB) complex (Thierry et al., 1990; Páp et al., 1999). Preliminary testing on the Vác remains suggested that the preservation of DNA was sufficiently robust to permit in-depth study of the mycobacterial DNA from different strains persisting in the bodies.

Contemporary written records were available for many of the individuals found in the crypt, including date of death, age, sex, family name, relatives and, in some cases, a brief description of the cause of death. These records enabled several bodies to be placed into family groups. Of particular interest in this respect was a small family group consisting of a mother and two daughters, all of whom had died over a four year period between 1793 and 1797. Body 28 was that of a 55-year-old woman who died on 16 December 1793, the mother of the two other individuals reported here. Her height was noted as 1·45 m. Soft tissue was taken from the tracheal region and abdomen of body 28, using Stortz endoscopes and an aseptic technique as precautions against cross-contamination. Body 68 was the daughter of the above, and she died on 25 December 1797 aged 28 years. No abnormalities were detected in a chest radiograph of body 68 but she was noticeably small for her age and of cachectic appearance. A tissue sample from the chest was examined. Body 72 was the younger daughter of body 28 and sister to body 68, and she died on 2 March 1795. Although the archive stated that body 72 was aged 14 years at death, when examined she was initially believed to be aged only eight or nine years due to her very small size and cachectic appearance. No abnormalities were seen in a chest radiograph. Samples of body 72 were taken from the abdomen, chest and from possible calcified pleura.

Several studies have examined skeletal and mummified remains for the presence of mycobacterial DNA (Spigelman & Lemma, 1993; Salo et al., 1994; Baron et al., 1996; Faerman et al., 1997; Nerlich et al., 1997; Braun et al., 1998; Crubézy et al., 1998; Donoghue et al., 1998; Taylor et al., 1996, 1999, 2001; Haas et al., 2000; Mays et al., 2001; Zink et al., 2001). However, the state of surviving DNA has only seldomly permitted detailed molecular examination of samples (Mays et al., 2001). Consequently, few population-based studies have been reported for ancient tuberculosis (Zink et al., 2001). However, the completion of the sequencing of the MTB genome (Cole et al., 1998) and comparative genomic studies of members of the MTB complex have revealed an increasing number of targets for the study of historical cases of tuberculosis. In the present study a number of PCR-based methods were applied to the MTB complex strains from the family group described above. The aim was to evaluate the use of these markers for the analysis of archaeological specimens and to assess the extent to which such studies may contribute to increasing our understanding of the evolution and molecular epidemiology of tuberculosis.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Samples.
Taking precautions against cross-contamination, samples of dehydrated soft tissue were collected from the selected family group, bodies 28, 68 and 72, respectively (Table 1). Samples were stored in sterile universal bottles at 4 °C until analysis.


View this table:
[in this window]
[in a new window]
 
Table 1. Samples collected from the family group studied here

 
Precautions against contamination.
Stringent precautions were taken against laboratory cross-contamination: clean protective clothing was worn; gloves were changed frequently; sterile tubes and aerosol-resistant tips were used. A two-laboratory, three-workstation strategy was used for DNA extraction and PCR amplification, which also served to verify the results. Pipettes, other equipment and surfaces were cleaned with concentrated household detergent or bleach, rinsed with ultrapure water and dried with ethanol. Molecular biology grade reagents (sterile where possible) were purchased and used only for the ancient DNA studies. Pre-aliquoted PCR master mix was used in laboratory one (Abgene). In laboratory two, the UNG-ready Excite core kit with hot-start Taq polymerase (BioGene) was used in later experiments for confirmation of genotyping experiments.

Reproducibility.
Data were verified by repeating PCRs on the same and on fresh extracts of samples. In addition, in several cases duplicate reactions were carried out independently in one of the other collaborating laboratories.

DNA extraction.
DNA extraction in laboratory one was performed using an adaptation of the Boom method (Boom et al., 1990) as described previously (Donoghue et al., 1998). Repeat extracts were performed using the Qiagen DNeasy Tissue Kit with the following adaptations: 25 mg of sample was added to a 1·5 ml tube containing 1·5–2 mm glass beads and incubated overnight at 56 °C with 200 µl digestion buffer [0·5 M EDTA, pH 8·0 (Promega), 0·4 mg proteinase K ml-1 (Finnzymes)]. Each sample was homogenized for 50 s at medium speed with a Mini Bead Beater (Stratech Scientific) and transferred to a column where it was further processed according to the manufacturer's instructions. In laboratory two DNA was extracted using a silica-based method as described previously (Taylor et al., 1999).

