Lyme borreliosis is the most prevalent tick-borne disease in humans and is caused by Borrelia burgdorferi sensu stricto and sensu lato. The disease is a multisystemic disorder involving mainly the skin, nervous system, and joints; this broad spectrum sometimes makes clinical diagnosis difficult and diagnosis strongly relies on the detection of specific antibodies, e.g., by enzyme-linked immunosorbent assay (ELISA) and Western blot (WB). Vectors in North America are Ixodes scapularis and Ixodes pacificus, whereas in Europe, I. ricinus and I. persulcatus may transmit the pathogen. The disease is endemic in most European countries, U.S., and in some Asian countries]. In Europe, prevalence of infection in forestry workers varies from 8 to 27%,whereas in enzootic zones of the U.S. prevalence varies from 1 to 10%. A disease similar to Lyme borreliosis has been reported in Australia, although B. burgdorferi has not been isolated from either ticks or human patients.

In Mexico the ticks of the genus I. scapularis and I. pacificus are found in the northern area of the country and along the coast of the Gulf of Mexico. Mammals susceptible to B. burgdorferi infection include Odocoileus virginianus (tailed deer) and Peromyscus leucopus (white-foot mouse) in these areas. Five cases with a clinical presentation suggestive of Lyme borreliosis were reported in Mexico at the beginning of the 1990s, and three were positive for anti-B. burgdorferi antibodies by an ELISA, although none were further confirmed with the more specific WB test. These results suggested the presence of Lyme borreliosis in our country. The proximity of Mexico to the U.S. and the large number of travelers across these countries enhance the possibility of infection with B. burgdorferi in Mexico. The aim of this study was to search for serological evidence of B. burgdorferi infection in a community-based sample population of individuals from Mexico.

Materials and Methods

Enzyme-linked immunosorbent assay (ELISA)

ELISA to detect IgG antibodies against B. burgdorferi was done as previously described, with some modifications. Microtiter plate wells were coated with 5 ?g/well of a whole-cell sonicated extract of B. burgdorferi strain B31 (kindly supplied by Dr. Alan Barbour, University of Texas Health Science Center) in phosphate buffer pH 7.6 at 37°C, overnight. Plates were blocked for 1 h with 5% skim milk in phosphate buffered saline (PBSM) pH 7.6. Serum samples were properly diluted (see below), and 60 ?l aliquots were plated and incubated for 1 h at 37°C. Next, monoclonal antibodies anti-human IgG conjugated to alkaline phosphatase (Southern Biotech, Birmingham, AL, USA) diluted 1:1000 with PBSM were incubated for 1 h at 37°C. p-Nitrophenylphosphate, 1 mg/ml (Southern Biotech) was used as substrate and absorbance was obtained at 405 nm (iEMS analyzer, Labsystems, Helsinki, Finland). The mean plus 3 standard deviations (SD) of the absorbance values (A.V.) from 25 B. burgdorferi uninfected Mexican subjects were used to define the cut-off point. A regression line was constructed plotting cut-off value against the mean A.V. of a positive control pool of serum samples which included five sera from infected patients from an endemic zone (kindly supplied by Dr. L. Magnarelli, The Connecticut Agricultural Experiment Station). During all assays, the positive pool was included in quadruplicate in every microplate and the mean of the four A.V. values was plotted in the regression line to calculate the threshold for that plate. The result for each serum sample was expressed as the ratio of the A.V. of the sample to the threshold value and expressed as ELISA units. Accordingly, serum samples with ELISA units >1.0 were considered seropositive. Each serum sample was tested in duplicate.

Usually, serum samples from patients are titrated and tested in twofold dilutions (commonly from 1:40 to 1:1,280). However, because of the number of samples for studying, we chose only one test dilution. For this purpose, 88 serum samples from healthy Mexican donors were tested at serial twofold dilutions (1:80 to 1:2,560), and a 1:640 dilution was found to be the maximal dilution with the lowest unspecific reaction. This was the dilution of sera used to construct the regression line with the controls and to test all samples. Positive samples were further tested at 1:1,280, 1:2560, and 1:5120.

