Structure-activity relationship studies of salubrinal lead to its active biotinylated derivative

Article abstract of DOI:10.1016/j.bmcl.2005.05.120

The synthesis and structure-activity relationships (SAR) of salubrinal, a small molecule that protects cells from apoptosis induced by endoplasmic reticulum (ER) stress, are described. It is revealed that the trichloromethyl group greatly contributes to the activity. Based on the SAR results, salubrinal was converted into a biotinylated derivative which retains activity and can be used as a biological tool for target identification.

Full text of DOI:10.1016/j.bmcl.2005.05.120

Bioorganic & Medicinal Chemistry Letters 15 (2005) 3849–3852
Structure–activity relationship studies of salubrinal lead
to its active biotinylated derivative
Kai Long,a Michael Boyce,b He Lin,a Junying Yuanb and Dawei Maa,*
aState Key Laboratory of Bioorganic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry,
Chinese Academy of Sciences, 354 Fenglin Road, Shanghai 200032, China
bDepartment of Cell Biology, Harvard Medical School, Boston, MA 02115, USA
Received 3 May 2005; revised 26 May 2005; accepted 30 May 2005
Available online 5 July 2005
Abstract—The synthesis and structure–activity relationships (SAR) of salubrinal, a small molecule that protects cells from apoptosis
induced by endoplasmic reticulum (ER) stress, are described. It is revealed that the trichloromethyl group greatly contributes to the
activity. Based on the SAR results, salubrinal was converted into a biotinylated derivative which retains activity and can be used as a
biological tool for target identification.
Ó 2005 Elsevier Ltd. All rights reserved.
The endoplasmic reticulum (ER) stress response, a sig-
naling pathway that connects ER with the nucleus, is
essential for cellular homeostasis of all eukaryotic cells.1
However, excessive and uncorrected ER stress can in-
duce cell apoptosis, which has been implicated in many
important pathologies, including diabetes, AlzheimerÕs
disease, and viral infection.2 Therefore, a better under-
standing of apoptotic pathway activated by ER stress
will provide insight into the development of therapeutics
for important human diseases. In a screen for com-
pounds that protect mammalian cells from apoptosis
induced by ER stress, we identified a small molecule,
termed salubrinal (1).3 Salubrinal protects cells against
ER stress-induced apoptosis by selectively inhibiting
the translation initiation factor eIF2a dephosphoryla-
tion. It blocks eIF2a dephosphorylation via inhibiting
the protein complex GADD34/PP1 consisting of the
general cellular serine/threonine phosphatase PP1 and
the non-enzymatic cofactor GADD34, which targets
eIF2a for dephosphorylation by PP1. As a specific
inhibitor of eIF2a dephosphorylation, salubrinal is also
useful in a disease model. It inhibits eIF2a dephospho-
rylation mediated by a herpes simplex virus encoded
protein and blocks viral replication in vitro and in vivo.3
These results suggest that the inhibition of eIF2a
dephosphorylation may be a potent therapeutic strategy
in diseases involving ER stress and viral infection, and
that such small molecule as salubrinal might find thera-
peutic use in these diseases.
Whether salubrinal inhibits the GADD34/PP1 complex
via direct binding or an indirect signaling event is yet un-
clear. A more detailed understanding of the molecular
mechanism of salubrinal action is currently under inves-
tigation. We describe herein our efforts to investigate the
structure–activity relationships (SAR) of salubrinal and
to convert it into a biotinylated derivative as a biological
tool for target identification.
The simplest modifications for salubrinal involves vary-
ing the substituents at its terminuses. The synthesis of
these analogs is outlined in Scheme 1. Reaction of
amides with anhydrous chloral gave chloralamides 2,
which were subsequently transformed into the corre-
sponding amines 3 via two successive substitution reac-
tions.4 Addition of these amines to separately prepared
isothiocyanates resulted in N,N0-disubstituted asym-
metric thioureas 4.5 Compound 4l was directly obtained
by reduction of 4k with FeCl2 in 79% yield. The struc-
tures of 4a–4q are shown in Table 1.
Keywords: Apoptosis; Inhibitor of eIF2a dephosphorylation;
Synthesis.
*
Corresponding author. Tel.: +86 21 64163300; fax: +86 21
64166128; e-mail: madw@mail.sioc.ac.cn
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.05.120