Screening methods.
Primer sequences, amplification conditions and details of where the work was carried out are given in Table 2. Samples were screened for an MTB-complex-specific region of the insertion sequence IS6110 (Eisenach et al., 1990) using a two-tube nested PCR as described previously (Taylor et al., 1996; Donoghue et al., 1998). Thereafter, target sequences in the 19 kDa antigen gene (Mustafa et al., 1995) and the dnaA–dnaN region were examined.


View this table:
[in this window]
[in a new window]
 
Table 2. Primer sequences and PCR conditions used in laboratories one, two and three

An initial denaturation step at 94 °C and a final extension step at 72 °C were used for all PCR amplifications.

 
Genotyping.
To place the strains into genotypic groups (Sreevatsan et al., 1997; Frothingham et al., 1999), polymorphisms at codons 203 and 463 of katG and codon 95 of gyrA were determined (Table 3). This was performed independently in laboratories one and two.


View this table:
[in this window]
[in a new window]
 
Table 3. Mutations in the katG and gyrA codons enabling division of MTB complex strains into genotypic groups 1A, 1B, 2 and 3

Data are from Sreevatsan et al. (1997) and Frothingham et al. (1999)*. Mutated nucleotides are underlined.

 
Distinction of MTB from Mycobacterium bovis.
Samples from the three family members were screened for the presence of the mtp40 element, located in the phospholipase A gene (plcA) (Del Portillo et al., 1991). This element is retained in around 90 % of MTB isolates but lost as part of a deletion event (RD5) in the majority of M. bovis strains (Del Portillo et al., 1996; Gordon et al., 1999). PCR for mtp40 was multiplexed with a second method, designed to amplify a region in the plcD-cutinase region of strain CDC 1551 known as MT1801. This region, deleted from the H37Rv laboratory strain of MTB (RvD2), is a hot-spot site for IS6110 insertion and is generally subject to deletion events. PCR for the MT1801 oxidoreductase DNA has been used as an indicator of deletion events in this region (Ho et al., 2000). The multiplex PCR method therefore yields information as to the presence of all four phospholipase gene copies (plcA, plcB, plcC and plcD). Amplification of these target sites (mtp40 and the plcD loci) was performed in laboratory two using a hot-start method as described previously (Taylor et al., 1999), as was determination of the polymorphism at nucleotide 285 in the oxyR pseudogene. In laboratory one, PCR of the plcD region was carried out using a shorter target sequence of 123 bp, which can be included in a nested reaction.

RD7 PCR.
As a further screening test to refute the presence of M. bovis DNA in the samples, flanking primers were used in a PCR to detect deletion region 7 (RD7). This deletion, of around 12·7 kb, extends from base pair 2 208 003 to base pair 2 220 721 in the MTB genome (Zumárraga et al., 1999; Gordon et al., 1999). A product of 211 bp is amplified from M. bovis isolates but under the conditions used no product is expected from MTB, particularly in fragmented archaeological material (Mays et al., 2001).

Molecular fingerprinting.
Spoligotyping (Kamerbeek et al., 1997) was carried out in laboratories two (Taylor et al., 1999) and three (van der Zanden et al., 1998).

Detection and analysis of amplified DNA.
In laboratories one and two, PCR products were electrophoresed in a 3·0 % (w/v) NuSieve (Flowgen) or agarose gel, respectively. DNA was visualized by ethidium bromide staining plus UV light. Images were recorded with a digital image capture system. Products for sequencing were run subsequently on 0·8 % low-melting-point agarose gels. DNA was purified and sequenced using in-house sequencing services or MWG-Biotech.

Quantification and amplification efficiency.
In laboratory one, the amount of amplified product obtained from the 18th century samples was determined in a single experiment by gel densiometric analysis using a DNA mass marker (Gibco) and LABWORKS software (UltraViolet Products). This was compared with DNA from an M. bovis culture, strain AN-5, which was prepared and set up in an isolated, totally separate laboratory, which was under negative air pressure. Four target sequences were used for this comparison, IS6110 using the inner primers (92 bp), the 19 kDa antigen gene (131 bp), the dnaA–dnaN region (159 bp) and the MPB70 antigen gene (372 bp). The quantity of amplicon was converted to the number of copies per microlitre using Avogadro's constant. Modern DNA was used to ascertain whether differences in yield were due to variations in amplification efficiency between the PCRs. A dilution curve (106 to 101) of each product was subjected to real-time PCR amplification (ABI 7000 SDS; Applied Biosystems) using Qiagen Quantitect master mix.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Evidence of MTB infection
All samples were strongly positive with both nested and single-stage PCR for IS6110. Positive results were also obtained consistently from PCRs based on the 19 kDa antigen and the dnaA–dnaN target sequence, and from spoligotyping.

