Identification of DOCK4 and its splicing variant as PIP3 binding proteins

Authors

  • Akinori Kanai,

    1. Division of Applied Biological Chemistry, Graduate School of Agricultural and Life Sciences, University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
    Current affiliation:
    1. Department of Medical Genome Sciences, Graduate School of Frontier Sciences, University of Tokyo, Kashiwa, Chiba 277-8562, Japan
    Search for more papers by this author
  • Sayoko Ihara,

    1. Division of Applied Biological Chemistry, Graduate School of Agricultural and Life Sciences, University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
    Search for more papers by this author
  • Tsutomu Ohdaira,

    1. Division of Applied Biological Chemistry, Graduate School of Agricultural and Life Sciences, University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
    Search for more papers by this author
  • Azusa Shinohara-Kanda,

    1. Protein-Research Network, Inc., 1-13-5 Fukuura, Kanazawa-ku, Yokohama-shi, Kanagawa 236-0004, Japan
    Search for more papers by this author
  • Akihiro Iwamatsu,

    1. Protein-Research Network, Inc., 1-13-5 Fukuura, Kanazawa-ku, Yokohama-shi, Kanagawa 236-0004, Japan
    Search for more papers by this author
  • Yasuhisa Fukui

    Corresponding author
    1. Division of Applied Biological Chemistry, Graduate School of Agricultural and Life Sciences, University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan
    • Laboratory of signal transduction, Hoshi University, 2-4-41 Ebara, Shinagawa-ku, Tokyo 152-8501, Japan
    Search for more papers by this author
    • Tel: +81-3-5498-6394. Fax: +81-3-5498-6394.


Abstract

DOCK4, a member of DOCK180 family proteins, was originally identified as a product of a gene deleted during tumor progression. Although its tumor suppression properties have been reported, the regulation mechanism of this protein has not been fully elucidated. DOCK4 shares two conserved domains called as DHR-1 and DHR-2 domain as other members including DOCK180. Although DHR-1 in DOCK180 is reported to bind to PIP3, whether that of DOCK4 exhibits similar function has yet not been examined. In a search for novel PIP3 binding proteins by the PIP3 analog beads binding assay, we found that DOCK4 and its novel splicing variant, whose exon1 and exon52 are different from the known one, bind to PIP3. Binding assay using deletion mutants of DOCK4 revealed that the binding region falls into the DHR-1 domain. These results raise the possibility that DOCK4 may be regulated by PIP3 to exert its function. © 2008 IUBMB IUBMB Life, 60(7): 467–472, 2008

INTRODUCTION

DOCK4, originally identified as a product of a gene which is deleted during tumor progression1, is a member of DOCK180 family proteins sharing 35% amino acid identity to DOCK180. The DOCK180 family proteins are relatively large-sized proteins, which share DOCK homology region (DHR)-1 and DHR-22. Several members of them were identified as a guanine nucleotide exchange factor (GEF) for Rho family GTPases2–4. DOCK180 was the first member of this family and shown to have critical roles in processes such as myoblast fusion, cell migration, and phagocytosis5–9. DOCK180 has been suggested to function as a GEF for Rac, a member of Rho family GTPases2, 3. On the other hand, DHR-1 domain of DOCK180 has been suggested to bind to PIP310. DOCK180 with substitution of PIP3 binding-PH domain for DHR-1 domain was reported to be functional, suggesting that the main function of DHR-1 domain is to bind to PIP3.

Although intensive studies were reported about DOCK180, comparatively little is known about DOCK4. Mutations in DOCK4 were found in human cancer cell lines, and it has been reported that DOCK4 exhibits tumor suppressor properties in vitro and in vivo1. In their study, DOCK4 has been demonstrated to activate small GTPase Rap1 and enhances the formation of adherence junction. In contrast, recent studies show that DOCK4 functions as an activator for Rac through its DHR-2 domain11, 12 and is involved in cell migration12. These contradictory results might be caused by complicated regulatory system of DOCK4, however, the mechanism has not been fully explored.

