ELISA kits for HSV1 IgG (Vir-Elisa Anti-HSV1-IgG, ref. 102) were obtained from VIRO-IMMUN, Oberursel, Germany. (Now: DIAsource ImmunoAssays S.A, Oberursel, Germany). We have performed a validation study where we found the sensitivity and specificity of this ELISA to be on par with the in house method [17], with antigen from HSV1-infected GMK cells, we use for routine analysis. Both methods show cross-reaction with anti-HSV2 IgG, as judged by a parallel run with the HSV1 & HSV2 type-specific HerpeSelect® IgG ELISA methods (Focus Diagnostics, California, USA) based on recombinant glycoprotein G. The sensitivity in scoring a serum positive or negative for presence of anti-HSV1 IgG was equal for these three methods [13].
Reagents were added manually with an 8-channel pipette and wells were washed with an automatic plate washer (Wellwash Versa Microplate Washer, ThermoScientific, Waltham, Massachusetts, USA). From a biobank collection (Betula) [18] of serum from adults (age 35–95), 28 random samples were tested for presence of anti-HSV1 IgG, according to the protocol provided by the manufacturer. In short, serum were diluted 1:101, where after 100 μl was added to each well and incubated at room-temperature for 30 min. Washing of wells were performed with 300 μl washing buffer 4 times, with 30 s soaking time. Peroxidase conjugate (100 μl) was added and incubated for 30 min at room-temperature, where after wells were washed with 300 μl washing buffer 4 times, with 30 s soaking time. TMB substrate (100 μl) was added and incubated, shielded from light, for 10 min at room temperature. Stop solution (100 μl) was added, and the absorbance was read at 450 nm (reference 620 nm) in a plate reader (model “Sunrise”, Tecan, Männedorf, Switzerland). The kit is provided with a cut-off control to which samples are related (positive: ODsample > 1.1x ODcut-off).
Sixteen samples were positive and subjected to avidity evaluation. Washing with 6 M Urea is earlier reported as a method for assessment of avidity for anti-HSV1 IgG [5, 6, 14, 15] and thus we modified the protocol to include such a washing after the serum incubation step. The Urea solution (150 μl) was allowed to soak the wells containing ag-bound anti-HSV1 IgG for 8 min, as reported earlier [6]. By dividing the values for Absorbance (Urea/no Urea) for doublets run with and without the Urea washing step, an Avidity Index (AI) with theoretical values between 0 and 100% is generated. AI for the 16 samples varied between 65 and 95%. However, when samples were reinvestigated, the reproducibility was poor (Intraclass correlation coefficient [ICC] < 0.8, data not shown). A number of variations to the protocol were performed; Urea concentration 2-8 M, incubation time 30s - 20 min, agitation of the plate while soaking, soaking at 4-37C, application of the urea washing step after an initial washing-buffer step, use of 0.75 M NaSCN. The resulting spread of avidity for the 16 samples varied between 42 and 100%. None of the protocol variants improved the reproducibility in a significant way (Intraclass correlation coefficient [ICC] < 0.8, data not shown).
This prompted us to instead challenge the initial binding of anti-HSV1 IgG to the antigen coated on the ELISA plate by dilution of sera with chaotropic salt (the dilution principle). For each sample we thus prepared additional samples diluted 1:101 in 2 M, 3 M, and 4 M Urea, respectively. Samples were applied to the standard ELISA protocol as described above. Spread of avidity for the samples varied between 69 and 88% (2 M), 51–82% (3 M), and 44–73% (4 M) (data not shown). Re-runs indicated good reproducibility, and the 4 M dilution protocol was chosen for further evaluation (see discussion). In the final method serum samples were clarified by centrifugation (1000 x g, 10 min) where after 10 μl was transferred to tubes containing: a) 1 ml sample diluent (provided in the kit), and b) 1 ml 4 M Urea (CAS Number 57–13-6, analysis grade, EMD Millipore) in sample diluent, prepared the same day. Diluted samples (100 μl), a & b for each serum, were immediately added to the ELISA plate wells and processed as described above, i.e. 30 min incubation followed by 4x washing etc. Background signal was subtracted from Absorbance (OD450) values, and an AI value was estimated for each serum sample (AbsUREA/AbsDiluent). Repeats were performed with fresh dilutions.