Halogen-substituted anthranilic acid derivatives provide a novel chemical platform for androgen receptor antagonists

Article abstract of DOI:10.1016/j.jsbmb.2018.12.005

Androgen receptor (AR) antagonists are used for hormone therapy of prostate cancer (PCa). However resistance to the treatment occurs eventually. One possible reason is the occurrence of AR mutations that prevent inhibition of AR-mediated transactivation by antagonists. To offer in future more options to inhibit AR signaling, novel chemical lead structures for new AR antagonists would be beneficial. Here we analyzed structure-activity relationships of a battery of 36 non-steroidal structural variants of methyl anthranilate including 23 synthesized compounds. We identified structural requirements that lead to more potent AR antagonists. Specific compounds inhibit the transactivation of wild-type AR as well as AR mutants that render treatment resistance to hydroxyflutamide, bicalutamide and the second-generation AR antagonist enzalutamide. This suggests a distinct mode of inhibiting the AR compared to the clinically used compounds. Competition assays suggest binding of these compounds to the AR ligand binding domain and inhibit PCa cell proliferation. Moreover, active compounds induce cellular senescence despite inhibition of AR-mediated transactivation indicating a transactivation-independent AR-pathway. In line with this, fluorescence resonance after photobleaching (FRAP) - assays reveal higher mobility of the AR in the cell nuclei. Mechanistically, fluorescence resonance energy transfer (FRET) - assays indicate that the amino-carboxy (N/C)-interaction of the AR is not affected, which is in contrast to known AR-antagonists. This suggests a mechanistically novel mode of AR-antagonism. Together, these findings indicate the identification of a novel chemical platform as a new lead structure that extends the diversity of known AR antagonists and possesses a distinct mode of antagonizing AR-function.

Full text of DOI:10.1016/j.jsbmb.2018.12.005

Accepted Manuscript
Title: Halogen-substituted anthranilic acid derivatives provide
a novel chemical platform for androgen receptor antagonists
Authors: Daniela Roell, Thomas W. Ro¨sler, Wiebke
Hessenkemper, Florian Kraft, Monique Hauschild, Sophie
Bartsch, Tsion E. Abraham, Adriaan B. Houtsmuller, Rudolf
Matusch, Martin E. van Royen, Aria Baniahmad
PII:
DOI:
Reference:
S0960-0760(18)30570-3
https://doi.org/10.1016/j.jsbmb.2018.12.005
SBMB 5269
To appear in:
Journal of Steroid Biochemistry & Molecular Biology
Received date:
Revised date:
Accepted date:
14 September 2018
4 December 2018
7 December 2018
Please cite this article as: Roell D, Ro¨sler TW, Hessenkemper W, Kraft F, Hauschild M,
Bartsch S, Abraham TE, Houtsmuller AB, Matusch R, van Royen ME, Baniahmad A,
Halogen-substituted anthranilic acid derivatives provide a novel chemical platform for
androgen receptor antagonists, Journal of Steroid Biochemistry and Molecular Biology
(2018), https://doi.org/10.1016/j.jsbmb.2018.12.005
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Halogen-substituted anthranilic acid derivatives provide a novel chemical
platform for androgen receptor antagonists
Daniela Roella,b,§, Thomas W. Röslerc,§, Wiebke Hessenkempera, Florian Krafta, Monique
Hauschilda, Sophie Bartscha, Tsion E. Abrahamd, Adriaan B. Houtsmullerd, Rudolf Matuschc,
Martin E. van Royend, Aria Baniahmada,*
aInstitute of Human Genetics, Jena University Hospital, Jena, Germany
bIntegrated Research and Treatment Center, Center for Sepsis Control and Care (CSCC), Jena
University Hospital, Jena, Germany
cInstitute of Pharmaceutical Chemistry, Philipps-University, Marburg, Germany
dDepartment of Pathology and Erasmus Optical Imaging Center OIC, Erasmus MC, Rotterdam, The
Netherlands
§These authors contributed equally.
*Corresponding author:
Jena University Hospital, Institute of Human Genetics, Am Klinikum 1, 07740 Jena, Germany
Ph.: +49-3641-9396820; Fax: +49-3641-99396822; aria.baniahmad@med.uni-jena.de

Graphical abstract
Highlights
 Identification of a novel chemical platform as a new lead structure that extends the
diversity of known AR antagonists
 Novel AR antagonists possess a distinct mode of antagonizing AR-functions
 Novel AR antagonists inhibit prostate cancer cell proliferation
 Novel AR antagonists inhibit chromatin accessibility
 Novel AR antagonists induce cellular senescence
 Novel AR antagonists inhibit AR mutants that mediate therapy resistance
2

Abstract
Androgen receptor (AR) antagonists are used for hormone therapy of prostate cancer (PCa).
However resistance to the treatment occur eventually. One possible reason is the occurrence of AR
mutations that prevent inhibition of AR-mediated transactivation by antagonists. To offer in future
more options to inhibit AR signaling, novel chemical lead structures for new AR antagonists would
be beneficial. Here we analyzed structure-activity relationships of a battery of 36 non-steroidal
structural variants of methyl anthranilate including 23 novel synthesized compounds. We identified
structural requirements that lead to more potent AR antagonists. Specific compounds inhibit the
transactivation of wild-type AR as well as AR mutants that render treatment resistance to
hydroxyflutamide, bicalutamide and the second-generation AR antagonist enzalutamide. This
suggests a distinct mode of inhibiting the AR compared to the clinically used compounds.
Competition assays suggest binding of these compounds to the AR ligand binding domain and
inhibit PCa cell proliferation. Moreover, active compounds induce cellular senescence despite
inhibition of AR-mediated transactivation indicating a transactivation-independent AR-pathway. In
line with this, fluorescence resonance after photobleaching (FRAP) - assays reveal higher mobility
of the AR in the cell nuclei. Mechanistically, fluorescence resonance energy transfer (FRET) -
assays indicate that the amino-carboxy (N/C)-interaction of the AR is not affected, which is in
contrast to known AR-antagonists. This suggests a mechanistically novel mode of AR-antagonism.
Together, these findings indicate the identification of a novel chemical platform as a new lead
structure that extends the diversity of known AR antagonists and possesses a distinct mode of
antagonizing AR-function.
Abbreviations: AR, androgen receptor; Cas, Casodex (bicalutamide); CRPCa, castration resistant PCa; CSS, charcoal stripped serum;
FACS, fluorescence-activated cell sorter; FRAP, fluorescence recovery after photobleaching, FRET, fluorescence resonance energy
transfer; GR, glucocorticoid receptor; N/C interaction, amino- / carboxy- terminal interaction; OH-F, hydroxy-flutamide; PCa, prostate
cancer; PR, progesterone receptor; PSA, prostate specific antigen; SA-β-Gal, senescence-associated β-galactosidase
Keywords: methyl anthranilate derivatives, antiandrogens, androgen receptor, antihormone, prostate
cancer, cellular senescence
3

