Lyme Borreliosis and Deficient Mannose-Binding Lectin Pathway of Complement Free

Lyme borreliosis (LB) is the most common pathogenic tick-borne infection in humans and is caused by spirochetes of the Borrelia burgdorferi sensu lato complex (referred to hereafter as “borrelia”) (1). The innate immune system acts as the first line of defense against borrelia infection because it controls dissemination of infection until a more specific adaptive immune response is developed (2). The innate immune response against borrelia involves host proteases; a wide range of immune cells, such as neutrophils and macrophages; and the complement system, which can be activated via three activation pathways: the classical pathway, the alternative pathway, and the lectin pathway (hereafter referred to as the mannose-binding lectin [MBL] pathway) (3). MBL is the key molecule in initiating the MBL pathway (4).

MBL is a circulating and tissue-residing pattern recognition molecule that binds to a broad range of pathogens and kills them through opsonophagocytosis and complement activation (5, 6). It was recently revealed that the tick salivary lectin pathway inhibitor effectively inhibits the binding of MBL to its ligand at the tick bite site, thereby interfering with complement-mediated killing of borrelia (7, 8). This finding highlights the importance of MBL-mediated protection against LB.

MBL pathway–mediated protection efficacy depends on the serum and tissue concentration of biologically active MBL (4, 9). Low MBL levels are primarily due to single-nucleotide polymorphisms residing in the mbl2 gene on chromosome 10. Mutant variants are unable to properly form the MBL molecule multimer; thus, the amount of functional MBL in blood and tissues decreases (10, 11). MBL pathway deficiency is considered the most common congenital immunodeficiency of humans; it affects approximately one fourth of the general population (12, 13).

Risk factors for LB are mostly unknown. In the current study, we first determined the cut-off values for serum MBL concentrations for diminished and deficient MBL pathway functions and then investigated whether impaired MBL pathway function predisposes individuals to disseminated Ab⁺ LB.

All 700 serum samples used for the analysis of serum MBL concentrations (350 LB patients and 350 non-LB controls) were collected with informed consent as a part of routine clinical practice from individuals (outpatients as well as inpatients) who were clinically suspected to have LB. All serum samples were analyzed in the diagnostic laboratory at the Department of Medical Microbiology and Immunology at the University of Turku. All samples were coded, and strict anonymity was maintained throughout the study. Except for age and gender, no other clinical data were used. According to the Finnish Medical Research Act (No. 488/1999), Chapter 1, Sections 1–3, the research in the current study was not medical research; thus, it was not necessary to obtain a separate approval from the local Ethics Committee to use the samples in the assays of the current study.

A serum MBL concentration cut-off for diminished MBL pathway function was determined by analyzing existing data for 201 serum samples from the period 2009–2013 that were available in our laboratory information management system. These samples were from patients with unknown LB status, whose serum MBL concentrations were measured together with MBL pathway function. The median age in this group was 10 y (age range, 0–74 y); 100 (50%) patients were males. The function of the MBL pathway was measured by a commercial ELISA-based procedure (Wieslab Total Complement system Screen Classical, MBL, Alternative Pathways; Euro-Diagnostica, Malmö, Sweden) using mannan as the activator of the MBL pathway (14). The test was performed, and the results were interpreted according to the manufacturer’s instructions (diminished function < 40%, deficient function < 10%).

LB patient serum samples were identified retrospectively from the information management system of our laboratory. The LB patient sample group included all 350 LB Ab⁺ cases identified in our laboratory in 2011 from ∼11,500 analyzed samples (positivity rate 3.0%). The patients were diagnosed according to the current Finnish and international guidelines (1). The median age was 62 y (range 2–89 y); 169 (48%) patients were males. The non-LB control group included 350 randomly selected serum samples from the ∼11,000 analyzed samples that tested negative for borrelia Abs. The non-LB control samples were selected so that there was an equal distribution over age and gender to obtain a baseline of MBL concentrations over age (i.e., non-LB samples were not matched to the LB patient samples); 177 (51%) controls were males, and the median age was 39 y (range, 1–92 y). Notably, the control group was purposely composed of an equal number of samples from each age group (0–10, 11–20, 21–30 y, and so forth)

