Characterization of Peripheral Human Cannabinoid Receptor (hCB2) Expression and Pharmacology Using a Novel Radioligand, [35S]Sch225336*

Studies to characterize the endogenous expression and pharmacology of peripheral human cannabinoid receptor (hCB2) have been hampered by the dearth of authentic anti-hCB2 antibodies and the lack of radioligands with CB2 selectivity. We recently described a novel CB2 inverse agonist, N-[1(S)-[4-[[4-methoxy-2-[(4methoxyphenyl)sulfonyl] phenyl]sulfonyl] phenyl]ethyl]methane-sulfonamide (Sch225336), that binds hCB2 with high affinity and excellent selectivity versus hCB1. The precursor primary amine of Sch225336 was prepared and reacted directly with [35S]mesyl chloride (synthesized from commercially obtained [35S]methane sulfonic acid) to generate [35S]Sch225336. [35S]Sch225336 has high specific activity (>1400 Ci/mmol) and affinity for hCB2 (65 pm). Using [35S]Sch225336, we assayed hemopoietic cells and cell lines to quantitate the expression and pharmacology of hCB2. Lastly, we used [35S]Sch225336 for detailed autoradiographic analysis of CB2 in lymphoid tissues. Based on these data, we conclude that [35S]Sch225336 represents a unique radioligand for the study of CB2 endogenously expressed in blood cells and tissues.

The endocannabinoids, anandamide and 2-arachidonyl glycerol, are ligands for two G protein-coupled cannabinoid receptors, CB1 2 and CB2. It has been hypothesized that human CB2 (hCB2) and its ligands are immune regulators due to the defined expression of hCB2 mRNA in peripheral immune cells and tissues. This hypothesis was bolstered by studies showing that 2-arachidonyl glycerol, a full agonist at hCB2 (1), stimulated chemotaxis of various hemopoietic cells, including differentiated monocytic and eosinophilic cells such as HL-60, THP-1, U937, and leukemia EoL-1 cells as well as human monocytes, eosinophils, and dendritic cells (2)(3)(4). CB2 was implicated in this response due to its sensitivity to blockade by the CB2 inverse agonist, SR144528. However, direct measurement of hCB2 protein expression in immune cells has been hampered by the lack of radioligands with CB2 selectivity and high specific activity. Binding analyses of CB2 have used tritiated agonists such as [ 3 H]CP55,940 or [ 3 H]WIN55,212 that bind with low nanomolar affinity but have low specific activities (specific activity, 20 -180 Ci/mmol) and little or no selectivity versus CB1. As a result, CB2 pharmacology has been almost completely restricted to studies with recombinantly expressed receptor (for review, see Ref. 5). Recently, we described a novel CB2 inverse agonist, N-[1(S)- [4-[[4-methoxy-2-[(4methoxyphenyl)sulfonyl]phenyl]sulfonyl]phenyl]ethyl] methanesulfonamide (compound 4j; Sch225336), that binds hCB2 with picomolar affinity and has excellent selectivity versus hCB1 (6). Using commercially obtained [ 35 S]methane sulfonic acid, [ 35 S]mesyl chloride was prepared and reacted directly with the precursor primary amine of Sch225336 to generate [ 35 S]Sch225336, a highly potent and CB2-selective radioligand with high specific activity (Ͼ1400 Ci/mmol). The studies presented herein describe the use of this unique radioligand to extensively characterize CB2 expressed endogenously in blood cells, cell lines, and tissue sections.

EXPERIMENTAL PROCEDURES
Cells and Cell Culture-The clonal CHO-hCB2 cell line was generated and cultured as previously described (1). Ba/F-hCB1 and -hCB2 cells were cultured as previously described (7). MonoMac6, Jurkat, U937, Hut78, and HL-60 cells were cultured in the presence or absence of the indicated differentiation factors as previously described (8 -11). Activated human peripheral blood lymphocytes were generated as previously described (12). Experimental cultures were used 1-5 days after seeding. Cell culture medium was purchased from Invitrogen.
