AKT1E17K‐Interacting lncRNA SVIL‐AS1 Promotes AKT1 Oncogenic Functions by Preferentially Blocking AKT1E17K Dephosphorylation

Abstract

AKT1^E17K is a gain‐of‐function mutation that constitutively activates the PI3K‐AKT pathway. However, how AKT1^E17K is regulated in cancer pathogenesis remains elusive. Here, RNA immunoprecipitation sequencing (RIP‐seq) is performed to interrogate the AKT1^E17K‐interacting lncRNAs and identify that SVIL‐AS1 preferentially binds to AKT1^E17K rather than AKT1^WT proteins. It is found that SVIL‐AS1 enhances AKT1 phosphorylation and downstream signaling. SVIL‐AS1 knockdown dramatically inhibits the growth of AKT1^E17K cells in vitro and in vivo. Notably, AKT1 and SVIL‐AS1 interaction is AKT1 phosphorylation‐dependent. SVIL‐AS1 also interacts with PPP2R2A, a subunit of phosphatase PP2A holoenzyme, and blocks the binding of PPP2R2A to AKT1^E17K to prevent AKT1 dephosphorylation. Moreover, AKT1^E17K cells are not effectively inhibited by the allosteric AKT inhibitor, whereas silencing SVIL‐AS1 sensitizes AKT1^E17K cells to AKT1 allosteric inhibitor, as well as the PI3Kα inhibitor. In breast cancer tissues, SVIL‐AS1 is highly expressed and associated with p‐AKT1 level and poor prognosis of patients. Together, the findings discover a novel lncRNA regulator of mutant oncoprotein which preferentially prevents AKT1^E17K dephosphorylation. Targeting SVIL‐AS1 may help to improve the responses to inhibitors of the PI3K‐AKT pathway, especially in AKT1^E17K mutant tumors.

Introduction

AKT1, also known as protein kinase B (PKB), is a key node in the PI3K-AKT pathway and regulates a variety of biological and pathological processes, such as cell proliferation, energy metabolism, and tumor progression.^[1,2] E17K is the most common mutation of AKT1, which is in the Pleckstrin Homology (PH) domain, the lipid-binding pocket of AKT1.^[3] This gain-of-function mutation alters the electrostatic interactions of the pocket and the lipid-binding capacity, with increased affinity to PI(3,4,5)P3 and broadened affinity to PI(4,5)P2, thereby constitutively targeting the PH domain to the plasma membrane and leading to abnormal activation of AKT1.^[4,5] Moreover, E17K mutation perturbates the interdomain interactions of the PH domain and the kinase domain, which disrupts the autoinhibition and results in the activation of AKT1 and downstream signaling.^[6]

AKT1^E17K occurs in breast cancer, bladder cancer, endometrial cancer, colorectal cancer and acute lymphoblastic leukemia, etc.^[7,8] In breast cancer, AKT1^E17K has been identified in 2% to 7% of patients and associated with poor prognosis.^[9] In non-small cell lung carcinoma, AKT1^E17K enhances K64 methylation by SETDB1, leading to its hyper-activation and promotes tumor development.^[10] However, the oncogenic potential of this mutation in different cell contexts is controversial. AKT1^E17K shows transforming capacity in fibroblasts in vitro and induces leukemia in vivo,^[4] whereas it is not sufficient for mammary epithelia cells to transform in vitro and in vivo,^[11,12] suggesting that AKT1^E17K may need cooperative genetic events to achieve its full oncogenic potential.

Long non-coding RNAs (lncRNA) are transcripts of more than 200 nucleotides which can form complex spatial structures.^[13] LncRNAs can bind to nucleic acids, proteins, and even lipids to exert their functions.^[13,14] Dysregulation of lncRNA promotes the occurrence and development of a wide range of tumors.^[15–19] Protein modifications and interactions can also be regulated by lncRNAs, as we and many groups have shown.^[20–22] More recently, lncRNA LINK-A has been revealed to regulate the phosphorylation of AKT1, which directly interacts with the PH domain of AKT1 and PIP3, promoting AKT1 activation in glioma and endometrial cancer.^[14,23,24] However, it remains unclear whether lncRNA can interact with mutant proteins in cancer to regulate tumor progression.

Here, we compared the lncRNAs that bind to wildtype and E17K mutant AKT1 protein and identified SVIL-AS1 as an AKT1^E17K-interacting lncRNA. We found that SVIL-AS1 could enhance the proliferation of cancer cells with AKT1^E17K mutation by interacting with PPP2R2A and preventing AKT1 dephosphorylation. Furthermore, we show that targeting SVIL-AS1 can sensitize AKT1^E17K cancer cells to AKT1 allosteric inhibitor and the PI3Kα inhibitor.

Result

Identification of lncRNAs Binding to the AKT1^E17K Mutant Protein

To identify lncRNAs binding to the AKT1^E17K mutant protein, we transfected Flag-tagged AKT1 wild-type (AKT1^WT) or E17K mutant (AKT1^E17K) plasmids into breast cancer cell line MDA-MB-231. RNA immunoprecipitation (RIP) and deep sequencing showed that several lncRNAs preferentially bound to the AKT1^E17K mutant protein (Figure 1A,B). The top 10 abundant lncRNAs enriched by AKT1^E17K were selected for validation using RIP followed by quantitative RT-PCR (RT-qPCR) in MDA-MB-231 and MCF-10A cells (Figure 1C,D). LncRNAs SVIL-AS1, AC107959.3, MINCR, PVT1, EXOC3-AS1, and NORAD were consistently enriched in the immunoprecipitates of Flag-AKT1^E17K, compared to Flag-AKT1^WT (Figure 1C,D).

