TRIB3 inhibition by palbociclib sensitizes prostate cancer to ferroptosis via SOX2-mediated SLC7A11 downregulation

Palbociclib is a CDK4/6 inhibitor approved for the treatment of breast cancer by suppressing cell proliferation. However, monotherapy with palbociclib was discouraging in prostate cancer, calling for a mechanism-based effective therapy. In this study, we reported in prostate cancer that palbociclib is a potent sensitizer of ferroptosis, which is worked out by downregulating the expression of TRIB3, a gene highly expressed in prostate cancer. Specifically, TRIB3 knockdown augmented the response of prostate cancer cells to ferroptosis inducers, whereas, TRIB3 overexpression rescued prostate cancer cells from palbociclib-induced ferroptosis. Mechanistically, TRIB3 inhibition by palbociclib resulted in downregulation of SOX2, which subsequently led to compromised expression of SLC7A11, a cystine/glutamate antiporter that counteracts ferroptosis. Functionally, a combined treatment of palbociclib with ferroptosis inducer significantly suppressed prostate cancer growth in a xenograft tumor model. Together, these results uncover an essential role of TRIB3/SOX2/SLC7A11 axis in palbociclib-induced ferroptosis, suggesting palbociclib a promising targeted therapy in combine with ferroptosis induction for the treatment of prostate cancer. Palbociclib were purchased from Selleck (S1116) and dissolved in sterile deionized water. LNCaP cells were treated with 0.5μM palbociclib for one week. Cell cycle assays were choose to identifity the successment of cell models. Total RNA was isolated from LNCaP control (Crtl) and LNCaP palbociclib-treated cells (pal), The RNA sequencing was performed with 27 Illumina HiSeq 6000 as instructed by the manufacturer (lc-bio, Hangzhou). Fastp 28 software (https://github.com/OpenGene/fastp) were used to remove the reads that 6 1 contained adaptor contamination, low quality bases and undetermined bases with 2 default parameter. Then sequence quality was also verified using fastp. We used 3 HISAT2 (https://ccb.jhu.edu/software/hisat2) to map reads to the reference genome of 4 Homo sapiens GRCh38). The mapped reads of each sample were assembled using 5 StringTie (https://ccb.jhu.edu/software/stringtie) with default parameters. Then, all 6 transcriptomes from all samples were merged to reconstruct a comprehensive 7 transcriptome using gffcompare (https://github.com/gpertea/gffcompare/). After the 8 final transcriptome was generated, StringTie and was used to estimate the expression 9 levels of all transcripts. StringTie was used to perform expression level for mRNAs by 10 calculating FPKM (FPKM = [total_exon_fragments / mapped_reads(millions) × 11 exon_length(kB)]). The differentially expressed mRNAs were selected with fold 12 change > 2 or fold change < 0.5 and with parametric F-test comparing nested linear 13 models (p value < 0.05) by R package edgeR 14 (https://bioconductor.org/packages/release/bioc/html/edgeR.html).