Human lymph node fibroblastic reticular cells maintain heterogeneous characteristics in culture

Summary Fibroblastic reticular cells (FRCs) are mesenchymal stromal cells in human lymph nodes (LNs) playing a pivotal role in adaptive immunity. Several FRC subsets have been identified, yet it remains to be elucidated if their heterogeneity is maintained upon culture. Here, we established a protocol to preserve and culture FRCs from human LNs and characterized their phenotypic profile in fresh LN suspensions and upon culture using multispectral flow cytometry. We found nine FRC subsets in fresh human LNs, independent of donor, of which four persisted in culture throughout several passages. Interestingly, the historically FRC-defining marker podoplanin (PDPN) was not present on all FRC subsets. Therefore, we propose that CD45negCD31neg human FRCs are not restricted by PDPN expression, as we found CD90, BST1, and CD146/MCAM to be more widely expressed. Together, our data provide insight into FRC heterogeneity in human LNs, enabling further investigation into the function of individual FRC subsets.

). B) Opt-SNE per human LN donor of the nine distinct clusters of fresh LNSC subsets.

Figure S1 .
Figure S1.Cell characteristics of improved digestion protocol, related to Figure 1.A) Bar graph visualising the percentage of CD31 neg PDPN + and CD31 neg PDPN neg cells from CD45 neg cells after 3 digestion rounds of 15 minutes or after 4 digestion rounds of 10 minutes.B) Geometric mean fluorescent intensity of PDPN on FRCs before or after CD45 neg enrichment.Shapes represent four different donors.C) FRC growth from passage 0, visualised as FRC cell number per cm 2 on day 0 within the lymph node cell suspensions and after 10 or 12 days in culture.The three different symbols correspond to three different donors.D) Bar graph visualising the percentage of endothelial stromal cells from viable CD45 neg cells within the fresh lymph node cell suspensions or after culture at passage 2, 4 and 6.For panels A and D, the mean is visualised with error bars representing standard deviations.

Figure S2 .
Figure S2.Gating strategy of clusters within human LN cell suspensions and cultured FRCs, related to Figures 3 & 4. A) Gating strategy and key membrane markers of the clusters identified within CD31 neg CD45 neg cells from human LN cell suspensions, related to Figure 3. Contour plots shown from one representative donor.B) Gating strategy and key membrane markers of the clusters identified within CD31 neg CD45 neg cells from cultured FRCs, related to Figure 4. Contour plots shown from one representative donor.

Figure S3 .
Figure S3.High dimensional analysis of FRCs in fresh human LN cell suspensions, related to Figure3.A) Based on Opt-SNE visualisation, we identified nine distinct clusters of fresh LNSC subsets, gated from live, CD45 neg CD31 neg cells.Scale bar represents median correction expression value (CEV) from -0.7 to 3.9.Data represents an overlay of four independent human LN donors (TableS1).B) Opt-SNE per human LN donor of the nine distinct clusters of fresh LNSC subsets.

Figure S4 .
Figure S4.High dimensional analysis of cultured FRCs, related to Figure 4. A) Based on Opt-SNE visualisation, we identified eight distinct clusters of FRC subsets throughout culture, gated from live, CD45 neg CD31 neg cells.Scale bar represents median correction expression value (CEV) from -0.7 to 3.9.Data represents an overlay of six human LN donors (Table S1), passages 2, 4 and 6 combined.B) Opt-SNE per human LN donor of the eight distinct clusters of cultured FRCs, passages 2, 4 and 6 combined.C) Significant different clusters between donors are represented per cluster.Data is visualised as mean of three passages per donor.Two-way ANOVA with Tukey's multiple comparison test, *p < 0.05 and **p < 0.01.D) Opt-SNE per passage of the eight distinct clusters of cultured FRCs, six human LN donors combined.

Figure S5 .
Figure S5.Spectra of the fluorophores of the antibody panel, related to STAR Methods.This image shows the spectra of fluorophores of the antibody panel used in this manuscript (made with Cytek ® Spectrum Viewer).The red arrows above the graph indicate areas in the spectrum with high autofluorescence of human FRCs, and the blue arrows indicate areas in the spectrum with low autofluorescence of human FRCs [S1].When adding extra markers to the panel, we recommend to choose fluorophores emitted in channels with low autofluorescence (blue arrows).

Figure S6 .
Figure S6.Morphology of cultured human FRCs affects the positioning of events on forward (FSC) and side (SSC) scatter, related to STAR Methods.Gating strategy for selection of human LN stromal cells (gate P1) in A) linear and B) logarithmic scale.C) Red arrow shows single cells deviating from linearity.D) Images of individual human FRCs acquired on CytPix flow cytometer.The blue squares indicate the area of zoom in of single events depicted by brightfield images on the right.FSC and SSC are in log scale.

Figure S7 .
Figure S7.Visualisation of batch correction, related to STAR Methods.A) UMAP visualisation of the different batches before and after iMUBAC batch correction.B) UMAP visualisation of the batches of fresh LN cell suspensions before and after iMUBAC batch correction.C) UMAP visualisation of the batches of cultured FRCs before and after iMUBAC batch correction.