The opioid ligand binding of human μ-opioid receptor is modulated by novel splice variants of the receptor

BBRC Biochemical and Biophysical Research Communications 343 (2006) 1132–1140 www.elsevier.com/locate/ybbrc

The opioid ligand binding of human l-opioid receptor is modulated by novel splice variants of the receptor Hack Sun Choi *, Chun Sung Kim, Cheol Kyu Hwang, Kyu Young Song, Wei Wang, Yu Qiu, Ping-Yee Law, Li-Na Wei, Horace H. Loh Department of Pharmacology, Medical School, University of Minnesota, Minneapolis, MN 55455, USA Received 3 March 2006 Available online 23 March 2006

Abstract The pharmacological actions of morphine and morphine-like drugs, such as heroin, mediate primarily through the l-opioid receptor (MOR). It has been proposed that the functional diversity of MOR may be related to alternative splicing of the MOR gene. Although a number of MOR mRNA splice variants have been reported, their biological function has been controversial. In this study, two novel splice variants of the human MOR gene were discovered. Splice variants 1 and 2 (here called the SV1 and SV2) retain different portions of intron I. In vitro translation of SV1 and SV2 produced proteins with the predicted molecular weights. The splice variant proteins were identical to the wild-type MOR-1 up to the first transmembrane domains, but were different after the first intracellular loop domains. SV1 and SV2 of hMOR were present in human neuroblastoma NMB cells and human whole brain confirmed by RT-PCR. In a receptor binding assay, cells expressing the SV1 and SV2 do not exhibit binding to [3H]diprenorphine. The formations of MOR Æ SV1 and MOR Æ SV2 heterodimers were demonstrated by co-immunoprecipitation and bioluminescence resonance energy transfer between MOR and splice variants. Co-transfection of MOR-GFP and SV-DsRed gene showed that MOR and SV protein co-localized at the cytoplasmic membrane. In NMB cells expressing human MOR gene, transfection of SV1 or SV2 reduced binding activity of the endogenous MOR. These data support a potential role of SV1 and SV2 proteins as possible biological modulator of human l-opioid receptor.  2006 Elsevier Inc. All rights reserved. Keywords: MOR, l-Opioid receptor; BRET, Bioluminescence resonance energy transfer; Rluc, Renilla luciferase; DsRed, Destabilized red fluorescence protein; SV, Splice variant; TM, Transmembrane domain

Morphine has been used as a potent clinical analgesic for pain killing, but has serious limitations such as tolerance and dependence. The opioid receptors are classified into three major types (l, d, and j) that were determined by numerous pharmacological reports and molecular cloning [1]. All three types of opioid receptors belong to the superfamily of G-protein-coupled receptors (GPCRs). Previous studies suggested that the l-opioid receptor (MOR) plays important roles in morphine-induced analgesia, tolerance, and dependence as indicated from pharmacological studies and analysis of MOR knockout mice [2,3]. Upon binding with opioids, MOR is able to couple to G-proteins and to regulate adenylyl cyclase, intracellular calcium, *

Corresponding author. Fax: +1 612 625 8408. E-mail address: [email protected] (H.S. Choi).

0006-291X/$ - see front matter  2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.03.084

inwardly rectifying potassium channels, MAP kinase, and other messengers, which further trigger a cascade of intracellular events [4]. MOR is expressed mainly in the central nervous system (CNS), but densities of MOR vary in different parts of the brain that may be related to differing roles of MOR in specific brain regions [5]. Furthermore, MOR is also expressed in other tissues, such as immune cells [6]. The presence of MOR in immune cells can elucidate the reason why drug abusers (opioid and heroin) become more susceptible to external pathogens after weakening the body’s immune system by chronic drug use, especially well described in HIVinfected opiate users [7–9]. The l-opioid receptor, MOR-1, was the first cloned l-opioid receptor cDNA and consisted of 4 exons [10–12]. Alternative splicing variants of MOR have been

H.S. Choi et al. / Biochemical and Biophysical Research Communications 343 (2006) 1132–1140

