Frequent functional activation of RAS signalling not explained by RAS/RAF mutations in relapsed/refractory multiple myeloma

RAS mutations are frequent in relapsed/refractory multiple myeloma (RRMM) but functional study in primary samples is scanty. Herein, in primary myeloma plasma cells of 17 suspected RRMM, functional activation of RAS signalling was studied by Western blot of phosphorylated ERK1/2 (phospho-ERK1/2). Moreover, activating mutations in KRAS, NRAS, BRAF, and ALK were studied by PCR and bidirectional direct sequencing. Furthermore, methylation of negative RAS signalling regulator genes, RASSF1A and RASD1, were analyzed by methylation-specific PCR. As evidenced by phospho-ERK1/2 over-expression, functional RAS activation was detected in 12 (75.0%) RRMM. Of patients with functional RAS activation, sequencing data showed only seven (58.3%) patients with one each had NRAS Q61H, NRAS Q61K, KRAS G12D, KRAS G12V, KRAS G13D, KRAS Q61P, or BRAF V600E mutation, whereas five (41.7%) patients had no RAS/RAF mutation. Conversely, patients without functional RAS activation had no RAS/RAF mutation. Moreover, none of the patients with functional RAS activation had ALK mutations, or methylation of RASSF1A and RASD1. Collectively, functional activation of RAS signalling was present in majority of RRMM but only about half (58.3%) accountable by RAS/RAF mutations. If verified in larger studies, clinical investigations of MEK inhibitors are warranted regardless of RAS/RAF mutations.


Results
Of the 17 patients, one patient (P5) had bone marrow examination for suspected myeloma relapse due to onset of pancytopenia, which revealed Philadelphia chromosome-negative acute lymphoblastic leukaemia but no evidence of myeloma, hence was excluded from analysis. By Western blot analysis on primary myeloma plasma cells, 12/16 (75.0%) RRMM patients showed phospho-ERK1/2, indicating constitutive activation of the RAS-RAF-MEK-ERK signalling pathway ( Fig. 1A and Table 1 Table 1). However, the other five patients with functional ERK activation had no RAS/RAF mutation at the selected mutation hotspots, including KRAS/NRAS (codon 12, 13 and 61) and BRAF (codon 469 and 600), which account for almost all RAS activation in cancers 1 . Moreover, in the remaining four patients without functional ERK activation, no RAS/RAF mutation was detected.
Furthermore, alternative mechanism of functional activation of RAS signalling by activating mutations of ALK (F1174, F1245, or R1275), which accounted for ERK activation in diagnostic and relapsed neuroblastoma 5 , was not found in our patients with functional activation of RAS (Fig. 1B). In addition, as amp(1q21) is an adverse cytogenetic aberration that may be acquired in myeloma patients at relapse 12 with overexpression of CKS1B 13 , association of amp(1q21) with activation of RAS signalling in myeloma was investigated ( Supplementary  Fig. S1). However, by FISH in marrow samples at relapse, amp(1q21) was present in seven with and two without phospho-ERK1/2 (P = 0.700; Table 1), hence not associated with functional activation of RAS. Finally, to investigate if methylation-mediated silencing of tumour suppressor genes negatively regulating RAS signalling pathway may account for functional activation of RAS, DNA methylation of RASSF1A and RASD1 promoters were studied by methylation-specific PCR (MSP). Similar to previous reports 6,7 , methylation of RASSF1A and RASD1 were absent in normal controls, including CD138-sorted normal bone marrow plasma cells (n = 8) and normal peripheral blood buffy coats (n = 10) ( Supplementary Fig. S2), but detected in six (60%) and one (10%) human myeloma cell lines respectively, hence tumour-specific ( Fig. 2A). Moreover, methylation of RASSF1A and RASD1 was associated with low expression in human myeloma cell lines ( Fig. 2A), thereby demonstrating methylation-mediated gene silencing. However, in primary samples, neither RASSF1A nor RASD1 methylation was detected (Fig. 2B).

