pubs.acs.org/jmc Article Orally Active Peptide Vector Allows Using Cannabis to Fight Pain While Avoiding Side Effects Maria Gallo, Estefanía Moreno, Sira Defaus, Antonio Ortega-Alvaro, Angel Gonzalez, Patricia Robledo, Marco Cavaco, Vera Neves, Miguel A. R. B. Castanho, Vicent Casadó, Leonardo Pardo, Rafael Maldonado,* and David Andreu* Cite This: J. Med. Chem. 2021, 64, 6937−6948 Read Online ACCESS Metrics & More Article Recommendations * sı Supporting Information ABSTRACT: The activation of cannabinoid CB1 receptors (CB1R) by Δ9-tetrahydrocannabinol (THC), the main component of Cannabis sativa, induces analgesia. CB1R activation, however, also causes cognitive impairment via the serotonin 5HT2A receptor (5HT2AR), a component of a CB1R−5HT2AR heteromer, posing a serious drawback for cannabinoid therapeutic use. We have shown that peptides reproducing CB1R transmembrane (TM) helices 5 and 6, fused to a cell-penetrating sequence (CPP), can alter the structure of the CB1R−5HT2AR heteromer and avert THC cognitive impairment while preserving analgesia. Here, we report the optimization of these prototypes into drug-like leads by (i) shortening the TM5, TM6, and CPP sequences, without losing the ability to disturb the CB1R−5HT2AR heteromer, and (ii) extensive sequence remodeling to achieve protease resistance and blood−brain barrier penetration. Our efforts have culminated in the identification of an ideal candidate for cannabis-based pain management, an orally active 16-residue peptide preserving THC- induced analgesia. ■ INTRODUCTION About 20% of the population suffers from chronic pain that in cannabis users.8,14−17 These adverse effects pose a serious hurdle to the therapeutic use of cannabinoids; hence, the seriously affects personal and professional life, highlighting the possibility to minimize their impact would constitute a major breakthrough. urgent need for analgesics that effectively alleviate pain with Behavioral studies in mice lacking the serotonin 5HT2A minimal side effects.1 Opioids, thus far the mainstay in severe receptor (5HT2AR) revealed a remarkable dissociation pain management, are under intense debate for their poor between the beneficial antinociceptive effects of THC and its safety profiles and high potential for abuse, as reflected in the detrimental amnesic properties.18 As CB1R and 5HT2AR are G opioid abuse disorder epidemic claiming about 145 deaths/day protein-coupled receptors (GPCRs), a class of proteins in the recent years in the USA.2,3 On the other hand, patients frequently forming oligomers in cells,19 we provided evidence with minor pain manifestations are usually dealt with that the underlying molecular basis for that observation was nonsteroidal anti-inflammatories (NSAIDs) that not only fail the formation of CB1R−5HT2AR heteromers.18 Both receptors quite often to achieve adequate relief but also have serious side are functionally active in specific brain regions; the heteromer effects (e.g., gastrointestinal and renal).4 Thus, there is a need shows specific biochemical signatures distinct from those of the to fill the therapeutic gap between opioids and NSAIDs under protomers, in artificial cell lines as well as native tissues; very multiple moderate pain conditions or in situations where importantly, while the cognitive impairment caused by THC is opioids are ineffective, such as severe neuropathic pain.5 due to the CB1R−5HT2AR heteromer, the antinociceptive The analgesic potential of cannabinoid compounds has long effects result solely from CB1R activation.18 The arrangement been recognized.6−8 The activation of cannabinoid CB1 receptors (CB1R) by agonists, such as Δ9-tetrahydrocannabi- nol (THC), the main psychoactive component of Cannabis Received: March 17, 2021 sativa, elicits therapeutically relevant responses for the Published: April 23, 2021 treatment of migraine,9,10 rheumatoid arthritis,10 osteoarthri- tis,11 neuropathic,12 or cancer-related pain,13 among many others. Unfortunately, CB1R activation is also linked to side effects, such as memory impairment, with major consequences © 2021 American Chemical Society https://doi.org/10.1021/acs.jmedchem.1c00484 6937 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Table 1. From TM5/6 Prototypes to a Drug-Like Finalist Peptidea peptide sequenceb residues 18 prototypes TM5-Tat ETYLMFWIGVTSVLLLFIVYAYMYILWGRKKRRQRRR 37 Tat-TM6 GRKKRRQRRRKTLVLILVVLIICWGPLLAIMVYDVF 36 optimized candidate wliymyayvaGilkrw 16 a A detailed list of all peptides used in the various rounds of the optimization process is given in Tables S1−S4. bUpper and lower case denote L- and D-residues, respectively. of CB1R and 5HT2AR protomers in the heteromer is via contrast to peptide TM7-Tat, used as a negative control transmembrane (TM) helices 5 and 618 as studied by (Figure 2a). Moreover, the biochemical properties of the bimolecular fluorescence complementation (BiFC) experi- heteromer differ from those of the protomers.18,23,24 In CB1R− ments with synthetic peptides TM5-Tat and Tat-TM6, 5HT2AR, in addition to a Gq to Gi switch in 5HT2AR coupling where TM5 and 6 helices are fused to cell-penetrating peptide preference, antagonist binding to one of the receptors blocks (CPP) GRKKRRQRRR, that is, HIV Tat(48−57),20 respec- signaling of the interacting receptor (known as cross- tively, at N- or C-termini, ensuring TM native alignment antagonism), and costimulation with both agonists does not relative to the membrane. produce an additive effect (known as negative cross-talk).23,24 The formation of a stable four-helix bundle between these Thus, we next determined signaling after receptor activation by TM5 and 6 helices of both protomers, as described for the measuring cAMP levels [decrease in forskolin (FK)-induced dimeric crystal of the μ-opioid receptor,21 was suggested to be cAMP as the result of adenylate cyclase inhibition by Gi] or the involved in the allosteric interactions between protomers and increase in pERK in the ERK1/2 phosphorylation pathway in the mechanisms behind the specific signaling signatures of (Figure 2d,g). Cells stimulated with FK and treated with the the heteromer.18,22 Moreover, interfering TM5-Tat and Tat- CB1R agonist WIN 55,212-2 (WIN) or 5HT2AR agonist 2,5- TM6 peptides were able to alter the structure of the heteromer dimethoxy-4-iodoamphetamine (DOI) showed reduced cAMP in vivo, leading to a selective abrogation of memory production in all groups, as expected for Gi-coupled receptors. impairments caused by the exposure to THC.18 From this, it Neither TM5-Tat, Tat-TM6, peptides 1 or 3, nor TM7-Tat was logical to assume that disturbing the formation of the (negative control) influenced G protein coupling preferences CB1R−5HT2AR heteromer might be an effective strategy for (5HT2AR remains Gi-coupled in the presence of interfering harnessing the therapeutic potential of THC while avoiding its peptides). The CB1R antagonist rimonabant (RIM) blocked side effects. TM5-Tat and Tat-TM6 peptides, however, were the decrease in FK-induced cAMP or the increase in pERK ungainly therapeutic leads with poor prospects related to their triggered by the 5HT2AR agonist DOI, and the 5HT2AR large size, low solubility, and inability to cross the blood−brain antagonist MDL 100,907 (MDL) also blocked the decrease in barrier (BBB). We now report on a comprehensive cAMP and increase in pERK induced by the CB1R agonist optimization effort leading from those prototypes to a WIN (bidirectional cross-antagonism). Notably, peptides 1 candidate with vastly improved, drug-like features (Table 1) and 3 eliminated this bidirectional cross-antagonism as including small size, proteolytic stability, BBB crossing ability, efficiently as TM5-Tat and Tat-TM6 peptides, an effect that and, remarkably, oral activity, thereby achieving the