Neuroactive Type-A γ-Aminobutyric Acid Receptor Allosteric Modulator Steroids from the Hypobranchial Gland of Marine Mollusk, Conus geographus

Mollusk, Conus geographus

Article

Neuroactive Type‑A γ‑Aminobutyric Acid Receptor Allosteric

Modulator Steroids from the Hypobranchial Gland of Marine

Changshan Niu, Lee S. Leavitt, Zhenjian Lin, Noemi D. Paguigan, Lili Sun, Jie Zhang, Joshua P. Torres,

Shrinivasan Raghuraman, Kevin Chase, Roberto Cadeddu, Manju Karthikeyan, Marco Bortolato,

Christopher A. Reilly, Ronald W. Hughen, Alan R. Light, Baldomero M. Olivera, and Eric W. Schmidt*

Cite This: J. Med. Chem. 2021, 64, 7033 −7043

Read Online

ACCESS

Metrics & More

Article Recommendations

sı Supporting Information

ABSTRACT: In a program to identify pain treatments with low addiction potential, we isolated ﬁve steroids, conosteroids A−E

(

1−5), from the hypobranchial gland of the mollusk Conus geographus. Compounds 1−5 were active in a mouse dorsal root ganglion

DRG) assay that suggested that they might be analgesic. A synthetic analogue 6 was used for a detailed pharmacological study.

Compound 6 signiﬁcantly increased the pain threshold in mice in the hot-plate test at 2 and 50 mg/kg. Compound 6 at 500 nM

antagonizes type-A γ-aminobutyric acid receptors (GABA Rs). In a patch-clamp experiment, out of the six subunit combinations

tested, 6 exhibited subtype selectivity, most strongly antagonizing α β γ and α β γ receptors (IC 1.5 and 1.0 μM, respectively).

1 2

3 2

Although the structures of 1−6 diﬀer from those of known neuroactive steroids, they are cell-type-selective modulators of GABA Rs,

expanding the known chemical space of neuroactive steroids.

. INTRODUCTION

which mediate their multiple biological eﬀects. The best-

characterized mechanism of allopregnanolone and other

neurosteroids such as pregnanolone and 3α,5α-tetrahydro-

deoxycorticosterone (THDOC) is their activity as positive

allosteric modulators (PAMs) of type-A γ-aminobutyric acid

The increase of opioid-related addictions and deaths under-

scores the crucial need for safer, eﬀective pain-management

therapies. Neuroactive steroids, such as neurosteroids, have

tremendous potential to address this need. They have a long

history of use as anesthetic/analgesic and neuromodulatory

agents. Progesterone analogues were used in surgery as

anesthetics, and more recently, the natural progesterone

derivative, allopregnanolone (3α,5α tetrahydroprogesterone),

was approved by the US Food and Drug Administration

receptors (GABA Rs). GABA Rs are pentameric ion

channels consisting most commonly of (α,β) γ subunits,

although other subtypes and combinations are known. With

at least 19 diﬀerent isoforms of various GABA R subunits,

there are many potential receptor combinations. PAM steroids

such as allopregnanolone act primarily at the interface between

(

FDA) for the treatment of postpartum depression. These

developments, in addition to better understanding of the role

of neurosteroids in biology, are leading to a renewed interest in

neuroactive steroids as drugs.

treatments for pain that have lower addiction potential than

currently approved agents.

Received: March 29, 2021

Published: May 5, 2021

This includes promising

The molecular mechanisms of key neuroactive steroids have

been described in great detail. Most known neurosteroids act

on a wide range of membranes and intracellular receptors,

2021 American Chemical Society

033

J. Med. Chem. 2021, 64, 7033−7043

Figure 1. Structures of compounds 1 −6 .

Article

Figure 2. DRG activity of the crude extract of the C. geographus HB gland (100 μg/mL) and compounds 1−5 (10 μg/mL), in response to (A) ATP

20 μM) and (B) KCl (30 mM). Each compound or extract was tested in a single well of a DRG assay, containing >2000 diﬀerentiated DRG cells.

(

The percentage of the total neuron population aﬀected by each compound is indicated. The compounds can have one of four possible eﬀects on

cytosolic Ca²⁺concentration: (1) no change (not shown); (2) increase in the concentration of cytosolic Ca concentration in response to KCl or

ATP relative to control (ampliﬁed); (3) decrease in the cytosolic Ca concentration in response to KCl or ATP relative to control (blocked); or

(

4) a direct eﬀect, wherein the cytosolic Ca concentration increases in direct response to the agent and in the absence of the addition of KCl or

ATP.

β and α subunits, although other sites may be important in

some cases. In general, most known neuroactive steroids

Here, we describe a series of natural steroids that add

signiﬁcantly to the known structural arsenal targeting

acting as GABA R PAMs are usually, but not always, deﬁned

GABA

R. We applied constellation pharmacology using

by the 3α conﬁguration of OH. The structures of chimeric

mouse dorsal root ganglion (DRG) neurons that are directly

6,27

GABA Rs bound to these PAMs reveal the interacting residues

relevant to pain

to discover initial hits that were

and demonstrate very similar docking poses, which are

adjacent to, but noncompetitive with, an important volatile

subsequently validated in animal models. Constellation

pharmacology is a platform for cell-based, high-content

phenotypic screening that can diﬀerentiate cell types during

the screening process. It was chosen because the heteroge-

neous system contains a large library of targets, which unlocks

the potential to unveil new molecular mechanisms and new hit

compounds simultaneously. As a source of new chemistry, we

investigated the oﬀensive/defensive and secretory glands of

12−14

anesthetic binding site.

Neurosteroids acting as GABA_AR

negative allosteric modulators (NAMs), such as pregnenolone

sulfate (PS), bind to diﬀerent but overlapping sites. Other

receptors and channels have also been implicated in

16,17

allopregnanolone-type neurosteroid activity,

and diﬀerent

structural families of neurosteroids can act on completely

28,29

,18,19

cone snails.

The cone secretory venom gland produces

diﬀerent pathways altogether.

Despite intensive study, there are still many unknowns about

thousands of bioactive peptides as well as bioactive small

molecules, which are much less explored than the pep-

GABA R modulation by steroids. By binding to and activating

30−33

tides.

Using an extract of the hypobranchial (HB) gland

GABA Rs, PAM steroids in concert with GABA or as direct

of C. geographus, we identiﬁed conosteroids A−E (1−5)

agonists hyperpolarize and inactivate neurons, leading to

(Figure 1) as the NAMs of GABA R. Although these were

observed physiological eﬀects. Counterintuitively, in some

aﬀorded in low amounts from tiny glands, semisynthesis led to

active analogue 6, which reduced pain in mice in some assays.

The key pharmacophore is an α,β-unsaturated ketone in C-5 to

C-7 positions, suggesting a new strategy to discover selective

neuroactive steroids for pain and other conditions.

neurons such as peripheral nociceptors, GABA R activation

21,22

leads instead to depolarization.

Thus, the contributions of

speciﬁc steroids to analgesia and anesthesia are quite complex.

In addition, there is an incomplete knowledge of the diﬀerent

GABA R subtypes that are targeted by various structural

23,24

classes of steroids.

This leaves a signiﬁcant opportunity to

2. RESULTS

discover neuroactive steroids with new chemical features that

are targeted more selectively toward speciﬁc disease states.

Moreover, neuroactive steroids target a plethora of diﬀerent

2.1. Isolation of Bioactive Conosteroids. Nineteen cone

snail species were collected in Cebu, Philippines, and dissected

to provide diﬀerent organs, which were extracted in methanol.

The dried methanol extracts were resuspended in dimethyl

receptors. For example, the GABA R NAM PS is also a weak

PAM of the N-methyl-D-aspartate (NMDA) receptor.

034

J. Med. Chem. 2021, 64, 7033−7043

Journal of Medicinal Chemistry

inspection of the H- H COSY spectra (Figure 3 ) revealed

Article

sulfoxide (DMSO) and screened using a constellation

pharmacology assay using mouse DRG cells. These cells

comprise many diﬀerent types of sensory neurons, including

several types of peptidergic and nonpeptidergic nociceptors,

the ring system (Figure 3). The absolute conﬁguration of C-24

was determined using the Mo (OAc) -induced electronic

35,36

circular dichroism (ECD) spectrum of 1.

A positive

Cotton eﬀect observed at 305 nm (Figure 4 A) suggested the S

mechanoreceptors, and glia. Using Fura2 to monitor calcium

inﬂux, the DRG cells were challenged with KCl and ATP.

Extracts were screened and scored according to whether they

blocked or ampliﬁed calcium inﬂux resulting from the

application of KCl or ATP or whether they directly depolarized

the cells. The extract from C. geographus HB glands was used in

calcium imaging assays at 100 μg/mL, where it blocked or

decreased responses to ATP (5% of neurons) and high-

concentration KCl (57% of neurons) (Figure 2 ). Also

promising was the low toxicity of the extract (Figure S54).