[IMAGE OMITTED. SEE PDF]

To further identify the most important lncRNA in regulating AKT1^E17K function, we performed western blot by silencing these lncRNAs in MDA-MB-231, MCF-10A and AKT1^E17K knock-in derivative cells (MCF-10A AKT1^E17K)^[11] (Figure S1A–C, Supporting Information). Knockdown of SVIL-AS1, AC107959.3, EXOC3-AS1, and NORAD decreased the phosphorylation levels of AKT1 and its downstream PRAS40 in all three cells (Figure 1E–G). Among these four lncRNAs, SVIL-AS1 knockdown was the most dramatic one to suppress cell proliferation and colony formation of these three cells (Figure 1I,J). Notably, the impact of SVIL-AS1 knockdown on AKT1 phosphorylation and cell growth was more obvious in MCF-10A AKT1^E17K than in AKT1 wildtype cells (Figure 1E–J).

SVIL-AS1 Binds to AKT1^E17K and is Essential for AKT1^E17K Phosphorylation

To characterize the sequence of SVIL-AS1, we performed 5′- and 3′- rapid amplification of cDNA ends (RACE) in MCF-10A AKT1^E17K cells (Figure S1D, Supporting Information). Sanger sequencing identified SVIL-AS1 as a 1247 nt transcript which was 4 nt longer in 3′ end than previous reported sequence (Ensemble ID: ENST00000414457) (Table S1, Supporting Information). To find out the subcellular localization of SVIL-AS1, we extracted the nuclear and cytoplasmic RNA fractions. RT-qPCR showed that the majority of SVIL-AS1 was distributed in the cytoplasm of MCF-10A and MCF-10A AKT1^E17K cells (Figure S1E, Supporting Information). Fluorescence in situ hybridization (FISH) also showed the cytoplasmic localization of SVIL-AS1 in these cells (Figure S1F, Supporting Information).

To verify the interaction of SVIL-AS1 with AKT1, we performed RIP and found that the binding capacity of SVIL-AS1 to Flag-AKT1^E17K was significantly stronger than that to Flag-AKT1^WT (Figure 2A). RNA pulldown also confirmed this result (Figure 2B). Then we overexpressed or silenced SVIL-AS1 in MCF-10A AKT1^E17K, IHH4 (papillary thyroid carcinoma cell), and KU-19-19 (bladder carcinoma cell), which all had the AKT1^E17K mutation (Figure 2C,F). We found that SVIL-AS1 overexpressing cells had higher phosphorylation levels of AKT1 and its downstream PRAS40 in cells with AKT1^E17K mutation, while the alterations were not obvious in MCF-10A cells (Figure 2D,E). SVIL-AS1 silenced cells had lower levels of p-AKT1 and p-PRAS40 in cells with AKT1^E17K mutation (Figure 2G,H). These results demonstrated that SVIL-AS1 interacted with AKT1^E17K protein and was essential for AKT1^E17K phosphorylation.

[IMAGE OMITTED. SEE PDF]

SVIL-AS1 Knockdown Impairs the Growth of AKT1^E17K Mutant Cells In Vitro and In Vivo

We compared the impact of SVIL-AS1 on the growth of wild-type AKT1 and E17K mutant cells. Colony formation assay showed that SVIL-AS1 knockdown reduced the clonogenicity more obviously in the MCF-10A AKT1^E17K than in MCF-10A cells (Figure 3A). Similarly, silencing SVIL-AS1 remarkably reduced the colony formation of IHH4 and KU-19-19 cells (Figure 3A). In parallel, MTS assays demonstrated that SVIL-AS1 knockdown impaired the proliferation of cells with AKT1^E17K mutation more dramatically than wild-type cells (Figure 3B).

[IMAGE OMITTED. SEE PDF]

To further explore the function of SVIL-AS1 in vivo, SVIL-AS1 knockdown cells and control cells were injected into nude mice. In line with previous reports,^[25] KRAS G12V-transfected MCF-10A AKT1^E17K cells formed bigger tumors than MCF-10A cells (Figure 3C–E). The growth rate, tumor volume, and tumor weight were reduced in SVIL-AS1 knockdown MCF-10A AKT1^E17K xenografts and the reduction was more obvious in AKT1^E17K than in wildtype xenografts (Figure 3C–E). Knockdown of SVIL-AS1 also decreased the tumor growth of KU-19-19 xenografts (Figure 3F–H). The efficiency of SVIL-AS1 knockdown in vivo was confirmed by RT-qPCR and ISH (Figure S2A–C, Supporting Information). The immunohistochemistry (IHC) of Ki67 also confirmed that SVIL-AS1 knockdown reduced cell proliferation in vivo (Figure 3I,J). Moreover, IHC results showed that SVIL-AS1 knockdown could inhibit AKT1 phosphorylation in both MCF-10A (KRAS G12V-transfected) and KU-19-19 xenografts (Figure S2D, Supporting Information). These results indicated that SVIL-AS1 promoted cell proliferation, especially in cells with AKT1^E17K mutation in vitro and in vivo.

