Synthesis and cellular effects of cycloterpenals: Cyclohexadienal-based activators of neurite outgrowth

Article abstract of DOI:10.1016/j.bmc.2008.07.030

An unusual class of diterpenoid natural products, 'cycloterpenals' (with a central cyclohexadienal core), that arise in nature by condensation of retinoids and other isoprenes, have been isolated from a variety of organisms including marine sponges as well as from the human eye. A milk whey protein has also demonstrated the formation of a cycloterpenal derived from β-ionylidineacetaldehyde. Here, we generate a synthetic library of these molecules where we detail reaction conditions required to effect cross condensation of α,β-unsaturated aldehydes as opposed to homodimerization. The ability of this class of molecules to activate neurite outgrowth activity is reported.

Full text of DOI:10.1016/j.bmc.2008.07.030

Bioorganic & Medicinal Chemistry 16 (2008) 7573–7581
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and cellular effects of cycloterpenals: Cyclohexadienal-based
activators of neurite outgrowth q
*
Bennie J. Bench, Shane E. Tichy, Lisa M. Perez, Jenna Benson, Coran M. H. Watanabe
Department of Chemistry, Texas A&M University, TAMU-3255, College Station, TX 77843, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 10 December 2007
Revised 5 July 2008
Accepted 14 July 2008
Available online 19 July 2008
An unusual class of diterpenoid natural products, ‘cycloterpenals’ (with a central cyclohexadienal core),
that arise in nature by condensation of retinoids and other isoprenes, have been isolated from a variety of
organisms including marine sponges as well as from the human eye. A milk whey protein has also dem-
onstrated the formation of a cycloterpenal derived from b-ionylidineacetaldehyde. Here, we generate a
synthetic library of these molecules where we detail reaction conditions required to effect cross conden-
sation of
a,b-unsaturated aldehydes as opposed to homodimerization. The ability of this class of mole-
Dedicated to Prof. Robert S. H. Liu at the
University of Hawaii, Manoa in honor of his
70th birthday
cules to activate neurite outgrowth activity is reported.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Terpene
Proline
Imine addition
Neurite outgrowth
1. Introduction
CO2H
CO2H
The carotenoids are responsible for such processes as light
absorption as well as for the inactivation of free radicals.1 The ret-
inoids, for example, the natural and synthetic derivatives of vita-
min A, serve as chromophores of the visual transduction system
(Fig. 1).2 Retinoids also exhibit a variety of other biological activi-
ties. They are used extensively in dermatology for a variety of con-
ditions, show activity as cancer preventative agents, and are potent
modulators of growth and differentiation.3–5 In this regard, all-
trans retinoic acid (1) has been shown to differentiate embryonic
stem cells into a neuron-like phenotype.6 It has also been em-
ployed as a cancer chemotherapeutic in treatment of promyelo-
cytic leukemia by causing terminal differentiation of the
malignant cells4 while retinoic acid analogs (TTNPB 2, azulenic
compounds 3, and 4) have been shown to possess biological activ-
ity as cancer chemopreventative agents.2
Retinoic Acid 1
TTNPB 2
CO2H
CO2H
3
4
Figure 1. Structures of representative retinoid compounds.
Interestingly, self-condensation of polyene aldehydes has been
observed in nature resulting in terpenes with cyclodimeric struc-
tures. We have defined these molecules as cycloterpenals as they
are composed of a cyclohexadiene core formed by the condensa-
tion of two terpene aldehydes. Cyclocitral (5), for example, was
O
O
q
A U.S. patent on this work has been filed (US-2007-0232813-A1). The patent
6
5
application publication can be accessed at http://www.uspto.gov/patft/.
* Corresponding author.
E-mail address: watanabe@mail.chem.tamu.edu (C.M.H. Watanabe).
Figure 2. Structures of citral and retinal self-condensation products.
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.07.030

7574
B. J. Bench et al. / Bioorg. Med. Chem. 16 (2008) 7573–7581
recently isolated from the North Sea bryozoan Flustra foliacea and
shown to exhibit antibiotic activity (Fig. 2),7–9 while cycloretinal
(6) has been suggested as a contributor of age-related macular
degeneration.10,11 Furthermore, incubation of the milk whey pro-
tein beta-lactoglobulin with b-ionylideneacetaldehyde has been
shown to result in cyclohexadienal formation.12
On these bases we broadly speculate that some of these cyclo-
terpenals or their structurally and functionally related analogs
could be important in developmental processes such as neurite
outgrowth. To begin to address this issue, we designed a synthetic
library of cyclohexadienals and tested the composite library (ring-
fused heterodimers and homodimers, Fig. 3)13 against PC12 cells,
Figure 3. Chemical structures of the cycloterpenal-based compound library.

B. J. Bench et al. / Bioorg. Med. Chem. 16 (2008) 7573–7581
7575
Table 1
H
N
COO-
Reactions involving a single b-methyl substituted aldehyde
R4
R1
R3
R2
+
R2
O
O
β
H
α
O
+
Donor
Acceptor
Homodimer
+
R1
R2
O
O
R1
R2
R2
R3
R2
R3
R4
CHO
CHO
CHO
CHO
R1a
R2a
Benzyl
Dimethylamino-benzyl
Nitrobenzyl
Benzyl
Dimethylamino-benzyl
Nitrobenzyl
Benzyl
Dimethylamino-benzyl
Nitrobenzyl
H–
Benzyl
Dimethylamino-benzyl
Nitrobenzyl
Benzyl
Dimethylamino-benzyl
Nitrobenzyl
Benzyl
Dimethylamino-benzyl
Nitrobenzyl
Benzyl
Dimethylamino-benzyl
Benzyl
Dimethylamino-benzyl
Nitrobenzyl
Comp #
Yieldb (%)
eec (%)
Thienyl
Thienyl
Thienyl
7
8
9
48
64
23
48
62
22
45
75
16
57
47
59
22
50
58
34
53
64
34
42
56
55
64
22
51
47
57
35
12
5
56
53
49
63
72
50
52
62
53
—
43
62
49
57
54
56
64
54
51
36
41
56
56
50
—
R1
R1
R1
R1
Figure 4. Substituted heterodimeric cyclohexadienals generated by proline-med-
iated condensation of ,b-unsaturated aldehydes.