Subsequent assay-guided puriﬁcation led to the discovery of

ﬁve active neuroactive steroids, conosteroids A−E (1−5) as

the active principles (Figure 1). Although there were many

minor constituents in the HB gland, the major compounds by

far according to diode array detection were the conosteroids,

with an isolated yield of 0.025% of the gland’s wet weight.

Other cone snails in our collection did not contain 1−5,

indicating that they are specialized metabolites with limited

distribution.

Figure 4. Absolute conﬁgurations of 1−3. (A) Mo (OAc) induced

the ECD spectrum of 1−3 in methanol. (B) Calculated and

experimental ECD spectra of 1.

conﬁguration of C-24. To determine the absolute conﬁguration

of 1, time-dependent density functional theory (TDDFT) at

B3LYP/6-31G(d,p) was used, comparing the calculated

ECD spectra of 1a (1S,8S,9S,10R,13R,14S,17R,20R,24S) with

that of its enantiomer 1b (1R,8R,9R,10S,13S,14R,17S,20S,24R)

.2. Structure Elucidation. Conosteroid A (1) was

obtained as white powder with a molecular formula of

C H O based upon the observed m/z 433.3300 [M +

(

Figure 4B). The experimental ECD spectrum of 1 agreed well

with the calculated ECD spectra of 1a. Thus, 1 was determined

as 24(S),3β,25-trihydroxycholest-5-en-7-one.

H] , indicating six degrees of unsaturation. The H, C, and

heteronuclear single quantum coherence nuclear magnetic

resonance (HSQC NMR) spectra (Figures S3, S4, and S7, and

Table S1 ) revealed characteristic signatures of cholestane

steroids, including 27 carbons with ﬁve methyl groups. Careful

Conosteroid B (2 ) was very similar to 1 . The comparison of

the NMR data (Figures S12 −S16 and Table S1 ) revealed that

the major diﬀerence between the NMR data of 1 and 2 was in

the downﬁeld shift of H-3 (from δ 3.56 in 1 to 4.26 in 2) and

C-3 (from δ 71.2 in 1 to 77.8 in 2), which indicated that 2 is

the epimer of 1 at C-3. The key NOESY correlations (Figure

S40 ) between CH −19/H-2a, H-3/H-2a, CH −19/H-4b, and

H-3/H-4b further supported the α-orientation of OH-3. The

Mo (OAc) -induced ECD spectrum of 2 showed a negative

Cotton eﬀect at 305 nm as 1 (Figure 4A), which indicated that

35,36

the absolute conﬁguration of C-24 is R.

Thus, 2 was

identiﬁed as 24(R)-3α,25-trihydroxycholest-5-en-7-one.

Conosteroid C (3) possessed a molecular formula C H O

as deduced from high-resolution electrospray ionization mass

spectroscopy (HRESIMS) (m/z 415.3204 [M + H] ),

indicating seven degrees of unsaturation. An analysis of the

NMR spectroscopic data (Figures S19 −S23 and Table S1)

revealed a very close structural resemblance to 1, with the

diﬀerence that the hydroxy group at C-3 in 1 was dehydrated

to yield a conjugated system. The extended spin system

Figure 3. Representative key COSY (

NOESY ( ) correlations.

), HMBC (

), and

C(1)H -C(2)H -C(3)H-C(4)H seen in H- H COSY and

HSQC spectra, combined with the key HMBC correlation

three extended spin systems, comprising most of the molecule,

supporting this assignment. The presence of an α,β-

unsaturated ketone moiety was supported by NMR data and

a λ_m_a_xat 240 nm (Figure S41). These structural features

revealed that 1 is an oxysterol similar to the known compound

from H-6 to C-4 (δ 128.4), veriﬁed this deduction (Figure 3),

which was further supported by the redshift of ultraviolet (UV)

maximum absorption from 240 nm in 1 to 280 nm in 3. The

NOESY correlations between CH -19/H-8 and H-8/CH -18

showed that these groups are β-orientation, while NOESY

correlations between H-9/H-14 and H-14/H-17 demonstrated

the α-orientation of these protons. The R conﬁguration of C-

24 was demonstrated by the negative Cotton eﬀect at 305 nm

β,25-dihydroxycholest-5-en-7-one, except for an extra

hydroxyl group substitution at C-24. The key heteronuclear

multiple bond correlations (HMBC) from gem-CH -26(27) to

35,36

C-24 (δ_C80.6) veriﬁed this deduction, and all HMBC

correlations were also consistent with it (Figure 3). Nuclear

Overhauser eﬀect spectroscopy (NOESY) correlations be-

tween CH -19/H-1a, CH -19/H-8, H-8/CH -18, and CH -

in the Mo (OAc) -induced ECD spectrum

of 3 (Figure

4A), revealing 3 as 24(R),25-dihydroxycholest-3,5-dien-7-one.

Conosteroids D (4) and E (5) diﬀered from 3 primarily by

having a ketone at C-24 (δ 217.7 in 4; 218.1 in 5) (Table S1).

8/H-16b demonstrated the β-orientation of these groups, and

In addition, C-25 harbored a tertiary −OH group in 4, whereas

5 lacked the hydroxyl. Thus, 4 and 5 were identiﬁed as 25-

the NOESY spectrum fully deﬁned the relative conﬁguration of

035

J. Med. Chem. 2021, 64, 7033−7043

Article

Figure 5. DRG activity of 1 (10 μM). (A) DRG cells loaded with Fura2-AM dye for calcium imaging. The yellow circles indicate the regions of

interest used to collect the traces. (B) Pseudo-colored image Fura2-AM dye reporting on the intracellular calcium level. Blue indicates low, and red

indicates high, cytosolic calcium. (C) Bright ﬁeld image (4×) overlaid with ﬂuorescent images obtained from isolectin B4 (IB4, red) staining, DAPI

(

blue), and GFP expression (green); these are some of the conditions that are used to diﬀerentiate the discrete cell types present in the DRG

preparation. (D) Selected traces from >1000 cells in a DRG assay. Each trace shows that the Ca -mediated ﬂuorescence of a single cell as

pharmacological agents is added and removed. These agents include compound 1 (10 μM), ATP (20 μM); KCl (30 mM); κM-RIIIJ (1 μM); allyl

isothiocyanate (AITC, 100 μM); menthol (400 μM); and capsaicin (300 mM). Overall, these results reveal the diverse eﬀects of 1 on DRG cells.

hydroxycholest-3,5-dien-7,24-dione and cholesta-3,5-dien-

,24-dione, respectively.

.3. Constellation Pharmacology of 1−5 and Syn-

thesis of Analogue 6. Puriﬁed compounds 1−5 exhibited

broadly similar activities when applied to mouse DRG cells at

0 μg/mL (Figure 2). All the compounds were active in the

assay, with diverse eﬀects in the subsets of neurons including

amplifying or blocking the eﬀects of KCl and ATP or directly

causing increases in cytosolic Ca²concentration (direct

eﬀects). These eﬀects were further explored by following 1

26,27

at 10 μM using constellation pharmacology.

Compound 1

blocked the response to ATP in 23.5% of glial cells and

exhibited complex eﬀects on neurons, mainly direct eﬀects or

amplifying the response to high-concentration extracellular

KCl (Figures 5 and 6). These eﬀects were not caused by the

general lipophilic or aggregation eﬀects (Figure S47 ) because

they still occurred in the presence of detergents and were not

caused by other sterols, such as cholesterol, ergosterol, and

cholic acid.

Figure 6. Eﬀects of 1 and 6 at 10 μM on DRG cells. DRG cells were

partitioned into four broad categories using constellation pharmacol-

ogy: glia, peptidergic nociceptors (PN), nonpeptidergic nociceptors

(

NPN), and other neuron subtypes. For each compound, more than

000 cells were used in the assay. The eﬀects were scored, as

described in Figure 2, and the responses to KCl and ATP were

combined.

We were initially prevented from further assessing the

biological activity of conosteroids because of the small size of

the C. geographus HB gland and consequently the very small

amount of available natural products. We hypothesized that

because compounds with diﬀerent 3-OH conﬁgurations

elicited similar eﬀects on neurons, the core pharmacophore

was likely to be dienone, which is reversibly formed from δ-

hydroxyenone. Therefore, we synthesized conosteroid ana-

logue cholesta-3,5-dien-7-one (6) from a readily available

starting point (Figure 1). Constellation pharmacology showed

that 6 at 10 μM exhibited similar activity to 1 (Figure 6). The

major observable diﬀerence in the DRG assay was that fewer

glial cells were inhibited by 6, but the increased Ca inﬂux

into neurons appeared to be very similar with 1−5 and 6. The

eﬀects in the DRG cells could be observed at doses as low as

500 nM, with saturating eﬀects at concentrations >30 μM.