Western blot (WB)

Two hundred micrograms of whole-cell sonicate antigen were submitted to electrophoresis in a preparative SDS-12.5% polyacrylamide gel at 200 V for 1 h. Proteins were then transferred to a nitrocellulose membrane at 400 mA for 2 h. Membrane was blocked with 5% skim milk in TBST (20 mM Tris, 500 mM NaCl, .05% Tween 20, pH 7.6) for 1 h, washed and mounted in a multiscreen apparatus (Bio-Rad, Hercules, CA, USA). A 1:100 dilution of each test sample was added into each row and incubated for 1 h. After washing, membranes were incubated with a mouse monoclonal anti-human IgG alkaline phosphatase conjugate for 1 h. Reaction was developed with NBT-BCIP (nitroblue tetrazolium-5-bromo-4-chloro-3-indolylphosphate toluidin (Gibco, Gaithersburg, MD, USA). The pool of sera from B. burgdorferi infected patients and a serum sample from a noninfected individual (same as for the ELISA) were included in each membrane as controls. Also, sera from patients with lupus, syphilis, rheumatoid arthritis, or anti-phospholipid antibodies were included to control for cross-reactions. A serum sample was considered positive for anti-B. burgdorferi if at least three of the following protein bands were recognized: 18, 24, 28, 29, 31, 34, 39, 41, 45, 58, 62, 66, and 93 kDa. Additionally, to be considered positive, sera should not react with any of the bands recognized by the cross-reaction controls (see previous text). Some sera positive by WB were further studied with the Borrelia immunodot-blot IgG test (GenBio, San Diego, CA, USA). This test has five antigen dots, one with a whole cell B. burgdorferi antigen, and four with recombinant or purified antigens: a high molecular weight protein (HMW) 83-100 kD, the flagellin, a 39 kDa protein (p39), and an outer surface protein C (Osp C).

Results

From the 2,890 serum samples studied, 34 were positive for IgG anti-B. burgdorferi by ELISA, representing 1.1% of the population studied (CI 95%, 0.74–1.45%). When titrated, 32 samples were positive at a 1:640 dilution and only two were positive at a 1:1,280 dilution. All 34 ELISA positive samples were tested in WB, and only nine fulfilled the criteria established by us for positivity; representing a confirmed seroprevalence of 0.3% (CI 95%, 0.13–0.51). The pattern of bands recognized by these nine sera is presented in Proteins of 28 kDa, 34 kDa (Osp B), and 45 kDa were recognized by sera from three of the nine samples; proteins of 39 kDa (BmpA), and 41 kDa (Fla) were recognized by five samples; protein of 58 kDa was recognized by six samples; and the protein of 66 kDa (Osp p66) was recognized by seven of the nine samples. The bands recognized by sera from a patient with syphilis were 29, 33, 49 and 62 kDa; none of the nine B. burgdorferi positive samples recognized any of these proteins. In four of the nine WB positive serum specimens, enough sample was available to further confirm anti-B. burgdorferi antibodies with the immuno-dot blot IgG test. These four sera were positive for antibodies against the whole cell, the HMW, the p41, and the OspC antigens.

Lyme borreliosis is the most prevalent tick-borne disease in humans and is caused by Borrelia burgdorferi sensu stricto and sensu lato. The disease is a multisystemic disorder involving mainly the skin, nervous system, and joints; this broad spectrum sometimes makes clinical diagnosis difficult and diagnosis strongly relies on the detection of specific antibodies, e.g., by enzyme-linked immunosorbent assay (ELISA) and Western blot (WB). Vectors in North America are Ixodes scapularis and Ixodes pacificus, whereas in Europe, I. ricinus and I. persulcatus may transmit the pathogen. The disease is endemic in most European countries, U.S., and in some Asian countries. In Europe, prevalence of infection in forestry workers varies from 8 to 27%,whereas in enzootic zones of the U.S. prevalence varies from 1 to 10%. A disease similar to Lyme borreliosis has been reported in Australia, although B. burgdorferi has not been isolated from either ticks or human patients.

