Hyperactive KRAS/MAPK signaling disrupts normal lymphatic vessel architecture and function

Complex lymphatic anomalies (CLAs) are sporadically occurring diseases caused by the maldevelopment of lymphatic vessels. We and others recently reported that somatic activating mutations in KRAS can cause CLAs. However, the mechanisms by which activating KRAS mutations cause CLAs are poorly understood. Here, we show that KRASG12D expression in lymphatic endothelial cells (LECs) during embryonic development impairs the formation of lymphovenous valves and causes the enlargement of lymphatic vessels. We demonstrate that KRASG12D expression in primary human LECs induces cell spindling, proliferation, and migration. It also increases AKT and ERK1/2 phosphorylation and decreases the expression of genes that regulate the maturation of lymphatic vessels. We show that MEK1/2 inhibition with the FDA-approved drug trametinib suppresses KRASG12D-induced morphological changes, proliferation, and migration. Trametinib also decreases ERK1/2 phosphorylation and increases the expression of genes that regulate the maturation of lymphatic vessels. We also show that trametinib and Cre-mediated expression of a dominant-negative form of MEK1 (Map2k1 K97M ) suppresses KRASG12D-induced lymphatic vessel hyperplasia in embryos. Last, we demonstrate that conditional knockout of wild-type Kras in LECs does not affect the formation or function of lymphatic vessels. Together, our data indicate that KRAS/MAPK signaling must be tightly regulated during embryonic development for the proper development of lymphatic vessels and further support the testing of MEK1/2 inhibitors for treating CLAs.

Next-generation sequencing studies have begun to shed light on the genetic basis of many CLAs.Somatic activating mutations in PIK3CA, ARAF, BRAF, CBL, HRAS, and NRAS have been found in sporadic CLAs (Manevitz-Mendelson et al., 2018;Barclay et al., 2019;Li et al., 2019;Ozeki et al., 2019;Rodriguez-Laguna et al., 2019;Foster et al., 2020;Grenier et al., 2023;Li et al., 2023).We and other groups recently identified somatic activating mutations in KRAS in CLA patients (Li et al., 2019;Nozawa et al., 2020;Homayun-Sepehr et al., 2021;Andreoti et al., 2023;Li et al., 2023;Sheppard et al., 2023).To date, five different activating KRAS mutations (p.G12V, p.G12D, p.G13D, p.Q61R, and p.A146T) have been identified in patients (Li et al., 2019;Nozawa et al., 2020;Homayun-Sepehr et al., 2021;Li et al., 2023;Sheppard et al., 2023).KRAS is a small GTPase that switches between inactive GDP-bound and active GTP-bound forms (Haigis, 2017).The KRAS mutations in CLA patients impair GTP hydrolysis, resulting in hyperactive downstream signaling (Haigis, 2017).To investigate the effect of hyperactive KRAS signaling on lymphatic vessels, we used the Cre-loxP system to conditionally express an active form of KRAS (Kras G12D ) in lymphatic endothelial cells (LECs) during postnatal development (Homayun-Sepehr et al., 2021).We found that Kras G12D -mutant mice developed intraosseous lymphatic vessels, a feature frequently seen in CLAs (Homayun-Sepehr et al., 2021).We also showed that the development of lymphatic valves and recruitment of lymphatic muscle cells were impaired in Kras G12D -mutant mice (Homayun-Sepehr et al., 2021).Additionally, we found that MEK1/2 inhibition with trametinib suppressed the loss of lymphatic valves in newborn Kras G12Dmutant mice (Homayun-Sepehr et al., 2021).Despite these recent advances, the mechanisms by which activating KRAS mutations cause CLAs remain incompletely understood.In the present study, we delineate the effect of hyperactive KRAS signaling on the development of the lymphatic system in mouse embryos.We also molecularly profile primary human LECs expressing an active form of KRAS and assess the effect of trametinib on KRAS mutant cells and embryos.Finally, we characterize Kras conditional knockout mice to determine the role of wild-type Kras in the formation and function of lymphatic vessels.

