Impaired cyclin D3 protein degradation contributes to trastuzumab resistance in HER2 positive breast cancer Authors: Zhuo Wang, Haiqi Lu, Yiming Zhong, Lifeng Feng, Hongchuan Jin, Xian Wang Published: 2024 ABSTRACT As the first anti-HER2 targeted agent approved by FDA in 1998, Trastuzumab has significantly improved the outcome of patients with HER2 positive metastatic breast cancer. Unfortunately, resistance to trastuzumab is a severe obstacle to its therapeutic efficacy in clinical application, and its mechanism has not yet been fully elucidated. In our study, we found that stabilization of cyclin D3 could be one reason for trastuzumab resistance. Trastuzumab could induce G1/G0 phase arrest by downregulating cyclin D3 protein expression. However, the protein expression of cyclin D3 was not affected in trastuzumab-resistant cells, which might be related to aberrant activation of ERK signaling pathway. Furthermore, degradation of cyclin D3 protein by trastuzumab was mainly resulted from ubiquitin-dependent proteasome mechanism instead of transcriptional regulation. In trastuzumab-resistant breast cancer cells, trastuzumab-induced degradation of cyclin D3 protein was abrogated. When the ubiquitin pathway was inhibited, cells would show a predisposition to resistance to trastuzumab. Further, CDK4/6 inhibitor can inhibit the proliferation of trastuzumab-resistant HER-2 positive breast cancer cells. Therefore, combination of CDK4/6 inhibitors and anti-HER2 targeted therapy may be an alternative and promising strategy to overcome trastuzumab resistance in the future. Introduction Breast cancer is the most common type of cancer diagnosed among women and a common leading cause of cancer-related mortality among women worldwide [1, 2]. In China the incidence of breast cancer also ranks the second among female malignancies according to the data published on the Chinese National Cancer Center [3]. As a kind of complex and highly heterogeneous diseases, breast cancer can be mainly classified into three types including hormone-receptor positive, human epidermal growth factor receptor-2(HER2)-positive and triple negative breast cancer (TNBC). HER2 is overexpressed in approximately 20%-30% of breast cancer patients' tissues. Trastuzumab (Herceptin), as the first commercially available targeted therapeutic agent against HER2, has dramatically improved the outcome of the majority of HER2 positive breast cancer patients and significantly changing the natural biology of HER2-overexpressed breast cancer [4–6]. In spite of its good therapeutic effect, nearly 50% of patients still develop resistance to trastuzumab, and the majority of patients develop resistance within one year [7]. Therefore, investigation of the mechanisms related to trastuzumab resistance and sequentially exploring alternative clinical therapeutic strategies to overcome this resistance is an urgent need. Cell cycle is a crucial fundamental to life. Three homologous cyclin D-type proteins including cyclin D1, cyclin D2 and cyclin D3 have been well-identified, which plays a pivotal function for the successful progression of cell cycle through G1/G0 phase [8–10]. It has been discovered that cyclin D3 is widespread expressed in various tissues, especially in some cancer cells or tissues such as leukemic T cell lines, breast cancer, gastric cancer and esophageal adenocarcinoma [11–14]. Remarkably, cyclin D3 has also been reported to correlate with the prognosis of breast cancer patients [15]. Cylin-dependent kinases (CDK) 4/6 forms complexes with cyclin D family, including cyclin D3, which are essential for the transition from G1 to S phase. Previous studies have demonstrated that CDK4/6 inhibitor can target this complex effectively suppress proliferation of breast cancer cell line[16]. PD0332991(Palbociclib) has been approved for use in hormone-receptor(HR)-positive breast cancer, and some clinical trials have shown promising efficacy in HER2 positive breast cancers as well [17, 18]. Ubiquitin proteasome system (UPS) is one of the most vital modes of protein degradation in cells in vivo and is an important post-translational modification process in organism, which plays an important role in some biological behaviors including cell proliferation, apoptosis, differentiation and response to stress [19, 20]. Culling-Ring Ligase (CRL) complex is a large class of multi-subunit ubiquitin ligase complexes in cells, which is involved in the regulation of many crucial physiological and pathological processes such as apoptosis, cell cycle, DNA replication and gene transcription [21]. There are currently seven members in Cullin protein family, including Cul1, Cul2, Cul3, Cul4A, Cul4B, Cul5, Cul7 and Cul9 [22]. Intriguingly, SCF complex (SKP, Cullin, F-box containing complex), consisting of SKP1(s phase kinase-associated protein 1), Cullin protein family, RBX/ROC (RING-box) proteins and the corresponding F-box proteins, is one of the most representative subgroups of CRL family. Cul1 usually acts as a scaffold to connect to the SKP1-F-box and Rbx1 subunits. Through this binding, the corresponding specific substrates can be recognized and then degraded simply by replacing the different F-box proteins. The Cullin protein family, which is an important component of CRL, is dependent on Neddylation modification by linking to Cullin through the ubiquitin-like small molecule NEDD8, which gives rise to a change in CRL conformation and consequently activating CRL and finally promoting substrate degradation [23]. Several reports have shown that cyclin D1 and cyclin D3 are regulated by the SCF complex-mediated protein degradation [24, 25]. Strikingly, it has been recently demonstrated that abnormal cyclin D1 degradation is associated with tamoxifen resistance in breast cancer [26]. In this study, we found that cyclin D3 protein expression can be degradated by