Discussion
We have demonstrated frequent functional activation of RAS signalling pathway in 75% of RRMM, but only about half (58.3%) accountable by RAS/RAF mutations. In majority of previous studies, RAS activation, defined and hence inferred by the presence of RAS/RAF DNA mutations, was present in about 25-55% of newly diagnosed myeloma, and about 45-81% of RRMM 9,10,14-17 . However, our study is the first functional study of RAS activation in primary myeloma samples by Western blot of phospho-ERK1/2.
There is a recent study, in which, functional RAS activation was assessed by immunohistochemistry (IHC) for phospho-ERK1/2, and revealed functional RAS activation in 41.7% RRMM, in which RAS/RAF mutations were found in 80% 18 . However, the discrepancy in the frequency of functional RAS activation between of both studies might be partly explained by the different definitions of ERK activation between IHC and Western blot. In the IHC study 18 , ERK activation was defined by the presence of median/strong IHC signal for phospho-ERK1/2 (intensity score >1) in ≥30% of tumour cells. On the contrary, in our study, ERK activation was defined by the presence of Western blot signal for phospho-ERK1/2 in protein lysate of CD138-sorted myeloma plasma cells of the patients' marrow and the positive (HeLa cells) but not negative (KMS-12-PE cells) controls. However, given that the incidence of myeloma is much higher in the Western countries with an incidence of 6.6/100,000/year in the US 19,20 , as compared to about 1.7/100,000/year in the Hong Kong Chinese 21 , a genuine difference in the frequency of RAS activation remains possible. On the other hand, in contrast to the IHC study showing the presence of NRAS G13R in cases without functional RAS activation, our data showed absence of RAS/RAF mutation in all RRMM patients without functional RAS activation, suggesting that functional RAS signaling activation defined by Western blotting of phosphor-ERK1/2 may be more specific. Collectively, our data showed that functional RAS activation was prevalent in RRMM that would otherwise be underestimated by the frequency of RAS/RAF DNA mutations.
Regarding the mechanism of RAS activation, RAS/RAF mutation is the most important cause. Indeed, various studies had shown frequent RAS/RAF mutations in particular in RRMM 9,15 . Other mechanisms of RAS activation include gain-of-function mutation of oncogenes including ALK [22][23][24] . In contrast to frequent ALK mutations accounting for RAS activation in neuroblastoma 5,25 , our study did not demonstrate any ALK mutation.
On the other hand, RAS activation may result from loss-of-function of tumour suppressor genes, either by loss of function mutation or promoter DNA methylation of negative regulators of RAS signalling. DNA methylation is an alternative mechanism of gene silencing mediated by addition of methyl groups to the cytosine rings of the CpG dinucleotides in the promoter-associated CpG islands of tumor suppressor genes 26 . In myeloma, multiple tumour suppressor genes and miRNAs have been shown to be associated with methylation-mediated gene silencing 27 . For instance, RASSF1A and RASD1 promoter DNA methylation have been shown in myeloma to be associated with activation of RAS signalling 6,7 . Herein, we showed tumour-specific methylation of both RASSF1A and RASD1 in myeloma cell lines, as evidenced by the presence of methylated MSP signals in myeloma cell lines but not normal control DNA. Moreover, the inverse correlation between RASD1 methylation and expression was consistent with methylation-mediated gene silencing. Indeed, in a genome-wide methylation study of 115 primary myeloma samples using Illumina 27 K 28 , RASD1 methylation was inversely correlated with expression, hence further testifying the role of methylation-mediated silencing for RASD1 in myeloma patients. However, in contrast to the previous reports of RASSF1A and RASD1 hypermethylation in primary myeloma samples 6,7 , RASSF1A or RASD1 methylation was absent in all samples, including those with functional RAS activation. Therefore, methylation of either RASSF1A or RASD1 may not be important for the activation of RAS in RRMM. However, in view of the limited number of samples herein, further studies with larger number of patients are required. Moreover, investigation into additional mechanisms of activation of RAS signalling is warranted. In addition to promoter DNA methylation, RASSF1A inactivation had been shown to be associated with a repressive chromatin configuration with histone deacetylation and H3K9 dimethylation in prostate cancer cell lines 29 .Nonetheless, in primary lung, liver and renal cancer tissues, RASSF1A inactivation has been consistently shown to be mediated by promoter DNA methylation, hence an important role in the regulation of RASSF1A expression 30 .
Finally, as our study showed functional activation of RAS in the majority (75%) of RRMM regardless of the presence of RAS/RAF mutation, activated RAS signaling poses a potential therapeutic target in RRMM. Moreover, as functional activation of RAS may occur with either RAS or RAF mutations, inhibition of RAS signalling appears more effective by the use of inhibitors to downstream signalling effectors such as MEK or ERK1/2. Indeed, there are ongoing clinical trials using MEK inhibitor, trametinib, in RRMM, which appeared promising 31,32 .