PIP3 is the product of PI-3 kinase, which is activated upon stimulation by a variety of growth factors and extracellular matrixes (ECMs), and plays key roles in various cellular functions13. Many downstream factors of PIP3 have been identified as PIP3 binding proteins, so far. Although DOCK180 has been suggested to bind to PIP3 through its DHR-1 domain, it remains to be open whether DOCK4 binds to PIP3 or not. Considering the fact that DOCK4 exhibits not necessarily the same GEF property as DOCK180 although they share DHR-2, it is possible that the role of the DHR-1 could be different.

In searching for PIP3 binding protein from lysate of glioblastoma cell line, we identified DOCK4 and its novel splicing variant as target proteins. In this article, we report novel splicing variants of DOCK4 and their binding properties for PIP3.

MATERIALS AND METHODS

Cell Culture and Transfection

SF295 and 293T cells were maintained in Dulbecco's modified eagle medium supplemented 10% calf serum under a 5% CO2 atmosphere. Transfection into 293T cells was done by the calcium phosphate precipitation as described before14.

MALDI TOF MASS analysis

Proteins eluted from PIP3 beads were subjected to SDS-PAGE and blotted on PVDF membrane. Each bands detected by CBB staining were excised and subjected to a Lys-C digestion and analyzed by MALDI-TOF-MASS.

Cloning of DOCK4 and its Variant

To obtain cDNA fragment common for DOCK4s, overlapping two fragments were amplified by PCR using primer sets, 5′-aaggcgatctcttccaccgg-3′, 5′-gccaggcaaact ctgcacca-3′ and 5′-gaccgaccatttcacaaagg-3′, 5′-ctttttcaggtgcagagagatt-3′ and joined utilizing restriction enzyme site. cDNA fragments unique to DOCK4 was obtained from SF295 cDNA by PCR reaction. N-terminal fragment was amplified with 5′-ccgtcgacatgtggatacc tacggagcacg-3′ used as a sense primer and 5′-tgttcc atcaatcggtagagc-3′ as an antisense primer. For C-terminal fragment, 5′-gaccgaccatttcacaaagg-3′ was used as a sense primer and 5′-cctctagattataactgaga gaccttgcgggg-3′ as an antisense primer. These fragments were joined to the rest of the clone utilizing the restriction enzyme site. To obtain cDNA fragment unique to DOCK4 variant, 3′ RACE was performed as described15 by use of primers 5′-gaccgaccatttcacaaagg-3′ and 5′-gactcg agtcgacatcgattttttttttttttttt-3′ (dT primer) for 1st PCR and 5′-caagttctgctccatcgagt-3′ and dT primer for 2nd PCR. Sequence for cDNA 5′ end was predicted by searching putative exon upstream of the known sequence in the genome. RT-PCR for confirmation of the sequence was performed with 5′-cggaactgttttaataattttgtgacc-3′ and 5′-ccgtcgaccatgaaagctggtcaagcagtc-3′ as sense primers for 1st PCR and 2nd PCR, respectively, and 5′-tgttccatcaatcggtagagc-3′ as an antisense primer for both PCR reactions.

