Discovery of novel benzo[b][1,4]oxazin-3(4H)-ones as poly(ADP-ribose) polymerase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2013.06.055

Structure based drug design of a series of novel 1,4-benzoxazin-3-one derived PARP-1 inhibitors are described. The synthesis, enzymatic & cellular activities and pharmacodynamic effects are described. Optimized analogs demonstrated inhibition of poly-ADP-ribosylation in SW620 tumor bearing nude mice through 24 h following a single dose.

Full text of DOI:10.1016/j.bmcl.2013.06.055

Bioorganic & Medicinal Chemistry Letters 23 (2013) 4501–4505
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel benzo[b][1,4]oxazin-3(4H)-ones
as poly(ADP-ribose)polymerase inhibitors
a
a
a
a
Anthony R. Gangloff a, , Jason Brown , Ron de Jong , Douglas R. Dougan , Charles E. Grimshaw ,
Mark Hixon a, Andy Jennings a, Ruhi Kamran a, Andre Kiryanov a, Shawn O’Connell b, Ewan Taylor a
Phong Vu a
⇑
a Takeda California, 10410 Science Center Dr, San Diego, CA 92121, USA
b Quanticel Pharmaceuticals, 9393 Towne Center Dr #110, San Diego, CA 92121, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Structure based drug design of a series of novel 1,4-benzoxazin-3-one derived PARP-1 inhibitors are
described. The synthesis, enzymatic & cellular activities and pharmacodynamic effects are described.
Optimized analogs demonstrated inhibition of poly-ADP-ribosylation in SW620 tumor bearing nude mice
through 24 h following a single dose.
Received 18 April 2013
Revised 13 June 2013
Accepted 17 June 2013
Available online 24 June 2013
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Poly(ADP-ribose)polymerase inhibitors
PARP-1
Structure-based drug design
1,4-Benzoxazin-3-one
Chemopotentiation
The enzyme poly(ADP-ribose)polymerase (PARP), previously
known as poly(ADP-ribose)synthase and poly(ADP-ribosyl)trans-
ferase, constitutes a super family of proteins containing PARP cat-
alytic domains involved in the repair of DNA, the regulation of
cellular energy pools, and the transcription of inflammatory
genes.1 PARP-1 is a nuclear enzyme which catalyzes DNA repair
through cleavage of NAD+ to nicotinamide and ADP-ribose to form
long, branched ADP-ribose polymers on target proteins.2 PARP-1
inhibition leads to sustained DNA damage and ultimately apopto-
sis when used in combination with DNA damaging agents such
as temozolomide (TMZ), thus providing for chemopotentiation
in vitro and in vivo.3 PARP-1 inhibitors have also been shown to
act as single agents through a form of synthetic lethality4 by
exploiting a DNA repair flaw seen in tumors expressing defective
BRCA1/2-genes.5 Therefore, inhibition of PARP-1 (which is over-ex-
pressed in tumors) is expected to hinder intracellular DNA repair
and enhance the antitumor effects of cancer therapies.6
inhibitors, from which the 1,4-benz-oxazin-3-one core emerged
as a promising chemotype. Using a published PARP-1 X-ray crystal
structure,8 benzoxazinone 2a was overlaid onto the bound ligand
present in the crystal structure (Fig. 2). Our analysis based on
alignment of the cis-amide indicated the 6-position of the benzox-
azinone to be a suitable initial starting point for extension into the
back pocket to pick up interactions with the protein, although the
7-position could offer an additional vector. In addition, we antici-
pated that interactions with the hydrophobic shelf could be
achieved through substitution on the 2- and 8-position of the ben-
zoxazinone core.
The initial compounds were synthesized as outlined in
Scheme 1. Alkylation of substituted o-nitrophenols was accom-
plished in >70% yield using
a-bromoacetates in DMF with K2CO3.
Reduction of the nitro group and subsequent cyclization of the
resultant amine was achieved in a single step using iron dust in
AcOH at 80 °C. Reductive amination when R3 is a formyl group
introduced substituents which could extend into the back pocket.