desired was not observed with TM7-Tat (negative control). effects, both in vitro and in vivo. Coadministration of WIN and DOI agonists, in the absence ■ RESULTS Shorter Peptides Mimicking CB1R TM Helices 5 and 6 of peptides, does not lead to a further statistically significant decrease in cAMP or increase in pERK (negative cross-talk). However, in the presence of 1 and 3, as well as TM5-Tat and and Altering CB1R−5HT2AR Heteromers. The design of Tat-TM6 (but not TM7-Tat), the simultaneous addition of these peptides required the identification of the amino acid both agonists results in a significant decrease in cAMP or an residues in TM5-Tat and Tat-TM6 that contribute more increase in pERK (absence of negative cross-talk). In sum, significantly to the binding to 5HT2AR. This was studied by interfering peptides 1 and 3 successfully blocked key several replicas of unbiased 1 μs molecular dynamic (MD) biochemical signatures (cross-antagonism and negative cross- simulations of TM5-Tat and Tat-TM6 in complex with talk) of the CB1R−5HT2AR heteromer as efficiently as TM5- 5HT2AR (see Figure 1a−c and Experimental Section). The Tat and Tat-TM6. simulations show that the intracellular part of the peptides We assumed that the segments of peptides TM5-Tat and remains more stable than the extracellular one (more Tat-TM6 replicating CB1R TM helices 5 and 6 are membrane- noticeable for Tat-TM6) due to specific interactions with embedded as in the receptor, whereas Tat(48−57) is in the 5HT2AR (mainly the V19-Y20-Y22-M23-L26-W27 stretch in intracellular space (Figure 1b,c). Peptides 2 and 4 (Table S1), TM5-Tat or the K11-V14-L17-V18 stretch in Tat-TM6, lacking the Tat(48−57) sequence, had no effect on decreasing Figure 1d,e). fluorescence complementation (Figure 2a) or in signaling Based on these predictions,18 peptides VYAYMYILW-Tat (Figure 2d−g), underscoring the essential role of the CPP and Tat-KTLVLILVV (1 and 3, Table S1), respectively, motif. mimicking the key interactions of TM5-Tat and Tat-TM6, HIV-Tat Replacement by Shorter CPPs. Inspired by were designed. Their ability to alter CB1R−5HT2AR heteromer recent studies,25−28 a second set of analogues were devised formation was first tested by BiFC assays,18,23,24 in HEK-293T (Table S2) in which four alternatives to 10-residue Tat(48− cells transfected with receptors fused to two complementary 57) were explored: (i) 7-residue PepH3 (AGILKRW), a halves of YFP (5HT2AR-cYFP and CB1R-nYFP). In this assay, reported BBB shuttle25 (peptides 5 and 9); (ii) 6-residue 1 and 3 caused a decrease in fluorescence complementation SKSKSK26 (6 and 10); (iii) 5-residue RLRWR27 (7 and 11), comparable to the TM5-Tat and Tat-TM6 prototypes, in and (iv) 3-residue RKR28 (8 and 12). The last two CPP motifs 6938 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Figure 1. (a−e) MD simulations of 5HT2AR in complex with TM5-Tat and Tat-TM6 peptides. The previously published computational model of the CB1R−5HT2AR heteromer18,20 (panel a) was used as the starting point for MD simulations (see Experimental Section) of a membrane- embedded (yellow spheres) 5HT2AR (green ribbons) in complex with peptides where the TM5 (blue ribbon, panel b) or TM6 (red ribbon, panel c) sequences are fused to HIV-Tat(48−57) (orange). These TM5-Tat and Tat-TM6 peptides (Table 1) were stable in the three replicas of 1 μs unbiased MD, as shown by root-mean-square deviations (Figure S1). Panels (d,e) show detailed views of the interactions between TM5-Tat and Tat-TM6, respectively, with TM helices 5 and 6 of 5HT2AR. Snapshots along the trajectories are shown as transparent ribbons, whereas representative structures are solid ribbons. Side chain residues involved in stable interactions of TM5-Tat and Tat-TM6 with 5HT2AR 6939 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Figure 1. continued (Ballesteros−Weinstein notation) during the MD simulations are labeled. Interaction frequency maps for frequencies above 50% in at least one of the three replicas (r1, r2, and r3) are depicted. TM5-Tat and Tat-TM6 residues involved in stable interactions with two or more 5HT2AR residues are highlighted in bold. Figure 2. (a−c) BiFC analysis of the effect of designed peptides on CB1R−5HT2AR heteromerization. Fluorescence (530 nm) of HEK-293T cells transfected with 5HT2AR-cYFP and CB1R-nYFP treated with a vehicle (no peptide) or 4 μM peptide for 4 h. Values are mean ± SEM of n = 10− 30; TM5-Tat and Tat-TM6 are positive controls, whereas TM7-Tat is a negative control; ns (no significant), *(p < 0.05), **(p < 0.01), and ***(p < 0.001) represent significantly different (two-tailed) values (one-way ANOVA followed by Tukey’s multiple comparison tests). (d−i) Effect of designed peptides on the decrease in FK-induced cAMP (d−f) or increase in ERK1/2 phosphorylation. (g−i) Transfected cells preincubated with a vehicle (no peptide) or with 5HT2AR (5HT2ARago, DOI, 100 nM) or CB1R (CB1Rago, WIN, 100 nM) agonists or with 5HT2AR (5HT2ARanta, MDL, 300 nM) or CB1R (CB1Ranta, RIM, 1 μM) antagonists and combinations thereof, in the presence or absence of 0.5 μM FK. Values in panels (d−f) are mean ± SEM of n = 3−6 of FK-treated cells. Quantification of phosphorylated ERK-1/2 was determined by the α-screen bead-based technology. Values in panels (g−i), expressed as a percentage of basal (nontreated cells), are mean ± SEM of n = 4−9. Two-way ANOVA followed by Tukey’s multiple comparison tests was used to analyze the data (Table S5). Cartoons depict the designed peptides (cylinders with color codes as in Figure 1) fused to different CPPs (lines). (7, 8, 11, and 12) were not as efficient as Tat(48−57) in ability (Figure 2c) and blocking of cross-antagonism and decreasing fluorescence complementation, as measured in negative cross-talk (Figure 2f,i), as somehow expected from the BiFC assays (Figure 2b), neither could 8 or 12 abolish the stereochemical nonmodification of the membrane-embedded cross-antagonism and negative cross-talk signatures of the TM segments. CB1R−5HT2AR heteromer (Figure 2e,h). In contrast, peptides BBB Translocation. In our earlier report,18 tacitly 5/9 and 6/10 were as effective as peptides 1/3 in fluorescence assuming the inability of TM5-Tat and Tat-TM6 to traverse complementation or signaling (Figure 2b,e,h) and were thus the BBB, these peptides were administered by intracerebroven- selected for further optimization. tricular (ICV) infusion, an invasive route of little interest for CPP Stereochemistry and Its Significance. Replace- human therapeutic application. However, for a peptide drug to ment of natural L-by D-amino acids results in CPP sequences be a successful candidate, BBB-crossing capability is with improved cell entry and metabolic stability.29,30 We thus mandatory. Therefore, to explore the BBB-crossing features switched the CPP motifs of peptides 5, 6, 9, and 10 to their of optimized peptides 13−16, their 5(6)-carboxyfluorescein- enantiomeric D-versions, while preserving the L configuration labeled versions (see Experimental Section) were tested for for the cell membrane-embedded sections replicating CB1R translocation in an in vitro BBB model using mouse b.End3 TM5 or 6 (peptides 13−16, Table S3). Trypsin incubation cells.30,31 Fluorescence intensity readings at the basolateral over 24 h confirmed our expectations of improved proteolytic chamber of the transwell device showed that 13 and 15, with stability in the 13−16 set, in contrast to poor survival for all-L the TM-altering sequence fused to D-CPP motif aGilkrw, counterparts (5, 6, 9, and 10) (Figure S2a). More relevantly, translocated more efficiently than 14 and 16, where the TM- 13−16 maintained the CB1R−5HT2AR heteromer-disturbing altering sequence was fused to the sksksk vector (Figure S2b). 