A novel ARMC5-containing ubiquitin ligase controls the degradation of HSPA1A and its client protein estrogen receptor alpha

A novel ARMC5-containing ubiquitin ligase controls the degradation of HSPA1A and its client protein estrogen receptor alpha | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article A novel ARMC5-containing ubiquitin ligase controls the degradation of HSPA1A and its client protein estrogen receptor alpha Jiangping Wu, Xiao He, Yuhan Ying, Junzheng Peng, Hongyu Luo This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8810312/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract ARMC5 is the substrate recognition subunit of a recently discovered multiple-subunit ubiquitin ligase (ARMC5-E3), which is composed of ARMC5, cullin3, and RBX1. ARMC5-E3 is the major E3 for RNA polymerase II. ARMC5 interacts with multiple heat shock proteins, but the biological significance of these interactions is not known. We discovered that HSPA1A and its client protein, estrogen receptor alpha (ERα), were novel substrates of ARMC5-E3. ARMC5 deletion led to the accumulation of these substrates, while ARMC5 overexpression reduced ERα protein levels. In the presence of estrogen, HSPA1A was no longer bound to ERα, and ARMC5-E3 no longer regulated ERα degradation. Based on transcriptome analysis, ARMC5 overexpression in the presence of estrogen significantly altered the expression of 247 genes in breast cancer MCF7 cells. While estrogen stimulation increased ERE-binding by ERα in MCF7 cells, ARMC5 overexpression reduced it, supporting the notion that ARMC5-E3 degrades ERα and decreases its availability to interact with EREs. In ER-positive human breast cancer patients, increased ARMC5 gene copy number was correlated to longer relapse-free survival time, probably in part due to ARMC5-E3’s effect on reducing HSPA1A-associated ERα levels. Therefore, ARMC5-E3 is a specific E3 that controls the degradation of HSPA1A and ERa. ARMC5 gene copy number variation is associated with tumor-free survival time of breast cancer patients. Hence, ARMC5 is a newly found regulator of estrogen signaling and function, and can serve as a prognostic parameter for breast cancer. Biological sciences/Cancer/Breast cancer Biological sciences/Cell biology/Proteolysis/Ubiquitylation Health sciences/Biomarkers/Prognostic markers Biological sciences/Cell biology/Mechanisms of disease ARMC5 ubiquitin ligase HSPA1A estrogen receptor α ARMC5 gene copy number variation breast cancer tumor-free survival Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Estrogen, a steroid hormone, plays a vital role in many, if not all, of our biological systems and has an outsized impact on many processes, such as reproduction, development, cardiovascular health, bone metabolism, immunity, brain function, and oncogenesis, particularly breast cancer oncogenesis 1 – 3 . Classical estrogen receptors (ER) are intracellular proteins with two subtypes: ERα and ERβ. They form homodimers or heterodimers 4 – 6 . The dominant form of dimers depends on tissue types. In breast cancer cells, the ERα homodimer is the major type 3 , 7 , 8 . In the absence of estrogen, ER dimers are mostly located in the cytosol and bound to heat shock proteins (HSPs) 9 – 11 . When ER dimers are bound to the ligand estrogen, HSPs will chaperone them into the nuclei. Although HSP90 is the major chaperone for ER, HSPA1A also participates in this function 9 – 11 . In the nuclei, ER dissociates from HSPs to bind estrogen-responsive elements (ERE) of various genes, functioning as transcription factors to regulate gene expression in an estrogen-dependent way. After ERs complete their function as transcription factors, they are recycled and reused, but eventually, they are degraded through the proteasome pathway in both the nuclei and cytosol 12 , 13 . E6AP has been identified as an ERα-specific ubiquitin ligase (E3) for ER degradation via the proteasome 14 . E6AP directly binds to ER, and its overexpression in cells leads to a decreased level of ER. A Hsc70 (heat shock cognate 70)-interacting protein, CHIP (carboxyl terminus of HSC70-interacting protein), has been documented as another E3 to ubiquitinate ERα, leading to its degradation 15 . In addition, a Cullin3-containing E3 called SPOP also acts on ERα 16 . HSPs are a family of proteins comprising small HSPs, HSP40, HSP60, HSP70, HSP90, and large HSPs. They are found in various cellular compartments and play vital physiological roles in the folding, stability, translocation, and degradation of their client proteins. During cellular stress, such as heat, oxidation, and inflammation, HSPs are upregulated to protect cells from adverse environments 17 , 18 . HSP70 is a well-studied family of HSP. It has 13 isoforms 19 , 20 , which are of similar size but are coded by different genes. HSP70-1, a.k.a. HSPA1A, is the dominant isoform. It binds to nascent polypeptides, preventing them from aggregating. It also chaperones its client proteins across bilayer membranes during protein translocation. HSPA1A is implicated in the degradation of its client proteins, such as p53, by associating with a p53-specific ubiquitin ligase (E3) CHIP (C-terminus of HSP70-interacting protein), enabling the latter to ubiquitinate p53 for the proteasome-mediated degradation 21 . ER is one of the HSPA1A client proteins. Whether HSPA1A plays a role in ER degradation has not been investigated. HSPA1A is eventually degraded via the proteasome pathway after being ubiquitinated by CHIP 15 , 21 , 22 , although it can also be delivered to the lysosome for degradation 23 , 24 . Thus, it seems that HSPA1A-associated E3s can control the degradation of both HSPA1A and its client proteins. Additional HSPA1A-specific E3s might exist that, in theory, could regulate the degradation of both HSPA1A and its client proteins. During our recent study of a novel multiple-subunit E3, ARMC5-CUL3-RBX1, which uses ARMC5 as its substrate recognition subunit, we employed proteomics to identify ARMC5-interacting proteins with a view to identifying additional substrates of this E3. We identified at least 10 members of the HSP family associated with ARMC5. This raises an intriguing possibility: these HSPs, along with their client proteins, may be potential substrates of this E3. We elected to investigate HSPA1A in detail for this possibility. In this study, we validated interactions between ARMC5 and HSPA1A, and between HSPA1A and its client protein ERα, using immunoprecipitation. HSPA1A and its client protein ERα are indeed substrates of the ARMC5-containing E3, as ARMC5 gene knockout caused the accumulation of these two proteins in various tissues of mice. Conversely, ARMC5 overexpression in human breast cancer MCF7 cells reduced the protein levels of HSPA1A and ERα. Functionally, according to RNA-seq, ARMC5 overexpression altered the expression of 247 genes, some of which are known to be involved in breast cancer tumorigenesis. ARMC5 overexpression also decreased ERα binding to EREs, as determined by ChIP-seq. Consistent with these findings at the cellular and molecular levels, human genetic studies showed a significant correlation between increased ARMC5 copy numbers and reduced incidences of relapse in breast cancer patients. Thus, our study identified a novel E3 ligase for ERα and demonstrated its involvement in regulating the expression of estrogen-responsive genes in breast cancer cells. ARMC5 mutations in breast cancer patients are implicated in the prognosis of breast cancer. Materials and methods Western blotting For mouse tissues, they were homogenized in RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate) in the presence of protease inhibitors (Roche). For cultured cells, they were lysed directly in RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate) supplemented with protease inhibitors. Protein concentrations were determined by BCA assays (Thermo Fisher). Equal amounts of total protein were loaded on 10% SDS-PAGE gels and transferred onto PVDF membranes (Bio-Rad). Membranes were blocked for 1 hour in 5% non-fat milk in TBS-T (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% Tween 20) and incubated overnight at 4°C with primary antibodies (1:1000 dilution for all the Abs used in Western blotting) in TBS-T. A list of Abs used in Western blotting is provided in Supplementary Table 1 (S-Table 1). After washing in TBS-T, membranes were incubated for 1 hour at room temperature with HRP-conjugated secondary antibodies (Cell Signaling). Immunoreactive bands were detected using enhanced chemiluminescence (ECL) and visualized on X-ray film. Band intensities were quantified using ImageJ (NIH) and normalized to β-actin or α-actinin to ensure equal loading. The Wilcoxon rank test was used for comparison. MCF7 cell culture To minimize the estrogenic activity of the culture medium, we cultured MCF7 cells in phenol red-free DMEM (Gibco) supplemented with 10% charcoal-stripped fetal bovine serum and 1% penicillin-streptomycin in a humidified incubator at 37°C with 5% CO₂ for 72 hours. One hundred nM 17β-estradiol (E2; final concentration) or vehicle (ethanol) was then added to some cultures for different periods of time. Co-immunoprecipitation To detect interactions between ARMC5 and HSP70A1A, HEK293 or MCF7 cells were transfected with HA-ARMC5-expressing plasmids. After 48 hours, cells were lysed with TNE buffer (50 mM Tris-HCl at pH 7.4, 100 mM NaCl, 0.1 mM EDTA, 1% Triton X-100) containing protease inhibitors. The lysate was incubated overnight at 4°C with an anti-HA antibody (1:100 dilution), then reacted with protein G-conjugated magnetic beads for 2 hours at room temperature. The beads were washed 3 times with TNE buffer, and the proteins were eluted with a 2 × SDS-loading buffer (4% SDS, 20% glycerol, 200 mM DTT, 0.01% bromophenol blue, and 0.1 M Tris HCl, pH 6.8) at 95°C for 15 minutes. The input lysates and eluates were subjected to Western blotting analysis for the presence of HSP70A1A. To detect the interaction between HSP70A1A and ERα, we cultured MCF7 cells in medium containing 10% stripped serum, without phenol red, for 72 hours. One hundred nM E2 was added to the culture for a given period. The cells were lysed in TNE buffer with protease inhibitors. The lysates were reacted with anti-human HSPA1A Ab (Cell Signaling Technology; 1:100 dilution). The remaining steps were the same as described above. The input lysates and eluates were subjected to Western blotting analysis for the presence of ERα. Detection of protein ubiquitination To detect HSP70A1 ubiquitination, mouse tissue lysates were incubated with anti-HSP70A1A Ab (1:100) overnight at 4°C. HSP701A1 was precipitated with protein G-conjugated magnetic beads at room temperature for 2 hours. HSP701A1 ubiquitination in the eluates was detected using Western blotting RT-qPCR Total RNA from mouse tissue was isolated with TRIzol. Total RNA from cells was extracted with the RNeasy Mini Kit (Qiagen). RNA was reverse transcribed into cDNA using the iScript cDNA Synthesis Kit (Bio-Rad) at 42°C for 30 minutes, followed by 85°C for 5 minutes to inactivate the enzyme. The resulting cDNA was amplified by PCR using gene-specific primers and Taq DNA polymerase (Thermo Scientific) under the following cycling conditions: 95°C for 2 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 60 seconds. A list of primers used for RT-qPCR is presented in S-Table 2. The levels of β-actin signals were used as internal controls. The CT ratios of the genes of interest and those of β-actin were presented as gene expression levels. The Wilcoxon rank test was used for statistical comparison. RNA-sequencing (RNA-Seq) MCF7 cells were transfected with a construct expressing HA-ARMC5 or an empty vector and cultured in regular medium for 24 hours. The cells were then switched to medium containing 10% stripped serum for an additional 48 hours. At that time, 100 nM E2 (final concentration) or vehicle (ethanol) was added to the medium, and the cells were cultured for an additional 2 hours. Total RNA was extracted from the cells using the RNeasy Mini Kit (Qiagen) following the manufacturer’s protocol. The integrity and quantity of the isolated RNA were assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies) with an RNA 6000 Pico kit. Libraries were prepared using an equal amount of total RNA per sample (150 ng). First, ribosomal RNAs were depleted using a KAPA RiboErase (HMR) kit, and then libraries were prepared using a KAPA RNA Hyperprep Kit (Roche Diagnostics) with 11 cycles of final amplification. Library size distribution was assessed on a 2100 bioanalyzer (Agilent Technologies) using a High Sensitivity DNA Kit, and libraries were quantified by qPCR. Equimolar libraries were sequenced in paired-end reads (PE100) on a Novaseq system (Illumina) at 50 M reads per library. RNA-seq data were analyzed using the Galaxy server 25 ( https://usegalaxy.org ). Quality control of the raw FASTQ files was performed using FastQC, followed by trimming low-quality reads and adapter sequences with Trimmomatic. Cleaned reads were then aligned to the reference genome (GRCh38) using the STAR aligner with default parameters. Gene-level expression was quantified from aligned reads with featureCounts, and differentially expressed genes were identified using edgeR. Graphical outputs, including heatmaps and volcano plots, were generated ggplot2 in R. ERα chromatin immunoprecipitation followed by sequencing (ChIP-seq) MCF7 cells were transfected with empty vectors or HA-ARMC5-expressing constructs as described above and cultured in regular medium for 24 hours. Then, they were cultured in medium containing 10% stripped FCS for an additional 48 hours. At that time, 100 nM E2 or vehicle (ethanol) was added to the culture. The cells were harvested at 0 and 45 minutes after the E2 treatment. The HA-ARMC5 and vector-transfected MCF7 cells from three independent experiments were used for ERα ChIP-seq. Approximately 8–10 × 106 cells per sample were crosslinked with 1% formaldehyde for 10 minutes at room temperature, then quenched with 125 mM glycine for 10 minutes while rotating. The cells were pelleted, washed with cold PBS, and resuspended in cell swelling buffer (25 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 0.1% NP-40, supplemented with protease and phosphatase inhibitors) on ice for 10 minutes to release nuclei. Nuclei were pelleted, resuspended in ChIP sonication buffer (50 mM HEPES pH 7.9, 140 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% Na-deoxycholate, 0.1% SDS) supplemented with protease and phosphatase inhibitors, and sonicated using a probe-based sonicator (FB120 with a CL-18 probe; ThermoFisher) at 25% amplitude with 30-second pulses at 30-second intervals for a total of 2.5 minutes. Sonicated chromatin was stored until immunoprecipitation. Chromatin fragmentation was assessed by retrieving 5% of sonicated nuclei and treating them with 10 µg RNase A for 45 minutes at 37°C, followed by 20 µg proteinase K for 1 hour at 65°C. Purified DNA (QIAquick PCR Purification Kit) was quantified using a Nanodrop 1000 Spectrometer (ThermoFisher), and fragment length was determined by electrophoresis. For immunoprecipitation, anti-human ERα Ab (0.6 µg per sample; S-Table 1) was added to each sonicated chromatin sample, and the mixture was incubated at 4°C overnight. The mixture was then reacted with 50 µl SureBeads™ Protein G Magnetic Beads (Bio-Rad) per sample for 2 hours at 4°C. Beads were washed sequentially with ChIP sonication buffer, LiCl wash buffer (20 mM Tris pH 8.0, 1 mM EDTA, 250 mM LiCl, 0.5% NP-40, 0.5% Na-deoxycholate), and TE buffer (10 mM Tris pH 8, 0.1 mM EDTA). Chromatin was eluted with ChIP elution buffer (50 mM Tris pH 8.0, 10 mM EDTA, 1% SDS) with agitation for 15 minutes at 1100 rpm. Immunoprecipitated chromatin was de-crosslinked at 65°C overnight, treated with RNase A (20 µg per sample) at 37°C for 1 hour and proteinase K (200 µg per sample) for 2 hours at 65°C. DNA was purified with the QIAquick PCR Purification Kit (QIAGEN). Size distribution and concentration of immunoprecipitated and input samples were evaluated on a 2100 bioanalyzer (Agilent Technologies) using a High Sensitivity DNA Kit. ChIP-seq libraries were prepared using the KAPA Hyperprep Library Kit (Roche Diagnostics). Normalization of the sample quantities was performed prior to final amplification based on qPCR quantification of the ligation products. Library size distribution was assessed on a 2100 bioanalyzer (Agilent Technologies) using a High Sensitivity DNA Kit, and libraries were quantified by qPCR. Equimolar libraries were sequenced in paired-end reads (PE100) on a Novaseq system (Illumina) at 50 million reads per library. ChIP–seq data from 12 EV/HA samples were processed on a SLURM-based HPC cluster using a standardized pipeline. Raw paired-end reads were quality-checked with FastQC and trimmed using cutadapt to remove Illumina TruSeq adapters (R1: AGATCGGAAGAGCACACGTCTGAACTCCAGTCA; R2: AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT), low-quality bases, and poly-A/N tails, requiring a minimum read length of 30 bp and at most 3 ambiguous bases per read. Trimmed reads were aligned to the human genome GRCh38 with Bowtie2 (v2.5.4, --sensitive), and only alignments with MAPQ ≥ 10 were retained. SAM files were converted to sorted BAM with samtools (v1.22.1). PCR duplicates were identified using Picard MarkDuplicates (v3.1.0), removed by filtering reads with the duplicate flag, then re-sorted and indexed. Deduplicated human read counts from samtools idxstats were used as effective library sizes. To generate genome-wide signal tracks, deduplicated BAM files were converted to CPM-normalized BigWig tracks using deepTools bamCoverage, with CPM normalization (--normalizeUsing CPM), --binSize 10, --smoothLength 40, --centerReads, --extendReads, and exclusion of chromosomes X, Y, and MT from normalization; an ENCODE-style black list was applied to mask problematic regions. Genome-wide ESR1-binding motifs were identified with FIMO (MEME Suite) using the JASPAR MA0112.4 ESR1 PWM on hg38 with a p -value threshold of 1×10⁻⁴. FIMO output was converted to BED format, and a custom Python script was used to rank sites by FIMO scores. We retained the top 20,000 highest-scoring estrogen-responsive elements (EREs), sorted by genomic coordinate, and assigned unique region IDs. ChIP–seq signal at these EREs was quantified for all samples using bedtools multicov, yielding a region-by-sample count matrix. Differential analysis was performed with edgeR (R v4.5.0). Regions with very low coverage were filtered out. The retaining regions all had ≥ 5 reads in at least one sample. The counts were converted to CPM and log₂(CPM + 1), based on Metadata Encoded Condition (ARMC5 vs vector-transfected) and Treatment (E2 vs vehicle). ARMC5 vs vector-transfected sample pairs were defined a priori, and a paired design was fitted separately for E2 vs vehicle using the model ~ PairID + Condition in limma, such that the Condition coefficient tested ARMC5 vs vector within the pairs. For each ERE, we obtained log₂ fold change (ARMC5 vs vector), moderated t -statistics, p -values and Benjamini–Hochberg–adjusted FDR. Genomic annotation of EREs was performed with ChIPseeker using the TxDb.Hsapiens.UCSC.hg38.knownGene and org.Hs.eg.db, classifying regions relative to promoters, UTRs, exons, introns, downstream and intergenic space, and assigning gene symbols. Volcano plots (log₂ fold change vs − log₁₀FDR) generated with ggplot2 and ggrepel. Motif-centered meta-profiles were generated using deepTools (computeMatrix and plotProfile) from CPM-normalized BigWigs. Locus-level ChIP–seq tracks were visualized and exported using Integrative Genomics Viewer. Breast cancer genome-wide association study (GWAS) analysis Publicly available breast cancer tumor mass GWAS data from the METABRIC study 26 , 27 were accessed from cBioPortal for Cancer Genomics 28 – 30 ( https://www.cbioportal.org ). The cohort has 2509 cases; among them, 2173 have ARMC5 gene copy number variation (CNV) information, while no other ARMC5 mutation information is available. The cases were further stratified by ER expression in the tumor mass by immunohistochemistry, and ER-positive cases (n = 1609) were selected for analysis of the correlation between ARMC5 CNV and relapse-free survival. The results were analyzed statistically with ANOVA followed by Tukey’s Honest Significant Difference Test. Results HSPA1A is a substrate of the novel ARMC5-containing E3 Our previous study showed that ARMC5 is the substrate recognition subunit of a multiple-subunit E3 complex that includes ARMC5, CUL3, and RBX1 31 . For the sake of convenience, we called it ARMC5-E3. Our proteomics results also indicate that ARMC5 interacts with a large number of HSPs 31 . This raises an intriguing question: are some of these HSPs also substrates of ARMC5-E3? As HSP70 is the predominant HSP family and HSPA1A is the dominant HSP70 isoform, we investigated whether HSPA1A is a substrate of ARMC5-E3. In HEK293 cells (Fig. 1A) and MCF7 breast cancer cells (Fig. 1B) transfected with HA-ARMC5-expressing plasmids, HSPA1A was detected in ARMC5 precipitates, demonstrating a physical interaction between the two. Such interaction provided a physical basis for HSPA1A being a substrate of ARMC5-E3. Indeed, in ARMC5 KO tissues such as the lung, brain, thymus, spleen, and adrenal glands, the HSPA1A protein levels were significantly elevated compared to the WT counterparts (Fig. 1C). On the other hand, the HSPA1A mRNA levels of these KO tissues were similar to that of the WT tissues (Fig. 1D). These results indicate that ARMC5 is essential for HSPA1A degradation, and HSPA1A is likely a substrate of ARMC5-E3. The ubiquitination of HSPA1 was significantly reduced in the KO tissue (Fig. 1E). Conversely, when V5-HSPA1A and HA-ARMC5 were co-expressed in HEK293 cells, HSPA1A ubiquitination was drastically increased, compared to V5-HSPA1A co-transfected with an empty vector (Fig. 1F). The results of this section collectively demonstrate that HSPA1A is a substrate of ARMC5-E3, which ubiquitinates it and is responsible for its degradation under physiological conditions (i.e., in the absence of exogenous intervention). ER a interacts with HSPA1A and is a substrate of ARMC5-E3 HSPA1A has multiple client proteins, ERa being one of them 9-11 . We demonstrated that in MCF7 breast cancer cells, the endogenous HSPA1A