PCAT‐1 promotes cell growth by sponging miR‐129 via MAP3K7/NF‐κB pathway in multiple myeloma

Abstract Loss of one or some specific miRNA‐mediated regulation is closely associated with malignant progression of multiple myeloma (MM). But how these miRNAs work and what role the specific miRNA plays in this process of malignant progression remain unclear. It was found in this study that the expression of miR‐129 was decreased in both MM cell lines and newly diagnosed MM patients. Further clinicopathological statistics showed that miR‐129 was correlated with the isotype of MM patients. MiR‐129 overexpression disturbed cell proliferation, cell cycle evolution and spurred apoptosis both in vitro and in vivo. MAP3K7, a kinase able to activate NF‐κB circuit, was found to be up‐regulated in MM and contain a binding target of miR‐129. In addition, lncRNA PCAT‐1 functioned to sponge miR‐129 and thereby lowered its expression. PCAT‐1 knockdown eliminated the tumour‐promoting effect caused by miR‐129 inhibition, probably through repressing MAP3K7 and subsequent NF‐κB activation. To the best of our knowledge, this is the first study to have discovered that increased expression of PCAT‐1 could augment cell proliferation and cycle procession and inhibit apoptosis by down‐regulating miR‐129 via the MAP3K7/NF‐κB pathway in MM.

phenotypes in MM, including proliferation, drug resistance, apoptosis and metastasis. Besides, some miRNAs involved in the regulation of several key tumour-related proteins (p53) or cytokines (IGF-1, NF-κB) could function as therapeutic targets in MM. 5 MiR-129 is a recently discovered tumour-associated miRNA found to be dysregulated in cancers. Diao et al 6

demonstrated that
miR-129 combined with TGIF2 could disturb glioma progression. In the case of glioblastoma multiforme, miR-129 blocked the ability of cell proliferation, invasion, migration, neurosphere formation, angiogenesis and temozolomide resistance via Wnt5a by blocking the PKC/ERK/NF-κB and JNK pathways. 7 In prostate cancer (PCa), miR-129 also played a tumour-suppressive role by inhibiting cell proliferation, invasion and migration via targeting ETS1. 8 Except for the above reports that identified miR-129 as a tumour suppressor, miR-129 was also documented to have oncogenic characteristics when targeting lamin A in triple-negative breast cancer. 9 In addition, elevated miR-129-induced ERβ loss in colorectal cancer (CRC) facilitated cell proliferation and migration. 10 The contradictory characteristics of miR-129 in different types of tumours add complexity to the research on miR-129, and its biological function along with the underlying molecular mechanisms remains largely unknown in MM.
In the present study, we targeted miR-129 and further explored its biological function and clinical application in MM. To explore its cellular mechanism, we artificially up-or down-regulated miR-129 in MM samples and cell lines, and found that miR-129 expression enhancement diminished cell proliferation and cell cycle progression, and spurred apoptosis both in vitro and in vivo.
Notably, we showed that MAP3K7 along with NF-κB signal could affect cell division and apoptosis under the regulation of miR-129.
Additionally, we identified that lncRNA PCAT-1 could potentially function as a competing endogenous RNA (ceRNA) of miR-129 to augment MAP3K7. These observations may provide novel targeted therapies for MM. The normal bone marrows from healthy donors were collected as controls. Following identification of all the samples (prior to initiating the treatment), the local ethics committee issued the approval of the protocols involved with requisite consent from the patients.

| RNA extraction and real-time PCR
Blood samples of patients and healthy controls were collected using plastic gel vacuum blood collection tubes and centrifuged at 1200 g for 10 minutes. The separated serum was placed in RNase-free centrifuge tubes and stored at 80°C for use. Serum RNA was obtained using the serum RNA extraction kit (Life Technologies), while total RNA in cells was lysed within TRIzol reagent (Takara). Reverse transcription to obtain complementary DNA was conducted with the reverse transcription kit (Thermo Fisher Scientific) in accordance with the prescribed protocol. All amplified procedures were performed in ABI 7500 PCR Detection System (ABI). The primers involved in this study were as follows: PCAT-1: F5′-GAGAGCTGACATAGGCACCC-3′   and  R5′-TCTCCACTGGTGTTCATGGC-3′;  GAPDH:  F5′-TGATGACATCAAGAAGGTGGTGAAG-3′  and  R5′-TCCTTGGAGGCCATGTGGGCCAT-3′. TaqMan miRNA assays were utilized for analysing miRNAs in adherence to prescribed procedures (RiboBio). All assays were done in triplicate. Relative quantitative method (2 −ΔΔCt ) was utilized to calculate the expression.