Genotypic grouping
Fragments of DNA spanning the regions of polymorphisms in the katG and gyrA genes (Table 3) were amplified and sequenced in laboratories one and two. Body 28 classified as group 2 with katG codon 463 CGG, katG 203 ACC and gyrA 95 ACC. Bodies 68 and 72 classified as group 3 with katG 463 CGG, katG 203 ACC and gyrA 95 AGC (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 4. Summary of results obtained for bodies 28, 68 and 72

All three samples were positive for the presence of IS6110 and negative for an IS6110 fragment in the dnaAdnaN region, indicating that they were members of the MTB complex but not of the group 1 subtype. An IS6110 fragment in the dnaAdnaN region is typical of members of the multi-drug-resistant Beijing family of strains. Also, the M. bovis-specific deletion region 7 (RD7) was not amplified from any of the samples.

 
oxyR pseudogene
Sequencing of both strands of the PCR products from bodies 28, 68 and 72 revealed a guanine residue at position 285 characteristic of MTB, as opposed to an adenine residue characteristic of M. bovis (Sreevatsan et al., 1996).

mtp40 and plcD-cutinase PCR
Samples from all three individuals were positive for the mtp40 fragment, which strongly indicated that the DNA was from MTB rather than M. bovis (Table 4). Samples from bodies 28 (abdomen and trachea) and 68 (chest) were consistently positive for plcD but body 72 yielded a negative result with single-stage PCR based on the larger target sequence (177 bp; Fig. 1a). However, in laboratory one, with single-stage PCR of the 123 bp target region, some positive results were obtained from body 72, although samples from the chest, pleura and abdomen differed in the quantity of amplifiable MTB DNA, and repeated DNA extracts also yielded differing amounts. Nested PCR of apparently negative PCR products of either the 177 or 123 bp target regions was required to give clear positive results from body 72 (Fig. 1b).



View larger version (96K):
[in this window]
[in a new window]
 
Fig. 1. (a) Multiplex PCR to screen for a deletion in plcD. Lanes: 1, 100 bp DNA size marker; 2, strain CDC 1551; 3, strain H37Rv; 4, extraction control; 5, PCR control; 6, body 72; 7, body 68; 8, body 28. (b) Examination of effect of PCR target size in plcD. Lanes: 1, 123 bp DNA ladder; 2, body 28 (177 bp); 3, body 28 (123 bp); 4, body 28 (nested PCR, 102 bp); 5, body 68 (177 bp); 6, body 68 (123 bp); 7, body 72 (177 bp); 8, body 72 (123 bp); 9, body 72 (nested PCR, 123 bp); 10, body 72 (nested PCR, 102 bp); 11, {phi}X174/HaeIII molecular markers.

 
RD7 PCR
No product was obtained from any of the samples, which was the result expected from MTB DNA.

Screening for IS6110 in the dnaA–dnaN region
A 159 bp product was amplified from the Vác samples with primers Pr1 and Pr2 and no product was obtained with primers Pr2 and Pr3 (Table 4), indicating that there was no insertion sequence in this region specific for the Beijing family of multi-drug-resistant MTB (Kurepina et al., 1998).

Spoligotyping
The spoligotyping patterns obtained from the extracts from bodies 28, 68 and 72 were remarkably complete (Fig. 2a). Two spoligotype patterns were observed for the three bodies, patterns 50 and 53 (Sola et al., 1999), which can also be described by an octal code as 777777777730771 and 777777777760771, respectively (Dale et al., 2001). Body 28 (mother) had pattern 53 and bodies 68 and 72 (daughters) had pattern 50. To confirm these results, spoligotyping was repeated in laboratory two using separate extracts from the bodies. In addition, a third set of extracts from the same bodies was taken to laboratory three for spoligotyping. Fig. 2(b) is a schematic representation of the results obtained from the two laboratories. The extracts from body 72 prepared in laboratory one and amplified in laboratory two gave extremely faint hybridization in spacers 29, 30 and 32, although these spacers were negative when the DNA was amplified according to the normal protocol used in laboratory three.