Cell Membrane Preparation-Sf9 membranes exogenously expressing hCB2 ϩ ␤ 1 ␥ 2 G␣ i1 or ␤ 1 ␥ 2 G␣ i1 alone and RB-hCB2 membranes were purchased from PerkinElmer Life Sciences. Other cell membrane preparations were generated as previously described (1). Briefly, adherent cells were harvested using cell dissociation buffer according to the manufacturer's instructions (Invitrogen). Dissociated cells and cells grown in * 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 suspension were collected by centrifugation and used immediately or stored at Ϫ80°C. Cell pellets were resuspended and incubated on ice for 30 min in homogenization buffer (10 mM Tris-HCl, 5 mM EDTA, 3 mM EGTA, pH 7.6) containing 1 mM phenylmethylsulfonyl fluoride as a protease and amidase inhibitor (13,14). Cells were then homogenized with 15-20 strokes at 900 rpm with a Dounce homogenizer using stirrer type RZR1 polytron homogenizer (Caframo, Wiarton, Ontario). Intact cells and nuclei were removed by low speed centrifugation (500 ϫ g for 5 min at 4°C). Membranes in the supernatant were pelleted by centrifugation at 100,000 ϫ g for 30 min at 4°C and then resuspended in gly-gly buffer (20 mM glycylglycine, 1 mM MgCl 2 , 250 mM sucrose, pH 7.2) and stored at Ϫ80°C. Protein determinations were performed using the Bradford method (15). Western Blot Analyses-Western blots were performed as previously described (16). CHO-hCB2, MonoMac6, Sf9-hCB2, Ba/F-hCB1, and Ba/F-hCB2 cell membranes were solubilized in sample buffer (62.5 mM Tris-HCl, 2% sodium dodecyl sulfate, 10% 2-mercaptoethanol (v/v), 6 M urea, 20% glycerol, pH 6.8) by incubation at 60°C for 15 min. Where indicated, membrane proteins were prepurified by wheat germ agglutinin-agarose chromatography prior to solubilization in sample buffer (16). Proteins were resolved on a 7.5% sodium dodecyl sulfate-polyacrylamide gel and transferred to polyvinylidene difluoride membrane. The membrane was then blocked for 2 h with "Blotto" (10 mM NaH 2 PO 4 , 10% nonfat dry milk, 10% glycerol, 0.2% Tween 20) and incubated overnight at 4°C with the indicated antibody dilutions. Following incubation with appropriate conjugated secondary antibodies, immunoreactive proteins were detected with the ECL chemiluminescent antibody detection system (Amersham Biosciences). 35  S]methane sulfonic acid (2.1) (5 mCi, 40 l, ϳ1400 Ci/mmol) was evaporated to dryness in a 10-ml V-Vial (Wheaton Scientific) and dissolved in anhydrous methylene chloride (1.5 ml). 150 l of oxalyl chloride was added, and the reaction was stirred for 30 min under Argon prior to the addition of 20 l of 10% dimethyl formamide. The reaction was stirred overnight at room temperature, diluted with methylene chloride (1.5 ml), cooled in an ice bath, and washed successively with ice-cooled 1% aqueous sodium bicarbonate (2 ϫ 2 ml), room temperature 1% aqueous sodium bicarbonate (2 ml), and 2% aqueous sodium bisulfate (1 ml) solutions. The methylene chloride solution was dried for 1 h over anhydrous sodium sulfate and concentrated to a volume of ϳ50 l via distillation for immediate use in the next step. In a 300-l V-Vial (Wheaton Scientific), 1-{4-[methoxy-2-(4-methoxy-benzenesulfonyl)benzenesulfonyl]-phenyl}-ethylamine (2.3) (10 mg) was dissolved in anhydrous methylene chloride (20 l) and triethylamine (10 l). The methylene chloride solution of [ 35 S]mesyl chloride (2.2) (ϳ50 l) was added and the reaction stirred vigorously for 1 h. The reaction was concentrated to dryness and the product purified by high pressure liquid chromatography on a 9.4 ϫ 250 mm Zorbax Extend C18 column with a mobile phase of 0.05 M pH 9 triethylammonium acetate (40:60) acetonitrile at a flow rate of 5 ml/min. Detection was at 254 nm. A total batch of 1.3 mCi of (2.4) at a specific activity of Ͼ1400 Ci/mmol at a radiochemical purity of 99.5% was isolated.
