Synthesis and antimicrobial activities of N6-hydroxyagelasine analogs and revision of the structure of ageloximes

Article abstract of DOI:10.1016/j.bmc.2019.01.002

(+)-N⁶-Hydroxyagelasine D, the enantiomer of the proposed structure of (?)-ageloxime D, as well as N⁶-hydroxyagelasine analogs were synthesized by selective N-7 alkylation of N⁶-[tert-butyl(dimethyl)silyloxy]-9-methyl-9H-purin-6-amine in order to install the terpenoid side chain, followed by fluoride mediated removal of the TBDMS-protecting group. N⁶-Hydroxyagelasine D and the analog carrying a geranylgeranyl side chain displayed profound antimicrobial activities against several pathogenic bacteria and protozoa and inhibited bacterial biofilm formation. However these compounds were also toxic towards mammalian fibroblast cells (MRC-5). The spectral data of N⁶-hydroxyagelasine D did not match those reported for ageloxime D before. Hence, a revised structure of ageloxime D was proposed. Basic hydrolysis of agelasine D gave (+)-N-[4-amino-6-(methylamino)pyrimidin-5-yl]-N-copalylformamide, a compound with spectral data in full agreement with those reported for (?)-ageloxime D.

Full text of DOI:10.1016/j.bmc.2019.01.002

Accepted Manuscript
Synthesis and antimicrobial activities of N6-hydroxyagelasine analogs and re-
vision of the structure of ageloximes
Britt Paulsen, Kim Alex Fredriksen, Dirk Petersen, Louis Maes, An
Matheeussen, Ali-Oddin Naemi, Anne Aamdal Scheie, Roger Simm, Rui Ma,
Baojie Wan, Scott Franzblau, Lise-Lotte Gundersen
PII:
S0968-0896(18)31794-2
https://doi.org/10.1016/j.bmc.2019.01.002
BMC 14694
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
18 October 2018
20 December 2018
3 January 2019
Please cite this article as: Paulsen, B., Fredriksen, K.A., Petersen, D., Maes, L., Matheeussen, A., Naemi, A-O.,
Scheie, A.A., Simm, R., Ma, R., Wan, B., Franzblau, S., Gundersen, L-L., Synthesis and antimicrobial activities of
N6-hydroxyagelasine analogs and revision of the structure of ageloximes, Bioorganic & Medicinal Chemistry
(2019), doi: https://doi.org/10.1016/j.bmc.2019.01.002
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis and antimicrobial activities of N6-hydroxyagelasine analogs and
revision of the structure of ageloximes
Britt Paulsen,a Kim Alex Fredriksen,a Dirk Petersen,a Louis Maes,b An Matheeussen,b
Ali-Oddin Naemic, Anne Aamdal Scheiec, Roger Simmc, Rui Ma,d Baojie Wan,d
Scott Franzblau,d Lise-Lotte Gundersena*
(a) Department of Chemistry, University of Oslo, P.O.Box 1033, Blindern, 0315
Oslo, Norway.
(b) Laboratory of Microbiology, Parasitology and Hygiene (LMPH), University of
Antwerp, Wilrijk, Belgium.
(c) Institute of Oral Biology, University of Oslo, P.O. Box 1052, Blindern, 0316
Oslo, Norway.
(d) Institute for Tuberculosis Research, University of Illinois at Chicago, 833 S.
Wood St., Chicago, IL, USA.
Abstract: (+)-N6-Hydroxyagelasine D, the enantiomer of the proposed structure of (–
)-ageloxime D, as well as N6-hydroxyagelasine analogs were synthesized by selective
N-7 alkylation of N6-[tert-butyl(dimethyl)silyloxy]-9-methyl-9H-purin-6-amine in
order to install the terpenoid side chain, followed by fluoride mediated removal of the
TBDMS-protecting group. N6-Hydroxyagelasine D and the analog carrying a
geranylgeranyl side chain displayed profound antimicrobial activities against several
pathogenic bacteria and protozoa and inhibited bacterial biofilm formation. However
these compounds were also toxic towards mammalian fibroblast cells (MRC-5). The
spectral data of N6-hydroxyagelasine D did not match those reported for ageloxime D
* To whom correspondence should be addressed. Tel.: +4722857019; E-mail:
l.l.gundersen@kjemi.uio.no
1

before. Hence, a revised structure of ageloxime D was proposed. Basic hydrolysis of
agelasine gave (+)-N-[4-amino-6-(methylamino)pyrimidin-5-yl]-N-
D
copalylformamide, a compound with spectral data in full agreement with those
reported for (–)-ageloxime D.
Key words: Agelasine, ageloxime, antimicrobial activity, structure elucidation
1. Introduction
Several bioactive secondary metabolites from marine sponges can be classified as
purine-terpene hybrids. Important sub-classes are agelasines and ageloximes (Fig. 1).
Agelasines are isolated from marine sponges of the genus Agelas and are associated
with bioactivities like antimicrobial and cytotoxic effects as well as contractive
responses of smooth muscles and inhibition of Na,K-ATPase.1 Ageloximes, also
isolated from Agelas sp., are claimed to be N6-hydroxyagelasines. To date isolation of
(–)-ageloxime B2 from A. mauritania and (–)-ageloxime D3 from A. nakamurai has
been reported (Fig. 1). The ageloximes are named after the corresponding agelasines,
purinium salts with a primary amino group in the purine 6-position.
Fig. 1. Proposed structures of ageloximes B and D and structures of agelasine B and
D.
(–)-Ageloxime D was reported to inhibit biofilm formation from Staphylococcus
epidermidis, but did not inhibit the growth of the planktonic bacteria (MIC > 45
M). (–)-Agelasine D, on the other hand, did not inhibit biofilm formation, but
2

displayed potent growth inhibition against S. epidermidis (MIC < 0.09 M) 3
indicating that the oxime substituent on C-6 in ageloximes is important for biofilm
inhibition and at the same time reduces toxicity towards the bacteria. However, it was
later reported that both (–)-ageloxime B and (–)-ageloxime D displayed antimicrobial
activity against other microorganisms (Cryptococcus neoformans and Staphylococcus
aureus; MICs ca. 10 g/mL (ca. 23 M calculated from the Mw's of the revised
structures).2 Bacteria capable of forming biofilms are often less sensitive to
antibiotics compared to planktonic organisms and there is a current need for efficient
and non-toxic biofilm inhibitors.4
We have previously showed that synthetic analogs of agelasines, carrying a simpler
terpenoid side chain (i.e. geranylgeranyl) and/or an alkoxy substituent at N6 display
just as high antimicrobial activities as the natural products.5 Thus we wanted to
develop an efficient synthetic route to ageloximes and analogs and compare their
antimicrobial activity as well as ability to inhibit biofilm formation with the
corresponding agelasines and N6-alkoxyagelasines. As a continuance of our synthetic
studies towards bioactive purine-terpene hybrids,5,6 we now report the first synthesis
of N6-alkoxyagelasine analogs and their antimicrobial activities. We also propose a
revised structure of ageloxime D.
2. Results and discussion
2.1. Chemistry
First we looked into methods for selective introduction of a side chain at N-7 on 6-
hydroxylamino-9-methylpurine (2). Benzyl bromide was used as alkylating agent in
these model reactions. Compound 2 was synthesized from the 6-chloropurine 1 by a
modified procedure from literature (Scheme 1).7 In this procedure, hydroxylamine
was generated in situ by treating HONH2HCl with an excess of KOH in ethanol.
However, we found that excess base resulted in partial hydrolysis of the starting
material to give 9-methylhypoxanthine8 whereby the desired product 2 could only be
isolated in ca. 30% yield. When hydroxylamine was generated using exactly equal
3

