Application of lineage tracing to signal cell sequencing (scSeq) to definitively discriminate retinal microglia from monocyte-derived macrophages (MDM)

Abstract

Purpose : scSeq is a powerful tool for the identification of immune cell populations and transcriptome alterations in them. Definitive identification of retinal microglia and infiltrated MDM in retinal disease models is problematic due to similarity of marker gene expression, phenotypic transformation of microglia, and the possibility that MDM adopt a microglia-like phenotype. Lineage tracing allows identification of microglia by transgenic protein expression but is rarely applied to scSeq. We hypothesized that lineage tracing and scSeq could be combined in order to help identify and compare microglial and MDM cell populations in a mouse model of retinal detachment (RD) injury.

Methods : Microglia were lineage traced utilizing a yellow fluorescent protein (EYFP) and tamoxifen (Tam)-inducible Cre (CreERT2) bicistronic transgene (EYFP-CreERT2) inserted into the Cx3cr1 allele, along with a floxed-STOP tdTomato (tdTom) transgene. A 4-week washout period after Tam treatment eliminated tdTom⁺ monocytes from circulation. Partial RD was created by subretinal injection of 1% hylaluronic acid. For scSeq, n=16 retinas were pooled at 7 days after injury (RD7), and CD45⁺ and CD11b⁺ immune cells were enriched by fluorescence activated cell sorting. Pooled naïve (n=12) and untreated fellow eyes (n=16) were controls. scSeq cell cluster differential gene expression was analyzed using Cellenics software. Retinal and subretinal cells were examined by immunofluorescence (IF) for EYFP, tdTom and endogenous proteins.

Results : Seventeen cell clusters were obtained. Sequence reads were successfully mapped to EYFP-CreERT2 and tdTom transgenes. Surprisingly, poor linear correlations were obtained for EYFP-CreERT2 and Cx3cr1 mRNA mean cluster expression levels (MCEL, R²=0.23) and for fraction of cells positive for expression in clusters (FCPE, R²=0.086). In contrast, good correlations were obtained for tdTom and Cx3cr1 MCEL (R²=0.73) and FCPE (R²=0.86), as well for tdTom with several other microglial marker mRNAs. Transgene mapping facilitated identification of subretinal tdTom⁺ microglia and EYFP⁺/tdTom^neg MDM clusters in RD. IF validated several DEG as protein markers for these subretinal populations.

Conclusions : Transgenic lineage tracing can add specificity to scSeq cluster cell type identification, aiding in the differentiation of cells of microglia and monocyte origins.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.