LEC KrasG12D embryos exhibit multiple lymphatic vessel defects
To characterize the effect of KRAS G12D on lymphatic vessels during embryonic development, we bred Lyve1-Cre mice with LSL-Kras G12D mice to create control (LEC Ctrl ) and mutant (LEC KrasG12D ) embryos (Figure 1A).Lyve1-Cre mice exhibit Cre activity in LECs, macrophages, and blood endothelial cells in the yolk sac, liver, and lung and have been used by many groups to induce recombination in LECs (Pham et al., 2010;Dellinger et al., 2013;Cha et al., 2016;Cha et al., 2020;Geng et al., 2020).LSL-Kras G12D mice express oncogenic Kras (KRAS G12D ) from the endogenous KRAS locus following Cremediated recombination (Jackson et al., 2001).LEC KrasG12D mice exhibited profound edema at embryonic day 14.5 (E14.5).Edema was displayed by 0/16 LEC Ctrl embryos and 17/17 LEC KrasG12D embryos (p < 0.0001; Fisher's exact test).Morphometric analysis of E14.5 embryos revealed that the jugular lymph sacs were significantly larger in LEC KrasG12D embryos than LEC Ctrl embryos (Figures 1C, D).Embryos with edema and enlarged jugular lymph sacs exhibit lymphovenous valve defects (Geng et al., 2016).Lymphovenous valves are located at the junction between the jugular lymph sacs and subclavian veins and prevent blood from entering lymphatic vessels (Geng et al., 2016).To determine whether LEC KrasG12D embryos display lymphovenous valve defects, we analyzed hematoxylin and eosin (H&E)-stained coronal sections of E14.5 embryos.LEC Ctrl embryos had lymphovenous valve leaflets that extended from the jugular lymph sacs into the veins (Figure 1E).In contrast, LEC KrasG12D embryos lacked well-defined valve leaflets at the lymphovenous junction (Figure 1E).Lymphovenous valves were observed in 6/6 LEC Ctrl embryos and 0/5 LEC KrasG12D embryos (p < 0.01 Fisher's exact test).Immunofluorescence staining of coronal sections of E14.5 embryos for Lyve1 and endomucin confirmed that LEC Ctrl embryos, but not LEC KrasG12D embryos, had well-defined lymphovenous valves (Figure 1F).Instead of lymphovenous valves, LEC KrasG12D embryos frequently had cell clusters at lymphovenous junctions (Figure 1F).Cell clusters were also present in other regions in the jugular lymph sacs of LEC KrasG12D embryos.Cell clusters were observed in the jugular lymph sacs of 0/6 LEC Ctrl embryos and 8/10 LEC KrasG12D embryos (p < 0.01 Fisher's exact test).Immunofluorescence staining revealed that the clusters contained numerous Lyve1 and Prox1 double-positive LECs (Figures 1G, H).It was recently reported that mouse embryos harboring mutations in an enhancer element located 11 kb upstream of Prox1 (Prox1-11 kb Δ/Δ ) exhibit similar cell clusters in their jugular lymph sacs (Kazenwadel et al., 2023).Immunofluorescence staining with a panel of markers revealed that the clusters in Prox1-11 kb Δ/Δ embryos contained Prox1-positive cells and cells expressing various hematopoietic markers (Kazenwadel et al., 2023).To further characterize the cellular landscape of clusters in the jugular lymph sacs of LEC KrasG12D embryos, we stained sections of E14.5 embryos with a panel of antibodies.This revealed that clusters in LEC KrasG12D embryos contained heterogeneous populations of CD45, F4/80, and RUNX1positive cells (Figures I-O).Therefore, the cell clusters in the jugular lymph sacs of LEC KrasG12D embryos include LECs and various hematopoietic cells.We next characterized the effect of KRAS G12D expression on the patterning of lymphatic vessels.Whole-mount immunofluorescence staining of E15.5 back skin for neuropilin-2 revealed that the dermal lymphatic network in LEC KrasG12D embryos was irregular and less complex than the network in LEC Ctrl embryos (Figures 2A, B).LEC KrasG12D embryos had significantly fewer lymphatic vessel branch points than LEC Ctrl embryos (Figure 2C).Additionally, the diameter of lymphatic vessels was significantly greater in LEC KrasG12D embryos than in LEC Ctrl embryos (Figure 2D).

Hyperactive KRAS signaling in LECs induces changes in cell shape, proliferation, and migration
To investigate the molecular mechanisms by which hyperactive KRAS signaling causes the maldevelopment of lymphatic vessels, we used lentiviral vectors to transiently express GFP (vector control) or KRAS G12D in primary human LECs.We based our approach on previously published studies that modeled KRAS-induced FIGURE 4 KRAS G12D expression in LECs induces PI3K and MAPK signaling and changes in gene expression.(A).Western blot results for phospho-AKT, AKT, phospho-ERK1/2, ERK1/2, GFP, KRAS G12D (mutation-specific antibody), and tubulin.Protein lysates were made 72 h after treating primary human LECs with GFP or KRAS G12D expressing lentivirus particles.(B).Graphs of Western blot results.Phospho-AKT and phospho-ERK1/2 levels were significantly higher in KRAS G12D -LECs than GFP-LECs.GFP-LECs specifically expressed GFP, and KRAS G12D -LECs specifically expressed KRAS G12D .(C).Volcano plot of RNA-Seq results comparing KRAS G12D -LECs to GFP-LECs.Two thousand and nine genes were upregulated (red dots), and 1,396 genes were downregulated (blue dots) by KRAS G12D (log 2 fold-change ≥1 or ≤ −1; FDR <0.01).D,E.Select GO terms associated with genes upregulated by KRAS G12D (D) and genes downregulated by KRAS G12D (E).(F,G).Heatmaps of genes that regulate angio/lymphangiogenesis (F) and genes that control lymphatic valve development or lymphatic muscle cell recruitment (G).Data are presented as mean ± SEM. * p < 0.05, ** p < 0.01; unpaired Student's t-tests.
arteriovenous malformations (AVMs) by treating cells with a control vector or a vector expressing mutant KRAS (Nikolaev et al., 2018;Fish et al., 2020).We found that KRAS WT -lentivirus had a similar effect on cell signaling in LECs across a range of MOIs as GFP-lentivirus (Supplementary Figure S1).Therefore, we believe that the GFP-lentivirus is a suitable control for our experiments.We found that GFP-LECs retained a standard cobblestone-like shape, whereas KRAS G12D -LECs exhibited an elongated/spindled shape (Figure 3A).Circularity index measurements revealed that KRAS G12D -LECs were significantly less circular than GFP-LECs (Figure 3B).We next assessed the effect of KRAS G12D on cellular processes associated with the growth of lymphatic vessels.MTS assays demonstrated that KRAS G12D stimulated LEC proliferation, and scratch assays showed that it promoted LEC migration (Figures 3C-E).These data indicate that hyperactive KRAS signaling in LECs promotes cellular behaviors associated with the growth of lymphatic vessels.

