1-(2,5-Dimethoxy-4-iodophenyl)-2-aminopropane (DOI): From an Obscure to Pivotal Member of the DOX Family of Serotonergic Psychedelic Agents – A Review

Abstract

1-(2,5-Dimethoxy-4-iodophenyl)-2-aminopropane (DOI, or DOX where X = −I) was first synthesized in 1973 in a structure–activity study to explore the effect of various aryl substituents on the then newly identified, and subsequently controlled, hallucinogenic agent 1-(2,5-dimethoxy-4-methylphenyl)-2-aminopropane (DOM, or DOX where X = −CH₃). Over time, DOI was found to be a serotonin (5-HT) receptor agonist using various peripheral 5-HT receptor tissue assays and later, following the identification of multiple families of central 5-HT receptors, an agonist at 5-HT₂ serotonin receptors in rat and, then, human brain. Today, classical hallucinogens, currently referred to as serotonergic psychedelic agents, are receiving considerable attention for their potential therapeutic application in various neuropsychiatric disorders including treatment-resistant depression. Here, we review, for the first time, the historical and current developments that led to DOI becoming a unique, perhaps a landmark, agent in 5-HT₂ receptor research.

There is currently a renewed and growing interest in serotonergic psychedelic agents or classical hallucinogens for use in various neuropsychiatric disorders including treatment-resistant depression (TRD).^e.g.1−10 A broad structural and mechanistic variety of antidepressants are currently being explored but, as a group, serotonergic psychedelic agents represent a large class of agents with psychotherapeutic potential.¹¹ In fact, recognizing their potential therapeutic utility, the U.S. Food and Drug Administration has recently published recommended guidelines for the investigation of such agents.¹² These agents can be broadly divided into two large structural classes: (i) indolealkylamines, exemplified by N,N-dimethyltryptamine-related (or DMT-related) analogues and lysergic acid diethylamide (LSD) that possess a tryptaminergic backbone, and (ii) phenylalkylamines (see below for examples); together, these agents can be collectively referred to as arylalkylamines.¹³ Nearly all of the recent literature on the psychotherapeutic potential of arylalkylamines has focused on indolealkylamines. Yet, phenylalkylamines, agents that include phenylethylamines, phenylisopropylamines, and other structurally related agents, represent an even larger class of compounds.

An interesting member of the psychoactive phenylalkylamine category lurking in the shadows of other better-known members of this class is the phenylisopropylamine 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane (occasionally referred to as 2,5-dimethoxy-4-iodoamphetamine) or DOI (1; Figure 1). Iodo-containing therapeutic agents are a rarity among clinically approved drugs.¹⁴ Interestingly, and rather oddly, many more psychoactive or so-called ”hallucinogenic” agents, or related psychoactive analogues investigated in animal or human structure–activity studies, bear an iodo substituent (for example, see the material below as well as the Shulgin Index¹⁵). DOI (1), although actually quite well-known nowadays from a research perspective, was once an obscure and under-appreciated member of this large family of phenylalkylamines that, over time, has had a profound impact on serotonin receptor research.

Figure 1.

Chemical structures of several phenylalkylamines: DOI (1), mescaline (2), and some representative examples of related DOX agents: 2,4,5-TMA (3), DOM (4), DOET (5), DOPR (6), and DOB (7). Other DOX agents will be mentioned (sans structures), but they retain the 2,5-dimethoxy-4-X substitution pattern shown here. The bolded portion of DOI (1) represents its parent phenylisopropylamine skeleton.

This is not intended to be an exhaustive review of the DOI literature; rather, it will highlight various studies–from their earliest history to the most recent findings, including some relevant and associated salient events–that eventually made DOI a pivotal member of this family of agents. The route taken here will be in chronological order as much as possible. However, many of the types of pharmacological studies being conducted that led to what we now know about DOI and related agents have overlapping considerations. That is, there was considerable chronological overlap in the development and use of the various assay systems to investigate these agents. Initial attempts were made to develop new agents and explain the results of structure–activity studies and to relate the results of certain (presumably centrally mediated) conditioned and unconditioned behavioral assays using peripheral neurotransmitter receptor information prior to the discovery and investigation of multiple types and subtypes of brain neurotransmitter receptors using radioligand binding techniques. Studies in one discipline impacted others. For example, the results of behavioral studies and peripheral tissue assays paved the way for the later development of novel radioligands for investigating central receptors (and, here, the primary focus will be on serotonin receptors). The results of these investigations with central receptors were subsequently used to explain the findings from earlier and concurrent behavioral assays. Combined, these investigations resulted in newer agents and further investigation of DOI. Hence, the journey is somewhat circuitous and requires several digressions; however, it might provide some insight into challenges that were faced and addressed by investigators in the field over the years.

A recent PubMed search for “1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane” or “2,5-dimethoxy-4-iodoamphetamine” revealed >800 publications since 1977.¹⁶ Although agents belonging to what are called the classical hallucinogens or serotonergic psychedelic agents (sometimes referred to as classical psychedelics), we use here the terms hallucinogen and serotonergic psychedelic interchangeably to refer to these agents. The terminology associated with these agents has been discussed and distinction between these agents and dissociative psychedelic agents has been described. ^e.g.(17−20) It is also quite likely that members of this large class of serotonergic psychedelic agents do not necessarily produce identical effects; this attribute, sometimes referred to as polypharmacology, implies that multiple mechanisms of action (i.e., multiple types of neurotransmitter receptors and, perhaps, other targets such as monoamine transporters and metabolic enzymes) might play a role in their actions depending upon the specific agent being considered; however, serotonergic psychedelic agents seem to share certain “common actions” that have allowed for their classification.

Mescaline (2; Figure 1), a phenylethylamine, is the oldest studied member of the hallucinogenic phenylalkylamine family; moving the methoxy substituent from the aryl 3-position to the 2-position and introduction of an α-methyl group results in a more potent hallucinogenic agent, the 2,4,5-trimethoxy analog 2,4,5-TMA (3; Figure 1), also referred to sometimes as TMA-2. Shulgin (for example, see page 633 in Shulgin and Shulgin)²¹ is credited with the acronymic nomenclature associated with many of these agents based, for the most part, on the structure of 2,4,5-TMA (3).

Structurally, DOM (4) can be envisioned as having arisen from 2,4,5-TMA (3) by removal of the 4-position methoxy group (hence, the desoxy or “DO” designation) and introduction of a methyl group (hence, the “M” in DOM). Accordingly, the ethyl analogue has been termed DOET (5), and its homologues would be named likewise, e.g. DOPR (6) and DOBU for the corresponding n-propyl and n-butyl compounds, respectively. Due to the incredibly large number of possible analogues, it is common to simply refer to this as the DOX series wherein the aryl 2,5-dimethoxy groups have been retained and the 4-position substituent has been varied. For example, DOHX is DOX where the 4-position substituent is an n-hexyl group. Obviously, DOI (1) is DOX where X = I; the 4-bromo counterpart of DOI is DOB (7). But, DOM (4) and LSD (8; Figure 2), perhaps due to their earlier appearance or enhanced potency relative to, for example, mescaline (2), have dominated this arylalkylamine class of agents for many years.

Figure 2.

Structures of selected indolealkylamines: lysergic acid diethylamide (LSD; 8), 4-, 5-, and 6-methoxy-N,N-dimethyltryptamine (4-, 5-, and 6-OMe DMT, 9a-9c, respectively), and serotonin (5-hydroxytryptamine, 5-HT, 10).

Both DOM (4) (initially referred to as “STP” – an acronym for “serenity, tranquility and peace”) and DOET (5), appeared on the clandestine market in the late 1960s.²² The earliest clinical studies were reported shortly thereafter by Snyder and colleagues²³⁻²⁶ and by Hollister et al.²⁷ DOM was also included in an early human structure–activity investigation.²⁸ Interestingly, Snyder and colleagues speculated that these agents might represent a potential treatment for mood disorders.²⁹

The first detailed synthesis of DOM (4) and DOET (5) was described in a patent by Shulgin³⁰ in 1970, and the optical isomers of 2,4,5-TMA (3), DOM (4), DOET (5), and DOB (7) were prepared soon thereafter by Nichols et al.³¹ As medicinal chemists are wont to do, the structure of DOM was manipulated to investigate its structure–activity relationships and DOI (1) was first synthesized in 1973 by Coutts and Malicky.³² Although most DOX analogs possess a chiral center at the α carbon atom resulting in a pair of optical isomers, the convention used here is that isomers will be labeled as R- and S-, but racemic mixtures will not use a prefix. For example, (±)DOI or racemic DOI will be referred to as DOI whereas its optical isomers will be identified as R(−)DOI and S(+)DOI.

Before continuing, it should be mentioned en passant that DOX agents might (and should) be viewed as aryl-substituted analogs of the phenylisopropylamine central stimulant 1-phenyl-2-aminopropane (i.e., amphetamine; see the bolded portion of DOI in Figure 1) or its N-monomethyl analogue (i.e., methamphetamine). As already noted above, DOI is sometimes referred to as 2,5-dimethoxy-4-iodoamphetamine; although not chemically incorrect from a trivial nomenclature perspective, the latter term is pharmacologically misleading. That is, although DOX and amphetamine/methamphetamine bear a common phenylalkylamine structural skeleton, very simple structural changes to this skeleton can dramatically alter pharmacology and attendant mechanism(s) of action.¹³ Supportive of this concept is that DOX-related agents produce behavioral effects, as will be further described below, that are distinct from those of amphetamine/methamphetamine;³³ for example, Quinteros-Muñoz et al.³⁴ showed distinct differences in the pharmacological actions of methamphetamine and DOI in rat behavioral assays. This is not to say that amphetamine and methamphetamine cannot produce a psychotogenic effect in humans (i.e., amphetamine psychosis or paranoid-hallucinatory psychosis) that might involve hallucinogenic episodes.^35,36 However, amphetamine psychosis is typically associated with chronic administration of high doses of these agents and involves (primarily catecholaminergic³⁵⁻³⁷) mechanisms that are distinguishable from that of DOX agents described here. In addition, whereas amphetamine-like stimulants produce an increase in rodent locomotor activity, agents such as DOI and LSD typically produce a decrease.³⁸ Moreover, it is the S-isomers of amphetamine-like stimulants (e.g., dextroamphetamine) that are more potent in animal and human studies than their R-enantiomers, whereas the reverse is true of phenylalkylamine hallucinogens as will be described below.

