RASA2 and NF1; two-negative regulators of Ras with complementary functions in melanoma.
RASA2 and NF1; two-negative regulators of Ras with complementary functions in melanoma.
复制标题
RASA2 和 NF1；
DOI：
10.1038/s41388-018-0578-4
复制
发表时间：
2019
期刊：
影响因子：
8
通讯作者：
Samuels,Yardena
中科院分区：
文献类型：
--
作者：
Arafeh,Rand;DiPizio,Antonella;Elkahloun,AbdelG;Dym,Orly;Niv,MashaY;Samuels,Yardena
RASA 2先前已被证明是黑色素瘤细胞中的功能性RasGAP [1]。RASA 2的突变或缺失促进黑素瘤中的RAS活化[1]。我们对RASA 2突变的遗传分析表明，RASA 2和NRAS突变是相互排斥的。（p= 0.002，Fisher精确检验），并且NF 1突变[2，3]与RASA 2突变显著共存（p= 0.000011，Fisher精确检验）在BRAF和NRAS野生型黑素瘤中，这表明RASA 2和NF 1的缺失具有互补的促肿瘤发生功能（图1A）。Maertens等人[3]表明NF 1对黑色素瘤中的KRAS和HRAS具有特异性。为了在我们的细胞中验证这一观察结果，我们在具有RASA 2和NF 1突变（CO 84）的黑素瘤细胞中过表达单独的载体（EV）或野生型NF 1，并发现与NRAS-GTP水平相比，野生型NF 1基本上抑制KRAS-GTP和HRAS-GTP水平（图1 B）。由于NF 1和RASA 2是RasGAP，它们的改变共同发生，并且同一肿瘤中共同发生的基因改变通常需要协同作用，因此我们假设RASA 2和NF 1可能通过作用于不同的RAS亚型而协同作用。事实上，我们发现RNAi介导的RASA 2抑制导致NRAS-GTP的激活，但不导致HRAS或KRAS的激活（图1C和补充图1A）。相反，野生型RASA 2的过表达基本上抑制了NRAS-GTP，但不抑制KRAS或HRAS，而RASA 2突变体则不能做到这一点（补充图1B、C）。重要的是，将野生型RASA 2重新引入到携带RASA 2突变的黑素瘤细胞中抑制NRAS活化（图1D和补充表）。1）。为了深入了解这种特异性，我们预测了RASA 2和NF 1与RAS蛋白的结合界面。与NF 1相比，RASA 2在其L 6c环555 SKSKSSFKE-T564 RASA 2（图1 E和补充图4）和片段345 SAAYILSEICRDK 357 RASA 2中具有不同的长度和氨基酸组成。由后者建立的相互作用可能对NRAS螺旋3具有特异性，因为它们在RASA 2-HRAS和RASA 2-KRAS中均未观察到。在RASA 2-NRAS复合物中，RASA 2 R355与NRAS D92建立H键，而在RASA 2-HRAS和RASA 2-KRAS中，R355通过分子内H键与D356相互作用，阻止与D92的相互作用，D92进而分别参与与HRAS和KRAS中的Q95和H95的分子内H键。在NF 1-HRAS和NF 1-KRAS复合物中，相同区域包括C1233与E91之间的氢键以及D92与K88和Q（H）95之间的分子内氢键，而在NF 1-NRAS复合物中，仅观察到C1233与D92之间的氢键。这些差异可能解释了RASA 2和NF 1对RAS蛋白的不同特异性。为了评估RASA 2和NF 1敲低对MAPK通路的影响，我们建立了稳定的合并克隆，其中我们在A375细胞中单独或组合敲低RASA 2和NF 1的表达，并检查对MAPK通路活化的影响。我们观察到当RASA 2和NF 1两者都被敲低时，ERK的磷酸化与当各自单独被敲低时相比增强（补充图5A）。此外，与其单独敲除相比，RASA 2和NF 1两者的敲除有效地增加了免疫受损小鼠中异种移植物的生长（补充图5 B）。因此，GAP、RASA 2和NF 1的缺失激活MAPK途径并增强体内肿瘤形成。
RASA2 has previously been shown to be a functional RasGAP in melanoma cells [1]. Mutation or loss of RASA2 promotes RAS activation in melanoma [1]. Our genetic analysis of RASA2 mutations identified that RASA2 and NRAS mutations are mutually exclusive (p= 0.002, Fisher, s exact test), and that NF1 mutations [2, 3] significantly cooccur with RASA2 mutations (p= 0.000011, Fisher, s exact test) in BRAF and NRAS wild-type melanomas, suggesting that loss of RASA2 and NF1 have complementary protumorigenic functions (Fig. 1 A). Maertens et al.[3] showed that NF1 is specific for KRAS and HRAS in melanoma. To validate this observation in our cells, we overexpressed vector alone (EV) or wild-type NF1 in melanoma cells that harbor RASA2 and NF1 mutations (CO84) and found that wild-type NF1 substantially suppressed KRAS-GTP and HRAS-GTP levels compared to NRAS-GTP levels (Fig. 1 B). As NF1 and RASA2 are RasGAPs and their alterations cooccur, and co-occurring gene alterations in the same tumor are often held to entail synergy, we hypothesized that RASA2 and NF1 may synergize by acting on different RAS isoforms. Indeed, we found that RNAi-mediated suppression of RASA2 led to activation of NRAS-GTP, but not HRAS or KRAS (Fig. 1 C and Supplementary Fig. 1A). Conversely, overexpression of wild-type RASA2 substantially suppressedNRAS-GTP, but not KRAS or HRAS, whereas RASA2 mutants failed to do so (Supplementary Fig. 1B, C). Importantly, re-introduction of wild-type RASA2 into melanoma cells harboring RASA2 mutations inhibited NRAS activation (Fig. 1 D and Supplementary Table. 1). To gain insight into this specificity, we predicted the binding interface of RASA2 and NF1 with RAS proteins. RASA2, compared to NF1, has a different length and amino acid composition in its L6c loop 555SKSKSSFKE-T564RASA2 (Fig. 1 E and Supplementary Fig. 4) and in the fragment 345SAAYILSEICRDK357RASA2. The interactions established by the latter may be specific for NRAS helix 3 as they are not observed in RASA2-HRAS nor in RASA2-KRAS. In RASA2-NRAS complex, RASA2 R355 establishes an H-bond with NRAS D92, while in both RASA2-HRAS and RASA2-KRAS, R355 interacts via intramolecular H-bond with D356, preventing the interaction with D92, which, in turn, is involved in intra-molecular H-bonds with Q95 and H95 in HRAS and KRAS, respectively. The same region, in NF1-HRAS and NF1-KRAS involves H-bond between C1233 and E91 and intra-molecular H-bonds between D92 and K88 and Q (H) 95, while in NF1-NRAS complex, only H-bond between C1233 and D92 is observed. These differences might explain the different specificity of RASA2 and NF1 toward the RAS proteins. To assess the effects of RASA2 and NF1 knockdown on the MAPK pathway, we established stable pooled clones in which we knocked-down the expression of RASA2 and NF1 individually or in combination in A375 cells and checked the effect on MAPK pathway activation. We observed that the phosphorylation of ERK is enhanced when both RASA2 and NF1 are knocked-down compared to when each is knocked-down individually (Supplementary Figure 5A). Additionally, knockdown of both RASA2 and NF1 potently increases the growth of xenografts in immunocompromised mice compared to their individual knockdowns (Supplementary Figure 5B). Therefore, loss of both GAPs, RASA2 and NF1 activates the MAPK pathway and enhances tumor formation in vivo.