3850
K. Long et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3849–3852
Scheme 1. Reagents and conditions: (i) CCl3CHO, toluene, 90 °C, 80–97%; (ii) (1) PCl5, Et2O, (2) NH3, Et2O, 0 °C, 60–80%; (iii) R2AN@C@S, THF,
60 °C, 60–95%.
Table 1. Structures of compounds 4a–4q and EC50 values of salubrinal
Table 1 (continued)
(1) and compounds 4–12
Compound R1
R2
EC50 (lM)
Compound R1
R2
EC50 (lM)
4q
41
1
15
5
TX
TX
TX
51
4a
4b
4c
4d
4e
4f
NA
40
6
7
8
9
75
10
11
12
NA
NA
NA
>100
24
NA: not active up to 100 lM. TX: toxic.
–n-Bu
In order to explore whether the trichloromethyl unit is
necessary for the activity, analogs 5–10 were synthe-
sized. As outlined in Scheme 2, compounds 5 and 6 were
readily prepared via condensation of cinnamic acid with
hydrazine and ethylenediamine respectively and subse-
quent addition to 2-isothiocyanatopyridine. Following
the same procedure as for the synthesis of 4, compounds
7–10 were obtained starting from the reaction of cinna-
mide with a range of aldehydes, as indicated in Scheme
3.6 Compound 10 was further transformed into 11 and
12 via reduction and subsequent cyclization as described
in Scheme 4.7
NA
NA
33
4g
4h
4i
29
Compounds were assayed for their ability to protect the
rat pheochromocytoma cell line PC12 from apoptosis
induced by ER stress. In this assay, PC12 cells were
treated with tunicamycin (Tm) and different doses of
compounds for 37 h and then cell viability was assessed
by cellular ATP content. Tm is an inhibitor of protein
glycosylation in the ER and thus results in the accumu-
lation of misfolded proteins, which will induce ER stress
and subsequent apoptosis.10 A treatment with 100 lM
of zVAD.fmk, a broad-spectrum caspase inhibitor,
served as a positive control. Salubrinal inhibited ER
stress-mediated apoptosis in a dose-dependent manner
with a median effective concen-tration (EC50, as defined
by the concentration at which the compound produces
half-maximal rescue from cell death versus the Tm-only
control) at about 15 lM and rescued about 35% of PC12
cells at 100 lM.3 The cytoprotection activity of other
analogues was assessed and the EC50 values are listed
in Table 1.
40
4j
38
4k
4l
28
56
4m
4n
4o
4p
50
NA
72
As shown in Table 1, Compounds 4a–4d and salubrinal
bear the same R1 group but different R2 groups. Struc-
turally compound 4a is almost the same as salubrinal
30

