Targeting c‐met receptor tyrosine kinase by the DNA aptamer SL1 as a potential novel therapeutic option for myeloma

Abstract Hepatocyte growth factor (HGF)/c‐met pathway activation has been implicated in the pathogenesis of multiple myeloma (MM), and blocking this pathway has been considered a rational therapeutic strategy for treating MM. Aptamers are single‐stranded nucleic acid molecules that fold into complex 3D structures and bind to a variety of targets. Recently, it was reported that DNA aptamer SL1 exhibited high specificity and affinity for c‐met and inhibited HGF/c‐met signaling in SNU‐5 cells. However, as the first c‐met‐targeted DNA aptamer to be identified, application of SL1 to myeloma treatment requires further investigation. Here, we explore the potential application of SL1 in MM. Our results indicated that c‐met expression is gradually increased in MM patients and contributes to poor outcomes. SL1 selectively bound to c‐met‐positive MM cells but not to normal B cells and suppressed the growth, migration and adhesion of MM cells in vitro in a co‐culture model performed with HS5 cells, wherein SL1 inhibited HGF‐induced activation of c‐met signaling. In vivo and ex vivo fluorescence imaging showed that SL1 accumulated in the c‐met positive tumour areas. In addition, SL1 was active against CD138+ primary MM cells and displayed a synergistic inhibition effect with bortezomib. Collectively, our data suggested that SL1 could be beneficial as a c‐met targeted antagonist in MM.

Most of MM development depends on the interactions between MM cells and components of the bone marrow microenvironment (BMME) and the signaling that results. 4 One promising candidate signaling pathway is the hepatocyte growth factor (HGF)/c-met pathway.
HGF is a constituent of the BMME and the specific ligand for the tyrosine kinase receptor c-met. 5 Upon HGF binding, c-met dimerizes, resulting in trans-phosphorylation of two tyrosine residues (Y1234 and Y1235) in the kinase domain followed by auto-phosphorylation of two tyrosine residues (Y1349 and Y1356) in the C-terminal region.
Phosphorylation of Y1349 and Y1356 creates a multisubstrate docking site that is necessary for the induction of downstream signaling cascades such as Ras/Raf/MAPK, PI3K/AKT/mTOR, and/or STAT3/5. These signaling cascades drive distinct biological responses including cell growth, inhibition of apoptosis, cell migration and angiogenesis. 6 MM cells express c-met and often simultaneously express HGF, thus creating an HGF/c-met autocrine loop. HGF is also secreted by BM stromal cells, which provides an additional paracrine loop to stimulate tumour-stromal crosstalk that favours MM cells growth and metastasis. 7 Clinically, c-met is not mutated in MM, but MM patients have high serum levels of HGF and high c-met expression and gene copy number, which are correlated with poor prognosis and advanced disease. [8][9][10][11] It has been demonstrated that abnormal activation of the HGF/c-met pathway supports MM cell survival, growth, angiogenesis, osteolytic lesions and drug resistance. 5,6 Thus, the HGF/c-met interaction has recently emerged as a promising target in MM therapy.
Recently, several antibodies/agents that interfere with HGF/c-met signaling have entered preclinical or clinical trials including ligand antagonists (monoclonal antibody), 12 receptor inhibitors (monoclonal antibody) 13 and receptor kinase inhibitors. 6 However, inherent limitations of these antibodies/inhibitors, 14,15 such as cellular cytotoxicity or off-target effects, limit their clinical use and prompted the development of a new class of therapeutic antagonists, namely, aptamers.
Aptamers are single-stranded oligonucleotides that are isolated from RNA or ssDNA libraries via systematic evolution of ligands by exponential enrichment (SELEX). 16 Similar to antibodies, aptamers bind to their targets with high affinity and selectivity due to their unique three-dimensional structures. However, aptamers are advantageous over antibodies due to their low potential for immunogenicity, efficient tissue penetration, relatively simple synthesis, etc. 17 To date, a small number of aptamers have been developed as therapeutic antagonists in MM, 18,19 but none target c-met.
Recently, DNA aptamer CLN0003 (CLN3) was isolated from Jurkat cells via Cell-SELEX and was found to bind c-met with high specificity and affinity. 20 Ueki et al identified the 50-mer minimal binding motif of CLN3 (SL1) that retained high c-met affinity and interfered with HGF binding to c-met in SNU-5 cells. 21 However, whether SL1 can become the first aptamer to target c-met in MM requires further investigation. In this work, we characterized the clinical significance of c-met in MM and studied the selectivity and binding properties of SL1 in MM via a series of in vitro, in vivo and ex vivo assays. Furthermore, we showed that SL1 has the potential for treating clinical MM cells that express CD138, a hallmark of malignant PC. Furthermore, we show that SL1 can be used in combination with the first-line drug, bortezomib (BTZ). In all, our data support SL1 as a promising molecular tool for developing new MM treatments.

