Simultaneous detection of DNA, RNA and proteins within the native tissue microenvironment is a trifecta that promises to deliver valuable insights into fundamental aspects of the biological processes underlying health and disease. Recent technical advances in both in situ hybridization and highly multiplexed imaging have enabled this combined analysis, yet many approaches are limited by poor detection of low copy numbers or difficulties owing to protein epitope loss during tissue processing. Our approach, termed PANINI (protein and nucleic acid in situ imaging), enables the detection of low-copy nucleic acids while preserving protein epitopes in intact tissues. This is achieved by bypassing the protease digestion step, which is used in various in situ hybridization assays to increase the accessibility of target nucleic acids by disrupting the packed architecture of tissue matrices and nucleic acid-binding proteins. PANINI starts with a protease-free antigen retrieval protocol before the application of a highly sensitive custom branched-chain amplification method in which horseradish peroxidase catalyses the deposition of tyramine-linked reporter haptens for nucleic acid labelling. These labelled nucleic acid targets, in conjunction with other protein targets, can then be detected using various antibody-based imaging platforms. We demonstrated that PANINI is compatible with multiplex spatial imaging platforms, including conventional microscopes, multiplexed ion beam imaging (MIBI), co-detection by indexing (CODEX) and multispectral imaging on the Vectra Polaris platforms1.

We developed PANINI initially to address fundamental questions regarding tissue reservoirs harbouring human and simian immunodeficiency virus (HIV and SIV, respectively): where are HIV-infected cells located within the tissues, and how does their tissue microenvironment affect the viral transcription state? This required an approach that was sensitive enough to detect even single copies of integrated viral DNA, while retaining the multiplexing capabilities of spatial omics platforms. Our results revealed an interesting communication between B cells and macrophages through an IL-10 feed-forward loop, which induces a localized immunosuppressive state in specific cellular neighbourhoods with HIV-infected cells. We also uncovered intracellular and extracellular features that influenced viral transcription states, including the expression of CD56, and numbers of regulatory T cells, neutrophils and CD8+ T cells2.