• Open Access

Identification of Annexin A1 interacting proteins in chronic myeloid leukemia KCL22 cells

Authors

  • Irene Colavita,

    1. CEINGE-Biotecnologie Avanzate scarl, Napoli, Italy
    2. Fondazione SDN-IRCCS, Napoli, Italy
    Search for more papers by this author
  • Nicola Esposito,

    1. CEINGE-Biotecnologie Avanzate scarl, Napoli, Italy
    2. Dipartimento di Medicina Molecolare e Biotecnologie Mediche, Università di Napoli “Federico II”, Napoli, Italy
    Search for more papers by this author
  • Concetta Quintarelli,

    1. CEINGE-Biotecnologie Avanzate scarl, Napoli, Italy
    2. Dipartimento di Medicina Molecolare e Biotecnologie Mediche, Università di Napoli “Federico II”, Napoli, Italy
    Search for more papers by this author
  • Ersilia Nigro,

    1. CEINGE-Biotecnologie Avanzate scarl, Napoli, Italy
    2. Fondazione SDN-IRCCS, Napoli, Italy
    Search for more papers by this author
  • Fabrizio Pane,

    1. CEINGE-Biotecnologie Avanzate scarl, Napoli, Italy
    2. Dipartimento di Medicina Molecolare e Biotecnologie Mediche, Università di Napoli “Federico II”, Napoli, Italy
    Search for more papers by this author
  • Margherita Ruoppolo,

    1. CEINGE-Biotecnologie Avanzate scarl, Napoli, Italy
    2. Dipartimento di Medicina Molecolare e Biotecnologie Mediche, Università di Napoli “Federico II”, Napoli, Italy
    Search for more papers by this author
  • Francesco Salvatore

    Corresponding author
    1. Fondazione SDN-IRCCS, Napoli, Italy
    • CEINGE-Biotecnologie Avanzate scarl, Napoli, Italy
    Search for more papers by this author

Correspondence: Professor Francesco Salvatore, CEINGE–Biotecnologie Avanzate Scarl c/o Università di Napoli “Federico II,” Via S. Pansini 5, 80131 Naples, Italy

E-mail: salvator@unina.it

Fax: +0039-081-7463650

Abstract

In the present study, we used a functional proteomic approach to identify Annexin A1 (Anxa1) interacting proteins in the Philadelphia-positive KCL22 cell line. We focused on Anxa1 because it is one of the major proteins upregulated in imatinib-sensitive KCL22S cells versus imatinib-resistant KCL22R. Our proteomic strategy revealed 21 interactors. Bioinformatic analysis showed that most of these proteins are involved in cell death processes. Among the proteins identified, we studied the interaction of Anxa1 with two phosphatases, Shp1 and Shp2, which were recently identified as biomarkers of imatinib sensitivity in patients affected by chronic myeloid leukemia. Our data open new perspectives in the search for annexin-mediated signaling pathways and may shed light on mechanisms of resistance to imatinib that are unrelated to Bcr-Abl activity. All mass spectrometry data have been deposited in the ProteomeXchange with identifier PXD000030.

Abbreviations
Anxa1

annexin A1

CML

chronic myeloid leukemia

Chronic myeloid leukemia (CML) therapy has changed dramatically in the last decade thanks to the advent of tyrosine kinase inhibitors, namely, imatinib, nilotinib, and dasatinib [1]. Notwithstanding the significant prolongation of overall survival of CML patients, there is still room for improvement. Approximately 20–25% of patients, initially treated with imatinib, will need alternative therapy due to drug resistance, which often results from the emergence of clones expressing mutant forms of breakpoint cluster region-abelson (Bcr-Abl) [2]. However, imatinib resistance may also be related to Bcr-Abl activity-independent mechanisms [3, 4], which have yet to be elucidated [5]. In this context, we previously described novel Bcr-Abl activity-independent mechanisms that involve the phosphatases Shp1 and Shp2 [6, 7].