1. Introduction
PCa is the most diagnosed cancer in men. In the initial stage, the growth of PCa depends on
androgens and the androgen receptor (AR) signaling [1, 2]. Therefore, blockage of androgen
signaling is the central in hormone therapy. This includes the androgen deprivation therapy, which
is combined with androgen antagonist (antiandrogen) treatment leading to the pharmacological
reduction of androgen levels and blockage of the AR-signaling. At the beginning of the treatment,
the lack of androgens leads to tumor regression. However, eventually during hormone therapy, PCa
develops into a castration-resistant stage but remains AR-dependent [3, 4]. Inactivation of the AR
also in castration-resistant PCa seems to be therefore a key target of treatment. Unfortunately, a
therapy resistance may develop, which can be caused by mutations in exons coding for the AR
ligand-binding domain (LBD) that can ultimately switch the action of antagonists into agonists. An
example is the AR mutant T877A at which hydroxyflutamide (OH-F), a therapeutically used
antiandrogen with strong AR inhibitory effect, acts as potent activator [5]. The AR T877A has been
identified in PCa tumor samples from patients treated with OH-F. Other occurring AR mutants,
such as AR-W741C or AR-F876L mediate resistance to bicalutamide and the second-generation
antiandrogen enzalutamide, respectively [6, 7]. Therefore, identifying distinct mechanisms
inhibiting the AR and AR mutants would be beneficial and interesting to obtain insights into
structural requirements for AR antagonism. Thus, there is a need for new chemical platforms acting
as AR antagonists.
The AR belongs to the subfamily of steroid receptors, which includes the glucocorticoid (GR) and
progesterone receptors (PR). They all are members of the large nuclear receptor superfamily of
ligand-inducible transcription factors. The AR consists of a carboxy- (C-) terminal LBD, a central
DNA binding domain and the amino (N-)-terminus harboring the major transactivation function.
Ligand binding induces the N/C terminal interaction of the AR. AR antagonists, so far analyzed, are
able to inhibit this interaction [8, 9]. The androgen-activated AR rapidly translocates into the
4

nucleus and binds to chromatin and to DNA at hormone response elements to modulate gene
expression [2]. Examples of direct AR target genes are FKBP5 and the prostate-specific antigen
(PSA/KLK3), which is an important diagnostic marker for PCa. AR may also repress some target
genes such as maspin (SERPINB5) [10].
In our previous works, we identified and characterized novel steroidal AR antagonists based on the
chemical platform of aminosteroids and dehydroepiandrosterones [11, 12] and also identified the
first natural occurring AR antagonist atraric acid [13]. In the present work, we focused on the
structural lead compound methyl anthranilate (MA) identified previously as an AR antagonist by
analyzing saw palmetto extracts [15] and used here as a basis for a novel class of small molecular
derivatives as selective AR antagonists. We analyzed in total 36 structural variants of MA, of which
23 compounds were synthesized. We carried out quantitative structure-activity relationships of AR
antagonism in cellular assay systems using the androgen-dependent human PCa cell line LNCaP.
Since clinically used AR antagonists are halogenated, effective halogen-substituted anthranilic acid
esters were investigated in more detail for their mechanistic effects. Here, we identified a novel
chemical platform that acts as a lead structure for new AR antagonists. This new chemical platform
might be used in future for further optimization.
Therefore, we suggest that these novel antiandrogenic compounds may serve as a basis and
chemical platforms for further analyses and designs in the context of structure-function relationship
between ligands and the AR, also in terms of enhancing the diversity of AR antagonistic compounds
that may be helpful for treating PCa in future.
2. Materials and methods
2.1. General methods for chemistry
The majority of synthesis were carried with modifications based on the reaction of isatoic anhydride
with the respective alcohol as found by Georg Schmidt in 1887 and modified later by Staiger &
Miller [14].
Melting points were determined with a heat table microscope HM-LUX (Leitz, Wetzlar, Germany)
and are corrected. The NMR spectra were recorded on a JEOL ECA-500 spectrometer (500 or 400
5

1
MHz for H and 125 or 100 MHz for 13C) and referenced to the solvent resonances in CDCl3 or
acetone-d6 respectively. Chemical shifts are reported in ppm and are assigned as singlets (s),
doublets (d), triplets (t), quartets (q), sextets, and multiplets (m); br abbreviates broad signal. The
1
2
3
coupling constants (J) are reported as J (direct coupling), J (germinal coupling), J (vicinal
4
coupling) or J (long-range coupling) in Hertz (Hz). The atom numberings of synthesized
compounds are shown in supplementary Fig. 1. EI-MS and HR-EI-MS spectra were measured with
a Micromass VG 7070H and a Micromass AutoSpec mass spectrometer using an ionizing energy of
70 eV.
Compound 18 and 19 were purified by preparative HPLC on a C18 reversed-phase column
(Nucleodur C18 Gravity, 5 m, ID 21 mm, L 250 mm; Macherey-Nagel, Düren, Germany) and
isocratic elution with acetonitrile/water (70:30, v/v for compound 18; 90:10, v/v for compound 19)
over 15 min at a flow rate of 22 mL/min. A Waters 990 diode-array detector was used for detection
at = 254 nm.
The following compounds were purchased commercially: Methyl anthranilate (MA, 1), methyl 2-
pyrrol-1-ylbenzoate (22), methyl 2-amino-5-chloro-3-iodobenzoate (28), methyl 2-amino-4,5-
methoxybenzoate (29), methyl-2-hydroxybenzoate (37) and methyl 2-(dimethylamino)benzoate (20)
were purchased from Alfa Aesar (Ward Hill, MA, USA), methyl 4-aminopyridine-3-carboxylate
(33) from Fluorochem (Derbyshire, UK). Phenyl 2-aminobenzoate (15), 2-amino-N-
methylbenzamide (24) and methyl 2-aminothiophene-3-carboxylate (32), methyl 3-aminothiophene-
2-carboxylate (31) and methyl 2-amino-5-chlorobenzoate (26) were purchased from ASDI Inc.
(Newark, DE, USA). Methyl 2-amino-4-chlorobenzoate (27) and methyl 2-acetamidobenzoate (23)
were purchased from Acros Organics (Thermo Fisher Scientific, Geel, Belgium), methyl 2-
morpholino-4-yl-benzoate (21) from Maybridge (Cambridge, UK). The purity of each compound
was ≥98 %.
2.2. Chemical synthesis
2.2.1. Ethyl-2-aminobenzoate (2)
6

Isatoic anhydride (4.08 g, 25 mmol) was dissolved in 40 mL ethanol and 0.05 g NaOH (1.25 mmol)
was added. The mixture was heated at 65 °C until the gas generation had ceased, subsequently
cooled and poured in 120 mL water. The product settled out as immiscible oil on the bottom. The
oil was separated and distilled under vacuum over a 20 cm Vigreux column. The pure product, 2.99
g of a colorless oil (yield 72 %) passed over at 111 °C (5 Torr). ¹H NMR (500 MHz, CDCl3) δ 1.38
(t, 3 H, 3J= 7.2 Hz, 2’-H); 4.33 (q, 2H, 3J= 7.2 Hz, 1’-H); 5.72 (br s, 2H, NH); 6.64 (m, 2H, 3-H, 5-
3
H); 7.25 (m, 1H, 4-H); 7.88 (d, 1H, J= 8.3 Hz, 6-H); ¹³C NMR (100 MHz, CDCl3) δ 14.4 (C2’);
60.4 (C1’); 111.2 (C1); 116.3 (C3); 116.8 (C5); 131.3 (C6); 134.1 (C4); 150.6 (C2); 168.3 (C7); EI-
MS (70eV): m/z (rel. int.) = 165 [M+] (78), 137 (12), 119 (100), 92 (28), 65 (9); HR-EI-MS: m/z =
165.0776; C9H11NO2 [M+] requires 165.0790.
2.2.2. Propyl-2-aminobenzoate (3)
Isatoic anhydride (4.08 g, 25 mmol) was dissolved in 40 mL 1-propanol and 0.05 g NaOH (1.25
mmol) was added. The mixture was heated at 65 °C until the gas generation had ceased,
subsequently cooled and poured in 120 mL water. The product settled out as immiscible oil on the
bottom. The oil was separated and distilled under vacuum over a 20 cm Vigreux column. The pure
product, 3.28 g of a clear, slightly yellowish oil (yield 73 %) passed over at 130 °C (5 Torr). ¹H
3
NMR (500 MHz, CDCl3) δ 1.03 (t, 6 H, J= 7.4 Hz, 3’-H); 1.78 (sextet, 2H, 2’-H); 4.23 (t, 2H, 3J=
6.7 Hz, 1’-H); 5.73 (br s, 2H, NH); 6.64 (m, 2H, 3-H, 5-H); 7.25 (m, 1H, 4-H); 7.87 (d, 1H, 3J= 8.0
Hz, 6-H); ¹³C NMR (125 MHz, CDCl3) δ 10.5 (C3’); 22.1 (C2’); 65.8 (C1’); 111.0 (C1); 116.3
(C3); 116.6 (C5); 131.1 (C6); 133.9 (C4); 150.4 (C2); 168.2 (C7); EI-MS (70eV): m/z (rel. int.) =
179 [M+] (61), 165 (29), 137 (25), 119 (100), 92 (24); HR-EI-MS: m/z = 179.0942; C10H13NO2
[M+] requires 179.0945.
2.2.3. Butyl-2-aminobenzoate (4)
Isatoic anhydride (8.16 g, 50 mmol) was dissolved in 70 mL 1-butanol and 0.1 g NaOH (2.5 mmol)
was added. The mixture was heated at 85 °C until the gas generation had ceased, subsequently
cooled and poured in 120 mL water. The product settled out as immiscible oil on the bottom
7

