Ureas with histamine H3-antagonist receptor activity-A new scaffold discovered by lead-hopping from cinnamic acid amides

Article abstract of DOI:10.1016/j.bmcl.2006.07.093

A group of tri and tetrasubstituted urea derivatives have been found to be hH₃-antagonists. The most potent compounds were found in the class of (piperazine-1-yl)-(piperidine-1-yl)-methanones which in addition showed negligible hERG inhibition.

Full text of DOI:10.1016/j.bmcl.2006.07.093

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5303–5308
Ureas with histamine H3-antagonist receptor activity—A new
scaffold discovered by lead-hopping from cinnamic acid amides
Jesper F. Lau,a,* Claus Bekker Jeppesen,b Karin Rimvallb and Rolf Hohlwega
a
˚
Medicinal Chemistry Research, Novo Nordisk A/S, Novo Nordisk Park, DK-2760 Maløv, Denmark
b
˚
Molecular Pharmacology, Novo Nordisk A/S, Novo Nordisk Park, DK-2760 Maløv, Denmark
Received 26 June 2006; revised 31 July 2006; accepted 31 July 2006
Abstract—A group of tri and tetrasubstituted urea derivatives have been found to be hH3-antagonists. The most potent compounds
were found in the class of (piperazine-1-yl)-(piperidine-1-yl)-methanones which in addition showed negligible hERG inhibition.
ꢀ 2006 Elsevier Ltd. All rights reserved.
In an attempt to overcome the hERG-channel inhibi-
tion related to the cinnamic acid amides, we decided
to explore the possibilities of scaffold-hopping from
this series. Information about the SAR from the old
series provided a good basis for the design of new
compounds. Since hERG inhibition is known to corre-
late with lipophilicity5,6 we proposed the replacement
of the C@C in the cinnamic acid amide with an
N–C fragment leading to a more polar scaffold
possessing a urea moiety.
During the past few years, a plethora of novel structures
with histamine H3-interactions have been published.1 In
contrast to early generations of H3-ligands which are de-
rived from histamine itself and still contain an imidazole
moiety, these new compounds are much more drug-like
and hold some promise to also be of use in a clinical
setting, for indications such as obesity and cognitive
disorders.2
However, while the H3-receptor is a target, for which
selective and in vitro potent ligands have been detected
quite readily, overcoming ADMET-issues is still a chal-
lenge for many of these substances,3 and only a few can-
didates have yet reached clinical phases of development.
Previously, urea containing H3-antagonists were report-
ed. However, the presence of the imidazole moiety was
crucial for H3-activity.7 During the preparation of this
manuscript two patent applications describing imidaz-
ole-free urea derivatives as H3-antagonists were
published.8
An earlier detected structural class of H3-antagonists,
the cinnamic acid amides, exemplified by NNC 0038-
0000-1202,4 bears the potential risk of chemical and
metabolic instability due to the reactivity of the double
bond. Moreover, many of these compounds have shown
substantial hERG-channel inhibition.
To facilitate high throughput chemistry, a solid-phase
parallel synthesis protocol was developed to generate li-
braries of trisubstituted ureas, bearing similar substitu-
tion as the original cinnamic acid amides.9
Commercially available 2-(3,5-dimethoxy-4-formyl-
phenoxy)ethoxymethyl polystyrene was treated with a
variety of amines under standard reductive amination
conditions10 to give the resin bound secondary amine
1 (Scheme 1). Subsequent treatment with triphosgene
and base followed by reaction with a secondary amine
yielded the resin bound urea 2. The products were
cleaved from the resin using trifluoroacetic acid, to give
the crude urea derivatives 3 in an average purity of
92%.11 The crude products were purified by preparative
HPLC to give the final products as their respective triflu-
Cl
O
N
O
N
N
N
O
CF3
0038-0000-1202
Ki hH3 = 4.7 nM ( 0.4)
0038-0000-1049
Ki hH3 = 1.2 nM ( 0.1)
Keywords: Histamine H3 receptor; hH3-receptor; Urea; hERG.
