Tracing footsteps: exploring friendly viruses to diagnose Lyme disease

Despite the progress that has been achieved in Lyme disease (LD) diagnostic development in recent years, lack of early and effective diagnosis of LD remains a major cause of misdiagnosis and long-term patient suffering. We invented a novel way of detecting LD-causing bacteria. The new technology could potentially transform LD diagnostics through the innovative use of quantitative polymerase chain reaction (qPCR) to target bacteriophage (also known as phage) DNA in blood samples. The phage-based qPCR can not only provide information on the presence/ absence of LD, but also differentiate between disease stages.

by Dr J. Shan, Dr Y. Jia, Dr L. Teulières and Prof. M. Clokie

Lyme disease is real Lyme disease

(LD) is the most commonly reported tick-borne disease in the United States of America and Europe. Approximately 476¦000 diagnosed cases in the USA (CDC estimation) were recorded annually during 2010–2018, which is approximately a 45% increase in comparison to the estimated annual number during the period of 2005 to 2010 [1]. In Europe, the number of LD cases has increased steadily over the last two decades with an estimation of more than 100¦000 cases every year. In England and Wales, it is estimated that around 3000 new cases of LD are diagnosed each year, but the actual number could be as high as 8000 [2].

LD is caused by a group of bacteria called Borrelia burgdorferi sensu lato (s.l.) (Fig. 1). To date, more than 20 species of B. burgdorferi s.l. have been identified. Among them, B. burgdorferi sensu stricto (s.s.) is mainly found in the USA, whereas B. garinii and B. afzelii are dominant in Europe and Asia. It is believed that the main route of transmission of LD-causing bacteria to humans is through the bite of infected ticks (Fig. 2). In addition to tick transmission, observations that Borrelia can survive and circulate in the human bloodstream raised concerns over Borrelia transmission through blood transfusion [3]. As there are no vaccines for LD, the prevention of LD relies heavily on risk communication, rapid diagnosis and personal protection measures.

LD is difficult to diagnose

Currently, LD diagnosis is largely based on the medical signs and symptoms (rather than diagnostic tests), including a distinctive erythema migrans skin lesion (known as a ‘bull’s-eye’ rash; Fig. 3), tiredness, muscle pain, headaches, severe arthritic, neurologic and cardiac conditions. Worryingly, clinical diagnosis of LD is often problematic because the symptoms of LD are easy to confuse with other diseases. To aggravate the situation, co-infection with other tick-borne bacteria can happen in LD. One example is co-infection with relapsing fever (RF) Borrelia strains, such as Borrelia miyamotoi [4]. The two Borrelia groups have caused great concern and clinical confusion, as they are morphologically similar and present with almost indistinguishable clinical symptoms [5]. Despite this, they respond to different antibiotics and treatment regimens. Another example of confusion surrounding LD is the co-infection caused by Bartonella species [6]. This group of bacteria is emerging as an increasingly common human infection.

Much of the controversy surrounding LD and co-infections with Bartonella and/or B. miyamotoi is due to the lack of a reliable and sensitive diagnostic method to detect and distinguish between the three groups of bacteria. Therefore, laboratory tests to determine and distinguish between LD and co-infections play a vital role in the correct diagnosis and consequent treatment with different antibiotics.

There are also antigen-based LD diagnostic methods. They provide useful information to help clinicians, although they are known to be ‘not very useful’. For example, two antigen detecting methods, Borrelia culture and PCR targeting Borrelia genomic DNA, are also employed in LD diagnosis. The culture method remains, debatably, the gold standard, but it can take up to 7|weeks to confirm the presence of Borrelia! PCR detection of Borrelia is considered to be a promising alternative because theoretically it should be able to detect Borrelia DNA from patients at any stage of LD, as long as the spirochete happens to be present in the samples collected for testing [7]. Although PCR can have very high specificity, it suffers from low sensitivity because the number of Borrelia in patient samples is normally very low, bearing in mind that evidence from published studies indicates that Borrelia presence in LD patients can range from 1 to 100|cells/ml. Consequently, only one-third of LD patients in the USA were positive by PCR when their cerebrospinal fluid samples were tested [8]. Another study revealed that PCR tests from half of patients in the early stage of LD gave negative results because of the sporadic presence of Borrelia in blood samples [9].

The low sensitivity of PCR methods

There are two reasons behind the poor sensitivity of the current PCR methods in LD diagnosis. First, the current PCR methods target bacterial genomic DNA regions that occur at only one copy in each bacterium, such as the bacterial 16S rRNA gene, recA gene, and 5S–23S intergenic regions [10]. Second, at least some Borrelia species are ‘tissue-bound’ and are only transiently found circulating in the blood [11]. To circumvent the bottleneck of single-copy PCR template and low numbers of Borrelia circulating in blood, the current practice is either to culture samples before PCR or to sample large volumes of blood to artificially increase the number of PCR templates. Both methods have obvious drawbacks of contamination susceptibility, long duration and increased patient suffering.

Viruses can help diagnose LD

To overcome these diagnostic challenges, we have been investigating viruses for detecting LD. Viruses are everywhere but each type behaves in an incredibly specific way. The viruses we are studying are viruses that infect bacteria, known as bacteriophages (phages; Fig. 4). Phages are very specific about which bacterial species they infect and, therefore, they can be seen as ‘markers’ to indicate the presence of their bacterial hosts.

We demonstrated that Borrelia phage genes are present in multiple copies and tightly correlate with their Borrelia hosts [12]. Intuitively, targeting endogenous multicopy genes offers an elegant and reliable way of increasing the amount of PCR template [13]. Moreover, targeting phage-encoded genes would provide additional sensitivity to detect Borrelia infections because phages can be released into the bloodstream, even though Borrelia cells may be tissue-bound and not circulating in the blood. This situation of detecting free phage DNA from human blood bears some resemblance to identifying cell-free circulating DNA (cfDNA) in cancer diagnostics [14]. The phage-based technology could potentially transform LD diagnostics through the use of qPCR to trace phage DNA in blood samples [14].

What the phage-based qPCR can and cannot do

The phage-based qPCR is highly specific to Borrelia species. It does not amplify human DNA or even closely related RF Borrelia species. The qPCR amplification efficiency is almost 100% according to serial dilution experiments. The qPCR is highly sensitive and can detect one single Borrelia cell from as little as 0.3|ml of blood, which is equivalent to 3.3 Borrelia cells per ml of blood. This is really promising, bearing in mind that Borrelia presence in LD patients was estimated to be in the range of 1 to 100|cells/ml.

In addition to analytical validation, the diagnostic potential of the qPCR was evaluated against clinical samples from 78 individuals, including 23 healthy volunteers with no ‘identifiable’ LD symptoms, 13 early-stage and 42 late-stage LD patients. Together, the data revealed significant differences between early-stage LD patients, late-stage LD patients and healthy volunteers in terms of the quantitative phage gene levels determined from whole blood, suggesting that the phage-based qPCR has power to improve successful detection and differentiation of LD. Along with providing information on the presence/absence of Borrelia, the novel test is also capable of estimating the ‘Borrelia load’ to infer the level of infection.


In a nutshell, the phage-based technology detects phage DNA in blood samples to diagnose bacterial infections. Our novel approach has higher sensitivity compared to the known methods and can distinguish between healthy, early- and late-stage LD patients. The idea of tracking down phage DNA for diagnosis of Borrelia infections could be applicable to detection of other bacterial infections in general.

The authors
Jinyu Shan*1 PhD, Ying Jia1 PhD, Louis Teulières2 MD and Martha
Clokie1 PhD
1 Department of Genetics and Genome Biology, University of Leicester,
Leicester, UK
2 PhelixRD Charity, 230 Rue du Faubourg Saint-Honoré, Paris, France

*Corresponding author

1. Kugeler K, Schwartz A, Delorey M, Mead P, Hinckley A. Estimating the frequency of Lyme disease diagnoses, United States, 2010–2018. Emerg
Infect Dis 2021; 27(2): 616–619.
2. Cairns V, Wallenhorst C, Rietbrock S, Martinez C. Incidence of Lyme disease in the UK: a population-based cohort study. BMJ Open 2019; 9(7): e025916.
3. Pavia CS, Plummer MM. Transfusion-associated Lyme disease – although unlikely, it is still a concern worth considering. Front Microbiol 2018; 9: 2070.
4. Wormser GP, Shapiro ED, Fish D. Borrelia miyamotoi: an emerging tick-borne pathogen. Am J Med 2019; 132(2): 136–137.
5. Bergström S, Normark J. Microbiological features distinguishing Lyme disease and relapsing fever spirochetes. Wien Klin Wochenschr 2018;
130(15–16): 484–490.
6. Anderson BE, Neuman MA. Bartonella spp. as emerging human pathogens. Clin Microbiol Rev 1997; 10(2): 203–219.
7. Schutzer SE, Body BA, Boyle J, Branson BM, Dattwyler RJ, et al. Direct diagnostic tests for Lyme disease. Clin Infect Dis 2019; 68(6): 1052–1057.
8. Aguero-Rosenfeld ME, Wang G, Schwartz I, Wormser GP. Diagnosis of Lyme borreliosis. Clin Microbiol Rev 2005; 18(3): 484–509.
9. Babady NE, Sloan LM, Vetter EA, Patel R, Binnicker MJ. Percent positive rate of Lyme real-time polymerase chain reaction in blood, cerebrospinal
fluid, synovial fluid, and tissue. Diagn Microbiol Infect Dis 2008; 62(4): 464–466.
10. Liveris D, Schwartz I, McKenna D, Nowakowski J, Nadelman R, et al. Comparison of five diagnostic modalities for direct detection of Borrelia
burgdorferi in patients with early Lyme disease. Diagn Microbiol Infect Dis 2012; 73(3): 243–245.
11. Liang L, Wang J, Schorter L, Nguyen Trong TP, Fell S, et al. Rapid clearance of Borrelia burgdorferi from the blood circulation. Parasit Vectors
2020; 13(1): 191.
12. Shan J, Clokie MR, Teulières L. Borrelia phage. International Patent Application Number: PCT/GB2017/053323 (International Publication
Number: WO2018/083491) 2018 (
13. Luo RF, Scahill MD, Banaei N. Comparison of single-copy and multicopy real-time PCR targets for detection of Mycobacterium tuberculosis in
paraffin-embedded tissue. J Clin Microbiol 2010; 48(7): 2569–2570.
14. Shan J, Jia Y, Teulières L, Patel F, Clokie MRJ. Targeting multicopy prophage genes for the increased detection of Borrelia burgdorferi
sensu lato (s.l.), the causative agents of Lyme disease, in blood. Front microbiol 2021; 12(464): doi: 10.3389/fmicb.2021.651217.
15. Burns N, James CE, Harrison E. Polylysogeny magnifies competitiveness of a bacterial pathogen in vivo. Evol Appl 2015; 8(4): 346–351.