Methylene blue staining of isolated mouse IALVs

Representative image of an isolated and cleaned IALV after methylene blue staining which revealed cells of various morphology. (B) is the zoomed in image of the yellow dotted box in A which contained large ovoid cells with granular staining (B, yellow asterisks). Fine cellular extensions (red asterisks) stained by methylene blue in some cells were visualized with color channel separation and division (C). (D, E) Similar as B and C, but in a separate vessel which stained with a higher density of methylene blue stained cells some of which had limited cellular processes. F) Focal reconstruction from imaging a methylene blue stained IALV using an upright microscope and immersion objective.

Staining Mouse IALVs for ICLC Markers

Representative immunofluorescent max projections of half vessel confocal image stacks imaged from mouse IALVs stained for ICLC markers. DAPI (A), cKit (B), and CD34 (C) and their merged image (D). Representative max projections of the intermediate filament vimentin (E), the intermediate filament desmin (F), CD34 (G) and their merged image (H). Representative max projection of vimentin (I), cKit (J), CD34 (K) and their merged image (L). Scale bar = 100 µm for all images.

Immunofluorescence Labeling of Mouse IALVs with Markers for ICLC, LMC, LEC, and Immune Cell Populations

We stained isolated mouse IALVs with cellular markers used to differentiate various cell types observed in cLVs. Half vessel image stacks were taken with confocal microscopy and the resulting representative max projections are shown. (A) CD34 stained cells and LMC staining with SMA (B) and calponin (C) and the corresponding merged (D) image. There was significant overlap in (E) CD34 staining along with the fibroblast marker PDGFRα compared to LMC staining with SMA (G) and the merged (H) image. The endothelial marker CD31 (I) to delineate LECs with PDGFRα staining (J), and the LMC marker calponin (K) with the merged image (L) revealed 3 separate populations of cells. PDGFRβ (O) stained many cells that were CD34 (M) and PDGFRα (N) positive, as seen in the merge imaged (P), in addition to PDGFRβ signal detected in the LMC layer (Q). Max projections of only the luminal frames of a z-stack at lymphatic valve locations revealed PDGFRβ, CD34, and PDGFRα labeling in bipolar shaped cells with long extensions that traveled throughout the valve leaflets (V, W). d Control IALV (Y) stained only with secondary antibody. Scale bar = 100 µm for all images.

iCre-ROSA26mTmG Labelling and Fidelity to Target Putative Pacemaker Cell Populations

Stitched montages of serial max projections of GFP and tdTomato signal from live IALVs isolated from PdgfrαCre-ROSA26mTmG (A), Ng2Cre-ROSA26mTmG (B), PdgfrαCreERTM-ROSA26mTmG (C), PdgfrβCreERT2-ROSA26mTmG (D), cKitCreERT2-ROSA26mTmG (E), and Myh11CreERT2-ROSA26mTmG (F). IALVs were digested into single cells and GFP+ cells were purified via FACS from Prox1-eGFP (G), Myh11CreERT2-ROSA26mTmG (H), PdgfrαCreERTM-ROSA26mTmG (I), and PdgfrβCreERT2-ROSA26mTmG (J) mice. Representative gels demonstrating RT-PCR products corresponding to the respective genes used in the promoter of each specific transgene employed to drive either eGFP or Cre mediated recombination of ROSA26mTmG from each GFP+ sorted population (K-N) to assess fidelity. Images are representative of IALVs from at least 3 separate mice. FACs and RT-PCR was repeated at least 3 times for each mouse.

scRNAseq analysis of mouse IALVs from ROSA26mTmG mice.

