The DNA of Borrelia burgdorferi spirochetes extracted by ammonium hydroxide was used as the template for nested polymerase chain reaction (PCR) amplification of the species-specific 16S ribosomal DNA (rDNA). The primers were those well known to be specific for signature sequence amplification of the B burgdorferi sensu lato 16S ribosomal RNA gene. The positive 293-base-pair nested PCR amplicon was subjected to routine direct automated Sanger sequencing. A 50-base sequence excised randomly from the sequencing electrophoretogram between the 2 nested PCR primer binding sites was sufficient for the Basic Local Alignment Search Tool (BLAST) analysis to validate the B burgdorferi sensu lato 16S rDNA without a reasonable doubt. Nested PCR increased the sensitivity of DNA detection by 100- to 1,000-fold. DNA sequence validation based on BLAST algorithms using the GenBank database practically eliminates any possibility of false-positive results due to molecular misidentification. This technology may be a valuable supplement to the current serologic tests for Lyme disease.
Lyme disease is the most common tick-borne human infection in the United States. It is a potentially debilitating, multisystemic illness caused by the spirochete, Borrelia burgdorferi. Timely, appropriate antibiotic treatment based on accurate early diagnosis is highly effective in preventing tissue damage and the infection from going into chronic phase.
Major laboratory tests available for support of the diagnosis of Lyme disease are the 2-tier IgM and IgG enzyme-linked immunosorbent assay and Western blot for the detection of antibodies against sonicated whole-cell B burgdorferi1 and the single-tier C6 peptide-based IgG immunoassay.2 However, only one third of the patients with erythema migrans, which is the hallmark of the acute phase of Lyme disease, are positive for IgM or IgG seroreactivity when tested by these 2 serologic methods.2 Seroreactivities increase to two thirds among patients with Lyme disease during convalescence 3 to 4 weeks later, and the patterns of seropositivity are influenced by the extent of disease dissemination and tissue involvement in the nervous system, the heart, or joints.2 Major limitations of the immunoblot assays include the visual scoring and subjective interpretation of band intensity that may lead to false-positive readings. False-positive IgM immunoblot results have been reported in patients with non-Lyme diseases, such as rheumatoid arthritis, infectious mononucleosis, and systemic lupus erythematosus.3 The performance of different commercial serologic test kits may also vary, generating false-positive and false-negative results.4
Several polymerase chain reaction (PCR)-based nucleic acid tests have been introduced for the detection of B burgdorferi DNA in clinical specimens as an adjunct to the serologic assays, including the DNA tests for chromosomally carried genes such as ribosomal RNA (rRNA) genes, flaB, recA, and p66, and the plasmid-carried ospA.5 In general, the sensitivity of these PCR tests for the detection of B burgdorferi DNA in blood, plasma, or serum samples from patients with Lyme disease is low, ranging from 0% to 59%,6–8 with an 18% mean detection sensitivity in the United States.9 The sensitivity for PCR detection of Lyme spirochete DNA in synovial fluid samples seems to be higher.10 False-positive PCR test results can sway physicians to making erroneous diagnoses of Lyme disease, which may lead to inappropriate treatment with potentially serious complications.11,12
Because DNA sequences of the 16S rRNA gene, or the 16S rDNA, of all bacteria contain their respective unique hypervariable regions, a proposal has been made to select certain segments of these hypervariable regions as species-specific or type-specific signature sequences for bacterial identification in clinical microbiology.13,14 PCR primers designed to amplify regions of the 16S rRNA gene have been used to distinguish various groups of B burgdorferi.15 However, this gene is also known to contain highly conserved regions that are shared among the various isolates of B burgdorferi and non-Lyme Borrelia.16 A new primer designated as TEC117 has been introduced to pair with a previously reported LD2 primer15 to improve the specificity of the 16S rDNA–based PCR, which would allow differentiation of B burgdorferi sensu lato from other Borrelia species that do not cause Lyme disease.