DOX Analogs and Peripheral Serotonin Receptors

Early on, there was speculation that hallucinogenic indolealkylamines such as LSD (8), 4-methoxy-N,N-dimethyltryptamine (4-OMe DMT, 9a), and 5-methoxy-N,N-dimethyltryptamine (5-OMe DMT, 9b; Figure 2) might produce their effects via a serotonergic mechanism; after all, such agents bear a structural similarity to the neurotransmitter serotonin (5-HT; 10).^e.g.39 But, central 5-HT receptors had yet to be understood or even identified. From the late 1950s through the early 1980s, one of the most sensitive and commonly employed in vitro assays for examining the actions of 5-HT was the peripheral rat fundus preparation; various indolealkylamines were found to bind and/or act as agonists.^e.g.40−44 For example, the affinities for a series of DMT analogues was 5-OMe DMT (9b) > 4-OMe DMT (9a) > 6-OMe DMT (9c).⁴⁵

Such studies were latter extended to phenylalkylamines including DOX-related agents, and a variety of isolated tissue preparations were explored. For example, Cheng et al.⁴⁶ found that DOM was active on vascular strips of dog dorsal metatarsal vein in activating 5-HT receptors in this preparation. Dyer et al.⁴⁷ found potencies of DOB > DOET > DOM using sheep umbilical artery preparations, with R(−)DOM being more potent than S(+)DOM. The QSAR of several DOX-related analogues as agonists in the sheep umbilical artery preparation was investigated;⁴⁸⁻⁵⁰ although DOI was not examined, Nichols at al.⁴⁸ introduced the idea of an optimal directional lipophilicity for the 4-position (i.e., the X of DOX) substituent as an important structural feature. In rat aortic strips, R(−)DOB was found more potent than racemic DOB.⁵¹ Contraction of guinea pig trachea by DOX agents via activation of 5-HT receptors has also been examined.⁵² The peripheral (i.e., non-CNS) actions and assays used to investigate 5-HT and various serotonergic agonists and antagonists has been reviewed.⁵³

Perhaps the most extensive amount of comparative data for phenylalkylamines at peripheral 5-HT receptors is derived from isolated rat fundus strips. Huang and Ho⁵⁴ were the first to demonstrate that DOM acts on receptors of the peripheral rat fundus preparation, and Standridge et al.⁵⁵ found R(−)DOM more potent than S(+)DOM. DOM, DOET, and DOB were among the higher-affinity agents at rat fundus 5-HT receptors with affinities comparable to that of 5-HT itself.⁵⁴⁻⁶⁰

None of these studies included DOI (1). Subsequently, DOI was shown to bind at fundus receptors with high affinity, and R(−)DOI displayed somewhat higher affinity than its racemate.⁶¹ When a large series of DOX-related phenylalkylamines was examined at fundus 5-HT receptors, the order of affinity was DOI (1) > DOB (7) > DOET (5) > DOM (4) > 2,4,5-TMA (3), and for the entire series, R(−)DOI was the highest-affinity member (reviewed⁴⁵). Also found was a significant correlation between rat fundus receptor affinity and hallucinogenic potency for 14 DOX-related compounds for which human data were then available.,^45,61

Drug Discrimination Studies

The relationship between animal drug discrimination studies and serotonin receptor action is tightly intertwined. Hence, the former will be discussed prior to turning attention to brain 5-HT receptors.

In drug discrimination studies, animals (most commonly, but not exclusively, rats) are trained to respond in one manner when administered a given dose of a training drug and to respond differently when administered a different drug, drug dose, or saline (i.e., nontraining drug conditions).⁶² For example, using a standard two-lever operant procedure, rats can be trained to reliably respond on one lever following administration of the training dose of a training drug, and to respond on the opposite lever when administered vehicle (e.g., saline), or even a lower dose of the training drug in a graded manner. Because the results are dose-responsive, a dose–response curve can be constructed, and an ED₅₀ value can be calculated for the training drug under extinction (i.e., nonreinforcement) conditions. Adding to the usefulness of this procedure, the trained animals can be administered doses of a test drug (i.e., a challenge drug) in tests of stimulus generalization or substitution (again, under extinction conditions). If a test drug substitutes for the training drug (i.e., if some dose of the test drug results in the animals making >80% of their responses on the training drug-appropriate lever), this is taken as evidence that the test drug can produce stimulus effects in common with (but not necessarily identical to) that of the training drug. Here too, an ED₅₀ dose can be calculated. In effect, the procedure can identify (i) whether or not the test drug produces stimulus effects common to the training drug–a qualitative comparison–and, if so, (ii) how potent the test drug is compared to the training drug–a quantitative comparison. The specific methodology, application, and limitations of the procedure have been reviewed.^13,62,63

As mentioned above, certain hallucinogenic indolealkylamines such as LSD (8) are structurally similar to the neurotransmitter serotonin (5-HT, 10). Hence, it was initially thought that these agents might act via a serotonergic mechanism.³⁹ Because of possible serotonergic involvement in their actions, mescaline (2),^e.g.64,65 LSD (8)^e.g.66,67 and 5-OMe DMT (9b),^e.g.68 as well as other hallucinogenic agents were examined as training drugs in rat drug discrimination studies. The literature on the use of hallucinogenic agents as training drugs has been reviewed.^{e.g.13,62,66,69,70}

Winter,⁶⁴ using the phenylethylamine mescaline (2) as training drug in rats, found that administration of the phenylisopropylamines DOM (4) and DOET (5) resulted in stimulus generalization. In other studies using rats trained to discriminate 5-OMe DMT (9b) from vehicle, the stimulus was attenuated by 5-HT antagonists, and stimulus generalization (substitution) was observed with various other indolealkylamine hallucinogens.^63,71 In fact, there was a positive relationship between their above-mentioned rat fundus-derived 5-HT receptor affinity and stimulus generalization potency.⁷¹ However, although certain other nonindolealkylamine hallucinogenic agents, such as DOM (4) substituted in the 5-OMe DMT-trained animals, it was subsequently demonstrated that (i) DOET (5) and DOB (7) failed to substitute, and (ii) the stimulus effects of 5-OMe DMT were training-dose dependent.⁷²⁻⁷⁴ Indeed, training dose sometimes is a confounding factor in the interpretation of drug discrimination findings.⁷⁵ It might be noted that the stimulus effects of LSD were also shown to be training-dose dependent (see White and Appel⁷⁶ and discussion therein). Other hallucinogenic (and possible hallucinogenic) agents were investigated as training drugs. Silverman and Ho⁷⁷ established DOM (4) as an effective training drug in rats, showed that its effects could be antagonized by the 5-HT receptor antagonists methysergide and cinanserin, and that stimulus generalization occurred upon administration of DOET (5). Later that same year, we followed this lead and trained rats to discriminate DOM from vehicle;⁷⁸ over the course of many years, we ultimately examined (i) the structure–activity relationships of a large variety of agents, including their optical isomers (where applicable), (ii) the relationship between stimulus generalization potency and human hallucinogenic potency, and (iii) the mechanistic underpinnings of these actions. Various indolealkylamine hallucinogenic agents [e.g., 4-OMe DMT (9a), 5-OMe DMT (9b), DMT, LSD (8)],⁷⁹ nonhallucinogenic agents, and “presumably” nonhallucinogenic agents were examined [e.g., there are no human data reported for 6-OMe DMT (9c), although even since its first synthesis, 6-OMe DMT has been found to be significantly less behaviorally active in rats than 5-OMe DMT,⁸⁰ but human data are still unavailable]. Figure 3 shows dose–response curves for several of these indolealkylamines, and the phenylethylamine mescaline and the phenylisopropylamine DOM for comparison (upper panel), and the results with some representative DOX agents (lower panel) in DOM-trained rats. Table 1 (see later) lists ED₅₀ doses for the DOX agents from Figure 3 and a few additional compounds. It might be noted that it was once thought, that because DOI substituted in DOM-trained animals, that DOI was being rapidly metabolized to 1-(2,5-dimethoxyphenyl)-2-aminopropane (2,5-DMA or DOX where X = H); this notion was dispelled when DOI was found to be >10-fold more potent than 2,5-DMA.⁶⁹

Figure 3.

Results of stimulus generalization studies with LSD (8), 4-methoxy-N,N-dimethyltryptamine (4-OMe DMT; 9a), 5-OMe DMT (9b), 6-methoxy-N,N-dimethyltryptamine (6-OMe DMT; 9c), DMT, and mescaline (2) relative to DOM (4) (upper panel) and for several DOX agents (lower panel) in rats trained to discriminate DOM (1.0 mg/kg, i.p.) from vehicle. The plots were generated from tabular data provided from the same laboratory but published separately;^61,82−84 standard errors, although reported in the cited literature, were not included here for purpose of clarity.

Table 1. 5-HT₂ Receptor Affinities (K_i) of Some Representative DOX Agents at [³H]Ketanserin-Labeled Sites, Their Stimulus Generalization Potencies (ED₅₀) in DOM-Trained Rats, and Human Hallucinogenic Doses for Comparison.

Agent	Ki (nM)a	ED50 (mg/kg, i.p.)b	Human Dose (mg, p.o.)c
2,5-DMA	5,200	5.51	80–160
2,4,5-TMA (3)	1,650	3.59	20–40
DOM (4)	100	0.44	3–10
R(−)DOM	60	0.21	<DOMd
DOET (5)	100	0.23	2–6
DOPR (6)	69	0.17	2.5–5
DOB (7)	63	0.20	1–3
R(−)DOB	60	0.10	<DOBd
S(+)DOB	120	0.81	>DOBd
DOI (1)	19	0.42	1.5–3.0
R(−)DOI	10	0.26	<DOId
S(+)DOI	35	1.70	>DOId
DOTBe	19	NSG	-f
DOAMe	7	NSG	>10g
DOBZe	7	NSG	-f

^aBinding data from rat brain cortical homogenates.^141−143

^bData using rats trained to discriminate DOM (1.0 mg/kg, i.p.) from vehicle.^69,83,84 For purpose of comparison, ED₅₀ values (mg/kg) for several related agents include S(+) DOM = 1.7, R(−)DOET = 0.09, S(+)DOET = 0.85, N-monomethyl DOM = 3.99, and LSD (8) = 0.05.

