Hsp74/14‐3‐3σ Complex Mediates Centrosome Amplification by High Glucose, Insulin, and Palmitic Acid

It has been reported recently that type 2 diabetes promotes centrosome amplification via 14‐3‐3σ/ROCK1 complex. In the present study, 14‐3‐3σ interacting proteins are characterized and their roles in the centrosome amplification by high glucose, insulin, and palmitic acid are investigated. Co‐immunoprecipitation in combination with MS analysis identified 134 proteins that interact with 14‐3‐3σ, which include heat shock 70 kDa protein 4 (Hsp74). Gene ontology analyses reveal that many of them are enriched in binding activity. Kyoto Encyclopedia of Genes and Genomes analysis shows that the top three enriched pathways are ribosome, carbon metabolism, and biosynthesis of amino acids. Molecular and functional investigations show that the high glucose, insulin, and palmitic acid increase the expression and binding of 14‐3‐3σ and Hsp74 as well as centrosome amplification, all of which are inhibited by knockdown of 14‐3‐3σ or Hsp74. Moreover, molecular docking analysis shows that the interaction between the 14‐3‐3σ and the Hsp74 is mainly through hydrophobic contacts and a lesser degree ionic interactions and hydrogen bond by different amino acids residues. In conclusion, the results suggest that the experimental treatment triggers centrosome amplification via upregulations of expression and binding of 14‐3‐3σ and Hsp74.


Introduction
Type 2 diabetes mellitus (T2DM) is a serious health problem worldwide, which can cause various chronic complications. [1] There is evidence that T2DM is associated with increased cancer risk and poor cancer prognosis. Currently, about 8-18% of all cancer patients have preexisting diabetes. [2] Individuals with diabetes who develop cancer have a 42% increased risk of death, a 21% increased risk of recurrence, and a significantly decreased 5-year overall and cancer-specific survival rate, as compared to individuals with cancer but free of diabetes. [3] However, little is known about the biological links between diabetes and cancer. It is speculated that deregulation of insulin and insulin-like growth factor signaling, obesity and inflammation, metabolic symbiosis, endoplasmic reticulum stress, and autophagy might play their roles.
Centrosome is the main microtubule organizing center. Each centrosome has two centrioles surrounded by the pericentriolar material, which plays important roles in cell division and the maintenance of genomic stability. [4] Centrosome amplification, acquisition of more than two centrosomes in each cell, is commonly found in various types of cancers, including solid tumors and hematological malignancies. [5] Recent experimental results suggest that centrosome amplification can initiate tumorigenesis [6] and increase cancer cell invasion potential. [7] Moreover, it is associated with poor cancer prognosis. [8] T2DM presents typical pathophysiological features that include hyperglycemia, hyperinsulinemia, and increased level of free fatty acids. Palmitic acid, the most common saturated free fatty acid, is often used to investigate the effects of free fatty acids, in particular the adverse effects. [9] We have recently reported that T2DM promotes cell centrosome amplification via www.advancedsciencenews.com www.proteomics-journal.com AKT-ROS-dependent signaling of 14-3-3σ and ROCK1, [10] and the pathophysiological factors in T2DM are the triggers. These results implicate that centrosome amplification is a candidate biological link between T2DM and cancer development. In a functional proteomic study, we identified nine proteins associated with centrosome amplification, which included 14-3-3σ , NPM, and PCNA, which were all confirmed to mediate the centrosome amplification by high glucose, insulin, and palmitic acid. [11] The results emphasize that 14-3-3σ and its binding partners play important roles in the occurrence of the diabetes-associated centrosome amplification.

Chemicals, Antibodies, and Cell
All chemicals were purchased from Sigma (St. Louis, MO, USA). Anti-γ -tubulin antibody (no. ab27074; mouse antibody) was purchased from Abcam (Cambridge, UK). Anti-14-3-3σ antibody (no. PLA0201; rabbit antibody) was purchased from Sigma. Anti-heat shock protein 70 kDa protein 4 (Hsp74) antibody (no. ab137631; rabbit antibody) was provided by Abcam. Other antibodies were provided by Cell Signaling Technology (Boston, MA, USA). HCT116 colon cancer cells were kindly provided by Dr. B. Vogelstein of the Johns Hopkins University School of Medicine. The culture medium and reagents were purchased from Gibco (Beijing, China). The palmitic acid stock was conjugated to fatty acidfree bovine albumin in a 3:1 molar ratio at 37°C for 1 h before use. Anti-gamma tubulin antibody was used to detect centrosome by immunofluorescent staining.

