Comparative Automated Quantification of Tertiary Lymphoid Structures in Two Distinct Mouse Models of Inflammation.

Abstract

Tertiary lymphoid structures (TLS) have been reported to form within nonlymphoid tissues upon various inflammatory conditions, such as tumors or bacterial infections. In particular, the lungs of patients with NSCLC or bacterial infections (such as tuberculosis or aspergillosis) are the sites of TLS neogenesis. An increasing number of preclinical models have been used over the years to recapitulate as accurately as possible the mechanisms and kinetics of TLS formation that are continuously reported in the clinic.We used herein two distinct murine models of pulmonary inflammation, relying on orthotopic lung tumor implantation, or intranasal administration of bacterial products, to generate TLS formation within mouse lung tissue. This review compares different workflows aiming at the detection and quantification of TLS using automated image analysis and artificial intelligence-assisted tissue classification. We also describe different methods for partitioning serial sections, depending on the tissue morphology and the experimental goal.