Compound 6 was thus a relatively potent neuroactive steroid.

2.4. Patch-Clamp Determination of GABA R Negative

Allosteric Modulation. Using constellation pharmacology,

neurons that were responsive or nonresponsive to 6 were

selected for single-cell transcriptome sequencing. As a result,

we identiﬁed key GABA R subtypes that we hypothesized to

036

J. Med. Chem. 2021, 64, 7033−7043

Journal of Medicinal Chemistry

Article

be important for the activity of 6 (Figure S49). We then used

the human versions of these subtypes in electrophysiology

experiments.

eceptors (C-LTMRs), and cold thermosensors. At higher doses

∼40 μM, 6 was more broadly active against many classes of

neurons. We therefore investigated the complex actions of 6 at

several diﬀerent doses and in comparison to other pharmaco-

logical agents. The bulk eﬀects of 6 on whole cellular classes

were calculated and statistically evaluated using empirical

cumulative distribution function (eCDF) plots (Figure 8).

Interestingly, Aδ-LTMRs were not aﬀected by 6.

Based on the observed DRG pharmacology and the fact that

many neuroactive steroids are GABA R PAMs, we hypothe-

sized that the actions of 6 might be explained by the PAM of

GABA receptors. To our surprise, a patch-clamp assay using six

diﬀerent human GABA Rs revealed that 6 is instead a

GABA R NAM (Figure 7 A, Figure S55 ). Compound 6

At a dose of 500 nM, it was diﬃcult to discern the direct

eﬀects of 6 on most cells. However, activity in the presence of

GABA was obvious, indicating that 6 acts as a NAM at low

doses (Figure 8B). At a dose of 340 μM, in nociceptors, C-

LTMRS and cold thermosensors GABA increased calcium

ﬂuorescence in response to KCl, indicating that GABA

increased the depolarization of cells. When either PS or 6

was added at 500 nM, the eﬀects of GABA were greatly

diminished, indicating that both act as NAMs under the

experimental conditions. In general, 6 caused a signiﬁcantly

greater decrease in depolarization than did PS, indicating its

greater potency as a NAM of GABA R.

Compound 6 was compared to other sterols at 10 μM

Figure 8C). At that dose, while common membranes and bile

(

acid sterols had no eﬀect, 6 increased depolarization in

response to KCl of nociceptors, C-LTMRs, and cold

thermosensors. The eﬀect was greatest in C-LTMRs.

At the highest dose tested of 40 μM, 6 very strongly

increased the depolarization of C-LTMRs in response to KCl,

and in general, gave a much stronger and more varied response

in several cell types. The ampliﬁcation in C-LTMRs and in

some other cells could be largely abolished by the application

of gabazine (340 μM), a GABA R antagonist. This likely

indicates that high doses of gabazine have additional targets

beyond GABA R, although these are not described in the

literature. At lower doses closer to gabazine IC , the eﬀects of

gabazine were not consistent. Overall, the diverse eﬀects of 6,

amplifying the DRG response at high doses and negatively

modulating GABA at low doses, suggest that 6 is both a

GABA R NAM and a positive modulator of other receptors

that are not deﬁned in this study.

.6. Pain Assays. Compound 6 increases the activity of the

Figure 7. Electrophysiology of compound 6 on selected human

GABA R subtypes. Six diﬀerent GABA R subtypes are used in the

GABA Rs in nociceptors, which can sometimes result in

decreased response to pain.

mg/kg, i.p.) in mice using several pain models, using 8 mice

per treatment group (Figure 9). Compound 6 was active in

17,38

We tested 6 (2, 10, and 50

presence of GABA at the EC₂₀for each channel. (A) Dose response of

at GABA EC . The y-axis shows the relative current amplitude of

each channel at GABA EC , at varying concentrations of 6 (x-axis). I

represents the measured current in the presence of 6, while I_c_o_n_t_r_o_l

represents the GABA EC₂₀in the absence of 6. IC₅₀is calculated using

curve-ﬁtting with nonlinear regression, while h represents the Hill

the hot-plate assay for thermal nociception (p < 0.05 at 2 and

50 mg/kg; p = 0.07 at 10 mg/kg). No signiﬁcant activity could

be discerned in the formalin and von Frey tests, but data

trending toward signiﬁcant could be observed in phase II of the

formalin test (10 mg/kg, p = 0.08). Moreover, 6 exhibited no

toxicity toward mice or human cells or cell lines at the tested

doses (up to 50 mg/kg in mice; up to 100 μM in a human cell

line cytotoxicity test). While the data are leading, further work

is required to determine the therapeutic potential of 6. The

following conclusions can be made from the pain data in hand:

6 is not an anesthetic steroid, as it does not impact the von

Frey test; modest activity in the formalin test indicates that 6

exhibits a minor impact on inﬂammatory pain in mice; activity

in the hot-plate assay may indicate that 6 in mice is an

analgesic via a central mechanism.

coeﬃcient. (B) Inhibition of GABA R current by 6 (50 μM) in the

^p^r^e^s^eⁿ^c^e^o^f^G^A^B^A^a^tⁱ^t^s^E^C²0 ⁱⁿ^c^o^m^p^a^rⁱ^s^oⁿ^t^o^G^A^B^A^E^C20 alone,

revealing the maximum eﬀect of 6 on each channel subtype.

exhibited receptor subtype selectivity, with a wide range of

potencies. The most potently antagonized receptors were

GABA R α β γ (IC 1.5 μM) and α β γ (IC 1.0 μM).

1 2

3 2

Moreover, when GABA was used as an agonist at its EC , the

activity of 6 as a NAM was readily observed (Figure 7B).

.5. Complex Eﬀects of 6 in DRG Cells Suggest

GABA R NAM and Also the Modulation of Other

Receptors. Compound 6 was used at a series of doses in

many diﬀerent DRG experiments. The sum of data after more

than 40 diﬀerent experiments revealed that at doses of ∼10 μM

and below, the activity of 6 was focused on nociceptors (both

peptidergic and nonpeptidergic), C-low-threshold mechanor-

2.7. Anxiety Tests. Neurosteroid PAMs that are

modulators of GABA Rs could have anxiolytic, anticonvulsant,

and sedative properties and can inﬂuence mood and behaviors

in various physiological situations, and conversely, NAMs

037

J. Med. Chem. 2021, 64, 7033−7043

Article

Figure 8. Compound 6 is a direct activator in some cells and a NAM in others. (A) Output of DRG/constellation pharmacology experiments was

analyzed using this equation, leading to the eCDF plots shown in (B−D). In the x-axis, the value of −1 to 1 indicates block to ampliﬁcation, with 0

indicating no eﬀect. The y-axis indicates the relative number of cells that fall below the given value (the empirical distribution function), from 0

(

bottom) to 1.0 (top). The number of cells in each statistics is provided as N = # (see the Methods for details). (B) Application of 6 at 0.5 μM

blue curve) decreases the activity of GABA (green curve), demonstrating the NAM activity. (C) At 10 μM, 6 ampliﬁes the KCl response,

indicating complex interactions likely with multiple receptors, but common membrane sterols do not. (D) At 40 μM, 6 causes even greater

ampliﬁcation of the KCl signal, which is blocked by the application of high-concentration gabazine.

potentially could have psychotropic or anxiogenic eﬀects. To

test this, compound 6 (2, 10, and 50 mg/kg, i.p.) was used in

6 or by the eﬀect of these compounds on the restoration of

presynaptic inhibition. Although modest in their antinoci-

ceptive eﬀects, these results, in addition to the lack of toxic,

psychotropic, or anxiogenic eﬀects, suggest that 6 may serve as

a scaﬀold for further development.

mouse open ﬁeld and elevated Plus-Maze tests. Compared to

its MCT oil vehicle, 6 did not exhibit signiﬁcant eﬀects on

these assays (Figures S51 and S52 ).

Compound 6 was synthesized based upon natural

. DISCUSSION

conosteroids 1−5, and indeed, it showed similar responses to

Signiﬁcant therapeutic interest is currently focused on

GABA R modulating steroids.

potent and cell-type-selective eﬀects on DRG neurons, which

suggested an activity that was in part mediated by GABA R.

Electrophysiology demonstrated that 6 is a potent GABA_AR

NAM with diﬀerent receptor subtype selectivity than other

neuroactive steroids. Moreover, it is well known that

neuroactive steroids often hit more than one target. Based

upon the diﬀerent activities observed when comparing 6 to

characterized NAM PS, compound 6 almost certainly has other

in the DRG assay. Compounds 1−5 clearly diﬀer from each

,7,11

Compound 6 exhibited

other in their pharmacological actions based upon the DRG

assay, but they share suﬃcient similarities that the modulation

of GABA R is at least a part of their mechanisms of action.

The natural and synthetic GABA R modulating steroids

characterized to date hold several structural features in

common, including the cholestane A−B−C−D ring structure.