interacted with endogenous ERa, according to immunoprecipitation (Fig. 2A). It is to be noted that in the IP blotting, the ERa band was at a slightly higher position than that in the lysate. This is a common occurrence due to the higher salt concentration in the IP samples. Interestingly, this interaction occurred only in the absence of E2 (lane 3 of the left panel). After 4 hours of culture in the presence of E2, such interaction was almost undetectable (lane 1 of the left panel). Considering that ARMC5 interacted with and ubiquitinated HSPA1A and that HSPA1A was in close contact with the ERa, is the ERa also a substrate of ARMC5-E3? We performed immunoblotting to assess ERα protein levels in various KO and WT tissues. The result of representative blots is shown in Figure 2B, and densitometric results of the ERa signals normalized with b-actin signals (means + SD) are provided in the right panel. The results indicate that ERα protein levels in the KO brain, kidney, spleen, liver, and lymph nodes were all significantly elevated compared to their WT counterparts. On the other hand, the ERa mRNA levels in the KO tissues were similar to those of the WT counterparts (Fig. 2C), indicating that the increased ERa protein levels in the KO tissues were due to post-transcriptional modulation. Considering that ARMC5-E3 ubiquitinates its substrates and channels the proteasome, we believe that the increased ERa protein levels in the KO tissue are caused by decreased degradation. To validate this notion, we overexpressed ARMC5 in MCF7 breast cancer cells. As expected, in cells with ARMC5 overexpression, the ERa protein level was decreased compared to the vector-transfected cells (Fig. 2D), consistent with the hypothesis that ARMC5-E3 is an E3 specific for ERa and controls its degradation. Interestingly, this E3 only functioned in the absence of estrogen. After 16 hours of culture in the presence of E2, the ERa protein level in the ARMC5-transfected cells became similar to that in the vector-transfected cells. The lack of ARMC5-E3’s effect after the E2 treatment is consistent with the fact that HSPA1A was no longer associated with ERa in the presence of estrogen (Fig. 2A), causing the loss of reach of ARMC5-E3 to ERa. The ERa mRNA levels in the ARMC5-overexpressing MCF7 cells were similar to those of vector-transfected cells (Fig. 2E), indicating the decreased ERa protein expression in the former is due to increased degradation. In the uterus and ovary, where estrogen levels are elevated 32,33 , the ERa protein levels were similar in the KO and WT mice (Fig. 2F), and this is reminiscent of the diminished difference in ARMC5-overexpressing MCF7 cells after estrogen treatment (Fig. 2D). The results above demonstrate that ARMC5 interacts with HSPA1A and its client protein ERa. A 3D model of this ARMC5-HSPA1A-ERα complex was constructed using AlphaFold3 (Fig. 2G). As most ERas exist as homo-dimers, two copies of ERa are included in this 3D model, which illustrates a physical basis for ARMC5-E3 in the degradation of both HSPA1A and ERa. Figure 2H illustrates the complex comprising ARMC5-E3, HSPA1A, and ERa. This shows the possibility that ARMC5-E3 embraces HSPA1A and ERa, and lets its enzymatically active RBX1 subunits ubiquitinate both HSPA1A and ERa. ARMC5-overexpression regulates the expression of genes related to ER signaling We performed RNA-seq on ARMC5-transfected and vector-transfected MCF7 cells cultured without E2 stimulation. Their transcriptomes were then compared with their respective counterparts that were stimulated with E2. The key information of the comparison is presented in Figure 3, and the datasets are appended in S-Tables 3 and 4. The volcano plots (Figs. 3A and 3B) show that in both comparisons (ARMC5-overexpression with versus without E2 stimulation, and vector-transfected MCF7 cells with versus without E2 stimulation), there are a large number of genes (808 genes and 703 genes, respectively) manifesting significant differential expression with FDR 2. The complete datasets are presented in S-Tables 3 and 4. The large number of DE genes indicates that the short 40-minute E2 stimulation was highly effective and powerful for both cell types, given that the mRNA extracted for RNA-seq was mostly pre-existing steady-state mRNA. Secondly, there are many more upregulated genes in both cell types than downregulated genes (Figs. 3C and 3D, bar graphs). Thus, in general, estrogen-responsive elements (EREs) function more as enhancers than as repressors. Lastly, between the ARMC5-overexpressing and vector-transfected groups, the majority of the differentially expressed (DE) genes are shared after E2 stimulation (Fig. 3E: 385 genes in the upregulated cohorts; Fig. 3F: 176 genes in the downregulated cohorts). For these common ones, logically, their DE has no relevance to the ARMC5 levels. Some genes are uniquely upregulated in the ARMC5-overexpressing cells but not in the vector-transfected cells (Fig. 3E; 143 genes). Similarly, there are some uniquely downregulated genes (Fig. 3F, 104 genes) in ARMC5-overexpressing cells. These unique genes, which are the ones in the volcano plot in Figure 3A minus those in the volcano plot in Figure 3B, are illustrated in Figure 4A. The expression changes in these uniquely DE genes are caused by higher ARMC5 levels (compared to vector-transfected cells) in the presence of estrogen. The fold changes and their directions (up- or downregulation) are shown in Figure 4B. Some GO biological process terms of these uniquely up- and downregulated DE genes are presented in Figures 4C and 4D. The full datasets of the terms are presented in S-Tables 5 and 6. Specific biological processes relevant to breast cancers, such as peptide tyrosine phosphorylation, hexose transmembrane transport, mammary gland development, ERK1/2 cascade activation, and leukocyte activation and inflammation, are among the significant terms associated with these DE genes. We channeled these unique DE genes to estrogen pathway-focused Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. The results are presented in S-Figure 1. Among the DE genes upregulated in ARMC5-overexpressing MCF7 cells after E2 stimulation, three of them, i.e ., HB-EGF, SHC, and KRT19, are in the estrogen signaling pathway. For DE genes downregulated in ARMC5-overexpressing MCF7 cells after E2 stimulation, none are in the E2 pathway, though (data not shown). These unique DE genes were also channeled to the breast cancer-focused KEGG pathway analysis, and the results are shown in S-Figure 2. Two genes (FGF and SHC) are found in this pathway. ARMC5-E3 overexpression reduced the estrogen binding to the ERE To assess whether ARMC5-E3 regulates ERa binding to ERE, we conducted ERa ChIP-seq of ARMC5-overexpressing MCF-7 cells, which were cultured in the absence or presence of E2 for 45 minutes. Vector-transfected cells were used as controls. The metagene profiles in Figure 5A illustrate ERα binding to the ERE. The addition of E2 to the cell culture increased the ERa binding to ERE in both the ARMC5-overexpressing cells and vector-transfected control cells (the upper row). Such an increase was less prominent in the former (left panel of the upper row) than in the latter (right panel of the upper row). The difference between the two can also be viewed in the left panel of the bottom row. On the other hand, in the absence of E2 stimulation, the levels of ERa binding to ERE were similar in the vector-transfected control and ARMC5-overexpressing cells (the right panel of the bottom row). This is compatible with our finding of AMC5-E3 as an ERa-specific ubiquitin ligase, which is essential for ERa degradation. The high level of ARMC5-E3 reduced ERα levels in the cytosol. As a consequence, after the E2 stimulation, fewer ERas enter the nuclei to bind to ERE. Volcano plots in Figures 5B and 5C show genes with significant differences (unadjusted p-values) and >2-fold changes in ERa peak density in cells with or without 40-minute E2 stimulation. Figure 5B was derived from ARMC5-overexpressing cells, while Figure 5C was from vector-transfected cells. The ChIP-seq datasets are presented in S-Tables 7 and 8. The increased peak density of EREs in these volcano plots (EREs on the right side of the volcano plots) reflects more ERa binding after the E2 stimulation in these cells. There are also some EREs with reduced peak density (EREs on the left side of the volcano plots). Such reduced density might be caused by a more complex and less understood indirect action of estrogen. We, therefore, are inclined to pay more attention to those EREs with increased peak density after the E2-stimulation. There are 34 such EREs in ARMC5-overexpressing cells (Fig. 5B) and 17 in the vector-transfected cells (Fig. 5C). Two EREs (THSD4 and LINC02038, marked in green in the volcano plots) are common ones between these two groups. If we subtract these two from the upregulated EREs in the ARMC5-overexpressing group, the remaining 32 will be the unique EREs caused by ARMC5 overexpression and E2 treatment. We have selected 4 of these 32 EREs and displayed their ERα peak density tracks (with E2 or vehicle treatment) in Figure 5D. It can readily appreciate the increased peak density centered on ERE (red oval dots) in the E2-treated samples compared to the vehicle-treated samples. In the ARMC5-overexpressing cells after the E2 treatment, their ERa peak density at EREs was lower than that of the vector-transfected cells, according to the metagene profile (Fig. 5A, left panel in the lower row). We retrieved some ERa density tracks of these EREs and presented them in Figure 5E, showing that the peak density around these ERE sites in cells with ARMC5-overexpression was apparently lower than that of cells transfected with vectors, supporting the notion that the overexpressed ARMC5-E3 aggravates ERa degradation, causing fewer ERa available to enter the nuclei for ERE association after the E2 stimulation. ARMC5 gene copy number dose-dependently prolonged the relapse-free survival time of breast cancer patients Considering that ARMC5-E3 is the major E3 responsible for the degradation of ERa, which is essential in breast cancer pathogenesis and prognosis, we queried the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) Genome-Wide Association Study (GWAS) dataset 26,27 to assess how ARMC5 mutations affect breast cancer prognosis. Among 2173 cases in the dataset, we selected 1609 cases that were ER-positive in the tumor mass and had ARMC5 gene mutation information available. In these cases, the only available ARMC5 mutation information was copy number variation (CNV), while no deletions, insertions, or point mutations were reported. The 1609 cases were further classified by pathological classification. Those that could not be classified were grouped under the name of “other.” For the CNV, it was divided into groups of deep deletion (bi-allele deletion, i.e ., 0 copy; n=1), shallow deletion (heterozygous deletion, i.e ., 1 copy; n=36), diploid (normal copy number, i.e ., 2 copies; n=976), gain ( i.e ., 3 copies; n=482), and amplification ( i.e ., more than 3 copies; n=114). The percentages of cases with different CNVs within a pathological category are presented in Figure 6A. In most cases across different categories, ARMC5 copy number was normal (diploid). The ARMC5 gene copy numbers showed a positive correlation with relapse-free survival time in ER-positive patients (Fig. 6B). We conducted a statistical analysis of all comparisons between CNV groups with different copy numbers. As there was only one case in the deep deletion group, this group was excluded from the comparison. The relapse-free survival (RFS) time of the shallow deletion group was significantly shorter than that of the diploid, gain, and amplification groups. Patients in the diploid group have a significantly shorter RFS than those in the amplification group. The RFS probability of different groups at various time points during the follow-up period (varying from 275 to 300 months, depending on the group) is plotted in Figure 6C. Compared with the diploid group, the shallow deletion group had a significantly lower RFS probability, whereas the gain and amplification groups had a considerably higher RFS probability. These results suggest that a higher ARMC5 copy number is positively associated with a longer RFS time and increases RFS probability in ER-positive breast cancer patients. Discussion In this study, we revealed that ARMC5 interacted with HSPA1A, and HSPA1A, in turn, interacted with its client protein ERα. Both were substrates of the novel ARMC5-E3 ubiquitin ligase, which was essential for their degradation under physiological conditions. The ARMC5 deletion mutation caused the accumulation of HSPA1A and ERα. In breast cancer MCF7 cells, ARMC5 overexpression decreased ERα protein levels. It reduced ERα binding to EREs, accompanied by altered transcription of multiple genes in the estrogen signaling and breast cancer tumorigenesis pathways. A tumor mass GWAS analysis of 1609 ER-positive human patients revealed that ARMC5 gene copy number was positively associated with RFS time and the probability of these patients. Elaboration and discussion of some findings in this study are presented below. A new and large repertoire of ARMC5-E3 substrates . Our previous studies demonstrate that ARMC5 is the substrate-recognition subunit of multiple-subunit ubiquitin ligases specific for the largest subunit, RBP1, of RNA polymerase II (Pol II) under physiological conditions 31 . Our immunoprecipitation experiments reveal that this E3 also comprises the scaffold protein CUL3 and the E3 enzymatic subunit RBX1. Furthermore, ARMC5 binds to itself and forms a homodimer in this E3 complex 31 . Self-dimerization implies that each ARMC5-E3 complex may contain two active RBX1 subunits. Very surprisingly, this E3 is not only essential for the RBP1 degradation but also for the degradation of all other 11 subunits of Pol II, even though some of them are pretty far away from RPB1. Our yeast 2-hybrid results detect only the interaction between ARMC5 and RPB1, not any other Pol II subunits, suggesting that RPB1 is the only contact molecule with which ARMC5-E3 interacts with Pol II. How do the other 11 subunits get ubiquitinated if they do not interact with the substrate recognition subunit ARMC5? There are two possible models. A. After ARMC5 binds to RPB1, the enzymatic subunit RBX1 at the other end of the complex can move around and briefly contact other Pol II subunits, ubiquitinating them. B. Since ARMC5-E3 is a dimer linked by two ARMC5 molecules, with one ARMC5 binding to its substrate, RBX1 attached to the other ARMC5 can be quite a distance away and can wrap around the Pol II complex and ubiquitinate any subunits in its reach. This is probably also the case in this study. We detected an interaction between ARMC5 and HSPA1A, but not between ARMC5 and ERa. It is possible that ARMC-E3 directly interacts with and ubiquitinates HSPA1A. While binding HSPA1A, the enzyme-active RBX1 in one or the other arm of the dimeric ARMC5-E3 might wrap around the HSPA1A-ERa complex and ubiquitinate a distant ERa (Fig. 2H). In addition to ERa, HSPA1A has many other client proteins, such as p53 34,35 , CFTR 36,37 , Tau 38-40 , Huntintin 41,42 , Akt 43,44 , etc., whose functions are across a wide range of cellular processes. This raises an interesting question of whether those client proteins are also substrates of ARMC5-E3 and depend, wholly or partially, on this E3 for degradation. On this note, we need to mention that in our immunoprecipitation using ARMC5 as bait, we identified many other members of the HSP family, including HSP90B1, HSP90AB1, HSPA4L, HSPA9, HSPA5, HSPA2, HSPD1, and HSPA1L 31 . These HSPs share some sequence homology, at least in their conserved chaperone function domains 18,45,46 . The sequence homology among the members of the HSP70 subfamily is even higher 20,47 . This finding begs two additional questions: a. whether ARMC5 binds to these HSPs due to their sequence homology, b. whether the client proteins of these HSPs are also substrates of this E3, which controls their degradation by and large 48 . If so, this ARMC5-E3 will be one of the most critical E3s in cell biology. Our additional investigation to address these questions is in progress. The current dogma is that ubiquitin ligases control the substrate specificity of the ubiquitination cascade. As 70% or more of the 20,000 proteins are degraded by the ubiquitin-proteasome pathway 49 , if E3 is very specific, one would expect to see a large number of E3s, probably in the range of more than ten thousand. However, only about 600 different E3s have been documented in humans 50 . Most of these 600 E3s are qualified based on substrate protein-level changes but lack more stringent proof from in vitro ubiquitination assays. How do we explain this vast discrepancy in the expected and discovered E3 numbers? One can, of course, argue that E3 lacks a conserved domain and is thus difficult to detect; many remain to be found. However, our finding related to ARMC5-E3 points to a different conclusion. This E3 has been shown to control the degradation of more than 15 proteins 31,51-54 . If it functions as an E3 for all the HSPs it associates with, and each of these HSPs has a dozen or so client proteins, which this E3 will also act on, then it can channel more than a hundred proteins to the proteasome. The puzzle of a paucity of discovered E3s to ubiquitinate 15,000 proteins can be solved because if each E3 can ubiquitinate 100 proteins, only 150 E3s will be required. This number is within the order of discovered E3. Based on our findings, we argue that the current dogma of E3 specificity needs to be revised: there is a degree of specificity in E3 binding to specific proteins. However, E3 will ubiquitinate anything in the vicinity of its binding partner. A large number of proteins exist as multi-component complexes, either transiently or persistently; if an E3 binds to one of the components, it will ubiquitinate any component in the complex, significantly increasing its substrate pool. Hence, E3 specificity is determined by how close a protein is to the E3, with no specific sequence or tertiary-structure requirements. In that sense, E3s are loosely specific, with one E3 capable of acting on dozens or even hundreds of proteins. This loose E3 specificity has been explored in “Pro-Tac,” which uses a linker to deliver an E3 (pVHL and CRBN are popular choices) to any target protein for ubiquitination and degradation in biotech and therapeutic applications (51-53). ARMC5-E3 maintains ER a homeostasis . Estrogen has profound roles in the reproductive system 55,56 , bone metabolism 57,58 , cardiovascular function 59,60 , cognitive function 61,62 , and mood regulation 63 , in addition to its function in breast tumor oncogenesis 3,64-66 . Its protein level homeostasis is thus of vital importance. Figure 7 illustrates a model based on the results of the current study about how ARMC5-E3 controls the homeostasis of the ERa. In WT cells (Fig. 7A), inactive ERa is associated with HSPA1A in the cytosol (left panel). ARMC5-E3 maintains the homeostasis of the level of these HSPA1A/ERa complexes by ubiquitinating some HSPA1A and ERa, and channeling them to the proteasome for degradation. When estrogen, such as E2, binds to ERa, ERa dissociates from HSPA1A (right panel). At this stage, ARMC5-E3 can no longer reach and degrade ERa, because the latter is pulled away from HSPA1A. The escaped ERa forms a complex with the E2 , enters the nucleus, and binds to estrogen-responsive elements (EREs). The ERa and E2 complex then plays its well-known role as a transcription factor, modulating the transcription of genes containing the EREs 67-70 . In cells with ARMC5 overexpression (Fig. 7B), in the absence of E2 (left panel), the increased level of ARMC5-E3 causes more degradation of its substrates, HSPA1A and ERa , resulting in fewer HSPA1A-ERa complexes. When E2 is presented to these ARMC5-overexpressing cells, due to the reduced availability of ERa-HSPA1A complexes, E2 can only find fewer ERas. Consequently, fewer E2-ERa complexes are formed and enter the nucleus to modulate genes with ERE. The net effect is that E2 becomes less potent in these cells. We observed that ARMC5 KO did not alter ERa levels in the ovary and uterus. How to explain such exceptions? These two organs produce estrogen 71 , and the latter even has specific E2-binding proteins to trap more E2 from the blood. Therefore, the local estrogen levels in these two organs are much higher than in other tissues, and most ERas are probably in the activated form and are not in association with HSPA1A. Hence, they are not subjected, or at least are less subjected to ARMC5-E3-mediated degradation. There are three other known specific E3 for ERa, i.e ., E6AP, CHIP, and SPOP, as reviewed in the introduction 15,16 . It is possible that one or several of these additional E3s are responsible for the degradation of ERa that are not associated with HSPA1A. ERa has a different association with chaperons and has different intracellular locations. It can associate with HSPA1A but can also dissociate from HSPA1A upon estrogen binding in the cytosol. While HSPA1A-associated ERa depends on ARMC5-E3 for its degradation, ERa complexed with other molecules ( e.g ., estrogen) or at different locations ( e.g ., nucleic) might use different E3s, hence the need for multiple E3s for ERa. Our ERa ChIP-seq identified many EREs (36 uniquely upregulated peaks in ARMC5-overexpressing MCF7 cells) with FC>2 and nominal p -value <0.05 in peak density upon E2 stimulation. It should be noted that their p -value significance did not survive the multiple-testing adjustment. Several reasons might contribute to a lack of significantly low adjusted p -values. First, ChIP-seq was performed in cells transiently overexpressing ARMC5. The transfection efficiency was about 25-30%. That translates into a high background noise at about 70%. Secondly, the cells were harvested within 45 minutes of E2 stimulation. These factors might significantly reduce the sensitivity of our statistical analysis. We used a loose criterion of a nominal p -value of 0.05 to construct the volcano plots in Figures 5B and 5C. It is therefore possible that this group contains some false-positive EREs. With that said, the ERE peak track display, we can still easily