JBC Origene Your Gene Company

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M305209200 on June 14, 2003

J. Biol. Chem., Vol. 278, Issue 36, 34445-34450, September 5, 2003
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
278/36/34445    most recent
M305209200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hanada, T.
Right arrow Articles by Chishti, A. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hanada, T.
Right arrow Articles by Chishti, A. H.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Protein 4.1-mediated Membrane Targeting of Human Discs Large in Epithelial Cells*

Toshihiko Hanada {ddagger}, Atsuko Takeuchi, Gautam Sondarva and Athar H. Chishti §

From the Departments of Medicine, Anatomy, and Cellular Biology, St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston, Massachusetts 02135

Received for publication, May 18, 2003 , and in revised form, June 4, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Human discs large (hDlg) protein binds to protein 4.1R via a motif encoded by an alternatively spliced exon located between the SH3 and the C-terminal guanylate kinase-like domains. To evaluate the functional significance of protein 4.1R binding for subcellular localization of hDlg in vivo, we expressed full-length recombinant constructs of two naturally occurring isoforms of hDlg termed hDlg-I2 and hDlg-I3. The hDlg-I3 but not the hDlg-I2 isoform binds to the FERM (Four.1-Ezrin-Radixin-Moesin) domain of protein 4.1R in vitro. Upon transient transfection into subconfluent Madine-Darby canine kidney (MDCK) epithelial cells, the hDlg-I3 fused with the green fluorescent protein accumulated predominantly at the plasma membrane of cell-cell contact sites, whereas the hDlg-I2 fusion protein distributed in the cytoplasm. In contrast, in stably transfected confluent MDCK cells, both hDlg-I2 and -I3 isoforms localized efficiently to the lateral membrane, consistent with the previous notion that the N-terminal domain of hDlg mediates its membrane targeting in polarized epithelial cells. We introduced a double mutation (I38A/I40A) into the N-terminal domain of hDlg, which disrupted its interaction with DLG2, a key event in the membrane targeting of hDlg. Interestingly, the hDlg-I2 isoform harboring the I38A/I40A mutation mislocalized from the membrane into cytoplasm. Importantly, the hDlg-I3 isoform with the same mutation localized efficiently to the membrane of confluent MDCK cells. Together, our results demonstrate that in addition to the N-terminal targeting domain, the alternatively spliced I3 insertion plays a critical role in recruiting hDlg to the lateral membrane in epithelial cells via its interaction with protein 4.1R.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Mutations in the Drosophila lethal (1) discs-large (dlg) tumor suppressor gene produce a neoplastic overgrowth phenotype in the imaginal discs, which give rise to adult epithelial structures and central nervous system (1). Dlg is a multidomain scaffolding protein composed of three PSD-95-Discs large-ZO-1 (PDZ)1 domains, an src homology 3 (SH3) domain, and a C-terminal guanylate kinase-like (GUK) domain (2, 3). Dlg is a membrane-associated guanylate kinase homologue (MAGUK) and is required for controlling growth and polarity of epithelial cells (4, 5) as well as asymmetric cell division of neuroblasts (6, 7). The human homologue of Dlg, termed hDlg, or its rat counterpart SAP97 is highly enriched in the epithelial cell-cell junction sites and is likely to play a role in epithelial polarity similar to that of its Drosophila counterpart (8, 9).

Alternative splicing in two regions within hDlg produces multiple isoforms (8, 10, 11). Region 1, which is located within the N-terminal region, contains two alternatively spliced exons termed I1A and I1B, and region 2, located between the SH3 and GUK domains, contains four alternatively spliced exons named I2, I3, I4, and I5 (11). The region between the SH3 and GUK domains, corresponding to region 2, also termed the HOOK domain in Drosophila Dlg, is critical for efficient membrane targeting of Dlg in epithelial cells presumably by mediating protein-protein interactions with the membrane cytoskeleton (12). The association of MAGUKs with the cortical actin cytoskeleton has been characterized previously using the erythroid p55 MAGUK, of which the HOOK domain binds to the FERM domain of protein 4.1R (13, 14). The sequence of the p55 HOOK domain, including its characteristic cluster of basic amino acid residues, is conserved in a subset of MAGUKs (13), and in the case of CASK (15, 16) and hDlg (8, 17), a similar mode of protein 4.1 binding has been established. The protein 4.1 binding site of hDlg is encoded by an alternatively spliced exon termed I3 (8, 17), raising the possibility that alternative splicing creates hDlg isoforms with distinct functional properties.

Although previous attempts have been made to elucidate the contribution of protein 4.1 binding to the membrane targeting of hDlg/SAP97 in epithelial cells, the role of the I3 insert remains controversial. Lue et al. (18) reported the presence of a second protein 4.1 binding site within the PDZ1 and -2 domains, and both the I3 insert and the PDZ1-PDZ2 module were found to be sufficient for lateral membrane targeting of hDlg in permeabilized epithelial cells. However, using transient transfection of truncated cDNA constructs, Wu et al. (19) demonstrated that the N-terminal 1–65 amino acid segment of SAP97 contains the membrane-targeting signal and that the PDZ domains and the I3 insert are neither necessary nor sufficient for membrane targeting. It was later discovered that the N-terminal domain of hDlg/SAP97, named the MRE (MAGUK recruitment) domain, binds to the L27 (Lin-2 Lin-7) domain (20) of CASK, and this interaction mediates the membrane targeting of hDlg/SAP97 in epithelial cells (21). The MRE domain of hDlg/SAP97 also binds to DLG2 and DLG3, two MAGUK proteins closely related to CASK, in their conserved L27 domains (22). MRE domains are found in the splice variants of Drosophila Dlg (23) and the mammalian PSD-95 (24) as well as in a non-MAGUK PDZ protein termed the PALS1-associated tight junction protein (PATJ) (25), and these domains appear to play a fundamental role in the membrane targeting of scaffolding proteins. In summary, the precise role of the protein 4.1 binding region and HOOK domain in the membrane targeting of MAGUKs remains uncertain.