ubiquitin–proteasome system in response to trastuzumab treatment. Impaired ubiquitination-mediated degradation of cyclin D3 led to trastuzumab resistance. Thus, combination of CDK4/6 inhibitors and anti-HER2 targeted therapy may be an alternative and promising strategy to overcome trastuzumab resistance in breast and other HER2-driven cancers. Materials and methods Cell culture HEK-293 and Human breast cancer cell lines including BT474 and SKBR-3 are obtained from American Type Culture Collection (ATCC, Manassas, VA, USA), which were cultured in RPIM-1640 medium with 10% FBS for BT474 and DMEM medium with 10% FBS for SKBR-3 and HEK-293 in 37 °C with 5% CO2. The construction of the BT474 trastuzumab-resistant cell line (BT474 HR) has been described previously [27]. Transfection Cells were transfected with different plasmids using the X-treme GENE HP DNA Transfection Reagent Kit (Roche, Switzerland). The siRNAs for CCND3, FBXL2, CUL1 and their negative control siRNAs used in this experiment were designed and synthesized from Gemma Pharmaceuticals (Shanghai, China). Cells were transfected with siRNA (20 nM) using Lipofectamine RNAiMAX transfection reagent (Invitrogen, USA). Antibodies and reagents Trastuzumab (Herceptin, Roche, Germany) is a commercially available drug, which was prepared by dissolving the drug into a 20 mg/mL master mix. List of antibodies used in this study are as follows: CUL-1 (Santa Cruz Biotechnology, sc-11384), FLAG (Sigma-Aldrich, F7425-2MG), Actin (Cell Signaling Technology, USA), Cyclin D3 (Cell Signaling Technology, USA), Phospho-ERK1/2 (Cell Signaling Technology, USA), ERK1/2 (Cell Signaling Technology, USA), AKT (ps473) (Epitomics, USA), AKT1 (Abcam, UK), Flag (Sigma aldrich, USA), UB (Santa Cruz Biotechnology, USA), Phospho-RB (Cell Signaling Technology, USA), CDK4 (Cell Signaling Technology, USA), MG132 (MedChem Express, HY-13259).MLN4924 (ApexBio, B1036) was as a gift from Takeda Pharmaceuticals. Plasmid construction The Flag-CCND3 plasmid was cloned to EX05-flag vector. The open reading frame (ORF) of human CCND3 was cloned into pCMVNeo05 (EX05) using Xbal and BstBI enzymatic sites. The HA-UB plasmid was presented from Yi Sun (Zhejiang University, Hangzhou, Zhejiang, China). Cell proliferation and flow cytometry Cell proliferation was performed using the standard method of 3-(4, 5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS). For cell cycle analysis, cell suspensions were obtained and added to ethanol solution for fixation, then suspended PI/RNase staining buffer (BD Pharmingen, USA). BrdU experiments were performed using Brdu Flow Kits (BD Pharmingen, USA), the instructions of which were followed. RNA extraction and RT-qPCR In this experiment, we used SYBR green real-time fluorescence quantitative RT-PCR to determine the expression level of mRNA. After extracting total cellular RNA with TRIzol Reagent (Invitrogen, USA), the concentration of RNA was determined using NanoDrop 2000 (Nanodrop, Wilmington, Del., USA). Following the instructions, 2 μg of RNA was taken and the reverse transcription reaction was performed using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems, USA). RT-PCR primers were designed using Primer Premier 5. Real-time quantitative RT-PCR was performed on a PCR Lightcycler480 (Roche, Germany). mRNA expression levels were determined using Actin primers as control internal reference (Tables 1 and 2). Table 1siRNA sequencesiRNASequenceCCND3 #1S:GCUGUGAUUGCACAUGAUUTTAS: AAUCAUGUGCAAUCACAGCTTCCND3 #2S: GACCAGCACUCCUACAGAUTTAS: AUCUGUAGGAGUGCUGGUCTTFBXL2 #1S: GGAUGGAAGCAACUGGCAATTAS: UUGCCAGUUGCUUCCAUCCTTFBXL2 #2S:CCUCACAGAUGCCUCUCUUTTAS:AAGAGAGGCAUCUGUGAGGTTCUL1 #1S: GCUCUACACUCAUGUUUAUTTAS: AUAAACAUGAGUGUAGAGCTTCUL1 #2S: GAACCCAGUUACUGAAUAUTTAS: AUAUUCAGUAACUGGGUUCTTTable 2Primers sequencePrimersSequenceActinF: CACCAACTGGGACGACATR: ACAGCCTGGATAGCAACGCCND3F: TAGGGAAGCTCAAGTGGGAR: ATCATGGATGGCGGGTACFBLX2F: CGAAACATTGAACATTTGAACCR: TCGACAACCTCGCACCAG Table 1siRNA sequencesiRNASequenceCCND3 #1S:GCUGUGAUUGCACAUGAUUTTAS: AAUCAUGUGCAAUCACAGCTTCCND3 #2S: GACCAGCACUCCUACAGAUTTAS: AUCUGUAGGAGUGCUGGUCTTFBXL2 #1S: GGAUGGAAGCAACUGGCAATTAS: UUGCCAGUUGCUUCCAUCCTTFBXL2 #2S:CCUCACAGAUGCCUCUCUUTTAS:AAGAGAGGCAUCUGUGAGGTTCUL1 #1S: GCUCUACACUCAUGUUUAUTTAS: AUAAACAUGAGUGUAGAGCTTCUL1 #2S: GAACCCAGUUACUGAAUAUTTAS: AUAUUCAGUAACUGGGUUCTT siRNA sequence Table 2Primers sequencePrimersSequenceActinF: CACCAACTGGGACGACATR: ACAGCCTGGATAGCAACGCCND3F: TAGGGAAGCTCAAGTGGGAR: ATCATGGATGGCGGGTACFBLX2F: CGAAACATTGAACATTTGAACCR: TCGACAACCTCGCACCAG Primers sequence Western blot and immunoprecipitation Cell lysis was performed using CytoBuster Protein Extraction (Novus Biologicals, USA) and the BCA Protein Concentration Assay Kit (Bio-Rad Laboratories, Hercules, CA, USA). In vitro ubiquitination assay, the cells were transfected with HA-Ub or Flag-CCND3 for 48 h, then the cells were treated with MG132, after which proteins were lysed. 1 ml ubiquitination lysis solution should contain N-Ethylamaleimide 5 mM and I-doacetamide 10 mM in addition to MG132 25 μM. To immunoprecipitate proteins, 500 μg of corresponding proteins were taken and incubated overnight with 5ug corresponding antibodies, then protein G sepharose 4 fast slow Beads (GE healthcare, USA) were added and incubated for 2–4 h, then centrifuged to obtain protein lysates. The lysate was denatured to SDS-PAGE, then transferred to PVDF membranes. Incubation with primary antibody was followed. PVDF membranes were subsequently washed with TBS-T (TBS containing 0.1% Tween-20) and the corresponding HRP-conjugated secondary antibody (Dako, Glostrup, Denmark) were used. The membranes were visualized on an Amersham Imager 600 system (GE Healthcare Life Sciences, Shanghai, China) using the developer ECL (EMD Millipore, Billerica, MA, USA). The grayscale values corresponding to the internal reference bands of each protein