Structure-activity relationship (SAR) studies of 3-(2-amino-ethyl)-5-(4-ethoxy-benzylidene)-thiazolidine-2,4-dione: Development of potential substrate-specific ERK1/2 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.09.057

A series of analogs of 3-(2-amino-ethyl)-5-(4-ethoxy-benzylidene)-thiazolidine-2,4-dione, a putative substrate-specific ERK1/2 inhibitor, were synthesized and biologically characterized in human leukemia U937 cells to define its pharmacophore. It was disc

Full text of DOI:10.1016/j.bmcl.2009.09.057

Bioorganic & Medicinal Chemistry Letters 19 (2009) 6042–6046
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure–activity relationship (SAR) studies of 3-(2-amino-ethyl)-5-(4-ethoxy-
benzylidene)-thiazolidine-2,4-dione: Development of potential substrate-
specific ERK1/2 inhibitors
Qianbin Li a, Adnan Al-Ayoubi b, Tailiang Guo c, Hui Zheng b, Aurijit Sarkar a, Tri Nguyen d, Scott T. Eblen b,
Steven Grant d, Glen E. Kellogg a, Shijun Zhang a,
*
a Department of Medicinal Chemistry, School of Pharmacy, Virginia Commonwealth University, Richmond, VA 23298-0540, USA
b Department of Cell and Molecular Pharmacology, Medical University of South Carolina, Charleston, South Carolina, USA
c Department of Pharmacology and Toxicology, Virginia Commonwealth University, Richmond, VA, USA
d Departments of Internal Medicine and Biochemistry, Virginia Commonwealth University, Richmond, VA, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 31 July 2009
Revised 8 September 2009
Accepted 14 September 2009
Available online 18 September 2009
A series of analogs of 3-(2-amino-ethyl)-5-(4-ethoxy-benzylidene)-thiazolidine-2,4-dione, a putative
substrate-specific ERK1/2 inhibitor, were synthesized and biologically characterized in human leukemia
U937 cells to define its pharmacophore. It was discovered that shift of ethoxy substitution from the 4- to
the 2-position on the phenyl ring significantly improved functional activities of inhibiting cell prolifera-
tion and inducing apoptosis. This may provide access to a new lead for developing ERK1/2 substrate-spe-
cific inhibitors.
Keywords:
ERK1/2
Published by Elsevier Ltd.
Substrate-specific inhibitor
Synthesis
SAR
Thiazolidine-2,4-dione
Rational design
Molecular modeling
The Raf/MEK/Extracellular Signal-Regulated Kinase (ERK1/2)
pathway is one of the evolutionarily conserved mitogen-activated
protein kinase (MAPK) pathways that play critical roles in driving
proliferation and preventing apoptosis. Signal transduction
through this pathway is regulated by a three-tiered sequential
phosphorylation cascade composed of MAPK (ERK1/2), MAPK ki-
nase (MEK1/2) and MAPK Kinase kinase (A-raf, B-raf, or Raf-1).1
Upon activation, ERK1/2 phosphorylates multiple substrates with-
in various cellular compartments, which in turn results in gene
expression changes and alterations in cell proliferation, differenti-
ation and survival.2 Substantial evidence has validated that dereg-
ulation of this pathway is involved in the oncogenesis of various
human cancers.3,4 For example, B-Raf mutations are found in about
66% of malignant melanomas and at lower frequency in many
other human malignancies including ovarian cancer.3–8 Recently,
it has also been reported that a high frequency of acute myeloid
leukemia (AML) and acute lymphocytic leukemia (ALL) display
constitutive activation of the Raf/MEK/ERK1/2 pathway.8–10 There-
fore, this pathway represents an attractive target for development
of potential anti-cancer drugs. Indeed, numerous structurally di-
verse small molecule inhibitors, such as Raf and MEK inhibitors,
have been developed-with Sorafenib being the most successful.
