Identification of KPNB1 as a Cellular Target of Aminothiazole Derivatives with Anticancer Activity

Article abstract of DOI:10.1002/cmdc.201600159

We found that aminothiazole derivative (E)-N-(5-benzylthiazol-2-yl)-3-(furan-2-yl)acrylamide (1) has strong anticancer activity, and undertook proteomics approaches to identify the target protein of compound 1, importin β1 (KPNB1). A competitive binding assay using fluorescein-labeled 1 showed that 1 has strong binding affinity for KPNB1 (K_d: ～20 nm). Furthermore, through western blotting assays for KPNB1, KPNA2, EGFR, ErbB2, and STAT3, we confirmed that 1 has inhibitory effects on the importin pathway. KPBN1 appears to be overexpressed in several cancer cells, and siRNA-induced inhibition of KPNB1 shows significant inhibition of cancer cell proliferation, while leaving non-cancerous cells unaffected. Therefore, compound 1 is a promising new lead for the development of KPNB1-targeted anticancer agents. Fluorescein-labeled 1 could be a useful quantitative probe for the development of novel KPNB1 inhibitors.

Full text of DOI:10.1002/cmdc.201600159

DOI: 10.1002/cmdc.201600159
Communications
Identification of KPNB1 as a Cellular Target of
Aminothiazole Derivatives with Anticancer Activity
Yong-Hak Kim,[b] Siyoung Ha,[a] Jungwon Kim,[a] and Seung Wook Ham*[a]
We found that aminothiazole derivative (E)-N-(5-benzylthiazol-
Table 1. IC50 values of compound 1 in various human cell lines.
2-yl)-3-(furan-2-yl)acrylamide (1) has strong anticancer activity,
and undertook proteomics approaches to identify the target
protein of compound 1, importin b1 (KPNB1). A competitive
binding assay using fluorescein-labeled 1 showed that 1 has
strong binding affinity for KPNB1 (Kd: ~20 nm). Furthermore,
through western blotting assays for KPNB1, KPNA2, EGFR,
ErbB2, and STAT3, we confirmed that 1 has inhibitory effects
on the importin pathway. KPBN1 appears to be overexpressed
in several cancer cells, and siRNA-induced inhibition of KPNB1
shows significant inhibition of cancer cell proliferation, while
leaving non-cancerous cells unaffected. Therefore, compound
1 is a promising new lead for the development of KPNB1-tar-
Compd
IC50 [nm][a]
SK-MEL2 HCT15
A549
SK-OV3
XF498
A431
1
38Æ2
1.3Æ0.6 11Æ1
5.4Æ0.5 22Æ2 43Æ2
Dox[b]
2Æ1 52Æ2
1.7Æ0.5 36Æ2 21Æ1 2.0Æ0.3
[a] Values are the meansÆSEM of n=3 experiments. [b] Reference com-
pound doxorubicin.
geted anticancer agents. Fluorescein-labeled
1
could be
several normal cell lines such as E. coli, yeasts, fungi, and
normal BJ cells. Our results are consistent with those reported
in a previous study, in which it was found that similar 2-amino-
thiazole analogues have selective effects on carcinoma cells
and exert antiproliferative effects specifically on DU-145
human prostate carcinoma cells.[2] Therefore, these results indi-
cate that aminothiazoles may exert antiproliferative effects via
a specific cellular target.
a useful quantitative probe for the development of novel
KPNB1 inhibitors.