*
Corresponding author. Tel.: +45 44 43 43 92; fax: +45 44 43 45 47;
e-mail: jrla@novonordisk.com
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.07.093

5304
J. F. Lau et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5303–5308
i
substituent (3n, 226 nM) did not change the H3-potency
ii
CHO
Pol
NH
R1
Pol
Pol
significantly, but constraining the phenylethyl group as
in the indanyl analog 3o gives rise to a more potent
compound (Ki = 52 nM). The most potent ureas of
pyrrolidinylmethylpyrrolidines were compounds 3k
and 3l with Ki of 34 and 47 nM, respectively. Unfortu-
nately, both showed unacceptable high inhibition of
the hERG-channel, which was assessed to be 62 and
40%, respectively.
1
O
O
R2
R2
iii
R1
N
N
N
N
R1
R3
H
R3
2
3
O
O
O
CHO
O
=
Pol
PS
O
H
We have previously been working on a series of piper-
azine amides14 represented by 0038-0000-1049. This
prompted us to incorporate an alkylpiperazine moiety
instead of the pyrrolidinylmethylpyrrolidine building
block. Representative examples of the second series
of ureas, bearing alkylpiperazine instead of pyrrolidi-
nylmethylpyrrolidine, are shown in Table 3. A com-
parison of these two series shows that the
alkylpiperazines generally are slightly more potent
than their equally substituted counterparts. Hence,
the two 3,4-dichlorobenzyl substituted cyclopentyl-
and isopropylpiperazine derivatives 3x and 3y (15
and 20 nM, respectively) are more potent than the pyr-
rolidinylmethylpyrrolidine bearing the same 3,4-dichlo-
ro-benzyl group (47 nM). Another observation was
that in all the piperazine ureas investigated, the poten-
cy is only minimally affected by various alkyl substitu-
tions on the piperazine ring.
Scheme 1. Reagents and conditions: (i) 10 equiv R1NH2, 15 equiv
NaBH3CN, NMP/MeOH, 10% AcOH; (ii) a—3 equiv CO(OCCl3)2,
DIPEA, DCM; b—10 equiv HNR2R3, NMP (iii) TFA/DCM (1:1).
oroacetic acid salts in overall yields typically between 35
and 75%.
Already in the first library, compounds with moderate
H3-potency12 were found. Early indications after com-
parison of the ureas and cinnamic acid amides data
showed the generally poorer potency of the ureas.
Hence, the most potent urea derivative 3d (Ki = 87 nM)
is 24 times less potent than the corresponding cinnamic
amide analog 9 (Ki = 3.7 nM) (Table 1).
We were however encouraged by the relative low
hERG-channel inhibition of the 4-trifluoromethyl
substituted urea derivative 3a (20%)13 compared to the
respective cinnamic acid amide analog NNC-0038-
0000-1202 (73%), therefore it was decided to further
explore this new series of H3-antagonists.
To further explore the scope of urea derivatives as
H3-antagonists, a parallel protocol for the synthesis of
tetrasubstituted ureas was established (Scheme 2). A
secondary amine was treated with CDI in DCM to give
the carbamoylimidazole 4.15 Methylation using methyl-
iodide gave the reactive intermediate 5 that subsequently
was treated with a secondary amine to give the tetrasub-
stituted urea derivative 6. Purification by HPLC yielded
the final products as their trifluoroacetate salts in yields
typically between 70 and 75%.16
More than 300 trisubstituted urea derivatives were made
but only minor improvement in H3-potency was ob-
tained. The SAR suggests a slight preference of 3- versus
4-substituted benzyl groups as exemplified by the 3-
methoxy-, 3-aminomethyl-, and 3-trifluoromethylbenzyl
derivatives that are all slightly more potent than the cor-
responding 4-substituted analogs (Table 2).
The simple introduction of an N-methyl group gave less
potent compounds, as exemplified by entry 6a (281 nM)
(Table 4) versus the non-methylated analog 3d (87 nM).
