Tuberculosis (TB) is one of the leading causes of death, with approximately 1.3 million people dying from TB infection each year1. Though not as lethal as tuberculosis, there has been a significant increase in pulmonary infections and subsequent disease caused by nontuberculous mycobacteria (NTM) in recent years2. At the tissue level, the host immune response to pulmonary Mycobacterium tuberculosis (Mtb) infection has been well studied and includes the formation of granulomas, heterogeneous aggregates of lymphocytes, dendritic cells, macrophages, and neutrophils3. While granulomatous inflammation is also a fundamental feature of host immunity to NTM infection, the immune cell populations and the spatial characteristics of this response remain understudied. Gaining a detailed understanding of the tissue components of immunopathology versus wound healing is critical for the development of treatments (e.g., host directed therapies) and preventative (i.e., vaccines) therapeutic strategies.
In this study, our team had access to rare, archived clinical samples acquired during lung resections or autopsies from patients with Mtb or NTM infection. The majority of these samples were over 10 years old and preserved as formalin-fixed, paraffin-embedded (FFPE) samples, restricting their use to histological assays. To overcome the marker limitations of traditional immunohistochemistry (IHC) and immunofluorescence (IF) imaging, we utilized the Iterative Bleaching Extends multi-pleXity (IBEX) method, an open-source, highly multiplexed imaging method4,5. Extension of the IBEX method to human lung FFPE samples required significant optimization due to high levels of endogenous fluorescence, a well described obstacle for multiplexed antibody-based imaging6.
Here, we provide a detailed protocol for reducing autofluorescence using a photoirradiation method optimized for lung FFPE tissue sections7,8. Importantly, the method described here utilizes an inexpensive, widely available light box that can be extended to other sample preparations with high autofluorescence. In addition to reducing autofluorescence prior to image acquisition, we provide guidance on antibody panel design such as pairing low abundance markers with bright fluorophore labels, evaluating the impact of photoirradiation on antibody labeling, and amplification with secondary antibodies in channels with high endogenous fluorescence. On average, we processed five slides corresponding to one tissue section, each ranging from 37 to 114 mm2. Finally, we demonstrate how to acquire whole-slide images using a widefield instrument. In summary, this protocol describes the adoption of IBEX to archived clinical samples and additionally offers guidance on how to troubleshoot issues commonly encountered with imaging highly autofluorescent samples.