Lymphatic disorders caused by mosaic, activating KRAS variants respond to MEK inhibition

Abstract

Central conducting lymphatic anomaly (CCLA) due to congenital maldevelopment of the lymphatics can result in debilitating and life-threatening disease with limited treatment options. We identified 4 individuals with CCLA, lymphedema, and microcystic lymphatic malformation due to pathogenic, mosaic variants in KRAS. To determine the functional impact of these variants and identify a targeted therapy for these individuals, we used primary human dermal lymphatic endothelial cells (HDLECs) and zebrafish larvae to model the lymphatic dysplasia. Expression of the p.Gly12Asp and p.Gly13Asp variants in HDLECs in a 2‑dimensional (2D) model and 3D organoid model led to increased ERK phosphorylation, demonstrating these variants activate the RAS/MAPK pathway. Expression of activating KRAS variants in the venous and lymphatic endothelium in zebrafish resulted in lymphatic dysplasia and edema similar to the individuals in the study. Treatment with MEK inhibition significantly reduced the phenotypes in both the organoid and the zebrafish model systems. In conclusion, we present the molecular characterization of the observed lymphatic anomalies due to pathogenic, somatic, activating KRAS variants in humans. Our preclinical studies suggest that MEK inhibition should be studied in future clinical trials for CCLA due to activating KRAS pathogenic variants.

Keywords: Cardiology; Cardiovascular disease; Genetics; Molecular biology; Molecular genetics.

Figure 1. DCMRL of individuals 1, 2, and 4.

(A) Intrahepatic DCMRL of individual 1 illustrating retrograde mesenteric perfusion (arrow) and pulmonary lymphatic perfusion with dilated mediastinal lymphatics (arrowhead). (B) Intranodal DCMRL of individual 1 shows retrograde into the mesentery (arrow), right renal, and dermal lymphatics (arrow), with an i.p. leak and intact thoracic duct coursing to the left venous angle (arrowhead). (C) Intrahepatic DCMRL of individual 2 shows retrograde flow into the lumbar and iliac lymphatics (arrow), splenic lymphatics (arrow), and i.p. leak (*). There is also bilateral pulmonary and mediastinal lymphatic perfusion (arrowhead) without a thoracic duct. (D) Intrahepatic and (E) intranodal DCMRL of individual 4 show a normal-appearing thoracic duct (arrowhead) with retrograde lumbar perfusion (arrow) and intraduodenal leak (arrow). Intranodal DCMRL shows dilated iliac lymphatics with retrograde dermal perfusion (scrotal and penile).

Figure 2. A 2D in vitro model of lymphatic dysplasia.

HDLECs were stained with VE-cadherin or actin. Scale bars: 50 μm. (A) KRAS WT. (B) KRAS p.Gly12Asp. Yellow circles show extensions from cells. (C) KRAS p.Gly12Asp treated with 30 nm trametinib. (D) KRAS p.Gly12Asp treated with 1 μm binimetinib. Yellow arrows show areas where abnormal extensions remain. (E) Cell lysates from HDLECs transduced with either KRAS WT or p.Gly12Asp were analyzed with IB for pERK at T202 and Y204 or pS6 at S235/236 with actin as control. (F) Quantification of IB, pERK, or pS6 normalized to actin, normalized to WT + DMSO sample. Data were quantitated from 4 independent experiments. Bars are means; error bars are SDs. One-sided Student’s t tests were performed to calculate significance. Bin, binimetinib; Tram, trametinib.

Figure 3. In vitro organoid model.

(A) Lymphatic organoids were transduced with KRAS WT, KRAS p.Gly12Asp, or KRAS p.G13D and treated with DMSO (control), 1 μM binimetinib, 3 μM binimetinib, 10 μM binimetinib, or 300 nM trametinib. Scale bars: 300 μm. (B) Quantitation of sprouting data from 3 independent experiments showing cumulative sprout length per sphere (top), mean sprout length per sphere (middle), and number of sprouts per sphere (bottom). In the box and whisker plots, the center line is the median, the lower and upper boundaries of the box are the 25% and 75% quartiles, and the whiskers extend to 1.5 times the interquartile range from the 25% and 75% quartiles. Two-sided Student’s t tests were performed to calculate significance. Comparisons were made between DMSO-treated WT and both DMSO-treated mutants, as well as each DMSO-treated mutant and all the drug treatments of that mutant; P values were corrected for multiple testing with the Benjamini and Hochberg FDR method. Bin, binimetinib; Tram, trametinib. (C) IB from in vitro organoid model. Cell lysates from HDLECs transduced with either KRAS WT, p.Gly12Asp, or p.Gly13Asp were analyzed with IB for pERK at T202 and Y204 or pS6 at S235/236 with actin as control and quantified. (D) Quantification of IB from 4 separate experiments, pERK or pS6 normalized to actin, normalized to WT + DMSO sample. Bars are means; error bars are SDs. Bin, binimetinib; Tram, trametinib.

Figure 4. In vivo zebrafish larvae modeling and therapeutic screening.

(A) Embryos were injected at 0-cell stage with either mrc1a:wt-hKRAS, mrc1a:hKRAS p.Gly12Asp, mrc1a:hKRAS p.Gly13Asp and tol2 transposase. Larvae at 7 dpf under light microscopy (2.5× magnification), GFP (2.5× magnification), mCherry (2.5× magnification), and confocal microscopy (20× magnification). Larvae injected with mrc1a:hKRAS p.Gly12Asp or p.Gly13Asp under light microscopy have edema around the heart and intestine (arrows). Mrc1a:wt-hKRAS have essentially normal vasculature of larvae injected with under confocal microscopy. Vasculature of larvae injected with mrc1a:hKRASp.Gly12Asp or mrc1a:hKRASp.Gly13Asp under confocal microscopy showing fusion of the thoracic duct with the cardinal vein (brackets). (B) In vivo zebrafish larvae therapeutic screening. Embryos were treated at 48 hpf and screened for edema at 5.5 dpf. (C) Fraction of larvae with edema by KRAS variant, drug, and concentration. Each dot represents a single experiment. *P < 0.05 by unpaired, 1-tailed Student’s t tests, after correction for multiple testing with the Benjamini and Hochberg FDR method. The mechanism of action for each drug can be seen in Supplemental Table 1. Due to figure legend space limitations, the number of zebrafish larvae for each experiment are in Supplemental Table 2.