However, incorporation of both urea nitrogens in rings
showed promising results, especially in the piperazine
series. Consequently, efforts were concentrated around
The presence of the aromatic ring is essential in this
series, as most of the H3-potency is lost when benzyl
is replaced by a cyclohexyl methyl group in 3m. Replac-
ing the benzyl (3e, 196 nM) with a 2-phenylethyl
Table 1. Comparison of hH3-potency and hERG inhibition of cinnamic acid amides and the corresponding urea analogs20
N
N
O
O
;
N
N
N
H
R
R
R
Entry
Ki (nM) SEMa
hERG-inhibitionb (%)
Entry
Ki (nM) SEMa
hERG-inhibitionb (%)
–CF3
–Cl
NNC 0038-0000-1201
4.7 ( 0.4)
11.2 ( 1.4)
13.4 ( 1.9)
3.7 ( 1.3)
73
3a
3b
3c
3d
162 ( 14)
165 ( 24)
121 ( 17)
87 ( 15)
20
7
8
9
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
–OCF3
–Ph
a hH3-[35S]GTPc[S] binding assay (n = 3).
b Astemizol binding, % inhibition (mean, n = 3) at 10 lM.

J. F. Lau et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5303–5308
Table 2. hH3-potency of pyrrolidinylmethylpyrrolidine-ureas20 Table 3. hH3-potency of alkylpiperazine-ureas20
5305
Entry
Kia (nM)
N
O
O
R
N
N
N
N
N
H
H
3p
94 ( 3)
29 ( 4)
30 ( 6)
44 ( 10)
40 ( 9)
N
N
O
Entry
R
Kia (nM)
3q
3r
3s
3t
N
H
3a
162 ( 14)
CF3
O
N
N
H
N
3e
3f
196 ( 29)
295 ( 19)
188 ( 31)
117 ( 11)
78 ( 9)
O
O
N
N
N
H
N
N
O
N
H
3g
3h
3i
O
O
N
N
H
N
CF3
3u
3v
38 ( 6)
34 ( 8)
O
O
N
N
H
N
NH2
NH2
O
O
Cl
Cl
N
O
N
H
3j
109 ( 10)
34 ( 4)
3x
3y
15 ( 4)
20 ( 2)
N
Cl
Cl
3k
3l
N
N
H
O
N
a hH3-[35S]GTPc[S] binding assay.
Cl
Cl
47 ( 7)
inhibition is also known to decrease with decreasing
basicity the observed effect is probably caused by the
lower basicity of the N-cyclopropylpiperazine.17,18
3m
3n
3o
458 ( 23)
226 ( 8)
52 ( 10)
A group of quite potent ureas were the (piperazine-1-yl)-
(piperidine-1-yl)-methanones. Surprisingly some of these
show potent H3-activity despite the lack of an aromatic
group, as exemplified in entries 6h and 6i (Ki = 17 and
O
R1
a hH3-[35S]GTPc[S] binding assay.
R1
ii
i
NH
R2
N
N
N
R2
4
this scaffold. A small series of compounds bearing a
dihydroindole moiety, entries 6d, 6e, and 6f, showed
notable increase in potency when incorporating the
alkylpiperazine moiety. H3-potency is highest with the
more lipophilic cyclopentyl substituted derivative 6f,
but much more important is the finding, that the cyclo-
propyl group causes 6e to be the one with the lowest
hERG-channel inhibition (16% inhibition). Since hERG
O
O
R1
R1
iii
R3
N+
N
N
N
N
R2
R2
R4
I
5
6
Scheme 2. Reagents and conditions: (i) 1.1 equiv CDI, DCM; (ii)
10 equiv MeI, MeCN, (iii) 1 equiv HNR3R4, DIPEA, DCM.

5306
J. F. Lau et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5303–5308
Table 4. hH3-potency and hERG inhibition of tetrasubstituted ureas20
Entry
Kia (nM)
hERG % inhibitionb
N
O
N
N
6a
281 ( 49)
66
N
N
N
6b
6c
37 ( 8)
55
26
N
O
N
N
94 ( 14)
O
N
N
6d
6e
59 ( 11)
45 ( 5)
34
16
N
O
N
N
N
O
N
N
6f
6h
6i
6j
23 ( 5)
17 ( 3)
14 ( 3)
25 ( 5)
45
6
N
O
O
N
N
N
N
N
N
O
N
N
N
N
H
8
N
O
O
O
O
9
O
O
N
N
N
N
6k
6l
14 ( 3)
21
13
N
N
47 ( 10)
O
O
O
N
N
N
N
N
N
6m
6n
73 ( 18)
16 ( 3)
8
8
N
N
N
O
O
6o
6p
86 ( 16)
27 ( 7)
16
31
O
N
N
N
a hH3-[35S]GTPc[S] binding assay (n = 3).
b Astemizol binding, % inhibition (mean, n = 3) at 10 lM.