K. Long et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3849–3852
3851
4c was much lower than salubrinal, whereas compound
4b and 4d had a close EC50 to that of salubrinal. Com-
pound 4b even demonstrated improved efficacy (the
maximum protection from cell death at the optimal con-
centration) over salubrinal, therefore we retained the
pyridine ring instead of quinoline ring in the following
structural modifications.
Next, the cinnamide system was modified and analogues
4e–4q were synthesized. Most of them displayed cytopro-
tection activity except compounds 4e, 4f, and 4n. Activity
results of compounds 4b, 4e, and 4h indicated that a cyclic
moiety as R1 might be favored for the activity. Compar-
ison of compounds 4b,4f, and4g suggested that a proper
distance between the cyclic moiety and the trichlorometh-
yl group might play some role for the activity. Activity re-
sults of compounds 4b, 4g, and 4m demonstrated that the
(E)-double bond in the leading compound might be
unimportant because it could be changed into a (Z)-dou-
ble bond or a single bond without much loss of activity.
Scheme 2. Reagents and conditions: (i) for 5: (1) oxalyl chloride,
DMF, CH2Cl2, (2) hydrazine, Et2O, 0 °C, 35%; for 6: (1) HOSu, DCC,
THF, (2) ethylenediamine, 60%; (ii) 2-isothiocyanatopyridine, EtOH,
60 °C, 44–78%.
except for the substitution position at the quinoline ring.
However, 4a displayed no activity. Replacement of the
quinoline ring by pyridyl, benzoyl, or alkyl groups
resulted in derivatives 4b–4d. The potency of compound
Scheme 3. Reagents and conditions: (i) for 7–9: cinnamide, toluene, 90 °C, 80–90%; for 10: cinnamide, THF, rt, 65%; (ii) (1) PCl5, Et2O, (2) NH3,
Et2O, 0 °C, 40–80%; (iii) 2-isothiocyanatopyridine, THF, 60 °C, 70–90%.
Scheme 4. Reagents and conditions: (i) NaBH4, LiCl, THF/EtOH, 85%; (ii) PPh3, DEAD, THF, 45%.
Scheme 5. Reagents and conditions: (i) (1) SOCl2, Et2O, (2) KSCN, acetone, 73%; (ii) for 14: HOCH2CH2NH2, CHCl3, 70%; for 15:
BocNH(CH2)4NH2, CHCl3, 80%; (iii) for 16: biotin, EDC, DMAP, DMF, 60%; for 17: (1) TFA, CH2Cl2, (2) biotin, EDC, HOBt, DIPEA, DMF,
60%.