CH3 furanyl
CH3 furanyl
CH3 furanyl
Biphenyl
Biphenyl
Biphenyl
Biphenyl
Fluorenyl
Fluorenyl
Fluorenyl
Naphthyl
Naphthyl
Naphthyl
Benzyl
Benzyl
Benzyl
2-Cl thienyl
2-Cl thienyl
Retinyl
Retinyl
Retinyl
Retinyl
b-Ionyl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
a
where promising effects were observed on neurite outgrowth. In
some instances, involving aromatic substitutions to the central
ring, the cyclohexadiene ring system could be aromatized to give
fluorescent molecules. In the future, some of these molecules or
their respective derivatives might find use as fluorescent probes
in protein target identification of a particular cyclohexadienal.
2. Results and discussion
2.1. Synthesis of asymmetric cyclohexadienals
The cycloterpenal-based compound library (Fig. 3) was gener-
ated by proline-promoted condensation of
hydes giving mono-, di-, tri-, and
cyclohexadienals (Fig. 4). The reaction involves condensation be-
tween a donor and acceptor aldehyde, where donor molecules re-
quire the presence of either a b-methyl- or b-methylene group.
Acceptor molecules exhibit a wider degree of flexibility, the single
a
,b-unsaturated alde-
tetra-substituted
H–
Benzyl
Dimethylamino-benzyl
Nitrobenzyl
Benzyl
Dimethylamino-benzyl
Benzyl
Benzyl
Dimethylamino-benzyl
Nitrobenzyl
56
59
53
52
40
62
56
48
54
b-Ionyl
b-Ionyl
Dimethylallyl
Dimethylallyl
Farnesenyl
Methyl
constraint owing to the presence of an
a,b-unsaturated aldehyde.
15
60
63
47
As demonstrated by NMR, MS, and LC/MS analyses (see Supple-
mental material), the reaction proceeds by complete conversion
of aldehyde to that of the Schiff base followed by a Diels–Alder
or Michael-like imine addition.13
Methyl
Methyl
a
Reaction conditions: 1 (donor): 1.5 (acceptor): 3.8 (L-proline); the acceptor was
dissolved in ethanol, reacted with 2 h with L-proline to which donor was added and
stirred at room temperature for 16–24 h.
2.2. Reaction optimization and scope
b
Isolated yield.
c
Determined by Pr(hfc)3 chiral shift reagent.
Here, we detail optimal reaction conditions to favor heterodi-
mer formation over competing homodimerization,13 which is
inherent to the reaction. The reaction is, therefore, not amenable
to a one-pot strategy (unlike homodimerization) as minimal yields
are obtained under such conditions. Moreover, we illustrate the
scope of the reaction by invoking substrates that vary in function-
ality and architecture.
appears at the base-line with concomitant loss of the starting alde-
hyde) prior to the addition of the donor aldehyde. Intuitively, this
should minimize formation of the cyclic homodimer 81, while
maximizing heterodimer 8 yields. To our satisfaction, an overall
enhancement in reaction yield (from 5% to 64%) was indeed ob-
tained. Such a strategy greatly facilitates reaction yields except in
the case of 4-nitrocinnamaldehyde, which undergoes unknown
side reactions after 13 min as detected by NMR (see Supplemental
material). With this method, we were able to limit the formation of
the homodimer product anywhere from 1% to 14% dependent upon
the substrate.
Other factors found to influence the cross-condensation reac-
tion included substrate (donor to acceptor) ratios and temperature
effects. For instance, varying the ratio of donor to that of the accep-
tor (in a 1:1, 1:1.5, 1:2, and 1:3), at a set catalyst ratio (1:1.5 sub-
strate to catalyst ratio, shown to be optimal from previous studies),
gave surprisingly higher yields at 1:1 (21.4%) and 1:1.5 (64%) molar
concentrations. At 1:2 and 1:3 molar concentrations, the overall
heterodimer yields were 14.0% and 11.2%, respectively. Based upon
2.2.1. Catalysis involving a single b-methyl substituted
substrate
Employing a one-pot strategy, reaction of 3-thiophen-2-yl-but-
2-enal with 4-(dimethyl-amino)cinnamaldehyde (in a 1:1 ratio)
gave a 5% and 43% yield, respectively, of heterodimer 8 to homodi-
mer 81 (Table 1). Likewise, skewing the ratio in favor of the accep-
tor aldehyde substrate over that of the b-methyl substituted
aldehyde donor in a one-pot format gave poor yields of the hetero-
dimer 8. Therefore, we implemented this reaction as a model sys-
tem and optimized its conditions in an effort to define
a
generalizable set of reaction conditions to maximize heterodimer
yields.
As
a,b-unsaturated aldehydes lacking a b-methyl group fail to
self-dimerize (confirmed through ESI mass spectroscopy) and the
reaction can be envisioned to proceed through formation of a Schiff
base,13,14 we examined the notion of pre-forming the proline Schiff
base of the acceptor aldehyde (as monitored by TLC, the Schiff base
these findings, we employed 1 (donor): 1.5 (acceptor): 3.8 (L-pro-
line) molar concentration ratios to effect synthesis of the remain-
der of the heterodimers involving a single b-methylenic substrate.