6940 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article The high in vitro BBB integrity observed after 24 h incubation ruled out paracellular transport (Figure S2c). In parallel, cytotoxicity studies on b.End3 cells were conducted to discard potential toxic effects modifying barrier permeability, that is, artifactual peptide ability to cross the BBB (Figure S2d). Applying Configurational Switch and Sequence Reversal for Last-Stage Lead Optimization. Although considerable progress had been achieved in terms of BBB- crossing, in the CPP-enabled constructs delivering short TM payloads to the CB1R−5HT2AR heteromer, the essential modifying segments (i.e., short TM5: VYAYMYILW and short TM6: KTLVLILVV) were still made up of natural L- amino acids, susceptible to endopeptidases such as chymo- trypsin, thermolysin, or pepsin, targeting preferentially hydro- phobic and/or aromatic residues (Ile, Leu, Val, Ala, Met, Trp, and Tyr) whose abundance in those TM sequences was an important concern. To boost protease resistance, both in body fluids and within the target cell, we developed two last-stage candidates (peptides 17−18, Table S4) where the optimal BBB-permeable shuttle (i.e., aGilkrw) was fused to the retro- enantio (r/e) versions of the TM5- and TM6-altering sequences (i.e., short TM5 and short TM6 composed of D- amino acids in the reversed sequence). In such an arrangement (Figure 3), if the all-L sequence is displayed in the conventional Figure 3. (a) Retro modification of a short TM5 peptide (VYAYMYILW) where the peptide sequence has been synthesized in the reverse order (WLIYMYAYV). (b) When the enantio Figure 4. (a) BiFC analysis of the effect of peptides 17 and 18 on modification (D-amino acids) is additionally applied (wliymyayv), CB1R−5HT2AR heteromerization. Fluorescence (530 nm) of HEK- the orientation of the side chains is similar to that of the original 293T cells transfected with 5HT2AR-cYFP and CB1R-nYFP treated peptide. with a vehicle or peptide. Values are mean ± SEM of n = 10−30. TM5-Tat and Tat-TM6 are positive controls, whereas TM7-Tat is a negative control; ns (no significant), (one-way ANOVA followed by Tukey’s multiple comparison tests). (b) Effect of peptides 17 and 18 N- to C-terminal fashion but the r/e analogue is laid out in the on cAMP production. Transfected cells preincubated with a vehicle opposite (C- to N-terminal) sense by way of a 180° rotation in (no peptide) or with 5HT2AR (5HT2ARago, DOI, 100 nM) or CB1R the plane, the side chains adopt similar orientations,32−35 (CB1Rago, WIN, 100 nM) agonists or with 5HT2AR (5HT2ARanta, although the amide bonds show inverted (CONH vs NHCO) MDL, 300 nM) or CB1R (CB1Ranta, RIM, 1 μM) antagonists and directions. In summary, peptides 17−18, recapitulating the combinations thereof, in the presence of FK. Values are mean ± SEM of n = 3−6 of FK-treated cells. Two-way ANOVA followed by various rounds of structural elaboration so far, met a priori all Tukey’s multiple comparison tests was used to analyze the data desired requirements, including (i) the same TM5/TM6 side (Table S5). (c) In vitro translocation of peptides 17 and 18. chain orientation than previous CB1R−5HT2AR-modifying Translocation (%) across the transwell BBB model quantified as analogues, (ii) high proteolytic resistance as expected from an fluorescence in the basolateral chamber after 24 h. A trans-BBB all-D composition,36 and (iii) ability to traverse the BBB.30 peptide was used as the positive control. Values are mean ± SEM of n When 17 and 18 were submitted to the same screens as = 3−6. *p < 0.05 vs positive control (one-way ANOVA followed by previous candidates (Figure 4), our expectations were fulfilled. Dunnett post hoc tests). (d) BBB integrity to peptides 17 and 18. The ability to interfere the CB1R−5HT2AR heterodimer Measured as permeability of fluorescent FD4 dextran upon peptide structure and signaling (Figure 4a,b) was preserved, serum exposure. Values are mean ± SEM of n = 3−6. No statistical degradation was successfully averted (Figure S3a), and despite significance difference was observed between samples (one-way ANOVA followed by Tukey’s multiple comparison test). (e) Confocal the extensive structural modifications undergone, BBB-crossing microscopy. Images showing peptide internalization into bEnd.3 cells. properties were also maintained (Figures 4c,d, S3b). This last feature was further confirmed by examining internalization of both 17 and 18 in bEnd.3 cells by confocal microscopy (Figure In Vivo Validation of Finalist Peptides 17 and 18. The 4e). promising in vitro data obtained for analogues 17 and 18 6941 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Figure 5. Prevention of THC-induced amnesic effects by peptide 17, while maintaining THC analgesic responses. (a) Analgesic effects of THC (10 mg/kg, IP) in the mouse hot-plate test were preserved after pretreatment with peptides 17 and 18 (both 20 mg/kg, IV). (b) Amnesic effects of THC (3 mg/kg, IP) observed in the mouse novel object recognition test were abrogated by pretreatment with 17 (20 mg/kg, IV) but not with 18 (20 mg/kg, IV). Data are presented as mean ± SEM (n = 10−12). *p < 0.05 vs vehicle + vehicle (Fisher LSD test). encouraged us to validate their in vivo efficacy in mice. Given their BBB-crossing features, we evaluated the efficacy of the peptides, intravenously (IV) administered, by examining the preservation of THC analgesic effects (hot-plate test) and the disruption of its amnesic effects (novel object recognition paradigm) (see Experimental Section for a detailed description of both assays). The in vitro-based expectations for 17 and 18 were partially fulfilled. Thus, the analgesic effects induced by THC (10 mg/ kg by intraperitoneal route, IP) were preserved when coadministered with either peptide (20 mg/kg, IV) (Figure 5a). In contrast, the memory impairment produced by THC (3 mg/kg, IP) was suppressed only when coadministered with 17 (20 mg/kg, IV), but amnesic effects remained unaffected when THC was coadministered with 18 (20 mg/kg, IV) (Figure 5b). Therefore, peptide 17 was chosen to evaluate the effects following oral administration (OR), the suitable route for therapeutic purposes. THC (10 mg/kg, IP) analgesic effects Figure 6. Prevention of THC-induced amnesic effects by peptide 17 remained unaffected by coadministration with 5, 10, and 20 administered orally, while maintaining THC analgesic responses. (a) mg/kg (OR) doses of 17 (Figure 6a), whereas amnesic effects Analgesic effects of THC (10 mg/kg, IP) revealed in the mouse hot- (3 mg/kg THC, IP) were reverted by the same doses of 17 plate test were preserved after pretreatment with 17 (5, 10, and 20 given by gavage (Figure 6b), demonstrating the effectiveness of mg/kg, OR). (b) Amnesic effects of THC (3 mg/kg, IP) observed in the peptide by the most desirable route of administration. the mouse novel object recognition test were abrogated by Non-immunogenicity Confirmed. After showing that pretreatment with 17 (5, 10, and 20 mg/kg, OR). Data are presented peptide 17 epitomized our wish list for an optimal Cannabis- as mean ± SEM (n = 5−8). **p < 0.01, ***p < 0.001 vs vehicle + complementing drug candidate, a final concern was its possible vehicle. #p < 0.05 vs vehicle + THC (Fisher LSD test). immunogenicity, despite expectations that a small 16-residue peptide would most likely go undetected by the immune that constitutes an optimal candidate for combined admin- system. To this end, we immunized mice using a repeated istration with THC, taking advantage of its analgesic effects but schedule with a high dosage (100 μg of 17 on 5 consecutive avoiding its associated cognitive impairment. days, and a boost at day 15). As shown in Figure 7, repeated administration of 17 in the free form did not evoke any detectable immune response. Antibodies