Novel components of the human metabolome: The identification, characterization and anti-inflammatory activity of two 5-androstene tetrols

Article abstract of DOI:10.1016/j.steroids.2010.10.005

Two natural 5-androstene steroid tetrols, androst-5-ene-3β,7β, 16α,17β-tetrol (HE3177) and androst-5-ene-3α,7β,16α, 17β-tetrol (HE3413), were discovered in human plasma and urine. These compounds had significant aqueous solubility, did not bind or transactivate steroid-binding nuclear hormone receptors, and were not immunosuppressive in murine mixed-lymphocyte studies. Both compounds appear to be metabolic end products, as they were resistant to primary and secondary metabolism. Both were orally bioavailable, and were very well tolerated in a two-week dose-intensive toxicity study in mice. Anti-inflammatory properties were found with exogenous administration of these compounds in rodent disease models of multiple sclerosis, lung injury, chronic prostatitis, and colitis.

Full text of DOI:10.1016/j.steroids.2010.10.005

Steroids 76 (2011) 145–155
Contents lists available at ScienceDirect
Steroids
journal homepage: www.elsevier.com/locate/steroids
Novel components of the human metabolome: The identification,
characterization and anti-inflammatory activity of two 5-androstene tetrols
Clarence N. Ahlema,∗ , Theodore M. Pagea , Dominick L. Aucia , Michael R. Kennedya , Katia Manganob ,
Ferdinando Nicolettib, Yu Gea, Yujin Huanga, Steven K. Whitea, Sonia Villegasa, Douglas Conradc,
Angela Wangc, Christopher L. Readinga, James M. Frinckea
a Harbor Biosciences, Inc., 9171 Towne Centre Drive, Suite 180, San Diego, CA 92122, United States
b Department of Biomedical Sciences, School of Medicine, University of Catania, Catania, Italy
c San Diego Veterans Administration Healthcare System, 3350 La Jolla Village Dr., San Diego, CA 92161, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 August 2010
Received in revised form 13 October 2010
Accepted 13 October 2010
Available online 23 October 2010
Two natural 5-androstene steroid tetrols, androst-5-ene-3␤,7␤,16␣,17␤-tetrol (HE3177) and androst-5-
ene-3␣,7␤,16␣,17␤-tetrol (HE3413), were discovered in human plasma and urine. These compounds had
significant aqueous solubility, did not bind or transactivate steroid-binding nuclear hormone receptors,
and were not immunosuppressive in murine mixed-lymphocyte studies. Both compounds appear to
be metabolic end products, as they were resistant to primary and secondary metabolism. Both were
orally bioavailable, and were very well tolerated in a two-week dose-intensive toxicity study in mice.
Anti-inflammatory properties were found with exogenous administration of these compounds in rodent
disease models of multiple sclerosis, lung injury, chronic prostatitis, and colitis.
Keywords:
Androstene
Tetrol
Metabolism
Inflammation
Metabolome
Human
© 2010 Elsevier Inc. All rights reserved.
1. Introduction
It is well established that rodents are an endocrine species
whereas humans have evolved to utilize a locally controlled
Dehydroepiandrosterone (DHEA) is not a significant product
of the rodent adrenal gland [1–4], but exogenous DHEA exhibits
anti-inflammatory activity in a wide range of metabolic or inflam-
matory disease models in rodents [5,6]. Unfortunately, clinical trials
evaluating DHEA supplementation in patients with inflammatory
conditions have reported undesired estrogenic and androgenic side
effects, but not significant reductions in disease parameters [7–9].
These clinical outcomes may be related to poor oral bioavailability
and rapid metabolism through the sex steroid pathways and into
inactive conjugates [6,10–12].
The historical focus on androgens and estrogens has left much
of the DHEA metabolome relatively unexplored. Reports describ-
ing the ability of exogenous DHEA or DHEA-sulfate (DHEA-S) to
rejuvenate rodents [13] and regulate stem cells [14,15] as well as
causal evidence linking adrenal dysfunction with inflammatory dis-
eases [16] have generated a high level of academic and industrial
interest, but although there are numerous publications describing
the effects of DHEA, the precise role of the DHEA metabolome in
biological systems remains something of an enigma.
paracrine–autocrine hormone system [17,18]. Accordingly,
steroidogenesis is profoundly different between the two species.
Exogenous DHEA in rodents, but not humans, is typically metab-
olized into an array of highly oxidized androstenes [19,11].
Several investigators have published evidence that many of the
anti-inflammatory properties initially attributed to DHEA reside in
these more highly oxidized metabolites [20–28]. These metabolites
include androst-5-ene-3␤,17␤-diol (5-AED) and 7-hydroxy mem-
bers of the series, such as androst-5-ene-3␤,7␤,17␤-triol (␤AET).
Although 5-AED and ␤AET have substantial anti-inflammatory
activity in rodents [20,29,30], these molecules do not have suitable
pharmaceutical properties, e.g., they lack sufficient metabolic
stability to allow for efficient oral delivery and sustained blood
concentrations. We have developed synthetic versions of several
of these compounds in our effort to improve their pharmaceutical
properties [29,31,32].
During the course of characterizing the stability of our steroid
library to microsomal metabolism, we observed several com-
pounds that appeared to be completely resistant to metabolism.
These compounds were 5-androstenetetrols, which had been
prepared by chemical synthesis, and the observed stability led
to their further characterization. Here we report the discovery
of androst-5-ene-3␤,7␤,16␣,17␤-tetrol (HE3177) and androst-
∗
Corresponding author. Tel.: +1 858 320 2571; fax: +1 858 558 6470.
E-mail address: cahlem@harborbiosciences.com (C.N. Ahlem).
0039-128X/$ – see front matter © 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.steroids.2010.10.005

146
C.N. Ahlem et al. / Steroids 76 (2011) 145–155
O
OH
OH
OH
OH
CH3
CH3
CH3
H
H
H
H
CH3
CH3
H
H
CH3
H
H
H
H
H
H
HO
HO
HO
OH
OH
HO
OH
AED
DHEA
AET
β
CH3
CH3
CH3
OH
OH
CH3
H
CH3
CH3
H
H
H
H
H
H
OH
HO
OH
OH
HO
OH
HE3413
AET
α
HE3177
CH3
CH3
OH
CH3
OH
H
CH3
H
H
H
H
OH
HO
OH
HO
HE3412
HE3354
Fig. 1. Structures of selected DHEA metabolites.
5-ene-3␣,7␤,16␣,17␤-tetrol (HE3413) in human plasma and
urine.
structures). Tetrols were synthesized from commercially available
dehydroepiandrosterone (DHEA). Reagents and anhydrous sol-
vents were obtained from commercial sources and used without
further purification. Synthetic procedures ranged from seven steps
with an overall yield of ∼15% (HE3177) to a 9 step process with
an overall yield of ∼2.5% (HE3413). NMR spectra were acquired
in d4-methanol with a Bruker 400 MHz NMR spectrometer. Purity
was measured by HPLC using a 4.6 mm × 150 mm Agilent XDB-C18,
3.5 ␮m column (Agilent Technologies) at ambient temperature,
with a mobile phase of 10–90% acetonitrile in water, 0.1% triflu-
oroacetic acid (TFA) (gradient completed in 10 min, 1 mL/min).
Eluted compounds were monitored with a diode array detector
(Agilent Technologies) and an evaporative light scattering detector
(SEDEX 85 LT-ELSD, SEDERE Sas, France). The molecular mass
was determined by LC–MS/MS using an Xbridge Phenyl column
(2.1 mm × 150 mm, 3.5 ␮m, Waters, Beverly, MA) eluted with a
mobile phase gradient of acetonitrile in water, 0.1% formic acid,
with the column temperature maintained at 40 ◦C. The column
eluent was subjected to positive-mode electrospray ionization
(ES+), and the analytes were detected with a tandem quadrupole
mass spectrometer (Waters, Beverly, MA), monitoring the M+1
(−1 or 2H2O) mass transitions (323.4 > 305.4 > 287.4 amu).
We report preliminary studies on their pharmacokinetics and
toxicology, and demonstrate their oral bioavailability. These com-
pounds appear to be end products of metabolism, as they were
highly resistant to primary and secondary metabolism in vitro and
in vivo. Neither compound appears to interact with steroid binding
nuclear hormone receptors or peroxisome proliferator activated
receptor (PPAR), and in preliminary high-dose evaluations, nei-
ther elicited acute toxicity, sex-hormone effects, or indications
of immunotoxicity. Next, we tested these novel tetrols in sev-
eral rodent models of chronic inflammatory disorders, and found
them to possess potent anti-inflammatory activities. In addition to
their potential as pharmaceuticals, since both are natural products,
either may be suitable for use as a dietary supplement to delay or
prevent diseases associated with chronic inflammation, including
cancer, autoimmune disease and metabolic syndrome.
2. Materials and methods
All in vivo experiments were conducted in accordance with
national and local animal welfare regulations and with Institutional
Animal Care and Use Committee (IACUC) approvals.