Another controversial issue pertains to the nuclear localization of hDlg. MAGUK proteins including CASK (26) and ZO-1 (27) are known to translocate to the nucleus and contribute to the transcriptional regulation of specific genes. In the case of hDlg, it has been shown that the truncated constructs of hDlg/SAP97 spanning the SH3-HOOK-GUK domains tend to accumulate in the nuclei of transfected cells (19, 28). In addition, the cluster of basic amino acid residues present in the I3 insert is believed to function as a nuclear localization signal (28). In contrast to this observation, the immunocytochemistry studies using I2- and I3-specific antibodies detected I2 insert-specific antigen in the nucleus and I3 insert-specific antigen in the plasma membrane, thus implicating the I2 insert in the nuclear targeting function (11).

Here, we have expressed two naturally occurring isoforms of hDlg as full-length polypeptides to elucidate the function of I2 and I3 inserts in vitro and in vivo. Under our experimental conditions, we did not observe any significant nuclear localization of either I2- or I3-containing isoforms of hDlg. Our data also indicate that the I3 insert is the sole protein 4.1 binding site within hDlg, and in addition to the MRE domain, the I3 insert is also sufficient for recruiting the full-length hDlg to the lateral membrane in polarized epithelial cells. These results implicate a fundamental role of the conserved HOOK domain of MAGUKs in mediating direct binding to the FERM domain of protein 4.1, thus regulating the membrane-targeting properties of the scaffolding complex in epithelial cells.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Generation of Expression Constructs and Production of Recombinant Proteins—Full-length hDlg cDNA clone hDlg-I2 contains I1A-I1B-I2-I5, and the hDlg-I3 construct contains I1A-I1B-I3-I5 inserts. Both hDlg-I2 and -I3 clones carry KpnI and SpeI adapters in their 5' and 3' ends, respectively (8), making them identical except for the I2 and I3 inserts. The GST-hDlg-I2 and -I3 constructs were produced by cloning the respective cDNAs into a modified pGEX2T vector (17) at KpnI and SpeI sites. The N-terminal half of DLG2 (1–221), which contains two L27 domains and a PDZ domain, was PCR-amplified from mouse DLG2 cDNA clone (GenBankTM/EBI accession number AF162685 [GenBank] ) with EcoRI adapters on both ends and cloned into the EcoRI site of pGEX2T plasmid. To generate the GFP-hDlg-I2 and -I3 constructs, the respective inserts were subcloned into pEGFP-C1 vector (Clontech) using KpnI and XbaI sites of the vector. The FERM domain of human protein 4.1R, which starts from the second ATG (nucleotide 801) in exon 4 to phenylalanine in exon 12 (nucleotide 1694), was PCR-amplified from the protein 4.1 cDNA (29) and cloned into pCITE-2a (Novagen) using EcoRI and NotI adapters.

Site-directed Mutagenesis—The double mutation (I38A/I40A) in the GFP-full-length hDlg constructs was introduced using the QuikChange site-directed mutagenesis kit (Stratagene). Primers used were 5'-GTTCCATAGAACGGGTTGCTAACGCATTTCAGAGCAACCTC (sense) and 5'-GAGGTTGCTCTGAAATGCGTTAGCAACCCGTTCTATGGAAC (antisense).

In Vitro Protein 4.1 Binding Assay—The FERM domain of human protein 4.1R was expressed using the rabbit reticulocyte lysate in vitro transcription and translation system STP3 (Novagen) in the presence of [35S]methionine (Amersham Biosciences). Binding of radiolabeled FERM domain to the GST fusion proteins was assessed by GST pulldown assay. After the completion of protein synthesis, lysate was diluted 10-fold in the binding buffer (phosphate-buffered saline with 1% Triton X-100), and the respective GST fusion proteins were added to each tube followed by incubation for 2 h at 4 °C. Beads were recovered by centrifugation and washed, and bound radiolabeled protein was analyzed by SDS-PAGE. The gel was stained by Coomassie Blue and exposed to x-ray film for fluorography at –70 °C.

DNA Transfection and Fluorescence Microscopy—MDCK cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. For immunofluorescence analysis of subconfluent MDCK cells, the MDCK cells were plated on glass coverslips 1 day before transfection to attain a density of ~40% at the time of transfection. Transient transfections were carried out using LipofectAMINE 2000 (Invitrogen). Twenty-four h after transfection, the cells were fixed with 4% paraformaldehyde for 10 min at room temperature, and the GFP signal was visualized by fluorescence microscopy. All the GFP-positive cells were counted and classified as M (membrane), C (cytosol), or M/C (membrane and cytosol) to distinguish the localization of GFP-hDlg protein in transfected cells. Reported data were collected from two independent experiments. For analysis of confluent polarized MDCK cells, stably transfected MDCK cells expressing the GFP-hDlg proteins were generated. After the transfection, the cells were selected by treating with 0.6 mg/ml of G418 (Invitrogen) for more than 2 weeks. After the transfected cells reached confluency and polarized morphology, they were fixed and imaged by fluorescence microscopy. Each experiment was repeated three times with identical results.