sample were determined using Image Studio software. TCGA database analysis We select the data set of breast cancer patients through TCGA Data Portal that contains both Level 3 expression microarray data and clinical data in XML format; apply the XML package in R software to parse the clinical data in XML format, extract the information necessary from the data and combine them into the corresponding expression microarray data, summarize all the expression microarray data together. Clone formation assay 2*1^5 cells were seeded into 6-well plates overnight, and replaced with culture medium containing the corresponding drug treatment for 72 h, then the medium was replaced with normal culture medium. After 2 weeks of incubation, crystalline violet staining was performed to fix the cells. Statistical analysis All experiments were repeated at least three times independently and statistical analysis was performed using student's t-test or one-way ANOVA, p < 0.05 was statistically significant. Results Trastuzumab failed to induce cell cycle arrest in trastuzumab-resistant breast cancer cells Firstly, BT474 HR (Herceptin Resistant) cell line has been established and demonstrated previously by culturing BT474 cells with trastuzumab [27]. To further determine the difference in the sensitivity of BT474 cells and BT474 HR cells to trastuzumab, MTS assay was used to detect cell viability after treatment with different concentrations of trastuzumab. According to the results, trastuzumab significantly inhibited the proliferation of BT474 cells with increasing concentrations of trastuzumab (using SKBR-3, another HER2 overexpressing cell line as a positive control), while the inhibitory effect was not as significant as that in BT474 HR cells (Fig. 1A). Since trastuzumab is a monoclonal antibody targeting HER2, it is appealing to assume that HER2 probably could resist trastuzumab binding by undergoing breakage of the extracellular segment. In order to explore the mechanism of Herceptin resistance in BT474 HR cells, we firstly detected the expression of HER2 protein both in BT474 cells and BT474 HR cells. However, western blot showed that there was no remarkable difference in the expression of HER2 protein both in BT474 cells and BT474 HR cells. Secondly, immunoprecipitation experiments revealed that trastuzumab was able to bind to HER2 protein in both trastuzumab-sensitive and trastuzumab-resistant cells. These results suggested that HER2 protein expression was not strikingly altered even in trastuzumab-resistant cells (Fig. 1B).Fig. 1Trastuzumab failed to induces cell cycle arrest in Trastuzumab-resistant breast cancer cells. A MTS assay to detect the activity of different concentrations of trastuzumab treated 72 h (left, * p < 0.05).BT474 cells and BT474 HR cells, (right). SKBR-3 cells. B Immunoprecipitation assay using Trastuzumab in BT474 and BT474 HR cells. C, D Cell cycle and BrdU assay performed by flow cytometry technique to detect variation of different concentrations of trastuzumab after 48 h treatment of BT474 cells and BT474 HR cells Trastuzumab failed to induces cell cycle arrest in Trastuzumab-resistant breast cancer cells. A MTS assay to detect the activity of different concentrations of trastuzumab treated 72 h (left, * p < 0.05).BT474 cells and BT474 HR cells, (right). SKBR-3 cells. B Immunoprecipitation assay using Trastuzumab in BT474 and BT474 HR cells. C, D Cell cycle and BrdU assay performed by flow cytometry technique to detect variation of different concentrations of trastuzumab after 48 h treatment of BT474 cells and BT474 HR cells To further clarify the causes for the trastuzumab-mediated inhibition of cell proliferation, cell cycle and BrdU assays were examined by flow cytometry (Fig. 1C, D). Cell cycle experiments showed a G1/G0 phase arrest in sensitive but not resistant cells with increasing trastuzumab concentrations (Fig. 1C). In addition, cell proliferation inhibition was further demonstrated by BrdU assay (Fig. 1D). Collectively, these results further indicated that trastuzumab could cause a G1/G0 phase cell cycle arrest in trastuzumab-sensitive but not trastuzumab-resistant cells. Loss of Cyclin D3 protein expression down regulation confers Trastuzumab resistance in breast cancer To further elucidate the mechanism of G1/G0 phase arrest by trastuzumab in sensitive cells, the alteration of cyclin D1, cyclin E1 and cyclin D3 proteins after trastuzumab treatment were detected by western blot (Fig. 2A). The results showed that only the expression of cyclin D3 protein was dramatically downregulated in sensitive but not resistant cells after trastuzumab treatment. Similarly, only the protein expression of cyclin D3 rather than cyclin D1 and cyclin E1 was found to be downregulated in SKBR-3 cells. These results suggested that the failure to downregulate cyclin D3 protein expression may be involved in trastuzumab resistance. It is well acknowledged that AKT and ERK are located in the downstream of HER2 signaling pathway, which are closely associated with cell growth, survival and proliferation. Therefore, we further detected the changes in these proteins’ expression after trastuzumab treatment (Fig. 2B). These results showed that p-AKT was significantly suppressed in trastuzumab-sensitive BT474 cells after trastuzumab treatment and partially reduced in BT474 HR cells, whereas p-ERK was only reduced in BT474 cells but not in BT474 HR cells after trastuzumab treatment. These data imply that phosphorylation of ERK may be probably associated with trastuzumab resistance.Fig. 2Loss of Cyclin D3 protein expression down regulation confers Trastuzumab resistance in breast cancer. A BT474, BT474 HR and SKBR-3 cells were treated with different concentrations of trastuzumab for 24 h to detect the expression of relevant cyclin proteins. Actin was used as the loading control, relative cyclin E1 