A notable observation was that a methyl group at the 2-position
increased both binding potency against PARP-1 and ligand-lipo-
philic efficiency10 (LLE) as demonstrated by the benzoxazinone
fragments 2a and 2d where the 2-methyl group confers an order
of magnitude improvement in inhibitory activity (Table 1). In the
case of 3d and 3e, the Kd was increased by two log units and LLE
by 1.5 log units. These observations can be explained by the
presence of a proximal narrow hydrophobic cleft (defined by
A key feature of PARP inhibitors is the ability to mimic nicotin-
amide binding to the backbone carbonyl and amide N–H of Gly202
as well as the O–H of Ser243 in the NAD pocket of PARP-1 (Fig. 1).7
Hence, inhibitors containing a cis-amide capable of competing with
NAD+ were considered to be a good starting point. A screening
campaign identified several small fragments as weak PARP
⇑
Corresponding author. Tel.: +1 858 731 3531; fax: +1 858 550 0526.
E-mail address: anthony.gangloff@takeda.com (A.R. Gangloff).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.06.055

4502
A. R. Gangloff et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4501–4505
Ser243
O
H
N
H
Gly202
O
O
H
-
O
P
N
H
O
N
O
O
O
O
+
N
H
N
H
O
O
O
OH OH
N
O
NH2
H
H
N
N
H
N
N
N
O
F
N
P
O
N
O
HN
N
-
N
NH
N
NH
MK-4827
OH OH
NAD
O
ABT-888
AZD-2281
Figure 1. Structures of reported PARP inhibitors and NAD+ with PARP-1 catalytic domain.
with the tetrahydropyridine moiety in 3j produces sub-10 nM po-
tency against PARP-1. The X-ray crystal structure of 3j clearly
shows both the hydrophobic shelf occupied by the C-2 methyl
and also the trajectory of the chlorophenyl group off the tetrahy-
dropyridine ring into the back pocket (Fig. 3). The importance of
the regiochemistry for the vector off the benzoxazinone is clearly
O
F
O
N
H
N
H
N
O
N
1UK0 Ligand
2a
demonstrated when comparing 6-substituted 3g (0.046
7-substituted 3h (10 M), and likewise with regioisomeric 3j
(0.003 M) and 3k (100
M). Larger substituents at R1 lead to a
greater than 30,000-fold loss of binding potency such as the ethyl
group in 3l (100 M) compared to its methyl analog 3j (0.003 M).
lM) and
l
l
l
Gly202
Ser243
l
l
Overall, efficiency was improved compared to the fragment start-
ing point 2d with LLE values ranging from 4.5 to 4.9.
Back
Pocket
Although sub-100 nM PARP-1 inhibitors had been identified,
cellular potentiation was lacking for all compounds in Table 1.
There are many possible reasons for the lack of cellular activity,
including the inability of the inhibitors to engage intranuclear
PARP since the nuclear membrane is less permeable to xenobiotics
than the cellular membrane.11 One strategy for improving cellular
potency was to design compounds with reduced LogD and lipohi-
licity (cLogD values for 3g–3i fell within the range of 3.3–3.9)
could be attenuated. One approach to reduce logD values was to
replace the piperidine or tetrahydropyridine moiety present in
compounds 3f–3l with a piperazine group and optimizing the phe-
nyl substituents, thus lowering the LogD to an lower value of <3.0.
Exploration of back pocket interactions seen in Fig. 3 was enabled
through appropriate substitutions on the phenyl ring with poten-
tial for hydrogen bonding with Ile879 and/or Arg878. The X-ray
co-crystal structure12 of 3j also showed there was room for substi-
tution at the 8-position of the benzoxazinone core to allow for
6-Position
Hydrophobic
Shelf
Figure 2. Overlay of 1UK0 ligand in PARP-1 with benzoxazinone 2a.