While small molecule kinase inhibitors have emerged as power-
ful tools and promising targeted cancer therapeutic approaches, a
continuing challenge faced with the majority of current inhibitors
is the lack of specificity to their targets.11 On the other hand, pro-
tein–protein interactions have been highlighted as powerful means
in coordinating signaling processes.12 Thus, novel substrate-spe-
cific kinase inhibitors that disrupt protein–protein interactions
would overcome the lack of specificity available when targeting
the highly conserved ATP binding sites. ERK1/2 represents a near
optimal target to develop such inhibitors given the following facts:
(1) The recent identification of the docking groove composed of a
common docking (CD) domain and an ED domain on ERK2
(Fig. 1) and the validation of their critical roles in controlling dock-
ing interactions with substrate proteins;13–16 (2) the important
position of ERK1/2 in the Raf/MEK/ERK1/2 pathway that controls
the distribution of signals from upstream; and (3) the recent dis-
closure of the crystal structures of ERK2.17,18 Recently, Shapiro’s
group has identified several such inhibitors by computer-aided
virtual screening of a small library of compounds (Fig. 2).19,20
* Corresponding author. Tel.: +1 804 6288266; fax: +1 804 8287625.
E-mail address: szhang2@vcu.edu (S. Zhang).
0960-894X/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2009.09.057

Q. Li et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6042–6046
6043
O
O
i.RCHO, piperidine
EtOH, reflux
NHBoc
Br
NH
N
NHBoc
S
S
Acetone, TBAI, reflux
ii. Dioxane-HCl
O
4
O
5
R=
O
Ph (1a)
4-MeO-Ph (1b) 2-EtO-Ph (1c)
4-F-Ph (1f)
4-NO2-Ph (1h) 3-NO2-Ph (1i)
R
3-EtO-Ph (1d) 4-Me-Ph (1e)
N
NH2
S
3-F-Ph (1g)
4-Cl-Ph-Ph (1j) Naphthalene (1k)
2,3-dihydrobenzo[b][1,4]dioxine-6-(1l)
Cinnamyl (3a)
O
1a-1l, 3a
O
O
O
i. HCHO, AcOH
Ac2O
N
N
N
S
76
NH
S
EtO
ii.NaBH(OAc)3
EtO
O
O
2b
2a
Scheme 1. Synthesis of 1a–1l, 2a, 2b and 3a.
Figure 1. CD and ED domains on ERK2.
we introduced two additional hydroxyl groups on this extended
analog to evaluate whether introduction of hydrogen bond interac-
tions at this domain can improve selectivity and potency (3b, 3c).
The synthetic protocol for 1a–1l is shown in Scheme 1. Briefly, N-
alkylation of thiazolidine-2,4-dione (4) with Boc-protected 2-bro-
moethylamine provided 5, which on Knoevenagel condensation
with various aldehydes followed by Boc deprotection afforded
1a–1l and 3a in good yields.24 Reductive amination of 76 with
formaldehyde gave the dimethylated analog 2a. Acetylation of 76
with acetic anhydride yielded 2b. The synthesis of 3b and 3c is
O2N
N
O
S
NH2
N
N
H2N
NH
N
N
O
O
O
O
O
76
81
HN
89
NH
O
NO2
Br
O
N
N
O
S
92
N
H
N
O
O
Cl
95
H
Figure 2. Putative substrate-specific ERK1/2 inhibitors.
shown in Scheme 2. Dihydroxylation of 6 with AD-mix-
a and
AD-mix-b gave 7 and 8, respectively. Protection of these two hy-
droxyl groups followed by diisobutylaluminium hydride (DIBALH)
reduction provided 9 and 10, respectively. Knoevenagel condensa-
tion followed by deprotection afforded 3b and 3c.
After synthesis, all of the compounds were initially evaluated
for their abilities to inhibit phosphorylation of Rsk1 and ERK1/2
by immunoblot analysis in human U937 leukemia cells under the
stimulation of phorbol 12-myristate 13-acetate (PMA) (Supple-
mentary data S3). Then, the active compounds were further tested
for relative selectivity by measuring their effects on phosphoryla-
tion of Rsk1, ERK1/2, Elk-1, MEK and p38 (Fig. 4). The parent com-
Among these structures, compound 76 represents an interesting
chemical scaffold. Compound 76 inhibited the phosphorylation of
Rsk1 and Elk-1, the downstream substrates of ERK1/2, in HeLa cells
with minimal effects on the phosphorylation of ERK1/2. It also
inhibited the growth and survival of several cancer cell lines. In
addition, small molecules containing the thiazolidine-2,4-dione
moiety, such as the anti-diabetic drug troglitazone, have been re-
cently reported to have anti-cancer activities.21–23 Thus, these re-
sults demonstrated the potential of compound 76 as a new
chemotype to develop substrate-specific ERK1/2 inhibitors. Be-
cause compound 76 was discovered as a lead through screening,
little information about the pharmacophore of compound 76 is
known and no further effort to modify 76 to improve its bioactivity
has been reported. In this study, we initiated preliminary struc-
ture–activity relationship (SAR) studies of compound 76 to define
its pharmacophore and to search for more potent and selective
substrate-specific ERK1/2 inhibitors.