expression (cyclin E1/Actin), cyclin D1 expression (cyclin D1/Actin) cyclin D3 expression (cyclin D3/Actin) was quantified and normalized to the relative protein expression in the control without trastuzumab. B BT474 and BT474 HR cells were treated with different concentrations of trastuzumab for 24 h to detect the protein expression of AKT and ERK. C BT474 and SKBR-3 cells were treated trastuzumab 10 μg/mL for 12 h, RT-qPCR to detect the mRNA expression of cyclin D3. D mRNA expression of cyclin D3 in BT474 cells and BT474 HR cells were detected by RT-qPCR. E BT474, BT474 HR and SKBR-3 cells were transfected with CCND3 siRNA for 72 h, western bolt was to validate knockdown effectiveness (above), and the cell cycle was detected by flow cytometry(down). F 409 cases of HER2 positivity were analyzed by TCGA public database. The relationship between CCND3 expression status and prognosis in breast cancer patients was determined by the median expression of CCND3 in the population as cutoff Loss of Cyclin D3 protein expression down regulation confers Trastuzumab resistance in breast cancer. A BT474, BT474 HR and SKBR-3 cells were treated with different concentrations of trastuzumab for 24 h to detect the expression of relevant cyclin proteins. Actin was used as the loading control, relative cyclin E1 expression (cyclin E1/Actin), cyclin D1 expression (cyclin D1/Actin) cyclin D3 expression (cyclin D3/Actin) was quantified and normalized to the relative protein expression in the control without trastuzumab. B BT474 and BT474 HR cells were treated with different concentrations of trastuzumab for 24 h to detect the protein expression of AKT and ERK. C BT474 and SKBR-3 cells were treated trastuzumab 10 μg/mL for 12 h, RT-qPCR to detect the mRNA expression of cyclin D3. D mRNA expression of cyclin D3 in BT474 cells and BT474 HR cells were detected by RT-qPCR. E BT474, BT474 HR and SKBR-3 cells were transfected with CCND3 siRNA for 72 h, western bolt was to validate knockdown effectiveness (above), and the cell cycle was detected by flow cytometry(down). F 409 cases of HER2 positivity were analyzed by TCGA public database. The relationship between CCND3 expression status and prognosis in breast cancer patients was determined by the median expression of CCND3 in the population as cutoff The protein expression level is generally influenced by mRNA expression level, protein stability and mRNA translocation efficiency. To explore the mechanism of downregulation of cyclin D3 protein expression by trastuzumab, firstly we detected the mRNA expression of cyclin D3 (CCND3) in trastuzumab-treated BT474 and SKBR-3 cells (Fig. 2C). Indeed, cyclin D3 mRNA expression has no dramatical change in both BT474 and SKBR-3 cells, indicating that trastuzumab actually did not affect cyclin D3 mRNA expression. In addition, we also investigated the mRNA expression of cyclin D3 in BT474 cells and BT474 HR cells (Fig. 2D) and found that there was no visible difference between their baseline of cyclin D3 mRNA expression. In contrast, knockdown of cyclin D3 using siRNA resulted in G1/G0 phase arrest in both BT474 and BT474 HR cells (Fig. 2E). Furthermore, we analyzed the prognostic influence of CCND3 mRNA expression levels in 409 HER2-overexpressed breast cancer patients in TCGA data base (Fig. 2F), with the median expression of CCND3 as cutoff value. This data illustrated that CCND3 expression level was not significant relevant to patients’ survival (p = 0.1). Taken together, these results suggested that we cannot directly determine the function of trastuzumab to cyclin D3 by simply detecting cyclin D3 mRNA expression level in tissues, we probably need to compare the changes in the expression of cyclin D3 protein levels before and after trastuzumab treatment. Further, these results suggested that trastuzumab-induced degradation of cyclin D3 protein expression may be associated with protein stability, and the failure of down-regulate cyclin D3 protein expression caused trastuzumab-resistant. Relevance of cyclin D3 protein expression to Trastuzumab-induced G1/G0 phase arrest To identify a further association between the timing of cyclin D3 protein changes and cell cycle arrest, BT474 cells and SKBR-3 cells were treated with trastuzumab for 8 h, 10 h, and 12 h, then we monitored the expression of cyclin D3 protein by western blot (Fig. 3A) and cell cycle distribution by flow cytometry (Fig. 3B). The results unraveled that trastuzumab treatment at these three time points did not show a significant G1/G0 phase arrest but cyclin D3 protein expression levels have already been downregulated. Further trastuzumab treatment of BT474 cells and SKBR-3 cells for 24 h, flow cytometric cell cycle detection demonstrated the appearance of G1/G0 phase arrest (Fig. 3C). Therefore, changes of cyclin D3 protein expression precedes cell cycle arrest, suggesting that aberrant cyclin D3 expression should be the cause rather than the consequence of G1/G0 phase cell cycle arrest.Fig. 3Relevance cyclin D3 protein expression to Trastuzumab-induced G1/G0 phase arrest. A. BT474 cells and SKBR-3 cells were treated with trastuzumab 10 μg/mL for 8 h, 10 h, and 12 h to monitor downregulation of cyclin D3 protein expression. Actin was used as the loading control, relative cyclin D3 expression (cyclin D3/Actin) was quantified and normalized to the relative cyclin D3 expression in the control without trastuzumab at the same time. B. BT474 cells and SKBR-3 cells were treated with trastuzumab 10 μg/mL for 8 h, 10 h, and 12 h, flow cytometry to assay the cell cycle. C. BT474 cells and SKBR-3 cells were treated with trastuzumab 10 μg/mL for 24 h, flow cytometry for cell cycle detection Relevance cyclin D3 protein expression to Trastuzumab-induced G1/G0 phase arrest. A. BT474 cells and SKBR-3 cells were treated with trastuzumab 10 μg/mL for 8 h, 10 h, and 12 h to monitor