SVIL-AS1 Interacts with PPP2R2A to Prevent AKT1 Dephosphorylation

To investigate how SVIL-AS1 specifically affected the function of AKT1^E17K mutant, we compared the proteins immunoprecipitated by Flag-AKT1^WT and Flag-AKT1^E17K in control and SVIL-AS1 knockdown MCF-10A cells (Figure 4A). We sent the differential protein bands for mass spectrometry (MS) analysis and compared the MS-identified proteins with the AKT1-interacting protein lists from KEGG, BioGRID, and UniProt databases (Figure 4A; Tables S2, S3, Supporting Information). Notably, two proteins, PPP2R2A and PPP2R5E, were immunoprecipitated by Flag in SVIL-AS1 knockdown cells but not in control AKT1^E17K and AKT1^WT cells (Figure 4A,B), suggesting that SVIL-AS1 prevent the interaction between AKT1 and these phosphatase proteins. Next, we performed RNA pulldown using SVIL-AS1 and its antisense RNA as a negative control. The protein band (30–70 kDa) enriched in the SVIL-AS1 pulldown complex in MCF-10A AKT1^E17K cells was cut for MS identification (Figure 4C). PPP2R2A was again characterized in SVIL-AS1 pulldown complex in AKT1^E17K MCF-10A cells (Figure 4D; Table S4, Supporting Information).

[IMAGE OMITTED. SEE PDF]

PP2A phosphatase is a heterotrimer protein complex composed of a catalytic subunit (PP2A Ca/b), a scaffold subunit (PP2A Aa/b), and a regulatory subunit (PP2A B). PPP2R2A belongs to the regulatory subunit (PP2A B). It has been reported that PP2A is a major phosphatase of AKT1.^[26,27] To verify whether SVIL-AS1 also interact with PP2A subunit, we performed RNA pulldown and RIP in MCF-10A AKT1^E17K cells. Both these experiments confirmed the interaction of SVIL-AS1 with PPP2R2A (Figure 4E-F). We then performed Co-IP in MCF-10A AKT1^E17K cells using PPP2R2A antibody and found that more AKT1 bound to PPP2R2A after SVIL-AS1 knockdown (Figure 4G). Reciprocal immunoprecipitation using AKT1 antibody confirmed this result (Figure 4G). Moreover, PPP2R2A knockdown could reverse the reduction of AKT phosphorylation caused by SVIL-AS1 knockdown in MCF-10A AKT1^E17K cells (Figure 4H). These results indicated that SVIL-AS1 specifically affected the phosphorylation level of AKT1^E17K mutant via PPP2R2A.

To further determine the interaction region of SVIL-AS to AKT1 and PPP2R2A, we generated a series of SVIL-AS1 fragments and performed RNA pulldown. The 1–500 and 250–750 nt fragment could bind to AKT1, and 500–1000 and 750–1247 nt fragment could bind to the PPP2R2A (Figure 4I), which indicated that SVIL-AS1 250–500 nt was the AKT1^E17K-interaction region and 750–1000 nt was the PPP2R2A-interaction region. The above results demonstrated that SVIL-AS1 specifically bound to PPP2R2A and prevents the interaction of AKT1^E17K with PPP2R2A.

AKT1 and SVIL-AS1 Interaction is AKT1 Phosphorylation-Dependent

Previous studies demonstrated that AKT1^E17K mutation could improve the affinity of AKT1 to cell membrane which made it easily to be phosphorylated.^[5] We altered AKT1 phosphorylation levels in MCF-10A and MCF-10A AKT1^E17K cells by serum starvation, insulin stimulation, and treatment of AKT1 inhibitor, and then explored the interaction of SVIL-AS1 with AKT1 and AKT1^E17K proteins. As expected, AKT1 phosphorylation level increased upon insulin stimulation and the treatment of the ATP-competitive inhibitor GSK690693, while the AKT1 phosphorylation level decreased upon serum starvation and the treatment of AKT1 allosteric inhibitor MK2206 (Figure 4J). Then we performed RNA pulldown and found that the extent of SVIL-AS1 interaction with AKT1 was in line with the level of AKT1 phosphorylation (Figure 4K), suggesting that this interaction was AKT1 phosphorylation-dependent.

To further determine which phosphorylation site was critical for the interaction of SVIL-AS1 with AKT1, we transfected Flag-tagged AKT1 T308A, S473A, and T308A+S473A constructs into MCF-10A cells. The constitutively active myristoylated AKT1 (myr-AKT1) was used as a positive control (Figure 4L). RIP experiments using the Flag antibody showed that SVIL-AS1 was enriched most in the myr-AKT1 transfected cells, while T308 mutation could obviously impair the interaction of AKT1 and SVIL-AS1 (Figure 4M). This result indicated that T308 phosphorylation was important for the interaction of AKT1 and SVIL-AS1.