3852
K. Long et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3849–3852
Substituents on the phenyl ring or replacement of it by
other heterocycles seemed to have little influence on the
activity except that a long chain attachment on the phen-
yl ring was unfavored (compound 4n).
pound 2a was converted into the corresponding
isothiocyanate 13 via two successive substitution reac-
tions.8 Addition of 2-hydroxy-ethylamine and mono-
protected 1,4-diaminobutane to 13 afforded thiourea
14 and 15, respectively. Condensation of 14 and depro-
tected 15 with biotin gave the corresponding biotinyla-
ted derivatives 16 and 17 (Scheme 5).9
Next, the role of the trichloromethyl group was investi-
gated. Removing the trichloromethyl unit resulted in
analogues 5 and 6, both of which were toxic to cells.
Replacement of the trichloromethyl group by other mul-
tihalomethyl groups resulted in compounds 7–9. Inter-
We tested compounds 16 and 17 for the ability to pro-
tect PC12 cells from Tm-induced apoptosis. Compound
16 was almost inactive up to 100 lM. However, com-
pound 17 exhibited considerable cytoprotection activity
at the concentration range from 50 to 100 lM, and the
efficacy was quite close to that of salubrinal, as shown in
Figure 1. These results encourage us to use compound
17 as an affinity reagent to identify the binding target
of salubrinal. The results will be reported elsewhere in
due course.
estingly,
a
trifluoromethyl substituent completely
eliminated the cytoprotection activity (compound 7),
whereas tribromomethyl (compound 8) and dichloro-
methyl (compound 9) substituents were tolerated. Other
modifications on the trichloromethyl group afforded
compounds 10–12, which were all inactive. By consider-
ing all these results together, it is clear that the trichlo-
romethyl group in the leading compound is crucial for
the activity.
Acknowledgments
As mentioned above, salubrinal protects cells against
ER stress-induced apoptosis via inhibiting eIF2a
dephosphorylation (namely inducing eIF2a phospho-
rylation). We tested the analogues that showed rescue
activity and found they also induced eIF2a phosphory-
lation,3 which suggested that these analogs protected
cells by the same mechanism.
The authors are grateful to the Chinese Academy of Sci-
ences (grant KGCX2-SW-209 to D.M.), National Nat-
ure Science Foundation of China (grant 20228205 to
D.M.) and National Institutes of Health, USA (grant
R37-AG012859 to J.Y.) for their financial support.
After getting a clear SAR pattern, we turned to intro-
duce a biotin moiety as an affinity tag into salubrinal,
in the hope of using such a tagged reagent to purify
the molecular target of salubrinal. According to the dif-
ferent activities of compounds 4d and 4n, we chose to
link biotin from the R2 side and used 2-hydroxyethyl-
amine or 1,4-diaminobutane as a linker. Thus, com-
References and notes
1. For a review, see: Chapman, R.; Sidrauski, C.; Walter, P.
Annu. Rev. Cell Dev. Biol. 1998, 14, 459.
2. For a review, see: Aridor, M.; Balch, W. E. Nature
Medicine 1999, 5, 745.
3. Boyce, M.; Bryant, K.; Jousse, C.; Long, K.; Harding, H.
P.; Scheuner, D.; Kaufman, R. J.; Ma, D.; Coen, D. M.;
Ron, D.; Yuan, J. Science 2005, 307, 935.
4. (a) Willard, J. R.; Hamilton, C. S. J. Am. Chem. Soc. 1953,
´
75, 2370; (b) Guirado, A.; Andreu, R.; Cerezo, A.; Galvez,
J. Tetrahedron 2001, 57, 4925.
5. For the preparation of isothiocyanates, see: (a) Yamada,
M.; Yura, T.; Morimoto, M.; Harada, T.; Yamada, K.;
Honma, Y.; Kinoshita, M.; Sugiura, M. J. Med. Chem.
1996, 39, 596–604; (b) Gulland, J. M.; Macey, P. E. J.
Chem. Soc. 1949, 1257–1259; (c) Frank, R. L.; Smith, P. V.
Org. Synth. Coll. Vol. 1955, 3, 735.
6. (a) Braid, M.; Iserson, H.; Lawlor, F. E. J. Am. Chem.
Soc. 1954, 76, 4027; (b) Ingrassia, L.; Mulliez, M.
Synthesis 1999, 10, 1731.
7. Kim, T. H.; Cha, M. H. Tetrahedron Lett 1999, 40, 3125.
8. Drach, B. S.; Sinitsa, A. D.; Kirsanov, A. V. Zh. Obshch.
Khim. 1970, 40, 934.
9. All the synthesized compounds were well charac-terized by
1H NMR, MS, IR and elemental analysis. Analytical data
for compound 17 are as follows: H NMR (CD3OD, 300
1
MHz) d : 1.28 (q, J = 7.2 Hz, 2H), 1.37–1.58 (m, 8H), 2.03
(t, J = 7.2 Hz, 2H), 2.54 (d, J = 13.2 Hz, 1H), 2.78 (dd,
J = 12.6, 5.1 Hz, 1H), 3.02–3.08 (m, 3H), 3.42 (m, 2H),
4.15 (dd, J = 7.8, 3.9 Hz, 1H), 4.35 (dd, J = 7.8, 4.5 Hz,
1H), 6.56 (d, J = 15.3 Hz, 1H), 7.24–7.52 (m, 6H), 7.88 (m,
1H); HRMS (ESI): Calcd for C26H35Cl3N6O3S2Na
(M+Na+): 671.1170. Found 671.1180.
Figure 1. Dose-dependent protection of PC12 cells treated with Tm by
salubrinal, compounds 16 and 17, assessed by cellular ATP content.
An amount of 100 lM of zVAD.fmk serves as a positive control. Error
bars represent standard deviation.
10. For a review, see: Kaufman, R. J. Genes Dev 1999, 13,
1211.