The PI3K–AKT signaling pathway is frequently dysregulated in cancer, and it is hyperactivated in approximately 50% of breast cancers. Although inhibitors directly targeting the PI3K–AKT axis have been developed, clinical efficacy has been limited to only a subset of patients. Identification of mechanisms underlying AKT-driven tumorigenesis could lead to alternative approaches to block pathway signaling and suppress breast tumor growth. Mass spectrometry–based analyses demonstrated that salt-inducible kinase 1 (SIK1) binds AKT and undergoes AKT-mediated phosphorylation, which compromises SIK1 tumor-suppressive functions. As a result, AKT relieved the binding and repression of STAT3 by SIK1 in a phosphorylation-dependent manner, resulting in breast cell tumorigenesis. Following AKT-mediated phosphorylation, SIK1 interacted with 14-3-3 and was translocated to the cytoplasm where the isomerase Pin1 facilitated SIK1 interaction with the E3 ligase ITCH to promote SIK1 ubiquitination and subsequent degradation. These findings indicate that SIK1 is a substrate of AKT that links AKT oncogenic function to STAT3 activation, highlighting targeting of the JAK2–STAT3 axis as a strategy to treat AKT-driven breast cancer.

Significance:

AKT binds and phosphorylates SIK1 to overcome SIK1-mediated repression of STAT3, indicating that STAT3 is a potential therapeutic target in breast cancer with hyperactive AKT signaling.

Breast cancer is the leading cause of tumor incidence and mortality for women worldwide (1, 2). In recent decades, although the canonical molecular classification has led to successful breast cancer therapies mediated by targeting estrogen receptor (ER) or human epidermal growth factor receptor 2 (HER2; refs. 3, 4), drug resistance and the lack of efficient strategies to combat triple-negative breast cancer (TNBC) are major challenges to current breast cancer therapies. With the development of multiomics, researchers have increasingly identified novel breast cancer subtypes on the basis of their genomic drivers, including the PI3K–AKT–mTOR pathway compounds, growth factor receptors (e.g., ERBB2), cell-cycle regulators (e.g., CCND1), ER signaling, and DNA repair (e.g., BRCA1/2; ref. 5). In breast cancers, the PI3K–AKT pathway is thought to be continuously activated in approximately 50% of patients (6–9) and is becoming a research focus and potential therapeutic target in breast cancer (10–12). However, due to its intolerable side effects, the clinical application of inhibitors directly targeting the PI3K–AKT pathway has been limited. Therefore, an alternative approach to regulating AKT downstream effectors in breast cancer is urgently needed.

Salt-inducible kinase 1 (SIK1), an AMPK-related kinase, regulates energy response-related physiopathologic processes, such as gluconeogenesis and lipid metabolism (13). Recently, SIK1, not AMPK, has been increasingly identified as a major player in tumor suppressor liver kinase 1 (LKB1)-deficient NSCLC (14, 15). However, the precise roles and upstream regulators of SIK1 in breast cancer are not well characterized. Signal transducers and activators of transcription (STAT) family transcription factors, including STAT1–6, play distinct and pivotal roles in physiologic and pathologic processes. Among these transcription factors, STAT3 probably plays the most significant role in tumorigenesis via its response to IL6-induced JAK2 activation (16, 17). Although diverse mechanisms have been identified that modulate STAT3 transcriptional activation, including phosphorylation, methylation, and palmitoylation (18, 19), whether classic signaling pathways such as the PI3K–AKT pathway activate STAT3 is not clear. Furthermore, although inhibitors antagonizing JAK2 or STAT3 have been developed and used as cancer therapies (16, 17, 20), whether these inhibitors can be successfully used for breast cancer therapies has not yet been determined. Considering these findings and research topics, we propose a model in which physiologic level of SIK1 maintains its tumor suppressor function by binding to STAT3 to repress STAT3 transcriptional activation and oncogenic function. Under certain conditions, activated AKT directly phosphorylates SIK1 to promote SIK1 cytoplasmic translocation, Pin1 binding, and ITCH recognition for degradation, resulting in the activation of STAT3 oncogenic functions and facilitating breast tumorigenesis. Our findings highlight a promising strategy to target the JAK2–STAT3 pathway to combat breast cancers driven by the hyperactivation of AKT.

Cell lines, transfection, and infection

MDA-MB-231, MDA-MB-468, MDA-MB-453, HEK293, and HEK293T cells were cultured in DMEM, and BT474 and BT549 cells were cultured in RPMI-1640 supplemented with 10% FBS and 1% penicillin/streptomycin (P/S). MCF-7 cells were cultured in MEM supplemented with 10% FBS, 1% P/S, and 0.01 mg/mL human recombinant insulin. MCF-10A were cultured in DMEM supplemented with 5% HS, 20 ng/mL EGF, 0.5 μg/mL hydrocortisone, 10 μg/mL insulin, 1% NEAA, and 1% P/S. All cell lines were maintained at 37°C in a humidified atmosphere with 5% CO2. Cell transfection was performed using Lipofectamine and Plus reagents. Packaging of lentiviral shRNA or cDNA expressing viruses were cotransfected with packaging vectors psPAX2 and pMD2G (Addgene) into HEK293T cells. Following viral infection, cells were maintained in the presence of hygromycin (200 μg/mL) or puromycin (1 μg/mL), depending on the viral vectors used to infect cells.

Cell fractionations were performed with the Cell Fractionation Kit (CST9038). Inhibitors or drugs such as MK2206 (Selleck, S1078), GSK690693 (Selleck, S1113), MG132 (Selleck, S2619), cycloheximide (MCE, HY-12320), 3-MA (Selleck, S2767), NH4Cl (Sigma, A9434), PiB (Sigma, B7688), ruxolitinib (Selleck, S1378), CYT387 (Selleck, S2219), and carboplatin (MCE, HY-17393) were used at the indicated doses.

DNA constructs and antibodies

pcDNA3-HA-SGK1, pcDNA3-HA-S6K1, pcDNA3-HA-AKT1, pcDNA3-HA-AKT2, pcDNA3-HA-AKT3, and pcDNA3-HA-myr-AKT1, pcDNA3-HA-myr-AKT2, pcDNA3-HA-S6K1-R3A, pcDNA3-HA-Δ60-SGK1, pcDNA3-HA-SIK1, pcDNA3-HA-14-3-3β, pcDNA3-HA-14-3-3γ, pcDNA3-HA-14-3-3φ, pcDNA3-HA-14-3-3ε, pcDNA3-HA-14-3-3ζ, pCMV-GST-14-3-3γ, pcDNA3-HA-Pin1, and pcDNA3-HA-ITCH were cloned into mammalian expression pCDNA3-HA vectors as previously described (21). pcDNA3-Flag-ITCH was cloned into mammalian expression pCDNA3-Flag vectors. pCMV-GST-SIK1, pCMV-GST-HDAC5, and pCMV-GST-Pin1 were cloned into mammalian expression pCMV-GST vectors. The pCMV-Myc-STAT3 was cloned into mammalian expression pCDNA3-Myc vectors. pGEX-4T1-SIK1 were cloned into mammalian expression pGEX-4T1 vectors. pLenti-puro-SIK1 was cloned into mammalian expression pLenti-puro vectors. SIK1-P187A/P188A/Y189A, SIK1-S65A, SIK1-S435A, SIK1-S435D, AKT1-E17K, Pin1-V34A/K63A, ITCH-C830A, and STAT3-P578A/E680A were obtained via Q5 Site-Directed Mutagenesis Kit (New England Biolabs) according to the manufacturer's instructions and the sequences were verified by DNA sequencing.