Many 2-aminothiazoles exert their antitumor activity through
the inhibition of various kinases.[1] In the search for new com-
pounds with anticancer activity, we prepared aminothiazole
derivatives by replacing the 2-amino position and the 4,5-sub-
stituted groups (Supporting Information). We examined the cy-
totoxicities of the resulting several hundred aminothiazole
compounds against tumor cells; (E)-N-(5-benzylthiazol-2-yl)-3-
(furan-2-yl)acrylamide (1) showed strong antiproliferative ef-
fects in the nanomolar concentration range, and the potency
of compound 1 against ovarian (SK-OV3) and colon (HCT15)
cancer cells was found to be respectively ~50- and 7-fold
higher than that of doxorubicin (Table 1). A number of 2-ami-
nothiazoles were also prepared in order to investigate the
structure–activity relationship (SAR) of the inhibitory potency
of compound 1. In evaluating the antiproliferative effects of
the compounds toward the SK-OV3 cell line in a 48 h cytotox-
icity assay, the 2-furylethenolyl group was found to clearly pro-
vide the best inhibitory potency among the various acyl
groups, and the benzyl group provided better results than the
other 5-substituted groups and 4-substituted benzyl group
(Supporting Information). At a concentration of up to 1 mm,
compound 1 had no (or only marginal) inhibitory effects on
Many 2-aminothiazole derivatives exert their inhibitory activ-
ity against kinases by competing for the ATP binding site;
therefore, we examined the effects of 1 on a panel of cancer-
related kinases using a kinase profiling service (www.emdmilli-
pore.com/Signaling). However, at 10 mm, compound 1 showed
no or <50% inhibition of the 109 different protein kinases ex-
amined, indicating that the inhibitory activity of 1 on cancer
cells is not due to the cytotoxicity caused by the inhibition of
the protein kinases. Therefore, although these aminothiazole
compounds are promising new leads for the development of
novel anticancer agents, no studies have provided information
about a molecular target.
It has been reported that immobilization of the widely used
kinase inhibitor and optimized biochemical conditions for affin-
ity chromatography permit the enrichment and identification
of several unknown protein targets of the kinase inhibitor.[3] To
enrich potential targets of 1, we also performed affinity chro-
matography; the affinity and control columns were prepared
by coupling compound 2 to epoxy-activated Sepharose 6B and
incubation of the solid support with ethanolamine, respectively
(Figure 1). Then, whole-cell lysates obtained from 2.5ꢀ108
HeLa and SK-OV3 cells used as starting materials were evenly
loaded on the affinity matrix and the control matrix, and the
binding proteins were separated under optimized adsorption
and elution conditions as described in the Supporting Informa-
tion. The affinity-purified proteins were identified using two
proteomic approaches: 2D gel-based mass spectrometry and
quantitative liquid chromatography tandem mass spectrometry
[a] S. Ha, J. Kim, Prof. Dr. S. W. Ham
Department of Chemistry, Chung-Ang University,
221 Heakseok-dong, Dongjak-gu, Seoul 156-756 (Republic of Korea)
E-mail: swham@cau.ac.kr
[b] Prof. Dr. Y.-H. Kim
Department, Catholic University of Daegu School of Medicine,
3056-6 Daemyeong 4-dong, Nam-gu, Daegu 705-718 (Republic of Korea)
Supporting information for this article can be found under http://
dx.doi.org/10.1002/cmdc.201600159.
ChemMedChem 2016, 11, 1 – 5
1
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
per minute. This reaction velocity was sufficient to
show the concentration–inhibition curve of com-
pound 1 that had little or no effect on NQO2 up to
a concentration of 1 mm, which exceeded the solu-
bility of the compound (Figure S2, Supporting Infor-
mation). This inhibitory effect of 1 was even lower
than that of dicoumarol, with a half-maximal inhibi-
tory concentration (IC50) of 708 mm, which is general-
ly resistant with NQO2.[5] The concentration–inhibi-
tion curve indicated that NQO2 is not an effective
target protein of 1, although it effectively bound to
Figure 1. Structures of compound 2 and the affinity matrices.
a sample affinity resin and 1 eluent by the optimized
buffer systems in the affinity purification procedure.