In our previous studies, based on DIGE plus MS, we characterized proteins whose expression differed between the imatinib-resistant KCL22R and imatinib-sensitive KCL22S cell lines. The analysis showed that Annexin A1 (Anxa1) is upregulated in KCL22S versus KCL22R cells [6]. In the present study, we used a functional proteomic approach to identify Anxa1-interacting proteins in KCL22 cells in the attempt to identify annexin-mediated signaling pathways in the Philadelphia-positive (Ph+) CML cell line KCL22 (Fig. 1). We showed that Anxa1 expression was unchanged when KCL22S or KCL22R cells were treated with imatinib (Fig. 2), in contrast to K562 cells [8], which are intrinsically more sensitive than KCL22 cells to imatinib [9].

Figure 1.

Separation of Annexin A1-interacting proteins in KCL22S cells. The gel was stained with CBB, and bands were excised and subjected to LC–MS/MS analysis. An irrelevant rabbit IgG was used as control. The bands of the interacting proteins lane, relative to the proteins identified by LC–MS/MS analysis, are indicated by annotations as reported in the figure and in Table 1.

Figure 2.

Western blot analysis of KCL22R and KCL22S protein lysates in the presence or absence of 5 μM imatinib (A and B). Protein lysates were used as input for the immunoprecipitation experiments with an anti-Annexin A1 (C). Proteins were separated on 10% SDS-PAGE and immunoblotted with antibodies against pBcr-Abl, Bcr-Abl, Shp2, phospho Shp2 (pTyr542), Dbc1, Shp1, Hsp70, and Anxa1. Gapdh and Shp2 served as loading control (A and B). The densitometric analysis was performed on three independent experiments; the results are shown as means ± SD (D and E).

Anxa1 is a member of a superfamily of closely related calcium- and membrane-binding proteins that have a wide variety of cellular functions including vesicle trafficking, cell division, apoptosis, calcium signaling, and growth regulation [10, 11]. Anxa1 is implicated in apoptosis induction [12], caspase-3 activation [13], and cell growth inhibition [14]. In addition, extracellular Anxa1 regulation of cell surface is mediated by signaling through the formyl peptide receptor signaling [15, 16]. Moreover, Anxa1 has also been associated with resistance of human solid cancers (breast, ovarian, and lung) and leukemias to several chemotherapeutic drugs [17-19]. In particular, of the 21 interacting proteins identified in this study, we investigated the interaction of Anxa1 with two phosphatases, Shp1 and Shp2, recently found to be biomarkers of imatinib sensitivity in patients affected by CML [7].

KCL22S and KCL22R cells were grown in RPMI 1640 medium (Gibco, Paisley, UK) containing 10% FBS, 1 mM l-glutamine, 100 U/mL penicillin, and 50 μg/mL streptomycin at 37°C in an atmosphere saturated with 5% CO2. Imatinib mesylate (5 μM) was added to the KCL22R cells.

Briefly, for the immunoprecipitation of endogenous Anxa1 in KCL22S cells (Fig. 1), which are characterized by a higher level of Anxa1 expression compared with KCL22R cells (Fig. 2), 2 × 107 cells were washed with cold phosphate-buffered saline three times and centrifuged at 1000 rpm for 10 min. The pellet was incubated with 1.5 mL of cold lysis buffer containing Tris-HCl 50 mM pH 7.5, 150 mM NaCl, 1mM NaF, 1 mM PMSF, 1% Nonidet P-40, 1 mM EDTA, 1 mM sodium orthovandate, and protease inhibitor cocktail (Complete mini EDTA-free, Roche Applied Science) for 30 min on ice and then cleared by centrifugation at 15 000 × g for 20 min at 4°C. To prevent nonspecific binding of proteins to Protein A/G-PLUS-Agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and to remove proteins that nonspecifically bind immunoglobulins, we first incubated lysates with the appropriate control, normal mouse IgG (Santa Cruz Biotechnology), which corresponds to the host species of the primary antibody used for immunoprecipitation, for 2 h at 4°C on a rocker platform. The solution was then incubated with Protein A/G-PLUS-Agarose beads for 1 h at 4°C on a rocker platform. Beads were subsequently centrifuged at 8000 × g for 5 min at 4°C. The supernatant was incubated with primary monoclonal Anxa1 antibody (Santa Cruz Biotechnology) overnight at 4°C on a rocker platform and then incubated with fresh Protein A/G-PLUS-Agarose beads for 5 h at 4°C. Beads were then centrifuged at 8000 × g for 5 min at 4°C. After extensive washing of the pellet beads with IP buffer (Tris-HCl 50 mM pH 7.5, NaCl 150 mM, NaF 1 mM, PMSF 1 mM, Nonidet P-40 1%, 1 mM EDTA, 1 mM sodium orthovandate, and protease inhibitor cocktail (Complete mini EDTA-free, Roche Applied Science), the resulting immune complexes were eluted from the beads with 2× electrophoresis sample buffer at 90°C for 10 min. The protein mixture was then resolved by electrophoresis on 10% SDS/polyacrylamide gels and stained with CBB (Pierce Rockford, Illinois, USA) (Fig. 1). Each lane of the gel was cut into pieces and processed for LC–MS/MS analysis.