Anti-AKT Substrate (RxxpS/T) antibody (9614), anti-phospho-Ser473-AKT antibody (4060), anti-phospho-Thr308-AKT antibody (2965), anti-AKT total antibody (4691), anti-phospho-Ser9-GSK3β antibody (5558), anti-GSK3β antibody (12456), anti-Tuberin/TSC2 antibody (4308), anti-phospho-Tuberin/TSC2 (Thr1462) antibody (3611), anti-PRAS40 antibody (2691), anti-phospho-PRAS40 (Thr246) antibody (2997), anti-pS6K1 (Thr389) antibody (9205), anti-S6K1 antibody (2708), anti-S6 antibody (2217), anti-pS240/244-S6 antibody (5364), anti-pntihospho-HDAC4 (Ser632)/HDAC5 (Ser661)/HDAC7 (Ser486) antibody (3424), anti-phospho-STAT3 (Tyr705) (9145) antibody, anti-STAT3 antibody (30835), anti-Myc antibody (2276), and anti-GST antibody (2625) were obtained from Cell Signaling Technology. Anti-GAPDH (10494–1-AP), anti-HA (51064-2-AP) were obtained from Proteintech. Anti-SIK1 (16864) were obtained from ELGBIO. Anti-Ki67 (GB111499) were obtained from Servicebio. Anti-SIK1-pS435 was generated by Abclonal with the peptides C-RPV(pS)PSSL.

Immunoprecipitation and immunoblotting analysis

Cells were lysed in EBC buffer (50 mmol/L Tris pH 7.5, 120 mmol/L NaCl, 0.5% NP-40) supplemented with protease inhibitors (cOmplete Protease Inhibitor Cocktail, Roche) and phosphatase inhibitors (PhosSTOP, Roche). The protein concentrations of whole-cell lysates were measured by the Pierce BCA Protein Assay Kit (23225). Equal amount of protein lysates was resolved by SDS-PAGE gels and then transferred on a PVDF membrane (Millipore). The membranes were incubated with a primary antibody at 4°C overnight and then hybridized with a secondary antibody at room temperature for 1 hour. For immunoprecipitation, whole cell lysates (WCL) were incubated with the indicated antibody (1–2 μg) for 3 to 4 hours at 4°C followed by 1 hour incubation with Protein A/G sepharose beads (GE Healthcare). For GST pulldown assays, 1,000 μg lysates were incubated with glutathione-sepharose 4B (GE Healthcare) for 2 hours at 4°C. The immunoprecipitants were washed five times with NETN buffer (20 mmol/L Tris, pH 8.0, 150 mmol/L NaCl, 1 mmol/L EDTA, and 0.5% NP-40) before being resolved by SDS-PAGE and immunoblotted with indicated antibodies. Quantification of the immunoblot band intensity was performed with ImageJ software as previously described (21).

Animal experiments

5×106 BT549 cells expressing WT or mutant forms of SIK1 and 10×106 of MCF7 cells stably knockdown SIK1 mixed with matrigel were injected into the flank of 8 BALB/c nude mice (University of Sun Yat-sen, 4–5 weeks of age). For treatment analysis, mice were randomized into different treatment arms and administrated with vehicle control, 50 mg/kg of ruxolitinib, 60 mg/kg of carboplatin, or combined 50 mg/kg of ruxolitinib and 60 mg/kg of carboplatin three times weekly for 3 to 4 weeks via intraperitoneal injection. Tumor size was recorded every other day over the course of the studies, and the tumor volume was determined with the formula: L (longer diameter) × W2 (shorter diameter) ×0.5. The mice were sacrificed after 5 weeks of injection. The xenograft tumor was removed and used to extract protein and RNA. The other part was fixed in paraformaldehyde for IHC detection.

Breast cancer tissues and IHC staining

The studies were performed with the approval from the Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University [(2019) No. 381]. Fresh and formalin-fixed paraffin-embedded breast tumor samples were collected from patients diagnosed with breast cancer at First Affiliated Hospital of Sun Yat-sen University using protocols approved by the institutional review boards and following the Declaration of Helsinki ethical guidelines. Written informed consent was obtained from all patients and samples were identified prior to transfer for experiments. Tumor histology and expression of ER, PR, and HER2 were justified for diagnosis. Tumor xenografts or surgical specimen tissue slides were fixed in 10% neutral buffered formalin, embedded in paraffin, and cut into 4 μm sections. These samples were deparaffinized and rehydrated, followed by antigen retrieval with sodium citrate buffer. The tissue sections were stained with hematoxylin and eosin. For IHC staining, the sections were incubated with 3% H2O2 for 15 minutes and blocked with 5% normal goat serum for 1 hour at room temperature. Following primary antibody incubation for SIK1, pS435-SIK1, pY705-STAT3, pS240/244-S6 overnight sections were incubated with monoclonal mouse anti-rabbit immunoglobulins (Dako#M0737) for 30 minutes at room temperature. Afterward, sections were incubated with Envision+ System-HRP Labeled Polymer Anti-Rabbit (Dako #K4003) for 30 minutes. All sections were developed using the DAB chromogen kit (Dako #K3468) and lightly counterstained with hematoxylin.

Dual luciferase reporter assay

Oligonucleotide fragments containing three tandems SIE-consensus (CTTCATTTCCCGTAAATCCCTA) motifs (22) with minor promoters were cloned into pGL3-basic vectors (Promega). The pGL3-3xSTAT3-Luc was transiently transfected using Lipofectamine and Plus regents into 293T cells along with CMV Renilla (pRL-CMV Renilla, Promega) as an internal control. Response ratios are expressed relative to the signal obtained for the empty vector control wells transfected with the pCDN3-HA vector. In evaluating the relative response ratios, the STAT3 transcriptional activity ratios regulated by other proteins, the amount of cDNA for CMV Renilla, pGL3-3xSTAT3-Luc, HA-SIK1, and mutants, other regulatory proteins were 10, 50, 200, and 100 ng, respectively. However, the cotransfected empty vector constructs amount varied to ensure the total amount of DNA transfected into each well was constant. Luminescence measurements were taken 40 hours after transfection. All results show means and standard deviation from experiments performed in biological triplicates (n = 3).

Immunostaining

Cells grown on a glass bottom cell culture dish were fixed in paraformaldehyde, blocked with 10% goat serum, and then incubated with primary antibodies targeting HA (1:200), washed thrice with PBS, and then incubated with 546 Alexa (red)–labeled secondary antibodies (Molecular Probes). The DNA dye DAPI was used to counterstain the nuclear DNA.

Peptide synthesis and dot immunoblot assays

SIK1-S435-WT and SIK1-S435-Phospho peptides used for dot blot assays were synthesized by Abclonal Technology. The sequences are as follows:

  •  SIK1-S435-WT: C-RPVSPSSL

  •  SIK1-S435-Phospho: C-RPV(pS)PSSL

Peptides were diluted into 2 mg/mL for further biochemical assays. For dot blot assays, peptides were diluted with PBS and spotted onto nitrocellulose membrane with the amount of 0.01, 0.03, 0.10, 0.30, and 0.50 μg. The membrane was dried and blocked by soaking in TBST buffer with 5% nonfat milk for immunoblot analysis.

In vivo ubiquitination assay

In vivo ubiquitination assays were performed as previously described (23). In brief, HEK293T cells were transfected with His-Ub and the indicated constructs. One day after transfection, cells were treated with 10 μmol/L MG132 for 12 hours and washed with PBS twice, and then were lysed in buffer A [6M guanidine-HCl, 0.1M Na2HPO4/NaH2PO4, and 10 mmol/L imidazole (pH 8.0)] and subjected to sonicate. After high-speed centrifugation, the supernatants were incubated with nickel-beads (Ni-NTA; QIAGEN) for 3 hours at room temperature. The products were washed twice with buffer A, twice with buffer A/TI (1 volume buffer A and 3 volumes buffer TI), and twice with buffer TI (25 mmol/L Tris-HCl and 20 mmol/L imidazole [pH 6.8]). The pulldown proteins were resolved in 6% SDS-PAGE for immunoblot analysis.

In vitro ubiquitination assay

In vitro ubiquitination assays were performed as described previously (24). Briefly, commercial E1, E2 (UbcH5a and UbcH3), ubiquitin (obtained from UBbiotech), 1 μg of bacterially purified GST-SIK1, and ITCH purified from HEK293T cells transfected with Flag-ITCH by Flag affinity precipitation were performed at 37°C for 2 hours and stopped by 2× SDS sample buffer and then resolved by SDS-PAGE for immunoblotting.