As shown in Figure 3, we mainly modified 76 in three domains:
the phenyl ring, the ethylamine tail, and the spacer domain be-
tween the phenyl ring and thiazolidine-2,4-dione. In the first case,
functional groups of varied lipophilicity, size and electronic prop-
erties were introduced in the phenyl ring to evaluate these effects
on biological activities (1a–1l). Next, dimethylated and acetylated
ethylamine tails were designed to evaluate whether a basic pri-
mary amine is needed for activity (2a, 2b). Finally, the aforemen-
tioned spacer was extended by two carbon atoms to gain insight
into the influence of molecular size on activity (3a). Furthermore,
pound 76 reduced the phosphorylation of Rsk1 at 50 lM with
minimal effects on the phosphorylation of ERK1/2, which is consis-
tent with the reported results.19,20 Substitution with F, NO2 at the
4- and 3-positions on the phenyl ring (1f–1i) abolished the inhibi-
tion of Rsk1 phosphorylation indicating that electron withdrawing
substituents are not favored at these two positions. However, 4-Cl
substitution (1j) exhibited similar effects as 76 (Fig. 4). The results
from 1a, 1b, 1d and 1e clearly indicate that steric effects and an
OH O
CH3SO2NH2
AD-mix-¦Á
O
BuOH, H2O
OH
O
7
O
OH
O
O
6
CH3SO2NH2
AD-mix-¦Â
BuOH, H2O
OH
8
O
O
O
O
i. acetone, DMP, PPTS
or
7 or 8
H
ii.Toluene, DIBALH, -78oC
Spacer domain
O
H
O
10
O
9
OH
O
R1
N
NHR2
i. 5, BuLi, THF, -78oC
S
N
NH2
9 or 10
S
O
OH
O
ii.MeOH, Conc. HCl
Ethylamine tail domain
Phneyl ring domain
3b (2S,3S) or 3c (2R,3R)
Figure 3. Structural modifications of compound 76.
Scheme 2. Synthesis of 3b and 3c.

6044
Q. Li et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6042–6046
Figure 4. Effects of test compounds on Rsk1, ERK1/2, Elk-1, MEK1/2 and p38 phosphorylation (A: immunoblot results; B: quantification of Rsk1 phosphorylation by
densitometry) in U937 cells.
oxygen atom at the 4-position are important factors for activity
since none of them showed inhibition activities (Supplementary
data S3). N-Dimethylation and N-acetylation (2a and 2b) also abol-
ished the inhibition of Rsk1 phosphorylation indicating that the
primary amine is essential to biological activity. The spacer ex-
tended analog 3a significantly inhibited phosphorylation of both
(EGF) (data not shown). To further prove that 76 and 1c do not af-
fect the catalytic activity of ERK1/2, in vitro cell free nonspecific
substrate myelin basic protein (MBP) phosphorylation assays were
performed using active recombinant ERK2 (Fig. 5). As expected, 1c
and 76 did not affect ERK1/2 catalytic activities. Taken together,
these results strongly suggest that 1c may serve as a new lead
structure and that modification at the 2-position and the adjacent
heterocycle ring (such as fused ring analogs) may lead to new can-
didate compounds with improved potency and selectivity profile.