J. F. Lau et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5303–5308
5307
(1·1 mL), and DCM/DIPEA (9:1, 1 mL). DIPEA (204 lL,
1.2 mmol) in DCP (0.5 mL) was added to the resin and the
mixture was shaken for 5 min before triphosgene (107 mg,
0.36 mmol) in DCP (1.0 mL) was added and the reaction
mixture was agitated for 16 h before the excess reagent
and solvent were removed by filtration. The resin was
washed with DCM (3·1 mL) before N-ethyl-piperazine
(137 mg, 1.2 mmol) was added. The mixture was agitated
for 18 h before it was filtered and washed in sequence with
NMP (3·1 mL), DCM (3·1 mL), MeOH (1·1 mL), and
DCM (3·1 mL). The resin was subsequently cleaved with
TFA/DCM (1:1, 1 mL) for 1 h at room temperature before
the cleavage mixture was filtered. The resin was washed
with DCM (1 mL) and filtered. The filtrates were com-
bined and concentrated in vacuo. The crude product was
purified by preparative HPLC to give the trifluoroacetate
salt of (4-ethyl-piperazin-1-yl)-(indane-2-amine-yl)-metha-
none (3s). Yield 23 mg (50%). 1H NMR (400 MHz,
DMSO-d6) d (ppm) 9.61 (br s, 1H), 7.22–7.17 (m, 2H),
7.16–7.10 (m, 2H), 6.92 (d, 1H, J = 6.7 Hz), 4.43–4.32 (m,
1H), 4.18–4.03 (m, 2H), 3.53–3.35 (m, 2H), 3.19–3.07 (m,
4H), 3.06–2.94 (m, 2H), 2.94–2.85 (m, 2H), 2.85–2.76
(2H), 1.21 (t, 3H, J = 6.9 Hz) 13C NMR (100.6 MHz,
14 nM, respectively). Moreover, all aliphatic derivatives
had a low propensity of hERG- as well as CYP-inhibi-
tion. Hence the aliphatic derivatives 6h, 6i, 6j, and 6l
show 6, 8, 9, and 13% hERG-channel inhibition, respec-
tively, at 10 lM. The same compounds all gave IC50 val-
ues >25 lM on CYP1A2, CYP3A4, and CYP2D6.19
The only exceptions were 6h and 6l that gave IC50 values
of 17 and 7 lM, respectively, on CYP2D6.
In conclusion, we have developed a new series of urea
H3-antagonists. The most potent compounds are found
in the class of (piperazine-1-yl)-(piperidine-1-yl)-metha-
nones which in addition give negligible hERG
interaction.
Acknowledgments
We thank Claus Bruun Jensen, Pernille Lund, Dorthe
Dreyer Andersen, and Sine L. Rosenfalck Døhn for
excellent technical support.
1
DMSO-d6) d (ppm) 158 (q, JCF = 31.5 Hz), 156.6, 141.3,
126.2, 124.3, 51.7, 50.7, 50.1, 40.6, 39.1, 8.8 HPLC–MS: m/
z = 274 (M+1) MA; calcd for C16H23N3OÆC2HF3O2: C
55.81%; H 6.24%; N 10.85%, found: C 55.74%; H 5.97%;
N 10.71%.
References and notes
10. Severinsen, R.; Kilburn, J. P.; Lau, J. F. Tetrahedron 2005,
61, 5565.
1. Leurs, R.; Bakker, R. A.; Timmerman, H.; de Esch, I. J. P.
Nat. Rev. Drug Disc. 2005, 4, 107.
11. The purity was calculated by ELS peak integration of the
HPLC–MS chromatograms.
2. Celanire, S.; Wijtmans, M.; Talaga, P.; Leurs, R.; de Esch,
I. J. P. Drug Discovery Today 2005, 10, 1613.
12. For a detailed description of determination of H3-binding
using a hH3-[35S]GTPc[S] binding assay (n = 3), see Ref. 4.
