Degradation of Trafficking-defective Long QT Syndrome Type II Mutant Channels by the Ubiquitin-Proteasome Pathway*

Mutations in the human ether-a-go-go-related gene (hERG) cause chromosome 7-linked long QT syndrome type II (LQT2). We have shown previously that LQT2 mutations lead to endoplasmic reticulum (ER) retention and rapid degradation of mutant hERG proteins. In this study we examined the role of the ubiquitin-proteasome pathway in the degradation of the LQT2 mutation Y611H. We showed that proteasome inhibitors N-acetyl-l-leucyl-l-leucyl-l-norleucinal and lactacystin but not lysosome inhibitor leupeptin inhibited the degradation of Y611H mutant channels. In addition, ER mannosidase I inhibitor kifunensine and down-regulation of EDEM (ER degradation-enhancing α-mannosidase-like protein) also suppressed the degradation of Y611H mutant channels. Proteasome inhibition but not mannosidase inhibition led to the accumulation of full-length hERG protein in the cytosol. The hERG protein accumulated in the cytosol was deglycosylated. Proteasome inhibition also resulted in the accumulation of polyubiquitinated hERG channels. These results suggest that the degradation of LQT2 mutant channels is mediated by the cytosolic proteasome in a process that involves man-nose trimming, polyubiquitination, and deglycosylation of mutant channels.

Mutations in the human ether-a-go-go-related gene (hERG) cause chromosome 7-linked long QT syndrome type II (LQT2). We have shown previously that LQT2 mutations lead to endoplasmic reticulum (ER) retention and rapid degradation of mutant hERG proteins. In this study we examined the role of the ubiquitin-proteasome pathway in the degradation of the LQT2 mutation Y611H. We showed that proteasome inhibitors N-acetyl-L-leucyl-L-leucyl-L-norleucinal and lactacystin but not lysosome inhibitor leupeptin inhibited the degradation of Y611H mutant channels. In addition, ER mannosidase I inhibitor kifunensine and down-regulation of EDEM (ER degradation-enhancing ␣-mannosidase-like protein) also suppressed the degradation of Y611H mutant channels. Proteasome inhibition but not mannosidase inhibition led to the accumulation of full-length hERG protein in the cytosol. The hERG protein accumulated in the cytosol was deglycosylated. Proteasome inhibition also resulted in the accumulation of polyubiquitinated hERG channels. These results suggest that the degradation of LQT2 mutant channels is mediated by the cytosolic proteasome in a process that involves mannose trimming, polyubiquitination, and deglycosylation of mutant channels.
Long QT syndrome is a cardiac disorder characterized by prolongation of QT intervals on the electrocardiogram and a high risk of sudden death due to ventricular arrhythmias. The chromosome 7-linked form of the inherited long QT syndrome (long QT syndrome type II (LQT2)) 1 is caused by mutations of human ether-a-go-go-related gene (hERG) (1). hERG encodes the pore-forming subunit of the rapidly activating-delayed rectifier K ϩ channel (I Kr ) in the heart (2,3). Studies of LQT2 mutant channels expressed heterologously in oocytes or mammalian cells have shown that LQT2 mutations cause hERG channel dysfunction by multiple mechanisms including defective protein trafficking, abnormal gating or permeation, and dominant negative suppression of wild type hERG channel function (4).
We have previously shown that the LQT2 mutation Y611H is trafficking-defective (5). This mutant expresses only the immature form of hERG channel protein, which fails to reach the plasma membrane. The Y611H mutant protein is retained in the endoplasmic reticulum (ER) and rapidly degraded. It is well recognized that newly synthesized proteins in the ER are under stringent surveillance by a quality control system (6 -8). The ER quality control system ensures that only properly folded and assembled proteins are exported from the ER to the Golgi. Misfolded and unassembled proteins are retained in the ER and eventually degraded by a process termed ER-associated degradation (ERAD). According to current models, ERAD substrates undergo retrotranslocation or dislocation from the ER to the cytosol, in which they are degraded by the proteasome (8). In most cases, the degradation by the proteasome is preceded by polyubiquitination. In addition, mannose trimming by ER mannosidase I and deglycosylation by cytosolic peptide:Nglycanase (PNGase) are important processes in proteasomal degradation of glycoproteins (8). Recently EDEM (ER degradation-enhancing ␣-mannosidase-like protein) has been shown to play a crucial role in the degradation of misfolded glycoproteins (9,10).