amounts of HONH2HCl and KOH, the oxime 2 was produced in 84% yield. We
also attempted the substitution with commercially available hydroxylamine in water,
but again hydrolysis of the starting material took place. Compound 2 existed as an
almost 1:1 mixture of the imino (oxime; shown in Scheme 1) and amino
(hydroxylamine) tautomer in DMSO-d6 at ambient temperature. Benzylation of the
oxime 2 resulted in at least three compounds; the desired product 3, the N6-
benzylated isomer 4 and the dibenzylated compound 5. Unfortunately we were not
able to isolate the ageloxime analog 3 in pure form from this mixture. We have
previously shown that a bulky alkoxy group at N6 prevents alkylation of the exocyclic
nitrogen,6b and we protected the hydroxyl group in compound 3 with the sterically
demanding tert-butyldimethylsilyl (TBDMS) group to give the O-silylated derivative
6. An attempt to synthesize compound 6 from chloropurine 1 by reaction with
ButMe2SiONH2 failed,9 probably due to decomposition of the hydroxylamine
derivate under reaction conditions normally used for similar reactions with
alkoxyamines.4,6b Compound 6 was selectively benzylated at N-7 and the O-
protecting group was easily removed under acidic conditions to give the desired
product 3a in high yield. Unfortunately, acidic deprotection was more complicated
when compounds with other N-7 side chains where the targets (vide infra), and we
searched for other ways to deprotect the benzylpurinium salt 7. TBAF mediated
cleavage turned out to be sluggish and purification of the salt 3 was more
complicated. CsF led to quantitative conversion into the target 3, but removal of
inorganic salts from the polar purinium salt 3 turned out to be difficult. NH4F was
initially introduced as an economical alternative to TBAF for cleavage of silyl
ethers10 and is often used for selective cleavage of tert-butyldiphenylsilyl ethers in
the presence of TBDMS-ethers.10,11 To our delight we found that the protecting group
in compound 7 could be removed by treatment with one equivalent NH4F in MeCN
and water and pure target compound 3b (bromide) was isolated in high yield after
crystallization from MeCN. The reaction proceeded readily at ambient temperatures
whereas most known examples of TBDMS ether cleavage using NH4F have been
performed at elevated temperatures (typically 40-75 oC).12
4

Scheme 1. Reagents and conditions: (a) HONH2HCl KOH, EtOH; (b) PhCH2Br,
o
DMA, 50 C; (c) ButMe2SiCl, imidazole, DMF; (d) ButMe2SiONH2, Et3N, n-BuOH,
; (e) HCl(aq), MeOH, ; (f) NH4F, MeCN, H2O.
Next we synthesized ageloxime analogs 10 with terpenoid side chains on N-7
(Scheme 2). Compounds 10a-c were easily formed by selective N-7 alkylation
employing geranyl bromide, geranylgeranyl bromide or anti-copalyl bromide (the
latter easily synthesized from commercially available (+)-manool via (+)-
copalol/anti-copalol).6a,6b The N-alkylations could be performed at lower reaction
temperatures and shorter reaction times compared to what has been used in related
reactions before,5,6a,6b,6e,6f,13 and in the syntheses of compounds 10b and 10c these
milder reaction conditions were required in order to avoid unwanted side reactions.
When compound 6 was reacted with anti-copalyl bromide for only 20 min at ambient
temperature, ca. 13% of the desilylated N6-substituted purine 9c was formed together
with the desired product 8c. If the reaction mixture was allowed to stir over night, the
yield of compound 8c decreased (23%) and the yield of the by-product 9c increased
(ca. 34%). Also the desilylated product 10c as well as unidentified compounds were
observed. Compounds 9 were probably formed in minor amounts in the syntheses of
compounds 8a and 8b, but were not isolated.
5

The N-7 benzylpurinium salt 7 was, as discussed above, quantitatively desilylated
when treated with HCl in refluxing methanol for a couple of hours. Unfortunately,
acidic deprotection of compounds 8 turned out to be more sluggish. For instance,
when the copalyl derivative 8c was reacted with acid in refluxing methanol for 70 h,
the ratio between starting material 8c and desired product 10c was ca. 3:7 and minor
amounts of an unknown product was observed. However, NH4F mediated
deprotections of compounds 8 gave all targets 10 in quantitative yields.
Scheme 2. Reagents and conditions: (a) R-Br, DMA, r.t. or 50 oC; (b) NH4F, MeCN,
H2O.
The structure of the target compound 10c was established by NMR spectroscopy
(Fig. 2 and 3, Table 1). The identity of the heterocyclic part of the molecule as well
as the attachment of the side chains were determined from HMBC (1H-13C as well as
1H-15N) spectroscopy (Fig. 2) and the relative stereochemistry of the terpene side
chain was established by 2D NOESY/1D SELNOE (Fig. 3). The heterocyclic NH
was found to be localized at N-1 and the 13C and 15N NMR shifts of the atoms in the
heterocycle were in good agreement with those reported before for 1H-imino
tautomers of previously synthesized N6-alkoxy agelasine analogs.5,6b Compounds 6, 8
and 10 are all drawed as syn imines, since calculations and spectroscopic studies have
shown the syn form of related molecules to be favored over the anti form.14
6

Furthermore, SELNOE irradiating the NH showed that compound 10c exists as syn
imine in DMSO-d6 (Fig. 3).
1
1
Fig. 2. Key correlations found in the H-13C HMBC spectrum (left) and H-15N
HMBC spectrum (right) of compound 10c.
Fig. 3. Key correlations found in the 2D NOESY/1D SELNOE spectra of compound
10c.
13
Table 1. NMR data of compound 10c. C NMR data in CD3OD reported for
ageloxime D is included for comparison.
No.a
DMSO-d6
1H NMRb 13C
NMR NMR
CD3OD
13C
NMR
15N
1H NMRb
13C NMR
Ageloxime
7

c
d
c
D3

Integration, 

Integration, 

DEP
T
DEP
T
multiplicity
,
multiplicity
,
J (Hz)
J (Hz)
1A, 1.65- 1H, mf
eqe
1.70
1.72
-
38.4,
CH2
40.3,
CH2
1.77 1H, mf
40.7
20.4
43.3
1B, 0.88 1H, td, 3.8,
axe
13.0
2A, 1.48- 1H, mf
axe
1.54
0.96 1H, td, 3.8,
13.0
1.57
-
1.63 1H, mf
18.8,
CH2
20.4,
CH2
2B, 1.41- 1H, mf
eqe
1.44
1.46
-
1.54 1H, mf
3A, 1.34 1H, br d,
eqe
12.9
1.39
-
41.6,
CH2
43.3,
CH2
1.41 1H, mf
3B, 1.10 1H, td, 4.0,
1.19 1H, td, 4.0,
13.5
axe
13.4
4
33.2,
C
34.5, 40.2
C
5
0.99 1H, dd, 2.6, 54.7,
1.05 1H, dd, 2.7, 57.0, 56.8
12.6
6A, 1.65- 1H, mf
CH
12.6
CH
1.72 1H, mf
25.6
23.9,
25.6,
8

eqe
1.70
CH2
-
CH2
1.77
6B, 1.22 1H, dq, 4.2,
axe
12.9
1.31 1H, dq, 4.2,
12.9
7A, 2.33 1H, ddd,
2.36 1H, ddd,
eqe
3.4, 4.0,
3.4,
4.0,
12.7
12.7
37.6,
CH2
39.4,
CH2
39.4
7B, 1.83- 1H, mf
1.88 1H, ddd,
axe
1.88
5.0, 13.0,
13.0
8
9
148.0
, C
149.7 149.8
, C
1.48- 1H, mf
1.54
1.57
-
58.3
57.1,
55.1,
CH
1.64 1H, mf
CH
10
39.2,
C
40.7, 34.5
C
11A 1.54- 1H, mf
1.59
1.65
-
1.70 1H, mf
20.9,
CH2
22.5,
41.6
CH2
11B 1.41- 1H, mf
1.44
1.46
-
1.54 1H, mf
12A 2.11 1H, ddd,
2.23 1H, ddd,
37.7,
CH2
39.2,
25.6
CH2
4.0, 9.4,
14.0
4.3,
9.1,
14.0
9