Phe897 and Tyr896) which exists in the active site, and the
2-methyl group is optimally sized and positioned to for Van der
Waal’s interactions with these residues. Bis-substitution on C-2
such as 2g resulted in 15-fold reduced binding potency. Occupa-
tion of the back pocket as in 3e, 3g and 3j resulted in sub-micromo-
lar activity and further increases in LLE. Optimizing substituents
for accessing the back pocket from the benzoxazinone 6-position
2 O
2 O
R
R
R1
R1
NO2
R3
O
NO2
NH
NH
iii
HO
O
i
ii
O
R1 R2
O
O
when R3 = CHO
53-74%
70-90%
61-88%
R3
R3
R3
1a R1, R2, R3 = H
2a R1, R2, R3 = H
1b R1, R2 = H, R3 = 6-CHO
1c R1, R2 = H, R3 = 6-C(O)Me
1d R1 = Me, R2, R3 = H
3
2b R1, R2 = H, R3 = 6-CHO
2c R1, R2 = H, R3 = 6-C(O)Me
2d R1 = Me, R2, R3 = H
1e R1 = Me, R2 = H, R3 = 6-CHO
1f R1, R2 = Me, R3 = H
iv
2e R1 = Me, R2 = H, R3 = 6-CHO
2f R1, R2 = H, R3 = Et
1g R1 = Me, R2 = H, R3 = 7-CHO
1h R1 = Et, R2 = H, R3 = 6-CHO
2g R1, R2 = Me, R3 = H
2h R1 = Me, R2 = H, R3 = 7-CHO
2i R1 = Et, R2 = H, R3 = 6-CHO
Scheme 1. Reagents and conditions: (i) BrCR1R2CO2Me, K2CO3, DMF, 80 °C; (ii) iron dust, AcOH, 80 °C; (iii) R3NH2, NaBH(OAc)3, AcOH, DCE, 23 °C; (iv) Et3SiH, TFA, 80 °C.

A. R. Gangloff et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4501–4505
4503
Table 1
PARP-1 binding potencies and efficiency metrics for benzoxazinones 2–3
O
R2
O
R1
NH
R3
a
Compound
R1
R2
R3
PARP-1bKd
(l
M)
PARP-1c LLE
2a
2d
2g
2b
2f
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
>500
32
>500
170
100
26
11
5.2
15
0.096
0.296
0.046
10
0.015
0.003
100
100
Me
Me
H
H
H
H
Me
H
Me
H
Me
Me
H
2.3
6-Formyl
6-Ethyl
6-PhenylNHCH2
6-BenzylNHCH2
6-BenzylNHCH2
6-(4-(Me2N)piperidine)CH2
6-(4-(Me2N)piperidine)CH2
6-(4-(Phenyl)piperidine)CH2
6-(4-(Phenyl)piperidine)CH2
7-(4-(Phenyl)piperidine)CH2
2.0
2.0
2.2
3.5
3.2
3.0
4.6
3.4
3.7
1.4
4.5
4.9
0.2
0.1
6-(4-(4-Chlorophenyl)-1,2,3,6-tetrahydropyridine)CH2
6-(4-(4-Chlorophenyl)-1,2,3,6-tetrahydropyridine)CH2
7-(4-(4-Chlorophenyl)-1,2,3,6-tetrahydropyridine)CH2
6-(4-(4-Chlorophenyl)-1,2,3,6-tetrahydropyridine)CH2
Me
Me
Et
a
b
c
When applicable, compounds are racemic.
Determined using SPR9, n = 2.
LLE = pKdÀcLogD.
(a)
(b)
Ile879
Hydrophobic Shelf
Back
Pocket
Arg878
Figure 3. X-ray crystal structure of 3j (S-enantiomer shown) (a) binding pocket view (b) back pocket view.
O
O
O
R3
R3
N
H
OH
N
H
i
ii
N
F
F
20-77%
53-89%
R2
R2
HN
R2
2
2 = H
= Cl
R
R
2 = H, R3 = Me
R
= H, R3 = Me
4a
4b
5a
5b
6a
6b
R
R
2
2
= H, R3 = Et
R
2 = H, R3 = Et
5c R2 = H, R3 = cPr
5d R2 = Cl, R3 = Me
5e R2 = Cl, R3 = Et
6c R2 = H, R3 = cPr
6d R2 = Cl, R3 = Me
6e R2 = Cl, R3 = Et
2 = Cl, R3 = cPr
2 = Cl, R3 = cPr
5f
6f
R
R
O
N
N
R3
N
H
ii, iii
iv, i, v
N
F
N
70%
83-91%
N
HN
Boc
3 = Me
3 = Et
4c
6g
6h
6i
R
R
R
7
3 = cPr
Scheme 2. Reagents and conditions: (i) R3NH2, EDC, HOBt, NMM, DMF, 23 °C; (ii) piperazine, DMSO, 120 °C; (iii) Boc2O, THF, MeOH, 23 °C; (iv) NaOH, EtOH, H2O, 90 °C; (v)
HCl, dioxane, 23 °C.