Gel pieces were placed in Eppendorf tubes and washed with 50 mM ammonium bicarbonate for 10 min, and destained twice with 50 mM ammonium bicarbonate followed by acetonitrile for 15 min. The gel pieces were dehydrated by incubation with acetonitrile for 15 min and vacuum-dried in a SpeedVac. The dried gel pieces were treated with proteomics grade trypsin (Sigma-Aldrich) at 37°C overnight. The samples were briefly spun, and supernatants were transferred to clean dry Eppendorf tubes. Peptides were further extracted from the gel once with 0.1% trifluoroacetic acid in 50% acetonitrile for 10 min followed by acetonitrile for 20 min at 37°C. Pooled supernatants were dried in a SpeedVac. Before LC–MS/MS analysis, the samples were reconstituted with 10 μL of 0.2% formic acid; then injection was performed with the LC/MSD Trap XCT Ultra system equipped with a 1100 HPLC-Chip Cube interface (Agilent Technologies, Palo Alto, CA, USA). We concentrated and washed the peptide mixture at 4 μL/min in a 40-nL enrichment column (Agilent Technologies chip); 0.1% formic acid served as eluent. We then fractionated the sample on a C18 reverse phase capillary column (75 μm × 43 mm) using a flow rate of 200 nL/min and a linear gradient of 5–60% acetonitrile in 2% formic acid. Subsequently, we analyzed the peptides using data-dependent acquisition of one MS scan (mass range from 400 to 2000 m/z) followed by MS/MS scans of the three most abundant ions. To obtain a more detailed survey of the peptides, we carried out dynamic exclusion using a procedure of automatic recognition and temporary exclusion (2 min) of ions from which definitive mass spectral data had previously been acquired.

For data analysis, we used the MASCOT software (version 2.4) Peptide Mass Fingerprinting search program (http://www.matrixscience.com) selecting NCBInr aug2010 database (11592891 sequences) (http://www.ncbi. nlm.nih.gov), and the following six parameters: specificity of the proteolytic enzyme used for hydrolysis (trypsin); up to 1 missed cleavage; cysteines such as S-carbamidomethylcysteines; unmodified N- and C-terminal ends; unmodified and oxidized methionines; putative Gln-induced pyroGlu formation; a precursor peptide maximum mass tolerance of 400 ppm, and a maximum fragment mass tolerance of 0.6 Da. Using the probability-based Mowse score, the ion score is −10 × Log(p), p being the probability that an eventual match is a random occurrence. All MS/MS spectra with a MASCOT score of or higher than 41 (p ≤ 0.05) had a good S/N, so the interpretation of the data was unambiguous. MS/MS spectra of peptides that had a MASCOT score below 41 were inspected manually and included in the statistics only in the presence of at least four continuous y or b ions. We also used the BLAST program (http://ncbi.nlm.nih.gov/blast) and the NCBI database to search for the peptide sequence. Peptides with an ambiguous identification were removed from the tables; specifically, the candidate protein was deleted from the list when it matched other proteins. Protein species identified by a single peptide were examined further. We manually reconstructed the peptide sequence stretch, and analyzed it and the precursor ion mass with the MASCOT software set in the sequence query mode. Supporting Information Fig. 1 shows the full MS and the MS/MS scans (annotated with masses and fragment assignments) of the ubiquitin carboxyl-terminal hydrolase CYLD isoform 1 protein, which was identified by a single peptide. We excluded nonspecific proteins found in control samples from the list of identified proteins. The detailed peptide data for identified proteins are listed in Supporting Information Table 1. All the available proteomic data are deposited in the PRIDE proteomics identification database and ProteomeXchange (accession numbers 27246–27249, dataset identifier PXD000030).