to 17 could only be ■ DISCUSSION CB1R is the most-abundant GPCR in the mammalian brain eventually raised by immunization with the keyhole limpet and the main target of THC.37−40 While originally thought to hemocyanin (KLH)-conjugated version, a positive control of exist exclusively as a monomer,41 strong evidence has immunogenicity that is, on the other hand, irrelevant with accumulated indicating its ability to form homo- and regard to administration of 17 as a therapeutic agent. heterodimer complexes42−48 with new pharmacological and In sum, our research has led to an orally active, functional properties.49 Of particular relevance in this regard is proteolytically stable, and nonimmunogenic short peptide the formation of the CB1R−5HT2AR heteromer, as previously 6942 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Figure 7. (a) Immunogenicity assessment. The procedure was devised to mimic a daily chronic pain therapy. Mice were treated for 5 consecutive days and boosted at day 15 with peptide 17 in phosphate-buffered saline (PBS) (group 1, 100 μg/dose), peptide 17-KLH conjugate in PBS (group 2, positive control, 30 μg), or PBS (group 3, negative control). The lower dose of the 17-KLH conjugate, compared to 17 alone, was chosen to prevent potential protein toxicity in the mice. Serum samples were collected on days 0, 9, and 19 and animals were euthanized on day 19. (b) ELISA of humoral response. Peptide-specific antibody response in sera collected at the indicated days. Each point depicts mean antibody titers ± SD. The dotted line indicates the detection limit. Free (nonconjugated) 17 induces no significant response; immunogenicity is observed only upon conjugation to a large-size carrier protein (KLH, MW ∼ 5 MDa) and boosting. revealed by (i) in vitro proximity-based biophysical techniques, bivalent ligands, that is, single chemical entities capable of (ii) specific biochemical signatures of heteromers distinct from interacting simultaneously with the orthosteric sites of a those of the protomers, (iii) in situ proximity ligation assay to (homo/hetero) dimer via two pharmacophore units covalently detect protein−protein interactions in native tissues, (iv) linked by an appropriate spacer.53,54 We have developed here a membrane-permeable peptides altering heteromeric interfaces, new approach in which the ligand/peptide binds at the TM and (v) mice expressing heteromerization-deficient receptors, dimerization interface of the heteromer. This rests on the all of them established criteria for the existence of assumption of allosteric binding sites at the extrahelical part of CB1R−5HT2AR heteromers in vivo.50 This CB1R−5HT2AR GPCRs, facing the membrane bilayer. The numerous recent heteromer, therefore, is not simply due to receptor over- structures of GPCRs bound to allosteric modulators have expression, but the consequence of an explicit molecular confirmed the presence of these sites for families A and B.55 mechanism of THC to dissociate beneficial antinociceptive Related to the GPCRs of this manuscript, the negative response and detrimental amnesic effects,50 hence defining it as allosteric modulator ORG27569 of CB1R binds to an a novel therapeutic target. extrahelical site, mostly interacting with TM4.56 The development of heteromer-specific drugs is more Protein−protein interfaces in general, and GPCR interfaces challenging than traditional drug-discovery programs, where in particular, consist of large and flat contact areas with no efforts are focused on compounds binding on orthosteric sites well-defined pockets.57 Here, we have taken advantage of our of monomeric GPCRs of endogenous ligands.51,52 One of the previous studies in which peptides with the TM5 and TM6 most widespread options to target heteromers is designing sequences of CB1R (TM5-Tat and Tat-TM6), fused to a HIV- 6943 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Tat CPP vector, can bind 5HT2AR and modify the quaternary were subject to 1000 cycles of energy minimization, followed by 20 ns structure of the CB1R−5HT2AR heteromer.18 MD simulations of gradual relaxation of positional restraints (corresponding to 100, of TM5-Tat and Tat-TM6 in complex with 5HT2AR have 50, 25, and 10 kJ mol−1 nm−2) at protein backbone coordinates before allowed to identify these “hotspots” near the intracellular part. the production phase in order to hydrate the receptor cavities and allow lipids to pack around the protein. The AMBER99SB force field The proteolytic instability of TM5-Tat or Tat-TM6 and as implemented in GROMACS and Berger parameters for POPC their inability to penetrate the BBB complicate any prospect of lipids were used for the MD simulations. After equilibration, three systemic administration. With this in mind, we have set out to replicas of 1 μs of unrestrained MD simulation were performed at a develop strategies aimed at improving the druggability of the constant temperature of 300 K and a time step of 2.0 fs. Lennard- TM5-Tat and Tat-TM6 prototypes by (i) downsizing the Jones interactions were computed using a cutoff of 10 Å, and the length of the interfering peptides to only include the essential electrostatic interactions were treated using PME with the same real- identified “hotspots”, (ii) enabling CNS delivery via a BBB- space cutoff under periodic boundary conditions. MD simulations permeating CPP shuttle, (iii) ensuring proteolytic robustness, were performed using GROMACS 2019. hence bioavailability, by judicious structural manipulation, Peptide Synthesis, Analysis, and Purification. Peptides were assembled in a Prelude instrument (Protein Technologies, Tucson, including the switch to enantio and retro-enantio versions of AZ) running optimized Fmoc synthesis protocols as described CPP and interfering motifs, respectively. To this end, we have before.60 For labeled peptides, the incorporation of 5(6)-carboxy- used in silico tools and peptide medicinal chemistry criteria fluorescein to the carboxy-terminal end was performed by adding an through a series of design/optimization rounds and a variety of additional Fmoc-Lys(Mtt)-OH residue to the sequence, the Mtt in vitro and in vivo screens that have eventually allowed to group selectively removable on-resin with 1% trifluoroacetic acid, after identify peptide wliymyayvaGilkrw (17) as a highly effective, which the fluorescent dye was coupled via its carboxyl function to the orally available candidate with encouraging therapeutic Lys free ε-NH2 group. Final deprotection and cleavage were prospects. performed with a CF3COOH/H2O/3,6-dioxa-1,8-octanedithiol/ ■ triisopropylsilane (94:2.5:2.5:1 v/v) cocktail for 90 min. Peptide analysis and purification were performed as previously detailed.60 CONCLUSIONS Fractions of satisfactory purity (>90%) by analytical high-performance Pain is a highly relevant pharmacological area with still liquid chromatography (HPLC) were pooled, lyophilized, and significant unmet needs. It has been suggested that the analyzed for identity by HPLC−MS (Supporting Information, section therapeutic gap between benign but rather feeble NSAID 3). For immunogenicity studies, conjugation to bovine serum albumin painkillers and effective but high-risk opioids could be ideally (BSA) or KLH was carried out with a peptide containing an extra C- terminal Cys residue. Both carrier proteins were preactivated with m- bridged by cannabinoids, although their important undesirable maleimidobenzoyl-N-hydroxysuccinimide ester (MBS). After Sepha- side effects represent a major limitation. In addition, new dex G25 gel filtration, purified KLH-MBS or BSA-MBS was added to markets are emerging, also aimed at using cannabis-derived the peptide (r.t., 5 h, pH 7.0). The unreacted peptide was removed by products with other (i.e., non pain-related) purposes, and