^cHuman psychoactive potency range from Shulgin and Shulgin.²¹

^dIsomers were either more potent (<) or less potent (>) than their racemates, but specific dose ranges were not provided.

^eDOX analogs where X = tert-butyl (DOTB), amyl (i.e., n-pentyl) (DOAM), and benzyl (DOBZ).

^fData not provided; agents not examined in detail.

^gThe >10 mg dose indicates that, should DOAM be active, higher doses might be required.

Almost immediately, some rank-order potency differences became obvious when comparing stimulus generalization data from 5-OMe DMT (1.5 mg/kg)-trained and DOM (1.0 mg/kg)-trained rats. For example, although LSD was the most potent agent in both group of animals, the order of potency in the 5-OMe DMT-trained animals was LSD > 5-OMe DMT > DOM > 4-OMe DMT > DMT > 6-OMe DMT > mescaline;⁸¹ this can be compared with the results from DOM-trained animals (Figure 3, upper panel) with the order of potency being LSD > DOM > 5-OMe DMT > 4-OMe DMT > DMT > mescaline, and where 6-OMe DMT produced saline-appropriate responding at the highest doses examined. A number of other agents were compared in both groups of animals, but only a few representative examples are provided here; for example 7-methoxy-N,N-dimethyltryptamine substituted in 5-OMe DMT-trained animals (with lower potency than 6-OMe DMT) but did not substitute in DOM-trained animals.⁸¹ Evidently, the 5-OMe DMT stimulus, although perhaps somewhat similar to the DOM stimulus, was most likely acting via a somewhat different or more complex receptor-mediated mechanism.

Of note is that there are six possible trimethoxy positional isomers of phenylisopropylamine, with one of them being 2,4,5-TMA (3); of these, five have been examined and all substituted in DOM-trained animals with 2,4,5-TMA (ED₅₀ = 3.59 mg/kg) being among the most potent. The other analogues (followed by ED₅₀ value) include 2,4,6- (3.69 mg/kg), 3,4,5- (or α-methyl mescaline; 6.34 mg/kg), 2,3,4- (7.8 mg/kg) and 2,3,5-trimethoxyphenylisopropylamine (16.5 mg/kg).³³ As mentioned in the Introduction, DOX agents are sometimes referred to as substituted amphetamines, but produce behavioral effects distinct from the latter. Neither DOM, R(−)DOM, S(+)DOM, DOET, nor any of the trimethoxy substituted phenylisopropylamines (e.g., 2,4,5-TMA) substituted in rats trained to discriminate (+)amphetamine from vehicle,^33,77,85,86 and neither amphetamine nor methamphetamine substituted in DOM-trained rats.⁸⁶ Also, there are four DOX analogues that bear a halogen substituent at the 4-position (i.e., DOX where X = -I, -Br, -Cl, -F); that is, in addition to DOI (1) and DOB (7), there are the 4-chloro (i.e., DOC) and 4-fluoro (i.e., DOF) analogues. All four halogenated DOX analogues substituted in DOM-trained rats with an order of potency, on a μMole/kg basis, of DOB > DOI ≅ DOC > DOF.⁸⁵ Rather than providing an extensive compilation of the large number of agents examined and a detailed structure–activity account here, such information has been summarized elsewhere.^{e.g.69,70,87−89} The discrimination-derived structure–activity relationships formulated for psychoactive arylalkylamines in DOM-trained rats can be compared with the structure–activity relationships derived/formulated from various other animal assays and human studies,^28,90−92 and are usually quite consistent depending upon the particular assay system (but, see later). QSAR or quantitative structure–activity relationship studies, to shed light on how physicochemical properties of various substituents might contribute to a given action, are rarely (almost never) conducted using in vivo data due to potential problems associated with absorption, distribution, and/or metabolism when varying structure types are compared. However, for a series of DOX agents varying in structure only by alteration of their X substituent, which might be viewed as a “matched set”, a relating equation was obtained (i.e., Log 1/ED₅₀ = 2.28π – 0.94π² + 4.81; n = 11, r = 0.977) in which substitution potency (ED₅₀) in DOM-trained animals was shown to be related to lipophilicity: π and π² of the 4-position substituent.⁸⁹ Occurrence of both π and π² terms in the equation reflects a parabolic relationship, with a maximum value related to the negative contribution of π. That is, potency increased as the lipophilicity (π value) of the X substituent increased, but decreased once an optimal lipophilicity was achieved (as determined from the negative contribution of the π² term to the relating equation). DOI (1), 2,4,5-TMA (3), DOM (4), DOET (5), DOPR (6), DOB (7), DOC and DOF were included in the study.⁸⁹ Involvement of central 5-HT₂ receptors in the actions of DOX-type agents, and the relationship between drug discrimination findings and human data will be discussed below.

Unconditioned Behavior

Hallucinogenic agents can elicit several conditioned and unconditioned behaviors in animals; one of the most popular unconditioned behaviors for investigation of these agents is the rodent head-twitch response (HTR),⁹³ sometimes referred to as head-shake behavior, that consists of rapid side-to-side rotational head movements (reviewed⁹⁴⁻⁹⁷). This behavior is exhibited by rats and, more commonly, mice after administration of serotonergic hallucinogens and certain other 5-HT receptor agonists. The HTR is not limited to hallucinogenic agents; but, where serotonergic agents have been examined, this effect has been blocked by a variety of 5-HT antagonists including ketanserin and pirenperone⁹⁴ (see below for further discussion on these two antagonists). DOI produced the mouse HTR^e.g.98−101 with R(−)DOI being twice as potent as S(+)DOI, and the response was antagonized by ketanserin.^98,99 A single dose of DOI resulted in tolerance to the DOI-induced HTR at 24 h postinjection but in supersensitivity at 48 h suggesting that the serotonergic system adapts to chronic exposure of agonists (and antagonists).^102,103 DOI also produced the HTR in rats, and the effect, although 5-HT₂-mediated, was found to be modulated by 5-HT_1A as well as D₁ and D₂ dopamine receptor involvement.¹⁰⁴ The HTR in mice and rats using DOI and other agents has been compared.¹⁰⁵ The shrew (Cryptotis parva) seems more sensitive than mice to agents producing the HTR, and DOI was found effective in this species.¹⁰⁶

Another assay used to examine DOX-related agents, although considerably less data are available than from the HTR assay, is the mouse ear-scratch response (ESR),¹⁰⁷ and the effect can be blocked with 5-HT antagonists.⁹⁵ The ESR is produced by DOM and DOET, but not by nonhallucinogenic phenylalkylamines, and appears to be stereoselective with the R(−)-enantiomers of active agents being the more potent.^107,108 DOI produced the ESR and R(−)DOI was about six times more potent that its S(+)-enantiomer; the DOI effect was blocked by very low doses of ketanserin.¹⁰⁹

A caveat is that the DOI-elicited HTR in mice is seemingly a 5-HT_2A agonist-mediated effect, but with possible modulation of the behavior that might involve a competing 5-HT_2C agonist component;^110−112 more will be discussed about these 5-HT receptors in the next section. Some have recently argued that activation of 5-HT_2C receptors, alone or in concert with 5-HT_2A receptor activation, yields comparable HTR behavior involving divergent receptor-mediated downstream signaling.¹¹³ Although the specific signaling cascades mediating the HTR have not been conclusively identified, Gq and β-arrestin2 have been implicated. Recent studies with different existing and novel agents, including DOI, found that the HTR was correlated with Gq efficacy but not with β-arrestin2 recruitment.¹¹⁴

In any event, the HTR behavior in mice has now become a rather routine application. Many have since examined the action of DOI on this behavior and have provided additional information and insight involving drug interactions, testing parameters and conditions, and novel methodologies. Although not a comprehensive listing, HTR studies involving DOI are becoming ever-increasingly used.^{e.g.115−125} Notably, Halberstadt et al.¹²⁶ have demonstrated that a significant correlation exists between the mouse HTR and the results of published drug discrimination studies with serotonergic hallucinogens using rats trained to discriminate either DOM or LSD from vehicle; a large number of agents was examined including DOM, DOET, DOB, and R(−)DOI. Given the reduced cost and labor associated with the HTR using mice relative to drug discrimination studies with rats, the former will probably see increased application in the future.

Mechanistically, 5-HT₂ receptor agonism is inarguably involved in the mouse HTR produced by hallucinogenic agents. This is a given, and this has been demonstrated time and time again. However, administration of the mGlu2/3 agonist LY354740 suppressed the HTR induced by DOI, whereas administration of the mGlu2/3 antagonist LY341495 enhanced the frequency of DOI-induced HTR, raising the possibility that the actions of DOI might be mediated, in part, via increased glutamate release.^125,127 Another study also found that the selective mGlu2/3 agonists LY354740 and LY379268 inhibited DOI-induced HTR in mice.¹²⁸ It has been suggested that Group II metabotropic glutamate receptor agonists are capable of modulating postsynaptic function preferentially in the limbic cortex under conditions of enhanced glutamate release.^129,130