Targeting SVIL-AS1 Sensitizes AKT1^E17K Cells to AKT1 Allosteric Inhibitor and PI3Kα Inhibitor

Preclinical evidence suggests that cancers with the AKT1^E17K mutation are more resistant to AKT1 allosteric inhibitors.^[28] To detect whether SVIL-AS1 affected the response to AKT1 allosteric inhibitor, we performed MTS assays and found that SVIL-AS1 knockdown sensitized both MCF-10A and AKT1^E17K cells to AKT1 allosteric inhibitor MK2206, and the effect was more dramatic in AKT1^E17K cells (Figure 5A). Moreover, SVIL-AS1 knockdown could also sensitize cells to PI3Kα inhibitor BYL719, especially the AKT1^E17K cells (Figure 5B).

[IMAGE OMITTED. SEE PDF]

Similar to the results in vitro, combining SVIL-AS1 knockdown and MK2206 or BYL719 therapy induced a significant reduction in tumor growth compared with that of MK2206 or BYL719 alone in MCF-10A-Ras AKT1^E17K and KU-19-19 xenografts (Figure 5C–H). Immunostaining in these xenografts demonstrated that the percentage of Ki67⁺ cells in SVIL-AS1 knockdown combining with MK2206 or BYL719 therapy decreased obviously comparing to SVIL-AS1 knockdown (Figure 5I). The level of p-AKT1 (S473 and T308) also reduced in SVIL-AS1 knockdown combining with MK2206 or BYL719 therapy (Figure 5J,K). Therefore, targeting SVIL-AS1 could further sensitize tumors with AKT1^E17K mutation to AKT1 allosteric inhibitor and PI3Kα inhibitor.

SVIL-AS1 is Highly Expressed in Breast Cancer and Associated with p-AKT1 Levels and the Poor Prognosis of Patients

We next performed ISH in breast cancer tissues to assess the association of SVIL-AS1 with clinical and pathological parameters. SVIL-AS1 was detected in 95.45% (105/110) of breast cancer tissues, which are mainly in the cytoplasm of breast cancer cells (Figure 6A). SVIL-AS1 expression was significantly associated with T stage (p = 0.027), N stage (p = 0.021), recurrence (p = 0.014), and metastasis failure (p < 0.001) of breast cancer patients (Figure 6B). Univariate Cox regression analysis showed that SVIL-AS1 expression was a prognostic factor for overall survival (OS) (p = 0.042) and disease-free survival (DFS) (p = 0.003) in breast cancer patients (Figure 6C,D). Similar results were obtained by Kaplan-Meier analysis in our cohort (Figure 6E). In the TCGA breast cancer database, the cumulative 5-year OS and 10-year OS were 93.2% and 80.4% in the SVIL-AS1^high group, whereas 96.6% and 90.2% in the SVIL-AS1^low group (Figure 6F). Kaplan-Meier analysis also demonstrated that high level of SVIL-AS1 significantly correlated with reduced OS of breast cancer patients in TCGA dataset (Figure 6F). Moreover, we analyzed SVIL-AS1 expression in breast cancer subtypes and other cancer types. There was no significant difference between luminal and HER2+ subtypes, but triple-negative breast tumors showed higher SVIL-AS1 expression (Figure S3A,B, Supporting Information). In TCGA dataset, overexpression of SVIL-AS1 was observed in head and neck squamous carcinoma and hepatocellular carcinoma when compared to normal tissues (Figure S3C, Supporting Information).

[IMAGE OMITTED. SEE PDF]

Further, we detected AKT1 phosphorylation in breast cancer tissues by immunostaining. A positive correlation was found between the expression of SVIL-AS1 and the level of p-AKT1(S473) (p < 0.001) (Figure 6G). Both SVIL-AS1^highp-AKT1(S473)^high and SVIL-AS1^highp-AKT1(T308)^high were significantly associated with T stage, N stage, recurrence, and metastasis failure of breast cancer patients (Figure 6H–I). SVIL-AS1^highp-AKT1(S473)^high and SVIL-AS1^high p-AKT1(T308)^high were also associated with shorter DFS time (Figure 6J,K). However, the combination of SVIL-AS1 and p-AKT1 did not predict the outcomes of breast cancer patients better than SVIL-AS1 alone, indicating that SVIL-AS1 is a potent biomarker of patient outcome.

Discussion

AKT1^E17K mutation can enhance the affinity of AKT1 to cell membrane, making it easily to be phosphorylated.^[4,5] Previous studies focused on the oncogenic mechanism of AKT1^E17K of increasing membrane localization and promoting phosphorylation. However, whether the phosphorylation status of AKT1^E17K is regulated by other factors is largely unknown. In this study, we discover that lncRNA SVIL-AS1 specifically interacts with AKT1^E17K mutant, which prevents PP2A holoenzyme from dephosphorylating AKT1^E17K by blocking the binding of PPP2R2A to AKT1^E17K (Figure 7). To our knowledge, this is the first study demonstrating the dephosphorylation of AKT1^E17K is regulated by a lncRNA.

[IMAGE OMITTED. SEE PDF]

Only a small number of lncRNAs have been functionally characterized in the human genome. The role of SVIL-AS1 in biological and pathological processes is little known. We identify SVIL-AS1 from AKT1 RIP-seq and validate it as an AKT1^E17K-interacting lncRNA using RIP-qPCR and RNA pulldown. Moreover, we uncover that SVIL-AS1 enhances AKT1^E17K phosphorylation and downstream signaling, thus promoting the proliferation of AKT1^E17K cells in vitro and in vivo. These data elucidate the critical role of SVIL-AS1 in sustaining AKT1^E17K activation and facilitating tumor growth. Of note, we reveal that SVIL-AS1 can also bind to wild-type AKT1. However, the function of SVIL-AS1 is dependent on AKT1 phosphorylation. Thus, SVIL-AS1 shows higher binding capacity and oncogenic activity in AKT1^E17K cells than in AKT1^WT cells.