again dialysis and the peptide−protein conjugates were lyophilized and entraining significant risks associated to these side effects. Our quantified by the AccQTag method. work has resulted in a novel tool to minimize the most Enzyme Digestion. Peptide samples, incubated with trypsin, were prominent of these adverse outcomes, namely cognitive collected and analyzed as before.60 For human serum digestion, impairment. We specifically propose a new approach based peptide [1000 μM, 2% dimethyl sulfoxide (DMSO) in filtered H2O] on the administration of a CB1R cannabinoid agonist (e.g., was added to human serum (Sigma, St Louis, Missouri) in a 1:1 ratio and incubated at 37 °C with gentle shaking. At 0 min and 24 h, THC) in combination with a CB1R−5HT2AR altering agent. aliquots were taken and proteolysis was stopped with CH3CN (80% This compound is a 16-residue peptide (17: wliymyayva- in filtered H2O), chilled at 0 °C for 15 min to precipitate serum Gilkrw) that provides a convincing proof of concept that proteins, and centrifuged at 13,000 rpm at 4 °C for 10 min. The appropriately modified peptides constitute valid therapeutic supernatant was collected and analyzed by analytical RP-HPLC and candidates for treating pain with cannabinoids minimizing LC−MS as described before,60 with a 0−95% linear gradient of their side effects. This novel pharmacological approach could CH3CN over 15 min. All experiments are performed in duplicate and also be of interest for attenuating the side effects associated to data were fitted with GraphPad Prism. the use of cannabis derivatives for other purposes. In Vitro Biochemical and Molecular Assays. Expression ■ Vectors, HEK-293T Cell Culture, and Transient Transfection. Sequences encoding YFP Venus protein amino acid residues 1−155 EXPERIMENTAL SECTION and 156−238 were subcloned in the pcDNA3.1 vector to obtain the Computational Models. The CB 1R−5-HT2A R heteromer YFP Venus hemitruncated proteins. The human cDNAs for 5HT2AR (Figure 1a) was built from the TM5/6 dimeric interface observed and CB1R, cloned into the pcDNA3.1, were amplified and subcloned in the crystal structure of the μ-opioid receptor (PDB id 4DKL),21 as described18 to give 5HT2AR-cYFP and CB1R-nYFP. Human using the crystal structure of CB1R (5TGZ)58 and a 5HT2B-based embryonic kidney (HEK-293T) cells obtained from ATCC were (4IB4)59 homology model of 5HT2AR as previously proposed.18,20 grown in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco) The molecular 3D-coordinates corresponding to CB1R TM helices 5 supplemented with 2 mM L-glutamine, 100 μg/mL sodium pyruvate, and 6 were extracted and fused with the HIV-Tat(48−58) motif. The 100 U/mL penicillin/streptomycin, MEM nonessential amino acid resulting TM5-Tat and Tat-TM6 peptides were capped at the N- and solution (1/100), and 5% (v/v) heat-inactivated fetal bovine serum C-termini with acetyl and carboxamide groups and energy-minimized (FBS) (all supplements from Invitrogen, Paisley, Scotland, UK). Cells using the Molecular Operating Environment (MOE) software were maintained at 37 °C in an atmosphere of 5% CO2. HEK-293T (Chemical Computing Group Inc., Montreal, Quebec, Canada). cells were transfected with the corresponding fusion protein cDNA by Molecular complexes of 5HT2AR with TM5-Tat and Tat-TM6 were the polyethylenimine (Sigma) method, as described before.18 embedded in a pre-equilibrated lipid bilayer box containing 1- BiFC Assay. HEK-293T cells, after 48 h transient cotransfection palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine (POPC), water with the cDNA encoding for 5HT2AR, fused to c-YFP, and CB1R molecules (TIP3P), and monoatomic Na+ and Cl− ions (0.2 M) fused to n-YFP (4 μg of cDNA for each construct), were treated or (Figure 1b,c). The assignment of ionization states and hydrogens at not with the indicated peptides (4 μM) for 4 h at 37 °C. To quantify physiological pH for the selected structures was conducted with the protein-reconstituted YFP Venus expression, cells (20 μg protein) Protonate3D method as implemented in MOE. Molecular systems were distributed in 96-well microplates (black plates with a 6944 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article transparent bottom, Porvair, King’s Lynn, UK), and emission The following day (Training), mice were placed in the V-maze for 9 fluorescence at 530 nm was read in a Fluo Star Optima fluorimeter min, two identical objects were presented and the time that the mice (BMG Labtechnologies, Offenburg, Germany) equipped with a high- spent exploring each object was recorded. Mice were again placed in energy Xe flash lamp, using a 10 nm bandwidth excitation filter at 400 the maze 24 h later for 9 min (Test), one of the familiar objects was nm reading. Protein fluorescence expression was determined as replaced with a novel object and the total time spent exploring each of fluorescence of the sample minus the fluorescence of cells not the two objects (novel and familiar) was recorded. Object exploration expressing the fusion proteins (basal). Cells expressing 5HT2AR- was defined as the orientation of the nose to the object at less than 2 cVenus and nVenus or CB1R-nVenus and cVenus showed similar cm. A discrimination index was calculated as the difference between fluorescence levels to nontransfected cells. the time spent exploring the novel or familiar object divided by the cAMP Production and ERK-1/2 Phosphorylation Assays. For total time exploring the two objects. A higher discrimination index is cAMP production, homogeneous time-resolved fluorescence energy considered to reflect greater memory retention for the familiar object. transfer (HTRF) assays were performed as previously described.18 Peptide or vehicle were administered by IV (20 mg/kg) or oral Cells (1000 cells/well) growing in medium containing 50 μM route (5, 10, and 20 mg/kg) immediately after the training session. zardeverine were pretreated with the CB1R antagonist, RIM (1 μM, Thirty min after peptide administration, mice received an IP injection RIM), and the 5-HT2AR antagonist, MDL 100,907 (300 nM, MDL), of vehicle (5% v/v of ethanol, 5% v/v of Cremophor-EL, and 90% v/v or the corresponding vehicle in white ProxiPlate 384-well microplates saline) or 3 mg/kg of THC dissolved in vehicle. (PerkinElmer, Waltham, Massachusetts, US) at 25 °C for 20 min and Hot-Plate Test. THC-induced analgesia was measured using a hot- stimulated with the CB1 agonist, WIN 55,212-2 (100 nM, WIN) and plate meter (Hot/Cold Plate Test, Bioseb, USA) 60 min after IP the 5HT2A agonist, DOI (100 nM, DOI) for 15 min before adding 0.5 injection of either THC dissolved in IP vehicle or vehicle alone. μM FK, or vehicle and incubated for an additional 15 min period. Peptide or vehicle were administered by IV (20 mg/kg) or oral route Fluorescence at 665 nm was analyzed on a PHERAstar Flagship (5, 10 and 20 mg/kg) 30 min before THC. The plate was heated to microplate reader equipped with an HTRF optical module (BMG Lab 52 ± 0.5 °C and the time (in s) until mice showed a jumping technologies, Offenburg, Germany). response on the plate was recorded. A cutoff time of 240 s was set to For ERK-1/2 phosphorylation assay, HEK-293 cells (30.000 cells/ prevent tissue damage. well) seeded in 96-well poly-D-lysine-coated plates (Sigma-Aldrich, Mice Immunization, Bleedings, and Serum Analysis by ELISA. Madrid, Spain) were pretreated at 25 °C for 15 min with RIM (1 μM, Experiments were performed at the ICTS “NANBIOSIS” Custom RIM) and MDL 100,907 (300 nM, MDL) or the corresponding Antibody Service (CAbS, CIBER-BBN, IQAC-CSIC). The immune vehicle and stimulated for an additional 7 min with the WIN 55,212-2 response to 17 was assessed in outbred Swiss ICR (CD-1) mice strain (100 nM, WIN) and DOI (100 nM, DOI). Phosphorylation was Hsd:ICR (CD-1) (Envigo RMS, The Netherlands), following the determined in white ProxiPlate 384-well microplates (PerkinElmer immunization schedule in Figure 7. Animals were housed under Life Sciences) by the α-screen bead-based technology using the standard conditions in the CID-CSIC animal facility. Animal amplified luminescent proximity homogeneous assay kit (PerkinElmer experimental procedures were conducted in accordance with Life Sciences) and using the Enspire multimode plate reader protocols approved by CSIC Committee on Ethics of Animal (PerkinElmer Life Sciences). Phosphorylation is expressed in arbitrary Experiments and Biosafety, as well as the Spanish National units, ALPHAcounts, as measured by light emission at 520−620 nm Committee on Ethics and Animal Welfare. Briefly, a first group of of the acceptor beads. 