Development and validation of a molecular size distribution method for polysaccharide vaccines.

Determination of the molecular size distribution of vaccine products by high performance size exclusion chromatography coupled to refractive index detection is important during the manufacturing process. Partial elution of high molecular weight compounds in the void volume of the chromatographic column is responsible for variation in the results obtained with a reference method using a TSK G5000PWXL chromatographic column. GlaxoSmithKline Vaccines has developed an alternative method relying on the selection of a different chromatographic column with a wider separation range and the generation of a dextran calibration curve to determine the optimal molecular weight cut-off values for all tested products. Validation of this method was performed according to The International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). The new method detected product degradation with the same sensitivity as that observed for the reference method. All validation parameters were within the pre-specified range. Precision (relative standard deviation (RSD) of mean values) was < 5 per cent (intra-assay) and < 10 per cent (inter-assay). Sample recovery was > 70 per cent for all polysaccharide conjugates and for the Haemophilus influenzae type B final container vaccine. All results obtained for robustness met the acceptance criteria defined in the validation protocol (≤ 2 times (RSD) or ≤ 2 per cent difference between the modified and the reference parameter value if RSD = 0 per cent). The new method was shown to be a suitable quality control method for the release and stability follow-up of polysaccharide-containing vaccines. The new method gave comparable results to the reference method, but with less intra- and inter-assay variability.

Validation of a new ELISA method for in vitro potency assay of hepatitis B-containing vaccines.

The discontinuation of the Auszyme kit used by vaccine manufacturers and national control laboratories to determine the Hepatitis B surface antigen (HBsAg) content of hepatitis B vaccines has led GlaxoSmithKline (GSK) to develop an alternative inhibition ELISA method. Validation of this ELISA was performed according to The International Conference of Harmonization and reproducibility was assessed in a feasibility study with four Official Medicines Control Laboratories (OMCLs). The dose response curve demonstrated linearity (R2>0.99) in the range of 60-360 ng/ml HBsAg. The repeatability (CV<7%), intermediate precision (CV<10%) and accuracy (91-113% recovery) were similar to the Auszyme method. The commercial antibodies used in the assay were shown to contain antibodies that bind to a protective epitope of HBsAg and the specificity of the method for HBsAg was demonstrated. There was a good concordance with the Auszyme method, although the ELISA yielded higher results (25.3 vs. 24.4 micro.g/ml for Engerix-B (n=64), 28.9 vs. 27.0 micro.g/ml for Twinrix (n= 69) and 25.5 vs. 21.6 micro.g/ml for Infanrix penta (n=62)). The method was successfully transferred to the four OMCLs. It has been demonstrated that the ELISA is suitable for its intended purpose with hepatitis B-containing vaccines from GSK and thus could be used for these vaccines by national control laboratories and authorities. Further validation studies should focus on the use of this ELISA with vaccines from other manufacturers.

Polymeric IgA binding to the human pIgR elicits intracellular signalling, but fails to stimulate pIgR-transcytosis.

The intracellular pathway of polymeric immunoglobulin receptor (pIgR) is governed by multiple signals that lead to constitutive transcytosis. In addition, in transfected polarized MDCK cells, polymeric immunoglobulin A (pIgA) binding stimulates rabbit pIgR-transcytosis, owing to phospholipase-C gamma 1 activation and increase of intracellular calcium. Transcytosis of rat pIgR across hepatocytes is similarly accelerated by pIgA injection. In contrast we show here that human Madrin-Darby Canine Kidney (pIgR)-transcytosis, in human Calu-3 and human pIgR-transfected MDCK cells, is not promoted by pIgA, as monitored by a continuous apical release of its secreted ectodomain. However, the incubation of cells expressing human or rabbit pIgR with pIgA induces a comparable IP3 production, and pIgR-transcytosis of either species is accelerated by the protein kinase C (PKC)-activator phorbol myristate acetate. Without pIgA, mimicking phospholipase-C activation by combining low concentrations of phorbol myristate acetate with ionomycin, or high concentrations of ionomycin alone, stimulates the rabbit, but not the human, pIgR transcytosis. These data suggest that the species difference in pIgA-induced pIgR-transcytosis does not stem from the defective production of second messengers, but from a different sensitivity of pIgR to intracellular calcium. Our results outline the danger of extrapolating to humans the abundant data obtained from mucosal vaccination of laboratory animals.

Lack of SC/pIgR-mediated epithelial transport of a human polymeric IgA devoid of J chain: in vitro and in vivo studies.

