Structure–Activity Relationships of 7-Substituted Dimethyltyrosine-Tetrahydroisoquinoline Opioid Peptidomimetics

The opioid receptors modulate a variety of biological functions, including pain, mood, and reward. As a result, opioid ligands are being explored as potential therapeutics for a variety of indications. Multifunctional opioid ligands, which act simultaneously at more than one type of opioid receptor, show promise for use in the treatment of addiction, pain, and other conditions. Previously, we reported the creation of bifunctional kappa opioid receptor (KOR) agonist/mu opioid receptor (MOR) partial agonist ligands from the classically delta opioid receptor (DOR) antagonist selective dimethyltyrosine-tetrahydroisoquinoline (Dmt-Tiq) scaffold through the addition of a 7-benzyl pendant on the tetrahydroisoquinoline ring. This study further explores the structure–activity relationships surrounding 7-position pendants on the Dmt-Tiq scaffold. Some analogues maintain a KOR agonist/MOR partial agonist profile, which is being explored in the development of a treatment for cocaine addiction. Others display a MOR agonist/DOR antagonist profile, which has potential to be used in the creation of a less addictive pain medication. Ultimately, we report the synthesis and in vitro evaluation of novel opioid ligands with a variety of multifunctional profiles.


Introduction
Opioids have one of the longest known histories of any drug class. The use of opium for ritual, medicinal, and/or recreational purposes dates back to ancient civilizations [1,2]. In the early 1800s, Friedrich Sertürner isolated the primary active ingredient of opium and named it morphine [1]. This began the chemical exploration of the opiates, and ultimately, led to discovery of the opioid receptors and their endogenous ligands. It is widely accepted that there are three major types of opioid receptors-the kappa opioid receptor (KOR), the mu opioid receptor (MOR), and the delta opioid receptor (DOR). These receptors have high sequence and structural homology, and they are all Class A GPCRs [3,4]. The structure and function of this type of receptor have been thoroughly reviewed [5,6]. Though most well-known for its role in regulating pain, the opioid system is also involved in many other biological processes, including mood [7-9] and reward [8,10]. As such, opioids remain an important and promising class of molecules for the development of therapeutics for a variety of indications.
Functions of the opioid receptors are modulated by both endogenous and exogenous opioid ligands. In the two centuries since the discovery of morphine, many semi-synthetic and synthetic opioids have been developed for this purpose. As the complex pharmacology of the opioid system continues to be revealed, it has been posited that unwanted effects and desired effects may result from the same interaction of an opioid agonist or antagonist with its target. As a result, the development of selective agents has declined, and the development of multifunctional ligands, compounds that act simultaneously at multiple opioid receptor types, has gained popularity as a strategy for the design of therapeutics [11,12]. The current state of multifunctional opioid ligands has recently been reviewed [13].
Our group [14] and others [15][16][17][18][19] have shown that the dimethyltyrosine-tetrahydroisoquinoline (Dmt-Tiq) scaffold can be used in the development of multifunctional opioid ligands ( Figure 1). This scaffold, originally developed as a selective DOR antagonist, has been extensively explored through synthesis of many analogues. However, the confines of traditional peptide synthesis have limited substitution on the tetrahydroisoquinoline (Tiq) ring. Nearly two decades ago, minor substitutions were reported at the 6-, 7-, and 8-positions, but all of these compounds displayed a DOR antagonist profile similar to that of the parent peptide [17,20]. Recently, we reported that installation of a 7-benzyl pendant on the Tiq could alter the profile of this series to KOR agonism/MOR partial agonism [14]. Functions of the opioid receptors are modulated by both endogenous and exogenous opioid ligands. In the two centuries since the discovery of morphine, many semi-synthetic and synthetic opioids have been developed for this purpose. As the complex pharmacology of the opioid system continues to be revealed, it has been posited that unwanted effects and desired effects may result from the same interaction of an opioid agonist or antagonist with its target. As a result, the development of selective agents has declined, and the development of multifunctional ligands, compounds that act simultaneously at multiple opioid receptor types, has gained popularity as a strategy for the design of therapeutics [11,12]. The current state of multifunctional opioid ligands has recently been reviewed [13].
Our group [14] and others [15][16][17][18][19] have shown that the dimethyltyrosine-tetrahydroisoquinoline (Dmt-Tiq) scaffold can be used in the development of multifunctional opioid ligands ( Figure 1). This scaffold, originally developed as a selective DOR antagonist, has been extensively explored through synthesis of many analogues. However, the confines of traditional peptide synthesis have limited substitution on the tetrahydroisoquinoline (Tiq) ring. Nearly two decades ago, minor substitutions were reported at the 6-, 7-, and 8-positions, but all of these compounds displayed a DOR antagonist profile similar to that of the parent peptide [17,20]. Recently, we reported that installation of a 7benzyl pendant on the Tiq could alter the profile of this series to KOR agonism/MOR partial agonism [14]. KOR agonists have shown potential for use in the treatment of cocaine addiction because of their reward-modulating properties. Specifically, administration of a KOR agonist can reduce cocaine selfadministration in non-human primates [21,22]. However, KOR agonism is also associated with dysphoria, an intense feeling of unease or dissatisfaction [7]. As a result, selective KOR agonists have limited therapeutic potential. It is well known that MOR agonism is associated with euphoria. Therefore, a bifunctional KOR agonist/MOR agonist offers a potential alternative to a selective KOR agonist that may result in a more favorable side effect profile. In fact, there is evidence to suggest that a KOR agonist/MOR agonist may be useful in the treatment of cocaine addiction [23,24].
The aim of this work was to explore structure-activity relationships around the 7-benzyl pendant which introduced KOR agonism to the Dmt-Tiq scaffold. Novel analogues reported here reveal that substitution on the benzyl ring can maintain a KOR agonist/MOR partial agonist profile while analogues with other 7-position pendants show varied results. Overall, this work demonstrates the development of novel Dmt-Tiq peptidomimetics that display a range of multifunctional opioid profiles.