10 female mice received IV doses of 17 (200 μL, 20 mg/kg, PBS + 2% bEnd.3 Cell Culture and Viability Assay. bEnd.3 brain endothelial DMSO), a second group of 5 mice received IV doses of 17-KLH cells (ATCC CRL-2299TM) were grown as a monolayer in DMEM conjugate (200 μL, 30 μg) and a third group of 5 mice received supplemented with 10% FBS and 1% penicillin/streptomycin solution IVdoses of PBS (200 μL, negative control). The 17-KLH conjugate, (Gibco, USA). Cells were cultured in a humidified atmosphere of 95% administered with Freund’s adjuvant, was also used to generate air and 5% CO2 at 37 °C (MCO-18AIC (UV), Sanyo, Japan), with polyclonal antibody (positive control) (data not shown). Serum daily medium replacement. samples were collected by tail vein bleeds on days 0, 9, and 19. The The cytotoxicity of peptides toward bEnd.3 cells seeded into a 96- animals were humanely sacrificed on day 19. The peptide-specific well plate (Corning, USA) at 10,000 cells/100 μL/well was tested humoral response was determined by indirect ELISA. Microtiter using CellTiter-Blue cell viability assay.31 IC50 values were determined plates (Nunc) were coated with the 17-BSA conjugate (1 μg/mL in with GraphPad Prism using a log(inhibitor) vs normalized response. coating buffer, 100 μL/well) overnight at 4 °C. The next day, plates Experiments were performed on different days using independently were washed and antiserum samples (diluted in PBST, serial dilutions grown cell cultures. starting at 1/100, 100 μL/well) were added and incubated for 30 min In Vitro BBB Translocation and Integrity Studies. The at r.t.. The plates were again washed and a solution of anti-IgG−HRP translocation capabilities of 5(6)-carboxyfluorescein-labeled peptides (1/4000 in 10 mM PBST) was added to the wells (100 μL/well) and were determined using a protocol previously described.30,31 After the incubated for 30 min at r.t.. After another wash, the tetramethylben- experiment, the integrity of the endothelial barrier was checked by zidine substrate solution was added (100 μL/well) to plates. Color measuring the permeability of a 4 kDa fluorescently labeled dextran development was stopped with 4 N H2SO4 (50 μL/well) after 30 min (FD4) as previously described.25 Experiments were carried out on at r.t., and absorbance read at 450 nm. Titers in a log10 scale were different days using independently grown cell cultures. expressed as the reciprocal of the last dilution giving the absorbance Confocal Microscopy. The internalization of 5(6)-carboxyfluor- recorded in the control wells (serum at day 0) plus 2 SD. escein-labeled peptides was evaluated using confocal microscopy as Statistical Analysis. Experimental data were managed and analyzed previously described.31 The acquisition was done on a confocal point- with GraphPad Prism software version 9 (San Diego, CA, USA) or scanning Zeiss LSM 880 microscope (Carl Zeiss, Germany) equipped IBM SPSS Statistics version 27.0 (IBM Corp., NY, USA). P-values lower than 0.05 were considered statistically significant. ■ with an alpha Plan-Apochromat X 63 oil immersion objective (1.40 numerical aperture). A diode 405−30 laser was used to excite Hoechst 33342 (Sigma-Aldrich, Spain) and CellMask Deep Red ASSOCIATED CONTENT plasma membrane stain (Thermo Fisher, USA). The 488 nm line from an Ar laser was used to excite labeled peptides. In the normal * Supporting Information sı confocal mode, X 0.6 zoom images were recorded at 1024 × 1024 The Supporting Information is available free of charge at resolution. ZEN software was used for image acquisition and Fiji https://pubs.acs.org/doi/10.1021/acs.jmedchem.1c00484. software for image processing. Design and synthesis of TMs 5 and 6 short analogues; In Vivo Behavior and Pain Response Assays. Memory Impairment Measurements. On day 1 (Habituation), mice were HIV-Tat replacement by shorter CPPs; HIV-Tat habituated for 9 min to the V-maze in which the task was performed. replacement by shorter CPPs; applying CPP stereo- 6945 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Author Contributions chemistry change; applying configurational switch and M.G. and E.M. contributed equally. V.C., L.P., R.M., and D.A. sequence reversal for last-stage lead optimization; are cosenior authors. statistical analyses used in signaling experiments; stabilities of the peptide-receptor complexes; effect of Funding CPP stereochemistry on trypsin digestion; HPLC Rhodes Pharmaceuticals (L.P., Coventry, R.I., USA; 2016− analysis of serum-incubated peptides 17 and 18; 2017), Marie Skłodowska-Curie Research and Innovation Staff analytical data; and molecular formula strings (PDF) Exchange (call H2020-MSCA-RISE-2014, 2015−2019), the Spanish Ministry of Economy and Innovation with FEDER Crystal structure of the μ-opioid receptor (PDB) funds (SAF2017-87629-R, PID2019-109240RB-I00), and the Fundació Bancària La Caixa (CaixaImpulse 2018 call; 2018− Molecular formula strings (CSV) 2020). Notes ■ AUTHOR INFORMATION The authors declare no competing financial interest. Authors will release the atomic coordinates upon article publication. ■ Corresponding Authors Rafael Maldonado − Department of Experimental and Health ACKNOWLEDGMENTS Sciences, Universitat Pompeu Fabra, 08003 Barcelona, Spain; Phone: (+34) 933160824; The authors thank Dr. Núria Pascual from Custom Antibody Email: rafael.maldonado@upf.edu Service (CAbS, CIBER-BBN, IQAC-CSIC) for help with immunizations. ■ David Andreu − Department of Experimental and Health Sciences, Universitat Pompeu Fabra, 08003 Barcelona, Spain; orcid.org/0000-0002-6317-6666; Phone: (+34) ABBREVIATIONS 933160868; Email: david.andreu@upf.edu BiFC, bimolecular fluorescence complementation; BBB, blood−brain barrier; CB1R, cannabinoid receptor 1; CNS, Authors central nervous system; CPP, cell-penetrating peptide; DOI, Maria Gallo − Department of Experimental and Health 2,5-dimethoxy-4-iodoamphetamine; FK, forskolin; GPCR, G- Sciences, Universitat Pompeu Fabra, 08003 Barcelona, protein-coupled receptor; ICV, intracerebroventricular; IP, Spain; orcid.org/0000-0001-9099-7235 intraperitoneal; IV, intravenous; KLH, keyhole limpet Estefanía Moreno − Department of Biochemistry and hemocyanin; MD, molecular dynamics; MDL, MDL Molecular Biomedicine, Institute of Biomedicine, University of 100,907; NSAID, nonsteroidal anti-inflammatory drug; r/e, Barcelona, 08028 Barcelona, Spain; orcid.org/0000- retro-enantio; 5HT2AR, serotonin (5-hydroxytryptamine, 5- 0002-2491-5753 HT) receptor 2A; THC, tetrahydrocannabinol; TM, trans- Sira Defaus − Department of Experimental and Health membrane; WIN, WIN 55,212-2; YFP, yellow fluorescent protein ■ Sciences, Universitat Pompeu Fabra, 08003 Barcelona, Spain; orcid.org/0000-0002-3106-458X Antonio Ortega-Alvaro − Department of Experimental and REFERENCES Health Sciences, Universitat Pompeu Fabra, 08003 (1) Dahlhamer, J.; Lucas, J.; Zelaya, C.; Nahin, R.; Mackey, S.; Barcelona, Spain DeBar, L.; Kerns, R.; Von Korff, M.; Porter, L.; Helmick, C. Angel Gonzalez − Laboratori de Medicina Computacional, Prevalence of Chronic Pain and High-Impact Chronic Pain Among Unitat de Bioestadística, Facultat de Medicina, Universitat Adults - United States, 2016. Morb. Mortal. Wkly. Rep. 2018, 67, 1001−1006. Autònoma de Barcelona, 08193 Bellaterra, Spain (2) Scholl, L.; Seth, P.; Kariisa, M.; Wilson, N.; Baldwin, G. Drug Patricia Robledo − Integrative Pharmacology and Systems and Opioid-Involved Overdose DeathsUnited States, 2013-2017. Neuroscience, IMIM-Hospital del Mar Research Institute, Morb. Mortal. Wkly. Rep. 2018, 67, 1419−1427. 