View larger version (48K):
[in this window]
[in a new window]
 
Fig. 2. (a) Spoligotyping of DNA from bodies 28, 68 and 72. Rows: 1, strain H37Rv; 2, PCR control; 3, body 28; 4, body 68; 5, body 72 (spacers 28, 29, 30 and 32 are present but faint). (b) Schematic representation of spoligotyping results from London and The Netherlands. Suffixes: L1, extract 1 typed in London; L2, extract 2 typed in London; N3, extract 3 typed in The Netherlands.

 
Quantification and amplification efficiency
From densiometric analysis the quantity of amplicon from modern M. bovis DNA was similar for each PCR and was not related to the size of target sequence (Table 5). The 18th century Hungarian samples gave a comparable yield for the IS6110 PCR (92 bp). Differences between samples were apparent with the other PCRs and, overall, there was a decline in copy number of amplicon per microlitre with the size of target sequence. Only one sample (body 68 chest) gave a comparable yield to modern DNA with the MPB70 PCR (372 bp), and two samples (body 28 trachea and body 72 abdomen) gave no detectable amplicon. Real-time amplification of dilution curves demonstrated that there was no relationship between amplification efficiency and PCR product size. In a plot of threshold cycle (detection level) versus log of the starting copy number, the greater the y value the greater the efficiency of amplification, yet values were as follows: IS6110 (92 bp) y=-3·8124; 19 kDa antigen gene (131 bp) y=-3·68439; dnaAdnaN region (159 bp) y=-3·5276; MPB70 antigen gene (372 bp) y=-3·9832.


View this table:
[in this window]
[in a new window]
 
Table 5. Quantity of amplified DNA (copies per microlitre) after 40 rounds of amplification

Data are from one experiment.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
This study of MTB DNA from a family of 18th century Hungarians has revealed several points of interest. It is apparent that the MTB DNA was remarkably well preserved. DNA fragments as it degrades, and the rate of degradation is dependent upon the environmental conditions in which the specimen is stored (Hoss et al., 1996). Light, moisture and temperature fluctuations are known to be particularly detrimental to DNA. Studies on ancient human DNA have shown that bones and teeth are the specimens most likely to yield amplifiable DNA and mummified tissue has been shown to be an extremely poor source (Lassen et al., 1994). However, MTB DNA was isolated consistently from the mummified tissues examined here. We have suggested elsewhere (Fletcher et al., 2003) that the properties of the MTB bacillus, which enable the organism to persist latently in the human lung, afford protection of the bacterial DNA post-mortem. This study demonstrates that naturally mummified tissue is likely to be a valuable resource for the palaeomicrobiologist. The reduction in the rate of PCR positives in relation to product size obtained from the mummified tissues was shown by real-time PCR to be due to DNA preservation and not to variation in amplification efficiency. The inverse relationship between PCR yield and target sequence size is a characteristic of ancient DNA (O'Rourke et al., 2000) and confirms that no contamination with modern MTB has occurred.

Bodies 68 and 72 were physically small in comparison to the majority of the remains found in the crypt, some of which were of individuals who lived to a great age and who showed no signs of malnourishment or wasting (Fletcher et al., 2003). Indeed, it is unlikely that the physical appearance of bodies 68 and 72 was due to malnutrition as it is known that only wealthy, middle class families were buried in the crypt. Therefore, the cachectic appearance of these bodies, in association with the age of death and ease of detection of MTB DNA, strongly suggests that death was as a result of active tuberculosis infection. Therefore, it can be concluded that the MTB characterized here was virulent.

Due to the close association of humans with domesticated animals, it was believed that many cases of tuberculosis in antiquity were due to M. bovis (Dankner et al., 1993). In addition, it has been suggested that as M. bovis can be transmitted via the consumption of dairy products it may be found infecting the abdomen. The abdomen of body 28 was found to be strongly positive for MTB; therefore, five different tests were performed that can differentiate between the DNA of MTB and M. bovis. All the results indicate that the DNA was from MTB and not M. bovis. M. bovis is a member of the MTB complex genotype 1, yet the strains we have analysed classify as genotypes 2 and 3. Finally, the presence of spacers 37–43 on the spoligotype confirm that these genotype 2 and 3 strains are MTB and not M. bovis. Initially, the failure to detect evidence of M. bovis infection, also reported by others (Mays et al., 2001), was surprising. However, M. bovis infections are zoonotic and person-to-person spread is thought to occur only in exceptional circumstances (Grange, 2001). Therefore, in a population with endemic MTB infection, M. bovis infections are always likely to comprise a small proportion of the total number of cases.