GST Pulldown Assay—COS-7 cells were plated on 10-cm dishes and transiently transfected with respective GFP-hDlg expression plasmids. Cell lysates were prepared 30 h after transfection using the lysis buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.1% Triton X-100, 1.0 mM phenylmethanesulfonyl fluoride). The GST-DLG2 (1–221) or control GST protein immobilized on glutathione-Sepharose 4B beads was added to the lysate and incubated overnight at 4 °C. Beads were recovered by centrifugation and washed, and bound proteins were processed for Western blotting using our anti-hDlg monoclonal antibody (2D11).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
The I3 Insertion of hDlg Contains the Sole Binding Site for Protein 4.1R—To assess the functional diversity of hDlg isoforms, we produced expression constructs of two naturally occurring full-length hDlg isoforms termed hDlg-I2 and -I3 (Fig. 1A). Other than the alternatively spliced exon encoding I2 and I3 insertions in the HOOK region, these two constructs have identical domain composition and amino acid sequence. To our knowledge, this is the first comparison of the I2 and I3 inserts in the context of full-length hDlg polypeptides. Both hDlg isoforms were expressed in bacteria as GST fusion proteins and used for in vitro protein 4.1 binding using a GST pulldown assay. The 35S-labeled FERM domain of protein 4.1R, which was expressed using the rabbit reticulocyte in vitro transcription-translation system, bound specifically to the GST-hDlg-I3 beads but not to the GST-hDlg-I2 or control GST beads (Fig. 1B). This result is consistent with our previous finding that the I3 insert is the only protein 4.1 binding site in hDlg (17). However, our results do not support the presence of the second high affinity protein 4.1 binding site located within the PDZ1-PDZ2 conformational unit as suggested previously (18).



View larger version (46K):
[in this window]
[in a new window]
  		FIG. 1.
Binding of hDlg isoforms to the FERM domain of human protein 4.1R. A, a schematic diagram of hDlg isoforms used in the binding experiments. Full-length constructs of hDlg containing either the I2 or I3 insert in the HOOK region were designated as hDlg-I2 and hDlg-I3, respectively. B, binding of hDlg isoforms with protein 4.1R. [35S]Methionine-labeled FERM domain of human protein 4.1R was expressed using an in vitro translation system. The lysate was incubated with GST-hDlg-I2, GST-hDlg-I3, or GST protein bound to glutathione-Sepharose beads. The beads were washed extensively, and bound proteins were analyzed by SDS-PAGE and fluorography. Coomassie-stained gel shows a comparative amount of the GST fusion proteins used in the binding assay. Fluorography shows that the FERM domain of protein 4.1R binds specifically to the GST-hDlg-I3 but not with the GST-hDlg-I2 or GST proteins.

 

I3 Insertion Mediates Membrane Targeting of hDlg in Subconfluent Epithelial Cells—Next we examined whether the isoform-specific binding of hDlg to protein 4.1 affects the subcellular localization of hDlg in epithelial cells. Protein 4.1R concentrates at the lateral membrane of epithelial cells (30), where the endogenous hDlg is also found (8, 9). Therefore, by virtue of its binding to protein 4.1R, the I3 insertion may provide a mechanism for the localization and stabilization of hDlg at the lateral plasma membrane. We transiently transfected GFP-fused hDlg-I2 and -I3 constructs in subconfluent MDCK cells, and GFP signal was observed by fluorescent microscopy after 24 h when cells were still subconfluent and did not display polarized morphology. The expression of the right size hDlg fusion proteins was confirmed by Western blot analysis in separate experiments (data not shown). We found that the GFP-hDlg-I2 isoform distributed diffusely throughout the cytoplasm with no protein accumulation detectable at the plasma membrane (Fig. 2A). On the other hand, the GFP-hDlg-I3 protein accumulated significantly at the plasma membrane of cell-cell contact sites (Fig. 2B). We introduced a semiquantitative criterion to evaluate the efficiency of hDlg localization in transiently transfected MDCK cells. We counted transfected cells that showed either an exclusive membrane localization (designated as M) or cytoplasmic localization along with significant concentration of the signal at the plasma membrane (designated as M/C), or exclusive localization in the cytoplasm without any membrane signal (designated as C). For example, the I2- and I3-expressing cells shown in Fig. 2, A and B, are classified as C and M/C, respectively. The GFP signal was visualized 24 h after transfection of MDCK cells under subconfluent culture conditions. The results show that the GFP-hDlg-I3 isoform is located predominantly at the plasma membrane with some signal detectable in the cytoplasm (Fig. 2C). In contrast, the GFP-hDlg-I2 isoform is found mostly in the cytoplasm. These observations suggest that the I3 insertion plays an important role in mediating the targeting of hDlg to the plasma membrane during the formation of initial cell-cell contact sites by non-polarized epithelial cells. It is noteworthy here that the cytoplasmic GFP-hDlg-I2 isoform partially migrates to the plasma membrane when transiently transfected MDCK cells are examined after 48 h post-transfection under conditions where the epithelial cells presumably begin to form mature cell-cell contact sites (data not shown).