7576
B. J. Bench et al. / Bioorg. Med. Chem. 16 (2008) 7573–7581
Table 3
Reaction with b-ethyl substituted donor aldehydes
R2
H
γ
O
O
O
+
Homodimer
+
R1
R2
O
O
R1
5*
5
aR1
aR2
Compound
bYield (%)
cee (%)
Thienyl
Thienyl
Thienyl
Benzyl
Dimethylamino benzyl
Nitrobenzyl
52
53
54
49
50
25
52
50
51
H
H
O
O
O
H
H
a
Reaction conditions: 1 (donor): 1.5 (acceptor): 3.8 (L-proline); the acceptor was
35
55
56
dissolved in ethanol, reacted for 2 h to which donor was added and stirred at room
temperature for 16–24 h.
Figure 5. Schematic representation of products resulting from reaction of cinna-
maldehyde with citral.
b
Isolated yield.
c
Determined by Pr(hfc)3 chiral shift reagent.
heterodimeric products, respectively (Table 2). To achieve these
cross-condensation reactions, reaction conditions were modified
slightly. A 1:1 ratio of each aldehyde substrate and 3 equivalents
of L-proline were optimal. Skewing the ratio (1:2) only enhanced
homodimer formation of the more prevalent substrate.
Temperature effects on the reaction paralleled that observed
with self-condensation reactions,13 whereby product yields were
highest at ambient temperature versus ꢀ20 to 0 °C or 50 to
100 °C. In all cases the concentration of catalyst, L-proline, in these
reactions was held constant at 3.8 equivalents since these amounts
Implementing these optimized conditions, we examined the
generality of the reaction. Sterics appeared to govern these cross
condensations, as the only heterodimer product formed gave the
bulkiest substituent at C-4 and the least hindered group at C-6
with the exception of 41 and 42, in which both heterodimeric
products were formed. Likewise, reaction of citral and crotonalde-
hyde gave rise to tri-substituted cyclohexadienal 51, with the
dimethylallyl-linkage at C-5 (Fig. 3). Reaction yields ranged from
14% to as high as 67%, with the highest yields arising from conden-
sation with an acceptor aldehyde where R2 is H– or CH3–. On the
contrary, the lowest yields were obtained from reaction with an
acceptor aldehyde where R2 is thienyl or methylfuranyl. While
one might expect high level of homodimer to be formed with this
one-pot strategy, the isolated yields ranged from 3% to 27% in each
case.
were shown to give the greatest extents of conversion in homodi-
mer reactions.13 Use of
L-proline (3.8 equivalents) maintains a sub-
strate to catalyst ratio of 1:1.5, where substrate equals donor plus
aldehyde.
Utilizing these optimized conditions, we investigated the over-
all scope of the reaction as illustrated in Table 1. In most cases the
heterodimers were generated in modest yields. The lowest yields
were obtained with heterodimers bearing dimethylallyl-35 and -
36 or farnesenyl-37 substituents giving reaction yields in the 5–
15% range, which presumably is attributed to the flexibility of
the side chains and competing side reactions (Fig. 5). Conversely,
the highest yields were achieved with heterodimers bearing rigid
substituents (aromatic or other conjugated systems). Such trends
are also observed with homodimerization.13 As expected, in almost
all cases the cross-condensation reaction was favored by electron
donating groups as opposed to electron withdrawing substituents.
Homodimer yields ranged from 1% to 14% in each instance.
2.2.3. Catalysis effected by the b-methylene group of the donor
aldehyde
Interestingly, catalysis is not limited to reaction with b-methyl
substituted aldehydes. Substitution at C-5 of the cyclohexadienal
was readily observed as a result of deprotonation occurring from
2.2.2. Catalysis between two b-methyl substituted substrates
Proline-promoted synthesis between two b-methylenic sub-
strates would be expected to yield a mixture of substituted cyclo-
hexadienals, corresponding to two homodimeric and two
the c-methylene position of the donor aldehyde (Table 3). For in-
stance, use of a b-ethyl substituted donor aldehyde gives methyl-
substitution at C-5 of the tri-substituted cyclohexadienal. In this
regard, reaction of 3-thiophen-2-yl-pent-2-enal with three differ-
Table 2
Reaction between two b-methyl substituted aldehydes
ent a,b-unsaturated aldehydes resulted in the formation of hetero-
R2
R1
dimers shown in Table 3 as the major product (yields ranging from
25% to 50%) and approximately 10% homodimer in each instance.
While it might be conceivable to generate the seven membered
ring system by proton abstraction of the d-methyl group, formation
was not observed. Reaction conditions were the same as detailed
previously and similar yields and ee’s were observed.
O
O
+
+
R2
R1
R2
O
O
R1
+
Homodimers
R1a
R2a
Compound
Yieldb (%)
eec (%)
Biphenyl
Methyl
Methyl
Biphenyl
Thienyl
CH3 furanyl
H–
H–
Methyl
H–
41
42
43
44
45
46
47
48
49
50
51
49
18
14
29
57
67
40
45
39
23
59
55
57
54
50
49
48
52
47
44
46
46
Likewise, reacting cinnamaldehyde with citral gave three het-
erodimer products (Fig. 5), the expected di-substituted product
35 and two tri-substituted cyclohexadienals, cis and trans diaste-
reomers (55 and 56), respectively. Previous work has actually
Biphenyl
Biphenyl
Naphthyl
Fluorenyl
CH3furanyl
Thienyl
b-Ionyl
Dimethylallyl
Dimethylallyl
*
shown that formation of homodimer 5 can be favored (95:5) in
the presence of strong base, NaH resulting in a mixture of homodi-
mers.15 The compounds were assigned on the basis of computa-
tional analyses and NMR spectroscopy (as described in
Supplemental material). While we limited the formation of self-
H–
CH3 furanyl
H–
*
condensation products (5 and 5 ) by allowing cinnamaldehyde to
a
Reaction conditions: 1 (donor): 1 (acceptor): 3 (
ethanol and stirred at room temperature for 16–24 h.
