Efficient synthesis and structure-activity relationship of honokiol, a neurotrophic biphenyl-type neolignan

Article abstract of DOI:10.1016/j.bmcl.2004.02.067

Honokiol, a biphenyl-type neolignan, which shows the remarkable neurotrophic effect in primary cultured rat cortical neurons, has been effectively synthesized in 21% yield over 14 steps starting from 5-bromosalicylic acid and p-hydroxybenzoic acid by utilizing Pd-catalyzed Suzuki-Miyaura coupling reaction as a key step. Additionally, the structure-activity relationship between neurite outgrowth-promoting activity and its O-methylated and/or its hydrogenated analogues was examined in the primary cultures of fetal rat cortical neurons, suggesting that 5-allyl and 4′-hydroxyl groups are essential for affecting the neurotrophic activity of honokiol.

Full text of DOI:10.1016/j.bmcl.2004.02.067

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2621–2625
Efficient synthesis and structure–activity relationship of honokiol,
a neurotrophic biphenyl-type neolignan
Tomoyuki Esumi,a Gouki Makado,a Haifeng Zhai,a Yasuhiro Shimizu,a
Yasuhide Mitsumotob and Yoshiyasu Fukuyamaa,*
aInstitute of Pharmacognosy, Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho,
Tokushima 770-8514, Japan
bSecond Institute of New Drug Research, Otsuka Pharmaceutical Co. Ltd, Tokushima 771-0192, Japan
Received 22 December 2003; accepted 17 February 2004
Abstract—Honokiol, a biphenyl-type neolignan, which shows the remarkable neurotrophic effect in primary cultured rat cortical
neurons, has been effectively synthesized in 21% yield over 14 steps starting from 5-bromosalicylic acid and p-hydroxybenzoic acid
by utilizing Pd-catalyzed Suzuki–Miyaura coupling reaction as a key step. Additionally, the structure–activity relationship between
neurite outgrowth-promoting activity and its O-methylated and/or its hydrogenated analogues was examined in the primary cultures
of fetal rat cortical neurons, suggesting that 5-allyl and 40-hydroxyl groups are essential for affecting the neurotrophic activity of
honokiol.
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
discovery of honokiol (1)7 together with several active
compounds from the bark of Magnolia obovata Thunb.8
In previous papers, we reported that 1 has the significant
neurotrophic property, that is, neurite outgrowth-pro-
moting and neuroprotective effects in the primary
cultures of fetal rat cortical neurons in two different
serum-free media.9 Moreover, we demonstrated that 1
increases the cytoplasmic free Ca2þ in rat cortical neu-
rons and neuroblastoma SH-SY5Y cells.10 These results
compelled us to supply enough amount of 1 not only to
clarify structure–activity relationships and the details of
its neurotrophic effects, but also to examine its neuro-
trophic action in some in vivo animal models.
Since neurodegenerative diseases such as AlzheimerÕs
disease have increasingly caused social problems as the
swell of aged populations, it is the strong demand to
develop some drugs for treatment of them. Several
endogenous neurotrophic factors, such as the family of
neurotrophins (NGF, BDNF, NT3, and NT4/5), are
known to play key roles in neuronal survival, process
outgrowth, and synaptic connectivity during develop-
ment and nervous system plasticity in adults. Thus, they
are likely to be useful agents for ameliorating neuro-
degeneration.1;2 However, the peptidyl properties of
neurotrophic factors have brought about some serious
issues to overcome, for example, unable to cross the
blood–brain barrier and to be easily metabolized, in
clinical uses.3;4 Small molecules, which can mimic
functions of neurotrophic factors, might be promising
alternatives for treatment of neurodegenerative dis-
eases.5;6 We have, therefore, been searching for neuro-
trophic small molecules in natural products by using the
primary cultures of rat cortical neurons, resulting in the
Herein, we report a highly efficient synthetic route11 of 1
based on the Pd-catalyzed Suzuki–Miyaura protocol
and structure–activity relationships between analogues
2a–c and 3a–c derived from 1 and their neurite out-
growth-promoting activity (Fig. 1).