08003 Barcelona, Spain (3) Kaye, A. D.; Jones, M. R.; Kaye, A. M.; Ripoll, J. G.; Galan, V.; Marco Cavaco − Instituto de Medicina Molecular, Faculdade Beakley, B. D.; Calixto, F.; Bolden, J. L.; Urman, R. D.; Manchikanti, de Medicina, Universidade de Lisboa, 1649-028 Lisboa, L. Prescription Opioid Abuse in Chronic Pain: An Updated Review of Portugal; orcid.org/0000-0002-0938-9038 Opioid Abuse Predictors and Strategies to Curb Opioid Abuse: Part Vera Neves − Instituto de Medicina Molecular, Faculdade de 1. Pain Physician 2017, 2, S93−S109. Medicina, Universidade de Lisboa, 1649-028 Lisboa, (4) Harirforoosh, S.; Asghar, W.; Jamali, F. Adverse Effects of Portugal; orcid.org/0000-0002-2989-7208 Nonsteroidal Antiinflammatory Drugs: An Update of Gastrointestinal, Cardiovascular and Renal Complications. J. Pharm. Pharm. Sci. 2013, Miguel A. R. B. Castanho − Instituto de Medicina Molecular, 16, 821−847. Faculdade de Medicina, Universidade de Lisboa, 1649-028 (5) Carter, G. T.; Duong, V.; Ho, S.; Ngo, K. C.; Greer, C. L.; Lisboa, Portugal; orcid.org/0000-0001-7891-7562 Weeks, D. L. Side Effects of Commonly Prescribed Analgesic Vicent Casadó − Department of Biochemistry and Molecular Medications. Phys. Med. Rehabil. Clin. N. Am. 2014, 25, 457−470. Biomedicine, Institute of Biomedicine, University of (6) Walker, J. M.; Hohmann, A. G. Cannabinoid Mechanisms of Barcelona, 08028 Barcelona, Spain; orcid.org/0000- Pain Suppression. Handb. Exp. Pharmacol. 2005, 168, 509−554. 0002-1764-3825 (7) Valjent, E.; Mitchell, J. M.; Besson, M.-J.; Caboche, J.; Leonardo Pardo − Laboratori de Medicina Computacional, Maldonado, R. Behavioural and biochemical evidence for interactions Unitat de Bioestadística, Facultat de Medicina, Universitat between Δ9-tetrahydrocannabinol and nicotine. Br. J. Pharmacol. Autònoma de Barcelona, 08193 Bellaterra, Spain 2002, 135, 564−578. (8) de Fonseca, F. R.; Rubio, P.; Menzaghi, F.; Merlo-Pich, E.; Complete contact information is available at: Rivier, J.; Koob, G. F.; Navarro, M. Corticotropin-Releasing Factor https://pubs.acs.org/10.1021/acs.jmedchem.1c00484 (CRF) Antagonist [D-Phe12,Nle21,38,C Alpha MeLeu37]CRF 6946 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article Attenuates the Acute Actions of the Highly Potent Cannabinoid (25) Neves, V.; Aires-da-Silva, F.; Morais, M.; Gano, L.; Ribeiro, E.; Receptor Agonist HU-210 on Defensive-Withdrawal Behavior in Rats. Pinto, A.; Aguiar, S.; Gaspar, D.; Fernandes, C.; Correia, J. D. G.; J. Pharmacol. Exp. Ther. 1996, 276, 56−64. Castanho, M. A. R. B. Novel Peptides Derived from Dengue Virus (9) Kandasamy, R.; Dawson, C. T.; Craft, R. M.; Morgan, M. M. Capsid Protein Translocate Reversibly the Blood-Brain Barrier Anti-Migraine Effect of (9)-Tetrahydrocannabinol in the Female Rat. through a Receptor-Free Mechanism. ACS Chem. Biol. 2017, 12, Eur. J. Pharmacol. 2018, 818, 271−277. 1257−1268. (10) Baron, E. P.; Lucas, P.; Eades, J.; Hogue, O. Patterns of (26) Jastrzebska, B.; Chen, Y.; Orban, T.; Jin, H.; Hofmann, L.; Medicinal Cannabis Use, Strain Analysis, and Substitution Effect Palczewski, K. Disruption of Rhodopsin Dimerization with Synthetic among Patients with Migraine, Headache, Arthritis, and Chronic Pain Peptides Targeting an Interaction Interface. J. Biol. Chem. 2015, 290, in a Medicinal Cannabis Cohort. J. Headache Pain 2018, 19, 37. 25728−25744. (11) La Porta, C.; Bura, S. A.; Llorente-Onaindia, J.; Pastor, A.; (27) Wang, Y. F.; Xu, X.; Fan, X.; Zhang, C.; Wei, Q.; Wang, X.; Navarrete, F.; García-Gutiérrez, M. S.; De la Torre, R.; Manzanares, Guo, W.; Xing, W.; Yu, J.; Yan, J.-L.; Liang, H.-P. A Cell-penetrating J.; Monfort, J.; Maldonado, R. Role of the Endocannabinoid System in Peptide Suppresses Inflammation by Inhibiting NF-κB Signaling. Mol. the Emotional Manifestations of Osteoarthritis Pain. Pain 2015, 156, Ther. 2011, 19, 1849−1857. 2001−2012. (28) Borroto-Escuela, D. O.; Romero-Fernandez, W.; Mudó, G.; (12) Maldonado, R.; Baños, J. E.; Cabañero, D. The Endocanna- Pérez-Alea, M.; Ciruela, F.; Tarakanov, A. O.; Narvaez, M.; Di binoid System and Neuropathic Pain. Pain 2016, 157, S23−S32. Liberto, V.; Agnati, L. F.; Belluardo, N.; Fuxe, K. Fibroblast Growth (13) Donovan, K. A.; Oberoi-Jassal, R.; Chang, Y. D.; Rajasekhara, S.; Haas, M. F.; Randich, A. L.; Portman, D. G. Cannabis Use in Factor Receptor 1- 5-Hydroxytryptamine 1A Heteroreceptor Young Adult Cancer Patients. J. Adolesc. Young Adult Oncol. 2020, 9, Complexes and Their Enhancement of Hippocampal Plasticity. Biol. 30−35. Psychiatry 2012, 71, 84−91. (14) Maldonado, R.; Berrendero, F.; Ozaita, A.; Robledo, P. (29) Najjar, K.; Erazo-Oliveras, A.; Brock, D. J.; Wang, T.-Y.; Pellois, Neurochemical Basis of Cannabis Addiction. Neuroscience 2011, J.-P. An L- to d-Amino Acid Conversion in an Endosomolytic Analog 181, 1−17. of the Cell-Penetrating Peptide TAT Influences Proteolytic Stability, (15) Moreira, F. A.; Lutz, B. The Endocannabinoid System: Endocytic Uptake, and Endosomal Escape. J. Biol. Chem. 2017, 292, Emotion, Learning and Addiction. Addict. Biol. 2008, 13, 196−212. 847−861. (16) Castellano, C.; Rossi-Arnaud, C.; Cestari, V.; Costanzi, M. (30) Cavaco, M.; Valle, J.; da Silva, R.; Correia, J. D. G.; Castanho, Cannabinoids and Memory; Animal Studies. Curr. Drug Targets: CNS M. A. R. B.; Andreu, D.; Neves, V. DPepH3, an Improved Peptide Neurol. Disord. 2003, 2, 389−402. Shuttle for Receptor-Independent Transport across the Blood-Brain (17) Puighermanal, E.; Busquets-Garcia, A.; Maldonado, R.; Ozaita, Barrier. Curr. Pharm. Des. 2020, 26, 1495−1506. A. Cellular and Intracellular Mechanisms Involved in the Cognitive (31) Cavaco, M.; Pérez-Peinado, C.; Valle, J.; Silva, R. D. M.; Impairment of Cannabinoids. Philos. Trans. R. Soc., B 2012, 367, Correia, J. D. G.; Andreu, D.; Castanho, M. A. R. B.; Neves, V. To 3254−3263. What Extent Do Fluorophores Bias the Biological Activity of (18) Viñals, X.; Moreno, E.; Lanfumey, L.; Cordomí, A.; Pastor, A.; Peptides? A Practical Approach Using Membrane-Active Peptides as de La Torre, R.; Gasperini, P.; Navarro, G.; Howell, L. A.; Pardo, L.; Models. Front. Bioeng. Biotechnol. 2020, 8, 552035. Lluís, C.; Canela, E. I.; McCormick, P. J.; Maldonado, R.; Robledo, P. (32) Merrifield, R. B.; Juvvadi, P.; Andreu, D.; Ubach, J.; Boman, A.; Cognitive Impairment Induced by Delta9-Tetrahydrocannabinol Boman, H. G. Retro and Retroenantio Analogs of Cecropin-Melittin Occurs through Heteromers between Cannabinoid CB1 and Hybrids. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 3449−3453. Serotonin 5-HT2A Receptors. PLoS Biol. 2015, 13, No. e1002194. (33) Guichard, G.; Benkirane, N.; Zeder-Lutz, G.; van Regenmortel, (19) Ferré, S.; Casadó, V.; Devi, L. A.; Filizola, M.; Jockers, R.; M. H.; Briand, J. P.; Muller, S. Antigenic Mimicry of Natural L- Lohse, M. J.; Milligan, G.; Pin, J.