Learning from mistakes: Taq polymerase contaminated with β-lactamase sequences results in false emergence of Streptococcus pneumoniae containing TEM

Sir,

In 2004, an Asiatic group published an article entitled ‘Study on the molecular epidemiology of β-lactamase TEM gene in isolated Streptococcus pneumoniae’.¹ This article can be found in the NCBI web site with the PMID number 15769331. The original article was written in Chinese, but in the published abstract, the authors reflected that 91.3% of their isolates had positive PCR amplifications for TEM. After sequencing these amplicons, the authors described new mutations in both TEM-1 and TEM-129 genes, which were submitted to the GenBank database and published under the accession numbers AY452662 and AY392531, respectively. The conclusion was that these results enriched the understanding of S. pneumoniae isolates exhibiting penicillin resistance.

The presence of bla_TEM genes had not been described in any Gram-positive bacteria until this report. Moreover, no type of β-lactamase had been found in S. pneumoniae; for these reasons, we were surprised that both the article and the report of these sequences in GenBank had no impact at all. The aim of the present work was to explore the putative presence and, if confirmed, the consequent prevalence of bla_TEM genes in a collection of well-characterized S. pneumoniae isolates with different levels of penicillin resistance.²

We analysed 33 high-level penicillin-resistant S. pneumoniae clones (including 7 Spain^23F-1, 11 Spain¹⁴-5, 13 Spain^6B-2 and 2 unique isolates) from different clinical sources showing penicillin MICs of 2–8 mg/L. The strains were serotyped at the National Centre of Microbiology (Instituto Carlos III, Majadahonda, Spain) and typed by PFGE-SmaI with the subsequent interpretation given by the Phoretix™ 5.0 software (Nonlinear Dynamics Ltd, UK).

Total pneumococcal DNA was extracted using the QIAamp^® kit (QIAGEN GmbH, Hilden, Germany), and the primers used for bla_TEM gene amplification were bla_TEM-F, 5′-ATGAGTATTCAACATTTCCG-3′ and bla_TEM-R, 5′-CTGACAGTTACCAATGCTTA-3′ following previously published conditions.³ Total DNA from bla_TEM-4-harbouring Klebsiella pneumoniae F40 strain and from S. pneumoniae R6 strain was used as PCR-positive and -negative controls, respectively. A negative control without bacterial DNA was included in each run.

The PCR experiments were performed with two different Taq polymerase enzymes: FastStart (Roche Diagnostics, Indianapolis, IN, USA) and Taq-Core (Qbiogene, Carlsbad, CA, USA). The nucleotide sequence of each positive PCR fragment obtained was determined with the ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction Kit and analysed with an ABI PRISM 377 automated sequencer (PE Applied Biosystems, Foster City, CA, USA). Hydrolysis assays using nitrocefin^® (Oxoid, Basingstoke, England) as substrate were also performed in crude extracts of all S. pneumoniae isolates.

When using the FastStart polymerase, a clear positive amplification with the same size as that of the bla_TEM gene was observed in all strains tested, including both negative controls. When the Taq-Core polymerase was used, positive amplifications were not detected in any of the isolates, except for the PCR-positive control corresponding to the K. pneumoniae F40 strain. Positive amplicons obtained with the FastStart polymerase were sequenced and after BLAST analysis, all of them corresponded to the previously described non-extended-spectrum β-lactamase (ESBL) TEM-116 (www.lahey.org).⁴ Nitrocefin test was consistently negative in all isolates.

This observation indicates that caution is required before accepting data about the presence of bla_TEM genes determined only with PCR-based experiments in the absence of adequate controls, particularly in species where this is an unusual or not previously described event. Moreover, the presence of bla_TEM gene sequences contaminating amplification reagents has been previously reported.^5,6 As PCR followed by nucleotide sequencing is the standard procedure for determining and characterizing bla genes, some epidemiological studies could be biased because of Taq enzyme contamination. Previous reports postulate that the contamination is frequently due to the presence of the bla_TEM-1a gene in the expression vectors commonly used to obtain recombinant amplification reagents.^5,6 In the particular case of the Taq polymerase purification process, remnant DNA may not be fully eliminated.

This observation suggests that it is quite desirable to ascertain and confirm any emergent unexpected result of potential importance, particularly in public health, by different research groups. Moreover, the publication of DNA sequences by different web sites should be submitted to a more strict control and referees should pay special attention when conflicting results such as ‘bla_TEM gene in S. pneumoniae’ nucleotide sequences are reported.

Acknowledgements

R. del C. has a research contract from the Fondo de Investigaciones Sanitarias of the Spanish Health Ministry (05/0137). The authors acknowledge partial funding and support to this research from the CIBER en Epidemiología y Salud Pública (CIBERESP) in Spain.

Transparency declarations

None to declare.

References