together with remaining 1-butanol. The oil was separated and distilled under vacuum over a 20 cm
Vigreux column. The pure product, 7.54 g of a clear, slightly yellowish oil (yield 78 %) passed over
at 137.5 °C (5.5 Torr). ¹H NMR (500 MHz, CDCl3) δ 0.97 (t, 3H, 3J= 7.4 Hz, 4’-H); 1.47 (sextet, 2
H, 3’-H); 1.73 (quintet, 2H, 2’-H); 4.27 (t, 2H, 3J= 6.6 Hz, 1’-H); 5.72 (br s, 2H, NH); 6.64 (m, 2H,
3
3-H, 5-H); 7.25 (dt, 1H, 3J= 7.7 Hz, 4J= 1.6 Hz, 4-H); 7.86 (dd, 1H, J= 7.9 Hz, 4J= 1.6 Hz, 6-H);
¹³C NMR (125 MHz, CDCl3) δ 13.8 (C4’); 19.4 (C3’); 30.9 (C2’); 64.3 (C1’); 111.2 (C1); 116.3
(C3); 116.8 (C5); 131.3 (C6); 134.0 (C4); 150.6 (C2); 168.2 (C7); EI-MS (70eV): m/z (rel. int.) =
193 [M+] (100), 137 (66), 119 (77), 92 (35), 65 (9); HR-EI-MS: m/z = 193.1103; C11H15NO2 [M+]
requires 193.1103.
2.2.4. Pentyl-2-aminobenzoate (5)
Isatoic anhydride (8.16 g, 50 mmol) was dissolved in 40 mL 1-pentanol and 0.1 g NaOH (2.5 mmol)
was added. The mixture was heated at 85 °C until the gas generation had ceased, subsequently
cooled and poured in 120 mL water. The product settled out as immiscible oil on the surface. The
oil was separated and distilled under vacuum over a 20 cm Vigreux column. The pure product, 6.26
g of a colorless oil (yield 60 %) passed over at 147.5 °C (5 Torr). ¹H NMR (500 MHz, CDCl3) δ
3
3
0.93 (t, 3H, J= 7.0 Hz, 5’-H); 1.40 (m, 4 H, 3’-H, 4’-H); 1.76 (quintet, 2H, 2’-H); 4.26 (t, 2H, J=
6.7 Hz, 1’-H); 5.72 (br s, 2H, NH); 6.64 (m, 2H, 3-H, 5-H); 7.25 (m, 1H, 4-H); 7.87 (d, 1H, 3J= 8.3
Hz, 6-H); ¹³C NMR (125 MHz, CDCl3) δ 14.1 (C5’); 22.5 (C4’); 28.3 (C3’); 28.6 (C2’); 64.6 (C1’);
111.2 (C1); 116.3 (C3); 116.7 (C5); 131.3 (C6); 134.0 (C4); 150.6 (C2); 168.3 (C7); EI-MS (70eV):
m/z (rel. int.) = 207 [M+] (100), 193 (6), 137 (66), 119 (75), 92 (21); HR-EI-MS: m/z = 207.1257;
C12H17NO2 [M+] requires 207.1259.
2.2.5. Propan-2-yl-2-aminobenzoate (6)
Isatoic anhydride (4.08 g, 25 mmol) was dissolved in 40 mL 2-propanol and 0.05 g NaOH (1.25
mmol) was added. The mixture was heated at 80 °C until the gas generation had ceased,
subsequently cooled and poured in 120 mL water. The product settled out as immiscible oil on the
bottom. The oil was separated and distilled under vacuum over a 20 cm Vigreux column. The pure
8

product, 2.45 g of a colorless oil (yield 55 %) passed over at 111 °C (0.7 Torr). ¹H NMR (500 MHz,
3
3
CDCl3) δ 1.35 (t, 6H, J= 6.2 Hz, 1’-H, 3’-H); 5.21 (septet, 1H, J= 6.2 Hz, 2’-H); 5.72 (br s, 2H,
3
NH); 6.64 (m, 2H, 3-H, 5-H); 7.25 (m, 1H, 4-H); 7.87 (d, 1H, J=8.2 Hz, 6-H); ¹³C NMR (125
MHz, CDCl3) δ 22.1 (C1’, C3’); 67.7 (C2’); 111.6 (C1); 116.3 (C3); 116.7 (C5); 131.2 (C6); 133.8
(C4); 150.5 (C2); 167.8 (C7); EI-MS (70eV): m/z (rel. int.) = 179 [M+] (100), 137 (78), 119 (86),
92 (36), 65 (11); HR-EI-MS: m/z = 179.0942; C10H13NO2 [M+] requires 179.0946.
2.2.6. Butan-2-yl-2-aminobenzoate (7)
Isatoic anhydride (4.08 g, 25 mmol) was dissolved in 40 mL 2-butanol and 0.05 g NaOH (1.25
mmol) was added. The mixture was heated at 100 °C until the gas generation had ceased,
subsequently cooled and poured in 120 mL water. The product settled out as immiscible oil on the
bottom. The oil was separated and distilled under vacuum over a 20 cm Vigreux column. The pure
product, 2.53 g of a colorless oil (yield 52 %) passed over at 126 °C (5 Torr). ¹H NMR (500 MHz,
3
3
CDCl3) δ 0.97 (t, 3H, J= 7.5 Hz, 4’-H); 1.32 (d, 3H, J= 6.2 Hz, 1’-H); 1.61-1.79 (m, 2H, 3’-H);
5.06 (sextet, 1H, 3J= 6.2 Hz, 2’-H); 5.72 (br s, 2H, NH); 6.64 (t, 2H, 3J= 7.8 Hz, 3-H, 5-H); 7.25 (dt,
1H, 3J= 7.4 Hz, 4J= 1.9 Hz, 4-H); 7.87 (d, 1H, 3J= 8.0 Hz, 6-H); ¹³C NMR (125 MHz, CDCl3) δ 9.7
(C4’); 19.6 (C1’); 28.9 (C3’); 72.1 (C2’); 111.5 (C1); 116.1 (C3); 116.6 (C5); 131.2 (C6); 133.8
(C4); 150.4 (C2); 167.8 (C7); EI-MS (70eV): m/z (rel. int.) = 193 [M+] (100), 179 (13), 137 (87),
119 (86); 92 (29); HR-EI-MS: m/z = 193.1108; C11H15NO2 [M+] requires 193.1103.
2.2.7. [(2S)-Butan-2-yl]-2-aminobenzoate (8)
To a mixture of isatoic anhydride (1.59 g, 9.8 mmol) and 2-(s)-butanol (0.73 g, 9.8 mmol) dissolved
in 10 mL 1,4-dioxane, NaOH was added (0.02 g, 0.5 mmol). The mixture was heated until moderate
evolution of CO2 gas occurred at approximately 90 °C. This temperature was maintained until gas
evolution had ceased. After cooling, the reaction mixture was poured into 30 mL water, whereas a
pale white precipitate was formed, which was filtered off and dried under vacuum. After drying it
remained 1.21 g (yield 64 %) of a pale yellow oil as product. ¹H NMR (400 MHz, CDCl3) δ 0.97 (t,
3H, 3J= 7.5 Hz, 4-H); 1.32 (d, 3H, 3J= 6.2 Hz, 1’-H); 1.55-1.78 (m, 2H, 3’-H); 5.05 (sextet, 1H, 3J=
9