[ 35 S]GTP␥S exchange was measured using a scintillation proximity assay as previously described (17). For each assay point, membranes (2 g/well in triplicate) were preincubated for 30 min at room temperature with 150 g of wheat germ agglutinin-coated scintillation proximity assay beads (Amersham Biosciences) in scintillation proximity assay binding buffer (20 mM HEPES, 100 mM NaCl, 5 mM MgCl 2 , and 0.2% (w/v) bovine serum albumin (Factor V, lipid free), pH 7.4) supplemented with 5 M GDP.). The beads and membranes were transferred to a 96-well Isoplate (Wallac, Gaithersburg, MD) and incubated for 60 min at room temperature with the indicated concentrations of cannabinoids. The reactions were then incubated for a further 60 min in the presence of 0.1 nM [ 35 S]GTP␥S (triethyl ammonium salt; specific activity, 1250 Ci/mmol; PerkinElmer Life Sciences). Membrane-bound [ 35 S]GTP␥S was then measured using a Trilux 1450 MicroBeta counter (Wallac, Turku, Finland).
Autoradiography and Cytochemical Analysis of Frozen Sections-Adjacent 20-m tissue sections were cut on a cryostat, mounted onto precleaned microscope glass slides (Colorfrost/Plus; Fisher), and stored at Ϫ80°C until the day of the assay. Before the experiments, slides were brought to a room temperature and then equilibrated in the assay buffer containing 50 mM HEPES, 5 mM MgCl 2 , 125 mM NaCl, 1 mM CaCl 2 , 0.0002% NaN 3 , 1% (w/v) bovine serum albumin (Factor V, lipid free, pH 7.4) for 5 min at 25°C. Slides then were transferred to a fresh assay buffer containing 0.1-0.3 nM [ 35 S]Sch225336 (1400 Ci/mmol) or 3 nM [ 3 H]CP55,940 in the presence or absence of cold cannabinoids and incubated for 2 h at 25°C. Slides were then rinsed three times for 5 min each in cold phosphate-buffered saline supplemented with 0.2% (w/v) bovine serum albumin, rinsed in deionized H 2 O, and gently dried under a stream of cold air. Sections were exposed to a tritium storage phosphor screen (GE Healthcare) for 24 h at 25°C. The phosphor screens were then scanned and digitized using a phosphorimager Storm 860 (GE Healthcare). The sections were exposed to either Kodak BioMax MS film (tritium) or Kodak BioMax MR film ([ 35 S]Sch225336) (Eastman Kodak, Rochester, NY) for 5-72 days at Ϫ80°C. In parallel, adjacent sections were either fixed in 10% phosphate-buffered paraformaldehyde and stained with hematoxylin/eosin or fixed with cold 100% acetone and used for immunohistochemistry with anti-CD79␣ and/or anti-CD3 (DakoCytomation, Carpinteria, CA). For immunohistochemistry, acetone-fixed slides were pretreated with 3% H 2 O 2 and 0.1% azide for 10 min before incubation with primary antibodies for 1 h at room temperature. Following the incubation with biotinylated anti-mouse antibody for 30 min, the binding of the antibodies was visualized with Vectastain Elite ABC kit plus 3,3Ј-diaminobenzidine tetrahydrochloride or Vectastain ABC-AP kit plus Vector Red (Vector, Burlingame, CA). When necessary, the slides were counterstained with hematoxylin.
Data Analysis-Data are reported as mean values Ϯ S.E. of at least three independent experiments, each performed in triplicate. Nonlinear regression analyses of saturation data and of concentration-response data were performed using Prism 2.0b software (GraphPad Software, San Diego, CA) to calculate K d , B max , IC 50 , and EC 50 . IC 50 values were converted to apparent K i values by the method of Cheng and Prusoff (18)

RESULTS
Immunoblot Analysis of hCB2 Protein-Anti-hCB2 antibodies were tested in Western blots for their utility in immunodetection of HA-tagged CB2 using an anti-HA antibody as a positive control. Membrane proteins from untransfected CHO-K1 cells or CHO-hCB2 cells overexpressing HA-tagged CB2 were separated by SDS-PAGE and electrophoretically transferred to polyvinylidene difluoride membrane. The membranes were incubated with an anti-HA monoclonal antibody or polyclonal antibodies raised against peptides representing unique amino acid sequences in hCB2. The anti-HA antibody recognized a diffuse 55-kDa protein and a faint but more defined 45-kDa protein in the CB2-transfected (but not parental) CHO cells (Fig. 1A). The appearance and M r of these proteins is reminiscent of that seen in the Western blots of CHO-hCB2 membranes generated by Carayon et al. (19) using their polyclonal antibody raised against a synthetic peptide derived from the predicted amino acid sequence of the carboxylic tail of hCB2. The diffuse appearance of the 55-kDa receptor protein is consistent with the migration patterns of a glycoprotein (20,21).