In vitro kinase assay

The in vitro kinase assay was performed as previously described (21). Briefly, 1 μg of the bacterially purified GST-HDAC fusion proteins was incubated with immunoprecipitated SIK1 from HEK293T cells transfected with various mutant SIK1 in the kinase reaction buffer (50 mmol/L Tris pH 7.5, 1 μmol/L MnCl2, 2 mmol/L DTT) for 30 minutes at 30°C. The reaction was subsequently stopped by adding 0.1 mmol/L EDTA. The reaction was resolved by SDS-PAGE and detected by pHDAC.

mRNA extraction, qPCR, and RNA sequencing

Total RNA was isolated from the cells using TRIzol Reagent (Invitrogen), and 1 mg of each RNA sample is reverse transcribed into cDNA, and then analyzed by quantitative PCR with three repeats. Relative gene expression was calculated using the 2-ΔΔCt method, which used the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as an internal control to standardize target mRNA expression. Primer sequences used were Oncostatin M Receptor (OSMR) gene, forward (5′-ATGGCTCTATTTGCAGTCTTTCA-3′) and reverse (5′-CACCCAGATGACATTGGATGTT-3′); Interferon Beta 1 (IFNB1) gene, forward (5′-TTGTTGAGAACCTCCTGGCT-3′) and reverse (5′-CCACACTGTATTTGGTGTCT-3′); interferon-stimulated gene 15 (ISG15) gene, forward (5′-TCATCAGGTCAAGGATAGTC-3′) and reverse (5′-CCACACTGTATTTGGTGTCT-3′); Myc gene, forward (5′- GGCTCCTGGCAAAAGGTCA-3′) and reverse (5′-CTGCGTAGTTGTGCTGATGT-3′); GAPDH gene, forward (5′-TGTGGGCATCAATGGATTTGG-3′) and reverse (5′- ACACCATGTATTCCGGGTCAAT-3′).

Ubiquitin-interaction domain RNA sequencing experiment and high-throughput sequencing and data analysis were conducted by Seqhealth Technology Co., Ltd as previously described (23). Briefly, total RNAs were extracted by using TRIzol Reagent (Invitrogen, cat. #15596026). DNA digestion was carried out after RNA extraction by DNaseI. RNA quality was determined by examining A260/A280 and confirmed by 1.5% agarose gel electrophoresis. Total RNAs (2 μg) were used for stranded RNA sequencing library preparation using the KC-Digital Stranded mRNA Library Prep Kit for Illumina (cat. #DR08502, Wuhan Seqhealth Co., Ltd. China) following the manufacturer's instruction. The kit eliminates duplication bias in PCR and sequencing steps, by using unique molecular identifier of 8 random bases to label the preamplified cDNA molecules. The library products corresponding to 200–500 bps were enriched, quantified, and finally sequenced on Novaseq 6000 sequencer (Illumina) with the PE150 model. The raw sequencing data were first filtered by Trimmomatic (version 0.36), low-quality reads were discarded. Clean Reads were further treated with in-house scripts to eliminate duplication bias introduced in library preparation and sequencing. After all subclusters were generated, multiple sequence alignment was performed to get one consensus sequence for each subcluster. After that, the deduplicated consensus sequences were used for standard RNA sequencing analysis. They were mapped to the reference genome of mouse using STAR software (version 2.5.3a) with default parameters. A P value cutoff of 0.05 and fold-change cutoff of 2 were used to judge the statistical significance of gene-expression differences. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis for differentially expressed genes was implemented by RStudio (version 1.2.1335) with a P value cutoff of 0.05 to judge statistically significant enrichment.

Mass spectrometry

For mass spectrometry analysis, GST-pulldown was performed with the WCL derived from three 10-cm dishes of HEK293T cells transfected with GST-SIK1 and HA-myr-AKT1 as previously described (21). Briefly, the proteins were resolved by SDS-PAGE and identified by Coomassie staining. The band containing SIK1 was reduced with 10 mmol/L DTT for 30 minutes, alkylated with 55 mmol/L iodoacetamide for 45 minutes, and in-gel digested with trypsin enzymes. The tryptically digested peptides were desalted with monospin C18 column (SHIMADZU-GL), and then analyzed on an Easy-nLC1200 system equipped with a homemade reverse phase C18 column (75 μm × 300 mm, 1.9 μm). The peptides were separated with a 120-minute gradient from 5% to 100% of buffer B (buffer A: 0.1% formic acid in water; buffer B: 0.1% formic acid in 80% acetonitrile) at 300 nL/minute. The eluted peptides were ionized and directly introduced into a Q-Exactive mass spectrometer (Thermo Scientific) using a nano-spray source with the application of a distal 2.5-kV spray voltage. A cycle of one full-scan mass spectrometry (MS) spectrum (m/z 300–1,500) was acquired followed by the top 20 MS/MS events, sequentially generated on the first to the twentieth most intense ions selected from the full MS spectrum at a 30% normalized collision energy.

The acquired MS/MS data were analyzed against a homemade database (including all target proteins) using PEAKS Studio 8.5. Cysteine alkylation by iodoacetamide was specified as fixed modification with mass shift 57.02146 and methionine oxidation, protein n-terminal acetylation as variable. Additionally, phosphorylation was set as dynamic modification with mass shift 79.9663. In order to accurately estimate peptide probabilities and false discovery rates, we used a decoy database containing the reversed sequences of all the proteins appended to the target database.

Purification of GST-tagged proteins from bacteria

The GST-tagged proteins were purified from bacteria as previously described (21). Briefly, recombinant GST-conjugated SIK1 and HDAC5 were generated by transforming the BL21 (DE3) E. coli strain. The cultured bacteria were grown at 37°C to an O.D. of 0.8, and then the protein expression was induced for 12 to 16 hours by adding 0.1 mmol/L IPTG at 16°C with vigorous shaking. Cell pellets were sonicated after resuspension in 5 mL EBC buffer and subjected for indicated protein purification. Insoluble proteins and cell debris were discarded, and the supernatant was incubated with 50 μL 50% glutathione-sepharose slurry (Pierce) for 3 hours at 4°C. The glutathione beads were washed 3 times with PBS buffer or eluted with elution buffer. The purified proteins were analyzed by Coomassie blue staining and quantified with BSA standards.

Colony formation assays

The cells were inoculated into a 6-well plate (300 or 600 cells/well) and left for 12 to 20 days until visible colonies were formed. The colonies were washed twice with PBS, fixed with 10% acetic acid/10% methanol for 20 minutes, and then stained with 0.4% crystal violet in 20% ethanol for 20 minutes. After staining, the plate was washed and air-dried, and the number of colonies was counted and quantified.

Annexin V/amino actinomycin D double staining assay

To detect cellular apoptosis, cells treated with the indicated concentration of drugs were costained with Annexin V-phycoerythrin (PE) and amino actinomycin D (7-AAD) from Annexin V–PE apoptosis detection kit I (BD Biosciences) according to the manufacturer's instructions. Stained cells were monitored with fluorescence-activated cell sorting.

Quantification and statistical analysis

The in vitro experiments were repeated at least three times unless specifically stated as previously described (21). The compared groups were set similarly in all procedures. For animal studies, we established the number of conditions to test the hypothesis, and two groups were randomly assigned. Results were collected and analyzed blindly. As indicated in the figure legends, all quantitative data are presented as the mean ± SD of three biologically independent experiments or samples; n and P values are indicated in every single figure. Two-way ANOVA was used for multiple-group comparisons and unpaired two-tailed t tests for two-group comparisons. Significant statistical differences between groups were indicated as *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Statistical analyses and graphics were carried out with GraphPad Prism software 7.0 and Microsoft Excel 16.0.

Study approval

Animal experiments were approved by the Laboratory Animal Center of the Sun Yat-sen University (IACUC) and complied with all relevant ethical regulations (SYSU-IACUC-2021-000647).

Data availability

The MS data generated in this study have been deposited in iProX/ProteomeXchange under the accession code IPX0005788000. The RNA sequencing data sets have been deposited to Gene Expression Omnibus with the accession number GSE223214. The genetic alterations of ITCH, PTEN, PIK3CA, AKT1, JAK2, and STAT3 in cBioPortal for Cancer Genomics data sets were integrated from www.cbioportal.org, with the query of according gene for both mutation and copy-number alterations in breast cancer types.