The importance of ERAD is underscored by the fact that proteasomal degradation of misfolded proteins has been shown in an increasing number of human diseases. The role of ERAD in the degradation of misfolded proteins has been studied extensively in cystic fibrosis and ␣1-antitrypsin deficiency (11,12). Defective trafficking of mutant channels to the plasma membrane has been shown as the most common mechanism of hERG channel dysfunction in LQT2. More than 15 LQT2 mutations have been reported to cause defective trafficking of mutant channels (13,14). These LQT2 mutations lead to misfolding, ER retention, and the degradation of mutant hERG channels (5,13,14). Although degradation of mutant hERG channels by the proteasome has been suggested (15,16), little is known about how mutant hERG channels are targeted to the cytosolic proteasome for degradation. In this study, we analyzed the role of the ubiquitin-proteasome pathway in the degradation of the LQT2 mutant Y611H. Our results suggest that the mutant hERG protein is targeted to the proteasome for degradation by dislocation from the ER membrane to the cytosol. The proteasome-dependent degradation involves mannose trimming, ubiquitination, and deglycosylation of mutant channels.

EXPERIMENTAL PROCEDURES
Reagents and Cell Transfection-The proteasome inhibitors lactacystin and N-acetyl-L-leucyl-L-leucyl-L-norleucinal (ALLN) were purchased from Kamiya Biomedical Co. (Seattle, WA) and Calbiochem, respectively. The ER mannosidase I inhibitor kifunensine was purchased from Toronto Research Chemicals, Inc. (North York, Ontario, Canada). Anti-hERG antibody was raised against the hERG-thioredoxon fusion protein as described previously (5). Anti-ubiquitin monoclonal antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anticalnexin antibody was purchased from Stressgen (Victoria, British Columbia, Canada). Expre 35 S 35 S protein labeling mix was purchased from PerkinElmer Life Sciences. HEK293 cells stably expressing wild type hERG or the Y611H mutant were cultured as described previously (5). For down-regulation of EDEM, HEK293 cells were cotransfected with 6 g of pcDNA3-Y611H and 10 g of pcDNA3-siEDEM or pcDNA3-siGFP utilizing Lipofectamine 2000 reagent (Invitrogen) (9).
Western Blot Analysis-Western blot procedures were described previously (5). Briefly, cells were lysed in lysis buffer containing 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, and proteinase inhibitors. After centrifugation at 14,000 ϫ g for 10 min at 4°C, cell lysates were subjected to 7.5% SDS-PAGE and electrophoretically transferred onto nitrocellulose membranes. hERG proteins were detected with anti-hERG antibody (1:10000 dilution) and visualized with a horseradish peroxidase-conjugated second antibody and ECL detection kit. Deglycosylation of hERG proteins by PNGase was performed as described previously (17).
Metabolic Labeling and Immunoprecipitation-Cells in 35-mm plates were pulse-labeled for 1 h in methionine-and cysteine-free Dulbecco's modified Eagle's medium containing 110 Ci/ml [ 35 S]methionine/cysteine and chased in Dulbecco's modified Eagle's medium with 2 mM unlabeled methionine and cysteine for time intervals up to 24 h. For drug treatment, 50 M ALLN, 20 M lactacystin, 100 M leupeptin, or 100 M kifunensine was added 1 h before pulse labeling and was continuously included in the culture medium during the pulse and the chase periods. At the end of chase period, cells were lysed in 600 l of immunoprecipitation buffer containing 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mg/ml bovine serum albumin, and protease inhibitors (100 M phenylmethylsulfonyl fluoride, 1 g/ml pepstatin A, 1 g/ml leupeptin, 4 l/ml aprotinin). After centrifugation at 14,000 ϫ g for 10 min at 4°C, cell lysates were precleared by incubation with protein A-agarose beads (Pierce). hERG antiserum (1:100 dilution) was then added, and the mixture was incubated at 4°C overnight. The antigen-antibody complexes were isolated with protein A-agarose beads, subjected to 7.5% SDS-PAGE, and visualized with autoradiography. For quantitative analysis, the dried gels were exposed to a K-screen and quantified by phosphorimaging analysis using Quantity One soft ware (Bio-Rad, Molecular Imager FX).