6.2 Hz, 2’-H); 5.72 (br s, 2H, NH); 6.64 (m, 2H, 3-H, 5-H); 7.25 (dt, 1H, 3J= 7.4 Hz, 4J= 1.6 Hz, 4-
H); 7.87 (dd, 1H, 3J= 8.0 Hz, 4J= 1.6 Hz, 6-H); ¹³C NMR (100 MHz, CDCl3) δ 9.9 (C4’); 19.7 (C1’);
29.1 (C3’); 72.2 (C2’); 111.6 (C1); 116.3 (C3); 116.7 (C-5); 131.3 (C6); 133.9 (C4); 150.5 (C2);
167.9 (C7); EI-MS (70eV): m/z (rel. int.) = 207 [M+] (100), 151 (24), 137 (99), 92 (54); HR-EI-
MS: m/z = 193.1088; C11H15NO2 [M+] requires 193.1103.
2.2.8. 3-Methylbutan-2-yl-2-aminobenzoate (9)
To a mixture of isatoic anhydride (4.08 g, 25 mmol) and DL-3-methyl-2-butanol (2.20 g, 25 mmol)
dissolved in 40 mL 1,4-dioxane, NaOH was added (0.05 g, 1.25 mmol). The mixture was heated at
100 °C under reflux for 12 h and subsequently poured into 120 mL water, where the product settled
as immiscible oil. The oil was separated and distilled over a 20 cm Vigreux column, whereas the
product, 4.01 g of a colorless oil (yield 77 %) passed over at 108-109 °C (0.8 Torr). ¹H NMR (400
MHz, CDCl3) δ 0.99 (t, 6H, 3J= 6.7 Hz, 4’-H, 5’-H); 1.27 (d, 3H, 3J= 6.4 Hz, 1’-H); 1.91 (m, 1H, 3’-
H); 4.95 (quintet, 1H, 3J= 6.4 Hz, 2’-H); 5.73 (br s, 2H, NH); 6.64 (m, 2H, 3-H, 5-H); 7.25 (m, 1H,
3
4-H); 7.87 (d, 1H, J= 8.0 Hz, 6-H); ¹³C NMR (125 MHz, CDCl3) δ 16.9 (C4’); 18.1 (C5'); 18.3
(C1’); 33.0 (C3’); 75.1 (C2’); 111.6 (C1); 116.3 (C3); 116.8 (C5); 131.2 (C6); 133.9 (C4); 150.6
(C2); 167.8 (C7); EI-MS (70eV): m/z (rel. int.) = 207 [M+] (100), 151 (24), 137 (99), 92 (54); HR-
EI-MS: m/z = 207.1250; C12H17NO2 [M+] requires 207.1259.
2.2.9. 2-Methoxyethyl-2-aminobenzoate (10)
Isatoic anhydride (8.16 g, 50 mmol) was dissolved in 60 mL 1,4-dioxane and 4.96 mL 1-methoxy-2-
ethanol (50 mmol) and 0.1 g NaOH (2.5 mmol) were added. The mixture was heated at 110 °C until
the gas generation had ceased, subsequently cooled and poured in 150 mL water. The product
settled out as immiscible oil on the bottom. The oil was separated and distilled under vacuum over a
20 cm Vigreux column. The pure product, 6.29 g of a clear, slightly yellowish oil (yield 64 %)
3
passed over at 136 °C (1 Torr). ¹H NMR (500 MHz, CDCl3) δ 3.40 (s, 3H, 3’-H); 3.69 (t, 2H, J=
4.8 Hz, 2’-H); 4.41 (t, 2H, 3J= 4.8 Hz, 1’-H); 5.71 (br s, 2H, NH); 6.61 (m, 2H, 3-H, 5-H); 7.23 (m,
1H, 4-H); 7.88 (d, 1H, 3J= 8.0 Hz, 6-H); ¹³C NMR (125 MHz, CDCl3) δ 58.9 (C3’); 63.2 (C2’); 70.5
10

(C1’); 110.5 (C1); 116.0 (C3); 116.5 (C5); 131.2 (C6); 134.0 (C4); 150.4 (C2); 167.9 (C7); EI-MS
(70eV): m/z (rel. int.) = 195 [M+] (100), 137 (63), 120 (78), 92 (47); 59 (21); HR-EI-MS: m/z =
195.0898; C10H13NO3 [M+] requires 195.0895.
2.2.10. 3-Methoxypropyl-2-aminobenzoate (11)
Isatoic anhydride (1.40 g, 8.6 mmol) was dissolved in 40 mL 1,4-dioxane and 0.772 g 3-methoxy-1-
propanol (8.6 mmol) and 0.05 g NaOH (1.25 mmol) were added. The mixture was heated at 90 °C
until the gas generation had ceased, subsequently cooled and poured in 120 mL water. The product
settled out as immiscible oil on the bottom. The oil was separated and distilled under vacuum over a
20 cm Vigreux column. The pure product, 0.52 g of a yellowish oil (yield 29 %) passed over at
169 °C (2 Torr). ¹H NMR (500 MHz, CDCl3) δ 2.02 (quintet, 2H, 2’-H); 3.35 (s, 3H, 4’-H); 3.53 (t,
2H, 3J= 6.2 Hz, 3’-H); 4.36 (t, 2H, 3J= 6.4 Hz, 1’-H); 5.71 (br s, 2H, NH); 6.64 (m, 2H, 3-H, 5-H);
7.25 (dt, 1H, 3J= 7.6 Hz, 4J= 1.6 Hz, 4-H); 7.85 (dd, 1H, 3J= 8.0 Hz, 4J= 1.6 Hz, 6-H); ¹³C NMR (125
MHz, CDCl3) δ 29.1 (C4’); 58.7 (C3’); 61.4 (C2’); 69.3 (C1’); 110.9 (C1); 116.2 (C3); 116.7 (C5);
131.2 (C6); 134.0 (C4); 150.5 (C2); 168.0 (C7); EI-MS (70eV): m/z (rel. int.) = 209 [M+] (100),
137 (65), 119 (72), 92 (32); 72 (26); HR-EI-MS: m/z = 209.1054; C11H15NO3 [M+] requires
209.1052.
2.2.11. 1-Methoxypropan-2-yl-2-aminobenzoate (12)
Isatoic anhydride (8.16 g, 50 mmol) was dissolved in 60 mL 1,4-dioxane and 4.65 mL 1-methoxy-2-
propanol (50 mmol) and 0.05 g NaOH (1.25 mmol) were added. The mixture was heated at 110 °C
until the gas generation had ceased, subsequently cooled and poured in 150 mL water. The product
settled out as immiscible oil on the bottom. The oil was separated and distilled under vacuum over a
20 cm Vigreux column. The pure product, 6.48 g of a clear, slightly yellowish oil (yield 62 %)
passed over at 124-125 °C (0.4 Torr). ¹H NMR (500 MHz, CDCl3) δ 1.34 (d, 3H, 3J= 6.4 Hz, 1’-H);
3.39 (s, 3H, 4’-H); 3.48-3.60 (m, 2H, 3’-H); 5.29 (m, 1H, 2’-H); 5.71 (br s, 2H, NH); 6.63 (m, 2H,
3
3-H, 5-H); 7.24 (m, 1H, 4-H); 7.87 (dd, 1H, J= 8.0 Hz, 4J= 1.5 Hz, 6-H); ¹³C NMR (125 MHz,
CDCl3) δ 16.8 (C1’); 59.2 (C4’); 69.2 (C2’); 75.2 (C3’); 111.1 (C1); 116.1 (C3); 116.3 (C5); 131.3
11