Central Serotonin Receptors

The year 1979 represented a turning point for serotonin receptor research in that, after quite a few years of investigation by several investigators, 5-HT₁ and 5-HT₂ receptors were identified in mammalian brain.¹³¹ Multiple families or populations of central 5-HT receptors have since been identified and these are referred to as 5-HT₁-5-HT₇ serotonin receptors with most, but not all, being G-protein-coupled (i.e., 5-HT₃ receptors are ligand-gated ion channel receptors) (reviewed¹³²). Today, three subpopulations of G-protein coupled 5-HT₂ receptors are recognized: 5-HT_2A, 5-HT_2B (initially termed 5-HT_2F receptors because they were first identified in the peripheral rat fundus), and 5-HT_2C receptors (initially termed 5-HT_1C) receptors. Earlier studies showing a correlation between rat fundus receptor affinity and human hallucinogenic potency was becoming clearer in that fundus serotonin receptors were now considered members of the 5-HT₂ receptor family. An early investigation using [³H]ketanserin and [³H]mesulergine to label rat 5-HT_2A and 5-HT_2C receptors, respectively, showed that for 34 DOX-related agents (including DOI, 2,4,5-TMA, DOM, DOET, DOPR and DOB) there was a significant correlation (r > 0.9) between their K_i values.¹³³ In fact, some years later, a significant correlation was demonstrated between human 5-HT_2A and both human 5-HT_2B and human 5-HT_2C receptor affinities.¹³⁴ The 5-HT_2A and 5-HT_2B functional activity of a series of DOX-related agents (including DOI) has been compared.¹³⁵ A binding profile for DOI at some common receptors has been published.¹³⁶

Digressing here, once 5-HT₁ and 5-HT₂ populations of brain 5-HT receptors were identified, many investigators working with peripheral 5-HT receptor tissue preparations at the time shifted their attention to these CNS receptors using rat brain cortical homogenates and, subsequently, human brain homogenates and, even later, to cloned rat and human receptor preparations. It was generally agreed that rat cortical 5-HT₂ receptors represented the rat version of what was later termed human 5-HT_2A receptors. In the mid to late 1980s, new 5-HT receptor types were still being identified (originally termed binding sites in the early literature but now termed receptor populations or subpopulations); 5-HT₁, 5-HT₂, and 5-HT₃ receptors had been proposed, and subpopulations of 5-HT₁ receptors were beginning to be reported. The selectivity of various agents for the different 5-HT receptor populations/subpopulations was a major and rather vexing problem at the time in that recognized 5-HT agonists (and antagonists), as well as newly developed serotonergic agents, required continual re-evaluation each time a new 5-HT receptor population was identified.¹³⁷ Moreover, given that many investigators were attempting to develop or identify in vitro and in vivo functional assays for various serotonergic agents, and to relate them to activation or antagonism of the different 5-HT receptor populations, this became a nearly Sisyphean chore.¹³⁸ In general, however, although some exceptions exist,^e.g.139 the greater the structural similarity between a serotonergic agent and that of 5-HT, the lower its selectivity.¹⁴⁰ For example, various indolealkylamines displayed affinity for 5-HT₁ and 5-HT₂ receptors; however, DOX-related and certain other phenylalkylamines displayed higher affinity for 5-HT₂ receptors. This made DOX agents–compounds lacking an indolealkylamine moiety–attractive targets for further examination of selectivity. One of the highest-affinity members of the DOX/phenylalkylamine family at 5-HT₂ receptors was DOB (7) (K_i = 63 nM);¹⁴¹ the DOB structure was “deconstructed” and it was found that removal of any of the aryl substituents (i.e., the bromo group, or either one or both of the methoxy groups, or a combination thereof) resulted in a substantial (i.e., 500- to >1,000-fold) decrease in affinity, and that the affinity of R(−)DOB was higher than that of S(+)DOB. Consequently, DOI (1) was examined in greater detail. Because S(+)DOI had not been previously reported, it was synthesized for comparison with R(−)DOI, and S(+)DOI (K_i = 35 nM) displayed only slightly lower affinity than did R(−)DOI (K_i = 10 nM).^141,142 In other words, binding was stereoselective, not stereospecific. See Table 1 for some representative binding data.

A large number of DOX-related phenylalkylamines was examined and structure–activity and QSAR studies were conducted.^51,144 Although 5-HT₂ receptor affinity could be accounted for by the lipophilic and electronic character of the 4-position substituent of DOX compounds, affinity and agonist action were not synonymous. That is, some high-affinity DOX analogues with sterically large/extended 4-position substituents unexpectedly resulted in antagonist action. For example, where the X substituent was benzyl (DOBZ, K_i = 7 nM), n-hexyl (DOHX, K_i = 2.5 nM), or n-octyl (K_i = 3 nM)] the compounds acted as antagonists, whereas DOM (4), DOET (5), DOPR (6), DOB (7), and DOI (1) displayed agonist action in a 5-HT₂-mediated inositol phosphate assay.^51,145 It would appear that there is a “sweet spot” for agonist action with compounds such as DOI and DOB (as well as DOET and DOPR) bearing 4-position substituents being among the optimal. In contrast, DOPP (11; Figure 4) is a high-affinity 5-HT_2A receptor antagonist,⁵¹ and the 2,5-dimethoxy substitution pattern of DOX compounds was not required for high affinity.

Figure 4.

Structures and 5-HT_2A receptor binding data for some representative examples of DOX-related compounds with a lipophilic aryl substituent X at the 4- or 5-position, and antagonist 19 (see text). Where two K_i values are provided; the first was obtained (where data were available) using [³H]DOB as radioligand and the second using [³H]ketanserin. See text for discussion. NA = data not available.

Today, there are a variety of high-affinity nonarylalkylamine chemotypes with 5-HT₂ antagonist action. However, the finding that certain DOX compounds acted as antagonists was unexpected. Later studies found that either one of the two methoxy groups of DOPP (11) could be removed without untoward effect on affinity (e.g., 12-14; Figure 4);¹⁴⁶ in contrast, removal of the 5-methoxy group of DOI resulted in a 10-fold decrease in affinity.¹³³ Also, removal of either the 2- or 5-methoxy group in a series of DOB-related agents resulted in a substantial decrease in 5-HT_2A receptor affinity and in a (although lesser) decrease in potency in the mouse head-twitch assay.¹⁴⁷ However, relocation of the methoxy substituents of DOPP actually enhanced affinity by as much as an order of magnitude. For example, the 2,3-, 3,5-, and 2,6-dimethoxy positional isomers of DOPP (11) bind with K_i values of 3–4 nM. In fact, the presence of a methoxy group is not essential for binding (i.e., 15).^146,148 This is in contrast to the structure–activity findings with DOB and DOI where both methoxy groups are nearly essential for binding, and that the parent structure of 15 (that is, the simplest phenylisopropylamine, amphetamine; see bolded portion of 1 in Figure 1) lacks 5-HT₂ receptor affinity (K_i > 40,000 nM).¹⁴¹ Furthermore, transposition of the 4-[3-(phenyl)propyl] group of 12 and 13 to the 5-position (i.e., 16 and 17) had little effect on affinity. The iodo counterpart of 17 (i.e., 18) displayed considerably reduced affinity. Where these high-affinity compounds were evaluated, they acted as antagonists.¹⁴⁸ It might be noted that removal of the two methoxy groups of DOI (1) results in an agent, 4-iodoamphetamine, that acts as a serotonin transporter reuptake inhibitor.¹⁴⁹ These results supported the concept, as mentioned above, that high affinity at 5-HT_2Areceptors for DOX-like compounds does not equate with agonist action, and indicates that there appear to be differences in the structure–activity relationships of phenylalkylamine agonists and antagonists.¹⁴⁸ In addition, where both [³H]DOB and [³H]ketanserin were used as radioligands in binding studies, it might be noted that these compounds showed minimal differences in affinity indicative of antagonists (see below discussion on the use of agonist radioligands such as [³H]DOB to label 5-HT₂ receptors).

Structural modification of 15 led to a series of novel 5-HT_2A antagonists 19 (Figure 4) and when X = -Br (19a; K_i = 1.3 nM) or −CH₂CH₂CH₂Ph (19b; K_i = 3.2 nM)¹⁵⁰ this once again showed the independence of binding on the presence of the DOX-type methoxy-group substitution pattern. Indeed, subsequent site-directed mutagenesis studies suggested that these agents (i.e., 19) bind differently at 5-HT₂ receptors than do DOX agonists.^150,151

Further suggestive of 5-HT₂ receptor agonist affinity as underlying the behavioral effects of DOX analogs are the results of drug discrimination studies using rats trained to discriminate DOM from vehicle.⁸⁷ Also, high-affinity agonists are among the agents found to be psychoactive in humans (see Table 1). In general, the R(−)-isomers typically bind at 5-HT₂ receptors with higher affinity than their opposite enantiomers, and are the more potent in drug discrimination studies.

DOX Compounds as Radioligands

Shortly following the discovery of brain 5-HT₁ and 5-HT₂ receptors, it was found that most serotonin antagonists available at the time generally bind with high(er) affinity for the latter and low(er) affinity for the former; the converse was true for serotonin agonists (e.g., for serotonin itself). For a while, it was thought that 5-HT₁ receptors might represent the site of action of agonists and 5-HT₂ receptors were the targets of antagonists. What was required at the time was an example(s) of a 5-HT₂ receptor agonist. Furthermore, brain 5-HT₂ receptors were being labeled using antagonist radioligands (e.g., radiolabeled spiperone and, later, ketanserin). To support or refute the thinking at that time, there was a pressing need to identify 5-HT₂ receptor agonists, and 5-HT₁ receptor antagonists. On the basis of earlier studies with peripheral 5-HT receptor preparations, it seemed possible that certain DOX-type compounds might represent the first examples of 5-HT₂ receptor agonists. Subsequent evaluation of a series of DOX and related agents indicated that several DOX compounds displayed high affinity for rat brain 5-HT₂ receptors and acted as agonists; ^e.g.(141) DOI, and particularly R(−)DOI, were among the higher affinity compounds and some data are provided in Table 1.

Subsequently, there was a flurry of interest from several laboratories to develop an agonist radioligand for labeling 5-HT₂ receptors, and [³H]DOB was developed as the first agonist radiolabel for 5-HT₂ receptor binding studies (Table 2).^152,153 Soon thereafter, [⁷⁷Br]-R(−)DOB was examined as a radioligand for labeling human cortical 5-HT_2A receptors.¹⁵⁴

Table 2. Representative 5-HT₂ Receptor Binding Data from Rat Brain Homogenates Using a Labeled Agonist (DOB) and Antagonist (ketanserin) as Radioligand^a.