The findings from the dnaA–dnaN spacer region are consistent with both the genotyping and spoligotyping data. Beijing strains are of genotype 1 and have a unique spoligotype. In addition, the multi-drug-resistant strain W has a characteristic IS6110 insertion in this target region. The Beijing/W group of strains is believed to be of recent origin (Glynn et al., 2002), which is supported by the failure to detect any members of this group in the mummified material from Vác examined so far (H. D. Donoghue, unpublished observations).

The spoligotypes obtained from the Hungarian MTB DNA samples were remarkably reproducible. Patterns 50 and 53 are the patterns most commonly found across the globe today and are non-discriminatory. Sola et al. (1999) have proposed a phylogenetic tree, the root of which is pattern 53. Although tuberculosis occurred in antiquity (Sola et al., 1999), it is thought that the modern tuberculosis epidemic began in Europe in the 1700s, then moved to the New World and Africa (Stead et al., 1995). Consistent with the above theories, we have demonstrated that patterns 50 and 53 were present in Europe in the 18th century. Spoligotypes 50 and 53 differ by the absence of one spacer (spacer 31). Recent studies have demonstrated that in some cases the absence of spacer 31 in spoligotype 50 may be due to the presence of IS6110 in this region (Filliol et al., 2000; Legrand et al., 2001). In an extensive genotyping study it has been shown that strains with spoligotype 53 are groups 2 or 3 and that strains with spoligotype 50 are group 2 (Soini et al., 2000). However, our spoligotype pattern 50 strains are group 3 rather than group 2. It is possible that the strains we are observing are indeed spoligotype 53 with an IS6110 element in spacer 31 as opposed to a true type 50 (Legrand et al., 2001). Screening for the presence of IS6110 in spacer 31 of the direct repeat (DR) region should determine if this is the case, although so far we have been unsuccessful in obtaining sequence data from this region.

The present study, based on multiple targets in the MTB genome, has confirmed that the amplified DNA from the Hungarian samples was that of MTB. The two-centre system used to analyse the Hungarian samples confirms the inter- and intra-laboratory reproducibility of the results. Using MTB-specific primers followed by sequencing or probing, DNA has been amplified from eight loci on the MTB genome, including both single-copy and multiple sites. The detection of small-scale genomic deletions has been shown to be a useful technique for exploring the molecular epidemiology, microbial evolution and pathogenesis of tuberculosis (Ho et al., 2000; Kato-Maeda et al., 2001). Using a novel multiplex PCR technique we have demonstrated the use of screening historical cases of tuberculosis for such deletions. Spoligotyping revealed a difference between the strain infecting the mother and the strain infecting her two daughters; the mother typing as type 53 and the daughters as type 50. Using multiplex PCR of the mtp40 and plcD loci we initially concluded that there was a further difference between the strains infecting the two daughters, the isolate from the younger daughter having undergone a deletion in the plcD-oxidoreductase region that was detected using the larger of the two target sequences used (Fig. 1a, lane 6). About one-third of clinical isolates of MTB studied appear to have undergone a deletion in this region (Ho et al., 2000). However, the use of a shorter target sequence and nested PCR demonstrated (Fig. 1b) that this initial assumption was incorrect and that the material from the younger daughter was less-well preserved, shown by the difficulty in obtaining positive results when larger target sequences were used. The spoligotyping data also suggest poor preservation of part of the DR spacer regions from MTB in this individual. As determination of deletion subsets in this field is likely to be increasingly attempted for phylogenetic purposes, it will be important, as a check on DNA preservation, to multiplex PCRs such as plcD with primers to another target region that produce a slightly longer product.

In conclusion, it appears that the three members of the family group studied here were infected with two distinct strains of MTB. This is the first demonstration of spoligotyping combined with screening of small-scale deletions for determining the molecular epidemiology of MTB from archaeological material. In comparison with the strain infecting the mother, there was one point mutation and one deletion event, the loss of spacer 31, detected in the MTB from the two daughters. This supports the theory that MTB undergoes more deletion events as it evolves (Ho et al., 2000; Kato-Maeda et al., 2001; Brosch et al., 2002). However, screening of the entire genomes of these strains would be needed to confirm this. The findings are also consistent with the major genotypic grouping described by Sreevatsan et al. (1997), which is thought to broadly represent an evolutionary scenario for the MTB complex. Genotypes 2 and 3 are believed to be younger in evolutionary terms than genotype 1 organisms, so their demonstration in 18th century Europe supports the latest hypothesis of MTB evolution based on deletion analysis of the genomes of MTB and M. bovis (Brosch et al., 2002). Future studies adopting the methods described here should contribute further to our knowledge of the evolution and epidemiology of tuberculosis.


