MDM2-p53 protein-protein interaction inhibitors: A-ring substituted isoindolinones

Article abstract of DOI:10.1016/j.bmcl.2011.07.084

Structure-activity relationships for the MDM2-p53 inhibitory activity of a series of A-ring substituted 2-N-benzyl-3-(4-chlorophenyl)-3-(1-(hydroxymethyl) cyclopropyl)methoxy)isoindolinones have been investigated, giving rise to compounds with improved potency over their unsubstituted counterparts. Isoindolinone A-ring substitution with a 4-chloro group for the 4-nitrobenzyl, 4-bromobenzyl and 4-cyanobenzyl derivatives (10a-c) and substitution with a 6-tert-butyl group for the 4-nitrobenzyl derivative (10j) were found to confer additional potency. Resolution of the enantiomers of 10a showed that potent MDM2-p53 activity resided in the (-)-enantiomer ((-)-10a; IC₅₀ = 44 ± 6 nM). The cellular activity of key compounds has been examined in cell lines with defined p53 and MDM2 status. Compounds 10a and (-)-10a increase p53 protein levels, activate p53-dependent MDM2 and p21 transcription in MDM2 amplified cells, and show improved selectivity for growth inhibition in wild type p53 cell lines over the parent compound.

Full text of DOI:10.1016/j.bmcl.2011.07.084

Bioorganic & Medicinal Chemistry Letters 21 (2011) 5916–5919
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
MDM2-p53 protein–protein interaction inhibitors: A-ring substituted
isoindolinones
Anna F. Watson a, Junfeng Liu b, Karim Bennaceur b, Catherine J. Drummond b, Jane A. Endicott c,
Bernard T. Golding a, Roger J. Griffin a, Karen Haggerty a, Xiaohong Lu b, James M. McDonnell c,
David R. Newell b, Martin E.M. Noble c, Charlotte H. Revill a, Christiane Riedinger c, Qing Xu b,
a,
Yan Zhao b, John Lunec b, , Ian R. Hardcastle
⇑
⇑
a Newcastle Cancer Centre at the Northern Institute for Cancer Research and School of Chemistry, Bedson Building, Newcastle University, Newcastle, NE1 7RU, UK
b Newcastle Cancer Centre at the Northern Institute for Cancer Research, Paul O’Gorman Building, Medical School, Framlington Place, Newcastle University, Newcastle, NE2 4HH, UK
c Laboratory of Molecular Biophysics, Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Structure–activity relationships for the MDM2-p53 inhibitory activity of a series of A-ring substituted
2-N-benzyl-3-(4-chlorophenyl)-3-(1-(hydroxymethyl)cyclopropyl)methoxy)isoindolinones have been
investigated, giving rise to compounds with improved potency over their unsubstituted counterparts.
Isoindolinone A-ring substitution with a 4-chloro group for the 4-nitrobenzyl, 4-bromobenzyl and 4-cya-
nobenzyl derivatives (10a–c) and substitution with a 6-tert-butyl group for the 4-nitrobenzyl derivative
(10j) were found to confer additional potency. Resolution of the enantiomers of 10a showed that potent
MDM2-p53 activity resided in the (ꢀ)-enantiomer ((ꢀ)-10a; IC50 = 44 6 nM). The cellular activity of
key compounds has been examined in cell lines with defined p53 and MDM2 status. Compounds 10a
and (ꢀ)-10a increase p53 protein levels, activate p53-dependent MDM2 and p21 transcription in MDM2
amplified cells, and show improved selectivity for growth inhibition in wild type p53 cell lines over the par-
ent compound.
Received 16 June 2011
Revised 20 July 2011
Accepted 22 July 2011
Available online 9 August 2011
Keywords:
MDM2
p53
Protein–protein interaction inhibitor
Isoindolinone
Ó 2011 Elsevier Ltd. All rights reserved.