| Western blot analysis
As described previously, 22 cells were lysed with RIPA buffer (Beyotime, Shanghai, China) that contained a protease and phosphatase inhibitor mixture (Roche, Mannheim, Germany) and cells membrane protein were extracted by membrane and cytosol protein extraction kit(P0033; Beyotime). Protein extraction (50 μg) were boiled and subjected to a 10% SDS-PAGE gel followed by immunoblotting with specific antibodies.

| Migration assay
As described previously, 23

| Adhesion assay
Cells were first labelled with the fluorescent dye Dio (Sigma-Aldrich, Burlington, VT, USA) for 1 hour at 37°C and were then washed with PBS three times. Cells were co-cultured directly with HS5 cells in 96-well plates (5 × 10 4 cells/well) in the presence of the 4 μM of aptamer for 4 hours at 37°C. Subsequently, the non-adherent cell fraction was removed by washing with PBS, and the remaining adherent cells were solubilized performed with 1% Triton X-100 (50 μL/well). Fluorescence intensity at 501 nm was measured performed with a multi-label plate reader.  After in vivo imaging, mice that had been injected with Cy5labelled SL1 or a control library were killed by cervical dislocation under narcosis at 1 hour post-injection. After anatomization, the dissected organs including liver, kidney, spleen, lung, heart and tumour tissue were imaged with the IVIS Lumina II in vivo imaging system.

| In vivo and ex vivo fluorescence imaging
A 640 nm (±15 nm) bandpass filter and a 695-770 nm bandpass filter were selected as the excitation filter and the emission filter respectively.

| Statistical analysis
Data are shown as the mean ± SD for three independent experiments. Statistical significance between groups was analysed by Student's t test. One-way analysis of variance and Fisher's least significant difference test were assessed performed with GraphPad Prism 7.0 (GraphPad software), and the results were used to compare different groups from the GEP dataset. P < 0.05 was considered significant.

| C-met is highly expressed in MM and correlates with poor patient outcomes
To assess whether c-met inhibition can be used for MM treatment,  Figure 1C). Therefore, c-met can be considered an ideal target for aptamerbased therapeutics in MM.

| SL1 in vitro binding specificity and affinity
To evaluate the binding specificity of SL1 for native cellular c-met,

| Fluorescence imaging of SL1 in vivo and ex vivo
To study whether SL1 retained binding specificity in vivo, Cy5-labelled SL1 or a Cy5-labelled control library was intravenously injected into ARP-1-bearing NCG mice for in vivo fluorescence imaging. After injection, the spatial and temporal biodistribution of the aptamer was determined at 10, 30, 60, 120 and 180 minutes. The results show that high levels of fluorescent signal could be observed in tumour sites from 10 to 60 minutes post injection, followed by a gradual decrease in signal thereafter. However, the mice injected with Cy5-labelled SL1 showed higher fluorescent signal than mice injected with the Cy5labelled control library, indicating that SL1 effectively targets MM tumours in vivo with high selectivity and sensitivity ( Figure 3A).
Whole body distribution of SL1 was assessed via ex vivo fluorescence imaging. Dissected tumour tissues and organs including heart, lung, spleen, liver and kidney were obtained 60 minutes after injecting Cy5-labelled SL1 or the Cy5-labelled control library. Imaging revealed high SL1 accumulation in tumour sites. A fluorescent signal was also observed in kidney tissue, which suggests that SL1 is excreted and cleared by the kidneys (Figure 3B). These results suggest that SL1 targets c-met expressing tumours in vivo and that SL1 could serve as a potential molecular probe for MM diagnosis and therapy.