RT-PCR Analysis for the DOCK Family mRNAs

RT-PCR for the DOCK family mRNAs was performed with primers, 5′-acgctcggtcaggatcttcatg-3′ and 5′-ctg accctgaagtaccccattg-3′ for β-actin, 5′-ggtgatgt tcgaaatgatatctatg-3′ and 5′-ctccatcatgatgtt gaagagg-3′ for DOCK180, 5′-ccgagagatgcctgact ttg-3′ and 5′-gaccttcttccgtgtgagtg-3′ for DOCK2, 5′-tcacatgcgtctagtgaagc-3′ and 5′-ctgagtgagt ggaactcagg-3′ for DOCK3, 5′-caagttctgctccatcg agt-3′ and 5′-ctttttcaggtgcagagagatttgg-3′ for DOCK4, 5′-tgctcagtggcatcgtggac-3′ and 5′-gctca accttctcctggtcttcag-3′ for DOCK5, 5′-aggaacc tgctgtacgtgtacc-3′ and 5′-gaccagacgcacgagct tgtc-3′ for DOCK6, 5′-gaacaccaggaggatcctg-3′ and 5′-gtctgcgggatcctgatgtgttgc-3′ for DOCK7, 5′-atgaggttcagttttggag-3′ and 5′-ttagctgccct gtgacaactg-3′ for DOCK8, 5′-aggtcaatgctggcc cactag-3′ and 5′-tcatcccagctgctcatgc-3′ for DOCK9, 5′-tggagtaccaggaagaactg-3′ and 5′-tcagactt cagcactagatg-3′ for DOCK10, 5′-ttgccttttatg gccaatct-3′ and 5′-gagcatattttggatcaagc-3′ for DOCK11.

Plasmids

cDNA for DOCK4 and its variant were cloned into pEGFPC1 (Clontech, Palo Alto, CA). Deletion mutants were generated by use of restriction enzyme sites. Expression vectors for myc-BD110 and myc-BDKN were described before16.

PIP3 Bead Binding Assay

A PIP3 binding assay was done as described previously17. The bound proteins were subjected to SDS-PAGE followed by Western blotting with an anti-GFP antibody (Santa Cruz Biotechnology, Santa Cruz, CA).

Confocal Microscopic Analysis

293T cells expressing GFP-DOCK4/DOCK4 (171-677) together with myc-BD110/BDKN were fixed with 3.7% formaldehyde for 5 min at room temperature, followed by treatment with 0.1% Triton X-100 in PBS for 5 min. Myc-BD110 and myc-BDKN was stained with Cy3-conjugated anti-myc antibody (Sigma, St. Louis, MO). Cells were analyzed by confocal fluorescence microscopy (Fluoview 300, Olympus).

RESULTS

Screening of PIP3 Binding Protein

As previously reported, we established PIP3 beads binding assay using PIP3 analog beads17. This method enabled us to screen PIP3 binding proteins in the lysates from various kinds of cell lines. When a cell lysate from a glioblastoma cell line, SF295, was used as a source, several bands were detected as candidates for the PIP3 binding proteins (Fig. 1A, left lane). When the cell lysate was incubated with PIP3 prior to binding to PIP3 beads, two bands became markedly faint (Fig. 1A, right lane), suggesting that they may bind to PIP3 specifically. These bands were excised and subjected to mass spectrometry analysis. Lower band revealed to be DOCK180. As for the upper band, two clones, AAB 83942.1 and AAB 83946.1 were suggested to be the corresponding genes. Because full-length cDNA clone covering them was not found at the time when the analysis was done, we attempted to obtain 5′ and 3′ terminal sequences by prediction from a genome database and 3′ RACE-PCR, respectively. 5′ sequence was obtained by RT-PCR by use of sense primer coding predicted 5′ sequence and anti-sense primer coding sequence found in AAB 83942.1. As shown in Fig. 1B, a clear band with the length of about 600 bp was detected specifically when cDNA of SF295 cells was used. cDNA of HeLa cells did not give such a band. Although this study was underway, another full-length cDNA covering the AAB 83942.1 and AAB 83946.1 was reported to be DOCK4. Comparison of the sequence we determined with that of DOCK4 revealed that they were only different in the first and the last exons (Fig. 1C), suggesting that our clone, designated as DOCK4-1, may be a splicing variant of DOCK4. In our assay, we detected only DOCK4 and DOCK180 as PIP3 binding proteins among 11 members of the DOCK family proteins. To examine whether the other members were also expressed in SF295 cell line, RT-PCR analysis for all the members of the DOCK family was performed with cDNAs synthesized using RNA from HeLa and SF295 cells. As shown in Fig. 1D, the specific bands were detected in any pairs of primers in the PCR analysis of the SF295 cDNA, suggesting that all the members of the DOCK family were expressed. This result implicates that DOCK4 and DOCK180 may have stronger affinity to PIP3 than the other members of the DOCK family.