The tumour suppressor protein p53 functions as a molecular
sensor in diverse signalling pathways resulting from cellular stres-
ses, such as hypoxia, DNA damage and oncogene activation.1 The
MDM2 protein, the gene for which is a transcriptional target of
p53, downregulates p53 in a negative feedback loop.2,3 MDM2
binds to the p53 transactivation domain and ubiquitylates the
MDM2-p53 complex, resulting in export from the nucleus and
proteasomal degradation.4,5 Amplification of the MDM2 gene,
resulting in overexpression of MDM2 and inactivation of p53, has
been reported in around 11% of all tumours.6,7
considerable interest, and has been reviewed recently.8–10 The
MDM2-p53 binding interaction has many attractive features for
small-molecule inhibition, as it consists of a relatively deep bind-
ing groove on the surface of the MDM2 protein into which an
amphipathic helix of p53 binds.11 The X-ray crystal structures of
MDM2 with a number of peptide and small-molecule ligands
bound have been reported, demonstrating the prevalent key
shape-filling and hydrophobic interactions.12–14 Potent MDM2-
p53 inhibitors, such as Nutlin-3 (1)12 and the spirooxindoles, for
example, MI-63 and MI-219 (2a and 2b),15,16 have demonstrated
cellular activity consistent with inhibition of MDM2-p53 binding
and have shown in vivo antitumor activity.12,17
The investigation of small-molecule inhibitors of protein–pro-
tein interactions as potential therapeutic agents has received
OH
OH
Cl
H
N
N
O
O
HN
O
Cl
F
F
O
N
N
O
NH
NH
Cl
F
N
HN
O
O
O
Cl
N
N
H
Cl
Cl
O
H
Nutlin-3 (1)
MI-63 (2a)
MI-219 (2b)
⇑
Corresponding authors.
E-mail address: i.r.hardcastle@ncl.ac.uk (I.R. Hardcastle).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.07.084

A. F. Watson et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5916–5919
5917
Table 1
and reacted with the appropriate benzylamine to give 3-hydroxy-
isoindolinones 8 as a separable mixture of regioisomers in moder-
ate to good yields (42–74%). 3-Chlorination under Vilsmeier
conditions gave the 3-chloro isoindolinones 9a–m which were re-
acted immediately with the appropriate alcohol giving isoindoli-
nones 10a–m, as previously described.19 The saturated derivative
10n was prepared from 4,5,6,7-tetrahydroisobenzofuran-1,3-dione
following the same method.
MDM2-p53 inhibitory activity of isoindolinones 10a–n and 1,5-dihydro-3,4-dimeth-
ylpyrrol-2-ones 11a–b
HO
HO
HO
Cl
Cl
Cl
O
O
O
H3C
H3C
N
N
N
R'
R
O
O
O
The synthesis of the 1,5-dihydro-3,4-dimethylpyrrol-2-ones
11a and 11b required addition of a Grignard reagent to the appro-
priately substituted maleimide (Scheme 2). The preparation of the
4-nitrobenzyl analogue was incompatible with the Grignard route
so the N-propyl and N-benzyl compounds (11a and 11b) were pre-
pared as examples. Thus, N-benzyl-3,4-dimethylmaleimide 12a
and N-propyl-3,4-dimethylmaleimide 12b were synthesised
according to the literature method.21 4-Chlorophenylmagnesium
bromide was added to the benzyl and propyl maleimides to give
the corresponding 3-hydroxy products (13a,b) in good yields.22 Fi-
nally, treatment with the Vilsmeier reagent followed by addition of
the alcohol and base gave the benzyl and propyl 3,4-dimethyl-1,5-
dihydropyrrol-2-ones 11a and 11b in excellent yield for both
analogues.
X
X
A
B
C
a
Compound
Scaffold
X
R
R0
IC50 (nM)
3
4
5
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
B
C
C
NO2
Br
CN
NO2
Br
H
H
H
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
225
9
1200 600
1800 700
94 10b
10a
10b
10c
10d
10e
10f
10g
10h
10i
10j
10k
10l
10m
10n
11a
11b
4-Cl
4-Cl
4-Cl
4-Cl
4-Me
5-F
5-Me
5-tBu
5-Br
6-tBu
6-F
6-Br
5,6-Cl2
—
—
—
169
3
CN
185 17
9000 950
274 35
295 65
492 35
733 29
902 71
152 27
852 90
1030 40
3670 1150
3470 10
3450 90
8900c
SO2CF3
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NO2
—
The MDM2-p53 inhibitory activity of each final compound was
determined as described previously.19 In comparison with the par-
ent compound 3, the addition of a 4-chloro substituent to the A-ring
resulted in a twofold improvement in potency. Interestingly, in
comparison with the unsubstituted 4-bromobenzyl and 4-cyanob-
enzyl parent compounds (4 and 5) introduction of the A-ring
4-chloro substituent resulted in an approximately 10-fold improve-
ment in potency (10b and 10c). A significant improvement in
potency was seen for the 4-chloro substituted trifluoromethylsulf-
one derivative 10d, compared with the unsubstituted derivative
CH2Ph
C3H7
—
a
b
c
Values are the mean of three determinations SE.