Detection of Ubiquitinated hERG Proteins-Cells were lysed in 600 l of immunoprecipitation buffer. Cell lysates containing an equal amount of protein were immunoprecipitated with anti-hERG antibody at 1:100 dilution as described above. The immunoprecipitates were subjected to 5.5% SDS-PAGE and electrophoretically transferred onto nitrocellulose membranes. The membranes were probed with monoclonal anti-ubiquitin antibody at 1:500 dilution. The membranes were also probed with anti-hERG antibody after stripping of the bound anti-ubiquitin antibodies.
Subcellular Fractionation-Cells were homogenized in 600 l of homogenization buffer containing 10 mM Tris-HCl. pH 7.4, 1 mM EDTA, and proteinase inhibitors with a Dounce homogenizer. Cells metabolically labeled by [ 35 S]methionine/cysteine were homogenized in 600 l of homogenization buffer by freezing and thawing the cell suspension and passing it 10 times through a 25-gauge needle as described previously (18). Homogenates were centrifuged at 1000 ϫ g for 10 min to remove unbroken cells and nuclei. The postnuclear supernatant was then centrifuged at 10,000 ϫ g for 30 min. The supernatant from the 10,000 ϫ g centrifugation was further centrifuged at 100,000 ϫ g for 60 min. Membrane pellets from 10,000 ϫ g and 100,000 ϫ g spins were washed with homogenization buffer and solubilized in a buffer containing 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mg/ml bovine serum albumin, and protease inhibitors. The supernatant from the 100,000 ϫ g centrifugation (cytosolic fraction) was adjusted to a final concentration of 150 mM NaCl, 1% Triton X-100, and 1 mg/ml bovine serum albumin. Solubilized membranes and cytosolic fractions were analyzed by immunoprecipitation and immunoblot as described above. Equal fractions from each pellet and supernatant were analyzed.
RNase Protection Assay-Twenty-four h after transfection with siRNAs, the cells were harvested and total RNA isolated using the RNeasy method (Qiagen, Valencia, CA). RNase protection probes were made by subcloning a 452-nucleotide cDNA fragment corresponding to 1386 -1838 bp of EDEM and a 290-nucleotide cDNA fragment corresponding to 466 -756 bp of actin into pCRII vector (Invitrogen). The antisense RNA probes were transcribed using a MAXIscript in vitro transcription kit (Ambion, Austin, TX) and biotin-16-UTP (Roche Applied Science). Thirty g of total RNA or yeast tRNA were analyzed with EDEM riboprobe, and 10 g of total RNA or yeast tRNA were analyzed by actin riboprobe using RPAII and BrightStart BioDetect kits (Ambion). Yeast tRNA was used as a control for the complete digestion of the probes by RNase.

RESULTS
The involvement of the proteasome in the degradation of hERG channel was initially assessed by examining the effect of proteasome inhibitors ALLN and lactacystin on the steady state protein levels of wild type hERG and the LQT2 mutant Y611H. Fig. 1 shows Western blot analysis of hERG channel in HEK293 cells stably transfected with wild type hERG or Y611H. Wild type hERG expressed two protein bands, an upper band of 155 kDa and a lower band of 135 kDa, whereas the Y611H mutant expressed only the 135-kDa lower band. We have previously shown that the upper band is the fully glycosylated, mature form of the channel protein located in the plasma membrane, and the lower band is a core-glycosylated, immature form of the channel protein located in the ER (5). Exposure to 50 M ALLN or 20 M lactacystin for 24 h significantly increased protein levels of the immature form without an apparent increase in the mature form of wild type hERG or Y611H. The lysosome inhibitor leupeptin had no effect on protein levels of wild type hERG or Y611H. These results strongly suggest that the proteasome is involved in the degradation of the immature form of hERG channels.