The fainter 45-kDa protein may be a receptor precursor similar to that described for other G protein-coupled receptors (20). The classification of hCB2 as a glycoprotein is verified by its successful purification with wheat germ agglutinin chromatography prior to Western blot analysis (Fig. 1B). None of the purported anti-hCB2 antibodies that we tested (Calbiochem 209552 and 209554, ABR PA1-744 and PA1-746, Alexis) successfully identified the 55-kDa (or 45-kDa) recep- tor protein using a panel of recombinant and tumor cell lines. Representative immunoblots are shown in Fig. 1, C-E. None of the purported anti-hCB2 antibodies immunoreacted with pro-teins consistent with the experiments with the HA-tagged receptor, i.e. a diffuse 55-kDa glycoprotein. With all the antibodies tested, immunoreactive bands were quite sharp, inconsistent with hCB2 (Fig. 1, A and B) and G protein-coupled receptors in general (16, 20, 21,). Moreover, the immunoreactive protein(s) were either too small or too large and/or were equally evident in samples from cells that do express hCB2 (Fig. 1, D  and E). With the Calbiochem 209552 antibody (Fig. 1C), a single sharp immunoreactive band (molecular mass ϳ60 kDa) was evident in the CHO-hCB2 sample but not in either the Sf9-hCB2 or the Monomac6 samples. The 746 antibody did not immunoreact with any specific protein. Taken together, these data indicate that these antibodies are less than optimal for the immunodetection of authentic hCB2. Therefore, we endeavored to develop a CB2-selective radioligand.  6, and data not shown). Scatchard analysis of the saturation binding data (Fig. 3A, inset) revealed a single binding site with a Hill ϳ1. Saturation analyses with Sf9 membranes exogenously expressing hCB2 ϩ ␤ 1 ␥ 2 G␣ i1 (Fig. 3B)    The utility of [ 35 S]Sch225336 relative to [ 3 H]CP55,940 in competition binding analyses was assessed in the CHO-hCB2 membranes. Radioligand binding was displaced in a concentration-dependent manner by unlabeled compound (Hill ϳ1) (Fig.  4A), though competition bindings with [ 35 S]Sch225336 conferred a much superior signal:noise (B/B 0 ) relative to the tritiated agonist (Fig. 4B). [ 35 S]Sch225336 binding was inhibited by structurally unrelated cannabinoid agonists, including CP55,940 and unlabeled Sch225336 (Fig. 4C), WIN55,212-2, HU210, anandamide, and 2-arachidonyl glycerol (data not shown) and the inverse agonist SR144528 (Fig. 4C). There was no measurable [ 35 S]Sch225336 binding in membranes expressing hCB1 (data not shown), consistent with the low affinity of Sch225336 for CB1 measured in competition binding analyses with [ 3 H]CP55,940 (6). Taken together, these data show that [ 35 S]Sch225336 represents a novel, high affinity CB2-specific radioligand with high specific activity.
Analyses of hCB2 Expression in Peripheral Blood Lymphocytes and Hematopoietic Cell Lines-CB2 expression in hematopoietic cells has been reported, largely based on the detection of its mRNA (23)(24)(25)(26) and/or by functional responses to cannabinoids that were attenuated with CB2-selective antagonists (2,23,26,27). As our efforts in CB2 immunodetection were unsuccessful (above), we assessed the utility of [ 35 S]Sch225336 to define and quantitate hCB2 expression in hemopoietic cells and cell lines. Saturation binding analyses were performed on mem-branes from human Jurkat T lymphoma cells (clone E6 -1), human T lymphoma HUT78 cells, human monocytic U937 and MonoMac6 cells, and activated human peripheral lymphocytes (PBL). All of these cells express measurable hCB2 with the highest expression in Mono-Mac6 cell membranes, whereas the PBL and HUT78 had the lowest expression (Fig. 5A). The effect of in vitro activation conditions on PBL expression of hCB2 is shown in Fig.  5B. Membranes were prepared from PBL isolated from two different individuals and treated with phytohemagglutinin in the presence of either IL-2 (to stimulated a TH1like phenotype) (28,29) or IL-4 (to stimulated a TH2-like phenotype) (29,30). There was no measurable difference in [ 35 S]Sch225336 binding correlating with the two differentiation conditions.