IALVs were cleaned and isolated from 8 ROSA26mTmG mice and digested into a single cell suspension for scRNAseq analysis with the 10X platform. A) UMAP of the various cell populations that compromise the mouse IALV though some mammary epithelia contamination was present (populations 18,19). B) Heat map of commonly used genes for cell identification for each of the cell clusters. Dot plots to assess cell cluster expression of the genes shown in Figure 4 using a dot plot for the LEC markers Prox1 (C) and Flt4 (D, VEGFR3), LMC markers Myh11 (E) and caponin1 (F, Cnn1), fibroblast markers Pdgfrα (G) and Lum (H, Lumican), ICC marker Kit (I), the pericyte and smooth muscle precursor marker (Pdgfrβ), and the hematopoietic marker Ptprc (K, CD45).

RT-PCR Profiling of FACs Purified Cells from iCre-ROSA26mTmG

Expanded RT-PCR profiling of genes to discriminate LECs, LMCs, and other cell types in our GFP+ sorted cells from Prox1-eGFP (A), Myh11CreERT2-ROSA26mTmG (B), PdgfrβCreERT2-ROSA26mTmG (C), and PdgfrαCreERTM-ROSA26mTmG (D). Dot plots for the genes assessed in A-D in our IALV scRNAseq analysis confirmed those results. In addition to a population of AdvCs expressing Cacna1c, we also noted expression of Cx45 (N) which was also observed in LECs) and Ano1 (O) in the AdvC clusters. We confirmed this expression using GFP+ cells sorted from PdgfrαCreERTM-ROSA26mTmG IALVs for RT-PCR (P) and ruled out hematopoietic or LEC contamination. All RT-PCRs were performed 2-4 times for each gene over each sorted cell population collected from different mice.

Isobaric contractile Assessment of popliteal cLV from PdgfrαCreERTM driven deletion of Ano1, CX45, and CaV1.2

Summary of the contractile parameters recorded from popliteal cLVs in PdgfrαCreERTM-Ano1fl/fl, PdgfrαCreERTM-Cx45fl/fl mice, PdgfrαCreERTM-Cav1.2fl/fl mice. Contraction frequency (A, D, G), ejection fraction (B, E, H), and vessel tone (C, F, I) were assessed. No statically significant differences observed in cLVs isolated from PdgfrαCreERTM-Ano1fl/fl and PdgfrαCreERTM-Cx45fl/fl mice across these three parameters. Mean and SEM shown, n=6 popliteal vessels from 3 mice PdgfrαCreERTM-Ano1fl/fl mice and n=10 popliteal vessels from 6 mice Ano1fl/fl mice. Mean and SEM shown, n=5 popliteal vessels from 3 mice PdgfrαCreERTM-CX45fl/fl mice and n=8 popliteal vessels from 11 mice CX45fl/fl mice. Mean and SEM shown, n=6 popliteal vessels from 3 mice PdgfrαCreERTM-Cav1.2fl/fl mice and n=9 popliteal vessels from 20 mice Cav1.2fl/fl mice. The contractile data from control Cav1.2fl/fl vessels was previously published but was separated by sex (Davis et al., 2022) while they are combined here. * Denotes significance at p <0.05 which 0.10 > p >0.05 are reported as text. Normalized contraction amplitude, fractional pump flow, end diastolic diameter can be found in Supp Figure 8.

ChR2-Mediated Depolarization Only in LMCs Triggers Contraction

Representative max projections of tdTomato-ChR2 signal in popliteal cLVs isolated from cKitCreERT2-ChR2-tdTomato (A), PdgfrαCreERTM-ChR2-tdTomato (C), and Myh11CreERT2-ChR2-tdTomato (E) with their corresponding brightfield image (B, D, F) respectively. Time-lapse brightfield images every 0.5 s starting at stimulation t=0 for cKitCreERT2-ChR2-tdTomato (G-J), PdgfrαCreERTM-ChR2-tdTomato (K-N), and Myh11CreERT2-ChR2-tdTomato (O-R). The I bar denotes the inner diameter at t=0 over time and white asterisks denote the contraction. Representative diameter trace for the popliteal cLV demonstrate spontaneous contractions with the dotted boxes indicating the optical stimulation event in the respective brightfield images of the time lapse images. Isolated cLVs from cKitCreERT2-ChR2-tdTomato (S), PdgfrαCreERTM-ChR2-tdTomato (T), and Myh11CreERT2-ChR2-tdTomato (U) were stimulated with light pulses (red dashed lines) and the summation of contraction triggering for each genotype (V). Mean and SEM are shown, **** denotes p<0.0001. Contraction recorded from at least 6 popliteal cLVs from 3 mice per genotype.

