Exosomal Linc00969 induces trastuzumab resistance in breast cancer by increasing HER-2 protein expression and mRNA stability by binding to HUR

Centrifuge the supernatant at 300×g 4 °C for 10 min, then harvest the supernatant and centrifuge it at 2000 ×g for 10 min, suck the supernatant and centrifuge at 10,000 ×g for 30 min, harvest the supernatant and 140,000×g overspeed centrifugation for 90 min, remove the supernatant, the sediment is exosomes. Wash the sediment with PBS buffer and centrifuge at 140,000 ×g for 90 min, then resuspend the precipitate with 100 μl PBS buffer and stored the samples at − 80 °C until use.

108 serum samples in total from HER-2+ breast cancer patients who received trastuzumab treatment were collected at Union Hospital, Tongji Medical College, Huazhong University of Science and Technology between June 2015 and June 2018. Samples of 5 ml venous blood from each participant were collected by venipuncture prior to starting trastuzumab treatment. Centrifuge the blood at 1600 × g for 10 min at room temperature within 2 h after collection, then second centrifuge the blood at 12,000 × g for 10 min at 4 °C to remove the residual cells debris. The serum supernatant was transferred into RNase free tubes and stored at -80 °C. All patients were pathologically confirmed, patients with breast benign disease, autoimmune diseases or other types of cancer were excluded.

We evaluated the efficacy of trastuzumab after 2 cycles of treatment. Tumor response was confirmed through computed tomography and evaluated according to the Response Evaluation Criteria In Solid Tumors (RECIST; version 1.1), complete response (CR), partial response (PR), stable disease (SD) and progressive disease (PD). We defined the patients who evaluated with PD were trastuzumab resistant patients, while patients with PR or CR were trastuzumab sensitive patients.

The quality control for each sample sequence was carried out by FastQC (http://​www.​bioinformatics.​babraham.​ac.​uk/​projects/​fastqc/​), the RNA-seq data were compared by using HISAT2 software, and the expression values from experimental group and control group were statistically calculated by DESeq2.0 algorithm. The calculation parameters mainly included: log2FC value, FDR value, P value. The screening criteria for significant difference factors were log2FC > 1 or < − 1, and FDR < 0.05. According to the results of significant difference genes, the cluster diagram was drawn.

The human breast cancer cell lines, BT474 and SKBR-3 were acquired from American Type Culture Collection (ATCC) and maintained in McCoy’s 5A with 10% fetal bovine serum and 1% penicillin/streptomycin. The trastuzumab resistant cell lines (BT474-TR and SKBR-3-TR) were established and also maintained in McCoy’s 5A with 10% fetal bovine serum and 1% penicillin/streptomycin. All above cells were cultured at 37 °C with 5% CO₂ condition.

Human breast cancer cell lines, SKBR-3 and BT474 cells were treated with trastuzumab (Roche) when the cells grew to 85 ~ 95% density. The initial concentration of trastuzumab is 10 μg/ml, after 24 h induction culture, the cell culture medium was changed to the regular medium without trastuzumab. When the cells grew to 85 ~ 95% density, the cells were stably subcultured for three times at this concentration. Then the breast cancer cells were treated with 20 μg/ml trastuzumab to conduct induction culture for 24 h similar with above steps, untill the cells could stably grow and pass on in the medium at this concentration. Further, the concentration of trastuzumab for induction culture was increased to 40 μg/ml, 60 μg/ml, 80 μg/ml and 100 μg/ml. In this way, we finally obtained trastuzumab resistant breast cancer cells (SKBR-3-TR and BT474-TR) that could stably grow, pass on, cryopreserved and recovered in the culture medium with an effective trastuzumab concentration at 100 μg/ml.

Cell viability was analysed by Cell Counting Kit-8 (CCK8, Beyotime, Shanghai, China) following to the manufacturer's protocols. The human breast cancer cells were seeded and cultured into 96-wells plates. Then, the cells were treated with trastuzumab. Add 10 μL of CCK-8 reagent to each well and culture the samples for 2 h. At last the absorbance was analysed at 450 nm by microplate reader. The wells without cells were treated as blanks.

The breast cancer cells with log phase growth were plated in 6-well plates. The cells were incubated at 37 °C overnight and treated with trastuzumab. Then the cells were fixed with a mixture of methanol and acetic acid (10:1 v/v) and stained with 1% crystal violet in methanol after 10–14 days of incubation in 6-well plates. At last the numbers of colonies with > 50 cells were counted and the surviving fractions were calculated.