Figure 1.

Identification of PIP3 binding proteins. (A) PIP3 binding assay was performed with cell lysate of SF295 using affinity beads bearing PIP3 analog. Prior to binding to the beads, cell lysate was preincubated with (+) or without (−) 4 μM of free PIP3. The proteins which bind to PIP3 specifically should be missing in the lane with preincubation compared with the one without preincubation. The proteins bound to the beads were eluted and detected by SDS-PAGE followed by silver staining. Two bands indicated by arrowheads were subjected to TOF-MASS analysis. (B) RT-PCR was performed by use of cDNA derived from HeLa cells or SF295 cells with a sense primer coding predicted 5′ sequence and an anti-sense primer coding sequence found in AAB 83942.1. The band indicated by arrowhead was cloned and subjected to sequence analysis. (C) Top panel: Comparison of the structures of cDNA clones for DOCK4 and DOCK4-1, which we cloned. Each exons were indicated by black boxes with exon numbers. Positions of primers used in the PCR reaction to clone DOCK4-1 were indicated by the arrows. Bottom panel: Sequences of exon 1′ and 52′ were shown. (D) Expression of the indicated mRNAs of the DOCK family in HeLa (H) and SF295 (S) cells was analyzed by RT-PCR. All the bands except for the case of DOCK9 corresponded to the predicted size of cDNA fragments. As for DOCK9, the specific bands given by PCR were larger than the predicted size, implicating that the DOCK9 gene was spliced differently from the way reported by others in these cells. β-actin mRNA, positive control.

DOCK4 and its Variant Bind to PIP3

To confirm the binding ability of DOCK4 and its variant to PIP3, cell lysates from 293T cells expressing them as GFP fusion proteins were subjected to the PIP3 beads binding assay. As shown in Fig. 2, preincubation with PIP3, but not with PI3,4P2 or PI4,5P2 abolished the binding of these proteins to PIP3 beads. These results suggest that DOCK4s bind to PIP3 specifically.

Figure 2.

DOCK4 and its splicing variant bind to PIP3 specifically. 293T cells were transfected with GFP-DOCK4 or GFP-DOCK4-1. The lysate were preincubated with or without (indicated as “none”) 4 μM of PIP3, PI3,4P2, PI4,5P2 prior to binding to the PIP3 beads to examine the specificity of the binding. The proteins were detected by Western blotting with an anti-GFP antibody.

DOCK4 Binds to PIP3 Through its DHR1 Domain

To determine the region responsible for the binding to the PIP3 analog beads, various types of deletion mutants were generated (Fig. 3A) and expressed as GFP fusion proteins in 293T cells. The cell lysates were subjected to the PIP3 binding assay. As shown in Fig. 3B, the mutants except for 1-334 lacking DHR-1 and the following C-terminal region exhibited the binding activity to the PIP3 beads. To determine the region required for the PIP3 binding more precisely, deletion mutants around DHR-1 region were examined (Fig. 3C). As shown in Fig. 3D, small fragments, 525–677, 617–677, 171–525, and 171–424, gave no signal or faint signal in the PIP3 binding assay. In contrast, larger fragments covering almost entire DHR-1 region were detected in the PIP3 beads bound fraction. These results suggest that PIP3 binding activity of DOCK4 may be mediated by the DHR-1 region. We then explored the effect of the binding of DOCK4 to PIP3in vivo. 293T cells were transfected with expression vectors for GFP-DOCK4 (171–677) covering entire region of DHR-1 and myc-BD110, an constitutively active form of PI-3 kinase16. As shown in Fig. 3E, clear membrane localization of GFP-DOCK4 (171–677) was observed in the myc-BD110 expressed cells but not in myc-BDKN, a kinase negative mutant of BD110, expressed cells by confocal microscopic analysis. These results suggest the possibility that DOCK4 could be regulated downstream of PI-3 kinase. Full length DOCK4 failed to translocate to the membrane even co-expressed with myc-BD110 (data not shown). The failure to see the effect of the binding of DOCK4 to PIP3 may be due to the fact that DOCK4 is also regulated by other factors such as RhoG12.