downregulation of cyclin D3 protein expression. Actin was used as the loading control, relative cyclin D3 expression (cyclin D3/Actin) was quantified and normalized to the relative cyclin D3 expression in the control without trastuzumab at the same time. B. BT474 cells and SKBR-3 cells were treated with trastuzumab 10 μg/mL for 8 h, 10 h, and 12 h, flow cytometry to assay the cell cycle. C. BT474 cells and SKBR-3 cells were treated with trastuzumab 10 μg/mL for 24 h, flow cytometry for cell cycle detection Trastuzumab induces ubiquitination-dependent proteasomal degradation of cyclinD3 in HER2 positive breast cancer cells Moreover, in order to investigate the influence of protein stability on cyclin D3 protein expression, we treated cells with several different autophagic lysosomal inhibitors including CQ, 3-MA, BafA1 and ubiquitin–proteasome inhibitor MG132. The results revealed that there is no significant change in cyclin D3 protein levels after inhibition of autophagy, whereas inhibition of the proteasome with MG132 obviously increased cyclin D3 protein expression (Fig. 4A), indicating that trastuzumab likely affected cyclin D3 protein stability by ubiquitin proteasome. Next, to further explore how the ubiquitin–proteasome system affects the degradation of cyclin D3 protein by trastuzumab, cells were co-treated with MG132 and trastuzumab. According to the results, we demonstrated that MG132 could reverse trastuzumab-induced downregulation of cyclin D3 protein expression, and its effect is concentration dependent in both BT474 and SKBR-3 cells. (Fig. 4B).Fig. 4Trastuzumab induces ubiquitination-dependent proteasomal degradation of cyclinD3 in HER2 positive breast cancer cells. A BT474 cells were treated with autophagic lysosomal inhibitors CQ 25 μM, 3-MA 1 mM and BafA1 100 nM for 24 h and ubiquitin proteasome inhibitor MG132 for 4 h, and the protein level expression of cyclin D3 was detected after different treatments. B BT474 cells and SKBR-3 cells were treated with trastuzumab 10 μg/mL in combination with MG132 at different concentrations for 24 h. The protein level expression of cyclin D3 was detected. Actin was used as the loading control, relative cyclin D3 expression (cyclin D3/Actin) was quantified and normalized to the relative cyclin D3 expression in the control without trastuzumab or MG132. C BT474 cells and SKBR-3 cells, were transfected with HA-UB plasmid for 72 h, MG132 10 μM with or without trastuzumab 10 μg/mL treatment for 12 h, IP assay was to detect the polyubiquitin status of cyclin D3. D. BT474 cells, were transfected with exogenous Flag-CCND3 plasmid for 48 h. Then the cells were treated with different concentrations of trastuzumab for 24 h, MLN-4924 for 24 h and MG132 for 4 h, respectively, to detect the changes in Flag protein expression Trastuzumab induces ubiquitination-dependent proteasomal degradation of cyclinD3 in HER2 positive breast cancer cells. A BT474 cells were treated with autophagic lysosomal inhibitors CQ 25 μM, 3-MA 1 mM and BafA1 100 nM for 24 h and ubiquitin proteasome inhibitor MG132 for 4 h, and the protein level expression of cyclin D3 was detected after different treatments. B BT474 cells and SKBR-3 cells were treated with trastuzumab 10 μg/mL in combination with MG132 at different concentrations for 24 h. The protein level expression of cyclin D3 was detected. Actin was used as the loading control, relative cyclin D3 expression (cyclin D3/Actin) was quantified and normalized to the relative cyclin D3 expression in the control without trastuzumab or MG132. C BT474 cells and SKBR-3 cells, were transfected with HA-UB plasmid for 72 h, MG132 10 μM with or without trastuzumab 10 μg/mL treatment for 12 h, IP assay was to detect the polyubiquitin status of cyclin D3. D. BT474 cells, were transfected with exogenous Flag-CCND3 plasmid for 48 h. Then the cells were treated with different concentrations of trastuzumab for 24 h, MLN-4924 for 24 h and MG132 for 4 h, respectively, to detect the changes in Flag protein expression To clarify the role of trastuzumab in ubiquitination-mediated degradation of cyclin D3 protein, BT474 and SKBR-3 cells overexpressed with exogenous HA-UB plasmid were treated with trastuzumab and MG132. Immunoprecipitation experiments were performed with cyclin D3 antibody to detect the interaction of trastuzumab on the binding of cyclin D3 to UB protein. According to the results, the level of polyubiquitination of cyclin D3 protein in the trastuzumab-treated cells was significantly elevated compared to MG132-treated control alone (Fig. 4C). Therefore, these results suggested that trastuzumab could promote the degradation of cyclin D3 protein by promoting the binding of cyclin D3 protein to UB protein. However, when the ubiquitination system was destroyed, trastuzumab would fail to degrade cyclin D3 protein. In addition, we observed that the expression of exogenous cyclin D3 protein was also downregulated after trastuzumab treatment but was upregulated after treatment with MLN-4924 (an inhibitor of SCF complex in CRL system) and MG132 after transferred with exogenous Flag-CCND3 plasmid into BT474 cells (Fig. 4D). Collectively, these data further confirmed that cyclin D3 protein can be degraded by trastuzumab via ubiquitin–proteasome system. Ubiquitination of cyclin D3 protein is regulated by SCF complex CRL (Culling-Ring Ligase) complex is an important family in ubiquitin–proteasome system. Several reports have confirmed that cyclin D3 can bind to CUL1 protein [28–30]. To determine the binding of cyclin D3 to CUL1 protein in in-vitro system, exogenous Flag-CCND3 plasmid was transfected in HEK-293 cells and immunoprecipitation experiments were performed. According to the results, we found that exogenous Flag-CCND3 can bind to CUL1 protein (Fig. 5A). Further studies revealed that knockdown of CUL1 counteracted partially