| Cell transfection
LncRNA PCAT-1 vectors, including the empty vector, PCAT-1 overexpression vector, PCAT-1 negative control and shRNA were synthesized and purchased from Shanghai GenePharma Co., Ltd. The vectors used for miR-129 overexpression and knockdown were synthesized by RiboBio. Six-well plates were used to seed 1 × 10 6 cells per well followed by transfection with the vectors packaged by Lipofectamine 3000 Reagent (Invitrogen) in adherence to the prescribed procedures.

| Measurement of cell proliferation
The cell proliferation ability was measured by Cell Counting Kit-8 (Beyotime). 3000 MM cells/well were seeded in a 96-well plate and cultured for 2 days, to which 10 μL CCK-8 reagents were added.
Following 2-hours incubation at 37°C, absorbance at 450 nm was recorded for each well using 650 nm as reference. Each group had 5 biological replicates, and each experiment was repeated three times.

| Cell cycle and apoptosis assay
After 48-hours transfection, a PE Annexin V apoptosis detection kit (BD Pharmingen) was used for FACS in adherence to the prescribed procedures. Propidium iodide cell cycle detected kits of the same company were used to analyse the proportion of cells in each stage according to the prescribed procedure. All assays were conducted in triplicate.

| Soft agar colony formation assay
Approximately 1000 cells per well diluted with RPMI-1640 medium plus 20% FBS, along with 3% low melting point agarose solution, were seeded in 6-well plates. Following 10-day culture, the number of colonies defined as >50 cells/colony was counted.

| Immunoblot analysis
Following extraction of total protein with RIPA, the lysate containing 1% PMSF was separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to the PVDF membrane, followed by application of 5% BSA or skin milk for blocking. The appropriate primary antibody was applied at 4°C overnight followed by the secondary antibody for 2 hours. Protein bands were detected using enhanced chemiluminescence (ECL, Amersham Pharmacia). Triplicate assays were conducted.

| RNA immunoprecipitation
The Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit (Millipore) was used in adherence to the prescribed procedures.

TA B L E 1 Correlations between the relative expression of serum miR-129 and clinicopathological features in 34 MM patients
While AGO2 antibody (Cell Signaling Technology) was utilized, the co-precipitated RNA was verified. AGO2-specific binding was confirmed with the use of controls for total RNAs (input controls) and IgG in simultaneous assays.
Transfection of psiCHECK-2-PCAT-1-wt or psiCHECK-2-PCAT-1-mut was done along cotransfected with miR-129-5p mimics or control with Lipofectamine 3000 (the procedure has already been mentioned above). A 96-well plate was used to seed 5000 HEK-293T cells per well. Following 48-hours transfection, Dual-Luciferase Reporter Assay System (Promega) was applied by adherence to the prescribed procedures. Similar to the above procedures, the putative miR-129-5p sequence for binding the 3′-UTR of MAP3K7 and MAPK1 was used for construction of vectors and transfection.
Triplicate assays were conducted.

| Xenograft construction in nude mice
Pathogen-free conditions were applied to maintain BALB/c female nude mice (4 weeks old) with the approval from the Animal Care Committee of Nantong University where the study was performed.
Approximately 3 × 10 6 U266 cells transfected with miR-129 mimics or NCs were injected into the nude mice subcutaneously. Tumour growth was monitored every 3 days, while the tumour volume was determined by length × width 2 × 0.5. Euthanization was done 35 days later, and at the same time, excision of tumours was done so as to image and get tissue sections for subsequent immunohistochemistry.

| Immunohistochemistry study
Tumour samples were paraffin-embedded and stained to observe the relative expression and location of the target proteins using the avidin-biotin-peroxidase approach. The degree of tumour proliferation was evaluated by observing Ki-67, and apoptosis was evaluated by caspase-3 under a microscope (Olympus) at 200× or 400×.