To test whether the immunoblot analysis results correlate with
their functional activities, the active compounds were further eval-
uated for their effects on cell viability and proliferation in U937
cells with the MTS [(3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxy-
methoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium] assay and
incorporation of [3H]-thymidine assay, respectively (Fig. 6). MTS
assay assesses the metabolic activity as an indicator of healthy
Rsk1 and ERK1/2 at 50 lM (Fig. 4). Surprisingly, it also significantly
inhibited the phosphorylation of MEK1/2 at this concentration,
thus indicating that 3a may target upstream activating kinases of
ERK1/2 such as MEK and Raf-1. These results also demonstrate that
the benzylidene moiety of 76 is necessary for its interaction with
ERK1/2 and molecular size is an important structural determinant
for target specificity. This speculation is further supported by 1k (a
structurally constrained analog of 3a) and 1l that both showed
similar activity profiles as 3a. Interestingly, 1k and 1l showed bet-
ter inhibition of Elk-1 phosphorylation at 50
tion, 1k and 1l exhibited inhibition for p38 phosphorylation at
50 M while 3a did not. This further indicates the importance of
lM than 3a. In addi-
l
molecular size on target specificity. Saturation of the extra double
bond of 3a and introduction of two addition hydroxy groups led to
3b and 3c. These structural modifications almost completely abol-
ished activities for Rsk1 and ERK1/2 except for weak inhibition of
Rsk1 phosphorylation at 50 lM by 3b (Supplementary data S3).
Thus, conjugation is especially important for the biological activity
of 3a. It is notable that the shift of the ethoxy substitution from the
4- to the 2-position on the phenyl ring (1c) exhibited better inhibi-
tion of Rsk1 phosphorylation than 76 at both 10 and 50
minimal effects on the phosphorylation of ERK1/2 and enhanced
inhibition on Elk-1 phosphorylation at 50 M (Fig. 4). This may
lM, had
l
lead to improved functional activity of 1c and suggest options for
optimizing selectivity. Compounds 76 and 1c did not show any
inhibition for p38, MEK and JNK (data not shown) indicating that
they primarily target ERK1/2. Similar effects of phosphorylation
of ERK1/2 and its downstream targets were observed after treat-
ment of ovarian cancer cells (OVCAR5 and SKOV3) with the test
compounds prior to stimulation with epidermal growth factor
Figure 5. Effects of 76 and 1c on MBP phosphorylation using active recombinant
ERK2.

Q. Li et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6042–6046
6045
Figure 6. Effects of 76 and 1c on the viability (MTS, closed bars) and proliferation ([3H]-thymidine incorporation, dense bars) of U937 cells. Each bar represents the
**
*
mean SEM of at least three different experiments. P <0.05 or P <0.01 indicates significant differences from control group.
cells. But this assay cannot distinguish whether the cells are ac-
tively dividing or quiescent. On the other hand, [3H]-thymidine
incorporation assay will assess actively dividing cells in a sample
by detecting DNA synthesis. The evaluation of active compounds
in these two assays might shed light on the possible mechanism
of actions. All compounds tested (measured by MTS assay and up
chondria membrane potential. This might be due to their inhibi-
tory effects on the phosphorylation of p-38 (Fig. 4) and JNK (data
not shown), that may play opposite roles in regulating apoptosis
and survival.
Since 1c retained activity as a putative ERK1/2 substrate-spe-
cific inhibitor like 76, but showed better functional activities, it
would be of interest to understand the nature of the interactions
of these two compounds with ERK2. 76 and 1c were docked into
the docking groove of active ERK2 (PDB code: 2ERK) as described
by Hancock et al.19 using GOLD 3.027 (Fig. 8). The docking results
suggest that 1c retains the ionic interactions with Asp316 and
to 30 lM) showed no effects on U937 cell viability. In the original
report, 76 dose-dependently inhibited cell growth of various can-
cer cell lines in a colony formation assay. This discrepancy might
be caused by the different assays (MTS) and cancer cells (U937)
used in this study. Interestingly, 1c dose-dependently reduced
the incorporation of [3H]-thymidine. These results suggest that
1c is able to reduce the proliferation rate without significant effects
on cell viability possibly due to its enhanced inhibition of Elk-1 and
Rsk1 phosphorylation. Again, the potential of 1c to serve as a new
lead structure is evident.