All samples were tested for IgM and IgG Ab levels by ELISA using sonicated borrelia whole-cell bacteria lysate as a coating Ag (15). Positive or equivocal serology was further analyzed by a flagella Ag–based ELISA (IDEIA Borrelia burgdorferi IgM/IgG; Oxoid, Cambridgeshire, U.K.) (n = 245, 70% of samples) and/or by a line immunoblot using recombinant borrelia Ags (recomLine Borrelia IgM/IgG; Mikrogen, Neuried, Germany) (n = 190, 54% of samples). IgG Abs against B. afzelii p18 Ag (Decorin binding protein A) were detected using the recomLine Borrelia IgG line immunoblot, according to the interpretation criteria of the manufacturer.

Double-Ab sandwich ELISA was used to measure the MBL concentrations, as described previously with minor modifications (16). Microtiter plates (Maxisorp; A/S Nunc, Roskilde, Denmark) were coated overnight at 4°C with Mouse Monoclonal Anti-Human Mannan-Binding Lectin Ig G1 Ab (Statens Serum Institut [SSI], Copenhagen, Denmark) diluted to a concentration of 8 μg/ml in 50 mM carbonate-bicarbonate buffer (Sigma-Aldrich, Steinheim, Germany). After the incubation, the wells were washed with PBS and then blocked with PBS containing 1% BSA (MP Biomedicals, Eschwege, Germany). The wells were incubated for 1 h at room temperature (RT). Samples were diluted 1:100 in PBS containing 0.05% Tween 20 (Merck, Darmstadt, Germany). Standard serum containing 3200 ng/ml oligomerized MBL (SSI) was used to create a standard curve. MBL-deficient serum (SSI) was used as a negative control in each assay. The samples were incubated for 1 h at 37°C. Biotinylated Mouse Monoclonal Anti-Human Mannan-Binding Lectin Ig G1 Ab (SSI) was diluted 1:10,000 in PBS–0.05% Tween 20 and used as the secondary Ab. The wells were incubated for 1 h at RT. Streptavidin peroxidase (Sigma-Aldrich, St. Louis, MO) used to detect the bound secondary Ab and diluted 1:8000 in PBS–0.05% Tween 20 was incubated for 1 h at RT. The wells were washed three times with PBS–0.05% Tween 20 after every incubation. Substrate (TMB Liquid Substrate System for ELISA; Sigma-Aldrich; 100 μl) was added to each well, and the wells were incubated for 20 min at RT. The reaction was stopped using Stop Reagent for TMB Substrate (Sigma-Aldrich; 100 μl/well). Absorbances were detected at 450 nm with a Multiskan EX spectrophotometer (Thermo Fisher Scientific, Vantaa, Finland). The detection limit of the assay was set as 50 ng/ml, and the concentration below this cut-off was given a value of 25 ng/ml. The upper limit of the assay was 6400 ng/ml. All samples were tested in duplicate.

Optimal serum MBL concentration cut-off values were determined using 10 and 40% cut-offs of the MBL pathway function, as described above. Receiver operating characteristic analyses were performed using MBL pathway function as a binary response and serum MBL concentration as an independent variable. Minimum sensitivity was set to 80%, which yielded the cut-off values of 445 and 787 ng/ml, with specificities of 97.3 and 93.8%, respectively.

The effect of age on serum MBL concentration, controlling for groups, was analyzed using analysis of covariance. The distribution of serum MBL concentrations was skewed; hence, a square-root transformation was applied. Model fit was confirmed with Q-Q plotted Pearson residuals. A Mann–Whitney U test was used to analyze the difference in the serum MBL concentrations between the genders and between the LB patient samples and the non-LB control samples.

Differences between LB patient samples and non-LB control samples were investigated using logistic regression with logit link. There were two parallel models for both MBL cut-off values for LB positivity as a response and the MBL cut-off, age, gender, and age and gender interaction as independent variables. Results are presented as odds ratios (ORs) and 95% confidence intervals (CIs). All analyses were conducted by a trained statistician (T.K.) using SAS System for Windows, version 9.4 (SAS Institute, Cary, NC), and all figures were drawn with R 3.1.0 (R Foundation for Statistical Computing, Vienna, Austria).