Phosphorylation at both T308 and S473 is required for full Akt activity.^[2] Dephosphorylation of AKT1 at Thr308 is mainly mediated by PP2A.^[26,27] Conformational change induced by dephosphorylation at T308 can expose S473 to phosphatases, thereby making it more susceptible to dephosphorylation as well.^[29] Thus, dephosphorylation of AKT1 by PP2A will lead to AKT1 inactivation. Access of PP2A to AKT1 can be restricted by proteins such as BTBD10,^[30] as well as AKT1 conformational change and intramolecular interactions.^[31,32] Here, our findings elucidate a previously unrecognized mechanism of lncRNA SVIL-AS1 in blocking the access of PP2A to phosphorylated AKT1, which prevents AKT1 dephosphorylation. LncRNAs have been demonstrated to bind to proteins and modulate the protein post-translational modifications, interactions, and activities.^[33] Unlike the regulation at protein levels, RNA biosynthesis costs much less nutrients and ATP.^[34] If lncRNAs can comply the same function as proteins to regulate post-translational modifications, it will be a more economical mechanism to regulate the cell signaling. Fast-growing cancer cells may take advantage of this mechanism and use more lncRNAs to fulfill such functions, thus optimizing the usage of cellular resources, although this notion still needs to be further validated and tested in different organisms.

The components of the PI3K-AKT pathway are important targets for drug development. Previous study showed that AKT1^E17K mutants were not effectively inhibited by allosteric AKT inhibitors.^[4,6] In our study, we found that SVIL-AS1 knockdown could decrease the IC50 of MK2206 more obviously in MCF-10A AKT1^E17K cells and significantly inhibited the proliferation of AKT1^E17K mutant subcutaneous tumors. We also tested the effects of knocking down SVIL-AS1 in combination with the most frequently used PI3K inhibitor BYL719 (Alpelisib). Knockdown of SVIL-AS1 could also increase the sensitivity of AKT1^E17K mutant cells to BYL719. These data suggest that SVIL-AS1 may serve as a novel therapeutic target toward tumors with AKT1^E17K mutation and reverse the resistance to allosteric AKT inhibitors and PI3K inhibitors. Recent translation studies have shown that lncRNAs can be knocked down by RNAi, antisense oligonucleotides (ASOs), CRISPR-based approaches, and small molecules that target RNA-protein interactions.^[35] In the future, technological advancements may enable SVIL-AS1 targeting in combination with PI3K-AKT1 inhibitors for cancer therapeutics.

In this study, breast cancer tissues were not categorized by AKT1 mutation status. Given the relatively low prevalence of AKT1 mutations in breast cancer, the majority of specimens are AKT1^WT tumors. Even so, high level of SVIL-AS1 significantly correlated with reduced survival of breast cancer patients in both TCGA dataset and our cohort, and SVIL-AS1 expression correlates with p-AKT1 (S473) in breast cancer tissues. This may be because SVIL-AS1 can also bind to AKT1^WT protein when AKT1 is phosphorylated, and SVIL-AS1 knockdown can reduce cell growth in AKT1^WT tumors, although the effect is not as dramatic as in AKT1^E17K tumors. Nevertheless, it remains crucial to distinguish AKT1^WT and AKT1^E17K status in further translational and clinical studies when targeting SVIL-AS1 in combination with PI3K-AKT1 inhibitors for more precise evaluation of treatment responses of cancer patients.

Together, our work discovered a novel lncRNA regulator of AKT1 dephosphorylation, which preferentially acts on the AKT1^E17K mutant to strengthen its oncogenic function. The development of strategies targeting SVIL-AS1 may help to improve the responses to inhibitors of the PI3K-AKT pathway, especially in AKT1^E17K mutant tumors.

Experimental Section

Patients and Tissue Samples

In our cohort, 110 breast cancer samples were collected from patients diagnosed between 2016 and 2018 in Sun Yat-sen Memorial Hospital, Sun Yat-sen University. The median follow-up time was 48 months (21–72 months). All tumor samples were collected from patients with informed consent, and approved by the Institute Research Ethics Committee of Sun Yat-sen Memorial Hospital, Sun Yat-sen University (SYSKY-2023-384-01).

Cell Cultures

MCF-10A, MDA-MB-231, KU-19-19, IHH4, and HEK293FT were obtained from the American Type Culture Collection (ATCC) and were cultured according to standard protocols. MCF-10A were cultured in DMEM/F12(Gibco, USA) supplemented with EGF (peprotech, USA), hydrocortisone (Sigma, USA), cholera toxin (Sigma, USA), insulin (Sigma, USA) and 5% horse serum (Invitrogen, USA). IHH4, HEK293FT, and MDA-MB-231 cell lines were cultured in DMEM (Gibco, USA) with 10% fetal bovine serum (Gibco, USA). KU-19-19 was cultured in RPMI1640 (Gibco, USA) with 10% fetal bovine serum (Gibco, USA). All the cell lines were cultured at 37 °C with 5% CO₂. The cells were free from mycoplasma contamination.