To further examine the role of the proteasome in the degradation of hERG channels, we performed pulse-chase analysis. In these experiments, cells expressing wild type hERG or Y611H were metabolically labeled with [ 35 S]methionine and [ 35 S]cysteine for 1 h and chased with unlabeled methionine and cysteine for up to 24 h under control conditions or in the presence of 50 M ALLN ( Fig. 2A). Under control conditions, wild type hERG was initially synthesized as the 135-kDa immature form, which was gradually converted to the 155-kDa mature form. The kinetics of hERG channel maturation and degradation were analyzed by phosphorimaging quantification. As shown in Fig. 2B, the conversion of wild type hERG from the immature to the mature form became obvious at 2 h of chase and reached a maximum at 8 h. About 60% of the immature form was converted to the mature form. In the presence of ALLN, there was a significant increase in the amount of labeled immature form of wild type hERG. The increase in the immature form, however, did not result in an increase in the mature form during the chase. In fact, ALLN significantly reduced the efficiency of hERG channel maturation, as only about 20% of the immature form was converted to the mature form in the presence of ALLN.
In pulse-chase experiments, the Y611H mutant was initially synthesized as the 135-kDa immature form and was not converted to the mature form during the chase (Fig. 2C). Rather, the mutant protein underwent progressive degradation with complete disappearance of the 135-kDa band by 24 h. Kinetic analysis shown in Fig. 2D revealed that the immature form of Y611H was degraded with an estimated half-life of 4.5 h. ALLN inhibited the degradation of the Y611H mutant channel, thereby prolonging its half-life to 12 h (Fig. 2, C and D). However, inhibition of the degradation did not cause conversion of the immature to the mature form. We also performed pulsechase experiments to study the effects of the specific proteasome inhibitor lactacystin and the lysosome inhibitor leupeptin on the degradation of the Y611H mutant channel. In these experiments, cells expressing Y611H were labeled for 1 h and chased for 8 h under control conditions or in the presence of 10 M lactacystin or 100 M leupeptin. As shown in Fig. 3, lactacystin inhibited the degradation of the Y611H mutant. After 8 h of chase, 26% of the 35 S-labeled hERG protein remained under control conditions, whereas in the presence of lactacystin 68% of the 35 S-labeled hERG protein remained. In contrast, leupeptin had no effect on the degradation of the Y611H mutant. After 8 h of chase, 17 and 21% of the 35 S-labeled hERG protein remained under control conditions and in the presence of leupeptin, respectively. Thus, these pulse-chase results are consistent with our Western blot data and suggest that the cytosolic proteasome is involved in the degradation of the Y611H mutant channel.
