The Myc-tagged YBX1 expression plasmid pDESTmycYBX1 [Addgene plasmid 19878, deposited by Landthaler (18)] was obtained from Addgene (Cambridge, MA, USA). The other expression plasmids (pHA-MAEL, pMyc-DDX39, pMyc-EIF4A1, pMyc-EIF3F, pMyc-ELAV1, pMyc-PABPC1, pMyc-SYNCRIP and pMyc-KHSRP) were constructed as follows. The coding sequence of each gene was amplified from the first strand cDNA of MDA-MB-231 or SW480 cells with the primers described in Table I. The amplified products were cloned into the pMD18-T vector (Takara, Dalian, China) by TA cloning method, and subcloned into the pCMV-HA or pCMV-Myc vector (Clontech Laboratories, Inc., Mountain View, CA, USA) with restriction enzyme sites in the primer (Table I).

The human breast cancer MDA-MB-231 and colorectal cancer SW480 cells [American Tissue Culture Collection (ATCC), Manassas, VA, USA] were cultured in L-15 media that was supplemented with glutamine, antibiotics and 10% fetal bovine serum (FBS) at 37°C in 5% CO₂. The cells were washed with phosphate-buffered saline (PBS) and incubated with 0.5 mM cross-linker DSP (Pierce, Rockford, IL, USA) for 30 min. After cross-linking, cells were sequentially washed with PBS, PBS containing 25 mM Tris-HCl buffer and PBS. Then, cells were lysed in RIPA buffer by sonication, and subjected to centrifugation at 12,000 × g for 15 min. The supernatants were transferred to a new tube for immunoprecipitation.

Immunoprecipitation was performed as described by Lin et al(19). In each 1.5-ml tube, 2.5 mg of lysates was mixed with 2 μg of polyclonal antibody against MAEL or preimmune rabbit IgG (Signalway Antibody Co., Ltd., Nanjing, China), and incubated overnight at 4°C with gentle shaking. Immune complexes were collected on protein A/G agarose beads and washed three times to remove non-specific binding using the same buffer as for the cell lyses. Proteins were eluted from the beads by boiling the beads in loading buffer for 10 min, and were then separated on SDS-PAGE gel and visualized with silver staining.

Each excised protein gel band was destained and digested with trypsin as described by Lin et al(19). Nano-LC MS/MS experiment was performed on an HPLC system composed of two LC-20AD nano-flow LC pumps, an SIL-20AC autosampler and an LC-20AB micro-flow LC pump (Shimadzu, Tokyo, Japan) connected to an LTQ Orbitrap mass spectrometer (Thermo Fisher Scientific, San Jose, CA, USA). Sample was loaded on a CapTrap column (0.5 × 2 mm; Michrom Bioresources, Inc., Auburn, CA, USA) for 6 min at a flow rate of 25 μl/min. The sample was subsequently separated by a C18 reverse-phase column (0.10 × 150 mm, packed with 3 μm Magic C18AQ particles; Michrom Bioresources) at a flow rate of 500 nl/min. The mobile phases were 2% acetonitrile with 0.1% formic acid (phase A and the loading phase) and 95% acenitrile with 0.1% formic acid (phase B). To achieve proper separation, a 60-min linear gradient from 0 to 80% phase B was employed. The separated sample was introduced into the mass spectrometer via an Advance 30-μm silica tip (Michrom Bioresources). The spray voltage was set at 1.6 kV and the heated capillary at 180°C. The mass spectrometer was operated in the data-dependent mode, and each cycle of duty consisted of one full-MS survey scan at the mass range 385–2000 Da with resolution power of 100,000 using the Orbitrap section, followed by MS/MS experiments for 10 strongest peaks using the LTQ section. The AGC expectation during full-MS and MS/MS were 1,000,000 and 10,000, respectively. Peptides were fragmented in the LTQ section using collision-induced dissociation with helium and the normalized collision energy value set at 35%. Previously fragmented peptides were excluded for 60 sec.

Tandem mass spectra were extracted by BioWorks version 3.3.1 SP1 (Thermo Fisher Scientific). All MS/MS samples were analyzed using Sequest (Thermo Fisher Scientific; version 28). Sequest was set up to search the human proteome database from UniProt release 2010_08 (downloaded from the official website of UniProtKB following this link: http://www.uniprot.org/uniprot/?query=organism:9606+keyword:181&format=*&compress=yes, downloaded at 2010-07-13) assuming the digestion enzyme trypsin. Sequest was searched with a fragment ion mass tolerance of 1.00 Da and a parent ion tolerance of 20 ppm. Oxidation of methionine (+15.99492 Da) and acetylation of lysine (+42.01057 Da) were specified as variable modifications. Trans-Proteomic Pipeline (20) was used to validate MS/MS based peptide and protein identifications. Peptide probability was specified by the Peptide Prophet algorithm (21). Protein probabilities were assigned by the Protein Prophet algorithm (22). Proteins that contained similar peptides and could not be differentiated based on MS/MS analysis alone were grouped to satisfy the principles of parsimony.

Gene Ontology (GO) enrichment analysis was performed using the Database for Annotation, Visualization and Integrated Discovery (DAVID) functional annotation tool (http://david.abcc.ncifcrf.gov) (23). UniProt accession numbers of MAEL-associated proteins were submitted to DAVID system and categorized based on biological process (BP), molecular function (MF) and cellular component (CC) using the software.