NAMs with both cis- and trans-A−B ring junctions have been

described. Eﬀective compounds usually contain an OH or a

charged group (such as sulfate in PS) in the 3-position, with

targets beyond GABA R. Thus, 6 is a GABA R NAM but one

one exception.

The PAM and NAM steroids have

that diﬀers from existing, well-characterized NAM steroids.

Because of the profound actions on nociceptors and C-

LTMRs, we tested 6 in a series of pain assays, demonstrating

modest antinociception in a formalin model and a hot-plate

model. Normally, GABA R PAMs, and not NAMs, are

associated with analgesia. Potentially, this discrepancy might

abbreviated sidechains at C-17 in comparison to cholesterol.

This is consistent with their metabolic origin largely from

pregnanes. The C-17 position often has H-bonding elements,

although this is not required. These features deﬁne a

structure−activity relationship (SAR) pattern for the steroid

NAMs.

be explained either by the diﬀerent receptor proﬁle targeted by

038

J. Med. Chem. 2021, 64, 7033−7043

Article

are similar to intermediates in cholic acid biosynthesis. In the

classical pathway, C-24 and C-25 are oxidized to the hydroxyl

groups, which provide a position for oxidative cleavage into

bile acids. Additional redox reactions around the A and B

rings of cholesterol lead to bile salts with diﬀerent oxidation

states.

Thus, it appears that 1−5 are caught at the intermediate

stage prior to cleavage into bile. Despite the fact that they have

not been observed before in nature, specialized enzymes do

not need to be invoked to explain their biosynthesis.

Overall, these results reinforce the usefulness of examining

specialized defensive/oﬀensive glands in animals as a source of

new neuroactive substances. By further investigating the

activity of 6 and the unexplored steroid diversity present in

marine animals, there is potential to uncover agents with novel

pharmacological proﬁles to treat unmet human health

challenges with greater precision.

. EXPERIMENTAL SECTION

Figure 9. Pain models used for the potential utility of 6 (2, 10, and 50

mg/kg, i.p.). In all the panels, error bars indicate the mean ± SE of the

mean, using eight mice per group. Statistical signiﬁcance was assessed

by comparing test compounds to the vehicle using one-way ANOVA

.1. General Experimental Procedures. CD spectra were

obtained on an AVIV circular dichroism spectrometer model 410.

Infrared (IR) spectra were recorded using a Nicolet iS50 FT-IR

spectrometer. NMR data were collected using either a Varian INOVA

(*p < 0.05; **p < 0.01; ***p < 0.001). (A,B) represent the response

00 ( H 500 MHz, C 125 MHz) NMR spectrometer with a 3 mm

to formalin in phase I (the ﬁrst 5 min postinjection) and phase II (the

time between 20 and 35 min postinjection). (C) In the von Frey test

for mechanical nociception, a 2 g stimulus was applied to each mouse

Nalorac MDBG probe or a Varian INOVA 600 ( H 600 MHz,

13 15

150 MHz) NMR spectrometer equipped with a 5 mm H[ C, N]

triple resonance cold probe with a z-axis gradient, utilizing residual

solvent signals for referencing. High-resolution electrospray ionization

mass spectra were obtained using a Waters Xevo G2-XS qTOF.

Phenomenex Luna C₁₈(4.6 × 250 mm, 5 μm) column was used for

analytical and preparative high-performance liquid chromatography

30 times, and the total number of responses (fast withdrawal,

ﬂinching, or licking/biting of the paw in response to the ﬁlament

contact) was recorded. (D) In the hot-plate test for thermal

nociception, animals were placed on a 52 °C hot plate, and the

amount of time prior to nocifensive behavior was recorded.

(HPLC), as conducted on a Hitachi Primaide System equipped with a

430 diode array detector. Adult male mice (C57BL/6, 9 weeks) used

Compounds 1−6 break the SAR model in several ways.

Compound 6 lacks OH or charge in the 3-position, and

for the von Frey and hot-plate tests were purchased from Charles

River Laboratories (Canada). Formalin-induced pain behavior assay

was performed by Charles River Laboratories (Canada). Locomotor

activity and elevated Plus-Maze experiments were performed using

C57BL6/J mice (The Jackson Laboratory, Bar Harbor, ME, USA)

weighing 25 g (nine-week-old). The purity of compounds was >95%,

as gauged by high-performance liquid chromatography−diode array

detection (HPLC−DAD), high-resolution mass spectrometry

instead, it has an unprecedented (for GABA R modulating

steroids) diene−ketone moiety. Compound 6 also contains a

longer sidechain, and yet, it is still an eﬀective neuroactive

steroid. These factors unveil new design principles for future

discovery of neuroactive steroids that have greater selectivity

and better pharmacological properties.

(HRMS), and NMR spectrometry.

The discovery of neuroactive steroids in the gland of a cone

snail suggests that it may be worth revisiting marine animal

steroids in neurological or neuronal assays. Marine inverte-

brates are an abundant source of unique sterols, covering

.2. Animal Collection. C. geographus and other cone snail

specimens were collected at Sogod, Cebu, Philippines, in 2017 and

identiﬁed based on morphology. The specimens were dissected, and

the glands were stored at −80 °C. The HB glands from all C.

geographus were pooled before storage.

45,46

immense chemical diversity.

For the most part, the

.3. Extraction and Puriﬁcation. Pooled HB glands (wet weight,

biological activities of these compounds are either unreported,

or sometimes when applicable, cytotoxicity is reported. In one

recent case, a soft coral steroid was initially reported as

8 g) were ground using a mortar and pestle and then extracted with

methanol (3 × 20 mL). The dried residue was then applied to a C₁₈

column and eluted with methanol. The residue was further separated

cytotoxic, but reinvestigation revealed that it is a potent

by C₁₈HPLC using 10 to 70% CH CN in H O with 0.1% TFA over

NMDA receptor NAM that is under further development.

0 min to obtain fractions 1−6. Fraction 3 was puriﬁed by C HPLC

Several interesting steroid derivatives have been previously

reported from cone snails, although neuroactivity was not

using 25 to 50% CH CN in H O with 0.1% TFA over 40 min to

obtain 1 (1.3 mg, t = 28 min) and 2 (0.7 mg, t = 37 min). Fraction

4 was puriﬁed by C₁₈HPLC using 60% to 90% CH CN in H O with

0.1% TFA over 40 min to obtain 3 (0.7 mg, t

mg, t = 25 min). Fraction 5 was puriﬁed by C HPLC using 60 to

3,49

investigated.

3 2

= 19 min) and 4 (0.8

Several of these sterols have oxygenated sidechains, as found

in 1−5, suggesting a potentially common biogenetic origin. Of

note, 1−5 are specialized metabolites found in the C.

geographus HB gland. The HB glands of diverse mollusks are

involved in defense and secretion, although the role of the HB

00% CH CN in H O with 0.1% TFA over 40 min to obtain 5 (1.0

3 2

mg, t = 36 min).

Conosteroid A (1): colorless amorphous powder; IR ν 3377,

max

−

8−30

2920, 2850, 1669 cm ; ECD (1.6 mM, MeOH), λmax (Δε) 220

gland in Conus spp. is unknown.

The isolation of

(

−13.9), 310 (0.4) nm; (+)-HRESIMS m/z 433.3300 [M + H] ,

calcd for 433.3312, C H O ); H NMR and C NMR (methanol-

compounds from discrete compartments is useful because

the compounds are present in low abundance elsewhere in C.

geographus such that we may not have detected them in whole-

animal extracts. In terms of biosynthetic origin, compounds 1−

+ 1 13

27 45 4

d₄), see Table S1.

Conosteroid B (2): colorless amorphous powder; IR ν 3491,

max

−

2923, 2840, 1670 cm ; (+)-HRESIMS m/z 433.3304 [M + H] ,

039

J. Med. Chem. 2021, 64, 7033−7043

Journal of Medicinal Chemistry

Article

(

calcd for 433.3312, C H O ); H NMR and ¹³C NMR (methanol-

experiments for each puriﬁed compound 1−5 (10 μg/mL) were

repeated twice using wells containing at least 1000 DRG cells.

4.9. Constellation Pharmacology Comparison of 1 and 6

(Figures 5 and 6). Compounds 1 and 6 (10 μM) were tested in

diﬀerent wells. This experiment was repeated twice for each

compound (34 day-old male mouse for 1 and 34 day-old male

mouse for 6) with at least 1000 DRG cells in each experiment.

d₄), see Table S1 .

Conosteroid C (3): colorless amorphous powder; IR ν 3407,

max

−

(

923, 2840, 1665 cm ; (+)-HRESIMS m/z 415.3204 [M + H] ,

calcd for 415.3207, C H O ); H NMR and C NMR (methanol-

27 43 3

d₄), see Table S1 .