L-proline) was dissolved in
go to complete Schiff base formation prior to the addition of citral,
homodimer 5 was still observed in 9% yield. Heterodimers 35, 55,
and 56 were formed in 12%, 25%, and 4% yield, respectively. As with
b
Isolated yield.
Determined by Pr(hfc)3 chiral shift reagent.
c

B. J. Bench et al. / Bioorg. Med. Chem. 16 (2008) 7573–7581
7577
Table 4
R2
R1
O
R1
+
R2
O
O
R1 and R2a
Product
Compound
Yieldb (%)
25
eec (%)
53
H
H
R1 dimethylallyl R2 benzyl
55
O
O
H
56
57
4
62
33
H
N
R1 dimethylallyl R2 dimethyl aminobenzyl
24d
H
H
O
O
H
N
58
59
41
44
H
O2N
R1 dimethylallyl R2 nitrobenzyl
18
H
O
H
R1 dimethylallyl R2 H-
R1 farnesenyl R2 benzyl
60
61
24d
18
45
63
O
H
O
H
N
R1 farnesenyl R2 dimethyl aminobenzyl
62
45
61
H
O
H
O2N
R1 farnesenyl R2 nitrobenzyl
R1 pentyl R2 benzyl
63
64
65
23
29
31
55
57
55
H
O
H
H
O
H
N
R1 pentyl R2 dimethyl aminobenzyl
H
O
H
a
Reaction conditions: 1 (donor): 1.5 (acceptor): 3.8 (
temperature for 16–24 h.
L-proline) the acceptor was dissolved in ethanol, reacted for 2 h to which donor was added and stirred at room
b
Isolated yield.
Determined by Pr(hfc)3 chiral shift reagent.
Diastereomers were inseparable by HPLC.
c
d

7578
B. J. Bench et al. / Bioorg. Med. Chem. 16 (2008) 7573–7581
previous systems, these reactions favored substituents with elec-
tron donating groups as opposed to those bearing electron with-
drawing substituents. In this regard, reactions involving
nitrobenzyl groups gave correspondingly lower yields than those
involving benzyl or dimethylamino-benzyl substituents.
differentiate into neurons. The PC12 assay is a widely implemented
method of identifying compounds that promote neurite outgrowth
or survival.17
The PC12 cells were plated onto collagen Type IV plates at
10,000 cells/well, allowed to adhere for 24 h, followed by the addi-
These b-methylene driven reactions were also demonstrated
with farnesenal, b-methyl substituted nonenal in combination
with acrolein, cinnamaldehyde and its derivatives (Table 4). In
most instances only the cis product was formed; citral was pur-
chased as a mixture of cis and trans isomers while all others were
obtained as a single (all-trans) isomeric form.
tion of compound at concentrations of 100, 10 and 1 lg/mL,
respectively. The cells were monitored over a period of two weeks,
where the most pronounced effects were observed at about one
week of exposure. From our library of cyclohexadienals, two het-
erodimers (44 and 21, Fig. 7) were able to successfully stimulate
neurite outgrowth, while one homodimer 7417 supported the
development of neural extensions. Therefore, neurite outgrowth
activity appears to be associated with aromatic substituents or ex-
tended hydrocarbon chains. Intriguingly, biphenyl groups have
been thought to serve as mimics of protein alpha helices.18 Long
aliphatic chains, on the other hand, are capable of inserting into li-
pid bilayers.
To assess the degree to which stereochemistry dictates pheno-
type in the neurite outgrowth assay, dimers 21, 44, and 74 were
derivatized into their corresponding menthylhydrazone deriva-
tives19 to allow separation of the stereoisomers by HPLC. Following
HPLC purification, the compounds were each hydrolyzed giving
each aldehyde in enantiomerically pure form. Upon reanalysis in
the PC12 assay, the R-configuration was shown to be responsible
for neurite outgrowth activity while the S-configuration exhibited
no activity (see Supplemental material for photos). In the future, it
will be interesting to establish whether heterodimers 21 and 44
share the same target as the aliphatic homodimer 74.
2.3. Aromatization and fluorescence
Interestingly, phenyl substituted cyclohexadienals (derived from
cinnamaldehyde and its substituted analogs) were found to sponta-
neously aromatize (Fig. 6) to give fluorescent molecules (over a per-
iod of one to several days) in about 5–15% conversion. Aromatization
could also be invoked directly by reaction with potassium perman-
ganate in quantitative yields.16 With further experimentation, some
of these molecules or their respective derivatives might therefore
find use as fluorescent probes to identify the protein binding partner
of a particular cyclohexadienal. Quantum yields20 are provided here
(Table 5), as measured in dichloromethane and methanol, for a sub-
set of the molecules including those that were most fluorescent as
judged by exposure to UV light (structures of additional compounds
isolated are provided as Supplemental material). In all cases, fluores-
cenceofmoleculeswas quenchedinmethanol. Thehighest quantum
yields were obtained for compounds 71 and 73 (in dichlorometh-
ane), where the substituent at C-4 is not conjugated to the central
aromatic ring. While nitro-substituted phenyl systems fluoresce
mildly, the nitro-group was shown to significantly quench the quan-
tum yield.
Table 5
Quantum yields of a representative set of fluorescent cyclohexadienals
Entry % Conv.
CH2Cl2
MeOH
kabs (nm) kem (nm) ua
kabs (nm) kem (nm) ua
66
67
68
69
70
71
72
73
10
14
7
13
14
11
15
8
368
368
368
374
368
363
368
368
490
559
474
520
526
502
519
501
0.025
0.016
0.042
0.112 384
0.046 368
0.699 354
0.088 368
0.312 368
360
368
374
503
413
598
434
497
396
515
522
0.008
0.002
0.056
0.005
0.044
0.002
0.005
0.010
2.4. Effects on neurite outgrowth and general cytotoxicity
Given the demonstrated ability of retinoic acid and its analogs
to modulate growth and cellular differentiation and the ability of
a milk whey protein to generate cyclo-b-ional, the cycloterpenal-
based library of compounds was evaluated for their ability to
stimulate neurite outgrowth in a PC12 assay. PC12 cells are a
pluripotent cell line that was originally cloned in 1976 from a
transplantable rat pheochromocytoma and exhibits the ability to
a
Measured as specified previously.20Quinine sulfate was used as standard for
quantum yield measurements (u = 0.55 in H2SO4).