12B 1.83- 1H, mf
1.88
1.96 1H, dt, 8.1,
14.0
13
144.1
, C
147.2 145.4/144.4
g
, C
14
5.37 1H, qt, 1.0, 116.5
5.43 1H, qt, 1.0, 117.4 117.9/118.4
7.2
, CH
7.2
, CH
15A 4.98 1H, dd, 7.2,
14.8
5.01
-
47.1,
CH2
49.0,
CH2
2H, m
45.9/41.6
15B 4.94 1H, dd, 7.2,
14.8
5.10
16
1.76 3H, d, 1.0
16.7,
CH3
1.84 3H, d, 1.0
4.82 1H, q, 1.5
17.0, 16.0/16.2
CH3
17A 4.81 1H, td, 1.1,
1.5
106.4
, CH2
107.0
107.0
, CH2
17B 4.48 1H, br s
4.52 1H, q 1.0
0.87 3H, s
18
19
20
0.83 3H, s
0.75 3H, s
0.61 3H, s
33.3,
CH3
34.1, 34.1
CH3
21.5,
CH3
0.81 3H, s
0.69 3H, s
22.1, 20.4
CH3
14.3,
CH3
15.0, 15.1
CH3
h
1'
11.0 1H, br s
0
-
_
NH
240.0
2'
7.79 1H, s
148.9
, CH
7.74 1H, s
149.9 157.7/157.3
, CH
3'
-
10

179.5
4'
5'
6'
140.7
, C
142.5 161.5/160.6
, C
110.6
, C
112.9 97.2/99.2
, C
136.9
, C
137.9 162.0/160.6
, C
h
6'-
NO
H
10.5 1H, br s
8
-
_
_
105.5
7'
8'
9'
-
208.5
h
9.26 1H, s
136.7
, CH
137.3 165.9/166.5
, Ci
-
218.5
9'-
3.78 3H, s
31.6,
CH3
32.1
28.2
CH3
a For numbering of the atoms, see Fig. 2.
b 1H NMR data were obtained at 600 MHz.
c 13C NMR data were obtained at 150 MHz.
d 15N NMR data were obtained at 60 MHz.
e
Equatorial/axial positions in a (double) chair conformation, obtained by 2D
NOESY/1D SELNOE.
f Unresolved or overlapping.
g Double set of signals were observed and explained by tautomery;3
11

h Exchangeable H not seen in CD3OD.
i The carbon resonance appeared as a triplet.
However, we were surprised to see that the NMR spectra of compound 10c did not
13
match those published for (–)-ageloxime D.3 Literature C NMR data (CD3OD) are
included in Table 1 for comparison. It was reported that ageloxime D exists as two
tautomers in CD3OD. Our 1H NMR and 13C NMR spectra of compound 10c did only
show one tautomer and looked substantially different from those published before,
especially the signals from the heterocyclic part. One interesting difference was the
signals for the purine H-8' / C-8'. The literature reported the H-8' to appear at 7.95,
but we could not see a signal for H-8' in the CD3OD 1H NMR spectrum of compound
10c. This indicated a very acidic proton at this site, being exchanged by deuterium.
13
We also observed a triplet (13C-2H-coupling) for C-8' in the C NMR spectrum, not
mentioned in the literature, which supported this observation. The acidity of H-8' in
the proposed imidazolium cation structure was also believed to be quite high.
Furthermore, ageloxime B and D were reported to be protonated at N-9 in CDCl3.2,3
We found protonation of an imidazolium cation quite unlikely in a relatively weak
acidic solution, and also that compound 10c was almost insoluble in CDCl3. Last but
not least, the molecular ions reported for ageloxime B and D were m/z 440,2,3 which
corresponds to C26H40N5O+2, whereas the molecular ion found for compound 10c
was 438 (M+).
We suspected that the previously isolated ageloximes were in fact not N6-
hydroxyagelasines. Ageloxime D2,3 seemed to have the same (enantiomeric) terpene
side chain as our compound 10c, but it was not likely that the heterocyclic parts were
identical. Capon et al.15 isolated formamides 12a and 12b (Fig. 4) and stated that they
were artifacts formed during silica gel chromatography of crude sponge extracts
containing the corresponding ageline A (agelasine F; 11a) and ageline B (11b).
Furthermore, they showed that ageline A and B could be converted to the
12

corresponding formamides 12 when treated with aqueous base. Also other agelasines
have been hydrolyzed to formamides using this protocol.16
Fig. 4. Structures of agelasines 11 and ring opened derivatives/axistatins 12.
Two formamidopyrimidines with terpene side chains, named axistatin 1 and 2 were
isolated from the marine sponge Agelas axifera Hentschel.17 The proposed structure
of axistatin 2 (12c) is shown in Fig. 4. We found the NMR data reported for axistatin
2 and ageloxime B2 to be essentially identical.* This led us to believe that the
previously isolated ageloxime B has the same structure as axistatin 2 and that
ageloxime D is the ring opened form of agelasine D. This would explain the
molecular ion found for the ageloximes and the coupling between the CH3N protons
and the NH previously explained by N-9 protonation. Furthermore, presence of
double NMR signals in CD3OD could be explained by restricted rotation around the
amide bond rather than the presence of two tautomers. The affected protons in the
1
CD3OD H NMR spectrum3 were those close to the formamide group earlier
mistaken to be the purine H-8'. Thus we hydrolized our previously synthesized (+)-
agelasine D (11d)6a,6b and obtained the formamide 12d (Scheme 3). The spectral data
for compound 12d were identical with those previously reported for (–)-ageloxime D3
and the optical rotation confirmed that this compound is the enatiomer of the
previously reported (–)-ageloxime D. Thus we have also confirmed the absolute
stereochemistry of the ageloxime reported by Hertiani et al. who just assumed that
*
A table with literature NMR data of ageloxime B and axistatin 2 can be found in the supplementary material.
13

the compound they named (–)-ageloxime D would have the same absolute
stereochemistry as (–)-agelasine D.3
Scheme 3. Reagents and conditions: (a) NaHCO3, H2O.
2.2. Biology
Even though the synthetic products 10 turned out not to be ent-ageloxime D and
analogs, we chose to examine their activities against a variety of microorganisms
(Table 2). The geranyl derivative 10a was essentially inactive against bacteria (S.
aureus, Escherichia coli and Mycobacterium tuberculosis) and yeast (Candida
albicans) and showed only a modest inhibitory activity against protozoa
(Trypanosoma cruzi, T. brucei, T. rhodesiense and Leishmania infantum). This is in
line with our previous findings; a certain size of the lipophilic substituent on N-7 is
required for antimicrobial activities of agelasine analogs.5,18 Compounds 10b and
10c, containing larger lipophilic side chains, displayed good to very good activities
towards all microorganisms examined except for E. coli. The IC50 or MIC values are
comparable with those found for other agelasine analogs with a diterpenoid side
chain. As also seen for related compounds, the N6-hydroxyagelasine analogs 10b and
10c were not only toxic to microorganisms but also to mammalian MRC-5sv2 (human
lung fibroblast) cells.5,18
Table 2. Antimicrobial activities of compounds 10.
Compound
Compound
Compound
14

10a
10b
10c
S. aureus, IC50 (M)a
E. coli, IC50 (M)b
>64.0
>64.0
>64.0
11.5
1.89
>64.0
2.00
0.54
2.01
0.50
2.16
11.3
1.84
>64.0
8.00
0.53
2.01
0.50
2.38
3.00
C. albicans, IC50 (M)c
T. cruzi, IC50 (M)d
T. brucei, IC50 (M)e
T. rhodesiense, IC50 (M)f
L. infantum IC50 (M)g
28.0
8.06
20.3
M. tuberculosis MIC (g/mL) >50.0
MABA assayh
M. tuberculosis MIC (g/mL) >50.0
15.3
1.90
5.50
1.59
LORA assayi
MRC-5 IC50 (M)j
>64.0
a IC50 doxycycline 0.03 M.
b IC50 doxycycline 0.58 M.
c IC50 flucytosine 0.41 M.
d IC50 benznidazol 2.69 M.
e IC50 suramin 0.02 M.
f IC50 suramine 0.03 M.
g IC50 miltefosine 11.8 M.
h MIC isoniazid 0.48 g/mL.
i MIC isoniazid >128 g/mL.
j IC50 tamoxifen 11.23 M.
15