View larger version (77K):
[in this window]
[in a new window]
  		FIG. 2.
Localization of GFP-hDlg isoforms in subconfluent MDCK epithelial cells. The GFP-hDlg-I2 and -I3 were transiently transfected in MDCK cells. Cells were harvested 24 h after transfection, fixed, and examined by fluorescence microscopy for GFP signal. A, the GFP-hDlg-I2 isoform is distributed diffusely in the cytoplasm without any enrichment at the plasma membrane. B, the GFP-hDlg-I3 isoform accumulates at the plasma membrane particularly at sites where cells form contacts. C, semiquantification of GFP-hDlg distribution in MDCK cells. Based on the localization of the GFP signal, the GFP-positive cells were classified either as M (exclusive membrane localization at the cell-cell contact sites), M/C (significant membrane localization as well as cytoplasmic staining), or C (diffuse cytoplasmic staining without any accumulation at the membrane). Statistical data represent mean and standard deviation from two independent experiments with a minimum of 80 cells counted for each sample.

 

Two Distinct Domains Are Responsible for Membrane Targeting of hDlg in Confluent MDCK Cells—To determine the subcellular localization of hDlg isoforms in MDCK cells under confluent culture conditions, we generated stable cell lines by selecting transfected cells using G418. When transfected cells became confluent and displayed polarized morphology, they were fixed, and the GFP signal was recorded. Under these conditions, both I2 and I3 isoforms of hDlg localized efficiently to the lateral membrane (Fig. 3, A and B). This result suggests that an I3-independent targeting mechanism exists that mediates membrane targeting of hDlg in confluent MDCK cells and is consistent with previous reports demonstrating that the N-terminal MRE domain recruits hDlg/SAP97 to the plasma membrane of polarized epithelial cells (19, 21).



View larger version (81K):
[in this window]
[in a new window]
  		FIG. 3.
Localization of GFP-hDlg isoforms in confluent MDCK cells. MDCK cells expressing (A) GFP-hDlg-I2, (B) GFP-hDlg-I3, (C) GFP-hDlg-I2 mutant (I38A/I40A), and (D) GFP-hDlg-I3 mutant (I38A/I40A) were selected with G418 to obtain stably transfected cells. After reaching confluency, cells were fixed on the coverslips, and the GFP signal was detected by fluorescence microscopy. E, schematic representation of the location of putative NES consensus within amino acids 30–40 of hDlg. The design of the combined mutation to disrupt the putative NES function (I38A/I40A) is shown above the hDlg schematic.

 

Using isoform-specific polyclonal antibodies, it was recently shown that the I2 isoform-specific antibodies detected hDlg in the nucleus, whereas the I3 isoform-specific antibodies detected the signal at the lateral plasma membrane (11). Occasionally, we have also observed the presence of hDlg isoforms in the nucleus (note the presence of GFP-hDlg-I3 in one of the transfected cells shown in Fig. 2B). However, the nuclear localization of hDlg was not isoform-specific under our experimental conditions. We therefore speculated that there might be an unidentified mechanism that specifically regulates the nuclear localization of hDlg protein in epithelial cells. While searching for potential nuclear targeting and/or nuclear export motifs in the hDlg coding sequence, we noticed a sequence, LRSSI-ERVINI40, which matches with the nuclear export signal (NES) consensus LX2–3LX2–3LXL, where L can also be either Val or Ile (Fig. 3E) (31, 32). Interestingly, this putative NES motif is located within the N-terminal MRE domain of hDlg/SAP97 that is reported to mediate the membrane targeting of hDlg/SAP97 in polarized epithelial cells (19, 21). We hypothesized that this putative NES motif regulates the nuclear localization cycle of hDlg, and therefore its disruption might result in the accumulation of hDlg inside the nucleus. To test this hypothesis, we mutated two of the conserved isoleucines within the NES-like motif of hDlg to alanines (I38A/I40A), a design consistent with the published mutagenesis approach that disrupts the NES activity in other proteins (32–34). The mutant hDlg isoforms were then tested for their localization in confluent MDCK cells. Contrary to our expectations, the I2 and I3 isoforms of hDlg carrying the double I38A/I40A mutation did not accumulate inside the nucleus (Fig. 3, C and D). Interestingly, however, the I2 isoform with the NES mutation almost completely lost its ability to localize to the lateral membrane and distributed diffusely in the cytoplasm (Fig. 3C). Although the precise relationship between the cytoplasmic localization of hDlg and the function of its putative NES sequence is not yet clear, it is conceivable that the double mutation I38A/I40A unexpectedly disrupted a key protein-protein interaction(s). Indeed, the N-terminal MRE domain of hDlg is known to interact with the CASK family of MAGUKs, and this interaction is necessary for targeting of hDlg to the lateral membrane (21). Importantly, the I3 isoform of hDlg with the I38A/I40A mutation localized efficiently to the lateral membrane of MDCK cells (Fig. 3D), suggesting that the I3 insertion is sufficient for membrane targeting in the absence of a functional N-terminal MRE domain. Together, our results indicate that the I3 insertion alone can mediate the membrane targeting of full-length hDlg in confluent epithelial cells.