n = 13.
n = 1.
(IC50 >50 l
M),19 however, in this series the trifluoromethylsulfone
Previously, we have reported MDM2-p53 inhibitors based on an
is not an effective nitro-bioisostere. The introduction of a 4-methyl
group to 3 did not result in an improvement in potency. Substitution
at the 5-position was not advantageous: the 5-fluoro derivative 10f
was equipotent with 3, whereas the larger methyl, tert-butyl and
bromo substituents resulted in a 2- to 4-fold loss of potency. The ef-
fect of substitution at the 6-position varied depending on the nature
of the substituent. Introduction of the sterically demanding tert-bu-
tyl group (10j) resulted in a modest improvement in activity,
whereas a fourfold loss of potency was seen for the 6-bromo com-
pound (10l). Surprisingly, 6-fluoro substitution (10k) was not well
tolerated. The 5,6-dichloro analogue (10m) was significantly less
potent than the parent isoindolinone (3). The 4,5,6,7-tetrahydroiso-
indolinone (10n) was greater than 10-fold less potent than the par-
ent, indicating a lack of tolerance for saturation in this ring. The N-
benzyl-1,5-dihydro-3,4-dimethylpyrrol-2-ones (11a) showed simi-
lar potency to 10n and the N-propyl analogue (11b) suffered a fur-
ther loss in potency.
isoindolinone scaffold.18,19 Structure–activity relationship (SAR)
studies guided by NMR binding experiments have demonstrated
the improved potency of 4-nitrobenzyl isoindolinones substituted
with sterically hindered 3-hydroxypropoxyl groups, for example,
Table 1 (3–5).20 In this paper, we describe SAR studies around
the isoindolinone aromatic or A-ring, leading to potent MDM2-
p53 inhibitors. The choice of A-ring substitution was limited by
the availability of the starting phthalic anhydrides for the chosen
route, thus the 3-chloro, 3-methyl, 4-fluoro, 4-tert-butyl, 4-bromo,
and 4,5-dichlorophthalic anhydrides were employed to investigate
SARs.
Substituted 2-benzoylbenzoic acids 7 were prepared from
phthalic anhydrides 6 via Friedel–Crafts acylation (Scheme 1).
3-Substituted phthalic anhydrides gave a mixture of 3- and 6-
substituted-2-(4-chlorobenzoyl) benzoic acids as an inseparable
mixture of regioisomers in a 20:1 ratio. Similarly, 4-substituted
phthalic anhydrides gave approximately equal mixtures of 4- and
5-substituted acids in moderate to good yields. The mixtures 7
The three most potent inhibitors 10a–c were resolved by chiral
HPLC and the inhibitory activity of the enantiomers was
determined (Table 2).20,23 In each case, the fastest eluting and
were converted into the corresponding mixture of
w
-acid chlorides
Cl
Cl
X
R
O
Y
O
O
e
a
b, c
R
N
N
R
R
R
O
Cl
CO2H
O
O
O
X
6
7
10a-m
Y = OH; 8
Y = Cl; 9
d
Scheme 1. Reagents and Conditions: (a) AlCl3, chlorobenzene, 90 °C; (b) SOCl2, DMF, THF; (c) substituted benzylamine, THF; (d) SOCl2, DMF, THF; (e) ROH, THF.