The activity of compounds 10a, 10b and 10c were also tested against biofilm
formation and compared to the effect on growth under the same conditions.
Compound 10a did not have any significant effect on biofilm formation or growth of
S. epidermidis, Pseudomonas aeruginosa or E. coli,19 with the exception of a 50%
reduction in biofilm formation for P. aeruginosa at the highest concentration tested
(1 mM) (Fig. 6). These results correlate with the antibacterial activity of the
compounds measured as reduced metabolic activity (Table 2). The N6-
hydroxyagelasine analogs 10b and 10c reduced biofilm formation of the Gram-
positive bacterium S. epidermidis by 90% at 125 M and 63 M respectively. In
comparison, growth was reduced by 50% at 125 M and 15% at 63 M for 10b and
10c, respectively.
Fig. 5. Effect of compounds 10a, 10b and 10c on biofilm (left) and growth (right) of
S. epidermidis. Shown are averages and 95% confidence intervals of at least two
independent experiments carried out in at least two technical replicates per
experiment. * indicates statistical significant difference compared to the control (two-
tailed students t-test, p < 0.05).
For the Gram-negative bacteria P. aeruginosa and E. coli, the effect on biofilm
formation was less pronounced. For P. aeruginosa (Fig. 6), compounds 10b and 10c
only inhibited biofilm formation by 40-50% at 250 M and reached 60% and 70%,
respectively at the highest concentration tested (1 mM). However, growth was only
inhibited by 20-30% at this concentration.
16

Fig. 6. Effect of compounds 10a, 10b and 10c on biofilm (left) and growth (right) of
P. aeruginosa. Shown are averages and 95% confidence intervals of at least two
independent experiments carried out in at least two technical replicates per
experiment. * indicates statistical significant difference compared to the control (two-
tailed students t-test, p < 0.05).
Similarly, compounds 10b and 10c inhibited biofilm formation of E. coli by 70% at
500 M but only reduced growth by approximately 20% at this concentration (Fig.
7). These results indicate a small inhibitory effect of compounds 10b and 10c on
biofilm formation. However, since the inhibition of biofilm formation is accompanied
by a reduction in growth, it is not possible to conclude from these experiments
whether it is a specific anti-biofilm effect or a more general growth mediated effect.
In addition, the results (Table 2 and Fig. 5) suggest that Gram-positive bacteria are
more susceptible to the antibacterial effects of the compounds.
Fig. 7. Effect of compounds 10a, 10b and 10c on biofilm (left) and growth (right) of
17

E. coli. Shown are averages and 95% confidence intervals of at least two independent
experiments carried out in at least two technical replicates per experiment. * indicates
statistical significant difference compared to the control (two-tailed students t-test, p
< 0.05).
3. Conclusions
We have synthesized (+)-N6-hydroxyagelasine D and other N6-hydroxyagelasine by
selective N-7 alkylation of N6-[tert-butyl(dimethyl)silyloxy]-9-methyl-9H-purin-6-
amine followed by efficient cleavage of the silyl ether by NH4F. However, the
spectral data of synthetically prepared (+)-N6-hydroxyagelasine D did not match
those reported for the naturally occurring (–)-ageloxime D. Instead we found that
when (+)-agelasine D was treated with aqueous base, ring opening of the imidazole
ring took place to give a 4-amino-6-(methylamino)pyrimidin-5-ylformamide with
spectral data essentially identical to those reported for (–)-ageloxime D before. This
led us to propose a revised structure for (–)-ageloxime D. This pyrimidinylformamide
may be found in Agelas sp. or alternatively the compound is an artifact from isolation
of (–)-agelasine D. The work presented herein demonstrates the value of total
synthesis in order to prove, or disprove, the structure of natural products.20 N6-
Hydroxyagelasine analogs carrying a diterpenoid side chain at N-7, displayed toxicity
towards both microorganisms as well as mammalian cells. These compounds also
inhibited bacterial biofilm formation, but the data indicate that this effect is likely a
result of the antibacterial properties of the compounds causing reduced bacterial
growth rather than a specific anti-biofilm property.
4. Experimental section
4.1. Chemistry
1H NMR spectra were recorded at 600 MHz on a Bruker AV 600 MHz or a Bruker
13
AVII 600 MHz, at 400 MHz on a Bruker AVII 400 instrument. The decoupled C
18

NMR spectra were recorded at 150 or 100 MHz using the instruments mentioned
above or at 125 MHz on a Bruker AVII 500 MHz. 15N NMR spectra were recorded at
1
60 or 50 MHz using instruments mentioned above. H chemical shift values are
calibrated relative to internal CD2HSOCD3 at 2.49 ppm, CD2HOD at 3.30 ppm or
13
CHCl3 at 7.24 ppm. C chemical shift values are calibrated relative to internal
(CD3)2SO at 39.50 ppm, CD3OD at 49.00 ppm or CDCl3 at 77.00 ppm. 15N chemical
shift values are calibrated relative to external CH3NO2/DMSO-d6 (9:1) at 0.00 ppm.
1
13
15
1
13
Assignments of H, C and N resonances are inferred from 1D H NMR, 1D C
NMR, DEPT or APT, and 2D NMR (COSY, HMQC, HMBC and/or NOESY)
spectroscopic data. ESI MS spectra were obtained on a Waters Micromass QTOF 2
W instrument. Melting points were determined on a Büchi Melting Point B-545
apparatus. Dry DMF were obtained from a solvent purification system, MB SPS-800
from MBraun, Garching, Germany. DMA was distilled over BaO prior to use, and n-
BuOH was distilled over BaO and stored over 3 Å molecular sieves. Triethylamine
was distilled from CaH2 and stored over molecular sieves (3 Å). Flash
chromatography was performed on silica gel Merck, Darmstadt, Germany (Merck
No. 09385). Compounds available by literature methods: Geranyl bromide,5a
geranylgeranyl bromide,5a 6-chloro-9-methyl-9H-purine (1),21 anti-copalyl
bromide,6a,6b (+)-agelasine D (11d).6a,6b
4.1.1. 6-Hydroxylamino-9-methyl-9H-purine (2).
A warm solution of KOH (3.270 g, 58.28 mmol) in EtOH (abs., 80 mL) was added to
a boiling solution of HONH2HCl (4.035 g, 58.07 mmol) in EtOH (abs., 80 mL).
Immediately upon mixing of the two solutions, precipitation of KCl was observed.
The resulting mixture was stirred and maintained near the boiling point for an
additional 10 min. The hot reaction mixture was filtered and 6-chloro-9-methyl-9H-
purine (1) (1.241 g, 8.027 mmol) in EtOH (abs., 70 mL) was added to the
hydroxylamine solution. The resulting solution was kept at reflux for 4 h. The
reaction mixture was cooled to ambient temperature and the precipitated product was
collected by filtration. The crystals were washed with EtOH (abs., 15 mL) and dried
19