The breast cancer cells were seeded into 96-well plates. The cells were incubated at 37 °C overnight and treated with trastuzumab. Then the cells were incubated with EdU solution for 2 h (1/1000, RiboBio, China). Remove EdU solution and fix the cells with 4% paraformaldehyde for 30 min, permeabilize the cells by 0.5% Triton X-100 for 10 min and stain the cells by Hoechst. At last the EdU positive cells were detected by fluorescent microscope and counted by ImageJ Software.

Centrifuge the breast cancer cells at 1000 rpm, 4 °C for 15 min and collect the cancer cells. Incubate the cells with 2.5% glutaraldehyde solution at 4 °C overnight. Then the cells undergo dehydrating, embedding, solidifying, ultrathin slicing, and staining. At last cell samples were observed and imaged by a transmission electron microscope.

The RNA extraction was harvested by using TRIzol reagent (Invitrogen). Then the RNA extraction was undergoing reverse transcription by using Prime RT reagent kit (Vazyme). PCR primer sequences (5'to3') are recorded as follows: human Gapdh-F primer sequence CCACATCGCTCAGACACCAT; human Gapdh-R primer sequence TGACAAGCTTCCCGTTCTCA; human Linc00969-F primer sequence ACGGATCACCACTGCAAGAG; human Linc00969-R primer sequence TAGGTGGAATCGGGCCTGTA; human HUR-F primer sequence GAAGACCACATGGCCGAAGA; human HUR-R primer sequence TGGTCACAAAGCCAAACCCT. Quantitative PCR was performed by using SYBR Green real-time PCR kit (Vazyme).

The whole cell lysates were harvested via cell lysis buffer and the protein concentration was detected with BCA Protein Assay Kit (Thermo). Then the proteins were undergoing separating by 8–12% gradient gels and transferred to PVDF (Polyvinylidene Fluoride) membranes. Membranes were blocked by blocking buffer and incubated with primary antibodies at 4 °C overnight. At last the membranes were incubated with secondary antibodies at room temperature for 1 h and scanned by infrared imaging system. The following primary antibodies were used: TSG101 (1:1000, ab133586, Abcam), CD81 (1:1000, ab109210, Abcam), HUR (1:1000, ab200342, Abcam), HER-2 (1:1000, ab134182, Abcam), GAPDH (1:1000, 60,004–1-Ig, Proteintech), p62 (1:1000, cat. no. 18420–1-AP), LC3 (1:1000, cat. no. 14600–1-AP), CD63 (1:1000, BD Bioscience, clone H5C6), CD9 (1:1000, Millipore, clone MM2/57), Fibronection (1:1200, ab285285, Abcam).

The breast cancer cells were fixed by 4% formaldehyde and underwent permeabilizing by PBS with 0.2% Triton X-100. Then the cells were blocked by blocking buffer and incubated with primary antibody at 4 °C overnight. Finally the cell samples were incubated with secondary antibody for 1 h, washed by PBS, mounted in DAPI (4',6-diamidino-2-phenylindole), and observed under confocal laser scanning fluorescence microscopy.

BT474 and SKBR-3 cells were transfected with Linc00969 overexpression plasmids. BT474-TR and SKBR-3-TR cells were transfected with siRNA-Linc00969. The transfection kit used was riboFECT™ CP kit as directed by manufacturer’s protocols and the breast cancer cells were used in following experiments after 24 h’ transfection. Overexpression or knockdown cells were confirmed by RT-PCR. (siRNA-Linc00969-1: CGAUUCCACCUACAGCAAAGC; siRNA-Linc00969-2: GGACGGAUCACCACUGCAAGA; siRNA-HUR: TCCAGATTTTTGAAAAATACAAT).

Breast cancer cells were fixed by 4% formaldehyde for 20 min, and washed by PBS on a shaker for 5 min × 3 times. Then cells were added protease K (20 μg/ml) to digest for 3 min, and washed by PBS for 5 min × 3 times. Droped the pre-hybridization solution and incubate for 1 h at 37 °C, removed pre-hybridization solution and incubated with probe hybridization solution at 37 °C overnight. Washed the cells by 2 × SSC for 10 min, 1 × SSC for 5 min twice and 0.5 × SSC at 37 °C for 10 min. Finally, cells were mounted in DAPI and observed under fluorescence microscopy.