KRAS G12D induces PI3K/AKT and MAPK signaling and changes in gene expression in LECs
KRAS G12D influences cell survival, proliferation, and migration by activating a diverse set of downstream signal transduction pathways.Two prominent pathways activated by KRAS G12D are the phosphatidylinositol 3-kinase (PI3K) and mitogen-activated protein kinase (MAPK) pathways.To determine whether KRAS G12D activated these pathways in LECs, we assessed the phosphorylation of AKT and ERK1/2 in GFP-LECs and KRAS G12D -LECs by immunoblotting.This revealed that the levels of phospho-AKT and phopsho-ERK1/2 were significantly higher in KRAS G12D -LECs than GFP-LECs (Figures 4A, B).Immunoblotting also demonstrated that GFP-LECs specifically expressed GFP and that KRAS G12D -LECs specifically expressed KRAS G12D (Figures 4A, B).
To determine the impact of KRAS G12D -induced alterations in cell signaling on gene expression, we analyzed the transcriptomes of GFP-LECs and KRAS G12D -LECs by bulk RNA-sequencing.We found that 2,009 genes were upregulated, and 1,396 genes were downregulated in KRAS G12D -LECs compared to GFP-LECs (Figure 4C).Analysis of gene ontology (GO) terms showed that genes associated with angiogenesis, cell proliferation, cell migration, metalloendopeptidase activity, and epithelial-to-mesenchymal transition were upregulated by KRAS G12D (Figure 4D).The RAS/ MAPK pathway is regulated by a series of negative feedback mechanisms (Lake et al., 2016).These negative feedback mechanisms include rapid responses mediated by protein kinases and transcriptionally induced mechanisms.Active RAS/MAPK signaling in cancer cells induces the expression of negative regulators of the RAS/MAPK pathway, such as MAPK phosphatases (e.g., DUSP4) and Sprouty (Spry) proteins (Lake et al., 2016).We found that KRAS G12D expression in LECs also induces the expression of negative regulators of the RAS/MAPK pathway (Figure 4D).GO term analysis also demonstrated that genes associated with cell adhesion, adherens junctions, and positive regulation of MAPK activity were downregulated by KRAS G12D (Figure 4E).We then looked closer at the expression of genes that promote angio/lymphangiogenesis and genes that regulate the maturation of lymphatic vessels (valve development and recruitment of lymphatic muscle cells).We found that the expression of several angio/lymphangiogenic genes was increased, whereas the expression of lymphatic maturation genes was decreased in KRAS G12D -LECs compared to GFP-LECs (Figures 4F, G).
The growth factor VEGF-C stimulates lymphangiogenesis by activating the receptor tyrosine kinase VEGFR3.VEGF-C/ VEGFR3 signaling was recently shown to promote the progression of PIK3CA-induced lymphatic malformations (Martinez-Corral et al., 2020).Because VEGF-C and VEGFR3 transcripts were elevated in KRAS G12D -LECs, we further characterized VEGF-C/VEGFR3 signaling in KRAS G12D -LECs.We found by ELISA that KRAS G12D -LECs expressed and secreted more VEGF-C than GFP-LECs (Supplementary Figure S2).We also found that phospho-VEGFR3 levels were greater in KRAS G12D -LECs than in GFP-LECs (Supplementary Figure S2).Although phospho-VEGFR3 levels were high in KRAS G12D -LECs, they could be driven even higher by the addition of recombinant VEGF-C.Additionally, phospho-AKT, but not phospho-ERK1/2 levels, increased in KRAS G12D -LECs following stimulation with recombinant VEGF-C (Supplementary Figure S2).These data suggest that KRAS G12D -LECs, like PIK3CA H1047R -LECs (Martinez-Corral et al., 2020), are responsive to exogenous VEGF-C.
Changes in global gene expression are mediated, in part, by the activity of transcription factors.We analyzed our bulk transcriptome data using Metascape (Zhou et al., 2019) to identify transcription regulatory networks altered by hyperactive KRAS signaling.We found that genes regulated by the transcription factors JUN, FOS, STAT3, ETS1, and ETS2 were upregulated by KRAS G12D , whereas genes regulated by TP53, E2F1, and HIF1α were downregulated by KRAS G12D (Supplementary Figure S3).
To characterize the effect of trametinib on cell signaling, we treated GFP-LECs and KRAS G12D -LECs with DMSO or trametinib, and then assessed the phosphorylation of AKT and ERK1/2 by immunoblotting.We found that trametinib significantly decreased ERK1/2 phosphorylation in GFP-LECs and KRAS G12D -LECs (Figures 6A, B).Interestingly, trametinib modestly increased AKT phosphorylation in GFP-LECs and significantly increased AKT phosphorylation in KRAS G12D -LECs (Figures 6A, B).Trametinib and other MEK1/2 inhibitors have also been shown to induce AKT phosphorylation in a variety of cancer cell lines (Chen et al., 2017;Tsubaki et al., 2019).Although trametinib affected ERK1/2 and AKT phosphorylation, it did not affect the expression of GFP in GFP-LECs or KRAS G12D in KRAS G12D -LECs (Figures 6A, B).
To determine the impact of trametinib on gene expression, we analyzed the transcriptomes of DMSO and trametinib-treated KRAS G12D -LECs by bulk RNA sequencing.We found that 841 genes were upregulated, and 424 genes were downregulated in trametinib-treated KRAS G12D -LECs compared to DMSO-treated KRAS G12D -LECs (Figure 6C).GO term analysis showed that genes associated with TNF signaling, cell adhesion, integrin binding, and cytokine-cytokine receptor signaling were upregulated by trametinib (Figure 6D).Additional GO term analysis demonstrated that genes associated with angiogenesis, GTPase activity, and cell migration were downregulated by trametinib (Figure 6E).Trametinib also decreased the expression of negative regulators of the MAPK pathway (Figure 6E).We then examined the expression of genes that promote angio/lymphangiogenesis and genes that regulate the maturation of lymphatic vessels.We found that trametinib decreased the expression of several angio/lymphangiogenic genes, and increased the expression of several genes that promote the maturation of lymphatic vessels (Figures 6F, G).To determine the effect of trametinib on transcription regulatory networks in KRAS G12D -LECs, we again analyzed our bulk transcriptome data using Metascape.We found that genes regulated by the transcription factors TP53, NFKB1, and IRF-1 were upregulated by trametinib, whereas genes regulated by CREB, JUN, FOS, ETS1, and ETS2 were downregulated by trametinib (Supplementary Figure S3).