In this article, we report a practical laboratory procedure by combining these 2 sets of primers to develop a nested (heminested) PCR assay for detection of Lyme Borrelia DNA in human body fluids and engorged Ixodes insects removed from skin bites. A target 293-base-pair (bp) nested PCR amplicon was further subjected to direct automated DNA sequencing to validate that the amplicon indeed consists of a signature DNA sequence of the B burgdorferi sensu lato 16S rRNA gene. Similar procedures have been used for human papillomavirus genotyping18–21 and for molecular testing of Chlamydia trachomatis and Neisseria gonorrhoeae.21,22
Materials and Methods
Standard Borrelia burgdorferi sensu stricto, strain B31 (ATCC 53210) in frozen liquid culture purchased from American Type Culture Collection, Manassas, VA, was used directly as the source of DNA for method development. A 100-μL aliquot of the thawed liquid Borrelia pure culture was mixed with 200 μL of 0.7 mol/L ammonium hydroxide in a 1.5-mL Eppendorf tube for DNA extraction.23 The mixture was heated at 95°C to 98°C for 5 minutes with a closed cap, followed by 10 minutes with an open cap. After the tube was cooled to room temperature, 700 μL of 95% ethanol and 30 μL of 3 mol/L sodium acetate were added to the mixture. The mixture was centrifuged at 13,000 rpm (~16,000g) for 5 minutes and the supernatant discarded. The precipitate was resuspended in 1 mL of cold 70% ethanol. Then the suspension was centrifuged at 13,000 rpm for 5 minutes. After all liquid was discarded, the pellet was air dried and resuspended in 100 μL tris(hydroxymethyl)aminomethane hydrochloride–EDTA buffer, with heating at 95°C to 98°C for 5 minutes. The heated suspension was finally centrifuged at 13,000 rpm for 5 minutes. One microliter of the supernatant was used for primary PCR to be followed by nested PCR amplification without further purification, using a ready-to-use HiFi DNA polymerase LoTemp PCR mix (HiFi DNA Tech, Trumbull, CT), as previously described.18–22
The primary PCR primers used were nucleotides LD1 (5′-ATGCACACTTGGTGTTAACTA) and LD2 (5′-GACTTATCACCGGCAGTCTTA).15 The nested PCR primers were nucleotides TEC1 (5′-CTGGGGAGTATGCTCGCA AGA)17 and LD2.15 The thermocycling steps were programmed to 30 cycles at 85°C for 30 seconds, 50°C for 30 seconds, and 65°C for 1 minute after an initial heating for 10 minutes at 85°C, with a final extension at 65°C for 10 minutes for primary and nested PCR in a TC-412 Thermal Cycler (Techne, Burlington, NJ).
All positive nested PCR products were subjected to direct automated DNA sequencing, using the TEC1 nucleotide as the sequencing primer according to the procedure previously published.22
For comparative quantitative studies, the standard B burgdorferi sensu stricto, strain B31, was subcultured in BSK-H Medium Complete (catalog No. B8291, Sigma Chemical, St Louis, MO) after 10-fold serial dilutions were made in 125 × 15-mm glass tubes, each containing 10 mL of medium and decreasing amounts of inoculum. Spirochete growth was indicated when the medium in the tubes turned acid without visible turbidity in 11 to 16 days. The acidic color change in the growing culture was sufficiently mild and gradual that to determine spirochete colony-forming units by culture was considered too imprecise. Formation of floccules of spirochetes was observed at the bottom of the liquid culture.
For bacterial counting, 1 μL of the growing culture without visible floccules was pipetted out and mixed with 3 μL of saturated aqueous Congo red solution on a clean microscopic slide. A thin smear of the mixture was made immediately, and the Congo red film was blued by exposure to the acid fume over a bottle of concentrated hydrochloric acid. The negatively stained spirochetes against a blue background on the smear were clearly observed microscopically under oil immersion, and the number of spirochetes on the entire smear was counted. The culture containing 1,000 spirochetes per microliter was selected for further quantitative studies.