RNA Immunoprecipitation (RIP)

Cell lysates were collected with IP lysis buffer supplementary with protease inhibitors, phosphatase inhibitors, RNase inhibitors, and EDTA. The cells were lysed at 4 °C for 20 min and centrifuged at 4 °C at 12 000 g for 20 min. The supernatant was transferred to a new EP tube. 30ul were reserved for transfection efficiency testing, and 5% of the volume was taken as input storing at −80 °C. The remaining supernatant was divided equally and incubated separately with magnetic beads coupled with specific protein antibodies or control IgG. The incubation was carried out overnight at 4 °C with rotation. The magnetic rack was used for collecting magnetic beads. The beads were washed by 500ul of TBST for 3–5 times. RNA was extracted by TRIzol reagent and was used for sequencing or PCR.

Fluorescence In Situ Hybridization (FISH) and ISH

The probe of 5′digoxin-labeled SVIL-AS1 was TGCAGAGTAGGACTGACATAA. FISH was performed by fluorescent in situ hybridization kit (C10910, RIBOBIO, China) according to the manufacturer's instructions. Confocal images were acquired using a confocal laser scanning microscope (Zeiss LSM710, Germany). ISH was performed on paraffin-embedded sections with in situ hybridization detection kit I (POD, Boster Biological Technology, Wuhan, China). Briefly, after dewaxing and rehydration, the samples were digested with proteinase K and hybridized with the 5′digoxin-labeled probe at 55 °C overnight. The sections were then incubated overnight at 4 °C with anti-digoxin monoclonal antibody (Roche, Basel, Switzerland) and stained with DAB (3,3′-diaminobenzidine) and hematoxylin.

Rapid Amplification of cDNA Ends (RACE)

RACE assay was performed using HiScript-TS 5′/3′ RACE Kit (634 923, Clontech, Mountain View, CA) according to the manufacturer's instructions. The primers were used as follows: (3′ RACE: 5′- GATTACGCCAAGCTTTGCAGCGGGGTTAGGATGGACG AGG-3′, 5′RACE: 5′- GATTACGCCAAGCTTCTCTGGGGCTGGGCTGAGCAAATG C-3′).

RNA Pulldown

The biotin-labeled SVIL-AS1 was transcribed with MEGAscript T7 in vitro transcription kit (AMB13345, Invitrogen, USA) according to the manufacturer's instructions. To allow proper secondary structure formation, 5 pmol of biotinylated RNA in RNA structure buffer was heated to 95 °C for 2 min, then put on ice for 3 min, and left at room temperature for 30 min. The cell lysates were incubated with folded RNA in RIP buffer at 4 °C overnight. The Dynabeads Streptavidin magnetic beads (M280, Invitrogen, USA) were added to each binding reaction and incubated at RT for 1h. After being washed by RIP buffer for 5 times, the samples were then collected for mass spectrometry analysis or western blot.

Animal Experiments

All the in vivo experiments were approved and supervised by the Animal Ethics Committee of Sun Yat-sen Memorial Hospital, Sun Yat-sen University (AP20220152). Female BALB/c nude mice (4-5 weeks old) were purchased from Gempharmatech (Guangdong). All the mice were raised under standard conditions (20–26 °C temperature, 40–60% humidity) with a 12 h light/12 h dark cycle at the specific-pathogen-free (SPF) animal facility in the Laboratory Animal Resource Center of Sun Yat-Sen University.

MCF-10A (KRAS G12 V) sh-NC or sh-SVIL-AS1 cells, MCF-10A AKT1^E17K (KRAS G12 V) sh-NC or sh-SVIL-AS1 cells were injected into the mammary fat pads of nude mice at a dose of 2 × 10⁶ 100 µL⁻¹ containing 20% Matrigel. KU-19-19 sh-NC cells and KU-19-19 sh-SVIL-AS1 cells were subcutaneously inoculated into nude mice at a dose of 2 × 10⁶ 100µL⁻¹ containing 20% Matrigel. To induce Tet-on shRNA expression, mice were fed with doxycycline (2000 ppm) via diet.

In the treatment experiments using MK2206 or BYL719, MCF-10A-Ras of sh-NC or sh-SVIL-AS1 cells, KU-19-19 sh-NC or sh-SVIL-AS1 cells were all subcutaneously inoculated into nude mice at a dose of 5 × 10⁶ 100µL⁻¹ containing 20% Matrigel. To induce Tet-on shRNA expression, mice were fed with doxycycline (2000 ppm) via diet. The mice were given BYL719 (25mg kg⁻¹, q3d), MK-2206 (240mg kg⁻¹, q3d) or normal saline by intragastric injection. The tumor volumes (mm³) were calculated according to the formula width² × length.

Statistical Analysis

The statistical analysis was performed using GraphPad Prism. The association of SVIL-AS1 with clinicopathological features was analyzed by either χ2 test or Fisher's exact test. 2-tailed Student's t-test was used to analyze differences between two groups of variables. Univariate regression analysis was performed using the Cox proportional hazard model. Survival curves were performed by Kaplan-Meier survival analysis and the log-rank test. p-value < 0.05 was considered statistically significant.