identify EREs with increased peak density after E2 stimulation in the ARMC5-overexpressing cells (Fig. 5D), and EREs with reduced peak density in the ARMC5-overexpressing cells compared to vector-transfected cells (Fig. 5D), indicating the meaningfulness and usefulness of EREs with nominal p -values of <0.05. The impact of ARMC5 on the estrogen signaling pathways and breast cancer To understand how ARMC5 regulates estrogen signaling pathways and impacts breast cancer oncogenesis, we conducted RNA-seq on ARMC5-overexpressing MCF7 breast cancer cells treated with E2 during the last 2 hours of culture. Among the 247 genes (143 upregulated and 104 downregulated) that were uniquely DE in the ARMC5-overexpressing cells with E2 treatment, according to GO analysis, 2 genes (AREG and FOXF1) are related to mammary gland alveolus and lobule development, 8 genes (IRS2, SOX3, AREG, OTP, FOXF1, INSM1, RCBTB2, and PCSK9) are relevant to gland development, 7 genes (CDH5, ZEB1, INHBA, HPGD, BMP6, CITED1, and BMPER) are related to cell surface serine/threonine receptor signaling, 7 genes (DUSP6, INHBA, CALCR, FGF19, BMPER, RGS14, and TLR9) are involved in the ERK1/ERK2 signaling pathway regulation, 7 genes (FGR, IL12A, FOXF1, CALHM6, CR2, KCNJ8, and TLR9 ) are implicated in leukocyte mediated immune and inflammatory responses. According to KEGG analysis, 3 of these genes (HB-EGF, SHC, and KRT19) are in the estrogen signaling pathway, and 2 (FGF and SHC) are in the breast cancer pathways. The genes in these GO terms and KEGG pathways are highly relevant to the breast cancer ontogenesis and prognosis. They are a trove of interesting candidates that need to be further investigated for their possible roles in prolonging the TFS time and rate in breast cancer patients. ARMC5-E3’s role in cancer . Our in silico search of GWAS datasets reveals that high ARMC5 CNV favors a longer RFS time and higher RFS probability in human ER-positive breast cancer patients. These patients are already on estrogen antagonist therapy and have relatively lower tissue E2 levels. Such a low E2 level creates a favorable environment for ARMC5-E3 to degrade inactive HSPA1A-associated ERa and reduce steady-state ERa levels. Such a reduction, in turn, reduces the tumorigenic effect of the estrogen and promotes the efficacy of estrogen antagonist treatment. This is probably one of the mechanisms by which patients with high ARMC5 CNV have better RFS. Of course, the broad effect of ARMC5-E3 across a large number of substrates can have additional benefits for patient survival. The huge proven and putative target protein repertoire of ARMC5-E3, including ERa and the transcription machinery, is a testament to the vital function of ARMC5. Such vital functions explain why the biallelic deletion of ARMC5 causes 60% perinatal lethality in mice with a 50% CD1 genetic background 72 . The surviving KO mice are dwarfs, whereas heterozygous ARMC5 KO mice have no obvious phenotype 73 . Humans with ARMC5 gene deletion probably have similar phenotypes. We are unlikely to find adult patients with biallelic germline ARMC5 deletions or detrimental biallelic ins/dels, because individuals with such mutations are eliminated early in life. In contrast, the germline ins/dels found in surviving individuals are probably inconsequential. Deleterious somatic biallelic ARMC5 deletion might be detectable in some cells, but they are under very unfavorable conditions to compete with cells with WT ARMC5, whether these cells are normal or malignant. Therefore, the cells or tissue with such a biallelic somatic dels/ins must be rare. Indeed, this is demonstrated in the breast cancer GWAS presented in Figure 6B. Only 1 of 1609 tumor masses with CNV information available harbors a biallelic ARMC5 deletion, which is likely somatic rather than germline. Based on these analyses, we believe it will be more fruitful to study human cohorts with high genomic ARMC5 CNVs rather than ins/dels for any unknown phenotypes related to ARMC5 mutations. Does the beneficial effect of high ARMC5 CNV qualify it as a tumor suppressor gene? There is no universally agreed-upon definition of a tumor suppressor gene, but, generally speaking, to qualify as a tumor suppressor gene, it should increase tumor risk across multiple tissue types. Except for adrenal hypertrophy, a lack of increased tumor incidence of any organs, including the breasts, in thousands of Armc5 KO mice we generated is a testament that ARMC5 is not a bona fide tumor suppressor gene. ARMC5 has profound roles in various cellular processes due to its large number of proven and putative substrates. Its functions are complex and context-dependent. Its roles, exemplified by its function as an E3 for all Pol II subunits, are essential for homeostasis and promote cell growth. Such functions of ARMC5 are the exact opposite of the criteria for a tumor suppressor. In the rare patient tissue sample with an ARMC5 biallelic deletion, as mentioned above, the deletion is likely somatic, and the tumor might harbor mutations in other genes to compensate for the deleterious effects of the ARMC5 deletion, allowing the tumor cells to survive. This patient case is a rare exception rather than the norm. Therefore, calling ARMC5 a tumor suppressor gene is a misnomer. In summary, we identified ARMC5-E3’s two new substrates, HSPA1A and ERα. ARMC5-E3 degrades inactivated HSPA1A-bound ERa. Increased ARMC5 expression reduced steady-state ERa levels. Such an effect probably contributes to a longer relapse-free survival time for breast cancer patients. Hence, ARMC5 CNV is a useful prognostic marker for breast cancer. Declarations Acknowledgments This work was supported by the Jean-Louis Levesque Foundation to J.W. It was also funded in part by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC; RGPIN-2017-04790), the Arthritis Society of Canada, the Canadian Institutes of Health Research (PJT-180284), and the Canadian Rare Disease Models and Mechanisms Network to J.W. Sequencing library construction and Illumina sequencing were performed by the Clinical Research Institute of Montreal Genome Platform and the Genome Quebec Innovation Centre, Montreal. The authors thank the professionals on these platforms for providing these essential services. Data Availability The RNA-seq and ChIP-seq datasets have been deposited in the Gene Expression Omnibus at NCBI under accession numbers GSE312622 (token: urkdimiuzlktncj) and GSE312623 (token: gdshiggcdhmlzcd), respectively. Author contribution X.H. and J.P. conducted experiments. X.H. and Y.H. performed data analysis. X.H., Y.H., and J.W. wrote the manuscript. H.L. and J. W. initiated, designed, and supervised the project . Competing financial interests The authors declare no competing financial interests. References Bentzon, N., Düring, M., Rasmussen, B. B., Mouridsen, H. & Kroman, N. Prognostic effect of estrogen receptor status across age in primary breast cancer. Int J Cancer 122, 1089–1094 (2008). https://doi.org/10.1002/ijc.22892 Jayasekara, H. et al. Mortality after breast cancer as a function of time since diagnosis by estrogen receptor status and age at diagnosis. Int J Cancer 145, 3207–3217 (2019). https://doi.org/10.1002/ijc.32214 Yu, K.-D., Wu, J., Shen, Z.-Z. & Shao, Z.-M. Hazard of Breast Cancer-Specific Mortality among Women with Estrogen Receptor-Positive Breast Cancer after Five Years from Diagnosis: Implication for Extended Endocrine Therapy. The Journal of Clinical Endocrinology & Metabolism 97, E2201–E2209 (2012). https://doi.org/10.1210/jc.2012-2423%J The Journal of Clinical Endocrinology & Metabolism Song, D. et al. ERα and ERβ Homodimers in the Same Cellular Context Regulate Distinct Transcriptomes and Functions. Front Endocrinol (Lausanne) 13, 930227 (2022). https://doi.org/10.3389/fendo.2022.930227 Kumar, R. et al. The dynamic structure of the estrogen receptor. J Amino Acids 2011, 812540 (2011). https://doi.org/10.4061/2011/812540 Tamrazi, A., Carlson, K. E., Daniels, J. R., Hurth, K. M. & Katzenellenbogen, J. A. Estrogen Receptor Dimerization: Ligand Binding Regulates Dimer Affinity and DimerDissociation Rate. Molecular Endocrinology 16, 2706–2719 (2002). https://doi.org/10.1210/me.2002-0250%J Molecular Endocrinology Irani, S. et al. Somatic estrogen receptor α mutations that induce dimerization promote receptor activity and breast cancer proliferation. J Clin Invest 134 (2024). https://doi.org/10.1172/jci163242 Weis, K. E., Ekena, K., Thomas, J. A., Lazennec, G. & Katzenellenbogen, B. S. Constitutively active human estrogen receptors containing amino acid substitutions for tyrosine 537 in the receptor protein. Mol Endocrinol 10, 1388–1398 (1996). https://doi.org/10.1210/mend.10.11.8923465 Dhamad, A. E., Zhou, Z., Zhou, J. & Du, Y. Systematic Proteomic Identification of the Heat Shock Proteins (Hsp) that Interact with Estrogen Receptor Alpha (ERα) and Biochemical Characterization of the ERα-Hsp70 Interaction. PLoS One 11, e0160312 (2016). https://doi.org/10.1371/journal.pone.0160312 Picard, D. Chaperoning steroid hormone action. Trends Endocrinol Metab 17, 229–235 (2006). https://doi.org/10.1016/j.tem.2006.06.003 Silveira, M. A. D. et al. HSP70 mediates a crosstalk between the estrogen and the heat shock response pathways. J Biol Chem 299, 102872 (2023). https://doi.org/10.1016/j.jbc.2023.102872 Zhou, W. & Slingerland, J. M. Links between oestrogen receptor activation and proteolysis: relevance to hormone-regulated cancer therapy. Nat Rev Cancer 14, 26–38 (2014). https://doi.org/10.1038/nrc3622 Helzer, K. T., Hooper, C., Miyamoto, S. & Alarid, E. T. Ubiquitylation of nuclear receptors: new linkages and therapeutic implications. J Mol Endocrinol 54, R151–167 (2015). https://doi.org/10.1530/jme-14-0308 Li, L., Li, Z., Howley, P. M. & Sacks, D. B. E6AP and calmodulin reciprocally regulate estrogen receptor stability. J Biol Chem 281, 1978–1985 (2006). https://doi.org/10.1074/jbc.M508545200 Fan, M., Park, A. & Nephew, K. P. CHIP (carboxyl terminus of Hsc70-interacting protein) promotes basal and geldanamycin-induced degradation of estrogen receptor-alpha. Mol Endocrinol 19, 2901–2914 (2005). https://doi.org/10.1210/me.2005-0111 Byun, B. & Jung, Y. Repression of transcriptional activity of estrogen receptor alpha by a Cullin3/SPOP ubiquitin E3 ligase complex. Mol Cells 25, 289–293 (2008). Milner, C. M. & Campbell, R. D. Structure and expression of the three MHC-linked HSP70 genes. Immunogenetics 32, 242–251 (1990). https://doi.org/10.1007/bf00187095 Serlidaki, D. et al. Functional diversity between HSP70 paralogs caused by variable interactions with specific co-chaperones. Journal of Biological Chemistry 295, 7301–7316 (2020). https://doi.org/10.1074/jbc.RA119.012449 Brocchieri, L., Conway de Macario, E. & Macario, A. J. hsp70 genes in the human genome: Conservation and differentiation patterns predict a wide array of overlapping and specialized functions. BMC Evol Biol 8, 19 (2008). https://doi.org/10.1186/1471-2148-8-19 Daugaard, M., Rohde, M. & Jäättelä, M. The heat shock protein 70 family: Highly homologous proteins with overlapping and distinct functions. FEBS Lett 581, 3702–3710 (2007). https://doi.org/10.1016/j.febslet.2007.05.039 Quintana-Gallardo, L. et al. The cochaperone CHIP marks Hsp70- and Hsp90-bound substrates for degradation through a very flexible mechanism. Scientific Reports 9, 5102 (2019). https://doi.org/10.1038/s41598-019-41060-0 McDonough, H. & Patterson, C. CHIP: a link between the chaperone and proteasome systems. Cell Stress Chaperones 8, 303–308 (2003). https://doi.org/10.1379/1466-1268(2003)008%3C0303:calbtc%3E2.0.co;2 Kaushik, S. & Cuervo, A. M. The coming of age of chaperone-mediated autophagy. Nature Reviews Molecular Cell Biology 19, 365–381 (2018). https://doi.org/10.1038/s41580-018-0001-6 Parzych, K. R. & Klionsky, D. J. An overview of autophagy: morphology, mechanism, and regulation. Antioxid Redox Signal 20, 460–473 (2014). https://doi.org/10.1089/ars.2013.5371 Community, G. The Galaxy platform for accessible, reproducible, and collaborative data analyses: 2024 update. Nucleic Acids Res 52, W83–w94 (2024). https://doi.org/10.1093/nar/gkae410 Pereira, B. et al. The somatic mutation profiles of 2,433 breast cancers refine their genomic and transcriptomic landscapes. Nature Communications 7, 11479 (2016). https://doi.org/10.1038/ncomms11479 Curtis, C. et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature 486, 346–352 (2012). https://doi.org/10.1038/nature10983 Cerami, E. et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov 2, 401–404 (2012). https://doi.org/10.1158/2159-8290.Cd-12-0095 Gao, J. et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal 6, pl1 (2013). https://doi.org/10.1126/scisignal.2004088 de Bruijn, I. et al. Analysis and Visualization of Longitudinal Genomic and Clinical Data from the AACR Project GENIE Biopharma Collaborative in cBioPortal. Cancer Res 83, 3861–3867 (2023). https://doi.org/10.1158/0008-5472.Can-23-0816 Lao, L. et al. ARMC5 is part of an RPB1-specific ubiquitin ligase implicated in adrenal hyperplasia. Nucleic Acids Research 50, 6343–6367 (2022). https://doi.org/10.1093/nar/gkac483%J Nucleic Acids Research Tetsuo, M., Eriksson, H. & Sjövall, J. Gas chromatographic-mass spectrometric analysis of endogenous levels of estradiol in plasma and in cytosol from rat uterus. J Chromatogr 239, 287–300 (1982). https://doi.org/10.1016/s0021-9673(00)81989-5 Giordano, M. V. et al. How long is the ovary relevant for synthesis of steroids after menopause? Gynecol Endocrinol 34, 536–539 (2018). https://doi.org/10.1080/09513590.2017.1417375 Zylicz, M., King, F. W. & Wawrzynow, A. Hsp70 interactions with the p53 tumour suppressor protein. 20, 4634–4638 (2001). https://doi.org/https://doi.org/10.1093/emboj/20.17.4634 Boysen, M., Kityk, R. & Mayer, M. P. Hsp70- and Hsp90-Mediated Regulation of the Conformation of p53 DNA Binding Domain and p53 Cancer Variants. Mol Cell 74, 831–843.e834 (2019). https://doi.org/10.1016/j.molcel.2019.03.032 Bagdany, M. et al. Chaperones rescue the energetic landscape of mutant CFTR at single molecule and in cell. Nature Communications 8, 398 (2017). https://doi.org/10.1038/s41467-017-00444-4 Bergbower, E., Boinot, C., Sabirzhanova, I., Guggino, W. & Cebotaru, L. The CFTR-Associated Ligand Arrests the Trafficking of the Mutant ∆F508 CFTR Channel in the ER Contributing to Cystic Fibrosis. Cell Physiol Biochem 45, 639–655 (2018). https://doi.org/10.1159/000487120 Voss, K., Combs, B., Patterson, K. R., Binder, L. I. & Gamblin, T. C. Hsp70 alters tau function and aggregation in an isoform specific manner. Biochemistry 51, 888–898 (2012). https://doi.org/10.1021/bi2018078 Ryder, B. D., Wydorski, P. M., Hou, Z. & Joachimiak, L. A. Chaperoning shape-shifting tau in disease. Trends in Biochemical Sciences 47, 301–313 (2022). https://doi.org/10.1016/j.tibs.2021.12.009 Moll, A. et al. Hsp multichaperone complex buffers pathologically modified Tau. Nature Communications 13, 3668 (2022). https://doi.org/10.1038/s41467-022-31396-z Ormsby, A. R., Ramdzan, Y. M., Mok, Y. F., Jovanoski, K. D. & Hatters, D. M. A platform to view huntingtin exon 1 aggregation flux in the cell reveals divergent influences from chaperones hsp40 and hsp70. J Biol Chem 288, 37192–37203 (2013). https://doi.org/10.1074/jbc.M113.486944 Rozales, K. et al. Differential roles for DNAJ isoforms in HTT-polyQ and FUS aggregation modulation revealed by chaperone screens. Nature Communications 13, 516 (2022). https://doi.org/10.1038/s41467-022-27982-w Koren, J. et al. Facilitating Akt Clearance via Manipulation of Hsp70 Activity and Levels*. Journal of Biological Chemistry 285, 2498–2505 (2010). https://doi.org /https://doi.org/10.1074/jbc.M109.057208 Yu, W. et al. PER2 interaction with HSP70 promotes cuproptosis in oral squamous carcinoma cells by decreasing AKT stability. Cell Death & Disease 16, 192 (2025). https://doi.org/10.1038/s41419-025-07523-1 Rosenzweig, R., Nillegoda, N. B., Mayer, M. P. & Bukau, B. The Hsp70 chaperone network. Nature Reviews Molecular Cell Biology 20, 665–680 (2019). https://doi.org/10.1038/s41580-019-0133-3 Zuiderweg, E. R., Hightower, L. E. & Gestwicki, J. E. The remarkable multivalency of the Hsp70 chaperones. Cell Stress Chaperones 22, 173–189 (2017). https://doi.org/10.1007/s12192-017-0776-y Maugeri, N., Radhakrishnan, J. & Knight, J. C. Genetic determinants of HSP70 gene expression following heat shock. Hum Mol Genet 19, 4939–4947 (2010). https://doi.org/10.1093/hmg/ddq418 Agopiantz, M., Carnot, M., Denis, C., Martin, E. & Gauchotte, G. Hormone Receptor Expression in Meningiomas: A Systematic Review. Cancers (Basel) 15 (2023). https://doi.org/10.3390/cancers15030980 Voges, D., Zwickl, P. & Baumeister, W. The 26S proteasome: a molecular machine designed for controlled proteolysis. Annu Rev Biochem 68, 1015–1068 (1999). https://doi.org/10.1146/annurev.biochem.68.1.1015 Li, W. et al. Genome-wide and functional annotation of human E3 ubiquitin ligases identifies MULAN, a mitochondrial E3 that regulates the organelle's dynamics and signaling. PLoS One 3, e1487 (2008). https://doi.org/10.1371/journal.pone.0001487 Aoi, Y. et al. SPT5 regulates RNA polymerase II stability via Cullin 3–ARMC5 recognition. 11, eadt5885 (2025). https://doi.org/doi:10.1126/sciadv.adt5885 Luo, H. et al. ARMC5 controls the degradation of most Pol II subunits, and ARMC5 mutation increases neural tube defect risks in mice and humans. Genome Biology 25, 19 (2024). https://doi.org/10.1186/s13059-023-03147-w Okuno, Y., Fukuhara, A., Otsuki, M. & Shimomura, I. ARMC5-CUL3 E3 ligase targets full-length SREBF in adrenocortical tumors. JCI Insight 7 (2022). https://doi.org/10.1172/jci.insight.151390 Uota, A. et al. ARMC5 selectively degrades SCAP-free SREBF1 and is essential for fatty acid desaturation in adipocytes. Journal of Biological Chemistry 300 (2024). https://doi.org/10.1016/j.jbc.2024.107953 Chantalat, E. et al. Estrogen Receptors and Endometriosis. Int J Mol Sci 21 (2020). https://doi.org/10.3390/ijms21082815 Fuentes, N. & Silveyra, P. Estrogen receptor signaling mechanisms. Adv Protein Chem Struct Biol 116, 135–170 (2019). https://doi.org/10.1016/bs.apcsb.2019.01.001 Jilka, R. L. et al. Increased osteoclast development after estrogen loss: mediation by interleukin-6. Science 257, 88–91 (1992). https://doi.org/10.1126/science.1621100 Pantschenko, A. G. et al. Effect of osteoblast-targeted expression of bcl-2 in bone: differential response in male and female mice. J Bone Miner Res 20, 1414–1429 (2005). https://doi.org/10.1359/jbmr.050315 Iorga, A. et al. The protective role of estrogen and estrogen receptors in cardiovascular disease and the controversial use of estrogen therapy. Biol Sex Differ 8, 33 (2017). https://doi.org/10.1186/s13293-017-0152-8 Aryan, L. et al. The Role of Estrogen Receptors in Cardiovascular Disease. Int J Mol Sci 21 (2020). https://doi.org/10.3390/ijms21124314 Bustamante-Barrientos, F. A. et al. The Impact of Estrogen and Estrogen-Like Molecules in Neurogenesis and Neurodegeneration: Beneficial or Harmful? Front Cell Neurosci 15, 636176 (2021). https://doi.org/10.3389/fncel.2021.636176 Yaffe, K., Haan, M., Byers, A., Tangen, C. & Kuller, L. Estrogen use, APOE, and cognitive decline: evidence of gene-environment interaction. Neurology 54, 1949–1954 (2000). https://doi.org/10.1212/wnl.54.10.1949 Albert, K. M. & Newhouse, P. A. Estrogen, Stress, and Depression: Cognitive and Biological Interactions. Annu Rev Clin Psychol 15, 399–423 (2019). https://doi.org/10.1146/annurev-clinpsy-050718-095557 Wang, Z. et al. Identification, cloning, and expression of human estrogen receptor-alpha36, a novel variant of human estrogen receptor-alpha66. Biochem Biophys Res Commun 336, 1023–1027 (2005). https://doi.org/10.1016/j.bbrc.2005.08.226 Toy, W. et al. Activating ESR1 Mutations Differentially Affect the Efficacy of ER Antagonists. Cancer Discov 7, 277–287 (2017). https://doi.org/10.1158/2159-8290.Cd-15-1523 Tang, J., Wu, Z., Tian, Z., Chen, W. & Wu, G. OTUD7B stabilizes estrogen receptor α and promotes breast cancer cell proliferation. Cell Death Dis 12, 534 (2021). https://doi.org/10.1038/s41419-021-03785-7 Seo, H., Seo, H., Lee, H. Y., Lee, S. H. & Park, Y. Comprehensive analysis of cellular estrogen signaling in representative estrogen receptor ligands. Chem Biol Interact 369, 110303 (2023). https://doi.org/10.1016/j.cbi.2022.110303 Helsen, C. & Claessens, F. Looking at nuclear receptors from a new angle. Mol Cell Endocrinol 382, 97–106 (2014). https://doi.org/10.1016/j.mce.2013.09.009 Hewitt, S. C. et al. Novel DNA motif binding activity observed in vivo with an estrogen receptor α mutant mouse. Mol Endocrinol 28, 899–911 (2014). https://doi.org/10.1210/me.2014-1051 Arao, Y., Hamilton, K. J., Coons, L. A. & Korach, K. S. Estrogen receptor α L543A,L544A mutation changes antagonists to agonists, correlating with the ligand binding domain dimerization associated with DNA binding activity. J Biol Chem 288, 21105–21116 (2013). https://doi.org/10.1074/jbc.M113.463455 Nelson, L. R. & Bulun, S. E. Estrogen production and action. J Am Acad Dermatol 45, S116–124 (2001). https://doi.org/10.1067/mjd.2001.117432 Luo, H. et al. ARMC5 controls the degradation of most Pol II subunits, and ARMC5 mutation increases neural tube defect risks in mice and humans. Genome Biol 25, 19 (2024). https://doi.org/10.1186/s13059-023-03147-w Hu, Y. et al. Armc5 deletion causes developmental defects and compromises T-cell immune responses. Nat Commun 8, 13834 (2017). https://doi.org/10.1038/ncomms13834 Additional Declarations There is NO conflict of interest to disclose. Supplementary Files STable1listofAbs20251113.docx List of antibodies used in this study STable2primersequences20251113.docx Sequences of PCR primers STable3RNAseqARMC5withE2vsvehicle20251205.xlsx Differentially expressed genes (FC>2 and FDR2, FDR<0.05) in MCF7 cells transfected with vector and treated with E2 vs vehicle according to RNA-seq STable5GOARMC5UPwithE22025121.xlsx GO analysis in terms of the biological processes for genes that were upregulated in MCF7 cells overexpressing ARMC5 in the presence versus absence of E2 STable6GOARMC5downwithE22025121.xlsx GO analysis in terms of the biological processes for genes that were downregulated in MCF7 cells overexpressing ARMC5 in the presence versus absence of E2 STable7ChIPseqE2vsNEARMC5significantgenes2025123.xlsx EREs with significantly different peak density (p2) in ERa ChIP-seq of ARMC5-overexpressing MCF7 cells in the presence of E2 versus vehicle STable8E2vsNEedgeRlog2CPMdedupLibEVsignificantgenes2025123.xlsx EREs with significantly different peak density (p2) in ERa ChIP-seq of vector-transfected MCF7 cells in the presence of E2 versus vehicle SFigure1KEGGestrogenARMC5specificUPRNAseq.tif Supplementary Figure 1 (S-Figure 1). KEGG estrogen signaling pathway analysis Unique DE genes (FDR2) in HA-ARMC5 overexpressing MCF7 cells with E2 versus vehicle treatment, after their counterparts from the vector-transfected MCF7 cells were subtracted, were subjected to KEGG analysis in terms of the estrogen signaling pathway. Genes framed in red are the ones from our list of the unique DE genes. SFigure2KEGGbreastcancerARMC5specificUPRNAseq.tif S-Figure 2. KEGG breast cancer oncogenesis pathway analysis Unique DE genes (FDR2) in HA-ARMC5 overexpressing MCF7 cells with E2 versus vehicle treatment, after their counterparts from the vector-transfected MCF7 cells were subtracted, were subjected to KEGG analysis in terms of the breast cancer tumorigenesis pathway. Genes in red are from our list of unique DE genes. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8810312","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":588385681,"identity":"adcacbc2-6712-4e0c-8adc-3ff167fa4ab1","order_by":0,"name":"Jiangping Wu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8klEQVRIiWNgGAWjYBACxmYQaQBhP4ALP8CmFosWZjDFAyISiLSRTYIoLcztzM8eFxTcsWuQyDGr5s2xybOXPnzwQwKDnTxuh7GZG88weJYM0nKbd1taMQ9fWrJEAkOyYQNuv5hJ8xgcTmaQBms5nNjDw2MA1HKAEbcW9m9wLcUQLfyffwC12OPWwgO2xQ6khRlqCxvIlkQ8WsqkZxgcTmCTf1YsOXdbWmLPGTYziwSD5GRcWgz7j2+TLvhz2J6f5/DGD2+32SS29zA/vvGhws4WpxagBDOQTmxDFTfAoR4IQKEP0mKPW8koGAWjYBSMeAAAqWdNwFtqQ+QAAAAASUVORK5CYII=","orcid":"","institution":"University of Montreal Department of Medicine","correspondingAuthor":true,"prefix":"","firstName":"Jiangping","middleName":"","lastName":"Wu","suffix":""},{"id":588385682,"identity":"0cb0b9d8-771a-44e2-85b5-756d0e06a946","order_by":1,"name":"Xiao He","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Xiao","middleName":"","lastName":"He","suffix":""},{"id":588385683,"identity":"dea816fb-53b9-4211-a434-722ea0e97617","order_by":2,"name":"Yuhan Ying","email":"","orcid":"https://orcid.org/0000-0003-1831-254X","institution":"the University of Montreal","correspondingAuthor":false,"prefix":"","firstName":"Yuhan","middleName":"","lastName":"Ying","suffix":""},{"id":588385684,"identity":"8758a644-8109-4b6a-9397-a0783d678dbc","order_by":3,"name":"Junzheng Peng","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Junzheng","middleName":"","lastName":"Peng","suffix":""},{"id":588385685,"identity":"122e073e-fffc-4952-ba02-8225eb2226d6","order_by":4,"name":"Hongyu Luo","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Hongyu","middleName":"","lastName":"Luo","suffix":""}],"badges":[],"createdAt":"2026-02-06 19:11:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8810312/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8810312/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103504920,"identity":"8c9d9af2-6977-4a51-805e-99519210292a","added_by":"auto","created_at":"2026-02-26 13:22:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":6074104,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eARMC5 interacted with HSPA1A and controlled its degradation via ubiquitination\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003eand \u003cstrong\u003eb\u003c/strong\u003e. ARMC5 interacted with HSPA1A. HEK293 (\u003cstrong\u003ea\u003c/strong\u003e) or MCF7 cells (\u003cstrong\u003eb\u003c/strong\u003e) were transfected with plasmids expressing ARMC5-HA. The cell lysates were precipitated with anti-HA Ab, and the precipitates were blotted with anti-HSPA1A Ab. Representative results of 3 experiments are shown. \u003cstrong\u003ec\u003c/strong\u003e. HSPA1A protein accumulation in KO mouse tissues as determined by Western blotting. β-actin was used as a loading control. The normalized HSPA1A expression was presented as the signal ratio of HSPA1A to β-actin, determined by densitometry (mean ± SD, n = 3). \u003cstrong\u003ed\u003c/strong\u003e. WT and KO tissues exhibited similar HSPA1A levels. \u003cem\u003eHSPA1A \u003c/em\u003emRNA levels of different WT and KO tissues were determined by RT-qPCR\u003cem\u003e. \u003c/em\u003eB-actin\u003cem\u003e \u003c/em\u003emRNA levels were used as internal controls. The results were expressed as signal ratios of \u003cem\u003eHSPA1A\u003c/em\u003e to β-actin (mean ± SD; n = 3). P-values of comparison between WT and KO samples are indicated in \u003cstrong\u003ec \u003c/strong\u003eand \u003cstrong\u003ed\u003c/strong\u003e (Wilcox rank tests). \u003cstrong\u003ee. \u003c/strong\u003eDecreased ubiquitination of HSPA1A in KO mouse tissues. Lysates from the WT and KO tissues were immunoprecipitated with anti-HSPA1A Ab. The precipitates were blotted with anti-total ubiquitin Ab. \u003cstrong\u003ef\u003c/strong\u003e. Increased SHP1A1 ubiquitination in HEK293 cells with ARMC5 overexpression. HEK293 was transfected with FLAG-ARMC5, HA-ubiquitin, and V5-HSP70A1A. The cell lysates were precipitated with anti-V5 Ab, and the precipitates were blotted with anti-HA Ab. The experiments were repeated 3 times, and representative data are shown.\u003c/p\u003e","description":"","filename":"OnlineFig1.png","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/f9c3d7c8aea986ed04686ba4.png"},{"id":103165894,"identity":"523d1500-0217-4471-a885-762ca6f7c2e9","added_by":"auto","created_at":"2026-02-22 12:35:52","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":8170961,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eHSPA1A interacted with ERaand ARMC5\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e. HSPA1A interacted with ERain the absence of estrogen. MCF7 cells were cultured in estrogen-deprived medium for 72 hours, and the cells were cultured for an additional 4 hours in the presence or absence of E2. The cell lysates were immunoprecipitated with anti-HSPA1A Ab or IgG. The precipitates were blotted with anti-ERa Ab. \u003cstrong\u003eb\u003c/strong\u003e. HSPA1A accumulation in KO mouse tissues according to immunoblotting (left panel). \u003cem\u003eβ\u003c/em\u003e-actin was used as a loading control. The ratios of HSPA1A versus b-actin signals were determined by densitometry (right panel; mean \u003cu\u003e+\u003c/u\u003e SD; n=3). \u003cstrong\u003ec\u003c/strong\u003e. WT and KO tissues presented similar \u003cem\u003eEsr1 \u003c/em\u003emRNA levels. \u003cem\u003eEsr1\u003c/em\u003e mRNA levels in WT and KO tissues were determined by RT-qPCR\u003cem\u003e. β\u003c/em\u003e-actin mRNA levels were used as internal controls. The results were expressed as signal ratios of \u003cem\u003eEsr1\u003c/em\u003e to β-actin (mean ± SD; n = 3). P-values of comparison between WT and KO samples are indicated in \u003cstrong\u003eb \u003c/strong\u003eand \u003cstrong\u003ec\u003c/strong\u003e (Wilcox rank tests). \u003cstrong\u003ed\u003c/strong\u003e. Overexpression of ARMC5 in MCF-7 cells reduces ERα levels. MCF7 cells were transfected with HA-ARMC5 and cultured for up to 24 hours in the presence or absence of E2 (100 nM). HA-ARMC5, ERa, and a-actinin proteins were detected by immunoblotting. a-Actinin was used as a loading control. The ratios of ERaversus a-actinin signals were determined by densitometry (right panel; mean \u003cu\u003e+\u003c/u\u003e SD; n=3). \u003cstrong\u003ee\u003c/strong\u003e. MCF7 cells transfected with HA-ARMC5 constructs or vectors showed similar levels of Esr1 mRNA. \u003cem\u003eEsr1\u003c/em\u003e mRNA levels in transfected cells were determined by RT-qPCR. P-values of comparison between HA-ARMC5- and vector-transfected MCF7 cells are indicated in \u003cstrong\u003ed\u003c/strong\u003e and \u003cstrong\u003ee\u003c/strong\u003e (Wilcox rank tests). \u003cstrong\u003ef.\u003c/strong\u003e KO mouse ovaries and uteruses showed no significant difference in ERα protein levels compared with their WT counterparts. \u003cstrong\u003eg\u003c/strong\u003e. A 3D model of a protein complex containing ARMC5, HSPA1A, and 2 copies of ERa, constructed by AlphaFold3. \u003cstrong\u003eh. \u003c/strong\u003eA diagram illustrating ARMC5-E3 interacting with and ubiquitinating HSPA1A and ERa.The components of ARMC5-E3, \u003cem\u003ei.e\u003c/em\u003e., ARMC5, CUL3, RBX1, ubiquitin-conjugating enzyme Ubc, and ubiquitin (Ub), are illustrated. Two copies of ARMC5-E3 are depicted in this diagram.\u003c/p\u003e","description":"","filename":"OnlineFig2.png","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/99170a25dad41bfa1000f0d0.png"},{"id":103505150,"identity":"43e4f5c6-73d8-4661-ab68-2994d3684a68","added_by":"auto","created_at":"2026-02-26 13:25:22","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":3688352,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eDifferentially expressed genes in MCF7 cells with and without ARMC5 overexpression according to RNA-seq\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eMCF7 cells were transfected with plasmids expressing HA-ARMC5 (ARMC5) or vectors. Seventy-two hours after the transfection, during the last 2 hours of culture, the cells were treated with E2 or vehicle. The cells were harvested, and their RNA was analyzed by RNA-seq.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eA \u003c/strong\u003eand\u003cstrong\u003e b\u003c/strong\u003e. Volcano plots of DE genes from cells with E2 versus vehicle treatment. Dashed vertical lines: -2 and +2-fold change (KO versus WT). A few notable genes are annotated. Horizontal line: FDR=0.05. Colored dots: genes with FDR \u0026lt;0.05. Grey dots: genes with FDR \u003cu\u003e\u0026gt;\u003c/u\u003e0.05. \u003cstrong\u003ea\u003c/strong\u003e. MCF7 cells overexpressing ARMC5 and treated with E2 versus vehicle. \u003cstrong\u003eb\u003c/strong\u003e. MCF7 cells were transfected with the vector and treated with E2 versus vehicle. \u003cstrong\u003ec\u003c/strong\u003e and \u003cstrong\u003ed\u003c/strong\u003e. Bar graphs illustrating the percentage and fold change of genes in MCF7 cells with E2 versus vehicle treatment. \u003cstrong\u003ec\u003c/strong\u003e. MCF7 cells with ARMC5-overexpression. \u003cstrong\u003ed\u003c/strong\u003e. MCF7 cells were transfected with the vector. \u003cstrong\u003ee \u003c/strong\u003eand \u003cstrong\u003ef\u003c/strong\u003e.\u003cstrong\u003e \u003c/strong\u003eVenn diagrams showing the overlap of significantly upregulated (\u003cstrong\u003ee\u003c/strong\u003e) or downregulated (\u003cstrong\u003ef\u003c/strong\u003e) genes (FDR\u0026lt;0.05 and FC\u0026gt;2) in MCF7 cells with ARMC5-overexpression and MCF7 cells transfected with a vector.\u003c/p\u003e","description":"","filename":"OnlineFig3.png","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/4fb7e2182d77bad8708f674d.png"},{"id":103165889,"identity":"515e3531-0637-4053-90c3-956274f586c7","added_by":"auto","created_at":"2026-02-22 12:35:52","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2226081,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eUnique DE genes (FDR\u0026lt;0.05 and FC\u0026gt;2) after E2 stimulation in ARMC5-overexpressing MCF7 cells\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAll the DE genes in the E2 versus vehicle-treated MCF7 cells with ARMC5 overexpression (i.e., those shown in Fig. 3A) and with vector transfection (those shown in Fig. 3B) were registered. Such genes in the latter group were subtracted from the former.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e. Volcano plot of the 143 upregulated ones and 104 downregulated ones (see Venn diagrams in Figure 3E and 3F). \u003cstrong\u003eb\u003c/strong\u003e. Bar graph of these genes showing the fold change and percentage of genes with up- and downregulation. \u003cstrong\u003ec\u003c/strong\u003e and \u003cstrong\u003ed\u003c/strong\u003e. GO analysis in terms of biological process for the upregulated genes (\u003cstrong\u003ec\u003c/strong\u003e; n=143) and downregulated genes (\u003cstrong\u003ed\u003c/strong\u003e; n=104).\u003c/p\u003e","description":"","filename":"OnlineFig4.png","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/cd64a63d2010f770e6e4e7cd.png"},{"id":104808212,"identity":"f8955215-73e2-432f-be1a-3748c57e6624","added_by":"auto","created_at":"2026-03-17 12:32:44","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":2634365,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eERE peak density in MCF7 cells with ARMC5 overexpression according to ERa ChIP-seq\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eMCF7 cells were transfected with an ARMC5-expressing construct or an empty vector. They were treated with E2 or vehicle for the last 45 minutes of culture.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e. Metagene profiles of ChIP-seq peak density centered in ERE. \u003cstrong\u003eb\u003c/strong\u003e and \u003cstrong\u003ec\u003c/strong\u003e. Volcano plots displaying the nominal \u003cem\u003ep\u003c/em\u003e-values and fold changes in the ARMC5-overexpressing cells (\u003cstrong\u003ea\u003c/strong\u003e) and vector-transfected cells (\u003cstrong\u003eb\u003c/strong\u003e) treated with E2 versus vehicle. Dashed vertical lines: -2 and +2-fold change (E2 versus vehicle treatment). Horizontal line: \u003cem\u003ep\u003c/em\u003e=0.05. Colored dots: genes with ERE peak density differences at \u003cem\u003ep\u003c/em\u003e \u0026lt;0.05 and FC\u0026gt;2. Gene names covered with green circles: genes with common DE EREs in the ARMC5-overexpressing cells and vector-transfected cells. \u003cstrong\u003ed\u003c/strong\u003e and \u003cstrong\u003ee\u003c/strong\u003e. ChIP-seq peak density tracks centered on EREs (red oval dots) of example genes. \u003cstrong\u003ed\u003c/strong\u003e. Four example genes with ERE-centered peak density in cells overexpressing ARMC5 after E2 versus vehicle treatment. \u003cstrong\u003ee\u003c/strong\u003e. Four example genes with ERE-centered peak density in the ARMC5-overexpressing cells versus vector-transfected cells, both of which were treated with E2. ARMC5 and vector: cells transfected with an ARMC5-expressing construct or an empty vector, respectively; E2 and vehicle: cells were treated with E2 or vehicle, respectively, for the last 45 minutes of culture.\u003c/p\u003e","description":"","filename":"OnlineFig5.png","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/86c942e5a74d4ec63a848164.png"},{"id":103165901,"identity":"f2ecd20c-c272-4192-b196-d90b8e03fbe9","added_by":"auto","created_at":"2026-02-22 12:35:53","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":2427679,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eHigh copy number variation of the ARMC5 gene was associated with longer relapse-free survival time of breast cancer patients\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe METABRIC GWAS of 2509 cases of breast cancer tumor masses was analyzed for their ARMC5 CNV and ER-expression. Among the 2173 cases with ARMC5 CNV information available, 1609 cases were ER-positive. These 1609 cases were analyzed to determine the patients’ relapse-free survival time. \u0026nbsp;\u003cstrong\u003ea\u003c/strong\u003e. percentage of different ARMC5 copy numbers found in various pathological types of breast cancers. \u003cstrong\u003eb\u003c/strong\u003e. Months of relapse-free survival time of breast cancer patients of all the pathological types with different ARMC5 copy numbers. Deep deletion: bi-allele deletion; shallow deletion: mono-allele deletion; diploid: no deletion nor increased copy number; gain: one additional copy; amplification: more than 3 copies. The results were analyzed using ANOVA followed by Tukey’s Honest Significant Difference Test. The deep deletion group (n = 1) was excluded from the comparison due to insufficient case numbers. Only significant \u003cem\u003eq\u003c/em\u003e-values between different groups are indicated. \u003cstrong\u003ec.\u003c/strong\u003e Relapse-free survival rates of different groups at different times during the follow-up period until 270-300 months. The hazard ratios (95% confidence intervals (CIs)) and \u003cem\u003eq\u003c/em\u003e-values for the different groups, with the diploid group as the control (hazard ratio = 1), are shown. The survival data were analyzed with the Cox proportional hazard model. Ductal NST: invasive ductal carcinoma of no special type (NST); Lobular: invasive lobular carcinoma of the breast; Medullary: medullary carcinoma of the breast; Mixed: mixed-type carcinoma of the breast; Mucinous: mucinous (or colloid) carcinoma of the breast; Tubular: tubular carcinoma of the breast; Tubular cribriform: Tubular cribriform carcinoma of the breast; RFS: relapse-free survival; Del: deletion.\u003c/p\u003e","description":"","filename":"OnlineFig6.png","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/f20e60762f6a588223ec0e50.png"},{"id":103165896,"identity":"7423bbcc-4f27-4f5b-a3e8-cee96ccdd160","added_by":"auto","created_at":"2026-02-22 12:35:52","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":2490804,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eA model depicting ARMC5-E3 controls homeostasis of steady-state ERa\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e. The role of AMRC-E3 in homeostasis of ERa in vector-transfected MCF7 cells in the absence (left panel) and presence of E2(right panel). \u003cstrong\u003eb\u003c/strong\u003e. The role of AMRC-E3 in homeostasis of ERain cells with ARMC5 overexpression in the absence (left panel) and presence of E2 (right panel).\u003c/p\u003e","description":"","filename":"OnlineFig7.png","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/4a0aac4428cc69e33b2389e8.png"},{"id":105564185,"identity":"40a832c0-05e5-4e32-a0f2-03203c9ce127","added_by":"auto","created_at":"2026-03-27 12:48:56","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6549826,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/1a1a179b-b0b6-4f57-85d8-7acd61ec5f8c.pdf"},{"id":103165900,"identity":"e1b55238-47f3-4ecb-846f-6bec5e8ac3af","added_by":"auto","created_at":"2026-02-22 12:35:53","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":13486,"visible":true,"origin":"","legend":"List of antibodies used in this study","description":"","filename":"STable1listofAbs20251113.docx","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/d40e3c29e9f1b230f91cd0e1.docx"},{"id":103165887,"identity":"9233e754-b941-4a97-aec6-b0d85d274560","added_by":"auto","created_at":"2026-02-22 12:35:52","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":13582,"visible":true,"origin":"","legend":"Sequences of PCR primers","description":"","filename":"STable2primersequences20251113.docx","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/4cb8e58f703e0568bd93d803.docx"},{"id":103165888,"identity":"9a1a41c2-6b63-45e3-b7d2-c379a89bf122","added_by":"auto","created_at":"2026-02-22 12:35:52","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":2088626,"visible":true,"origin":"","legend":"Differentially expressed genes (FC\u0026gt;2 and FDR\u0026lt;0.05) in MCF7 cells overexpressing ARMC5 and treated with E2 vs vehicle according to RNA-seq","description":"","filename":"STable3RNAseqARMC5withE2vsvehicle20251205.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/1301cfd258d1f8e1f5e3962e.xlsx"},{"id":103504361,"identity":"0da2e448-cef9-4822-8e46-f5a74dbaec2a","added_by":"auto","created_at":"2026-02-26 13:19:30","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":2057853,"visible":true,"origin":"","legend":"\u003cp\u003eDifferentially expressed genes (FC\u0026gt;2, FDR\u0026lt;0.05) in MCF7 cells transfected with vector and treated with E2 vs vehicle according to RNA-seq\u003c/p\u003e","description":"","filename":"STable4RNAseqEmptyVectorwithE2vsvehicle20251205.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/7bba2e2cf7a09d390edc94f9.xlsx"},{"id":103505036,"identity":"0c599944-aed5-4c9e-a933-ec04ab8e0aef","added_by":"auto","created_at":"2026-02-26 13:22:34","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":202580,"visible":true,"origin":"","legend":"\u003cp\u003eGO analysis in terms of the biological processes for genes that were upregulated in MCF7 cells overexpressing ARMC5 in the presence versus absence of E2\u003c/p\u003e","description":"","filename":"STable5GOARMC5UPwithE22025121.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/63d5678acb4c5c6cbf6c91f1.xlsx"},{"id":103504430,"identity":"0a4ef6d1-7bd0-4faf-a24a-f8c504a6df42","added_by":"auto","created_at":"2026-02-26 13:19:52","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":135539,"visible":true,"origin":"","legend":"\u003cp\u003eGO analysis in terms of the biological processes for genes that were downregulated in MCF7 cells overexpressing ARMC5 in the presence versus absence of E2\u003c/p\u003e","description":"","filename":"STable6GOARMC5downwithE22025121.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/83a1339519dc7db244844709.xlsx"},{"id":103165891,"identity":"e503a769-056c-4012-aec1-23994d361cae","added_by":"auto","created_at":"2026-02-22 12:35:52","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":19421,"visible":true,"origin":"","legend":"\u003cp\u003eEREs with significantly different peak density (p\u0026lt;0.05 and FC\u0026gt;2) in ERa ChIP-seq of ARMC5-overexpressing MCF7 cells in the presence of E2 versus vehicle\u003c/p\u003e","description":"","filename":"STable7ChIPseqE2vsNEARMC5significantgenes2025123.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/0f50b3bfdbe8ef98e5ae4e3c.xlsx"},{"id":103505151,"identity":"b418c7ef-4a2c-49ef-b29b-6d54c9e4f277","added_by":"auto","created_at":"2026-02-26 13:25:22","extension":"xlsx","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":17922,"visible":true,"origin":"","legend":"EREs with significantly different peak density (p\u0026lt;0.05 and FC\u0026gt;2) in ERa ChIP-seq of vector-transfected MCF7 cells in the presence of E2 versus vehicle","description":"","filename":"STable8E2vsNEedgeRlog2CPMdedupLibEVsignificantgenes2025123.