identified in human [13,14], rat [15], and mouse tissues [16– 19]. It has been reported that the mouse MOR genome is greater than 250 kb long, contains at least 17 exons, and undergoes extensive alternative splicing [20]. Antisense mapping studies in both mice [21] and rats [22] supported the possibility that the various subtypes of MOR could result from alternative splicing of the MOR gene. Several splice variants of human MOR have been identified. MOR-1A is an intron-retention variant which lacks exon 4 [13]. MOR-1O and MOR-1X contain exons 1, 2, and 3 of the original hMOR-1, and exon O or exon X as the alternative fourth exon, respectively [14]. They share the same first three exons that encode all seven transmembrane domains. The only difference among these variants is in a carboxyl terminal portion of the receptors. All three variants were detected in human neuroblastoma BE (2) C cells and human brain [23]. Despite the report of these MOR splicing variants, it remains largely unknown as to how the generation of variants is regulated. Previous studies have reported that MORknockout mice lacking exons 2 and 3 of the MOR gene exhibit a loss of MOR-mediated G-protein activation in the pons/medulla [24]. MOR agonists (such as DAMGO, endomorphin-1, and endomorphin-2) could not produce an increasing of GTP binding to the pons/medulla in MOR (exon 2 and 3)-knockout mice, but b-endorphin could increase GTP binding to the pons/medulla in MOR (exon 2 and 3)-knockout mice, suggesting that the portion of only exon 1 containing transmembrane domain 1 (TM1) may have a functional role in opioid signaling. However, this is largely controversial. GPCRs have been shown to dimerize/oligomerize. Receptor oligomerization is essential for receptor function, i.e., for the GABAB [25], metabotropic glutamate [26], taste receptor [27], and rhodopsin [28] as well as opioid receptors, which has been shown to form heterodimers/oligomeric structures in native disk membrane [29]. Moreover, oligomerization has been shown to play an important modulatory role. Dimerization of opioid receptors has been shown to alter opioid ligand properties and affect receptor trafficking both in vitro [30–32] and in vivo [33]. In our study, two novel human MOR splice variants that contain exon 1 and different size of intron 1 were identified. The SV1 and SV2 were detected in human neuroblastoma NMB cells and human brain. Furthermore, we report here that they hetero-dimerized with and modulate the binding activity of the wild-type MOR. Materials and methods Materials. Mammalian expression vectors (pCMV4 and pCMV5) and fluorescent protein vector (pDsRed2) were from Stratagene (La Jolla, CA) and BD Bioscience (Palo Alto, CA). GFP2 and Renilla luciferase protein expression vectors (GFP-N1 and pRL-SV40) were from Perkin-Elmer Life Sciences (Torrance, CA) and Promega (Madison, WI), respectively. [3H]Diprenorphine (60 Ci/mmol) was from Amersham Pharmacia Biotech (Piscataway, NJ). DeepBlue C was from Perkin-Elmer Life Sciences (Torrance, CA). Anti-FLAG monoclonal antibody (M2) was from Sigma.

1133

Materials for tissue culture were from Life Technologies (Rockville, MD). All other reagents were from Sigma (St. Louise, MO). Cell culture and transfection. Human neuroblastoma NMB cells were cultured as previously described [34]. HEK-293 cells were cultured in minimal essential medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 10 lg/ml streptomycin. For transfection, cells were plated in 6-well dishes at a concentration of 1 · 106 cells/well and cultured overnight before transfection. Various plasmids at equimolar concentrations were used with Effectene transfection reagent (Qiagen) according to the manufacturer’s manual [35]. Reverse transcription-PCR of human MOR and splice variants (SV1 and SV2). Total RNA isolated from human neuroblastoma NMB cells according to the supplier’s protocol (TRI Reagent, Molecular Research Center) and human brain total RNA (Ambion) were analyzed by reverse transcription-PCR (RT-PCR) using primers, as follows. Primers specific to MOR mRNA were primer P1 (5 0 -GATCATGGCCCTCTACTCCA-3 0 , located at position 216 in exon 1) and primer P2 (5 0 -GCATTTCGGG GAGTACGGAA-3 0 , located at position 557 in exon 2 to avoid amplification of genomic DNA). The PCR cycle conditions for human MOR consisted of 95 C for 60 s, 60 C for 60 s, and 72 C for 60 s followed by a 10-min extension at 72 C (34 cycles) and for splice variants consisted of 95 C for 60 s, 60 C for 60 s, and 72 C for 60 s, followed by a 10-min extension at 72 C (30 cycles). The PCR products of partial human MOR and splice variants (SV1 and SV2) gene were cloned into the pCRII-TOPO vector (Invitrogen) and sequenced. Identification and cloning of human MOR, SV1, and SV2. To obtain a full-length cDNA clone of human MOR (hMOR), a sense primer P3 covering start codon in exon 1 (5 0 -TCAGTACCATGGACAGCAGC-3 0 ) and an antisense primer containing the MOR gene termination site (5 0 -TGTTAGGGCAACGGAGCA-3 0 ) were used in a PCR with the firststrand cDNA reverse-transcribed from total RNA of human neuroblastoma NMB cells as the template. To obtain cDNA clones of SV1 and SV2, primers specific to human SV1 and SV2 mRNA, the above sense primer P3 and an antisense primer P4 (5 0 -TATCAGTCTTCTTGGTAGCAT GTCA-3 0 , located at position 146 in intron1) were used. RT-PCRs were performed using the QIAGEN OneStep RT-PCR kit. The PCR products of the full-length hMOR (1.2 kb), SV1 (400 bp), and SV2 (450 bp) were amplified, cloned into the pCRII-TOPO vector, and sequenced. In vitro transcription/translation of human SV1 and SV2. The cDNA fragments containing the human MOR, SV1, and SV2 in pCRII-TOPO were subcloned into the EcoRI site of pcDNA3.1 (Invitrogen). The resulting plasmids, pcDNA3.1-hMOR, pcDNA3.1-SV1, and pcDNA3.1SV2, were transcribed and translated in vitro with a TNT-coupled reticulocyte lysate kit (Promega). Briefly, the plasmids were incubated with T7 RNA polymerase containing the above TNT kit and 0.04 mCi [35S]methionine at 30 C for 1 h. The translation products were separated on a 14% SDS–polyacrylamide gel, and the gel was visualized by autoradiography. Receptor constructs. The human MOR, SV1, and SV2 coding sequences without their stop codon were amplified from the receptor expression plasmids using sense and antisense primers harboring unique cloning site (HindIII and SalI for MOR, SV1, and SV2). The PCR fragments were then inserted in-frame into pCMV4 (5) and GFP-N1 to yield constructs named hMOR (pCMV4), SV1 (pCMV5), SV2 (pCMV5), and hMOR-GFP. Renilla luciferase gene of pRL-SV40 was inserted in-frame into hMOR (pCMV4), SV1 (pCMV5), and SV2 (pCMV5) to yield constructs named hMOR-Rluc, SV1-Rluc, and SV2-Rluc. The sequences of all constructs were confirmed by DNA sequencing. Radioligand binding. NMB cells containing endogenous opioid receptor and HEK cells containing hMOR, SV1, and SV2 gene were harvested and resuspended in 25 mM Hepes buffer (pH 7.6). After the protein concentrations of the cells were determined by the Lowry method, [3H]diprenorphine (2 nM) binding in the presence or absence of 1 lM CTOP (selective antagonist for MOR) was carried out so as to determine the specific binding. Prism 3 program (GraphPad, San Diego, CA) was used to analyze. All binding experiments were performed in triplicate. Co-immunoprecipitation of FLAG-tagged hMOR and Rluc-fusion receptors. Cells in six-well plates were co-transfected with C-terminal