cKitCreERT2 Drives GCaMP6f Expression Primarily in Mast Cells in Mouse IALVs

Representative max projection of GCaMP6f signal over time in an IALV isolated from a cKitCreERT2-GCaMP6f mouse with ROI indicated around individual cells, primarily large ovoid cells, but also including a circumferential LMC (Cell10) and a horizontal LEC (Cell 11). Of cells 1-9, only cell 7 had any Ca2+ activity (red arrows) during the recording time as indicated by the STMs from each ROI (B) and their normalized F/F0 plots in (C). In contrast, the LMC in ROI 10 had both rhythmic global Ca2+ events (D) that spanned the cell axis (vertical axis) in the STM (E) in addition to localized Ca2+ events intervening the time between global events (green arrows). Representative max projection of GCaMP6f signal over time after stimulation with C48-80 (F) with many large ovoid cells displaying long lasting global Ca2+ events (G, H) while not immediately affecting the LMC Ca2+ dynamics (I, J).

Lack of coordinated Ca2+ Activity Across Contraction Cycle in PDGFRα Cells

Representative max projections of GCaMP6f signal over time in an IALVs isolated from PdgfrαCreERTM-GCaMP6f mice (A, D). ROIs were made around cells and GCaMP6f recorded over time to generate the corresponding STMs (B, E) for each cell and plots (C, F) respectively. Once again, incidental recombination occurred in a LMC which displayed rhythmic Ca2+ flashes (C) while the slight undulation in the other cells is due to movement artifact (B). Red arrows indicate the limited local Ca2+ activity observed in two cells from a PdgfrαCreERTM-GCaMP6f IALV.

Heterogeneous Diastolic Ca2+ Transient Activity in LMCs

Representative max projections of GCaMP6f signal over time in an IALVs isolated from Myh11CreERT2-GCaMP6f mice (A). LMCs were outlined with ROIs to assess GCaMp6F signal over time. Rhythmic global flashes (B) were entrained across all the LMCs in the FOV (C) with many cells exhibiting diastolic Ca2+ release events. Cells exhibiting at least one diastolic Ca2+ event, within the context of our focal plane constraints, over the recorded time were denoted by the red asterisks. The plot in (D) magnifies the first diastolic period, seconds 1-3 of C to assist in visualizing the lack of coordination of the diastolic events. (D) Max projection of the pseudo-linescan analysis across the axis of the vessel to highlight diastolic Ca2+ transients in all cells in the field of view and their lack of coordination across the cells (x-axis). The white dotted box shows the first diastolic period plotted in (D).

Pressure Dependency of Mouse LMC Diastolic Ca2+ Transients

Representative max projection of GCaMP6f signal over 20 s in an IALVs isolated from Myh11CreERT2-GCaMP6f mice in the presence of the L-type blocker nifedipine (1μM) (A) pressurized to 0.5 cmH2O, 2 cmH2O, 5 cmH2O. The local diastolic Ca2+ transients persist in the presence of nifedipine and increase with increasing pressure as demonstrated in the whole vessel STMs (B). Particle occurrence maps highlight the Ca2+ activity in each LMC as pressure is raised (C). Representative particle analysis plots for particle area (D) and particle counts/frame at each pressure (E). Summary files for particle area (F) and count /frame (G0. * Denotes p<0.05, Mean and SEM shown with n=12 separate IALVs from 8 MYH11-CreERT2-GCaMP6f