2. Synthesis of honokiol (1) and its analogues
2a–c and 3a–c
Keywords: Neurotrophic activity; Honokiol; Neolignan; Palladium;
Rat cortical neurons.
As depicted in Scheme 1, a convergent synthetic route
was envisioned to effectively produce some versatile
analogues of 1 in future. Thus, two allyl groups would
* Corresponding author. Tel.: +81-886-22-9611; fax: +81-886-55-3051;
e-mail: fukuyama@ph.bunri-u.ac.jp
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.02.067

2622
T. Esumi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2621–2625
OH
4'
OR1
OH
3'
R1
1'
1
HO
R2O
HO
2
R2
5
honokiol (1)
2a; R1 = H, R2 = Me
2b; R1 = Me, R2 = H
2c; R1 = R2 = Me
3a; R1 = Et, R2 = CH=CH2
3b; R1 = CH=CH2, R2 = Et
3c; R1 = R2 = Et
Figure 1. Structure of honokiol (1) and its analogues 2a–c and 3a–c.
OMOM
OH
MeO2C
honokiol (1)
HO2C
B
O
O
Br
5-bromosalichilic acid (7)
OMOM
5
Pd(0)
MeO2C
MOMO
Br
HO
MOMO
OTBS
CO2H
OTBS
4
4-hydroxybenzoic acid (8)
6
Scheme 1. Retrosynthetic analysis of honokiol (1).
be elaborated from an ester and a hydroxymethyl
groups of biaryl 4 at later stage of synthesis, respec-
tively, so as to introduce some versatile functional
groups for enhancing activity as well as preparing
molecular probes needed to analyze the mechanism of
action. The Pd-catalyzed Suzuki–Miyaura reaction12
would be applied to construct biaryl 4 from borate 5 and
bromide 6. Both the borate 5 and the bromide 6 would
be easily obtained from 5-bromosalicylic acid (7) and 4-
hydroxybenzoic acid (8), respectively.
Br
MOMO
a-c)
HO
d,e)
6
CO2Me
CO2H
8
9
Scheme 2. Preparation of one unit 6. Reagents and conditions: (a)
concd H2SO4, MeOH, reflux, overnight, 99%; (b) HBF4, NBS, MeCN,
ꢀ20 °C to rt, 80%; (c) MOMCl, i-Pr2NEt, DMF, rt, 3 h, 98%; (d)
DIBAL-H, CH2Cl2, ꢀ78 °C, 2 h, 100%; (e) TBSCl, imid., DMF, rt,
100%.
The synthesis of 1 commenced with the Fischer esteri-
fication of 4-hydroxybenzoic acid (8) followed by reg-
ioselective bromination13 using N-bromosuccinimide
and HBF4 in MeCN and protection of phenolic
hydroxyl group with methoxymethyl chloride to give the
bromide 9 in good yield. Reduction of 9 with DIBAL-H
in CH2Cl2 at ꢀ78 °C followed by protection of a
hydroxyl group with tert-butyldimethylsilyl chloride
gave aryl bromide 6 quantitatively over two steps
(Scheme 2).
dium catalysts, none of 5 was obtained. Under the
particular conditions, that is, bis(pinacolato)diborane
(1.0 equiv)14 with PdCl2(dppf) (10 mol %), dppf
(10 mol %), and AcOK (3.0 equiv) in 1,4-dioxane at
80 °C, the coupling reaction smoothly proceeded to
provide the desired arylboronate 5 in 86% yield. The
critical coupling reaction between aryl bromide 6 and
arylboronate 5 based on the Suzuki–Miyaura protocol12
was accomplished by using 10 mol % of PdCl2(dppf)
with dppf (10 mol %) and K3PO4 (3.0 equiv) in 1,4-
dioxane at 108 °C, giving rise to the desired coupling
adduct 4 in 87% yield. The methyl ester moiety of 4 was
reduced to a hydroxymethyl group with LiAlH4, and
then tert-butyldimethylsilyl group was removed by 47%
HF/pyridine/MeCN (1:3:5) in a high yield. Both the
hydroxyl groups of diol 11 were converted to bromides
The other unit 5 was prepared as followings: Fischer
esterification of 5-bromosalicylic acid (7) followed by
protection of a phenolic hydroxyl group with meth-
oxymethyl chloride gave methoxymethyl ether 10 in
quantitative yield and then 10 was subjected to Pd-cat-
alyzed transformation into pinacol borate 5. However,
in the case of using pinacolborane with several palla-

T. Esumi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2621–2625
2623
by treatment of N-bromosuccinimide and triphenyl-
phosphine, followed by the reaction with vinylcuprate to
give diallyl compound 12 in a moderate yield. Finally,
treatment of 12 with 2 M HCl afforded honokiol (1), all
spectra data15 (1H NMR, 13C NMR, IR, HR-EIMS) of
which were superimposed on those of natural product.