o
o
1
in vacuo; yield 992 mg (84%), colorless crystals, mp 233-234 C. (lit.8 244 C). H
NMR (400 MHz, DMSO-d6) δ 10.96, 10.24 (s, 1H, OH taut. A and B), 9.49, 9.00 (s,
H, NH taut. A and B), 8.25, 8.11, 7.71, 7.52 (s, 1H, H-2 and H-8, taut A and B), 3.73,
3.62 (br s, 3H, CH3 taut. A and B); HRMS (ESI) found 166.0732, calcd for
C6H8N5O+ 166.0723.
4.1.2. N6-Benzyl-N6-hydroxy-9-methyl-9H-purin-6-amine (4) and N6-benzyl-N6-
benzyloxy-9-methyl-9H-purin-6-amine (5).
Compound 2 (81 mg, 0.49 mmol) was dissolved in dry DMA (8 mL) at ambient
temperature under Ar atm before benzyl bromide (0.09 mL, 0.7 mmol) was added.
The mixture was stirred at 50 ˚C for 18 h and evaporated in vacuo. The products were
separated by flash chromatography eluting with 5-10% MeOH saturated with NH3 in
CH2Cl2; yield 13 mg (ca. 11%, contained 1-2 mol% DMA) of compound 4, purple
waxy material, and 29 mg (ca. 16%, contained ca. 17 mol% DMA) of compound 5,
pale greenish oil. 4; 1H NMR (400 MHz, CDCl3) δ 8.22 (s, 1H, H-2), 7.39 (s, 1H, H-
8), 7.49-7.32 (s, 5H, Ph), 6.24 (s, 1H, OH), 5.32 (s, 2H, CH2), 3.57 (s, 3H, CH3); MS
+
1
ESI found 240.2, calcd for C13H14N5 -OH 240.1. 5; H NMR (400 MHz, CDCl3) δ
8.54 (s, 1H, H-8), 7.85 (s, 1H, H-2), 7.47-7.25 (m, 10H, Ph), 5.23 (s, 2H, NCH2),
5.01 (s, 2H, OCH2), 3.85 (s, 3H, NCH3); HRMS (ESI) found 346.1683, calcd for
C20H20N5O+ 346.1662. Spectral data were in good agreement with those reported
before.6b
4.1.3. N6-[tert-Butyl(dimethyl)silyloxy]-9-methyl-9H-purin-6-amine (6).
A mixture of 6-hydroxylamino-9-methyl-9H-purine (2) (302 mg, 1.94 mmol),
imidazole (329 mg, 4.84 mmol) and dry DMF (2.7 mL) was stirred under Ar atm
before tert-butyldimethylsilyl chloride (307 mg, 2.03 mmol) was added. The reaction
mixture was stirred at ambient temperature for 22 h, diluted with EtOAc (150 mL)
and washed with water (330 mL) and brine (40 mL). The organic phase was dried
(Na2SO4) and evaporated in vacuo; yield 317 mg (93%); colorless solid, mp 216-217
1
oC. H NMR (400 MHz, DMSO-d6) δ 10.91 (br s, 1H, NH), 7.78 (s, 1H, H-8), 7.65
20

(s, 1H, H-2), 3.64 (s, 3H, NCH3), 0.95 (s, 9H, But), 0.15 [s, 6H, Si(CH3)2]; 13C NMR
(100 MHz, DMSO-d6) δ 144.3 (C-6), 144.1 (C-2), 141.6 (C-4), 138.8 (C-8), 118.1
(C-5), 29.5 (NCH3), 26.2 (CH3 in But), 18.0 (C in But), -5.0 [Si(CH3)2]; HRMS (ESI)
found 280.1597, calcd for C12H22N5OSi+ 280.1588.
4.1.4.
7-Benzyl-6-tert-butyl(dimethyl)silyloxyamino-9-methyl-7H-purinium
bromide (7).
Benzyl bromide (0.037 mL, 0.31 mmol) was added to a stirring solution of compound
6 (100 mg, 0.283 mmol) in dry DMA (5 mL) under Ar atm. The resulting mixture
was stirred at ambient temperature for 1 h, and at 50 ˚C for 6 h and evaporated in
vacuo. The crude product was purified by flash chromatography eluting with 5-10%
1
MeOH in CH2Cl2; yield 104 mg (82%), colorless solid, mp 249-251 ˚C. H NMR
(400 MHz, DMSO-d6) δ 11.87 (br s, 1H, NH), 9.33 (s, 1H, H-8), 7.86 (s, 1H, H-2),
7.39-7.34 (m, 5H, Ph), 5.65 (s, 2H, CH2), 3.79 (s, 3H, NCH3), 0.89 (s, 9H, But), 0.08
[s, 6H, Si(CH3)2]; 13C NMR (100 MHz, DMSO-d6) δ 148.9 (C-2), 141.4 (C-4), 139.9
(C-6), 138.5 (C-8), 134.5 (C in Ph), 128.7 (CH in Ph), 128.4 (CH in Ph), 127.4 (CH
in Ph), 110.6 (C-5), 52.1 (CH2), 31.9 (NCH3), 25.9 (CH3 in But), 17.9 (C in But), -5.3
[Si(CH3)2]; HRMS (ESI) found 370.2057, calcd for C19H28N5OSi+ 370.2058.
4.1.5. 7-Benzyl-6-hydroxylamine-9-methyl-9H-purin-7-ium chloride (3a).
Compound 7 (127 mg, 0.283 mmol) was dissolved in methanol (2 mL) and 1M HCl
(aq) (1 drop) was added. The solution was stirred at reflux for 2.5 h, cooled to
ambient temperature, washed with hexane (33 mL) and evaporated in vacuo; yield
1
92 mg (97%); colorless solid, mp 219-222 ˚C. H NMR (400 MHz, DMSO-d6) δ
11.96 (br s, 1H, NH), 10.65 (s, 1H, OH), 9.39 (s, 1H, H-8), 7.82 (s, 1H, H-2), 7.41-
13
7.37 (m, 5H, Ph), 5.63 (s, 2H, CH2), 3.81 (s, 3H, NCH3); C NMR (100 MHz,
DMSO-d6) δ 148.9 (C-2), 140.8 (C-4), 137.5 (C-8), 136.6 (C-6), 134.5 (C in Ph),
128.88 (CH in Ph), 128.58 (CH in Ph), 127.92 (CH in Ph), 110.8 (C-5), 51.9 (CH2),
31.79 (NCH3); HRMS (ESI) found 256.1193, calcd for C13H14N5O+ 256.1193.
21

4.1.6. 7-Benzyl-6-hydroxylamine-9-methyl-9H-purin-7-ium bromide (3b).
Compound 7 (20.5 mg, 0.0453 mmol) was dissolved in MeCN-H2O (9:1, 0.5 mL)
and transferred to a vial containing NH4F (1.3 mg, 0.035 mmol). Additional MeCN-
H2O (9:1, 0.5 mL) was used to transfer the starting material to the reaction vial. The
clear and colorless solution stirred for 60 min at ambient temperature and was
evaporated in vacuo. Dry MeCN (2 mL) was added and the resulting heterogeneous
mixture was agitated by a spatula for 2 min while gently heating the solution with a
heating gun. The crystals were allowed to settle by gravity (2 min) before the MeCN-
phase was removed by a glass pipette. This process was repeated once more with
MeCN (1 mL). The crystals were dried in vacuo; yield 12.7 mg (83% based on the
amount of comp. 7), colorless solid, mp 244-245 oC. Spectral data as reported for the
chloride 3a above.
4.1.7. 6-(tert-Butyldimethylsilyloxyimino)-7-[(E)-3,7-dimethylocta-2,6-dienyl]-9-
methyl-6,9-dihydro-1H-purin-7-ium bromide (8a).
Geranyl bromide (265 mg, 1.12 mmol) was added dropwise over 1 min to a stirring
solution of compound 6 (283 mg, 1.02 mmol) in dry DMA (10 mL) under Ar atm.
The resulting mixture was stirred at ambient temperature for 10 min and at 50 ˚C for
18 h and evaporated in vacuo. The crude product was purified by flash
chromatography eluting with 5-10% MeOH in CH2Cl2; yield 348 mg (70%),
colorless solid, mp 196-197 ˚C. 1H NMR (400 MHz, DMSO-d6) δ 11.90 (br s, 1H,
NH), 9.37 (s, 1H, H-8 in purine), 7.92 (s, 1H, H-2 in purine), 5.49-5.46 (m, 1H, H-2),
5.07-5.03 (m, 1H, H-6), 4.97 (d, J = 7.0 Hz, 2H, H-1), 3.81 (s, 3H, NCH3), 2.10-2.05
(m, 2H, H-5), 2.05-2.01 (m, 2H, H-4), 1.77 (s, 3H, CH3 at C-3), 1.62 (s, 3H, H-8),
13
1.54 (s, 3H, CH3 at C-7), 0.96 (s, 9H, But), 0.20 [s, 6H, Si(CH3)2]; C NMR (100
MHz, DMSO-d6) δ 148.8 (C-2 in purine), 143.0 (C-3), 141.2 (C-4 in purine), 140.0
(C-6 in purine), 137.6 (C-8 in purine), 131.2 (C-7), 123.5 (C-6), 116.8 (C-2), 110.1
(C-5 in purine), 47.3 (C-1), 38.9 (C-4), 31.7 (NCH3), 26.0 (CH3 in But), 25.7 (C-5),
25.4 (C-8), 18.0 (C in But), 17.5 (CH3 at C-7), 16.5 (CH3 at C-3), -5.2 [(Si(CH3)2];
HRMS (ESI) found 416.2840, calcd for C22H38N5OSi+ 416.2840.
22