Three human polymeric IgA (pIgA) myeloma proteins of tetrameric size were compared for their J-chain content, their in vitro secretory component (SC)-binding ability, and their capacity to be transcytosed by polymeric immunoglobulin receptor (pIgR)-expressing epithelial cells in vitro and rat hepatocytes in vivo. One of the three pIgA preparations, pIgA-L, was shown to lack J chain and was unable to combine with purified free human and rat SC, whereas pIgA-G and pIgA-C contained J chain and combined readily with SC. Furthermore, pIgA-L was not transferred into rat bile after intravenous injection, and was hardly transported apically by polarized Madin-Darbey canine kidney cell monolayers expressing the human pIgR, whereas pIgA-G and pIgA-C were efficiently transported in both test systems. Together with our recent demonstration that antibodies to human J chain block the SC/pIgR-mediated epithelial transport of pIgA, these data unanimously confirm the proposed key role of J chain in the epithelial transport of polymeric immunoglobulins into exocrine secretions.

Secretory component production by human bronchial epithelial cells is upregulated by interferon gamma.

Secretory immunoglobulin A (S-IgA) participates in the first noninflammatory line of defence of the respiratory tract. S-IgA consists of dimeric IgA (dIgA) produced by plasma cells and secretory component (SC) produced by epithelial cells. This study compared SC production by primary cultures of human bronchial epithelial cells (HBEC) and by respiratory epithelial cell lines. Among the cell lines, A549 did not produce detectable SC, 16HBE produced very low levels of SC, while CALU-3 produced significant levels of SC. HBEC produced SC in nonpolarized and polarized primary cultures, where it was secreted apically. Polarized HBEC transcytosed radiolabelled and cold dIgA, resulting in the presence of S-IgA in their apical media. SC production and IgA transcytosis by polarized HBEC were upregulated by interferon-gamma (IFN-gamma) after 48 h. By reverse transcription-polymerase chain reaction, no SC messenger ribonucleic acid (mRNA) was detected in A549 and 16HBE, while SC mRNA in CALU-3 was comparable to that of HBEC incubated for 48 h with IFN-gamma. By immunocytochemistry, HBEC expressed SC immunostaining and its intensity increased after 48 h with IFN-gamma. It is concluded that human bronchial epithelial cells produce secretory component and transcytose dimeric immunoglobulin A in vitro. These processes were apically polarized and upregulated by interferon-gamma. Among the cell lines studied, only CALU-3 expressed secretory component-messenger ribonucleic acid and produced detectable secretory component.

In vivo stimulation of polymeric Ig receptor transcytosis by circulating polymeric IgA in rat liver.

Binding of human polymeric IgA ligand to its epithelial cell polymeric Ig receptor, pIgR, has been shown to stimulate pIgR apical transcytosis in an in vitro system, based on polarized confluent MDCK cells expressing rabbit pIgR. The present study aimed at testing whether such a stimulation also occurs in vivo. Transcytosis of pIgR was monitored by rat liver output of total secretory component (SC) into bile, measured by radial immunodiffusion as the sum of free SC and pIgA-bound SC. Whereas in the perfused rat liver system addition of pIgA to the perfusate showed no effect, i.v. injection of human and rat pIgA, but not of monomeric IgA nor PBS, in living rats significantly increased total bile SC output for more than 1 h. Furthermore, depletion of the normal pIgA level circulating in the liver before injecting more pIgA was not required to show the stimulation. Our data thus strongly suggest that stimulation of liver pIgR transcytosis by pIgA ligand binding is physiologically relevant, helping to quickly adjust pIgA transport into bile to increase circulating pIgA levels, without need for increased SC/pIgR synthesis.

Antibody against the human J chain inhibits polymeric Ig receptor-mediated biliary and epithelial transport of human polymeric IgA.

To emphasize the requirement for a J chain in native polymeric immunoglobulins for their selective transport into exocrine secretions, IgG, purified from two different antisera specific for the human J chain, was shown to: (i) bind in vitro to human polymeric IgA (pIgA) by density gradient ultracentrifugation; (ii) inhibit binding in vitro of rat secretory component to human pIgA; (iii) inhibit hepatic transport of human pIgA into rat bile in vivo; and (iv) inhibit apical transcytosis of pIgA in vitro by polarized human polymeric immunoglobulin receptor (pIgR)-expressing Madin-Darby canine kidney cells. Inhibition of biliary transport increased with the molar ratio of anti-J chain antibodies against pIgA and their incubation time. Anti-J chain F(ab')2 and Fab fragments also inhibited biliary transport, excluding a role for phagocytic clearance or excessive size of the immune complexes. Anti-human-Fc alpha Fab, bound to human pIgA in complexes of larger size than those with anti-J chain Fab, did not inhibit biliary transport of human pIgA. Propionic acid-denatured human pIgA, although containing J chains, was very poorly transported into rat bile. Altogether, our data strongly support, now also by in vivo experiments, the crucial role of the J chain of native pIgA in its selective pIgR-mediated transport into secretions, as suggested long ago by in vitro data only. Recent data on J chain-knockout mice, with low IgA levels in bile and feces, cannot explain the role of the J chain in contributing to the secretory component/pIgR-binding site of normal pIgA, but otherwise agree with our study.