Results
A series of novel Dmt-Tiq compounds with substitution at the Tiq 7-position were prepared and evaluated in vitro for opioid activity.

Synthesis
All compounds were prepared from commercial starting materials according to one of the synthetic routes shown in Scheme 1. In the first route, commercially available carboxylic acid 1 was KOR agonists have shown potential for use in the treatment of cocaine addiction because of their reward-modulating properties. Specifically, administration of a KOR agonist can reduce cocaine self-administration in non-human primates [21,22]. However, KOR agonism is also associated with dysphoria, an intense feeling of unease or dissatisfaction [7]. As a result, selective KOR agonists have limited therapeutic potential. It is well known that MOR agonism is associated with euphoria. Therefore, a bifunctional KOR agonist/MOR agonist offers a potential alternative to a selective KOR agonist that may result in a more favorable side effect profile. In fact, there is evidence to suggest that a KOR agonist/MOR agonist may be useful in the treatment of cocaine addiction [23,24].
The aim of this work was to explore structure-activity relationships around the 7-benzyl pendant which introduced KOR agonism to the Dmt-Tiq scaffold. Novel analogues reported here reveal that substitution on the benzyl ring can maintain a KOR agonist/MOR partial agonist profile while analogues with other 7-position pendants show varied results. Overall, this work demonstrates the development of novel Dmt-Tiq peptidomimetics that display a range of multifunctional opioid profiles.

Results
A series of novel Dmt-Tiq compounds with substitution at the Tiq 7-position were prepared and evaluated in vitro for opioid activity.

Synthesis
All compounds were prepared from commercial starting materials according to one of the synthetic routes shown in Scheme 1. In the first route, commercially available carboxylic acid 1 was reduced to the corresponding secondary alcohol 2 using borane dimethylsulfide. An Appel reaction was performed to convert alcohol 2 to benzyl bromide 3. The pendant was then attached via Suzuki coupling of intermediate 3 with the corresponding boronic acid or S N 2 reaction with the corresponding nucleophile.