4.1.8.
6-(tert-Butyldimethylsilyloxyimino)-9-methyl-7-[(2E,6E,10E)-3,7,11,15-
tetramethylhexadeca-2,6,10,14-tetraenyl]-6,9-dihydro-1H-purin-7-ium bromide
(8b).
Geranylgeranyl bromide (95 mg, 0.27 mmol) was dissolved in dry DMA (3 mL)
under Ar atm and compound 6 (68 mg, 0.24 mmol) was added. The resulting mixture
was stirred at ambient temperature for 1 h and evaporated in vacuo. The crude
product was purified by flash chromatography eluting with 5-10% MeOH in CH2Cl2;
yield 140 mg (91%), colorless solid, mp 170-171 ˚C. 1H NMR (600 MHz, DMSO-d6)
δ 11.88 (br s, 1H, NH), 9.34 (s, 1H, H-8 in purine), 7.91 (s, 1H, H-2 in purine), 5.46
(qt, J = 7.0 and 1.1, Hz, 1H, H-2), 5.05-5.02 (m, 3H, H-6, H-10, H-14), 4.96 (d, J =
7.0 Hz, 1H, H-1), 3.79 (s, 3H, NCH3), 2.08-1.96 (m, 8H, H-4, H-5, H-9, H-13), 1.92-
1.88 (m, 4H, H-8, H-12), 1.76 (s, 3H, CH3 at C-3), 1.61 (s, 3H, H-16), 1.53 (s, 6H,
CH3 at C-7 and C-15), 1.52 (s, 3H, CH3 at C-11) 0.95 (s, 9H, But), 0.18 [s, 6H,
13
Si(CH3)2]; C NMR (150 MHz, DMSO-d6) δ 148.7 (C-2 in purine), 143.0 (C-3),
141.2 (C-4 in purine), 139.9 (C-6 in purine), 137.6 (C-8 in purine), 134.8 (C-7),
134.3 (C-11), 130.6 (C-15), 124.0 (C-14), 123.8 (C-10), 123.3 (C-6), 116.7 (C-2),
110.1 (C-5 in purine), 47.2 (C-1), 39.2 (2C, C-8 and C-12), 38.9 (C-4), 31.7
(NCH3), 26.2 (2C, C-9 and C-13), 26.0 (CH3 in But), 25.6 (C-5), 25.5 (C-16), 18.0
(C in But), 17.5 (CH3 at C-15), 16.5 (CH3 at C-3), 15.8 (2C, CH3 at C-7 and C-11), -
5.3 [(Si(CH3)2]; HRMS (ESI) 552.4090, calcd for C32H54N5O+ 552.4092.
4.1.9.
(+)-6-(tert-Butyldimethylsilyloxyimino)-9-methyl-7-{(E)-3-methyl-6-
[(1S,4aS,8aS)-5,5,8a-trimethyl-2-methylenedecahydronaphthalen-1-yl]hex-2-
enyl}-6,7-dihydro-1H-purin-9-ium bromide (8c) and (E)-N-{3-methyl-5-
[(1S,4aS,8aS)-1,5,5,8a-tetramethyl-2-methylenedecahydronaphthalen-1-yl]pent-
2-enyl}-N-(9-methyl-9H-purin-6-yl)hydroxylamine (9c).
anti-Copalyl bromide (93 mg, 0.26 mmol) was dissolved in dry DMA (3.3 mL) under
Ar atm and compound 6 (67 mg, 0.24 mmol) was added. The resulting mixture was
stirred at ambient temperature for 20 min and evaporated in vacuo. The crude product
23

was purified by flash chromatography eluting with 5-10% MeOH in CH2Cl2; yield
108 mg (70%) 8c, colorless solid, and 15 mg (ca. 13%, contained ca. 30 mol% DMA)
1
9c, waxy purple solid. 8c; mp 194-196 ˚C. H NMR (400 MHz, DMSO-d6) δ 11.89
(s, 1H, NH), 9.36 (s, 1H, H-8 in purine), 7.92 (s, 1H, H-2 in purine), 5.43 (qt, J = 7.2,
1.0 Hz, 1H, H-14), 5.01-4.91 (m, 2H, H-15), 4.79 (s, 1H, H-17A), 4.47 (s, 1H, H-17B),
3.80 (s, 3H, NCH3), 2.32 (ddd, J = 12.7, 4.0, 3.4 Hz, 1H, H-7A), 2.10 (ddd, J = 14.0,
9.4, 4.0 Hz, 1H, H-12A), 1.88-1.78 (m, 2H, H-7B, H-12B), 1.75 (d, J = 1.0 Hz, 3H, H-
16), 1.69-1.63 (m, 2H, H-1A, H-6A), 1.56-1.47 (m, 3H, H-11A, H-2A, H-9), 1.43-1.39
(m, 2H, H-2B, H-11B), 1.35 (br d, 1H, J = 12.9 Hz, H-3A), 1.22 (dq, J = 12.9, 4.2 Hz,
1H, H-6B), 1.09 (dt, J = 13.4, 4.0 Hz, 1H, H-3B), 0.90-1.00 (m, 1H, H-5), 0.95 (s, 9H,
But), 0.82-0.87 (m, 1H, H-1B), 0.83 (s, 3H, H-18), 0.75 (s, 3H, H-19), 0.61 (s, 3H, H-
20), 0.18 [s, 6H, Si(CH3)2]; 13C NMR (100 MHz, DMSO-d6) δ 148.8 (C-2 in purine),
147.9 (C-8), 143.7 (C-13), 141.3 (C-4 in purine), 140.0 (C-6 in purine), 137.6 (C-8 in
purine), 116.7 (C-14), 110.0 (C-5 in purine), 106.4 (C-17), 55.2 (C-9), 54.8 (C-5),
47.2 (C-15), 41.7 (C-3), 39.1 (C-10), 38.5 (C-1) 37.7 (C-12), 37.6 (C-7), 33.3 (C-18),
33.2 (C-4), 31.7 (NCH3), 26.0 (CH3 in But), 23.9 (C-6), 21.5 (C-19), 20.9 (C-11),
15
18.8 (C-2), 18.0 (C in But), 16.7 (C-16), 14.3 (C-20), -5.2 [Si(CH3)2]; N NMR (50
MHz, DMSO-d6) δ -239.5 (N-1), -218.5 (N-9), -208.5 (N-7), -176.0 (N-3), N6 was
[훼]2D5
hidden; HRMS (ESI) found 552.4092, calcd for C32H54N5OSi+ 552.4092,
(c 0.26 MeOH). 9c; 1H NMR (600 MHz, DMSO-d ) δ 9.81 (br s, 1H, OH),
=+ 4.2
6
8.27 (s, 1H, H-2 in purine), 8.15 (s, 1H, H-8 in purine), 5.27-5.29 (m, 1H, H-14),
4.77-4.81 (m, 1H, H-15A), 4.76 (br s, 1H, H-17A), 4.61-4.65 (m, 1H, H-15B), 4.44 (br
s, 1H, H-17B), 3.72 (s, 3H, NCH3), 2.25 (br d, J = 11.7 Hz, 1H, H-7A), 2.01-2.04 (m,
1H, H-12A), 1.71-1.79 (m, 2H, H-12B, H-7B), 1.69 (s, 3H, H-16), 1.58-1.64 (m, 2H,
H-1A, H-6A), 1.46-1.52 (m, 2H, H-11A, H-9), 1.29-1.37 (m, 3H, H-2B, H-3A, H-11B),
1.17 (q, J = 12.6 Hz, 1H, H-6B), 1.05 (t, J = 12.5 Hz, 1H, H-3B), 0.90 (d, J = 12.4 Hz,
13
1H, H-5), 0.81 (s, 3H, H-18), 0.73 (s, 3H, H-19), 0.58 (s, 3H, H-20); C NMR (125
MHz, DMSO-d6) δ 155.1 (C-6 in purine), 151.5 (C-2 in purine), 150.7 (C-4 in
purine), 148.0 (C-8), 141.3 (C-8 in purine), 138.8 (C-13), 119.2 (C-14), 117.8 (C-5 in
purine), 106.2 (C-17), 54.7 (C-9), 54.5 (C-5), 49.7 (C-15), 41.5 (C-3), 39.0 (C-10),
24