	Ki (nM)
	[3H]DOB as Radioligand	[3H]Ketanserin as Radioligand	Fold Difference
DOB (7)	0.8	41	51
R(−)DOB	0.4	24	60
S(+)DOB	2.3	146	63
DOI (1)	0.7	19	27
2,4,5-TMA (3)	81	1,250	15
DOM (4)	8.0	100	12
R(−)DOM	1.8	60	33
DOET (5)	1.5	100	67
DOPR (6)	0.9	69	77
5-HT (10)	7.8	928	119
Ketanserin	1.3	1.2	1

^aBinding data are from the same laboratory but were published separately.^{51,152,153,155}

Agonists (e.g., 5-HT) displayed substantially higher affinity for the agonist-labeled (i.e., [³H]DOB-labeled) 5-HT₂ receptors than for the antagonist-labeled receptors (Table 2). Similar results were later obtained with [⁷⁷Br]-R(−)DOB-labeled receptors; for example, DOI (IC₅₀ = 0.4 nM) and DOB (IC₅₀ = 0.2 nM) displayed 80-fold and 200-fold higher affinity, respectively, when compared with their affinities measured using an antagonist (i.e., [³H]spiperone) ragioligand.¹⁵⁴ These findings resulted in greater interest to develop other DOX compounds as radioligands. Several laboratories joined the fray. We¹⁵⁶ prepared and evaluated [¹²⁵I]DOI as a radioligand (Table 3) that eventually became commercially available, and Nichols and co-workers¹⁵⁷ prepared both optical isomers of [¹²⁵I]DOI. Due to the higher specific activity of ¹²⁵I over tritium, radioiodinated compounds could be particularly useful in autoradiographic studies. McKenna and Peroutka¹⁵⁸ examined the binding of a number of agents using [¹²⁵I]-R(−)DOI (see Table 3). Here too, as with [³H]DOB as radioligand, agonists (as opposed to antagonists) tended to display enhanced affinity relative to their affinity measured using [³H]ketanserin as radioligand, and the R-isomers represented the eutomers.

Table 3. Affinities of Selected Agents for Rat Cortical 5-HT_2A Receptors Labeled with Different Radioligands.

	[125I]DOIa (Ki, nM)	[3H]Ketanserina (Ki, nM)	Fold Differenceb	[125I]-R(−)DOIcKi, nM)
R(−)DOM	-d	-		4.8
R(−)DOET	-	-		0.54
R(−)DOB	0.4	25	62.5	0.4
S(+)DOB	-	-		1.0
DOI (1)	0.7	20	28.5	-
R(−)DOI	-	-		0.53
S(+)DOI	-	-		0.89
LSD (8)	-	-		3.5
5-HT (10)	6.1	600	98.4	3.0
Ketanserin	1.3	1.2	1.0	1.2

^aData from Glennon et al.¹⁵⁶

^bAgonists exhibited higher affinity for sites labeled by the agonist radioligand over the antagonist radioligand by x-fold.

^cData from McKenna et al.^158,159

^dData not reported in the cited studies.

Using [¹²⁵I]-R(−)DOI as radioligand, McKenna and Peroutka¹⁵⁸ suggested that it labels a 5-HT receptor (that they tentatively designated 5HT_2A receptors) in rat cortex that is either absent or minimally present in bovine cortex. In contrast, [³H]ketanserin labeled both the putative 5-HT_2A site in rat cortex as well as a separate and distinct recognition site present both in rat and bovine cortex that they tentatively designated “5-HT_2B sites”.¹⁵⁸ This led to some controversy referred to as the “two-site versus two-state” hypothesis. In contrast to two different receptor types, some investigators argued that 5-HT₂ receptors exist in high-affinity and low-affinity states. Subsequent studies with cloned human 5-HT_2A receptors by Hartig and co-workers¹⁶⁰ using [³H]DOB, and Titeler and co-workers^156,161,162 using [¹²⁵I]DOI, argued for the latter and this is now accepted.¹⁶³ That is, radiolabeled agonists label the high-affinity state of 5-HT_2A receptors whereas tritiated ketanserin labels both states. Today, 5-HT_2B receptors, initially referred to as 5-HT_2F receptors, are synonymous with the 5-HT₂ receptors found in the rat fundus preparation,¹³² and are distinct from the above-mentioned “5-HT_2B sites”. DOI (1) binds at [¹²⁵I]DOI-labeled h5-HT_2A receptors (K_i = 0.7 nM) with 3-fold higher affinity than at similarly labeled 5-HT_2C receptors (K_i = 2.4 nM) and with 30 times higher affinity than for [³H]5-HT-labeled h5-HT_2B receptors (K_i = 20 nM).¹³⁴

Autoradiographic studies have utilized [¹²⁵I]DOI ^e.g.(164−166) and both radiolabeled isomers.^167,168 It should be mentioned that [¹³¹I]DOI and [¹²³I]DOI were developed for dog and monkey imaging studies before 5-HT₂ receptors were ever identified,¹⁶⁹ and we later synthesized and examined [¹²³I]DOI as a SPECT (single-photon emission computed tomography) imaging agent in Rhesus monkeys in collaboration with Dr. Kan Sam Lee (NIH)¹⁷⁰ (and see Figure 7–5 in reference (62)). However, although [¹²³I]-R(−)DOI displayed good brain uptake and localized in serotonergic areas of baboon brain, its target to nontarget ratio and its relative insensitivity to ketanserin displacement suggested high nonspecific uptake limiting its usefulness for 5-HT₂ receptor imaging studies by SPECT.¹⁷¹

DOX Agents, Drug Discrimination, and 5-HT_2A Serotonin Receptors

Prior to 1979, a variety of 5-HT receptor agonists and antagonists had been identified, mostly from earlier peripheral tissue studies, and were available for pharmacological investigation. With the subsequent identification of an ever-growing list of multiple brain binding sites as potential 5-HT receptors, it was becoming increasingly difficult to associate a particular physiological response or pharmacological action with one 5-HT receptor population over another, or the involvement of several types of 5-HT receptors, as being involved in the mechanism of action of serotonergic agents.^e.g.172,173 With respect to drug discrimination studies to investigate hallucinogens, a variety of agents were examined as training drugs,¹⁷⁴ and the results were, initially, quite confusing. A good example of this problem is our use of the 5-HT receptor agonist and hallucinogenic agent 5-OMe DMT (9b).¹⁷⁵ Depending upon the training dose (using rats trained to discriminate various i.p. training doses of between 1.0 and 3.0 mg/kg), the 5-OMe DMT stimulus was blocked by certain recognized (at the time) 5-HT antagonists but not by others, and substitution occurred with certain DOX agents but not with some that had previously shown agonist action in peripheral (see above) tissue assays. It was speculated that 5-HT₁ receptor agonism might be playing a role in the 5-OMe DMT stimulus depending upon its training dose,¹⁷⁵ and Spencer et al.¹⁷⁶ later demonstrated, using rats trained to a low dose of 5-OMe DMT (1.25 mg/kg) that, at this dose, its stimulus effects were primarily mediated by 5-HT_1A receptors but involved a 5-HT₂ receptor component. Since that time, the 5-OMe DMT stimulus has been suggested to be nonselective between 5-HT₂ and 5-HT_1A receptors (depending upon the training dose),¹⁷⁷ and was later abandoned. Hence, a different “hallucinogen” would be required to further investigate the stimulus effects of other hallucinogenic agents.

Eventually, we found that the discriminative stimulus effect of DOM as a training drug in rats was not dose-dependent, was attenuated by pretreatment of the animals with certain 5-HT antagonists (later found to be somewhat 5-HT₂ receptor “selective”), but not by other less-selective antagonists, and that DOM-stimulus generalization failed to occur with certain other 5-HT receptor agonists (e.g., 8-hydroxy-N,N-di-n-propylaminotetralin or 8-OH DPAT – now considered a prototypical 5-HT_1A receptor agonist). Conversely, DOM¹⁷⁸ (as well as DOI¹⁷⁹) failed to substitute in rats trained to discriminate 8-OH DPAT from vehicle. Later, DOI was shown to display low affinity (i.e., K_i = 19,000 nM) for [³H]8-OH DPAT-labeled 5-HT_1A receptors.¹⁵⁴ In contrast, although 5-OMe DMT only resulted in partial generalization (i.e., 50% drug-appropriate responding) in rats trained to discriminate 8-OH DPAT from vehicle,¹⁸⁰ generalization was seen when the animals were pretreated with a 5-HT₂ receptor antagonist prior to administration of 5-OMe DMT.¹⁸¹ Hence, DOM, unlike 5-OMe DMT, seemed to lack a 5-HT_1A serotonin receptor component in its actions. Nevertheless, coadministration of a low dose of 8-OH DPAT (50 μg/kg) in combination with DOM to DOM-trained rats resulted in a leftward shift of the DOM dose–response curve and, when administered in combination with the ED₅₀ dose of DOM, resulted in stimulus generalization.⁸⁸ It would appear that although DOM and 8-OH DPAT produce different stimulus effects, and that DOM does not bind at 5-HT_1A receptors and 8-OH DPAT does not bind at 5-HT₂ receptors, 8-OH DPAT is capable of modulating the actions of DOM. Similarly, 8-OH DPAT did not substitute in rats trained to discriminate LSD from vehicle;¹⁸² however, pretreatment with 8-OH DPAT produced a leftward shift in the LSD dose–response curve.¹⁸³ It has been suggested that 5-HT_1A receptor stimulation seems to be a “non-essential component” of the LSD stimulus because administration of the training dose of LSD was unaffected by a 5-HT_1A receptor antagonist.¹⁸³

Ketanserin (initially referred to as R 41 468) and the structurally related pirenperone were identified as the first 5-HT₂- versus 5-HT₁-selective antagonists in the early 1980s,^184,185 and Colpaert et al.¹⁸⁶ found that pirenperone was a potent antagonist of LSD in rats trained to discriminate LSD from vehicle. Both ketanserin and pirenperone were soon found to very potently antagonize the stimulus effects of DOM, and DOM-stimulus generalization to, for example, mescaline (2), LSD (8), and 5-OMe DMT (9b), suggesting a common stimulus action via activation of 5-HT₂ receptors.¹⁸⁷ Later, the stimulus effect of DOM in monkeys, to which substitution occurred with DOI, R(−)DOM, and LSD,¹⁸⁸ was also blocked with 5-HT₂ antagonists including ketanserin.¹⁸⁹ AMI-193 or 8-[3-(4-fluorophenoxy)propyl]-1-phenyl-1,3,8-triazaspiro[4.5]-decanone is a combined 5-HT_2A/dopamine receptor antagonist with >2,000-fold selectivity over 5-HT_2C receptors; this compound potently blocked the stimulus actions of DOM in rats at an antagonist dose of 0.04 mg/kg suggesting that the DOM stimulus is quite likely a 5-HT_2A- rather than 5-HT_2C-mediated event.¹⁹⁰ Fiorella et al.,¹⁹¹ using rats trained to discriminate R(−)DOM (0.4 mg/kg) from vehicle found that stimulus antagonism was better correlated to the 5-HT_2A (r = 0.95) rather than 5-HT_2C (r = 0.25) receptor affinity of the 11 antagonists examined.