To determine the optimum method for preparing blood samples for Lyme spirochete DNA detection, a concentrated B burgdorferi culture containing 106 spirochetes per microliter was used to spike Lyme disease–negative fresh whole blood samples with and without EDTA to reach a final concentration of 1,000 spirochetes per microliter of whole blood. After being mixed well, the whole blood without EDTA was allowed to clot at room temperature, and the serum was collected before centrifugation and after centrifugation for B burgdorferi DNA recovery by primary PCR. The EDTA blood was divided into 2 parts, 1 part being centrifuged at 2,398g for plasma separation and 1 part left standing at room temperature for 2 to 8 hours in a test tube so that the RBCs were separated from the plasma by gravitation only. Aliquots of 100 μL of plasma from the centrifuged blood and from the unspun blood were pipetted for B burgdorferi DNA recovery.
To determine the Lyme spirochete carrier status of the deer ticks removed from the skin of patients at the bite site, the engorged insect, fresh or archived, was first rinsed in 0.1 mL of 0.5% sodium hypochlorite and washed in 1 mL of water in a 1.5-mL Eppendorf tube. Then the insect was heated in 300 μL of 0.7 mol/L ammonium hydroxide in an Eppendorf tube at 95°C to 98°C for 20 minutes with a closed cap, followed by 10 minutes with an open cap. After the tube was cooled to room temperature and the insect carcass was discarded, 700 μL of 95% ethanol and 30 μL of 3 mol/L sodium acetate were added to the ammonium hydroxide digestate. The precipitated crude DNA was washed in 1 mL of 70% ethanol, air dried, and redissolved in 100 μL of tris(hydroxymethyl)aminomethane hydrochloride–EDTA buffer. One microliter of the digestate was used for primary and nested PCR for the detection of 16S rDNA, and all positive nested PCR target amplicons were validated by a segment of signature sequence.
As expected, the primary PCR amplicon flanked by a pair of LD1/LD2 primers and the nested PCR amplicon flanked by a pair of TEC1/LD2 primers were 351 bp and 293 bp, respectively, when the standard B burgdorferi culture was used as the source of DNA template.15,17 Successful demonstration of a 293-bp nested PCR amplicon in gel electrophoresis was used as the presumptive evidence for the presence of B burgdorferi sensu lato 16S rDNA, pending validation by DNA sequencing.
With 1 μL of crude DNA extracted from 1 μL of liquid culture containing 1,000 spirochetes of B burgdorferi per reaction, a primary PCR was able to detect a 351-bp amplicon visualized as a strong band on the gel plate after 30 cycles of amplification. When the digestate was diluted to 1/10, the band of the amplicon became very faint, sometimes barely visible. No visible primary PCR amplicons were detected when the digestate was further diluted beyond 1/10 Image 1. In the corresponding nested PCR setting with the TEC1 and LD2 primers, however, a clearly defined 293-bp nested PCR amplicon was generated in all PCR tubes using 1 μL each of the 1/1, 1/10, 1/100, and 1/1,000 digestate dilutions of the primary PCR products as template, indicating that nested PCR increased the sensitivity of the test by 100 to 1,000 times in detecting B burgdorferi 16S rDNA. The limit of detection by the nested PCR method was the amount of 16S rDNA derived from a single B burgdorferi spirochete. In comparison, the sensitivity of detection by primary PCR alone was 100 to 1,000 spirochetes or the DNA equivalent.
To validate that the 293-bp nested PCR amplicons observed in gel electrophoresis indeed indicate the presence of B burgdorferi 16S rDNA, a trace of the products from the nested PCR tube was subjected to direct automated Sanger DNA sequencing, using the TEC1 nucleotide as the sequencing primer. A 50-base segment randomly excised from the DNA base-calling tracing on an electrophoretogram flanked by the TEC1 and LD2 binding sites for a GenBank BLAST (Basic Local Alignment Search Tool) sequence match algorithm was found to be sufficient for validation of the species-specific 16S rRNA gene for B burgdorferi sensu lato. The sequence match was 100%, exclusive and unique as indicated in the GenBank BLAST online report Image 2.