Pharmacological Evaluation
Each novel compound was evaluated for binding to and stimulation of KOR, MOR, and DOR. Binding affinity (K i ) was determined by competitive displacement of [ 3 H]-diprenorphine, a non-selective opioid receptor antagonist with similar affinity for each of the three receptors. Potency (EC 50 ) and efficacy, expressed as percent stimulation compared to a standard agonist at each receptor, were determined by a [ 35 S]-GTPγS binding assay.
Building on our previous work [14], several ortho and meta substituents on the benzyl ring were investigated as well as the o-,m-dimethyl analogue. The results of the pharmacological evaluation of these compounds are shown in Table 1. Data for the previously reported 7-benzyl analogue 4c is shown for comparison. Previously, this compound was evaluated at human KOR, rat MOR, and rat DOR. The profile shown here differs slightly from that previously reported because all compounds in this study were evaluated only at human receptors. All ortho-and meta-substituted analogues reported here display single digit nanomolar or subnanomolar binding at all three opioid receptors. In general, ortho analogues show the highest affinity for KOR compared to the other receptors, while most meta analogues show the highest affinity for DOR. Compared to standard agonists, each of these analogues retains moderate (54%) to high (89%) efficacy at KOR and low (29%) to high efficacy (81%) at MOR. Most analogues show no DOR agonism, but the ortho trifluoromethyl analogue 8c shows low DOR efficacy and potency. Potency for these compounds remains primarily in the double or triple digit nanomolar range. The balance of potencies varies for ortho analogues, while meta analogues and the disubstituted analogue are consistently more potent at MOR than KOR. (81%) at MOR. Most analogues show no DOR agonism, but the ortho trifluoromethyl analogue 8c shows low DOR efficacy and potency. Potency for these compounds remains primarily in the double or triple digit nanomolar range. The balance of potencies varies for ortho analogues, while meta analogues and the disubstituted analogue are consistently more potent at MOR than KOR. (81%) at MOR. Most analogues show no DOR agonism, but the ortho trifluoromethyl analogue 8c shows low DOR efficacy and potency. Potency for these compounds remains primarily in the double or triple digit nanomolar range. The balance of potencies varies for ortho analogues, while meta analogues and the disubstituted analogue are consistently more potent at MOR than KOR. shows low DOR efficacy and potency. Potency for these compounds remains primarily in the double or triple digit nanomolar range. The balance of potencies varies for ortho analogues, while meta analogues and the disubstituted analogue are consistently more potent at MOR than KOR. or triple digit nanomolar range. The balance of potencies varies for ortho analogues, while meta analogues and the disubstituted analogue are consistently more potent at MOR than KOR. analogues and the disubstituted analogue are consistently more potent at MOR than KOR.   Next, we explored the incorporation of nitrogen into the aromatic ring of the pendant. In place of the benzyl pendant, 3-and 4-pyridyl pendants were added at the 7-position of the tetrahydroisoquinoline ring with a methylene spacer ( Table 2). Due to well-known synthetic difficulties [25], the 2-pyridyl analogue was not successfully synthesized. Evaluation of the pyridyl analogues revealed a loss in binding and a drastic loss of potency at KOR with low efficacy at MOR and no agonism at DOR, compared to their carbocyclic counterparts.
Non-aromatic pendants were also explored. Table 3 shows pharmacological data for analogues with saturated, cyclic amine pendants. For this series, all KOR agonism was lost, and only weak potency and low efficacy at MOR were observed. Next, we explored the incorporation of nitrogen into the aromatic ring of the pendant. In place of the benzyl pendant, 3-and 4-pyridyl pendants were added at the 7-position of the tetrahydroisoquinoline ring with a methylene spacer ( Table 2). Due to well-known synthetic difficulties [25], the 2-pyridyl analogue was not successfully synthesized. Evaluation of the pyridyl analogues revealed a loss in binding and a drastic loss of potency at KOR with low efficacy at MOR and no agonism at DOR, compared to their carbocyclic counterparts. Non-aromatic pendants were also explored. Table 3 shows pharmacological data for analogues with saturated, cyclic amine pendants. For this series, all KOR agonism was lost, and only weak potency and low efficacy at MOR were observed. of the benzyl pendant, 3-and 4-pyridyl pendants were added at the 7-position of the tetrahydroisoquinoline ring with a methylene spacer ( Table 2). Due to well-known synthetic difficulties [25], the 2-pyridyl analogue was not successfully synthesized. Evaluation of the pyridyl analogues revealed a loss in binding and a drastic loss of potency at KOR with low efficacy at MOR and no agonism at DOR, compared to their carbocyclic counterparts. Non-aromatic pendants were also explored. Table 3 shows pharmacological data for analogues with saturated, cyclic amine pendants. For this series, all KOR agonism was lost, and only weak potency and low efficacy at MOR were observed.  Non-aromatic pendants were also explored. Table 3 shows pharmacological data for analogues with saturated, cyclic amine pendants. For this series, all KOR agonism was lost, and only weak potency and low efficacy at MOR were observed. Non-aromatic pendants were also explored. Table 3 shows pharmacological data for analogues with saturated, cyclic amine pendants. For this series, all KOR agonism was lost, and only weak potency and low efficacy at MOR were observed. Non-aromatic pendants were also explored. Table 3 shows pharmacological data for analogues with saturated, cyclic amine pendants. For this series, all KOR agonism was lost, and only weak potency and low efficacy at MOR were observed. To test whether opioid activity could be maintained in the presence of larger pendants, 1-and 2-naphthyl analogues were synthesized. These results are shown in Table 4. The 1-naphthyl pendant displays a KOR agonist/MOR partial agonist profile, while the 2-naphthyl pendant results in a drastic loss in KOR binding and a complete loss of KOR agonism.
Given the MOR/KOR profile of analogue 8o, a nitrogen scan was conducted to further explore the structure-activity relationships around this pendant (Table 5). Similar to the pyridyl analogues, synthetic difficulties prevented the synthesis and evaluation of the 1-isoquinolinyl analogue. With few exceptions, single digit nanomolar or stronger binding is observed at all three receptors, and these analogues favor binding to DOR over MOR and KOR. All of these compounds show MOR agonism, and all except analogue 8u display partial to full KOR agonism. However, only compounds 8q and 8v shows DOR agonism. Ultimately, the addition of a single nitrogen to this ring results in compounds that show a range of multifunctional opioid profiles.
Finally, bicyclic pendants with one saturated ring and one aromatic ring were explored. This subset of analogues displays two distinct profiles (Table 6). Compounds 8x and 8y show balanced affinity and efficacy at KOR and MOR, while compounds 8w and 8z display a loss in KOR binding and no KOR agonism. However, the latter two compounds show potent, moderate to high efficacy at MOR and strong binding but no agonism at DOR. To test whether opioid activity could be maintained in the presence of larger pendants, 1-and 2-naphthyl analogues were synthesized. These results are shown in Table 4. The 1-naphthyl pendant displays a KOR agonist/MOR partial agonist profile, while the 2-naphthyl pendant results in a drastic loss in KOR binding and a complete loss of KOR agonism. Given the MOR/KOR profile of analogue 8o, a nitrogen scan was conducted to further explore the structure-activity relationships around this pendant (Table 5). Similar to the pyridyl analogues, synthetic difficulties prevented the synthesis and evaluation of the 1-isoquinolinyl analogue. With few exceptions, single digit nanomolar or stronger binding is observed at all three receptors, and these analogues favor binding to DOR over MOR and KOR. All of these compounds show MOR agonism, and all except analogue 8u display partial to full KOR agonism. However, only compounds 8q and 8v shows DOR agonism. Ultimately, the addition of a single nitrogen to this ring results in compounds that show a range of multifunctional opioid profiles. 4 To test whether opioid activity could be maintained in the presence of larger pendants, 1-and 2-naphthyl analogues were synthesized. These results are shown in Table 4. The 1-naphthyl pendant displays a KOR agonist/MOR partial agonist profile, while the 2-naphthyl pendant results in a drastic loss in KOR binding and a complete loss of KOR agonism. Given the MOR/KOR profile of analogue 8o, a nitrogen scan was conducted to further explore the structure-activity relationships around this pendant (Table 5). Similar to the pyridyl analogues, synthetic difficulties prevented the synthesis and evaluation of the 1-isoquinolinyl analogue. With few exceptions, single digit nanomolar or stronger binding is observed at all three receptors, and these analogues favor binding to DOR over MOR and KOR. All of these compounds show MOR agonism, and all except analogue 8u display partial to full KOR agonism. However, only compounds 8q and 8v shows DOR agonism. Ultimately, the addition of a single nitrogen to this ring results in compounds that show a range of multifunctional opioid profiles.       Finally, bicyclic pendants with one saturated ring and one aromatic ring were explored. This subset of analogues displays two distinct profiles (  Finally, bicyclic pendants with one saturated ring and one aromatic ring were explored. This subset of analogues displays two distinct profiles (Table 6). Compounds 8x and 8y show balanced affinity and efficacy at KOR and MOR, while compounds 8w and 8z display a loss in KOR binding