Dimethylsulfoxide
(trimethylsilyl)trifluoroacetamide (MSTFA), methyl-tert-butyl
ether (MTBE), human sulfotransferase, human UDP-
glucuronosyltransferase, NADPH, glucose-6-phosphate,
(DMSO),
N-methyl-N-
2.1. Steroids, chemicals, and reagents
glucose-6-phosphate dehydrogenase, octanol, phosphate-buffered
saline, dehydroepiandrosterone (DHEA), androst-5-ene-3␤,17␤-
diol (5-AED), and dexamethasone were purchased from Sigma
Chemical Company (St. Louis, MO). Androst-5-ene-3␣,17␤-diol
(DHA, HE3209), 3␤,7␤-diol-androst-5-en-17-one (HE3129),
The tetrol compounds, androst-5-ene-3␣,7␤,16␣,17␤-tetrol
(HE3413),
androst-5-ene-3␤,7␤,16␣,17␤-tetrol
(HE3177),
androst-5-ene-3␤,7␣,16␣,17␤-tetrol (HE3354), androst-5-ene-
3␣,7␣,16␣,17␤-tetrol (HE3412), androst-5-ene-3␤,7␤,17␤-triol
(␤AET), and androst-5-ene-3␤,7␣,17␤-triol (␣AET) were synthe-
sized at Harbor Biosciences (San Diego, CA) (see Fig. 1 for chemical
3␤-ol-androst-5-ene-7,17-dione
(7-keto-DHEA),
androst-5-

C.N. Ahlem et al. / Steroids 76 (2011) 145–155
147
ene-3␣,16␣,17␤-triol (HE3411), androst-5-ene-3␣,7␤,17␤-triol
2.3. Biopharmaceutical characterization
(HE3593), and 3␣,7␤-diol-androst-5-en-17-one (HE3768) were
purchased from Steraloids (Newport, RI). ␤-Cyclodextrin sulfobutyl
ether (Captisol®) was purchased from CyDex, Inc. (Overland Park,
KS). Vehicles for oral administration of compounds were prepared
by Harbor Biosciences: 300 mg/mL ␤-cyclodextrin sulfobutyl ether
and 1 mg/mL carboxymethylcellulose (Spectrum, Gardena, CA),
9 mg/mL NaCl (Sigma, St. Louis, MO), 20 mg/mL polysorbate 80
(Spectrum, Gardena, CA), and 0.5 mg/mL phenol (Sigma, St. Louis,
MO).
2.3.1. Pharmacokinetics assessments in mice
Male CD-1 mice (6–8 weeks of age) received a single oral gav-
age of the test compound formulated as an aqueous solution in
␤-cyclodextrin sulfobutyl ether. Cohorts of 3 animals were sac-
rificed by CO2 asphyxiation and cardiac puncture prior to dosing
(t = 0), and at different times after dosing up to 8 h. Serum was
prepared and stored at −20 ◦C until analysis. Test compound con-
centrations in serum were measured by LC–MS/MS, and the mean
concentration of the cohort was used to report PK parameters.
2.3.2. Pharmacokinetics assessments in monkeys
2.2. Analytical methods for tetrol determinations
The absolute oral bioavailability of HE3413 was determined in
cynomolgus monkeys (Macaca fascicularis). HE3413, 10 mg/kg, in
an aqueous ␤-cyclodextrin sulfobutyl ether solution was admin-
istered both orally and intravenously to groups of 4 males (body
weight 3.7–4.5 kg). Blood was collected serially for 24 h (0, 0.25,
0.5, 1, 2, 4, 8, and 24 h for oral and 0, 0.05, 0.25, 0.5, 0.75, 1,
2, 4, 8, and 24 h for IV) after drug administration and analyzed
for by LC–MS/MS. The absolute oral bioavailability (% F) was
expressed as the dose mass adjusted ratio of the integrated drug
plasma concentration, AUC(0–infinity), for the oral relative to the IV
dose.
2.2.1. Structure determination using a gas
chromatography–mass spectrometry method
Human plasma and urine (1 mL) was extracted with 10 mL
MTBE, dried under nitrogen, and derivatized with MSTFA. The
derivative (1 ␮L) was analyzed on a Varian 3800 gas chromatograph
coupled to a 1200 L mass spectrometer in electron impact mode (EI)
with a Varian CP-Sil column (0.15 mm × 10 m) eluted with helium
at 1 mL/min (Varian, Palo Alto, CA). Tetrols were identified by a
combination of their retention time, mass transitions (610 > 520 @-
9 V, and 520 > 415 @-13 V), and by the ratio of these two transitions,
and by comparison to authentic standards, when applicable.
2.3.3. Microsomal metabolism
The test compounds (10 ␮M) were incubated with 0.25 mg
human liver microsomal extract (In Vitro Technologies), 10 mM
glucose-6-phosphate, 10 mM NADP, and 5 U glucose-6-phosphate
dehydrogenase in 0.5 mL PBS for 75 min at 37 ◦C. An equivalent
amount of androst-5-ene-3␤,17␤-diol was incubated separately as
an enzyme activity control. Steroids were extracted with 5 mL ethyl
acetate and analyzed by LC–MS.
2.2.2. Structure determination using a liquid
chromatography–mass spectrometry method
Human plasma (10 mL) was washed with 50 mL hexane. The
aqueous phase was extracted with 150 mL ethyl acetate, dried
under nitrogen, and derivatized with 500 ␮L of 50 mg/mL nicotinyl
chloride in pyridine at 80 ◦C for 1 h. The reaction was cooled to
room temperature and quenched with 1 mL 5% sodium bicarbonate.
The derivatized steroids were extracted with 10 mL MTBE, dried
under nitrogen, dissolved in 100 ␮L acetonitrile/water (45:65, v/v),
and analyzed on a Varian ProStar HPLC coupled to a Varian 1200 L
mass spectrometer in MS/MS mode (Varian, Palo Alto, CA), using a
Phenomenex Polar RP C18 column (2 mm × 250 mm, Phenomonex,
Torrance, CA) eluted with a water/acetonitrile gradient. Some anal-
yses were performed with two columns in tandem. Steroid tetrols
were identified by their 743 > 497 @-14 V mass transition and
their retention time. For the determination of tetrol concentrations
in biopharmaceutical assays (solubility, Caco-2 transport, in vitro
metabolism), tetrols were extracted with the specified volume
of ethyl acetate, dried under nitrogen, dissolved in 100 ␮L ace-
tonitrile/water (45/65), and analyzed by LC–MS/MS as above, but
without derivatization, and detected by their retention time and
287 > 269 @-7 V mass transition.
2.3.4. In vitro tetrol formation from 5-androstene precursors
Steroid procursors (10 ␮M) were incubated for 90 min at 37 ◦C
with 0.25 mg liver microsomal extract with 10 mM glucose-6-
phosphate, 10 mM NADP, and 10 U/mL glucose-6-phosphate dehy-
drogenase. The reaction mixtures were subsequently extracted
with ethyl acetate, derivatized with nicotinyl chloride, and ana-
lyzed using a high sensitivity tandem column LC–MS/MS method.
Peak areas were determined for responses with the appropriate
retention time of androstenetetrols and 743 > 497 amu mass tran-
sition.
2.3.5. Measurement of sulfation and glucuronidation
The susceptibility of test compounds to sulfation and glu-
curonidation was measured in vitro by comparing the disappear-
ance of test compounds in the presence and absence of activated
cofactor. Test compounds (10 ␮M) were incubated with human
liver S9 fraction (1 mg/mL) in 0.5 mL PBS containing 5 mM mag-
nesium chloride in the presence or absence of 100 ␮M adenosine
3ꢀ-phosphate-5ꢀ-phosphosulfate (for sulfation assays), or 800 ␮M
UDP-glucuronic acid (for glucuronidation assays). The assays were
incubated at 37 ◦C for 15 min (sulfation) or 1 h (glucuronidation).
5-AED (sulfation) and DHA (glucuronidation) were incubated in
parallel as controls. Following incubation, steroids were extracted
with 5 mL ethyl acetate and analyzed by LC–MS.
2.2.3. Pharmacokinetics determinations using a liquid
chromatography–mass spectrometry method
Serum or plasma samples were extracted with MTBE. The
organic phase was collected and evaporated to dryness, and
dissolved in HPLC mobile phase (water/acetonitrile, 90:10, v/v).
Samples were analyzed on a Waters Xbridge Phenyl column by
reversed-phase high-performance liquid chromatography (Agilent,
Palo Alto, CA and Leap Technologies, Carrboro, NC) coupled with a
tandem mass spectrometer (Waters, Beverly, MA). Concentrations
were determined from a calibration curve using Masslynx analysis
software (Waters, Beverly, MA). Analytes below the quantifiable
assay limit (BQL) were assigned values of zero for calculation of
pharmacokinetics parameters (WinNonlin 5.2; Pharsight, Moun-
tain View, CA).
2.3.6. Partition constant
Compound solubility in water was measured using Millipore
MultiScreen HTS solubility plates (Millipore, Billerica, MA) accord-
ing to the manufacturer’s instructions. Compounds were dissolved
in dimethyl sulfoxide, and diluted to 250 ␮M in water. The con-
centration of soluble compound in the filtrate was determined

148
C.N. Ahlem et al. / Steroids 76 (2011) 145–155
by LC–MS, and the log10 of the partition constant in PBS (pH
7.4)/octanol was calculated [33].