38.3 (C-1), 37.6 (C-12), 37.5 (C-7), 33.2 (C-18), 33.1 (C-4), 29.4 (NCH3), 23.9 (C-6),
15
21.5 (C-19), 20.9 (C-11), 18.8 (C-2), 16.4 (C-16), 14.3 (C-20); N NMR (50 MHz,
DMSO-d6) δ -229.6 (N-9), -152.2 (N-3), -148.6 (N-1), -136.8 (N-7), N6 was hidden;
HRMS (ESI) found 438.3227, calcd for C26H40N5O+ 438.3227.
4.1.10.
7-[(E)-3,7-Dimethylocta-2,6-dienyl]-6-(hydroxyimino)-9-methyl-6,9-
dihydro-1H-purin-7-ium bromide (10a).
Compound 8a (54 mg, 0.11 mmol) was dissolved in MeCN-H2O (9:1, 1 mL) and
transferred to a vial containing NH4F (4 mg, 0.1 mmol) in MeCN-H2O (9:1, 1 mL).
Additional MeCN-H2O (9:1, 0.5 mL) was used to transfer the starting material to the
reaction vial. The resulting solution was stirred for 22 h at ambient temperature and
1
evaporated in vacuo; yield 41 mg (99%), colorless solid, mp 187-188 ˚C (dec.). H
NMR (400 MHz, DMSO-d6) δ 11.92 (br s, 1H, NH), 10.62 (s, 1H, OH), 9.22 (s, 1H,
H-8 in purine), 7.80 (s, 1H, H-2 in purine), 5.44-5.41 (m, 1H, H-2), 5.08-5.05 (m, 1H,
H-6), 4.97 (d, J = 7.1 Hz, 2H, H-1), 3.79 (s, 3H, NCH3), 2.10-2.06 (m, 2H, H-5),
2.04-2.02 (m, 2H, H-4), 1.77 (s, 3H, CH3 at C-3), 1.62 (s, 3H, H-8), 1.56 (s, 3H, CH3
at C-7); 13C NMR (100 MHz, DMSO-d6) δ 148.8 (C-2 in purine), 143.5 (C-3), 140.6
(C-4 in purine), 136.7 (C-6 in purine), 131.2 (C-7), 123.5 (C-6), 116.5 (C-2), 110.7
(C-5 in purine), 47.1 (C-1), 38.9 (C-4) 31.6 (NCH3), 25.7 (C-5), 25.4 (C-8), 17.6
15
(CH3 at C-7), 16.5 (CH3 at C-3); N NMR (50 MHz, DMSO-d6) δ -240.0 (N-1), -
218.4 (N-9), -209.2 (N-7), -179.4 (N-3), -105.5 (N6); HRMS (ESI) found 302.1975,
calcd for C16H24N5O+ 302.1975.
4.1.11.
6-(Hydroxyimino)-9-methyl-7-[(2E,6E,10E)-3,7,11,15-
tetramethylhexadeca-2,6,10,14-tetraenyl]-6,9-dihydro-1H-purin-7-ium bromide
(10b).
Compound 8b (118 mg, 0.19 mmol) was dissolved in MeCN-H2O (9:1, 4 mL) and
NH4F (6.9 mg, 0.19 mmol) was added as a powder. The vial previously containing
NH4F was rinsed with MeCN-H2O (9:1, 2 mL). The resulting solution was stirred for
40 min and evaporated in vacuo; yield 99 mg (>99%), colorless solid, mp 188-189
25

1
˚C. H NMR (600 MHz, DMSO-d6) δ 11.70 (br s, 1H, NH), 9.25 (s, 1H, H-8 in
purine), 7.78 (s, 1H, H-2 in purine), 5.41 (qt, J = 7.1, 1.1 Hz, 1H, H-2), 5.03-5.08 (m,
3H, H-6, H-10, H-14), 4.97 (d, J = 7.1 Hz, 1H, H-1), 3.78 (s, 3H, NCH3), 2.10-2.07
(m, 2H, H-5), 2.05-2.02 (m, 2H, H-4), 2.00-1.96 (m, 4H, H-9, H-13), 1.91-1.89 (m,
4H, H-8, H-12), 1.77 (s, 3H, CH3 at C-3), 1.62 (s, 3H, H-16), 1.54 (s, 3H, CH3 at C-
7), 1.53 (s, 3H, CH3 at C-15), 1.52 (s, 3H, H-18); 13C NMR (125 MHz, DMSO-d6) δ
148.9 (C-2 in purine), 143.5 (C-3), 140.7 (C-4 in purine), 136.8 (C-6 in purine),
136.7 (C-8 in purine), 134.8 (C-7), 134.3 (C-11), 130.6 (C-15), 124.1 (C-14), 123.8
(C-10), 123.4 (C-6), 116.5 (C-2), 110.7 (C-5 in purine), 47.1 (C-1), 39.2 (C-12), 39.1
(C-8), 38.9 (C-4), 31.6 (NCH3), 26.2 (C-13), 26.0 (C-9), 25.6 (C-5), 25.5 (C-16), 17.5
15
(CH3 at C-15), 16.6 (CH3 at C-3), 15.8 (CH3 at C-7), 15.7 (CH3 at C-11); N NMR
(50 MHz, DMSO-d6) δ -239.9 (N-1), -218.4 (N-9), -208.7 (N-7), -178.5 (N-3), -105.5
(N6); HRMS (ESI) 438.3226, calcd for C26H40N5O+ 438.3227.
4.1.12. (+)-6-(Hydroxyimino)-9-methyl-7-{(E)-3-methyl-6-[(1S,4aS,8aS)-5,5,8a-
trimethyl-2-methylenedecahydronaphthalen-1-yl]hex-2-enyl}-6,7-dihydro-1H-
purin-9-ium bromide [(+)-N6-Hydroxyagelasine D bromide‚ 10c].
Compound 8c (108 mg, 0.171 mmol) was dissolved in MeCN-H2O (9:1, 3 mL) and
NH4F (6.3 mg, 0.17 mmol) was added as a powder. The vial previously containing
NH4F was rinsed with MeCN-H2O (9:1, 2 mL). The resulting solution was stirred for
1 h 30 min and evaporated in vacuo; yield 88.5 mg (>99%), colorless solid, mp 188-
1
13
15
189 ˚C. H NMR, C NMR and N NMR data are presented in Table 1. HRMS
[훼]2D0 =+ 6.2
(ESI) found 438.3228, calcd for C26H40N5O+ 438.3227;
(c 0.26
MeOH).
4.1.13.
N-[4-Amino-6-(methylamino)pyrimidin-5-yl]-N-{(E)-3-methyl-5-
[(1S,4aS,8aS)-5,5,8a-trimethyl-2-methylenedecahydronaphthalen-1-yl]pent-2-
enyl}formamide (12d).
(+)-Agelasine D (11d) (25 mg, 0.050 mmol) was treated with 2M NaHCO3 solution
(12.5 mL) for 20 min at ambient temperature, before the mixture was extracted with
26