Discussion
Previous work by our group [14] and others [15][16][17][18][19] has indicated that the classically DOR antagonist selective dimethyltyrosine-tetrahydroisoquinoline (Dmt-Tiq) scaffold can be used as a starting point for the development of multifunctional opioid ligands. Building on our previous work, this study further explores installation of a 7-position pendant on the tetrahydroisoquinoline ring as a means of developing ligands with pharmacologically useful, multifunctional profiles. Previously, we reported that introduction of ortho and meta substituents onto a 7-benzyl pendant could produce ligands that demonstrate KOR agonism and MOR partial agonism [14], a bifunctional profile which has shown promise for the treatment of addiction to cocaine and other drugs of abuse. Here, we further explore the structure-activity relationships surrounding this novel series of opioid ligands and report compounds with this and other multifunctional opioid profiles.
Based on previously reported initial results from this series [14], we believed ortho and meta substituents on a 7-benzyl pendant to be promising structural modifications for the development of KOR agonist/MOR partial agonist ligands. A series of additional ortho and meta substitutions were evaluated to confirm whether they would exhibit the anticipated profile ( Table 1). As expected, ortho and meta substitutions on the 7-benzyl pendant are favorable for the development of KOR/MOR ligands. Because di-substitution (compound 8i) results in a notable drop in KOR potency, it shows no advantage over a single ortho or meta substituent. A few of the ligands in this series, including analogue 8c, show DOR agonism, which represents a problem for the development of a therapeutically useful KOR/MOR ligand because DOR agonism is associated with problematic side effects, including convulsions [9,26,27]. On the other hand, DOR antagonism may be beneficial for the development of a treatment for addiction since it has been shown to lower the addiction potential of MOR agonists in preclinical models [28][29][30]