1134

H.S. Choi et al. / Biochemical and Biophysical Research Communications 343 (2006) 1132–1140

FLAG-tagged hMOR and hMOR, or SV1, or SV2 with Rluc-fusion receptors in a ratio of 1:1. Forty-eight hours after transfection, cells were washed with ice-cold phosphate-buffered saline (PBS) and lysed with buffer containing 50 mM Tris–HCl, pH 7.4, 300 mM NaCl, 10% glycerol, 1.5 mM CaCl2, and 1 mM MgCl2 plus protease inhibitors on ice for 30 min. The lysates were centrifuged at 27,000g for 10 min. The supernatants were incubated with anti-FLAG antibodies for 2 h at 4 C. Then, 30 ll of 50% protein G beads was added, and the mixture was incubated for another hour at 4 C. The beads were washed three times with lysis buffer and then suspended in Dulbecco’s PBS and distributed into 96-well microplates for luciferase activity assay using Renilla luciferase assay system (Promega). Rluc without a fused receptor was served as background control.

Confocal microscopy. Cells in 6-well plates were transfected with 1 lg GFP and DsRed2 fused opioid receptors. Twenty-four hours after transfection, cells were trypsinized and plated into glass-bottomed culture dishes (MatTek, Ashland, MA) at a density of 0.5 · 104 cell/dish and cultured for another 24 h. Fixing cells were observed using a laser scanning confocal microscope. Bioluminescence resonance energy transfer assay. HEK-293 cells were co-transfected with vectors expressing the GFP- and Rluc-fusion proteins in a ratio of 4:1. Twenty-four hours after transfection, cells were harvested and washed once with PBS. The cells were then suspended in Dulbecco’s PBS (PBS + 0.1% glucose + 0.01% CaCl2 + 0.01% MgCl2) and distributed into 96-well microplates (white Optiplate; Perkin-Elmer Life and Analytical Sciences) at a density of

Fig. 1. Alternative splice variants of human l-opioid receptor (MOR) gene. (A) The mRNA of hMOR is alternatively spliced. Coding exons are indicated by black boxes. Diagram depicts pre-mRNA and alternatively spliced products. An asterisk (*) indicates the premature stop codon. Arrows with P1 and P2 indicate RT-PCR primers for MOR gene and P4 is RT-PCR primer for the specific SV1 and SV2. Primer P3 is for both MOR and splice variants. Numbers with K (as kilobases) between exons indicate distance kilobases between each exons and/or coding introns. Capital letters A and B indicate the novel coding introns. (B) RT-PCR pattern of human neuroblastoma NMB cells using MOR gene specific primers. Lane 1, molecular markers on a 100-bp ladder; lane 2, hMOR, SV1, and SV2. (C) RT-PCR pattern of human neuroblastoma NMB cells using splice variant gene specific primers. Lane 1, molecular markers on a 100-bp ladder, lane 2, SV1, and SV2. Total RNA was reverse-transcribed and PCR was performed using MOR gene specific primers or splice variant specific primers as indicated. The products of RT-PCR were separated by to 2% agarose gel. (D) Schematic of partial hMOR-1 gene structure and MOR specific primers (P1 and P2) and splice variant specific primers (P3 and P4). RT-PCR products of human brain total RNA using MOR gene specific primers (lane 2) and splice variant specific primers (lane 3). Lane 1, 100-bp DNA ladder (Invitrogen); lane 2, hMOR; lane 3, SV1 and SV2.

H.S. Choi et al. / Biochemical and Biophysical Research Communications 343 (2006) 1132–1140 50,000 cells per well. DeepBlue C was added at a final concentration of 5 lM, and the readings at 410 and 515 nm were measured simultaneously using Fusion system (Perkin-Elmer Life and Analytical Sciences). The bioluminescence resonance energy transfer (BRET) signal was determined by the ratio of the light emitted by the receptor-GFP (515 nm) over the light emitted by the receptor-Rluc (410 nm). The background signal was determined by transfection of receptor-Rluc or GFP construct lacking the receptor sequence and was subtracted from total BRET signal.