4504
A. R. Gangloff et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4501–4505
O
O
O
R3
NO2
O
NO2
NH
NH
N
H
HO
R1
O
i
ii
iii
O
O
O
N
73-89%
73-90%
20-45%
R2
CHO
R1
CHO
N
R1
CHO
R1
8a R1 = Me
9a R1 = Me
10
8b R1 = OMe
9b R1 = OMe
iv
9c R1 = OH
v
9d
R
1 = OEt
Scheme 3. Reagents and conditions: (i) BrCHMeCO2Me, K2CO3, DMF, 80 °C; (ii) iron dust, AcOH, 80 °C; (iii) 6, NaBH(OAc)3, AcOH, DCE, 23 °C; (iv) BBr3, DCM, À10 °C; (v) Et-I,
CsCO3, DMF, 0 °C.
Table 2
Enzymatic and cellular activity for benzoxazinones 10
O
O
R3
NH
N
H
O
N
N
R2
R1
Compounda
R1
R2
R3
PARP-1b IC50 (nM)
Cell EC50 w/TMZ (
lM)
Cell EC50 w/o TMZ (lM)
Cell PF50
c
10a
10b
10c
10d
10e
10f
10g
10h
10i
10j
10k
10l
10m
10n
10o
10p
10q
10r
10s
10t
10u
ABT-888
AZD-2281
MK-4827
H
H
H
H
H
Cl
H
H
H
H
H
H
Cl
Cl
Cl
Me
Me
Me
Me
Me
Me
Me
Me
Me
—
Me
Et
Et
Me
Et
cPr
Me
Et
cPr
Me
Et
cPr
Me
Et
cPr
Me
Et
cPr
Me
Et
29
22
18
17
6
15
15
4
11
9
5
27
16
16
14
14
0.01
9
14
42
100
75
94
100
251
151
162
45
47
43
78
55
100
100
44
77
26
1.0
2.8
20
6.8
10.4
20
9.3
9.4
10.1
3.2
3.4
4300
8.7
6.1
5.6
1667
4400
3850
1300
5000
5000
56
Me
Me
Me
OMe
OMe
OMe
Me
Me
Me
Me
Me
Me
OMe
OMe
OMe
OEt
OEt
OEt
—
33
29
23
30
49
89
76
38
56
52
36
32
147
35
55
41
16
6
9
18
0.06
0.01
0.02
0.02
0.02
0.02
3
100
100
160
16
cPr
—
—
—
—
—
—
0.06
0.20
660
160
—
35
31
a
b
c
Compounds are racemic.
Determined using Trevigen assay, n = 2, av SD = 8 nM.
TMZ = 100 M.
l
interaction with the hydrophobic shelf and further optimization of
physicochemical properties.
The desired piperazines 6a–i were prepared using coupling and
nucleophilic aromatic substitution reactions (Scheme 2). Scheme 3
outlines the preparation of compounds 10 exemplified in Table 2 in
which all three substituent possibilities are exemplified: substitu-
tion at the benzoxazinone core (8-position R1), and back-pocket
phenyl group substitution (R2 and R3).