Animal experiments were authorized by Medical Ethics Committee of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, under national standard guidelines for animal welfare. Nude mice (4 weeks old, 16-18 g, BALB/c Nude) were randomly grouped and five nude mice each group. A suspension of 1–5 × 10⁷ human breast cancer cells in 1 ml PBS was perpared, then the medium was mixed with matrigel at the ratio 1:1 for injection. 1–5 × 10⁶ (100 μl) human breast cancer cells were injected subcutaneously into middle posterior part of axilla of each nude mouse and the mice were treated with trastuzumab. Tumor volume was monitered twice a week and calculated by the formula: V = 1/2 × a × b², where a = length (mm), and b = width (mm).

The subcutaneous trasnplanted tumores were submitted in cassettes for paraffin embedding and sectioning. The tumor Sects. (4 µm) were incubated with primary antibodies at 4 °C overnight, then underwent incubating with secondary antibody and Streptavidin-Avidin–Biotin. Finally, peroxidase reaction was performed by diaminobenzidine tetrahydrochloride and the sections were counterstained by haematoxylin. The sections were visualized under microscope in five independent high magnification fields.

The nuclear and cytosolic samples of breast cancer cells were separated by utilizing PARIS kit (Am1921, Thermo Fisher Scientific, USA) according to the manufacturer’s protocols. The U1, GAPDH and Linc00969 expression levels in nuclear and cytoplasm of breast cancer cells were detected by qRT-PCR.

Firstly, the control RNA, target RNA and the probe labeling reaction system were prepared by using Pierce™ RNA 3' End Desthiobiotinylation Kit according to the instructions. The probe labeling reaction system was added into PCR instrument at 16 °C for more than 4 h or overnight, 400 μl nuclease free-water was added into each sample afer reaction. Then 300 μl phenol chloroform was added into samples to extract successfully labeled RNA, centrifuge the samples at fastest speed for 15 min after vibrate, transfer the supernatant to a new EP tube. Secondly, 10 μl 5 M NaCl, 2 μl glycogen and 600 μl pre-colded 100% ethanol were added into the supernatant, then the samples were deposited overnight at -20 °C or -80 °C and centrifuged at fastest speed for 30 min at 4 °C, the precipitate was the RNA sample. Remove the supernatant, wash the RNA by 70% ethanol, centrifuge the samples for 10 min, remove the supernatant again and dry the RNA in air. Finally, 20 μl nuclease free water was added into each sample to dissolve the RNA, then the RNA was added into RNA instrument and denatured for 5 min at 95 °C for following experiments.

The magnetic beads were also need to prepared. Put 400 μl magnetic beads (400 μl for control RNA, 400 μl for target RNA) on the magnetic frame, remove the supernatant, wash the beads by 800 μl 1 × binding & washing buffer 3 times. Then, 400 μl 2 × binding & washing buffer, 20 μl RNA and 380 μl DEPC water were added into the magnetic beads, rotate the samples slowly at room temperature for 20 min, so that the beads could fully bind with RNA. Transfer the samples on the magnetic frame and remove the supernatant, wash the samples by 800 μl 1 × binding & washing buffer 3 times. Finally, the RNA binded beads were washed by cell lysis buffer A for following experiments.

The protein extraction samples were harvested by cell lysis buffer and mixed with the prepared beads (1 U/μl RNase inhibitor was also added), rotate the samples slowly at 4 °C for 2 h, so that the beads could fully combine with protein. Transfer the samples on the magnetic frame, remove the supernatant and wash the samples by 400 μl cell lysis buffer A 5 times. Then, the beads were suspended by 25 μl pre-colded 0.1% SDS solution, added 6.25 μl 5 × protein loading buffer, boiled at 100 °C for 10 min and placed on ice immediately for 5 min. At last the beads were placed on magnetic frame, the supernatant was transfered into a new EP tube for western blotting detection.

Firstly, we should prepare the magnetic beads and antibody. The magnetic beads coated with protein-A/G were fully suspended and washed by NT-2 buffer twice, then the magnetic beads were suspended by 100 μl NT-2 buffer and mixed with 5 μg target antibody in room temperature for 1 h. Centrifuge the magnetic beads at 5000 × g for 15 s, add magnetic base to absorb magnetic beads and remove the supernatant, then use 1 ml NT-2 buffer to wash the magnetic beads 5 times. At last the magnetic beads were suspended by 900 μl NET-2 buffer for following experiments.

