Discovery of 2-pyrimidyl-5-amidothiophenes as potent inhibitors for AKT: Synthesis and SAR studies

Article abstract of DOI:10.1016/j.bmcl.2006.05.092

A series of 2-pyrimidyl-5-amidothiophenes has been synthesized and evaluated for AKT inhibition. SAR studies resulted in potent inhibitors of AKT with IC₅₀ values as low as single digit nanomolar as represented by compound 2aa. Compound 2aa sho

Full text of DOI:10.1016/j.bmcl.2006.05.092

Bioorganic & Medicinal Chemistry Letters 16 (2006) 4163–4168
Discovery of 2-pyrimidyl-5-amidothiophenes as potent inhibitors
for AKT: Synthesis and SAR studies
Xiaodong Lin,* Jeremy M. Murray, Alice C. Rico, Michael X. Wang, Daniel T. Chu,
Yasheen Zhou, Merci Del Rosario, Susan Kaufman, Sylvia Ma, Eric Fang,
Kenneth Crawford and A. B. Jefferson
Small Molecule Drug Discovery, Biopharma Research, Chiron Corporation, 4560 Horton Street, Emeryville, CA, USA
Received 14 April 2006; revised 25 May 2006; accepted 30 May 2006
Available online 9 June 2006
Abstract—A series of 2-pyrimidyl-5-amidothiophenes has been synthesized and evaluated for AKT inhibition. SAR studies resulted
in potent inhibitors of AKT with IC50 values as low as single digit nanomolar as represented by compound 2aa. Compound 2aa
showed cellular activity including antiproliferation and downstream target modulation. Selectivity profile is described. A co-crystal
of 2aa with PKA is determined and discussed.
Ó 2006 Elsevier Ltd. All rights reserved.
AKT is a serine/threonine kinase that plays a central
role in diverse cellular functions including cell growth
and proliferation, metabolism, and apoptosis. A number
of growth factors and cytokines activate phosphatidylin-
ositol-3 kinase (PI3K), which then phosphorylates
membrane phosphatidylinositols. The most highly phos-
phorylated of these, PI(3,4,5)P3, essentially serves as a
tethered ligand, directing the membrane localization
and activation of a number of cellular enzymes including
PDK1 and AKT. Full activation of AKT requires phos-
phorylation by PDK1 and an undefined PDK2. AKT, in
turn, phosphorylates a growing list of substrates includ-
ing GSK3, FOXO transcription factors, MDM2, TSC1/
2, and BAD.1,2
and PI3K has also been shown to have activating muta-
tions in tumors. Finally, each of the three AKT family
members (AKT1, AKT2, and AKT3)4–7 has been shown
to be amplified or overexpressed in at least one cancer
type. Thus, there are a number of patient populations
who might benefit from inhibition of signaling through
this pathway at the level of AKT.
Multiple approaches are being taken to target
AKT.3,4,8,9 AKT binds to PI(3,4,5)P3 via a pleckstrin
homology (PH) domain suggesting the possibility of
inhibiting AKT by interfering with this association.
Another strategy takes advantage of the sequence diver-
sity in the hinge region between the PH domain and ki-
nase domains of AKT1,2,3 to create isoform-selective
allosteric inhibitors. But by far the most highly explored
strategy is to target AKT’s kinase domain with ATP-
competitive inhibitors. While this strategy has a high
likelihood of generating potent inhibitors, it presents
the challenge of achieving selectivity, since the ATP-
binding site of AKT has high homology to other AGC
group kinases, particularly PKA.10 Small molecule
AKT inhibitors have been reported in recent literature
including ATP-competitive inhibitors11–14 as well as
allosteric inhibitors.15
Not surprisingly, given the many cellular functions reg-
ulated by AKT signaling, tumor cells have selected for
multiple circumventions of the normal checks on this
pathway.3 PTEN, a phosphatidylinositol phosphatase
that negatively regulates this pathway, is one of the most
frequently inactivated genes in tumors. Several receptor
tyrosine kinases such as EGFR and Her2 that signal
through the AKT pathway are overexpressed in cancer,
Keywords: Protein kinase B/AKT; kinase inhibitor; Substituted pyrim-
idine; Cancer.