1135

Results Identification of two new splice variants of the MOR-1 gene and their structure To identify splice variants from the human MOR gene, we performed RT-PCR using MOR exon 1 and

Fig. 2. Schematic of splice variant (SV1) and splice variant (SV2) gene structure and in vitro translation of SV1 and SV2 genes. (A) The nucleotide sequence of human splice variant (SV1) and its predicted amino acid sequence. (B) The nucleotide sequence of human splice variant (SV2) and its predicted amino acid sequence. The exon sequences are shown in capital letters and intron sequences in lowercase letters. An asterisk (*) indicates the premature stop codon. (C) In vitro translated hMOR-1, SV1, and SV2 proteins were radiolabelled with [35S]methionine using the Promega’s in vitro translation kit. These proteins were separated on a 14% SDS–PAGE gel and the gel was dried and visualized by autoradiography. Lanes 1 and 3 are in vitro translated hMOR-1 protein. Lanes 2 (SV1) and 4 (SV2) are in vitro translated products.

1136

H.S. Choi et al. / Biochemical and Biophysical Research Communications 343 (2006) 1132–1140

exon 2 specific primers from total RNA of human neuroblastoma NMB cells. In addition to the expected fragment (340 bp) corresponding to the wild-type MOR1 gene, we observed two new bands, 515 and 562 bp (Fig. 1B). We designated these as splice variant 1 (SV1) and splice variant 2 (SV2). In order to isolate SV1 and SV2 only, we used a splice specific primer (P4 in Fig. 1A), antisense primer specific to the retained intron 1 sequence and we isolated SV1 and SV2 transcripts in NMB cells (Fig. 1C). Sequencing of the splice variants revealed new 173 bp (SV1) and 220 bp (SV2) insertions between exon 1 and exon 2 (Figs. 2A and B). The new intronic coding sequences for SV1 and SV2 were located between exon 1 and exon 2 (Fig. 1A). Sequence analysis of SV1 and SV2 indicated that they encoded a premature stop codon (TGA) and typical splicing junctions (GT/AG). The transcripts of SV1 were more abundant than transcripts of SV2, which suggest the major splice variant of MOR in NMB cells might be SV1. Expression of human MOR, SV1, and SV2 in human brain To address whether the two SV1 and SV2 transcripts are present in human brain, RT-PCRs were performed using total RNA from human brain (purchased from Ambion). When using primers P1 and P2 specific to exon 1 and exon 2, only MOR-1 was detected effectively (Fig. 1D). However, by using a specific primer set (primers P3 and P4) to the retained intron 1 sequence, both SV1 and SV2 transcripts were detected in the brain with the similar quantity. Therefore, both SV1 and SV2 transcripts were also expressed in human brain. In NMB cells, SV1 was more expressed than SV2 (Fig. 1C), suggesting that these human neuroblastoma cells (NMB) may have a different splicing system for MOR gene compared to normal human brain. In vitro translations of human SV1 and SV2 Translation of SV1 is predicted to generate total 128 amino acids including extra 32 amino acids from the new coding intron B (Fig. 2A). SV2 is predicted to generate total 101 amino acids including extra 5 amino acids from the new coding introns A and B (Fig. 2B). To determine whether the two SV1 and SV2 transcripts could be translated into proteins, we carried out in vitro translation of SV1 and SV2. In vitro translated products

confirmed the correct molecular weight as predicted from their sequences. The molecular weights of the SV1 and SV2 proteins were approximately 14 and 11 kDa, respectively (Fig. 2C). Heterodimerization of hMOR and splice variants (SV1 and SV2) To determine whether splice variants (SV1 and SV2) dimerize with hMOR, we performed the co-immunoprecipitation and BRET assay. Hetero-oligomerization of DOR-KOR and MOR-DOR had been reported by using co-immunoprecipitation of epitope tag, whereas the interaction between MOR-KOR was uncertain [30]. Jordan and Devi were unable to show interaction between MOR-KOR by co-immunoprecipitation of different epitope-tagged MOR and KOR, but Wang and Sadee [36] showed interaction between MOR and KOR using co-immunoprecipitation of Rluc fusion receptors. So to determine whether hMOR and splice variants (SV1 and SV2) dimerize, we performed the co-immunoprecipitation of C-terminal FLAG-tagged hMOR and Rluc fusion receptors, followed by Rluc activity. Rluc without a fused receptor served as background control. Shown in Figs. 3A and B, Rluc-fused hMOR, SV1, and SV2 co-immunoprecipitate with C-terminal FLAG-tagged hMOR in each combination, whereas no Rluc activity was recovered with the Rluc construct alone. The result indicated that the interactions between hMOR and splice variants (SV1 and SV2) did not differ from other receptor interactions, supporting direct physical interaction between hMOR and splice variants. To further confirm hMOR-SV1 (and SV2) heterodimer formation, we performed BRET assay (Fig. 3C), which is based on energy transfer from a bioluminescent donor to a fluorescent protein [36]. In this study, we used a modified form of the Rluc coelenterazine substrate, called DeepBlue C, which emits the blue light between 390 and 400 nm upon activation of Renilla luciferase. Stimulated at this wavelength, the GFP variant, GFP2 emits green light at 505–508 nm, when in close molecular proximity to the Rluc. The BRET signal was measured as ratio of green light emitted by GFP2 over blue light emitted by Rluc. To measure hMOR-SV1 (SV2) heterodimerization by BRET, hMOR with GFP2 and SV1 (SV2) with Rluc protein were fused at the C-terminal. Cells expressing hMOR-SV1 and hMOR-SV2 had a 4-fold higher BRET signal than cells expression hMOR-GFP and Rluc