Human promyelocytic leukemia HL-60 cells have been reported to express functional hCB2 (2,19,26). Saturation analyses in membranes from HL-60 cells differentiated toward a more neutrophil-like phenotype by prolonged incubation in the presence of Me 2 SO or db-cAMP reveal saturable, high affinity [ 35 S]Sch225336 binding (Fig. 5C). [ 35 S]Sch225336 binding in membranes from monocytic (phorbol 12-myristate 13-acetate-treated) HL-60 cells (Fig. 5D) is considerably less than that measured in Me 2 SO-treated HL membranes or in the monocyte-like MonoMac6 cells (Fig. 5A). There was no measurable [ 35 S]Sch225336 binding in membranes prepared from either retinoic acid-pretreated HL-60 cells, which stimulates monocytic differentiation (10) (data not shown), or in human monocyte THP-1 cell membranes (Fig.  5D). Taken together, these data demonstrate that [ 35 S]Sch225336 has excellent utility for detection of CB2 at physiological expression levels.

The Effect of hCB2 Coupling State on [ 35 S]Sch225336
Pharmacology-We have previously established that Sch225336 is an inverse agonist (7) based on its ability to inhibit constitutive CB2 signaling and its higher affinity for uncoupled hCB2 conformation(s). To further define the pharmacology of Sch225336, we measured the effect of the nonhydrolyzable GTP analog, GppNHp, on [ 35 S]Sch225336 binding in CB2 transfectants (RB-hCB2, Sf9-hCB2) and hemopoietic cell lines (HL-60(Me 2 SO) and U937). GppNHp increased [ 35 S]Sch225336 binding in a concentration-responsive manner although the extent of the increase varied between lines (Fig. 6,  A and B). In RB-hCB2 membranes, which have the most [ 35 S]Sch225336 binding sites, binding increased only slightly with uncoupling, much less than the increase measured either in HL-60(Me 2 SO) membranes or in Sf9-hCB2 membranes that overexpress both hCB2 and its G proteins (1). GppNHp had the largest relative effect on [ 35 S]Sch225336 binding in U937 membranes (Fig. 6B), suggesting that although CB2 expression is relatively low, a greater proportion of the receptor in these membranes is coupled to G proteins. As would be expected, CB2 uncoupling with GppNHp decreased the binding of the agonist [ 3 H]CP55,940 in the Sf9-hCB2 and U937 membranes, whereas [ 35 S]Sch225336 binding increased (Fig. 6C). Note that [ 3 H]CP55,940 binding in U937 membranes is barely measurable. To better assess the impact of receptor coupling in the U937 membranes, we performed saturation analyses with [ 35 S]Sch225336 in these membranes in the presence or absence of another nonhydrolyzable GTP analogue, GTP␥S (Fig. 6D). In the absence of the uncoupling agent, [ 35 S]Sch225336 bound with a K d ϭ 0.13 Ϯ 0.04 nM. Consistent with the effect of GppNHp (above), co-incubation with GTP␥S increased the affinity of [ 35 S]Sch225336 (K d ϭ 0.066 Ϯ 0.022 nM) in the U937 membranes without affecting the B max . The increased affinity associated with CB2 uncoupling is readily apparent in the Scatchard analyses of the data (Fig. 6D, inset).

Autoradiographic Analyses of CB2 Expression in Spleen and Lymphoid
Tissues-We assessed the utility of [ 35 S]Sch225336 or [ 3 H]CP55,940 for autoradiographic analysis of CB2 expression in human spleen (Fig. 7, A  and B) or human peripheral lymph node (Fig. 7C). Incubation of the spleen slices with either radioligand results in a distinct, punctate labeling of areas representing splenic white pulp based on lymphocyte surface marker staining (Fig. 8). The images generated with [ 35 S]Sch225336 are much more detailed than those generated with the tritiated material and require relatively short film exposure times (5-8 days) versus that required with [ 3 H]CP55,940 (72 days). Based on superior signal:noise, [ 35 S]Sch225336 was used for analysis of human lymph node sections. As can be seen in Fig. 7C, [ 35 S]Sch225336 binding relative to nonspecific binding is apparent, although the displaceable binding is too small to be conclusive. We can conclude, however, that CB2 expression in the lymph node section is significantly lower than in the spleen.
To identify the cell types accounting for [ 35 S]Sch225336 binding in the spleen (Fig. 8, A and B), adjacent sections were stained with the B and T cell surface markers, CD79␣ and CD3 (Fig. 8, C and D). The pattern of [ 35 S]Sch225336 binding corresponds predominantly to B lymphocytes in the primary lymphoid structures of the spleen (Fig. 8C). [ 35 S]Sch225336 binding to T lymphocytes appears fairly low compared with the B cells (Fig. 8D), although we cannot conclusively rule out T cell binding. Interestingly, there are CD79␣-positive B cells scattered in the red pulp (Fig. 8D, arrows), although no significant [ 35 S]Sch225336 binding localized with these cells.