Compound 1 (Fig. 1) was identified as an AKT3 inhib-
itor via high throughput screening of our compound col-
lection. It is an ATP-competitive inhibitor with an IC50
*
Corresponding author. Tel.: +1 510 923 7752; fax: +1 510 923
3360; e-mail: xiaodong_lin@chiron.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.05.092

4164
X. Lin et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4163–4168
Me
N
Me
H
N
Me
N
HN
HN
Me
N
HN
N
N
2
N
N
3
1
O
N
NH2
Cl
S
S
O
S
O
S
NH
O
Cl
F
O
N
F
H
OEt
N
5a
H
12
11
1
2aa
f
a, b
Me
N
H
Me
N
IC50 = 3.0 µM
IC50 = 2.6 nM
HN
N
N
Figure 1. From a low micromolar hit 1 to a potent inhibitor 2aa.
N
O
S
S
H
N
NH
O
e
Cl
N3
8
of 3.0 lM against AKT3. We report herein the structur-
al modifications of this compound, which led to a series
of novel and potent AKT inhibitors exemplified by com-
pound 2aa.
Cl
c
Me
N
H
Me
N
HN
N
N
d
N
Synthesis of analogs of 2-pyrimidyl-5-amidothiophene is
depicted in Scheme 1. Starting with 5-acetylthiophene-2-
carboxylic acid 3, the aminopyrimidine ring was formed
by a two-step sequence to give ethyl ester 5. The subse-
quent amide bond formation with various amines then
yielded analogs 2. Alternatively, 2-methylthiopyrimidine
6 was built from vinylogous amide 4 and S-meth-
ylisothiourea. Amide bond formation and oxidation
then gave sulfone 7. Finally SNAr reaction with amines
gave analogs 2. These two sequences enabled efficient
variation of amide position and 2-amino position of
the pyrimidine.
S
S
NH
O
F
NH2
9
10
Scheme 2. Reagents and conditions: (a) hydrazine, MeOH, 70 °C,
16 h, 87%; (b) 0.6 M HCl(aq), NaNO2, HOAc, 0 °C, 30 min, 91%; (c)
1.5 M HCl(aq), 100 °C, 16 h, 30%; (d) 3-(4-fluorophenyl)propionic acid,
CDI, THF, 50 °C, 30 min; then 9, 2 h, 50 °C, 30%; (e) xylenes, 120 °C,
25 min; then 2,4-dichlorobenzylamine, 120 °C, 30 min, 12%; (f)
xylenes, 120 °C, 25 min; then 4-fluorobenzylalcohol, xylenes, 120 °C,
45 min, 11%.
was formed by treating 5a with hydrazine followed by
diazatization. Subsequent Curtius rearrangement using
2,4-dichlorobenzylamine and 4-fluorobenzyl alcohol
yielded urea 11 and carbamate 12, respectively. Curtius
rearrangement using HCl gave aminothiophene 9, which
was converted to desired reverse amide 10.
The amide function in 1 can be modified into reverse
amide unit, urea and carbamate, using 5 as the interme-
diate. As shown in Scheme 2, intermediate acylazide 8
O
O
O
O
a
S
Scheme 3 summarizes the synthesis of 15 and 17. Analog
15 was prepared using commercially available thiophe-
S
Me2N
Me
OMe
OH
4
3
Cl
b
d
O
a, b
BrZn
S
N
N
R1
N
R2
N
SMe
N
O
N
S
O
N
H
13
O
S
14
O
S
c
Me
N
HN
OMe
c, d
OEt
5
N
6
S
Cl
c,e
N
H
R1
R2
N
15
Cl
N
Me
N
SO2Me
N
HN
SMe
N
f
N
N
O
e, f, g
SO2Cl
S
N
N
O
O
S
S
O
S
S
NHR3
Cl
NHR3
N
H
7
2
17
16
Cl
Scheme 1. Reagents and conditions: (a) N,N-dimethylformamide
dimethyl acetal, toluene, 110 °C, 16 h, 91%; (b) 1-methylguanidine
hydrochloride or guanidine hydrochloride, NaOEt, EtOH, 70 °C, 40 h,
45–58%; (c) substituted phenethylamine R3NH2, NaOMe, MeOH,
44 h, 7–55%; (d) S-methylisothiourea hemisulfate, pyridine, H2O,
100 °C, 68 h, 69%; (e) m-CPBA, DCM, rt, 100%; (f) amines R1R2NH,
1,4-dioxane, 110 °C, 3–16 h, 40–55%.