Pressure Dependent Diastolic Depolarization in LMCs

Intracellular recordings of LMC action potentials (AP) were confirmed by loading (greater than 10minutes) the impaling electrode with 1M KCl 100ug/ml AF488-Biocytin while recording APs followed by imaging on a spinning disk confocal microscope. 3D reconstruction of the z-stack confirmed the circumferential pattern of the impaled LMC that was strongly labeled by AF488-Biocytin (A, B), which also labeled neighboring LMCs, likely through gap junctions as AF488-Biocytin is <1kDa. In a separate set of experiments APs were recorded at 3 different pressures, 0.5 cmH2O, 2 cmH2O, and 5cmH2O. We plotted the representative recordings from 1 cell at each pressure (C). AP frequency was significantly increased with pressure (D) as was the diastolic depolarization rate. Plotting the AP frequency and diastolic depolarization rate from all recordings at each pressure (F) highlights the significant effect diastolic depolarization rate has on the AP frequency. Minimum membrane potential (G), threshold membrane potential of AP initiation (H), upstroke constant (I), peak membrane potential (J), plateau membrane potential (K), and time over threshold (L) are also reported, although not significant.

Primer list for RT-PCR

Colocalization of CD34 and PDGFRα

Representative max projections and their corresponding threshold adjusted image for colocalization analysis for PDGFRα (A), CD34 (B), and their colocalized signal (C) and for comparison we tested Myh11 (D) and PDGFRα (E) colocalization (F) using the FIJI BIOP-JACoP colocalization plugin on the z-stacks acquired by confocal microscopy. Pearson’s coefficient (G) and Mander’s coefficients (H) were calculated from n=3 separate stained IALVS, each from a separate mouse for CD34 and PDGFRα and n=4 for Myh11 and PDGFRα. Magnification for A-C 40X and 25x for D-F. Significant differences in colocalization below 0.05 are signified by the overhead lines.

PDGFRα+ Cells Reside Primarily in the Mouse Lymphatic Collecting Vessel Adventitia and Some in the Subendothelial Space

Max projection of confocal imaging of an IALV stained for LECs with CD31 (A), LMCs with MYH11(B), and for PDGFRα (C) with the corresponding merge file (D). Orthogonal views of the z-stack with (E) showing a single slice in the z stack and E’ and E” the orthogonal views. White dotted boxes outline locations where PDGFRα signal is observed between LMC and LEC layers. Scale bar is 100 µm in (D) and 50 µm in (E).

scRNASeq Analysis of the mouse IALV cell populations.

Heatmap of top 4-5 differentially expressed genes, based on p value, for each major cell cluster identified. LECs (Clusters 0,1,2, 11), LMCs (Cluster 5,6), and IALV adventitial cells (AdvC, 3,7,8,9,10,13) were comprised of multiple clusters. B) Bubble plot of common identification genes reveal that the previous reported LMC transcriptome markers Dpt, Pi16, and Ackr3 are specific for a sub population of the Adv and not LMCs.

Subclusters of IALV LECs revealed by scRNAseq.

The LECs were further sub-clustered to reveal 10 putative LEC subclusters (0-9) as shown in the UMAP (A) and the top differentially expressed genes amongst those sub-clusters are provided in the adjacent heatmap (B). (C) Bubble plot showing sub-cluster 8 was significantly enriched for previously documented lymphatic endothelial cell up valve genes including Itga9, Cldn11, and Neo1 and Cluster 6 had down valve gene signature including Clu and Adm. The top 30 differentially expressed genes in cluster 8, both positive and negative fold change regulated, are labeled in the volcano plot(D).

Subclusters of IALV LMCs revealed by scRNAseq.

The LMCs could be subclustered into 4 putative subclusters (0-3) as shown in the UMAP (A). We profiled these subclusters based on their expression of the typical smooth muscle markers (B), SR associated genes (C), voltage gated Ca2+ channels, (D) Voltage gated Na+ channels and Na+ transporters implicated in lymphatic pacemaking (E), voltage gated K+ channels (F), Ca2+ activated K+ channels (G), inward rectifying K+ channels and two-pore domain K+ channels (H), and Cl- channels (I).