N
NO2
N
N
N
N
N
O
O
O
O
O
O
O
O
66
67
68
69
70
71
72
73
Figure 6. Representative set of fluorescing aromatized heterodimers.

B. J. Bench et al. / Bioorg. Med. Chem. 16 (2008) 7573–7581
7579
A Control, EtOH
B 1μg/mL, 6-days
O
O
44
N
D 1 μg/mL 5-days
C 10 μg/mL, 7-days
O
O
74
21
Figure 7. Stimulation of neurite outgrowth by cyclohexadienal heterodimers and homodimers: (A) Control, cells treated with absolute EtOH; (B) methylfuranyl, biphenyl-
cyclohexadienal, 1 g/mL at 6 days (C) dimethylamino-phenyl, naphthyl-cyclohexadienal; 10 g/mL 7 days (D) nonenal-cyclohexadienal, 1 g/mL 5 days.
l
l
l
As citral dimer 5 has been reported to possess antibiotic activ-
ity,7–9 we also surveyed our cyclohexadienal library of compounds
against bacterial strains (Escherichia coli, Bacillus subtilis) and yeast
(Saccharomyces cerevisiae). See Supplemental material for graphical
representation of data. All compounds were tested at three concen-
regeneration, which plays a significant role in the treatment of
neurodegenerative diseases or CNS injuries.
Experiments are now underway to investigate the enzymatic
reaction in the biosynthesis of cycloterpenals.
trations, 100, 10, and 1
bition was observed, where compounds exhibited MIC values
>100 g/mL. Three compounds (71, 59, and 43) were reasonably
active against B. subtilis with MIC values within the 10–25 g/mL
lg/mL. In most cases, moderate to no inhi-
3. Experimental
l
3.1. General methods
l
range. Against Jurkats, a human T-cell Leukemia cell line, while
the effects were again moderate for the majority of compounds
tested, a few (44, 15, 19, and 27) demonstrated complete inhibition
Fresh THF was dried by passing it through an MBRAUN solvent
purification system and stored over molecular sieves. All other sol-
vents and reagents were purchased and used without further purifi-
cation. Chemical shifts for 1H and 13C NMR spectra are reported in
ppm referenced to TMS (0 ppm) and coupling constants are reported
in hertz (Hz). MS spectra were recorded on an API QSTAR PULSAR
(ES) apparatus. IR spectra were recorded with a FT-IR spectrometer.
Chromatographic separations were achieved by flash silica chroma-
tography (silica gel 60 mesh, EMD Biosciences). Determination of
ee’s was achieved by Pr(hfc)3 chemical shift reagent. b-Methylenic
aldehydes (101–111) were synthesized by reduction of their corre-
sponding nitriles (91–100),21 which were derived from their respec-
tive methyl-ketone analogs.22 In all cases, the all-trans isomer was
utilized except in the case of citral which was purchased as a mix-
ture. Ring-fused homodimers were prepared by self-condensation
of the b-methylenic aldehyde substrates.13
of growth at 1
lg/mL.
2.5. Conclusion
The retinoids have achieved prominence for their therapeutic
benefits ranging from their ability to treat acne and skin blemishes
to their ability to differentiate stem cells.1–6 An unusual class of
diterpenoid natural products, ‘cycloterpenals’, have been isolated
from a variety of organisms including marine sponges and the hu-
man eye.7–11 Moreover, incubation of the milk whey protein beta-
lactoglobulin with b-ionylideneacetaldehyde has been shown to
give cyclo-b-ional (a C-30 ring-fused dimer).12 As such it is plausi-
ble that some of these cycloterpenals or their structurally and
functionally related analogs could be important in developmental
processes such as neurite outgrowth.
3.2. Organisms
In this investigation, we implemented a proline-promoted con-
densation reaction that enabled diversification of the cycloterpenal
skeleton to give a library of molecules with varying degrees of sub-
stitution. Specific reaction conditions required to favor cross con-
densation as opposed to self-condensation were detailed. In the
case of aromatic substituents, the cyclohexadiene ring system
could be aromatized to give fluorescent molecules, which in the
long term could lead to use of these molecules as probes in protein
target identification of a particular cyclohexadienal.
PC12 neuronal cells (CRL 1721) and B. subtilis strains 6633 were
obtained from the American Type Culture Collection, Rockville,
MD. The yeast S. cerevisiae wild-type strain [#404; BY4741; MATa
his3D1 leu2D0 met15D0 ura3D0] was obtained from Dr. Michael
Kladde at the Department of Biochemistry, Texas A&M University.
The E. coli strain DH10B was obtained from Prof. Dennis Gross,
Texas A&M University, Department of Plant Pathology.