13. The ability of test compounds to bind to the hERG-
channel was assessed by [3H]Astemizole binding to hERG
transfected HEK293 cell membranes essentially as
described by Chiu et al. (J. Pharmacol. Sci. 2004, 95,
311–319). All binding assays were performed in a total
volume of 100 lL: 60 lL buffer (10 mM Hepes, 5 mM
KCl, 130 mM NaCl, 0.8 mM MgCl2, 1 mM EGTA,
10 mM glucose, and 0.01% BSA, pH 7.4) containing
10 lg membrane, 20 lL test drug or vehicle (in buffer
containing 5% DMSO), and 20 lL [3H]Astemizole (15 nM
in buffer). Non-specific binding (NSB) was defined by
10 lM Astemizole (FAC). Incubation was conducted in
96-well polypropylene plates at 25ꢁC for 60 min. Binding
was terminated by rapid filtration using a FilterMate
Harvester (Packard) onto GF/B filters, presoaked with
0.3% polyethyleneimine, followed by rapid washing with
10 · 300 lL ice-cold washing buffer (25 mM Tris–HCl,
5 mM KCl, 130 mM NaCl, 0.8 mM MgCl2, 0.05 mM
CaCl2, and 0.01% BSA, pH 7.4). After drying of plates
and addition of 50 lL MicroScint 0 (Perkin-Elmer),
captured radiolabel was detected using a Perkin-Elmer
TopCount NXT. Results are presented as percent inhibi-
tion of [3H]Astemizole binding at 10 lM. Using the
conditions described above the IC50 values of the follow-
ing compounds were determined; Bepridil 444 ( 39) nM;
Pimozide 42 ( 2) nM; Haloperidol 512 ( 72) nM, and
Thioridazine 869 ( 170) nM.
3. Hancock, A. A. Biochem. Pharmacol. 2006, 71, 1103.
4. Peschke, B.; Bak, S.; Hohlweg, R.; Pettersson, I.; Refsg-
aard, H. H. F.; Viuff, D.; Rimvall, K. Bioorg. Med. Chem.
2004, 12, 2603.
5. Bilodeau, M. T.; Balitza, A. E.; Koester, T. J.; Manley, P.
J.; Rodman, L. D.; Buser-Doepner, C.; Coll, K. E.;
Fernandes, C.; Gibbs, J. B.; Heimbrook, D. C.; Huckel,
W. R.; Kohl, N.; Lynch, J. J.; Mao, X. Z.; McFall, R. C.;
McLoughlin, D.; Miller-Stein, C. M.; Rickert, K. W.;
Sepp-Lorenzino, L.; Shipman, J. M.; Subramanian, R.;
Thomas, K. A.; Wong, B. K.; Yu, S.; Hartman, G. D.
J. Med. Chem. 2004, 47, 6363.
6. Friesen, R. W.; Ducharme, Y.; Ball, R. G.; Blouin, M.;
Boulet, L.; Cote, B.; Frenette, R.; Girard, M.; Guay, D.;
Huang, Z.; Jones, T. R.; Laliberte, F.; Lynch, J. J.;
Mancini, J.; Martins, E.; Masson, P.; Muise, E.; Pon, D.
J.; Siegl, P. K. S.; Styhler, A.; Tsou, N. N.; Turner, M. J.;
Young, R. N.; Girard, Y. J. Med. Chem. 2003, 46, 2413.
7. Aslanian, R.; Mutahi, M. W.; Shih, N-Y.; McCormick, K.
D.; Piwinsky, J. J.; Ting, P. C.; Albanese, M. M.; Berlin,
M. Y.; Zhu, X.; Wong, S. C.; Rosenblum, S. B.; Jiang, Y.;
West, R.; She, S.; Williams, S. M.; Bryant, M.; Hey, J. A.
Bioorg. Med. Chem. Lett. 2002, 12, 937.
8. Folmer, J.; Hunt, S. F.; Hamley, P.; Wesolowski, S. Pat.
Appl. WO 2006/014135 and WO 2006/014136.
