The 11 pneumococcal strains used in this assay are listed in Table 1 (provided by Moon H. Nahm, University of Alabama at Birmingham, Birmingham, AL, USA). Strains were selected for resistance to optochin (OREP), spectinomycin (SPEC), streptomycin (STREP), or trimethoprim (TREP) (Table 1). To create working stocks, ~10 µL of each bacterial master stock was streaked onto blood agar plates (Hanil Komed, Seongnam, Korea) and cultured overnight at 37°C with 5% CO₂. Bacterial colonies were subcultured in filter-sterilized Todd-Hewitt broth with 0.5% yeast extract (THY; Becton Dickinson, Sparks, MD, USA) in a 37°C water bath until reaching log-phase growth (OD₆₀₀ = ~0.8). Multiple working assay stocks were then cryopreserved with ~15% glycerol and stored at −80ºC until use.

Assay stocks were deemed acceptable when the following criteria were met: 1) purity, as determined by streaking on blood agar plates and overnight incubation; 2) resistance to the appropriate antibiotic at a 2× selection concentration and sensitive to the other three antibiotics at 0.5×; and 3) an assay dilution factor of ≥ 10³ to yield the ideal number of colonies per spot (80–150 CFU/spot).

Three quality control (QC) serum pools were prepared by mixing equal volumes of serum from four healthy young adults (12 total donors, aged 22–34 years) 1 month after vaccination with PPV23 (Pneumovax; Merck & Co., Inc., Whitehouse Station, NJ, USA). The serum pools were expected to have variable opsonic indices (OIs) for each serotype.

A total of 42 serum samples were collected from healthy adults aged 20–50 years. Of these samples, 35 samples were obtained 14 to 27 days after immunization with PPV23 (Prodiax23^®; Merck & Co., Inc.) by intramuscular injection (group 1). Seven samples were obtained from subjects immunized previously (46 to 123 months prior) with PPV23 (Prodiax23^®) (group 2). OIs for 11 additional non-PCV13 serotypes were studied for 42 serum samples. The accuracy of MOPA was studied using 20 selected sera because of the limited volume for analysis.

Collected samples were stored frozen at −70°C until testing. All serum samples were heat-inactivated at 56°C for 30 minutes to inactivate endogenous complement activity and confirmed as antibiotic-free in test cultures with a rough pneumococcal strain (R36A).

One technical difficulty with MOPAs is the nonspecific bacterial killing by the rabbit serum used as a source of complement. As such, multiple lots of 3–4-week-old baby rabbit complement (PelFreeze Biologicals; Rogers, AR, USA) were screened for potency and nonspecific killing defined as: 100 × (1 − [CFU from rabbit serum/HL-60 reaction]/[CFU from heat-inactivated rabbit serum/HL-60 reaction]). FBS (HyClone, Logan, UT, USA) was heat-inactivated and screened in OPA reactions. Lots with equal to or less than 30% nonspecific killing for all 11 serotypes in the absence of human serum were deemed acceptable for use.

An HL-60 promyelocytic leukemia cell line derived from a 36-year-old female patient was obtained from American Type Culture Collection (CCL-240; American Type Culture Collection [ATCC], Manassas, VA, USA).31 Cells were expanded and cryopreserved immediately upon receipt. Cell aliquots were thawed and propagated in RPMI 1640 media supplemented with 10% FBS and 1% l-glutamine. To generate phagocytes for the OPA, HL-60 cells were differentiated for 5 days in the presence of 0.8% dimethylformamide at a starting density of 4 × 10⁵ cells/mL. Differentiated HL-60 cells have been shown to express the markers of mature neutrophils when differentiated into granulocytes.3233 It was important to demonstrate the presence of these receptors under the differentiation conditions used in this particular assay, since complement-dependent opsonophagocytosis depends on the presence of the appropriate cell receptors on the effector cell. Differentiated HL-60 cells were considered acceptable as effector cells if the following criteria were met: 1) viability ≥ 90% (trypan blue), 2) CD35, also known as CR1 or C3b receptor which presents on granulocytes, expressed on ≥ 55% of cells, 3) CD71, presenting on proliferating cells as transferring receptor, expressed on ≤ 20% of cells, 4) ≤ 25% apoptotic cells, defined as Annexin V⁺/PI^– by flow cytometry using a FACS Calibur (Becton-Dickinson, San Jose, CA, USA).