Thus, we have established the efficient synthesis of
honokiol (1) in a 21% overall yield (Scheme 3).
stration of morphology, the neurons were incubated in
primary antibody anti-MAP2 (1:1000) overnight in 4 °C
followed by incubation with horse peroxidase-conju-
gated second antibody [rabbit anti-mouse IGg (1:2)] for
1 h, then peroxidase was developed with 200 lL sub-
strate Simple Stain DAB solution. Images of neuron in
random fields were captured with inverse microscope
equipped with Macintosh computer and imaging soft-
ware LuminaVision 1.0 (Mitani Corp., Fukui, Japan).
The length of the primary (longest) neurite on each
Next, six analogues 2a–c and 3a–c were prepared as
follows: Treatment of 1 with TMS–diazomethane in
MeOH yielded a mixture of methylated products, which
were separated by a flash silica gel chromatography to
give methylated analogues 2a–c in 13%, 13%, and 47%
yield, respectively. On the other hand, 1 was subjected to
hydrogenation with WilkinsonÕs catalyst in EtOH to
afford dihydrogenated analogues 3a–c in 12%, 7%, and
18% yield, respectively.
neuron was manually measured with software
MacScop
2.6 (Mitani Corp., Fukui, Japan). The average length of
at least 100 neurons in each group was expressed as
mean ꢁ SEM. Statistical analysis was performed with
O
rigin 7.0 (OriginLab, USA).
The morphological effects of honokiol (1) and its ana-
logues 2a–c and 3a–c on the cultured rat cortical neu-
rons were evaluated by the anti-MAP2 staining method.
Honokiol (1) (Fig. 2b), 2-O-methylhonokiol (2a) (Fig.
2c), and 80-dihydrohonokiol (3a) (Fig. 2d) had striking
effects on the morphology of cortical neurons in com-
parison with the cortical culture (Fig. 2a) as shown in
Figure 2. The other analogues 2b, 2c, 3b, and 3c showed
no enhancement of the neurite extension. As shown in
Figure 3, the quantitative analysis of the longest neurite
length affected by each compound at concentrations of
0.1 and 1 lM indicates that 2a (neurite length:
132.7 ꢁ 5.7 and 122.6 ꢁ 6.2 lM) and 3a (127.5 ꢁ 5.6 and
150.5 ꢁ 8.1 lM) also have as high potential of enhancing
the neurite extension in the cultured rat cortical neurons
as 1 (138.9 ꢁ 6.3 and 142.1 ꢁ 5.3 lM), whereas other
analogues 2b (113.1 ꢁ 4.5 and 114.4 ꢁ 6.5 lM), 2c
3. Structure–activity relationship
The neuronal cultures and neurite outgrowth-promoting
assay were performed with slight modifications in
serum-free medium and cell density according to the
way as previously described.9 Briefly, the neurons were
isolated from cerebral cortex of pregnant day 18 SD
rat fetal (SLC, Japan) and cultured in DulbeccoÕs
modified eagleÕs medium (DMEM) (Gibco BRL, USA)
supplemented with 10% of fetal bovine serum (FBS)
and 50 IU mLꢀ1 penicillin-50 lg mLꢀ1 streptomycin at a
density of 5000 cells cmꢀ2 on poly-
L-lysine-coated 24-
well plates for 24 h. Then the medium was changed to
the serum-free Neurobasale medium (NB) plus 2% B27
supplement containing the tested sample. The treatment
continued for 144 h. Then the neurons were fixed with
4% paraformaldehyde (in PBS) for 20 min, and per-
meabilized with 0.1% Triton X-100 in PBS for 20 min
after endogenous peroxidase activity was blocked by
freshly prepared 0.3% H2O2 for 20 min. For demon-
(111.6 ꢁ 5.8,
111.0 ꢁ 5.4 lM),
3b
(102.6 ꢁ 5.0,
120.1 ꢁ 6.1 lM), and 3c (100.9 ꢁ 5.3, 102.3 ꢁ 6.1 lM)
diminish neurotrophic efficiency. Thus, these results
suggest that 40-hydroxy group and 5-allyl group are
responsible for honokiol-mediated neurite outgrowth-
promoting activity, but 30-allyl group is not essential for
that.
