A comparison of the direct and reporter antigen popliteal lymph node assay for the detection of immunomodulation by low molecular weight compounds

Abstract

The direct popliteal lymph node assay (PLNA) is a predictive test used to detect the immune-stimulating potential of pharmaceuticals and other low molecular weight compounds (LMWCs) with known autoimmunogenic or sensitizing properties. Two limitations in the PLNA are the existence of false negatives and the inability of the assay to provide mechanistic information. Recently the direct PLNA was modified by incorporating reporter antigens (RA), either TNP-Ficoll or TNP-OVA. In the RA-PLNA, immune stimulation is detected by measuring IgM or IgG TNP-specific antibody-forming cells (AFC) using an enzyme-linked immunospot (ELISPOT) assay. The RA-PLNA, when using potent, known autoimmunogenic compounds, may provide greater sensitivity compared to the direct PLNA and might distinguish LMWCs that have intrinsic adjuvant activity from those that create neo-antigens, using TNP-OVA and TNP-Ficoll, respectively. The purpose of this study was to rigorously compare the two assays. Our first objective was to investigate the interlaboratory reproducibility of the RA-PLNA using four autoimmunogenic LMWC models, plus one negative control LMWC. Subsequently, we tested seven LMWCs with known sensitizing properties and compared the results from the direct and modified assay. The test group included LMWCs thought to be mechanistically distinct and similar to compounds typically encountered in preclinical safety assessment. All control and treatment AFC plaques were collected (76 total), pooled, coded to conceal their source, and counted. The interlaboratory reproducibility of the RA-PLNA was demonstrated with the model autoimmunogenic compounds HgCl2, diphenylhydantoin, D-penicillamine, and the negative control compound phenobarbital, by detecting TNP-specific IgM and polyclonal IgG production to both reporter antigens. Additionally, the sensitizing effects of streptozotocin were identified using an IgG2a ELISPOT with both TNP-OVA and TNP-Ficoll. With the extended test group, the sensitizing effects of aniline, a false negative LMWC in the direct PLNA, was not detected in this study when using the direct PLNA. However, there was an increase of IgG1 AFCs using TNP-OVA, when compared to control (508 +/- 113 vs. 12 +/- 4 respectively). Glafenine, diclofenac, and ibuprofen, all associated with drug-induced anaphylaxis in humans, produced significant increases in IgG1 production to TNP-OVA. Of these three LMWCs, only diclofenac, which has been documented to induce neo-antigen formation, was detected with TNP-Ficoll. Hydralazine immunomodulation could be detected only with the direct PLNA although significant increases in IgM were identified with the co-injection of either reporter antigen. Isoniazid and methyldopa consistently produced negative responses in both assays. In summary, this study has demonstrated acceptable interlaboratory reproducibility of the RA-PLNA, using model autoimmunogenic LMWCs. Additionally, it demonstrated that an advantage of the RA-PLNA was that it identified all anaphylactic-associated LMWCs tested, detected the false negative compound aniline, and revealed what is thought to be the mechanism(s) associated with diclofenac-induced immunostimulation.