Opsonic activity was evaluated by MOPA as previously described.30 Briefly, frozen aliquots of target pneumococci were thawed, washed twice, and diluted to the proper bacterial density (~5 × 10⁴ CFU/mL for each serotype). Serum samples were plated in 3-fold serial dilutions (20 µL/well) in round-bottom 96-well plates (Corning Inc., Corning, NY, USA). Then, 10 µL of bacterial suspension was added to each well in duplicate and incubated for 30 minutes at room temperature (25°C) with shaking (Mini Orbital Shaker; Bellco Biotechnology, Vineland, NJ, USA) at 700 rpm. Next, 10 µL of rabbit complement and 40 µL of differentiated HL-60 cells (4 × 10⁵ cells) were added to each well. Plates were incubated in a tissue culture incubator (37°C, 5% CO₂) with mechanical agitation at 700 rpm for 45 minutes and transferred to ice for 10–15 minutes. Next, an aliquot of the final reaction mixture (10 µL) was spotted onto four different THY agar plates (THY in 1.5% agar). When the fluid was absorbed into the agar, an equal volume of overlay agar (THY with 0.75% agar and 25 mg/L 2,3,5-triphenyltetrazolium chloride) containing one of the four antibiotics was added to each plate. After overnight incubation at 37°C, bacterial colonies were counted using the National Institute of Standards and Technology (NIST) Integrated Colony Enumerator (NICE), freely available in the US NIST (Gaithersburg, MD, USA; http://www.vaccine.uab.edu/NICE.pdf), and analyzed using Opsotiter 3.0 (http://www.vaccine.uab.edu). The OI was defined as the serum dilution that killed 50% of bacteria by linear interpolation, as previously described.30

MOPAs were subjected to accuracy, specificity, and reliability analyses to validate their application as an analytic methodology to evaluate pneumococcal vaccine immunogenicity. Since there is no standard or reference serum with assigned OPA values, OIs from the MOPA were compared with those from single-serotype OPAs (SOPAs) with 20 serum samples from vaccinated adults to determine assay accuracy. Assay specificity was examined by neutralizing QC sera in reactions with homologous PS (ATCC), three heterologous capsular PSs, or with no PS before being evaluated separately in the OPA. In an additional assay for serotype 20B, another capsular PS of serotype 20, known as serotype 20B, was used (SSISP20/2; Staten Serum Institute [SSI], Copenhagen, Denmark).34 Equal volumes of diluted serum and PS were mixed and incubated at room temperature (25°C) for 15 minutes. The sera were then serially diluted with opsonization buffer B (OBB), and the reactions were performed as described above. Intra-assay variation for the 11 serotypes was determined based on the coefficient of variation (CV) for a serum pool estimated 10 times in a single assay run. Similarly, inter-assay precision was estimated by testing several sera in 10 independent experiments over a 2-week interval. For these studies, the same lots of target bacteria and complement were used for each assay.

To apply this expanded MOPA to immune sera, the functional antibodies against 11 serotypes of pneumococcus (2, 8, 9N, 10A, 11A, 12F, 15B, 17F, 20B, 22F, and 33F) were evaluated in 42 sera immunized with PPV23 previously. Sera samples were assigned to two groups according to their intervals between immunization and blood sampling such as group 1 and group 2.

Correlation analysis of OIs from MOPAs and SOPAs was performed to assess accuracy, and correlation coefficient (r) values for each pneumococcal serotype were calculated by the Pearson's correlation coefficient using SPSS statistical software (version 23.0; IBM Corp., New York, NY, USA). The overall mean ± standard deviation (SD) and the CV of OIs were calculated to assess precision. OIs of 42 samples were reported as geometric mean indices (GMIs) and 95% confidence interval. GMIs between two groups were compared using the Mann-Whitney U nonparametric test on the 5% significance level.

The study protocol was reviewed and approved by the Institutional Review Board of Ewha Womans University Mokdong Hospital (ECT 11-13-43 and EUMC 2016-01-012-001). The study was conducted in accordance with clinical best-practices (national regulations and ICH E6) and the principles of the Helsinki Declaration. Written informed consent was obtained from all participants following a detailed explanation of the study.

Antibiotic-resistant strains were assessed with respect to resistance and sensitivity to the four antibiotics utilized in the MOPA. As expected, all 11 resistant strains showed 20% or less diminished growth when cultured in the presence of a 2× concentration of the appropriate antibiotic, but were sensitive to the other three antibiotics at half the concentration used in the assay with reduction rates ranging from 96% to 100% (data not shown). All 11 bacterial strains met the assay dilution cut-off of ≥ 10³.

The phenotype of HL-60 cells that differentiated into phagocytes over 5 days was assessed by flow cytometry (Fig. 1). The percentage of apoptotic cells (Annexin V⁺/PI^–) was 20.2%, which met the ≤ 25% acceptance criteria. Moreover, 83.0% and 11.8% of cells expressed CD35 (acceptance criteria, ≥ 55%) and CD71 (acceptance criteria, ≤ 20%), respectively, indicating sufficient phagocytic differentiation.

The nonspecific killing attributed to complement and FBS ranged from −14% to 12% and from −16% to 9%, respectively, which met the necessary criterion of equal to or less than 30% (Supplementary Table 1).

The median age of total volunteers was 35 years (range, 20–50 years), and that of group 1 and group 2 was 35 years (20–48 years) and 33 years (30–50 years), respectively. The mean ages were 35.0 ± 7.4 years for total, 34.5 ± 7.4 years for group 1, and 37.0 ± 8.5 years for group 2. The total number of males was 27, and 26 of them were assigned to group 1. The range between immunization and bleeding of group 1 was 14–22 days, and that of group 2 was 49–123 months.

The OIs obtained from forty-two clinical samples using this expanded MOPA are described in Table 3. The GMIs of group 1 for all serotypes were significantly higher than those of group 2 except for serotype 15B (P < 0.05).