2.7. Anti-inflammatory assays
The anti-inflammatory activity of HE3177 and HE3413 was
assessed in rodent (SJL/J mice and Dark Agouti rat) models of
experimental autoimmune encephalomyelitis (EAE), experimental
autoimmune prostatitis (EAP), LPS induced lung injury, and dextran
sodium sulfate induced colitis (DSSIC).
2.3.7. Caco-2 membrane transport measurement
The apical to basolateral transport of tetrols and reference
compounds was measured in Caco-2 cells (In Vitro Technolo-
gies, Baltimore, MD) as described [34]. Test compounds were
dissolved in apical buffer at 10 ␮M, and samples of basolateral
buffer were taken at 0, 20, 40, 80, and 120 min, extracted with 5 mL
of ethyl acetate, and analyzed for the concentration of compound
by LC/MS–MS as described above.
2.7.1. SJL/J mouse EAE
Mice, 6–8 weeks of age (Charles River Italia, Udine, Italy) were
adapted (at least 7 days) and maintained in a pathogen-free vivar-
ium (University of Catania, School of Medicine, Catania, Italy), and
had access to sterilized food and water ad libitum. Relapsing and
remitting EAE was induced with a subcutaneous 200 ␮L injection
containing 75 ␮g proteolipid protein (PLP) as previously described
[36]. Body weight and disease symptoms were monitored daily
and scored (blinded): 0 = no illness; 1 = flaccid tail; 2 = moderate
paraparesis; 3 = severe paraparesis; 4 = moribund state; 5 = death.
Before immunization mice were randomized to each experimental
group (n = 7–8) and treated daily (6 days/week) with either 100 ␮L
of HE3413 (4 mg/kg), or HE3177 (4 mg/kg) by oral gavage. Treat-
ment was initiated at disease onset (12 days after immunization,
disease scores 2–3 in approximately 25% of mice) and continued
for 18 days. Mice were monitored for an additional 3 weeks after
treatment. A two-way ANOVA was used to determine statistically
significant differences in disease severity among the treatment
groups. The Mann–Whitney U test was used to determine the sig-
nificance between the tetrol- and vehicle-treated groups.
2.4. Nuclear hormone interactions
2.4.1. Nuclear receptor binding assessment
Assessment of binding activity for various nuclear receptors
was performed by homogeneous competition assays using the
PolarScreenTM fluorescence polarization system (InVitrogen, Carls-
bad, CA) as previously described [32]. The highest concentration
tested was 10,000 nM.
2.4.2. Nuclear receptor transactivation assays
Transactivation of androgen (AR), estrogen (ER␣, and ER␤),
and glucocorticoid (GR) receptors were measured as previously
described [32]. Human mineralocorticoid receptor (MR) and per-
oxisome proliferator-activated receptor (PPAR-␣, PPAR-␥ and
PPAR-␦) transactivation was measured with the Gene Blazer® ␤-
lactamase assay system (InVitrogen, Carlsbad, CA) following the
manufacturer’s instructions. The highest concentration tested was
10,000 nM.
2.7.2. Dark Agouti rat EAE
Female rats, 8–10 weeks old, 150–175 g (Scanbur AB, Sweden),
maintained under standard (non-specific pathogen free) laboratory
conditions (Department of Biomedical Sciences, Section of Gen-
eral Pathology, Catania, Italy), were adapted for at least 7 days
with free access to food and water. Rats were subcutaneously
immunized (day = 0) at the base of the tail with 200 ␮L of a homog-
enized emulsion containing 50 mg of whole DA rat spinal cord
in Complete Freund’s Adjuvant containing 10 mg/mL Mycobac-
terium tuberculosis (Difco, Detroit, USA). A progressive paralysis
resulted approximately 8–9 days after immunization, initially in
the tail, ascending up to the forelimbs. Body weight and disease
symptoms were monitored daily beginning on day 7, and severity
scored in a blinded manner according to the same scale used in
mice (see above). Rats were randomized to experimental groups
before immunization. When an animal’s clinical EAE score ranged
between 1 and 2 (days 10–12), once daily oral gavage was initiated
with either 125 ␮L of cyclodextrin vehicle (n = 10) or HE3413 in
cyclodextrin (10 mg/kg; n = 7), 6 days per week for 20 days. Animals
were subsequently monitored daily until day 34. Daily severity
scores from individual animals were used to determine group dif-
ferences. The Mann–Whitney U test was used to determine the
significance between the tetrol- and vehicle-treated groups.
2.5. Acute toxicity in mice
Experiments to determine the acute toxicity and estrogenic
potential were performed at MPI Research (Matawan, MI). Three
groups of 10 female Crl:CD1®(Icr) mice, approximately 6 weeks
of age, received 10 mL/kg twice daily oral gavage for 14 consec-
utive days of either 30% ␤-cyclodextrin sulfobutyl ether (vehicle
control) or 20 mg/mL HE3413 or HE3177 (400 mg/kg/day) as a solu-
tion in cyclodextrin. The mice were observed twice daily for clinical
signs of toxicity and weighed on days 1, 7, and 14. On the fifteenth
day, the animals were sacrificed, and subjected to gross necropsy.
Blood from 5 animals per group was used to measure hematological
parameters, and blood from the remaining 5 per group was used to
measure a limited panel of clinical chemistry parameters (empha-
sizing liver function tests). The liver, uterus, and adrenal glands
from each animal were weighed and evaluated for histopathology.
2.6. Immune suppression assessment by mixed lymphocyte
response
2.7.3. Murine experimental autoimmune prostatitis
The classic mixed lymphocyte reaction (MLR) was used to assess
immune suppressive potential essentially as previously described
[35]. Peripheral blood mononuclear cells (PBMC) were isolated
from three healthy donors (2 females, 1 male, between 18 and 65
years of age) after informed consent. Each of the three donors was
tested as responders in an allogeneic MLR assay that used allogenic
stimulator cells pretreated with mitomycin C (25 ␮g/mL for 30 min
at 37 ◦C in a 5% CO2 atmosphere and washed) in lieu of radiation.
Compounds were tested at 100 and 1000 ng/mL in cultures incu-
bated for six days. The data are expressed as mean SEM of the
normalized (% DMSO control) quadruplicate values obtained from
the three different responders.
Animals were adapted at least 7 days and kept under standard
(non-specific pathogen free) laboratory conditions (University of
Catania, Catania, IT) with free access to food and water. Male NOD
mice (n = 8 per group), 6–8 weeks old (Jackson Laboratories, ME)
were immunized with a subcutaneous injection of 300 ␮g CD-1
mouse prostate homogenate in CFA containing 1 mg/mL M. tubercu-
losis (Difco, Detroit, MI) [37]. A total volume of 150 ␮L administered
for immunization was divided between four sites: right and left
foot pad (25 ␮L); tail base and shoulder (50 ␮L). Animals received
a daily oral gavage of either 100 ␮L cyclodextrin vehicle or HE3413
(80 mg/kg) on days 14–30 after immunization. Mice were sacrificed
4 h after the final treatment, and prostates excised and snap-frozen

C.N. Ahlem et al. / Steroids 76 (2011) 145–155
149
in Tissue Tek (Miles Laboratories). Fifty cryostatic sections covering
the entire prostate were prepared, stained with H&E and scored by a
pathologist (blinded to treatment) for organized intraprostatic lym-
phomononuclear cell infiltrate nodules and reported as the number
of infiltrates/cross-section. The significance between groups was
tested using Student’s t test.
(ELS detector), and molecular mass (LC–MS/MS) are reported for
HE3177, HE3354, HE3412, and HE3413.
Androst-5-ene-3␤,7␤,16␣,17␤-tetrol
(HE3177),
mp
229.1–231.4 ◦C. 1HNMR (CD3OD, 400 MHz): ı 5.24 (s, 1H), 4.00 (m,
1H), 3.74 (d, 1H, J = 6.9 Hz), 3.42 (m, 1H), 3.36 (d, 1H, J = 6.3 Hz),
2.28–1.06 (m, 15H), 1.07 (s, 3H), 0.77 (s, 3H). HPLC: tR = 4.26 min
(98.7%). LC–MS/MS: tR = 3.84 min, 305.2 amu (M+1 −H2O).
Androst-5-ene-3␤,7␣,16␣,17␤-tetrol
(HE3354),
mp
2.7.4. LPS induced lung injury
202.1–204.4 ◦C. 1H NMR peaks (CD3OD, 400 MHz): ı 5.53 (d,
1H, J = 3.7 Hz), 4.03 (m, 1H), 3.73 (m, 1H), 3.47 (m, 1H), 3.42 (d, 1H,
J = 6.3 Hz), 2.33–2.18 (m, 2H), 1.88–1.28 (m, 11H), 1.21–1.08 (m,
2H), 1.00 (s, 3H), 0.76 (s, 3H). HPLC: tR = 4.84 min (100%, ELSD).
LC–MS/MS: not analyzed.