Double Mutation I38A/I40A of hDlg Disrupts Its Interaction with DLG2—To test whether the combined I38A/I40A mutation interferes with functional protein-protein interaction(s) of the MRE domain, we examined whether the I38A/I40A mutation of hDlg disrupts its interaction with the CASK family of proteins. The L27 domain is a conserved feature of the CASK subfamily of the MAGUKs, including the DLG2 and -3, which are known to interact with hDlg/SAP97 via a similar mechanism (Fig. 4A) (22). The N-terminal segment of human DLG2 (amino acids 1–221) containing two L27 domains and a PDZ domain was expressed as a GST fusion protein (Fig. 4B), and its interaction with hDlg was measured by a GST pulldown assay. Consistent with a previous report (22), the DLG2 specifically interacted with GFP-hDlg-I2 expressed in the COS-7 cells (Fig. 4C, upper panel). In contrast, the GFP-hDlg-I2 (I38A/I40A) mutant protein did not bind to DLG2 (Fig. 4C, upper panel). An identical result was obtained with the GFP-hDlg-I3 (I38A/I40A) mutant protein (Fig. 4C, lower panel). Together, these results demonstrate that the combined I38A/I40A mutation introduced in the N terminus of hDlg completely abrogates its interaction with the CASK family of MAGUKs. This fortuitous finding begins to offer a molecular framework for the membrane targeting of hDlg in polarized epithelial cells and underscores the importance of the I3 insertion as an alternate membrane-targeting motif that functions independently of the N-terminal domain of hDlg in polarized epithelia.



View larger version (34K):
[in this window]
[in a new window]
  		FIG. 4.
Combined I38A/I40A mutation disrupts hDlg interaction with DLG2. A, schematic representation of CASK family of MAGUKs that bind to the MRE domain of hDlg. Conserved L27 domains mediate the interaction. B, GST-DLG2 (1–221) and control GST proteins used for the hDlg binding experiment. C, binding of hDlg with GST-DLG2 (1–221). Lysates were prepared from COS-7 cells expressing GFP-hDlg-I2 and the GFP-hDlg-I2 mutant (I38A/I40A) (upper panel) and GFP-hDlg-I3 and the GFP-hDlg-I3 mutant (I38A/I40A) (lower panel). The GST pulldown experiment was performed using GST-DLG2 (1–221) and control GST proteins immobilized to beads. Bound hDlg proteins were resolved by SDS-PAGE, transferred to the nitrocellulose membrane, and detected by Western blotting using the anti-hDlg monoclonal 2D11. Input represents 4% of each lysate used for the pulldown experiment.

 


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The data presented in this manuscript provide evidence for the following. (1) Alternatively spliced I3 insert in the HOOK region is the sole protein 4.1 binding site in hDlg; (2) the I3 insert mediates recruitment of hDlg to the plasma membrane at an early stage of cell-cell contact formation in subconfluent epithelial cells; and (3) the I3 insert is sufficient to recruit hDlg to the lateral membrane in polarized epithelial cells by a mechanism that is independent of the membrane-targeting function of the N-terminal MRE domain. These results are summarized in Fig. 5. Alternative splicing produces the I2 and I3 hDlg isoforms that are believed to encode distinct functions. At the initial stage of cell contact formation, only the I3 isoform is recruited to the membrane, presumably via its interaction with protein 4.1R. At the later stages when epithelial cells become polarized and form mature junctional structures, both isoforms of hDlg are recruited to the lateral membrane via an interaction with their N-terminal MRE domain to CASK or related family members. The I3 insert alone is also sufficient to recruit hDlg to the lateral membrane, and this additional binding site might contribute to the formation of a higher order scaffolding complex in the mature cell junctions.



View larger version (36K):
[in this window]
[in a new window]
  		FIG. 5.
A proposed model for isoform-specific targeting of hDlg in epithelial cells. In non-polarized epithelial cells, such as subconfluent MDCK cells, the I3 insert directs membrane targeting of hDlg by interacting with protein 4.1. The I2 isoform of hDlg, which does not bind to protein 4.1, stays in the cytoplasm. In polarized epithelial cells, another MAGUK protein of the CASK subfamily provides the binding site for the N-terminal (NT) MRE domain of hDlg. This I3-independent interaction recruits both I2 and I3 isoforms of hDlg to the lateral membrane of epithelial cells. Upon mutation of the N-terminal MRE domain, the I3 insert is sufficient to target the hDlg to the membrane in an N-terminal MRE domain-independent manner.

 

It is relevant to emphasize here that our experiments utilized the full-length naturally occurring isoforms of hDlg, which is in contrast to the use of truncated constructs or individual domains used by others to assess their subcellular distribution in epithelial cells (18, 19). MAGUK proteins are multidomain proteins and participate in various intramolecular domain interactions that contribute to the protein stability and regulate their function as effective scaffolding molecules (35, 36). Therefore, there is a possibility that the truncation of any domain could unexpectedly disrupt one of these intramolecular interactions and lead to an artifactual mislocalization of the expressed protein in heterologous cells. Furthermore, it has been proposed that the targeting of Drosophila Dlg to the synaptic membrane involves a stepwise contribution of various domains in a sequential manner (37). Guided by these observations, it is perhaps more prudent to utilize full-length hDlg polypeptides that are capable of participating in all protein-protein interactions at various stages of intracellular transport and targeting processes. Indeed, our recent demonstration of a direct interaction between GAKIN (guanylate kinase domain-associated kinesin) and the GUK domain of hDlg (38, 39) further underscores the importance of the full-length protein in the intracellular transport of hDlg and its associated partners to correct cellular destinations in mammalian cells.