No generally accepted cut-off MBL concentration that correlates with impaired MBL pathway function exists. The purpose of this analysis was to determine cut-off MBL concentrations applicable to the Finnish population. The functional results were plotted against the MBL concentrations. With a minimum level of sensitivity of 80%, diminished (<40%; 2Materials and Methods) and deficient (<10%) MBL pathway functions were associated with MBL concentrations of 787 ng/ml (specificity 93.8%) and 445 ng/ml (specificity 97.3%) (Fig. 1). Therefore, we used the MBL concentration of 787 ng/ml as the cut-off for diminished MBL pathway function and 445 ng/ml as the cut-off for deficient MBL pathway function.

MBL concentrations were measured in the LB group and the non-LB control group; each included 350 subjects. Among both groups, there was a declining trend in the MBL concentration with age (β = −0.17, p = 0.0021 and β = −0.11, p = 0.0226, respectively) (Fig. 2), whereas there was no difference in the MBL concentrations between genders (p = 0.9166). Diminished MBL pathway function was observed in 145/350 (41.4%) of the LB samples and in 96/350 (27.4%) of the non-LB controls. These data were analyzed statistically using logistic regression, controlling for age, gender, and the age and gender interaction. Diminished MBL pathway function was significantly more frequent (p = 0.0027; OR 1.67, 95% CI: 1.19–2.34) in the LB samples than in the control samples (Fig. 3). Deficient MBL pathway function was observed in 92/350 (26.3%) of the LB samples and in 60/350 (17.1%) of the non-LB controls, which also was significantly different (p = 0.0361; OR 1.51, 95% CI: 1.03–2.23) (Fig. 3).

In general, the serum MBL concentrations were significantly lower in the LB samples than in the non-LB controls (Fig. 4). The median MBL concentration was 1088 ng/ml (interquartile range, 418–2166 ng/ml) in LB samples and 1988 ng/ml (interquartile range 687-4220 ng/ml) in non-LB samples (p < 0.0001).

The presence of IgG Abs against p18 Ag may correspond to a more severe course of LB (17). In our study, B. afzelii p18 IgG Abs were detected in 87/190 (45.8%) LB patients, reflecting the fact that B. afzelii is the most common borrelia genospecies in Finland (18). In 40 of these 87 patients (46.0%), MBL pathway function was diminished. Thus, the frequency of diminished MBL pathway function was even higher in LB patients who had B. afzelii p18 IgG Abs in comparison with all LB patients (46.0 versus 41.4%). Compared with the non-LB controls, this subset of patients showed significant differences in the frequency of diminished MBL pathway function (46.0 versus 27.4%, p = 0.0016, OR 2.34, 95% CI: 1.38–3.97) and deficient MBL pathway function (35.6 versus 17.1%, p = 0.0253, OR 1.96, 95% CI: 1.09–3.52) (Fig. 3). There was no difference in the MBL concentration between genders (p = 0.7026).

Furthermore, the age-adjusted median MBL concentration was lower in the LB patient group with B. afzelii p18 IgG Abs than in the control group (968 versus 1988 ng/ml, p < 0.0001) (Fig. 4).

These results show that both diminished and deficient MBL pathway functions are more frequent in LB patients than in non-LB controls and suggest that the proper function of the MBL pathway, in the complement cascade, is protective against disseminated LB.

A generally accepted cut-off serum MBL concentration required for a normally activated MBL pathway has not been determined, and the cut-off levels for low MBL concentrations vary among studies (19). Based on the functional assessment of the MBL pathway, we determined that serum MBL levels < 787 and < 445 ng/ml are associated with diminished and deficient functions of the MBL pathway, respectively. In our study, diminished MBL pathway function was clearly more frequent in LB patients in comparison with non-LB controls. Also, the median serum MBL concentration was lower in LB patients than in controls. These findings suggest that impaired function of the MBL pathway increases susceptibility to LB. It must be noted that the serum cut-off levels for diminished and deficient MBL pathway functions were based on an analysis that was performed in a rather young patient group. However, in all age groups in our control population, the frequency of MBL deficiency was consistently 25–30%. This suggests that the frequency of diminished serum MBL level does not change substantially with age. Generally, with age, there is a decreasing trend in serum MBL concentrations. This finding is also in line with previous studies which show that the frequency of low serum MBL level (using the cut-off < 500 ng/ml) is ∼25% in the general population (12, 20).