Cellular potency was determined in duplicate in the Jurkat cell
line using a cell viability assay that measured the conversion of
MTS after 96 h treatment with serially diluted compound in the
presence or absence of TMZ. Chemopotentiation factor (PF50) was
calculated as the ratio of EC50 values of cells co-treated with and
without TMZ, respectively.13
8.0 hr
-
8.0 hr
-
8.0 hr
24 hr
Biological evaluation showed R2 substitution maintained enzy-
matic potency (10b–c,Table 2).14 However, 10c where R2 is chloro
shows the first hint of chemopotentiation (PF50 = 20). When R2 = H
(10d–i), cellular chemopotentiation showed no improvement over
10c and the optimal substituent for R2 is methyl. The first
10l – 25 mpk
10l – 25 mpk
V
EH (0.5% MC)
TMZ - 50 MPK TMZ -50 MPK
TMX- 50 MPK
Figure 4. Pharmacodynamic effects of 10l in SW-620 tumor xenografts.15

A. R. Gangloff et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4501–4505
4505
compound with >100-fold cellular chemopotentiation was 10l
Supplementary data
where all three positions are substituted. Compounds 10p through
10u have an alkoxy substitutent at R1 and methyl at R2. Chemopot-
entiation was improved when R3 is ethyl or cyclopropyl compared
to methyl. Such optimal substitution patterns resulted in chemo-
potentiation factors above 1000. When compared to the clinical
compounds in Figure 1, the benzoxazinones showed chemopoten-
tiation factors with an order of magnitude improvement in many
cases despite being 3- to 6-fold less potent in the enzymatic assay.
Our objective to design higher quality compounds (cLogD <3.0)
which maintained the enzymatic potencies seen in Table 1 but also
displayed chemopotentiation in cellular systems had been
demonstrated.
The increased chemopotentiation seen with 10l resulted in a
pharmacodynamic effect in tumor bearing mice, preventing poly-
ADP-ribosylation in SW-620 tumor xenografts at 8 and 24 h post
dose (Fig. 4).15 SW-620 (a p53À/À human colorectal adenocarci-
noma cell line) was implanted into nude mice and allowed to grow
to ꢀ325 75 mg. The PARP inhibitor 10l was dosed orally with TMZ
(50mpk). Tumors were removed 8 or 24 h after dosing, homoge-
nized, normalized for total protein and the lysates were analyzed
using Trevigen ELISA (Cat. #: 4510-096-K). A strong and sustained
tumor pharmacodynamic response was observed with 10l main-
taining inhibition of poly-ADP-ribosylation through 24 h.
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.bmcl.2013.06.055.
References and notes
1. Rouleau, M.; Patel, A.; Hendzel, M. J.; Kaufmann, S. H.; Poirier, G. G. Nat. Rev.
Cancer 2010, 10, 293.
2. Chambon, P.; Weil, J. D.; Mandel, P. Biochem. Biophys. Res. Commun. 1963, 11,
39.
3. Delaney, C. A.; Wang, L. Z.; Kyle, S.; White, A. W.; Calvert, A. H.; Curtin, N. J.;
Durkacz, B. W.; Hostomsky, Z.; Newell, D. R. Clin. Cancer Res. 2000, 6, 2860;
Miknyoczki, S. J.; Jones-Bolin, S.; Pritchard, S.; Hunter, K.; Zhao, H.; Wan, W.;
Ator, M.; Bihovsky, R.; Hudkins, R.; Chatterjee, S.; Klein-Szanto, A.; Dionne, C.;
Ruggeri, B. Mol. Cancer Ther. 2003, 2, 371; Calabrese, C. R.; Batey, M. A.; Thomas,
H. D.; Durkacz, B. W.; Wang, L. Z.; Kyle, S.; Skalitzky, D.; Li, J.; Zhang, C.;
Boritzki, T.; Maegley, K.; Calvert, A. H.; Hostomsky, Z.; Newell, D. R.; Curtin, N. J.
Clin. Cancer Res. 2003, 9, 2711; Donawho, C. K.; Luo, Y.; Penning, T. D.; Bauch, J.
L.; Bouska, J. J.; Bontcheva-Diaz, V. D.; Cox, B. F.; DeWeese, T. L.; Dillehay, L. E.;
Ferguson, D. C.; Ghoreishi-Haack, N. S.; Grimm, D. R.; Guan, R.; Han, E. K.;
Holley-Shanks, R. R.; Hristov, B.; Idler, K. B.; Jarvis, K.; Johnson, E. F.; Kleinberg,
L. R.; Klinghofer, V.; Lasko, L. M.; Liu, X.; Marsh, K. C.; McGoniga, T. P.;
Meulbroek, J. A.; Olson, A. M.; Palma, J. P.; Rodriguez, L. E.; Shi, Y.; Stavropoulos,
J. A.; Tsurutani, A. C.; Zhu, G. D.; Rosenberg, S. H.; Giranda, V. L.; Frost, D. J. Clin.
Cancer Res. 2007, 13, 2728.