A previous study has shown that the truncated constructs spanning the C-terminal segment of hDlg/SAP97 tend to accumulate in the nuclear compartment (19). Based on this observation, we initially surmised that the N terminus of hDlg might contain a functional NES motif that prevents the accumulation of full-length hDlg in the nucleus. In fact, the presence of a functional NES motif in the N terminus of hDlg would be akin to the case of APC (adenomatous polyposis coli) tumor suppressor protein, where the nuclear localization is dynamically regulated by the CRM1-dependent nuclear export system utilizing the NES motifs (33, 34). However, under our experimental conditions, the disruption of the putative NES motif in the N terminus of hDlg did not result in the accumulation of hDlg mutants in the nucleus. One possible explanation of our observation is that the NES-like motif in hDlg does not function as an active nuclear export signal under these experimental conditions. Contrary to the previous reports that have assigned the I3 (28) and I2 (11) inserts as potential nuclear targeting motifs, we could not detect any significant nuclear localization of hDlg isoforms either with or without the double mutation in the putative NES (I38A/I40A) consensus. Therefore, the authenticity of the nuclear localization signal and the possible function of hDlg in the nucleus remain a topic of future investigation.

Our results confirm the importance of the N-terminal MRE domain for membrane targeting of hDlg in polarized epithelial cells (21), and this targeting is known to be mediated through its binding to the CASK or other related MAGUKs such as DLG2 (22). An unexpected finding, however, was our observation that the combined mutation I38A/I40A of hDlg disrupted the DLG2 binding and the MRE domain-dependent membrane targeting of hDlg in epithelial cells. Similar point mutations of conserved hydrophobic residues in the L27 domains are known to disrupt specific protein-protein interactions (22, 40). The mutant hDlg constructs used in our study are likely to serve as useful experimental tools for elucidating the mode of interactions between the MRE domain and its binding partners and providing a biochemical rationale for future structural studies of this conserved protein domain. In addition, the N-terminal segment of hDlg is known to self-associate into homo-oligomers (41), binds tyrosine kinase Lck (42), a motor protein myosin VI (43), and appears to contain multiple protein-protein interaction motifs that facilitate the scaffolding function(s) of hDlg in vivo. The availability of full-length hDlg constructs carrying a combined mutation in the N-terminal MRE domain would permit testing of the role of this domain in a myriad of hDlg functions.

A notable feature among p55, CASK, and the I3 sequence of hDlg is the conserved cluster of basic amino acid residues (13, 15, 17). However, the crystal structure of the FERM domain of protein 4.1R did not reveal any specific pocket of acidic residues within the binding interface of p55 (44). In addition, the VAM1/Pals2 protein displays a similar patch of basic residues in the HOOK domain but failed to interact with protein 4.1R in vitro (45). Clearly, more studies are needed to define the precise residues within the HOOK domain that mediate its binding to the FERM domain of protein 4.1R and govern the specificity of binding between various HOOK and FERM domains. In conclusion, our results suggest an important function of the protein 4.1R binding domain in the membrane-targeting events likely to be widely conserved among various MAGUKs.


   FOOTNOTES
 
* This work was supported in part by National Institutes of Health Grants CA94414 and HL 60755. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

{ddagger} Recipient of the 2002 Tufts University Earl P. Charlton Award. Back

§ To whom correspondence should be addressed: St. Elizabeth's Medical Center, CBR 404, 736 Cambridge St., Boston, MA 02135. Tel.: 617-789-3118; Fax: 617-789-3111; E-mail: Athar.Chishti{at}Tufts.edu.

1 The abbreviations used are: PDZ, PSD-95-Discs large-ZO-1; hDlg, human discs large; FERM, Four.1-Ezrin-Radixin-Moesin; MDCK, Madine-Darby canine kidney; GFP, green fluorescent protein; MAGUK, membrane-associated guanylate kinase homologue; SH3, src homology 3; GUK, guanylate kinase-like; MRE, MAGUK recruitment; L27, Lin-2 Lin-7; GST, glutathione S-transferase; NES, nuclear export signal. Back


