|Home | About | Journals | Submit | Contact Us | Français|
Conjugate vaccines against group B Streptococcus (GBS), which is a leading cause of bacterial disease among newborns and the elderly with underlying illnesses, have progressed from animal studies to phase 1 and 2 clinical trials in healthy adults. Due to the wide-spread use of antibiotics to treat at-risk deliveries, a phase 3 efficacy trial of a GBS vaccine to prevent neonatal disease in the United States is unlikely. A viable approach to assess a vaccine's efficacy is to use a surrogate of protection which in the case of GBS is the opsonizing activity of serum antibody. The opsonophagocytosis assay (OPA) measures the ability of serum antibody to opsonize GBS for killing by effector cells in the presence of complement. In this report we demonstrate that differentiated HL-60 cells can substitute for human peripheral blood leukocytes (hPMNLs) in the OPA. Antisera to GBS type Ia CPS and type III CPS conjugate vaccines opsonized homologous GBS for killing at effector cells to GBS ratios of 2–4:1 regardless of whether HL-60 or hPMNLs were used. These results represent the first important step in developing a standardized, high-throughput OPA that could be used to assess the functional activity of vaccine-induced antibody and potentially serve as a surrogate of efficacy.
Vaccine development passes through several stages beginning with identification of the candidate antigen, the establishment of the proof-of-concept in relevant animal models of disease then through phase 1 and phase 2 clinical trails to assess safety and immunogenicity in relatively small numbers of healthy adults.1 Prior to licensure, the efficacy of a vaccine is assessed in a clinical phase 3 trial on a population that is at-risk of contracting the disease against which the vaccine was developed.
Recent amendments to the U.S. 21 Code of Federal Regulations 314 and 610, initially drafted in an effort to improve emergency responses to agents of bioterrorism, allows approval of certain drugs and biologics based on animal efficacy data when clinical phase 3 trials in humans cannot be conducted ethically.2 In addition, it has been suggested that a convincing argument to the FDA for licensure of a vaccine can be based not on a traditional efficacy study per se but on a surrogate of protection.3 As stated by Qin et al.3 the “surrogate of protection is a correlate of risk that reliably predicts a vaccine's level of protective efficacy on the basis of contrasts in the vaccinated and unvaccinated groups.”
This change to the Code of Federal Regulations and discussions of surrogates of protection in lieu of pivotal phase 3 trials have particular importance to the development of a maternal vaccine to prevent neonatal group B streptococcal disease. Group B Streptococcus (GBS) remains a leading cause of neonatal- and perinatal maternal-morbidity despite dramatic reductions in cases due to intrapartum antibiotic prophylaxis measures.4 Viable GBS vaccine candidates, namely conjugated capsular polysaccharide (CPS) vaccines, have safely advanced to phase 1 and 2 testing in healthy adults5 including pregnant women at 30 to 32 weeks of gestation.6 However, a traditional phase 3 efficacy study of a maternal GBS vaccine will likely not be performed in the United States or elsewhere because of ethical issues and high antibiotic use. Therefore, the development of a high throughput in vitro assay to quantitatively and reproducibly measure functional antibody to GBS in a standardized fashion is required. The opsonophagocytosis assay (OPA) is a functional measurement of the ability of antibody to opsonize viable GBS for killing by effector cells.7-9 In the past, our laboratory has used human peripheral polymorphonuclear leukocytes (hPMNLs) from healthy donors as effector cells. While effective, humans effector cells are not a logical for use in a standardized OPA compared with differentiated HL-60 cells which can be cultured in vitro to high yields. Differentiated HL-60 cells are the effector cells in OPAs in routine use by the pneumococcal vaccine field.10-12
To the best of our knowledge, only two extended abstracts13,14 and a recent publication from our laboratory15 are the only reports of the use of differentiated HL-60 cells to assess the functional activity of GBS antibody in an OPA. Absent from these publications, however, were studies comparing HL-60 cells with the ‘gold standard’ hPMNLs as effector cells to determine the validity of their use in an OPA. Validation of an OPA that uses differentiated HL-60 cells is an important first step in the development of a standardized, high-throughput OPA to assess the functionality of GBS vaccine-induced antibody.
This study was undertaken to determine if differentiated HL-60 cells (Fig. 1) could replace human peripheral blood leukocytes in an OPA designed to measure the functional activity of antibody against GBS. An important variable in the OPA is the effector cell to target cell ratio (E:T). The E:T in the GBS ‘gold standard’ OPA assay,7 in which enriched human polymorphonuclear leukocytes were used as phagocytes, was 2 to 4.9 Because this low ratio is in contrast to the E:T ratio of 200 to 400 used in the standardized HL-60 OPA to measure opsonic capacity of pneumococcal antibodies,16,17 we compared E:T ratios varying from 90 to 2 to 4 in the HL-60 GBS OPA to measure opsonic capacity of standard rabbit reference sera Ia.