Figure 3.

DOCK4 binds to PIP3 through its DHR-1 domain (A) The structure of DOCK4 and its truncation mutants are shown. All constructs were tagged with GFP at the N-termini. (B) Lystate of 293T cells expressing mutant DOCK4s shown in (A) was subjected to the PIP3 binding assay. The proteins were detected by Western blotting with an anti-GFP antibody. (C) The structures of truncation mutants for the analysis of the PIP3 binding activity of the DHR-1 domains are shown. All constructs were tagged with GFP at the N-termini. (D) The same experiment was done as in (B) using constructs in (C). (E) Myc-BD110 or myc-BDKN was expressed in 293T cells together with GFP-DOCK4 (171–677). Expression of myc-BD110 or myc-BDKN was detected by immunostaining with anti-myc antibody. Localization of GFP-DOCK4 (171–677) was observed by confocal microscopy. Images shown were the xy sections dissected at the middle of the cells. Bar, 10 μm.

DISCUSSION

In this study, we identified DOCK4 as a PIP3 binding protein with use of PIP3 analog bead binding assay, which we established before. In this system, we detected many proteins as specific PIP3 binding proteins including centaurin α/PIP3BP18, SWAP-7019, DOCK18020, so far. In SF295 cell lysates, the band of DOCK4 was detected in the PIP3 bead-bound fraction clearly as well as that of DOCK180 (Fig. 1A), implicating that DOCK4 may exhibit the similar binding affinity for PIP3 as DOCK180. The previous study suggested that DHR-1 domain in DOCK180 family proteins may be conserved domain for PIP3 binding2. However, only the lipid binding ability of DOCK3, 6, 9 was referred without the description about the preference for phospholipids in the report and there was no information about DOCK4. Our study provided the first evidence for their prediction, showing that DOCK4 had binding activity specific for PIP3 clearly.

DOCK4 has been shown to exhibit tumor suppression properties in vitro and in vivo1. Despite its potentially significant role in tumor progression, the regulation mechanism has not been fully examined. In this regard, our study suggested the possible mechanism that DOCK4 may be regulated by PIP3. The activity of many other PIP3 binding proteins is regulated by PIP3, which caused the conformational change of the protein into active form21 or translocation of the protein to the proper position13. How PIP3 affects the activity of DOCK4 is an attractive problem for future studies.

We identified a novel putative splicing variant of DOCK4 from SF295 cell lysates with different exons at the first and last. The alternative first exon found in DOCK4-1 may suggest that the expression of DOCK4-1 is regulated by a different promoter from that of DOCK4. These isoforms may also have functionally different roles. Although SH3 domain was conserved in DOCK4-1, part of amino acid sequence of the domain coded by the first exon was different from that of DOCK4, suggesting the possibility that these SH3 domains may have different functions. So far, another splicing variant of DOCK4, the exon 49th of which is lacked has been reported11. Although its DHR-2 domain was identical to that of DOCK4, the variant exhibited stronger activity of Rac GEF compared with the case of DOCK4. The finding of these isoforms of DOCK4 including DOCK4-1 may suggest the diverged roles of the DOCK4s.

Acknowledgements

This work was supported by grants-in-aid for YF and SI from the Ministry of Education, Culture, Sports, Science, and Technology, and by a grant to YF from the National Institute of Biomedical Innovation, Japan.

Ancillary