Conosteroid D (4): colorless amorphous powder; IR ν 2950,

max

−

(

860, 1715, 1684 cm ; (+)-HRESIMS m/z 413.3065 [M + H] ,

4.10. Constellation Pharmacology to Test the GABA

Receptor Negative Allosteric Modulator Activity of Com-

pound 6 (Figure 8B). Neurons were depolarized every 5 min using

KCl (30 mM) for 10 s. The cells were incubated for 4 min with

GABA (340 μM), GABA (340 μM) + PS (pregnolone sulfate sodium

salt, 0.5 μM), GABA (340 μM) + 6 (0.5 μM), and GABA (340 μM)

at times 12, 32, and 42 min. After each incubation, another KCl pulse

was used to evaluate compound activity by comparing the cell

response to KCl pulse before compound incubation. This experiment

used a 40-day-old male mouse with 1600 DRG neurons.

calcd for 413.3050, C H O ); H NMR and C NMR (methanol-

27 41 3

d₄), see Table S1 .

Conosteroid E (5): colorless amorphous powder; IR ν 3491,

max

−

(

923, 2840, 1656 cm ; (+)-HRESIMS m/z 397.3106 [M + H] ,

calcd for 397.3101, C H O ); H NMR and C NMR (methanol-

27 41 2

d₄), see Table S1 .

.4. Synthesis of Compound 6. To a solution of 7-

ketocholesterol (250 mg) dissolved in MeOH (4 mL) was added

TFA (0.5 mL). The reaction solution was stirred at 50 °C for 2 h. The

reaction mixture was dried in vacuo and resuspended in MeOH.

Compound 6 was puriﬁed by HPLC using a C₁₈column with 100%

acetonitrile. The purity (>95%) and identity of compound 6 were

conﬁrmed by HRMS, HPLC and NMR spectra (Figures S42 −S45 ).

.11. Activity of Other Sterols in Constellation Pharmacol-

ogy (Figure 8C). Neurons were depolarized every 5 min using KCl

30 mM) for 10 s. The cells were incubated for 4 min with 10 μM

(

each cholesterol, ergosterol, cholic acid, and 6 at times 12, 22, 32, 42,

and 52 min. After each incubation, another KCl pulse was used to

evaluate compound activity by comparing the cell response to KCl

pulse before compound incubation. This experiment was done once

.5. ECD Calculation. Conformational searches were performed

by the Spartan 14 package using the MMFF molecular mechanics

force ﬁeld. The geometries with a Boltzmann distribution >5% were

optimized with DFT calculations at the B3LYP/6-31G(d,p) level

using the Gaussian 09 program. The calculations of ECD spectra were

performed at the TDDFT level of theory using CAM-B3LYP/SVP.

Theoretical ECD spectra were obtained as the weighted averages of

Boltzmann populations using the program SpecDis.

.6. General Procedure for the DRG Assay. Lumbar DRGs

L1−L6 from CGRP-GFP (calcitonin gene related peptide and green

ﬂuorescent protein) mice were used as described.

DRG neurons from CGRP-GFP mice were dissociated by treating

DRGs with 0.25% trypsin at 37 °C for 20 min followed by mechanical

trituration with decreasing diameters of ﬁre-polished Pasteur pipettes.

Neurons were then plated into the center of a silicone ring that was

previously attached to the ﬂoor of a 24-well poly-D-lysine coated plate.

After allowing the cells to adhere to the ﬂoor of the plate for

approximately 1 h, medium (0.7 mL) was added to each well,

consisting of minimal essential medium (Invitrogen) supplemented

with 10% fetal bovine serum, 1× penicillin/streptomycin, 10 mM

HEPES, and 0.4% (w/v) glucose, pH 7.4. The plated cells were placed

(

32-day-old male mouse) using 1200 DRG neurons.

.12. Evaluating 6 in the Presence of Gabazine (Figure 8D).

Neurons were depolarized every 5 min using KCl (30 mM) for 10 s.

We performed the DRG assay, as described above, except using

gabazine (3.4 μM), gabazine (34 μM), gabazine (340 μM), 6 (40

μM), gabazine (340 μM) + 6 (40 μM), and 6 (40 μM) for 4 min at

times 12, 22, 32, 42, 52, and 62 min, respectively. This experiment

was repeated twice (26 and 32 day-old male mice) with 2500 DRG

neurons.

26,27

Brieﬂy, lumbar

.13. Formulation of 6. A panel of diﬀerent formulations were

attempted, leading to the choice of medium-chain triglyceride (MCT)

oil as a diluent that completely dissolved 6, which was biocompatible

and was nontoxic in mouse tests. MCT oil (100% USP grade) was

purchased from MCT Lean.

.14. von Frey and Hot-Plate Tests. Animals were randomly

divided into a total of ﬁve groups with eight mice per group. All

groups were intraperitoneally injected 0.25 mL/mouse with vehicle

(

MCT oil), buprenorphine (2 mg/kg), or diﬀerent doses of

compound 6 (2, 10, and 50 mg/kg) prepared in MCT oil 15 min

in a 5% CO incubator at 37 °C overnight. The cells were incubated

before the von Frey test. For the von Frey test, animals were

individually placed in Plexiglas chambers on a wire mesh grid

with Fura2-AM dye for 1 h at 37 °C and 0.5 h at room temperature.

Using a ﬂuorescence microscope, the ﬂuorescence ratio (340 nm/380

nm excitation; 510 nm emission) was used as an indicator of the

relative level of intracellular calcium in each cell while applying a set

of pharmacological agents over a 2 h time course at room

temperature. The cells were exposed to each agent for 200−400

ms, and images were recorded every 2 s for the duration of the

experiment. Pharmacological agents used to diﬀerentiate cell types

were present in each experiment and included allyl isothiocyanate

(Bioseb). A total of three sets of stimuli (10 applications each) were

applied to the plantar surface in the left hind paw, with at least a 5 s

interval between two consecutive applications and at least 5 min

between two consecutive sets of stimuli. The response was deﬁned as

fast withdrawal, ﬂinching, or licking/biting of the paw in response to

the ﬁlament contact. The total number of responses from the 30

stimuli applied was used to calculate the response percentage of each

animal. For the hot-plate test, after the von Frey test, the animal was

individually put on a metal plate surface, which was set at 52 °C, to

evaluate thermal pain hypersensitivity. The time was recorded when

the mouse exhibited nocifensive behaviors such as hind paw

withdrawal, licking, or jumping. Each mouse was tested three times

with 5 min interval between each test. The average time each mouse

stayed on the 52 °C plate was calculated. The compound eﬀect (*p <

0.05, **p < 0.01, ***p < 0.001) was compared to the vehicle group

with one-way analysis of variance (ANOVA) using the GraphPad

Prism.

(

100 μM), menthol (400 μM), capsaicin (300 nM), ATP (20 μM),

KCl (20, 30, and 40 mM), and conotoxin κM-RIIIJ (1 μM).

Following the calcium imaging experiments, the cells were incubated

with Alexa-Fluor 568 isolectin B4 (2.5 μg/mL) for 5 min at room

temperature. The cells were then washed 4 times, incubated for 4 or 5

min, and washed three more times. Images were acquired using a

rhodamine ﬁlter set, and trace data were analyzed with an in-house

software built in the R language.

.7. Cell-Type Deﬁnition by Constellation Pharmacology.

Peptidergic nociceptors were ﬂuorescent cells expressing CGRP-GFP,

while nonpeptidergic nociceptors were stained with the isolectin B4

antibody. C-LTMRs were small-diameter neurons that were not

CGRP-GFP or isolectin B4 positive and that did not respond to

capsaicin.

4.15. Formalin-Induced Pain Assay. The assay was performed

by Charles River Laboratories. Brieﬂy, a total of 42 animals (C57BL/

6, 6−8 weeks) were randomly divided into ﬁve groups with eight mice

per group based on the body weight. Each group was treated 5 mL/kg

with vehicle (MCT oil), duloxetine (10 mg/kg), or diﬀerent doses of

compound 6 (2, 10, and 50 mg/kg) in MCT oil 10 min before the

intraplantar subcutaneous injection of 30 μL of 5% formalin into the

left hind paw. For all groups, animals were acclimated to the

.8. Constellation Pharmacology-Guided Isolation of 1−5

(

Figure 1). The C. geographus HB gland extract (100 μg/mL) and

compounds 1−5 (10 μg/mL) were tested in a series of DRG cell

cultures over a period of months (mice from 25 to 35 days old). The

040

J. Med. Chem. 2021, 64, 7033−7043

Journal of Medicinal Chemistry

The Supporting Information is available free of charge at

Article

observation chamber for about 10 min immediately prior to formalin

injection and then placed in the observation chamber immediately.