Trametinib suppresses KRAS G12D -induced lymphatic vessel hyperplasia
We previously showed that trametinib suppresses the loss of lymphatic valves in newborn mice that express Kras G12D in their LECs.We did not assess the effect of trametinib on the diameter of lymphatic vessels.Evaluating the impact of trametinib on the size of lymphatic vessels is clinically relevant because patients with RAS pathway mutations have dilated lymphatic channels (Liu et al., 2022).To assess the effect of trametinib on lymphatic hyperplasia in LEC KrasG12D embryos, we treated embryos with vehicle or trametinib from E11.5 to E13.5 and stained back skin from E14.5 embryos with an antibody against neuropilin-2 (Figures 7A-D; Supplementary Figure S4).We found that trametinib did not significantly affect the number of lymphatic branch points in LEC KrasG12D embryos (Figure 7B).However, trametinib significantly decreased the diameter of lymphatic vessels in LEC KrasG12D embryos (Figure 7C).These data suggest that pharmacologic inhibition of MAPK signaling can partially suppress Kras G12D -induced lymphatic hyperplasia in vivo.

Genetic inhibition of MAPK signaling in LECs partially suppresses KRAS G12D -induced lymphatic vessel hyperplasia
Trametinib blocks MEK1/2 activity throughout the body.To determine whether specifically blocking MAPK activation in LECs could suppress KRAS G12D -induced lymphatic vessel hyperplasia, we obtained LSL-Map2k1 K97M transgenic mice.LSL-Map2k1 K97M transgenic mice express a dominant-negative form of human MEK1 following Cre-mediated removal of an upstream transcriptional stop sequence (Figure 8A) (Kelleher et al., 2004).The K97M mutation abolishes MEK1's kinase activity but does not affect its ability to interact with ERK1 and ERK2 (Kelleher et al., 2004).To assess the effect of the Map2k1 K97M mutation on lymphatic vessel development, we collected back skin from E15.5 embryos and stained it for neuropilin-2 (Figures 8B-D; Supplementary Figure S5).We found that the number of lymphatic vessel branch points was not significantly different between LEC Ctrl and LEC Map2k1K97M embryos.The number of branch points was also not significantly different between LEC KrasG12D and LEC KrasG12D;Map2k1K97M embryos (Figure 8B).The diameter of lymphatic vessels was slightly lower in LEC Map2k1K97M embryos compared to LEC Ctrl embryos.However, this did not reach statistical significance.In contrast, lymphatic vessel diameter was significantly smaller in LEC KrasG12D;Map2k1K97M embryos compared to LEC KrasG12D embryos (Figure 8C).These results suggest that blocking MAPK activation in LECs partially suppresses KRAS G12D -induced lymphatic vessel hyperplasia.

Conditional knockout of Kras in LECs does not impair the development or function of lymphatic vessels
Through gain-of-function experiments, we found that mutant KRAS causes lymphatic vessel hyperplasia and defects in the maturation of lymphatic vessels.However, little is known about the role of wild-type KRAS in the development of the lymphatic system.This is partly because global Kras knockout mice die during embryonic development before the lymphatic network fully forms (Johnson et al., 1997;Koera et al., 1997).Therefore, we set out to conditionally knockout Kras in LECs to identify its function in the development of lymphatic vessels.To conditionally knockout Kras in LECs, we bred Lyve1-Cre mice with Kras loxp mice to generate control (LEC Ctrl ) and Lyve1-Cre;Kras loxp/loxp (LEC ΔKras ) mice (Figure 9A).We found that LEC ΔKras mice were viable and born at their expected Mendelian frequency.To assess the efficiency of Kras deletion in LEC ΔKras mice, we first isolated RNA from CD31-positive cells collected from the lungs of 3 to 5-week-old LEC Ctrl and LEC ΔKras mice.We then performed reverse transcription (RT)-PCR for Kras and Gapdh (Figure 9B).This experiment demonstrated that Kras was efficiently deleted in LEC ΔKras mice.To determine whether LEC ΔKras mice have abnormal lymphatic vessels, we stained ear skin from 3 to 5week-old mice with an antibody against Lyve1 (Figure 9C).We found that the number of lymphatic branch points and the diameter of lymphatic vessels were not significantly different between LEC Ctrl and LEC ΔKras mice (Figures 9D, E).We also stained ears for Lyve1, CD31, and VEGFR3 to identify lymphatic valves.We counted the number of CD31-positive valves in VEGFR3-positive-Lyve1-negative vessels and found that the number of lymphatic valves was not significantly different between LEC Ctrl and LEC ΔKras mice (Figures 9F,G).To assess lymphatic function in LEC ΔKras mice, we injected the hind paws and forepaws of mice with Evans blue dye (EBD).EBD was effectively transported from the hind paws to the iliac lymph nodes, then to the thoracic duct in LEC Ctrl and LEC ΔKras mice (Figure 9H).It was also transported from the forepaws to the axillary lymph nodes in mice (Figure 9H).Notably, none of the LEC ΔKras mice exhibited chylous ascites or chylothorax.Together, these data suggest that loss of wild-type Kras in LECs does not affect the formation or function of lymphatic vessels.