| Statistical analysis
Data analysis was completed using SPSS software 18.0 and GraphPad Prism 6.0. All data are expressed as the means ± SD obtained in triplicate assays. Mann-Whiney U test was used to compare the differences between two groups. Kruskal-Wallis H test was used for multiple comparisons between the groups. P < .05 was considered statistically significant.

| miR-129 is decreased significantly in MM patients and cells
We compared miR-129 expression in fresh bone marrow obtained from 16 MM patients with that from 7 healthy controls and found that miR-129 expression was significantly lower than that in the healthy controls ( Figure 1A). Additionally, similar to the bone marrow samples, the serum miR-129 level in the 34 MM patients was lower than that in the healthy controls ( Figure 1B). Serum miR-129 expression was also associated with the clinicopathological features of the MM patients. Statistics showed that there was no significant difference in serum miR-129 in terms of sex, age and the ISS stage in the MM patients (P > .05), but there was significant difference between different isotopes (P = .024) ( Table 1). Furthermore, we also detected miR-129 expression in MM cell lines vs. normal bone marrow-derived plasma cells (nPCs) and found that miR-129 expression was lower in MM cells ( Figure 1C).

| Overexpression of miR-129 reduces cell division, arrests cell cycle and induces cell apoptosis
The role of miR-129 in MM was explored through the transfec-

| miR-129 overexpression suppresses tumour growth in mice
To To summarize, an inhibitory effect on tumour growth was observed in the nude mouse xenograft model following miR-129 overexpression.

| MAP3K7 directly interacts with miR-129
Several bioinformatics databases and software systems concerning TargetScan, microRNA.org and DIANA-miT were utilized to scan for the targets of miR-129. MAP3K7 and MAPK1 were thought to be the targets due to their match in the miRNA "seed sequence" ( Figure 4A). Then, enzyme reporter assay was applied to confirm their interaction. It was found that the enzyme activity was inhibited significantly at a higher amount of miR-129 when the 3′ UTR of MAP3K7 was wild-type as against the mutated sequence (P < .05) ( Figure 4B). Additionally, MAP3K7 was inhibited markedly both at the transcriptional and translational levels by miR-129 overexpression in NCI-H929 and U266 lines. An opposite tendency was observed with higher RNA and protein levels of MAP3K7 when miR-129 was suppressed (P < .05) ( Figure 4C-D).
The observations were suggestive of modulation of MAP3K7 by miR-129.

| The NF-κB pathway is involved in MAP3K7 and miR-129 functions in MM
To examine the mechanisms of the expressed biological function, we analysed signalling pathways through pathDIP to find the significantly enriched pathways of MAP3K7 gene. As shown in Figure 5A, several pathways were associated with MAP3K7, including the "NF-κB pathway" and "MAPK pathway," and activated TAK1 mediated "p38 MAPK activation," "WNT signalling" and "SAPK/JNK signalling," suggesting that MAP3K7 played a role in activating NF-κB by acting on IKK complex via phosphorylation, causing ubiquitination degradation of NFKB1A or IκBα, resulting in freeing NF-κB that was translocated to the nucleus after activation. 12 We therefore inferred that this pathway may be inhibited by miR-129 overexpression. Western blot showed that

| PCAT-1 acts as a sponge of miR-129
Research has shown that most lncRNAs work as a molecular sponge to reduce the ability of miRNAs to bind with their targets and the miRNA activity at a post-transcriptional level. According to the online software MicroInspector (http://bioin fo.uni-plovd iv.bg/micro inspe ctor/), we observed that PCAT-1 contained predicted targeting sites of miR-129 ( Figure 6A). Knowing that Ago2 protein is a vital part of the RNAinduced silencing complex (RISC), we conducted a RIP experiment in U266 lines using Ago2 antibody to investigate potential involvement of PCAT-1 and miR-129 complex. The precipitates were then checked by real-time PCR. The Ago2 fraction contained enrichment for PCAT-1 and miR-129 by a factor of 224 and 83, respectively, as compared with the controls in lieu of our expectations. This was suggestive of the presence of PCAT-1 in Ago2 RISC via sponging miR-129 ( Figure 6B). This was further confirmed by luciferase reporters with both wild-type and mutated sites for binding miR-129. The enzyme activity was lowered when miR-129 mimics were transfected for PCAT-1-wt while that of PCAT-1mut and vector controls showed no change ( Figure 6C). All these results suggest that PCAT-1 played a role in down-regulating miR-129.