Scheme 3. Reagents and conditions: (a) 2,4-dichloropyrimidine,
(PPh3)4Pd(0), THF, 70 °C, 16 h; (b) 1.2 M HCl(aq), acetone, 16 h, rt;
(c) 2,4-dichlorophenethylamine, toluene, 110 °C, 4 h, removed vola-
tiles, MeOH, NaBH4, 2 h, rt, 54%; (d) MeNH2, THF, 75 °C, 16 h,
47%; (e) 2,4-dichlorophenethylamine, TEA, DCM, 3 h, rt; (f) m-
CPBA, 16 h, 100%; (g) MeNH2, THF, 75 °C, 16 h, 25%.

X. Lin et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4163–4168
4165
nyl zinc bromide 13. Negishi cross-coupling with 2,4-
dichloropyrimidine followed by hydrolysis gave inter-
mediate 14. Reductive amination and SNAr with
methylamine yielded desired secondary amine analog.
Sulfonamide analog 17 was synthesized using commer-
cially available starting material 16 via a straightforward
three-step sequence.
important for potency to maintain N1 and 2-NH
(Fig. 1) as H-bond acceptor and donor, respectively.
As shown in Table 1, when the 2-amino group is
bis-substituted (2q), the potency against AKT3 is com-
pletely eliminated. Replacing the methyl group on the
2-amino position with groups bigger than ethyl (2o vs
2p) leads to much less potent inhibitors, indicating
restricted protein space in this region. Other substituents
all proved to be detrimental to the potency. Throughout
our study, we have observed that 2-amino and 2-amino-
methyl substituent gave similar potency as exemplified
by 2n and 2m.
We started SAR studies by changing the fluorophenyl
ring in 1, as well as evaluating the tether length between
the phenyl ring and the amide nitrogen. As summarized
in Table 1, the tether length is very important to potency
(1, 2a–c). The optimal length appears to be two carbon
atoms (1). Using commercially available substituted
phenethylamines, we found that both the position and
the nature of substituents on the phenyl ring are crucial
to potency. Although it is clear that para-halogens are
beneficial (2g vs 2d), a chloro group (2g) is only margin-
ally superior to fluoro (1) and bromo (2h) groups. Sub-
stituents other than halogens such as methyl, methoxyl,
and aminosulfonyl are detrimental to the activity (2i–k).
para-Pyridyl is also detrimental (2l). It was also
observed that ortho- and meta-chloro groups are less
potent than corresponding para-analog (2f, 2e vs 2g).
Among several dihalogen substitutions, 2,4-dichloro
appears the best (2m).
Next, we decided to probe the effect of the amide group
in 1. As shown in Table 2, changing the amide group to
its reversed amide analog 10 or urea analog 11 does not
affect binding potency. Changing the amide to carba-
mate 12 and less constrained amine analog 15 leads to
slightly decreased potency. However, sulfonamide ana-
log 17 lost potency completely. These data suggest that
the amide bond is important sterically to place the
substituted phenyl group in a proper orientation for
potency. In contrast to s-trans conformation of an
Table 2. Activity of AKT inhibitorsa
Compound
IC50 (nM, AKT3)
At this point, we decided to fix the right-hand side as
2,4-dichlorophenethyl group in order to evaluate the ef-
fect of substitution on 2-amino group of the pyrimidine
ring. Based upon a homology model, we have postulated
(later confirmed by X-ray co-crystal structure, vide in-
fra) that the aminopyrimidine portion of molecule 1
interacts with the hinge domain of the kinase. It is
1
10
11
12
15
17
3000
2300
2100
7400
5200
>25,000
a AKT3 as assayed at 0.2 lM ATP in the presence of MnCl2.