Female C57 black/6 mice (6–8 weeks old; Harlan, San Diego,
CA) were used in these studies. Animals were acclimated for 3
days prior to the experiment in a controlled laboratory environ-
ment (San Diego Veterans Administration Healthcare System, La
Jolla, CA) with free access to standard rodent chow and water.
Lipopolysaccharide (LPS) from Escherichia coli 055:B5 was pur-
chased from Sigma, St. Louis, MO. Animals received an oral gavage
of 100 ␮L of cyclodextrin vehicle (n = 7), 40 mg/kg HE3413 (n = 7),
or HE3177 (n = 8) formulated as a solution in cyclodextrin, 24 h and
1 h before LPS challenge. Challenge was performed by lightly anes-
thetizing the mice with isofluorane and then directly administering
50 ␮L LPS solution (1 mg/mL in sterile saline) into the trachea with
mice in a vertical position using a gel loading pipette under direct
observation through a medical otoscope. Forty-eight hours after the
LPS challenge, the animals were sacrificed, bronchoalveolar lavage
(BAL) samples were collected (BAL performed 3× using sterile PBS;
1.3 mL were typically recovered) and myeloperoxidase (MPO) lev-
els in lungs measured as previously described [38]. TNF␣ and IL-6
ELISA kits were purchased from Assay Designs (Ann Arbor, MI) and
used according to the manufacturer’s instructions. Significance was
determined using a two-sided Mann–Whitney exact test.
Androst-5-ene-3␣,7␣,16␣,17␤-tetrol
(HE3412),
mp
5.54 (d,
245–248.3 ◦C. 1H NMR peaks (CD3OD, 400 MHz):
ı
1H, J = 5.4 Hz), 4.02 (m, 2H), 3.70 (m, 1H), 3.43 (d, 1H, J = 6.1 Hz),
2.55 (m, 1H), 2.10 (m, 1H), 1.83–1.39 (m, 12H), 1.16 (m, 1H),
1.01 (s, 3H), 0.76 (s, 3H). HPLC: tR = 5.05 min (100%, ELSD) LC–MS:
tR = 6.58 min, 305.2 amu (M+1 −H2O).
Androst-5-ene-3␣,7␤,16␣,17␤-tetrol
(HE3413),
mp
220–222 ◦C. 1H NMR peaks: ı 5.23 (s, 1H), 4.00 (m, 2H), 3.78
(m, 1H), 3.37 (d, 1H, J = 6.2 Hz), 2.53 (m, 1H), 2.09 (m, 2H),
1.83–1.39 (m, 10H), 1.16 (m, 2H), 1.08 (s, 3H), 0.77 (s, 3H). HPLC:
tR = 4.72 min (98.6%, ELSD) LC–MS: tR = 3.81 min, 305.2 amu (M+1
−H2O).
3.2. Identification of tetrols in plasma and urine
Rat liver metabolism of androst-5-ene-3␤,7␤,17␤-triol (␤AET),
a natural hormone with anti-inflammatory activity [29] produced
an ensemble of unknown tetrol derivatives. We hypothesized 16-
hydroxylation could produce certain of these through the action of
CYP2C11, and focused on four structurally related tetrol deriva-
tives as possibilities: HE3177 (androst-5-ene-3␤,7␤,16␣,17␤-
tetrol), HE3354 (androst-5-ene-3␤,7␣,16␣,17␤-tetrol), HE3412
(androst-5-ene-3␣,7␣,16␣,17␤-tetrol) and HE3413 (androst-5-
ene-3␣,7␤,16␣,17␤-tetrol) (Fig. 1). These compounds were used
as analytical standards to determine their natural occurrence. Our
analysis revealed HE3413 and HE3177 in human plasma (Fig. 2A).
The specificity of the peaks corresponding to standards was con-
firmed by admixing the sample with a low (2 pg/mL, Fig. 2B) and a
high (5 pg/mL, Fig. 2C) concentration of the respective analytical
standards. An increased spectral intensity established the iden-
tity of two naturally occurring plasma compounds, HE3413 and
HE3177. The HE3354 and HE3412 tetrols were not detected. The
identity of HE3413 was confirmed using a second high sensitiv-
ity method with higher chromatographic resolution (Fig. 3). The
endogenous concentration was estimated to be 2 pg/mL. Additional
unidentified putative tetrols were detected in human plasma sam-
ple (Fig. 4). HE3413 was also observed un-conjugated in human
urine (data not shown). In a similar manner, HE3177 in human
plasma was discovered and confirmed by GC-MS/MS (data not
shown).
2.7.5. Dextran sodium sulfate (DSS)-induced colitis
Male C57BL/6 mice, 6–8 weeks old, 25–30 g (Charles River,
Wilmington, MA) were maintained in a pathogen-free ambient
environment with ad libitum sterilized food and water and adapted
for at least 7 days before induction. Prior to dextran sodium sul-
fate (DSS) treatment, the mice were randomized into groups of
five to receive either daily oral gavage of 100 ␮L of cyclodextrin
vehicle or HE3413 (10 and 40 mg/kg) for 5 or 15 days, starting
immediately prior to DSS administration. Colitis was induced with
DSS as previously described [39]. Briefly, mice were administered
3% DSS ad libitum in drinking water for 5 days, after which, mice
received DSS-free water. On days 5 and 15 the respective groups
were anesthetized with isoflurane for blood (serum) collection via
cardiac puncture, and immediately euthanized by CO2 asphyx-
iation. The colons were scored visually for diarrhea (0 = normal
fecal pellets, 1 = slightly loose feces, 2 = loose feces, and 3 = watery
feces) and colon inflammation (0 = normal, 1 = slight inflammation,
2 = moderate inflammation and/or edema, and 3 = heavy inflamma-
tion, edema and/or ulcerations). The colon samples were fixed in
buffered formalin and paraffin-embedded sections were prepared
and stained with Hematoxylin and Eosin (H&E) for histological
analysis of inflammation, edema, epithelial defects, gland atrophy,
hyperplasia and dysplasia. Haptoglobin levels in serum samples
were measured by ELISA (GenWay, San Diego, CA). Significant dif-
ferences were determined using a two-sided Student’s t-test.
3.3. Human liver microsome studies
3. Results
Androstene tetrols were found in in vitro metabolism studies
using human liver microsomes in which common naturally occur-
ring 5-androstenes produced tetrols (Table 1), including HE3177
and HE3413. Both the 3␣- and 3␤-hydroxy precursors were metab-
olized into both 3␣- and 3␤-hydroxy tetrols suggesting formation
through several different steroidogenic pathways. Greater amounts
of the tetrols were derived from triols, which required only a single
carbon oxidative reaction for tetrol formation.
3.1. Characterization of synthesized 5-androstenetetrols
5-Androstenetetrols used as analytical standards and/or phar-
macological agents were prepared by unambiguous chemical
synthesis, and were greater than 98% pure using HPLC analysis with
evaporative light scattering detection. The melting range (mp), pro-
ton NMR spectrum, HPLC retention time (tR) and percent purity

150
C.N. Ahlem et al. / Steroids 76 (2011) 145–155
MCounts
MCounts
6
1172-46-03 7-31-2008 12-08-38 PM.xms 743.0>497.0 [-14.0V] Filter
5
4
3
2
1
0
743.0>497.0 [-14.0V]
plasma 011209 1-12-2009 4-51-02 PM.xms 743.0>497.0 [-14.0V] Filtered
743.0>497.0 [-14.0V]
A
5
4
3
2
1
A
MCounts
1172-46-04 7-31-2008 2-11-11 PM.xms 743.0>497.0 [-14.0V] Filter
743.0>497.0 [-14.0V]
5
4
3
2
1
B
0
GCounts
743.0>497.0 [-14.0V]
1172-92-02 1-13-2009 2-00-12 PM.xms 743.0>497.0 [-14.0V] Filtered
B
1.25
1.00
0.75
0.50
0.25
0.00
0
MCounts
1172-46-06 7-31-2008 3-11-13 PM.xms 743.0>497.0 [-14.0V] Filter
743.0>497.0 [-14.0V]
HE3177
6
5
4
3
2
1
0
HE3354
HE3412
C
HE3413
9
10
11
12
13
minutes
20.0
22.5
25.0
27.5
minutes
Fig. 2. Identification of androstene tetrols in human plasma using LC–MS/MS.
Plasma was extracted with ethyl acetate, and steroids were derivatized with
nicotinyl chloride, and identified as tetrols by LC–MS/MS. (A) LC–MS/MS chro-
matogram of tetrols in human plasma, (B) human plasma spiked with 2 pg/mL tetrol
standards for HE3177, HE3354, HE3412, and HE3413, (C) human plasma spiked with
5 pg/mL tetrol standards of HE3177, HE3354, HE3412, and HE3413.
Fig. 3. Identification of androstene tetrols in human plasma using a modified high
sensitivity tandem column LC–MS/MS method. (A) LC–MS/MS chromatogram of
nicotinyl chloride derivatized tetrols in human plasma analyzed with a tandem col-
umn chromatographic system, (B) LC–MS/MS chromatogram of nicotinyl-HE3413
standard.