Conclusions
Collectively, functional activation of RAS signalling, as evidenced by ERK activation, was present in the majority of RRMM, but only accountable by known RAS/RAF mutations in half. In view of prevalent functional RAS activation regardless of RAS/RAF mutations, a clinical trial of MEK inhibitors is warranted in RRMM.

Methods
Patients. Seventeen patients with suspected RRMM, including those with "relapsed" or "relapsed-and-refractory myeloma" were included for functional study of activation of RAS-RAF-MEK-ERK cascade by Western blot of phospho-ERK1/2 in primary myeloma plasma cells from the CD138-sorted marrow or nodal plasma cells.
Relapsed myeloma is defined as previously treated myeloma that progresses and requires the initiation of salvage therapy but does not meet criteria for either "primary refractory myeloma" or "relapsed-and-refractory myeloma" categories 33 . Relapsed and refractory myeloma is defined as disease that is nonresponsive while on salvage therapy, or progresses within 60 days of last therapy in patients who have achieved minimal response (MR) or better at some point previously before then progressing in their disease course 33 .
Patient demographics were described in Supplementary Table S1. This study has been approved by the Institutional Review Board of Queen Mary Hospital with informed consent in accordance with the Declaration of Helsinki. All study methods were performed in accordance with relevant guidelines and regulations. Western blotting for ERK1/2 activation. Purified CD138+ plasma cells obtained from magneticactivated cell sorting (Miltenyi Biotec, Cologne, Germany) were lysed in RIPA buffer supplemented with phosphatase inhibitor cocktail and 1 mM PMSF. Cell debris was removed by centrifugation at 10,000 × g for 5 min at 4 °C. Protein lysate was heated in an equal volume of blue loading buffer at 95 °C for 5 min, and separated on 10% SDS-PAGE. Separated samples were then transferred to a 0.45 μm PVDF membrane (Amersham Biosciences, Buckinghamshire, UK). The membrane was blocked at room temperature for 1 hour in 5% skim milk diluted in PBS-Tween 20 (0.5% v/v). The membrane was then incubated with specific primary antibody (1:1000) (Cell signalling, Danvers, MA, USA) at 4 °C overnight with shaking. After washing 3 times of 15 minutes each in PBS-Tween 20 (0.5% v/v), the membrane was incubated with specific horseradish peroxidase conjugate secondary antibody (1:1000) (Bio-Rad) at room temperature for 1 hour. After washing 3 times of 15 minutes each in PBS-Tween 20 (0.5% v/v), signals were detected by ECL Western blotting detection reagents (Amersham Biosciences, Buckinghamshire, UK) and exposed to X-ray film.
Mutation analysis of KRAS, NRAS, BRAF, and ALK. Direct sequencing analysis was employed to study codons 12, 13, and 61 of KRAS and NRAS, which encompass more than 96% of all RAS mutations in human cancers 34 ; to examine codons 469, 600 and 601 of BRAF, which represent more than 85% of all BRAF mutations in human cancers 35 ; and to study codons 1174, 1245, 1275 of ALK, which account for 85% of ALK mutations in diagnostic neuroblastoma 36 . In brief, genomic DNA extracted from CD138-sorted plasma cells was amplified by PCR, followed by bidirectional direct sequencing. Primer sets specific to these regions adopted from literatures were shown in Supplementary Table S2 5  products were loaded onto 6% non-denaturing polyacrylamide gels, electrophoresed, and visualized under ultraviolet light after staining with ethidium bromide.

Quantitative real-time reverse transcription-PCR (qRT-PCR). In myeloma cell lines, expression of
RASSF1A or RASD1 was studied by SYBR Green-based qRT-PCR. In brief, total RNA was isolated using mir-Vana ™ miRNA Isolation Kit (Ambion, Austin, TX, USA), followed by reverse transcription to cDNA using QuantiTect Reverse Transcription Kit (Qiagen), according to the manufacturers' instructions. The resulting cDNA was used as template for qRT-PCR (iQ SYBR Green Supermix, Bio-Rad), with GAPDH as endogenous control. Primer sequences were listed in Supplementary Table S2. Correlation between methylation and expression was studied by Student's t-test (two-tailed), whereas P < 0.05 was regarded as statistical significant.
All data generated or analysed during this study are included in this published article (and its Supplementary  Information files).
Ethics approval and informed consent. The study has been approved by Institutional Review Board of Queen Mary Hosital (UW 05-269T/932), and written informed consent was obtained from patient for publication of this study and any accompanying data or images.