(C6); 134.0 (C4); 150.5 (C2); 167.5 (C7); EI-MS (70eV): m/z (rel. int.) = 209 [M+] (100), 137 (76),
120 (77), 92 (26); 73 (32); HR-EI-MS: m/z = 209.1044; C11H15NO3 [M+] requires 209.1052.
2.2.12. Cyclohexyl 2-aminobenzoate (13)
To a mixture of isatoic anhydride (5.00 g, 31 mmol) and cyclohexanol (6.21 g, 62 mmol) dissolved
in 40 mL 1,4-dioxane, NaOH (0.10 g, 2.5 mmol) was added. The mixture was heated under reflux
for 12 h, subsequently cooled and poured in 120 mL water. The remaining solution was three times
extracted with 100 mL diethyl ether and the combined organic phases were dried over Na2SO4.
After removal of the diethyl ether, a reddish oil remained which was distilled under vacuum over a
20 cm Vigreux column. The pure product, 4.84 g of a pale yellow oil (yield 71 %) passed over at
145 °C (0.38 Torr). ¹H NMR (400 MHz, CDCl3) δ 1.20-2.00 (m, 10 H, 2’-H, 3’-H, 4’-H, 5’-H, 6’-
3
H); 5.00 (septet, 1H, 1’-H); 6.04 (br s, NH); 6.64 (m, 2H, 3-H, 5-H); 7.25 (dt, 1H, J= 7.6 Hz, 4J=
1.6 Hz, 4-H); 7.89 (dd, 1H, 3J= 8.2 Hz, 4J= 1.6 Hz, 6-H); ¹³C NMR (100 MHz, CDCl3) δ 23.8 (C3’,
C5’); 25.6 (C4’); 31.8 (C2’, C6’); 72.4 (C1’); 111.7 (C1); 116.3 (C3); 116.8 (C5); 131.4 (C6); 134.0
(C4); 150.5 (C2); 167.1 (C7); EI-MS (70eV): m/z (rel. int.) = 219 [M+] (62), 137 (100), 119 (78),
92 (10); HR-EI-MS: m/z = 219.1245; C13H17NO2 [M+] requires 219.1259.
2.2.13. Cyclohexylmethyl 2-aminobenzoate (14)
To a mixture of isatoic anhydride (5.00 g, 31 mmol) and cyclohexyl methanol (7.08 g, 62 mmol)
dissolved in 35 mL 1,4-dioxane, NaOH was added (0.10 g, 2.5 mmol). The mixture was heated at
100 ºC under reflux for 12 h and subsequently poured into 120 mL water. The remaining solution
was extracted three times with 100 mL diethyl ether and the combined organic phases were dried
over Na2SO4. After evaporation of the diethyl ether, the remaining reddish oil was distilled under
vacuum over a 20 cm Vigreux column. The product passed over at 152 °C (0.15 Torr). At room
temperature, the product crystalized to give 6.42 g (yield 89 %) of a white powder. mp: 42-43 °C;
¹H NMR (400 MHz, CDCl3) δ 1.06 (m, 2H, 4’-Hax, 6’-Hax); 1.24 (m, 3H, 4’-Heq, 5’-Hax, 6’-Hax);
3
1.58-1.99 (m, 6H, 2’-Heq, 3’-H, 5’-Heq, 7’-H); 4.08 (d, 2H, J= 6.4 Hz, 1’-H); 5.72 (br s, 2H, NH);
6.64 (m, 2H, 3-H, 5-H); 7.25 (dt, 1H, 3J= 7.6 Hz, 4J= 1.6 Hz, 4-H); 7.88 (dd, 1H, 3J= 8.2 Hz, 4J= 1.6
12

Hz, 6-H); ¹³C NMR (100 MHz, CDCl3) δ 25.9 (C4’, C6’); 26.5 (C5’); 29.9 (C3’, C7’); 37.4 (C2’);
69.5 (C1’); 111.1 (C1); 116.3 (C3); 116.8 (C5); 131.3 (C6); 134.1 (C4); 150.7 (C2); 168.3 (C7); EI-
MS (70eV): m/z (rel. int.) = 233 [M+] (100), 137 (73), 119 (59), 92 (14); HR-EI-MS: m/z =
233.1415; C14H19NO2 [M+] requires 233.1416.
2.2.14. [(E)-3-Phenylprop-2-enyl]-2-aminobenzoate (16)
Isatoic anhydride (9.00 g, 55 mmol) was dissolved in 70 mL 1,4-dioxane and 7.38 g cinnamyl
alcohol (55 mmol) and 0.11 g NaOH (2.75 mmol) were added. The mixture was heated at 90 °C for
6 h, cooled and subsequently poured into 200 mL water. The product precipitated, was filtered off
and recrystallized in methanol to give the pure product, 9.31 g of a pale yellow powder (yield 67
%). mp: 61 °C; ¹H NMR (500 MHz, CDCl3) δ 4.94 (dd, 2H, 3J= 6.3 Hz, 4J= 1.4 Hz, 1’-H); 5.73 (br
s, 2H, NH); 6.42 (m, 1 H, 2’-H); 6.65 (m, 2H, 3-H, 5-H); 6.73 (d, 1H, 3J= 16.1 Hz, 3’-H); 7.24-7.44
(m, 6H, 4-H, 2’’-H, 3’’-H, 4’’-H, 5’’-H, 6’’-H); 7.92 (d, 1H, 3J= 8.2 Hz, 6-H); ¹³C NMR (125 MHz,
CDCl3) δ 64.9 (C1’); 110.7 (C1); 116.3 (C3); 116.7 (C5); 123.6 (C1’’); 126.6 (C2’’, C6’’); 128.0
(C2’); 128.6 (C3’’, C5’’); 131.3 (C6); 133.9 (C4’’); 134.2 (C4); 136.3 (C3’); 150.6 (C2); 167.8
(C7); EI-MS (70eV): m/z (rel. int.) = 253 [M+] (63), 117 (100), 115 (16), 91 (8); HR-EI-MS: m/z =
253.1071; C16H15NO2 [M+] requires 253.1103.
2.2.15. Methyl-2-methylaminobenzoate (17)
N-methylisatoic anhydride (5.0 g, 28 mmol) was dissolved in 40 mL methanol and 0.1 g NaOH (2.5
mmol) was added. The mixture was heated at 60 °C until the gas generation had ceased,
subsequently cooled and poured in 120 mL water. The product settled out as immiscible oil on the
bottom. The oil was separated and distilled under vacuum over a 20 cm Vigreux column. The pure
product, 3.10 g of a clear, slightly yellowish oil (yield 67 %) passed over at 94 °C (1 Torr). ¹H NMR
3
(500 MHz, CDCl3) δ 2.91 (s, 3 H, 1’’-H); 3.85 (s, 3H, 1’-H); 6.58 (m, 1H, 5-H); 6.66 (d, 1H, J=
3
3
8.5 Hz, 3-H); 7.38 (dt, 1H, J= 7.6 Hz, 4J= 1.6 Hz, 4-H); 7.63 (br s, 1H, NH); 7.89 (dd, 1H, J= 8.0
Hz, 4J= 1.6 Hz, 6-H); ¹³C NMR (100 MHz, CDCl3) δ 29.6 (C1’’); 51.5 (C1’); 109.9 (C1); 110.8
(C3); 114.4 (C5); 131.6 (C6); 134.7 (C4); 152.1 (C2); 169.2 (C7); EI-MS (70eV): m/z (rel. int.) =
13

