Mutant KRAS Enhances Stress Granules and Resistance to Proteasome Inhibition Via 15-d-PGJ2 in Multiple Myeloma

Introduction: Mutant KRAS leads to activation of the MEK/ERK pathway in approximately 20% of multiple myeloma (MM) patients. Stress granules (SGs) are cytosolic non-membranous structures composed of translational mRNAs, ribosomal proteins, and RNA-binding proteins, which form in response to different stress stimuli and chemotherapeutic treatment. We have previously shown that mutant KRAS upregulates SG formation. The molecular mechanisms underlying this process are unclear. The purpose of this study was to elucidate a) the mechanism by which mutant KRAS upregulates SG formation and b) to provide a molecular rationale for the targeting the cascades for SG formation in MM patients.

Methods. We determined whether mutant KRAS is necessary for SG upregulation. To this purpose a panel of MM cell lines including mutant KRAS (KRAS^M), mutant NRAS (NRAS^m) and wildtype KRAS (KRAS^WT) was used. The gain- and loss-of-function of mutant KRAS was generated to suppress (KRAS^M/KO) in KRAS^M or switch on by expression of mutant KRAS G12A (KRASG12A) in KRAS^WT cells by a lentiviral vector system. SG formation was assessed by measuring the cellular distribution of SG resident proteins by immunofluorescence using anti-G3BP and anti-eIF4G antibodies. The quantitative readout for SG formation was analyzed by ImageJ analyze-particle tool as the SG index. Immunoblotting was used to detect inactivation of eIF2a or activity of MAPK. qPCR analysis was employed to detect the mRNA of COX2. MTT assays were used to detect the effect of SGs on the survival of MM cells alone or in combination with diclofenac sodium (DS), or bortezomib (Bzb).

Results: We compared the induction of SGs in response to oxidative stress across a panel of MM cells with different Ras mutation status (KRAS^M, NRAS^M and KRAS^WT)to investigate the role mutant KRAS in GS formation. The level of SGs were significantly higher (p<0.01) in six KRAS^M, compared with four KRAS^WT cell lines.We assessed SG formation in KRAS^M, KRAS^M/KO and KRAS/G12A cells to determine whether mutant KRAS is a causative factor leading to elevated SG levels. SG formation in KRAS^M/KO cells was significantly attenuated, by approximately 5-fold (p<0.001), compared to KRAS^M in response to oxidative stress. Conversely, SG levels were 3.5-fold elevated in KRAS/G12A cells (P<0.01), compared to KRAS^WT cells. Upregulation of SGs by mutant KRAS resulted in a larger average size and number of SGs per cells. Together, these results indicate that mutant KRAS is necessary and sufficient for SG upregulation under oxidative stress condition.

We investigated whether 15-d-PGJ2-depedent inactivation of elF4A is involved in SG upregulation to determine the mechanism through which mutant KRAS promotes SG formation. To do this, we assessed the effect of 15-d-PGJ2 on SG formation because that elF4A inactivation is controlled by 15-d-PGJ2, which is one of COX2-derived products. Treatment of KRAS^M cells with 15-d-PGJ2 significantly induced accumulation of SGs compared with untreated cells (40-fold increase vs control cells, p<0.001). To determine whether COX2, which controls the rate of biosynthesis of 15-d-PGJ2, is regulated by mutant KRAS, we first compared COX2 mRNA of KRAS^M/KO with KRAS^M and showed that a decrease in COX2 expression in KRAS^M/KO (2-fould decrease, P< 0.01). Conversely, increased COX2 mRNA was seen in KRAS/G12A, compared with KRAS^WT cells. Furthermore, treatment of KRAS^M cells with DS led to a decrease in SG levels and this effect was rescued by the addition of 15-d-PGJ2. These results indicate that increase in prostaglandin levels by upregulating prostaglandin biosynthesis is essential for mutant KRAS-induced up-regulation of SGs. Finally, we determined if inhibition of COX2 activity by DS was able to overcome Bzb resistance in KRAS^M or in KRAS^M co-occurring t(14:16)/MAF. A significant decrease of cell viability (P<0.001) was observed when treatment of the cells with Bzb and DS were combined, compared with DS treatment alone. Taken together, these results indicate that inhibition of COX2 activity enhances mutant KRAS MM cell sensitivity to proteasome inhibition.

Conclusion: Our results suggest that mutant KRAS mutant upregulates SG formation by 15-d-PGJ2 and inhibition of COX2 activity enhances KRAS sensitivity to proteasome inhibition and targeting SG formation might be represent an effective strategy to treatment of KRAS mutant myeloma.