DISCUSSION
Based on initial in situ hybridization studies, it was apparent that hCB2 has a rather defined expression in peripheral immune cells and multiple lymphoid organs (31,32). CB2 mRNA is found in spleen, tonsils, bone marrow, mast cells, peripheral blood leukocytes, and a variety of hemopoietic cell lines, including the myeloid cell line U937 and HL-60 cells (33)(34)(35)(36)(37)(38)(39). Based on these data, it has been suggested that hCB2 may have a direct role in immune function and/or mediate the immunosuppressive effects of cannabinoids. Subsequently, it was shown that 2-arachidonyl glycerol, a full agonist at hCB2 (1), stimulated chemotaxis of hemopoietic cells (monocytic HL60, THP-1, U937, human monocytes, leukemia EoL-1 cells,   and human peripheral blood eosinophils) via an endogenous receptor (2, 3) The authors reasonably credited this response to CB2 due to its sensitivity to blockade by the CB2 inverse agonist SR144528 and immunoblot data suggesting that CB2 protein was expressed in eosinophils. Indeed, direct demonstration of CB2 expression has been difficult due to the scarcity of commercially available antibodies with substantial and/or authentic CB2 immunoreactivity. Also, in contrast to the relatively high expression of CB1 in the brain (Ն1 pmol/mg; (40), CB2 expression in immune cells/tissues is more moderate (10 -300 fmol/mg; Fig. 5). Lastly, there are no CB2-selective radioligands with high specific activity. Not surprisingly, this has resulted in the almost exclusive use of recombinant expression systems for CB2 binding studies (5 The primary advantage of [ 35 S]Sch225336 over other cannabinoid radioligands may be its utility for autoradiography. This application is important as the commercially available anti-hCB2 antibodies that we tested were ineffective in detecting authentic receptor protein in immunoblot analyses of samples prepared from cells that express hCB2 at high levels either recombinantly (ϳ7-20 pmol/mg) or endogenously in Mono-mac6 cells (ϳ0.3 pmol/mg hCB2). Using a cell line expressing HA-tagged hCB2, we showed that the receptor migrates as a diffuse 55-kDa protein that is absent in parental CHO cells. The diffuse nature of the band and its absorption to wheat germ agglutinin-agarose is consistent with the expression of hCB2 as a glycoprotein. The appearance of the receptor in our Western blots is very reminiscent of that seen in immunoblots of CHO-hCB2 membranes by Carayon (19) using their polyclonal antibody raised against a peptide sequence within the carboxylic tail of the receptor. The authors performed very nice studies on endogenous CB2 expression with this antibody, demonstrating expression of the receptor on CD4/CD8 T cells, natural killer cells, and B cells. Unfortunately, this reagent is not commercially available.
As an alternative (or additional) approach to immunodetection of cannabinoid receptors, tritiated radioligands have been used successfully for autoradiographic studies of CB1 expression in the central nervous system wherein CB1 is highly expressed (41,42 (44). [ 3 H]CP55,940 has been used with some success for autoradiography studies to detect CB2 (presumably) in immune tissues from rat (31,45) and human (Fig. 7B). However, in our studies, the image quality arising from studies with [ 35 S]Sch225336 are much superior to that generated with the tritiated agonist, and the necessity of using emulsions or specialized film to detect the lower energy emissions from tritium is avoided. Indeed, because of its higher specific activity relative to tritiated cannabinoids (Ͼ1400 versus ϳ20 Ci/mmol), experiments with [ 35 S]Sch225336 required much shorter exposure times to film (5-8 days versus 72 days for tritium). Based on immunohistochemistry in parallel sections, CB2 appears to be expressed predominantly on B cells in the white pulp with little expression in the red pulp or T cells. This observation is consistent with the results published by Rayman et al. (46). These authors used two different specific anti-CB2 antibodies in immunohistochemistry and showed that CB2 receptor was predominantly expressed by follicular B cells in the primary follicles of human spleen.
In conclusion, [ 35 S]Sch225336 is a CB2-selective radioligand with unparalleled specific activity (Ͼ1400 Ci/mmol) and binding affinity. As such, [ 35 S]Sch225336 is uniquely suited for the study of endogenously expressed CB2 both by equilibrium binding analyses and autoradiography.