165 [M+] (100), 132 (28), 119 (17), 105 (46), 40 (32); HR-EI-MS: m/z = 165.0792; C9H11NO2
[M+] requires 165.0790.
2.2.16. Methyl-2-ethylaminobenzoate (18)
3.5 g NaOH (88 mmol) were suspended in 75 mL dimethylformamide (DMF) and a solution of 9.6
g isatoic anhydride (59 mmol) solved in 35 mL DMF was added at 30 °C dropwise over 30 min.
Subsequently 7.2 mL ethyl iodide (88 mmol) were added dropwise over 20 min. The temperature
never exceeded 44 °C. The mixture was stirred for 10 min and 50 mL methanol were added and
subsequently stirred at 25 °C for 15 min. The methanol was evaporated and 200 mL water were
added. The mixture was then extracted with hexane and washed several times with water. The
hexane extract was evaporated and the remaining oil, distilled and subsequently purified by
preparative HPLC to give the pure product, 1.78 g of a yellow oil (yield 17 %). ¹H NMR (500
MHz, CDCl3) δ 1.31 (t, 3H, 3J= 7.1 Hz, 2’’-H); 3,23 (m, 2 H, 3J= 7.1 Hz, 1’’-H); 3.84 (s, 3H, 1’-H);
3
6.56 (m, 1H, 5-H); 6.68 (d, 1H, J= 8.5 Hz, 3-H); 7.34 (m, 1H, 4-H); 7.59 (br s, 1H, NH); 7.89 (d,
3
1H, J= 8.0 Hz, 6-H); ¹³C NMR (125 MHz, CDCl3) δ 14.5 (C2’’); 37.4 (C1’’); 51.4 (C1’); 109.6
(C1); 111.1 (C3); 114.2 (C5); 131.6 (C6); 134.6 (C4); 151.2 (C2); 169.1 (C7); EI-MS (70eV): m/z
(rel. int.) = 179 [M+] (100), 164 (70), 146 (47), 132 (95), 119 (15); HR-EI-MS: m/z = 179.0956;
C10H13NO2 [M+] requires 179.0946.
2.2.17. Methyl-2-propylaminobenzoate (19)
3.5 g NaOH (88 mmol) were suspended in 75 mL dimethylformamide (DMF) and a solution of 9.6
g isatoic anhydride (59 mmol) solved in 35 mL DMF was added at 30 °C dropwise over 30 min.
Subsequently 8.05 mL 1-bromopropane (88 mmol) were added dropwise over 20 min. The
temperature never exceeded 44 °C. The mixture was stirred for 10 min and 50 mL methanol were
added and subsequently stirred at 25°C for 15 min. The methanol was evaporated and 200 mL water
were added. The mixture was then extracted with hexane and washed several times with water. The
hexane extract was evaporated and the remaining oil, distilled and subsequently purified by
preparative HPLC to give the pure product, 1.47 g of a yellow oil (yield 13 %). ¹H NMR (500
14

3
MHz, CDCl3) δ 1.03 (t, 3H, J= 7.3 Hz, 3’’-H); 1.71 (sextet, 2H, 3J= 7.3 Hz, 2’’-H); 3.16 (q, 2 H,
3
3J= 6.2 Hz, 1’’-H); 3.85 (s, 3H, 1’-H); 6.56 (m, 1H, 5-H); 6.68 (d, 1H, J= 8.5 Hz, 3-H); 7.34 (m,
3
1H, 4-H); 7.70 (br s, 1H, NH); 7.89 (d, 1H, J= 8.0 Hz, 6-H); ¹³C NMR (125 MHz, CDCl3) δ 11.7
(C3’’); 22.4 (C2’’); 44.6 (C1’’); 51.4 (C1’); 109.6 (C1); 111.1 (C3); 114.1 (C5); 131.6 (C6); 134.5
(C4); 151.3 (C2); 169.1 (C7); EI-MS (70eV): m/z (rel. int.) = 193 [M+] (91), 164 (85), 132 (100),
119 (27), 105 (11); HR-EI-MS: m/z = 193.1191; C11H15NO2 [M+] requires 193.1103.
2.2.18. 2-[(2-Aminobenzoyl)amino]benzoic acid (25)
Isatoic anhydride (2.5 g, 15 mmol) and anthranilic acid (2.3 g, 17 mmol) were suspended in 50 mL
water. The mixture was heated at 100 °C for 4 h. After cooling, crystals were formed which were
filtered off and washed with water. The product, pale yellow acicular crystals were purified by
recrystallization in ethanol/chloroform (50:50) to give 2.92 g of pure product (yield 76 %). mp: 201-
202 °C; ¹H NMR (400 MHz, acetone-D6) δ 2.82 (br s, 2H, NH); 6.62-7.76 (m, 6H, 3-H, 4-H, 5-H,
4
6-H, 4’-H, 5’-H); 8.15 (dd, 1H, 3J= 8.0 Hz, 4J= 1.6 Hz, 6’-H); 8.87 (dd, 1H, 3J= 8.6 Hz, J= 1.1 Hz,
3’-H); 12.04 (s, 1H, NH); ¹³C NMR (400 MHz, acetone-D6) δ 115.0 (C1); 115.6 (C2’); 117.4 (C3);
117.5 (C5); 120.0 (C6’); 122.2 (C4’); 127.5 (C3’); 131.5 (C6); 132.8 (C5’); 134.6 (C4); 142.6
(C1’); 151.1 (C2); 168.0 (C7); 170.0 (C7’); EI-MS (70eV): m/z (rel. int.) = 256 [M+] (100), 238
(16), 120 (82), 92 (17); HR-EI-MS: m/z = 256.0818; C14H12N2O3 [M+] requires 256.0848.
2.2.19. Methyl 2-amino-4-(trifluoromethyl)benzoate (30)
2-amino-4-(trifluoromethyl)benzoic acid (0.50 g, 2.44 mmol) was dissolved in 40 mL methanol and
8 mL HClc were added slowly. The mixture was stirred and heated under reflux for 12 h.
Subsequently, the methanol was evaporated and the remaining residue suspended in water. This
mixture was adjusted to an alkaline pH by adding a 10 % NaOH solution and extracted with diethyl
ether. The combined organic phases were dried over Na2SO4. After evaporation of the diethyl ether,
the remaining solid was recrystalized in isohexane to give 0.50 g (yield 93 %) of a yellowish
powder. mp: 62 °C; ¹H NMR (400 MHz, CDCl3) δ 3.89 (s, 3H, 1’-H); 5.90 (br s, 2H, NH); 6.84 (dd,
3
3
1H, J= 8.2 Hz, 4J= 1.6 Hz, 5-H); 6.89 (s, 1H, 3-H); 7.94 (dd, 1H, J= 8.2 Hz, 4J= 1.6 Hz, 6-H); ¹³C
15