c Fig. 3. Co-immunoprecipitation of FLAG-tagged hMOR with Rluc-SV1 (and SV2) and heterodimerization of human opioid receptor and splice variants using BRET and co-localization of human opioid receptor with splice variants at cytoplasmic membrane. (A) Schematic diagram representing the assay system of homo/heterodimerization of opioid receptor protein. (B) HEK 293 cells co-expressing C-terminal FLAG-tagged human opioid receptor and Rluc-fusion splice variants (SV1 and SV2) were lysed and immunoprecipitated with anti-FLAG antibody. Renilla luciferase activity was measured in precipitates. Values are presented as percentage of total Renilla luciferase activity in the cell lysates. All values correspond to means ± SEM calculated from at least three independent experiments. (C) HEK-293 cells were transiently co-transfected with GFP-hMOR and Rluc-fusion splice variants (SV1 and SV2). A transfection with MOR-Rluc fusion protein was used as positive control. Twenty-four hours after transfection, BRET (luminescence/ fluorescence) ratios were measured. All values correspond to means ± SEM calculated from at least three independent experiments. (D) HEK-293 cells were transiently co-transfected with MOR-GFP and SV1-DsRed2 or HEK-293 cells co-transfected with MOR-GFP and SV2-DsRed2. Forty-eight hours after transfection, images were taken by confocal laser microscopy.

H.S. Choi et al. / Biochemical and Biophysical Research Communications 343 (2006) 1132–1140

A

B

MOR-Rluc

MOR-FLAG

Luciferase assay

SV-Rluc

5

0

0.16

BRET Ratio

0.14 0.12 0.10 0.08 0.06 0.04

D

O

M

M

O RG

FP

/M

0.00

O RRl Ruc G FP /S V1 -R M lu O Rc G FP /S V2 -R lu c M O RG FP /R lu c

0.02

hM

FhM

O

0.18

F-

R

0.20

O

/H M

C

R /S

O

R -R

V1 -R

lu c

Luciferase assay

10

R /R lu c

MOR-FLAG

IP: anti-FLAG

15

M O

IP: anti-FLAG

SV-Rluc

Fh

MOR-FLAG

2R lu c

MOR-Rluc

20

/S V

MOR-FLAG

SV-Rluc gene co-transfection

lu c

co-transfection

FhM O R

MOR -FLAG gene MOR-Rluc gene

25

Luciferase activity (% total)

MOR-FLAG gene

1137

1138

H.S. Choi et al. / Biochemical and Biophysical Research Communications 343 (2006) 1132–1140

antagonist CTOP for MOR. When the splice variants (SV1 and SV2) were transfected into HEK-293 cells, they had no binding activity as compared to wild-type hMOR transfected cells, indicating that these two variants have no opioid binding ability by themselves (Fig. 4A). Interestingly, in NMB cells expressing endogeneous hMOR protein, specific surface binding of hMOR was reduced by the over-expression of splice variants (SV1 and SV2) (Fig. 4B). These data indicate that the splice variant and hMOR protein interact to each other and then co-localized at the cell membrane. Dimerized splice variant and hMOR inhibit ligand binding of human opioid receptor on the surface. Discussion

Fig. 4. The splice variants do not display high binding to [3H]diprenorphine and reduced the binding activity of human opioid receptor. (A) Binding of [3H]diprenorphine from HEK-293 cells expressing hMOR (2 lg) and splice variants (2 lg each, SV1 and SV2) separately. All values of binding assay correspond to means ± SEM calculated from at least three independent experiments. (B) NMB cell, opioid receptor expression cells were transfected with 2 lg of splice variant (SV1 and SV2) cDNA. Binding activity of [3H]diprenorphine to NMB cells was carried out as described under Materials and methods. All values correspond to means ± SEM calculated from at least three independent experiments.

(Fig. 3C). Taken together, these results demonstrate that formation of hMOR and splice variant (SV1 and SV2) heterodimers occurs at HEK-293 cells. Membrane co-localization of human MOR and splice variants (SV1 and SV2) Since the splice variants (SV1 and SV2) form heterodimers with hMOR inside the cell (Figs. 3B and C), we further investigated where the location of these hMOR-SV1, hMOR-SV2 heterodimers was localized. HEK-293 cells were co-transfected with GFP-MOR and DsRed2-SV1 (and SV2), the cellular distribution of hMOR and splice variants (SV1 and SV2) was examined by confocal laser microscopy. Both splice variants and opioid receptors were observed on the cell surface (Fig. 3D). So they are co-localized at cell membrane. Splice variants modulate the binding activity of human opioid receptor The presence of splice variants in vitro and in vivo intrigued us to investigate their biological role to wild-type MOR function. After various different approaches, we finalized to have a meaningful result with the opioid binding assays as follows (Fig. 4). The following binding activity of opioid receptor was monitored by measuring the [3H]diprenorphine specific binding with or without selective