Discussion
Previous work by our group [14] and others [15][16][17][18][19] has indicated that the classically DOR antagonist selective dimethyltyrosine-tetrahydroisoquinoline (Dmt-Tiq) scaffold can be used as a starting point for the development of multifunctional opioid ligands. Building on our previous work, this study further explores installation of a 7-position pendant on the tetrahydroisoquinoline ring as a means of developing ligands with pharmacologically useful, multifunctional profiles. Previously, we reported that introduction of ortho and meta substituents onto a 7-benzyl pendant could produce ligands that demonstrate KOR agonism and MOR partial agonism [14], a bifunctional profile which has shown promise for the treatment of addiction to cocaine and other drugs of abuse. Here, we further explore the structure-activity relationships surrounding this novel series of opioid ligands and report compounds with this and other multifunctional opioid profiles.
Based on previously reported initial results from this series [14], we believed ortho and meta substituents on a 7-benzyl pendant to be promising structural modifications for the development of KOR agonist/MOR partial agonist ligands. A series of additional ortho and meta substitutions were evaluated to confirm whether they would exhibit the anticipated profile ( Table 1). As expected, ortho and meta substitutions on the 7-benzyl pendant are favorable for the development of KOR/MOR ligands. Because di-substitution (compound 8i) results in a notable drop in KOR potency, it shows no advantage over a single ortho or meta substituent. A few of the ligands in this series, including analogue 8c, show DOR agonism, which represents a problem for the development of a therapeutically useful KOR/MOR ligand because DOR agonism is associated with problematic side effects, including convulsions [9,26,27]. On the other hand, DOR antagonism may be beneficial for the development of a treatment for addiction since it has been shown to lower the addiction potential of MOR agonists in preclinical models [28][29][30]. The strong MOR agonism of some compounds in this

Discussion
Previous work by our group [14] and others [15][16][17][18][19] has indicated that the classically DOR antagonist selective dimethyltyrosine-tetrahydroisoquinoline (Dmt-Tiq) scaffold can be used as a starting point for the development of multifunctional opioid ligands. Building on our previous work, this study further explores installation of a 7-position pendant on the tetrahydroisoquinoline ring as a means of developing ligands with pharmacologically useful, multifunctional profiles. Previously, we reported that introduction of ortho and meta substituents onto a 7-benzyl pendant could produce ligands that demonstrate KOR agonism and MOR partial agonism [14], a bifunctional profile which has shown promise for the treatment of addiction to cocaine and other drugs of abuse. Here, we further explore the structure-activity relationships surrounding this novel series of opioid ligands and report compounds with this and other multifunctional opioid profiles.
Based on previously reported initial results from this series [14], we believed ortho and meta substituents on a 7-benzyl pendant to be promising structural modifications for the development of KOR agonist/MOR partial agonist ligands. A series of additional ortho and meta substitutions were evaluated to confirm whether they would exhibit the anticipated profile ( Table 1). As expected, ortho and meta substitutions on the 7-benzyl pendant are favorable for the development of KOR/MOR ligands. Because di-substitution (compound 8i) results in a notable drop in KOR potency, it shows no advantage over a single ortho or meta substituent. A few of the ligands in this series, including analogue 8c, show DOR agonism, which represents a problem for the development of a therapeutically useful KOR/MOR ligand because DOR agonism is associated with problematic side effects, including convulsions [9,26,27]. On the other hand, DOR antagonism may be beneficial for the development of a treatment for addiction since it has been shown to lower the addiction potential of MOR agonists in preclinical models [28][29][30]

Discussion
Previous work by our group [14] and others [15][16][17][18][19] has indicated that the classically DOR antagonist selective dimethyltyrosine-tetrahydroisoquinoline (Dmt-Tiq) scaffold can be used as a starting point for the development of multifunctional opioid ligands. Building on our previous work, this study further explores installation of a 7-position pendant on the tetrahydroisoquinoline ring as a means of developing ligands with pharmacologically useful, multifunctional profiles. Previously, we reported that introduction of ortho and meta substituents onto a 7-benzyl pendant could produce ligands that demonstrate KOR agonism and MOR partial agonism [14], a bifunctional profile which has shown promise for the treatment of addiction to cocaine and other drugs of abuse. Here, we further explore the structure-activity relationships surrounding this novel series of opioid ligands and report compounds with this and other multifunctional opioid profiles.
Based on previously reported initial results from this series [14], we believed ortho and meta substituents on a 7-benzyl pendant to be promising structural modifications for the development of KOR agonist/MOR partial agonist ligands. A series of additional ortho and meta substitutions were evaluated to confirm whether they would exhibit the anticipated profile ( Table 1). As expected, ortho and meta substitutions on the 7-benzyl pendant are favorable for the development of KOR/MOR ligands. Because di-substitution (compound 8i) results in a notable drop in KOR potency, it shows no advantage over a single ortho or meta substituent. A few of the ligands in this series, including analogue 8c, show DOR agonism, which represents a problem for the development of a therapeutically useful KOR/MOR ligand because DOR agonism is associated with problematic side effects, including convulsions [9,26,27]. On the other hand, DOR antagonism may be beneficial for the development of a treatment for addiction since it has been shown to lower the addiction potential of MOR agonists in preclinical models [28][29][30]