CH2Cl2 (350 mL), and evaporated in vacuo. The crude product was purified by flash
chromatography eluting with 2-10% MeOH in CH2Cl2. Due to a low yield, the
extraction was performed a second time on the basic water solution with CH2Cl2 (24
1
mL) yielding more product; total yield 13 mg (57%), pale yellow waxy solid. H
NMR (600 MHz, CDCl3) δ 8.14 (s, 1H, H-2 in pyrimidine), 7.95 (s, 1H, NCHO),
5.30 (t, J = 7.7 Hz, 1H, H-14), 4.95 (br s, 2H, NH2), 4.87 (br s, 1H, NH), 4.79 (s, 1H,
H-17A), 4.40 (s, 1H, H-17B), 4.14-4.12 (m, 2H, H-15), 2.98-2.96 (m, 3H, NCH3),
2.37-2.34 (m, 1H, H-7A), 2.12-2.07 (m, 1H, H-12A), 1.92 (ddd, J = 14.0, 13.4, 4.8 Hz,
1H, H-7A), 1.76-1.67 (m, 3H, H-6A, H-1A, H-12B), 1.60 (d, J = 5.0 Hz, 3H, H-16),
1.55-1.44 (m, 4H, H-9, H-11A, H-2A, H-2B), 1.38-1.36 (m, 1H, H-3A), 1.34-1.25 (m,
2H, H-11B, 1H, H-6B), 1.15 (ddd, J = 13.5, 13.4, 4.1 Hz, 13.4, 1H, H-3B), 1.04 (dd, J
= 12.7, 2.6 Hz, 1H, H-5), 0.93-0.96 (m, 1H, H-1B), 0.85 (s, 3H, H-18), 0.77 (s, 3H,
H-19), 0.64 (s, 3H, H-20); 13C NMR (125 MHz, CDCl3) δ 164.3 (NCHO), 160.5 (C-4
in pyrimidine), 158.9 (C-6 in pyrimidine), 156.4 (C-2 in pyrimidine), 148.5 (C-8),
144.2 (C-13), 117.0 (C-14), 106.2 (C-17), 99.3 (C-5 in pyrimidine), 56.4 (C-9), 55.5
(C-5), 42.1 (C-3), 41.4 (C-15), 39.7 (C-10), 39.1 (C-1), 38.6 (C-12), 38.3 (C-7), 33.6
(2C, C-4, C-18), 28.1 (NCH3), 24.4 (C-6), 21.9 (C-11), 21.7 (C-19), 19.4 (C-2), 16.4
(C-16), 14.5 (C-20); HRMS (ESI) found 440.3385, calcd for C26H42N5O+ 440.3384;
[훼]2D5 =+ 5.6
(c 0.50 MeOH). The spectral data and absolute value for optical
rotation were in good agreement with those reported for (–)-ageloxime D before.3
4.2. Biology
4.2.1. Activity against M. tuberculosis
MICs against replicating and non-replicating M. tuberculosis H37Rv were
determined using the Microplate Alamar Blue Assay and Low Oxygen Recovery
Assay, respectively.22
4.2.2. Activity against S. aureus, E. coli, C. albicans, L. infantum, T. cruzi, T.
brucei, T. rhodesiense and MRC-5 cells
27

4.2.2.1. Compounds and reagents. Compounds stock solutions were prepared in
100% DMSO at 20 mM. The compounds were serially pre-diluted (2-fold or 4-fold)
in DMSO followed by a further dilution in demineralized water to assure a final in-
test DMSO concentration of <1%.
4.2.2.2. Bacteria, fungi, parasite and cell cultures. S. aureus ATCC6538 and E.
coli ATCC8739 were cultured in MHT (Mueller Hinton Broth) and maintained on
TSA (Tryptone Soy Agar). C. albicans (azole resistant) was cultured in RPMI-1640
medium supplemented with Mops buffer and glucose and maintained on PDA (Potato
Dextrose Agar). L. infantum MHOM/MA(BE)/67 was maintained in the golden
hamster. Amastigotes were collected from the spleen of an infected donor hamster
using three centrifugation purification steps and spleen parasite burdens are assessed
using the Stauber technique.23 Primary peritoneal mouse macrophages were used as
host cell and were collected 2 days after peritoneal stimulation with 2% potato starch
suspension. T. cruzi, Tulahuen CL2 galactosidase strain (nifurtimox-sensitive),24 was
maintained on MRC-5sv2 (human lung fibroblast) cells in MEM medium,
supplemented with 200 mM L-glutamine, 16.5 mM NaHCO3, and 5% inactivated
fetal calf serum. T. b. brusei Squib 427 strain (suramine-sensitive) and T. b.
rhodesiense STIB-900 strain were maintained in Hirumi (HMI-9) medium,
supplemented with 10% inactivated fetal calf serum. MRC-5sv2 cells were cultured in
MEM + Earl's salts-medium, supplemented with L-glutamine, NaHCO3, and 5%
o
inactivated fetal calf serum. All cultures were conducted at 37 C under an
atmosphere of 5% CO2.
4.2.2.3. Activity against S. aureus, E. coli, C. albicans and MRC-5 cells. The
o
assays were performed at 37 C in sterile 96-well microtiter plates, each well
containing the aqueous compound (10 L) dilution together with
bacterial/fungal/MRC-5 inoculum (190 L; 5103 CFU/mL). The MRC-5 assay was
performed under an atmosphere of 5% CO2. The compounds were tested at 64, 16, 4,
1 and 0.25 M. Bacterial/fungal/MRC-5 growth was compared to untreated-control
28

wells (100% growth) and medium-control wells (0% cell growth). After 17 h (S.
aureus and E. coli), 24 h (C. albicans) or 3 days (MRC-5) incubation, viability was
assessed fluorimetrically after addition of resazurin. After 0.5 h (S. aureus, E. coli
o
and C. albicans) or 4 h (MRC-5) at 37 C, fluorescence was measured (ex 550 nm,
em 590 nm). The results are expressed as % reduction in growth/viability compared
to control wells and an IC50 (50% inhibitory concentration) was calculated.
Doxycycline was used as positive control for S. aureus and E. coli, flucytosine for C.
albicans and tamoxifen for MRC-5 cells.
o
4.2.2.4. Activity against L. infantum. The assay was performed at 37 C under an
atmosphere of 5% CO2 in sterile 96-well microtiter plates, each well containing the
aqueous compound (10 L) dilution together with macrophage/parasite inoculum
(190 L; 3.104 cells + 4,5.105 parasites/well). The inoculum was prepared in RPMI-
1640 medium, supplemented with 200 mM L-glutamine, 16.5 mM NaHCO3, and 5%
inactivated fetal calf serum. The macrophages were infected after 48 h and the
compounds were added after 2 h of infection. The compounds were tested at 64, 16,
4, 1 and 0.25 M. Parasite multiplication was compared to untreated-infected
controls and uninfected controls. After 5 days, parasite burdens (number of
amastigotes/macrophage) were microscopically assessed after staining with a 10%
Giemsa solution. The results were expressed as % reduction in parasite burden
compared to control wells and an IC50 was calculated. Miltefosine was used as
positive control.
o
4.2.2.5. Activity against T. cruzi. The assay was performed at 37 C under an
atmosphere of 5% CO2 in sterile 96-well microtiter plates, each well containing the
aqueous compound (10 L) dilution together with MRC-5 cell/parasite inoculum
(190 L; 4.103 cells/well + 4.104 parasites/well). The compounds were tested at 64,
16, 4, 1 and 0.25 M. Parasite growth was compared to untreated-infected controls
and uninfected controls. After 7 days incubation, parasite burdens were assessed after
adding the substrate CPRG [chlorophenolred -D-galactopyranoside; 50 L/well of
29