Acknowledgements

This work was supported by grants from National Key Research and Development Program of China (2021YFA1300602), Natural Science Foundation of China (82073048 and 82472775), China Postdoctoral Science Foundation (2022M713608), Guangdong S&T Program (2023B1111030006), Key-Area Research and Development Program of Guangdong Province (2023B1111020007), and Guangdong Science and Technology Department (2023A1515111108, 2023B1212060013, 2020B1212030004, and 2022A1515010168).

Conflict of Interest

The authors declare no conflict of interest.

Author Contributions

J.W. and W.C. contributed equally to this work. M.L.L. and H.H. conceived ideas and designed the study; M.L.L. and X.L. supervised the study; J.W. and W.C. performed most of the experiments and analysis; J.W., R.Y., and Q.L. performed in vivo study; P.H. and X.H. collected the patients’ information; J.W. drafted the manuscript; M.L.L. edited the manuscript. All authors reviewed the manuscript.

Data Availability Statement

The RIP-seq data have been deposited in GEO and are publicly available on the date of publication. Accession number is GSE268685 with the secure token szwxqsomxxgjrwn for review.

References

G. Hoxhaj, B. D. Manning, Nat. Rev. Cancer 2020, 20, 74.

B. D. Manning, A. Toker, Cell 2017, 169, 381.

F. E. Bleeker, L. Felicioni, F. Buttitta, S. Lamba, L. Cardone, M. Rodolfo, A. Scarpa, S. Leenstra, M. Frattini, M. Barbareschi, M. D. Grammastro, M. G. Sciarrotta, C. Zanon, A. Marchetti, A. Bardelli, Oncogene 2008, 27, 5648.

J. D. Carpten, A. L. Faber, C. Horn, G. P. Donoho, S. L. Briggs, C. M. Robbins, G. Hostetter, S. Boguslawski, T. Y. Moses, S. Savage, M. Uhlik, A. Lin, Du J., Y. W. Qian, D. J. Zeckner, G. Tucker‐Kellogg, J. Touchman, K. Patel, S. Mousses, M. Bittner, R. Schevitz, M. H. Lai, K. L. Blanchard, J. E. Thomas, Nature 2007, 448, 439.

K. E. Landgraf, C. Pilling, J. J. Falke, Biochemistry 2008, 47, 12260.

C. Parikh, V. Janakiraman, W. I. Wu, C. K. Foo, N. M. Kljavin, S. Chaudhuri, E. Stawiski, B. Lee, J. Lin, H. Li, M. N. Lorenzo, W. Yuan, J. Guillory, M. Jackson, J. Rondon, Y. Franke, K. K. Bowman, M. Sagolla, J. Stinson, T. D. Wu, J. Wu, D. Stokoe, H. M. Stern, B. J. Brandhuber, K. Lin, N. J. Skelton, S. Seshagiri, Proc. Natl. Acad. Sci. USA 2012, 109, 19368.

Y. Cohen, B. Shalmon, J. Korach, I. Barshack, E. Fridman, G. Rechavi, Gynecol. Oncol. 2010, 116, 88.

M. S. Kim, E. G. Jeong, N. J. Yoo, S. H. Lee, Br. J. Cancer 2008, 98, 1533.

K. Kalinsky, F. Hong, C. K. Mccourt, J. C. Sachdev, E. P. Mitchell, J. A. Zwiebel, L. A. Doyle, L. M. Mcshane, S. Li, R. J. Gray, L. V. Rubinstein, D. Patton, P. M. Williams, S. R. Hamilton, B. A. Conley, P. J. O'Dwyer, L. N. Harris, C. L. Arteaga, A. P. Chen, K. T. Flaherty, JAMA Oncol 2021, 7, 271.

G. Wang, J. Long, Y. Gao, W. Zhang, F. Han, C. Xu, L. Sun, S. C. Yang, J. Lan, Z. Hou, Z. Cai, G. Jin, C. C. Hsu, Y. H. Wang, J. Hu, T. Y. Chen, H. Li, M. G. Lee, H. K. Lin, Nat. Cell Biol. 2019, 21, 214.

J. Lauring, D. P. Cosgrove, S. Fontana, J. P. Gustin, H. Konishi, A. M. Abukhdeir, J. P. Garay, M. Mohseni, G. M. Wang, M. J. Higgins, D. Gorkin, M. Reis, B. Vogelstein, K. Polyak, M. Cowherd, P. J. Buckhaults, B. H. Park, Oncogene 2010, 29, 2337.

M. L. Mancini, E. C. Lien, A. Toker, Oncotarget 2016, 7, 17301.

K. Nemeth, R. Bayraktar, M. Ferracin, G. A. Calin, Nat. Rev. Genet. 2024, 25, 211.

A. Bhan, M. Soleimani, S. S. Mandal, Cancer Res. 2017, 77, 3965.

R. A. Gupta, N. Shah, K. C. Wang, J. Kim, H. M. Horlings, D. J. Wong, M. C. Tsai, T. Hung, P. Argani, J. L. Rinn, Y. Wang, P. Brzoska, B. Kong, R. Li, R. B. West, M. J. van de Vijver, S. Sukumar, H. Y. Chang, Nature 2010, 464, 1071.