-P.; Guitart, X. G Protein-Coupled Peptides with Retro-Inverso-Peptidomimetics. Proc. Natl. Acad. Sci. Receptor Oligomerization Revisited: Functional and Pharmacological U.S.A. 1994, 91, 9765−9769. Perspectives. Pharmacol. Rev. 2014, 66, 413−434. (34) Prades, R.; Oller-Salvia, B.; Schwarzmaier, S. M.; Selva, J.; (20) Schwarze, S. R.; Ho, A.; Vocero-Akbani, A.; Dowdy, S. F. In Moros, M.; Balbi, M.; Grazú, V.; de La Fuente, J. M.; Egea, G.; Vivo Protein Transduction: Delivery of a Biologically Active Protein Plesnila, N.; Teixidó, M.; Giralt, E. Applying the Retro-Enantio into the Mouse. Science 1999, 285, 1569−1572. Approach to Obtain a Peptide Capable of Overcoming the Blood- (21) Manglik, A.; Kruse, A. C.; Kobilka, T. S.; Thian, F. S.; Brain Barrier. Angew. Chem., Int. Ed. 2015, 54, 3967−3972. Mathiesen, J. M.; Sunahara, R. K.; Pardo, L.; Weis, W. I.; Kobilka, B. (35) Vaissière, A.; Aldrian, G.; Konate, K.; Lindberg, M. F.; Jourdan, K.; Granier, S. Crystal Structure of the Μ-Opioid Receptor Bound to C.; Telmar, A.; Seisel, Q.; Fernandez, F.; Viguier, V.; Genevois, C.; a Morphinan Antagonist. Nature 2012, 485, 321−326. Couillaud, F.; Boisguerin, P.; Deshayes, S. A Retro-Inverso Cell- (22) Ferré, S.; Ciruela, F.; Casadó, V.; Pardo, L. Oligomerization of G protein-coupled receptors: Still doubted?. In Oligomerization in Penetrating Peptide for SiRNA Delivery. J. Nanobiotechnol. 2017, 15, Health and Disease: From Enzymes to G Protein-Coupled Receptors; 34. Giraldo, J., Ciruela, F., Eds.; Academic Press, 2020; Vol. 169, pp 297− (36) Almiñana, N.; Alsina, M. A.; Ortiz, A.; Reig, F. Comparative 321. Physicochemical Study of SIKVAV Peptide and Its Retro and Retro- (23) Guinart, D.; Moreno, E.; Galindo, L.; Cuenca-Royo, A.; Enantio Analogues. Colloids Surf., A 2004, 249, 19−24. Barrera-Conde, M.; Pérez, E. J.; Fernández-Avilés, C.; Correll, C. U.; (37) Howlett, A. C.; Bidaut-Russell, M.; Devane, W. A.; Melvin, L. Canela, E. I.; Casadó, V.; Cordomi, A.; Pardo, L.; de la Torre, R.; S.; Johnson, M. R.; Herkenham, M. The Cannabinoid Receptor: Pérez, V.; Robledo, P. Altered Signaling in CB1R-5-HT2AR Biochemical, Anatomical and Behavioral Characterization. Trends Heteromers in Olfactory Neuroepithelium Cells of Schizophrenia Neurosci. 1990, 13, 420−423. Patients Is Modulated by Cannabis Use. Schizophr. Bull. 2020, 46, (38) Pertwee, R. G. Cannabinoid Pharmacology: The First 66 Years. 1547−1557. Br. J. Pharmacol. 2006, 147, S163−S171. (24) Galindo, L.; Moreno, E.; López-Armenta, F.; Guinart, D.; (39) Gaoni, Y.; Mechoulam, R. Isolation, Structure, and Partial Cuenca-Royo, A.; Izquierdo-Serra, M.; Xicota, L.; Fernandez, C.; Synthesis of an Active Constituent of Hashish. J. Am. Chem. Soc. 1964, Menoyo, E.; Fernández-Fernández, J. M.; Benítez-King, G.; Canela, E. 86, 1646−1647. I.; Casadó, V.; Pérez, V.; de la Torre, R.; Robledo, P. Cannabis Users (40) Manzanares, J.; Julian, M.; Carrascosa, A. Role of the Show Enhanced Expression of CB1-5HT2A Receptor Heteromers in Cannabinoid System in Pain Control and Therapeutic Implications Olfactory Neuroepithelium Cells. Mol. Neurobiol. 2018, 55, 6347− for the Management of Acute and Chronic Pain Episodes. Curr. 6361. Neuropharmacol. 2006, 4, 239−257. 6947 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948 Journal of Medicinal Chemistry pubs.acs.org/jmc Article (41) Matsuda, L. A.; Lolait, S. J.; Brownstein, M. J.; Young, A. C.; (59) Wacker, D.; Wang, C.; Katritch, V.; Han, G. W.; Huang, X.-P.; Bonner, T. I. Structure of a Cannabinoid Receptor and Functional Vardy, E.; McCorvy, J. D.; Jiang, Y.; Chu, M.; Siu, F. Y.; Liu, W.; Xu, Expression of the Cloned CDNA. Nature 1990, 346, 561−564. H. E.; Cherezov, V.; Roth, B. L.; Stevens, R. C. Structural Features for (42) Hudson, B. D.; Hébert, T. E.; Kelly, M. E. Physical and Functional Selectivity at Serotonin Receptors. Science 2013, 340, Functional Interaction between CB1 Cannabinoid Receptors and 615−619. Beta2-Adrenoceptors. Br. J. Pharmacol. 2010, 160, 627−642. (60) Gallo, M.; Navarro, G.; Franco, R.; Andreu, D. A2A Receptor (43) Bagher, A. M.; Laprairie, R. B.; Toguri, J. T.; Kelly, M. E. M.; Homodimer-Disrupting Sequence Efficiently Delivered by a Protease- Resistant, Cyclic CPP Vector. Int. J. Mol. Sci. 2019, 20, 4937. ■ Denovan-Wright, E. M. Bidirectional Allosteric Interactions between Cannabinoid Receptor 1 (CB1) and Dopamine Receptor 2 Long (D2L) Heterotetramers. Eur. J. Pharmacol. 2017, 813, 66−83. NOTE ADDED AFTER ASAP PUBLICATION (44) Rios, C.; Gomes, I.; Devi, L. A. μ Opioid and CB1 Cannabinoid This paper was published on the Web on April 23, 2021. Receptor Interactions: Reciprocal Inhibition of Receptor Signaling Additional corrections were received post ASAP, and the and Neuritogenesis. Br. J. Pharmacol. 2006, 148, 387−395. corrected version was reposted on April 23, 2021. (45) Flores, A.; Maldonado, R.; Berrendero, F. Cannabinoid- Hypocretin Cross-Talk in the Central Nervous System: What We Know so Far. Front. Neurosci. 2013, 7, 256. (46) Mackie, K. Cannabinoid Receptor Homo- and Heterodimeriza- tion. Life Sci. 2005, 77, 1667−1673. (47) Navarro, G.; Carriba, P.; Gandí, J.; Ciruela, F.; Casadó, V.; Cortés, A.; Mallol, J.; Canela, E. I.; Lluis, C.; Franco, R.; Gandí, J.; Ciruela, F.; Casadó, V.; Cortés, A.; Mallol, J.; Canela, E. I.; Lluis, C.; Franco, R. Detection of Heteromers Formed by Cannabinoid CB1, Dopamine D2, and Adenosine A2AG-Protein-Coupled Receptors by Combining Bimolecular Fluorescence Complementation and Bio- luminescence Energy Transfer. Sci. World J. 2008, 8, 1088−1097. (48) Köfalvi, A.; Moreno, E.; Cordomí, A.; Cai, N.-S.; Fernández- Dueñas, V.; Ferreira, S. G.; Guixà-González, R.; Sánchez-Soto, M.; Yano, H.; Casadó-Anguera, V.; Cunha, R. A.; Sebastião, A. M.; Ciruela, F.; Pardo, L.; Casadó, V.; Ferré, S. Control of Glutamate Release by Complexes of Adenosine and Cannabinoid Receptors. BMC Biol. 2020, 18, 9. (49) Casadó-Anguera, V.; Cortés, A.; Casadó, V.; Moreno, E. Targeting the Receptor-Based Interactome of the Dopamine D1 Receptor: Looking for Heteromer-Selective Drugs. Expert Opin. Drug Discovery 2019, 14, 1297−1312. (50) Gomes, I.; Ayoub, M. A.; Fujita, W.; Jaeger, W. C.; Pfleger, K. D. G.; Devi, L. A. G Protein-Coupled Receptor Heteromers. Annu. Rev. Pharmacol. Toxicol. 2016, 56, 403−425. (51) Hauser, A. S.; Attwood, M. M.; Rask-Andersen, M.; Schiöth, H. B.; Gloriam, D. E. Trends in GPCR Drug Discovery: New Agents, Targets and Indications. Nat. Rev. Drug Discovery 2017, 16, 829−842. (52) Wacker, D.; Stevens, R. C.; Roth, B. L. How Ligands Illuminate GPCR Molecular Pharmacology. Cell 2017, 170, 414−427. (53) Pérez-Benito, L.; Henry, A.; Matsoukas, M.-T.; Lopez, L.; Pulido, D.; Royo, M.; Cordomí, A.; Tresadern, G.; Pardo, L. The Size Matters? A Computational Tool to Design Bivalent Ligands. Bioinformatics 2018, 34, 3857−3863. (54) Pulido, D.; Casadó-Anguera, V.; Pérez-Benito, L.; Moreno, E.; Cordomí, A.; López, L.; Cortés, A.; Ferré, S.; Pardo, L.; Casadó, V.; Royo, M. Design of a True Bivalent Ligand with Picomolar Binding Affinity for a G Protein-Coupled Receptor Homodimer. J. Med. Chem. 2018, 61, 9335−9346. (55) Pérez-Benito, L.; Llinas Del Torrent, C.; Pardo, L.; Tresadern, G. The Computational Modeling of Allosteric Modulation of Metabotropic Glutamate Receptors. Adv. Pharmacol. 2020, 88, 1−33. (56) Shao, Z.; Yan, W.; Chapman, K.; Ramesh, K.; Ferrell, A. J.; Yin, J.; Wang, X.; Xu, Q.; Rosenbaum, D. M. Structure of an Allosteric Modulator Bound to the CB1 Cannabinoid Receptor. Nat. Chem. Biol. 2019, 15, 1199−1205. (57) Filizola, M. Increasingly Accurate Dynamic Molecular Models of G-Protein Coupled Receptor Oligomers: Panacea or Pandora’s Box for Novel Drug Discovery? Life Sci. 2010, 86, 590−597. (58) Hua, T.; Vemuri, K.; Pu, M.; Qu, L.; Han, G. W.; Wu, Y.; Zhao, S.; Shui, W.; Li, S.; Korde, A.; Laprairie, R. B.; Stahl, E. L.; Ho, J.-H.; Zvonok, N.; Zhou, H.; Kufareva, I.; Wu, B.; Zhao, Q.; Hanson, M. A.; Bohn, L. M.; Makriyannis, A.; Stevens, R. C.; Liu, Z.-J. Crystal Structure of the Human Cannabinoid Receptor CB1. Cell 2016, 167, 750−762.e14. 6948 https://doi.org/10.1021/acs.jmedchem.1c00484 J. Med. Chem. 2021, 64, 6937−6948