Interaction of the growth and tumour suppressor NORE1A with microtubules is not required for its growth-suppressive function
© Moshnikova et al; licensee BioMed Central Ltd. 2008
Received: 18 April 2008
Accepted: 15 May 2008
Published: 15 May 2008
The NORE1 protein was identified in a yeast two-hybrid screen as a Ras effector that binds Ras protein in a GTP-dependent manner. NORE1A is a growth and tumour suppressor that is inactivated in a variety of cancers. In transformed human cells, both full-length NORE1A protein and its effector domain alone (amino acids 191–363) are localized to microtubules and centrosomes. However, the mechanism by which NORE1A associates with these cytoskeletal elements is not known; furthermore, whether centrosomally-associated or microtubule-associated NORE1A suppresses tumour cell growth has not been yet established.
We have shown that purified NORE1A fails to bind to microtubules in vitro suggesting that other protein(s) mediate NORE1A-microtubule association. Using mass-spectrometry, we identified the Microtubule-Associated Protein 1B (MAP1B) and its homologue C19ORF5 as NORE1A interaction partners. Suppression of C19ORF5 expression by RNA interference (RNAi) and immunodepletion of C19ORF5 protein from cell extracts showed that binding of NORE1A to microtubules is not dependent on C19ORF5. Conversely, RNAi suppression of MAP1B revealed that MAP1B is required for association of NORE1A with microtubules. RNAi-mediated depletion of C19ORF5 or MAP1B did not prevent centrosomal localization of NORE1A. Moreover, the depletion of C19ORF5 or MAP1B did not prevent NORE1A's ability to suppress tumour cell growth.
The interaction of NORE1A with microtubules is mediated by MAP1B, but not C19ORF5 protein. Interaction of NORE1A with centrosomes is not dependent on C19ORF5 or MAP1B, and appears to involve a different mechanism independent of binding to microtubules. The NORE1A microtubular localization is not required for growth suppression.
NORE1A interacts with C19ORF5 and MAP1B proteins
C19ORF5 interacts with NORE1A but does not mediate its binding to microtubules
Both full-length NORE1A and its effector domain alone are capable of interacting with endogenous C19ORF5 protein in HEK293 cells (Fig. 1A). The N-terminal portion (amino acids 1–268) and the C-terminal portion (amino acids 250–416) of NORE1A failed to interact with C19ORF5 (Fig. 1A), suggesting that the intact effector domain, amino acids 191–363, mediates NORE1A-C19ORF5 interaction.
To map the NORE1A-C19ORF5 interaction interface more precisely, we made several small deletions in the NORE1A effector domain and tested these proteins for interaction with C19ORF5. Figure 1C shows that the deletion of amino acids 227–253 completely abolished NORE1A-C19ORF5 interaction. The smaller deletion mutant, NORE1A Δ236–253, also failed to interact with C19ORF5; however, the deletion mutant NORE1A Δ246–253 retained binding activity. Thus, the region 227–246 within the NORE1A effector domain is responsible for interaction with the C19ORF5 protein.
MAP1B interacts with NORE1A and is required for its binding to microtubules
MAP1B is a large (ca. 2500 amino acids) microtubule-associated protein consisting of several domains. Figures 1B and 1D show that NORE1A and the effector domain can interact with endogenous MAP1B protein in HEK293 and A549 cells, respectively. Furthermore, interaction required the intact NORE1A effector domain (amino acids 191–363) since neither the N-terminal portion (aa 1–268) or the C-terminal portion (aa 250–416) of NORE1A could interact with MAP1B (Fig. 1B).
To determine whether MAP1B protein is required for targeting of NORE1A to microtubules in living cells, we examined the distribution of GFP-tagged NORE1A or its effector domain in A549 cells in which MAP1B has been depleted by RNAi. Depletion of MAP1B resulted in loss of microtubule localization of full-length NORE1A (Figure 4C) and its effector domain, aa 191–363 (See additional file 1: Supplemental Figure 2). Probing of A549 cell lysates where MAP1B has been depleted for NORE1A indicated that the NORE1A level was not changed, ruling out the effect of MAP1B on NORE1A stability (Fig. 4A, lower panel). Thus, our results indicate that the MAP1B protein is essential for interaction of NORE1A with microtubules in vitro and in vivo.
The reason why binding of NORE1A to microtubules is mediated by MAP1B, but not by C19ORF5 is not clear at the present time. For example, it is possible that the C19ORF5 protein mediates NORE1A interaction with microtubules under specific conditions or in specific cell types not tested here.