9. General procedure for the solid-phase parallel synthesis of
trisubstituted ureas exemplified with the synthesis of the
trifluoroacetate salt of (4-ethyl-piperazin-1-yl)-(indane-2-
amine-yl)-methanone (3s). A solution of 2-aminoindane
(200 mg, 1.5 mmol) in NMP (1.0 mL) was added to 2-(3,5-
dimethoxy-4-formyl-phenoxy)-ethoxymethyl polystyrene
resin (100 mg, 0.12 mmol) pre-swollen in NMP followed
by a solution of NaBH3CN (94 mg, 1.5 mmol) in MeOH
(0.5 mL). After addition of AcOH (150 lL), the mixture
was agitated for 18 h at room temperature. Excess reagent
and solvent were removed by filtration and the resin was
washed with NMP (3·1 mL), MeOH (1·1 mL), DCM
14. Zaragoza, F.; Stephensen, H.; Knudsen, S. M.; Pridal, L.;
Wulff, B. S.; Rimvall, K. J. Med. Chem. 2004, 47, 2833.
15. Grzyb, J. A.; Shen, M.; Yoshina-Ishii, C.; Chi, W.; Brown,
R. S.; Batey, R. A. Tetrahedron 2005, 61, 7153.
16. General procedure for the parallel synthesis of tetrasub-
stituted ureas exemplified with the synthesis of the
trifluoroacetate salt of (4-isopropyl-piperazin-1-yl)-(4-
phenyl-piperidin-1-yl)-methanone (6o). 4-Phenyl-piperi-

5308
J. F. Lau et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5303–5308
dine (81 mg, 0.5 mmol) in DCM (5 ml) in a test tube with
3.92–3.70 (m, 4H), 3.68–3.27 (m, 5H), 3.00–2.80 (m, 4H),
2.77–2.61 (m, 1H), 1.93–1.83 (m, 2H), 1.75–1.57 (m, 2H),
1.36 (d, 6H, J = 6.8) 13C NMR (100.6 MHz, CDCl3) d
(ppm) 162.7 (q, 1JCF = 35.1 Hz), 162.8, 145.2, 128.6, 126.7,
126.5, 57.8, 47.4, 47.0, 43.8, 42.7, 33.0, 16.5 HPLC–MS:
m/z = 316 (M+1) MA; calcd for C19H29N3OÆC2HF3O2:
C 58.73%; H 7.04%; N 9.78%, found: C 58.69%; H 6.99%;
N 9.69%.
magnetic stirring was added CDI (89 mg, 0.55 mmol). The
reaction mixture was stirred for 18 h at room temperature
before it was washed with water (2·2 mL). The phases
were separated using phase separation tubes from Radley.
The volatiles were removed in vacuo and the residue was
re-dissolved in MeCN (10 ml). MeI (5 mmol, 709.7 mg)
was added and the reaction mixture was stirred for 18 h at
room temperature. The volatiles were removed in vacuo
and the residue was re-dissolved in DCM (10 ml) where-
upon 1-isopropyl-piperazine (64 mg, 0.5 mmol) and tri-
ethylamine (70 ml, 0.5 mmol) were added. The reaction
mixture was stirred for 18 h at room temperature before
the solvent was removed in vacuo. THF (3 mL) and water
300 lL were added and the mixture was purified on
preparative HPLC to give the trifluoroacetate salt of (4-
isopropyl-piperazin-1-yl)-(4-phenyl-piperidin-1-yl)-metha-
none (6o). Yield 144 mg (67%). 1H NMR (400 MHz,
CDCl3) d (ppm) 7.36–7.27 (m, 2 H), 7.25–7.16 (m, 3H),
17. Armour, D. R.; de Groot, M. J.; Price, D. A.; Stammen,
B. L. C.; Wood, A.; Perros, M.; Burt, C. Chem. Biol. Drug
Des. 2006, 67, 305.
18. Bourrain, S.; Collins, I.; Neduvelil, J. G.; Rowley, M.;
Leeson, P. D.; Patel, S.; Patel, S.; Emms, F.; Marwood,
R.; Chapman, K. L.; Fletcher, A. E.; Showell, G. A.
Bioorg. Med. Chem. 1998, 6, 1731.
19. The cytochrome P450 screening was performed by Cypro-
tex in human liver microsomes.
20. All urea derivatives were isolated and characterized as
trifluoroacetate salts.