Degradation of ERAD substrate proteins by the cytosolic proteasome implies that they must be dislocated from the ER to the cytosol. It has been reported that some membrane proteins can be dislocated from the ER to the cytosol in an intact form when proteasome activity is inhibited (18 -20). To test whether hERG channel proteins undergo similar dislocation, we performed subcellular fractionation experiments (Fig. 4A). After removal of unbroken cells and nuclei by centrifugation at 1000 ϫ g, postnuclear supernatant was separated into three fractions: 10,000 ϫ g pellet (crude membranes), 100,000 ϫ g pellet (residual microsome membranes), and 100,000 ϫ g supernatant (cytosolic fraction) (18). Western blot analysis showed that in the absence of ALLN, both wild type hERG and Y611H mutant proteins were recovered only from the 10,000 ϫ g pellet membrane fraction. In the presence of ALLN, they were also present in the 100,000 ϫ g supernatant cytosolic fraction. As a control, the ER membrane protein calnexin was shown to be present only in the 10,000 ϫ g pellet membrane fraction. Thus, these results suggest that an intact form of hERG channel can be dislocated from the ER to the cytosol. Because ERAD substrate proteins dislocated to the cytosol have been reported to undergo deglycosylation prior to degradation by the proteasome (18,(21)(22)(23), we examined whether the hERG protein recovered from the cytosolic fraction represents the deglycosylated form. As shown in Fig. 4B, the Y611H mutant protein recovered from the cytosolic fraction had a molecular mass of 132 kDa and was resistant to PNGase treatment. In contrast, the mutant protein recovered from membrane fraction had a molecular mass of 135 kDa and was sensitive to PNGase, with its molecular mass being reduced to 132 kDa. These results suggest that Y611H mutant protein recovered from the cytosolic fraction represents the deglycosylated form.
To demonstrate that the presence of hERG channels in the cytosol is due to dislocation of those retained in the ER, rather than the failure of incorporation of newly synthesized hERG channels into the ER, we performed pulse-chase experiments followed by cell fractionation. Cells expressing the Y611H mutant were labeled for 1 h and chased for up to 24 h in the presence of 50 M ALLN. hERG proteins were immunoprecipitated from the 10,000 ϫ g pellet membrane fraction and the 100,000 ϫ g supernatant cytosolic fraction. As shown in Fig. 5, nearly all 35 S-labeled hERG protein was present in the membrane fraction at the end of the pulse labeling. The amount of hERG protein in the membrane fraction gradually decreased during the chase, whereas the amount of hERG protein in the cytosolic fraction gradually increased, reaching a maximum at 8 h of chase. These pulse-chase experiments strongly suggest that newly synthesized hERG channels are incorporated into the ER membrane and then dislocated from the ER into the cytosol for proteasomal degradation. The results also show that despite the presence of hERG protein in the cytosolic fraction, the majority of hERG protein was still present in the membrane fraction during the chase, indicating that the rate of dislocation from the membrane to the cytosol is reduced in the presence of proteasome inhibitors.
Many ERAD substrate proteins are conjugated with polyubiquitin chains that serve as a signal for proteasomal degradation. To demonstrate that the hERG channel is modified by polyubiquitin chains, cell lysates were immunoprecipitated with anti-hERG antibody and then immunoblotted with antiubiquitin antibody. In these experiments, untransfected HEK cells and cells expressing wild type hERG or Y611H were treated with or without ALLN for 24 h. As shown in Fig. 6A, treatment of cells with ALLN resulted in the accumulation of high molecular weight ubiquitinated hERG channels in wild type hERG and in Y611H-transfected cells but not in untransfected cells. Ubiquitinated hERG channels were barely detectable in untreated cells, suggesting that ubiquitinated hERG channel protein is normally degraded by the proteasome in the absence of proteasome inhibitor. To determine the subcellular localization of the ubiquitinated hERG channels, cell fractionation experiments were carried out in which hERG channels were immunoprecipitated from membrane and cytosolic fractions of ALLN-treated cells and then immunoblotted with antiubiquitin antibody. As shown in Fig. 6B, ubiquitinated hERG channels were present in both membrane and cytosolic fractions. This result suggests that the ubiquitination of hERG channels starts before they are released into the cytosol.
The hERG channel protein is modified by N-linked glycosylation (17,24). Several studies have shown that the trimming of mannose by ER mannosidase I is required for proteasomal degradation of glycoproteins (25)(26)(27)(28)(29). We examined the role of mannose trimming in the degradation of the Y611H mutant channel by using the ER mannosidase I inhibitor kifunensine. As shown in Fig. 7 accumulation of hERG channel protein in the cytosol (Fig. 7B). These results suggest that mannose trimming is an important process in the degradation of the Y611H mutant channel.
