SGM Journals
Research Article

Molecular detection of Yersinia pestis in dental pulp

Drancourt, Michel, Raoult, Didier

Microbiology 2004; 150(2):263 · https://doi.org/10.1099/mic.0.26885-0

Download PDF View at publisher PubMed

Summary auto-generated

This comment addresses a conflicting study by Gilbert et al. (2004) that failed to detect Yersinia pestis DNA in 108 teeth from suspected northern European plague sites, contradicting earlier findings by Drancourt and Raoult from southern Europe. The authors defend their methodology for detecting Y. pestis in medieval plague victims using dental pulp as a biological sample. They explain that dental pulp is superior to dentine or bone because it is well-vascularized, sterile in non-bacteremic patients, and equivalent to a small blood sample. They argue that Gilbert et al. did not accurately reproduce their techniques: only 7 of 108 samples had dental pulp tested, and most used dentine instead. Additionally, the authors identify technical differences in sample extraction methods and PCR protocols, noting that Gilbert et al. used less stringent PCR conditions and obtained non-specific amplicons from negative controls and dentine samples, suggesting contamination issues. The authors emphasize their use of nested PCR and proper dental pulp extraction techniques yielded consistent negative controls. They conclude that the conflicting results stem from methodological differences rather than failed reproduction of their work.

Key findings

  • Drancourt and Raoult successfully detected Y. pestis DNA in dental pulp from three southern European plague sites, confirming the pathogen's role in medieval plague in those regions.
  • Dental pulp is superior to dentine or bone for molecular detection of pathogens because it is well-vascularized, sterile in non-bacteremic patients, and provides reliable microbial DNA recovery.
  • Gilbert et al.'s study did not properly replicate the authors' techniques, testing dental pulp in only 7 of 108 specimens and using inferior extraction methods and less stringent PCR conditions.
  • Technical differences in PCR procedures and specimen contamination in Gilbert et al.'s work, including non-specific amplicons from negative controls, suggest their negative results reflect methodological limitations rather than absence of Y. pestis.
  • The authors propose that different microorganisms may have caused medieval plague epidemics in northern versus southern Europe, but more standardized work with proper techniques is needed to resolve this question.

This summary was generated automatically from the article PDF and is not part of the original publication. Refer to the PDF for the authoritative text.

Abstract

Gilbert et al. (2004) report on the lack of Yersinia pestis DNA detection in 108 teeth collected in five northern Europe suspected plague sites, thus challenging our discoveries in three southern Europe sites (Drancourt et al., 1998; Raoult et al., 2000). The interpretation of these negative results is either the absence of Y. pestis in tested specimens or a failure to detect it. Particularities in the epidemiology of Black Death in medieval northern Europe including the absence of the suitable rat species in some countries led to a controversy regarding the aetiology of Black Death in these countries. It is possible that different microbes were responsible for epidemics during the Middle Ages but present work does not help to resolve this issue. We detected specific Y. pestis sequences in three different sites in southern Europe, thus clearly demonstrating that Y. pestis was involved in historical plague epidemics in southern countries (Drancourt et al., 1998; Raoult et al., 2000). Basically, our strategy was based on the use of dental pulp as the sample and specific Y. pestis primers for detection.

The reason why we choose dental pulp is that it is a well vascularized soft tissue in contrast to bone or dentine (Drancourt et al., 1998). It is sterile in patients without bacteraemia (Potsch et al., 1992). It may be equivalent to a small blood sample. We were able to recover viable Coxiella burnetii after apparent cure from the dental pulp of experimentally infected animals (Aboudharam, 2001). That dental pulp is a suitable tissue on which to base molecular detection of microbial nucleic acid was demonstrated not only by our work on ancient plague (Drancourt et al., 1998; Raoult et al., 2000) but also in an experimental animal model (Aboudharam et al., 2000) and detection of specific RNA sequences in HIV-infected patients (Glick et al., 1989, 1991). In contrast, dentine has never been tested experimentally for this purpose. We doubt that it may be superior to bone as it is not a soft, well vascularized tissue. To use dentine and not the dental pulp makes the use of teeth a nonsense.

The work of Gilbert et al. (2004) did not test systematically our dental technique and in only seven suspected specimens was the dental pulp tested. This included five teeth from Verdun. We also tested three teeth from this place sent by E. Carniel and found them to be negative (unpublished data). We reported our findings on this matter a long time before the current study of Gilbert et al. (2004). Therefore, this result did not surprise us. Moreover, we found that the recovery of dental pulp via the tooth apex, as described by M. T. P. Gilbert, is almost ineffective. In our work, we always recovered dental pulp after complete opening of the dental pulp cavity. This approach is more difficult to perform and a dentist is probably necessary here. This allowed the recovery of all the pulp remnants. When this technique was applied, only two dental pulps, recovered by a different technique, from patients sampled in Copenhagen gave results different from ours. The use of the less satisfactory technique makes a very weak basis to contest our data.

As for PCR techniques, our approaches and results were different as well. PCR results varied depending on the source of Taq polymerase and the PCR conditions. It was significant that the results from samples examined by Gilbert et al. (2004) differed in laboratory 1 and laboratory 2. In their present work, a non-specific amplicon was obtained in five samples, including four dentine samples, using specific pla primers that we also used in our initial work. Also, two amplicons obtained from negative controls were not cloned or sequenced. This raises the question of specificity and stringency of the PCRs in the work of Gilbert et al. (2004). It is very different from what we reported. In our experience, negative controls were negative! We therefore believe that our test conditions were not reproduced. In our experiments with Black Death samples, we also used specific nested PCR which is, indeed, much more sensitive than regular PCR (Nolte & Caliendo, 2003). Gilbert et al. (2004) used on a second occasion universal primers and, of course, detected contaminating organisms, and non-specific amplicons were obtained using the universal rpob or 16S rRNA primers and dentine.

We therefore disagree with the authors' conclusion that they failed to reproduce our results since they did not use our techniques to obtain material or our PCR testing techniques. However, Gilbert et al. (2004) provide interesting approaches such as silicon embedding to limit external contamination of teeth. There is a need for more work on dental pulp specimens from northern Europe in order to resolve the aetiology(gies) of Black Death in these countries. We agree with Gilbert and colleagues on the usefulness of parallel work in two trained laboratories to ensure the validity of positive results. We were ready to do such work but were not invited. As ever, works performed in a single place are challenged until they can be reproduced independently. Our laboratory is open to train investigators so that they may be able to reproduce our work in their laboratories by reproducing our techniques.