Transfection and co-IP were performed as previously described (24). Hek293 cells were co-transfected with HA-tagged MAEL expression plasmids and expression plasmids of each Myc-tagged SG gene. At 24 h after transfection, cell lysates were prepared and immunoprecipitated with rabbit anti-HA polyclonal antibody or preimmune rabbit IgG, and the precipitated protein was detected by western blot analysis using the anti-Myc monoclonal antibody. The anti-HA, anti-Myc and IgG antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA).

SW480 cells were grown on glass coverslips, and treated with 0.5 mM hydrogen peroxide (H₂O₂) for 1 h to induce the formation of SGs. Treated cells and control cells were fixed with 4% paraformaldehyde and permeabilized in 0.1% Triton X-100, respectively. To reduce the unspecific binding, the cells were incubated in 1% BSA overnight the 4°C. Then, cells were incubated with anti-MAEL polyclonal antibody and anti-PABPC1 monoclonal antibody (Pierce, Rockford, IL, USA) for 2 h at room temperature, followed by incubation with Texas Red-conjugated anti-rabbit IgG (red) and FITC-conjugated anti-mouse IgG (green) for 1 h. The nuclei were labeled by DAPI. Images were obtained using a Zeiss LSM 510 confocal microscope (Carl Zeiss Imaging, Oberkochen, Germany). Texas Red-conjugated anti-rabbit IgG and FITC-conjugated anti-mouse IgG were purchased from Santa Cruz Biotechnology.

To identify the potential interacting partners of MAEL protein, we performed the co-IP experiments to isolate the MAEL complex in human breast cancer MDA-MB-231 and colorectal cancer SW480 cells. The immune complexes of the anti-MAEL antibody and IgG were resolved on SDS-PAGE gel followed by silver staining. The protein bands which only existed in the complex of the anti-MAEL antibody were excised (Fig. 1), digested with trypsin and analyzed by Nano-LC MS/MS method. All MS/MS samples were analyzed using Sequest. According to the database search results, we identified 178 non-redundant proteins in the three bands from MDA-MB-231 cells and 167 proteins in the four bands from SW480 cells. Among these proteins, 78 proteins were common to both MDA-MB-231 and SW480 cells.

To find whether these 78 proteins share some particular features, we performed GO enrichment analysis with DAVID (23). Table II shows the top 5 significantly enriched terms in the MAEL-associated proteins. In the molecular function (MF) category, the RNA binding term is the most significant term (with the lowest P-value) and contains the largest number of proteins (32.1% of the 77 proteins assigned to MF category). In the cellular component (CC) category, the term with the lowest P-value was the non-membrane-bounded organelle, and that with the second lowest P-value was the ribonucleoprotein complex term. Additionally, a large number of MAEL-associated proteins were enriched in the nuclear lumen, consisting of 23 proteins (29.5% of the 70 proteins assigned to CC category). In the biological process (BP) category, the top three terms with the lowest P-value were RNA process, chromatin assembly and macromolecular complex assembly. When considering all three categories, we found that most MAEL-associated proteins were RNA-binding proteins existing in the ribonucleoprotein complex.

MAEL protein is a component of nuage (3,6,7), a germline-unique ribonucleoprotein complex particle. According to the above results (Table II), MAEL associates with RNA-binding proteins, possibly a component of the ribonucleoprotein complex in cancer cells. However, we did not find other nuage components in the MAEL-associated proteins. Notably, the MAEL-associated proteins contained 14 components of SGs: PABPC1 (25), FUBP1/FBP1 (26), KHSRP/FBP2 (26), YBX1/YB-1 (27), SYNCRIP/hnRNP Q (28), EIF3F (29), EIF4A1 (30), EIF4B (31), ELAVL1 (32), HNRNPA1 (33), HNRNPA2B1 (34), DDX1 (27), DDX3X (35) and DDX39 (36). Among these, 8 proteins (PABPC1, FUBP1, KHSRP, YBX1, SYNCRIP, EIF4B, HNRNPA1 and HNRNPA2B1) were identified from both MDA-MB-231 and SW480 cells (Table III); 6 proteins were found only in MDA-MB-231 (EIF4A1, EIF3F and DDX3X) or SW480 cells (DDX39, ELAVL1 and DDX1) (data not shown).

Then, 8 (PABPC1, KHSRP, YBX1, SYNCRIP, EIF4A1, DDX39, ELAVL1 and EIF3F) of the identified SG proteins were selected for further confirmation of their interactions with MAEL protein by co-IP experiments. The coding regions of these 8 genes were cloned into mammalian expression vector pCMV-Myc for constructing Myc-tagged expression plasmids. The Myc-tagged expression plasmid of each SG protein was co-transfected with HA-tagged MAEL expression plasmid into Hek293 cells, respectively. The anti-tag immunoprecipitation assays showed that all these 8 SG proteins were able to interact with the MAEL protein (Fig. 2).

As stated above, MAEL protein interacts with SG proteins. Therefore, we performed immunofluorescence analysis to investigate whether the MAEL protein localizes in SGs. In the untreated SW480 cells, MAEL proteins were distributed homogeneously in both the cytoplasm and the nucleus (Fig. 3). After induction with H₂O₂, some cytoplasmic MAEL proteins displayed a distribution pattern of distinct speckles, and co-localization with SG marker PABPC1 (25) in the speckle (Fig. 3), suggesting that MAEL proteins localize to SGs. Strangely, after treatment with H₂O₂, PABPC1 did not condense into apparent speckles as it did in the cells treated with arsenite (37).