Formalin-evoked spontaneous nociceptive behaviors were continu-

ously recorded for 0−60 min using a commercial camcorder. Scoring

from the recorded video ﬁles were performed oﬄine using a PC by an

observer who has been validated to score such nociceptive behavior in

rodents. The total time spent in a 5 min bin was recorded using a

stopwatch for the following nociceptive behavior: ﬂinching, shaking,

biting, and licking of the injected paw. The eﬀects of the drugs were

assessed in the following bins: 0−5 min from the early phase (phase I)

and 20−35 min from the late phase (phase II).

4.19. Elevated Plus-Maze Test. The test was performed under

dim (10 lux) environmental light for 5 min after treatment with

compound 6 (2, 10, and 50 mg/kg) or its vehicle (100% MCT oil).

Brieﬂy, the apparatus was prepared from black Plexiglas with a light

gray ﬂoor and consisted of two open (25 × 5 cm) and two closed

arms (25 × 5 × 5 cm), which extended from a central platform (5 × 5

cm) at 60 cm from the ground. Mice (n = 8/treatment) were

individually placed on the central platform facing an open arm, and

their behavior was observed for 5 min by an experimenter unaware of

the treatment. An arm entry was counted when all four paws were

inside the arm. Behavioral measures included time spent, latencies,

and entries into each partition of the elevated Plus-Maze apparatus;

head dipping and body stretch events as a further indication of

exploratory behaviors; and number of fecal boli.

4.16. Single-Cell Transcriptomics (Figure S49 ). Experiments

were performed as described. Brieﬂy, after the constellation

pharmacology experiments were performed, individual cells were

picked using ﬁre-polished glass pipettes. RNA was prepared for

sequencing using the QIAseq FX Single-Cell RNA library kit

according to the manufacturer’s protocol (Qiagen). The cells were

lysed and mRNA was puriﬁed and reverse-transcribed to generate

cDNA, which then underwent whole-transcriptome ampliﬁcation.

The ampliﬁed cDNA was fragmented to ∼300 bp, end-repaired, and

ligated to adapters. A ﬁnal cleanup was performed using Agencourt

AMPure XP magnetic beads (Beckman Coulter). The cDNA library

was submitted to the HCI high throughput genomics shared resource

for library quality control and sequencing. Sequencing data were

.20. Animals. All animal care and experimental procedures were

performed in accordance with the current guidelines of the National

Institutes of Health (NIH), Charles River Laboratories-Montreal

IACUC and the USA National Research Council and the Canadian

Council on Animal Care (CCAC). Animal experiments were

approved by the Institutional Animal Care and Use Committee

(

IACUC) of the University of Utah.

.21. Statistics. For in vitro DRG assays, the eﬀect of the tested

compound was deﬁned, as shown in Figure 8A, using eCDF:

f(x) = (K_t_e_s_t− K_c_o_n_t_r_o_l)/(K_c_o_n_t_r_o_l+ K_t_e_s_t), f(x): the eﬀect of the

tested compound; Kcontrol : peak height, the cell control response to

analyzed using in-house R scripts, as previously described.

DRG observation solutions before incubation with the test

^.¹⁷^.^G^A^B^A^R^e^c^e^p^t^o^r^E^l^e^c^t^r^o^p^h^y^sⁱ^o^l^o^g^y^S^t^u^d^y^.^Sⁱ^x^G^A^B^AA

compound; K_t_e_s_t: peak height, the cell response to DRG observation

subtypes α1β1γ2, α1β3γ2, α2β3γ2, α3β3γ2, α4β3γ2, and α5β3γ2,

were individually expressed in human HEK cell line. Compound 6 (50

nM, 100 nM, 500 nM, 1 μM, 10 μM, and 50 μM) was assessed using

these expressed channels in the PAM/NAM mode. Positive controls

diazepam (1 μM) and allopregnanolone (100 nM, 300 nM, 1 μM, and

solutions after incubation with the test compound; if f(x) < 0, the test

compound showed decreased intracellular Ca concentration in

response to the test solutions (“blocked”); if f(x) = 0, the test

compound did not aﬀect the intracellular Ca concentration; if f(x) >

, the test compound increased the intracellular Ca concentration,

μM) were tested on each plate, while negative control DMSO

either in response to the observation solution compounds such as KCl

“ampliﬁed”) or directly in response to the test compound (“direct

eﬀect”).

remained constant (0.5%) in the external solution throughout the

assay. A minimum of two cells were used at each concentration of

compound. To test for the PAM or NAM activity, the agonist GABA

at its EC₂₀was applied three times (with wash steps in between) as a

control and to show activation reproducibility followed by 1−2 min

preincubation of 6 and then reapplication of GABA EC in the

(

In GABA receptor electrophysiology studies, the results from each

condition were averaged, and the standard error of mean (SEM) was

calculated for each point (see Table S2 for the n and statistics for each

condition).

For in vivo assays, the treated groups (compound 6 or

buprenorphine) were compared to the vehicle group (n = 8 for

each group). Statistically signiﬁcant responses (*p < 0.05; **p < 0.01;

presence of 6. Finally, following a further wash step, maximum GABA

10 mM) was applied. EC20s for each channel subtype were directly

(

measured as follows: α1β1γ2 (GABA 17 μM), α1β3γ2 (34 μM),

α2β3γ2 (32 μM), α3β3γ2 (38 μM), α4β3γ2 (13 μM), α5β3γ2 (13

μM). Automated patch-clamp recordings were performed at room

temperature using SyncroPatch 384i. The voltage protocol generation

and data collection were performed with PatchController384 V1.6.6

and Data Controller V1.8.0. The data from each experiment for a

**p < 0.001) were calculated in comparison to the vehicle group

using one-way ANOVA.

ASSOCIATED CONTENT

sı Supporting Information

https://pubs.acs.org/doi/10.1021/acs.jmedchem.1c00562.

■

given GABA R were pooled using SyncroPatch. Subsequently, to

show all the results in a single Figure 7, the raw averaged data were

transferred to GraphPad Prism and reanalyzed. Current amplitude

ⁱⁿ^c^r^e^a^s^e^f^o^l^d^w^a^s^g^eⁿ^e^r^a^t^e^d^u^sⁱⁿ^g^t^h^e^f^o^l^l^o^wⁱⁿ^g^e^q^u^a^tⁱ^oⁿ^:⁽^I^/^Icontrol

−

Supporting ﬁgures and tables: Spectral data (1D and 2D

NMR, IR, HRESIMS) for compounds 1−6 and

additional assays (PDF).

), where I is the current amplitude in the presence of 6 and GABA at

EC₂₀for an individual receptor; I_c_o_n_t_r_o_lis the current amplitude in the

presence of GABA alone at EC . If I/I

− 1 > 0, the test

compound showed PAM activity, and while I/I_c_o_n_t_r_o_l− 1 < 0, the test

control

Molecular Formula Strings (CSV)

compound showed NAM activity.

.18. Open Field Test. Locomotion was tested in individual

AUTHOR INFORMATION

Corresponding Author

■

dimly lit (10 lux) gray Plexiglas cubic boxes (20 × 20 × 20 cm) for 5

min after treatment with compound 6 (2, 10, and 50 mg/kg) or its

vehicle (100% MCT oil). To avoid potential anxiety-related

confounds, animals were previously acclimatized to room and light

conditions. On the day of testing, mice (n = 8/treatment) were placed

individually in the center of the box, and their behavior was

videotaped for the following 5 min. Analyses were performed by

experimenters blinded to treatment. Videos were extracted, converted,

and analyzed using Ethovision XT video Tracking Software (Noldus).

Locomotor behaviors measured included cumulative total distance

covered and velocity/s; time spent, latencies, frequencies of

movement, or freezing episodes; and number of fecal boli and

presence of urine in each ﬂoor box.