3.4.2. Metabolism
3.4. Pharmacokinetics and metabolism
HE3177 and HE3413 appear to resist primary metabolism since
neither was transformed by mouse or human liver microsomes
(Table 3). Similar metabolic resistance was found for HE3354 and
HE3412 (data not shown). This outcome was in sharp contrast to
results found with 5-AED and ␤AET. The in vitro metabolic stability
of HE3177 and HE3413 was reflected in a low abundance of oxi-
dized (hydroxy to keto) and epimerized metabolites in mouse and
monkey PK studies (data not shown). HE3177 and HE3413 were not
susceptible to the activity of CYP19 (aromatase) under conditions
that completely convert testosterone to estradiol (data not shown).
3.4.1. Pharmacokinetics
The pharmacokinetics of HE3177 and HE3413 were assessed
in mice following oral gavage of 40 mg/kg aqueous solutions and
results were compared to 5-AED (Fig. 5). Serum levels were ele-
vated relative to the endogenous plasma levels found in humans
(∼2 pg/mL) and persisted relative to 5-AED. HE3413 pharmacoki-
netics parameters in cynomolgus monkeys are given in Table 2. The
absolute oral bioavailability (% F) was 15.8%.
Table 1
Metabolic conversion of androst-5-ene precursors to tetrols by human liver microsomes.
Precursor structure
HE3412 (3␣, 7␤, 16␣, 17␤)
HE3413 (3␣, 7␤, 16␣, 17␤)
HE3354 (3␤, 7␣, 16␣, 17␤)
HE3177 (3␤, 7␤, 16␣, 17␤)
422
115
0
3␤-ol-androst-5-en-17-one
Androst-5-ene-3␤,17␤-diol
0
0
0
0
0
49
43
0
0
0
0
94
80
4199
88
0
0
3␤-ol-androst-5-ene-7,17-dione
3␤,7␤-Diol-androst-5-en-17-one
Androst-5-ene-3␤,7␣,17␤-triol
Androst-5-ene-3␤,7␤,17␤-triol
3␣-ol-androst-5-en-17-one
0
0
1090
780
34
192
546
0
47
0
Androst-5-ene-3␣,17␤-diol
51
0
3␣,7␤-Diol-androst-5-en-17-one
Androst-5-ene-3␣,7␤,17␤-triol
Androst-5-ene-3␣,16␣,17␤-triol
274
2058
19
201
2167
431
278
895
21
0
0
0
Steroid procursors (10 ␮M) were incubated for 90 min at 37 ◦C with 1 mg/mL human liver microsomal extract with 10 mM glucose-6-phosphate, 10 mM NADP, and 10 U/mL
glucose-6-phosphate dehydrogenase. After the incubation, the reaction mixtures were extracted with ethyl acetate, derivatized with nicotinyl chloride, and analyzed using
a high sensitivity tandem column LC/MS–MS method. Peak areas were determined for peaks with appropriate retention times androstenetetrols and 743 > 497 amu mass
transitions; values represent the areas of each peak.

C.N. Ahlem et al. / Steroids 76 (2011) 145–155
151
kCoun7t4s3.0>497.0 [-14.0V]
1172-41-1 7-3-2008.xms743.0>497.0 [-14.0V] Filtered
Compound
HE3177
HE3413
Cmax, ng/mL
571
1,457
Tmax, hr
0.5
0.5
AUC(0-8), ng*hr/mL
765
2,181
150
5,000
HE3413
HE3177
AED
500
50
5
100
50
0.5
0
1
2
3
4
5
6
7
8
9
0
Time (h)
MCoun7t4s3.0>497.0 [-14.0V]
1172-41-2 7-4-2008.xms743.0>497.0 [-14.0V] Filtered
HE3412
12.5
10.0
7.5
HE3177
Fig. 5. Pharmacokinetics of the 5-androstenetetrols, HE3413, HE3177, and 5-AED in
CD-1 mice. Male CD-1 mice received a single oral gavage of HE3413, HE3177 (aque-
ous solutions), or 5-AED (aqueous solution in cyclodextrin). Cohorts of 3 animals
were sacrificed at each time point. Steroid concentrations in serum were measured
by LC–MS/MS.
HE3354
HE3413
oxidized members of this structure class (Table 3), but the solubility
of HE3413 was surprisingly high, 28 mM. The low partition constant
was apparently reflected in relatively poor transport across Caco-2
cells (Table 3).
5.0
2.5
3.4.4. Nuclear hormone receptor interactions
0.0
Nuclear receptor binding experiments demonstrated that
HE3413 and HE3177 do not competitively bind any of the steroid
receptors, AR, ER␣, ER␤, PR, or GR, and did not transactivate the
nuclear hormone receptors for AR, ER␣ or ER␤, GR, MR, or PPAR-
␣ (Transactivation EC50 values were all greater than 10,000 nM,
except for HE3413 PPAR-␣ EC50 = 4000 nM). These results suggest
that HE3413 and HE3177 do not exert their activity through classic
steroid hormone nuclear receptor binding or through PPAR-␣/␥/␦.
8
9
10
11
12
13
14
minutes
Fig. 4. Unidentified (putative) tetrols in human plasma. LC–MS/MS chro-
matogram of human plasma, indicating the presence of numerous unidentified
androstenetetrols (top panel); LC–MS/MS chromatogram of androstenetetrol stan-
dards (bottom panel).
Tetrols appear to be very resistant to secondary metabolism, a prop-
erty that is presumably related to their intrinsically high water
solubility. Their metabolic stability probably contributes to their
slow rate of clearance compared to the more readily conjugated di-
and tri-oxygenated C-19 androstene DHEA metabolites.
3.5. Systemic toxicity in mice
The systemic toxicity and estrogenicity of HE3413 and HE3177
(400 mg/kg) were assessed in CD-1 female mice. Both compounds
were well tolerated. There were no clinical signs of toxicity or
changes in body weight. There were no significant treatment effects
on organ weights, histopathology (adrenal glands, liver, uterus),
hematology parameters, and blood chemistry (data not shown).
The absence of peroxisome proliferation in vivo was consistent with
3.4.3. Partition constant and Caco2 transport
The octanol:PBS partition constants (Log D) of HE3413 and
HE3177 were similar, and considerably lower than diols and triols.
The tetrols, as expected, were more water soluble relative to less
Table 2
HE3413 pharmacokinetics parameters (SD) in cynomolgus monkeys.
Route of administration
Cmax (ng/mL)
Tmax (h)
AUC(0–24) (nga h/mL)
AUC(0–infinity) (nga h/mL)
T1/2 (h)
IV
18,609 (4,438)
292 (167)
0.05 (0.0)
1.5 (0.6)
5,881 (454)
877 (625)
5,888 (452)
929 (618)
Oral-gastric
7.5*
a
Harmonic mean.
Table 3
Comparison of biopharmaceutical properties of HE3177 and HE3413 with 5-AED and ␤AET.
Compound
Solubilitya
Log Db
Metabolismc
Sulfationd
Glucuronidatione
Caco-2f
5-AED
␤AET
HE3177
HE3413
5
164
1,740
28,900
3.3
1.8
0.53
0.43
3.89
0.583
0
9.75
ND
0
1.02
ND
0
ND
21.5
1.57
2.74
0
0
0
ND: not determined.
a
Solubility (␮M) in H2O at 22 ◦C, pH 7.4.
b
c
d
e
f
Log of the distribution coefficient in octanol/phosphate-buffered saline pH 7.4 at 22 ◦C.
Metabolism of compound by human liver microsomes in nanomoles compound metabolized/mg microsomal protein/h.
Sulfation of compound by human liver S9 fraction expressed in nanomoles compound sulfated/mg protein/h.
Glucuronidation of test compound by human liver S9 fraction expressed as nanomoles of compound glucuronidated per mg protein/h.
Apical to basolateral transport by Caco-2 cells, expressed as picomoles transported/well/h.

152
C.N. Ahlem et al. / Steroids 76 (2011) 145–155
the in vitro results, which again differentiates the pharmacology of
these tetrols from those members of this structure class with lower
oxidation states.
Vehicle
HE3413
HE3177
end Tx
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
A
Tx
3.6. Immune suppression assessment by mixed lymphocyte
reaction
The immunosuppressive potential of HE3177 and HE3413 was
assessed using an allogeneic MLR. Blood was obtained from three
individual donors (1 male, 2 females) and each of two concen-
trations (100 and 1000 ng/mL) tested in quadruplicate. Neither
concentration of either tetrol suppressed lymphocyte prolifer-
ation (high concentration: 102% 14 and 116% 18 of vehicle
control respectively). In contrast, dexamethasone (1 ␮M) signif-
icantly (p = 0.0004) suppressed (>85%) proliferation in this same
human MLR assay.
