Fig. 2: CD56^bright NK cells can access the tissue parenchyma in steady-state conditions.

a, Schema of experimental workflow for MISTRG humanized mouse experiments and representative flow cytometry plots of i.v. labeling. b, Representative flow cytometry plots showing CD56⁺CD16⁺ (red) and CD56⁺CD16⁻ (orange) NK cells, i.v. CD45-labeled NK cells versus CD16 expression, i.v. CD45-labeled cells versus granzyme B expression and granzyme B-expressing CD16⁺ and CD16⁻ NK cells (left to right) in PB and example tissues. c,d, Summary data for NK cell subsets from indicated tissues (c) and their expression of granzyme B (d). Each dot represents an individual mouse (n = 4 for uterus; n = 6 for PBMCs, AT, liver and kidney). The line in c indicates the median and the error bars display the IQR. The lines in d connect matched values from the same mouse. e, Summary data for frequency of extravasated cells within NK cell subsets, shown as proportion of non-i.v. CD45-labeled NK cells. Each dot represents an individual mouse (n = 4 for uterus; n = 6 for PBMCs, AT, liver and kidney). The lines connect CD16⁺ and CD16⁻ NK cell values from the same mouse. f, Sample collection scheme for afferent (portal venous) and efferent (hepatic venous and central venous) blood of the human liver and liver tissue. g,h, Summary data for NK cell subset distribution (g) and expression of CD69 within NK cell subsets (h) from indicated sources: CD56^bright (left) and CD56^dim (right) NK cells. Each dot represents one individual donor (n = 3 for portal venous blood, hepatic venous blood and central venous blood; n = 5 for liver tissue and sinusoidal blood; n = 9 for PBMCs). The line indicates the median and the bars depict the IQR. Illustrations in a from Servier Medical Art (https://smart.servier.com/); panel a and f created using BioRender.com.