Silencing of carbonic anhydrase I enhances the malignant potential of exosomes secreted by prostatic tumour cells

Abstract We report results showing that the silencing of carbonic anhydrase I (siCA1) in prostatic (PC3) tumour cells has a significant impact on exosome formation. An increased diameter, concentration and diversity of the produced exosomes were noticed as a consequence of this knock‐down. The protein composition of the exosomes' cargo was also altered. Liquid chromatography and mass spectrometry analyses identified 42 proteins significantly altered in PC3 siCA1 exosomes compared with controls. The affected proteins are mainly involved in metabolic processes, biogenesis, cell component organization and defense/immunity. Interestingly, almost all of them have been described as ‘enhancers' of tumour development through the promotion of cell proliferation, migration and invasion. Thus, our results indicate that the reduced expression of the CA1 protein enhances the malignant potential of PC3 cells.

the increased abundance of some specific exosomal miRNAs and proteins has been positively correlated with the stage and degree of prostate cancer progression. 10,11 Since exosomes reflect the pathological state of the secretory cells, they have become attractive new biomarkers for the diagnosis and prognosis of cancer. 1,10 Furthermore, they represent a novel therapeutic strategy for the treatment of cancers. 1,10 Besides being ubiquitous in all life forms, it was found that carbonic anhydrases perform numerous activities in a variety of organisms. Primarily they catalyse the hydration of carbon dioxide to a bicarbonate ion and a proton. 12 This reaction is reversible and the metalloenzymes can accomplish it in both directions, forward and reverse. 13 This interconversion is essential for many biological processes, which require acid-base balance and depend on spatially and temporally regulated ion transport in various subcellular compartments and across the plasma membrane. Abnormal levels or activities of these enzymes have been associated with many disorders such as obesity, gastric ulcers, glaucoma, acid-base imbalances, cancer and epilepsy. 14 Carbonic anhydrase I (CA I) is a zinc metalloenzyme belonging to the α CA family. 12 It is involved in pH homeostasis, respiration, erythroid differentiation and some pathological processes such as anaemia, chronical acidosis, proliferative diabetic retinopathy and diabetic macular and vasogenic oedemas. 14,15 Variations in the expression of CA I have recently been associated with some malignancies. A low level of CA I in colonic epithelial cells was found to be a specific marker for the prediction of colorectal cancer. 18 On the other hand, CA I was highly expressed in the sera of patients with stage I non-small cell lung cancer (NSCLC). 19 Even the plasma of patients with prostate cancer contains an increased level of CA I compared with healthy controls. 20 Thus, the elevated level of the CA I protein in the plasma or serum may represent a promising biomarker for both prostate cancer and early stage NSCLC.
Interestingly, the autoantibodies against CA I were also linked to the progress of malignant diseases. Lakota et al have reported a significant increase in these autoantibodies in the sera of patients with tumours that spontaneously regressed after high-dose therapy and autologous stem cell transplantation. 21 Thus, the presence of autoantibodies against CA I in patients' sera could be put forward as a marker of a good prognosis. Further in vitro research on tumour cells showed that treatment with these sera may result in the upregulation of the CA1 mRNA expression, which can be linked to the down-regulation of the mRNAs encoding structural proteins of basal lamina, the cytoskeleton, WNT7B and collagen triple helix repeat containing 1 (CTHRC1). 22 On the other hand, CA1 mRNA silencing via the RNA interference system in PC3 tumour cells enhanced the expression of some of the extracellular matrix (ECM) proteins. 23 To reveal the role of CA I in prostatic cancer development, a more thorough study is required. The mRNA CA1 was silenced in PC3 cells and the secreted exosomes were subsequently isolated and characterized using different methods. A comprehensive proteomic analysis was then performed using mass spectrometry to identify the differences in the protein composition of the exosomal cargo of these cells developed due to the changes in mRNA CA1 expression.
Taken together, the comprehensive characterization of exosomes derived from PC3 prostate cells, which have different mRNA CA1 expression, shows that the knock-down of CA1 mRNA in PC3 cells alters the exosomal pattern of cancer cells and enhances their malignant potential.