Cell Culture and Experimental Treatment
HCT116 cells were maintained in the DMEM (glucose, 5 mm) supplemented with 10% fetal bovine serum, and 1% penicillinstreptomycin in a humidified incubator with 5% CO 2 at 37°C. Cells from the cultures at 70% confluence were used for all

Significance Statement
Type 2 diabetes increases the risk of all-site cancer, except prostate cancer. Centrosome amplification is sufficient to initiate tumorigenesis. We have recently reported that type 2 diabetes promotes centrosome amplification in vivo and protein 14-3-3σ is a signal mediator, which provides a candidate biological link between diabetes and cancer. However, the molecular mechanisms underlying the diabetes-associated centrosome amplification remain unknown. The present study has shown that 14-3-3σ interacts with various proteins through co-immunoprecipitation in combination with proteomic analysis, which occur concomitantly with the centrosome amplification. More specifically, 14-3-3σ interacts with Hsp74 mainly via hydrophobic contacts to form a complex that mediates the centrosome amplification. The results provide the directions for preventing the centrosome amplification and its adverse consequences in patients with type 2 diabetes by targeting at the Hsp74/14-3-3σ protein complex. The results also highlight the advantages of analyzing protein-protein interactions using proteomic analysis strategy. experimental treatments. Cells treated for 48 h were used for quantification of centrosome number. Time course assays were performed and the time point was chosen, since this time point produced the significant level of differences for centrosome amplification between the control and the treated samples. Cells treated for 30 h were used for CoIP and Western blot analysis. The experimental treatment included high glucose (15 mm), insulin (150 nm), and palmitic acid (150 μm).

Confocal Microscopy
A cover slip was placed in a well of a six-well plate. Cells were plated at a density of 50 000 cells per well. Cells grown on the cover slips were fixed in cold methanol and acetone (1:1; v/v) for 6 min at -20°C, followed by three washes with PBS (10 min each time). Then, the cells were incubated with 0.1% Triton X-100 for 15 min and 3% BSA for 1 h. The cells were incubated with a primary antibody in 3% BSA in PBS overnight at 4°C, washed twice with PBS, and incubated with an FITC-conjugated secondary antibody in 3% BSA in PBS for 1 h at room temperature in the dark. Finally, the cells were mounted with mounting medium. Confocal microscopy was performed using the Zeiss LSM880 microscope (Oberkochen, Germany) with a 1.4 NA oilimmersion lens, and image processing was performed with Zen software (Oberkochen, Germany).

Filter-Aided Sample Preparation
The eluted proteins were then digested according to the filteraided sample preparation (FASP) procedure described. Briefly, 200 μg of proteins for each sample (supernatant) were incorporated into 30 μL SDT buffer (4% SDS, 100 mm DTT, 150 mm Tris-HCl pH 8.0) at 90°C for 5 min. The detergent DTT and other lowmolecular-weight components were removed using 200 μL UA buffer (8 m urea, 150 mm Tris-HCl pH 8.0) by repeated ultrafiltration (Microcon units, 30 kDa). Then 100 μL 0.05 m iodoacetamide in UA buffer was added to block reduced cysteine residues and the samples were incubated for 20 min in darkness. The filter was washed with 100 μL UA buffer three times and then twice with 100 μL 25 mm NH 4 HCO 3 . Finally, the protein suspension was digested with 2 μg trypsin (Promega) in 40 μL 25 mm NH 4 HCO 3 overnight at 37°C, and the resulting peptides were collected as a filtrate.

LC-ESI-MS/MS Analysis by Q Exactive
Experiments were performed on a Q Exactive mass spectrometer that was coupled to Easy nLC (Proxeon Biosystems, now Thermo Fisher Scientific). Six microliters of each fraction was injected for nanoLC-MS/MS analysis. The peptide mixture (5 μg) was loaded onto a the C18-reversed phase column (Thermo Scientific Easy Column, 10 cm long, 75 μm inner diameter, 3μm resin) in buffer A (0.1% formic acid) and separated with a linear gradient of buffer B (80% acetonitrile and 0.1% formic acid) at a flow rate of 250 nL min -1 controlled by IntelliFlow technology over 140 min. MS data were acquired using a data-dependent top 10 method dynamically choosing the most abundant precursor ions from the survey scan (300-1800 m/z) for high-energy collisioninduced dissociation fragmentation. Determination of the target value is based on predictive automatic gain control. Dynamic exclusion duration was 60 s. Survey scans were acquired at a resolution of 70 000 at m/z 200 and resolution for HCD spectra was set to 17 500 at m/z 200. Normalized collision energy was 30 eV and the underfill ratio, which specifies the minimum percentage of the target value likely to be reached at maximum fill time, was defined as 0.1%. The instrument was run with peptide recognition mode enabled.

Sequence Database Searching and Data Analysis
All MS/MS spectra were searched using Mascot software v2.2.2 software (Matrix Science, London, UK), against the Homo sapiens UniProtKB database (www.uniprot.org). For protein identification, the following options were used: Peptide mass tolerance, 20 ppm; MS/MS tolerance, 0.1 Da; enzyme, trypsin; and missed cleavage, 2. Fixed modification was carbamidomethyl. Variable modification: oxidation.

Bioinformatic Analysis
To examine the biological and functional properties of the identified proteins, Gene ontology (GO) annotation was conducted by searching the GO database (http://www.geneontology.org). Functional category analysis was performed with protein2go and go2protein for annotation. Visualization and Integrated Discovery v6.7 was used for functional enrichment analysis of GO terms and KEGG pathways. A false discovery rate of <0.01 was selected as the cut-off criterion.

Western Blot Analysis
The cells were lysed in RIPA buffer. Proteins were separated by PAGE and transferred onto polyvinylidene fluoride membrane. After blocking for 1 h at room temperature with TBST containing 0.05% (v/v) Tween-20 and 5% (w/v) non-fat milk, the membranes were incubated with primary antibodies overnight at 4°C, followed by washes with TBST containing 0.05% Tween-20. The membranes were then incubated with a horseradish peroxidaseconjugated secondary antibody for 1 h at room temperature. ECL reagents (Thermo Biosciences, Massachusetts, USA) were used to visualize the protein bands which were captured on X-ray film.

Homology Modeling and Molecular Docking
To further elucidate the functional relationships between 14-3-3σ and interacting proteins, a protein of interests was chosen, which is Hsp74. The crystallographic structure of the Hsp74 has not been published yet. In order to expose the binding mode between human 14-3-3σ protein and Hsp74 at the molecular level, the 3D structure of the Hsp74 was www.advancedsciencenews.com www.proteomics-journal.com built by means of modeler 9.19 homology modeling software (http://salilab.org/modeller/). The sequence in FASTA format of Hsp74 was retrieved from NCBI (Accession: P34932.4). The crystallographic structure of Saccharomyces cerevisiae Hsp110 (PDB ID: 3C7N) was selected as the templates for modeling. Molecular docking were performed to investigate the binding mode between the human 14-3-3σ protein and the Hsp74 using the ZDOCK server (http://zdock.umassmed.edu/). The 3D structure of the human 14-3-3σ (PDB ID: 6FCP) was downloaded from Protein Data Bank (http://www.rcsb.org/pdb/home/home.do), while the 3D structure of the Hsp74 was built by modeler 9.19. For docking, the default parameters were used as described in the ZDOCK server. The top ranked pose as judged by the docking score was using PyMoL 1.7.6 software (http://www.pymol.org/).

Statistical Analysis
All the experiments were performed in triplicate. Data were expressed as the mean ± SD. Student's t-test was performed for comparison between two groups. The statistical analysis software package SPSS 21.0 was employed for the statistical comparisons. A p-value < 0.05 was considered significant.

High Glucose, Insulin, and Palmitic Acid Induce Centrosome Amplification
We have recently reported that the level of centrosome amplification is increased in peripheral blood mononuclear cells from patients with type 2 diabetes. AKT-ROS-dependent upregulation of 14-3-3σ and ROCk1 as well as their binding and translocation to centrosome is the underlying signal transduction pathway for the diabetes-associated centrosome amplification. [10] In the present study, we further investigated the molecular basis of centrosome amplification associated with T2DM using colon cancer cells as an experimental model, which were treated with high glucose, insulin, and palmitic acid. As shown in Figure 1A,B, high glucose, insulin, and palmitic acid were able to induce moderate centrosome amplification in the cells. Under the experimental conditions, most cells with centrosome amplification had three to five centrosomes per cell ( Figure 1A). Glucose, insulin, and palmitic acid were used at 15 mm, 5 nm, and 150 μm, respectively, which were close to their pathophysiological levels.

Identification of 14-3-3σ -Interacting Proteins
To characterize proteins attached to the 14-3-3σ , we performed CoIP using 14-3-3σ antibody and identified the partner proteins using MS. As shown in Table 1, a total of 165 proteins were identified, among which 134 protein were identified from the treated samples, 19 proteins were identified from the control samples and 12 proteins were identified from both control and treated samples (Figure 2). Thus, 153 proteins were responsive to the experimental treatment.

Bioinformatics Analysis of the 14-3-3σ Binding Proteins
GO clustering analysis was performed to provide relevant information about their biological processes, molecular functions, and cellular components. Totally, there were 134 14-3-3σ interacting proteins. One-hundred thirty-one (97.76%) proteins were categorized into biological process, 133 (99.25%) proteins were involved in molecular functions, and 133 (99.25%) proteins were grouped into cellular components. Within the biological processes category, the majority of the proteins were involved in cellular process and metabolic process. For molecular functions, the data indicated that most of the proteins were linked to binding, catalytic activity, and structural molecule activity. Regarding cellular components, cell, organelle, and cell part were the top-ranked www.advancedsciencenews.com www.proteomics-journal.com  (Figure 3A-C). KEGG pathway enrichment analysis revealed that the 14-3-3σ -interacting proteins were related to 149 pathways. The 20 highly enriched pathways are shown in Figure 3D. The five most significantly enriched were ribosome, carbon metabolism, biosynthesis of amino acids, PI3K-AKT signaling pathway, and protein processing in ER.

Hsp74 Mediates the Centrosome Amplification
From the proteins pulled down using 14-3-3σ antibody, we were interested in Hsp74. Why we targeted at Hsp74? The protein came to our attention, as several Hsp proteins are present in centrosome, [19] which suggests that Hsp proteins may play roles in centrosome homeostasis. Moreover, Hsp proteins are known binding partners of 14-3-3 proteins. [20] Thus, we investigated whether Hsp74 contributed to the centrosome amplification. Indeed, we found that the expression level of Hsp74 was increased by high glucose, insulin, and palmitic acid, which was inhibited by Hsp74 specific siRNA ( Figure 4A). Knockdown of Hsp74 using their siRNA downregulated the treatment-induced centrosome amplification ( Figure 4B).

5 14-3-3σ -Hsp74 Complex is Required for the Centrosome Amplification
We next performed experiments to confirm the binding between 14-3-3σ and Hsp74, and to examine whether 14-3-3σ and Hsp74 complex was required for the centrosome amplification. As expected, Hsp74 was pulled down by 14-3-3σ antibody ( Figure 5A). Importantly, the binding between Hsp74 and 14-3-3σ was increased by high glucose, insulin, and palmitic acid ( Figure 5A). If the Hsp74 /14-3-3σ complex mediated the treatment-induced centrosome amplification, inhibition or disruption of the complex would inhibit the centrosome amplification. siRNA technology was used to inhibit or disrupt the protein complex via protein knockdown of Hsp74 or 14-3-3σ . Indeed, individual knockdown of Hsp74 or 14-3-3σ protein level ( Figure 5B) attenuated the treatment-induced centrosome amplification ( Figures 4A,B  and 5C). In addition, when Hsp74 was knocked down using siRNA, although Hsp74 was pulled down by 14-3-3σ antibody, the level was very low ( Figure 5D). These data proved that the complex was reduced using siRNA of Hsp74.

Discussion
In the present study, we showed that high glucose, insulin, and palmitic acid could induce centrosome amplification ( Figure 1A,B) and 14-3-3σ was a signal mediator, which is in agreement with our previous report that the experimental treatment induces centrosome amplification via ROCK1/14-3-3σ complex. [10] Moreover, previous report indicated 14-3-3 proteins interacted with over 200 human phosphoproteins in HeLa cells using 14-3-3 affinity chromatography. [21] Here, we identified 31 and 146 14-3-3σ binding proteins in the untreated and treated HCT116 cells, respectively, which included Hsp74 (Table 1, Figure 2). Compared with previous study, the reasons for the difference in the amount of interacted protein were: 1) different cell types and 2) different antibody used. Compared with untreated cells, the amount of treated cells proteins almost increased fivefold. The experimental treatment increased the abundance to several proteins in the centrosomes, which may be due to the treatment-promoted translocation of these proteins to the centrosomes, the role of which in the centrosome amplification require further examinations (Lu et al., unpublished data; Figure 5B). Hsp74, also called Apg-2 (ATP and peptide-binding protein in germ cells-2), [22] is a member of the Hsp110 family. It was first described as Hsp70 RY from B cells [23] and is encoded by HSPA4 gene. [24] Hsp74 is inducible under various conditions, including cancer, [25] chronic inflammation, [26] and acidic pH stress. [27] It is overexpressed in various cancer cells and intestinal cells. Chen and co-workers reported that Hsp74 was highly expressed in bladder cancer and distributed into cytoplasm, which was associated with keratin 1. [28] Hsp74 also upregulates Bcl-2 and IL-17 in the gut, which controls cell apoptosis as well as immune response. [26] In addition, the synthesis of Hsp74 affects microtubules stability, which is related to cytoskeletal stability. [29] Centrosome comprises of a pair of centrioles surrounded by the pericentriolar material. [4] Centrioles comprise nine triplet microtubules, which are arranged into a cylinder with a diameter of 250 nm. [30] Whether Hsp74 modifies the stability of centriole for centrosome amplification remains unknown. In the present study, we showed that the overexpression of Hsp74 mediated centrosome amplification ( Figure 4A,B). Furthermore, Hsp74 and 14-3-3σ formed a complex to promote centrosome amplification ( Figure 5A-D) triggered by the experimental treatment, as disruption of the complex attenuated the treatment-elicited centrosome amplification.
large aromatic side chains of ChREBP are intimately involved in both hydrogen bonding and van der Waals stacking interactions with 14-3-3. [35] In 14-3-3-Exos complex, the binding sequence of ExoS was ten residues ( 421 GLLDALDLAS 430 ). The interaction mostly relies on hydrophobic interactions and a lesser degree of electrostatic interactions. [36] In this study, we predicted the interactions of the 14-3-3σ -Hsp74 complex using molecular docking analysis. 14-3-3σ binds to Hsp74 through hydrophobic contacts, electrostatic interactions, and hydrogen bond interactions ( Figure 6A,B). Compare with Mode I-III, we do not find the basic cluster of 14-3-3 residues (Lys-Arg-Arg). However, compared with 14-3-3-Exos and 14-3-3-ChREBP complex, 14-3-3σ binding to Hsp74 is also mainly through hydrophobic interactions. These data provide insight into the structural basis for the affinity binding between 14-3-3σ and Hsp74. Ideally, the binding between 14-3-3σ and Hsp74 is confirmed in vitro using purified proteins, which can specify whether phosphorylation is required for the binding or not. If yes, required phosphorylation site(s) can also be found. Mutant proteins can be created to identify the binding domain(s).
In conclusion, high glucose, insulin, and palmitic acid promote centrosome amplification by increasing expressions of 14-3-3σ and Hsp74 in HCT116 cells. The results from CoIP assay, proteomic analysis, and functional studies show the experimental treatment increase the formation of 14-3-3σ /Hsp74 complex that mediates the centrosome amplification.