Recently, it has been shown that EDEM plays an important role in proteasomal degradation of misfolded glycoproteins (9,10). EDEM is an ER type II membrane protein homologous to ␣-ER mannosidase but lacking mannosidase activity. To test whether EDEM is involved in the degradation of mutant hERG channels, we used RNA-mediated interference to reduce the intracellular concentration of EDEM. In these experiments, we cotransfected the Y611H mutant with an EDEM-directed siRNA (siEDEM) or a control GFP-directed siRNA (siGFP) (9) into HEK293 cells and performed pulse-chase analysis 24 h after transfection. A RNase protection assay revealed a decrease in EDEM mRNA in siEDEM-transfected cells (57 Ϯ 5%) compared with siGFP control (n ϭ 3) (Fig. 8A). Pulse-chase experiments showed that down-regulation of EDEM inhibited the degradation of the Y611H mutant channel (Fig. 8, B and C). Down-regulation of EDEM led to a prolongation of the half-life of the Y611H mutant channel from 3.2 to 9.5 h, similar to that found in ALLN-treated samples. DISCUSSION The present experiments demonstrate that degradation of LQT2 mutant hERG channels is mediated by the ubiquitinproteasome pathway. We show that the misfolded hERG protein is targeted to the proteasome for degradation by dislocation from the ER membrane to the cytosol. The proteasomedependent degradation involves mannose trimming, ubiquitination, and deglycosylation of mutant channels.
Our pulse-chase experiments revealed important information about the kinetics of the biogenesis of the wild type hERG channel. About 60% of the core-glycosylated immature form of wild type hERG is processed to the complex-glycosylated mature form. Compared with other polytopic membrane proteins, the hERG channel has intermediate maturation efficiency. For example, the maturation efficiency of the Shaker potassium channel is near 100%, whereas less than 40% of newly synthesized wild type cystic fibrosis transmembrane conductance and delta opioid receptor are processed to the mature forms, and the remaining immature forms are degraded by the ubiquitinproteasome pathway (18,30,31). Our data also suggest that a fraction of wild type hERG protein that fails to mature is degraded by the ubiquitin-proteasome pathway. In the presence of proteasome inhibitors, there is an increase in the immature form of wild type hERG protein. In addition, proteasome inhibition results in the accumulation of ubiquitinated wild type hERG channels. In the case of the delta opioid receptor, proteasome inhibition leads to an increase in the cell surface expression of functional receptors (18). In hERG, however, inhibition of the proteasome does not result in an increase in the mature form. Similarly, proteasome inhibition increases the immature but not the mature form of wild type cystic fibrosis transmembrane conductance (11).
Degradation of ER luminal and membrane proteins by the cytosolic proteasome requires dislocation of substrates from the ER to the cytosol. This retrograde transport is believed to occur through the Sec61p ER translocation complex (32,33). However, the mechanism of dislocation and the link between dislocation and proteolysis by the proteasome are controversial. It has been reported that dislocation and degradation by the proteasome are tightly coupled events in several ERAD substrates, suggesting that active proteasome is required for the extraction of ERAD substrates from the ER (23, 34 -36). In this model, ER proteins are directly dislocated from the ER membrane into the membrane-bound proteasome for degradation. The 19 S cap ATPases may provide the driving force for the membrane extraction. However, some ERAD substrates have been reported to accumulate in the cytosol when proteasome activity is inhibited (18 -23). This suggests that proteasome activity is not required for the dislocation of these ERAD substrates from the ER to the cytosol. Recently, it has been shown that the cdc48/p97⅐Ufd1⅐Npl4 complex is involved in the dislocation of ERAD substrates from the ER to the cytosol (37,38). It is proposed that the ATPase activity of cdc48/p97 is involved in extracting proteins from the ER membrane (39). In our experiments, a fraction of the hERG channel protein accumu- lates in the cytosol following inhibition of the proteasome. However, the fact that the rate at which the hERG channel protein accumulates in the cytosol is much slower than the rate of degradation in the absence of proteasome inhibitors indicates that the efficiency of dislocation is reduced in the presence of proteasome inhibitors. A similar reduction in dislocation efficiency by proteasome inhibitors has been reported for other ERAD substrates (18,20,23). Thus, although proteasome activity is not obligatory for the dislocation of hERG channel, inhibition of the proteasome affects the efficiency of hERG channel dislocation.