5918
A. F. Watson et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5916–5919
HO
O
Cl
Cl
O
O
HO
Me
Me
Me
Me
Me
Me
Me
Me
c
b
a
O
N
R'
N
R'
N
R'
O
O
O
O
12a; R' = CH2Ph
11a; R' = CH2Ph
13a; R' = CH2Ph
12b; R' = n-Pr
11b; R' = n-Pr
13b; R' = n-Pr
Scheme 2. Reagents and conditions: (a) R0NH2, THF, reflux, 16 h; (b) 4-chlorophenylmagnesium bromide, THF, ꢀ78 °C, 2 h; (c) (i) SOCl2, DMF, THF, rt, 4 h; (ii) 1,1-
bis(hydroxymethyl) cyclopropane, K2CO3, THF, rt, 16 h.
Table 2
Table 3
MDM2-p53 inhibitory activity of enantiomers of isoindolinones 10a–c
Growth inhibitory activity of (rac)-10a and enantiomers in MDM2 amplified cell lines
a
HO
Cell Lines
SRB assay GI50
Nutlin-3 10a
0.55 0.01 1.3 0.2 6.0 0.8 3.7 0.6 12.2 1.1
1.3 0.2 2.1 0.6 6.0 0.7 3.7 0.6 12.3 3.0
(
l
M)
MI-63
(ꢀ)-10a
(+)-10a
Cl
SJSA-1
HCT-116
Cl
O
*
HCT-116 p53 (ꢀ/ꢀ) 11.8 1.4 25.6 1.8 14.9 1.9 12.5 3.0 14.4 2.8
N
A2780
1.19 0.04 1.6 0.9 4.2 1.9 3.3 2.2 7.7 1.5
15.1 2.5 20.4 1.7 15.9 1.4 13.6 3.5 14.5 2.3
A2780 CP70
O
a
Values are the mean of at least three determinations SE.
10a-c
X
a
Compound
Rotation
X
IC50 (nM)
10a
10a
10b
10b
10c
10c
(+)
(ꢀ)
(+)
(ꢀ)
(+)
(ꢀ)
NO2
NO2
Br
Br
CN
CN
763 400b
44 18c
4015 285
137 67
4077 75
136 38
cells (MDM2 amplified, p53 (wt)) were treated with increasing
concentrations (1–20
l
M) of 10a, (+)-10a, and (ꢀ)-10a. Nutlin-3a
(1) was included for comparison (Fig. 1). Both racemic 10a and
(ꢀ)-10a induced a dose-dependent increase in MDM2, p53, and
p21 protein levels, consistent with cellular activation of p53. As ex-
pected, the less potent (+)-10a did not activate p53 at the highest
concentration, suggesting that the p53 dependent cellular activity
of (rac)-10a resides solely in the (+)-enantiomer.
a
b
c
Values are the mean of three determinations SE.
n = 6.
n = 7.
The growth inhibitory activity of compounds 10a, (+)-10a, and
(ꢀ)-10a was investigated in cell-lines with known p53 and
MDM2 status (Table 3). Nutlin-3 and MI-63 were included as posi-
tive controls for comparison. The GI50 values for 10a, (+)-10a, and
(ꢀ)-10a in p53 wild type, the MDM2 amplified cell line (SJSA-1)
showed that the least potent enantiomer (+)-10a is 2- to 3-fold less
growth inhibitory than the more potent (ꢀ)-10a. Compounds 10a
and (ꢀ)-10a were 2.5- to 4-fold more growth inhibitory in p53
wild-type cells (A2780 and HCT116 p53(+/+)) compared with their
isogenically paired p53 mutant (A2780-derived CP70) and p53 null
cell lines (HCT116 p53(ꢀ/ꢀ)), whereas a less than twofold differen-
tial was observed for the less potent (+)-10a. These results suggest
that this series possess growth inhibitory activity which is depen-
dent on p53-MDM2 inhibition, but also a component of p53-inde-
pendent activity is observed. Importantly, all classes of MDM2-p53
inhibitors tested showed similar growth inhibitory activities in
p53-null cell-lines at an equivalent potency. Isoindolinone (ꢀ)-
10a is 3- to 6-fold less growth inhibitory than MI-63 in p53 func-
tioning cells, consistent with the inhibitor’s 10-fold lower potency.
Similarly, (ꢀ)-10a 2- to 3-fold less growth inhibitory than Nutlin-3.