3
3
NMR (100 MHz, CDCl3) δ 52.0 (C1’); 112.4 (q, J (C,F)= 3.8 Hz, C3); 113.1 (C1); 113.5 (q, J
(C,F)= 4.8 Hz, C5); 123.6 (q, 1J (C,F)= 273.2 Hz, C8); 132.2 (C6); 135.5 (q, 2J (C,F)= 31.6 Hz, C4);
150.3 (C2); 113.1 (C1); 167.8 (C7); EI-MS (70eV): m/z (rel. int.) = 219 [M+] (100), 200 (25), 187
(91), 160 (81); HR-EI-MS: m/z = 219.0507; C9H8F3NO2 [M+] requires 219.0494.
2.2.20. Butan-2-yl-2-amino-4-(trifluoromethyl)benzoate (30a)
2-amino-4-(trifluoromethyl)benzoic acid (0.50 g, 2.44 mmol) was dissolved in 40 mL sec-butanol
and 8 mL concentrated HCl were added slowly. The mixture was stirred and heated under reflux for
12 h. Subsequently, the sec-butanol was evaporated and the remaining residue suspended in water.
This mixture was adjusted to an alkaline pH by adding a 10 % NaOH solution and extracted with
diethyl ether. The combined organic phases were washed with water and subsequently dried over
Na2SO4. After evaporation of the diethyl ether, it remained 0.17 g (yield 27 %) of an orange colored
oil as product. ¹H NMR (400 MHz, CDCl3) δ 0.96 (t, 3H, 3J= 7.4 Hz, 4’-H); 1.33 (d, 3H, 3J= 6.2 Hz,
3
1’-H); 1.60-1.81 (m, 2H, 3’-H); 5.07 (sextet, 1H, J= 6.2 Hz, 2’-H); 5.91 (br s, 2H, NH); 6.83 (dd,
1H, 3J= 8.2 Hz, 4J= 1.2 Hz, 5-H); 6.89 (s, 1H, 3-H); 7.96 (d, 1H, 3J= 8.2 Hz, 6-H); ¹³C NMR (100
MHz, CDCl3) δ 9.8 (C4’); 19.6 (C1’); 29.0 (C3’); 72.9 (C2’); 112.3 (q, 3J (C,F)= 3.8 Hz, C3); 113.5
3
1
2
(q, J (C,F)= 3.8 Hz, C5); 113.9 (C1); 123.7 (q, J (C,F)= 273.1 Hz, C8); 132.2 (C6); 135.5 (q, J
(C,F)= 31.7 Hz, C4); 150.3 (C2); 167.8 (C7); EI-MS (70eV): m/z (rel. int.) = 261 [M+] (61), 239
(48), 205 (100), 187 (99); HR-EI-MS: m/z = 261.0977; C12H14F3NO2 [M+] requires 261.0947.
2.2.21. Methyl 3-aminonaphthalene-2-carboxylate (34)
3-Amino-2-naphthoic acid (1.00 g, 0.5 mmol) was slowly added to a mixture of 15 mL H2SO4c and
30 mL methanol and allowed to react for 5 h under reflux. During this time another mixture of
methanol/H2SO4c (35 mL/10 mL) was gradually added. After cooling, the reaction mixture was
stirred another 12 h at room temperature, subsequently evaporated and the remaining solution was
poured on ice. When neutralizing with saturated Na3CO3 solution, a yellow precipitate was formed
which was filtered off, solved in dichlormethane and washed several times with water. After drying
of the organic phase over Na2SO4, the solvent was removed and 0.62 g (yield 61 %) of pure product
16

remained. mp: 97-98 °C; ¹H NMR (400 MHz, CDCl3) δ 3.95 (s, 3H, 1’-H); 5.61 (br s, 2H, NH);
6.97 (s, 1H, 3-H); 7.18 (m, 1H, 6-H); 7.40 (m, 1H, 7-H); 7.52 (d, 1H, 3J= 8.2 Hz, 5-H); 7.71 (d, 1H,
3J= 8.5 Hz, 8-H), 8.49 (s, 1H, 10-H); ¹³C NMR (100 MHz, CDCl3) δ 52.0 (C1’); 110.1 (C3); 114.8
(C1); 122.6 (C7); 125.2 (C5); 126.1 (C9); 128.9 (C6); 129.3 (C8); 133.5 (C10); 137.4 (C4); 146.1
(C2); 168.4 (C11); EI-MS (70eV): m/z (rel. int.) = 201 [M+] (100), 169 (73), 142 (45), 115 (17);
HR-EI-MS: m/z = 201.0790; C12H11NO2 [M+] requires 201.0789.
2.2.22. Methyl-3-aminobenzoate (35)
3-Aminobenzoic acid (5.48 g, 40 mmol) was dissolved in 60 mL methanol and 4.36 mL thionyl
chloride (60 mmol) were added dropwise with stirring and cooling on ice. When the reaction had
stopped, the methanol was evaporated and the remaining residue was neutralized with saturated
NaHCO3 solution and extensively extracted with ethyl acetate. After evaporation of the ethyl
acetate, a reddish oil remained which was distilled under vacuum over a 20 cm Vigreux column.
The pure product passed over at 118 °C (0.3 Torr) and crystalized by cooling to give 5.46 g of a
white powder (yield 90 %). mp: 31-32 °C; ¹H NMR (500 MHz, CDCl3) δ 3.82 (br s, 2H, NH); 3.87
(s, 3H, 1’-H); 6.84 (d, 1H, 3J= 8.4 Hz, 4-H); 7.19 (m, 1H, 5-H); 7.34 (s, 1H, 2-H); 7.40 (d, 1H, 3J=
7.5 Hz, 6-H); ¹³C NMR (100 MHz, CDCl3) δ 51.9 (C1’); 115.6 (C2); 119.3 (C4); 119.5 (C6); 129.1
(C5); 131.0 (C1); 146.5 (C3); 167.2 (C7); EI-MS (70eV): m/z (rel. int.) = 151 [M+] (100), 120 (72),
92 (45), 65 (13).
2.2.23. Methyl-4-aminobenzoate (36)
4-Aminobenzoic acid (2.74 g, 20 mmol) was dissolved in 30 mL methanol and 2.18 mL thionyl
chloride (30 mmol) were added dropwise with stirring and cooling on ice. When the reaction had
stopped, the methanol was evaporated and the remaining residue was neutralized with saturated
NaHCO3 solution and extensively extracted with ethyl acetate. The organic phase was dried over
Na2SO4 and subsequently evaporated. The pure product remained, 2.60 g of slightly yellowish
acicular crystals (yield 86 %). mp: 107-108 °C ¹H NMR (500 MHz, CDCl3) δ 3.85 (s, 3H, 1’-H);
3
3
4.04 (br s, 2H, NH); 6.64 (d, 2H, J= 8.7 Hz, 3-H, 5-H); 7.85 (d, 2H, J= 8.7 Hz, 2-H, 6-H); ¹³C
17