   ACKNOWLEDGEMENTS
 
We are grateful to Dr Ildikó Pap and her colleagues at the Anthropology Department, Hungarian Natural History Museum, Budapest, for assisting us in sampling the remains of the three individuals. Dr Barry Krieswerth kindly designed the primers to screen for a Beijing-specific IS6110 insertion. Sheima Abdulla Ahmed, Anneli Cooper and Farah Aladin gave excellent technical assistance. This research was funded primarily by the UK Wellcome Trust under award number 0516/Z/97/Z to M. S.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Baron, H., Hummel, S. & Herrmann, B. (1996). Mycobacterium tuberculosis complex DNA in ancient human bones. J Archaeol Sci 23, 667–671.[CrossRef]

Boom, R., Sol, C. J., Salimans, M. M., Jansen, C. L., Wertheim-van Dillen, P. M. & van der Noordaa, J. (1990). Rapid and simple method for purification of nucleic acids. J Clin Microbiol 28, 495–503.[Medline]

Braun, M., Cook, D. C. & Pfeiffer, S. (1998). DNA from Mycobacterium tuberculosis complex identified in North American, pre-Colombian human skeletal remains. J Archaeol Sci 25, 271–277.[CrossRef]

Brosch, R., Gordon, S. V., Marmiesse, M. & 12 other authors (2002). A new evolutionary scenario for the Mycobacterium tuberculosis complex. Proc Natl Acad Sci U S A 99, 3684–3689.[Abstract/Free Full Text]

Cole, S. T., Brosch, R., Parkhill, J. & 39 other authors (1998). Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393, 537–544.[CrossRef][Medline]

Crubézy, É., Ludes, B., Poveda, J.-D., Clayton, J., Crouau-Roy, B. & Montagnon, D. (1998). Identification of Mycobacterium DNA in an Egyptian Pott's case of 5,400 years old. C R Acad Sci III 321, 941–951.[Medline]

Dale, J. W., Brittain, D., Cataldi, A. A. & 11 other authors (2001). Spacer oligonucleotide typing of bacteria of the Mycobacterium tuberculosis complex: recommendations for standardised nomenclature. Int J Tuberc Lung Dis 5, 216–219.[Medline]

Dankner, W. M., Waecker, N. J., Essey, M. A., Moser, K., Thompson, M. & Davis, C. E. (1993). Mycobacterium bovis infections in San Diego: a clinicoepidemiologic study of 73 patients and a historical review of a forgotten pathogen. Medicine (Baltim) 72, 11–37.[Medline]

Del Portillo, P., Murillo, L. A. & Patarroyo, M. E. (1991). Amplification of a species-specific DNA fragment of Mycobacterium tuberculosis and its possible use in diagnosis. J Clin Microbiol 29, 2163–2168.[Medline]

Del Portillo, P., Thomas, M. C., Martinez, E., Maranon, C., Valladares, B., Patarroyo, M. E. & Carlos Lopez, M. (1996). Multiprimer PCR system for differential identification of mycobacteria in clinical samples. J Clin Microbiol 34, 324–328.[Abstract]

Donoghue, H. D., Spigelman, M., Zias, J., Gernaey-Child, A. M. & Minnikin, D. E. (1998). Mycobacterium tuberculosis complex DNA in calcified pleura from remains 1400 years old. Lett Appl Microbiol 27, 265–269.[CrossRef][Medline]

Dubos, R. & Dubos, J. (1952). The White Plague, Tuberculosis, Man and Society. Boston: Little Brown.

Eisenach, K. D., Cave, M. D., Bates, J. H. & Crawford, J. T. (1990). Polymerase chain reaction amplification of a repetitive DNA sequence specific for Mycobacterium tuberculosis. J Infect Dis 161, 977–981.[Medline]

Faerman, M., Jankauskas, R., Gorski, A., Bercovier, H. & Greenblatt, C. L. (1997). Prevalence of human tuberculosis in a medieval population of Lithuania studied by ancient DNA analysis. Ancient Biomolecules 1, 205–214.