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/1f5aaca40892ea58e2a806a6.xlsx"},{"id":103505035,"identity":"833273f3-722b-4046-8c51-573caa5b97da","added_by":"auto","created_at":"2026-02-26 13:22:34","extension":"tif","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":8112784,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eSupplementary Figure 1 (S-Figure 1). KEGG estrogen signaling pathway analysis\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eUnique DE genes (FDR\u0026lt;0.05, FC\u0026gt;2) in HA-ARMC5 overexpressing MCF7 cells with E2 versus vehicle treatment, after their counterparts from the vector-transfected MCF7 cells were subtracted, were subjected to KEGG analysis in terms of the estrogen signaling pathway. Genes framed in red are the ones from our list of the unique DE genes.\u003c/p\u003e","description":"","filename":"SFigure1KEGGestrogenARMC5specificUPRNAseq.tif","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/9efbbf61936260be9cfa9922.tif"},{"id":103165897,"identity":"8f5c4fb4-34db-442e-b5e3-d14bbc92e9e8","added_by":"auto","created_at":"2026-02-22 12:35:52","extension":"tif","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":8718768,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eS-Figure 2. KEGG breast cancer oncogenesis pathway analysis\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eUnique DE genes (FDR\u0026lt;0.05, FC\u0026gt;2) in HA-ARMC5 overexpressing MCF7 cells with E2 versus vehicle treatment, after their counterparts from the vector-transfected MCF7 cells were subtracted, were subjected to KEGG analysis in terms of the breast cancer tumorigenesis pathway. Genes in red are from our list of unique DE genes.\u003c/p\u003e","description":"","filename":"SFigure2KEGGbreastcancerARMC5specificUPRNAseq.tif","url":"https://assets-eu.researchsquare.com/files/rs-8810312/v1/0c5387c9f7d63124ee85fc88.tif"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose.","formattedTitle":"A novel ARMC5-containing ubiquitin ligase controls the degradation of HSPA1A and its client protein estrogen receptor alpha","fulltext":[{"header":"Introduction","content":"\u003cp\u003eEstrogen, a steroid hormone, plays a vital role in many, if not all, of our biological systems and has an outsized impact on many processes, such as reproduction, development, cardiovascular health, bone metabolism, immunity, brain function, and oncogenesis, particularly breast cancer oncogenesis\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eClassical estrogen receptors (ER) are intracellular proteins with two subtypes: ERα and ERβ. They form homodimers or heterodimers\u003csup\u003e\u003cspan additionalcitationids=\"CR5\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. The dominant form of dimers depends on tissue types. In breast cancer cells, the ERα homodimer is the major type\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn the absence of estrogen, ER dimers are mostly located in the cytosol and bound to heat shock proteins (HSPs)\u003csup\u003e\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. When ER dimers are bound to the ligand estrogen, HSPs will chaperone them into the nuclei. Although HSP90 is the major chaperone for ER, HSPA1A also participates in this function\u003csup\u003e\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. In the nuclei, ER dissociates from HSPs to bind estrogen-responsive elements (ERE) of various genes, functioning as transcription factors to regulate gene expression in an estrogen-dependent way.\u003c/p\u003e \u003cp\u003eAfter ERs complete their function as transcription factors, they are recycled and reused, but eventually, they are degraded through the proteasome pathway in both the nuclei and cytosol\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. E6AP has been identified as an ERα-specific ubiquitin ligase (E3) for ER degradation via the proteasome\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. E6AP directly binds to ER, and its overexpression in cells leads to a decreased level of ER. A Hsc70 (heat shock cognate 70)-interacting protein, CHIP (carboxyl terminus of HSC70-interacting protein), has been documented as another E3 to ubiquitinate ERα, leading to its degradation\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. In addition, a Cullin3-containing E3 called SPOP also acts on ERα \u003csup\u003e16\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHSPs are a family of proteins comprising small HSPs, HSP40, HSP60, HSP70, HSP90, and large HSPs. They are found in various cellular compartments and play vital physiological roles in the folding, stability, translocation, and degradation of their client proteins. During cellular stress, such as heat, oxidation, and inflammation, HSPs are upregulated to protect cells from adverse environments\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHSP70 is a well-studied family of HSP. It has 13 isoforms\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, which are of similar size but are coded by different genes. HSP70-1, a.k.a. HSPA1A, is the dominant isoform. It binds to nascent polypeptides, preventing them from aggregating. It also chaperones its client proteins across bilayer membranes during protein translocation. HSPA1A is implicated in the degradation of its client proteins, such as p53, by associating with a p53-specific ubiquitin ligase (E3) CHIP (C-terminus of HSP70-interacting protein), enabling the latter to ubiquitinate p53 for the proteasome-mediated degradation\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. ER is one of the HSPA1A client proteins. Whether HSPA1A plays a role in ER degradation has not been investigated.\u003c/p\u003e \u003cp\u003eHSPA1A is eventually degraded via the proteasome pathway after being ubiquitinated by CHIP\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, although it can also be delivered to the lysosome for degradation\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Thus, it seems that HSPA1A-associated E3s can control the degradation of both HSPA1A and its client proteins. Additional HSPA1A-specific E3s might exist that, in theory, could regulate the degradation of both HSPA1A and its client proteins.\u003c/p\u003e \u003cp\u003eDuring our recent study of a novel multiple-subunit E3, ARMC5-CUL3-RBX1, which uses ARMC5 as its substrate recognition subunit, we employed proteomics to identify ARMC5-interacting proteins with a view to identifying additional substrates of this E3. We identified at least 10 members of the HSP family associated with ARMC5. This raises an intriguing possibility: these HSPs, along with their client proteins, may be potential substrates of this E3. We elected to investigate HSPA1A in detail for this possibility.\u003c/p\u003e \u003cp\u003eIn this study, we validated interactions between ARMC5 and HSPA1A, and between HSPA1A and its client protein ERα, using immunoprecipitation. HSPA1A and its client protein ERα are indeed substrates of the ARMC5-containing E3, as ARMC5 gene knockout caused the accumulation of these two proteins in various tissues of mice. Conversely, ARMC5 overexpression in human breast cancer MCF7 cells reduced the protein levels of HSPA1A and ERα. Functionally, according to RNA-seq, ARMC5 overexpression altered the expression of 247 genes, some of which are known to be involved in breast cancer tumorigenesis. ARMC5 overexpression also decreased ERα binding to EREs, as determined by ChIP-seq.\u0026nbsp;Consistent with these findings at the cellular and molecular levels, human genetic studies showed a significant correlation between increased ARMC5 copy numbers and reduced incidences of relapse in breast cancer patients. Thus, our study identified a novel E3 ligase for ERα and demonstrated its involvement in regulating the expression of estrogen-responsive genes in breast cancer cells. ARMC5 mutations in breast cancer patients are implicated in the prognosis of breast cancer.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eWestern blotting\u003c/h2\u003e \u003cp\u003eFor mouse tissues, they were homogenized in RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate) in the presence of protease inhibitors (Roche). For cultured cells, they were lysed directly in RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate) supplemented with protease inhibitors. Protein concentrations were determined by BCA assays (Thermo Fisher). Equal amounts of total protein were loaded on 10% SDS-PAGE gels and transferred onto PVDF membranes (Bio-Rad). Membranes were blocked for 1 hour in 5% non-fat milk in TBS-T (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% Tween 20) and incubated overnight at 4\u0026deg;C with primary antibodies (1:1000 dilution for all the Abs used in Western blotting) in TBS-T. A list of Abs used in Western blotting is provided in Supplementary Table\u0026nbsp;1 (S-Table\u0026nbsp;1). After washing in TBS-T, membranes were incubated for 1 hour at room temperature with HRP-conjugated secondary antibodies (Cell Signaling). Immunoreactive bands were detected using enhanced chemiluminescence (ECL) and visualized on X-ray film. Band intensities were quantified using ImageJ (NIH) and normalized to β-actin or α-actinin to ensure equal loading. The Wilcoxon rank test was used for comparison.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eMCF7 cell culture\u003c/h3\u003e\n\u003cp\u003eTo minimize the estrogenic activity of the culture medium, we cultured MCF7 cells in phenol red-free DMEM (Gibco) supplemented with 10% charcoal-stripped fetal bovine serum and 1% penicillin-streptomycin in a humidified incubator at 37\u0026deg;C with 5% CO₂ for 72 hours. One hundred nM 17β-estradiol (E2; final concentration) or vehicle (ethanol) was then added to some cultures for different periods of time.\u003c/p\u003e\n\u003ch3\u003eCo-immunoprecipitation\u003c/h3\u003e\n\u003cp\u003eTo detect interactions between ARMC5 and HSP70A1A, HEK293 or MCF7 cells were transfected with HA-ARMC5-expressing plasmids. After 48 hours, cells were lysed with TNE buffer (50 mM Tris-HCl at pH 7.4, 100 mM NaCl, 0.1 mM EDTA, 1% Triton X-100) containing protease inhibitors. The lysate was incubated overnight at 4\u0026deg;C with an anti-HA antibody (1:100 dilution), then reacted with protein G-conjugated magnetic beads for 2 hours at room temperature. The beads were washed 3 times with TNE buffer, and the proteins were eluted with a 2\u003cem\u003e\u0026times;\u003c/em\u003e SDS-loading buffer (4% SDS, 20% glycerol, 200 mM DTT, 0.01% bromophenol blue, and 0.1 M Tris HCl, pH 6.8) at 95\u0026deg;C for 15 minutes. The input lysates and eluates were subjected to Western blotting analysis for the presence of HSP70A1A.\u003c/p\u003e \u003cp\u003eTo detect the interaction between HSP70A1A and ERα, we cultured MCF7 cells in medium containing 10% stripped serum, without phenol red, for 72 hours. One hundred nM E2 was added to the culture for a given period. The cells were lysed in TNE buffer with protease inhibitors. The lysates were reacted with anti-human HSPA1A Ab (Cell Signaling Technology; 1:100 dilution). The remaining steps were the same as described above. The input lysates and eluates were subjected to Western blotting analysis for the presence of ERα.\u003c/p\u003e\n\u003ch3\u003eDetection of protein ubiquitination\u003c/h3\u003e\n\u003cp\u003eTo detect HSP70A1 ubiquitination, mouse tissue lysates were incubated with anti-HSP70A1A Ab (1:100) overnight at 4\u0026deg;C. HSP701A1 was precipitated with protein G-conjugated magnetic beads at room temperature for 2 hours. HSP701A1 ubiquitination in the eluates was detected using Western blotting\u003c/p\u003e\n\u003ch3\u003eRT-qPCR\u003c/h3\u003e\n\u003cp\u003eTotal RNA from mouse tissue was isolated with TRIzol. Total RNA from cells was extracted with the RNeasy Mini Kit (Qiagen). RNA was reverse transcribed into cDNA using the iScript cDNA Synthesis Kit (Bio-Rad) at 42\u0026deg;C for 30 minutes, followed by 85\u0026deg;C for 5 minutes to inactivate the enzyme. The resulting cDNA was amplified by PCR using gene-specific primers and Taq DNA polymerase (Thermo Scientific) under the following cycling conditions: 95\u0026deg;C for 2 minutes, followed by 40 cycles of 95\u0026deg;C for 15 seconds and 60\u0026deg;C for 60 seconds. A list of primers used for RT-qPCR is presented in S-Table\u0026nbsp;2. The levels of β-actin signals were used as internal controls. The CT ratios of the genes of interest and those of β-actin were presented as gene expression levels. The Wilcoxon rank test was used for statistical comparison.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eRNA-sequencing (RNA-Seq)\u003c/h2\u003e \u003cp\u003eMCF7 cells were transfected with a construct expressing HA-ARMC5 or an empty vector and cultured in regular medium for 24 hours. The cells were then switched to medium containing 10% stripped serum for an additional 48 hours. At that time, 100 nM E2 (final concentration) or vehicle (ethanol) was added to the medium, and the cells were cultured for an additional 2 hours. Total RNA was extracted from the cells using the RNeasy Mini Kit (Qiagen) following the manufacturer\u0026rsquo;s protocol. The integrity and quantity of the isolated RNA were assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies) with an RNA 6000 Pico kit. Libraries were prepared using an equal amount of total RNA per sample (150 ng). First, ribosomal RNAs were depleted using a KAPA RiboErase (HMR) kit, and then libraries were prepared using a KAPA RNA Hyperprep Kit (Roche Diagnostics) with 11 cycles of final amplification.\u003c/p\u003e \u003cp\u003eLibrary size distribution was assessed on a 2100 bioanalyzer (Agilent Technologies) using a High Sensitivity DNA Kit, and libraries were quantified by qPCR. Equimolar libraries were sequenced in paired-end reads (PE100) on a Novaseq system (Illumina) at 50 M reads per library.\u003c/p\u003e \u003cp\u003eRNA-seq data were analyzed using the Galaxy server\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://usegalaxy.org\u003c/span\u003e\u003cspan address=\"https://usegalaxy.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Quality control of the raw FASTQ files was performed using FastQC, followed by trimming low-quality reads and adapter sequences with Trimmomatic. Cleaned reads were then aligned to the reference genome (GRCh38) using the STAR aligner with default parameters. Gene-level expression was quantified from aligned reads with featureCounts, and differentially expressed genes were identified using edgeR. Graphical outputs, including heatmaps and volcano plots, were generated ggplot2 in R.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eERα chromatin immunoprecipitation followed by sequencing (ChIP-seq)\u003c/h3\u003e\n\u003cp\u003eMCF7 cells were transfected with empty vectors or HA-ARMC5-expressing constructs as described above and cultured in regular medium for 24 hours. Then, they were cultured in medium containing 10% stripped FCS for an additional 48 hours. At that time, 100 nM E2 or vehicle (ethanol) was added to the culture. The cells were harvested at 0 and 45 minutes after the E2 treatment.\u003c/p\u003e \u003cp\u003eThe HA-ARMC5 and vector-transfected MCF7 cells from three independent experiments were used for ERα ChIP-seq.\u0026nbsp;Approximately 8\u0026ndash;10 \u0026times; 106 cells per sample were crosslinked with 1% formaldehyde for 10 minutes at room temperature, then quenched with 125 mM glycine for 10 minutes while rotating. The cells were pelleted, washed with cold PBS, and resuspended in cell swelling buffer (25 mM HEPES, pH 7.9, 1.5 mM MgCl\u003csub\u003e2\u003c/sub\u003e, 10 mM KCl, 0.1% NP-40, supplemented with protease and phosphatase inhibitors) on ice for 10 minutes to release nuclei. Nuclei were pelleted, resuspended in ChIP sonication buffer (50 mM HEPES pH 7.9, 140 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% Na-deoxycholate, 0.1% SDS) supplemented with protease and phosphatase inhibitors, and sonicated using a probe-based sonicator (FB120 with a CL-18 probe; ThermoFisher) at 25% amplitude with 30-second pulses at 30-second intervals for a total of 2.5 minutes. Sonicated chromatin was stored until immunoprecipitation.\u003c/p\u003e \u003cp\u003eChromatin fragmentation was assessed by retrieving 5% of sonicated nuclei and treating them with 10 \u0026micro;g RNase A for 45 minutes at 37\u0026deg;C, followed by 20 \u0026micro;g proteinase K for 1 hour at 65\u0026deg;C. Purified DNA (QIAquick PCR Purification Kit) was quantified using a Nanodrop 1000 Spectrometer (ThermoFisher), and fragment length was determined by electrophoresis.\u003c/p\u003e \u003cp\u003eFor immunoprecipitation, anti-human ERα Ab (0.6 \u0026micro;g per sample; S-Table\u0026nbsp;1) was added to each sonicated chromatin sample, and the mixture was incubated at 4\u0026deg;C overnight. The mixture was then reacted with 50 \u0026micro;l SureBeads\u0026trade; Protein G Magnetic Beads (Bio-Rad) per sample for 2 hours at 4\u0026deg;C. Beads were washed sequentially with ChIP sonication buffer, LiCl wash buffer (20 mM Tris pH 8.0, 1 mM EDTA, 250 mM LiCl, 0.5% NP-40, 0.5% Na-deoxycholate), and TE buffer (10 mM Tris pH 8, 0.1 mM EDTA). Chromatin was eluted with ChIP elution buffer (50 mM Tris pH 8.0, 10 mM EDTA, 1% SDS) with agitation for 15 minutes at 1100 rpm. Immunoprecipitated chromatin was de-crosslinked at 65\u0026deg;C overnight, treated with RNase A (20 \u0026micro;g per sample) at 37\u0026deg;C for 1 hour and proteinase K (200 \u0026micro;g per sample) for 2 hours at 65\u0026deg;C. DNA was purified with the QIAquick PCR Purification Kit (QIAGEN). Size distribution and concentration of immunoprecipitated and input samples were evaluated on a 2100 bioanalyzer (Agilent Technologies) using a High Sensitivity DNA Kit. ChIP-seq libraries were prepared using the KAPA Hyperprep Library Kit (Roche Diagnostics). Normalization of the sample quantities was performed prior to final amplification based on qPCR quantification of the ligation products. Library size distribution was assessed on a 2100 bioanalyzer (Agilent Technologies) using a High Sensitivity DNA Kit, and libraries were quantified by qPCR. Equimolar libraries were sequenced in paired-end reads (PE100) on a Novaseq system (Illumina) at 50\u0026nbsp;million reads per library.\u003c/p\u003e \u003cp\u003eChIP\u0026ndash;seq data from 12 EV/HA samples were processed on a SLURM-based HPC cluster using a standardized pipeline. Raw paired-end reads were quality-checked with FastQC and trimmed using cutadapt to remove Illumina TruSeq adapters (R1: AGATCGGAAGAGCACACGTCTGAACTCCAGTCA; R2: AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT), low-quality bases, and poly-A/N tails, requiring a minimum read length of 30 bp and at most 3 ambiguous bases per read. Trimmed reads were aligned to the human genome GRCh38 with Bowtie2 (v2.5.4, --sensitive), and only alignments with MAPQ\u0026thinsp;\u0026ge;\u0026thinsp;10 were retained. SAM files were converted to sorted BAM with samtools (v1.22.1). PCR duplicates were identified using Picard MarkDuplicates (v3.1.0), removed by filtering reads with the duplicate flag, then re-sorted and indexed. Deduplicated human read counts from samtools idxstats were used as effective library sizes.\u003c/p\u003e \u003cp\u003eTo generate genome-wide signal tracks, deduplicated BAM files were converted to CPM-normalized BigWig tracks using deepTools bamCoverage, with CPM normalization (--normalizeUsing CPM), --binSize 10, --smoothLength 40, --centerReads, --extendReads, and exclusion of chromosomes X, Y, and MT from normalization; an ENCODE-style black list was applied to mask problematic regions.\u003c/p\u003e \u003cp\u003eGenome-wide ESR1-binding motifs were identified with FIMO (MEME Suite) using the JASPAR MA0112.4 ESR1 PWM on hg38 with a \u003cem\u003ep\u003c/em\u003e-value threshold of 1\u0026times;10⁻⁴. FIMO output was converted to BED format, and a custom Python script was used to rank sites by FIMO scores. We retained the top 20,000 highest-scoring estrogen-responsive elements (EREs), sorted by genomic coordinate, and assigned unique region IDs. ChIP\u0026ndash;seq signal at these EREs was quantified for all samples using bedtools multicov, yielding a region-by-sample count matrix.\u003c/p\u003e \u003cp\u003eDifferential analysis was performed with edgeR (R v4.5.0). Regions with very low coverage were filtered out. The retaining regions all had\u0026thinsp;\u0026ge;\u0026thinsp;5 reads in at least one sample. The counts were converted to CPM and log₂(CPM\u0026thinsp;+\u0026thinsp;1), based on Metadata Encoded Condition (ARMC5 vs vector-transfected) and Treatment (E2 vs vehicle). ARMC5 vs vector-transfected sample pairs were defined a priori, and a paired design was fitted separately for E2 vs vehicle using the model\u0026thinsp;~\u0026thinsp;PairID\u0026thinsp;+\u0026thinsp;Condition in limma, such that the Condition coefficient tested ARMC5 vs vector within the pairs. For each ERE, we obtained log₂ fold change (ARMC5 vs vector), moderated \u003cem\u003et\u003c/em\u003e-statistics, \u003cem\u003ep\u003c/em\u003e-values and Benjamini\u0026ndash;Hochberg\u0026ndash;adjusted FDR.\u003c/p\u003e \u003cp\u003eGenomic annotation of EREs was performed with ChIPseeker using the TxDb.Hsapiens.UCSC.hg38.knownGene and org.Hs.eg.db, classifying regions relative to promoters, UTRs, exons, introns, downstream and intergenic space, and assigning gene symbols. Volcano plots (log₂ fold change vs\u0026thinsp;\u0026minus;\u0026thinsp;log₁₀FDR) generated with ggplot2 and ggrepel. Motif-centered meta-profiles were generated using deepTools (computeMatrix and plotProfile) from CPM-normalized BigWigs. Locus-level ChIP\u0026ndash;seq tracks were visualized and exported using Integrative Genomics Viewer.\u003c/p\u003e\n\u003ch3\u003eBreast cancer genome-wide association study (GWAS) analysis\u003c/h3\u003e\n\u003cp\u003ePublicly available breast cancer tumor mass GWAS data from the METABRIC study\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e were accessed from cBioPortal for Cancer Genomics\u003csup\u003e\u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cbioportal.org\u003c/span\u003e\u003cspan address=\"https://www.cbioportal.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The cohort has 2509 cases; among them, 2173 have ARMC5 gene copy number variation (CNV) information, while no other ARMC5 mutation information is available. The cases were further stratified by ER expression in the tumor mass by immunohistochemistry, and ER-positive cases (n\u0026thinsp;=\u0026thinsp;1609) were selected for analysis of the correlation between ARMC5 CNV and relapse-free survival. The results were analyzed statistically with ANOVA followed by Tukey\u0026rsquo;s Honest Significant Difference Test.