The cell lysates were harvested via cell lysis buffer and centrifuged at 4 °C 20000 × g for 10 min. Mix 100 μl cell lysate supernatant with 900 μl NET-2 buffer suspended magnetic beads to carry out antibody incubation. Reserve 10 μl sample for “Input” copy and store it at -80 °C. Mix the other sample by vertical mixer at 4 °C for more than 3 h or overnight. Centrifuge the sample for a short time, and put the sample on the magnetic base upon the ice, remove the supernatant after 1 min at 4 °C, then use 1 ml NT-2 buffer to wash the sample 5 times, the precipitate was the final sample got from RIP assay. The RNA sample could be further extracted after digestion by proteinase K for subsequent analysis.

Exosomes were isolated from the plasma of trastuzumab-resistant BC patients (R-exo) and trastuzumab-sensitive BC patients (S-exo) and observed by electron microscopy, and the morphology and size conformed to the characteristics of exosomes (Additional file 1: Fig. S1A). Then, we used western blotting to measure the expression of biomarkers of exosomes. We found that R-exo and S-exo both had high levels of proteins (CD63, CD81 and CD9) (Additional file 1: Fig. S1B). The diameter of the exosomes was also determined by particle size identification, and the results confirmed that the diameter of exosomes we derived from patients conformed to the exosome parameters (Additional file 1: Fig. S1C). The exosomes derived from BC cells were also observed by electron microscopy (Additional file 1: Fig. S1D), and the expression of exosome biomarkers (TSG-101, CD81, CD9, CD63 and fibronectin) was extremely high in exosomes derived from BC cells (Additional file 1: Fig. S1E). The diameter and number were also examined by particle size identification, confirming again that we had isolated exosomes (Additional file 1: Fig. S1F). Therefore, we successfully isolated and identified exosomes derived from BC patients and cells.

First, we established trastuzumab-resistant BC cell lines: BT474-TR and SKBR-3-TR (Additional file 1: Fig. S1G). The results showed that exosomes from BT474-TR and SKBR-3-TR cells can be taken up by recipient BC cells within 24 h (Fig. 1A). We further examined whether trastuzumab-resistant cell-derived exosomes could confer trastuzumab resistance in recipient BC cells. Figure 1B shows that parental cancer cells incubated with exosomes derived from BT474-TR or SKBR-3-TR cells exhibited increased viability after trastuzumab treatment. Similar results were also obtained via colony formation assay and EdU assay (Fig. 1C, D). To determine whether exosomes from BC cells were involved in this effect, we reduced exosome production with GW4869 (Fig. 1E). GW4869 is a specific noncompetitive neutral sphingomyelinase (N-SMase) inhibitor with cell permeability. It can block the sprouting of multivesicular bodies mediated by ceramide, thereby inhibiting the biogenesis or release of exosomes. GW4869 is commonly used to inhibit the generation of exosomes [28]. Multiple assays, including CCK8, colony formation and EdU assays, revealed that incubation with culture medium from BT474-TR or SKBR-3-TR cells failed to confer trastuzumab resistance in recipient BC cells after treatment with GW4869 (Fig. 1F–H). The above results proved that trastuzumab-resistant BC cell-derived exosomes can enhance trastuzumab resistance in recipient BC cells.

To identify the underlying mechanism by which exosomes mediate trastuzumab resistance, deep RNA sequencing of exosomal lncRNA and cirRNA were performed, which included samples from four patients sensitive to trastuzumab treatment and samples from four patients resistant to trastuzumab treatment. The sequencing analysis revealed 4785 lncRNAs in total, including 306 upregulated lncRNAs and 4479 downregulated lncRNAs, that were dysregulated in exosomes from trastuzumab-sensitive BC tissues compared with trastuzumab-resistant BC tissues (Additional file 5: Table S1, Fig. 2A). However, we did not find any differences in the expressions of cirRNA (Additional file 6: Table S2). The ceRNA network map of differentially expressed RNAs (DERs) in these transcriptome data was enriched in the PI3K-Akt pathway. Because the PI3K-Akt signaling pathway plays a critical role in the occurrence and development of BC, we screened lncRNA candidates that participate in the PI3K-Akt pathway and were the most upregulated (Fig. 2B). The results showed that the expression trends of MROH4P, REXO1L1P, LOC346296, and LINC00969 were consistent with the sequencing results. LINC00969 was the most significantly different lncRNA found in sequencing test.

