Development and Validation of a Protein Array for Detection of Antibodies against the Tick-Borne Pathogen Borrelia miyamotoi

Abstract

Current serological tests for the emerging tick-borne pathogen Borrelia miyamotoi lack diagnostic accuracy. To improve serodiagnosis, we investigated a protein array simultaneously screening for IgM and IgG reactivity against multiple recombinant B. miyamotoi antigens. The array included six B. miyamotoi antigens: glycerophosphodiester phosphodiesterase (GlpQ), multiple variable major proteins (Vmps), and flagellin. Sera included samples from cases of PCR-proven Borrelia miyamotoi disease (BMD), multiple potentially cross-reactive control groups (including patients with culture-proven Lyme borreliosis, confirmed Epstein-Barr virus, cytomegalovirus, or other spirochetal infections), and several healthy control groups from regions where Ixodes is endemic and regions where it is nonendemic. Based on receiver operating characteristic (ROC) analyses, the cutoff for reactivity per antigen was set at 5 μg/mL for IgM and IgG. The individual antigens demonstrated high sensitivity but relatively low specificity for both IgM and IgG. The best-performing single antigen (GlpQ) showed a sensitivity of 88.0% (95% confidence interval [CI], 78.9 to 93.5) and a specificity of 94.2% (95% CI, 92.7 to 95.6) for IgM/IgG. Applying the previous published diagnostic algorithm-defining seroreactivity as reactivity against GlpQ and any Vmp-revealed a significantly higher specificity of 98.5% (95% CI, 97.6 to 99.2) but a significantly lower sensitivity of 79.5% (95% CI, 69.3 to 87.0) for IgM/IgG compared to GlpQ alone. Therefore, we propose to define seroreactivity as reactivity against GlpQ and any Vmp or flagellin which resulted in a comparable sensitivity of 84.3% (95% CI, 74.7 to 90.8) and a significantly higher specificity of 97.9% (95% CI, 96.9 to 98.7) for IgM/IgG compared to GlpQ alone. In conclusion, we have developed and validated a novel serological tool to diagnose BMD that could be implemented in clinical practice and epidemiological studies. IMPORTANCE This paper describes the protein array as a novel serological test for the diagnosis of Borrelia miyamotoi disease (BMD), by reporting the methodology, the development of a diagnostic algorithm, and its extensive validation. With rising numbers of ticks and tick bites, tick-borne diseases, such as BMD, urgently deserve further societal and medical attention. B. miyamotoi is prevalent in Ixodes ticks across the northern hemisphere. Humans are exposed to, and infected by, B. miyamotoi and develop BMD in Asia, in North America, and to a lesser extent in Europe. However, the burden of infection and disease remains largely unknown, due to the noncharacteristic clinical presentation, together with the lack of awareness and availability of diagnostic tools. With this paper, we offer a novel diagnostic tool which will assist in assessing the burden of disease and could be implemented in clinical care.

Keywords: Borrelia miyamotoi; Borrelia miyamotoi disease; hard-tick-borne relapsing fever; relapsing fever Borrelia; serology.

FIG 1

Absolute antibody response in the protein array against recombinant B. miyamotoi antigens in BMD patients and control groups. Mean values with standard errors of the means for IgM and IgG concentrations in acute-phase (0 to 4 days after disease onset), early-convalescent-phase (5 to 25 days after disease onset), late-convalescent-phase (26 to 149 days after disease onset), and reconvalescent-phase (150 to 499 days after disease onset) BMD samples, compared to samples from disease controls (culture-proven Lyme borreliosis [LB] patients), cross-reactive controls with other infectious diseases (ID; patients with serologically proven leptospirosis, syphilis, CMV, EBV, or HSV I/II infection or culture-proven H. influenzae bacteremia), healthy blood donors from regions (in Russia and The Netherlands) where Ixodes ticks are endemic, and healthy controls from regions (in Russia and Norway) where the ticks are not endemic. B., Borrelia; BMD, Borrelia miyamotoi disease; CMV, cytomegalovirus; E, endemic; EBV, Epstein-Barr virus; GlpQ, glycerophosphodiester phosphodiesterase; HBD, healthy blood donors; HSV, herpes simplex virus; Ig, immunoglobulin; ID, infectious diseases; LB, Lyme borreliosis; NE, nonendemic; Reconv., reconvalescent; Vlp, variable large protein; Vsp, variable small protein.

FIG 2

Sensitivity of individual antigens and combinations of antigens in the protein array for BMD patients over time. Sensitivity with standard error of the mean for IgM and IgG is given for individual antigens and combinations of antigens. The sensitivity is shown for BMD patients for each time window. Cutoffs for both IgM and IgG were set at 5 μg/mL with a specificity of >95% for IgM and a specificity of >93% for IgG. Combinations of antigens are shown with either “or” or “and.” “Or” indicates that either antigen-specific-antibody level was increased; “and” indicates that both antigen-specific-antibody levels were increased. BMD, Borrelia miyamotoi disease; GlpQ, glycerophosphodiester phosphodiesterase; Ig, immunoglobulin; Vmp, variable major protein.