   ACKNOWLEDGMENTS
 
We are grateful to Drs. S.-C. Huang and E. J. Benz, Jr., of the Dana Farber Cancer Institute of Boston for sharing the human protein 4.1R cDNA clone.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Woods, D. F., and Bryant, P. J. (1991) Cell 66, 451–464[CrossRef][Medline] [Order article via Infotrieve]
  2. Anderson, J. M. (1996) Curr. Biol. 6, 382–384[CrossRef][Medline] [Order article via Infotrieve]
  3. Dimitratos, S. D., Woods, D. F., Stathakis, D. G., and Bryant, P. J. (1999) Bioessays 21, 912–921[CrossRef][Medline] [Order article via Infotrieve]
  4. Woods, D. F., Hough, C., Peel, D., Callaini, G., and Bryant, P. J. (1996) J. Cell Biol. 134, 1469–1482[Abstract/Free Full Text]
  5. Bilder, D., Li, M., and Perrimon, N. (2000) Science 289, 113–116[Abstract/Free Full Text]
  6. Ohshiro, T., Yagami, T., Zhang, C., and Matsuzaki, F. (2000) Nature 408, 593–596[CrossRef][Medline] [Order article via Infotrieve]
  7. Peng, C.-Y., Manning, L., Albertson, R., and Doe, C. Q. (2000) Nature 408, 596–600[CrossRef][Medline] [Order article via Infotrieve]
  8. Lue, R. A., Marfatia, S. M., Branton, D., and Chishti, A. H. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 9818–9822[Abstract/Free Full Text]
  9. Muller, B. M., Kistner, U., Veh, R. W., Cases-Langhoff, C., Becker, B., Gundelfinger, E. D., and Garner, C. C. (1995) J. Neurosci. 15, 2354–2366[Abstract]
  10. Mori, K., Iwao, K., Miyoshi, Y., Nakagawara, A., Kofu, K., Akiyama, T., Arita, N., Hayakawa, T., and Nakamura, Y. (1998) J. Hum. Genet. 43, 123–127[CrossRef][Medline] [Order article via Infotrieve]
  11. McLaughlin, M., Hale, R., Ellston, D., Gaudet, S., Lue, R. A., and Viel, A. (2002) J. Biol. Chem. 277, 6406–6412[Abstract/Free Full Text]
  12. Hough, C. D., Woods, D. F., Park, S., and Bryant, P. J. (1997) Genes Dev. 11, 3242–3253[Abstract/Free Full Text]
  13. Marfatia, S. M., Lue, R. A., Branton, D., and Chishti, A. H. (1995) J. Biol. Chem. 270, 715–719[Abstract/Free Full Text]
  14. Nunomura, W., Takakuwa, Y., Parra, M., Conboy, J., and Mohandas, N. (2000) J. Biol. Chem. 275, 24540–24546[Abstract/Free Full Text]
  15. Cohen, A. R., Woods, D. F., Marfatia, S. M., Walther, Z., Chishti, A. H., and Anderson, J. M. (1998) J. Cell Biol. 142, 129–138[Abstract/Free Full Text]
  16. Biederer, T., and Sudhof, T. C. (2001) J. Biol. Chem. 276, 47869–47876[Abstract/Free Full Text]
  17. Marfatia, S. M., Morais-Cabral, J. H., Lin, L., Hough, C., Bryant, P. J., Stolz, L., and Chishti, A. H. (1996) J. Cell Biol. 135, 753–766[Abstract/Free Full Text]
  18. Lue, R. A., Brandin, E., Chan, E. P., and Branton, D. (1996) J. Cell Biol. 135, 1125–1137[Abstract/Free Full Text]
  19. Wu, H., Reuver, S. M., Kuhlendahl, S., Chung, W. J., and Garner, C. C. (1998) J. Cell Sci. 111, 2365–2376[Abstract]
  20. Doerks, T., Bork, P., Kamberov, E., Makarova, O., Muecke, S., and Margolis, B. (2000) Trends Biochem. Sci. 25, 317–318[CrossRef][Medline] [Order article via Infotrieve]
  21. Lee, S., Fan, S., Makarova, O., Straight, S., and Margolis, B. (2002) Mol. Cell. Biol. 22, 1778–1791[Abstract/Free Full Text]
  22. Karnak, D., Lee, S., and Margolis, B. (2002) J. Biol. Chem. 277, 46730–46735[Abstract/Free Full Text]
  23. Mendoza, C., Olguin, P., Lafferte, G., Thomas, U., Ebitsch, S., Gundelfinger, E. D., Kukuljan, M., and Sierralta, J. (2003) J. Neurosci. 23, 2093–2101[Abstract/Free Full Text]
  24. Chetkovich, D. M., Bunn, R. C., Kuo, S.-H., Kawasaki, Y., Kohwi, M., and Bredt, D. S. (2002) J. Neurosci. 22, 6415–6425[Abstract/Free Full Text]
  25. Roh, M. H., Makarova, O., Liu, C.-J., Shin, K. Y., Lee, S., Laurinec, S., Goyal, M., Wiggins, R., and Margolis, B. (2002) J. Cell Biol. 157, 161–172[Abstract/Free Full Text]
  26. Hsueh, Y.-P., Wang, T.-F., Yang, F.-C., and Sheng, M. (2000) Nature 404, 298–302[CrossRef][Medline] [Order article via Infotrieve]
  27. Balda, M. S., and Matter, K. (2000) EMBO J. 19, 2024–2033[CrossRef][Medline] [Order article via Infotrieve]
  28. Kohu, K., Ogawa, F., and Akiyama, T. (2002) Genes Cells 7, 707–715[Abstract]
  29. Baklouti, F., Huang, S.-C., Vulliamy, T. J., Delaunay, J., and Benz, E. J., Jr. (1997) Genomics 39, 289–302[CrossRef][Medline] [Order article via Infotrieve]
  30. Mattagajasingh, S. N., Huang, S.-C., Hartenstein, J. S., and Benz, E. J., Jr. (2000) J. Biol. Chem. 275, 30573–30585[Abstract/Free Full Text]
  31. Fukuda, M., Asano, S., Nakamura, T., Adachi, M., Yoshida, M., Yanagida, M., and Nishida, E. (1997) Nature 390, 308–311[CrossRef][Medline] [Order article via Infotrieve]
  32. Gaubatz, S., Lees, J. A., Lindeman, G. J., and Livingston, D. M. (2001) Mol. Cell. Biol. 21, 1384–1392[Abstract/Free Full Text]
  33. Henderson, B. R. (2000) Nat. Cell Biol. 2, 653–660[CrossRef][Medline] [Order article via Infotrieve]
  34. Rosin-Arbesfeld, R., Townsley, F., and Bienz, M. (2000) Nature 406, 1009–1012[CrossRef][Medline] [Order article via Infotrieve]
  35. McGee, A. W., and Bredt, D. S. (1999) J. Biol. Chem. 274, 17431–17436[Abstract/Free Full Text]
  36. Wu, H., Reissner, C., Kuhlendahl, S., Coblentz, B., Reuver, S., Kindler, S., Gundelfinger, E. D., and Garner, C. C. (2000) EMBO J. 19, 5740–5751[CrossRef][Medline] [Order article via Infotrieve]
  37. Thomas, U., Ebitsch, S., Gorczyca, M., Koh, Y. H., Hough, C. D., Woods, D., Gundelfinger, E. D., and Budnik, V. (2000) Curr. Biol. 10, 1108–1117[CrossRef][Medline] [Order article via Infotrieve]
  38. Hanada, T., Lin, L., Tibaldi, E. V., Reinherz, E. L., and Chishti, A. H. (2000) J. Biol. Chem. 275, 28774–28784[Abstract/Free Full Text]
  39. Asaba, N., Hanada, T., Takeuchi, A., and Chishti, A. H. (2003) J. Biol. Chem. 278, 8395–8400[Abstract/Free Full Text]
  40. Kamberov, E., Makarova, O., Roh, M., Liu, A., Karnak, D., Straight, S., and Margolis, B. (2000) J. Biol. Chem. 275, 11425–11431[Abstract/Free Full Text]
  41. Marfatia, S. M., Byron, O., Campbell, G., Liu, S. C., and Chishti, A. H. (2000) J. Biol. Chem. 275, 13759–13770[Abstract/Free Full Text]
  42. Hanada, T., Lin, L., Chandy, K. G., Oh, S. S., and Chishti, A. H. (1997) J. Biol. Chem. 272, 26899–26904[Abstract/Free Full Text]
  43. Wu, H., Nash, J. E., Zamorano, P., and Garner, C. C. (2002) J. Biol. Chem. 277, 30928–30934[Abstract/Free Full Text]
  44. Han, B.-G., Nunomura, W., Takakuwa, Y., Mohandas, N., and Jap, B. K. (2000) Nat. Struct. Biol. 7, 871–875[CrossRef][Medline] [Order article via Infotrieve]
  45. Tseng, T.-C., Marfatia, S. M., Bryant, P. J., Pack, S., Zhuang, Z., O'Brien, J. E., Lin, L., Hanada, T., and Chishti, A. H. (2001) Biochim. Biophys. Acta 1518, 249–259[Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
J. Bohl, N. Brimer, C. Lyons, and S. B. Vande Pol
The Stardust Family Protein MPP7 Forms a Tripartite Complex with LIN7 and DLG1 That Regulates the Stability and Localization of DLG1 to Cell Junctions
J. Biol. Chem., March 30, 2007; 282(13): 9392 - 9400.
[Abstract] [Full Text] [PDF]