| Transient silencing of CA1 gene
For the transient silencing of the CA1 gene, PC3 cells were seeded at a density of 1.5 × 10 6 cells in T25 cell culture flasks. Transient CA1 knockdown cells were produced by transfection with SmartPool CA1 siRNA specific oligonucleotides (Dharmacon, GE Healthcare, USA) using DharmaFECT ™ (GE Healthcare, USA) according to the manufacturer's recommendations, while siMock oligonucleotides were used as a control. Eighteen hours after transfection, the medium was changed to a (FCS)/antibiotic (ATB) free medium and the cells were incubated for 48 hours at 37°C. Conditioned media were collected for exosome isolation and the cells were lysed in a RIPA buffer for Western blot analysis.

| Exosome preparation and purification
For the isolation of PC3 exosomes, PC3 silencing of carbonic anhydrase I (siCA1) as well as PC3 siMock (1.5 × 10 6 each) cells were cultured in T25 cell culture flasks 50 mL (two flasks with 25 mL medium each) of FCS/ATB free medium. After 48 hours (max 80% confluency), the media were collected (100 mL from PC3 siCA1 and siMock cells), centrifuged (300g for 10 minutes) to remove cell debris and filtered through a 0.22 µm filter (Merck Millipore, USA).
The pre-cleared medium was concentrated to 2 mL using a 100 kDa MWCO Amicon Ultra Centrifugal Filter (Merck Millipore, USA). The concentrated media samples (2 mL) were centrifuged at 3 000 g for 15 minutes at 4°C, and subsequently, the exosomes were precipitated using a ExoQuick-TC exosome precipitation solution (System Biosciences, USA) according to the manufacturer's instructions. The medium was then briefly mixed with 1/5 V (400 µL) of an ExoQuick-TC exosome precipitation solution and refrigerated overnight (at least 12 hours) at 4°C. The next day, an ExoQuick-TC/medium mixture was centrifuged at 1 500 g for 30 minutes. After centrifugation, the exosomes appeared as beige pellets and were washed once in a phosphate-buffered saline (PBS) solution, and resuspended in an appropriate volume of PBS.

| Nanoparticle tracking analysis
A NanoSight NS500 (Malvern Instruments Ltd., Malvern, UK) equipped with a sCMOS Trigger camera and a 405 nm laser was used to measure the concentration and size distribution of the exosomes isolated from PC3 siCA1 and PC3 siMock cells. The measured data were analysed using the noparticle tracking analysis (NTA) 2.3 analytical software. NTA is based on capturing the Brownian motion and light scattering properties to obtain particle size distributions and concentrations. Each sample was diluted in PBS prior to the measurements to optimize the number of particles (from 60x to 3600x dilution; used dilution:1800x). Samples were measured in triplicates in 60-second videos with manual shutter and gain adjustments. All measurements were performed at room temperature.

| Transmission electron microscopy
Five mocrolitres of PC3 siCA1 and siMock exosome samples were placed onto glow-discharge activated carbon/formvar grids and were allowed to adsorb for 60 seconds at room temperature. 24 After adsorption, the grids were negatively stained with 1% ammonium molybdate + 0.1% trehalose for 30 seconds. 25

| Gene expression analyses
For reverse transcriptase quantitative PCR (RT-qPCR), the total RNA was extracted from PC3 siCA1 and PC3 siMock cells with a NucleoSpin RNA II kit (Macherey-Nagel, Dueren, Germany). The RNA was depleted from genomic DNA using DNase treatment
Enzymatic cleavage was stopped with the addition of 5% formic acid (FA, Sigma Aldrich, USA). Subsequently, each sample was desalted using a MicroTrap Peptide 6 PK cartridge (Bruker, Germany). The extracts were concentrated in the SpeedVac (Eppendorf, Germany) to 0.5 μg/μL.
Samples were loaded onto a Symmetry C18 trap column (20 mm length, 180 μm diameter, 5 μm particles size). After 3 minutes of desalting/concentration by 1%, acetonitrile containing 0.1% formic acid at a flow rate 10 μL/min, peptides were introduced to a BEH130 C18 analytical column (200 mm length, 75 μm diameter, 1.7 μm particle size). For the thorough separation, a 60-min gradient of 5%-40% acetonitrile with 0.1% formic acid was applied at a flow rate of 300 nL/min. The column outlet was connected to a PicoTip emitter (360 μm outer diameter, 20 μm inner diameter, 10 μm tip diameter) and samples were nanosprayed (3.4 kV capillary voltage) to the quadrupole time-of-flight mass spectrometer

Q-TOF Premier (Waters).
Spectra were recorded in a data-independent manner in MSE mode. This mode uses alternate scans at low (4 eV) and high (20-40 eV ramp) collision energies to obtain full-scan mass data for both precursors and fragments in a single run. Ions with 50-1950 m/z were detected in both channels. The spectral acquisition scan rate was 1.2 seconds, with a 0.05 seconds inter-scan delay. The external mass calibrant Glu1-Fibrinopeptide B (500 fmol/mL) was infused through the reference line at a flow rate of 500 nL/min and employed for mass correction.

| Efficacy of the CA1 silencing
The knockdown of the CA1 gene was performed using the siRNA-SMARTpool system targeting the CA1 mRNA. The CA1 siRNAtransfected PC3 cells did not show any differences in morphology ( Figure 1A) compared to the negative control (PC3 siMock). The silencing efficiency of the CA1 mRNA was confirmed by qRT-PCR and Western blot analysis. The result showed a 75% reduction in mRNA expression in the PC3 siCA1-transfected cells ( Figure 1B).
Accordingly, the abundance of CA I protein was also decreased ( Figure 1C).
A NanoSight system was also used for the nano-sized particle quantity determination in the suspension of isolated exosomes. The concentration of the nano-size particles derived from PC3 siCA1 cells was significantly higher than from PC3 siMock control cells (unpaired t test, PC3 siCA1 vs. PC3 siMock, P = 0.007, P < 0.01**).
The average concentration was established as 3.02E+11 particles/ mL (σ = 5.48E+10 particles/mL) and 1.75E+11 particles/mL (σ = 1.81E+10 particles/mL) for the suspensions isolated from the silenced and the control cells, respectively. The values are presented as means ± σ of three biological and technical replicates ( Figure 3B). Isolated exosomes were diluted to a suitable concentration with phosphate-buffered saline (PBS) (1800x; siCA1 as well as control siMock), and the size distribution was analysed by NTA using a NanoSight NS500 (for each sample 3 × 60 second runs; for error bars indicating ±standard error of the mean/mode and final nanoparticle concentration see Figure 3) Furthermore, the amount of total protein was quantified in the isolated nanovesicles by BCA assay and the presence of exosome specific markers CD9, CD63 (both tetraspanins) and TSG101 (tumour susceptibility gene 101 protein) was verified using Western Blot analyses and ELISA assay. As shown in Figure 3C,D, the level of these markers as well as the total protein content ( Figure 3E) were higher in the exosomal suspension derived from CA1 siRNA transfected cells compared with control PC3 siMock cells. Significant differences were also found between protein patterns of CA1 silenced and control cells ( Figure 3E) separated in 12% SDS-PAGE.
Generally, these molecular and biophysical measurements demonstrate that the silencing of the CA1 gene in PC3 prostate cells has a profound effect on the production of exosomes and their secretion into the cultivation medium. It is also noteworthy that the method of centrifugation, filtration, concentration of the culture medium and final isolation using an ExoQuick seems to be reliable for the purification of high-quality exosomes.  (Table S1). Among them, 42 proteins demonstrated statistically significant (P < 0.05, ≥1.5-fold change)

| CA1 silencing in prostatic cancer cells alters the protein cargo of the exosomes
differences (Table 1). Interestingly, almost all of them (41 proteins) were more abundant in PC3 siCA1-derived exosomes compared with controls.
A gene ontology (GO) database search was carried out to reveal molecular functions and cellular localization of the identified proteins as well as the biological processes in which they are involved.  Here it is important to mention that GO annotations often provide