| PCAT-1 is a modulator in miR-129/MAP3K7 regulatory network
The relative expression between miR-129 and PCAT-1 expressions was found to be inversely correlated in the serum samples of MM patients ( Figure 7A). Transfection of miR-129 mimics into U266 and NCI-

| The inhibitory effect of miR-129 is reversed by PCAT-1
As miR-129 expression could be modulated by PCAT-1, an association in terms of functional phenotypes was also examined.
Cotransfection with pcDNA-PCAT-1 and miR-129 mimic in U266 and NCI-H929 cells augmented the viability of cells and restored their growth as compared with mimics NC/pcDNA-NC ( Figure 8A).

| D ISCUSS I ON
The involvement of miRNAs in several malignancies including MM suggests their potential as biomarkers and therapeutic targets. [13][14][15][16] Yet, the miRNAs involved in MM have not been clearly defined. In this study, we for the first time demonstrated that the This protein is also found to regulate several pathways, such as JNK and NF-κB pathway, that are commonly reported to participate in carcinogenesis. 20 MAP3K7 inhibition is associated with the tumour survival rate. Hence, it is of interest to explore its role in therapeutic interventions. 21 It was found in our luciferase reporter assay that there was a binding site between MAP3K7 and miR-129, and an inverse correlation between MAP3K7 and miR-129 in the serum samples of MM patients, which discloses the role of MAP3K7 as a direct target of miR-129.
Earlier work 22,23 showed inhibition of MAP3K7 compromised tumour growth and metastasis in breast cancer and stalled the progression of leukaemia in an AML xenograft model. 24 In addition, MAP3K7 participated in the evolution of inflammation, tumour-associated immunoreactivity and apoptosis inhibition via NF-κB activation. 25 In fact, NF-κB was elevated in many malignancies (including MM) where its activation was associated with fast division, tumour growth, diminished apoptosis and wide drug resistance. 26 This study showed that overexpression of miR-129 markedly lowered the enrichment of p65 and p52 protein in the nucleus. To summarize, the involvement of MAP3K7 in controlling the NF-κB circuit, which was further regulated by miR-129, compromised carcinogenic manifestations.
The diminished expression of miR-129 in MM cell lines was further explored. MiR-129 was reported previously to be regulated by lncRNAs. 27,28 In this study, we for the first time demonstrated a direct link between PCAT-1 and miR-129 in MM and that the PCAT-1 level in MM-derived PCs was significantly higher than that in the control and negatively associated with the level of serum miR-129.
Altering the levels of PCAT-1 was pursued to explore its functions.
PCAT-1 overexpression was found to augment cell growth and division, while PCAT-1 inhibition produced an opposite effect, suggesting that PCAT-1 worked as an oncogene o in MM. Triplicate assays were conducted, mean ± SD, *P < .05 Previous studies 29,30 have reported a mode of lncRNA as ceRNA to silence miRNAs. RIP and luciferase reporter assay explored potential sites of PCAT-1-associated miR-129 and found a direct association between miR-129 and PCAT-1. It was also found that PCAT-1 knockdown or miR-129 mimics alone reduced the expression of MAP3K7, while combination of pDNA-PCAT-1 with miR-129 mimics restored the reduction of MAP3K7 expression significantly.
To summarize, we demonstrated that PCAT-1 worked a ceRNA to regulate MAP3K7 and negatively modulate miR-129 in MM.

CO N FLI C T O F I NTE R E S T
None.

AUTH O R CO NTR I B UTI O N S
XJS participated in cell culture and drafted the manuscript. QY and QQY carried out the immunoassays. HC and XDW carried out cell functional tests and PCR assay. SQJ and SK conceived, co-ordinated the study and analysed data. All authors reviewed and edited the manuscript and approved the final version.

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
The local ethics committee issued approval of the protocols involved with requisite consent from the patients. The animal experiments were applied according to the guidelines of the Animal Care Committee of Nantong University.