Filliol, I., Sola, C. & Rastogi, N. (2000). Detection of a previously unamplified spacer within the DR locus of Mycobacterium tuberculosis: epidemiological implications. J Clin Microbiol 38, 1231–1234.[Abstract/Free Full Text]

Fletcher, H. A., Donoghue, H. D., Holton, J., Pap, I. & Spigelman, M. (2003). Widespread occurrence of Mycobacterium tuberculosis DNA from 18th–19th century Hungarians. Am J Phys Anthropol (in press

Frothingham, R., Strickland, P. L., Bretzel, G., Ramaswamy, S., Musser, J. M. & Williams, D. L. (1999). Phenotypic and genotypic characterization of Mycobacterium africanum isolates from West Africa. J Clin Microbiol 37, 1921–1926.[Abstract/Free Full Text]

Glynn, J. R., Whiteley, J., Bifani, P. J., Kremer, K. & van Soolingen, D. (2002). Worldwide occurrence of Beijing/W strains of Mycobacterium tuberculosis: a systematic review. Emerg Infect Dis 8, 843–849.[Medline]

Gordon, S. V., Brosch, R., Billaut, A., Garnier, T., Eiglemeier, K. & Cole, S. T. (1999). Identification of variable regions in the genomes of tubercle bacilli using bacterial artificial chromosome arrays. Mol Microbiol 32, 643–655.[CrossRef][Medline]

Grange, J. M. (2001). Mycobacterium bovis infection in human beings. Tuberculosis 81, 71–77.[CrossRef][Medline]

Haas, C. J., Zink, A., Molnar, E. & 7 other authors (2000). Molecular evidence for different stages of tuberculosis in ancient bone samples from Hungary. Am J Phys Anthropol 113, 293–304.[CrossRef][Medline]

Ho, T. B., Robertson, B. D., Taylor, G. M., Shaw, R. J. & Young, D. B. (2000). Comparison of Mycobacterium tuberculosis genomes reveals frequent deletions in a 20 kb variable region in clinical isolates. Yeast 4, 272–282.[CrossRef]

Hoss, M., Jaruga, P., Zastawny, T. H., Dizdaroglu, M. & Pääbo, S. (1996). DNA damage and DNA sequence retrieval from ancient tissues. Nucleic Acids Res 24, 1304–1307.[Abstract/Free Full Text]

Kamerbeek, J., Schouls, L., Kolk, A. & 8 other authors (1997). Simultaneous detection and strain differentiation of Mycobacterium tuberculosis for diagnosis and epidemiology. J Clin Microbiol 35, 907–914.[Abstract]

Kato-Maeda, M., Rhee, J. T., Gingeras, T. R., Salamon, H., Drenkow, J., Smittipat, N. & Small, P. M. (2001). Comparing genomes within the species Mycobacterium tuberculosis. Genome Res 4, 547–554.[CrossRef]

Kurepina, N. E., Sreevatsan, S., Plikaytis, B. B., Bifani, P. J., Connell, N. D., Donnelly, R. J., van Soolingen, D., Musser, J. M. & Kreiswirth, B. N. (1998). Characterization of the phylogenetic distribution and chromosomal insertion sites of five IS6110 elements in Mycobacterium tuberculosis: non-random integration in the dnaA–dnaN region. Tuber Lung Dis 79, 31–42.[CrossRef][Medline]

Lassen, C., Hummel, S. & Herrmann, B. (1994). Comparison of DNA extraction and amplification from ancient human bone and mummified soft tissue. Int J Leg Med 107, 152–155.[Medline]

Legrand, E., Filliol, I., Sola, C. & Rastogi, N. (2001). Use of spoligotyping to study the evolution of the direct repeat locus by IS6110 transposition in Mycobacterium tuberculosis. J Clin Microbiol 39, 1595–1599.[Abstract/Free Full Text]

Mays, S., Taylor, G. M., Legge, A. J., Young, D. B. & Turner-Walker, G. (2001). Paleopathological and biomolecular study of tuberculosis in a medieval skeletal collection from England. Am J Phys Anthropol 114, 298–311.[CrossRef][Medline]

Mustafa, A. S., Ahmed, A., Abal, A. T. & Chugh, T. D. (1995). Establishment and evaluation of a multiplex polymerase chain reaction for detection of mycobacteria and specific identification of Mycobacterium tuberculosis complex. Tuber Lung Dis 76, 336–343.[Medline]

Nerlich, A. G., Haas, C. J., Zink, A., Szeimies, U. & Hagedorn, H. G. (1997). Molecular evidence for tuberculosis in an ancient Egyptian mummy. Lancet 350, 1404.[Medline]

O'Rourke, D. H., Hayes, M. G. & Carlyle, S. W. (2000). Ancient DNA studies in physical anthropology. Annu Rev Anthropol 29, 60–66.