SIK1 plays a tumor suppressor role in breast cancer

Although recent studies have demonstrated the potential tumor-suppressive functions of SIK1 in breast cancer (25, 26), systematic analyses of SIK1 in breast cancer have not been reported. Therefore, we initially analyzed SIK1 expression in the breast cancer context via bioinformatics and observed that SIK1 was highly expressed in normal tissues compared with breast tumor tissues and that it tended to be expressed at lower levels in metastatic tissues and advanced-stage cancers (Fig. 1A and B). Moreover, lower SIK1 expression indicated a low survival rate for breast cancer patients (Fig. 1C; Supplementary Fig. S1A). Aligned with these findings, SIK1 was expressed at lower levels in breast tumor tissues than in adjacent normal samples (Supplementary Fig. S1B). Notably, tumor tissues derived from HER2-positive and TNBC patients displayed lower SIK1 expression than those derived from luminal (ER+/PR+) breast tumors (Fig. 1D and E), and this decreased expression was accompanied by relatively high rates of AKT phosphorylation and activation (Fig. 1D and E; Supplementary Fig. S1C). Similar results were observed with a panel of breast cancer cell lines (Fig. 1F). To further explore the potential roles of SIK1 in breast cancer, we knocked down SIK1 in breast cancer cells (Fig. 1G) and observed that depletion of SIK1 enhanced the colony formation rate (Fig. 1H) and tumor growth rate in vivo (Fig. 1IL; Supplementary Fig. S1D). Collectively, these findings suggest that SIK1 is a tumor suppressor and prognostic marker for breast cancer.

AKT interacts with SIK1 to promote its phosphorylation

To explore the upstream regulation of SIK1, we performed an MS-based protein identification assay, and the AKT1, ITCH, STAT3, and 14-3-3 proteins were found to interact with SIK1 (Supplementary Fig. S1E). Among these SIK1-binding proteins, we confirmed the interaction between SIK1 and AKT in cells in which SIK1 was expressed at both ectopic and physiologic levels (Fig. 2A; Supplementary Fig. S1F), and the SIK1 C-terminus was identified as the AKT-interacting region (Supplementary Fig. S1F). To identify the phosphorylated residues in SIK1, an MS analysis of phosphorylated SIK1 in the presence of active AKT (myr-AKT1) was performed, and several phosphorylated residues in SIK1 were thus identified (Supplementary Fig. S1G). Among these residues, S65 and S435 were the foci of our investigation because of their location within a classic AGC phosphorylation motif (RxRxxS/T) and high degree of conservation among species (Fig. 2B). To assess whether other AGC kinases phosphorylate SIK1 at these residues, we screened a panel of AGC kinases, including constitutively active Myr-AKT1, Myr-AKT2, Myr-AKT3, S6K1-R3A, and SGK1-Δ60. The results showed that AKT1, and to a lesser degree, AKT2 and AKT3, but not the other kinases we assessed, significantly promoted SIK1 phosphorylation (Fig. 2C), which was markedly attenuated by the administration of AKT antagonists (Supplementary Fig. S1H). Furthermore, the C-terminus of SIK1, with which AKT interacted and where S435 resided (Supplementary Fig. S1F), was observed to undergo AKT1-mediated phosphorylation (Supplementary Fig. S1I). To confirm that AKT phosphorylates SIK1 at S435, an unphosphorylated mutant mimic, SIK1-S435A, was generated to inhibit AKT-mediated phosphorylation (Fig. 2D). Moreover, in vitro kinase assays coupled with MS analyses were performed to confirm that AKT directly phosphorylated SIK1 at S435 (Supplementary Fig. S1J).

Next, to study the phosphorylation of SIK1 under physiologic conditions, we generated antibodies against SIK1-pS435 and used them to detect the specific phosphorylation of SIK1 S435 both in vitro and in vivo (Fig. 2E and F; Supplementary Fig. S1K). Notably, the phosphorylation of SIK1-S435 was markedly enhanced by the expression of tumor-associated AKT1 (E17K; Fig. 2E) or via stimulation with insulin (Fig. 2F), whereas S435 SIK1 phosphorylation was markedly attenuated after AKT was repressed pharmacologically or genetically (Fig. 2F and G). In addition, a positive connection between AKT activity and pS435-SIK1 was observed in both breast cancer cell lines and tumor tissues (Fig. 1DF). Together, these findings suggest that SIK1 is a bona fide substrate of AKT.

AKT inhibits SIK1 kinase activity and tumor suppressor functions

To investigate the biological function of AKT-mediated SIK1 phosphorylation, we initially observed that the ectopic expression of Myr-AKT1 repressed SIK1-mediated HDAC5 phosphorylation, a well-characterized SIK1 substrate (Supplementary Fig. S2A–S2B). Furthermore, compared with WT or S435A variant, the phosphomimic SIK1-S435D mutant impaired HDAC5 phosphorylation both in cells and in vitro (Fig. 2H and I; Supplementary Fig. S2C), indicating that AKT phosphorylates SIK1 and attenuates SIK1 kinase activity. Aligned with this finding, in contrast to WT and S435A SIK1, the S435D mutant did not reduce either the breast cancer cell colony formation (Fig. 2J; Supplementary Fig. S2D–S2E) or tumor growth rate (Fig. 2K and L; Supplementary Fig. S2F). Together, these findings imply that AKT-mediated SIK1 phosphorylation negatively regulates SIK1 kinase and tumor suppressor functions.

SIK1 physiologically represses the JAK2–STAT3 signaling pathway

Although we and other researchers have identified SIK1 as a tumor suppressor in breast and lung cancers, the underlying mechanism of SIK1-induced tumor suppression has not been well characterized to date. Therefore, RNA sequencing with WT SIK1 and the S435D mutant in BT549 cells was performed, and enrichment of the JAK2–STAT3 pathway was found to be significantly higher in SIK1-S435D–expressing cells than in SIK1-WT–expressing cells (Fig. 3A; Supplementary Fig. S3A—S3B). To confirm this finding, we investigated whether SIK1 influences the phosphorylation of STAT3 at Y705, an indicator of STAT3 activity (27), and observed that the expression of WT- and S435A-mutant SIK1 but not that of S435D-mutant SIK1 decreased (Fig. 2H; Supplementary Fig. S3C), whereas the depletion of SIK1 elevated the phosphorylation rate of STAT3 (Figs. 1G, L, and 3B; Supplementary Fig. S1D). In contrast, the expression of S435D-mutant SIK1 but not that of WT or S435A-mutant SIK1 markedly enhanced STAT3 transcriptional activity, as measured by luciferase reporter assay (Fig. 3C). As a result, SIK1-S435D elevated the expression of downstream STAT3 target genes, such as c-MYC, ISG15, and IFNB1 (Fig. 3D). Notably, we observed that SIK1 knockdown clearly enhanced STAT3 transcriptional activity by regulating target genes (Supplementary Fig. S3D). In support of these findings, ectopically expressed AKT1-E17K enhanced the influence of WT-SIK1 but not that of S435A- or S435D-mutant SIK1 on pSTAT3, whereas administration of the AKT inhibitor MK2206 attenuated the effect of WT SIK1 on pSTAT3 (Fig. 3E and F). Moreover, Myr-AKT1, but not DN-AKT1, promoted STAT3 transcriptional activity and target gene expression at least partially by inhibiting SIK1 activity (Fig. 3G and H), suggesting that AKT1 blocks SIK1-mediated repression of STAT3 transcriptional activity.