Aghajanian and colleagues,^192,193 using electrophysiological techniques to examine various rat brain sites that might mediate the actions of arylalkylamine hallucinogens (including DOI, DOM, DOB, mescaline, and LSD) found that common sites of action were located in the neocortex or in subcortical areas with efferent projections throughout the neocortex and implicated 5-HT_2A receptors as playing a role. Activation of these receptors suppressed the firing rate of 5-HT-containing neurons in the midbrain dorsal raphe nucleus.¹⁹³ Although it cannot be assumed that the raphe nucleus is the sole site of action of LSD, using rats trained to discriminate intraperitoneal administration of LSD, stimulus generalization occurred when LSD was administered centrally via stereotaxically implanted indwelling cannulae in the raphe nucleus.¹⁹⁴ In a pair of rather unique studies, rats trained to discriminate electrical stimulation of the dorsal raphe nucleus generalized to LSD and DOI.^195,196

With the finding that DOB (7) is a high-affinity 5-HT_2A receptor agonist, the higher-affinity isomer, R(−)DOB, was established as a training drug (training dose = 0.2 mg/kg, i.p.) in rat drug discrimination studies.^197,198R(−)DOB (ED₅₀ = 0.05 mg/kg) stimulus generalization occurred with DOM (ED₅₀ = 0.24 mg/kg) and LSD (ED₅₀ = 0.04 mg/kg) (Figure 5),¹⁹⁸ and was potently blocked by pirenperone (at 0.03 mg/kg).¹⁷⁹ Consistent with their affinities for [³H]DOB-labeled 5-HT_2A receptors (K_i), the order of substitution potency (ED₅₀) for several DOB-related agents was: R(−)DOB (K_i = 0.39 nM/ED₅₀ = 0.05 mg/kg) > S(+)DOB (2.3 nM/0.56 mg/kg) > N-monomethyl DOB (7.7 nM/0.82 mg/kg) > N,N-dimethyl DOB (94 nM/5.36 mg/kg) (Figure 5). The N,N,N-trimethyl quaternary amine analog of DOB (K_i = 8,250 nM) failed to substitute, due perhaps to its rather low affinity and/or inability to penetrate the blood-brain barrier.¹⁹⁷ This provided some very useful and applicable in vitro and in vivo structure–activity and structure-affinity (i.e., SAFIR) information.

Figure 5.

Representative results of stimulus generalization studies using rats trained to discriminate R(−)DOB (0.2 mg/kg, i.p.) from saline vehicle. Agents shown elicited >80% R(−)DOB-appropriate responding. Data are plotted primarily from tabular results that were previously reported by us.^197,198 Although standard errors (SEM) were provided in the cited literature, they are not included here for purpose of clarity.

Being another high-affinity agonist at 5-HT_2A receptors, DOI also was examined as a training drug in rats (training dose = 0.5 mg/kg, i.p.; ED₅₀ = 0.16 mg/kg).¹⁹⁹ The DOI stimulus was attenuated by pretreatment of the animals with a very low dose (0.05 mg/kg) of the 5-HT₂ receptor antagonist ketanserin, and DOI stimulus generalization occurred to R(−)DOI, S(+)DOI, DOM, and LSD (ED₅₀ = 0.15, 0.34, 0.49, and 0.05 mg/kg, respectively) (see Figure 6).¹⁹⁹ Stimulus generalization also occurred with DOB, but not with the 5-HT_1A receptor agonist 8-OH DPAT.¹⁷⁹ Subsequently, others employed DOI (1) as a training drug in drug discrimination studies using different training doses and/or schedules of reinforcement. A DOI training dose of 0.63 mg/kg in rats²⁰⁰ was blocked by 5-HT_2A antagonists, and with a DOI training dose of 0.35 mg/kg, LSD substituted for DOI and the stimulus was completely antagonized by ketanserin (0.5 mg/kg) and cyproheptadine (0.5 mg/kg).²⁰¹ Using a training dose of 0.75 mg/kg, Smith et al.²⁰² found that chronic administration of DOI resulted in behavioral tolerance and suggested on the basis of autoradiographic studies, that neuroadaptive changes in 5-HT_2A, not 5-HT_2C, receptors accounted for its stimulus effect. Marona-Lewicka et al.²⁰⁴ reported that unlike an LSD stimulus that involves a role for 5-HT₂ and dopamine D₄ receptors, the DOI stimulus does not involve a dopaminergic component. Overall, the various results were in agreement with the conclusion that DOI serves as an effective discriminative stimulus, that DOI-stimulus generalization can occur with certain other arylalkylamine hallucinogens, and that the stimulus is 5-HT₂-receptor (likely 5-HT_2A rather than 5-HT_2C) mediated. However, far fewer agents have been examined in DOB- or DOI-trained animals than in DOM-trained animals; nevertheless, using either DOM, R(−)DOB, or DOI as training drug in drug discrimination studies, the results were quite consistent in that a 5-HT_2A receptor agonist mechanism primarily underlies their common stimulus effects.

Figure 6.

Dose–response curves for several agents in rats trained to discriminate DOI (0.5 mg/kg) from vehicle. The graph was generated from previously published tabular data.¹⁹⁹

We found for a series of serotonergic hallucinogens that their human potency was correlated with their stimulus generalization potency in DOM-trained animals (reviewed⁶²). For 20 such agents (including, for example, DOI, DOM, DOET, DOPR, and 2,4,5-TMA) that their 5-HT_2A receptor affinity was also significantly correlated (r = 0.94) with their human hallucinogenic potency (reviewed^70,89). Interestingly, there was also a correlation between rat 5-HT_2A and 5-HT_2C (termed 5-HT_1C at the time) receptor affinity for DOX-type agents.¹³³ Later, 5-HT_2A and 5-HT_2C receptor affinity was found to be less robustly correlated (r = 0.66), but achieved significance (r = 0.89) if two outliers [i.e., 1-(2,4-dimethoxyphenyl)-2-aminopropane and 1(2,4-dimethoxy-5-ethoxyphenyl)-2-aminopropane] were removed.⁷⁰ Perhaps, then, correlations between behavioral actions and 5-HT_2A/2C receptor affinity are not surprising given that their 5-HT_2A and 5-HT_2C receptor affinity are intercorrelated.¹³⁴ More recently, Luethi and Liechti²⁰³ demonstrated a significant Spearman rank-order correlation between the human hallucinogenic potency of 35 agents both with 5-HT_2A and 5-HT_2C receptor affinity even though there was only a single agent (i.e., mescaline) common to the two^70,203 investigations. Other findings of the latter study was that there was no correlation between human data and 5-HT_1A or 5-HT_2B receptor affinity, that there was a significant intercorrelation between 5-HT_2A and 5-HT_2C receptor affinity, and that although all agents were 5-HT_2A agonists, there was no correlation between their human potency and agonist action as measured using a calcium mobilization assay.²⁰³

DOI Analogues

Removal or introduction of substituents to the DOX structure has not only provided structure–activity data but has also resulted in a number of useful agents. To illustrate the trend, analogs of DOM and DOB will be described for comparison followed by analogs of DOI.

α-Desmethyl Analogs

Removal of the α-methyl group of phenylisopropylamine DOX-related agents typically results in a slight (usually <10-fold) decrease in in vivo potency but in little change in 5-HT₂ receptor affinity. An explanation for this is that the α-desmethyl analogues are less readily able to penetrate the blood-brain barrier and/or are more quickly metabolized in vivo. The α-desmethyl counterparts of DOX-related agents are commonly referred to as 2C-X compounds. That is, the 3-atom carbon chain of phenylisopropylamines such as DOX is shortened to a 2-atom phenylethylamine moiety (hence, the “2C”, with X representing the 4-position substituent as it does with DOX agents); alternatively, these agents might be viewed as DOX compounds where the α-methyl group has been replaced by a hydrogen atom. α-Desmethyl DOB (or 2C–B, 20; Figure 7) binds at 5-HT_2A receptors (K_i = 32 nM and 1.0 nM at [³H]ketanserin-labeled and [³H]DOB-labeled receptors, respectively), and DOB (K_i = 40 nM and 0.79 nM), binds with comparable affinity. In drug discrimination studies using rats, α-desmethyl DOB substituted in DOM-trained and R(−)DOB-trained animals with potencies (ED₅₀ values) of 0.67 and 0.48 mg/kg; in the latter group of animals, the potencies (ED₅₀ values) of DOM and R(−)DOB were 0.24 and 0.05 mg/kg, respectively.^197,205 Likewise, α-desmethyl DOM (also known as 2C-M) (ED₅₀ = 1.31 mg/kg) substituted in DOM-trained rats (DOM ED₅₀ = 0.44 mg/kg) and was four times less potent than DOM;⁸⁴ its affinity for [³H]ketanserin-labeled rat cortical 5-HT₂ receptors was similar to that of DOM (K_i = 110 nM and 100 nM, respectively) but less than the affinity of R(−)DOI (K_i = 9.9 nM).¹⁴¹ An advantage of studying α-desmethyl analogues is that they lack a chiral center and avoid the synthesis and evaluation of the individual optical isomers.

Figure 7.

DOX-related agents described in the text.