Discussion
Previous work by our group [14] and others [15][16][17][18][19] has indicated that the classically DOR antagonist selective dimethyltyrosine-tetrahydroisoquinoline (Dmt-Tiq) scaffold can be used as a starting point for the development of multifunctional opioid ligands. Building on our previous work, this study further explores installation of a 7-position pendant on the tetrahydroisoquinoline ring as a means of developing ligands with pharmacologically useful, multifunctional profiles. Previously, we reported that introduction of ortho and meta substituents onto a 7-benzyl pendant could produce ligands that demonstrate KOR agonism and MOR partial agonism [14], a bifunctional profile which has shown promise for the treatment of addiction to cocaine and other drugs of abuse. Here, we further explore the structure-activity relationships surrounding this novel series of opioid ligands and report compounds with this and other multifunctional opioid profiles.
Based on previously reported initial results from this series [14], we believed ortho and meta substituents on a 7-benzyl pendant to be promising structural modifications for the development of KOR agonist/MOR partial agonist ligands. A series of additional ortho and meta substitutions were evaluated to confirm whether they would exhibit the anticipated profile ( Table 1). As expected, ortho and meta substitutions on the 7-benzyl pendant are favorable for the development of KOR/MOR ligands. Because di-substitution (compound 8i) results in a notable drop in KOR potency, it shows no advantage over a single ortho or meta substituent. A few of the ligands in this series, including analogue 8c, show DOR agonism, which represents a problem for the development of a therapeutically useful KOR/MOR ligand because DOR agonism is associated with problematic side effects, including convulsions [9,26,27]. On the other hand, DOR antagonism may be beneficial for the development of a treatment for addiction since it has been shown to lower the addiction potential of MOR agonists in preclinical models [28][29][30]. The strong MOR agonism of some compounds in this series (compounds 8c and 8f) is also a concern for the development of a therapy, as this activity would likely impart greater abuse potential. The most promising compound in this series for the development of a KOR agonist/MOR partial agonist for treatment of cocaine addiction, analogue 8b, shows high potency and efficacy at KOR, high potency and low efficacy at MOR, and is devoid of DOR agonism. This compound also has higher affinity for KOR and MOR than for DOR (eight-fold and two-fold, respectively), making it a promising candidate for further evaluation.
The introduction of a nitrogen to the 7-benzyl ring was not favorable for the development of a KOR/MOR ligand (Table 2). Rather, these analogues are selective for DOR over KOR and MOR and display low potency and efficacy at KOR and MOR. Replacement of the benzyl pendant with a saturated, cyclic amine pendant likewise decreases MOR potency drastically and eliminates KOR agonism altogether (Table 3). Unlike many of the compounds reported here, these analogues do not show particularly useful opioid profiles.
To explore the potential of installing larger pendants at the Tiq 7-position, we first synthesized analogues with 1-and 2-naphthyl pendants ( Table 4). Though the high clogP (5.6) and associated insolubility of these compounds is a problem for the ultimate development of a therapeutic, they were prepared as useful probes to further explore what might be tolerated in this series. Based on our previous observations from ortho, meta, and para substitutions [14], we hypothesized that the 1-naphthyl pendant would be favorable for the development of a KOR agonist while the 2-naphthyl pendant would not. The 1-naphthyl pendant points in the same direction as ortho and meta substituents, where there is room for additional steric bulk to be accommodated in the active configuration of the KOR orthosteric site. The 2-naphthyl pendant, on the other hand, points in the direction of meta and para substituents, where it clashes with the receptor. As expected, the 1-naphthyl analogue 8o shows the desired KOR/MOR profile, while the 2-naphthyl analogue 8p shows a drastic decrease in KOR binding and a complete loss of KOR agonism.
Based on these findings, we conducted a nitrogen scan of the 1-naphthyl pendant. The introduction of a single nitrogen drops the clogP by approximately 1.5 units, making these analogues much more promising candidates for use in animal studies and clinical settings. The profile of these analogues differed based on the placement of the nitrogen. Most of these compounds (analogues 8q-8v) show some degree of KOR agonism and MOR agonism, but only analogue 8t displays the desired KOR agonist/MOR partial agonist profile. Notably, this compound is equipotent at KOR and MOR and is a promising candidate for further study. It is approximately 3-fold selective for KOR and DOR over MOR which may lower the abuse potential of such a compound. Analogue 8u displays partial agonism only at MOR but high affinity for DOR, a profile most similar to that of previously reported Dmt-Tiq compounds. On the other hand, compound 8v has a potent KOR agonist/MOR agonist/DOR partial agonist profile and is weakly selective for KOR and DOR over MOR. While interesting, this profile is likely clinically irrelevant.
Finally, we explored bicyclic pendants with a cyclic amine attached to an aromatic ring. The profile differs, likely due to the placement of the second ring within the receptor binding site. As expected, those that would most closely mimic the 1-naphthyl pendant, compounds 8w and 8z, exhibit a KOR agonist/MOR agonist profile. The binding and efficacy profile at KOR and MOR for these two compounds is remarkably balanced, though they are more potent at MOR (five-fold and two-fold, respectively). As discussed above, the higher MOR efficacy and potency of these compounds compared to others would likely impart greater addiction potential. On the other hand, those compounds which more closely mimic the 2-naphthyl pendant, analogues 8w and 8z, show no KOR agonism, as expected. However, these analogues exhibit MOR agonism and DOR antagonism, a bifunctional profile being explored in the development of a less addictive treatment for pain [13,31,32]. Both compounds display potent MOR agonism and selectivity for MOR and DOR over KOR (18-fold and 43-fold, respectively).
In addition, compound 8z shows balanced affinity at MOR and DOR, a quality previously explored by our group as a way to mitigate addiction potential [33]. These compounds represent a starting point for further study for the development of a MOR agonist/DOR antagonist.
In conclusion, this work reports novel opioid ligands with a variety of multifunctional profiles. We have further elucidated structure-activity relationships surrounding the 7-position pendant on the tetrahydroisoquinoline ring of the classically DOR antagonist selective Dmt-Tiq scaffold. Ortho substituted analogues and select bicyclic pendant analogues show promise for the development of a KOR agonist/MOR partial agonist, a profile being investigated for the treatment of cocaine addiction. Compounds 8w and 8z exhibit a balanced MOR agonist/DOR antagonist profile and have potential to be investigated as a treatment for pain with lowered addiction potential. Future work will examine the pharmacokinetic properties of these compounds and explore the in vivo activity of interesting compounds from this series.

