To genetically manipulate KCs efficiently we developed an in vivo CRISPR pipeline. We genetically modify hematopoietic stem cells and use these cells to reconstitute the bone-marrow of mice (liver protected from irradiation). After depletion of the KCs, the genetically modified bone marrow-derived monocytes engraft in the liver, colonize the KC niche and differentiate into CRISPR-edited KCs. As proof of concept, we have reconstituted Clec4f-DTR mice with stem cells in which we used CRISPR to knock-down F4/80 or CD64, two macrophage surface markers expressed by KCs. We mixed these cells with 98% of wild-type stem cells. Confocal microscopy showed the presence of small islands of F4/80-deficient KCs and CD64-deficient KCs in these mice.

By combining the CRISPR technology with single-cell sequencing (PERTURB-seq), we are able to study the role of dozens of genes in the regulation of KC differentiation and activation in parallel. Since we can isolate one million KCs per liver this means we can perturb dozens of genes in parallel in one single mouse (2000 cells per guideRNA is enough to obtain an in-depth single-cell transcriptome profile). As proof-of-concept data, we studied the capacity of CRISPR-edited monocytes to differentiate into KCs. We previously reported that monocytes lacking the receptor for BMP9, called ALK1, cannot develop into KCs (Guilliams et al. Cell 2022). We however did not know at which stage of the KC development the ALK1-KO monocytes would be blocked. We therefore reconstituted KC-DTR mice with a mix of F4/80-KO, CD64-KO, ALK1-KO and control LSKs (scrambled gRNA), depleted the KCs, and performed single-cell RNA-seq on the pool of monocytes developing into KCs. We found that monocyte-derived macrophages lacking ALK1 clustered separately and lacked KC genes and could perform a trajectory analysis.

Protein staining for surface markers has the advantage to highlight the whole surface of the cell. RNA detection has the huge advantage that it does not rely on the availability of particular antibody clones and that RNA probes can be generated against virtually any RNA species identified through scRNA-seq. However, RNA imaging yields a punctate profile and it is very difficult to correctly annotate each RNA dot to a specific cell in the tissue. As such the ideal solution is co-detection of RNA molecules (versatile RNA probes allowing to identify any cell type or activation state) and surface proteins (the antibodies should stain the surface of the cells of interest, but do not have to be ultra-specific to a particular cell type or activation state). We are therefore developing novel wet-lab protocols that will allow the co-detection of surface protein markers on the MERSCOPE instrument. In parallel, we utilize the Miltenyi MACSima instrument to detect hundreds of proteins and combine it with RNAsky to detect dozens of RNA molecules.

We use VisiumHD as untargeted spatial transcriptomics platform. 

To investigate the chromatin accessibility landscape of cells we use single-cell ATAC, mini-bulk ATAC and Cut&Run pipelines for the main histone marks and for transcription factors.

(best viewed with Chrome)
We use cell-specific fluorescent reporter mice to track the in vivo behavior of our cells of interest, such as Kupffer cells or recruited monocytes.

(best viewed with Chrome)
We use light microscopy to identify our cells of interest combined with electron microscopy to gain subcellular resolution.

Using barcoded antibodies, we perform single-cell CITE-seq and detect the expression of hundreds of surface proteins together with the transcriptome of each individual cell.

We use confocal microscopy to identify our cells of interest and study tissue architecture.