trastuzumab-induced downregulation of cyclin D3 protein levels (Fig. 5B), implying that trastuzumab-induced downregulation of cyclin D3 protein could be related to CRL family. Furthermore, it has been well documented that FBXL2 is an E3 ligase of cyclin D3 protein. However, there was no significant increase in cyclin D3 protein expression after knockdown of FBXL2 using siRNA, thus FBXL2 probably did not affect the degradation of cyclin D3 protein in our in-vitro experimental system (Fig. 5C). Importantly, SCF complex is one of the most crucial subgroups of CRL family. So aiming to explore whether trastuzumab-induced downregulation of cyclin D3 protein could also be associated with SCF complex, MLN-4924, an inhibitor of SCF complex in the CRL system was used [31]. We observed an increased expression of cyclin D3 protein after treatment with MLN-4924 in BT474 and SKBR-3 cells, however trastuzumab cannot be able to downregulate the expression of cyclin D3 protein in MLN-4924-treated cells (Fig. 5D). These data further suggested that degradation of cyclin D3 by trastuzumab was mediated by SCF complex. In addition, in order to clarify that destruction in the SCF complex affected trastuzumab-induced cyclin D3 degradation, immunoprecipitation experiments were performed in BT474 cells after treated with trastuzumab, MLN-4924 and MG132. The results showed that trastuzumab effectively promoted cyclinD3 polyubiquitination, whereas the level of polyubiquitination was downregulated after MLN-4924 treatment. The above results suggested that MLN-4924 treatment markedly disturbed the combination of UB protein and cyclin D3, and consequently destroyed the degradation of cyclin D3 (Fig. 5E). Indeed, trastuzumab treatment alone or in combination with MLN-4924 treatment in BT474 cells by cell counting revealed that the trastuzumab-sensitive cells can develop into resistance to trastuzumab after MLN-4924 treatment (5F). These results suggested that destruction in the ubiquitination system (especially of SCF complex) of cyclin D3 degradation can lead to trastuzumab resistance in HER2 positive breast cancer cells.Fig. 5Ubiquitination of cyclin D3 protein is regulated by SCF complex and associated with Trastuzumab resistance. A Overexpression of Flag-CCND3 in HEK-293 cells 72 h after addition of MG132 treatment for 4 h followed by IP experiments, western blot was to observe the binding of exogenous cyclin D3 and CUL1. B Trastuzumab was added to BT474 cells 72 h after transfection with CUL1 siRNA, and the expression of CUL1 and cyclin D3 was detected by western blot. Actin was used as the loading control, relative cyclin D3 expression (cyclin D3/actin) was quantified and normalized to the relative cyclin D3 expression in the control. C BT474 cells were transfected with FBXL2 siRNA for 72 h, western blot was performed to detect the protein expression of cyclin D3(above), and RT-qPCR was to validate knockdown effectiveness. D BT474 cells and SKBR-3 cells were treated with or without trastuzumab 10 μg/mL in combination with MLN-4924. Actin was used as the loading control, relative cyclin D3 expression (cyclin D3/Actin) was quantified and normalized to the relative cyclin D3 expression in the control. E BT474 cells were overexpressed Flag-CCND3 and HA-UB, and after 72 h they were treated with trastuzumab 10 μg/mL and MLN-4924 100 nM for 24 h, respectively, before protein collection. MG132 25 μM treated cells for 4 h and IP experiments were performed. The polyubiquitin status of Flag-CCND3 was detected. F. Trastuzumab 10 μg/ml was added to BT474 cells and MLN4924 100 nM was added or not at the same time and cell counting was performed after 72 h (*p < 0.01) Ubiquitination of cyclin D3 protein is regulated by SCF complex and associated with Trastuzumab resistance. A Overexpression of Flag-CCND3 in HEK-293 cells 72 h after addition of MG132 treatment for 4 h followed by IP experiments, western blot was to observe the binding of exogenous cyclin D3 and CUL1. B Trastuzumab was added to BT474 cells 72 h after transfection with CUL1 siRNA, and the expression of CUL1 and cyclin D3 was detected by western blot. Actin was used as the loading control, relative cyclin D3 expression (cyclin D3/actin) was quantified and normalized to the relative cyclin D3 expression in the control. C BT474 cells were transfected with FBXL2 siRNA for 72 h, western blot was performed to detect the protein expression of cyclin D3(above), and RT-qPCR was to validate knockdown effectiveness. D BT474 cells and SKBR-3 cells were treated with or without trastuzumab 10 μg/mL in combination with MLN-4924. Actin was used as the loading control, relative cyclin D3 expression (cyclin D3/Actin) was quantified and normalized to the relative cyclin D3 expression in the control. E BT474 cells were overexpressed Flag-CCND3 and HA-UB, and after 72 h they were treated with trastuzumab 10 μg/mL and MLN-4924 100 nM for 24 h, respectively, before protein collection. MG132 25 μM treated cells for 4 h and IP experiments were performed. The polyubiquitin status of Flag-CCND3 was detected. F. Trastuzumab 10 μg/ml was added to BT474 cells and MLN4924 100 nM was added or not at the same time and cell counting was performed after 72 h (*p < 0.01) CDK4/6 inhibitor PD0332991 suppresses the proliferation of Trastuzumab-resistant breast cancer cells As cyclin D3 was stabilized to confer trastuzumab resistance, CDK4/6 inhibitor was therefore applied to overcome therapeutic resistance in HER-2 positive breast cancer [32]. Indeed, trastuzumab-resistant cells remained sensitive to CDK4/6 inhibitor PD0332991 (Fig. 6A). Moreover, treatment of BT474 cells and BT474 HR cells with PD0332991 revealed a significant inhibition of cyclin D3 protein expression along with prolonged treated time (Fig. 6B). Further treatment of BT474 cells and BT474 HR cells with different concentrations of PD0332991 