*
*
0
10
20
30
40
50
60
Post-Induction of Disease (Days)
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
B
Vehicle
3.7. Anti-inflammatory activity of HE3177 and HE3413 in mouse
inflammation models
HE3413
3.7.1. Experimental autoimmune encephalomyelitis (EAE)
HE3413 was assessed in both the female SJL/J mouse and
the Dark Agouti (DA) rat EAE models of neuro-inflammation
resembling multiple sclerosis. In both models, disease scores in
tetrol-treated animals were significantly (p < 0.0001) reduced com-
pared to controls (Fig. 6A). Mice received daily doses of HE3177 or
HE3413 (4 mg/kg) or vehicle beginning on day 12 (first disease inci-
dence) and continuing through day 33. Cumulative disease index
(CDI) for this period was reduced in the tetrol-treated groups com-
pared to vehicle-treated animals. ANOVA analysis showed that CDI
from groups receiving HE3177 and HE3413 were significantly dif-
ferent from those receiving vehicle (p < 0.005). Benefit appeared to
be sustained for greater than three weeks after cessation of dosing.
In the DA rat model, where only HE3413 was tested (Fig. 6B), the
compound significantly (p = 0.018) reduced CDI and blunted peak
relapse (day 29) compared to vehicle (CDI 1.1 1.1 versus 2.0 0.0
respectively; p = 0.02).
0
10
20
30
40
Post-Induction of Disease (Days)
Fig. 6. Efficacy of the 5-androstenetetrols HE3413 and HE3177 in experimental
autoimmune encephalomyelitis (EAE). (A) EAE was induced in female SJL/J mice
on day 0 as in our previous studies [44]. Mice received daily oral gavage of vehicle,
HE3413 (4 mg/kg), or HE3177 (4 mg/kg) administered at the first disease incidence
(day 12). Tx indicates the beginning and end of therapy. *Significantly different from
vehicle-treated group (Mann–Whitney test; p < 0.0001). (B) Female DA rats received
s.c. injections of 0.2 mL (50 mg) of homogenized emulsion of whole DA rat spinal
cord in complete Freund’s Adjuvant on day 1. Rats received daily oral gavage of
vehicle (0.125 mL; n = 10) or HE3413 (10 mg/kg; n = 7) (6 days per week) beginning
when animals developed a clinical score of 1–2 (∼days 10–12). +Significant dif-
ference between HE3413- and vehicle-treated groups total daily disease scores of
individual animals (Mann–Whitney test; p < 0.02).
3.7.2. Experimental autoimmune prostatitis (EAP)
NOD mice develop a dramatic inflammatory, non-infectious,
prostatitis upon autoimmunization with gland extracts [40]. This
EAP model has been used to test agents for potential activity
in human chronic prostatitis [37]. Daily treatment (days 14–30)
with HE3413 (80 mg/kg) provided clinical benefit in this model
as determined by a significant (p = 0.034) decrease in the num-
ber of infiltrating lymphocyte nodules per prostate tissue section
(2.23 0.65 in the vehicle group versus 1.50 0.58 in the HE3413-
treated group) on day 30 (data not shown).
solved infection, chronic non-productive inflammation and a
pro-inflammatory cycle that resists resolution [43]. HE3413 was
evaluated in the DSS-induced colitis mouse model. The published
literature indicates that in this model, five days of DSS exposure
typically produces a colitis that continues well beyond 30 days
[39], and in our experiments colitis was readily observed in colon
samples assessed for diarrhea and inflammation on days 5 and 15
(Fig. 7). While no significant treatment effects were observed on day
5, HE3413 (10 and 40 mg/kg) significantly suppressed DSS-induced
diarrhea (p = 0.03) and colon inflammation (p = 0.04) on day 15 in a
dose-dependent manner. Histological analysis of the H&E stained
colon segments demonstrated that HE3413 suppressed edema,
leukocytic infiltration, epithelial defects, and glandular atrophy
(Fig. 7D). In addition, a dose-dependent suppression of serum hap-
toglobin, an inflammatory acute phase protein associated with
tissue deterioration in this condition [39], was observed on day 15
(Fig. 7C).
3.7.3. LPS-induced lung injury
Neutrophil mediated inflammation is implicated in the tis-
sue damage and fibrosis associated with several forms of chronic
obstructive pulmonary disease, and has been suggested as a
potential target for therapeutic action [41]. The murine model of
LPS-induced lung injury has been proposed to model neutrophil-
mediated lung inflammation and tissue damage [42]. Treatment
with either 40 mg/kg HE3413 or HE3177 significantly decreased
the myeloperoxidase activity in BAL fluid compared to the vehicle
control group (44%, p = 0.038 and 51%, p = 0.014 respectively). The
levels of TNF␣ and IL-6 in BAL fluid were also decreased approx-
imately 50%, but did not achieve statistical significance (data not
shown).
4. Discussion
3.7.4. DSS-induced colitis
Inflammatory bowel diseases (IBD) such as ulcerative colitis
(UC) and Crohn’s disease have been associated with unre-
A survey of human plasma revealed an ensemble of uniden-
tified tetrols, which we investigated starting with the synthesis
of four structurally related derivatives of the 7-hydroxy 5-

C.N. Ahlem et al. / Steroids 76 (2011) 145–155
153
Fig. 7. Activity of HE3413, in the mouse DSS-induced colitis model of IBD. Colitis was induced in C57BL/6 male mice with 3% (w/v) dextran sodium sulfate (DSS) in drinking
water initiated on days 1–5. The mice received daily oral gavage of vehicle or HE3413 (10 and 40 mg/kg) beginning on day 1. Groups of mice (n = 5) were sacrificed on day
5 and 15. Colons were removed and visually scored for the degree of (A) diarrhea and (B) colon inflammation. (C) Haptoglobin levels were measured in serum by ELISA.
*Statistically different from vehicle control mean (p < 0.05; Student’s t test). (D) Histological analysis of one representative colon sample (H&E staining) from each treatment
group. Vehicle sample shows moderate colitis with epithelial attenuation (arrow), mucosal inflammation and glandular loss (*), and ectasia (#). All results are representative
of two experiments that had similar results.
androstenetetrols (HE3413, HE3177, HE3354, and HE3412). Of
these tetrols, the 7␤-hydroxy epimers, HE3413 and HE3177, but
not the analogous 7␣-epimers, HE3354 and HE3412 were detected
in human plasma. Human liver microsomes generated both the
7␣- and 7␤-hydroxy epimers from common 5-androstene pre-
cursors, such as DHEA, 5-AED, and DHA, which suggests that
both epimers are produced, but the 7␣-epimer, reported to be
the metabolic precursor to the 7␤-epimer, is rapidly and com-
pletely converted to the 7␤ form by action of 11␤-hydroxysteroid
dehydrogenase (11␤-HSD) [44,45]. Thus, HE3413 and HE3177
were discovered and identified as new members of the human
metabolome. Both are highly resistant to secondary metabolism
and are orally bioavailable. Neither interacts with steroid binding
nuclear hormone receptors or PPAR, nor are they metabolized into
sex steroids. Although they have anti-inflammatory activity, they
are not immunotoxic, and do not appear to be immunosuppressive
in the manner of a glucocorticoid, although a thorough assess-
ment of potential immunosuppressive activity was not conducted.
In rodents, their potent anti-inflammatory activities are reminis-
cent of several other 5-androstene metabolites of DHEA [22,29].
Although poorly transported in the Caco-2 assay, in mice both
HE3413 and HE3177 had greater Cmax and AUC values following

154
C.N. Ahlem et al. / Steroids 76 (2011) 145–155
oral administration when compared to 5-AED. This is consistent
with their resistance to primary and secondary metabolism in vitro
in contrast to the high metabolic susceptibility displayed by 5-AED
and ␤AET. Oral HE3413 administration to cynomolgus monkeys
resulted in a sustained plasma drug exposure profile.
ing to speculate that they may also be the tetrol target. The
murine models used in the present studies all possess a significant
inflammatory macrophage component to their pathology [39,43].
However, other cell types, such as lymphocytes [56], hepatocytes
[57], adipocytes [58], endothelial cells [59] may also be involved
given the broad spectrum of DHEA activities. While a dedicated
receptor has not been identified for the 7-hydroxy C-19 steroids, an
ethynylated androstenetriol, HE3286, is reported to attenuate the
NF-␬B, p38, JNK, and ERK-2 pro-inflammatory pathways [32,47].
We are actively investigating the possibility that the molecularly
related tetrols have a similar mode of action.
To our knowledge, this is the first report of the natural existence
of the 5-androstene tetrols, HE3413 and HE3177 in human tissues.
Their potent anti-inflammatory properties are described and their
target tissue and mode of action inferred from that of known molec-
ularly related agents. The discovery of these remarkably stable,
anti-inflammatory components of the DHEA metabolome intro-
duces them as naturally occurring agents that may be useful for
prevention or resolution of chronic inflammatory conditions.
The discovery that these novel tetrols are highly resistant to
metabolism in their natural form was unexpected because nat-
ural C-19 steroids generally are highly susceptible to metabolic
transformation, and have very poor oral bioavailability, proper-
ties that limit their clinical usefulness [20]. We have previously
explored various strategies to overcome these limitations. For
example, we have developed a depot formulation of AED (an ER␤
biased agonist) that maintains serum levels of this hormone that
are associated with biological activity in humans [46]. Our expe-
rience with immune regulating hormones of the 5-androstene
series has been that 17␣-substitution to prevent oxidation by 17␤-
hydroxysteroid dehydrogenase frequently results in the loss of
biological activity, although 17␣-ethynylation was implemented
with ␤AET with surprisingly good results [21,29]. The result-
ing 17␣-ethynyl-5-androsten-3␤,7␤,17␤-triol (HE3286) retained
broad-based biological activity and has not been immunosuppres-
sive in vivo or in vitro [24,47]. In the present studies, we tested and
confirmed the anti-inflammatory activity of two tetrols in rodent
models of acute and chronic inflammation.