The phenotypic effects of these cycloterpenal-based molecules
were examined in a PC12 assay, where dramatic effects were ob-
served on neurite outgrowth. These results suggest that molecules
with this cyclohexadienal motif might have applications in neurite
3.3. Neurite outgrowth assay
PC12 cells were maintained in 100 mm Collagen Type IV dishes
at 37 °C in a 5% CO2/air temperature. The cell culture medium

7580
B. J. Bench et al. / Bioorg. Med. Chem. 16 (2008) 7573–7581
consisted of RPMI 1640, with 10% heat-inactivated horse serum
and 5% fetal bovine serum. Cells were removed from the culture
flasks by flushing with fresh culture medium. Cells were then
transferred to a 15 mL falcon tube and centrifuged for 15 min at
1000 rpm. The medium was removed and the cells were resus-
pended in fresh culture medium. The cells were counted with a
hemocytometer and replated in a 96-well Collagen Type IV plate
at 10,000 cells/well with enough fresh medium to bring the vol-
and was washed with brine. The organic layer was dried over mag-
nesium sulfate (MgSO4) and concentrated under vacuum. The
crude product was purified by column chromatography (10% ethyl
acetate/hexanes) to yield 12.7 mg of the final product (48% yield)
as a red oil. 1H NMR (300 MHz, CDCl3, 25 °C), d = 9.54 (s, 1H),
7.12–7.28 (m, 5H), 7.10 (d, J = 6.3 Hz, 1H), 6.68 (dd, J = 2.6, 6.2 Hz,
1H), 6.45 (d, J = 3.3 Hz, 1H), 6.05–6.06 (m, 1H), 4.19 (dd, J = 1.8,
9.3 Hz, 1H), 2.87–3.09 (m, 2H), 2.34 (s, 3H); 13C NMR (75 MHz,
CDCl3, 25 °C) d = 191.9 (CH), 155.2 (C), 151.9 (C), 144.1 (CH),
142.9 (C), 137.9 (C), 134.0 (C), 128.7 (2 ꢁ CH), 127.4 (2 ꢁ CH),
127.0 (CH), 115.1 (CH), 112.6 (CH), 108.9 (CH), 34.0 (CH), 31.5
ume to 200 lL. The cells were allowed to adhere for 24 h prior to
the addition of compound. Each cyclohexadienal compound was
examined at three concentrations in duplicate, to give final concen-
trations of 100, 10, and 1
l
g/mL, respectively. Each well was visu-
(CH2), 14.2 (CH3); IR (neat) m 3012, 2962, 2941, 2326, 1735,
ally monitored over a period of 2 weeks for neurite outgrowth and
intercellular connections. Fresh medium was added when neces-
sary. Photographs were taken with an Olympus SP-310 camera
equipped with a modified lens piece that fits over the microscope
eyepiece.
1431, 1365, 1229, 1211, 1206, 913 cmꢀ1
; HRMS (ESI) for
C18H16O2Li (M+Li)+: calcd 271.1310, found 271.1365.
3.7. 4-(9H-fluoren-2-yl)-6-methylcyclohexa-1,3-
dienecarbaldehyde (Table 2, Entry 46): general procedure for
the proline-mediated condensation of two b-methylenic
aldehydes
3.4. Cytotoxicity assay against the human T-cell leukemia,
Jurkat cell line
In a dried 100 mL flask, 3-(9H-fluoren-2-yl)-but-2-enal (100 mg,
1 equivalent) and crotonaldehyde (30 mg, 1 equivalent) were dis-
solved in 15 mL of 200 proof ethanol. L-Proline (300 mg, 3 equiva-
Jurkat cells were maintained in 75 cm2 culture flasks at 37 °C
under a 5% CO2 atmosphere. The cell culture medium consisted
of RPMI 1640 with 10% fetal bovine serum. Cells were harvested
by centrifugation (20 min at 1000 rpm), resuspended in fresh cul-
ture medium and counted on a hemocytometer. The cells (1 mL)
were arrayed (200,000 cells/well) in 48-well plates to which com-
lents) was added to facilitate the reaction. The reaction mixture
was stirred for 16 h at room temperature before being quenched
by the addition of 20 mL of water. The reaction mixture was ex-
tracted with a 1:1 ethyl acetate/hexanes solution (3 ꢁ 50 mL) and
waswashedwithbrine. Theorganiclayerwasdriedover magnesium
sulfate (MgSO4) and concentrated under vacuum. The crude product
was purified by column chromatography (10% ethyl acetate/hex-
anes)to yield57.6 mg of thefinalproduct(67%yield)as a greensolid.
1H NMR (300 MHz, CDCl3, 25 °C), d = 9.56 (s, 1H), 7.77–7.90 (m, 2H),
7.55–7.59 (m, 2H), 7.25–7.44 (m, 3H), 6.90 (d, J = 6.0 Hz, 1H), 6.65
(dd, J = 2.7, 6.0 Hz, 1H), 3.93 (s, 2H), 3.05–3.15 (m, 1H), 2.95 (dq,
J = 2.7, 8.4 Hz, 1H), 2.76–2.78 (m, 1H), 1.04 (d, J = 7.2 Hz, 3H); 13C
NMR (75 MHz, CDCl3, 25 °C) d = 192.6 (CH), 145.7 (C), 143.9 (C),
143.8 (C), 142.8 (C), 142.7 (CH), 141.6 (C), 141.3 (C), 138.7 (C),
127.4 (CH), 127.2 (CH), 125.4 (CH), 125.0 (CH), 122.6 (CH), 120.4
(CH), 120.2 (CH), 119.1 (CH), 37.2 (CH2), 33.9 (CH2), 24.4 (CH), 18.1
pound was added to give final concentrations of 100, 10, and 1 lg/
mL, respectively. Cells were incubated for 24 h, transferred to
1.5 mL eppendorf tubes and centrifuged for 15 min at
14,000 rpm. The medium was removed and the cell pellet resus-
pended in 1 mL of phosphate buffer saline (PBS) to wash the cells.
The cells were again centrifuged for 15 min at 14,000 rpm. The PBS
was removed and the cell pellet resuspended in a lysis buffer. The
lysed cells were then placed in a black 96-well plate and analyzed
using a Bio-Tek fluorometer to measure relative fluorescence.
3.5. Anti-bacterial and anti-fungal assays
(CH3); IR (neat)
m 3053, 2962, 2925, 2868, 1733, 1663, 1546, 1423,
Microbes (Saccharomyces cerevisiae, Escherichia coli, Bacillus sub-
tilis 6633 and Bacillus subtilis 21332) were cultured and maintained
on agar plates (see Supplemental material for media conditions). A
single colony was used to inoculate 3 mL of culture medium and
1265, 1183, 1046, 825, 733, 703 cmꢀ1 HRMS (ESI) for C21H18
(M+H)+: calcd 287.1436, found 287.1463.