In Mexico the ticks of the genus I. scapularis and I. pacificus are found in the northern area of the country and along the coast of the Gulf of Mexico. Mammals susceptible to B. burgdorferi infection include Odocoileus virginianus (tailed deer) and Peromyscus leucopus (white-foot mouse) in these areas. Five cases with a clinical presentation suggestive of Lyme borreliosis were reported in Mexico at the beginning of the 1990s, and three were positive for anti-B. burgdorferi antibodies by an ELISA [, although none were further confirmed with the more specific WB test. These results suggested the presence of Lyme borreliosis in our country. The proximity of Mexico to the U.S. and the large number of travelers across these countries enhance the possibility of infection with B. burgdorferi in Mexico. The aim of this study was to search for serological evidence of B. burgdorferi infection in a community-based sample population of individuals from Mexico.

Materials and Methods

Population

During 1987–1988, the Ministry of Health of Mexico performed a national serologic survey to create a National Serum Bank. Serum samples were collected according to a master sampling frame based on the General Population Census data. The survey included all 32 states of Mexico and was designed to include individuals with ages ranging from 1- to 90-years-old. All socioeconomic levels and all geographic zones of the country were included. A total of 32,200 households were visited and more than 70,000 serum samples were collected. The overall response rate was 78.4% of the surveyed homes. An urban population was defined as an area with more than 2,500 inhabitants, whereas a rural population was defined as an area with less than 2,500 inhabitants.

Study population

A sample representing urban and rural communities from the 32 states of the country, including individuals 1- to 90-years old from all socioeconomic levels were selected by a simple random design. The sample size was calculated estimating a seroprevalence for B. burgdorferi of 5% (range 1–9%; 99% confidence), resulting in a sample of 3,016 serum specimens.

Aliquots of 100 ?l of selected serum samples located in the National Serum Bank were transported to our laboratory and frozen at ?20°C until tested. Some of the originally selected samples were not recovered because they were contaminated, dried, or missing. Ultimately, 2,890 serum samples were studied, corresponding to 95.8% of the sample size selected.

Enzyme-linked immunosorbent assay (ELISA)

ELISA to detect IgG antibodies against B. burgdorferi was done as previously described, with some modifications. Microtiter plate wells were coated with 5 ?g/well of a whole-cell sonicated extract of B. burgdorferi strain B31 (kindly supplied by Dr. Alan Barbour, University of Texas Health Science Center) in phosphate buffer pH 7.6 at 37°C, overnight. Plates were blocked for 1 h with 5% skim milk in phosphate buffered saline (PBSM) pH 7.6. Serum samples were properly diluted (see below), and 60 ?l aliquots were plated and incubated for 1 h at 37°C. Next, monoclonal antibodies anti-human IgG conjugated to alkaline phosphatase (Southern Biotech, Birmingham, AL, USA) diluted 1:1000 with PBSM were incubated for 1 h at 37°C. p-Nitrophenylphosphate, 1 mg/ml (Southern Biotech) was used as substrate and absorbance was obtained at 405 nm (iEMS analyzer, Labsystems, Helsinki, Finland). The mean plus 3 standard deviations (SD) of the absorbance values (A.V.) from 25 B. burgdorferi uninfected Mexican subjects were used to define the cut-off point. A regression line was constructed plotting cut-off value against the mean A.V. of a positive control pool of serum samples which included five sera from infected patients from an endemic zone (kindly supplied by Dr. L. Magnarelli, The Connecticut Agricultural Experiment Station). During all assays, the positive pool was included in quadruplicate in every microplate and the mean of the four A.V. values was plotted in the regression line to calculate the threshold for that plate. The result for each serum sample was expressed as the ratio of the A.V. of the sample to the threshold value and expressed as ELISA units. Accordingly, serum samples with ELISA units >1.0 were considered seropositive. Each serum sample was tested in duplicate.