Increasing the E:T ratio compared to that used in our ‘gold standard’ assay offered no advantage in our HL-60 GBS OPA. The mean ± standard deviation log10 reduction in CFU using HL-60 cells was 1.1 ± 0.1, irrespective of the E:T ratio; 90 (Fig. 2A), 9 (Fig. 2B), or 3 (Fig. 2C) using a 1:1,000 dilution (i.e., 49 ng type Ia CPS-specific antibody in the reaction mixture) of standard rabbit reference serum Ia. Further dilution of standard rabbit reference serum Ia resulted in no GBS killing irrespective of the E:T ratio. The difference between the mean reduction in GBS CFU with and without specific antibody in the reaction mixture was 1.8 ± 0.1 (Fig. 2A–C). Reaction mixtures without antibody and/or complement, or with pre-vaccination rabbit serum resulted in GBS growth (Fig. 2A–C).
To directly compare the results obtained in the GBS HL-60 OPA to those obtained with our ‘gold standard’ OPA,9 the ‘gold standard’ GBS OPA was performed with standard rabbit reference serum Ia using freshly isolated hPMNLs. The reduction in GBS CFU obtained with HL-60 cells was identical to that obtained with hPMNLs a the same E:T ratio of 3 (Fig. 2D), a result that suggests that HL-60 cells can replace hPMNLs for evaluation of functional activity of vaccine-induced antibody.
To insure that the results obtained were not limited to standard rabbit reference serum Ia, we tested a second standardized rabbit sera with specificity to type III CPS, standard rabbit reference serum III, raised to type III CPS-tetanus toxoid conjugate.18 We compared the degree of antibody-mediated opsonic killing of GBS type III strain M781 obtained with differentiated HL-60 cells with that obtained with use of hPMNLs. The levels of GBS CFU reduction were similar whether differentiated HL-60 cells or hPMNLs were used; mean ± standard deviation in the log10 reduction in CFU/ml using standard rabbit reference serum III diluted 1:1,000 was 0.73 ± 0.4 (Fig. 3A) and 0.89 ± 0.6 (Fig. 3B), respectively. Further dilution of the standard rabbit reference serum III resulted in GBS growth as did controls that lacked antibody and/or complement or that contained normal rabbit serum. The difference in the mean and standard deviation of the log10 reduction in CFU/ml in reaction mixtures containing either differentiated HL-60 effector cells or hPMNLs with type III CPS-specific IgG (diluted 1/100) and complement compared to mixtures lacking antibody was 1.37 ± 0.21 and 1.88 ± 0.30, respectively (p = 0.20). Thus, HL-60 cells can substitute for hPMNLs in the evaluation of functional activity of vaccine-induced GBS CPS-specific antibody.
While CPS-specific rabbit sera are excellent for use in developing assays and as internal standards, we sought to verify that CPS-specific human sera would opsonize GBS for complement mediated killing by HL-60 cells as have been shown previously for hPMNLs.8,9,19 Standard streptococcal human reference serum for type III diluted 1:100 (21 ng of type III CPS-specific IgG in the reaction mixture) resulted in a mean log10 reduction in CFU/ml of 0.9 (range 0.7–1.2) and 1.2 (range 0.9–1.5) with HL-60 cells and hPMNLs, respectively (Fig. 4). The difference in mean log10 reduction in CFU/ml with and without antibody was 1.35 and 1.39 when HL-60 cells and hPMNLs, respectively, were effector cells. Both types of effector cells did not kill GBS in the presence of human sera without specific antibody. The difference in the mean and standard deviation of the log10 reduction in CFU/ml in reaction mixtures containing either differentiated HL-60 effector cells or hPMNLs with type III CPS-specific IgG (diluted 1/100) and complement compared to mixtures lacking antibody was 1.35 ± 0.29 and 1.23 ± 0.42, respectively (p = 0.60). These results suggests that differentiated HL-60 cells could substitute for human effector cells in a OPA when the opsonic activity of human antibody is sought.
An effective GBS vaccine ought to be multivalent in formulation to provide coverage against the five GBS serotypes (Ia, Ib, II, III and V) currently responsible for the majority of disease in the United States and throughout most of the world. Accurate measurements of GBS CPS-specific, vaccine-induced IgG20 combined with a validated OPA that reflects functionality of those antibodies can serve as surrogates of vaccine efficacy. Recently, a four-fold multiplex OPA was validated and standardized to determine the opsonophagocytic antibody titers to 13 pneumococcal serotypes in a high-throughput fashion using HL-60 cells as effector cells.12 Our results using differentiated HL-60 cells to assess functional activity of CPS-specific antibody and the report of a pneumococcal multiplex OPA provide the framework necessary to develop a similar high throughput method specific for GBS.