Our present results show that mannose trimming by ER mannosidase I is required for the degradation of Y611H mutant channels. We have previously shown that the hERG channel undergoes N-linked glycosylation at residue Asp-598 in the extracellular domain between transmembrane segments 5 and 6 (24). N-Linked glycosylation begins with the en bloc transfer of a preassembled oligosaccharide chain, Glu 3 Man 9 GluNA 2 , to asparagine residues of glycoproteins in the ER (40). The oligosaccharide chain is subject to the trimming of glucose and mannose residues by a number of glycosidases including glucosidases I and II and ER mannosidases I and II. Several reports have indicated the prerequisite of mannose trimming by ER mannosidase I in proteasomal degradation of glycoproteins (25)(26)(27)(28)(29). In our experiments, the ER mannosidase I inhibitor kifunensine suppresses the degradation of Y611H mutant channels. Unlike proteasome inhibitor treatment, in which a fraction of hERG protein accumulates in the cytosol, kifunensine treatment does not result in the cytosolic accumulation of hERG protein. The results suggest that mannose trimming contributes to the initial targeting process that is upstream of the dislocation of the hERG protein in the ubiquitin-proteasome pathway. This process may involve the interaction of the mannose-trimmed hERG channel with EDEM because down-regulation of EDEM by RNA-mediated interference suppresses the degradation of Y611H mutant channels. Our results support the proposal that EDEM functions as a lectin that recognizes the mannose-trimmed misfolded glycoproteins and promotes their degradation by the cytosolic proteasome (9,10,41).
The present experiments show that the hERG channel is modified by ubiquitination. The requirement of ubiquitination for proteasomal degradation has been shown in many ERAD substrates (42). Recent studies also suggest that attachment of polyubiquitin chains to ERAD substrates is required for their dislocation from the ER to the cytosol (43)(44)(45). In our experiments, ubiquitinated hERG channels were recovered from both membrane and cytosolic fractions, suggesting that ubiquitination of hERG channels starts prior to their dislocation from the ER to the cytosol. Thus, it is possible that ubiquitin modification may contribute to dislocation as well as proteasomal degradation of hERG channels. In addition to ubiquitin modification, hERG channel accumulated in the cytosol was also found to be deglycosylated. The presence of deglycosylated intermediates in the cytosol has been reported for other glycosylated ERAD substrates (18,(21)(22)(23). These intermediates are believed to be the result of deglycosylation of ERAD substrates by a cytoplasmic PNGase prior to their degradation by the proteasome (21,22). Recently, genes encoding yeast cytoplasmic PNGase PNG1 and its mouse homologue mPNG1 have been identified (46). Both Png1p and mPng1p display N-glycanase activity toward intact glycoproteins and are believed to be responsible for the deglycosylation of ERAD substrates in the cytosol (47). The cytoplasmic PNGase has been shown to interact with the proteasome and may play an important role in proteasome-mediated degradation of misfolded glycoproteins (48,49).
Defective trafficking of mutant channels to the plasma membrane has been increasingly recognized as an important mechanism of hERG channel dysfunction in LQT2. Trafficking-defective LQT2 mutations lead to misfolding, ER retention, and degradation of mutant hERG channels. Our present study suggests that the ubiquitin-proteasome pathway plays an important role in the ER retention and degradation of LQT2 mutant channels. Because hERG belongs to a large superfamily of voltage-gated potassium channels and defective trafficking of disease-causing mutations has been reported in several voltage-gated potassium channels such as Kv1.1 and KvLQT1 (50 -52), elucidating the mechanisms underlying the degradation of LQT2 mutant channels will increase our knowledge of the pathogenesis of a variety of human diseases involving voltagegated potassium channels.