EDTA-treated plasma collected after the erythrocytes were separated by gravitation provided the highest yield of the spiked Lyme spirochete DNA. The specific 16S rDNA was detected by primary PCR only in the unspun EDTA plasma after the original whole blood sample was spiked with 1,000 spirochetes per microliter. When the plasma was separated from the RBCs by routine centrifugation, the spiked spirochete DNA was barely detectable by primary PCR Image 3. Primary PCR failed to detect spirochete DNA in the serum separated from the blood clot with or without centrifugation.
In 50 blind-coded EDTA normal human plasma samples, 16 of which were spiked with 5 to 1,000 Lyme spirochetes per microliter, the species-specific 16S rDNA was detected in all 16 spiked samples by the nested PCR method. The 34 unspiked samples were found to be negative by primary and nested PCR, indicating a robust detection rate of 100% by this protocol with no cross-contamination.
Using unspun plasma and synovial fluid for the detection of B burgdorferi 16S rDNA, nonspecific amplification of human genomic DNA might be observed in the nested PCR gel when the samples were negative for Lyme disease spirochete DNA. However, amplification of nontarget DNA in the human plasma samples positive for B burgdorferi was negligible Image 4. When the target DNA template for the PCR primers was present, nontarget DNA amplifications were largely suppressed. The suppression of nontarget DNA amplification by target DNA with fully matched complementary sequences at the PCR primer binding sites was demonstrated by artificially introducing a trace of preamplified primary PCR products of the 16S rDNA into the otherwise negative nested PCR system. The result was a suppression of all visible nonspecific amplicons in the nested PCR Image 5. Publication of laboratory data with blinded patient identities was approved by the Milford Hospital Institutional Review Board (Milford, CT).
B burgdorferi 16S rDNA was readily eluted from the engorged deer tick specimens by ammonium hydroxide solution at high temperature even when the insects had been archived and dry in storage for more than 6 months. In 1 random sampling, Lyme spirochete DNA was detected in 5 of 10 dried archived engorged insects Image 6 that had been grossly identified as Ixodes species.
The purpose of this study was to optimize a nested (heminested) PCR-based protocol for detection of B burgdorferi sensu lato or its species-specific DNA in human body fluids. Nested PCR has been used to increase the sensitivity in Lyme disease DNA testing in clinical samples.24 We have confirmed that the sensitivity of a nested PCR is 100 to 1,000 times higher than that of a nonnested PCR in detecting the species-specific 16S rDNA. Although nested PCR can detect a single spirochete or its DNA equivalent, a nonnested PCR requires 100 to 1,000 spirochetes to generate a visible band on the gel electrophoresis plate (Image 1). Because the digestate used for each PCR amplification is derived from 1 μL of a liquid sample and there is a single gene for 16S rRNA25 in one B burgdorferi spirochete, our data suggest that the threshold for 16S rDNA nested PCR detection is 1,000 spirochetes per milliliter of body fluids.
All nucleic acid–based molecular tests involve determination of the order, namely the sequence, of the 4 nucleotide bases in the target DNA or RNA molecule in question, directly or indirectly. The technology of PCR, nested or non-nested, is a means of replicating the target DNA, although this process of in vitro replication of DNA also depends on sequence-specific complementary binding between the nucleotide primer and the template to initiate its enzymatic primer extension reaction. It is not exactly a tool for sequence determination of the target DNA.
In the current protocol, the purpose of PCR is to generate a template suitable for direct automated Sanger reaction, which is a technology to determine DNA sequence precisely.26 When a nested PCR amplicon is proven to contain a segment of a 50-base DNA sequence fully matched with the signature sequence of B burgdorferi sensu lato 16S rDNA stored in the GenBank between the TEC1 and LD2 primer binding sites (Image 2), molecular validation of B burgdorferi genetic materials in the clinical specimen is established without a reasonable doubt.
Nested PCR has not been commonly used in clinical laboratories because of the general concern that amplicon carryover may cause cross-contamination and potential false-positive PCR results.27 However, cross-contamination is not an inherent part of the nested PCR technology. It is rather a function of the clinical laboratory that performs PCR. Cross-contamination in a PCR laboratory can be readily monitored and largely eliminated by meticulous selection of technical staff, proper training, implementation of strict operational rules, and elimination of all micropipetting procedures that may induce aerosol of PCR amplicons.19–22 In this study, we demonstrated a robust no-cross-contamination system by spiking 16 of 50 blind-coded plasma samples with B burgdorferi pure culture as unknown samples for nested PCR detection. The results showed that even at a positive rate of 32%, the sensitivity of nested PCR detection of 16S rDNA is 100%, with no cross-contamination observed in the remaining 68% of the negative samples tested in the same batch.
Because in our local patient population the clinical samples truly positive for Lyme Borrelia DNA are still uncommon, repeated PCR amplification is performed on all patients’ body fluid digestates shown to be positive for B burgdorferi 16S rDNA to verify that the positive result is not due to DNA cross-contamination because of failing good laboratory practice. The practicality of general application of the nested PCR technology in community hospital clinical microbiology laboratories needs further field testing.
Hematogenous dissemination of the causative agents28 may be an important pathogenetic event in early Lyme borreliosis, before any specific antibodies are detectable. However, the traditional PCR6–9 and culture methods28 are not sensitive enough for early detection of Lyme disease Borrelia in blood samples. It requires 9 mL of plasma for culture if a more than 40% detection rate is expected. It is also known that plasma is a better source of culture material than serum or whole blood.28 We have demonstrated that EDTA plasma without centrifugation is the material of choice for nested PCR amplification of B burgdorferi 16S rDNA (Image 3). Blood clotting and centrifugation at 2,398g may reduce 90% of the spirochetes in the blood supernatant.
The cost of sample preparation and DNA purification is a prohibitory factor in introducing PCR and direct automated Sanger sequencing technologies to clinical laboratories. We used a LoTemp HiFi nested PCR for preparation of the templates for Sanger reaction without any purification steps. The HiFi DNA polymerase is highly processive. In the presence of target DNA with well-matched sequences perfectly complementary to the PCR primers, there is preferential amplification of the target DNA to the exclusion of other contaminating DNAs, on a competitive basis. However, in the absence of target DNA with no preferred template for the PCR primers, certain nontarget human genomic DNA molecules may be amplified instead, as demonstrated in 2 of 4 non-Lyme disease plasma samples (lanes 28 and 29, lower part of Image 4). These nonspecific amplifications that may be observed in the nested PCR gel are largely suppressed when the sample is positive for B burgdorferi 16S rDNA (lane 30, lower part of Image 4) or when the negative nested PCR mixture is spiked with a trace of the positive control primary PCR amplicon (Image 5). The presence of human genomic DNA does not seem to interfere with nested PCR detection of the 16S rDNA of B burgdorferi. Our experience in using this technology for the molecular diagnosis of early or atypical Lyme Borrelia infections will be presented as the number of positive cases being accumulated permits a meaningful analysis.
In this study, we also demonstrated that the nested PCR method may be adapted for detection of B burgdorferi in Ixodes insects that may carry the infective agents. The 16S rDNA is readily extracted by ammonium hydroxide solution from archived engorged insects (Image 6). Our preliminary data show that 20% to 50% of the engorged Ixodes insects removed from patient skin were positive for B burgdorferi in this community, depending on the season of the year.
The nested PCR testing validated by Sanger DNA sequencing provides reliable in vitro information for physicians in the management of early Lyme disease and possible chronic Borrelia infections, especially when the results of the serologic tests are ambiguous. However, this PCR-based test may fail to identify B burgdorferi–infected samples when the number of spirochetes in the sample is less than 1,000/mL. Therefore, it does not replace the established serologic tests that have proven useful in support of the clinical diagnosis of Lyme disease in most cases.