showed that neither cylinD1 nor CDK4 protein expression could be downregulated by PD0332991, while cyclin D3 protein expression was significantly downregulated with increasing concentration of PD0332991. Importantly, the efficacy of PD0332991 in both cell lines was further verified by the cell variability (Fig. 6D) and plate cloning formation assay (Fig. 6E). BT474 cells often grew into clusters and could not easily form monoclonal clones, therefore we can only observe the change of morphology and cell density. These results showed that both trastuzumab and PD0332991 could significantly inhibit the proliferation of BT474 cells, and combination of these two drugs has a synergistic effect in the suppression of cell clusters and density. Therefore, trastuzumab combined with PD0332991 might be a rational strategy for managing trastuzumab resistance.Fig. 6CDK4/6 inhibitor PD0332991 suppresses the proliferation of Trastuzumab-resistant breast cancer cells. A BT474 cells and BT474 HR cells were treated with different concentrations of PD0332991, and cell proliferation was detected by MTS. B BT474 cells and BT474 HR cells were treated with PD0332991 5um for different times, and the change of cyclinD3 expression was observed by western blot. C BT474 cells and BT474 HR cells were treated with PD0332991 for different concentrations, and protein expression was observed by western blot. D BT474 cells and BT474 HR cells were treated with trastuzumab 10ug/ml and PD0332991 5 μm alone and in combination treated, then cell counting was performed after 72 h. E Trastuzumab-sensitive and resistant cells plate cloning assay, PD0332991 5 μm and trastuzumab 10 μg/ml alone and in combination treated BT474 cells or BT474 HR cells for 72 h, then replacement of normal medium, and cells proliferation were observed 2 weeks thereafter. FThe general process of cyclin D3 degradation by trastuzumab CDK4/6 inhibitor PD0332991 suppresses the proliferation of Trastuzumab-resistant breast cancer cells. A BT474 cells and BT474 HR cells were treated with different concentrations of PD0332991, and cell proliferation was detected by MTS. B BT474 cells and BT474 HR cells were treated with PD0332991 5um for different times, and the change of cyclinD3 expression was observed by western blot. C BT474 cells and BT474 HR cells were treated with PD0332991 for different concentrations, and protein expression was observed by western blot. D BT474 cells and BT474 HR cells were treated with trastuzumab 10ug/ml and PD0332991 5 μm alone and in combination treated, then cell counting was performed after 72 h. E Trastuzumab-sensitive and resistant cells plate cloning assay, PD0332991 5 μm and trastuzumab 10 μg/ml alone and in combination treated BT474 cells or BT474 HR cells for 72 h, then replacement of normal medium, and cells proliferation were observed 2 weeks thereafter. FThe general process of cyclin D3 degradation by trastuzumab Discussion Trastuzumab remains the standard first-line anti-HER2 targeted therapy regimen for patients with HER2 positive breast cancer. Patients who have been previously treated with trastuzumab can still benefit from trastuzumab in combination with other drugs even when they experience recurrence or metastasis, eventually improving their prognosis [33, 34]. It is believed that trastuzumab resistance mechanisms in breast cancer mainly include the inability of trastuzumab to bind effectively to HER2 or the activation of other tyrosine kinase receptor-related bypass alternative signaling pathways, which leads to the continuous activation of HER2 downstream PI3K signaling pathways [35]. Mechanically, Trastuzumab acts on the extracellular segment of HER2 that is a transmembrane receptor. When the extracellular segment of HER2 is cleaved by metalloproteinase, the binding site of trastuzumab to HER2 is blocked and eventually resulting in the decreased sensitivity of trastuzumab [36]. In this study, we found that the expression of HER2 protein was not altered both in trastuzumab sensitive and resistant cells. Furthermore, immunoprecipitation experiments showed that trastuzumab could bind to HER2 protein and inhibit downstream AKT phosphorylation in both sensitive and resistant cells, indicating that there are no changes in the conformation of HER2 protein and trastuzumab was still able to bind to HER2 protein both in trastuzumab sensitive and resistant cells. Nevertheless, although we demonstrated that abnormal alterations in the ERK pathway are also involved in the trastuzumab resistance process, the exact mechanism remains obscure and still needs to be further elucidated. Some studies have shown that tyrosine kinase receptor inhibitor lapatinib can partially alleviate trastuzumab resistance, but some clinical studies revealed that lapatinib is only effective in some breast cancer patients, indicating that other mechanisms of trastuzumab resistance may still exist. In summary, we suggest that trastuzumab resistance is closely associated with its downstream signaling pathway transduction. Cell cycle is integral to the maintenance of normal cell survival [37]. We found that trastuzumab could induce sensitive cells to arrest in G1/G0 phase, however resistant cells can overcome trastuzumab-induced G1/G0 phase arrest. Among G1-phase cell cycle-associated proteins, cyclin D1 and cyclin E1 have been reported to be involved in trastuzumab resistance [38, 39]. Cyclin D1 and cyclin E1 were found to be downregulated in sensitive cells but not in resistant cells after trastuzumab treatment. Intriguingly, we firstly confirmed that cyclin D3 protein expression was decreased after trastuzumab treatment in sensitive cells (both in BT474 and SKBR-3 cell lines), but was invariant with trastuzumab treatment in trastuzumab-resistant cells. However, as downstream of Her2, mTOR inhibitor was found to be unable to downregulate cyclin