Furthermore, we measured the expression levels of the Linc00969 in 108 patients blood samples and 47 tissue samples. We found that linc00969 was significantly increased in trastuzumab-resistant patients blood samples, both in 44 early-stage HER-2+ BC patients and 64 metastatic HER-2+ BC patients (Fig. 2C–E). In total 108 patients blood samples, the patients were divided into high expression group and low expression group based on the median value of Linc00969 expression (0.476) in exosomes. In the trastuzumab sensitive patients (n = 62), the median level of Linc00969 was 0.1267 (SD = 0.8632) and 1.5802 (SD = 0.355) in the trastuzumab resistant patients (n = 36). The Linc00969 expression level in exosome was signifcantly upregulated in the trastuzumab resistant compared with the sensitive patients (p < 0.001). Besides, In tissue samples, the median level of Linc00969 was 0.2693 (SD = 0.3102) in the trastuzumab sensitive patients (n = 28) and 1.1430 (SD = 0.9723) in the trastuzumab resistant patients (n = 19). The Linc00969 expression level in tissue was also signifcantly upregulated in the trastuzumab resistant compared with the sensitive patients (p < 0.001) (Additional file 1: Fig. S1H).

Additionally, we evaluated the association between exosomal Linc00969 and clinical characteristics in BC. As shown in Additional file 7: Table S3, elevated expression levels of exosomal Linc00969 were significantly associated with positive nodal status, higher histological grade (2/3), higher Ki67 score (≥ 40%) and distant metastasis (p < 0.05). The above results showed that high expression level of exosomal Linc00969 in BC patient serum have close correlation with trastuzumab resistance and poor clinical characteristics in BC. Therefore, we focused on Linc00969 in subsequent experiments.

To further investigate whether exosomal Linc00969 was involved in HER-2+ BC trastuzumab resistance, we designed and performed relevant experiments in vitro. First, we stably knocked down Linc00969 expression in trastuzumab-resistant BC cells with siRNA and overexpressed Linc00969 in parental trastuzumab-sensitive cells with plasmids (Additional file 2: Fig. S2A-B). As shown in Fig. 3A-C, CCK8, colony formation and EdU assays indicated that Linc00969 knockdown notably inhibited cell proliferation after trastuzumab treatment, while Linc00969 overexpression obviously increased viability (Additional file 2: Fig. S2C-E). Furthermore, as shown in Fig. 3D, exosomal Linc00969 was markedly upregulated in the trastuzumab-resistant subgroups compared to the parental cell lines. FISH assays with the Linc00969 probe showed that Linc00969 was distributed mainly in the cytoplasm (Fig. 3E). Moreover, the Linc00969 level in the culture medium was nearly unchanged after treatment with RNase alone but sharply decreased after treatment with both RNase and Triton X100. This result indicated that Linc00969 was encapsulated by a membrane instead of being directly released (Fig. 3F). Moreover, the Linc00969 expression levels in exosomes of trastuzumab-resistant cells were almost equivalent to those in culture medium. However, extracellular Linc00969 levels were sharply reduced after removing the exosomes in culture medium (Fig. 3G), which indicated that the exosomes were the main carrier of extracellular Linc00969. More importantly, silencing Linc00969 sharply inhibited the ability of cocultured BC cells to acquire resistance to trastuzumab (Fig. 3H–J). These in vitro experiments suggest that Linc00969 mediates trastuzumab resistance in HER-2+ BC cells. Additionally, Linc00969 can be transported to recipient cells through exosome encapsulation and cause recipient cells to acquire trastuzumab resistance.

Following our in vitro experiments, we also wanted to validate our observations in vivo by establishing a xenograft BALB/c nude mouse model. We found that the volumes of subcutaneously transplanted tumors were significantly higher in the Linc00969 overexpression groups than in the regular groups after trastuzumab treatment (Fig. 4A), while the si-Linc00969 groups were more sensitive to trastuzumab than the control groups (Fig. 4B).

Then, we performed IHC staining on a mouse model to evaluate the expression levels of Ki67 and PCNA, which represented the proliferation capacity of the tumors. The expression levels of Ki67 and PCNA were markedly decreased by silencing Linc00969 (Fig. 4D) and significantly increased by Linc00969 overexpression (Fig. 4C). All of the above results proved that Linc00969 contributed to trastuzumab resistance and that silencing Linc00969 reversed trastuzumab resistance in HER-2+ BC in vivo.