Eric W. Schmidt − Department of Medicinal Chemistry,

University of Utah, Salt Lake City, Utah 84112, United

States; orcid.org/0000-0001-5839-694X; Phone: 801-

585-5234; Email: ews1@utah.edu

Authors

Changshan Niu − Department of Medicinal Chemistry,

University of Utah, Salt Lake City, Utah 84112, United

States

Lee S. Leavitt − School of Biological Sciences, University of

Utah, Salt Lake City, Utah 84112, United States

041

J. Med. Chem. 2021, 64, 7033−7043

Journal of Medicinal Chemistry

Rasmussen, K. The opioid crisis and the future of addiction and pain

Article

Zhenjian Lin − Department of Medicinal Chemistry,

University of Utah, Salt Lake City, Utah 84112, United

States

Noemi D. Paguigan − Department of Medicinal Chemistry,

University of Utah, Salt Lake City, Utah 84112, United

States

Lili Sun − Department of Pharmacology and Toxicology,

University of Utah, Salt Lake City, Utah 84112, United

States

Jie Zhang − Department of Anesthesiology, School of Medicine,

University of Utah, Salt Lake City, Utah 84112, United

States

Joshua P. Torres − Department of Medicinal Chemistry,

University of Utah, Salt Lake City, Utah 84112, United

States

Shrinivasan Raghuraman − School of Biological Sciences,

University of Utah, Salt Lake City, Utah 84112, United

States

Kevin Chase − School of Biological Sciences, University of

Utah, Salt Lake City, Utah 84112, United States

Roberto Cadeddu − Department of Pharmacology and

Toxicology, University of Utah, Salt Lake City, Utah 84112,

United States

Manju Karthikeyan − School of Biological Sciences, University

of Utah, Salt Lake City, Utah 84112, United States

Marco Bortolato − Department of Pharmacology and

Toxicology, University of Utah, Salt Lake City, Utah 84112,

United States

cumulative distribution function; GABA, γ-aminobutyric acid;

GABAAR, γ-aminobutyric acid type-A receptor; HB, hypo-

branchial; NAMs, negative allosteric modulator; NPN, non-

peptidergic nociceptor; PAM, positive allosteric modulator;

PN, peptidergic nociceptor; PS, pregnanolone sulfate sodium

salt; SAR, structure−activity relationship; THDOC, tetrahy-

drodeoxycorticosterone

REFERENCES

■

(

1) Coussens, N. P.; Sittampalam, G. S.; Jonson, S. G.; Hall, M. D.;

Gorby, H. E.; Tamiz, A. P.; McManus, O. B.; Felder, C. C.;

therapeutics. J. Pharmacol. Exp. Ther. 2019, 371, 396−408.

(

2) Davis, K. D.; Aghaeepour, N.; Ahn, A. H.; Angst, M. S.; Borsook,

D.; Brenton, A.; Burczynski, M. E.; Crean, C.; Edwards, R.;

Gaudilliere, B.; Hergenroeder, G. W.; Iadarola, M. J.; Iyengar, S.;

Jiang, Y.; Kong, J. T.; Mackey, S.; Saab, C. Y.; Sang, C. N.; Scholz, J.;

Segerdahl, M.; Tracey, I.; Veasley, C.; Wang, J.; Wager, T. D.; Wasan,

A. D.; Pelleymounter, M. A. Discovery and validation of biomarkers to

aid the development of safe and effective pain therapeutics: challenges

and opportunities. Nat. Rev. Neurol. 2020, 16, 381−400.

(3) Belelli, D.; Hogenkamp, D.; Gee, K. W.; Lambert, J. J. Realising

the therapeutic potential of neuroactive steroid modulators of the

GABAA receptor. Neurobiol. Stress 2020 , 12 , No. 100207.

M.; Harrison, B. L.; Salituro, F. G. Breakthroughs in neuroactive

4) Kharasch, E. D.; Hollmann, M. W. Steroid anesthesia revisited:

again. Anesth. Analg. 2015 , 120 , 983 −984.

5) Scott, L. J. Brexanolone: first global approval. Drugs 2019, 79,

79−783.

6) Blanco, M. J.; La, D.; Coughlin, Q.; Newman, C. A.; Griffin, A.

(

Christopher A. Reilly − Department of Pharmacology and

Toxicology, University of Utah, Salt Lake City, Utah 84112,

United States; orcid.org/0000-0002-5006-1982

Ronald W. Hughen − Department of Anesthesiology, School of

Medicine, University of Utah, Salt Lake City, Utah 84112,

United States

Alan R. Light − Department of Anesthesiology, School of

Medicine, University of Utah, Salt Lake City, Utah 84112,

United States

steroid drug discovery. Bioorg. Med. Chem. Lett. 2018, 28, 61−70.

(7) Martinez Botella, G.; Salituro, F. G.; Harrison, B. L.; Beresis, R.

T.; Bai, Z.; Blanco, M. J.; Belfort, G. M.; Dai, J.; Loya, C. M.; Ackley,

M. A.; Althaus, A. L.; Grossman, S. J.; Hoffmann, E.; Doherty, J. J.;

Robichaud, A. J. Neuroactive steroids. 2. 3 α -Hydroxy-3 β -methyl-21-

(4-cyano-1H-pyrazol-1 ’-yl)-19-nor-5 β -pregnan-20-one (SAGE-217):

a clinical next generation neuroactive steroid positive allosteric

modulator of the (γ -aminobutyric Acid)_A receptor. J. Med. Chem.

(

017, 60, 7810−7819.

Baldomero M. Olivera − School of Biological Sciences,

University of Utah, Salt Lake City, Utah 84112, United

States

8) Joksimovic, S. L.; Covey, D. F.; Jevtovic-Todorovic, V.;

Todorovic, S. M. Neurosteroids in pain management: a new

perspective on an old player. Front. Pharmacol. 2018, 9, No. 1127.

(

9) Naylor, J. C.; Kilts, J. D.; Shampine, L. J.; Parke, G. J.; Wagner,

Complete contact information is available at:

https://pubs.acs.org/10.1021/acs.jmedchem.1c00562

H. R.; Szabo, S. T.; Smith, K. D.; Allen, T. B.; Telford-Marx, E. G.;

Dunn, C. E.; Cuffe, B. T.; O ’Loughlin, S. H.; Marx, C. E. Effect of

pregnenolone vs placebo on self-reported chronic low back pain

among US military veterans: a randomized clinical trial. JAMA Netw.

Open 2020, 3, No. e200287.

Author Contributions

The manuscript was written through contributions of all

authors. All authors have contributed to writing the manu-

script.

(11) Alvarez, L. D.; Pecci, A.; Estrin, D. A. In search of GABA

10) Olsen, R. W. GABAA receptor: Positive and negative allosteric

modulators. Neuropharmacology 2018, 136, 10−22.

Notes

receptor ’s neurosteroid binding sites. J. Med. Chem. 2019, 62, 5250−

260.

12) Laverty, D.; Thomas, P.; Field, M.; Andersen, O. J.; Gold, M.

The authors declare no competing ﬁnancial interest.

(

ACKNOWLEDGMENTS

■

G.; Biggin, P. C.; Gielen, M.; Smart, T. G. Crystal structures of a

GABAA-receptor chimera reveal new endogenous neurosteroid-

binding sites. Nat. Struct. Mol. Biol. 2017, 24, 977−985.

We thank the government of the Solomon Islands and the

Solomon Islands National University for facilitating the

collection of materials, following the Nagoya Protocol. SB

Drug Discovery performed the GABA_Areceptor electro-

physiology study. This work was funded by the Department

of Defense CDMRP W81XWH-17-1-0413.

(13) Miller, P. S.; Scott, S.; Masiulis, S.; De Colibus, L.; Pardon, E.;

Steyaert, J.; Aricescu, A. R. Structural basis for GABAA receptor

potentiation by neurosteroids. Nat. Struct. Mol. Biol. 2017, 24, 986−

(

92.

14) Chen, Q.; Wells, M. M.; Arjunan, P.; Tillman, T. S.; Cohen, A.

E.; Xu, Y.; Tang, P. Structural basis of neurosteroid anesthetic action

on GABAA receptors. Nat. Commun. 2018, 9, 3972.

ABBREVIATIONS

■

AITC, allyl isothiocyanate; Aδ-LTMR, Aδ type low-threshold

mechanoreceptor; C-LTMR, C type low-threshold mechanor-

eceptor; DRG, dorsal root ganglion; eCDF, empirical

(15) Sugasawa, Y.; Cheng, W. W.; Bracamontes, J. R.; Chen, Z. W.;

Wang, L.; Germann, A. L.; Pierce, S. R.; Senneff, T. C.; Krishnan, K.;

Reichert, D. E.; Covey, D. F.; Akk, G.; Evers, A. S. Site-specific effects

042

J. Med. Chem. 2021, 64, 7033−7043

Journal of Medicinal Chemistry

of neurosteroids on GABAA receptor activation and desensitization.

Article

Elife 2020, 9, No. e55331.

(35) Snatzke, G.; Wagner, U.; Wolff, H. P. Circular dichroism 

LXXV: Cottonogenic derivatives of chiral bidentate ligands with the

(

16) Pathirathna, S.; Brimelow, B. C.; Jagodic, M. M.; Krishnan, K.;

complex [Mo

(36) Frelek, J.; Ikekawa, N.; Takatsuto, S.; Snatzke, G. Application of

[Mo (OAc) ] for determination of absolute configuration of

brassinosteroid vic-diols by circular dichroism. Chirality 1997, 9,

78−582.

37) Lin, Z.; Phadke, S.; Lu, Z.; Beyhan, S.; Abdel Aziz, M. H.;

2 2 3 4

(O CCH ) ]. Tetrahedron 1981, 37, 349 −361.

Jiang, X.; Zorumski, C. F.; Mennerick, S.; Covey, D. F.; Todorovic, S.

M.; Jevtovic-Todorovic, V. New evidence that both T-type calcium

channels and GABAA channels are responsible for the potent

peripheral analgesic effects of 5 α -reduced neuroactive steroids. Pain

(

005, 114, 429−443.