These orally bioavailable, metabolically stable, biologically
active 5-androstene tetrols pose the possibility that a natural
product in the human metabolome may be able to assert pharma-
ceutically relevant anti-inflammatory activity in target tissues.
Pharmacokinetics, drug metabolism and toxicology observa-
tions provided rationale and incentive for evaluation of these
newly discovered C-19 steroid tetrols in mouse inflammation mod-
els. Anti-inflammatory activity was observed in the LPS-induced
lung injury and DSS-induced colitis models, where oral admin-
istration was associated with reductions in pro-inflammatory
cells and cytokines as well tissue inflammation (respectively).
Anti-inflammatory activity was evidenced by reduced cellular
infiltrates was also observed in experimental autoimmune pro-
statitis, and the symptoms of EAE were suppressed in the female
SJL mouse MS model. In the later model, benefit was observed
when administration was initiated after disease onset that suggests
an anti-inflammatory and/or neuroprotective effect of the agent.
These newly profiled androstenetetrols possess a highly potent
anti-inflammatory activity that was observable at 4 mg/kg in the
EAE model, compared to ␤AET, which required a much higher dose
(120 mg/kg injected subcutaneously) to achieve a similar biological
effect [24].
Disclosure statement
Authors Ahlem, Page, Acui, Kennedy, Ge, Huang, White, Villegas,
Reading, and Frincke acknowledge potential conflict of interest as
employees of Harbor Biosciences, Inc. Authors Mangano, Nicoletti,
Conrad, and Wang declare no conflict of interest.
Role of the funding source
All experimental work reported in this article was sponsored
by Harbor Biosciences, Inc. Author representatives from Harbor
Biosciences participated in study design, data analysis and inter-
pretation, report writing, and the decision to publish.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.steroids.2010.10.005.
References
[1] Cutler Jr GB, Glenn M, Bush M, Hodgen GD, Graham CE, Loriaux DL. Adrenar-
che: a survey of rodents, domestic animals, and primates. Endocrinology
1978;103:2112–8.
[2] Hobe G, Hillesheim HG, Schon R, Reddersen G, Knappe R, Bannasch P, et al. Sex
differences in dehydroepiandrosterone metabolism in the rat: different plasma
levels following ingestion of DHEA-supplemented diet and different metabolite
patterns in plasma, bile and urine. Horm Metab Res 1994;26:326–9.
[3] Wichmann U, Wichmann G, Krause W. Serum levels of testosterone precur-
sors, testosterone and estradiol in 10 animal species. Exp Clin Endocrinol
1984;83:283–90.
[4] Belanger B, Belanger A, Labrie F, Dupont A, Cusan L, Monfette G. Comparison of
residual C-19 steroids in plasma and prostatic tissue of human, rat and guinea
pig after castration: unique importance of extratesticular androgens in men. J
Steroid Biochem 1989;32:695–8.
[5] Dillon JS. Dehydroepiandrosterone, dehydroepiandrosterone sulfate and
related steroids: their role in inflammatory, allergic and immunological dis-
orders. Curr Drug Targets Inflamm Allergy 2005;4:377–85.
[6] Svec F, Porter JR. The actions of exogenous dehydroepiandrosterone in experi-
mental animals and humans. Proc Soc Exp Biol Med 1998;218:174–91.
[7] Giltay EJ, van Schaardenburg D, Gooren LJ, von Blomberg BM, Fonk JC, Touw DJ,
et al. Effects of dehydroepiandrosterone administration on disease activity in
patients with rheumatoid arthritis [letter]. Br J Rheumatol 1998;37:705–6.
[8] Sirrs SM, Bebb RA. DHEA: panacea or snake oil? Can Fam Physician
1999;45:1723–8.
[9] Raven PW, Hinson JP. Dehydroepiandrosterone (DHEA) and the menopause:
an update. Menopause Int 2007;13:75–8.
[10] Buster JE, Casson PR, Straughn AB, Dale D, Umstot ES, Chiamori N, et al. Post-
menopausal steroid replacement with micronized dehydroepiandrosterone:
preliminary oral bioavailability and dose proportionality studies. Am J Obstet
Gynecol 1992;166:1163–8 [discussion 8–70].
There are descriptions of C-18 steroid tetrols in humans in the
published literature [48–53], but we could find no prior reference
to C-19 tetrols in humans, perhaps because they are present at such
low plasma concentrations. This suggests that the tetrols described
here may be intracrine members of the human metabolome. While
others have described the chemical synthesis or natural occurrence
of certain C-18 tetrols in human tissues [50–54], their biological
relevance was not explored. The anti-inflammatory properties of
DHEA, 5-AED and ␤AET have been demonstrated in animal models
[5,6,22,24,30]. Rodents readily metabolize exogenous androstenes
into more highly oxidized compounds, and one such compound,
HE3177, has been previously described in rat liver digests in vitro
[55]. Thus, these naturally occurring androstene tetrols may con-
tribute to biological activity observed from administration of
androstenes. We have yet to explore the relationship between
tetrol structure and anti-inflammatory activity.
The cellular and molecular targets relevant to the biologi-
cal activity of HE3413 and HE3177 remain unknown. Since our
recent studies have implicated macrophages as the target cell
in the androstenetriol biological mechanism [32,47], it is tempt-
[11] Fitzpatrick JL, Ripp SL, Smith NB, Pierce Jr WM, Prough RA. Metabolism of
DHEA by cytochromes P450 in rat and human liver microsomal fractions. Arch
Biochem Biophys 2001;389:278–87.

C.N. Ahlem et al. / Steroids 76 (2011) 145–155
155
[12] Leblanc M, Labrie C, Belanger A, Candas B, Labrie F. Bioavailability and phar-
macokinetics of dehydroepiandrosterone in the cynomolgus monkey. J Clin
Endocrinol Metab 2003;88:4293–302.
[36] Ahlem C, Auci D, Mangano K, Reading C, Frincke J, Stickney D, et al. HE3286: a
novel synthetic steroid as an oral treatment for autoimmune disease. Ann N Y
Acad Sci 2009;1173:781–90.
[13] Poynter ME, Daynes RA. Restoration of immunocompetence in aging and other
inflammatory disease states by dehydroepiandrosterone-3␤-sulfate, an activa-
tor of the peroxisome proliferator-activated receptor alpha (PPAR␣). In: Kalimi
MY, Regelson W, editors. Dehydroepiandrosterone (DHEA): Biochemical, Phys-
iological and Clinical Aspects. Berlin; New York: Walter de Gruyter; 1999. p.
65–117.
[14] Suzuki M, Wright LS, Marwah P, Lardy HA, Svendsen CN. Mitotic and neurogenic
effects of dehydroepiandrosterone (DHEA) on human neural stem cell cultures
derived from the fetal cortex. Proc Natl Acad Sci USA 2004;101:3202–7.
[15] Risdon G, Kumar V, Bennett M. Differential effects of dehydroepiandros-
terone (DHEA) on murine lymphopoiesis and myelopoiesis. Exp Hematol
1991;19:128–31.
[16] Roberge C, Carpentier AC, Langlois MF, Baillargeon JP, Ardilouze JL, Maheux P,
et al. Adrenocortical dysregulation as a major player in insulin resistance and
onset of obesity. Am J Physiol Endocrinol Metab 2007;293:E1465–78.
[17] Labrie F, Belanger A, Cusan L, Candas B. Physiological changes in dehy-
droepiandrosterone are not reflected by serum levels of active androgens and
estrogens but of their metabolites: intracrinology. J Clin Endocrinol Metab
1997;82:2403–9.
[18] Labrie F, Luu-The V, Belanger A, Lin SX, Simard J, Pelletier G, et al. Is dehy-
droepiandrosterone a hormone? J Endocrinol 2005;187:169–96.
[19] Marwah A, Marwah P, Lardy H, Ergosteroids VI. Metabolism of dehy-
[37] Penna G, Amuchastegui S, Cossetti C, Aquilano F, Mariani R, Sanvito F, et al.
Treatment of experimental autoimmune prostatitis in nonobese diabetic
mice by the vitamin D receptor agonist elocalcitol. J Immunol 2006;177:
8504–11.
[38] Krawisz JE, Sharon P, Stenson WF. Quantitative assay for acute intestinal inflam-
mation based on myeloperoxidase activity. Assessment of inflammation in rat
and hamster models. Gastroenterology 1984;87:1344–50.
[39] Melgar S, Karlsson A, Michaelsson E. Acute colitis induced by dextran sulfate
sodium progresses to chronicity in C57BL/6 but not in BALB/c mice: correlation
between symptoms and inflammation. Am J Physiol Gastrointest Liver Physiol
2005;288:G1328–38.
