Response to Esteve-Gassent et al.: flaB sequences obtained from Texas PCR products are identical to the positive control strain Borrelia burgdorferi B31
© Norris et al. 2015
Received: 21 March 2015
Accepted: 15 May 2015
Published: 9 June 2015
Feria-Arroyo et al. had reported previously that, based on PCR analysis, 45 % of Ixodes scapularis ticks collected in Texas and Mexico were infected with the Lyme disease spirochete Borrelia burgdorferi (Parasit. Vectors 2014, 7:199). However, our analyses of their initial data (Parasit. Vectors 2014, 7:467) and a recent response by Esteve-Gassent et al. (Parasit. Vectors 2015, 8:129) provide evidence that the positive PCR results obtained from both ribosomal RNA intergenic sequences and the flagellin gene flaB are highly likely due to contamination by the B. burgdorferi B31 positive control strain.
KeywordsBorrelia burgdorferi Lyme disease Ixodes scapularis Texas Mexico
In 2014, Feria-Arroyo et al.  reported that 45 % of Ixodes scapularis ticks collected in Texas and Mexico were infected with the Lyme disease spirochete Borrelia burgdorferi, based on nested PCR amplification and sequencing of the 16S rDNA-23S rDNA intergenic spacer region (IGS). Positive PCR results were also reported for the B. burgdorferi flagellin gene flaB, but no flaB sequences were provided in the article. The results reported by Feria-Arroyo et al.  were highly questionable in that all prior studies of ticks from Texas indicated that less than two percent were infected with Borrelia of any kind. In a comprehensive reanalysis of their data , we demonstrated previously that the reported IGS sequences had a high degree of identity with the IGS region of the positive control strain, B. burgdorferi B31, used in their study. This commonly used strain was the first isolate of B. burgdorferi, and was obtained from ticks collected on Shelter Island, New York in 1981 . Based on our analysis, we concluded that the high frequency of positive PCR results reported by Feria-Arroyo et al.  were due to contamination of their tick specimens with DNA from the positive control strain. In a recent response published on 27 February 2015, corresponding author Dr Maria Esteve-Gassent and co-authors  provided flaB sequences from several of the tick specimens from Texas and claimed that “Infection levels using a second genetic marker (flaB), confirmed the results originally obtained by the 16S rRNA-23S rRNA gene intergenic spacer (IGS) of B. burgdorferi”. However, a simple comparison of these flaB sequences to existing genomic sequences indicated that they are essentially identical to the B. burgdorferi B31 sequence, contrary to their conclusions . In addition, the author’s recent response  did not address the high sequence identity between the IGS sequences obtained from the tick specimens with that of the positive control strain (see ), and contained several additional errors and misstatements. If anything, the additional data and analysis of sequences provided by Esteve-Gassent et al.  further strengthen our critique and deepen our doubts about the accuracy of the observations. Therefore, our assessment and conclusions remain the same, namely that: (1) The findings reported by Feria-Arroyo and coworkers [1, 4] of B. burgdorferi in their sample of ticks from Texas and Mexico are erroneous; and (2) The inaccurate information, most plausibly, was the consequence of laboratory contamination of the samples in the chain of possession and faulty analysis of their results and the scientific literature. Our presumption is that the contamination was not intentional but inadvertent. (Indeed, the avoidance of contamination through the scrupulous preparation and handling of specimens is an important and difficult challenge for all laboratories carrying out the highly sensitive PCR procedure ). The inaccurate and unreliable information reported is not a trivial matter or of little consequence for public health. Mistaken allegations of the new presence of B. burgdorferi in a geographical area, particularly at such high prevalence as Esteve-Gassent and co-authors [1, 4] reported, could lead to clinical misdiagnoses and misdirected prevention, treatment and control efforts.
The Esteve-Gassent et al. response does not address the IGS sequence identity between the Texas samples and the positive control strain
The flaB sequences from the tick samples from Texas are also identical to those of the positive control strain B31
“Norris et al. argued that the infected ticks reported in our study were found infected with B. burgdorferi likely due to contamination of the PCR reactions with DNA from the strain B31 of B. burgdorferi, the positive control used in the study. Nevertheless, B. burgdorferi B31 flaB has a cytosine (C) at position 75 in this alignment (Fig. one) while the Texas isolates had an adenine (A). The A in the Texas isolate makes them more similar to strains N40 and 297 than to B31. Contamination of our samples with strains N40 and 297 is impossible, since these strains are not present in the laboratory in which molecular analyses were carried out”.
We were puzzled by this statement in that the widely accepted B. burgdorferi B31 chromosome sequence (GenBank Accession No. AE000783.1) has an adenine (A) at the indicated position. A recent resequencing of the B31 genome (CP009656) has the same sequence. In fact, nearly all of the reported B. burgdorferi flaB sequences (over 100) have an adenine at this position. The only exceptions are three reported sequences: two from Dr. Reinhardt Wallich in Germany (X15661 and X16833) and one from a group in the United Kingdom (Y15088). Because the sequence difference is restricted to this one nucleotide, the particular clone used by these groups likely had a point mutation at this position.