O
3.8. Determination of quantum yields
allowed to grow for 16 h. The cells (500
50 mL of fresh medium to an OD of 0.1 and aliquoted into a 96-well
plate (100 L/well). The assays were set up in duplicate and the
compounds were tested to give final concentrations of 100, 10,
and 1 g/mL. Plates were incubated (E. coli and B. subtilis, 37 °C;
lL) were diluted with
Relative quantum yields20 were obtained with a photon count-
ing spectrofluorometer equipped with a photomultiplier tube,
which is sensitive up to 850–900 nm. The slit width was 0.5 nm
for both excitation and emission sources. The quantum yield was
calculated with the following equation:
l
l
yeast, 30 °C) and shaken (250 rpm) for 16 h, and cell density was
(OD) measured with a Bio-Tek microplate reader.
/x ¼ /stðIx=IstÞðAst=AxÞðgx2
=
g2st
Þ
3.6. 4-(5-Methyl-furan-2-yl)-6-phenyl-cyclohexa-1,3-
where the x subscript denotes unknown, st denotes standard, ust is
the reported quantum yield of the standard, I is the integrated emis-
dienecarbaldehyde (Table 1, Entry 10): general procedure for
the proline-mediated condensation of the
aldehydes
a,b-unsaturated
sion spectra, A is the absorbance at the excitation wavelength and
g
is the refractive index of the solvent used. Quinine sulfate (u = 0.55
in 0.1 M H2SO4) was used as the standard.
The
equivalents) was dissolved in 200 proof ethanol (8 mL) with 3.8
equivalents (72.0 mg) of -proline. After approximately 2 h, at
which point Schiff base formation is complete as monitored via
thin layer chromatography (TLC), one equivalent of b-methylenic
aldehyde 3-(5-methylfuran-2-yl)-but-2-enal (25 mg, 1 equivalent)
was added to the reaction mixture and stirred at ambient temper-
atures for 18 h. The reaction was monitored by TLC. Upon comple-
tion, the reaction was quenched by the addition of water (20 mL),
extracted with a 1:1 ethyl acetate/hexanes solution (3 ꢁ 50 mL),
a,b-unsaturated compound cinnamaldehyde (33.0 mg, 1.5
3.9. Synthesis of 2-isopropyl-5-methylcyclohexyl
hydrazinecarboxylate (112)19
L
To synthesize 2-isopropyl-5-methylcyclohexyl hydrazinecarb-
oxylate 112, ethyl 2-isopropyl-5-methylcyclohexyl carbonate was
first prepared by reacting 25.0 g (1 equivalent) of L-menthol with
75 mL of ethyl chloroformate (7.3 equivalents) in the presence of
pyridine (1 mL). The reaction was refluxed for 48 h before the light

B. J. Bench et al. / Bioorg. Med. Chem. 16 (2008) 7573–7581
7581
brown product was poured over ice followed by dilution with
20 mL of water. The mixture was then extracted with ethyl ether
(5 ꢁ 100 mL) with the organic layers combined, dried over MgSO4,
and concentrated under vacuum. Assuming 100% conversion of the
ethyl 2-isopropyl-5-methylcyclohexyl carbonate, 75 mL of 2-eth-
oxy ethanol was used to solubilize the starting material. Once com-
pletely dissolved, a twofold excess of hydrazine monohydrate
(20 mL) was being added, and then the reaction mixture was
poured over ice and extracted with ethyl ether (5 ꢁ 100 mL) with
the organic layers combined, dried over MgSO4, and concentrated
under vacuum. Once the ethyl ether was evaporated, 50 mL of hex-
anes was added and the solution was heated and left at ꢀ20 °C for
48 h in which white fluffy crystals were formed and were subse-
quently collected via vacuum filtration. The recrystallization was
performed three more times yielding 12.7 g (51% yield) of 112.
1H NMR (300 MHz, CDCl3, 25 °C), d = 6.14 (s, 1H), 4.56 (dt, J = 4.5,
21.9 Hz, 1H), 3.74 (s, 2H), 1.97–2.02 (m, 1H), 1.82–1.92 (m, 1H),
1.61–1.66 (m, 2H), 1.40–1.52 (m, 1H), 1.30 (t, J = 10.8 Hz, 1H),
1.00–1.10 (m, 1H), 0.91–0.99 (m, 2H), 0.85–0.89 (m, 6H), 0.75 (d,
J = 6.9 Hz, 3H); 13C NMR (75 MHz, CDCl3, 25 °C) d = 159.0 (C), 75.7
(CH), 47.5 (CH), 41.6 (CH2), 34.4 (CH2), 31.6 (CH), 26.4 (CH), 23.7
(CH2), 22.2 (CH3), 21.0 (CH3), 16.6 (CH3); HRMS (ESI) for
C11H22N2O2 (M+H)+: calcd 215.1760, found 215.1752.
ethyl ether (2 ꢁ 25 mL). The organic layer was then dried with
MgSO4 and concentrated under vacuum. The crude mixture was
then passed through a silica plug with a mobile phase of 10% ethyl
acetate/hexanes. HRMS (ESI) for C20H34O (M+H)+: calcd 291.2688,
found 291.2667.
Acknowledgments
We thank the Welch Foundation (A-1587), Texas A&M Univer-
sity, and the NIH CBI Training Program (T32 GM008523) for sup-
port of this work. We also thank Ms. Vanessa Santiago for mass
spectrometry analyses, Jiney Jose for help with quantum yield
determination, and Dr. Joseph Reibenspies for X-ray crystal struc-
ture determination.
Supplementary data
Experimental procedures, additional characterization data, 1H
and 13C NMR spectra, X-ray data, LC/MS data, NMR studies on
Schiff base formation are provided. Supplementary data associated
with this article can be found, in the online version, at doi:10.1016/
j.bmc.2008.07.030.