For Western blot analysis, proteins were first resolved on a 10% SDS-PAGE gel, and then transferred onto nitrocellulose membranes (GE Healthcare) by Mini Trans-Blot electrophoretic transfer (Bio-rad). Membranes were blocked using 5% nonfat milk in PBS pH 7.5 for 2 h, and then immunoblotted with antibodies against Dbc1 (1:500), Shp2 (1:1000), phospho Shp2 (pTyr542) (1:1000), Shp1 (1:1000), Hsp70 (1:1000) (all from Santa Cruz Biotechnology, Heidelberg, Germany), Anxa1 (1:5000) (BD Biosciences, Erembodegem, Belgium), Bcr-Abl (2.5 μg/mL) (Calbiochem, Nottingham, UK), and pBcr-Abl (1:1000) (Cell Signaling Technology Danvers, MA, USA). For immunoblot detection, we used HRP-conjugated anti-mouse (1:5000) or anti-rabbit (1:10000) secondary antibodies (GE Healthcare). Immunoblots were detected by chemiluminescence (ECL-Advance Western Blotting Detection kit; GE Healthcare). The Western blot images obtained were scanned by PDquest 7.1 software (Bio-Rad). The protein band images on X-ray films were acquired with the Chemidoc XRS system (Bio-Rad). Shp2 and Gapdh were used as loading control since they are equally expressed in all the cell lines considered [6, 7]. We used the Quantity One 4.5 tool (Bio-Rad) for densitometric measurements.

We measured pBcr-Abl, Bcr-Abl, pShp2(Tyr542), and Shp2 expression by Western blot analysis in KCL22R and KCL22S cells in the presence or absence of 5 μM imatinib. As shown in Fig. 2A, imatinib (5 μM for 24 h) inhibited Bcr-Abl in both sensitive and resistant KCL22 cells. Evaluation of Shp-2 phosphorylation at Tyr542 showed no difference between KCL22R and KCL22S cell lines whereas it was decreased in imatinib-treated KCL22S cells (Fig. 2A). These findings suggest that the KCL22R cell line lacks factors that are required to downregulate Shp2-activating signals after imatinib treatment.

In summary, using LC–MS/MS and Western blotting, we found 21 Anxa1 interacting proteins (Table 1 and Fig. 2B and C). The proteins identified were analyzed with the Ingenuity Pathway Analysis software 9.0 (IPA) (Ingenuity Systems, Inc. www.ingenuity.com). This system produces a series of networks constituted by a maximum of 70 genes or gene products. Each of these networks is assigned a score that is computed based on the fit of the user's set of focus genes/gene sproducts with all the genes/gene products stored in the knowledge database. Finally, biological functions are assigned to each network. The system created four protein networks. The network with the highest score, i.e., 38, which includes 15 of the 21 proteins we identified, is related to cell death, cellular growth, and proliferation (Supporting Information Fig. 2).

Table 1. Annexin A1 interacting proteins in KCL22S cells identified by LC–MS/MS
BandsProteinGene IDa
  1. a

    Accession numbers are from NCBI database.