The other candidate, KPNB1, is involved in the nu-
(LC–MS/MS). 2D gel electrophoresis was performed under reg-
ular polyacrylamide gel electrophoresis (PAGE) conditions in
the first dimension followed by sodium dodecyl sulfate (SDS)-
PAGE in the second dimension. We identified 23 protein spots
from the affinity-purified sample of HeLa cells, and 17 spots
clear translocation of proteins.[6] KPNB1 recognizes a nuclear lo-
calization sequence (NLS) on cargo proteins to be transported
to the nucleus and interacts with KPNA (importin a1) to
import proteins containing a classical NLS.[7] KPNB1 can also di-
rectly interact with the cargo proteins and transport them into
the nucleus via a non-classical transport pathway.[8] In cancer
cells, KPNB1 is reported to play a role in the nuclear transloca-
tion of phosphorylated STAT3 (pSTAT3), cell membrane recep-
tors, and receptor tyrosine kinases (RTK) containing NLS se-
quences, such as ErbB2 and epidermal growth factor receptor
(EGFR), from the cell surface to the nucleus.[9] Then, nuclear
EGFR and ErbB2 interact with several proteins including cy-
clin D1, cyclooxygenase-2 (COX-2), aurora kinase A, c-Myc,
breast cancer resistant protein (BCRP), STAT1, and B-Myb, acti-
vating their transcription, thus leading to accelerated cell-cycle
progression.[10]
To confirm the inhibitory effects of 1 on the function of
KPNB1, we performed western blotting using EGFR, ErbB2,
STAT3, and pSTAT3 in whole-cell lysates and nuclear fractions
of SK-OV3 treated with various concentrations of 1. Whereas
the nuclear levels of EGFR, ErbB2, STAT3, pSTAT3, KPNB1, and
KPNA decreased in a concentration-dependent manner after
treatment with 1, their levels in whole-cell lysates were nearly
constant (Figure 2). Results of western blotting showed that
compound 1 inhibits the classical importin pathway of KPNB1
and KPNA, and in turn decreased the nuclear transport of
EGFR, ErbB2, STAT3, and pSTAT3. The EGFR superfamily has
a significant impact on tumor initiation, cell-cycle progression,
tumor progression, and drug resistance. These membrane-
bound proteins have been intensively studied to block RTK-
centered signaling transduction networks, including the down-
stream pathways of STAT3. However, the internalization and
nuclear translocation of these membrane receptors induced by
specific agents highlight the need to modify the strategy for
the discovery of new anticancer agents, considering the nucle-
ar levels of RTKs as important markers in cancer cells.[11]
Regarding the binding effect of 1 on KPNB1, methods for
determining small-molecule binding affinity and analyzing
direct molecular interactions include surface plasmon reso-
nance (SPR) and isothermal titration calorimetry (ITC). However,
we were unable to obtain any information about binding affin-
ity using these methods because these are easily influenced by
exogenous factors. Therefore, we prepared fluorescein-labeled
from SK-OV3 cells (Figure S1, Supporting Information). To de-
termine the extent of purification, we performed quantitative
LC–MS/MS using an accurate and sensitive nanoLC-Orbitrap
LTQ tandem mass spectrometer, which enables a repeat analy-
sis of each sample of whole-cell lysates, affinity-purified pro-
teins, and control eluents to statistically compare the peptide
spectrum matches (PSMs) of the identified proteins, as de-
scribed in the Supporting Information. After normalization of
the PSMs of each sample, we calculated protein purification
folds (Tables S1 and S2, Supporting Information). From the pro-
teins highly purified using affinity chromatography, two com-
pound-1-binding proteins, KPNB1 (importin b1) and NQO2
(NRH:quinone oxidoreductase 2), were found in HeLa and SK-
OV3 cells; these were considered to be the target protein can-
didates of 1.
The first candidate, NQO2, is known to reduce quinones and
nitroaromatic compounds using a shortened nicotinamide de-
rivative, nicotinamide riboside, which is not present in cells.[4]
We examined the activity of a reconstituted NQO2 protein on
the reduction of dichlorophenolindophenol (DCPIP) with NRH
synthesized by the addition of calf intestinal alkaline phospha-
tase to nicotinamide mononucleotide in phosphate-buffered
saline, pH 7.4, as described previously.[4] Reoxidation-reduced
DCPIP occurred within a few minutes after saturation of the
enzyme reaction; therefore, the concentration of NQO2 was
limited to produce a 0.1-a.u. change in absorbance difference
&
ChemMedChem 2016, 11, 1 – 5
www.chemmedchem.org
2
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
a decrease in fluorescence (Figure 3c), giving a Kd value of
~20 nm.