3.10. 2-isopropyl-5-methylcyclohexyl 2-((4,6-dihexyl-6-
methylcyclohexa-1,3-dienyl)methylene) hydrazinecarboxylate
(115): general procedure for the synthesis of the
menthylhydrazones19
References and notes
1. Krinsky, N. I. Pure Appl. Chem. 1994, 66, 1003–1010.
2. Asato, A. E.; Peng, A.; Hossain, M. Z.; Mirzadegan, T.; Bertram, J. S. J. Med. Chem.
1993, 36, 3137–3147.
3. Ganguly, J. Biochemistry of Vitamin A; Springer: Boca Raton, 1989. pp 195–207.
4. Bertram, J. S.; Kolonel, L. N.; Meyakens, F. L. Cancer Res. 1987, 47, 3012–3031.
5. Huang, M.-E.; Ye, Y.-C.; Chen, S.-R.; Chai, J.-R.; Lu, J.-X.; Zhoa, L.; Gu, L.-J.; Wang,
Z. –Y. Blood 1988, 72, 567–572.
6. Jones-Villeneuve, E. M. V.; McBurney, M. W.; Rogers, K. A.; Kanins, V. I. J. Cell
Biol. 1982, 94, 253–262.
7. Peters, L.; König, G. M.; Wright, A. D.; Pukall, R.; Stackebrandt, E.; Eberl, L.;
Reidel, K. Appl. Environ. Microbiol. 2003, 69, 3369–3475.
8. Peters, L.; Wright, A. D.; Krick, A.; König, G. M. J. Chem. Ecol. 2004, 30,
1165–1181.
9. Peters, L.; Wright, A. D.; Kehraus, S.; Gündisch, D.; Tilotta, M. C.; König, G. M.
Planta Med. 2004, 70, 883–886.
10. Fishkin, N.; Pescitelli, G.; Sparrow, J. R.; Nakanishi, K.; Berova, N. Chirality 2004,
16, 637–641.
11. Fishkin, N.; Sparrow, J. R.; Allikmets, R.; Nakanishi, K. PNAS 2005, 102,
7091–7096.
12. Li, X.; Asato, A. E.; Liu, R. S. H. Tetrahedron Lett. 1990, 31, 4841–4844.
13. Bench, B. J.; Liu, C.; Evett, C. R.; Watanabe, C. M. H. J. Org. Chem. 2006, 71,
9458–9463.
14. Asato, A. E.; Watanabe, C.; Li, X.-Y.; Liu, R. S. H. Tetrahedron Lett. 1992, 33,
3105–3108.
Homodimer 4,6-dihexyl-6-methylcyclohexa-1,3-dienecarbalde-
hyde 74 (0.100 g, 1 equivalent) was dissolved in 200 proof ethanol
(20 mL) with 1.5 equivalents (0.111 g) of 2-isopropyl-5-meth-
ylcyclohexyl hydrazinecarboxylate 112 at room temperature. After
72 h, the reaction was extracted with 1:1 ethyl acetate/hexanes
solution (2 ꢁ 100 mL), and was washed with brine. The organic
layer was dried over magnesium sulfate (MgSO4) and concentrated
under vacuum. The crude material was passed through a silica plug
with 10% ethyl acetate/hexanes as the mobile phase. The mixture
was then subjected to high pressure liquid chromatography (HPLC)
with a mobile phase of 96% ethyl acetate/4% hexanes with a flow
rate of 1.5 mL/min. Two fractions were collected at 11.524 and
12.501 min corresponding to the R and S stereoisomers, respec-
tively. HRMS (ESI) for C31H54N2O2 (M+H)+: calcd 487.4264, found
487.4238.
15. Taneja, S. C.; Koul, S. K.; Dhar, K. L. Indian J. Chem. 1988, 27B, 769–770.
16. Reddy, T. R. K.; Mutter, R.; Heal, W.; Guo, K.; Gillet, V. J.; Pratt, S.; Chen, B. J.
Med. Chem. 2006, 49, 607–615.
17. Culturing Nerve Cells; Banker, G., Goslin, K., Eds., 2nd ed.; MIT Press, 2002.
18. Okuyama, M.; Laman, H.; Kingsbury, S. R.; Visintin, C.; Leo, E.; Eward, K. L.;
Stoeber, K.; Boshoff, C.; Williams, G. H.; Selwood, D. L. Nat. Methods 2007, 4,
153–159.
3.11. General procedure for the conversion of 2-isopropyl-5-
methylcyclohexyl 2-((4,6-dihexyl-6-methylcyclohexa-1,3-
dienyl)methylene) hydrazinecarboxylate (115) back to 4,6-
dihexyl-6-methylcyclohexa-1,3-dienecarbaldehyde (74)19
19. Woodward, R. B.; Kohman, T. P.; Harris, G. C. J. Am. Chem. Soc. 1941, 63, 120–
124.
Once the R and S stereoisomers were purified, each was dis-
solved in 20 mL of 1:1 water/ethanol with 5 mL of 10% sulfuric acid
and boiled for 2.5 h in which a color change from clear to pink oc-
curred within 2 h. The reaction mixture was removed from the
heat source and allowed to cool when it was then extracted with
20. Rhys Williams, A. T.; Winfield, S. A. Analyst 1983, 108, 1067–1071.
21. Taber, D. F.; Raman, K.; Gaul, M. D. J. Org. Chem. 1987, 52, 28–34.
22. Uchikawa, O.; Fukatsu, K.; Tokunoh, R.; Kawada, M.; Matsumoto, K.; Imai, Y.;
Hinuma, S.; Kato, K.; Nishikawa, H.; Hirai, K.; Miyamoto, M.; Ohkawa, S. J. Med.
Chem. 2002, 45, 4222–4239.