D1 and cyclin E but were able to downregulate cyclin D3 in the study [40]. In addition, knockdown of cyclin D3 expression by siRNA in both trastuzumab-sensitive and trastuzumab-resistant cells could induce G1/G0 phase blockade. The above results imply that decreased expression of cyclin D3 may be responsible for trastuzumab-induced G1/G0 phase arrest of the cell cycle, but when the process is abnormal it can lead to trastuzumab-resistant. Further studies revealed that cyclin D3 mRNA level was not be affected by trastuzumab. Epigenetics are also involved in cyclin D3 expression and degradation, Cyclin D3 protein synthesis in lymphocytes is in part regulated by the RNA binding protein HuR [41]. H3K27me3 affects cyclin D3 expression in non-muscle genes through epigenetics [42]. It was reported that cAMP can induce the degradation of Cyclin D3 through glycogen synthase kinase-3beta (GSK-3beta) [43]. The most researched pathway of cyclin D3 degradation is still ubiquitination. Through the screening of protein degradation inhibitors, our study demonstrated that ubiquitin–proteasome pathway was involved in the degradation of cyclin D3 protein. These results indicated that trastuzumab may induce cyclin D3 protein degradation by ubiquitylation, thus inducing cell cycle arrest. Additional studies illustrated that trastuzumab could enhance the binding of UB protein to cyclin D3 protein, which subsequently promoted the polyubiquitination of cyclin D3 protein. CRL (Culling-Ring Ligase) complex is a crucial family in the ubiquitin–proteasome system. It has been confirmed in our experiments that cyclin D3 protein can bind to CUL1 protein, consistent with previous reports [28–30]. Further research revealed that SCF complex inhibitor MLN-4924 was identified to antagonize trastuzumab-induced downregulation of cyclin D3 protein, which established the correlation between cyclin D3 protein degradation and ubiquitylation by trastuzumab. Indeed, damage in ubiquitin proteasome system of cyclin D3 can cause trastuzumab resistance. In addition, MG132 and MLN-4924 were able to inhibit the degradation of exogenous cyclin D3 protein. Knockdown of CUL1 expression with siRNA also antagonized trastuzumab-induced downregulation of cyclin D3 protein expression. However, knockdown of FBXL2 with siRNA, an well-defined E3 ligase of cyclin D3 protein, didn’t affect cyclin D3 protein degradation [44]. Unfortunately, we are not currently able to find a definite E3 ligase associated with cyclin D3 protein. Therefore, future research will be urgently needed to investigate new E3 ligases for cyclin D3 protein which involved in trastuzumab treatment. Since we have already established the interaction of cyclin D3 and CUL1, we can narrow the scope of our search for F-boxes that can bind to CUL1, from which we can further screen for E3 ligases that can bind to cyclin D3 protein and affect the regulation of cyclin D3 protein by trastuzumab. CDK4/6 inhibitors have become the new standard of first-line treatment for hormone-receptor positive advanced breast cancer [45]. However, by now there are few studies to explore the role of CDK4/6 inhibitors in HER2-overexpressed patients except one paper reported that CDK4/6 inhibitor abemaciclib in combination with lapatinib can overcome trastuzumab resistance [46]. Recently, monarcHer clinical study showed that the median progression-free survival in patients treated with trastuzumab in combination with CDK4/6 inhibitor was not different from the trastuzumab in combination with chemotherapy in patients treated with multilinear anti-Her2 therapy, both were nearly 5.7 months [47]. However, the most remarkable clinical benefit in PFS was seen with three-drug combination regimen of trastuzumab, CDK4/6 inhibitor and fulvestrant. These data suggest that CDK4/6 inhibitor combined with trastuzumab may be an alternative option for patients treated with other anti-HER2 targeted therapy. In our experimental research, it was illustrated that the trastuzumab-resistant cells still showed high sensitivity to CDK4/6 inhibitor PD0332991, whereas the combination of trastuzumab did not further increase their sensitivity. These results indicate that CDK4/6 inhibitor is a potential therapeutic strategy after trastuzumab resistance. Due to the poor neoplasia of BT474 cells in our experiments, we did not successfully construct an in vivo animal xenograft tumor model. Although cyclin D3 protein and mRNA level has been reported to correlate with the prognosis of breast cancer patients [15], we found that the mRNA expression levels of cyclin D3 were not significantly related to the prognosis of HER2-overexpressed breast cancer patients by data analysis of TCGA public database. Based on the variation of cyclin D3 protein levels, further studies should be directed to dissect cyclin D3 protein expression in more breast cancer tissue samples before and after trastuzumab treatment. According to our current results, it is hypothesized that some patients can be considered as sensitive to trastuzumab if their cyclin D3 protein expressions in tissues are downregulated after trastuzumab treatment, indicating that cyclin D3 can functions as a biomarker to predict patients’ sensitivity to trastuzumab. In addition, search for definite E3 ligase of cyclin D3 protein, which is associated with trastuzumab sensitivity, identification of the effect of CDK4/6 inhibitors on cyclin D3 degradation in trastuzumab-resistant cells still need to be further investigated. In summary, our study firstly reveals the molecular regulatory mechanism of cyclin D3 expression by trastuzumab and the role of cyclin D3 stabilization in trastuzumab-resistance breast cancer cells, thus paving the way for further investigating drug resistance mechanisms and reversal treatment strategies in relapsed/metastatic breast cancer patients after trastuzumab treatment.