The mouse MOR gene has been reported to undergo extensive splicing at both its 5 0 - and 3 0 -ends. Now several human MOR splice variants [hMOR-1A, hMOR-1B (1– 5), hMOR-1O, hMOR-1X, and hMOR-1Y] have been reported [14,23]. All these human MOR variants share exons 1, 2, and 3 of the original hMOR-1. As the process of splicing is not nearly as precise as one might imagine, splicing small exons in a sea of large introns is so difficult that the splicing machinery is error prone [37]. We have identified two novel splice variants of the human MOR gene, each retaining different portions of intron 1 and a premature stop codon (Figs. 1 and 2). These variants, as well as wild-type, can be translated into proteins of the predicted molecular weights and also expressed in human neuronal cell and brain. Alternative splicing has been shown to affect more than one-third of human genes. In eukaryotes, elaborate sets of mechanisms have evolved to ensure that the multistep process of gene expression is accurately executed. One of these elaborate mechanisms is mRNA surveillance for mRNA quality control [38]. The mRNA surveillance facilitates the detection and destruction of mRNA that contains premature stop codons by a process called nonsense-mediated decay (NMD), presumably to prevent the production of potentially deleterious proteins [39]. SV1 and SV2, which contain premature stop codons, could be candidates for NMD. If this were the case, transcripts of SV1 and SV2 would be rapidly degraded without being accumulated. However, we have confirmed that the two variants are present not only in NMB cells but also in human brain (Fig. 1). Each has the same nucleotide sequence in NMB cells and in human brain. The mRNAs of SV1 and SV2 that are not degraded by nonsense-mediated decay may produce abnormal MOR proteins and may interfere with MOR signaling. Therefore, they could have pharmacological meaning. There has been some controversy regarding the biological function of MOR splice variants. One hypothesis is that changes in protein conformation rather than alternative splicing cause the functional diversity of MORs [40]. However, experimental data are controversial due in part to their very low abundance and a general lack of validation from independent laboratories [20]. The 3 0 terminal splicing of

H.S. Choi et al. / Biochemical and Biophysical Research Communications 343 (2006) 1132–1140

rat and mouse MORs influences agonist-mediated internalization and receptor resensitization [41,42]. l-Opioid ligands show significant binding differences to different splice variants, and their regional expression patterns vary considerably [16,43–45]. The different splice variants could be responsible for morphine and morphine-6-glucuronide-mediated analgesia [21,22]. The MOR agonist b-endorphin induces G-protein activation in MOR (exon 2 and 3)knockout mice, but DAMGO, endomorphine-1, and endomorphine-2 do not [24]. Since the gene structures of SV1 and SV2 are very similar to that of MOR (exon 2 and 3)-knockout mice, proteins encoded by SV1 and SV2 might play a role in G-protein activation through b-endorphin. We found that splice variant-hMOR heterodimer complex was demonstrated by BRET assay and co-immunoprecipitation experiment. The exact nature of the opioid receptor homo- and heterodimerization has not been defined [46]. Two general schemes for the dimerization of GPCRs have been proposed by Reynolds and co-workers [47]; (1) the 1:1 stoichiometric molecular complexes of the receptors or contact dimers and (2) the swapping of the transmembrane domains of the GPCRs. Using computational three-dimensional models of the opioid receptors, Filizola and Weinstein [48] suggested the most likely interfaces between the opioid receptor homodimers are TM4TM4, TM4-TM5, and TM5-TM5 for DOR, TM1-TM1 for MOR, and TM5-TM5 for j-opioid receptor. We assumed that hMOR and splice variants (SV1 and SV2) might dimerize through TM1 of hMOR and TM1 of splice variant and then co-localized at the cell membrane. The heteromeric receptor interaction likely occurs during or soon following translation and then transported to the cell surface [49]. Dimerized splice variants inhibit ligand binding of human MOR (Fig. 4B). These data support a potential role of splice variant protein as possible biological modulator of hMOR ligands. Acknowledgments This work was supported by National Institutes of Health research Grants DA00564, DA01583, DA11806, DA11190, K05-DA00153, and K02-DA13926 and by the F&A Stark Fund of the Minnesota Medical Foundation. We also thank Vida G for helpful manuscript review. References [1] B.H. Min, L.B. Augustin, R.F. Felsheim, J.A. Fuchs, H.H. Loh, Genomic structure analysis of promoter sequence of a mouse mu opioid receptor gene, Proc. Natl. Acad. Sci. USA 91 (1994) 9081– 9085. [2] B.L. Kieffer, Recent advances in molecular recognition and signal transduction of active peptides: receptors for opioid peptides, Cell Mol. Neurobiol. 15 (1995) 615–635. [3] B.L. Kieffer, Opioids: first lessons from knockout mice, Trends Pharmacol. Sci. 20 (1999) 19–26. [4] P.Y. Law, Y.H. Wong, H.H. Loh, Molecular mechanisms and regulation of opioid receptor signaling, Annu. Rev. Pharmacol. Toxicol. 40 (2000) 389–430.