Active site binding modes of compound 1 were then investi-
gated using ICM-Pro 3.6-1d molecular docking software (Mol-
soft) to gain structural insight into mechanisms by which 1 in-
hibits KPNB1. As shown in Figure 4, the docking of compound
1 into the pocket around S476 has the strongest binding
energy at 14.41 kcalmolÀ1. This model has three hydrogen
bonding sites between 1 and KPNB1, each of which have a dis-
tance within 2.6 ꢂ (dotted lines in Figure 4b, S476···O: 2.58 ꢂ,
E479···O: 2.63 ꢂ, and S475···O: 3.09 ꢂ). Participating in hydro-
gen bonding are the amide oxygen atom and the hydroxy
group of S476, the amide NH group of E479, and the furan
ring oxygen atom and hydroxy group of S475. These three hy-
drogen bonding interactions fix the binding direction of 1 by
placing the three positions in a nearly planar structure deter-
mined by central extended p bonding in compound 1. Anoth-
er weak interaction (3.54 ꢂ) is present between the phenyl ring
in compound 1 and the side chain of P440 through CH–p hy-
drogen bonds. This unconventional hydrogen bonding is
known to be important in the recognition of proline resi-
dues.[12]
Figure 2. Effect of compound 1 on in vitro nuclear import in SK-OV3 cells:
Cells were treated with compound 1 at various concentrations for 24 h at
378C, after which time nuclear protein extraction was performed using an
NE-PER nuclear extraction reagent. Proteins from the nuclear fraction and
total lysates were then analyzed by western blotting with antibodies to
ERBB-2, EGFR, KPNB1, KPNA2, STAT3, pSTAT3, Lamin A, and GAPDH.
1 for competitive binding assays.[13] To determine the Kd value
of the fluorophore probe, we incubated compound 1 (100 nm)
for 30 min with increasing concentrations of KPNB1 (0–2 mm).
The fluorescence intensity of fluorescein-1 increased by ~50%
upon addition of KPNB1 (Figure 3a). The affinity of fluorescein-
1 for KPNB1 was ~10 nm, as determined by nonlinear curve fit
of data according to the 1:1 binding model (Figure 3b). We
used a competition assay to determine the value of Kd for the
1–KPNB1 complex by using the Kd value of the complex be-
tween fluorescein-1 and KPNB1. Upon titrating the fluorescein-
1–KPNB1 complex with various concentrations of 1, the dis-
placement of fluorescein-1 by compound 1 was observed as
KPBN1 appears to be overexpressed in several cancer cell
types, and siRNA-induced inhibition of KPNB1 shows significant
inhibition of cancer cell proliferation, but does not affect
normal cells.[13] Results from these studies indicate that KPNB1
is a potential target, and inhibition of KPNB1 can be used as
a novel therapeutic approach for the treatment of cancer. A
pyrrole compound, karyostatin 1A, importazole, peptides, and
small protein-like peptidomimetics bind to KPNB1 and inhibit
the classical importin pathway.[14] However, some inhibitors are
not cell-permeable and their Kd values range single-digit micro-
molar to several hundred micromolar. Because its cytotoxicity
Figure 3. Competitive binding assays: a) Emission spectra of 100 nm fluorescein-1 excited at 491 nm with the addition of 0–2 mm KPNB1. b) Graphical analysis
of the fluorescence emission peak (520 nm) of fluorescein-1 with increasing concentrations of KPNB1. c) Titration curve of the fluorescein-1–KPNB1 complex
with various concentrations of compound 1. The obtained polarization values were analyzed by a nonlinear regression fitting program using GraphPad Prism
software (version 5.0).
ChemMedChem 2016, 11, 1 – 5
www.chemmedchem.org
3
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
M. S. Bednarz, K. A. Kellar, J. G. Mulheron, R. Batorsky, U. Roongta, A.
Kamath, P. Marathe, S. A. Ranadive, J. S. Sack, J. S. Tokarski, N. P. Lee, F. Y.
Pavletich, K. R. Webster, S. D. Kimball, J. Med. Chem. 2004, 47, 1719–
1728; b) J. Das, P. Chen, D. Norris, R. Padmanabha, J. Lin, R. V. Moquin,
Z. Shen, L. S. Cook, A. M. Doweyko, S. Pitt, S. Pang, D. R. Shen, Q. Fang,
H. F. de Fex, K. W. McIntyre, D. J. Shuster, K. M. Gillooly, K. Behnia, G. L.
Schieven, J. Wityak, J. C. Barrish, J. Med. Chem. 2006, 49, 6819–6832;
c) H. Hirai, T. Shimomura, M. Kobayashi, T. Eguchi, E. Taniguchi, K. Fuka-
sawa, T. Machida, H. Oki, T. Arai, K. Ichikawa, S. Hasako, K. Haze, T.