Pap, I., Józsa, L., Repa, I., Bajzik, G., Lakhani, S. R., Donoghue, H. D. & Spigelman, M. (1999). 18–19th century tuberculosis in naturally mummified individuals (Vác, Hungary). In: Tuberculosis: Past and Present, pp. 421–428. Edited by G. Pálfi, O. Dutour, J. Deák & I. Hutás. TB Foundation, Budapest: Golden Book.

Salo, W. L., Aufderheide, A. C., Buikstra, J. & Holcomb, T. A. (1994). Identification of Mycobacterium tuberculosis DNA in a pre-Columbian Peruvian mummy. Proc Natl Acad Sci U S A 6, 2091–2094.

Soini, H., Pan, X., Amin, A., Graviss, E. A., Siddiqui, A. & Musser, J. M. (2000). Characterization of Mycobacterium tuberculosis isolates from patients in Houston, Texas, by spoligotyping. J Clin Microbiol 38, 669–676.[Abstract/Free Full Text]

Sola, C., Devallois, A., Horgen, L., Maisetti, J., Filliol, I., Legrand, E. & Rastogi, N. (1999). Tuberculosis in the Caribbean: using spacer oligonucleotide typing to understand strain origin and transmission. Emerg Infect Dis 5, 404–414.[Medline]

Spigelman, M. & Lemma, E. (1993). The use of the polymerase chain reaction to detect Mycobacterium tuberculosis in ancient skeletons. Int J Osteoarchaeol 3, 137–143.

Sreevatsan, S., Escalante, P., Pan, X. & 11 other authors (1996). Identification of a polymorphic nucleotide in oxyR specific for Mycobacterium bovis. J Clin Microbiol 34, 2007–2010.[Abstract]

Sreevatsan, S., Pan, X., Stockbauer, K. E., Connell, N. D., Kreiswirth, B. N., Whittam, T. S. & Musser, J. M. (1997). Restricted structural gene polymorphism in the Mycobacterium tuberculosis complex indicates recent global dissemination. Proc Natl Acad Sci U S A 94, 9869–9874.[Abstract/Free Full Text]

Stead, W. W., Eisenach, K. D., Cave, M. D., Beggs, M. L., Templeton, G. L., Thoen, C. O. & Bates, J. H. (1995). When did Mycobacterium tuberculosis infection first occur in the New World? An important question with public health implications. Am J Respir Crit Care Med 151, 1267–1268.[Medline]

Taylor, G. M., Crossey, M., Salanha, J. A. & Waldron, T. (1996). Detection of Mycobacterium tuberculosis bacterial DNA in medieval human skeletal remains using polymerase chain reaction. J Archaeol Sci 23, 789–798.[CrossRef]

Taylor, G. M., Goyal, M., Legge, A. J., Shaw, R. J. & Young, D. (1999). Genotypic analysis of Mycobacterium tuberculosis from medieval human remains. Microbiology 145, 899–904.[Abstract]

Taylor, G. M., Mays, S., Legge, A. J., Ho, T. B. L. & Young, D. (2001). Genetic analysis of tuberculosis in human remains. Ancient Biomolecules 3, 267–280.

Thierry, D., Brisson-Noel, A., Vincent-Levy-Frebault, V., Nguyen, S., Guesdon, J. L. & Gicquel, B. (1990). Characterization of a Mycobacterium tuberculosis insertion sequence, IS6110, and its application in diagnosis. J Clin Microbiol 28, 2668–2673.[Medline]

van der Zanden, A. G., Hoentjen, A. H., Heilmann, F. G., Weltevreden, E. F., Schouls, L. M. & van Embden, J. D. (1998). Simultaneous detection and strain differentiation of Mycobacterium tuberculosis complex in paraffin wax embedded tissues and in stained microscopic preparations. Mol Pathol 51, 209–214.[Abstract]

Zink, A., Haas, C. J., Reischl, U., Szeimies, U. & Nerlich, A. G. (2001). Molecular analysis of skeletal tuberculosis in an ancient Egyptian population. J Med Microbiol 50, 355–366.[Abstract/Free Full Text]

Zumárraga, M., Bigi, F., Alito, A., Romano, M. I. & Cataldi, A. (1999). A 12·7 kb fragment of the Mycobacterium tuberculosis genome is not present in Mycobacterium bovis. Microbiology 145, 893–897.[Abstract]

Received 23 August 2002; revised 9 October 2002; accepted 10 October 2002.