STAT3 was observed to be an SIK1 interaction partner via MS analysis (Supplementary Fig. S1E), and this finding was confirmed when SIK1 was expressed at an endogenous level (Fig. 3I). Furthermore, the DBD and SH2 domain of STAT3 was identified as the interaction sites for the SIK1 UBA domain (Supplementary Fig. S3E–S3F). Although SIK1 typically acts as a kinase, the serine/threonine phosphorylation of STAT3 was not altered by the forced expression of SIK1, as determined via MS. Hence, we proposed that SIK1 likely directly binds STAT3 to and abrogates STAT3 function, similar to the mechanism in which FoxA1 binds STAT2 to block its function and thus regulate the immune response (28). Due to playing the critical roles of the SH2 domain in STAT3 dimerization and DBD domain in DNA binding, we found that the interplay between SIK1 and STAT3 markedly blocked STAT3 homodimer formation (Fig. 3JL; Supplementary Fig. S3G–S3H) and subsequent binding with its response elements (Fig. 3M and N). Moreover, this interplay was antagonized by the forced expression of AKT1 or the SIK1-S435D mutant. Moreover, AKT1 knockdown decreased the STAT3 phosphorylation rate, which was restored by further depletion of SIK1 (Fig. 3O), indicating that AKT1 promoted STAT3 transcriptional activity by phosphorylating SIK1. Therefore, our results suggest that, via direct interaction, SIK1 blocks STAT3 dimerization and DNA binding, resulting in decreased STAT3 transcriptional activity in an AKT-mediated phosphorylation-dependent manner.

14-3-3 facilitates SIK1 cytoplasmic translocation

Next, we examined how AKT-mediated phosphorylation affects SIK1 functions to repress STAT3. It had been previously reported that 14-3-3, an adaptor protein, plays pivotal roles in cell signal transduction by binding phosphorylated proteins, such as FOXO3a and YAP, thereby affecting their localization (29, 30). Generally, 14–3-3 recognizes the phosphorylated motif RxxpS(x)P, which is conserved in the AKT-phosphorylated SIK1 region (Supplementary Fig. S4A). Therefore, we speculated that AKT possibly regulates the SIK1/14-3-3 interaction, thereby affecting SIK1 cellular localization. Consistent with a previous finding and our MS results (Supplementary Fig. S1E; ref. 31), SIK1 bound to multiple 14-3-3 family members, including 14-3-3γ, θ, and η (Supplementary Fig. S4B). Importantly, forcing the expression of Myr-AKT1 but not that of DN-AKT1 markedly enhanced SIK1 interaction with 14-3-3 (Supplementary Fig. S4C–S4D), whereas the interaction rate of endogenous SIK1 with 14-3-3 was decreased by treatment with AKT inhibitor (Fig. 4A), indicating that AKT regulates the SIK1/14-3-3 interaction. In addition, the S435D mutant induced a markedly elevated SIK1 interaction with 14-3-3 (Fig. 4B; Supplementary Fig. S4E–S4F). Interestingly, WT and S435A-mutant SIK1 displayed nuclear localization, which was largely mediated by active AKT or the S435D mutation (Fig. 4CE). Notably, when proline in the 14-3-3 binding motif was replaced with an alanine (P436A), the 14-3-3 recognition of SIK1 was profoundly attenuated (Supplementary Fig. S4G). Together, these data suggest that AKT-mediated SIK1 phosphorylation enhances SIK1 binding with 14-3-3, which in turn increases the rate of SIK1 transfer to the cytoplasm.

AKT cooperates with pin1 to degrade SIK1

Notably, in experiments with pS435-SIK1, an inverse correlation between AKT activation and SIK1 abundance was identified (Fig. 1DF). Confirming this finding, depletion of AKT1 elevated the SIK1 protein level, whereas the ectopic expression of AKT reduced it (Fig. 2E, G). To explore the pathways involved in AKT-mediated SIK1 degradation, the proteasomal inhibitor MG132, but not another degradation-related inhibitor, such as the lysosome inhibitor NH4Cl or autophagy inhibitor 3-methyladenine (3-MA), clearly blocked AKT-mediated SIK1 degradation (Fig. 4F). Furthermore, the ectopic expression of Myr-AKT1 shortened the half-life of endogenous SIK1, whereas the depletion of AKT1 prolonged it (Fig. 4GJ). Therefore, the ubiquitination of SIK1 was markedly increased by Myr-AKT1, but not by DN-AKT (Fig. 4K). Consistent with these findings, the S435A mutant, but not the S435D mutant, resisted AKT-mediated SIK1 degradation (Fig. 4L; Supplementary Fig. S5A–S5B), and this effect was accompanied by a decrease in AKT-induced SIK1 ubiquitination (Fig. 4M). Together, these findings suggest that AKT promotes SIK1 degradation in a phosphorylation- and proteasome-dependent manner (Fig. 4N).

Next, we sought to determine how AKT promotes SIK1 degradation. Recent studies have reported that Pin1 plays a critical role in regulating the stability of phosphorylated proteins by selectively recognizing serine/threonine-proline motifs (32, 33). Therefore, we investigated whether Pin1 is involved in AKT-mediated SIK1 degradation in a phosphorylation-dependent manner. We found that the ectopic expression of WT but not that of an enzymatically inactive Pin1 mutant (W34A/K63A) shortened the SIK1 half-life (Supplementary Fig. S5C–S5E). In addition, PiB, a specific Pin1 inhibitor previously utilized in a clinical trial for cancer therapies (33), markedly attenuated Pin1-induced SIK1 degradation (Fig. 5A). Notably, Pin1, but not its inactive mutant, promoted SIK1 ubiquitination, which was antagonized by Pin1 inhibitor PiB (Fig. 5B). In support of SIK1 as a potential target of Pin1, the interaction between endogenous Pin1 and SIK1 was observed (Fig. 5C). Notably, the AKT-phosphorylated C-terminus of SIK1 was found to be the region of Pin1 binding (Supplementary Fig. S5F), and activation of AKT increased the SIK1–Pin1 interaction (Supplementary Fig. S5G). Specifically, the S435A or P436A mutation repressed the SIK1 interaction with Pin1, whereas the S435D mutation enhanced their interaction (Supplementary Fig. S5H–S5I), indicating that Pin1 binds to SIK1 in an AKT-mediated phosphorylation-dependent manner. Taken together, our results suggest that Pin1 facilitates AKT activity to promote SIK1 ubiquitination and subsequent degradation.

ITCH is a bona fide E3 ubiquitin ligase for SIK1

To explore the E3 ubiquitin ligase for SIK1, we analyzed SIK1 binding partners derived from MS, from which the E3 ligase ITCH was observed in the top rank (Supplementary Fig. S1E). The results confirmed the interaction of endogenous ITCH and SIK1 (Fig. 5D), and the HECT domain of ITCH was shown to interact with SIK1 (Supplementary Fig. S6A). Moreover, the UBA domain in SIK1 was mapped for its interaction with ITCH (Supplementary Fig. S6B). Notably, WT, but not a catalytically deficient ITCH mutant (C830A), efficiently reduced SIK1 abundance in a dose-dependent manner (Fig. 5E; Supplementary Fig. S6C), and this reduction was accompanied with a decrease in the SIK1 half-life (Supplementary Fig. S6D–S6E). In contrast, depletion of ITCH markedly induced endogenous SIK1 expression (Fig. 5F) and was accompanied by an increased half-life (Fig. 5G and H). Consistent with these findings, WT but not C830A-mutant ITCH promoted SIK1 ubiquitination both in cells and in vitro (Fig. 5I and J). Additionally, SIK1 ubiquitin conjugation was markedly repressed after ITCH depletion (Fig. 5K). It has been reported that the NEDD4 family of E3 ubiquitin ligases binds their substrates via a specifically recognizing motif such as PP(x)Y (34, 35). Therefore, we analyzed the SIK1 protein sequence and identified an evolutionarily conserved putative ITCH recognition motif at residues 187–189 (PPY) (Supplementary Fig. S6F). Consistent with this finding, SIK1 carrying a motif that underwent a substitution mutation (PPY to AAA) failed to interact with ITCH (Fig. 5L) and underwent ITCH-mediated ubiquitination (Fig. 5M) and subsequent degradation (Fig. 5N). These findings imply that ITCH is a bona fide E3 ligase of SIK1.