W. L. Hu, L. Jin, A. Xu, Y. F. Wang, R. F. Thorne, X. D. Zhang, M. Wu, Nat. Cell Biol. 2018, 20, 492.

E. Leucci, R. Vendramin, M. Spinazzi, P. Laurette, M. Fiers, J. Wouters, E. Radaelli, S. Eyckerman, C. Leonelli, K. Vanderheyden, A. Rogiers, E. Hermans, P. Baatsen, S. Aerts, F. Amant, S. Van Aelst, J. van den Oord, B. de Strooper, I. Davidson, D. L. Lafontaine, K. Gevaert, J. Vandesompele, P. Mestdagh, J. C. Marine, Nature 2016, 531, 518.

G. Li, I. Kryczek, J. Nam, X. Li, S. Li, J. Li, S. Wei, S. Grove, L. Vatan, J. Zhou, Du W., H. Lin, T. Wang, C. Subramanian, J. J. Moon, M. Cieslik, M. Cohen, W. Zou, Nat. Cell Biol. 2021, 23, 526.

A. M. Schmitt, J. T. Garcia, T. Hung, R. A. Flynn, Y. Shen, K. Qu, A. Y. Payumo, A. Peres‐Da‐Silva, D. K. Broz, R. Baum, S. Guo, J. K. Chen, L. D. Attardi, H. Y. Chang, Nat. Genet. 2016, 48, 1370.

Q. Guo, Y. Li, Y. Zhang, L. Shen, H. Lin, J. Chen, E. Song, M. L. Luo, EMBO J. 2023, 42, 112414.

M. L. Luo, J. Li, L. Shen, J. Chu, Q. Guo, G. Liang, W. Wu, J. Chen, R. Chen, E. Song, J. Natl. Cancer Inst. 2020, 112, 356.

L. Statello, C. J. Guo, L. L. Chen, M. Huarte, Nat. Rev. Mol. Cell Biol. 2021, 22, 96.

D. M. Hyman, L. M. Smyth, M. Donoghue, S. N. Westin, P. L. Bedard, E. J. Dean, H. Bando, A. B. El‐Khoueiry, J. A. Perez‐Fidalgo, A. Mita, J. Schellens, M. T. Chang, J. B. Reichel, N. Bouvier, S. D. Selcuklu, T. E. Soumerai, J. Torrisi, J. P. Erinjeri, H. Ambrose, J. C. Barrett, B. Dougherty, A. Foxley, J. Lindemann, R. Mcewen, M. Pass, G. Schiavon, M. F. Berger, S. Chandarlapaty, D. B. Solit, U. Banerji, et al., J. Clin. Oncol. 2017, 35, 2251.

L. M. Smyth, Q. Zhou, B. Nguyen, C. Yu, E. M. Lepisto, M. Arnedos, M. J. Hasset, M. L. Lenoue‐Newton, N. Blauvelt, S. Dogan, C. M. Micheel, C. Wathoo, H. Horlings, J. Hudecek, B. E. Gross, R. Kundra, S. M. Sweeney, J. Gao, N. Schultz, A. Zarski, S. M. Gardos, J. Lee, S. Sheffler‐Collins, B. H. Park, C. L. Sawyers, F. Andre, M. Levy, F. Meric‐Bernstam, P. L. Bedard, A. Iasonos, et al., Cancer Discovery 2020, 10, 526.

R. Li, J. Wei, C. Jiang, D. Liu, L. Deng, K. Zhang, P. Wang, Cancer Res. 2013, 73, 5742.

S. Padmanabhan, A. Mukhopadhyay, S. D. Narasimhan, G. Tesz, M. P. Czech, H. A. Tissenbaum, Cell 2009, 136, 939.

S. Resjo, O. Goransson, L. Harndahl, S. Zolnierowicz, V. Manganiello, E. Degerman, Cell. Signalling 2002, 14, 231.

B. T. Shrestha, T. Shamu, A. N. Gorelick, M. T. Chang, D. Chakravarty, E. I. Gavrila, M. Donoghue, J. Gao, S. Patel, S. P. Gao, M. H. Reynolds, S. M. Phillips, T. Soumerai, W. Abida, D. M. Hyman, A. M. Schram, D. B. Solit, L. M. Smyth, B. S. Taylor, Nat. Commun. 2022, 13, 2111.

M. Ebner, I. Lucic, T. A. Leonard, I. Yudushkin, Mol. Cell 2017, 65, 416.

M. Nawa, K. Kanekura, Y. Hashimoto, S. Aiso, M. Matsuoka, Cell. Signalling 2008, 20, 493.

T. O. Chan, J. Zhang, U. Rodeck, J. M. Pascal, R. S. Armen, M. Spring, C. D. Dumitru, V. Myers, X. Li, J. Y. Cheung, A. M. Feldman, Proc. Natl. Acad. Sci. USA 2011, 108, E1120.

T. O. Chan, J. Zhang, B. C. Tiegs, B. Blumhof, L. Yan, N. Keny, M. Penny, X. Li, J. M. Pascal, R. S. Armen, U. Rodeck, R. B. Penn, Biochem. J. 2015, 471, 37.

X. Lin, M. L. Luo, E. Song, Cell Insight 2022, 1, 100004.

Y. Liu, A. Beyer, R. Aebersold, Cell 2016, 165, 535.

M. Winkle, S. M. El‐Daly, M. Fabbri, G. A. Calin, Nat. Rev. Drug Discovery 2021, 20, 629.