Reilly, C.; Schmidt, E. W. Onydecalins, fungal polyketides with Anti-

17) Colciago, A.; Bonalume, V.; Melfi, V.; Magnaghi, V. Genomic

Histoplasma and anti-TRP activity. J. Nat. Prod. 2018, 81, 2605−2611.

and non-genomic action of neurosteroids in the peripheral nervous

system. Front. Neurosci. 2020 , 14 , No. 796.

(38) Du, X.; Hao, H.; Yang, Y.; Huang, S.; Wang, C.; Gigout, S.;

Ramli, R.; Li, X.; Jaworska, E.; Edwards, I.; Deuchars, J.; Yanagawa, Y.;

Qi, J.; Guan, B.; Jaffe, D. B.; Zhang, H.; Gamper, N. Local GABAergic

signaling within sensory ganglia controls peripheral nociceptive

transmission. J. Clin. Invest. 2017, 127, 1741−1756.

18) Vallee, M. Neurosteroids and potential therapeutics: focus on

pregnenolone. J. Steroid Biochem. Mol. Biol. 2016 , 160 , 78 −87.

19) Ackerman, H. D.; Gerhard, G. S. Bile acids in neuro-

degenerative disorders. Front. Aging Neurosci. 2016 , 8 , 263.

20) Farrant, M.; Nusser, Z. Variations on an inhibitory theme:

phasic and tonic activation of GABA(A) receptors. Nat. Rev. Neurosci.

005, 6, 215−229.

21) Tillman, L.; Zhang, J. Crossing the chloride channel: the

(39) Deuis, J. R.; Dvorakova, L. S.; Vetter, I. Methods used to

evaluate pain behaviors in rodents. Front. Mol. Neurosci. 2017, 10,

No. 284.

41) Ramos, A.; Pereira, E.; Martins, G. C.; Wehrmeister, T. D.;

40) Belelli, D.; Lambert, J. J. Neurosteroids: endogenous regulators

of the GABA A receptor. Nat. Rev. Neurosci. 2005, 6, 565−575.

−

current and potential therapeutic value of the neuronal K -Cl

cotransporter KCC2. BioMed Res. Int. 2019, 2019, No. 1.

(

Izídio, G. S. Integrating the open field, elevated plus maze and light/

dark box to assess different types of emotional behaviors in one single

trial. Behav. Brain Res. 2008, 193, 277−288.

(42) Munro, G.; Erichsen, H. K.; Rae, M. G.; Mirza, N. R. A

22) Wilke, B. U.; Kummer, K. K.; Leitner, M. G.; Kress, M.

Chloride the underrated ion in nociceptors. Front. Neurosci. 2020,

4, No. 287.

23) Borghese, C. M.; Herman, M.; Snell, L. D.; Lawrence, K. J.;

(

question of balance positive versus negative allosteric modulation of

GABA(A) receptor subtypes as a driver of analgesic efficacy in rat

models of inflammatory and neuropathic pain. Neuropharmacology

Lee, H. Y.; Backos, D. S.; Vanderlinden, L. A.; Harris, R. A.; Roberto,

M.; Hoffman, P. L.; Tabakoff, B. Novel molecule exhibiting selective

affinity for GABAA receptor subtypes. Sci. Rep. 2017 , 7 , 6230.

(

011, 61, 121−132.

24) Lu, X.; Zorumski, C. F.; Mennerick, S. Lack of Neurosteroid

43) Akk, G.; Li, P.; Bracamontes, J.; Reichert, D. E.; Covey, D. F.;

selectivity at delta vs. gamma2-containing GABAA receptors in

dentate granule neurons. Front. Mol. Neurosci. 2020, 13, 6.

(

Smejkalova, T.; Ladislav, M.; Balik, A.; Korinek, M.; Chodounska, H.;

Kudova, E.; Vyklicky, L. Site of action of brain neurosteroid

pregnenolone sulfate at the N -methyl-D -aspartate receptor. J. Neurosci.

Steinbach, J. H. Mutations of the GABA-A receptor alpha1 subunit

M1 domain reveal unexpected complexity for modulation by

neuroactive steroids. Mol. Pharmacol. 2008, 74, 614−627.

25) Hrcka Krausova, B.; Kysilov, B.; Cerny, J.; Vyklicky, V.;

(44) Li, P.; Bandyopadhyaya, A. K.; Covey, D. F.; Steinbach, J. H.;

Akk, G. Hydrogen bonding between the 17beta-substituent of a

neurosteroid and the GABA(A) receptor is not obligatory for channel

potentiation. Br. J. Pharmacol. 2009 , 158 , 1322 −1329.

(

020, 40, 5922−5936.

26) Giacobassi, M. J.; Leavitt, L. S.; Raghuraman, S.; Alluri, R.;

(45) Giner, J. L. Biosynthesis of marine sterol side chains. Chem. Rev.

Chase, K.; Finol-Urdaneta, R. K.; Terlau, H.; Teichert, R. W.; Olivera,

B. M. An integrative approach to the facile functional classification of

dorsal root ganglion neuronal subclasses. Proc. Natl. Acad. Sci. U. S. A.

1993, 93, 1735−1752.

(46) Kerr, R. G.; Baker, B. J. Marine sterols. Nat. Prod. Rep. 1991, 8,

465−497.

27) Teichert, R. W.; Schmidt, E. W.; Olivera, B. M. Constellation

020, 117, 5494−5501.

(47) Duh, C. Y.; Wang, S. K.; Chu, M. J.; Sheu, J. H. Cytotoxic

sterols from the soft coral Nephthea erecta . J. Nat. Prod. 1998, 61,

1022−1024.

pharmacology: a new paradigm for drug discovery. Annu. Rev.

Pharmacol. Toxicol. 2015, 55, 573 −589.

(48) Leng, T.; Liu, A.; Wang, Y.; Chen, X.; Zhou, S.; Li, Q.; Zhu,

W.; Zhou, Y.; Su, X.; Huang, Y.; Yin, W.; Qiu, P.; Hu, H.; Xiong, Z.

G.; Zhang, J.; Yan, G. Naturally occurring marine steroid 24-

methylenecholestane-3 β ,5 α ,6 β ,19-tetraol functions as a novel neuro-

protectant. Steroids 2016, 105, 96−105.

(28) Ponte, G.; Modica, M. V. Salivary glands in predatory mollusks:

evolutionary considerations. Front. Physiol. 2017, 8, No. 580.

29) Turner, A. H.; Craik, D. J.; Kaas, Q.; Schroeder, C. I. Bioactive

compounds isolated from neglected predatory marine gastropods.

Mar. Drugs 2018, 16, No. 118.

(49) Aknin, M.; Viracaoundin, I.; Faure, R.; Gaydou, E. M. 5 α ,8 α -

Epidioxycholest-6-en-3-β -ol from three cone snails of the Indian

ocean. J. Am. Oil Chem. Soc. 1998, 75, 1679−1681.

30) Kohn, A. J.; Saunders, P. R.; Wiener, S. Preliminary studies on

the venom of the marine snail Conus. Ann. N. Y. Acad. Sci. 1960, 90,

06−725.

31) Neves, J. L.; Lin, Z.; Imperial, J. S.; Antunes, A.; Vasconcelos,

(50) Shefer, S.; Cheng, F. W.; Dayal, B.; Hauser, S.; Tint, G. S.;

(

Salen, G.; Mosbach, E. H. A 25-hydroxylation pathway of cholic acid

biosynthesis in man and rat. J. Clin. Invest. 1976, 57, 897−903.

V.; Olivera, B. M.; Schmidt, E. W. Small molecules in the cone snail

arsenal. Org. Lett. 2015, 17, 4933−4935.

(32) Lin, Z.; Torres, J. P.; Ammon, M. A.; Marett, L.; Teichert, R.

W.; Reilly, C. A.; Kwan, J. C.; Hughen, R. W.; Flores, M.; Tianero, M.

D.; Peraud, O.; Cox, J. E.; Light, A. R.; Villaraza, A. J.; Haygood, M.

G.; Concepcion, G. P.; Olivera, B. M.; Schmidt, E. W. A bacterial

source for mollusk pyrone polyketides. Chem. Biol. 2013, 20, 73−81.

(33) Lee, Y. J.; Han, S.; Kim, S. H.; Lee, H. S.; Shin, H. J.; Lee, J. S.;

Lee, J. Three new cytotoxic steroidal glycosides isolated from Conus

pulicarius collected in Kosrae. Micronesia. Mar. Drugs 2017, 15,

No. 379.

(34) Ohgane, K.; Karaki, F.; Noguchi-Yachide, T.; Dodo, K.;

Hashimoto, Y. Structure-activity relationships of oxysterol-derived

pharmacological chaperones for Niemann-Pick type C1 protein.

Bioorg. Med. Chem. Lett. 2014, 24, 3480−3485.

043