In conclusion, we have discovered that modifications to the A-
ring of the isoindolinones has significant impact on their MDM2-
p53 activity. In particular, introduction of a 4-chloro group re-
sults in an approximately 4-fold improvement in potency for
(ꢀ)-10a compared with (+)-3. Structural studies to rationalise
the observed SAR are ongoing. These results demonstrate the
versatility of the isoindolinone scaffold as MDM2-p53 inhibitors
and show that significant improvements in potency may be
gained by modest structural modifications. Studies are ongoing
to optimise the potency, selectivity and pharmaceutical proper-
ties of this series.
laevo-rotatory enantiomer was the most potent of the pair, while
the antipode displayed only modest inhibitory activity. Interest-
ingly, the (ꢀ)-enantiomer of the 4-nitro-isoindolinone 10a was
more potent than the either the 4-bromo or 4-cyano analogues
(10b and 10c). To date, an X-ray structure has not been obtained
for these enantiomers and so the absolute stereochemistry has
not been assigned, efforts are continuing. The eutomer of the par-
ent 3 was also the fastest eluting from the chiral HPLC column and
determined to have an (R)-configuration, but has the opposite (+)-
rotation.20 Racemisation under the assay conditions was not
observed.
The cellular activity of racemic 10a and its enantiomers was
determined by Western blotting as previously described.20 SJSA-1
Figure 1. Western blot analysis of SJSA-1 cells treated for 6 h with (rac)-10a and
enantiomers.

A. F. Watson et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5916–5919
5919
Lattanze, J.; Franks, C. F.; Zhao, S. Y.; Ramachandren, K.; Bylebyl, G. R.; Zhang,
M.; Manthey, C. L.; Petrella, E. C.; Pantoliano, M. W.; Deckman, I. C.; Spurlino, J.
C.; Maroney, A. C.; Tomczuk, B. E.; Molloy, C. J.; Bone, R. F. J. Med. Chem. 2005,
48, 909.
Acknowledgements
We thank Cancer Research UK, the EU (FP6, DePPICT), China
Scholarship Council (UK/China Scholarships for Excellence), MRC,
and the Wellcome Trust for funding. Use of the EPSRC National
Mass Spectrometry Service at the University of Wales (Swansea)
is gratefully acknowledged.
14. Allen, J. G.; Bourbeau, M. P.; Wohlhieter, G. E.; Bartberger, M. D.; Michelsen, K.;
Hungate, R.; Gadwood, R. C.; Gaston, R. D.; Evans, B.; Mann, L. W.; Matison, M.
E.; Schneider, S.; Huang, X.; Yu, D.; Andrews, P. S.; Reichelt, A.; Long, A. M.;
Yakowec, P.; Yang, E. Y.; Lee, T. A.; Oliner, J. D. J. Med. Chem. 2009, 52, 7044.
15. Shangary, S.; Qin, D.; McEachern, D.; Liu, M.; Miller, R. S.; Qiu, S.; Nikolovska-
Coleska, Z.; Ding, K.; Wang, G.; Chen, J.; Bernard, D.; Zhang, J.; Lu, Y.; Gu, Q.;
Shah, R. B.; Pienta, K. J.; Ling, X.; Kang, S.; Guo, M.; Sun, Y.; Yang, D.; Wang, S.
Proc. Natl. Acad. Sci. 2008, 105, 3933.
16. Yu, S.; Qin, D.; Shangary, S.; Chen, J.; Wang, G.; Ding, K.; McEachern, D.; Qiu, S.;
Nikolovska-Coleska, Z.; Miller, R.; Kang, S.; Yang, D.; Wang, S. J. Med. Chem.
2009, 52, 7970.
17. Mohammad, R. M.; Wu, J.; Azmi, A. S.; Aboukameel, A.; Sosin, A.; Wu, S.; Yang,
D. J.; Wang, S. M.; Al-Katib, A. M. Mol. Cancer 2009, 8.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.07.084.
18. Hardcastle, I. R.; Ahmed, S. U.; Atkins, H.; Farnie, G.; Golding, B. T.; Griffin, R. J.;
Guyenne, S.; Hutton, C.; Källblad, P.; Kemp, S. J.; Kitching, M. S.; Newell, D. R.;
Norbedo, S.; Northen, J. S.; Reid, R. J.; Saravanan, K.; Willems, H. M. G.; Lunec, J.