Esteve-Gassent et al.  apparently selected this rare sequence because the GenBank entry X15661 is annotated as flaB. In the B31 genomic sequence, the gene is annotated as BB0147 and the descriptor used is “P41”, one of the initial descriptions of this protein as a 41-kDa protein antigen. Esteve-Gassent et al.  apparently did not recognize that BB0147 is the same as flaB, although a simple BLAST search would have demonstrated this fact.
The reverse complement of the BWTX12-16 DNA sequence is mistakenly displayed in Fig. one
“BWTX12-16, a questing tick, has a significantly different sequence from either of the controls or the other Texan samples, suggesting that the degree of genetic variation of B. burgdorferi in the regions sampled likely exceed the values found by Feria-Arroyo et al. study , which only sampled a limited number of potential vertebrate hosts“.
Figure two of the Esteve-Gassent et al. response reinforces sequence identity with B31 in most tick samples from Texas, but also indicates the presence of sequence errors in some of the sequences from Texas samples
Figure three of the response by Esteve-Gassent et al.  inappropriately displays phylogenetic trees based on sequence errors and incorrect sequence orientation
The conclusions of Esteve-Gassent et al.  are contrary to all prior published reports regarding the heterogeneity of IGS sequences in B. burgdorferi strains
The following paragraph indicates their resistance to consider the possibility of DNA contamination and sequence errors in their results, with the conclusion that samples acquired from ticks feeding on white-tailed deer, gembok and dogs in a large geographical area harbour B. burgdorferi with identical or nearly identical IGS genotypes.
“Norris et al. stated in their letter that due to the low variability observed in the IGS from the Texas samples most, if not all of them, were likely to have been originated from the same clone which they assume could be the product of contamination with the B31 strain. We disagree with the interpretation put forward by Norris et al. and instead think it is more likely that the lack of variability reported in Feria-Arroyo et al. reflects the level of B. burgdorferi variability present in the Texas-Mexico transboundary region. Several of the ticks included in the Feria-Arroyo study were collected from white-tailed deer, gemsbok and dog. These mammalian hosts, particularly white-tailed deer, harbour ticks from several lineages. Thus, ticks collected from white-tailed deer, even if collected from the same individual, are likely to carry a representation of the B. burgdorferi strains present in a particular location. Thus, the B. burgdorferi genetic diversity reported by Feria-Arroyo et al., likely represents the genetic variation present in the Texas-Mexico transboundary region”.
The article by Feria-Arroyo et al.  provides false and misleading information regarding the risk of B. burgdorferi infection in Texas and Mexico
To quote the Feria-Arroyo et al. article , “Infection with B. burgdorferi was detected in 45 % of I. scapularis ticks…”.
“Norris et al. suggest that the Feria-Arroyo et al.  publication is advocating a high LD risk in Texas and Mexico but this cannot be further from the truth”.
It is in fact a natural conclusion that a high tick infection rate connotes a heightened risk of human disease.
We conclude that, as in the case of the article by Feria-Arroyo et al. , the data and interpretations presented in the response from Esteve-Gassent et al.  are unreliable and are not valid scientifically. The Committee of Publication Ethics guidelines (http://publicationethics.org/files/retraction%20guidelines.pdf), to which BioMedCentral adheres, states that “Journal editors should consider retracting a publication if they have clear evidence that the findings are unreliable, either as a result of misconduct (e.g., data fabrication) or honest error (e.g., miscalculation or experimental error).” We believe that the latter assessment is the case. Therefore, we recommend the retraction of the Feria-Arroyo et al. article  from Parasites & Vectors.
- Feria-Arroyo T, Castro-Arellano I, Gordillo-Perez G, Cavazos A, Vargas-Sandoval M, Grover A, et al. Implications of climate change on the distribution of the tick vector Ixodes scapularis and risk for Lyme disease in the Texas-Mexico transboundary region. Parasit Vectors. 2014;7:199.PubMed CentralPubMedView ArticleGoogle Scholar
- Norris SJ, Barbour AG, Fish D. Diuk-Wasser MA. Analysis of the intergenic sequences provided by Feria-Arroyo et al does not support the claim of high Borrelia burgdorferi tick infection rates in Texas and northeastern Mexico Parasit Vectors. 2014;7:467.PubMedGoogle Scholar
- Burgdorfer W, Barbour AG, Hayes SF, Benach JL, Grunwaldt E, Davis JP. Lyme disease, a tick-borne spirochetosis? Science. 1982;216:1317–9.PubMedView ArticleGoogle Scholar
- Esteve-Gassent MD, Grover A, Feria-Arroyo TP, Castro-Arellano I, Medina RF, Gordillo-Perez G, et al. Prevalence of Borrelia burgdorferi-infected ticks from wildlife hosts, a response to Norris et al. Parasit Vectors. 2015;8:129.PubMed CentralPubMedView ArticleGoogle Scholar
- Kwok S, Higuchi R. Avoiding false positives with PCR. Nature. 1989;339:237–8.PubMedView ArticleGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.