1139

[5] A. Mansour, C.A. Fox, H. Akil, S.J. Watson, Opioid-receptor mRNA expression in the rat CNS: anatomical and functional implications, Trends Neurosci. 18 (1995) 22–29. [6] M. Sedqi, S. Roy, S. Ramakrishnan, R. Elde, H.H. Loh, Complementary DNA cloning of a mu-opioid receptor from rat peritoneal macrophages, Biochem. Biophys. Res. Commun. 209 (1995) 563–574. [7] R.M. Donahoe, D. Vlahov, Opiates as potential cofactors in progression of HIV-1 infections to AIDS, J. Neuroimmunol. 83 (1998) 77–87. [8] S.B. Nyland, S. Specter, J. Im-Sin, K.E. Ugen, Opiate effects on in vitro human retroviral infection, Adv. Exp. Med. Biol. 437 (1998) 91–100. [9] W.S. Sheng, S. Hu, G. Gekker, S. Zhu, P.K. Peterson, C.C. Chao, Immunomodulatory role of opioids in the central nervous system, Arch. Immunol. Ther. Exp. (Warsz) 45 (1997) 359–366. [10] J.B. Wang, Y. Imai, C.M. Eppler, P. Gregor, C.E. Spivak, G.R. Uhl, mu opiate receptor: cDNA cloning and expression, Proc. Natl. Acad. Sci. USA 90 (1993) 10230–10234. [11] Y. Chen, A. Mestek, J. Liu, J.A. Hurley, L. Yu, Molecular cloning and functional expression of a mu-opioid receptor from rat brain, Mol. Pharmacol. 44 (1993) 8–12. [12] K. Fukuda, S. Kato, K. Mori, M. Nishi, H. Takeshima, Primary structures and expression from cDNAs of rat opioid receptor deltaand mu-subtypes, FEBS Lett. 327 (1993) 311–314. [13] L.A. Bare, E. Mansson, D. Yang, Expression of two variants of the human mu opioid receptor mRNA in SK-N-SH cells and human brain, FEBS Lett. 354 (1994) 213–216. [14] Y.X. Pan, J. Xu, L. Mahurter, M. Xu, A.K. Gilbert, G.W. Pasternak, Identification and characterization of two new human mu opioid receptor splice variants, hMOR-1O and hMOR-1X, Biochem. Biophys. Res. Commun. 301 (2003) 1057–1061. [15] A. Zimprich, T. Simon, V. Hollt, Cloning and expression of an isoform of the rat mu opioid receptor (rMOR1B) which differs in agonist induced desensitization from rMOR1, FEBS Lett. 359 (1995) 142–146. [16] Y.X. Pan, J. Xu, E. Bolan, C. Abbadie, A. Chang, A. Zuckerman, G. Rossi, G.W. Pasternak, Identification and characterization of three new alternatively spliced mu-opioid receptor isoforms, Mol. Pharmacol. 56 (1999) 396–403. [17] Y.X. Pan, J. Xu, E. Bolan, A. Chang, L. Mahurter, G. Rossi, G.W. Pasternak, Isolation and expression of a novel alternatively spliced mu opioid receptor isoform, MOR-1F, FEBS Lett. 466 (2000) 337–340. [18] Y.X. Pan, J. Xu, L. Mahurter, E. Bolan, M. Xu, G.W. Pasternak, Generation of the mu opioid receptor (MOR-1) protein by three new splice variants of the Oprm gene, Proc. Natl. Acad. Sci. USA 98 (2001) 14084–14089. [19] J.H. Kwon, S. Keates, S. Simeonidis, F. Grall, T.A. Libermann, A.C. Keates, ESE-1, an enterocyte-specific Ets transcription factor, regulates MIP-3alpha gene expression in Caco-2 human colonic epithelial cells, J. Biol. Chem. 278 (2003) 875–884. [20] T.M. Kvam, C. Baar, T.T. Rakvag, S. Kaasa, H.E. Krokan, F. Skorpen, Genetic analysis of the murine micro opioid receptor: increased complexity of Oprm gene splicing, J. Mol. Med. 82 (2004) 250–255. [21] G.C. Rossi, Y.X. Pan, G.P. Brown, G.W. Pasternak, Antisense mapping the MOR-1 opioid receptor: evidence for alternative splicing and a novel morphine-6 beta-glucuronide receptor, FEBS Lett. 369 (1995) 192–196. [22] G.C. Rossi, L. Leventhal, Y.X. Pan, J. Cole, W. Su, R.J. Bodnar, G.W. Pasternak, Antisense mapping of MOR-1 in rats: distinguishing between morphine and morphine-6beta-glucuronide antinociception, J. Pharmacol. Exp. Ther. 281 (1997) 109–114. [23] L. Pan, J. Xu, R. Yu, M.M. Xu, Y.X. Pan, G.W. Pasternak, Identification and characterization of six new alternatively spliced variants of the human mu opioid receptor gene, Oprm, Neuroscience 133 (2005) 209–220.

1140

H.S. Choi et al. / Biochemical and Biophysical Research Communications 343 (2006) 1132–1140