| D ISCUSS I ON
We have previously shown that the presence of autoantibodies against CA I can be correlated with the stage of the malignant disease. It was noticed, that high titers of these autoantibodies in the sera of patients can be associated with a good prognosis which is usually accompanied with spontaneous regression of the tumour after high dose therapy and autologous stem cell transplantation. 21 To shed more light onto this phenomenon, we have treated several cancer cell lines derived from prostatic, colon or breast carcinomas with the sera of these patients. Significant modifications in the morphology of these cells were noticed in contrast to controls.
The mRNA levels for proteins associated with basal lamina assembly, cytoskeleton, WNT7B and CTHRC1 were down-regulated, whereas the expression of CA1 mRNA was up-regulated in the tumour cell lines treated with the anti-CA I autoantibody positive sera. 22 To

ANOVA (p)
Fold log 2 (fold) Proteomic profiles from PC3 siCA1 and control PC3 siMock derived exosomes were compared. Proteins with more than 2 matching peptides, statistically significant (p < 0.05) and with more than 1. with breast cancer along with higher proteasome activity. 32 The overexpression of proteasome subunit S10 accompanied by an increased proteasome activity has also been reported in melanoma. 33 It is noteworthy, that the application of a proteasome inhibitor suppresses the transactivation of the androgen receptor (AR) in an androgen-dependent manner in prostate cancer LNCaP and PC3 cell lines. Thus, the proteasome system seems to play an important role in the regulation of AR activity and can be proposed as a unique target for the development of therapeutic drugs blocking androgen/ AR-mediated prostate tumour growth. 34 Because translation is the final step in the production of a functional protein, alterations in translational control may represent an 'oncogenic' node which may serve as a potential target for tumour suppression. Ultimately, the target of specific translational components in cancer represents the most promising therapeutic approaches for clinical trials. 35 In PC3 siCA1 derived exosomes we also found an increased level Molecular chaperones and heat shock proteins (HSP) represented by endoplasmin, heat shock 70 kDa protein 1B, heat shock protein beta-1, T-complex protein 1 subunit beta, T-complex protein 1 subunit eta and T-complex protein 1 subunit epsilon were also identified in this study as differentially altered due to reduced levels of CA I.
HSPs are known to be expressed at the high amount in a wide range of tumours. They are in close association with a poor prognosis and resistance to therapy. The HSP family members participate in autonomous cell proliferation and inhibit death pathways. 51 Also, 14-3-3 protein zeta/delta which was more abundant in PC3 siCA1 derived exosomes promotes cell proliferation, adhesion and survival and inhibits apoptosis in multiple cancers. For this function represents a novel molecular agent for targeted cancer therapy. 52 In addition to the above-mentioned proteins, we observed an increase in the abundance of nucleic acid binding proteins (histone H3.2, histone H2B type 1-C/E/F/G/I, X-ray repair cross-complementing protein 5, heterogeneous nuclear ribonucleoprotein H).
These are key players in chromatin structure shaping. They regulate fundamental cellular processes such as chromosome segregation and gene expression. For these reasons, histone variants represent potential drivers of cancer initiation and/or progression. Thus, targeting histone deposition or the chromatin remodelling machinery may have a therapeutic value. 53 Lastly, among the differentially abundant proteins, are the constituents of the cell cytoskeleton (actin, alpha cardiac muscle 1, tubulin, beta-6 chain, plectin), defense/immunity protein (complement C3) and transfer/carrier (importin subunit beta-1).
Mutations and the abnormal expression of cytoskeletal and cytoskeletal-associated proteins play an important role in the ability of cancer cells to resist chemotherapy and metastasize. The dynamic reorganization of the actin cytoskeleton is a prerequisite for the morphology, migration and invasion of cancer cells. 54 The complement 3 (C3), as a central protein of the complement system, is expressed in numerous cancer tissues (lung, colorectal, esophageal and gastric) indicating that C3 may be a suitable biomarker for the outcome of malignancies. 55 Moreover, up-regulation of importin subunit beta-1 promotes tumour cell proliferation and predicts poor prognosis in non-small lung, cervical, glioma and gastric cancer. 56,57 All these proteins are known as building blocks for cell formation and survival.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.