Next, we explored how Linc00969 induced trastuzumab resistance. Since Linc00969 was overexpressed in HER-2+ BC cells, we first investigated the relationship between Linc00969 and HER-2 expression. Based on the results of immunofluorescence staining, we found that the HER-2 protein was mainly located in the nucleus of BC cells and upregulated in the resistant subgroups compared to the parental BC cell lines (Fig. 5A–C). Then, we found that HER-2 expression was sharply decreased by silencing Linc00969 (Fig. 5D), while overexpressing Linc00969 upregulated HER-2 protein levels (Fig. 5F). However, the expression of ERBB2 (ERBB2 indicated the coding RNA of HER-2 protein) measured by qRT‒PCR did not change when Linc00969 was silenced or overexpressed (Fig. 5E, G). Together, these results suggested that Linc00969 regulates trastuzumab resistance in HER-2+ BC cells through HER-2 expression only at the protein level but not at the transcription level.

However, the regulatory mechanism of Linc00969 in the upregulation of HER-2 expression is still unknown and needs further study. The subcellular location of Linc00969 was identified by nucleus-cytoplasm fraction qPCR, and the results demonstrated that Linc00969 was mainly located in the cytoplasm (Fig. 6A). According to the evaluation of the online minimum free energy (MFE) (http://​rna.​tbi.​univie.​ac.​at/​), we predicted that the Linc00969 transcript at the 1058–1305 nt locus formed a stem‒loop structure (Fig. 6B), which is the key structure related to the targeted RNA-binding proteins. It has been reported that the upregulation of HER-2 expression may occur due to HUR, the RNA-binding protein of Linc00969, which increases HER-2 mRNA stability and expression in hepatocellular carcinoma [29]. Moreover, prediction by POSTAR3 showed that the chr3:195,689,368–195,689,390 nt region of Linc00969 contained a binding motif for HUR (Fig. 6C), which indicated that Linc00969 could be a scaffold that mediates the interaction of HER-2 and HUR. Consistently, immunofluorescence assays showed that Linc00969 and HUR were colocalized mostly in the cytoplasm (Additional file 3: Fig. S3C). These results suggest that Linc00969 may interact with HUR proteins. Furthermore, we designed a Linc00969 probe, performed an RNA pull-down assay, and found that the HUR protein was enriched by Linc00969 (Fig. 6D). RIP assays were also performed to prove the direct interaction between Linc00969 and HUR (Fig. 6E). However, HUR was not affected by Linc00969 knockdown at the transcriptional level (Additional file 3: Fig. S3D). Silencing HUR with si-HUR (Additional file 3: Fig. S3A, B) significantly abrogated the Linc00969-induced increase in HER-2 protein (Fig. 6F). Moreover, overexpression of Linc00969 increased the stability of HER2 mRNA; however, this effect was significantly reversed in HUR-knockdown cells (Fig. 6G), indicating that HUR was essential for Linc00969-induced HER-2 mRNA stability. CCK8, colony formation and EdU assays showed that the trastuzumab resistance effect after overexpression of Linc00969 could be partially restored by blocking HUR in both SKBR-3 and BT474 cells (Fig. 6H–J). The above results indicated that Linc00969 promoted trastuzumab resistance in HER-2+ BC cells by maintaining HER-2 mRNA stability by binding to HUR.

Autophagy might induce drug resistance by decreasing the cytotoxicity of drugs to cancer cells. To investigate whether exosomal Linc00969 regulates the resistance of trastuzumab via autophagy, we analyzed the level of autophagy in BC cells. We found that the trastuzumab-resistant BC cells showed markedly higher levels of autophagy than the parental cells, as evidenced by increased formation of autophagosomes (Fig. 7A), increased autophagic flux (Fig. 7B), increased LC3-II (Light Chain-3) expression (the biomarker of autophagy) and decreased expression of p62 protein (biomarker of autophagy suppression) (Fig. 7I), confirming the close relationship between trastuzumab resistance and autophagy in BC cells. Then, we explored whether Linc00969 influences the expression level of autophagy in BC cells. In contrast to the control group, silencing Linc00969 reduced the number of autophagosomes (Fig. 7C), decreased the autophagic flux (Fig. 7D), decreased LC3-II expression and elevated p62 expression (Fig. 7J) in the presence of trastuzumab treatment. More importantly, parental BC cells cultured with exosomes derived from trastuzumab-resistant cells exhibited increased autophagy activity (Fig. 7E, F, K). However, silencing Linc00969 sharply inhibited the ability of cocultured BC cells to increase autophagy activity (Fig. 7G, H, L). Furthermore, we found that inhibition of autophagy with HCQ can reverse linc00969 trastuzumab resistance (Additional file 4: Fig. S4). In summary, we confirmed that exosomal Linc00969 could transmit trastuzumab resistance by regulating autophagy levels.