Discussion
KRAS G12D expression during postnatal development was previously shown to cause the enlargement of lymphatic vessels and impair the development and maintenance of lymphatic valves (Homayun-Sepehr et al., 2021).In the present study, we show that hyperactive KRAS signaling in embryos impairs the morphogenesis of lymphovenous valves and causes lymphatic vessel hyperplasia.We also show that pharmacologic or genetic inhibition of MAPK activation can partially suppress lymphatic vessel hyperplasia in KRAS G12D -mutant embryos.These data extend our previous findings and indicate that KRAS/MAPK signaling must be tightly regulated during embryonic development for the proper development of lymphatic vessels.
The identification of causal genetic mutations for CLAs is facilitating the development of in vitro and in vivo models of CLAs.Recently, transient expression of KRAS G12D or KRAS G13D in LECs was reported to induce edema and the enlargement of lymphatic vessels in zebrafish larvae (Sheppard et al., 2023).Likewise, we show that KRAS G12D expression in LECs causes edema and the enlargement of lymphatic vessels in mouse embryos.These findings highlight the similarities between the two model systems and suggest that KRAS serves a critical role in regulating the size of lymphatic vessels during embryonic development.
Lymphovenous valves are the first lymphatic valves to form, and they develop in a stepwise manner during embryonic development (Geng et al., 2016).Lymphovenous valve cells are specified at E12.0, begin to extend into veins at E12.5, and form well-defined valve leaflets by E14.5 (Geng et al., 2016).We found that E14.5 iLEC KrasG12D embryos lacked well-defined lymphovenous valve leaflets.The Ras pathway was previously shown to serve a critical role in the development of lymphovenous valves.The Rasa1 gene encodes for p120, which is a Ras GTPase-activating protein (GAP) that negatively regulates RAS signaling by promoting GTP hydrolysis by RAS proteins (Chen et al., 2020).Conditional knockout of Rasa1 in endothelial cells increases RAS/MAPK signaling and impairs the development of lymphovenous valves (Chen et al., 2020).Thus, our findings further highlight the importance of the RAS pathway in the formation of lymphovenous valves, which are essential for the normal physiology of the lymphatic system.
We found that the jugular lymph sacs of LEC KrasG12D embryos contained irregular clusters of LECs and hematopoietic cells.Similar cell clusters are present in Prox1-11 kb Δ/Δ embryos, which harbor a mutation in an enhancer upstream of Prox1 (Kazenwadel et al., 2023).The cell clusters in Prox1-11 kb Δ/Δ embryos include Runx1positive cells (Kazenwadel et al., 2023).Runx1 is expressed by hemogenic endothelium, a specialized endothelium that gives rise to hematopoietic cells (Kazenwadel et al., 2023).The discovery of Runx1-positive cells in Prox1-11 kb Δ/Δ embryos, and ex vivo colonyforming data, led investigators to propose that Prox1 represses hemogenic activity by LECs (Kazenwadel et al., 2023).Like the clusters in Prox1-11 kb Δ/Δ embryos, the clusters in LEC KrasG12D embryos had Runx1-positive cells.This result raises the possibility that hyperactive KRAS signaling also induces hemogenic activity by LECs.However, a subset of hematopoietic stem cells come from Lyve1-positive cells (Lee et al., 2016).Thus, the irregular cell clusters in LEC KrasG12D embryos could be due to hyperactive KRAS signaling in the hematopoietic lineage.Future experiments with a more LEC-restricted Cre (e.g., Prox1-CreER T2 ) are required to determine whether Kras G12D signaling in LECs or the hematopoietic lineage triggers the formation of hematopoietic cell clusters in jugular lymph sacs.
Mutant LECs have been isolated from fluid collected from patients with macrocystic lymphatic malformations and effusion fluid harvested from patients with chylothorax (Osborn et al., 2015;Glaser et al., 2018;Li et al., 2023).These data suggest that malformations in patients shed LECs.Like patients, LEC KrasG12D embryos have free-floating LECs in their macrocystic malformations.The process by which mutant LECs delaminate from vessels is not understood.We found that KRAS G12D expression in primary human LECs causes cells to adopt a mesenchymal morphology.This phenomenon has also been observed by others expressing RAS pathway mutations in endothelial cells (Nikolaev et al., 2018;Li et al., 2019;Boscolo et al., 2022;Sheppard et al., 2023).Our bulk RNA-Seq data shows that KRAS G12D -LECs express genes associated with epithelial-to-mesenchymal transition (EMT).Cells undergoing EMT lose their cell-cell contacts and gain migratory and invasive capabilities.EMT is driven by specific transcription factors, including FOXC2, Twist, Zeb1, SNAI1 (Snail), and SNAI2 (Slug).Of interest, we found that SNAI2 was approximately 20-fold higher in KRAS G12D -LECs compared to GFP-LECs.SNAI1 and ZEB1 were also elevated in KRAS G12D -LECs.These results raise the possibility that EMT could drive the delamination of mutant LECs into the lumens of lymphatic vessels and cysts.Future studies aimed at blocking SNAI2 or other EMT-inducing transcription factors could reveal the role of this process in the pathogenesis of complex lymphatic anomalies and other vascular malformations.