Next, we investigated whether AKT-mediated phosphorylation and Pin1-mediated SIK1–ITCH interaction affected ITCH-induced SIK1 ubiquitination and degradation. We initially observed that Myr-AKT1 enhanced the interaction of ITHC with SIK1 (Supplementary Fig. S6G), whereas SIK1-S435A abrogated AKT-mediated SIK1 interaction with ITCH (Supplementary Fig. S6H). Notably, Pin1 enhanced, whereas its inhibitor repressed, the interaction between ITCH and SIK1 (Supplementary Fig. S6I). In addition, knocking down ITCH elevated the protein levels of WT but not that of S435A- or S435D-mutant SIK1 in breast BT549 cancer cells (Fig. 5O). Moreover, a mutant SIK1 (PPY-AAA) antagonized AKT-mediated degradation (Fig. 5P). Together, these observations demonstrate that ITCH interacts with and degrades SIK1 in an AKT-mediated phosphorylation- and Pin1-mediated SIK1-ITCH interaction-dependent manner (Fig. 5Q).

JAK2/STAT3 inhibitors antagonize hyperactivated AKT-driven breast tumorigenesis

To further investigate the connections among AKT, ITCH, SIK1, and STAT3, we performed bioinformatics analysis and observed that the amplification of ITCH was mutually exclusive to PI3K–AKT pathway alterations (including PTEN deletion/mutations and PIK3CA and AKT1 mutations) and JAK2/STAT3 amplification (Supplementary Fig. S7A), indicating a potential connection between ITCH and the PI3K–AKT pathway and JAK2/STAT3 activation. Furthermore, we stained select proteins in breast cancer cell lines and breast cancer tissues and observed that AKT activation (pS473-AKT, pGSK3β, and pS6 levels) was positively correlated with SIK1-pS435 and STAT3-pY705 but negatively correlated with SIK1 protein levels (Fig. 1DF).

To determine whether the JAK2–STAT3 axis is the major effector of the AKT–SIK1 pathway in breast tumorigenesis, ruxolitinib, a clinically approved JAK2 inhibitor used in certain cancer therapies and clinical trials for breast cancer (36), was used to treat breast cancer cell lines. The sensitivity of breast cancer cell lines to ruxolitinib was measured, and basal cell lines such as MDA-MB-231, MDA-MB-468, and BT549 were shown to exhibit lower SIK1 expression than the other cell lines, indicating greater sensitivity to ruxolitinib (Supplementary Figs. S7B and S8A). To determine whether the cell response to ruxolitinib depended on SIK1, we depleted SIK1 and found that SIK1 depletion significantly decreased the cell colony formation rate and elevated apoptosis rate after ruxolitinib administration (Fig. 6A and B; Supplementary Fig. S8B–S8C). Aligned with these findings, breast cancer cells ectopically expressing wild-type (WT) or S435A-mutant SIK1 displayed resistance to ruxolitinib compared with S435D-mutant SIK1-expressing cancer cells (Fig. 6C and D; Supplementary Fig. S8D–S8I). Similar results were observed with another JAK2 inhibitor, momelotinib (CYT387; Supplementary Fig. S8J–S8K). Next, xenograft mouse models were analyzed, and the results showed that ruxolitinib treatment significantly attenuated SIK1 S435D-induced tumor growth (Fig. 6EG), and this effect accompanied by a decreased pSTAT3 level (Fig. 6H; Supplementary Fig. S8L). To directly evaluate the effect of PI3K–AKT pathway on JAK2–STAT3 inhibitors, we observed that depletion of AKT markedly resisted breast cancer cells to JAK2 inhibitors, including both ruxolitinib and momelotinib (Supplementary Fig. S9A–S9D). Moreover, pharmacologic inhibition of AKT potentially enhanced breast cancer cell resistance to JAK2 inhibitors (Supplementary Fig. S9E), suggesting that repression of AKT decreases STAT3 activity, which leads to breast cancer cell resistance to JAK2 inhibitors. Together, these findings indicate that the JAK2–STAT3 axis is a potential clinical target in hyperactivated AKT1- or SIK1 deletion-induced breast cancer malignancies.

In addition to targeting therapies by antagonizing ER or HER2, platinum-based drugs, including carboplatin (CBP), have been broadly used in the clinic for breast cancer intervention (37–39). Thus, we aimed to determine whether JAK2–STAT3 inhibitors show synergistic effects when combined with CBP for breast cancer therapy. To this end, we measured the synergistic effect of CBP and ruxolitinib in breast cancer cells and observed that the combination of these two drugs achieved a greater synergetic effect than treatment with either drug alone (Supplementary Fig. S10A). In line with this observation, the combination of CBP with ruxolitinib not only abrogated breast cancer cell colony formation (Fig. 6I; Supplementary Fig. S10B) but also significantly retarded tumor growth in vivo compared with either individual treatment (Fig. 6JL; Supplementary Fig. S10C–S10D), highlighting the promising strategy of combining CBP and ruxolitinib for combating hyperactivated AKT-driven breast cancers.

The roles of the PI3K–AKT axis in breast cancer have garnered considerable attention in recent years due to its frequent genetic alterations and aberrant activation (6–9, 40). Therefore, many studies have been devoted to developing inhibitors to target the PI3K–AKT pathway for breast cancer therapy; however, this treatment is not promising (10–12, 41), mainly due to the higher probability that this treatment induces toxicity and causes intolerable side effects (42, 43). Therefore, both upstream regulators and downstream effectors of AKT have been subjected to extensive investigation (44, 45). Via MS-based interaction protein identification, we identified AKT as a partner and upstream kinase of SIK1 to attenuate the tumor-suppressive roles of SIK1. We also found that AKT directly phosphorylates SIK1, a physiologic repressor of STAT3, to boost STAT3 oncogenic functions, facilitating breast tumorigenesis. Hence, we propose a strategy in which the JAK2–STAT33 axis is targeted to combat AKT-driven breast cancers. LKB1, a canonical upstream kinase of SIK1 (46), can also phosphorylate and activate SIK1, but whether and how LKB1-mediated phosphorylation and AKT-mediated regulation of SIK1 antagonize each other or act in a stimulus-dependent manner to maintain SIK1 biological functions remain unclear. Moreover, to determine whether LKB1 affects the SIK1–STAT3 axis under physiologic conditions to repress tumorigenesis (Fig. 7A), further investigation is needed.

In contrast to AMPK, the regulation of which has been extensively studied, SIK1 functions and upstream regulators have not yet been extensively characterized. In this study, via MS-based interaction protein identification, we identified SIK1-binding proteins, such as 14-3-3, AKT, ITCH, and STAT3, and their potential interplay. Of note, AKT directly phosphorylates SIK1, which promotes 14-3-3 binding with and transfer of SIK1 to the cytoplasm, where SIK1 is recognized and subsequently degraded by ITCH as facilitated by Pin1 (Fig. 7B). Thus, impairment of any links in this process, including those involving AKT, 14-3-3, Pin1, or ITCH, would attenuate SIK1 degradation and its tumor-suppressive functions. Therefore, the frequent activation of AKT induced by genomic alterations, such as mutation/amplification of PIK3CA, HER2, PTEN, or AKT (6), contributes to SIK1 phosphorylation and degradation to promote breast tumorigenesis. However, in vivo studies are warranted to confirm whether SIK1 is critical for mediating AKT oncogenic functions; these studies can be performed by generating Sik1-S435A knock-in mice, which can be further subjected to crossbreeding with mammary-specific erbb2 transgenic mouse models (47, 48).

The predominant roles played by JAK2–STAT3 in tumorigenesis have been established in various cancers (16, 20, 27, 49), and specific inhibitors targeting JAK2 or STAT3 have been developed for evaluation in clinical trials of different cancer therapies, including breast cancer (20, 27). Notably, a special type of breast cancer (inflammatory breast cancer) has been shown to be sensitive to JAK2 inhibitors (50). In this study, we revealed that hyperactivated AKT activates STAT3 signaling by repressing SIK1 activity. Due to the frequent activation of the PI3K–AKT pathway in breast cancer (43), targeting the JAK2–STAT3 pathway may be an alternative treatment that partially reduces AKT oncogenic function by restoring SIK1 activity. Moreover, other events involved in regulating SIK1, such as deletion of LKB1 and amplification of ITCH or Pin1, would also contribute to repressing SIK1 and possibly result in STAT3 activation. Therefore, these tumors may also be reduced by JAK2/STAT3-targeted therapies, which are therefore worthy of further investigation. Interestingly, the combination of JAK2 inhibitors with the clinically approved chemodrug CBP has been confirmed to reduce AKT activation or SIK1 repression-driven breast cancers in a synergistic manner, which is consistent with a recent finding showing that the combination treatment of paclitaxel and JAK2/STAT3 inhibitors prevented breast cancer chemotherapy resistance (50). In summary, in this study, we identify a novel negative regulatory role for the tumor suppressor SIK1, which is activated via AKT-induced phosphorylation, that mediates STAT3-related breast oncogenesis and highlights a strategy to combat hyperactivated AKT-driven breast cancers by utilizing JAK2/STAT3 inhibitors.