Kodera, N. Kawanishi, I. Takahashi-Suziki, Y. Nakatsuru, H. Kotani, Y. Iwa-
sawa, Cell Cycle 2010, 9, 1590–1600; d) I. Bruce, M. Akhlaq, G. C. Bloom-
field, E. Budd, B. Cox, B. Cuenoud, P. Finan, P. Gedeck, J. Hatto, J. F.
Hayler, D. Head, T. Keller, L. Kirman, C. Leblanc, D. Le Grand, C. McCar-
thy, D. O’Connor, C. Owen, M. S. Oza, G. Pilgrim, N. E. Press, L. Sviriden-
ko, L. Whitehead, Bioorg. Med. Chem. Lett. 2012, 22, 5445–5450; e) D.
Vogt, J. Weber, K. Ihlefeld, A. Brꢃggerhoff, E. Proschak, H. Stark, Bioorg.
Med. Chem. 2014, 22, 5354–5367; f) C. Fritsch, A. Huang, C. Chatenay-
Rivauday, C. Schnell, A. Reddy, M. Liu, A. Kauffmann, D. Guthy, D. Erd-
mann, A. De Pover, P. Furet, H. Gao, S. Ferretti, Y. Wang, J. Trappe, S. M.
Brachmann, S. M. Maira, C. Wilson, M. Boehm, C. Garcia-Echeverria, P.
Chene, M. Wiesmann, R. Cozens, J. Lehar, R. Schlegel, G. Caravatti, F.
Hofmann, W. R. Sellers, Mol. Cancer Ther. 2014, 13, 1117–1129; g) C. M.
Francini, A. L. Fallacara, R. Artusi, L. Mennuni, A. Calgani, A. Angelucci, S.
Schenone, M. Botta, ChemMedChem 2015, 10, 2027–2041; h) D. Das, P.
Sikdar, M. Bairaqi, Eur. J. Med. Chem. 2016, 109, 89–98.
[2] M. Krasavin, R. Karapetian, I. Konstantinov, Y. Gezentsvey, K. Bukhryakov,
E. Godovykh, O. Soldatkina, Y. Lavrovsky, A. V. Sosnov, A. Gakh, Arch.
Pharm. 2009, 342, 420–427.
[3] K. Godl, J. Wissing, A. Kurtenbach, P. Habenberger, S. Blencke, H. Gut-
brod, K. Salassidis, M. Stein-Gerlach, A. Missio, M. Cotton, H. Daub, Proc.
Natl. Acad. Sci. USA 2003, 100, 15434–15439.
[4] U. Oppermann, Annu. Rev. Pharmacol. Toxicol. 2007, 47, 293.
[5] K. Wu, R. Knox, X. Z. Sun, P. Joseph, A. K. Jaiswal, D. Zhang, P. S. Deng, S.
Chen, Arch. Biochem. Biophys. 1997, 347, 221–228.
[6] D. Gçrlich, S. Kostka, R. Kraft, C. Dingwall, R. A. Laskey, E. Hartmann, S.
Prehn, Curr. Biol. 1995, 5, 383–392.
Figure 4. Binding modes of compound 1 in KPNB1. The initial structure of
compound 1 in water was optimized with density functional theory (DFT)
methods implemented in Gaussian 09. The flexible 1 was docked to a grid
representation of KPNB1 (PDB ID: 2QNA) as receptor and assigned a score
for complex quality according to the iterative conditional modes (ICM)
method using ICM-Pro 3.6-1d docking software at Molsoft (www.molsoft.-
com). a) Binding position of 1 in KPNB1. b) Binding modes of 1 at the bind-
ing sites in KPNB1.
[7] D. Xu, A. Farmer, Y. M. Chook, Curr. Opin. Struct. Biol. 2010, 20, 782–790.
[8] G. Cingolani, J. Bednenko, M. T. Gillespie, L. Gerace, Mol. Cell 2002, 10,
1345–1353.