During embryonic and postnatal development, subsets of lymphatic vessels mature into collecting lymphatic vessels.During this maturation phase, lymphatic vessels recruit lymphatic muscle cells and acquire valves.The lymphovenous junction also matures and remodels to have valves.We previously showed that hyperactive Kras signaling in LECs affects the maturation of lymphatic vessels (Homayun-Sepehr et al., 2021).We found that Kras G12D suppressed the recruitment of lymphatic muscle cells to lymphatic vessels and the formation of lymphatic valves (Homayun-Sepehr et al., 2021).Here, we show that Kras G12D also affects the formation of lymphovenous valves.Our bulk transcriptome analysis of KRAS G12D -LECs and GFP-LECs shows that KRAS G12D decreases the expression of genes involved in the recruitment of lymphatic muscle cells and the development of lymphatic valves.Additionally, we demonstrate that trametinib increases the expression of several of these genes in KRAS G12D -LECs.These results suggest that one mechanism by which hyperactive KRAS/MAPK signaling impairs the maturation of lymphatic vessels is by suppressing the expression of genes critical to this process.However, the downstream transcription factors mediating these changes in global gene expression remain unclear.
The discovery of activating RAS/MAPK pathway mutations in CLA patients has sparked interest in repurposing MEK1/2 inhibitors for CLAs.Of the FDA-approved MEK1/2 inhibitors, trametinib has been the most widely used in CLA patients and is reported to reduce patient symptoms (Li et al., 2019;Foster et al., 2020;Chowers et al., 2022).We previously reported that trametinib suppresses the loss of lymphatic valves in Kras G12D -mutant mice during postnatal development (Homayun-Sepehr et al., 2021).In the present study, we characterized the effect of trametinib on dermal lymphatic vessels in LEC KrasG12D embryos.We focused on dermal lymphatic vessels because the diameter of these vessels was dramatically increased in LEC KrasG12D embryos, with some areas appearing as continuous lymphatic sheets.Additionally, CLA patients with RAS pathway mutations have dilated lymphatic vessels (Liu et al., 2022).We found that blocking ERK1/ 2 activation with trametinib significantly decreased the diameter of lymphatic vessels in LEC KrasG12D embryos.Similarly, suppressing ERK1/2 activation by expressing a dominant-negative form of MEK1 (Map2k1 K97M ) in LECs also decreased the diameter of lymphatic vessels in LEC KrasG12D embryos.However, lymphatic vessels in trametinib-treated LEC KrasG12D embryos and LEC KrasG12D; Map2k1K97M embryos were still larger than normal lymphatic vessels.This could be due to incomplete inhibition of the MAPK pathway by trametinib or Map2k1 K97M .Alternatively, sustained or enhanced PI3K/AKT signaling in the face of MEK1/2 inhibition could cause the enlargement of lymphatic vessels in LEC KrasG12D embryos.Future experiments conditionally knocking out Map2k1/Map2k2 (MEK1 and MEK2) in LEC KrasG12D mice or combining PI3K inhibitors (e.g., alpelisib) with trametinib could help distinguish between these possibilities.KRAS, HRAS, and NRAS are highly related small GTPases encoded by three separate genes and somatic activating mutations in these genes have been identified in complex lymphatic anomaly patients (Manevitz-Mendelson et al., 2018;Barclay et al., 2019;Ozeki et al., 2019;Nozawa et al., 2020;Homayun-Sepehr et al., 2021;Li et al., 2023;Sheppard et al., 2023).This has fueled intense research efforts to identify the mechanisms by which mutant RAS genes cause lymphatic anomalies.Despite the growing interest in the function of mutant RAS genes in LECs, relatively little is known about the role of wild-type KRAS, HRAS, and NRAS in the development of the lymphatic system.Kras knockout mice die during embryonic development, whereas Hras and Nras knockout mice are viable (Ichise et al., 2010).Therefore, to investigate the function of RAS genes in the development of the lymphatic system, various Kras, Hras, and Nras compound mutant lines of mice were generated and analyzed (Ichise et al., 2010).This revealed that mice heterozygous for Kras and lacking either Hras or Nras develop chylous ascites and lymphatic hypoplasia (Ichise et al., 2010).However, the effect of the compound mutations on collecting vessels, valves, and lymphatic function was not delineated, and the specific role Kras serves in the development of lymphatic vessels was not investigated.We found that conditional deletion of Kras in LECs did not affect the formation of lymphatic capillaries, collecting lymphatic vessels, or valves.Interestingly, our transcriptome analysis of primary human LECs shows that LECs express HRAS and NRAS.Therefore, the lack of a lymphatic phenotype in LEC ΔKras mice could be due to compensation by Hras or Nras.Together, our results suggest that wild-type Kras is dispensable for the development and function of lymphatic vessels.
In conclusion, we show that hyperactive KRAS signaling in LECs during embryonic development causes lymphatic vessel defects.We also demonstrate that inhibition of the MAPK pathway can suppress changes to cellular processes and lymphatic vessel growth induced by mutant KRAS.Together, these data further support the testing of MEK1/2 inhibitors for the treatment of complex lymphatic anomalies.

