Abstract: Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors in the gastrointestinal tract, and they have been refractory to conventional chemotherapy with a median survival of only 12 to 19 months in patients with unresectable or metastatic diseases. Most GISTs express KIT, a receptor tyrosine kinase encoded by proto-oncogene KIT. Approximately 90% of GISTs have somatic gain-of-function mutations of the KIT and half of GISTs without KIT mutations have been demonstrated to have gain-of-function mutations in the PDGFRA (platelet-derived growth factor receptor-a) that encodes another receptor tyrosine kinase. lmatinib mesylate (imatinib) is a specific tyrosine kinase inhibitor that acts with BCR-ABL, PDGFRA, KIT, and has been used to treat GISTs that have constitutive activating mutations in KIT. Although more than 80% of inoperable GISTs patients have dramatic clinical benefits from imatinib, most of these patients will eventually progress. The effect of imatinib is different in various types of KIT and PDGFRA mutations, and the secondary resistance against imatinib is often acquired by the secondary mutation or amplification of the KIT or PDGFRA. Since only about half of GISTs with secondary resistance to imatinib have secondary mutations in KIT or PDGFRA, the mechanisms of the remaining secondary resistance have not yet been fully elucidated.

To explore additional mechanisms of imatinib-resistant GISTs, I generated resistant cells from imatinib-sensitive GIST-T1 cells with heterozygous del(V560-Y578) in exon 11 by exposing them to increasing concentrations of imatinib for 6 months. The resultant cell line, GIST-T1 IR showed resistant to imatinib in vitro with IC(50) 5-7 uM of imatinib, and phosphorylated KIT and its downstream intermediates such as AKT and JAK2 with the presence of 1 uM imatinib (Figure 1). To see whether there were any new mutations occurring in KIT, PDGFRA, PKCO, and JAK2, I sequenced all exons of these genes of GIST-T1 IR cells; however, I did not find any new mutation. I next used DNA micro-array to check the expression profile of GIST-T1 IR cells, I found over-expression of Cas-L in the resistant cells with 513 fold higher than that in the parental cells. The over-expression of Cas-L was also observed in GIST-T1 IR cells at protein level (Figure 2). Cas-L is a scaffolding protein at focal adhesion sites; its action involves regulated assembly of protein complexes. The interaction of Cas-L, FAK (focal adhesion kinase) and SRC kinases plays an important role at adhesion sites, overexpression or hyper-activation of one of these proteins leads to the activation of the others. I therefore checked the activations of SRC and FAK in GIST-T1 IR cells. As expected, I found hyperactivation of SRC and FAK in GIST-T1 IR cells comparing with those in GIST-T1 cells (Figure 2).

To verify whether the increased SRC signaling underlies the acquired resistance to imatinib of GISTT1 IR cells, I used MTT assay to examine the effect of PP1, a SRC inhibitor, on cellular proliferation of GIST-T1 IR cells. PP1 (10 uM) or imatinib (1 uM) alone could suppress the proliferation of GISTT1 IR by 40% and 33%, respectively. However, the combination of them showed strong inhibitory effect on the growth of the resistant cells by suppressing 84% of cellular proliferation. Additionally,the combination of 10 uM PP1 and 1 uM imatinib drastically suppressed the activation of KIT and SRC kinases in GIST-T1 IR cells (Figure 3).

Tyr568, a SRC binding site of KIT, was deleted in the parental and the resistant cells and it is likely that SRC kinase did not have a crucial function in the parental cells, but it turned out to be activated in the resistant cells. The next important question is whether or not the SRC signaling activities are dependent on KIT signaling pathway. In order to examine the relation between KIT, Cas-L and SRC kinases, I transfected KIT siRNA and Cas-L siRNA into GIST-T1 IR cells. The expression of Cas-L and the phosphorylation of SRC were almost vanished in KIT siRNA transfected GIST-T1 IR cells,suggesting that the expression of Cas-L and the activation of SRC kinase were dependent on KIT signaling. Additionally, the phosphorylation of SRC was decreased in corresponding to the decreased expression of Cas-L, suggesting that Cas-L expression was important for SRC activation.To scrutinize more the important role of Cas-L and SRC kinase in inducing imatinib-resistance of GIST-T1 IR cells, I next checked the effect of imatinib on GIST-T1 IR cells transfected with Cas-L siRNA. As expected, Cas-L siRNA transfected GIST-T1 IR cells turned out to become again sensitive to imatinib with IC50 < 0.1 uM (Figure 4)

I herein reported for the first time the over-expression of Cas-L and its important role in imatinibresistance GISTs. This should be of interests to oncologists to further investigate signaling pathways related to Cas-L/SRC in GISTs.

Figure 1: GIST-T1 IR cells showed resistant to imatinib. The phosphorylated of KIT and its downstream intermediates remained activated in the presence of 1 uM imatinib.

Figure 2: Cas-L was over-expressed; SRC and FAK were hyper-activated in GIST-T1 IR cells

Figure 3: The combination of PP1 and imatinib showed synergistic inhibitory effects on the cellular proliferation as well as the phosphorylation of KIT and SRC in GIST-T1 IR cells

Figure 4: Cas-L and SRC were dependent on KIT signaling. Cas-L depletion sensitized the resistant GIST-T1 IR cells to imatinib.

消化管間質腫瘍(GIST)は消化管の間葉系腫瘍の中では最も頻度が高いものであるが、通常の化学療法には抵抗性で、手術不能の患者の生存中央値は12ヶ月から19ヶ月である。大部分のGISTに発現する受容体型チロシンキナーゼKITには、90%位の頻度でgain-of-functionの変異が認められる。また、KITの変異が無いGISTでは受容体型チロシンキナーゼPDGFレセプタ-α(PDGFRα)にgain-of-functionの変異がある。これらの変異を持ったチロシンキナーゼは校正的に活性化されており、GISTの増殖に必須である。分子標的治療薬imatinibがGISTの抗腫瘍薬として用いられて著効を示す事が知られている。しかし、imatinib耐性細胞の出現が問題となっている。本研究では、imatinib耐性GISTの薬剤耐性の分子機構を明らかにする事を目的として解析を行った。GIST細胞株GIST-T1を用いてin vitroで耐性株を確立し、耐性獲得機構を検討した。この細胞株におけるKIT,PDGFRA,PKCq,JAK2などには変異は認められなかった。しかし、発現アレイ解析を行った結果、Cas-Lの過剰発現が認められた。Cas-Lは接着部位のドッキング蛋白質として知られており、近年ではがん細胞の転移への関与が報告されている。imatinib耐性とCas-L発現の関係を明らかにするために、siRNAを用いてCas-Lのノックダウンを行ったところ、imatinib感受性が回復したことから、Cas-L過剰発現がimatinib耐性に関与している事を初めて明らかにした。

本研究は、GISTにおける分子標的薬imatinib耐性機構の検討からCas-Lの役割を初めて明らかにして、耐性克服への新たな方向性を示したものであり、薬剤耐性獲得の分子機構の理解に重要な貢献をするものであると考えられる。したがって、博士(生命科学)の学位を授与できると認める。