| SL1 inhibits MM cell growth in vitro in an HS5 co-culture model
In addition to binding c-met, we further explored the therapeutic potential of SL1 in MM performed with an in vitro co-culture model.
Given the supportive role of BMMEs in disease progression in malignant PC, a co-culture system performed with the human BM stromal cell line HS5 was explored. Co-culture with HS5 stromal cells was used to model HGF production by the BMME in vitro. First, the proliferation of B cells, MM1.S cells and ARP-1 cells in the presence of SL1 and HS5 stromal cells was evaluated. As shown in Figure 4A, SL1 inhibited the proliferation of two MM cell lines in a dose-dependent manner, particularly at 2 and 4 μM but had almost no effect on To further understand the anti-proliferative effects of SL1, we investigated the impact of SL1 (4 μM) on cell cycle and apoptosis at 48 hours via PI-staining and annexin V-FITC/PI staining assays respectively. As shown in Figure 4B, after serum deprivation for 24 hours to allow for cell cycle synchronization and then re-feeding with 10% serum in a co-culture system for 48 hours, SL1 treated MM.1S or ARP-1 cells showed a decrease in the G2/M phase population compared to library-treated control cells. Accordingly, SL1treated cells showed a significant increase in the percentage of cells in the G0/G1 phase. Moreover, cell apoptosis was dramatically F I G U R E 4 Effects of SL1 on MM cell proliferation, cell cycle and apoptosis in an HS5 co-culture model. Cells were plated in the upper chamber of a Transwell co-culture system with the specific concentration of SL1 indicated or a DNA control library (4 μM). A, The co-culture continued for 24, 48 or 72 hours. Cells were harvested and assessed for cell viability by manual cell counting. Data are presented from three independent experiments as the mean ± SD, *P < 0.05. B, Cells were synchronized at G0/G1 phase by serum starvation for 24 hours. After coculturing with HS5 cells for 48 hours, cells were harvested, stained with PI, and cell populations were sorted by flow cytometry into G0/G1, S and G2/M phases. Data are shown as the mean ± SD, P < 0.05. C, After co-culturing with HS5 cells for 48 hours, cells were harvested, stained with annexing V-FITC/PI, and the percentage of apoptotic cells was determined by flow cytometry. Data are shown as the mean ± SD from three independent experiments, *P < 0.05. D, Serum stability assays were performed to assess the stability of SL1 in cell medium containing 10% FBS. While SL1 gradually degraded after 1 hour, but it was still detected after 72 hours in 10% FBS increased in SL1-treated cells compared to library-treated control cells ( Figure 4C). These data indicate that inhibition of cell cycle progression and induction of cell death both contribute to the observed SL1-mediated anti-proliferative effect.

| SL1 prevents cell migration and adhesion in an HS5 co-culture model, and suppresses HGFinduced activation of c-met signaling in vitro
Multiple myeloma has a disseminated growth pattern throughout the BM and is dependent on the migration of cells across endothelial barriers and on adhesion to other cells and BMME components. 24 As HGF/c-met signaling elicits MM cell migration and adhesion, we investigated the effects of SL1 on MM cell migration and adhesion via Transwell migration and cell adhesion assays in a co-culture system. As shown in Figure 5A, HS5 cells were added to the bottom compartment of transwell chambers to give a confluent monolayer.
The migration of cells across an 8 μm pore filter was significantly inhibited in the presence of SL1 relative to library treated control cells. In addition, SL1 treatment significantly prevented adhesion of MM.1S or ARP-1 cells to a confluent monolayer of HS5 cells (Figure 5B).
It is known that HGF/c-met signaling promotes the adhesion and migration of myeloma cells by stimulating multiple downstream pathways including PI3K/AKT and Ras/ERK. 25 To confirm that SL1-mediated effects are a result of HGF/c-met pathway inhibition, we investigated the effects of SL1 on the expression levels of p-c-met (Y1349), p-ERK and p-AKT after stimulating with HGF. The results show that HGF-stimulated phosphorylation of c-met, ERK and AKT was dramatically reduced in the presence of SL1, while their total protein levels remained unchanged (Figure 5C). Therefore, SL1 blocks HGF/c-met signaling and the phenotypic effects that typically result from HGF/c-met pathway stimulation.
F I G U R E 5 Effects of SL1 on MM cell migration and adhesion in an HS5 co-culture model and HGF-induced c-met signaling. A, HS5 cells were pre-seeded in the bottom wells of Transwell migration chambers for 24 hours. Then, MM1.S or ARP-1 cells were seeded in the top wells of chambers containing serum-free media with 4 μM of SL1 or 4 μM of library control. After 48 hours, cell migration was measured as described in the Section 2. Data are means ± SD from three independent experiments, *P < 0.05. B, Adhesion of MM1.S or ARP-1 cells to HS5 monolayers was evaluated by a cell adhesion assay as described in the Section 2. Data are means ± SD from three independent experiments, *P < 0.05. C, ARP-1 cells were pre-stimulated with HGF (110 pmol/L) for 20 minutes before the addition of 4 μM of SL1 or 4 μM of library control for 48 hours. After 48 hours, expression analysis was performed for p-c-met/c-met, p-AKT/AKT and p-ERK/ERK by Western blotting. GAPDH served as the loading control. Right panel: densitometric analysis and statistical analysis (n = 3 independent experiments) of protein bands performed with ImageJ software. *P < 0.05 versus the loading control 3.6 | Aptamer SL1 specifically recognizes and suppresses primary CD138 PC from MM patients CD138 is a well-known surface antigen for MM and PC in BM. 26 To determine whether SL1 can recognize MM cells from clinical specimens, BM samples from patients (n = 4) were stained with CD138-APC antibody and co-stained with FAM-labelled SL1 or the DNA control library. The results show that SL1 specifically binds CD138+ cells but not CD138-cells ( Figure 6A).
We isolated primary CD138+ cells from the BM of MM patients