Z. Sun reports other support from the National Natural Science Foundation of China during the conduct of the study. B. Gao reports grants from the National Natural Science Foundation of China during the conduct of the study. No disclosures were reported by the other authors.

Z. Sun: Data curation, formal analysis, methodology, writing–original draft, writing–review and editing. Q. Jiang: Data curation, validation, investigation, methodology, writing–review and editing. B. Gao: Data curation, validation, visualization, methodology, writing–review and editing. X. Zhang: Validation, investigation, methodology. L. Bu: Validation, investigation, methodology. L. Wang: Resources, methodology. Y. Lin: Resources, formal analysis, validation, methodology. W. Xie: Validation, investigation, visualization, methodology. J. Li: Resources, supervision, project administration, writing–review and editing. J. Guo: Conceptualization, supervision, funding acquisition, project administration.

The authors thank members of the Guo laboratory and Li laboratory for critical reading and kind suggestions for the manuscript. The authors thank Xiaolin Tian and Dr. Haiteng Deng at the Center of Protein Analysis Technology, Tsinghua University for MS analysis. This work was supported by the National Natural Science Foundation of China to J. Guo (32070767 and 31871410), China Postdoctoral Science Foundation (Q. Jiang 2020M683035).

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

1.
Siegel
RL
,
Miller
KD
,
Fuchs
HE
,
Jemal
A
.
Cancer statistics, 2021
.
CA Cancer J Clin
2021
;
71
:
7
33
.
2.
Siegel
RL
,
Miller
KD
,
Jemal
A
.
Cancer statistics, 2020
.
CA Cancer J Clin
2020
;
70
:
7
30
.
3.
Waks
AG
,
Winer
EP
.
Breast cancer treatment: a review
.
JAMA
2019
;
321
:
288
300
.
4.
Harbeck
N
,
Penault-Llorca
F
,
Cortes
J
,
Gnant
M
,
Houssami
N
,
Poortmans
P
, et al
.
Breast cancer
.
Nat Rev Dis Primers
2019
;
5
:
66
.
5.
Ross
JS
,
Gay
LM
.
Comprehensive genomic sequencing and the molecular profiles of clinically advanced breast cancer
.
Pathology (Phila)
2017
;
49
:
120
32
.
6.
Chen
L
,
Yang
L
,
Yao
L
,
Kuang
XY
,
Zuo
WJ
,
Li
S
, et al
.
Characterization of PIK3CA and PIK3R1 somatic mutations in Chinese breast cancer patients
.
Nat Commun
2018
;
9
:
1357
.
7.
Wang
X
,
Tokheim
C
,
Gu
SS
,
Wang
B
,
Tang
Q
,
Li
Y
, et al
.
In vivo CRISPR screens identify the E3 ligase Cop1 as a modulator of macrophage infiltration and cancer immunotherapy target
.
Cell
2021
;
184
:
5357
74
.
8.
Ciriello
G
,
Gatza
ML
,
Beck
AH
,
Wilkerson
MD
,
Rhie
SK
,
Pastore
A
, et al
.
Comprehensive molecular portraits of invasive lobular breast cancer
.
Cell
2015
;
163
:
506
19
.
9.
Hyman
DM
,
Smyth
LM
,
Donoghue
MTA
,
Westin
SN
,
Bedard
PL
,
Dean
EJ
, et al
.
AKT inhibition in solid tumors with AKT1 mutations
.
J Clin Oncol
2017
;
35
:
2251
9
.
10.
Kim
SB
,
Dent
R
,
Im
SA
,
Espié
M
,
Blau
S
,
Tan
AR
, et al
.
Ipatasertib plus paclitaxel versus placebo plus paclitaxel as first-line therapy for metastatic triple-negative breast cancer (LOTUS): a multicentre, randomised, double-blind, placebo-controlled, phase 2 trial
.
Lancet Oncol
2017
;
18
:
1360
72
.
11.
Sweeney
C
,
Bracarda
S
,
Sternberg
CN
,
Chi
KN
,
Olmos
D
,
Sandhu
S
, et al
.
Ipatasertib plus abiraterone and prednisolone in metastatic castration-resistant prostate cancer (IPATential150): a multicentre, randomised, double-blind, phase 3 trial
.
Lancet
2021
;
398
:
131
42
.
12.
André
F
,
Ciruelos
E
,
Rubovszky
G
,
Campone
M
,
Loibl
S
,
Rugo
HS
, et al
.
Alpelisib for PIK3CA-mutated, hormone receptor-positive advanced breast cancer
.
N Engl J Med
2019
;
380
:
1929
40
.
13.
Sun
Z
,
Jiang
Q
,
Li
J
,
Guo
J
.
The potent roles of salt-inducible kinases (SIKs) in metabolic homeostasis and tumorigenesis
.
Signal Transduct Target Ther
2020
;
5
:
150
.
14.
Hollstein
PE
,
Eichner
LJ
,
Brun
SN
,
Kamireddy
A
,
Svensson
RU
,
Vera
LI
, et al
.
The AMPK-related kinases SIK1 and SIK3 mediate key tumor-suppressive effects of LKB1 in NSCLC
.
Cancer Discov
2019
;
9
:
1606
27
.
15.
Murray
CW
,
Brady
JJ
,
Tsai
MK
,
Li
C
,
Winters
IP
,
Tang
R
, et al
.
An LKB1-SIK axis suppresses lung tumor growth and controls differentiation
.
Cancer Discov
2019
;
9
:
1590
605
.
16.
Verhoeven
Y
,
Tilborghs
S
,
Jacobs
J
,
De Waele
J
,
Quatannens
D
,
Deben
C
, et al
.
The potential and controversy of targeting STAT family members in cancer
.
Semin Cancer Biol
2020
;
60
:
41
56
.
17.
Miklossy
G
,
Hilliard
TS
,
Turkson
J
.
Therapeutic modulators of STAT signalling for human diseases
.
Nat Rev Drug Discov
2013
;
12
:
611
29
.
18.
Zhang
M
,
Zhou
L
,
Xu
Y
,
Yang
M
,
Xu
Y
,
Komaniecki
GP
, et al
.
A STAT3 palmitoylation cycle promotes T(H)17 differentiation and colitis
.
Nature
2020
;
586
:
434
9
.
19.
Kim
E
,
Kim
M
,
Woo
DH
,
Shin
Y
,
Shin
J
,
Chang
N
, et al
.
Phosphorylation of EZH2 activates STAT3 signaling via STAT3 methylation and promotes tumorigenicity of glioblastoma stem-like cells
.
Cancer Cell
2013
;
23
:
839
52
.
20.
Johnson
DE
,
O'Keefe
RA
,
Grandis
JR
.
Targeting the IL-6/JAK/STAT3 signalling axis in cancer
.
Nat Rev Clin Oncol
2018
;
15
:
234
48
.
21.
Jiang
Q
,
Zhang
X
,
Dai
X
,
Han
S
,
Wu
X
,
Wang
L
, et al
.
S6K1-mediated phosphorylation of PDK1 impairs AKT kinase activity and oncogenic functions
.
Nat Commun
2022
;
13
:
1548
.
22.
Adams
EJ
,
Karthaus
WR
,
Hoover
E
,
Liu
D
,
Gruet
A
,
Zhang
Z
, et al
.
FOXA1 mutations alter pioneering activity, differentiation and prostate cancer phenotypes
.
Nature
2019
;
571
:
408
12
.
23.
Guo
J
,
Cheng
J
,
Zheng
N
,
Zhang
X
,
Dai
X
,
Zhang
L
, et al
.