Mitochondria have been shown to play an important role in cell
death and the loss of mitochondria membrane potential is an early
event in mitochondrially mediated apoptosis.25,26 To investigate
whether the active analogs induce early apoptotic effects, the
Asp319 and the
p-cation interactions with Arg133 are also the
same as 76. An additional hydrogen bond interaction was observed
between the 2-ethoxy group and Asn80 that may contribute to its
better Elk-1 inhibition and functional activities in cell proliferation
and apoptosis assays. Additional studies such as X-ray crystallogra-
phy are needed to verify these predicted interactions.
In summary, a series of analogs of compound 76 with different
substitutions on the phenyl ring, modified ethylamine tail and ex-
tended spacer between phenyl ring and thiazolidine-2,4-dione
were synthesized to conduct SAR studies and to define the phar-
macophore of 76. The immunoblot analysis assay of ERK1/2 down-
mitochondria membrane potential change (Dwm) in U937 cells
was measured using the 3,30-dihexyloxocarbocyanine iodide
(DiOC6) uptake assay. As shown in Figure 7, only 1c dose-depen-
dently increased the number of cells exhibiting mitochondria
membrane potential loss, which suggests inductive effects of 1c
on early apoptosis of U937 cells. Compounds 1k and 1l both exhib-
ited stronger inhibition of ERK1/2 and Rsk1 phosphorylation than
1c, but they, especially 1k, exhibited minimum effects on mito-
Figure 7. Effects of active compounds on the mitochondria membrane potential
loss (Dwm) of U937 cells. Each bar represents the mean SEM of at least three
*
**
different experiments. P <0.05 or P <0.01 indicates significant differences from
control group.
Figure 8. Predicted binding of 76 and 1c to active ERK2.

6046
Q. Li et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6042–6046
4. Downward, J. Nat. Rev. Cancer 2003, 3, 11.
5. Garnett, M. J.; Marais, R. Cancer Cell 2004, 6, 313.
stream substrate phosphorylation established that an oxygen atom
attached at 4-position of phenyl ring is important for its activity.
Steric effects in the phenyl ring domain are important factors in
determining target specificity since introduction of bulky moieties
or extension of the spacer between the phenyl ring and thiazoli-
dine-2,4-dione changes the targets, possibly upstream of ERK1/2.
We also discovered that the shift of 4-ethoxy substitution from
the 4-position to the 2-position on the phenyl ring of 76 can en-
hance Rsk1 and Elk-1 phosphorylation inhibition activity. More
importantly, 1c has no effects on the phosphorylation of ERK1/2
and their catalytic activities. Cell-based in vitro assays further
demonstrated that 1c can reduce the proliferation of U937 cells
without effects on cell viability and also induces mitochondria
membrane potential loss in a dose-dependent manner. These re-
sults strongly encourage further investigation of 1c and analogs
to develop more potent substrate-specific ERK1/2 inhibitors as
chemical probes and potential anti-cancer agents.
6. Davies, H.; Bignell, G. R.; Cox, C.; Stephens, P.; Edkins, S.; Clegg, S.; Teague, J.;
Woffendin, H.; Garnett, M. J.; Bottomley, W.; Davis, N.; Dicks, E.; Ewing, R.;
Floyd, Y.; Gray, K.; Hall, S.; Hawes, R.; Hughes, J.; Kosmidou, V.; Menzies, A.;
Mould, C.; Parker, A.; Stevens, C.; Watt, S.; Hooper, S.; Wilson, R.; Jayatilake, H.;
Gusterson, B. A.; Cooper, C.; Shipley, J.; Hargrave, D.; Pritchard-Jones, K.;
Maitland, N.; Chenevix-Trench, G.; Riggins, G. J.; Bigner, D. D.; Palmieri, G.;
Cossu, A.; Flanagan, A.; Nicholson, A.; Ho, J. W.; Leung, S. Y.; Yuen, S. T.; Weber,
B. L.; Seigler, H. F.; Darrow, T. L.; Paterson, H.; Marais, R.; Marshall, C. J.;
Wooster, R.; Stratton, M. R.; Futreal, P. A. Nature 2002, 417, 949.
7. Fransen, K.; Klintenas, M.; Osterstrom, A.; Dimberg, J.; Monstein, H. J.;
Soderkvist, P. Carcinogenesis 2004, 25, 527.