Chemistry
The chemical methods used were the same as those previously described [14] with any changes noted below. Unless otherwise noted, all reagents and solvents were purchased from commercial sources and used without additional purification. DiBoc-DMT was prepared from commercially available DMT according to standard procedures as previously reported [14]. Microwave reactions were performed in a Discover SP microwave synthesizer (CEM Corp., Matthews, NC, USA) in a closed vessel with maximum power input of 300 W. Column chromatography was carried out on silica gel cartridges using an Isolera One flash purification system (Biotage AB, Uppsala, Sweden) with a linear gradient of 100% hexanes to 100% ethyl acetate. Before chromatographic purification, crude reaction mixtures were analyzed by thin layer chromatography in hexanes/ethyl acetate. Purification of final compounds was performed using asemipreparative HPLC (Waters Technologies Corp., Milford, MA, USA) with a Vydac protein and peptide C18 reverse phase column using a linear gradient of 100% solvent A (water with 0.1% TFA) to 100% solvent B (acetonitrile with 0.1% TFA) at a rate of 1% per minute with UV absorbance monitored at 230 nm. Purity of final compounds was determined on an Alliance 2690 analytical HPLC (Waters Technologies Corp., Milford, MA, USA) with a Vydac protein and peptide C18 reverse phase column using the same gradient with UV absorbance monitored at 230 nm. Purity of final compounds used for testing was ≥95% as determined by HPLC. 1 H-NMR data for intermediates and final compounds in CDCl 3 or CD 3 OD was obtained on a 400 MHz or 500 MHz Varian spectrometer (Agilent Technologies Inc., Santa Clara, CA, USA). EIMS data was obtained using an Agilent 6130 HPLC-MS (Agilent Technologies Inc., Santa Clara, CA, USA) in positive ion mode. HREIMS data was obtained using an Agilent QTOF HPLC-MS (Agilent Technologies Inc., Santa Clara, CA, USA) in positive ion mode.

General Procedure A for Microwave Suzuki Coupling of Benzyl Bromide 3 and Pendant Boronic Acid
Benzyl bromide 3 (1.0 eq), the appropriate boronic acid (1.5 eq), Pd(dppf)Cl 2 (0.1 eq), and K 2 CO 3 (3.0 eq) were combined in a microwave vessel equipped with a teflon stirbar. The system was flushed with argon. A degassed mixture of 3:1 acetone:water (3 mL) was added, and the reaction was heated in a microwave to 100 • C for 30 min. The product was purified via silica gel chromatography in ethyl acetate/hexanes.