Usually, serum samples from patients are titrated and tested in twofold dilutions (commonly from 1:40 to 1:1,280). However, because of the number of samples for studying, we chose only one test dilution. For this purpose, 88 serum samples from healthy Mexican donors were tested at serial twofold dilutions (1:80 to 1:2,560), and a 1:640 dilution was found to be the maximal dilution with the lowest unspecific reaction. This was the dilution of sera used to construct the regression line with the controls and to test all samples. Positive samples were further tested at 1:1,280, 1:2560, and 1:5120.

Western blot (WB)

Two hundred micrograms of whole-cell sonicate antigen were submitted to electrophoresis in a preparative SDS-12.5% polyacrylamide gel at 200 V for 1 h. Proteins were then transferred to a nitrocellulose membrane at 400 mA for 2 h. Membrane was blocked with 5% skim milk in TBST (20 mM Tris, 500 mM NaCl, .05% Tween 20, pH 7.6) for 1 h, washed and mounted in a multiscreen apparatus (Bio-Rad, Hercules, CA, USA). A 1:100 dilution of each test sample was added into each row and incubated for 1 h. After washing, membranes were incubated with a mouse monoclonal anti-human IgG alkaline phosphatase conjugate for 1 h. Reaction was developed with NBT-BCIP (nitroblue tetrazolium-5-bromo-4-chloro-3-indolylphosphate toluidin (Gibco, Gaithersburg, MD, USA). The pool of sera from B. burgdorferi infected patients and a serum sample from a noninfected individual (same as for the ELISA) were included in each membrane as controls. Also, sera from patients with lupus, syphilis, rheumatoid arthritis, or anti-phospholipid antibodies were included to control for cross-reactions. A serum sample was considered positive for anti-B. burgdorferi if at least three of the following protein bands were recognized: 18, 24, 28, 29, 31, 34, 39, 41, 45, 58, 62, 66, and 93 kDa. Additionally, to be considered positive, sera should not react with any of the bands recognized by the cross-reaction controls (see previous text). Some sera positive by WB were further studied with the Borrelia immunodot-blot IgG test (GenBio, San Diego, CA, USA). This test has five antigen dots, one with a whole cell B. burgdorferi antigen, and four with recombinant or purified antigens: a high molecular weight protein (HMW) 83-100 kD, the flagellin, a 39 kDa protein (p39), and an outer surface protein C (Osp C)

Statistics

Results were expressed as the ratio of seropositive samples to the total number of samples studied (%); with a 95% confidence interval.

Results

From the 2,890 serum samples studied, 34 were positive for IgG anti-B. burgdorferi by ELISA, representing 1.1% of the population studied (CI 95%, 0.74–1.45%). When titrated, 32 samples were positive at a 1:640 dilution and only two were positive at a 1:1,280 dilution. All 34 ELISA positive samples were tested in WB, and only nine fulfilled the criteria established by us for positivity; representing a confirmed seroprevalence of 0.3% (CI 95%, 0.13–0.51). The pattern of bands recognized by these nine sera.. Proteins of 28 kDa, 34 kDa (Osp B), and 45 kDa were recognized by sera from three of the nine samples; proteins of 39 kDa (BmpA), and 41 kDa (Fla) were recognized by five samples; protein of 58 kDa was recognized by six samples; and the protein of 66 kDa (Osp p66) was recognized by seven of the nine samples. The bands recognized by sera from a patient with syphilis were 29, 33, 49 and 62 kDa; none of the nine B. burgdorferi positive samples recognized any of these proteins. In four of the nine WB positive serum specimens, enough sample was available to further confirm anti-B. burgdorferi antibodies with the immuno-dot blot IgG test. These four sera were positive for antibodies against the whole cell, the HMW, the p41, and the OspC antigens.