8. Steelman, L. S.; Abrams, S. L.; Whelan, J.; Bertrand, F. E.; Ludwig, D. E.; Basecke,
J.; Libra, M.; Stivala, F.; Milella, M.; Tafuri, A.; Lunghi, P.; Bonati, A.; Martelli, A.
M.; McCubrey, J. A. Leukemia 2008, 22, 686.
9. Ricciardi, M. R.; McQueen, T.; Chism, D.; Milella, M.; Estey, E.; Kaldjian, E.;
Sebolt-Leopold, J.; Konopleva, M.; Andreeff, M. Leukemia 2005, 19, 1543.
10. Kornblau, S. M.; Womble, M.; Qiu, Y. H.; Jackson, C. E.; Chen, W.; Konopleva, M.;
Estey, E. H.; Andreeff, M. Blood 2006, 108, 2358.
11. Bain, J.; McLauchlan, H.; Elliott, M.; Cohen, P. Biochem. J. 2003, 371, 199.
12. Kolch, W. Nat. Rev. Mol. Cell Biol. 2005, 6, 827.
13. Tanoue, T.; Maeda, R.; Adachi, M.; Nishida, E. EMBO J. 2001, 20, 466.
14. Zhang, J.; Zhou, B.; Zheng, C. F.; Zhang, Z. Y. J. Biol. Chem. 2003, 278, 29901.
15. Biondi, R. M.; Nebreda, A. R. Biochem. J. 2003, 372, 1.
Acknowledgments
16. Tanoue, T.; Adachi, M.; Moriguchi, T.; Nishida, E. Nat. Cell Biol. 2000, 2, 110.
17. Zhou, T.; Sun, L.; Humphreys, J.; Goldsmith, E. J. Structure 2006, 14, 1011.
18. Canagarajah, B. J.; Khokhlatchev, A.; Cobb, M. H.; Goldsmith, E. J. Cell 1997, 90,
859.
19. Hancock, C. N.; Macias, A.; Lee, E. K.; Yu, S. Y.; Mackerell, A. D., Jr.; Shapiro, P. J.
Med. Chem. 2005, 48, 4586.
The work was supported in part by Grant# IRG-73-001-34 from
the American Cancer Society (S.Z.), CA 10086 from the NIH (S.G.)
and new faculty start-up funds from Virginia Commonwealth Uni-
versity (S.Z.).
20. Chen, F.; Hancock, C. N.; Macias, A. T.; Joh, J.; Still, K.; Zhong, S.; MacKerell, A. D.,
Jr.; Shapiro, P. Bioorg. Med. Chem. Lett. 2006, 16, 6281.
21. Yang, J.; Wei, S.; Wang, D. S.; Wang, Y. C.; Kulp, S. K.; Chen, C. S. J. Med. Chem.
2008, 51, 2100.
Supplementary data
22. Motomura, W.; Tanno, S.; Takahashi, N.; Nagamine, M.; Fukuda, M.; Kohgo, Y.;
Okumura, T. Biochem. Biophys. Res. Commun. 2005, 332, 89.
23. Sugimura, A.; Kiriyama, Y.; Nochi, H.; Tsuchiya, H.; Tamoto, K.; Sakurada, Y.; Ui,
M.; Tokumitsu, Y. Biochem. Biophys. Res. Commun. 1999, 261, 833.
24. Momose, Y.; Maekawa, T.; Yamano, T.; Kawada, M.; Odaka, H.; Ikeda, H.; Sohda,
T. J. Med. Chem. 2002, 45, 1518.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.09.057.
References and notes
25. Norvell, S. M.; Green, K. J. J. Cell Sci. 1998, 111, 1305.
1. Raman, M.; Chen, W.; Cobb, M. H. Oncogene 2007, 26, 3100.
2. Yoon, S.; Seger, R. Growth Factors 2006, 24, 21.
3. Dhillon, A. S.; Hagan, S.; Rath, O.; Kolch, W. Oncogene 2007, 26, 3279.
26. Johnson, B. W.; Boise, L. H. J. Biol. Chem. 1999, 274, 18552.
27. Jones, G.; Willett, P.; Robert, G. C.; Leach, A. R.; Taylor, R. J. Mol. Biol. 1997, 267,
727.