Table 1. Activity of AKT inhibitorsa
R1
N
R2
N
N
O
N
S
R4
n
H
Compound
R1
R2
R4
n
IC50 (nM, AKT3)
1
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
H
4-Fluorophenyl
4-Fluorophenyl
4-Fluorophenyl
4-Fluorophenyl
Phenyl
2
1
0
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3000
4700
2a
2b
2c
2d
2e
2f
>25,000
18,000
5000
3-Chlorophenyl
2-Chlorophenyl
4-Chlorophenyl
4-Bromophenyl
4-Methylphenyl
4-Methoxyphenyl
4-Sulfamoylphenyl
4-Pyridyl
5400
2900
2g
2h
2i
1100
2600
>25,000
>25,000
19,000
>25,000
990
2j
2k
2l
2m
2n
2o
2p
2q
2,4-Dichlorophenyl
2,4-Dichlorophenyl
2,4-Dichlorophenyl
2,4-Dichlorophenyl
2,4-Dichlorophenyl
610
1500
Et
Pr
>25,000
>25,000
Me
a AKT3 as assayed at 0.2 lM ATP in the presence of MnCl2.

4166
X. Lin et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4163–4168
amide bond, sulfonamides prefer s-cis ground state con-
formation. Therefore, the phenyl group in the sulfon-
amide analog is no longer projected with the right
conformation for potency.
Lys72
Lys177
Met120
Met225
Phe54
Phe159
Glu121
Glu226
Substitution on the two-carbon chain between the phen-
yl ring and amide nitrogen of 1 proved to be very impor-
tant for AKT3 potency. Table 3 summarizes the effect of
carbon chain substitution. It was clear that the substitu-
tent R6 at the benzylic position (when R7 is phenyl or
substituted phenyl) can tolerate small alkyl groups (2r
and 2s) and basic amino group (2u). The bigger phenyl
group is not tolerated as R6 (2t). Small groups such as
methyl (2v) and methyl ester (2w) are tolerated at
homobenzylic position R5 (when R7 is substituted phen-
yl). Polar group hydroxymethyl (2x) improved potency
about 4-fold (compared to 2m). To our excitement, basic
aminomethyl group boosted potency to 4.8 nM (2y).
The S-enantiomer 2aa proved to have the optimal con-
figuration (compared to R-enantiomer 2z) for the bind-
ing. Substitutions on the primary amine (2bb) as well as
moving the amino group further away (2cc) turned out
to be less potent. Removal of the benzylic group resulted
in more than 500-fold potency loss (2dd vs 2aa and 2ee
vs 2bb). Therefore, both substituted phenyl ring and a
properly positioned primary amine are crucial for the
potency.
Val123
Val228
Asn171
Asn276
Leu173
Met278
Asp166
Asp269
Figure 2. X-ray crystal structure of 2aa in complex with PKA in the
active site with key interactions highlighted. The amino acids
highlighted in black are from PKA and red are from AKT3.
pyrimidine N1 and the backbone nitrogen of Val123.
The effects of the substitutions at the 2-amino position
of the pyrimidine are due to the close packing of resi-
˚
dues around the methyl group which is 3.5 A from the
˚
Ce of Met120. The residues within 4 A of the pyrimidine
moiety are mainly hydrophobic in nature (Val57, Ala70,
Val104, Met120, and Leu173), and the thiophene moiety
is sandwiched above and below the plane by the side-
chains of Val57 and Leu173. The oxygen atom of the
central amide group makes a water-mediated interaction
with the catalytic lysine (Lys72) indicating that the exact
placement of the hydrogen bonding atoms in this por-
tion of the molecule is less important than the role that
this portion plays in ensuring the correct orientation for
the substituted phenyl group.
The co-crystal structure of 2aa in complex with PKA
was determined by X-ray crystallography (Fig. 2, 3).16
PKA has previously been used as a surrogate for AKT
in structure-based drug design.12–14 The co-crystal struc-
ture of 2aa with PKA rationalizes much of the SAR.