| DISCUSSION
The development of MM depends not only on oncogenic events occurring in MM cells but also on the extracellular BMME, which plays a key role in MM cell growth, survival, homing and drug resistance.
"Normalizing" the tumour microenvironment or inhibiting communication between MM cells and their surrounding microenvironment is an important therapeutic strategy. 28 HGF/c-met signaling is a key pathogenic factor in the BMME. HGF is frequently present in the BMME where it activates the c-met receptor in MM cells in a paracrine or autocrine fashion. 5 We evaluated the mRNA expression of c-met in two independent publicly available data sets and showed that c-met  29 Their potential immunogenicity and high production costs are of particular concern for this modality.
In addition, selective c-met small-molecule inhibitors have failed to gain approval due to adverse effects, dose-limiting toxicities or acquired resistance; therefore, novel approaches that suppress HGF/ c-met signaling are needed. 29 In the last decade, aptamers have emerged as attractive alternatives to antibodies and small molecules and have found use in diagnostic, therapeutic, imaging and targeting applications. 30 A handful of aptamers have been developed to function as therapeutic agonists in cancer. For example, DNA aptamers targeting nucleolin, PDGF, and HER2 and RNA aptamers targeting VEGF, HER3, and EGFR have been explored. 16 In myeloma, two aptamers have been developed as antagonists to inhibit the interaction between protein with its ligand. NOX-A12, an RNA oligonucleotide with L-configuration that binds and neutralizes the C-X-C motif chemokine ligand 12 (CXCL12), is one such aptamer. NOX-A12 has been shown to decrease tumour metastasis and drug resistance caused by cancer cell homing. 31 A CD38-specific ssDNA aptamer drug conjugate (ApDC) targets drug delivery to CD38expressing MM cells and releases the drug payload intracellularly. 18 However, the use of this ApDC to target c-met in MM has yet to be reported. The aptamer CLN3 is the only known c-met-binding DNA aptamer. The minimal binding domain of CLN3, SL1, was found to retain high binding affinity for c-met and blocked the HGF/c-met interaction and c-met signaling in SNU-5 cells. 21 In this study, we Importantly, we found that SL1 has a much greater affinity for CD138+ cells than for CD138-cells isolated from the BM of MM patients. Furthermore, SL1 showed antiproliferative effects on primary CD138+ cells. CD138 is a specific surface marker of MM PC. As a co-receptor, CD138 strongly promotes HGF-induced signaling by promoting the dimerization of c-met. 33 c-met mRNA and protein expression were found to be higher in CD138+ cells than in CD138-cells. 11 These results support the potential clinical use of SL1 by only targeting MM cells as opposed to healthy cells.
SL1 and BTZ were found to act synergistically in vitro as a combination of SL1 and BTZ gave enhanced inhibition of MM cell proliferation compared to either compound alone. In the clinic, c-met expression levels can affect a patient's response to BTZ. For example, higher c-met levels are associated with poor response and outcome in myeloma patients treated with BTZ-based therapies. 11 Knockdown of c-met by shRNA in vitro increases sensitivity to BTZ in MM U266 cells. 34 Furthermore, SU11274, a novel selective c-met inhibitor, is known to induce apoptosis and necrosis, and it can reverse BTZ resistance in R5 cells. 27 Our results suggest that SL1 is comparable to other BTZ-based combination treatments for MM. SL1 in combination with BTZ significantly increased BTZ's therapeutic activity against MM and has the potential to improve patient survival.
In conclusion, SL1 is the first c-met-specific DNA aptamer with potential application in targeting and treating MM. SL1 can serve not only as a molecular probe to selectively recognize cellular c-met with high affinity in vitro and in vivo, but it can also serve as a therapeutic antagonist for c-met positive MM cell lines and primary CD138+ MM cells from clinical samples through the inhibition of HGF-induced c-met signaling. Furthermore, in combination with SL1, BTZ exhibits increased cytotoxicity. Given that no other aptamer has been explored as a c-met-targeting antagonist in MM, our data warrant further clinical development of this novel therapeutic aptamer of c-met.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE
As for the experiments on the use of human clinic samples were col-

CONFLI CT OF INTEREST
The authors confirm that there are no conflicts of interest.