Home page
 Mol. Biol. CellHome page
A. S. Fanning, B. P. Little, C. Rahner, D. Utepbergenov, Z. Walther, and J. M. Anderson
The Unique-5 and -6 Motifs of ZO-1 Regulate Tight Junction Strand Localization and Scaffolding Properties
Mol. Biol. Cell, March 1, 2007; 18(3): 721 - 731.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurosci.Home page
M. P. Regalado, R. T. Terry-Lorenzo, C. L. Waites, C. C. Garner, and R. C. Malenka
Transsynaptic signaling by postsynaptic synapse-associated protein 97.
J. Neurosci., February 22, 2006; 26(8): 2343 - 2357.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
K. J. Ralston, S. L. Hird, X. Zhang, J. L. Scott, B. Jin, R. F. Thorne, M. C. Berndt, A. W. Boyd, and G. F. Burns
The LFA-1-associated Molecule PTA-1 (CD226) on T Cells Forms a Dynamic Molecular Complex with Protein 4.1G and Human Discs Large
J. Biol. Chem., August 6, 2004; 279(32): 33816 - 33828.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
D. Godreau, R. Vranckx, A. Maguy, C. Goyenvalle, and S. N. Hatem
Different Isoforms of Synapse-associated Protein, SAP97, Are Expressed in the Heart and Have Distinct Effects on the Voltage-gated K+ Channel Kv1.5
J. Biol. Chem., November 21, 2003; 278(47): 47046 - 47052.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
278/36/34445    most recent
M305209200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hanada, T.
Right arrow Articles by Chishti, A. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hanada, T.
Right arrow Articles by Chishti, A. H.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals