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


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
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 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
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fortier, J. M.
Right arrow Articles by Kornbluth, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fortier, J. M.
Right arrow Articles by Kornbluth, J.
The Journal of Immunology, 2006, 176: 6454-6463.
Copyright © 2006 by The American Association of Immunologists

NK Lytic-Associated Molecule, Involved in NK Cytotoxic Function, Is an E3 Ligase1

Julie M. Fortier* and Jacki Kornbluth2,*,{dagger}

* Department of Pathology, St. Louis University School of Medicine, St. Louis, MO 63104; and {dagger} Veterans Administration Medical Center, St. Louis, MO 63106


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
NK lytic-associated molecule (NKLAM) is a protein involved in the cytolytic function of NK cells and CTLs. It has been localized to the cytolytic granules in NK cells and is up-regulated when cells are exposed to cytokines IL-2 or IFN-beta. We report in this study that NKLAM contains a cysteine-rich really interesting new gene (RING) in between RING-RING domain, and that this domain possesses strong homology to the RING domain of the known E3 ubiquitin ligase, Dorfin. To determine whether NKLAM functions as an E3 ligase, we performed coimmunoprecipitation binding assays with ubiquitin conjugates (Ubcs) UbcH7, UbcH8, and UbcH10. We demonstrated that both UbcH7 and UbcH8 bind to full-length NKLAM. We then performed a similar binding assay using endogenous NKLAM and UbcH8 expressed by human NK clone NK3.3 to show that the protein interaction occurs in vivo. Using the yeast two-hybrid system, we identified uridine kinase like-1 (URKL-1) protein as a substrate for NKLAM. We confirmed that NKLAM and URKL-1 interact in mammalian cells by using both immunoprecipitation and confocal microscopy. We demonstrated decreased protein expression and enhanced ubiquitination of URKL-1 in the presence of NKLAM. These data indicate that NKLAM is a RING finger protein that binds Ubcs and has as one of its substrates, URKL-1, thus defining this cytolytic protein as an E3 ubiquitin ligase.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
The regulation of cell-mediated cytotoxicity is an avenue of research being investigated to aid in the body’s fight against infection and cancer. CTLs and NK cells are lymphocytes that participate in the body’s immune system. NK cells, unlike CTLs, are able to kill tumors as well as virally infected cells without prior sensitization. However, like CTLs, the killing activity of NK cells is enhanced by cytokines (1). NK cells and CTLs kill by a similar mechanism involving perforin and granzyme B, which are located in the cells’ cytoplasmic granules. Once the target cell is identified, the granules migrate to the region of target cell contact. The granules then fuse with the cell’s outer membrane and release perforin and granzymes into the intracellular junction of the immune cell and target cell. Perforin and granzymes stimulate apoptosis of the target cell (1, 2).

Our laboratory set out to identify additional proteins that were involved in the cytolytic process. We used the human NK clone NK3.3 as our model to isolate these cytolytic proteins. NK3.3 cells are phenotypically and functionally representative of peripheral blood NK cells, including enhanced cytolytic activity in the presence of cytokines IL-2 or IFN-beta. Using a cDNA library derived from IFN-beta-stimulated NK3.3 cells, we identified a novel protein whose expression increased in the presence of IFN-beta as well as IL-2; it was named NK lytic-associated molecule (NKLAM)3 (3). Once the cDNA was identified, NKLAM characterization began by determining whether it played a role in NK cytolytic activity. In one study, NKLAM-specific antisense oligonucleotides were electroporated into NK cells, specifically decreasing the level of NKLAM protein expression in these cells. The study showed that decreasing NKLAM protein expression diminished NK lysis of tumor cells. The decrease in killing was seen without affecting NK cell growth or viability, thus implicating NKLAM in the cytolytic function of NK cells (3). Subcellular fractionation studies were then performed to identify the intracellular location of NKLAM. The results obtained from these studies localized NKLAM to the cytolytic granules of NK cells, reinforcing its involvement in the cytolytic activity of these cells (3). To determine what function NKLAM plays in cytolytic killing, we focused our attention on the physical characteristics of NKLAM protein. Nucleic acid analysis of NKLAM identified a 2824-bp cDNA encoding a 731-aa protein (4). The amino acid sequence revealed two important functional structures; one of these is the three transmembrane domains (TMD) located toward the C terminus of the protein, between aa 358 and 531, and the second is a really interesting new gene (RING) in between RING (IBR)-RING domain, located in the central portion of NKLAM. The RING region of the protein contains three cysteine-rich domains (CRDs); two of these domains possess the classical RING finger motif of C3HC4 and are located in aa 115–161 and aa 282–333 (5, 6). The third domain motif is classified as an IBR and possesses the motif of C6HC; it is located in aa 202–252 (7).

The identification of the CRDs provided us with insight into a possible function of NKLAM. CRDs have been found to be important in protein-protein interactions. They provide a cross-braced zinc ligation system that allows for the binding of one cysteine to another. The cross-bracing provides a unique shape for the protein, making specific amino acids accessible for binding to other proteins (5). When we compared NKLAM to other RING finger proteins, we noted strong homology between the CRDs of NKLAM and the CRDs of the Dorfin protein. Dorfin has been defined as an E3 ligase, a protein involved in the process of ubiquitination (8, 9, 10). Parkin is another E3 ligase whose RING-IBR-RING region was found to have homology to NKLAM in the RING region (11, 12). These homologies suggest NKLAM could also participate in the ubiquitination process as an E3 ubiquitin ligase.

E3 ligases have been found to be important proteins involved in the ubiquitination of substrate proteins. In the ubiquitination process, a thiol-ester bond is formed between the C terminus of the ubiquitin protein and the active site of a ubiquitin-activating enzyme known as an E1. The ubiquitin protein is then transferred to a ubiquitin conjugate (Ubc) protein E2 using a thiol-ester linkage. Using its RING fingers, an E3 ligase is able to bind to the ubiquitin-conjugated E2s as well as their substrates, facilitating the transfer of the ubiquitin from the E2 conjugates to the substrates, tagging the substrates with ubiquitin (13). Ubiquitination is one method that cells use to regulate cellular processes. It has been found to be involved in ribosomal function, postreplication DNA repair, initiation of the inflammatory response, and the function of some transcription factors (14). One of the better characterized functions of ubiquitination, however, is its involvement in the degradation of cellular proteins by the 26 S proteasome. Cell functions that have been found to be regulated by this method of degradation include progression of the cell cycle, induction of the inflammatory response, and Ag presentation (14).

In the ubiquitination process, it is the E3 ligase that regulates the ubiquitination of the substrates by selectively binding to specific proteins. Dorfin was confirmed as an E3 ligase by identifying its binding to UbcH7 and UbcH8. The substrates that have been identified for Dorfin include mutant superoxide dismutase-1 (SOD1), a protein found in inclusion bodies in the brain of patients with amyotrophic lateral sclerosis, and synphilin-1, a protein found in the neuronal inclusions of Parkinson’s patients known as Lewy bodies (8, 9, 10). The binding of Dorfin to SOD1 mutants and to synphilin-1 has been shown to ubiquitinate and degrade these proteins. It is speculated that the ubiquitination of SOD1 mutants and synphilin-1 is a regulatory mechanism that helps to prevent neuronal inclusions formed in patients with amyotrophic lateral sclerosis and Parkinson’s disease, respectively. The E3 ligase Parkin is localized to the Lewy bodies of sporadic Parkinson’s disease, and like Dorfin, was found to interact with synphilin-1 (15, 16, 17, 18). Mutations in Parkin, which impair its ubiquitination function, have been linked to autosomal recessive juvenile parkinsonism (16).

E3 ligases have recently been found to have important roles in immune cells. Cbl and GRAIL (gene related to anergy in lymphocytes) are E3 ligases that are crucial for the correct function of T cells. Cbl ubiquitinates activated Src and Syk protein tyrosine kinases, thereby down-regulating Ag signaling. Another substrate identified for the E3 ligase Cbl is p85, a regulatory subunit of PI3K (19, 20). The interaction between Cbl and p85 brings about the ubiquitination of p85 but does not signal it for degradation; instead, the ubiquitination appears to affect p85 interaction with the cell surface receptor CD28 or the TCR-CD3 complex, affecting the internalization of TCRs (19, 20, 21, 22). GRAIL is an E3 ligase, whose function has been found to be important in anergic CD4+ T cells. When hybridomas were transduced with either wild-type GRAIL or mutant GRAIL lacking RING fingers, it was found that T cells expressing the mutant GRAIL failed to become anergic (23). Another E3 ligase involved in T cell function is the T cell RING protein identified in activation screen (TRAC-1). This E3 ligase has been found to positively regulate T cell expression of the cell surface activation marker CD69 upon TCR signaling (24). TLRs involved in proinflammatory responses are also affected by an E3 ligase, Triad3A. Triad3A enhances the ubiquitination of TLR4 and TLR9, leading to their proteolytic degradation (25). Cbl, GRAIL, TRAC-1, and Triad3A are all examples of E3 ligases with important roles in the immune system, where NKLAM has been found to be relevant.

In this report we describe the structure of NKLAM’s RING domain and observe its homology with the E3 ligase Dorfin. Using immunoprecipitation studies, we identify Ubcs that are able bind to NKLAM and investigate the CRDs responsible for these interactions. Additionally we determine that the UbcH8-NKLAM interaction we observed in an overexpression system is seen in NK cells in vivo. We use the yeast two-hybrid system to identify uridine kinase like-1 (URKL-1) protein as a substrate for NKLAM. We demonstrate that the NKLAM-URKL-1 protein interaction occurs not only in yeast but also in mammalian cells. We show that the RING fingers of NKLAM are involved in URKL-1 interaction and that the level of URKL-1 protein decreases in the presence of NKLAM. We also show that URKL-1 ubiquitination is enhanced in the presence of NKLAM. Thus our results indicate that NKLAM is an E3 ligase, with URKL-1 as its substrate, resulting in its ubiquitination and degradation.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Abs and reagents

NKLAM mAbs 14, 76, and 35 were produced in our laboratory. Monoclonal hemagglutinin (HA) and myc Abs, and polyclonal Flag Abs, proteasome inhibitor MG132, FBS, and iodoacetamide were purchased from Sigma-Aldrich. Lipofectamine was purchased from Invitrogen Life Technologies. HRP-conjugated donkey anti-rabbit IgG and Pierce Supersignal Chemiluminescence were purchased from Amersham. Protein G-agarose was purchased from Boehringer Mannheim. Peroxidase-conjugated goat anti-mouse IgG was purchased from Bio-Rad. The pFlag-CMV vector was purchased from Kodak. Alexa Fluor 488-labeled goat anti-mouse Ig, Alexa Fluor 586-labeled goat anti-rabbit Ig, and TOPRO3 were purchased from Molecular Probes. Polyclonal UbcH8 Ab was purchased from Boston Biochem.

Constructs

NKLAM constructs lacking the TMDs, as well as one or more CRDs, were designed using a combination of cDNA fragments as well as PCR products. NKLAM’s cDNA served as a template for both restriction digestion and PCR. All constructs contained the entire 5' coding region of NKLAM encompassing bp 240-544. This fragment was obtained using NcoI/TaqI restriction digestion. This fragment was then ligated to the appropriate PCR product to obtain the fragment of DNA required to produce the constructs. The resulting DNA fragments were flanked by NcoI and XhoI restriction sites on their 5' and 3' ends, respectively, and ligated into the mammalian expression vector CS2-MT, tagging the NKLAM N terminus with five myc epitopes. All plasmids were sequenced for confirmation. The constructs were named for the CRDs they possessed. CRD1, 2, 3 contains all three CRDs, encompassing the cDNA bp 240-1218. CRD1, 2 encompasses cDNA bp 240-1014. CRD1 contains only cDNA bp 240-775. CRD2, 3 contains the 5' region of NKLAM bp 240-544 along with the second and third CRDs bp 739-1223. CRD1, 3 contains NKLAM bp 240-802 along with bp 994-1223. CRD2 contains NKLAM bp 240-544 and 739-1086. CRD3 contains NKLAM bp 240-544 and 994-1244.

PCR products of full-length UbcH7, UbcH8, and UbcH10 were made using cDNA obtained from Clontech Laboratories yeast two-hybrid human spleen cDNA library as the template. Flag-tagged conjugates were made by directionally ligating the PCR products into the KpnI and BamHI sites of pcDNA3-Flag, tagging the N terminus of the proteins with Flag.

The NKLAM construct used in the yeast two-hybrid system encoded the first aa 349 of the NKLAM protein and contained all three CRDS. This corresponded to the bp 240-1259 and was designed using the NcoI/EcoRI fragment of NKLAM cDNA (accession no. AK074486). This fragment was ligated in frame into the binding domain Gal4BD Pas-1 plasmid.

The URKL-1 pFlag construct was designed by removing the 1.8-kb EcoRI/XhoI fragment containing the URKL-1 cDNA from the yeast two-hybrid pACT2 vector. The digested ends were blunted and the fragment was ligated into the EcoRV site of the pFlag vector, placing URKL-1 protein in frame with the Flag tag. Orientation of this clone was determined by restriction enzyme digestion. The HA-tagged ubiquitin plasmid was graciously provided by Dr. D. Bohmann of the University of Rochester Medical Center (Rochester, NY).

Cell culture and transfections

The human embryonic kidney epithelial cells, HEK 293, were purchased from American Type Culture Collection and maintained at 1.5 x 105 cells/ml in DMEM supplemented with 10% FBS and 1% L-glutamine. Lipofectamine was used to transfect plasmids into HEK 293 cells as per the manufacturer’s protocol. Briefly, 1.5 x 106 HEK 293 cells were grown in 35-mm petri dishes overnight. The 1–6 µg of plasmid DNA was mixed with 100 µl of DMEM and 6 µl of Plus reagent followed by 15 min incubation at room temperature. The 4 µl of the Lipofectamine reagent was mixed with 100 µl of DMEM, and this mixture was then added to the DNA mixture and incubated for 15 min at room temperature. The 293 cells were gently washed with 1 ml of DMEM and covered with 800 µl of fresh DMEM. The Lipofectamine/DNA mixture was then added to the cells for 3 h incubation at 37°C. The Lipofectamine medium mixture was then replaced with 3 ml of fresh 293 cell culture media and the cells were incubated at 37°C for 24 h. Cells were then treated with 2–25 µM MG132, a 26 S proteasome inhibitor, for 16 h. NK3.3 cells were maintained at 3 x 105 cells/ml in RPMI 1640 containing 15% FBS, 15% T-STIM (BD Biosciences), 200 U/ml rIL-2 (a gift from Chiron), and 1% L-glutamine.

Western blot analysis

Cell lysates were obtained from transfected 293 cells that were treated with MG132. The cells were pelleted at 1000 rpm for 5 min, washed in 5 ml of 1x PBS, and pelleted a second time. The pellets were resuspended in lysis buffer (20 µM PIPES (pH 6.8), 4% sucrose, 100 mM NaCl, 0.5% Nonidet P-40) containing protease inhibitors (25 µg/ml aprotinin, leupeptin, pepstatin, 1 mM PMSF, 2 mM benzamidine), at a concentration of 2 x 106/100 µl of lysis buffer. The cells were then sonicated twice for 10 s each and then rotated slowly for 1 h at 4°C, followed by centrifugation for 3 min at 3000 rpm. Lysates were then stored at –80°C until used.

The 5–30 µl of lysates was mixed with an equal volume of 2x SDS loading buffer. The samples were boiled for 20 min and loaded onto a 7.5% or 10% SDS-PAGE at 100–150 V. The gels were transferred 1 h onto polyvinylidene difluoride membranes using an ice block at 100 V in transfer buffer containing 20% methanol and 0.01% SDS. The membranes were then blocked overnight in 5% blotto, and 0.2% Tween, followed by 1 h incubation in 20 ml blotto containing 2–8 µg of primary Ab. The membranes were washed three times for 10 min each in wash buffer (150 mM NaCl, 10 mM Tris (pH 8), 0.2% Tween), followed by a 1 h incubation in 20 ml blotto containing a 1/5000 dilution of goat anti-mouse secondary Ab conjugated to HRP or a 1/3000 dilution of the donkey anti-rabbit secondary Ab conjugated to HRP. The membranes were then washed five times for 10 min each in wash buffer, twice for 1 min each in rinse buffer (150 mM NaCl, 10 mM Tris (pH 8)) then placed in Supersignal solution for 2 min and exposed to film for 2–30 min.

Immunoprecipitation for identifying interacting proteins

The 25–50 µl of the lysates was mixed with 2–5 µg of desired Ab to precipitate the protein of interest. Lysates used for Ubc immunoprecipitation were precleared using 25 µl of washed protein G-agarose in 100 µl of lysis buffer and rotated for 1 h at 4°C. The protein G-agarose was pelleted 30 s at 3000 rpm and the supernatants were mixed with the appropriate Ab for the precipitations. All other lysates were mixed directly with precipitating Ab. Samples were then rotated at 4°C for 2 h followed by the addition of 25 µl of protein G-agarose in 100 µl lysis buffer and incubated 24 h at 4°C while rotating slowly. The samples were then centrifuged 30 s at 3000 x g and the pellets washed three times with 1 ml of PBS buffer containing protease inhibitors (2.5 µg/ml aprotinin, leupeptin, pepstatin, 0.1 mM PMSF, 0.2 mM benzamidine). The final pellets were resuspended in 40 µl of 1x SDS loading buffer, boiled 20 min, run on SDS-PAGE gels, and then Western blotting was performed.

Coprecipitation of endogenous UbcH8 with NKLAM

NK3.3 cells were stimulated with 10,000 U/ml IFN-beta and 200 U/ml IL-2 for 12 h followed by the addition of 2 µM MG132 for 6 h. NK3.3 cells were lysed 1 h at 4°C while rotating in lysis buffer containing 20 mM PIPES (pH 6.8), 4% sucrose, 100 mM NaCl, 0.5% Nonidet P-40 along with 15 mM iodoacetamide and protease inhibitors (25 µg/ml aprotinin, leupeptin, pepstatin, 1 mM PMSF, 2 mM benzamidine). The supernatants were obtained by centrifugation at 10,000 rpm for 3 min. The 300 µl of the lysates were mixed with NKLAM mAbs (5 µg of Ab14, 2 µg of Ab76, and 2 µg of Ab35) and incubated at 4°C while rotating. Following the incubation, 50 µl of washed protein G in 100 µl of buffer was added to the lysates and Ab mixture and allowed to rotate at 4°C overnight. The samples were spun and the resulting pellets were washed three times with 1 ml of PBS containing the protease inhibitors. The final pellet was mixed with 40 µl of 1x SDS loading buffer, boiled for 10 min and run on SDS-PAGE gels. Western blot analysis was performed as described above using 8 µg of polyclonal UbcH8 Ab as the primary Ab in 20 ml of blotto solution.

Yeast two-hybrid system

The yeast two-hybrid system was purchased from Clontech and the assay was performed as per protocol. Briefly, human spleen cDNA cloned into the Pact2 vector, and an NKLAM construct cloned into the Pas-1 vector, were cotransformed into yeast Y153. The transformed cells were plated onto minimal media YPD plates designed for yeast growth; the plates contained 20 mM 3-amino-1,2,4 triazole and lacked leucine, tryptophan, and histidine. Successful transformation with both plasmids would enable the yeast to grow on these plates. Colonies were lifted from plates using nitrocellulose filters and then placed into liquid nitrogen to expose the DNA. The filters were then treated with X-Gal to identify the presence of lacZ gene. Positive blue colonies, also indicative of protein-protein interactions, were identified and DNA preparations were made from the yeast clones. Plasmid DNA was then transformed into bacteria and DNA preparations were made for sequence and restriction digest analysis. Restriction digests were performed on the DNA using HindIII/XhoI to remove the insert from the Pact2 clones.

Sequencing

DNA preparations of the Pact2 clones were used for sequence analysis of the cDNA. The DNA was sequenced using primers specific to the Pact2 vector (5'-gat ctg tat ggc tta ccc-3'). Sequence results were run against GenBank database (available at <www.ncbi.nlm.nih.gov>) for homology analysis.

URKL-1 ubiquitination assay using HA-tagged ubiquitin

HEK 293 cells were cotransfected with 2, 4, or 6 µg of an HA-tagged ubiquitin plasmid along with 1 µg of pFlag-URKL-1 plasmid with and without the presence of a vector expressing NKLAM. The transfections were performed as per the Lipofectamine protocol and 24 h posttransfection, 25 µM MG132 was added to the cells. At 16 h post MG132 addition, cells were pelleted, washed with PBS, and the lysates were obtained as described using a lysis buffer (20 mM PIPES (pH 6.8), 4% sucrose, 100 mM NaCl, 0.5% Nonidet P-40) containing protease inhibitors (25 µg/ml aprotinin, leupeptin, pepstatin, 1 mM PMSF, 2 mM benzamidine), at a concentration of 2 x 106/100 µl of lysis buffer. Some 50 µl of the lysates were mixed with 2 µg of poly-Flag Ab and incubated for 2 h at 4°C while rotating. Washed protein G (25 µl), which was resuspended in 100 µl of lysis buffer, was then added to the lysate Ab mixtures and allowed to rotate overnight at 4°C. Pellets were then obtained by centrifugation, and washed three times with 1 ml of PBS containing the protease inhibitors. The final pellets were resuspended in 40 µl 1x SDS loading buffer. The samples were boiled for 20 min and then subjected to Western blot analysis as described, using 6 µg of polyclonal Flag Ab in 20 ml of blotto to identify URKL-1.

Immunocytochemistry

HEK 293 cells transfected with NKLAM, URKL-1, or the combination, were plated overnight onto 35-mm petri dishes containing a glass coverslip. The slips were then washed briefly in PBS and then fixed 10 min in ice-cold methanol and 1 min in ice-cold acetone. The slips were washed twice in PBS with 1% BSA for 5 min each, and then incubated for 1 h in PBS containing either 1 µg/ml NKLAM mAb14 or 2.5 µg/ml polyclonal Flag Ab. Combinations of the monoclonal and polyclonal Abs were used for colocalization studies. The slips were then washed three times for 5 min each in PBS and then incubated 30 min in PBS with 1% BSA containing either 2.5 µg/ml Alexa Fluor 488-labeled goat anti-mouse Ig to identify the mAbs, or 2.5 µg/ml Alexa Fluor 568-labeled goat anti-rabbit Ig to identify the presence of the polyclonal Abs. A combination of both Fluors 488 and 568 was used in the colocalization studies. The slips were washed three times 5 min each in PBS and then placed on slides for viewing using the confocal microscope MRC 1024. As a background control, cells were treated with only secondary Abs.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
NKLAM interaction with ubiquitin-conjugating enzymes UbcH7 and UbcH8

We observed that one category of E3 ligases possesses CRDs in a RING-IBR-RING structure, the same structure found in NKLAM. This observation led us to determine whether NKLAM may in fact be an E3 ligase. To determine whether NKLAM CRDs had any homology with any other known E3 ligases, we performed BLAST (available at website) searches using the amino acid sequences of NKLAM’s three CRDs. The searches identified one E3 ligase whose CRDs possessed a strong homology to that of NKLAM. That E3 ligase is a protein known as Dorfin. When comparing the RING fingers of Dorfin to those of NKLAM, we found the levels of homology to be 62, 93, and 97% for CRD1, CRD2, and CRD3, respectively (Fig. 1a). Studies involving Dorfin had identified two Ubcs that it is able to bind, UbcH7 and UbcH8 (8, 9, 10). Based upon the high level of homology we observed between the CRDs of NKLAM and Dorfin, we investigated whether NKLAM would also be able to interact with UbcH7 and UbcH8. We used UbcH10 as our negative control for an E2 conjugate. This study was performed by cotransfecting 293 cells with a plasmid encoding Flag-tagged UbcH7, UbcH8, or UbcH10 along with a plasmid encoding full-length NKLAM. To decrease protein degradation, the cells were treated with 2 µM of the 26 S proteasome inhibitor MG132 24 h posttransfection. At 16 h after the addition of the inhibitor, cell lysates were obtained as described. Coimmunoprecipitations were then performed using 1 µg of NKLAM-specific mAb 14 followed by SDS-PAGE and immunoblotting using the polyclonal Flag Ab to identify the presence of the Ubcs. Similar to the results seen with the Dorfin protein, UbcH7 and UbcH8 coimmunoprecipitated with NKLAM (Fig. 1b). UbcH10, our negative control, was not coprecipitated with NKLAM (Fig. 1b). In our control samples that contained the Ubcs but lacked NKLAM, we did not observe precipitation of the conjugates (Fig. 1b), ensuring that NKLAM’s presence was necessary for the coimmunoprecipitation.


Figure 1
View larger version (28K):
[in this window]
[in a new window]
  		FIGURE 1. Comparison of the CRDs of NKLAM and Dorfin and coimmunoprecipitation of Ubcs with NKLAM. a, The amino acid sequence of NKLAM’s three CRDs was aligned with the respective CRDs of Dorfin. The shaded areas represent common amino acids between the two proteins. b, HEK 293 cells were cotransfected with a plasmid encoding full-length NKLAM and a plasmid encoding Flag-tagged UbcH7, UbcH8, or UbcH10. Following transfection, cells were treated with proteasome inhibitor MG132 and lysates were obtained. Lysates were mixed with 5 µg of NKLAM-specific mAb 14 and precipitated using protein G-agarose. Precipitates were resolved by SDS-PAGE and then subjected to immunoblotting using polyclonal Flag Ab to identify the Ubcs. To ensure that NKLAM and the Ubcs were expressed in the correct cells, cell lysates were subjected to immunoblotting using NKLAM mAb 14 to identify the presence of NKLAM and polyclonal Flag Ab to identify the presence of the conjugates.

 
Identification of RING fingers responsible for UbcH binding

To further characterize NKLAM, we investigated which of NKLAM CRDs are responsible for its interaction with UbcH7 and UbcH8. Dorfin studies have shown that all three CRDs are required for its interaction with UbcH7. When the C terminus or the N terminus of the Dorfin protein was removed, but the entire CRD stayed intact, UbcH7 was still able to bind to Dorfin (8). Parkin, however, requires only its third CRD to bind to UbcH8 (11). To determine which CRDs NKLAM uses for conjugate binding, we designed constructs that were lacking one or more NKLAM CRDs. All of these NKLAM constructs are myc-tagged and contain the entire 5' region of the protein, lack the TMD, but possess different CRDs. The constructs were named for the CRDs they contain and are: CRD1, 2, 3, CRD1, 2, CRD2, 3, CRD1, CRD2, CRD3, and CRD1, 3 (Fig. 2a). Interactions between the Ubcs and the CRDs of NKLAM were evaluated in 293 cells that were cotransfected with a plasmid expressing a myc-tagged NKLAM CRD along with the Flag-tagged UbcH7 or UbcH8 plasmid. Cells were transfected for 24 h, followed by the addition of 2 µM of the proteasome inhibitor MG132, then incubated for an additional 16 h. Cell lysates were then obtained, precleared using washed protein G-agarose, and mixed with 2 µg of the myc mAb and the washed protein G to precipitate NKLAM. Western blot analysis followed using the polyclonal Flag Ab to identify the presence of any bound UbcH7 or UbcH8. Both UbcH7 and UbcH8 coprecipitated with the NKLAM construct that contained all three CRDs but lacked the TMD, indicating that the TMD is not involved in these protein interactions. UbcH7 was coimmunoprecipitated with NKLAM constructs expressing CRD1, 2, 3, CRD1, 2, CRD2, 3, and CRD2. In comparison to the three constructs listed, a lesser amount of UbcH7 coprecipitated with construct CRD1, 3. Constructs CRD1 and CRD3, however, were unable to bind to UbcH7 (Fig. 2b). Similar to that of UbcH7, UbcH8 showed a reduced level of binding to CRD3 (Fig. 2c) compared with the binding to all other NKLAM constructs.


Figure 2
View larger version (40K):
[in this window]
[in a new window]
  		FIGURE 2. Coimmunoprecipitation of UbcH7 and UbcH8 with NKLAM CRDs. a, NKLAM constructs containing one or more CRDs were designed using DNA fragments and PCR products obtained from NKLAM cDNA. The 5' region of NKLAM was obtained by restriction digestion of the cDNA. This was ligated to the appropriate PCR products containing the desired CRDs. The ligated fragments were then inserted into the CS2-MT vector, myc tagging the NKLAM constructs. The CRD present in each construct is indicated. All CRD NKLAM constructs lack the TMD. b and c, 293 cells were cotransfected with a plasmid encoding a myc-tagged NKLAM CRD construct and either Flag-tagged UbcH7 (b) or UbcH8 (c). At 24 h posttransfection, cells were treated with MG132 and 16 h later lysates were obtained. The lysates were mixed with 2 µg of myc mAb to precipitate NKLAM. The precipitates were resolved by SDS-PAGE and subjected to Western blot (WB) analysis using polyclonal Flag Ab to identify the Ubcs. Samples of the cell lysates were also subjected to Western blot analysis to ensure the presence of NKLAM and the conjugate proteins in the lysates. myc mAb was used to identify NKLAM; polyclonal Flag Ab was used to identify both UbcH7 and UbcH8.

 
The coprecipitations performed using UbcH7 and the NKLAM CRD constructs show UbcH7 is unable to bind to NKLAM that contains only the first or the third CRD. There is also less binding to NKLAM that possesses both the first and the third domain (Fig. 2b). Similar to that of UbcH7, UbcH8 showed a reduced level of binding to the NKLAM constructs that contained only the third CRD (Fig. 2c). HEK 293 cells that were transfected with Flag-UbcH7 or Flag-UbcH8 alone served as our negative controls.

Precipitation of endogenous UbcH8 with endogenous NKLAM in NK3.3 cells

Because we were able to show that UbcH8 was able to be precipitated with NKLAM in an overexpression system, we wanted to determine whether the same interaction occurred in NK3.3 cells where NKLAM is endogenous. With a commercially available polyclonal Ab to UbcH8, we identified the presence of UbcH8 in NK3.3 cells by Western blot analysis using lysates from IL-2 and IFN-beta-stimulated NK3.3 cells. UbcH8 has been shown to be transcriptionally induced by IFN-beta (26). Similarly, NKLAM levels are strongly increased by stimulation with IFN-beta and IL-2 (3). Therefore, we treated NK3.3 cells with both IL-2 and IFN-beta to maximize expression of both NKLAM and UbcH8 for the coimmunoprecipitation studies. Without stimulation, the levels of NKLAM and UbcH8 were not sufficient to carry out the endogenous immunoprecipitation experiments (data not shown). Once UbcH8 was confirmed as being present in cytokine-activated NK3.3 cells, we performed immunoprecipitations on cell lysates from NK3.3 cells that had been stimulated for 12 h with IL-2 and IFN-beta followed by 6 h incubation in media containing 2 µM MG132. The cell lysates were obtained as described, and immunoprecipitated with NKLAM mAbs 14, 35, and 76 for 2 h followed by the addition of protein G for an overnight incubation. The precipitates were subjected to Western blot analysis using the UbcH8 specific Ab. The results of the Western blot show that UbcH8 is present in the cell lysates and it is able to be pulled down in the presence of NKLAM (Fig. 3), indicating that the interaction we observed in the overexpression system could also be seen in vivo in NK3.3 cells.


Figure 3
View larger version (54K):
[in this window]
[in a new window]
  		FIGURE 3. Coimmunoprecipitation of endogenous UbcH8 with endogenous NKLAM in NK3.3 cells. Cell lysates of NK3.3 cells that had been stimulated with IL-2 and IFN-beta were subjected to NKLAM immunoprecipitation using NKLAM-specific mAbs 14, 35, and 76. The cell lysates as well as the precipitates were subjected to Western blot analysis using an Ab-specific to UbcH8. The presence of NKLAM and UbcH8 (first lane) in NK3.3 cell lysates and coimmunoprecipitation of UbcH8 (second lane) with NKLAM are shown.

 
Yeast two-hybrid system

We used the yeast two-hybrid system to identify a potential substrate for NKLAM. We screened ~30,000 cDNA clones using a construct of NKLAM that encoded the N terminus of the protein and expressed all three of the CRDs in their entirety. Using a high stringency screening protocol, we identified three clones whose proteins strongly interacted with NKLAM.

Once the positive yeast clones were identified, plasmid DNA was isolated and transformed into bacterial cells to isolate the Pact2 plasmid containing the cDNA of interest. We repeated the cotransformation of NKLAM containing all three CRDs and the pACT2 positive clones into yeast to ensure the isolated clones would again prove to be positive for protein interaction. All of the positive clones were sequenced using a primer internal to the Pact2 vector. The sequences were run through GenBank database to search for known sequences. The analysis of the three clones showed that only one, identified as URKL-1 (GenBank accession no. NM_017589), was in the correct frame to produce a known protein. These results provided us with a protein that interacted with NKLAM in the yeast two-hybrid system, thus providing us a potential substrate to investigate further.

Immunoprecipitation of URKL-1 with NKLAM

Because we were able to show that NKLAM and URKL-1 interact in a yeast system, we wanted to determine whether this protein interaction could also be seen in a mammalian expression system. To determine whether URKL-1 and NKLAM interact in 293 cells, first we removed URKL-1 from the Gal4AD tag of the pACT2 vector and placed it in an expression vector that would tag URKL-1 on the N terminus with the Flag epitope. We then used a polyclonal Flag Ab for Western analysis of URKL-1 expression.

NKLAM was immunoprecipitated from 293 cells that were cotransfected with 1 µg of a plasmid encoding Flag-tagged URKL-1 and 1 µg of a plasmid encoding NKLAM. Cells were transfected using Lipofectamine as described and 24 h posttransfection, the cells were treated with 2 µM of the proteasome inhibitor MG132 to inhibit protein degradation by the 26 S proteasome. At 16 h after the addition of MG132, cell lysates were obtained and subjected to immunoprecipitation using 5 µg of NKLAM-specific mAb 14 and protein G agarose. We performed Western analysis using a polyclonal Flag Ab to identify the presence of URKL-1. The results showed that when NKLAM is precipitated, URKL-1 is also pulled down, indicating a protein-protein interaction (Fig. 4a, lane 1). URKL-1, however, was not precipitated by NKLAM mAb 14 alone when NKLAM was not present in the lysates (Fig. 4a, lane 2), indicating that the NKLAM mAb alone was not responsible for the precipitation of URKL-1. Thus, we were able to determine that the protein interaction between NKLAM and URKL-1 we observed in the yeast system could also be observed in a mammalian expression system.


Figure 4
View larger version (39K):
[in this window]
[in a new window]
  		FIGURE 4. Coimmunoprecipitation of NKLAM and URKL-1. a, 293 cells were cotransfected with plasmids encoding Flag-tagged URKL-1 and NKLAM. At 24 h after transfection, the cells were treated with 2 µM of the proteasome inhibitor MG132 for 16 h and lysates were obtained. The lysates were incubated with NKLAM monoclonal Ab14 for 2 h, followed by the addition of protein G beads. After 24 h incubation, the pellets were washed with PBS, and Western blot (WB) analysis was performed using a polyclonal Flag Ab to identify the presence of URKL-1. The lysates alone were used in Western blot analysis to show the levels of URKL-1 and NKLAM expression using polyclonal Flag Ab and NKLAM monoclonal Ab14, respectively. b, 293 cells were cotransfected with URKL-1 along with one of the myc-tagged NKLAM constructs. At 24 h posttransfection the cells were treated with 2 µM MG132, and 16 h later the lysates were obtained and mixed with myc mAb followed by the addition of protein G. The precipitated proteins were then subjected to Western blot analysis using polyclonal Flag Ab to identify the presence of URKL-1. Western blot analysis was also performed on the lysates alone to indicate the presence of the shortened forms of NKLAM using the myc Ab, or URKL-1 using the Flag Ab.

 
Identification of CRD interaction with URKL-1

With the observation that NKLAM was able to bind URKL-1 in 293 cells, we set out to identify which NKLAM RING fingers were responsible for the interaction with URKL-1. To perform this experiment we used the NKLAM CRD constructs shown in Fig. 2a. We cotransfected 293 cells with 1 µg of a myc-NKLAM construct along with 1 µg of a plasmid encoding Flag-tagged URKL-1 using Lipofectamine as described. 24 h post transfection, the cells were treated with 2 µM MG132 and 16 h later lysates were obtained. The lysates were mixed with 2 µg of the myc Ab and protein G-agarose to precipitate the shortened forms of NKLAM and the precipitates were then subjected to Western blot analysis using polyclonal Flag Ab to identify the presence of URKL-1. The results show that URKL-1 interaction with NKLAM is greatest with the second CRD (Fig. 4b). The first domain, however, also binds URKL-1, although a smaller amount of protein was present compared with the amount bound to CRD2. When just the third CRD of NKLAM was present, only a small amount of URKL-1 could be precipitated, indicating a weaker interaction with CRD3 than with the first and second domains. The presence of the third domain, however, did not impede the binding of URKL-1 to the first or the second domain when they were also present. We conclude from these binding results that the second CRD is the most important in NKLAM interaction with URKL-1, as it can stand independently in its ability to bind URKL-1.

Decreasing levels of URKL-1 in the presence of NKLAM

In an attempt to classify NKLAM as an E3 ligase, we investigated whether expression of URKL-1 was affected by the presence of NKLAM. We examined proteins levels of URKL-1 in the presence of full-length NKLAM, CRD1, 2, 3, CRD1, CRD2, and CRD3. This experiment was done by cotransfecting 293 cells with myc-tagged NKLAM, myc-tagged CRD1, 2, 3, CRD1, CRD2, or CRD3 along with Flag-tagged URKL-1. As a negative control, we transfected Flag-URKL-1 alone into 293 cells. After 24 h incubation, cell lysates were obtained as described and Western blot analysis was performed using polyclonal Flag Ab to identify the amounts of URKL-1 present. The results show that the levels of URKL-1 are much lower when it is cotransfected with full-length NKLAM compared with when it is cotransfected with CRD1, 2, 3, CRD1, CRD2, CRD3, or transfected alone (Fig. 5). These results indicate that URKL-1 levels are reduced in the presence of NKLAM. This supports the hypothesis that NKLAM ubiquitinates URKL-1, resulting in its degradation.


Figure 5
View larger version (38K):
[in this window]
[in a new window]
  		FIGURE 5. URKL-1 protein expression is decreased in the presence of NKLAM. HEK 293 cells were cotransfected with Flag-URKL-1 (all lanes), and either myc-tagged full-length NKLAM, myc-CRD1, 2, 3, myc-CRD1, myc-CRD2, and myc-CRD3, or with Flag-URKL-1 alone. At 24 h posttransfection cell lysates were obtained, and the lysates were subjected to Western blot analysis using polyclonal Flag Ab to identify the expression level of URKL-1. Western blot analysis was also performed on the lysates using monoclonal myc Ab to identify the presence of the NKLAM forms. A background band from the myc Ab appears below the full-length NKLAM band.

 
Ubiquitination of URKL-1 in the presence of NKLAM

To further test the hypothesis that NKLAM is an E3 ligase with URKL-1 as its substrate, we performed ubiquitination assays. These assays were performed using 293 cells that were cotransfected with 2 µg of plasmid encoding NKLAM along with 1 µg of the Flag-URKL-1 plasmid and 2–6 µg of a plasmid encoding HA-tagged ubiquitin. Increasing amounts of HA-tagged ubiquitin were transfected to ensure that ubiquitin would not be rate limiting in the reaction. At 24 h posttransfection, the cells were treated with 25 µM MG132 and 16 h later cell lysates were obtained. The lysates were mixed with Flag Ab and protein G to precipitate URKL-1. Western blot analysis followed to identify the presence of ubiquitin using a polyclonal HA Ab. Positive ubiquitination results have the appearance of a smear, indicating the presence of ubiquitin chains. The results of three independent experiments with different amounts of HA-tagged ubiquitin transfected all show increased URKL-1 ubiquitination in the presence of NKLAM (Fig. 6).


Figure 6
View larger version (51K):
[in this window]
[in a new window]
  		FIGURE 6. Enhanced ubiquitination of URKL-1 in the presence of NKLAM. HEK 293 cells were cotransfected with plasmids encoding Flag-URKL-1 and NKLAM and varying amounts of HA-tagged ubiquitin. The cells were treated with MG132 following transfection, and cell lysates were obtained. The lysates were mixed with Flag Ab and protein G-agarose to precipitate URKL-1. The precipitates were then subjected to Western blot analysis using HA Ab to indicate the presence of the ubiquitin chains that appear as smears. Enhanced ubiquitination of URKL-1 is seen in the presence of NKLAM. Results shown are from three independent experiments, all demonstrating increased ubiquitination of URKL-1 in the presence of NKLAM.

 
Colocalization of NKLAM and URKL-1 in 293 cells

To confirm that the protein interaction we observed between NKLAM and URKL-1 in 293 cells was due to the proteins being present in the same region of the cells, we performed colocalization studies using confocal microscopy. This experiment was done by transfecting 293 cells, together and separately, with 1 µg of a plasmid encoding NKLAM and 1 µg of a plasmid encoding Flag-URKL-1. The plasmids were also transfected individually to determine the location of the proteins in 293 cells when they are expressed by themselves. At 24 h posttransfection, the cells were fixed onto coverslips as described. The cells were incubated with either NKLAM mAb14 or polyclonal Flag Ab, or a combination of both to indicate the location of NKLAM and URKL-1, respectively. The coverslips were then treated with the secondary Abs, either Alexa Fluor 488-labeled goat anti-mouse Ig or Alexa Fluor 568-labeled goat anti-rabbit Ig for NKLAM-transfected and URKL-1-transfected cells, respectively. A combination of the two primary and secondary Abs was used on the cells that were cotransfected. The cells were then treated with the nuclear dye TOPRO3 to stain the nucleus (Fig. 7). The green appearance of the cells shows the location of NKLAM, and the location of URKL-1 is seen in red; the nuclear region is seen in blue. The cells individually transfected with NKLAM show that the NKLAM protein is located in the cytosol and predominantly toward the outer membrane, as seen in green (Fig. 7, a and b). No NKLAM is found in the nucleus, which appears blue due to the TOPRO3 staining (Fig. 7b). In cells transfected with URKL-1 alone, URKL-1 is primarily found in the nuclear region, as seen in red (Fig. 7c). The nuclear presence of URKL-1 was confirmed with the costaining of the nucleus by the nuclear stain TOPRO3. The overlapping staining of URKL-1 and TOPRO3 appears purple (Fig. 7d), indicating URKL-1 is present in the nucleus. In 293 cells that were cotransfected with NKLAM and URKL-1 (Fig. 7, e–g), NKLAM was still found in the cytosol of the cells. URKL-1, however, which was seen in the nucleus when transfected alone, is now seen in the cytosol of cotransfected cells, in the same location as NKLAM; the yellow color indicates an overlap of NKLAM and URKL-1 staining (Fig. 7g). In three independent experiments, counting 200 cotransfected cells for each experiment, we found that URKL-1 was in the cytosol 65% of the time when it was cotransfected with NKLAM, compared with 4% of the time when it was transfected individually. These results not only show that the location of URKL-1 changes in the presence of NKLAM, but they also show that when both proteins are expressed, they are found in the same region of the cell, allowing for the protein interactions we confirmed in Western blot analysis.


Figure 7
View larger version (27K):
[in this window]
[in a new window]
  		FIGURE 7. Localization of NKLAM and URKL-1 in 293 cells. HEK 293 cells that were transfected with NKLAM alone (a and b), URKL-1 alone (c and d), or cotransfected with both NKLAM and URKL-1 (e–g) were fixed onto microscope coverslips. The coverslips were then incubated with either NKLAM mAb (a and b), Flag Ab (c and d), or a combination of both Abs (e–g), followed by incubation with secondary Abs labeled with Alexa Fluor 488 (a and b), Alexa Fluor 568 (c and d), or a combination of both Abs (e–g). The cells were then treated for 10 min with TOPRO3, a nuclear staining reagent. a, NKLAM (green) is located near the cell membrane. b, The location of the nuclei (blue). c, URKL-1 (red) is located in the nuclear region of the cell. d, The location of URKL-1 (red) that coincides with the TOPRO3 (blue) staining of the nuclei is shown (purple). e, The presence of NKLAM (green). f, URKL-1 (red) found in the cytoplasm. g, Colocalization of NKLAM and URKL-1 (yellow). TOPRO3 staining of nuclei (blue). Scale bar, 10 µM.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
NKLAM possesses three CRDs that form a RING-IBR-RING structure (7, 27, 28, 29). Because RING finger motifs are often involved in protein-protein interactions, we focused on this property for NKLAM (5, 6, 7, 27, 28, 29). In an attempt to gain insight into the function of NKLAM as well as proteins that may interact with it, we performed a homology search using the CRDs of NKLAM to identify proteins that had similar binding domains. Our search led us to Dorfin, a RING-IBR-RING finger protein whose CRDs possess a strong homology with the CRDs of NKLAM. When the amino acid sequences of the first, second, and third CRDs of NKLAM were aligned with the corresponding CRDs of Dorfin, we found they possessed a 62, 93, and 97% homology, respectively. Through binding studies with UbcH7 and UbcH8, the Dorfin protein has been identified as an E3 ligase (8). Dorfin binds UbcH7 and UbcH8 through its CRDs. The high level of homology between NKLAM’s and Dorfin’s CRDs prompted us to investigate whether the similarity would permit the same interactions between NKLAM and UbcH7 or UbcH8 as were seen with Dorfin. We tested this theory by cotransfecting HEK 293 cells with both NKLAM and either UbcH7, UbcH8, or UbcH10. The addition of the proteasome inhibitor MG132 was necessary to prevent the degradation of the NKLAM protein by the 26 S proteasome. Through immunoprecipitation of NKLAM followed by Western blot analysis to determine the presence of the Ubc, we were able to determine that similar to Dorfin, NKLAM was able to bind both UbcH7 as well as UbcH8. Although NKLAM was able to bind to both of the conjugates, we noted that there appeared to be a greater interaction with UbcH8 than with UbcH7. We believe that the stronger interaction we observed may be due to a higher affinity between NKLAM and UbcH8, or due to the fact that UbcH8 expression was greater than that of UbcH7, allowing more of the UbcH8 protein to be available for protein interaction with NKLAM.

We expanded our studies to investigate which of NKLAM’s CRDs were involved in the interactions with UbcH7 and UbcH8. We began by using a myc-tagged form of NKLAM lacking the TMD cotransfected with UbcH7 or UbcH8. We found that both UbcH7 and UbcH8 were able to bind to this truncated form of NKLAM, indicating that the N terminus of the protein, and/or the CRDs, are responsible for these interactions. We continued our studies using various constructs of the NKLAM protein that were lacking one or more RING domains. Unlike Dorfin, which required all three of the CRDs to be present for its interaction with UbcH7 (8), experiments with NKLAM indicated that UbcH7 would not bind to constructs that possessed only the first or only the third CRD. We also observed less UbcH7 binding to the construct that contained both the first and the third domain but lacked the second (IBR). The absence of UbcH7 binding seen with the NKLAM forms containing only the first and third domains indicated to us the importance of these domains for UbcH7 interaction with NKLAM. It also indicated that the region of the NKLAM protein that is N-terminal to the CRD region is not involved in these protein interactions because that region was present in all of the constructs and UbcH7 was unable to bind CRD1 or CRD3. UbcH7 was, however, able to bind to NKLAM that contained the second CRD alone as well as to NKLAM possessing either the first and second CRDs or the second and third CRDs. When we performed identical studies with UbcH8, we determined that UbcH8 also does not associate with NKLAM’s third CRD as strongly as it does the other CRDs. The binding studies showed that when NKLAM possessed only the third CRD, the interaction with UbcH8 was greatly reduced. Similar to UbcH7, however, UbcH8 was able to interact with NKLAM that contained both the first and second CRDs alone, as well as in combination with other domains. The results then as a whole show that the third CRD is not involved in UbcH7 interaction with NKLAM, and with UbcH8 it is involved to a lesser extent than both the first and second domain. It is apparent that the second domain, the IBR, is the most crucial in Ubc interactions with NKLAM because UbcH7 will bind to it alone and not to the first or third domains when they are by themselves. We also observed that in the presence of NKLAM, the protein expression of both UbcH7 as well as UbcH8 decreases when compared with protein obtained from cells that were transfected with the conjugates alone. Because 2 µM MG132 may not completely inhibit the 26 S proteasome, we speculate that in the presence of NKLAM the level of conjugate ubiquitination may be enhanced, thus increasing its rate of degradation.

In addition to the binding, we observed between NKLAM and the Ubcs in the overexpression system, our studies have also shown that the interaction exists in vivo. We are able to coprecipitate NKLAM and UbcH8 expressed endogenously in NK3.3 cells. This observation confirms the interaction that we witnessed in the overexpression system and provides strong evidence that NKLAM functions as an E3 ligase.

Because an E3 ligase possesses a substrate that it binds to and ubiquitinates, we sought to identify a protein that interacts with NKLAM. Using the yeast two-hybrid system, we identified URKL-1 as a possible substrate for NKLAM. Our analysis of URKL-1 identified a 1847-bp cDNA that yields a 548-aa protein. To date, very little research has been done on URKL-1. One study identified URKL-1 as a protein that interacts with Epstein Barr nuclear Ag 3 (EBNA3) when the cDNAs encoding both proteins were transfected into CV-1 cells (30). This study also determined that URKL-1 mRNA was ubiquitous throughout the body, including in the brain, placenta, lung, liver, and pancreas. However, higher levels of URKL-1 mRNA were found in the skeletal muscle, heart, and kidney (30). No studies to date have been performed on endogenous URKL-1 protein expression or function. The crystal structure of the URKL-1 protein suggests homology to a uracil phosphoribosyltransferase in its C terminus, a protein with the ability to salvage endogenous uracil to uridine monophosphate for pyrimidine synthesis (31). The N terminus however, has ATP/GTP binding sites that are similar to binding sites found in uridine kinases (30). Uridine kinases are also involved in pyrimidine synthesis, using ATP to phosphorylate uridine to uridine monophosphate (31, 32). Because of this involvement with pyrimidine synthesis, uridine kinases are involved in cell proliferation and cell survival.

We have also shown with overexpression studies in 293 cells that URKL-1 coprecipitates with NKLAM, indicating that the protein-protein interaction we observed in the yeast two-hybrid system is able to occur in a mammalian system. Additionally, our findings document that the RING fingers of NKLAM interact with URKL-1, with the strongest interaction occurring in the presence of CRD2. This finding is noteworthy because not all substrates of E3 ligases bind to the RING domains of their E3 ligases (33, 34, 35). As examples, both Parkin and Dorfin bind one of their substrates outside of their RING domains; Parkin binds {alpha}-synuclein using its N terminus and Dorfin binds synphilin-1 via its C terminus (9, 12). NKLAM as an E3 ligase, however, is not alone in binding its substrate using its CRDs. The E3 ligase Parkin does bind one of its substrates, CDCrel-1, using its RING finger domains. Not only are the RING fingers of Parkin involved in CDCrel-1 interaction, but like NKLAM and URKL-1, all CRDs of Parkin are able to bind to CDCrel-1, with a preferential binding occurring with CRD3, whereas CRD2 is preferential with URKL-1 binding to NKLAM. In addition, the studies involving Parkin showed enhanced binding of CDCrel-1 when all of the CRDs are present; a similar finding was observed with NKLAM and URKL-1 binding (11).

We have shown that URKL-1 interaction with NKLAM leads to decreased expression of URKL-1. Maximal effects on URKL-1 levels are seen by cotransfecting full-length NKLAM or a form containing all three CRDs, which have the strongest interaction with URKL-1. These results are consistent with the hypothesis that NKLAM ubiquitinates URKL-1, resulting in its enhanced degradation. Ubiquitination studies were then performed, confirming increased ubiquitination of URKL-1 in the presence of NKLAM. We are confident that a significant amount of the ubiquitin smears seen in these blots represents URKL-1, although it is possible that there is some ubiquitinated NKLAM coprecipitating with URKL-1. NKLAM is a much larger protein than URKL-1, with a molecular mass of 150 kDa. Therefore, all of the ubiquitinated material below 150 kDa is most likely URKL-1. The pattern of ubiquitination where NKLAM is present and absent is the same, with increased ubiquitination in the presence of NKLAM, and therefore the majority of the ubiquitinated proteins are most likely to be URKL-1.

Colocalization studies affirmed that both NKLAM and URKL-1 are found in the same region of a cell when they were cotransfected, allowing a protein-protein interaction to occur. This colocalization data is consistent with and support the biochemical data showing a physical interaction between NKLAM and URKL-1. Because we observed URKL-1 in the nucleus when it was transfected alone into 293 cells, we believe its presence in the cytosol, when cotransfected with NKLAM, is due to NKLAM binding to and sequestering URKL-1, prohibiting URKL-1 from entering the nucleus. The prohibition of URKL-1 from entering the nucleus suggests it may be being suppressed from performing its function. Previous studies, involving an overexpression system expressing URKL-1 and EBNA3, also found that URKL-1 entered the nucleus when it was transfected alone, but when it was cotransfected with EBNA3, URKL-1 was sequestered in the cytoplasm (30).

Thus, we have presented in this study that the RING fingers of NKLAM bind to UbcH7 and UbcH8, identifying NKLAM as an E3 ligase. The identification of NKLAM as an E3 ligase suggests that NKLAM interacts with a substrate. As an E3 ligase, this interaction should facilitate the ubiquitination of the substrate protein.

We have shown that a substrate for NKLAM is URKL-1, with this interaction leading to enhanced ubiquitination and subsequent degradation of the URKL-1 protein. We speculate that the interaction of NKLAM and URKL-1 most likely occurs in the target cell. Preliminary data suggest that upon NK-target cell interaction, resulting in granule exocytosis, NKLAM is released from NK cells. Based on preliminary results, it likely enters the target cell and interacts with its substrate, URKL-1. This interaction would result in the ubiquitination and degradation of URKL-1. If URKL-1 is an essential protein in cell proliferation and cell survival, as its homology to uridine kinases may suggest, then its binding to NKLAM in the target cell, resulting in its degradation, would assist in the inhibition of proliferative or survival functions of that cell. This observation is also consistent with the fact that uridine kinases are known to be up-regulated in tumor cells. To prove that NKLAM and URKL-1 interact in target cells, we need to confirm that NKLAM is released from NK cells and migrates into the target cell. We also need to confirm that URKL-1 is present in the target cells and that subsequent entry of NKLAM into the target cell decreases the amount of URKL-1 in that cell.

Previous studies have demonstrated the importance of NKLAM in NK-mediated killing of tumor cells (3). The results presented in our study show that NKLAM functions as an E3 ubiquitin ligase and identifies URKL-1 as a substrate of NKLAM. Additional studies are ongoing to identify the precise mechanism by which NKLAM and URKL-1 function.


   Acknowledgments
 
We thank Dr. Dirk Bohmann of the University of Rochester Medical Center for providing the HA-tagged ubiquitin construct and Chiron for providing the rIL-2.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The content of this research does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred. The authors have no financial conflict of interest.


   Footnotes
 
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.

1 This work was supported by grants from the Veterans Administration, American Heart Association, and the Department of the Army Grant DAMD17-02-1-0571. The U.S. Army Medical Research Acquisition Activity is the awarding and administering acquisition office. Back

2 Address correspondence and reprint requests to Dr. Jacki Kornbluth, Department of Pathology, St. Louis University School of Medicine, 1402 South Grand Boulevard, St. Louis, MO 63104-1004. E-mail address: kornblut{at}slu.edu Back

3 Abbreviations used in this paper: NKLAM, NK lytic-associated molecule; TMD, transmembrane domain; RING, really interesting new gene; IBR, in between RING; CRD, cysteine-rich domain; Ubc, ubiquitin conjugate; URKL-1, uridine kinase like-1; SOD1, superoxide dismutase-1; GRAIL, gene related to anergy in lymphocytes; HA, hemagglutinin. Back

Received for publication August 15, 2005. Accepted for publication March 9, 2006.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Russell, J., T. Ley. 2002. Lymphocyte-mediated cytotoxicity. Annu. Rev. Immunol. 20: 323-370. [Medline]
  2. Lehmann, C., M. Zeis, L. Uharek. 2001. Activation of natural killer cells with interleukin 2 (IL-2) and IL-12 increases perforin binding and subsequent lysis of tumour cells. Br. J. Haematol. 114: 660-665. [Medline]
  3. Kozlowski, M., J. Schorey, T. Portis, V. Grigoriev, J. Kornbluth. 1999. NK lytic associated molecule: a novel gene selectively expressed in cells with cytolytic function. J. Immunol. 163: 1775-1785. [Abstract/Free Full Text]
  4. Portis, T., J. Anderson, A. Esposito, J. Kornbluth. 2000. Gene structure of human and mouse NKLAM, a gene associated with cellular cytotoxicity. Immunogenetics 51: 546-555. [Medline]
  5. Borden, K.. 1998. RING fingers and B-boxes: zinc binding protein-protein interaction domains. Biochem. Cell Biol. 76: 351-358. [Medline]
  6. Saurin, A., K. Borden, M. Boddy, P. Freemont. 1996. Does this have a familiar RING?. Trends Biochem. Sci. 21: 208-214. [Medline]
  7. Martinez-Noel, G., U. Müller, K. Harbers. 2001. Identification of molecular determinants required for interaction of ubiquitin-conjugating enzymes and RING finger proteins. Eur. J. Biochem. 268: 5912-5919. [Medline]
  8. Niwa, J., S. Ishigaki, M. Doyu, T. Suzuki, K. Tanka, G. Sobue. 2001. A novel centrosomal RING-finger protein, Dorfin, mediates ubiquitin activity. Biochem. Biophys. Res. Commun. 281: 706-713. [Medline]
  9. Ita, T., J. Niwa, N. Hishikawa, S. Ishigaki, M. Doyu, G. Sobue. 2003. Dorfin localizes to Lewy bodies and ubiquitylates synphilin-1. J. Biol. Chem. 278: 29106-29114. [Abstract/Free Full Text]
  10. Hishikawa, N., J. Niwa, D. Manabu, T. Ito, S. Ishigaki, Y. Hashizume, G. Sobue. 2003. Dorfin localizes to the ubiquitylated inclusions in Parkinson’s disease, dementia with Lewy bodies, multiple system atrophy, and amyotrophic lateral sclerosis. Am. J. Pathol. 163: 609-619. [Abstract/Free Full Text]
  11. Zang, Y., J. Gao, K. Chung, H. Huang, V. Dawson, T. Dawson. 2000. Parkin functions as an E2-dependent ubiquitin-protein ligase and promotes the degradation of the synaptic vesicle-associated protein, CDCrel-1. Proc. Natl. Acad. Sci. USA 97: 13354-13359. [Abstract/Free Full Text]
  12. Shimura, H., M. Schlossmacher, N. Hattori, M. Frosch, A. Trockenbacher, R. Schneider, Y. Mizuno, K. Kosik, D. Selkoe. 2001. Ubiquitination of a new form of {alpha}-synuclein by Parkin from human brain: implications for Parkinson’s disease. Science 293: 263-269. [Abstract/Free Full Text]
  13. Joazeiro, C., A. Weissman. 2000. RING finger proteins: mediators of ubiquitin ligase activity. Cell 102: 549-552. [Medline]
  14. Pickart, C. M.. 2001. Mechanisms underlying ubiquitination. Annu. Rev. Biochem. 70: 503-533. [Medline]
  15. Fallon, L., F. Moreau, B. Croft, N. Labib, W. Gu, E. Fon. 2002. Parkin and CASK/LIN-2 associate via a PDZ-mediated interaction and are co-localized in lipid rafts and postsynaptic densities in brain. J. Biol. Chem. 277: 486-491. [Abstract/Free Full Text]
  16. Kitada, T., S. Asakawa, N. Hattori, H. Matsumine, Y. Yamamura, S. Minoshima, M. Yokochi, Y. Mizuno, N. Shimizu. 1998. Mutations in the Parkin gene cause autosomal recessive juvenile parkinsonism. Nature 392: 605-608. [Medline]
  17. Shimura, H., N. Hattori, S. Kubo, Y. Mizuno, S. Asakawa, S. Minoshima, N. Shimizu, K. Iwai, T. Chiba, K. Tanaka, T. Suzuki. 2000. Familial Parkinson disease gene product, Parkin, is a ubiquitin-protein ligase. Nat. Genet. 25: 302-305. [Medline]
  18. Imai, Y., M. Soda, R. Takahashi. 2000. Parkin suppresses unfolded protein stress-induced cell death through its E3 ubiquitin-protein ligase activity. J. Biol. Chem. 275: 35661-35664. [Abstract/Free Full Text]
  19. Liu, Y. C.. 2004. Ubiquitin ligases and the immune response. Annu. Rev. Immunol. 22: 81-127. [Medline]
  20. Fang, D., H.-Y. Wang, N. Fang, Y. Altman, C. Elly, Y. C. Liu. 2001. Cbl-b, a RING-type E3 ubiquitin ligase, targets phosphatidylinositol 3-kinase for ubiquitination in T cells. J. Biol. Chem. 276: 4872-4878. [Abstract/Free Full Text]
  21. Fang, D., Y. C. Liu. 2001. Proteolysis-independent regulation of phosphatidylinositol 3-kinase by Cbl-b-mediated ubiquitination in T cells. Nat. Immunol. 2: 870-875. [Medline]
  22. Panigada, M., S. Porcellini, E. Barbier, S. Hoeflinger, P. A. Cazenave, H. Gu, H. Band, H. von Boehmer, F. Grassi. 2002. Constitutive endocytosis and degradation of the pre-T cell receptor. J. Exp. Med. 195: 1585-1597. [Abstract/Free Full Text]
  23. Anandasabapathy, N., G. S. Ford, B. Bloom, C. Holness, V. Paragas, C. Seroogy, H. Skrenta, M. Hollenhorst, C. G. Fathman, L. Soares. 2003. GRAIL: an E3 ubiquitin ligase that inhibits cytokine gene transcription is expressed in anergic CD4+ T cells. Immunity 18: 535-547. [Medline]
  24. Zhao, H., C. C. Li, J. Pardo, P. C. Chu, C. X. Liao, J. Huang, J. G. Dong, X. Zhou, Q. Huang, B. Huang, et al 2005. A novel E3 ubiquitin ligase TRAC-1 positively regulates T cell activation. J. Immunol. 174: 5288-5297. [Abstract/Free Full Text]
  25. Chuang, T., R. J. Ulevitch. 2004. Triad3A, an E3 ubiquitin-protein ligase regulating Toll-like receptors. Nat. Immunol. 5: 495-502. [Medline]
  26. Zhao, C., S. L. Beaudenon, M. L. Kelley, M. B. Waddell, W. Yuan, B. S. Schulman, J. M. Huilbregtse, R. M. Krug. 2004. The UbcH8 ubiquitin E2 enzyme is also the E2 enzyme for ISG15, an IFN-{alpha}beta-induced ubiquitin-like protein. Proc. Natl. Acad. Sci. USA 101: 7578-7582. [Abstract/Free Full Text]
  27. Van Der Reijden, B. A., C. A. J. Erpelinck-Verschueren, B. Lowenberg, J. H. Jansen. 1999. TRIADs: a new class of proteins with a novel cysteine-rich signature. Protein Sci. 8: 1557-1561. [Abstract]
  28. Berg, J., Y. Shi. 1996. The galvanization of biology: a growing appreciation of the roles of zinc. Science 271: 1081-1085. [Abstract]
  29. Freemont, P.. 1993. The RING finger, a novel protein sequence motif related to the zinc finger. Annu. NY Acad. Sci. 684: 174-192. [Medline]
  30. Kashuba, E., V. Kashuba, T. Sandalova, G. Klein, L. Szekely. 2002. Epstein-Barr virus encoded nuclear protein EBNA-3 binds a novel human uridine kinase/uracil phosphoribosyltransferase. BMC Cell Biol. 3: 23[Medline]
  31. Mitterbauer, R., T. Karl, G. Adam. 2002. Saccharomyces cerevisiae URH1 (encoding uridine-cytidine N-ribohydrolase): functional complementation by a nucleoside hydrolase from a protozoan parasite and by a mammalian uridine phosphorylase. Appl. Environ. Microbiol. 68: 1336-1343. [Abstract/Free Full Text]
  32. Van Rompay, A. R., A. Norda, K. Lindén, M. Johansson, A. Karlsson. 2001. Phosphorylation of uridine and cytidine nucleoside analogs by two human uridine-cytidine kinases. Mol. Pharmacol. 59: 1181-1186. [Abstract/Free Full Text]
  33. Moilanen, A., H. Poukka, U. Karvonen, M. Hakli, O. A. Janne, J. J. Palvimo. 1998. Identification of a novel RING finger protein as a coregulator in steroid receptor-mediated gene transcription. Mol. Cell. Biol. 18: 5128-5139. [Abstract/Free Full Text]
  34. Asamitsu, K., T. Tetsuka, S. Kanazawa, T. Okamoto. 2003. RING finger protein AO7 supports NF-{kappa}B-mediated transcription by interacting with transactivation domain of the p65 subunit. J. Biol. Chem. 278: 26879-26887. [Abstract/Free Full Text]
  35. Nie, J., M. A. McGill, M. Dermer, S. E. Dho, C. D. Wolting, C. J. McGlade. 2002. LNX functions as a RING type E3 ubiquitin ligase that targets the cell fate determinant Numb for ubiquitin-dependent degradation. EMBO J. 21: 93-102. [Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
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 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
Citing Articles
Right arrow