Mice and genotyping
The animal experiments described in this manuscript were carried out in accordance with an animal protocol approved by the Institutional Animal Care and Use Committee of UT Southwestern Medical Center.Mice were maintained in ventilated microisolator cages and were fed a standard diet.Mice were provided igloos and nestlets as enrichment items.Male and female mice were used in experiments.Lyve1-Cre, LSL-Kras G12D , LSL-Map2k1 K97M , and Kras loxp mice were previously described (Jackson et al., 2001;Kelleher et al., 2004;Pham et al., 2010;Damnernsawad et al., 2016).

Immunofluorescence staining of tissue sections
Slides were deparaffinized with xylene and rehydrated through a descending ethanol series.Non-specific binding was blocked by incubating slides with TBST +3% donkey serum for 30 min.Slides were then incubated overnight with primary antibodies diluted in TBST +5% BSA.Slides were washed with TBST and then incubated with fluorophore-conjugated secondary antibodies diluted in TBST +5% BSA.Following washes with TBST, coverslips were mounted with ProLong Gold + DAPI (Invitrogen, #36935).

Whole-mount immunofluorescence staining
Back skin was harvested from E14.5 and E15.5 embryos for whole-mount immunofluorescence staining.Fixed samples were washed with PBS, permeabilized with PBS +0.3% TX-100 (PBST), and then blocked overnight with PBST +3% donkey serum.Samples were then incubated overnight with primary antibody diluted in PBST.After washing samples with PBST (3 × 40 min), they were incubated overnight with a secondary antibody diluted in PBST.Following washes with PBST (3 × 40 min), samples were placed on glass slides, and coverslips were mounted with ProLong Gold (Invitrogen, #36934).

Analysis of lymphatic branch points and vessel diameters
Analysis of back skin whole-mounts: Images of back skin were captured at ×10 magnification, and the number of lymphatic branch points was manually counted.The same images were used to measure vessel diameters.We placed a 12 × 12 grid over the images and measured the diameter of vessels located on intersecting grid lines.We used NIS Elements software (version 5.30.02) to assess branch points and diameters.

Analysis of ear skin whole-mounts
To assess branch points, we manually counted the number of lymphatic branch points per 10X image.To measure vessel diameters, we placed a 12 × 12 grid over images and measured the diameter of vessels located on intersecting grid lines.To assess lymphatic valves per mm vessel, we measured the length of the lymphatic network and then manually counted the number of lymphatic valves.We used NIS Elements software (version 5.30.02) to manually count branch points and valves and to measure vessel lengths and diameters.

Tissue culture and lentivirus treatment
Primary human LECs were purchased from PromoCell (C-12216) and cultured in EGM-2MV media (LONZA CC-3125).Cells were used between passages 3 and 6.GFP-expressing, KRAS WT -expressing, and KRAS G12D -expressing lentivirus particles were purchased from Vector Builder.We expressed the 4A-splice variant of KRAS (KRAS4A).To infect cells, we treated LECs with polybrene (final concentration 8 μg/mL) and lentivirus particles (MOI = 12.5).We selected an MOI of 12.5 because we found that the infection efficiency of the CMV-GFP lentivirus was greater than 90% when we used an MOI of 12.5 (data not shown).The tissue culture media was replaced with normal growth media 24 h after infecting the cells.

Proliferation/viability assays
To assess cell proliferation/viability, we plated LECs (3,000 cells per well) into the wells of a 96-well plate.Cells were infected with lentivirus the following day.Twenty-4 hours later, we replaced the media with regular growth media or with growth media containing DMSO or trametinib (10 nM).For experiments with regular growth media, we measured proliferation/viability 72 h after treating cells with lentivirus.For experiments with DMSO and trametinib, we measured proliferation/viability 72 h after treating cells with DMSO or trametinib.To measure proliferation/viability, we added MTS reagent (abcam, #ab197010) to the wells and read absorbance (490 nm) with a plate reader.

Scratch assays
LECs were plated into the wells of a 6-well or 12-well plate (50,000 cells/ml).Cells were infected with lentivirus the following day.Twenty-4 hours later, we replaced the media with regular growth media and cultured the cells until confluent.We used a p200 pipet tip to wound the monolayers.We then replaced the media with regular growth media or with growth media containing DMSO or trametinib (10 nM) and immediately imaged the cells.We reimaged the cells approximately 14 hours later.FIJI (Figi Is Just ImageJ) was used to measure the wounded area at both timepoints.Mitomycin C was not added to the scratch assays.Therefore, cell proliferation could contribute to the closure of the wounded areas.

Immunoblotting
Cells were scraped in lysis buffer [mPER (Thermo Scientific, #78501) +Protease Inhibitor (Thermo Scientific, #78425) +Phosphatase Inhibitors I and II (Sigma-Aldrich, P2850 and P5726)], spun for 10 min at 4 °C, and then supernatants were transferred to new tubes.Protein concentrations were determined with a BCA kit (Thermo Scientific, #23227).Proteins were separated by SDS-PAGE and then transferred to PVDF membranes.Nonspecific binding was blocked by incubating membranes with TBST +5% non-fat milk for 30 min.Membranes were then incubated overnight with primary antibodies.After washing with TBST, membranes were incubated with the appropriate HRP-conjugated secondary antibodies.Bound antibodies were detected with SuperSignal West Dura Extended Duration Substrate (Thermo Scientific, #34075).

RNA isolation, sequencing, and bioinformatic analysis
To examine the effect of hyperactive KRAS signaling on gene expression, primary human LECs were infected with lentivirus particles, as described above.Twenty-4 hours later, the media was replaced with standard growth media.RNA was isolated with a Qiagen RNeasy Plus Mini Kit (Qiagen, #74314) 72 h after infecting the cells.To assess the impact of trametinib on KRAS G12D -LECs, primary human LECs were infected with lentivirus particles, and the media was replaced with normal growth media 24 h later.Forty-8 hours after infecting cells, they were treated with DMSO or trametinib (10 nM) for approximately 16 h.RNA was then isolated with a Qiagen RNeasy Plus Mini Kit (Qiagen, #74314).Samples were submitted to the Next-Generation Sequencing Core at UT Southwestern Medical Center for library preparation with the TruSeq RNA library prep kit and sequencing with an Illumina NextSeq 2000 system.Gene expression analyses were performed using Partek Flow software.Single-end reads were aligned to the human reference genome (hg38) using STAR, and gene expression was quantified using Ensembl (v109) protein-coding transcript annotations.Transcripts with no detectable expression in any sample were excluded.Expression counts were normalized using the Counts per Million mapped reads (CPM) method for subsequent differential gene expression analysis using the Partek Gene Specific Analysis (GSA) method.GO term analysis was performed with DAVID (https://david.ncifcrf.gov).Transcription regulatory networks were identified with Metascape (www.metascape.org).GraphPad Prism software (Version 9.5.1) was used to generate volcano plots and heat maps.