Copper promotes tumorigenesis by activating the PDK1-AKT oncogenic pathway in a copper transporter 1 dependent manner
.
Adv Sci (Weinh)
2021
;
8
:
e2004303
.
24.
Jiang
Q
,
Zheng
N
,
Bu
L
,
Zhang
X
,
Zhang
X
,
Wu
Y
, et al
.
SPOP-mediated ubiquitination and degradation of PDK1 suppresses AKT kinase activity and oncogenic functions
.
Mol Cancer
2021
;
20
:
100
.
25.
Xin
L
,
Liu
C
,
Liu
Y
,
Mansel
RE
,
Ruge
F
,
Davies
E
, et al
.
SIKs suppress tumor function and regulate drug resistance in breast cancer
.
Am J Cancer Res
2021
;
11
:
3537
57
.
26.
Ponnusamy
L
,
Manoharan
R
.
Distinctive role of SIK1 and SIK3 isoforms in aerobic glycolysis and cell growth of breast cancer through the regulation of p53 and mTOR signaling pathways
.
Biochim Biophys Acta Mol Cell Res
2021
;
1868
:
118975
.
27.
Zou
S
,
Tong
Q
,
Liu
B
,
Huang
W
,
Tian
Y
,
Fu
X
.
Targeting STAT3 in cancer immunotherapy
.
Mol Cancer
2020
;
19
:
145
.
28.
He
Y
,
Wang
L
,
Wei
T
,
Xiao
YT
,
Sheng
H
,
Su
H
, et al
.
FOXA1 overexpression suppresses interferon signaling and immune response in cancer
.
J Clin Invest
2021
;
131
:
e147025
.
29.
Huang
Y
,
Yang
M
,
Huang
W
.
14–3-3 σ: A potential biomolecule for cancer therapy
.
Clin Chim Acta
2020
;
511
:
50
8
.
30.
Pennington
KL
,
Chan
TY
,
Torres
MP
,
Andersen
JL
.
The dynamic and stress-adaptive signaling hub of 14–3-3: emerging mechanisms of regulation and context-dependent protein-protein interactions
.
Oncogene
2018
;
37
:
5587
604
.
31.
Sonntag
T
,
Vaughan
JM
,
Montminy
M
.
14–3-3 proteins mediate inhibitory effects of cAMP on salt-inducible kinases (SIKs)
.
Febs j
2018
;
285
:
467
80
.
32.
Koikawa
K
,
Kibe
S
,
Suizu
F
,
Sekino
N
,
Kim
N
,
Manz
TD
, et al
.
Targeting Pin1 renders pancreatic cancer eradicable by synergizing with immunochemotherapy
.
Cell
2021
;
184
:
4753
71
.
33.
Zhang
J
,
Chen
M
,
Zhu
Y
,
Dai
X
,
Dang
F
,
Ren
J
, et al
.
SPOP promotes nanog destruction to suppress stem cell traits and prostate cancer progression
.
Dev Cell
2019
;
48
:
329
44
.
34.
Ahmed
N
,
Zeng
M
,
Sinha
I
,
Polin
L
,
Wei
WZ
,
Rathinam
C
, et al
.
The E3 ligase Itch and deubiquitinase Cyld act together to regulate Tak1 and inflammation
.
Nat Immunol
2011
;
12
:
1176
83
.
35.
Suryaraja
R
,
Anitha
M
,
Anbarasu
K
,
Kumari
G
,
Mahalingam
S
.
The E3 ubiquitin ligase Itch regulates tumor suppressor protein RASSF5/NORE1 stability in an acetylation-dependent manner
.
Cell Death Dis
2013
;
4
:
e565
.
36.
Luo
Y
,
Alexander
M
,
Gadina
M
,
O'Shea
JJ
,
Meylan
F
,
Schwartz
DM
.
JAK-STAT signaling in human disease: from genetic syndromes to clinical inhibition
.
J Allergy Clin Immunol
2021
;
148
:
911
25
.
37.
von Minckwitz
G
,
Schneeweiss
A
,
Loibl
S
,
Salat
C
,
Denkert
C
,
Rezai
M
, et al
.
Neoadjuvant carboplatin in patients with triple-negative and HER2-positive early breast cancer (GeparSixto; GBG 66): a randomised phase 2 trial
.
Lancet Oncol
2014
;
15
:
747
56
.
38.
Sikov
WM
,
Berry
DA
,
Perou
CM
,
Singh
B
,
Cirrincione
CT
,
Tolaney
SM
, et al
.
Impact of the addition of carboplatin and/or bevacizumab to neoadjuvant once-per-week paclitaxel followed by dose-dense doxorubicin and cyclophosphamide on pathologic complete response rates in stage II to III triple-negative breast cancer: CALGB 40603 (Alliance)
.
J Clin Oncol
2015
;
33
:
13
21
.
39.
Loibl
S
,
O'Shaughnessy
J
,
Untch
M
,
Sikov
WM
,
Rugo
HS
,
McKee
MD
, et al
.
Addition of the PARP inhibitor veliparib plus carboplatin or carboplatin alone to standard neoadjuvant chemotherapy in triple-negative breast cancer (BrighTNess): a randomised, phase 3 trial
.
Lancet Oncol
2018
;
19
:
497
509
.
40.
Xu
J
,
Yu
X
,
Martin
TC
,
Bansal
A
,
Cheung
K
,
Lubin
A
, et al
.
AKT degradation selectively inhibits the growth of PI3K/PTEN pathway-mutant cancers with Wild-Type KRAS and BRAF by destabilizing aurora kinase B
.
Cancer Discov
2021
;
11
:
3064
89
.
41.
André
F
,
Ciruelos
EM
,
Juric
D
,
Loibl
S
,
Campone
M
,
Mayer
IA
, et al
.
Alpelisib plus fulvestrant for PIK3CA-mutated, hormone receptor-positive, human epidermal growth factor receptor-2-negative advanced breast cancer: final overall survival results from SOLAR-1
.
Ann Oncol
2021
;
32
:
208
17
.
42.
Verret
B
,
Cortes
J
,
Bachelot
T
,
Andre
F
,
Arnedos
M
.
Efficacy of PI3K inhibitors in advanced breast cancer
.
Ann Oncol
2019
;
30
(
Suppl_10
):
x12
20
.
43.
Alzahrani
AS
.
PI3K/Akt/mTOR inhibitors in cancer: At the bench and bedside
.
Semin Cancer Biol
2019
;
59
:
125
32
.
44.
Manning
BD
,
Toker
A
.
AKT/PKB signaling: navigating the network
.
Cell
2017
;
169
:
381
405
.
45.
Hoxhaj
G
,
Manning
BD
.
The PI3K-AKT network at the interface of oncogenic signalling and cancer metabolism
.
Nat Rev Cancer
2020
;
20
:
74
88
.
46.
Liu
L
,
Siu
FM
,
Che
CM
,
Xu
A
,
Wang
Y
.
Akt blocks the tumor suppressor activity of LKB1 by promoting phosphorylation-dependent nuclear retention through 14–3-3 proteins
.
Am J Transl Res
2012
;
4
:
175
86
.
47.
Ursini-Siegel
J
,
Schade
B
,
Cardiff
RD
,
Muller
WJ
.
Insights from transgenic mouse models of ERBB2-induced breast cancer
.
Nat Rev Cancer
2007
;
7
:
389
97
.
48.
Young
CD
,
Nolte
EC
,
Lewis
A
,
Serkova
NJ
,
Anderson
SM
.
Activated Akt1 accelerates MMTV-c-ErbB2 mammary tumourigenesis in mice without activation of ErbB3
.
Breast Cancer Res
2008
;
10
:
R70
.
49.
Yu
H
,
Lee
H
,
Herrmann
A
,
Buettner
R
,
Jove
R
.
Revisiting STAT3 signalling in cancer: new and unexpected biological functions
.
Nat Rev Cancer
2014
;
14
:
736
46
.
50.
Stevens
LE
,
Peluffo
G
,
Qiu
X
,
Temko
D
,
Fassl
A
,
Li
Z
, et al
.
JAK-STAT signaling in inflammatory breast cancer enables chemotherapy-resistant cell states
.
Cancer Res
2022
.