4. Leung, M.; Rosen, D.; Fields, S.; Cesano, A.; Budman, D. R. Mol. Med. 2011, 17,
854.
5. Bryant, H. E.; Schultz, N.; Thomas, H. D.; Parker, K. M.; Flower, D.; Lopez, E.;
Kyle, S.; Meuth, M.; Curtin, N. J.; Helleday, T. Nature 2005, 434, 913; Farmer, H.;
McCabe, N.; Lord, C. J.; Tutt, A. N.; Johnson, D. A.; Richardson, T. B.; Santarosa,
M.; Dillon, K. J.; Hickson, I.; Knights, C.; Martin, N. M.; Jackson, S. P.; Smith, G.
C.; Ashworth, A. Nature 2005, 434, 917; Balmana, J.; Domchek, S. M.; Tutt, A.;
Garber, J. E. Cancer Disc. 2011, 1, 29.
6. Ferraris, D. V. J. Med. Chem. 2010, 53, 4561.
7. Residues numbered as found in the PDB file 1UK0.
8. Kinoshita, T.; Nakanishi, I.; Warizaya, M.; Iwashita, A.; Kido, Y.; Hattori, K.;
Fujii, T. FEBS Lett. 2004, 556, 43.
9. Complete details for SPR determination can be found in the Supplementary
data.
10. Edwards, M. P.; Price, D. A. Annu. Rep. Med. Chem. 2010, 45, 380.
11. Chi, Y.-H.; Chen, Z.-J.; Jeang, K.-T. J. Biomed. Sci. 2009, 16.
12. The coordinates of compound 3j have been deposited in the RCSB Protein Data
Bank (PDB ID code 4L6S). To obtain the co-crystal structure, racemic 3j was
utilized but only the S-enantiomer was seen in the resulting structure. Based
on modeling, both enantiomers are possible but S- is preferred.
13. Jurkat cells were seeded in 96-well tissue culture microplates at 10,000 cells
per well and cultured for 24 h prior to addition of compounds, TMZ
(Temozolomide) or DMSO (dimethylsulfoxide) vehicle. After 96 h, conversion
of MTS by metabolically active cells was determined. Cells were treated in
duplicate with a range of serial compound dilutions in the absence or presence
The major liability of the PARP inhibitors described above which
prevented further development was hERG inhibition. Compounds
10p, 10q, and 10s all exhibited >97% inhibition of the hERG chan-
nel in an electrophysiology assay at 10 lM.
In summary, PARP inhibitors have been developed from a frag-
ment screening campaign utilizing structure-based drug design
principles to yield novel benzoxazinones displaying sub-100 nM
potency in enzymatic assays and cellular potentiation factors
greater than 1000. When optimally substituted, the compounds
show an improvement in synergy over PARP inhibitors currently
in clinical development. The increased cellular activity translated
to an in vivo pharmacodynamic response with sustained inhibition
for 24 h. Further development of the benzoxazinone scaffold into
drug-like inhibitors which demonstrate in vivo efficacy and over-
come a hERG liability seen in the above compounds has been pur-
sued. The inhibitors described here provide additional tools to
explore the role of PARP-1 in cancer and other diseases.
of 100 lM TMZ. Complete details can be found in the Supplementary data.
Acknowledgments
14. Inhibition of PARP catalytic activity was determined by use of an ELISA-based
colorimetric PARP/Apoptosis Assay kit (part number 4684-096-K HT, Trevigen).
Complete details can be found in the Supplementary data.
15. 3 animals per group. Vehicle for all groups was 0.5% methylcellulose. Average
standard deviation was 0.14 (Vehicle), 0.36 (TMZ), 0.81 (10l, 8 h), 0.14 (10l,
24 h).
We thank Dr. Matthew Marx and Dr. Gavin Hirst for their crit-
ical reading of the manuscript, Dr. Clifford D. Mol for crystallogra-
phy work and Philip Erickson for chemical synthesis.