The effect of MAP1B or C19ORF5 on centrosomal localization of NORE1A
We showed previously that NORE1A is localized to centrosomes in addition to microtubules . This localization was observed for endogenous NORE1A, and for ectopically expressed NORE1A and its effector domain. Song and co-workers showed that the C19ORF5 protein was present on centrosomes in interphase and mitotic HeLa cells , while a recent report from the Latif group described centrosomal localization for C19ORF5 only in mitotic cells . The MAP1B protein was also detected on centrosomes in interphase and mitotic cells [4, 5]. We tested whether C19ORF5 or MAP1B play a role in the centrosomal localization of NORE1A.
In A549 cells into which NORE1A was introduced by retroviral-mediated gene transfer, depletion of the C19ORF5 protein did not result in dissociation of NORE1A from centrosomes in interphase (Fig. 3B, middle row) or mitotic (Fig. 3B, bottom row) cells. Depletion of MAP1B protein induced dissociation of full-length NORE1A (Fig. 4C, arrows) and its effector domain (See additional file 1: Supplemental Figure 3) from microtubules, but not from centrosomes. Thus, centrosomal localization of NORE1A is independent of its microtubular localization and does not require C19ORF5 or MAP1B proteins.
In contrast to NORE1A, the centrosomal localization of the closely related tumour suppressor RASSF1A was found to be dependent on the C19ORF5 protein , possibly reflecting differences in the regulation or mechanisms of action of NORE1A and RASSF1A. Alternatively, this difference could be caused by the use of different cell systems: HeLa cells in ref. , compared with A549 cells in this study.
MAP1B or C19ORF5 are not required for NORE1A to suppress tumour cell growth
Our previous work identified endogenous NORE1A as centrosomal protein in nontransformed human lung cells and as a centrosomal and microtubular protein in transformed lung cells . However, this study shows that the microtubular NORE1A association is not required for its growth-suppressive activities. This raise the possibility that it is centrosomal NORE1A that is important for growth suppression. Further studies are required to determine the mechanism of NORE1A centrosomal localization and the importance of its targeting to centrosomes for suppression of tumour cell growth.
Pools of small-interfering RNA duplexes, cat.# L-016881-00 against C19ORF5 and cat.# L-010348-00 against MAP1B were from Dharmacon, Inc. (Lafayette, CO). The Alexa Fluor 488-labelled AllStars Negative control siRNA, cat # 1027284, were from Qiagen (Valencia, CA). RNA duplex labelled with CY3 dye, cat # 4621, were from Ambion (Austin, TX). The anti-C19ORF5 4G1 antibody were from A&G Pharmaceutical, Inc. (Columbia, MD). The monoclonal anti-NORE1A antibody 10F10 was described in ref . Anti-pericentrin antibody, cat.# ab-4448, were from Abcam (Cambridge, MA), the anti-α-tubulin DM-1-A and anti-γ-tubulin GTU-88 antibodies were from Sigma (St Louis, MO), and the anti-MAP1B antibody, cat # 612679 was from BD Biosciences. Secondary antibody conjugates were from Jackson ImmunoResearch (West Grove, PA). All other reagents were previously described by Moshnikova et al. .
Plasmids encoding NORE1A were described previously . The human C19ORF5 protein (IMAGE clone; ID 5925583) was purchased from ATCC (Manassas, VA) and subcloned into pCMV5 vector with the N-terminal FLAG tag. This and all other constructs were prepared using standard molecular biology techniques  and verified by sequencing.
Cell Lines, Transfection and Retrovirus Production
Cultivation of A549 human lung adenocarcinoma and HEK293 human embryonic kidney cells (ATCC), HEK293 cells transfection and retroviral supernatants production was described previously [1, 7]. The siRNA duplexes were transfected using DharmaFECT1 reagent (Dharmacon) according to the manufacturer's instructions.
Immunodepletion, Immunoprecipitation and Western Blot Analysis
HEK293 cells were lysed in buffer A (30 mM HEPES, pH 7.4, 20 mM KCl, 1mM NaVO4, 20mM NaF, 20 mM β-glycerophosphate, 7.5 mM MgCl2, 1% Triton X-100, 2 mM EGTA, 0.1% 2-mercaptoethanol and protease inhibitors) and the C19ORF5 protein was immunodepleted by overnight incubation with 4G1 antibody adsorbed on Protein A/G+ beads. The beads were pelleted and the supernatant was subjected to a second immunodepletion as described above. For mock immunodepletion, normal mouse IgG was used instead of anti-C19ORF5 antibody. Other procedures were performed as described previously [1, 7].