Evans blue dye lymphangiography
Mice were anesthetized with an intraperitoneal injection of avertin.Evans blue dye (EBD) (1% w/v) was serially injected intradermally into the hind paws and forepaws of control and mutant mice.Approximately 5 min later, the iliac, thoracic, and axillary regions were dissected and visualized under a dissecting microscope.

Microscope and camera information
Back skin samples and fluorescent tissue sections were imaged with a Nikon DS-Qi2 camera attached to a Nikon Eclipse E600 microscope.Tissue sections stained by immunohistochemistry were imaged with an AmScope MU1000 camera attached to a Nikon Eclipse E600 microscope.Embryos were imaged with an AmScope MU1000 camera attached to an AmScope dissecting microscope.Cells were imaged with an AmScope MU1000 camera attached to a Nikon Eclipse TS100 microscope.

Statistical analysis
Data were analyzed using GraphPad Prism statistical analysis software (Version 9.5.1).All results are expressed as mean ± SEM.The number of mice in each group is indicated in the figure legends (n = number of mice).For experiments with two groups, unpaired Student's t-tests were performed to test means for significance.For experiments with more than two groups, group differences were assessed by ANOVA followed by Dunnett's multiple comparisons test or Tukey's multiple comparisons test.We performed Fisher's exact tests to analyze contingency tables for the incidence of edema, lymphovenous valves, and cell clusters in LEC Ctrl and LEC KrasG12D embryos.Data were considered significant at p < 0.05.

FIGURE 1
FIGURE 1 LEC KrasG12D embryos have enlarged jugular lymph sacs and malformed lymphovenous valves.(A).Schematics of the Lyve1-Cre and Kras LSL-G12D alleles.(B).Representative images of E14.5 LEC Ctrl and LEC KrasG12D embryos.The LEC KrasG12D embryo has edema.(C).Transverse sections of E14.5 embryos stained with an anti-Lyve1 antibody (brown) and hematoxylin (purple).(D).Jugular lymph sac area was significantly greater in LEC KrasG12D embryos (273391 ± 47694; n = 6 mice) than LEC Ctrl embryos (76949 ± 17566; n = 5 mice).(E).Coronal sections of E14.5 embryos stained with hematoxylin and eosin (H&E).The arrow points to a lymphovenous valve in a LEC Ctrl embryo.The arrow points to a cluster of cells in the lymphovenous valve region in a LEC KrasG12D embryo.Six LEC Ctrl and five LEC KrasG12D embryos were analyzed.(F).Coronal sections of E14.5 embryos stained with DAPI (blue) and antibodies against Lyve1 (yellow) and endomucin (red).The arrow points to a lymphovenous valve in a LEC Ctrl embryo.The arrow points to a cluster of Lyve1-positive cells in the lymphovenous valve region in a LEC KrasG12D embryo.Five LEC Ctrl and three LEC KrasG12D embryos were analyzed.(G).Immunostaining revealed that cell clusters in the jugular lymph sacs of LEC KrasG12D embryos contained Prox1 and Lyve-1 double-positive LECs.Six LEC KrasG12D embryos were analyzed.(H).Higher magnification image of cluster in panel (G).(I,J).Clusters also contained F4/80-positive cells (I) and CD45-positive cells (J).(K).Immunostaining revealed that clusters in LEC KrasG12D embryos contained Runx1-positive cells.(L-O).Higher magnification images of the separate channels in panel (K).Data are presented as mean ± SEM. **p < 0.01; unpaired Student's t-tests.Scale bar in panel C = 200 µm.Scale bars in panels E, G, I, J, and K = 100 µm.

FIGURE 9
FIGURE 9 Conditional knockout of Kras in LECs does not impair the development or function of lymphatic vessels.(A).Schematics of the Lyve1-Cre and Kras loxp alleles.(B).RT-PCR results for Kras and Gapdh using cDNA generated from RNA isolated from pulmonary endothelial cells from LEC Ctrl and LEC ΔKras mice.(C).Ear skin whole-mount preparations from LEC Ctrl and LEC ΔKras mice stained for Lyve1.(D).The number of lymphatic vessel branch points was not significantly different between LEC Ctrl (21.68 ± 1.049; n = 6 mice) and LEC ΔKras mice (24.10 ± 0.9766; n = 5 mice).(E).The diameter of lymphatic vessels was not significantly different between LEC Ctrl (43.05 ± 1.669; n = 6 mice) and LEC ΔKras mice (47.47 ± 1.141; n = 5 mice).(F).Ear skin whole-mount preparations from LEC Ctrl and LEC ΔKras mice stained for Lyve1, CD31, and Vegfr3.(G).The number of lymphatic valves/mm vessel was not significantly different between LEC Ctrl (0.4812 ± 0.06282; n = 5 mice) and LEC ΔKras mice (0.5495 ± 0.02791; n = 6 mice).(H).Intradermally injected Evans blue dye was effectively transported from injection sites in LEC Ctrl (n = 4 mice) and LEC ΔKras mice (n = 4 mice).Data are presented as mean ± SEM. ns = not significant; unpaired Student's t-tests.Scale bar in panel C = 250 µm.Scale bar in panel F = 100 µm.