General Procedure B for HCl Boc Deprotection, Peptide Coupling, and TFA Boc Deprotection
The appropriate Boc-protected amine intermediate was dissolved in 1,4-dioxane (2-5 mL) and excess concentrated HCl (100-500 µL) was added. The reaction mixture stirred at room temperature for 1-3.5 h. The solvent was removed under vacuum to yield the deprotected amine. The amine intermediate (1.0 eq), diBoc-DMT (1.05 eq), PyBOP (1.0 eq), and 6Cl-HOBt (1.0 eq) were combined, and the reaction flask was flushed with argon. Dry DMF (3-12 mL) and DIEA (10 eq) were added. The reaction mixture stirred at room temperature for 6-24 h. The solvent was removed under vacuum, and the coupled product was purified via silica gel chromatography in ethyl acetate/hexanes. The Boc-protected compound was dissolved in DCM (2-2.5 mL). An equal volume of TFA was added, and the reaction mixture stirred at room temperature for 1-1.5 h. The solvent was removed under vacuum, and the product was purified by semi-preparative HPLC and lyophilized.

General Procedure C for Microwave Suzuki Coupling of Boronic Ester 6 and Pendant Benzyl Bromide
The appropriate benzyl bromide (1.5-2.0 eq), intermediate 6 (1.0 eq), Pd(dppf)Cl 2 (0.1 eq), and K 2 CO 3 (3.0 eq) were combined in a microwave vessel equipped with a teflon stirbar. The system was flushed with argon. A degassed mixture of 3:1 acetone:water (2-3 mL) was added, and the reaction was heated in a microwave to 100 • C for 30 min. The product was purified via silica gel chromatography in ethyl acetate/hexanes. The appropriate benzyl bromide (1.0 eq), intermediate 6 (1.5 eq), Pd(dppf)Cl 2 (0.1 eq), and K 2 CO 3 (3.0 eq) were combined in a microwave vessel equipped with a teflon stirbar. The system was flushed with argon. A degassed mixture of 3:1 acetone:water (2-3 mL) was added, and the reaction was heated in a microwave to 100 • C for 30 min. The product was purified via silica gel chromatography in ethyl acetate/hexanes.

General Procedure E for S N 2 reaction of Benzyl Bromide 3 and Pendant Amine
Benzyl bromide 3 (1.0 eq), the appropriate nucleophile (1.2 eq), and K 2 CO 3 (1.2 eq) were dissolved in dry DMF (3 mL) under an inert atmosphere. The reaction stirred at room temperature overnight. The reaction mixture was partitioned between 2 M NaOH and ethyl acetate. The aqueous layer was extracted with additional ethyl acetate. Combined organic layers were dried over MgSO 4, filtered, and concentrated under vacuum to obtain the product.

General Procedure F for TFA Boc Deprotection, Peptide Coupling, and TFA Boc Deprotection
The appropriate Boc-protected amine intermediate was dissolved in DCM (1-3 mL). An equal volume of TFA was added, and the reaction mixture stirred at room temperature for 1-1.5 h. The solvent was removed under vacuum to yield the deprotected amine. The amine intermediate (1.0 eq), diBoc-DMT (1.05 eq), and PyBOP (1.0 eq) were combined, and the reaction flask was flushed with argon. Dry DMF (3-12 mL) and DIEA (10 eq) were added. The reaction mixture stirred at room temperature for 6-24 h. The solvent was removed under vacuum, and the coupled product was purified via silica gel chromatography in ethyl acetate/hexanes. The Boc-protected compound was dissolved in DCM (2-2.5 mL). An equal volume of TFA was added, and the reaction mixture stirred at room temperature for 1-1.5 h. The solvent was removed under vacuum, and the product was purified by semi-preparative HPLC and lyophilized.

Stimulation of [ 35 S]-GTPγS Binding
Agonist stimulation of KOR, MOR, and DOR by all test compounds was determined by [ 35 S]-guanosine 5 -O-[γ-thio]triphosphate ([ 35 S]-GTPγS) binding assays as previously reported [35][36][37][38]. In a 96-well plate format, membranes from cells expressing opioid receptors as described above (10 µg of protein), [ 35 S]-GTPγS (0.1 nM), and guanosine diphosphate (30 µM) were incubated in GTPγS buffer (50 mM Tris-HCl, 100 mM NaCl, 5 mM MgCl 2 , 1 mM EDTA, pH 7.4) with various concentrations of test compound at 25 • C for 1 h. Basal stimulation was determined by incubation in the absence of any ligand. After incubation, membranes were filtered through Whatman GF/C 1.2 micron glass fiber filters and washed with GTPγS buffer with no EDTA. The radioactivity remaining on the filters was then quantified by liquid scintillation counting after saturation with EcoLume liquid scintillation cocktail in a Perkin-Elmer Microbeta 2450. Data are reported as percent stimulation compared to the effects of 10 µM standard agonists-U69,593 (KOR), DAMGO (MOR), or DPDPE (DOR). Percent stimulation and EC 50 values were determined via nonlinear regression analysis using GraphPad Prism software from at least three separate assays performed in duplicate. Efficacy is expressed as percent stimulation relative to standard agonists.