[24] H. Mizoguchi, H.E. Wu, M. Narita, H.H. Loh, H. Nagase, L.F. Tseng, Loss of mu-opioid receptor-mediated G-protein activation in the pons/medulla of mice lacking the exons 2 and 3 of mu-opioid receptor gene, Neurosci. Lett. 335 (2002) 91–94. [25] J.P. Pin, J. Kniazeff, V. Binet, J. Liu, D. Maurel, T. Galvez, B. Duthey, M. Havlickova, J. Blahos, L. Prezeau, P. Rondard, Activation mechanism of the heterodimeric GABA(B) receptor, Biochem. Pharmacol. 68 (2004) 1565–1572. [26] J. Kniazeff, A.S. Bessis, D. Maurel, H. Ansanay, L. Prezeau, J.P. Pin, Closed state of both binding domains of homodimeric mGlu receptors is required for full activity, Nat. Struct. Mol. Biol. 11 (2004) 706–713. [27] G. Nelson, J. Chandrashekar, M.A. Hoon, L. Feng, G. Zhao, N.J. Ryba, C.S. Zuker, An amino-acid taste receptor, Nature 416 (2002) 199–202. [28] S. Filipek, K.A. Krzysko, D. Fotiadis, Y. Liang, D.A. Saperstein, A. Engel, K. Palczewski, A concept for G protein activation by G protein-coupled receptor dimers: the transducin/rhodopsin interface, Photochem. Photobiol. Sci. 3 (2004) 628–638. [29] D. Fotiadis, Y. Liang, S. Filipek, D.A. Saperstein, A. Engel, K. Palczewski, The G protein-coupled receptor rhodopsin in the native membrane, FEBS Lett. 564 (2004) 281–288. [30] B.A. Jordan, L.A. Devi, G-protein-coupled receptor heterodimerization modulates receptor function, Nature 399 (1999) 697–700. [31] S.R. George, T. Fan, Z. Xie, R. Tse, V. Tam, G. Varghese, B.F. O’Dowd, Oligomerization of mu- and delta-opioid receptors. Generation of novel functional properties, J. Biol. Chem. 275 (2000) 26128– 26135. [32] I. Gomes, A. Gupta, J. Filipovska, H.H. Szeto, J.E. Pintar, L.A. Devi, A role for heterodimerization of mu and delta opiate receptors in enhancing morphine analgesia, Proc. Natl. Acad. Sci. USA 101 (2004) 5135–5139. [33] L. He, J. Fong, M. von Zastrow, J.L. Whistler, Regulation of opioid receptor trafficking and morphine tolerance by receptor oligomerization, Cell 108 (2002) 271–282. [34] C.K. Hwang, C.S. Kim, H.S. Choi, S.R. McKercher, H.H. Loh, Transcriptional regulation of mouse mu opioid receptor gene by PU.1, J. Biol. Chem. 279 (2004) 19764–19774. [35] H.S. Choi, C.K. Hwang, C.S. Kim, K.Y. Song, P.Y. Law, L.N. Wei, H.H. Loh, Transcriptional regulation of mouse mu opioid receptor gene: Sp3 isoforms (M1, M2) function as repressors in neuronal cells to regulate the mu opioid receptor gene, Mol. Pharmacol. 67 (2005) 1674–1683. [36] D. Wang, X. Sun, L.M. Bohn, W. Sadee, Opioid receptor homo- and heterodimerization in living cells by quantitative bioluminescence resonance energy transfer, Mol. Pharmacol. 67 (2005) 2173–2184.

[37] B.P. Lewis, R.E. Green, S.E. Brenner, Evidence for the widespread coupling of alternative splicing and nonsense-mediated mRNA decay in humans, Proc. Natl. Acad. Sci. USA 100 (2003) 189–192. [38] E. Wagner, J. Lykke-Andersen, mRNA surveillance: the perfect persist, J. Cell Sci. 115 (2002) 3033–3038. [39] L.E. Maquat, Nonsense-mediated mRNA decay, Curr. Biol. 12 (2002) R196–R197. [40] K.B. Gaveriaux-Ruff C, Opioid receptor: gene structure and function, in: C. Stein (Ed.), Opioids in Pain Control, Cambridge University Press, Cambridge, 1999. [41] T. Koch, S. Schulz, H. Schroder, R. Wolf, E. Raulf, V. Hollt, Carboxyl-terminal splicing of the rat mu opioid receptor modulates agonist-mediated internalization and receptor resensitization, J. Biol. Chem. 273 (1998) 13652–13657. [42] T. Koch, S. Schulz, M. Pfeiffer, M. Klutzny, H. Schroder, E. Kahl, V. Hollt, C-terminal splice variants of the mouse mu-opioid receptor differ in morphine-induced internalization and receptor resensitization, J. Biol. Chem. 276 (2001) 31408–31414. [43] C. Abbadie, Y.X. Pan, G.W. Pasternak, Differential distribution in rat brain of mu opioid receptor carboxy terminal splice variants MOR-1C-like and MOR-1-like immunoreactivity: evidence for region-specific processing, J. Comp. Neurol. 419 (2000) 244–256. [44] C. Abbadie, Y. Pan, C.T. Drake, G.W. Pasternak, Comparative immunohistochemical distributions of carboxy terminus epitopes from the mu-opioid receptor splice variants MOR-1D, MOR-1 and MOR-1C in the mouse and rat CNS, Neuroscience 100 (2000) 141– 153. [45] C. Abbadie, S.H. Gultekin, G.W. Pasternak, Immunohistochemical localization of the carboxy terminus of the novel mu opioid receptor splice variant MOR-1C within the human spinal cord, Neuroreport 11 (2000) 1953–1957. [46] L. Xin, Z.J. Wang, Bioinformatic analysis of the human mu opioid receptor (OPRM1) splice and polymorphic variants, AAPS PharmSci. 4 (2002) E23. [47] M.K. Dean, C. Higgs, R.E. Smith, R.P. Bywater, C.R. Snell, P.D. Scott, G.J. Upton, T.J. Howe, C.A. Reynolds, Dimerization of G-protein-coupled receptors, J. Med. Chem. 44 (2001) 4595– 4614. [48] M. Filizola, H. Weinstein, Structural models for dimerization of Gprotein coupled receptors: the opioid receptor homodimers, Biopolymers 66 (2002) 317–325. [49] T. Fan, G. Varghese, T. Nguyen, R. Tse, B.F. O’Dowd, S.R. George, A role for the distal carboxyl tails in generating the novel pharmacology and G protein activation profile of mu and delta opioid receptor hetero-oligomers, J. Biol. Chem. 280 (2005) 38478– 38488.