α-Desmethyl DOI or 2C–I (21) behaved in like manner to its α-desmethyl DOM and DOB counterparts. An early QSAR study found that α-desmethyl DOB and α-desmethyl DOI should be the most potent in an examined series of hallucinogenic agents in human studies;^49,206 the QSAR study was performed prior to obtaining human data on α-desmethyl DOI. However, this was found to be the case once human data with α-desmethyl DOI were obtained.^206,207 α-Desmethyl DOI binds at [¹²⁵I]DOI-labeled 5-HT_2A receptors with high affinity (K_i = 0.62 nM),²⁰⁸ and [¹²⁵I]α-desmethyl DOI has been demonstrated to label 5-HT_2A receptors.²⁰⁹ The [¹³¹I]-radiolabeled version also has been used in animal imaging studies.²¹⁰ α-Desmethyl DOI substituted for LSD in drug discrimination studies and produced the HTR in mice; its potency relative to other agents in these assays has been compared.¹²⁶ Moya et al.²¹¹ found, upon comparison of the HTR of a series of 4-substituted DOX analogs bearing or lacking an α-methyl group, that subtle changes in ligand structure can result in significant differences in the cellular (arachidonic acid release and inositol phosphate accumulation) signaling profile, and that the presence or absence of an α-methyl group (e.g., of DOI, DOM, DOB) influences their efficacy at 5-HT_2A (and 5-HT_2C) receptors.²¹²

Simple N-Alkyl Analogues

Introduction of simple terminal-amine substituents such as an N-monomethyl, N,N-dimethyl, or a related short alkyl group to phenylalkylamine DOX agonists results in a small decrease in their 5-HT_2A receptor affinity and human (where data are available) and animal behavioral potency that progressively worsens as the alkyl group is extended in length. This was described above, for example, for N-substituted analogs of DOB;¹⁹⁷ N-monomethylation of DOM or R(−)DOM has resulted in reduced 5-HT_2A receptor affinity (K_i = 415 nM and 260 nM, respectively, relative to 100 nM for DOM),¹⁴¹ and in about a 10-fold reduction in substitution potency in drug discrimination studies with DOM-trained rats.^84,87 N,N-Dimethylation of α-desmethyl DOM (i.e., N,N-dimethyl 2C-M) (ED₅₀ = 5.37 mg/kg) resulted in a 4-fold decease in potency in these same animals.⁸⁴ Various N-monosubstituted and N,N-disubstituted analogs of DOI have been described (e.g., 22a-22h; Figure 7),¹⁵ but pharmacological data are minimal or nonexistent. Human data have been reported for the N,N-dimethyl analogue 22e, which has been referred to as IDNNA (see below). The ¹¹C N-monomethyl and N,N-dimethyl analogues of DOI have been prepared and examined as PET imaging agents.²¹³

α-Ethyl Homologues

Standridge et al.^55,214 in a “program directed at potential non-hallucinogenic performance-enhancing agents” first synthesized and examined the α-ethyl homologue of DOM (i.e., BL-3912; 23) and its optical isomers, followed by an investigation of a series of DOX-related agents. The latter study included the α-ethyl homologues of DOM, DOET, DOPR and, in some cases, their optical isomers; the α-ethyl homologue of R(−)DOI was also prepared and examined.²¹⁴ Some preliminary animal pharmacological data were provided revealing that most of the analogs possess a “low hallucinogenic potential” relative to DOM. The ¹³¹I-substituted compound (i.e., the radioiodinated analogue of the 4-iodo counterpart of BL-3912, or the α-ethyl homologue of DOI, has been reported.²¹⁰

The α-ethyl homologue of DOM, 23, substituted in LSD-trained rats.²¹⁵ Confirming its stimulus actions, in DOM-trained rats the α-ethyl homologue of DOM and its R(−)-isomer substituted for DOM.⁸⁴ The corresponding DOI homologue 24 was not investigated in these studies. A more recent study examined both the α-ethyl (i.e., 24) and α-propyl homologues of DOI (and other compounds) in various assays including activation of 5-HT_2A-mediated second messenger systems.²¹⁶ Because these α-ethyl compounds reportedly lack human hallucinogenic action (where this was investigated; see below), a compelling argument was made that they might be effective for the treatment of certain neuropsychiatric disorders without producing the undesired psychoactive actions of their hallucinogenic DOX counterparts.²¹⁶

N-Benzyl Analogs

Sixty years ago, Ariëns and colleagues^217,218 found that introduction of lengthy/bulky substituents to the terminal amine of certain agonists converted them to antagonists. Today, it is thought that these added substituents might interact within the orthosteric binding site of a target protein but utilize a receptor feature not required by agonists. Hence, we endeavored to use this concept to develop novel 5-HT_2A receptor antagonists; that is, will adding the appropriate terminal amine substituent to DOX-type agonist agents offer a new approach to the development of novel 5-HT_2A receptor antagonists? We began this study with simple N-alkyl substituents (see above discussion with simple N-alkyl groups) and then increased the bulk and length of the substituent. What we found was that certain lengthy/bulky amine substituents (to a degree) enhanced the affinity of α-desmethyl DOB. The study began with small N-monoalkyl groups (e.g., methyl, n-propyl) that were then extended to longer or more bulky groups such as the N-phenylethyl analogue 25 and longer substituents. Interestingly, it was found that N-benzyl analogs of phenylalkylamines (as well as that of indolealkylamines–compounds that will not be discussed here) actually resulted in enhanced affinity.²¹⁹ N-Benzyl analogs of α-desmethyl DOB (e.g., 26) displayed significantly enhanced affinity for 5-HT_2A receptors (and selectivity for 5-HT_2A over 5-HT_2C receptors), and an affinity higher than their longer-chain analogs.²¹⁹ Because they displayed higher affinity for agonist-labeled versus antagonist-labeled 5-HT_2A receptors (see Table 4 for some representative data), we speculated that they would behave as agonists. That is, 5-HT_2A receptor agonists typically display higher affinity for agonist-labeled 5-HT_2A receptors than they do for radiolabeled antagonist-labeled receptors. These studies were never followed up by us. However, others have examined tritiated versions of N-benzyl 21.²²⁰

Table 4. Affinities of Selected N-Alkyl Analogs of 11 at [³H]Ketanserin- and [¹²⁵I]DOI-Labeled 5-HT_2A Receptors²¹⁹.

R	[3H]Ketanserin Ki (nM)	[125I]DOI Ki (nM)
-H	34	1.0
-CH2Ph (26)	16	0.3
-(CH2)2Ph (25)	83	23
-(CH2)3Ph	-	23
-(CH2)4Ph	-	27
-(CH2)2Ph(4′-I)	22	0.4

In a very productive series of studies, Nichols and colleagues prepared and examined a number of conformationally constrained DOX-related analogs. Whereas conformationally constrained compound 27 displayed low affinity for 5-HT_2A receptors, compound 28 acted as an agonist.²²¹ Constraining both methoxy substituents of DOX analogs as dihydrofurans (i.e., 29) was found to be favorable. For example, 29 where X = Br displayed high affinity for [³H]ketanserin- and [¹²⁵I]DOI-labeled 5-HT_2A receptors (K_i = 10.7 and 0.48 nM, respectively).^221,222 This compound (ED₅₀ = 0.06 mg/kg) was nearly equipotent with LSD (ED₅₀ = 0.04 mg/kg) in rats trained to discriminate LSD from vehicle.

Subsequent studies examined the optical isomers of a series of 29 analogs, their fully aromatized counterparts 30, and their individual optical isomers.²²³ For example, 30 where X = I acted as an agonist with the R-isomer (30, X = I; K_i values for [³H]DOB- and [¹²⁵I]DOI-labeled receptors of 0.31 and 0.11 nM, respectively) binding with slightly higher affinity than its S-enantiomer (K_i = 0.68 and 0.25 nM, respectively). Continued effort led to a series of N-substituted analogs 31 (where R′ = H or −CH₃, and X was varied). For example, 32 (25I-NBOMe; K_i = 0.087 nM at [³H]DOB-labeled 5-HT_2A receptors), behaved as an agonist and displayed higher affinity than 2C–I (12; K_i = 0.62 nM), substituted in LSD-trained rats, and was very potent in producing the mouse HTR.¹²⁶ These compounds represent some of the highest affinity 5-HT_2A agonists yet reported.^208,222 In studies focused on developing an agent for positron-emission imaging, Hansen et al.²²⁴ examined 48 α-desmethyl analogues of 2,5-DMA that varied in substituents at the 2′- and 3′-positions of the N-benzyl group and possessed a dozen different aryl 4-position substituents (i.e., 33, Figure 8) with respect to their h5-HT_2A versus r5-HT_2C receptor affinity (using [³H]ketanserin and [³H]mesulergine, respectively, as radioligands) and functional activity as determined using an inositol phosphate turnover assay. Compound 33 where X = Et and R = 2′–OH displayed the highest 5-HT_2A receptor affinity (K_i = 0.29 nM) whereas iodo compound 33 (where X = I and R = 2′–OH) was the most potent (EC₅₀ = 0.074 nM) and 400-fold 5-HT_2A-selective in the functional assay.²²⁴

Figure 8.

Chemical structures of compounds 26-33.

Human Studies

Apart from anecdotal reports that lack citable documentation and, most frequently, lack authentication as to the exact identity of what was being evaluated, relatively little reliable human data have been reported on DOI or its analogues. Some of the most cited and comparative human data have been described by Shulgin and co-workers;^15,21,225 the identity of the agents was rigorously established and other parameters provided (e.g., dose, route of administration, time of onset, duration of action). Unfortunately, there is often little information on the number of subjects in which an agent was examined. DOI was evaluated at various oral doses and found to be psychoactive at 1.5–3.0 mg with an extended (16–30 h) duration of action. R(−)DOI (at 1.0 and 2.3 mg) was found to be more potent than the S(+)-isomer (examined at 6.3 mg). Hence, the human findings with the DOI isomers are consistent with the results of radioligand binding studies and various in vivo conditioned and unconditioned animal behaviors (vide supra) in that the isomers produce a stereoselective, not stereospecific, effect. For purpose of comparison, the potency range for its 4-methyl counterpart, DOM, is stated to be 3–10 mg.²¹ Data for several other DOX analogs can be found in Table 1 for comparison.

