Association of pulmonary lymphocytes with radiation-induced lung disease in a mouse model

Background Specific congenic mice derived from inbred C3H/HeJ mice, which present early onset pneumonitis, and C57BL/6J mice manifesting later onset pneumonitis with pulmonary fibrosis, vary in time to respiratory distress from these responses following whole thorax irradiation. Methods Herein, to investigate a potential adaptive immunity contribution to lung disease in this model, we used flow cytometry to enumerate the pulmonary lymphocytes of C3H/HeJ, C57BL/6J and three lines of sub/congenic mice, at respiratory distress from 18 Gy whole thorax irradiation and in strain matched controls. Results CD4 + lymphocyte % of C3H/HeJ mice exceeded that of C57BL/6J mice at both radiation-induced respiratory distress and in unirradiated controls ( P < 0.04) and pulmonary CD4+% was reduced, at distress relative to control levels, in fibrosis responding strains. CD8 + lymphocyte % was reduced in distressed mice, compared to controls, for 8 of 10 comparisons by strain and sex including those of both pneumonitis and pneumonitis with fibrosis responses. γδ + lymphocyte % was largely unchanged at distress from control levels. Time to radiation-induced respiratory distress, and strain fibrosis score, were each negatively correlated with pulmonary CD4+%, and with the CD4/CD8 ratio, measured in distressed mice ( r <-0.76; P < 0.01) or in unirradiated controls ( r <-0.75; P < 0.02). Inclusion of the lymphocyte profile of unirradiated mice of a separate C3H/HeJ and C57BL6/J-derived congenic line, named Radpf1 and known to be spared radiation-induced lung disease, yielded a pulmonary CD4+% correlation with time to respiratory distress of ( r = 0.16, P = 0.76) and of ( r =-0.83; P = 0.04) for strain fibrosis score. Conclusions Pulmonary lymphocyte profiling revealed strain-dependent %CD4 + lymphocytes, measured in mice manifesting radiation-induced respiratory distress or in untreated controls, to correlate with fibrotic lung disease in this mouse model.