OMOM
OH
MeO2C
HO2C
d)
c)
e,f)
a,b)
4
5
Br
Br
10
7
OMOM
OMOM
HO
i)
g,h)
honokiol (1)
MOMO
MOMO
OH
11
12
Scheme 3. Total synthesis of honokiol (1). Reagents and conditions: (a) concd H2SO4, MeOH, reflux, overnight, 100%; (b) MOMCl, i-Pr2NEt,
DMF, rt, 3 h, 100%; (c) bis(pinacolato)diborane, 10 mol % PdCl2(dppf), dppf, AcOK, 1,4-dioxane, 80 °C, 86%; (d) 6, 10 mol % PdCl2(dppf), dppf,
K3PO4, dioxane, reflux, 87%; (e) LiAlH4, THF, 0 °C, 99%, (f) 47% HF/pyridine/MeCN (1/3/5), rt, 1 h, 98%; (g) NBS, PPh3, CH2Cl2, 80%; (h)
H2C@C(H)MgBr, CuI, THF, ꢀ26 °C, 52%; (i) 2 M HCl, MeOH (5/1), rt, 44 h, 89%.

2624
T. Esumi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2621–2625
Figure 2. Morphology of cultured rat cortical neurons demonstrated with anti-MAP2 immunochemical staining. After 1-day pre-culture, neurons
were treated for 3 days, and then stained by anti-MAP2 as described in the text. (a) Neurons in the presence of 0.5% ethanol as vehicle control; (b)
neurons in the presence of 0.1 lM 1; (c) neurons in the presence of 1 lM 2a; (d) neurons in the presence of 0.1 lM 3a.
for enhancing and clarifying neurotrophic potential and
mechanism caused by honokiol currently is undergoing
based on the present data.
Acknowledgements
This work is supported by the Open Research Center
Fund from the Promotion and Mutual Aid Corporation
of Private Schools of Japan.
Figure 3. Quantitative analysis of anti-MAP2 immunochemically
References and notes
stained processes affected by honokiol (1) and its analogues 2a–c and
3a–c. In each group, the average length of the primary processes were
determined from 100 neurons selected in random fields. ꢂꢂ, P < 0:01;
ꢂꢂꢂ, P < 0:001 compared with control. Data presented here are derived
from one of two repeated experiments with similar results.
1. Sofroniew, M. V.; Howe, C. L.; Mobley, W. C. Ann. Rev.
Neurosci. 2001, 24, 1217.
2. Hefti, F. J. Neurobiol. 1994, 25, 1417.
3. Pardridge, W. M. Curr. Opin. Invest. Drugs 2002, 3, 1753.
4. Thoenen, H.; Sendtner, M. Nat. Neurosci. 2002, 5, 1046.
5. Massa, S. M.; Xie, Y.; Longo, F. M. J. Mol. Neurosci.
2002, 19, 107.