Subclusters of IALV AdvCs revealed by scRNAseq.

AdvCs also could be further subclustered into multiple populations as shown in the UMAP (A). Bubble plot of genes used as Cre drivers and genes associated with pacemaking revealed subcluster 10 had expression of Ano1, Cx45, and Cacna1c (CaV1.2) but with only minimal evidence of LMC contamination as indicated by muscle signature genes Myh11, Kcnma1, and Tagln. C) Heatmap of the top differentially expressed genes among each of the subclusters. We assessed co-expression of Pdgfrα with CD34 (D) to confirm our immunofluorescence imaging (Sugg Figure 1), and assessed the co-expression of Pdgfrα with the pericyte markers Pdgfrβ (E) and Cspg4 (F). We further assessed co-expression of Pdgfrα the genes linked with contractile dysfunction Ano1 (G), Gcj1 (H), and Cacna1c (I) to ensure PdgfrαCreERTM would target the AdvCs expressing these genes. The cyan colored slice of the pie chart indicates the minor population of cells expressing these genes that did not express Pdgfrα.

Immune cell populations associated with the mouse IALV.

Lymphatic vessels are host to numerous immune cell populations, including monocyte, macrophage, and dendritic cell populations are revealed by immunofluorescent staining for eGFP in the “Macgreen” (Csf1r-eGFP) reporter mice (A). Staining for Pdgfrα (B) demonstrates that AdvCs are distinct from the GFP+ cells nor do they stain for the hematopoietic marker Ptprc (CD45) (C, D). Bubble plot of our scRNASeq analysis of IALVs revealed macrophages (cluster 4), moDCs (cluster14) and cDC1 cells (17) based off identifying gene markers (B). C) Bubble plot of T-cell markers revealed multiple populations of T cells including naive double negative T-cells (Yang et al., 2021) and naive CD4+ and CD8+ T-cells. A bubble plot for B-cell markers showed that cluster 15 had an expression profile for immature and mature B2 B-cells (D)(Luo et al., 2022b).

Contractile indices from isobaric myography on cLVs from PdgfrαCreERTM driven deletion of Ano1, CX45, and CaV1.2

Summary of the contractile parameters recorded from popliteal cLVs in PdgfrαCreERTM-Ano1fl/fl, PdgfrαCreERTM-Cx45fl/fl mice, PdgfrαCreERTM-Cav1.2fl/fl mice. No differences in normalized contraction amplitude (A, D, G), fractional pump flow (B, E, H), or end diastolic diameter (C, F, I) were observed. The contractile data from control Cav1.2fl/fl vessels was previously published but was separated by sex (Davis et al., 2022) while they are combined here.

PDGFRα AdvCs Include Multipotent Cell

Representative RT-PCR results profiling purified GFP+ cells purified from IALVs isolated from PdgfrαCreERTM-ROSA26mTmG via FACS. PDGFRα cells expressed the multipotent markers Klf4, Sca1, Gli1, CD29, CD105, and CD44 (A) with total brain cDNA serving as a positive control (B). Representative RT-PCR results showing lack of expression of some of these markers in the GFP+ cells purified from Myh11CreERT2--ROSA26mTmG (C) or Prox1-eGFP mice, in contrast to the RFP+ population from Myh11CreERT2--ROSA26mTmG mice (D). RT-PCRs were repeated at least 2 times from separate purified cells populations from different mice. Dot plots of only the AdvCs cluster highlights populations of cells that express genes associated with multipotency such as Ly6a (E), Klf4 (F), Gli1 (G), Itgb1 (H, CD29), Eng (I, CD105), CD44 (J). Expression of protein for Ly6a was confirmed with immunofluorescence. Representative max projections of IALVs stained for Sca1 (K), PDGFRα (L), Myh11 (M) and the corresponding merged file (N). Scale bar is 100 µm.