The 2-aminopyrimidyl group binds to the hinge via
hydrogen bond interactions between the 2-amino group
and the backbone carbonyl group of Glu121, and the
Table 3. Activity of AKT inhibitorsa
H
N
Me
N
N
O
S
R5
HN
R7
R6
Compound
R5
R6
R7
IC50 (nM, AKT3)
2r
H
Me
Et
2,4-Dichlorophenyl
2,4-Dichlorophenyl
Phenyl
660
1000
>25,000
2800
820
2s
H
2t
2u
H
Phenyl
–NH2 (S)
H
–Me (R)
Phenyl
2v
H
H
H
H
H
H
H
H
H
H
2,4-Dichlorophenyl
2,4-Dichlorophenyl
2,4-Dichlorophenyl
2,4-Dichlorophenyl
2,4-Dichlorophenyl
2,4-Dichlorophenyl
2,4-Dichlorophenyl
2,4-Dichlorophenyl
NH2
2w
2x
2y
–CO2Me (S)
–CH2OH (S)
–CH2NH2
–CH2NH2 (R)
–CH2NH2 (S)
–CH2NMe2 (S)
–CH2CH2NH2 (S)
H
2400
280
4.8
2z
240
2aa
2bb
2cc
2dd
2ee
2.6
11
27
1300
7000
H
NMe2
a AKT3 as assayed at 0.2 lM ATP in the presence of MnCl2.

X. Lin et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4163–4168
4167
Figure 4. Compound 2aa inhibits PKA more potently than AKT in
UACC903 cells. Cells were treated with 2aa for 45 min prior to
stimulation with 7 lM forskolin for 15 min. Cell lysates were Western
blotted for phosphorylation of VASP by PKA (cell signaling #3111) or
MDM2 by AKT (cell signaling #3521).
Figure 3. X-ray crystal structure of compound 2aa bound to the ATP
binding site of PKA.
against other kinases with the exception of PKA, and
it shows good inhibition of proliferation and AKT tar-
get phosphorylation in the ovarian carcinoma cell line,
DOV13.
The primary amino group of the aminomethyl makes
important interactions as it projects into one of the
Mg2+ binding pockets of PKA, and it makes two direct
hydrogen bond interactions with protein sidechains, one
with Asn171 and the other with Asp184 from the DFG.
There is also a water-mediated interaction with Asp166.
Comparison of 2z and 2aa in Table 3 shows that the pre-
cise positioning of this group has a 100-fold effect on
potency. The tip of the Gly-rich loop clampsdown around
the disubstituted phenyl moiety, with Phe54 swinging
down to exclude bulk solvent from the active site
(Fig. 3). The para-Cl group points up into a predominant-
ly hydrophobic environment created by the sidechain
methylene groups of Phe54, Val57, Leu74, and Lys72.
The ortho-Cl group projects into a space between Val57
and the mainchain carbon atoms of Gly50 and Thr51.
While this compound appears to inhibit PKA at least
30x more potently than AKT, we sought to confirm this
in vitro potency differential in a cell line in which both
kinases were active concurrently. Thus, we pre-incubat-
ed UACC903 melanoma cells with various concentra-
tions of compound 2aa before stimulation with the
PKA activator, forskolin, for 15 min prior to cell lysis
(Fig. 4). PKA activity could be monitored through its
phosphorylation of VASP on Ser157 in response to for-
skolin stimulation, while AKT activity was monitored
by basal MDM2 phosphorylation on Ser166. Com-
pound 2aa inhibited cellular PKA activity at between
10 and 33 nM concentration, while AKT activity was
inhibited at 330 nM to 1 lM of 2aa. Thus, the apparent
ratio of inhibition of the two kinases observed in vitro
was confirmed in cells.
The selectivity and cellular activity of 2aa are listed in
Table 4. The compound has some degree of selectivity
In summary, we have developed a series of 2-pyrimidyl-
5-amidothiophene as potent AKT inhibitors. Com-
pounds in the series showed inhibition of cancer cell line
DOV13. One of the most potent inhibitors 2aa showed
moderate to excellent selectivity against a range of ki-
nases. The selectivity against AGC family kinases, espe-
cially PKA, is poor. The co-crystal structure with PKA
will guide the structure-based inhibitor design. Effort to
develop more selective inhibitors for AKT is underway
and the results will be reported in due course.
Table 4. Selectivity profile and cellular activity of AKT inhibitor 2aa
Test
Name
2aa (nM) Staurosporin
(nM)
Kinasesa
AKT1
AKT2
AKT3
PKA
6
23
3
11
9
5
0.9
0.1
PKCa
Kit
Met
66
110
1
>1
1100
130
19
EGFR
CaMK2a
ERK1
GSK3
RAF1
>10,000
8000
400
>1
References and notes
4900
12
>10,000
>10,000
1. Scheid, M. P.; Woodgett, J. R. FEBS Lett. 2003, 546, 108.
2. Bellacosa, A.; Testa, J. R.; Moore, R.; Larue, L. Cancer
Biol. Ther. 2004, 3, 268.