A2Prepro-alpha1(I) collagen1418928
C1p30 DBC protein193788217
C3Ubiquitin carboxyl-terminal hydrolase CYLD isoform 114165258
C5DOC-2 (disabled homolog-2)1063686
C5DEAH (Asp-Glu-Ala-His) box polypeptide 152696613
D1TLE3 protein27469815
D3WD repeat-containing protein C2orf4413376798
D3Poly(A) binding protein II119612225
D3DEAD (Asp-Glu-Ala-Asp) box polypeptide 416118555
D4Heat shock protein 1A4529893
D4RNA binding motif protein 145454064
D4DEAD (Asp-Glu-Ala-Asp) box helicase 17397501955
D5Dihydrolipoamide S-acetyltransferase119587578
E2Pyruvate kinase35505
F1Flotillin 294538362
F2Actin, gamma 1 propeptide4501887
F2hnRNP G protein (RBMX)3256007
G1Ribosomal protein P04506667
I5Ribosomal protein S145032051

We verified the proteomic analysis by examining the interaction of Anxa1 with proteins known to be involved in signaling transduction (Shp1 and Shp2) [20], apoptosis (Hsp70) [21], and tumorigenesis (Dbc1) [22]. To this aim, we performed three independent immunoprecipitation experiments using an antibody against Anxa1 on KCL22S and KCL22R cellular lysates. We found that Anxa1 co-immunoprecipitates with all the investigated proteins in both KCL22S and KCL22R cell lines in an imatinib-independent manner (Fig. 2C). Moreover, we observed that imatinib treatment may affect Anxa1/Hsp70 interactions in the KCL22 imatinib-sensitive cell line (Fig. 2C and E). We also evaluated the interaction of Anxa1 with the previously selected proteins in the imatinib-resistant model of KCL22 cells. We observed that Dbc1, Shp2, and Hsp70 interacted with Anxa1 in KCL22R cells, irrespective of imatinib treatment, whereas the interaction between Anxa1 and Shp1 was not detectable in KCL22R cells, probably due to very low Shp1 expression in resistant cells, as shown in Fig. 2B and D.

In this study, we have identified new Anxa1-interacting proteins involved in cell death, cellular growth and proliferation. Interestingly, one of these proteins, namely Hsp70, belongs to a group of stress response and chaperone proteins, whose reduced expression in KCL22R cells could be associated to imatinib resistance [6]. Importantly, we also observed another Anxa1 partner in KCL22 cells, namely Dbc1, which is a protein involved in cell proliferation, apoptosis, and histone modification, all processes crucial in the regulation of tumorigenesis [22].

Our data also demonstrate that protein Anxa1 interacts with key elements that are involved in imatinib resistance. In this context, our group previously demonstrated that reduced Shp1 expression in KCL22R cells contributes to continuous Shp2 activation thereby sustaining a Bcr-Abl activity-independent pathway of survival and resistance to imatinib treatment [6, 7]. We also showed that low Shp1 levels are associated with a reduced response to or failure of imatinib treatment in patients affected by CML. Our current results suggest that Anxa1 may be a new actor in the regulation of imatinib resistance in concert with Shp1 and Shp2. These data could be of relevance in studies on the search for new biomarkers implicated in the modulation of imatinib resistance in CML.

Acknowledgments

The mass spectrometry proteomics data in this paper have been deposited in the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository [23]: dataset identifier PXD000030. All the available data sets are also deposited in the PRIDE proteomics identification database (accession numbers 27250–27265 and 27246–27249). The authors thank Jean Ann Gilder (Scientific Communication srl., Naples, Italy) for writing assistance. In addition, we are very grateful to Prof. Junia V. Melo (Institute of Medical & Veterinary Science, Adelaide, Australia) for providing KCL22S and KCL22R cell lines.

This work was supported by the Ministero della Salute (Roma), Convenzione CEINGE-MIUR (2000) art 5.2 (to F.S.), grants from Regione Campania (Convenzione CEINGE-Regione Campania, G.R. 20/12/2004 n.2495 and LR n.5/2002 year 2005), from a CEINGE-Regione Campania Contract [DGRC 1901/2009] (to F.S), Progetto S.co.Pe, Centro di eccellenza riconosciuto dal MIUR ex dm 11/2000, Prin 2006 to M.R and from the Istituto di Ricovero e Cura a carattere scientifico (IRCCS)-Fondazione SDN, Naples, Italy (to F.S.)

The authors have declared no conflict of interest.

Ancillary