In summary, differentiated HL-60 cells can replace hPMNLs in the OPA to measure the functional activity of vaccine-induced CPS-specific antibody to GBS. A validated and standardized high-throughput OPA with HL-60 cells combined with accurate measurement of specific antibody may be used as a surrogates of efficacy for GBS vaccines.
GBS type Ia strain 515 and type III strain M781 were obtained from the Channing laboratory culture collection and grown in sealed tubes containing Todd-Hewitt broth to mid-exponential growth (A650 nm = 0.3; ~3 × 107 colony-forming units (CFU) per milliliter). GBS were pelleted by centrifugation, washed with saline, suspended in modified Eagles medium and used in the assay.
Promyelocytic HL-60 cells16 were obtained from the American Type Culture Collection (CCL-240). Cells were thawed according to ATCC specifications in RPMI 1640 (Difco Laboratories, Detroit, MI) containing 10% FBS (RPMI-10) and grown in upright vented tissue culture flasks at 37°C with 5% CO2 in single cell suspension cultures and expanded to a density of 6 × 105 cells/ml. HL-60 cells were then harvested by centrifugation (160 × g for 10 min at room temperature), and the cell density and viability verified by trypan blue exclusion; cells were split and expanded from a density of 2 × 105 cells/ml to 5–7 × 107 cells/ml before harvesting by centrifugation. The HL-60 cells were adjusted to a cell density of 2 × 105 cells/ml in RPMI-10 and differentiated into granulocyte-like cells by the addition of 100 mM N,N dimethylformamide to the growth medium.10 After 4 to 5 days of differentiation ≥95% of the cells expressed CD11b (iC3b receptor)16 as verified by flow cytometry following binding with FITC-labeled anti-CD11b antibodies (BD Pharmingen) (Fig. 1). HL-60 seed stock from the initial expansion (cell density of 5–6 × 106 cells/ml as per ATCC recommendations) was maintained at -80°C and subsequent expansions were prepared from this stock. For consistency, HL-60 cells from Channing Laboratory passage numbers 1, 2 or 3 were expanded for differentiation and use in the OPA.
OPAs were performed with rabbit and human sera. The type III CPS standard rabbit reference serum was raised to type III CPS-tetanus toxoid conjugate (III-TT) vaccine;18 it contained 0.93 mg/ml of type III CPS-specific antibody. The type Ia CPS standard rabbit reference serum was raised to Ia-TT vaccine; it contained 0.98 mg/ml of type Ia CPS-specific antibody.21,22 Normal rabbit sera was used as a negative control.
A type III CPS standard human reference serum contained 83.5 μg/ml of type III CPS-specific IgG and 4.5 μg/ml of specific IgA and no detectible levels of specific IgM was also used as a test serum.8,19 Serum from a healthy, non-vaccinated individual containing <0.05 μg/ml of type III CPS-specific IgG was used as a negative control.
The OPA combines viable GBS with effector cells to test the ability of an antiserum to opsonize the bacteria for killing in the presence of complement. Reaction mixtures (250 μl total in modified Eagle's medium) consisted of heat-inactivated (56°C for 30 min) test serum (50 μl of rabbit sera or 25 μl of human sera), differentiated HL-60 cells or freshly prepared hPMNLs (150 μl), GBS cells (25 μl), and 25 μl of 10% baby rabbit complement (Cedarlane, Burlington, NC). Control reactions lacked complement and/or antibody, effector cells or all components except GBS. The effector cell to GBS cell ratio varied from 90:1 to 2 to 4:1. Reaction mixtures were incubated at 37°C for 1h with end-over-end mixing. Aliquots were removed prior to and after incubation, serially-diluted in 0.9% saline and samples plated on tryptic soy agar or blood agar plates.
To insure that streptococcal chains were adequately disrupted, each reaction mixture was vortexed for 4 s prior to plating. This allowed for sample- to sample and day-to-day comparisons of results. In a pilot experiment, microscopic evaluation of 1,300 Gram-stained GBS sampled before and after the 1 h incubation at 37°C from all test and control mixtures revealed an overall average and standard deviation of 1.6 ± 0.7 cocci per chain, a result that indicates substantial and consistent disruption of the streptococcal chain.
The plates were incubated overnight at 37°C and GBS CFUs enumerated by standard plate count. The results are expressed as the difference in the log10 reduction in GBS CFU/ml before and after 1 h incubation. At a minimum, all samples and controls were tested in duplicate with several experiments performed at least twice as described in the figure legends.
Statistical analysis was performed with use of the two-tailed Mann-Whitney test.
We thank Stacie Bush and Julia C. West for excellent technical assistance. This work was supported by NIH-NIAID grant AI-060603.