Removal of the α-methyl group of DOX-related agents reduces human potency by “sometimes up to an order of magnitude”.²⁰⁷ α-Desmethyl DOM (or 2C-M) is active in humans with an oral dose range of 20–60 mg;²¹ that is, it is approximately six to seven times less potent than DOM. Likewise, α-desmethyl DOB is about ten times less potent than DOB.²¹ α-Desmethyl DOI (i.e., 21) was found to be approximately eight times less potent (dose range = 14–22 mg, p.o.) and of shorter duration (6–12 h) than DOI itself.^21,49,207 The N,N-dimethyl analogue of DOI (or IDNNA, 22e) showed “no activity” at a dose where DOI is active (i.e., at 2.6 mg, p.o), and the α-ethyl homologue of DOI (i.e., 24) was without central effect at oral doses of up to 4.0 mg.^15,21

Due perhaps to their very high affinity as 5-HT_2A receptor agonists, a dose of 50 μg of 32 being sufficient to produce psychoactive effects, abuse of 32 and several related agents has led to toxicity problems including death.²²⁶ See also Pelletier et al.²²⁷ for a recent summary.

In general, it would be gratifying if data from animal behavioral studies were in agreement with human findings, and this is very often the case. For example, the results of human studies are generally consistent with the above-mentioned drug discrimination and HTR data in that DOI is more potent than DOM, the R-isomers of DOX-type agents are more potent than their S-enantiomers (with action being stereoselective rather than stereospecific), deletion of the α-methyl group of DOX agents results in somewhat diminished potency, and N-alkylation (except for the N-benzyl analogs described above) decreases or abolishes activity. However, it should be mentioned that exceptions exist. Whereas BL-3912 (23) substituted in LSD-trained animals²¹⁵ and both BL-3912 and its R(−)isomer substituted in DOM-trained rats⁸⁴ – suggesting that BL-3912 might represent a hallucinogenic agent–human results, although lacking detail–do not seem to support this. Moreover, BL-3912 was active in the HTR assay, with the R-isomer being more potent than the S-isomer (although the response was “markedly attenuated” compared to active control).²¹⁶ In humans, oral doses of the R(−)isomer of 23 “increased mental alertness” at 25–50 mg, and 50–100 mg produced “relief of manic depression”.¹⁵ Doses of 23 up to 270 mg in normal human subjects produced “euphoria and other LSD-like effects but neither perceptual changes nor hallucinations”.²¹⁵ To put this in perspective, the parent structure, DOM, is active in human subjects at doses of 3–10 mg.²¹

Other Studies Utilizing DOI

Although by no means comprehensive, Table 5 provides representative examples of other types of studies that have been conducted with DOI; admittedly, hundreds more studies could have been cited.¹⁶ Early studies simply examined its in vitro and in vivo pharmacology whereas later studies, once it was realized that DOI is a 5-HT₂ receptor agonist, utilized DOI or a radiolabeled version thereof to implicate or eliminate a role for 5-HT₂ receptor involvement in a specific action. In many early cases, the effects of DOI were blocked by various nonselective serotonin antagonists and, later, by more selective 5-HT₂ antagonists such as ketanserin and pirenperone following their discovery. The breadth, scope, and applicational utility of DOI in various fields of investigation are exemplified.

Table 5. Some Representative Studies Utilizing DOI^a.

Topic	Reference
Anorectic actions	(228−232)
Anxiety, depression, schizophrenia	(233−246)
Auditory filtering	(247)
Blood pressure and heart rate	(248−250)
Body shakes and head-bobs in rabbits	(251−253)
Brain-derived neurotrophic factor (BDNF) regulation	(254−258)
Circadian rhythm	(259)
Cocaine (rat hyperactivity, self-administration)	(260, 261)
Compulsive behavior (mice)	(262)
Conditioned taste aversion in rats	(263)
Corticosterone and rodent behavior	(264, 265)
Cyclic AMP response element binding protein (CREB)	(266)
Dorsal raphe neuronal firing inhibited in rat	(267)
Ethanol consumption by mice/rats	(101, 268)
Forepaw treading by rats	(269)
Glutamate–involvement in DOI actions	(270, 271)
Head-twitch response, tonic activation	(272)
Hyperglycemia in rats	(273)
Hyperthermia in rats	(264, 274−277)
Inflammation	(278, 279)
Intracranial self-stimulation (ICSS)	(280)
Intraocular pressure (glaucoma)	(281)
Locomotor activity in rodents	(282−286)
Malignant hyperthermia	(287−289)
PI Turnover	(271, 290, 291)
Platelet aggregation inhibition	(292−295)
Prolactin secretion in rodents	(296, 297)
Self-administration (lack of by Rhesus monkeys)	(298)
Serotonin syndrome in rats	(299)
Sexual performance of male and female rats	(300−305)
Signaling (DOI versus lisuride)	(306)
Spinal 5-HT2 receptors involved in rat motor behavior and nociception	(307−309)
Visual sensorimotor gating	(310)
5-HT2 receptor down-regulation	(311, 312)

^aA number of the studies mentioned here have seen follow-up or related investigations that are not necessarily cited.

Because serotonergic psychedelics are currently receiving renewed interest as potential psychotherapeutic agents it is, perhaps, logical that phenylalkylamines such as DOI be considered. With respect to DOI, very low doses (0.1 mg/kg) showed an antidepressant-like effect in animals (e.g., mouse forced-swim and tail-suspension tests) without inducing the HTR, and a role for 5-HT_2A receptors was implicated.³¹³ However, there has been some controversy regarding the use of serotonergic psychedelic agents for treatment-resistant depression and anxiety and whether or not this beneficial therapeutic effect is inextricably linked to their attendant hallucinogenic actions (for example, see: Jaster and González-Maeso,¹³⁶ Takaba et al.,³¹³ Lewis et al.,³¹⁴ Cameron et al.,³¹⁵ and Wallach et al.³¹⁶ for extended discussions). Nearly 40 years ago, it was speculated that if such agents act via a 5-HT₂ agonist mechanism, “appropriate structural modifications could give rise to 5-HT₂agonists that are not hallucinogenic. In other words, 5-HT₂agonist action might simply be a predisposing factor for such activity; hallucinogens might be 5-HT₂agonists, but will all 5-HT₂agonists be hallucinogenic?”.⁴⁵ Recent studies^{216,314,317,318} now have identified several 5-HT₂ agonists that seemingly lack the human hallucinogenic actions, or hallucinogen-like actions in animals, associated with serotonergic psychedelic agents. by virtue of their different effects involving biased signaling, synaptic plasticity, agonism versus partial agonism, and possibly their specific brain-site(s) of action.^{e.g.4,114,314−316,318−320}

Conclusion

It has been 50 years since its inauspicious beginnings as part of a DOM-related structure–activity study on hallucinogenic agents by Coutts and Malicky,³² but regardless of whether or not DOI achieves clinical ascendency, it deserves acclaim for the considerable role it has played in understanding serotonin receptor pharmacology and function. Being among one of the first recognized 5-HT₂ receptor agonists (although with little selectivity among the three 5-HT₂ receptor subfamilies), DOI and DOI analogues assisted, together with antagonists such as ketanserin, pirenperone, and other antagonists since developed, in unraveling some of the mysteries of 5-HT receptor subtypes. DOI did so, in part, by being recognized as a 5-HT₂ receptor agonist, contributing to formulation of structure–activity (and QSAR) relationships, leading to the development novel agonist radioligands for binding and imaging studies, aiding animal behavioral studies (e.g., drug discrimination, the HTR) to classify serotonergic psychedelic and other agents, assisting the identification of novel 5-HT₂ receptor antagonists, and the introduction of some of the most potent 5-HT₂ receptor agonists reported to date. It has been, and continues to be, especially valuable in achieving and extending the current state-of-affairs in 5-HT receptor research and contributing as a useful tool toward the goal of developing novel and safe psychotherapeutic agents. Nevertheless, future studies using this agent might be hampered somewhat because DOI is now being considered for control as a Schedule I substance by the U.S. DEA.³²¹

Glossary

Abbreviations

AMI-193

8-[3-(4-fluorophenoxy)propyl]-1-phenyl-1,3,8-triazaspiro[4.5]-decanone

DEA

Drug Enforcement Administration

DMT

N,N-dimethyltryptamine

DOB

1-(4-bromo-2,5-dimethoxyphenyl)-2-aminopropane

DOC

1-(4-chloro-2,5-dimethoxyphenyl)-2-aminopropane

DOET

1-(2,5-dimethoxy-4-ethyl-phenyl)-2-aminopropane

DOF

1-(2,5-dimethoxy-4-fluorophenyl)-2-aminopropane)

DOI

1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane

DOM

1-(2,5-dimethoxy-4-methyl-phenyl)-2-aminopropane

DOPP

1-[2,5-dimethoxy-4-(n-3-phenylpropy)phenyl)-2-aminopropane

DOPR

1-(2,5-dimethoxy-4-n-propylphenyl)-2-aminopropane

DOX

1-(2,5-dimethoxy-4-X-phenyl)-2-aminopropane (a general structure where the X substituent is varied)

ESR

ear-scratch response

HTR

head-twitch response

IDNNA

the N,N-dimethyl analogue of DOM

LSD

(+)lysergic acid diethylamide

PET

positron emission tomography

QSAR

quantitative structure–activity relationships

SPECT

single-photon emission computed tomography

TRD

treatment-resistant depression

2C–B

α-desmethyl DOB

2C–I

α-desmethyl DOI

2C-M

α-desmethyl DOM

2,5-DMA

1-(2,5-dimethoxyphenyl)-2-aminopropane

4-OMe DMT

4-methoxy-N,N-dimethyltryptamine

5-HT

5-hydroxytrptamine (serotonin)

5-OMe DMT

5-methoxy-N,N-dimethyltryptamine

6-OMe DMT

6-methoxy-N,N-dimethylytryptamine

8-OH DPAT

8-hydroxy-2-(di-n-propylamino)tetralin

Although some data provided here from the authors’ laboratories were previously supported by NIH grants from NIDA and the NIMH, no funds were available for the preparation of this article.

The authors declare no competing financial interest.

Notes

No new data are provided here. All data have been previously reported and are cited.