Immunofluorescent Microscopy and cell preparation
were performed as described previously [1, 7]. In some cases, cells were transfected with siRNA duplexes prior to plating on coverslips. Forty-eight hours after plating, cells were fixed in 100% methanol at -20°C for 5 minutes and processed for immunostaining as described . Digital images were taken under UV illumination with appropriate filter sets at 100× magnification using a LEICA DMIRE2 microscope and captured using IPLab 3.6.5a software (Scanalytics, Rockville, MD). To allow comparison of image intensity within each experiment, images for GFP-tagged NORE1A and α-tubulin were taken at the same exposure and processed identically in the Adobe Photoshop program.
Tubulin-Binding (microtubule cosedimentation in vitro) Assays
Microtubule cosedimentation assay was done with the Microtubule Binding Protein Spin Down Assay Kit (Cytoskeleton Inc. Denver, CO, cat.# BK029), according to the manufacturer's instructions. After assay, the pellet was adjusted to the same volume as the supernatant with GTB. Both supernatant and pellet were adjusted to 1 × SDS gel loading buffer with 4× stock solution. Equal volumes of supernatant and pellet were separated by SDS PAGE and analyzed by western blotting. In some experiments, purified FLAG-NORE1A was preincubated for 1 hour at 37°C with C19ORF5-immunodepleted or mock-immunodepleted HEK293 cell extract, or purified FLAG-C19ORF5 prior to incubation with microtubules.
Cell Cycle Distribution Assay
A549 cells were transfected with anti-MAP1B siRNA pool, or anti-C19ORF5 siRNA pool or control siRNA as descried above. After one day cells were replated in triplicate on 6-well plates and 8 hours after were infected with retrovirus expressing GFP or GFP-NORE1A effector domain (aa 191–363). Two days later, cells were fixed with 1% formaldehyde in PBS, stained with 0.05 mg/ml propidium iodide and analyzed by FACSvantage (Becton-Dickinson, San Jose, CA). Cell cycle distributions of GFP-positive cells were determined by ModFIT software.
- The abbreviations are:
NORE1, novel Ras effector 1
Ras-association domain family 1 isoform A
peptide tag DYKDDDDK
Renilla reniformis Green Fluorescent Protein
general tubulin buffer
bovine serum albumin
We thank Dr. Friedrich Propst (University of Vienna, Austria) for providing MAP1B plasmids and Charles Lyons for assistance with Mass Spectrometry. This work was supported by University of Virginia Cancer Centre Grant GF10574 and by a grant from the Wendy Will Case Cancer Fund to A.V.K.
- Moshnikova A, Frye J, Shay JW, Minna JD, Khokhlatchev AV: The growth and tumor suppressor NORE1A is a cytoskeletal protein that suppresses growth by inhibition of the ERK pathway. J Biol Chem. 2006, 281: 8143-8152. 10.1074/jbc.M511837200.View ArticlePubMedGoogle Scholar
- Song MS, Chang JS, Song SJ, Yang TH, Lee H, Lim DS: The centrosomal protein RAS association domain family protein 1A (RASSF1A)-binding protein 1 regulates mitotic progression by recruiting RASSF1A to spindle poles. J Biol Chem. 2005, 280: 3920-3927. 10.1074/jbc.M409115200.View ArticlePubMedGoogle Scholar
- Dallol A, Cooper WN, Al-Mulla F, Agathanggelou A, Maher ER, Latif F: Depletion of the Ras association domain family 1, isoform A-associated novel microtubule-associated protein, C19ORF5/MAP1S, causes mitotic abnormalities. Cancer Res. 2007, 67: 492-500. 10.1158/0008-5472.CAN-06-3604.View ArticlePubMedGoogle Scholar
- Dominguez JE, Buendia B, Lopez-Otin C, Antony C, Karsenti E, Avila J: A protein related to brain microtubule-associated protein MAP1B is a component of the mammalian centrosome. J Cell Sci. 1994, 107 ( Pt 2): 601-611.Google Scholar
- Vandre DD, Feng Y, Ding M: Cell cycle-dependent phosphorylation of centrosomes: localization of phosphopeptide specific antibodies to the centrosome. Microsc Res Tech. 2000, 49: 458-466. 10.1002/(SICI)1097-0029(20000601)49:5<458::AID-JEMT8>3.0.CO;2-#.View ArticlePubMedGoogle Scholar
- Sambrook J and Russell, D.W.: Molecular Cloning: A Laboratory Manual. 2001, Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press, Third EditionGoogle Scholar
- Khokhlatchev A, Rabizadeh S, Xavier R, Nedwidek M, Chen T, Zhang XF, Seed B, Avruch J: Identification of a novel Ras-regulated proapoptotic pathway. Curr Biol. 2002, 12: 253-265. 10.1016/S0960-9822(02)00683-8.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.