[40] Rivero VE, Cailleau C, Depiante-Depaoli M, Riera CM, Carnaud C. Non-obese
diabetic (NOD) mice are genetically susceptible to experimental autoimmune
prostatitis (EAP). J Autoimmun 1998;11:603–10.
[41] Kim S, Nadel JA. Role of neutrophils in mucus hypersecretion in COPD and
implications for therapy. Treat Respir Med 2004;3:147–59.
[42] Asti C, Ruggieri V, Porzio S, Chiusaroli R, Melillo G, Caselli GF.
Lipopolysaccharide-induced lung injury in mice. I. Concomitant evalua-
tion of inflammatory cells and haemorrhagic lung damage. Pulm Pharmacol
Ther 2000;13:61–9.
[43] Torres MI, Rios A. Current view of the immunopathogenesis in inflamma-
tory bowel disease and its implications for therapy. World J Gastroenterol
2008;14:1972–80.
droepiandrosterone by rat liver in vitro:
a liquid chromatographic-mass
spectrometric study. J Chromatogr B Biomed Sci Appl 2002;767:285–99.
[20] Auci DL, Ahlem C, Li M, Trauger R, Dowding C, Paillard F, et al. The immunobi-
ology and therapeutic potential of androstene hormones and their synthetic
derivatives: novel anti-inflammatory and immune regulating steroid hor-
mones. Mod Asp Immunobiol 2003;3:64–70.
[21] Auci D, Kaler L, Subramanian S, Huang Y, Frincke J, Reading C, et al. A new
orally bioavailable synthetic androstene inhibits collagen-induced arthritis in
the mouse. Ann N Y Acad Sci 2007;1110:630–40.
[22] Auci D, Nicoletti F, Mangano K, Pieters R, Nierkens S, Morgan L, et al. Anti-
inflammatory and immune regulatory properties of 5-androsten-3{beta},
17{beta}-diol (HE2100), and synthetic analogue HE3204: implications for
treatment of autoimmune diseases. Ann N Y Acad Sci 2005;1051:730–42.
[23] Offner H, Firestein GS, Boyle DL, Pieters R, Frincke JM, Garsd A, et al. An orally
bioavailable synthetic analog of an active dehydroepiandrosterone metabo-
lite reduces established disease in rodent models of rheumatoid arthritis. J
Pharmacol Exp Ther 2009;329:1100–9.
[44] Muller C, Pompon D, Urban P, Morfin R. Inter-conversion of 7alpha- and 7beta-
hydroxy-dehydroepiandrosterone by the human 11beta-hydroxysteroid
dehydrogenase type 1. J Steroid Biochem Mol Biol 2006;99:215–22.
[45] Hennebert O, Le Mee S, Pernelle C, Morfin R. 5Alpha-androstane-
3beta,7alpha,17beta-triol and 5alpha-androstane-3beta,7beta,17beta-triol as
substrates for the human 11beta-hydroxysteroid dehydrogenase type 1.
Steroids 2007;72:855–64.
[46] Stickney DR, Dowding C, Garsd A, Ahlem C, Whitnall M, McKeon M,
et al. 5-Androstenediol stimulates multilineage hematopoiesis in rhesus
monkeys with radiation-induced myelosuppression. Int Immunopharmacol
2006;6:1706–13.
[47] Lu M, Patsouris D, Li P, Flores-Riveros J, Frincke JM, Watkins S, et al. A new antidi-
abetic compound attenuates inflammation and insulin resistance in Zucker
diabetic fatty rats. Am J Physiol Endocrinol Metab 2010;298:E1036–48.
[48] Elce JS, Lai PP. Water-soluble metabolites of the estrogens. Quantitation of C-18
tetrols in rat feces. Steroids 1976;27:335–52.
[24] Offner H, Zamora A, Drought H, Matejuk A, Auci D, Morgan E, et al. A synthetic
androstene derivative and a natural androstene metabolite inhibit relapsing-
remitting EAE. J Neuroimmunol 2002;130:128.
[49] Badeau M, Adlercreutz H, Kaihovaara P, Tikkanen MJ. Estrogen A-ring struc-
ture and antioxidative effect on lipoproteins. J Steroid Biochem Mol Biol
2005;96:271–8.
[25] Hennebert O, Pelissier MA, Le Mee S, Wulfert E, Morfin R. Anti-
inflammatory effects and changes in prostaglandin patterns induced by
7beta-hydroxy-epiandrosterone in rats with colitis. J Steroid Biochem Mol Biol
2008;110:255–62.
[50] Caspi E, Njar VC. Concerning the pathway from 19-oxoandrost-4-ene-3,17-
dione to estrone. Steroids 1987;50:347–62.
[51] Adlercreutz H, Jarvenpaa P. Assay of estrogens in human feces. J Steroid
Biochem 1982;17:639–45.
[26] Le Mee S, Hennebert O, Ferrec C, Wulfert E, Morfin R. 7beta-Hydroxy-
epiandrosterone-mediated regulation of the prostaglandin synthesis pathway
in human peripheral blood monocytes. Steroids 2008;73:1148–59.
[27] Loria RM, Padgett DA. Control of the immune response by DHEA and its metabo-
lites. Rinsho Byori 1998;46:505–17.
[28] Padgett DA, Loria RM. Endocrine regulation of murine macrophage function:
effects of dehydroepiandrosterone, androstenediol, and androstenetriol. J Neu-
roimmunol 1998;84:61–8.
[52] Gachancard-Bouya JL, Begue RJ. Urinary steroids from a newborn human
infant. Identification of 2 alpha-hydroxy-4-pregnene-3,20-dione, 3 beta,15
beta-dihydroxy-5-pregnen-20-one and 3 beta,15 alpha-dihydroxy-5-pregnen-
20-one. J Steroid Biochem Mol Biol 1994;49:213–26.
[53] Homoki J, Roitman E, Shackleton CH. Characterization of the major steroids
present in amniotic fluid obtained between the 15th and 17th weeks of gesta-
tion. J Steroid Biochem 1983;19:1061–8.
[54] Numazawa M, Mutsumi A, Ogata M, Osawa Y. Synthesis of 16 alpha,19-
[29] Auci DL, Reading CL, Frincke JM. 7-Hydroxy androstene steroids and a novel
synthetic analogue with reduced side effects as a potential agent to treat
autoimmune diseases. Autoimmun Rev 2009;8:369–72.
dihydroxy-4-androstene-3,17-dione and 3 beta,16 alpha,19-trihydroxy-5-
androsten-17-one and their 17 beta-hydroxy-16-keto isomers. Steroids
1987;49:247–57.
[30] Marcu AC, Kielar ND, Paccione KE, Barbee RW, Carter H, Ivatury RR, et al.
Androstenetriol improves survival in a rodent model of traumatic shock. Resus-
citation 2006;71:379–86.
[31] Trauger R, Corey E, Bell D, White S, Garsd A, Stickney D, et al. Inhibition of
androstenediol-dependent LNCaP tumour growth by 17alpha-ethynyl-5alpha-
androstane-3alpha, 17beta-diol (HE3235). Br J Cancer 2009;100:1068–72.
[32] Wang T, Villegas S, Huang Y, White SK, Ahlem C, Lu M, et al. Amelioration of
glucose intolerance by the synthetic androstene HE3286: link to inflammatory
pathways. J Pharmacol Exp Ther 2010;333:70–80.
[33] Smith DA, Humphrey MJ, Charuel C. Design of toxicokinetic studies. Xenobiotica
1990;20:1187–99.
[34] van Breemen RB, Li Y. Caco-2 cell permeability assays to measure drug absorp-
tion. Expert Opin Drug Metab Toxicol 2005;1:175–85.
[55] Lardy H, Marwah A, Marwah P. Transformations of DHEA and its metabolites
by rat liver. Lipids 2002;37:1187–91.
[56] Solano ME, Sander V, Wald MR, Motta AB. Dehydroepiandrosterone and met-
formin regulate proliferation of murine T lymphocytes. Clin Exp Immunol
2008;153:289–96.
[57] Peters JM, Zhou YC, Ram PA, Lee SS, Gonzalez FJ, Waxman DJ. Peroxi-
some proliferator-activated receptor alpha required for gene induction by
dehydroepiandrosterone-3 beta-sulfate. Mol Pharmacol 1996;50:67–74.
[58] Rice SP, Zhang L, Grennan-Jones F, Agarwal N, Lewis MD, Rees DA, et al.
Dehydroepiandrosterone (DHEA) treatment in vitro inhibits adipogenesis in
human omental but not subcutaneous adipose tissue. Mol Cell Endocrinol
2010;320:51–7.
[59] Liu D, Iruthayanathan M, Homan LL, Wang Y, Yang L, Wang Y, et al.
Dehydroepiandrosterone stimulates endothelial proliferation and angiogen-
esis through extracellular signal-regulated kinase 1/2-mediated mechanisms.
Endocrinology 2008;149:889–98.
[35] Auci D, Mangano K, Destiche D, White SK, Haung Y, Boyle D, et al. Oral treatment
with HE3286 ameliorates disease in rodent models of rheumatoid arthritis. Int
J Mol Med 2010:625–33.