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eHSPA1A is a substrate of the novel ARMC5-containing E3\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur previous study showed that ARMC5 is the substrate recognition subunit of a multiple-subunit E3 complex that includes ARMC5, CUL3, and RBX1\u003csup\u003e31\u003c/sup\u003e. For the sake of convenience, we called it ARMC5-E3. Our proteomics results also indicate that ARMC5 interacts with a large number of HSPs\u003csup\u003e31\u003c/sup\u003e. This raises an intriguing question: are some of these HSPs also substrates of ARMC5-E3? As HSP70 is the predominant HSP family and HSPA1A is the dominant HSP70 isoform, we investigated whether HSPA1A is a substrate of ARMC5-E3.\u003c/p\u003e\n\u003cp\u003eIn HEK293 cells (Fig. 1A) and MCF7 breast cancer cells (Fig. 1B) transfected with HA-ARMC5-expressing plasmids, HSPA1A was detected in ARMC5 precipitates, demonstrating a physical interaction between the two. Such interaction provided a physical basis for HSPA1A being a substrate of ARMC5-E3.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIndeed, in ARMC5 KO tissues such as the lung, brain, thymus, spleen, and adrenal glands, the HSPA1A protein levels were significantly elevated compared to the WT counterparts (Fig. 1C). On the other hand, the HSPA1A mRNA levels of these KO tissues were similar to that of the WT tissues (Fig. 1D). These results indicate that ARMC5 is essential for HSPA1A degradation, and HSPA1A is likely a substrate of ARMC5-E3.\u003c/p\u003e\n\u003cp\u003eThe ubiquitination of HSPA1 was significantly reduced in the KO tissue (Fig. 1E). Conversely, when V5-HSPA1A and HA-ARMC5 were co-expressed in HEK293 cells, HSPA1A ubiquitination was drastically increased, compared to V5-HSPA1A co-transfected with an empty vector (Fig. 1F).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe results of this section collectively demonstrate that HSPA1A is a substrate of ARMC5-E3, which ubiquitinates it and is responsible for its degradation under physiological conditions (i.e., in the absence of exogenous intervention).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eER\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003ea\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003e\u0026nbsp;interacts with HSPA1A and is a substrate of ARMC5-E3\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHSPA1A has multiple client proteins, ERa\u0026nbsp;being one of them \u003csup\u003e9-11\u003c/sup\u003e. We demonstrated that in MCF7 breast cancer cells, the endogenous HSPA1A interacted with endogenous ERa, according to immunoprecipitation (Fig. 2A). It is to be noted that in the IP blotting, the ERa\u0026nbsp;band was at a slightly higher position than that in the lysate. This is a common occurrence due to the higher salt concentration in the IP samples. Interestingly, this interaction occurred only in the absence of E2 (lane 3 of the left panel). After 4 hours of culture in the presence of E2, such interaction was almost undetectable (lane 1 of the left panel).\u003c/p\u003e\n\u003cp\u003eConsidering that ARMC5 interacted with and ubiquitinated HSPA1A and that HSPA1A was in close contact with the ERa, is the ERa\u0026nbsp;also a substrate of ARMC5-E3? We performed immunoblotting to assess ERα protein levels in various KO and WT tissues. The result of representative blots is shown in Figure 2B, and densitometric results of the ERa\u0026nbsp;signals normalized with\u0026nbsp;b-actin signals (means \u003cu\u003e+\u003c/u\u003e SD) are provided in the right panel. The results indicate that ERα protein levels in the KO brain, kidney, spleen, liver, and lymph nodes were all significantly elevated compared to their WT counterparts. On the other hand, the ERa\u0026nbsp;mRNA levels in the KO tissues were similar to those of the WT counterparts (Fig. 2C), indicating that the increased ERa\u0026nbsp;protein levels in the KO tissues were due to post-transcriptional modulation. Considering that ARMC5-E3 ubiquitinates its substrates and channels the proteasome, we believe that the increased ERa\u0026nbsp;protein levels in the KO tissue are caused by decreased degradation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo validate this notion, we overexpressed ARMC5 in MCF7 breast cancer cells. As expected, in cells with ARMC5 overexpression, the ERa\u0026nbsp;protein level was decreased compared to the vector-transfected cells (Fig. 2D), consistent with the hypothesis that ARMC5-E3 is an E3 specific for ERa\u0026nbsp;and controls its degradation. Interestingly, this E3 only functioned in the absence of estrogen. After 16 hours of culture in the presence of E2, the ERa\u0026nbsp;protein level in the ARMC5-transfected cells became similar to that in the vector-transfected cells. The lack of ARMC5-E3’s effect after the E2 treatment is consistent with the fact that HSPA1A was no longer associated with ERa\u0026nbsp;in the presence of estrogen (Fig. 2A), causing the loss of reach of ARMC5-E3 to ERa. The ERa\u0026nbsp;mRNA levels in the ARMC5-overexpressing MCF7 cells were similar to those of vector-transfected cells (Fig. 2E), indicating the decreased ERa\u0026nbsp;protein expression in the former is due to increased degradation.\u003c/p\u003e\n\u003cp\u003eIn the uterus and ovary, where estrogen levels are elevated\u003csup\u003e32,33\u003c/sup\u003e, the ERa\u0026nbsp;protein levels were similar in the KO and WT mice (Fig. 2F), and this is reminiscent of the diminished difference in ARMC5-overexpressing MCF7 cells after estrogen treatment (Fig. 2D).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe results above demonstrate that ARMC5 interacts with HSPA1A and its client protein ERa. A\u0026nbsp;3D model of this ARMC5-HSPA1A-ERα complex was constructed using AlphaFold3 (Fig. 2G). As most ERas exist as homo-dimers, two copies of ERa\u0026nbsp;are included in this 3D model, which illustrates a physical basis for ARMC5-E3 in the degradation of both HSPA1A and ERa. Figure 2H illustrates the complex comprising ARMC5-E3, HSPA1A, and ERa. This shows the possibility that ARMC5-E3 embraces HSPA1A and ERa, \u0026nbsp;and lets its enzymatically active RBX1 subunits ubiquitinate both HSPA1A and ERa.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eARMC5-overexpression regulates the expression of genes related to ER signaling\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe performed RNA-seq on ARMC5-transfected and vector-transfected MCF7 cells cultured without E2 stimulation. Their transcriptomes were then compared with their respective counterparts that were stimulated with E2. The key information of the comparison is presented in Figure 3, and the datasets are appended in S-Tables 3 and 4. The volcano plots (Figs. 3A and 3B) show that in both comparisons (ARMC5-overexpression with versus without E2 stimulation, and vector-transfected MCF7 cells with versus without E2 stimulation), there are a large number of genes (808 genes and 703 genes, respectively) manifesting significant differential expression with FDR \u0026lt;0.05 and fold change (FC) \u0026gt; 2. The complete datasets are presented in S-Tables 3 and 4. The large number of DE genes indicates that the short 40-minute E2 stimulation was highly effective and powerful for both cell types, given that the mRNA extracted for RNA-seq was mostly pre-existing steady-state mRNA. Secondly, there are many more upregulated genes in both cell types than downregulated genes (Figs. 3C and 3D, bar graphs). Thus, in general, estrogen-responsive elements (EREs) function more as enhancers than as repressors. Lastly, between the ARMC5-overexpressing and vector-transfected groups, the majority of the differentially expressed (DE) genes are shared after E2 stimulation (Fig. 3E: 385 genes in the upregulated cohorts; Fig. 3F: 176 genes in the downregulated cohorts). For these common ones, logically, their DE has no relevance to the ARMC5 levels. Some genes are uniquely upregulated in the ARMC5-overexpressing cells but not in the vector-transfected cells (Fig. 3E; 143 genes). Similarly, there are some uniquely downregulated genes (Fig. 3F, 104 genes) in ARMC5-overexpressing cells. These unique genes, which are the ones in the volcano plot in Figure 3A minus those in the volcano plot in Figure 3B, are illustrated in Figure 4A. The expression changes in these uniquely DE genes are caused by higher ARMC5 levels (compared to vector-transfected cells) in the presence of estrogen. The fold changes and their directions (up- or downregulation) are shown in Figure 4B. Some GO biological process terms of these uniquely up- and downregulated DE genes are presented in Figures 4C and 4D. The full datasets of the terms are presented in S-Tables 5 and 6. Specific biological processes relevant to breast cancers, such as peptide tyrosine phosphorylation, hexose transmembrane transport, mammary gland development, ERK1/2 cascade activation, and leukocyte activation and inflammation, are among the significant terms associated with these DE genes.\u0026nbsp;We channeled these unique DE genes to estrogen pathway-focused Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. The results are presented in S-Figure 1. Among the DE genes upregulated in ARMC5-overexpressing MCF7 cells after E2 stimulation, three of them, \u003cem\u003ei.e\u003c/em\u003e., HB-EGF, SHC, and KRT19, are in the estrogen signaling pathway. For DE genes downregulated in ARMC5-overexpressing MCF7 cells after E2 stimulation, none are in the E2 pathway, though (data not shown). These unique DE genes were also channeled to the breast cancer-focused KEGG pathway analysis, and the results are shown in S-Figure 2. Two genes (FGF and SHC) are found in this pathway.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eARMC5-E3 overexpression reduced the estrogen binding to the ERE\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo assess whether ARMC5-E3 regulates ERa\u0026nbsp;binding to ERE, we conducted ERa\u0026nbsp;ChIP-seq of ARMC5-overexpressing MCF-7 cells, which were cultured in the absence or presence of E2 for 45 minutes. Vector-transfected cells were used as controls. The metagene profiles in Figure 5A illustrate ERα binding to the ERE. The addition of E2 to the cell culture increased the ERa\u0026nbsp;binding to ERE in both the ARMC5-overexpressing cells and vector-transfected control cells (the upper row). Such an increase was less prominent in the former (left panel of the upper row) than in the latter (right panel of the upper row). The difference between the two can also be viewed in the left panel of the bottom row. On the other hand, in the absence of E2 stimulation, the levels of ERa\u0026nbsp;binding to ERE were similar in the vector-transfected control and ARMC5-overexpressing cells (the right panel of the bottom row). This is compatible with our finding of AMC5-E3 as an ERa-specific ubiquitin ligase, which is essential for ERa\u0026nbsp;degradation. The high level of ARMC5-E3 reduced ERα levels in the cytosol. As a consequence, after the E2 stimulation, fewer ERas enter the nuclei to bind to ERE.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eVolcano plots in Figures 5B and 5C show genes with significant differences (unadjusted p-values) and \u0026gt;2-fold changes in ERa\u0026nbsp;peak density in cells with or without 40-minute E2 stimulation. Figure 5B was derived from ARMC5-overexpressing cells, while Figure 5C was from vector-transfected cells. The ChIP-seq datasets are presented in S-Tables 7 and 8. The increased peak density of EREs in these volcano plots (EREs on the right side of the volcano plots) reflects more ERa\u0026nbsp;binding after the E2 stimulation in these cells. There are also some EREs with reduced peak density (EREs on the left side of the volcano plots). Such reduced density might be caused by a more complex and less understood indirect action of estrogen. We, therefore, are inclined to pay more attention to those EREs with increased peak density after the E2-stimulation. There are 34 such EREs in ARMC5-overexpressing cells (Fig. 5B) and 17 in the vector-transfected cells (Fig. 5C). Two EREs (THSD4 and LINC02038, marked in green in the volcano plots) are common ones between these two groups. If we subtract these two from the upregulated EREs in the ARMC5-overexpressing group, the remaining 32 will be the unique EREs caused by ARMC5 overexpression and E2 treatment. We have selected 4 of these 32 EREs and displayed their ERα peak density tracks (with E2 or vehicle treatment) in Figure 5D. It can readily appreciate the increased peak density centered on ERE (red oval dots) in the E2-treated samples compared to the vehicle-treated samples.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn the ARMC5-overexpressing cells after the E2 treatment, their ERa\u0026nbsp;peak density at EREs was lower than that of the vector-transfected cells, according to the metagene profile (Fig. 5A, left panel in the lower row). We retrieved some ERa\u0026nbsp;density tracks of these EREs and presented them in Figure 5E, showing that the peak density around these ERE sites in cells with ARMC5-overexpression was apparently lower than that of cells transfected with vectors, supporting the notion that the overexpressed ARMC5-E3 aggravates ERa\u0026nbsp;degradation, causing fewer ERa\u0026nbsp;available to enter the nuclei for ERE association after the E2 stimulation. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eARMC5 gene copy number dose-dependently prolonged the relapse-free survival time of breast cancer patients\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConsidering that ARMC5-E3 is the major E3 responsible for the degradation of ERa, which is essential in breast cancer pathogenesis and prognosis, we queried the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) Genome-Wide Association Study (GWAS) dataset\u003csup\u003e26,27\u003c/sup\u003e to assess how ARMC5 mutations affect breast cancer prognosis. Among 2173 cases in the dataset, we selected 1609 cases that were ER-positive in the tumor mass and had ARMC5 gene mutation information available. In these cases, the only available ARMC5 mutation information was copy number variation (CNV), while no deletions, insertions, or point mutations were reported. The 1609 cases were further classified by pathological classification. Those that could not be classified were grouped under the name of “other.” For the CNV, it was divided into groups of deep deletion (bi-allele deletion, \u003cem\u003ei.e\u003c/em\u003e., 0 copy; n=1), shallow deletion (heterozygous deletion, \u003cem\u003ei.e\u003c/em\u003e., 1 copy; n=36), diploid (normal copy number, \u003cem\u003ei.e\u003c/em\u003e., 2 copies; n=976), gain (\u003cem\u003ei.e\u003c/em\u003e., 3 copies; n=482), and amplification (\u003cem\u003ei.e\u003c/em\u003e., more than 3 copies; n=114). The percentages of cases with different CNVs within a pathological category are presented in Figure 6A. In most cases across different categories, ARMC5 copy number was normal (diploid). \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe ARMC5 gene copy numbers showed a positive correlation with relapse-free survival time in ER-positive patients (Fig. 6B). We conducted a statistical analysis of all comparisons between CNV groups with different copy numbers. As there was only one case in the deep deletion group, this group was excluded from the comparison. The relapse-free survival (RFS) time of the shallow deletion group was significantly shorter than that of the diploid, gain, and amplification groups. Patients in the diploid group have a significantly shorter RFS than those in the amplification group. The RFS probability of different groups at various time points during the follow-up period (varying from 275 to 300 months, depending on the group) is plotted in Figure 6C. Compared with the diploid group, the shallow deletion group had a significantly lower RFS probability, whereas the gain and amplification groups had a considerably higher RFS probability.\u003c/p\u003e\n\u003cp\u003eThese results suggest that a higher ARMC5 copy number is positively associated with a longer RFS time and increases RFS probability in ER-positive breast cancer patients.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we revealed that ARMC5 interacted with HSPA1A, and HSPA1A, in turn, interacted with its client protein ERα.\u0026nbsp;Both were substrates of the novel ARMC5-E3 ubiquitin ligase, which was essential for their degradation under physiological conditions. The ARMC5 deletion mutation caused the accumulation of HSPA1A and ERα. In breast cancer MCF7 cells, ARMC5 overexpression decreased ERα protein levels. It reduced ERα binding to EREs, accompanied by altered transcription of multiple genes in the estrogen signaling and breast cancer tumorigenesis pathways. A tumor mass GWAS analysis of 1609 ER-positive human patients revealed that ARMC5 gene copy number was positively associated with RFS time and the probability of these patients. Elaboration and discussion of some findings in this study are presented below.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eA new and large repertoire of ARMC5-E3 substrates\u003c/em\u003e. Our previous studies demonstrate that ARMC5 is the substrate-recognition subunit of multiple-subunit ubiquitin ligases specific for the largest subunit, RBP1, of RNA polymerase II (Pol II) under physiological conditions \u003csup\u003e31\u003c/sup\u003e. Our immunoprecipitation experiments reveal that this E3 also comprises the scaffold protein CUL3 and the E3 enzymatic subunit RBX1. Furthermore, ARMC5 binds to itself and forms a homodimer in this E3 complex\u003csup\u003e31\u003c/sup\u003e. Self-dimerization implies that each ARMC5-E3 complex may contain two active RBX1 subunits. Very surprisingly, this E3 is not only essential for the RBP1 degradation but also for the degradation of all other 11 subunits of Pol II, even though some of them are pretty far away from RPB1. Our yeast 2-hybrid results detect only the interaction between ARMC5 and RPB1, not any other Pol II subunits, suggesting that RPB1 is the only contact molecule with which ARMC5-E3 interacts with Pol II. How do the other 11 subunits get ubiquitinated if they do not interact with the substrate recognition subunit ARMC5? There are two possible models. A. After ARMC5 binds to RPB1, the enzymatic subunit RBX1 at the other end of the complex can move around and briefly contact other Pol II subunits, ubiquitinating them. B. Since ARMC5-E3 is a dimer linked by two ARMC5 molecules, with one ARMC5 binding to its substrate, RBX1 attached to the other ARMC5 can be quite a distance away and can wrap around the Pol II complex and ubiquitinate any subunits in its reach. This is probably also the case in this study. We detected an interaction between ARMC5 and HSPA1A, but not between ARMC5 and ERa. It is possible that ARMC-E3 directly interacts with and ubiquitinates HSPA1A. While binding HSPA1A, the enzyme-active RBX1 in one or the other arm of the dimeric ARMC5-E3 might wrap around the HSPA1A-ERa\u0026nbsp;complex and ubiquitinate a distant ERa (Fig. 2H). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn addition to ERa, HSPA1A has many other client proteins, such as p53\u003csup\u003e34,35\u003c/sup\u003e, CFTR\u003csup\u003e36,37\u003c/sup\u003e, Tau\u003csup\u003e38-40\u003c/sup\u003e, Huntintin\u003csup\u003e41,42\u003c/sup\u003e, Akt\u003csup\u003e43,44\u003c/sup\u003e, etc., whose functions are across a wide range of cellular processes. This raises an interesting question of whether those client proteins are also substrates of ARMC5-E3 and depend, wholly or partially, on this E3 for degradation. On this note, we need to mention that in our immunoprecipitation using ARMC5 as bait, we identified many other members of the HSP family, including HSP90B1, HSP90AB1, HSPA4L, HSPA9, HSPA5, HSPA2, HSPD1, and HSPA1L\u003csup\u003e31\u003c/sup\u003e. These HSPs share some sequence homology, at least in their conserved chaperone function domains\u003csup\u003e18,45,46\u003c/sup\u003e. The sequence homology among the members of the HSP70 subfamily is even higher\u003csup\u003e20,47\u003c/sup\u003e. This finding begs two additional questions: a. whether ARMC5 binds to these HSPs due to their sequence homology, b. whether the client proteins of these HSPs are also substrates of this E3, which controls their degradation by and large\u003csup\u003e48\u003c/sup\u003e. If so, this ARMC5-E3 will be one of the most critical E3s in cell biology. Our additional investigation to address these questions is in progress.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe current dogma is that ubiquitin ligases control the substrate specificity of the ubiquitination cascade. As 70% or more of the 20,000 proteins are degraded by the ubiquitin-proteasome pathway\u003csup\u003e49\u003c/sup\u003e, if E3 is very specific, one would expect to see a large number of E3s, probably in the range of more than ten thousand. However, only about 600 different E3s have been documented in humans\u003csup\u003e50\u003c/sup\u003e. Most of these 600 E3s are qualified based on substrate protein-level changes but lack more stringent proof from in vitro ubiquitination assays. How do we explain this vast discrepancy in the expected and discovered E3 numbers? One can, of course, argue that E3 lacks a conserved domain and is thus difficult to detect; many remain to be found. However, our finding related to ARMC5-E3 points to a different conclusion. This E3 has been shown to control the degradation of more than 15 proteins\u003csup\u003e31,51-54\u003c/sup\u003e. If it functions as an E3 for all the HSPs it associates with, and each of these HSPs has a dozen or so client proteins, which this E3 will also act on, then it can channel more than a hundred proteins to the proteasome. The puzzle of a paucity of discovered E3s to ubiquitinate 15,000 proteins can be solved because if each E3 can ubiquitinate 100 proteins, only 150 E3s will be required. This number is within the order of discovered E3. Based on our findings, we argue that the current dogma of E3 specificity needs to be revised: there is a degree of specificity in E3 binding to specific proteins. However, E3 will ubiquitinate anything in the vicinity of its binding partner. A large number of proteins exist as multi-component complexes, either transiently or persistently; if an E3 binds to one of the components, it will ubiquitinate any component in the complex, significantly increasing its substrate pool. Hence, E3 specificity is determined by how close a protein is to the E3, with no specific sequence or tertiary-structure requirements. In that sense, E3s are loosely specific, with one E3 capable of acting on dozens or even hundreds of proteins. This loose E3 specificity has been explored in “Pro-Tac,” which uses a linker to deliver an E3 (pVHL and CRBN are popular choices) to any target protein for ubiquitination and degradation in biotech and therapeutic applications (51-53).