Bioorg. Med. Chem. Lett. 2005, 15, 1515.
19. Hardcastle, I. R.; Ahmed, S. U.; Atkins, H.; Farnie, G.; Golding, B. T.; Griffin, R. J.;
Guyenne, S.; Hutton, C.; Kallblad, P.; Kemp, S. J.; Kitching, M. S.; Newell, D. R.;
Norbedo, S.; Northen, J. S.; Reid, R. J.; Saravanan, K.; Willems, H. M. G.; Lunec, J.
J. Med. Chem. 2006, 49, 6209.
References and notes
1. Vousden, K. H.; Lu, X. Nat. Rev. Cancer 2002, 2, 594.
2. Momand, J.; Zambetti, G. P.; Olson, D. C.; George, D.; Levine, A. Cell 1992, 69,
1237.
3. Fuchs, S. Y.; Adler, V.; Buschmann, T.; Wu, X. W.; Ronai, Z. Oncogene 1998, 17,
2543.
4. Haupt, Y.; Maya, R.; Kazaz, A.; Oren, M. Nature 1997, 387, 296.
5. Kubbutat, M. H. G.; Jones, S. N.; Vousden, K. H. Nature 1997, 387, 299.
6. Oliner, J. D.; Kinzler, K. W.; Meltzer, P. S.; George, D. L.; Vogelstein, B. Nature
1992, 358, 80.
7. Toledo, F.; Wahl, G. M. Nat. Rev. Cancer 2006, 6, 909.
8. Shangary, S.; Wang, S. Clin. Cancer Res. 2008, 14, 5318.
9. Weber, L. Exp. Opin. Ther. Patents 2010, 20, 179.
10. Khoury, K.; Popowicz, G. M.; Holak, T. A.; Domling, A. MedChemComm 2011, 2,
246.
11. Kussie, P. H.; Gorina, S.; Marechal, V.; Elenbaas, B.; Moreau, J.; Levine, A. J.;
Pavletich, N. P. Science 1996, 274, 948.
20. Hardcastle, I. R.; Liu, J.; Valeur, E.; Watson, A.; Ahmed, S. U.; Blackburn, T. J.;
Bennaceur, K.; Clegg, W.; Drummond, C.; Endicott, J. A.; Golding, B. T.; Griffin,
R. J.; Gruber, J.; Haggerty, K.; Harrington, R. W.; Hutton, C.; Kemp, S.; Lu, X.;
McDonnell, J. M.; Newell, D. R.; Noble, M. E. M.; Payne, S. L.; Revill, C. H.;
Riedinger, C.; Xu, Q.; Lunec, J. J. Med. Chem. 2011, 54, 1233.
21. Punniyamurthy, T.; Katsuki, T. Tetrahedron 1999, 55, 9439.
22. Chen, M.-D.; He, M.-Z.; Zhou, X.; Huang, L.-Q.; Ruan, Y.-P.; Huang, P.-Q.
Tetrahedron 2005, 61, 1335.
23. Chiral separations: (10a) Diacel Chiralpak Ad-H, 250 ꢁ 10 mm; 80:20 hexane,
ethanol, 1.6 mL/min; rt, (ꢀ) 11.1 min, (+) 15.4 min; (10b) Diacel Chiralpac IA,
250 ꢁ 10 mm; 85:15 heptane, propan-2-ol, 3.75 mL/min; rt, (ꢀ) 17.8 min, (+)
24.9 min; (10c) Diacel Chiralpac IC, 250 ꢁ 4.6 mm; 95:5 heptane, ethanol,
1.0 mL/min; rt, (ꢀ) 17.4 min, (+) 23.3 min.
12. Vassilev, L. T.; Vu, B. T.; Graves, B.; Carvajal, D.; Podlaski, F.; Filipovic, Z.; Kong, N.;
Kammlott, U.; Lukacs, C.; Klein, C.; Fotouhi, N.; Liu, E. A. Science 2004, 303, 844.
13. Grasberger, B. L.; Lu, T. B.; Schubert, C.; Parks, D. J.; Carver, T. E.; Koblish, H. K.;
Cummings, M. D.; LaFrance, L. V.; Milkiewicz, K. L.; Calvo, R. R.; Maguire, D.;