3. Cheng, J. Q.; Lindsley, C. W.; Cheng, G. Z.; Yang, H.;
Nicosia, S. V. Oncogene 2005, 24, 7482.
4. Kumar, C. C.; Madison, V. Oncogene 2005, 24, 7493.
5. Altomare, D. A.; Testa, J. R. Oncogene 2005, 24, 7455.
6. Nakatani, K.; Thompson, D. A.; Barthel, A.; Sakaue, H.;
Liu, W.; Weigel, R. J.; Roth, R. A. J. Biol. Chem. 1999,
274, 21528.
140
Cellular activity DOV13b
GSK3 S9Pc
1000
320
160
370
ND
ND
ND
ND
PRAS40 T246Pc
S6RP S235/236Pc
a All kinase IC50s were assayed at or below Km for ATP.
b Cell proliferation EC50 determined by Promega Cell TiterGlo.
c Cell phosphorylation EC50 determined by scanned Western blot in
DOV13 ovarian carcinoma cells.

4168
X. Lin et al. / Bioorg. Med. Chem. Lett. 16 (2006) 4163–4168
Rosenberg, S. H.; Giranda, V. L. Bioorg. Med. Chem.
Lett. 2006, 16, 1679.
7. Stahl, J. M.; Sharma, A.; Cheung, M.; Zimmerman, M.;
Cheng, J. Q.; Bosenberg, M. W.; Kester, M.; Sandirase-
garane, L.; Robertson, G. P. Cancer Res. 2004, 64, 7002.
13. Collins, I.; Caldwell, J.; Fonseca, T.; Donald, A.; Bave-
tsias, V.; Hunter, L. J.; Garrett, M. D.; Rowlands, M. G.;
Aherne, G. W.; Davies, T. G.; Berdini, V.; Woodhead, S.
J.; Davis, D.; Seavers, L. C.; Wyatt, P. G.; Workman, P.;
McDonald, E. Bioorg. Med. Chem. 2006, 14, 1255.
14. Breitenlechner, C. B.; Friebe, W. G.; Brunet, E.; Werner,
G.; Graul, K.; Thomas, U.; Kunkele, K. P.; Schafer, W.;
Gassel, M.; Bossemeyer, D.; Huber, R.; Engh, R. A.;
Masjost, B. J. Med. Chem. 2005, 48, 163.
15. Lindsley, C. W.; Zhao, Z.; Leister, W. H.; Robinson, R.
G.; Barnett, S. F.; Defeo-Jones, D.; Jones, R. E.;
Hartman, G. D.; Huff, J. R.; Huber, H. E.; Duggan, M.
E. Bioorg. Med. Chem. Lett. 2005, 15, 761.
8. Li, Q.; Zhu, G. D. Curr. Top. Med. Chem. 2002, 2, 939.
9. Machajewski, T.; Lin, X.; Jefferson, A. B.; Gao, Z. Annu.
Rep. Med. Chem. 2005, 40, 263.
10. Hanks, S. K.; Hunter, T. FASEB J. 1995, 9, 576.
11. Li, Q. W.; Keith, W.; Thomas, S.; Zhu, G.; Packard, G.;
Fisher, J.; Li, T.; Gong, J.; Dinges, J.; Song, X.; Abrams,
J.; Luo, Y.; Johnson, E.; Shi, Y.; Liu, X.; Klinghofer, V.;
Des Jong, R.; Oltersdorf, T.; Stoll, V.; Jakob, C.;
Rosenberg, S.; Giranda, V. Bioorg. Med. Chem. Lett.
2006, 16, 2000.
12. Li, Q.; Li, T.; Zhu, G. D.; Gong, J.; Claibone, A.; Dalton,
C.; Luo, Y.; Johnson, E. F.; Shi, Y.; Liu, X.; Klinghofer,
V.; Bauch, J. L.; Marsh, K. C.; Bouska, J. J.; Arries, S.;
Jong, R. D.; Oltersdorf, T.; Stoll, V. S.; Jakob, C. G.;
16. The deposition number is 2GU8 with the Protein Data
Bank (PDB) database.