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eARMC5-E3 maintains ER\u003c/em\u003e\u003cem\u003ea\u003c/em\u003e\u003cem\u003e\u0026nbsp;homeostasis\u003c/em\u003e. Estrogen has profound roles in the reproductive system\u003csup\u003e55,56\u003c/sup\u003e, bone metabolism\u003csup\u003e57,58\u003c/sup\u003e, cardiovascular function\u003csup\u003e59,60\u003c/sup\u003e, cognitive function\u003csup\u003e61,62\u003c/sup\u003e, and mood regulation\u003csup\u003e63\u003c/sup\u003e, in addition to its function in breast tumor oncogenesis\u003csup\u003e3,64-66\u003c/sup\u003e. Its protein level homeostasis is thus of vital importance. Figure 7 illustrates a model based on the results of the current study about how ARMC5-E3 controls the homeostasis of the ERa.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn WT cells (Fig. 7A), inactive ERa\u0026nbsp;is associated with HSPA1A in the cytosol (left panel). ARMC5-E3 maintains the homeostasis of the level of these HSPA1A/ERa\u0026nbsp;complexes by ubiquitinating some HSPA1A and ERa, and channeling them to the proteasome for degradation. When estrogen, such as E2, binds to ERa, ERa\u0026nbsp;dissociates from HSPA1A (right panel). At this stage, ARMC5-E3 can no longer reach and degrade ERa, because the latter is pulled away from HSPA1A. The escaped ERa\u0026nbsp;forms a complex with the E2 , enters the nucleus, and binds to estrogen-responsive elements (EREs). The ERa\u0026nbsp;and E2 complex then plays its well-known role as a transcription factor, modulating the transcription of genes containing the EREs\u003csup\u003e67-70\u003c/sup\u003e. In cells with ARMC5 overexpression (Fig. 7B), in the absence of E2 (left panel), the increased level of ARMC5-E3 causes more degradation of its substrates, HSPA1A and ERa\u0026nbsp;, resulting in fewer HSPA1A-ERa\u0026nbsp;complexes. When E2 is presented to these ARMC5-overexpressing cells, due to the reduced availability of ERa-HSPA1A complexes, E2 can only find fewer ERas. Consequently, fewer E2-ERa\u0026nbsp;complexes are formed and enter the nucleus to modulate genes with ERE. The net effect is that E2 becomes less potent in these cells. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe observed that ARMC5 KO did not alter ERa\u0026nbsp;levels in the ovary and uterus. How to explain such exceptions? These two organs produce estrogen\u003csup\u003e71\u003c/sup\u003e, and the latter even has specific E2-binding proteins to trap more E2 from the blood. Therefore, the local estrogen levels in these two organs are much higher than in other tissues, and most ERas are probably in the activated form and are not in association with HSPA1A. Hence, they are not subjected, or at least are less subjected to ARMC5-E3-mediated degradation. There are three other known specific E3 for ERa, \u003cem\u003ei.e\u003c/em\u003e., E6AP, CHIP, and SPOP, as reviewed in the introduction\u003csup\u003e15,16\u003c/sup\u003e. It is possible that one or several of these additional E3s are responsible for the degradation of ERa\u0026nbsp;that are not associated with HSPA1A. ERa\u0026nbsp;has a different association with chaperons and has different intracellular locations. It can associate with HSPA1A but can also dissociate from HSPA1A upon estrogen binding in the cytosol. While HSPA1A-associated ERa\u0026nbsp;depends on ARMC5-E3 for its degradation, ERa\u0026nbsp;complexed with other molecules (\u003cem\u003ee.g\u003c/em\u003e., estrogen) or at different locations (\u003cem\u003ee.g\u003c/em\u003e., nucleic) might use different E3s, hence the need for multiple E3s for ERa. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOur ERa\u0026nbsp;ChIP-seq identified many EREs (36 uniquely upregulated peaks in ARMC5-overexpressing MCF7 cells) with FC\u0026gt;2 and nominal \u003cem\u003ep\u003c/em\u003e-value \u0026lt;0.05 in peak density upon E2 stimulation. It should be noted that their \u003cem\u003ep\u003c/em\u003e-value significance did not survive the multiple-testing adjustment. Several reasons might contribute to a lack of significantly low adjusted \u003cem\u003ep\u003c/em\u003e-values. First, ChIP-seq was performed in cells transiently overexpressing ARMC5. The transfection efficiency was about 25-30%. That translates into a high background noise at about 70%. Secondly, the cells were harvested within 45 minutes of E2 stimulation. These factors might significantly reduce the sensitivity of our statistical analysis. We used a loose criterion of a nominal \u003cem\u003ep\u003c/em\u003e-value of 0.05 to construct the volcano plots in Figures 5B and 5C. It is therefore possible that this group contains some false-positive EREs. With that said, the ERE peak track display, we can still easily identify EREs with increased peak density after E2 stimulation in the ARMC5-overexpressing cells (Fig. 5D), and EREs with reduced peak density in the ARMC5-overexpressing cells compared to vector-transfected cells (Fig. 5D), indicating the meaningfulness and usefulness of EREs with nominal \u003cem\u003ep\u003c/em\u003e-values of \u0026lt;0.05.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eThe impact of ARMC5 on the estrogen signaling pathways and breast cancer\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo understand how ARMC5 regulates estrogen signaling pathways and impacts breast cancer oncogenesis, we conducted RNA-seq on ARMC5-overexpressing MCF7 breast cancer cells treated with E2 during the last 2 hours of culture. Among the 247 genes (143 upregulated and 104 downregulated) that were uniquely DE in the ARMC5-overexpressing cells with E2 treatment, according to GO analysis, 2 genes (AREG and FOXF1) are related to mammary gland alveolus and lobule development, 8 genes (IRS2, SOX3, AREG, OTP, FOXF1, INSM1, RCBTB2, and PCSK9) are relevant to gland development, 7 genes (CDH5, ZEB1, INHBA, HPGD, BMP6, CITED1, and BMPER) are related to cell surface serine/threonine receptor signaling, 7 genes (DUSP6, INHBA, CALCR, FGF19, BMPER, RGS14, and TLR9) are involved in the ERK1/ERK2 signaling pathway regulation, \u0026nbsp;7 genes (FGR, IL12A, FOXF1, CALHM6, CR2, KCNJ8, and TLR9 ) are implicated in leukocyte mediated immune and inflammatory responses. According to KEGG analysis, 3 of these genes (HB-EGF, SHC, and KRT19) are in the estrogen signaling pathway, and 2 (FGF and SHC) are in the breast cancer pathways.\u0026nbsp;The genes in these GO terms and KEGG pathways are highly relevant to the breast cancer ontogenesis and prognosis. They are a trove of interesting candidates that need to be further investigated for their possible roles in prolonging the TFS time and rate in breast cancer patients.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eARMC5-E3’s role in cancer\u003c/em\u003e. Our \u003cem\u003ein silico\u003c/em\u003e search of GWAS datasets reveals that high ARMC5 CNV favors a longer RFS time and higher RFS probability in human ER-positive breast cancer patients. These patients are already on estrogen antagonist therapy and have relatively lower tissue E2 levels. Such a low E2 level creates a favorable environment for ARMC5-E3 to degrade inactive HSPA1A-associated ERa\u0026nbsp;and reduce steady-state ERa\u0026nbsp;levels. Such a reduction, in turn, reduces the tumorigenic effect of the estrogen and promotes the efficacy of estrogen antagonist treatment. This is probably one of the mechanisms by which patients with high ARMC5 CNV have better RFS. Of course, the broad effect of ARMC5-E3 across a large number of substrates can have additional benefits for patient survival. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe huge proven and putative target protein repertoire of ARMC5-E3, including ERa\u0026nbsp;and the transcription machinery, is a testament to the vital function of ARMC5. Such vital functions explain why the biallelic deletion of ARMC5 causes 60% perinatal lethality in mice with a 50% CD1 genetic background\u003csup\u003e72\u003c/sup\u003e. The surviving KO mice are dwarfs, whereas heterozygous ARMC5 KO mice have no obvious phenotype \u003csup\u003e73\u003c/sup\u003e. Humans with ARMC5 gene deletion probably have similar phenotypes. We are unlikely to find adult patients with biallelic germline ARMC5 deletions or detrimental biallelic ins/dels, because individuals with such mutations are eliminated early in life.\u003c/p\u003e\n\u003cp\u003eIn contrast, the germline ins/dels found in surviving individuals are probably inconsequential. Deleterious somatic biallelic ARMC5 deletion might be detectable in some cells, but they are under very unfavorable conditions to compete with cells with WT ARMC5, whether these cells are normal or malignant. Therefore, the cells or tissue with such a biallelic somatic dels/ins must be rare. Indeed, this is demonstrated in the breast cancer GWAS presented in Figure 6B. Only 1 of 1609 tumor masses with CNV information available harbors a biallelic ARMC5 deletion, which is likely somatic rather than germline. Based on these analyses, we believe it will be more fruitful to study human cohorts with high genomic ARMC5 CNVs rather than ins/dels for any unknown phenotypes related to ARMC5 mutations.\u003c/p\u003e\n\u003cp\u003eDoes the beneficial effect of high \u003cem\u003eARMC5\u003c/em\u003e CNV qualify it as a tumor suppressor gene? There is no universally agreed-upon definition of a tumor suppressor gene, but, generally speaking, to qualify as a tumor suppressor gene, it should increase tumor risk across multiple tissue types. Except for adrenal hypertrophy, a lack of increased tumor incidence of any organs, including the breasts, in thousands of \u003cem\u003eArmc5\u003c/em\u003e KO mice we generated is a testament that ARMC5 is not a \u003cem\u003ebona fide\u003c/em\u003e tumor suppressor gene. ARMC5 has profound roles in various cellular processes due to its large number of proven and putative substrates. Its functions are complex and context-dependent. Its roles, exemplified by its function as an E3 for all Pol II subunits, are essential for homeostasis and promote cell growth. Such functions of ARMC5 are the exact opposite of the criteria for a tumor suppressor. In the rare patient tissue sample with an ARMC5 biallelic deletion, as mentioned above, the deletion is likely somatic, and the tumor might harbor mutations in other genes to compensate for the deleterious effects of the ARMC5 deletion, allowing the tumor cells to survive. This patient case is a rare exception rather than the norm. Therefore, calling \u003cem\u003eARMC5\u003c/em\u003e a tumor suppressor gene is a misnomer.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn summary, we identified ARMC5-E3’s two new substrates, HSPA1A and ERα. ARMC5-E3 degrades inactivated HSPA1A-bound ERa. Increased ARMC5 expression reduced steady-state ERa\u0026nbsp;levels. Such an effect probably contributes to a longer relapse-free survival time for breast cancer patients. Hence, ARMC5 CNV is a useful prognostic marker for breast cancer. \u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the Jean-Louis Levesque Foundation to J.W. It was also funded in part by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC; RGPIN-2017-04790), the Arthritis Society of Canada, the Canadian Institutes of Health Research (PJT-180284), and the Canadian Rare Disease Models and Mechanisms Network to J.W. Sequencing library construction and Illumina sequencing were performed by the Clinical Research Institute of Montreal Genome Platform and the Genome Quebec Innovation Centre, Montreal. The authors thank the professionals on these platforms for providing these essential services.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe RNA-seq and ChIP-seq datasets have been deposited in the Gene Expression Omnibus at NCBI under accession numbers GSE312622 (token: urkdimiuzlktncj) and GSE312623 (token: gdshiggcdhmlzcd), respectively.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contribution\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eX.H. and J.P. conducted experiments. X.H. and Y.H. performed data analysis. X.H., Y.H., and J.W. wrote the manuscript. H.L. and J. W. initiated, designed, and supervised the project\u003csup\u003e.\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting financial interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing financial interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBentzon, N., D\u0026uuml;ring, M., Rasmussen, B. B., Mouridsen, H. \u0026amp; Kroman, N. Prognostic effect of estrogen receptor status across age in primary breast cancer. \u003cem\u003eInt J Cancer\u003c/em\u003e 122, 1089\u0026ndash;1094 (2008). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ijc.22892\u003c/span\u003e\u003cspan address=\"10.1002/ijc.22892\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJayasekara, H. \u003cem\u003eet al.\u003c/em\u003e Mortality after breast cancer as a function of time since diagnosis by estrogen receptor status and age at diagnosis. \u003cem\u003eInt J Cancer\u003c/em\u003e 145, 3207\u0026ndash;3217 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ijc.32214\u003c/span\u003e\u003cspan address=\"10.1002/ijc.32214\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, K.-D., Wu, J., Shen, Z.-Z. \u0026amp; Shao, Z.-M. Hazard of Breast Cancer-Specific Mortality among Women with Estrogen Receptor-Positive Breast Cancer after Five Years from Diagnosis: Implication for Extended Endocrine Therapy. \u003cem\u003eThe Journal of Clinical Endocrinology \u0026amp; Metabolism\u003c/em\u003e 97, E2201\u0026ndash;E2209 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1210/jc.2012-2423%J\u003c/span\u003e\u003cspan address=\"10.1210/jc.2012-2423%J\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e The Journal of Clinical Endocrinology \u0026amp; Metabolism\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong, D. \u003cem\u003eet al.\u003c/em\u003e ERα and ERβ Homodimers in the Same Cellular Context Regulate Distinct Transcriptomes and Functions. \u003cem\u003eFront Endocrinol (Lausanne)\u003c/em\u003e 13, 930227 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fendo.2022.930227\u003c/span\u003e\u003cspan address=\"10.3389/fendo.2022.930227\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKumar, R. \u003cem\u003eet al.\u003c/em\u003e The dynamic structure of the estrogen receptor. \u003cem\u003eJ Amino Acids\u003c/em\u003e 2011, 812540 (2011). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4061/2011/812540\u003c/span\u003e\u003cspan address=\"10.4061/2011/812540\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTamrazi, A., Carlson, K. E., Daniels, J. R., Hurth, K. M. \u0026amp; Katzenellenbogen, J. A. Estrogen Receptor Dimerization: Ligand Binding Regulates Dimer Affinity and DimerDissociation Rate. \u003cem\u003eMolecular Endocrinology\u003c/em\u003e 16, 2706\u0026ndash;2719 (2002). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1210/me.2002-0250%J\u003c/span\u003e\u003cspan address=\"10.1210/me.2002-0250%J\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Molecular Endocrinology\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIrani, S. \u003cem\u003eet al.\u003c/em\u003e Somatic estrogen receptor α mutations that induce dimerization promote receptor activity and breast cancer proliferation. \u003cem\u003eJ Clin Invest\u003c/em\u003e 134 (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1172/jci163242\u003c/span\u003e\u003cspan address=\"10.1172/jci163242\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeis, K. E., Ekena, K., Thomas, J. A., Lazennec, G. \u0026amp; Katzenellenbogen, B. S. Constitutively active human estrogen receptors containing amino acid substitutions for tyrosine 537 in the receptor protein. \u003cem\u003eMol Endocrinol\u003c/em\u003e 10, 1388\u0026ndash;1398 (1996). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1210/mend.10.11.8923465\u003c/span\u003e\u003cspan address=\"10.1210/mend.10.11.8923465\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDhamad, A. E., Zhou, Z., Zhou, J. \u0026amp; Du, Y. Systematic Proteomic Identification of the Heat Shock Proteins (Hsp) that Interact with Estrogen Receptor Alpha (ERα) and Biochemical Characterization of the ERα-Hsp70 Interaction. \u003cem\u003ePLoS One\u003c/em\u003e 11, e0160312 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0160312\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0160312\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePicard, D. Chaperoning steroid hormone action. \u003cem\u003eTrends Endocrinol Metab\u003c/em\u003e 17, 229\u0026ndash;235 (2006). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.tem.2006.06.003\u003c/span\u003e\u003cspan address=\"10.1016/j.tem.2006.06.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSilveira, M. A. D. \u003cem\u003eet al.\u003c/em\u003e HSP70 mediates a crosstalk between the estrogen and the heat shock response pathways. \u003cem\u003eJ Biol Chem\u003c/em\u003e 299, 102872 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jbc.2023.102872\u003c/span\u003e\u003cspan address=\"10.1016/j.jbc.2023.102872\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou, W. \u0026amp; Slingerland, J. M. Links between oestrogen receptor activation and proteolysis: relevance to hormone-regulated cancer therapy. \u003cem\u003eNat Rev Cancer\u003c/em\u003e 14, 26\u0026ndash;38 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nrc3622\u003c/span\u003e\u003cspan address=\"10.1038/nrc3622\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHelzer, K. T., Hooper, C., Miyamoto, S. \u0026amp; Alarid, E. T. Ubiquitylation of nuclear receptors: new linkages and therapeutic implications. \u003cem\u003eJ Mol Endocrinol\u003c/em\u003e 54, R151\u0026ndash;167 (2015). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1530/jme-14-0308\u003c/span\u003e\u003cspan address=\"10.1530/jme-14-0308\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, L., Li, Z., Howley, P. M. \u0026amp; Sacks, D. B. E6AP and calmodulin reciprocally regulate estrogen receptor stability. \u003cem\u003eJ Biol Chem\u003c/em\u003e 281, 1978\u0026ndash;1985 (2006). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1074/jbc.M508545200\u003c/span\u003e\u003cspan address=\"10.1074/jbc.M508545200\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFan, M., Park, A. \u0026amp; Nephew, K. P. CHIP (carboxyl terminus of Hsc70-interacting protein) promotes basal and geldanamycin-induced degradation of estrogen receptor-alpha. \u003cem\u003eMol Endocrinol\u003c/em\u003e 19, 2901\u0026ndash;2914 (2005). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1210/me.2005-0111\u003c/span\u003e\u003cspan address=\"10.1210/me.2005-0111\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eByun, B. \u0026amp; Jung, Y. Repression of transcriptional activity of estrogen receptor alpha by a Cullin3/SPOP ubiquitin E3 ligase complex. \u003cem\u003eMol Cells\u003c/em\u003e 25, 289\u0026ndash;293 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMilner, C. M. \u0026amp; Campbell, R. D. Structure and expression of the three MHC-linked HSP70 genes. \u003cem\u003eImmunogenetics\u003c/em\u003e 32, 242\u0026ndash;251 (1990). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/bf00187095\u003c/span\u003e\u003cspan address=\"10.1007/bf00187095\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSerlidaki, D. \u003cem\u003eet al.\u003c/em\u003e Functional diversity between HSP70 paralogs caused by variable interactions with specific co-chaperones. \u003cem\u003eJournal of Biological Chemistry\u003c/em\u003e 295, 7301\u0026ndash;7316 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1074/jbc.RA119.012449\u003c/span\u003e\u003cspan address=\"10.1074/jbc.RA119.012449\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrocchieri, L., Conway de Macario, E. \u0026amp; Macario, A. J. hsp70 genes in the human genome: Conservation and differentiation patterns predict a wide array of overlapping and specialized functions. \u003cem\u003eBMC Evol Biol\u003c/em\u003e 8, 19 (2008). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/1471-2148-8-19\u003c/span\u003e\u003cspan address=\"10.1186/1471-2148-8-19\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDaugaard, M., Rohde, M. \u0026amp; J\u0026auml;\u0026auml;ttel\u0026auml;, M. The heat shock protein 70 family: Highly homologous proteins with overlapping and distinct functions. \u003cem\u003eFEBS Lett\u003c/em\u003e 581, 3702\u0026ndash;3710 (2007). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.febslet.2007.05.039\u003c/span\u003e\u003cspan address=\"10.1016/j.febslet.2007.05.039\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQuintana-Gallardo, L. \u003cem\u003eet al.\u003c/em\u003e The cochaperone CHIP marks Hsp70- and Hsp90-bound substrates for degradation through a very flexible mechanism. \u003cem\u003eScientific Reports\u003c/em\u003e 9, 5102 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-019-41060-0\u003c/span\u003e\u003cspan address=\"10.1038/s41598-019-41060-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcDonough, H. \u0026amp; Patterson, C. CHIP: a link between the chaperone and proteasome systems. \u003cem\u003eCell Stress Chaperones\u003c/em\u003e 8, 303\u0026ndash;308 (2003). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1379/1466-1268(2003)008%3C0303:calbtc%3E2.0.co;2\u003c/span\u003e\u003cspan address=\"10.1379/1466-1268(2003)008%3C0303:calbtc%3E2.0.co;2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaushik, S. \u0026amp; Cuervo, A. M. The coming of age of chaperone-mediated autophagy. \u003cem\u003eNature Reviews Molecular Cell Biology\u003c/em\u003e 19, 365\u0026ndash;381 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41580-018-0001-6\u003c/span\u003e\u003cspan address=\"10.1038/s41580-018-0001-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParzych, K. R. \u0026amp; Klionsky, D. J. An overview of autophagy: morphology, mechanism, and regulation. \u003cem\u003eAntioxid Redox Signal\u003c/em\u003e 20, 460\u0026ndash;473 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1089/ars.2013.5371\u003c/span\u003e\u003cspan address=\"10.1089/ars.2013.5371\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCommunity, G. The Galaxy platform for accessible, reproducible, and collaborative data analyses: 2024 update. \u003cem\u003eNucleic Acids Res\u003c/em\u003e 52, W83\u0026ndash;w94 (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/nar/gkae410\u003c/span\u003e\u003cspan address=\"10.1093/nar/gkae410\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePereira, B. \u003cem\u003eet al.\u003c/em\u003e The somatic mutation profiles of 2,433 breast cancers refine their genomic and transcriptomic landscapes. \u003cem\u003eNature Communications\u003c/em\u003e 7, 11479 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/ncomms11479\u003c/span\u003e\u003cspan address=\"10.1038/ncomms11479\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCurtis, C. \u003cem\u003eet al.\u003c/em\u003e The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. \u003cem\u003eNature\u003c/em\u003e 486, 346\u0026ndash;352 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nature10983\u003c/span\u003e\u003cspan address=\"10.1038/nature10983\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCerami, E. \u003cem\u003eet al.\u003c/em\u003e The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. \u003cem\u003eCancer Discov\u003c/em\u003e 2, 401\u0026ndash;404 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1158/2159-8290.Cd-12-0095\u003c/span\u003e\u003cspan address=\"10.1158/2159-8290.Cd-12-0095\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGao, J. \u003cem\u003eet al.\u003c/em\u003e Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. \u003cem\u003eSci Signal\u003c/em\u003e 6, pl1 (2013). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/scisignal.2004088\u003c/span\u003e\u003cspan address=\"10.1126/scisignal.2004088\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Bruijn, I. \u003cem\u003eet al.\u003c/em\u003e Analysis and Visualization of Longitudinal Genomic and Clinical Data from the AACR Project GENIE Biopharma Collaborative in cBioPortal. \u003cem\u003eCancer Res\u003c/em\u003e 83, 3861\u0026ndash;3867 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1158/0008-5472.Can-23-0816\u003c/span\u003e\u003cspan address=\"10.1158/0008-5472.Can-23-0816\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLao, L. \u003cem\u003eet al.\u003c/em\u003e ARMC5 is part of an RPB1-specific ubiquitin ligase implicated in adrenal hyperplasia. \u003cem\u003eNucleic Acids Research\u003c/em\u003e 50, 6343\u0026ndash;6367 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/nar/gkac483%J\u003c/span\u003e\u003cspan address=\"10.1093/nar/gkac483%J\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Nucleic Acids Research\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTetsuo, M., Eriksson, H. \u0026amp; Sj\u0026ouml;vall, J. Gas chromatographic-mass spectrometric analysis of endogenous levels of estradiol in plasma and in cytosol from rat uterus. \u003cem\u003eJ Chromatogr\u003c/em\u003e 239, 287\u0026ndash;300 (1982). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/s0021-9673(00)81989-5\u003c/span\u003e\u003cspan address=\"10.1016/s0021-9673(00)81989-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiordano, M. V. \u003cem\u003eet al.\u003c/em\u003e How long is the ovary relevant for synthesis of steroids after menopause? \u003cem\u003eGynecol Endocrinol\u003c/em\u003e 34, 536\u0026ndash;539 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/09513590.2017.1417375\u003c/span\u003e\u003cspan address=\"10.1080/09513590.2017.1417375\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZylicz, M., King, F. W. \u0026amp; Wawrzynow, A. Hsp70 interactions with the p53 tumour suppressor protein. 20, 4634\u0026ndash;4638 (2001). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/https://doi.org/10.1093/emboj/20.17.4634\u003c/span\u003e\u003cspan address=\"10.1093/emboj/20.17.4634\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoysen, M., Kityk, R. \u0026amp; Mayer, M. P. Hsp70- and Hsp90-Mediated Regulation of the Conformation of p53 DNA Binding Domain and p53 Cancer Variants. \u003cem\u003eMol Cell\u003c/em\u003e 74, 831\u0026ndash;843.e834 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.molcel.2019.03.032\u003c/span\u003e\u003cspan address=\"10.1016/j.molcel.2019.03.032\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBagdany, M. \u003cem\u003eet al.\u003c/em\u003e Chaperones rescue the energetic landscape of mutant CFTR at single molecule and in cell. \u003cem\u003eNature Communications\u003c/em\u003e 8, 398 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-017-00444-4\u003c/span\u003e\u003cspan address=\"10.1038/s41467-017-00444-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBergbower, E., Boinot, C., Sabirzhanova, I., Guggino, W. \u0026amp; Cebotaru, L. The CFTR-Associated Ligand Arrests the Trafficking of the Mutant ∆F508 CFTR Channel in the ER Contributing to Cystic Fibrosis. \u003cem\u003eCell Physiol Biochem\u003c/em\u003e 45, 639\u0026ndash;655 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1159/000487120\u003c/span\u003e\u003cspan address=\"10.1159/000487120\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVoss, K., Combs, B., Patterson, K. R., Binder, L. I. \u0026amp; Gamblin, T. C. Hsp70 alters tau function and aggregation in an isoform specific manner. \u003cem\u003eBiochemistry\u003c/em\u003e 51, 888\u0026ndash;898 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1021/bi2018078\u003c/span\u003e\u003cspan address=\"10.1021/bi2018078\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRyder, B. D., Wydorski, P. M., Hou, Z. \u0026amp; Joachimiak, L. A. Chaperoning shape-shifting tau in disease. \u003cem\u003eTrends in Biochemical Sciences\u003c/em\u003e 47, 301\u0026ndash;313 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.tibs.2021.12.009\u003c/span\u003e\u003cspan address=\"10.1016/j.tibs.2021.12.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoll, A. \u003cem\u003eet al.\u003c/em\u003e Hsp multichaperone complex buffers pathologically modified Tau. \u003cem\u003eNature Communications\u003c/em\u003e 13, 3668 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-022-31396-z\u003c/span\u003e\u003cspan address=\"10.1038/s41467-022-31396-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrmsby, A. R., Ramdzan, Y. M., Mok, Y. F., Jovanoski, K. D. \u0026amp; Hatters, D. M. A platform to view huntingtin exon 1 aggregation flux in the cell reveals divergent influences from chaperones hsp40 and hsp70. \u003cem\u003eJ Biol Chem\u003c/em\u003e 288, 37192\u0026ndash;37203 (2013). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1074/jbc.M113.486944\u003c/span\u003e\u003cspan address=\"10.1074/jbc.M113.486944\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRozales, K. \u003cem\u003eet al.\u003c/em\u003e Differential roles for DNAJ isoforms in HTT-polyQ and FUS aggregation modulation revealed by chaperone screens. \u003cem\u003eNature Communications\u003c/em\u003e 13, 516 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-022-27982-w\u003c/span\u003e\u003cspan address=\"10.1038/s41467-022-27982-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoren, J. \u003cem\u003eet al.\u003c/em\u003e Facilitating Akt Clearance via Manipulation of Hsp70 Activity and Levels*. \u003cem\u003eJournal of Biological Chemistry\u003c/em\u003e 285, 2498\u0026ndash;2505 (2010). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org\u003c/span\u003e\u003cspan address=\"https://doi.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e/https://doi.org/10.1074/jbc.M109.057208\u003c/span\u003e\u003cspan address=\"/10.1074/jbc.M109.057208\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, W. \u003cem\u003eet al.\u003c/em\u003e PER2 interaction with HSP70 promotes cuproptosis in oral squamous carcinoma cells by decreasing AKT stability. \u003cem\u003eCell Death \u0026amp; Disease\u003c/em\u003e 16, 192 (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41419-025-07523-1\u003c/span\u003e\u003cspan address=\"10.1038/s41419-025-07523-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRosenzweig, R., Nillegoda, N. B., Mayer, M. P. \u0026amp; Bukau, B. The Hsp70 chaperone network. \u003cem\u003eNature Reviews Molecular Cell Biology\u003c/em\u003e 20, 665\u0026ndash;680 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41580-019-0133-3\u003c/span\u003e\u003cspan address=\"10.1038/s41580-019-0133-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZuiderweg, E. R., Hightower, L. E. \u0026amp; Gestwicki, J. E. The remarkable multivalency of the Hsp70 chaperones. \u003cem\u003eCell Stress Chaperones\u003c/em\u003e 22, 173\u0026ndash;189 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s12192-017-0776-y\u003c/span\u003e\u003cspan address=\"10.1007/s12192-017-0776-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaugeri, N., Radhakrishnan, J. \u0026amp; Knight, J. C. Genetic determinants of HSP70 gene expression following heat shock. \u003cem\u003eHum Mol Genet\u003c/em\u003e 19, 4939\u0026ndash;4947 (2010). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/hmg/ddq418\u003c/span\u003e\u003cspan address=\"10.1093/hmg/ddq418\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAgopiantz, M., Carnot, M., Denis, C., Martin, E. \u0026amp; Gauchotte, G. Hormone Receptor Expression in Meningiomas: A Systematic Review. \u003cem\u003eCancers (Basel)\u003c/em\u003e 15 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/cancers15030980\u003c/span\u003e\u003cspan address=\"10.3390/cancers15030980\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVoges, D., Zwickl, P. \u0026amp; Baumeister, W. The 26S proteasome: a molecular machine designed for controlled proteolysis. \u003cem\u003eAnnu Rev Biochem\u003c/em\u003e 68, 1015\u0026ndash;1068 (1999). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1146/annurev.biochem.68.1.1015\u003c/span\u003e\u003cspan address=\"10.1146/annurev.biochem.68.1.1015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, W. \u003cem\u003eet al.\u003c/em\u003e Genome-wide and functional annotation of human E3 ubiquitin ligases identifies MULAN, a mitochondrial E3 that regulates the organelle's dynamics and signaling. \u003cem\u003ePLoS One\u003c/em\u003e 3, e1487 (2008). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0001487\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0001487\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAoi, Y. \u003cem\u003eet al.\u003c/em\u003e SPT5 regulates RNA polymerase II stability via Cullin 3\u0026ndash;ARMC5 recognition. 11, eadt5885 (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/doi:10.1126/sciadv.adt5885\u003c/span\u003e\u003cspan address=\"doi:10.1126/sciadv.adt5885\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo, H. \u003cem\u003eet al.\u003c/em\u003e ARMC5 controls the degradation of most Pol II subunits, and ARMC5 mutation increases neural tube defect risks in mice and humans. \u003cem\u003eGenome Biology\u003c/em\u003e 25, 19 (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13059-023-03147-w\u003c/span\u003e\u003cspan address=\"10.1186/s13059-023-03147-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkuno, Y., Fukuhara, A., Otsuki, M. \u0026amp; Shimomura, I. ARMC5-CUL3 E3 ligase targets full-length SREBF in adrenocortical tumors. \u003cem\u003eJCI Insight\u003c/em\u003e 7 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1172/jci.insight.151390\u003c/span\u003e\u003cspan address=\"10.1172/jci.insight.151390\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUota, A. \u003cem\u003eet al.\u003c/em\u003e ARMC5 selectively degrades SCAP-free SREBF1 and is essential for fatty acid desaturation in adipocytes. \u003cem\u003eJournal of Biological Chemistry\u003c/em\u003e 300 (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jbc.2024.107953\u003c/span\u003e\u003cspan address=\"10.1016/j.jbc.2024.107953\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChantalat, E. \u003cem\u003eet al.\u003c/em\u003e Estrogen Receptors and Endometriosis. \u003cem\u003eInt J Mol Sci\u003c/em\u003e 21 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ijms21082815\u003c/span\u003e\u003cspan address=\"10.3390/ijms21082815\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFuentes, N. \u0026amp; Silveyra, P. Estrogen receptor signaling mechanisms. \u003cem\u003eAdv Protein Chem Struct Biol\u003c/em\u003e 116, 135\u0026ndash;170 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/bs.apcsb.2019.01.001\u003c/span\u003e\u003cspan address=\"10.1016/bs.apcsb.2019.01.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJilka, R. L. \u003cem\u003eet al.\u003c/em\u003e Increased osteoclast development after estrogen loss: mediation by interleukin-6. \u003cem\u003eScience\u003c/em\u003e 257, 88\u0026ndash;91 (1992). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.1621100\u003c/span\u003e\u003cspan address=\"10.1126/science.1621100\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePantschenko, A. G. \u003cem\u003eet al.\u003c/em\u003e Effect of osteoblast-targeted expression of bcl-2 in bone: differential response in male and female mice. \u003cem\u003eJ Bone Miner Res\u003c/em\u003e 20, 1414\u0026ndash;1429 (2005). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1359/jbmr.050315\u003c/span\u003e\u003cspan address=\"10.1359/jbmr.050315\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIorga, A. \u003cem\u003eet al.\u003c/em\u003e The protective role of estrogen and estrogen receptors in cardiovascular disease and the controversial use of estrogen therapy. \u003cem\u003eBiol Sex Differ\u003c/em\u003e 8, 33 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13293-017-0152-8\u003c/span\u003e\u003cspan address=\"10.1186/s13293-017-0152-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAryan, L. \u003cem\u003eet al.\u003c/em\u003e The Role of Estrogen Receptors in Cardiovascular Disease. \u003cem\u003eInt J Mol Sci\u003c/em\u003e 21 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ijms21124314\u003c/span\u003e\u003cspan address=\"10.3390/ijms21124314\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBustamante-Barrientos, F. A. \u003cem\u003eet al.\u003c/em\u003e The Impact of Estrogen and Estrogen-Like Molecules in Neurogenesis and Neurodegeneration: Beneficial or Harmful? \u003cem\u003eFront Cell Neurosci\u003c/em\u003e 15, 636176 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fncel.2021.636176\u003c/span\u003e\u003cspan address=\"10.3389/fncel.2021.636176\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYaffe, K., Haan, M., Byers, A., Tangen, C. \u0026amp; Kuller, L. Estrogen use, APOE, and cognitive decline: evidence of gene-environment interaction. \u003cem\u003eNeurology\u003c/em\u003e 54, 1949\u0026ndash;1954 (2000). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1212/wnl.54.10.1949\u003c/span\u003e\u003cspan address=\"10.1212/wnl.54.10.1949\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlbert, K. M. \u0026amp; Newhouse, P. A. Estrogen, Stress, and Depression: Cognitive and Biological Interactions. \u003cem\u003eAnnu Rev Clin Psychol\u003c/em\u003e 15, 399\u0026ndash;423 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1146/annurev-clinpsy-050718-095557\u003c/span\u003e\u003cspan address=\"10.1146/annurev-clinpsy-050718-095557\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, Z. \u003cem\u003eet al.\u003c/em\u003e Identification, cloning, and expression of human estrogen receptor-alpha36, a novel variant of human estrogen receptor-alpha66. \u003cem\u003eBiochem Biophys Res Commun\u003c/em\u003e 336, 1023\u0026ndash;1027 (2005). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.bbrc.2005.08.226\u003c/span\u003e\u003cspan address=\"10.1016/j.bbrc.2005.08.226\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eToy, W. \u003cem\u003eet al.\u003c/em\u003e Activating ESR1 Mutations Differentially Affect the Efficacy of ER Antagonists. \u003cem\u003eCancer Discov\u003c/em\u003e 7, 277\u0026ndash;287 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1158/2159-8290.Cd-15-1523\u003c/span\u003e\u003cspan address=\"10.1158/2159-8290.Cd-15-1523\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang, J., Wu, Z., Tian, Z., Chen, W. \u0026amp; Wu, G. OTUD7B stabilizes estrogen receptor α and promotes breast cancer cell proliferation. \u003cem\u003eCell Death Dis\u003c/em\u003e 12, 534 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41419-021-03785-7\u003c/span\u003e\u003cspan address=\"10.1038/s41419-021-03785-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeo, H., Seo, H., Lee, H. Y., Lee, S. H. \u0026amp; Park, Y. Comprehensive analysis of cellular estrogen signaling in representative estrogen receptor ligands. \u003cem\u003eChem Biol Interact\u003c/em\u003e 369, 110303 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cbi.2022.110303\u003c/span\u003e\u003cspan address=\"10.1016/j.cbi.2022.110303\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHelsen, C. \u0026amp; Claessens, F. Looking at nuclear receptors from a new angle. \u003cem\u003eMol Cell Endocrinol\u003c/em\u003e 382, 97\u0026ndash;106 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.mce.2013.09.009\u003c/span\u003e\u003cspan address=\"10.1016/j.mce.2013.09.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHewitt, S. C. \u003cem\u003eet al.\u003c/em\u003e Novel DNA motif binding activity observed in vivo with an estrogen receptor α mutant mouse. \u003cem\u003eMol Endocrinol\u003c/em\u003e 28, 899\u0026ndash;911 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1210/me.2014-1051\u003c/span\u003e\u003cspan address=\"10.1210/me.2014-1051\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArao, Y., Hamilton, K. J., Coons, L. A. \u0026amp; Korach, K. S. Estrogen receptor α L543A,L544A mutation changes antagonists to agonists, correlating with the ligand binding domain dimerization associated with DNA binding activity. \u003cem\u003eJ Biol Chem\u003c/em\u003e 288, 21105\u0026ndash;21116 (2013). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1074/jbc.M113.463455\u003c/span\u003e\u003cspan address=\"10.1074/jbc.M113.463455\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNelson, L. R. \u0026amp; Bulun, S. E. Estrogen production and action. \u003cem\u003eJ Am Acad Dermatol\u003c/em\u003e 45, S116\u0026ndash;124 (2001). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1067/mjd.2001.117432\u003c/span\u003e\u003cspan address=\"10.1067/mjd.2001.117432\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo, H. \u003cem\u003eet al.\u003c/em\u003e ARMC5 controls the degradation of most Pol II subunits, and ARMC5 mutation increases neural tube defect risks in mice and humans. \u003cem\u003eGenome Biol\u003c/em\u003e 25, 19 (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s13059-023-03147-w\u003c/span\u003e\u003cspan address=\"10.1186/s13059-023-03147-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu, Y. \u003cem\u003eet al.\u003c/em\u003e Armc5 deletion causes developmental defects and compromises T-cell immune responses. \u003cem\u003eNat Commun\u003c/em\u003e 8, 13834 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/ncomms13834\u003c/span\u003e\u003cspan address=\"10.1038/ncomms13834\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"ARMC5, ubiquitin ligase, HSPA1A, estrogen receptor α, ARMC5 gene copy number variation, breast cancer tumor-free survival","lastPublishedDoi":"10.21203/rs.3.rs-8810312/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8810312/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eARMC5 is the substrate recognition subunit of a recently discovered multiple-subunit ubiquitin ligase (ARMC5-E3), which is composed of ARMC5, cullin3, and RBX1. ARMC5-E3 is the major E3 for RNA polymerase II. ARMC5 interacts with multiple heat shock proteins, but the biological significance of these interactions is not known. We discovered that HSPA1A and its client protein, estrogen receptor alpha (ERα), were novel substrates of ARMC5-E3. ARMC5 deletion led to the accumulation of these substrates, while ARMC5 overexpression reduced ERα protein levels. In the presence of estrogen, HSPA1A was no longer bound to ERα, and ARMC5-E3 no longer regulated ERα degradation. Based on transcriptome analysis, ARMC5 overexpression in the presence of estrogen significantly altered the expression of 247 genes in breast cancer MCF7 cells. While estrogen stimulation increased ERE-binding by ERα in MCF7 cells, ARMC5 overexpression reduced it, supporting the notion that ARMC5-E3 degrades ERα and decreases its availability to interact with EREs. In ER-positive human breast cancer patients, increased \u003cem\u003eARMC5\u003c/em\u003e gene copy number was correlated to longer relapse-free survival time, probably in part due to ARMC5-E3’s effect on reducing HSPA1A-associated ERα levels. Therefore, ARMC5-E3 is a specific E3 that controls the degradation of HSPA1A and ERa. ARMC5 gene copy number variation is associated with tumor-free survival time of breast cancer patients. Hence, ARMC5 is a newly found regulator of estrogen signaling and function, and can serve as a prognostic parameter for breast cancer.\u003c/p\u003e","manuscriptTitle":"A novel ARMC5-containing ubiquitin ligase controls the degradation of HSPA1A and its client protein estrogen receptor alpha","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-22 12:35:45","doi":"10.21203/rs.3.rs-8810312/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a59f0e32-a547-4355-989e-709f4a9434d5","owner":[],"postedDate":"February 22nd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":62597455,"name":"Biological sciences/Cancer/Breast cancer"},{"id":62597456,"name":"Biological sciences/Cell biology/Proteolysis/Ubiquitylation"},{"id":62597457,"name":"Health sciences/Biomarkers/Prognostic markers"},{"id":62597458,"name":"Biological sciences/Cell biology/Mechanisms of disease"}],"tags":[],"updatedAt":"2026-03-23T09:59:14+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-22 12:35:45","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8810312","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8810312","identity":"rs-8810312","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}