6. Xie, Y.; Longo, F. M. Prog. Brain Res. 2000, 128, 333.
7. Fukuyama, Y.; Otoshi, Y.; Miyoshi, K.; Nakamura, K.;
Kodama, M.; Nagasawa, M.; Hasegawa, T.; Okazaki, H.;
Sugawara, M. Tetrahedron 1992, 48, 377.
8. Fujita, M.; Itokawa, H.; Sashida, Y. Chem. Pharm. Bull.
1972, 20, 212.
9. Fukuyama, Y.; Nakade, K.; Minoshima, Y.; Yokoyama,
R.; Zhai, H.; Mitsumoto, Y. Bioorg. Med. Chem. Lett.
2002, 12, 1163.
In conclusion, we have achieved the efficient synthesis of
honokiol (1) by utilizing Suzuki–Miyaura coupling
reaction in 21% yield over 14 steps and then have pre-
pared the six analogues of 1. According to the mor-
phological analyses of these analogues in the primary
cultured rata cortical neurons together with the previous
results,16 the biaryl structure bearing a hydroxyl and a
allyl groups at the 40- and 5-positions, respectively, is
shown to be at least essential for honokiol-mediated
neurite outgrowth-promoting activity. Further research

T. Esumi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2621–2625
2625
10. Zhai, H.; Nakade, K.; Mitsumoto, Y.; Fukuyama, Y. Eur.
J. Pharmacol. 2003, 474, 199.
11. The first synthesis of honokiol (1) was reported in the
following paper. Takeya, T.; Okubo, T.; Tobinaga, S.
Chem. Pharm. Bull. 1986, 34, 2066.
12. (a) Suzuki, A. Pure Appl. Chem. 1985, 57, 1749; (b)
Miyaura, N.; Suzuki, N. Chem. Rev. 1995, 95, 2457; (c)
Miyaura, N.; Ishiyama, T.; Sasaki, H.; Ishikawa, M.;
Sato, M.; Suzuki, A. J. Am. Chem. Soc. 1989, 111, 314; (d)
Ishihara, T.; Murata, M.; Miyaura, N. J. Org. Chem.
1995, 60, 7508.
13. Oberhauser, T. J. Org. Chem. 1997, 62, 4504.
14. Takagi, J.; Takahashi, K.; Ishiyama, T.; Miyaura, N.
J. Am. Chem. Soc. 2002, 124, 8001.
15. Spectral data of 1: 1H NMR (300 MHz, CDCl3): d 7.22
(1H, dd, J ¼ 7:8, 2.1 Hz), 7.21 (1H, s), 7.20 (1H, d,
J ¼ 2:1 Hz), 7.05 (1H, dd, J ¼ 7:8, 2.1 Hz), 6.92 (1H, d,
J ¼ 7:8 Hz), 6.90 (1H, d, J ¼ 7:8 Hz), 6.04 (1H, ddt,
J ¼ 16:8, 10.2, 6.6 Hz), 5.97 (1H, ddt, J ¼ 16:8, 9.9,
6.9 Hz), 5.2 (1H, dq, J ¼ 9:9, 1.5 Hz), 5.14 (1H, br s),
5.15 (1H, dq, J ¼ 16:8, 1.5 Hz), 5.10 (1H, br s), 5.08 (1H,
dq, J ¼ 16:8, 1.5 Hz), 5.05 (1H, dq, J ¼ 10:2, 1.2 Hz), 3.46
(1H, d, J ¼ 6:6 Hz), 3.35 (1H, d, J ¼ 6:3 Hz); 13C
NMR (75 MHz, CDCl3): d 153.9, 150.7, 137.8, 135.9,
132.2, 131.1, 130.2, 129.6, 128.8, 128.5, 127.7, 126.3, 116.9,
116.6, 115.5, 39.4, 35.2; FT-IR (neat) 3434, 1637,
1607 cmꢀ1
;
266.1307; found 266.1294.
HR-EIMS calcd for C18H18O2 (Mþ)
16. 2-Allyphenol and 4-allyphenol are already known to be
toxic against the rat cortical neurons at 1 lM, indicating
that the biaryl structure in honokiol makes a contribution
to express its neurotrophic activity.