[9] a) H. W. Lo, M. Ali-Seyed, Y. Wu, G. Bartholomeusz, S. C. Hsu, M. C. Hung,
J. Cell. Biochem. 2006, 98, 1570–1583; b) J. Bromberg, J. E. Darnell Jr.,
Oncogene 2000, 19, 2468–2473.
[10] T. M. Brand, M. Iida, N. Luthar, M. M. Starr, E. J. Huppert, D. L. Wheeler,
Radiother. Oncol. 2013, 108, 370–377.
[11] a) S. C. Wang, M. C. Hung, Clin. Cancer Res. 2009, 15, 6484–6489; b) N. J.
Moerke, Curr. Protoc. Chem. Biol. 2009, 1, 1–15.
[12] T. Ozawa, K. Okazaki, K. Kitaura, J. Comput. Chem. 2011, 32, 2774–2782.
[13] a) P. J. van der Watt, C. L. Stowell, V. D. Leaner, Crit. Rev. Eukaryotic Gene
Expression 2013, 23, 1–10; b) P. J. van der Watt, C. P. Maske, D. T. Hen-
dricks, M. I. Parker, L. Denny, D. Govender, M. J. Birrer, V. D. Leaner, Int. J.
Cancer 2009, 124, 1829.
(~1–5 nm) is more potent than its binding affinity for KPNB1 (~
20 nm), compound 1 may also elicit cytotoxicity through other
targets. Nevertheless, our results indicate that 1 is a promising
new lead compound with higher potency for the development
of KPNB1-targeted anticancer agents. Moreover, fluorescein-
1 could be a useful quantitative probe for the identification of
a novel KPNB1 inhibitor by competition assays with compound
libraries.
Acknowledgements
[14] a) S. Kosugi, M. Hasebe, T. Entani, S. Takayama, M. Tomita, H. Yanagawa,
Chem. Biol. 2008, 15, 940–949; b) M. Hintersteiner, G. Ambrus, J. Bed-
nenko, M. Schmied, A. J. Knox, N. C. Meisner, H. Gstach, J. M. Seifert,
E. L. Singer, L. Gerace, M. Auer, ACS Chem. Biol. 2010, 5, 967–979; c) G.
Ambrus, L. R. Whitby, E. L. Singer, O. Trott, E. Choi, A. J. Olson, D. L.
Boger, L. Gerace, Bioorg. Med. Chem. 2010, 18, 7611–7620; d) J. F. Soder-
holm, S. L. Bird, P. Kalab, Y. Sampathkumar, K. Hasegawa, M. Uehara-
Bingen, K. Weis, R. Heald, ACS Chem. Biol. 2011, 6, 700–708.
This research was supported by the Chung-Ang University Excel-
lent Student Scholarship in 2014 and the Basic Science Research
Program (2013R1A1A2061369) of the National Research Founda-
tion of Korea.
Keywords: anticancer agents · competitive binding · KPNB1 ·
LC–MS/MS · proteomics
Received: March 18, 2016
Revised: April 11, 2016
[1] a) R. N. Misra, H. Y. Xiao, K. S. Kim, S. Lu, W. C. Han, S. A. Barbosa, J. T.
Hunt, D. B. Rawlins, W. Shan, S. Z. Ahmed, L. Qian, B. C. Chen, R. Zhao,
Published online on && &&, 0000
&
ChemMedChem 2016, 11, 1 – 5
www.chemmedchem.org
4
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

COMMUNICATIONS
An importin discovery: A proteomics
approach was used to identify the
target protein (KPNB1) of compounds
with anticancer activity. A competitive
binding assay using a fluorescein-la-
beled compound showed that the com-
pound has strong binding affinity for
KPNB1, suggesting that the KPNB1-tar-
geted anticancer agents and fluores-
cein-labeled compound could be useful
quantitative probes for the develop-
ment of novel KPNB1 inhibitors.
Y.-H. Kim, S. Ha, J. Kim, S. W. Ham*
&& – &&
Identification of KPNB1 as a Cellular
Target of Aminothiazole Derivatives
with Anticancer Activity
ChemMedChem 2016, 11, 1 – 5
www.chemmedchem.org
5
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