Furthermore, we found that the frequencies of diminished and deficient MBL pathway functions were even higher in LB patients who had mounted an IgG response against B. afzelii p18 Ag compared with all LB patients. Oschmann et al. (17) showed that the presence of IgG Abs against p18 Ag corresponds to a more severe course of LB. It also was shown that p18 (DbpA) IgG positivity increases as the infection progresses from the early erythema migrans stage to neuroborreliosis or Lyme arthritis (21). Thus, it can be speculated that a diminished MBL pathway function leads to an impaired early-phase immune defense at the tick bite site; this could be an underlying factor allowing for a wider dissemination of the spirochetes, which is indicated by the presence of p18 IgG Abs.

The LB patient group and the non-LB control group were not matched. As explained in 2Materials and Methods, the age distribution was intentionally different between the groups; therefore, the statistical analyses performed were adjusted for age. According to our data, low serum MBL levels occurred in ∼25–30% of the non-LB control samples throughout the age groups, which is in agreement with previous studies in the general population (12, 20). According to a survey conducted in a general Finnish population by Aittoniemi et al. (22), the highest serum MBL concentrations were detected in early childhood, after which there was a decline in serum MBL levels with age. We also observed this trend in both of our study groups.

MBL deficiency is known to predispose individuals to certain autoimmune diseases, such as systemic lupus erythematosus (23), and the symptoms of autoimmune diseases might be similar to those of LB. This leads to the question of whether low serum MBL levels in the LB patient group may, in fact, reflect a potential autoimmune predisposition in the patients whose serum samples are subjected to borrelia serology. For this reason, we also selected non-LB control samples from sera that were sent to our laboratory for testing of borrelia serology instead of using data from blood donor samples. Among the non-LB control samples, the frequency of low MBL concentration in each age group corresponded with the occurrence of low MBL concentration in the general population (12, 20). This indicates that low serum MBL levels among the LB patient samples are indeed associated with LB infection and are not due to any bias in sample selection.

MBL is considered an acute-phase reactant, and its concentration may change in acute infection (24). To investigate whether our LB patients’ serum MBL concentrations were low in comparison with non-LB controls as a result of the acute borrelia infection and consumption of MBL, we identified 20 samples (MBL concentration 600–800 ng/ml) from LB patients who were retested for borrelia Abs 2–16 mo later and measured the MBL concentrations in the follow-up samples. There was no systematic change in MBL concentrations, which suggests that, in LB, the serum MBL level is stable, and MBL does not behave as an acute-phase protein (E.M. Sajanti and J. Hytönen, unpublished observations).

Three limitations of the study must be noted. First, genotyping was not performed because only retrospectively identified serum samples were available. Second, we also lacked MBL pathway functional data for the LB patient samples because of the retrospective nature of the study. However, our analysis showed an obvious association between MBL pathway function and serum MBL concentration. In addition, based on previous studies, it is recommended to use either a low serum MBL level or a direct functional assessment of the MBL pathway as an indicator of diminished MBL pathway function, instead of genotyping, when examining disease associations (12, 20). Third, because the borrelia-expressed ligand that interacts with MBL has not been identified and because LB in Europe is caused mainly by B. garinii and B. afzelii, a similar study should be performed in the United States to investigate whether deficient MBL pathway of the complement cascade also is a risk factor in patients in an area in which B. burgdorferi sensu stricto is the only encountered genospecies. However, the study that revealed the essential role of the MBL pathway in the eradication of borrelia was performed with B. burgdorferi sensu stricto, which suggests that similar results would be obtained in the United States (7).

Taken together, our study suggests that proper function of the MBL pathway of the complement cascade is protective against disseminated LB; thus, people whose complement-mediated defense is incomplete as a result of impaired MBL pathway function seem to be at a higher risk for developing disseminated and more severe LB.

We thank Päivi Haaranen and Anna Karvonen for technical assistance. We are also grateful to Prof. Sakari Suominen for advice concerning data interpretation and Robert Badeau Jr. for editing the manuscript.

The authors have no financial conflicts of interest.