NMR (100 MHz, CDCl3) δ 51.6 (C1’); 113.8 (C3, C5); 119.8 (C1); 131.6 (C2, C6); 150.8 (C4);
167.1 (C7); EI-MS (70eV): m/z (rel. int.) = 151 [M+] (92), 120 (100), 92 (23), 65 (10).
2.3. Cell culture and functional analysis of derivatives
Monkey kidney CV-1 and COS-7 cells lacking endogenous functional steroid receptors as well as
HeLa cells were maintained in Dulbecco´s modified Eagle´s medium (DMEM), supplemented with
10 % (v/v) fetal calf serum (FCS) (Invitrogen) and penicillin (100 IU/mL), and streptomycin (100
IU/mL) at 37°C and 5 % CO2. The androgend dependent PCa cell line LNCaP was cultured in
(RPMI), supplemented with 10 % (v/v) FCS, 1 % (v/v) sodium pyruvate and penicillin (100
IU/mL), and streptomycin (100 IU/mL) at 37°C and 5 % CO2. For growth assays 5 % FCS was
used. Hep3B cells stably expressing GFP-AR-wt or YFP-AR-CFP were cultured in α-MEM (Lonza)
supplemented with 5 % FCS, 2 mM L-glutamine, 100 U/mL penicillin, 100 µg/mL streptomycin
and 600 µg/mL G418.
The plasmid pMMTV-luc, which contains a luciferase reporter gene driven by the mouse mammary
tumor virus long terminal repeats that is responsive to androgens [15]. The expression vectors for
the human AR, pSG5-hAR, pSG-hAR T877A, W741C, GR and PR have been described earlier
[9][16]. The expression vector for AR-F876L was kindly provided by Dr. Charles Sawyers. For the
transfection experiments, CV-1 cells were seeded onto 6-well tissue culture plates (Greiner Bio-
One, Germany) at 105 cells per well, and grown in DMEM supplemented with 10 % (v/v) charcoal-
stripped serum [17]. Six hours later, cells were transfected using the CaPO4-method essentially as
described earlier [17]. The DNA mixture per well for transfections consisted of 1 µg of the MMTV-
luciferase reporter construct, 0.2 µg of the pSG5-hAR expression vector, and 0.2 µg of the
cytomegalovirus (CMV)-driven β-galactosidase expression vector (pCMV-lacZ), which was used as
an internal control for transfection efficiency. After 14 hours, media were replaced either with or
without the addition of androgens together with the indicated compounds. After additional 48 hours,
cells were lysed and assayed for luciferase and β-galactosidase activity. Obtained luciferase units
were normalized to those obtained with β-galactosidase. All transfection assays shown were
18

performed in duplicate and were repeated at least three times. Growth analyses were performed as
described previously [13].
2.4. Whole-cell binding assay
The competitive whole-cell binding assay using tritium-labeled mibolerone has been described
elsewhere,[13] and was used here with one modification: COS-7 cells were transfected using the
CaPO4-method with 0.75 µg of the pSG5-hARwt plasmid, together with 0.2 µg of the pCMV-lacZ
expression vector for normalization/control per well.
2.5. Real time RT-PCR
The reverse transcription quantitative real-time PCR (RT-qPCR) was performed essentially as
described previously with specific primers for detection of PSA, FKBP5 and maspin mRNA and β-
actin mRNA for normalization for the indicated cell via the ΔΔCt method using CFX manager
software from Bio-Rad.[13] As modifications 2 x 105 cells were seeded out directly in media
containing 10 % charcoal stripped FCS. After two days cells were treated with the indicated
compounds or solvent (DMSO) or androgen (R1881 100 pM). RNA was isolated from the cells
using peqGOLD TriFast (Peqlab, Erlangen, Germany) according to the manufacturer’s protocol.
One-step RT-qPCR was conducted using the SuperScript™ III One-Step RT-PCR System with
Platinum™ Taq DNA Polymerase kit (Invitrogen, Carlsbad, USA), gene specific primers and
Biorad CFX96TM Real Time PCR detection system. RT-qPCR results were analyzed via ΔΔCt
method using CFX manager software from Bio-Rad.
2.6. Detection of the SA-β-galactosidase
Senescence-associated β-galactosidase (SA-β-gal) staining was carried out as described
previously[18, 19] using 40,000 cells per well in a 6-well plate. The treatment was performed for
three days. The percentage of stained cells was determined following counting at least 400 cells per
well under a light microscope.
19

2.7. FACS, FRET and FRAP assays
FACS analyses were performed as described earlier.[16] Acceptor photobleaching FRET and FRAP
assays were performed as described earlier.[16]
2.8. Translocation studies
AR nuclear translocation studies were performed in general as described earlier.[20] As
modification the transfected cells were treated for 90 min with the first indicated compound and
additional 90 min with either androgen (R1881 100 pM) or DMSO as solvent control, resulting in
an overall incubation time of 180 min. Further for specific translocation analysis the protein
biosynthesis inhibitor cycloheximide (50 µg/mL) was added to all dishes at the beginning of the
treatment.
3. Results
3.2. Quantitative structure-activity relationship analysis.
Previously we identified MA (compound 1) to be a potent inhibitor of the AR [15]. In order to
further characterize this lead structure, we synthesized and analyzed derivatives of MA with
different structural modifications. To detect specific antiandrogenic potential we used the
established AR-dependent reporter gene assay using CV1 cells that lack endogenous AR as well as
the closely related GR and PR, which could potentially interfere with the assay system. The cells
were cotransfected with an expression vector encoding the AR-wt and the androgen-responsive
promoter-luciferase reporter (pMMTV-luc). Cells were treated with the indicated compounds either
alone to analyze their role in activating the AR, or together with the specific synthetic AR agonist
R1881 to analyze AR antagonism of the derivatives (Fig. 1; see the Supporting Information for
details). Treatment with DMSO alone was used as solvent control. Antagonistic activity of the
tested compounds resulted in reduced AR-mediated transactivation induced by androgen treatment.
Within this test system, a repression of the activating effect of R1881 is equivalent to the AR
inhibitory potential of the tested compound.
20

We observed that an elongation of the alkyl ester chain strongly increased the inhibitory effect upon
the androgen receptor (Fig. 1A). At a compound concentration of 300 µM, the addition of one
methylene group (-CH2-) to the ester chain increased the repression of the androgenic effect from
8.7-fold (compound 1) to about 35-fold (compound 2) and the addition of a propyl, butyl, or pentyl
group to more than 53-fold (compound 3, 4, 5).
Fig. 1. Schematic overview of tested structural changes in the lead structure methyl-anthranilate (MA,
compound 1). Modified derivatives (structural changes labeled in red) were tested for their inhibitory effect
upon the AR in an androgen receptor dependent MMTV-luc reporter gene assays using CV-1 cells lacking
endogenous AR as well as the closely related GR and PR, which could potentially interfere with the assay
system. The fold repression of compound concentrations of 300 µM (300) and 30 µM (30) were calculated by
correlating normalized relative light units (RLU) of treated and untreated cells (Supplemental Fig. 3) in the
presence or absence of the synthetic androgen R1881 (30 pM). Fold repression indicates the maximum
potential of each compound to repress the androgenic effect of R1881. (A) Unbranched alkyl esters. (B)
Branched alkyl esters. (C) Alkoxy esters. (D) Esters with ring structure. (E) Modifications of the amino group.
(F) Derivatives with amide group. (G) Halogen substitutions at the benzene ring. (H) Variations of the ring
system. (I) Shifting of the amino group and exchange to hydroxyl group. n.d.: not determined. *negative
inhibition values were set as 0x repression.
21

The introduction of brunched alkyl ester groups also led to a stronger repression of the androgenic
effect of R1881 (compound 6 and 8, Fig. 1B and Supplemental Fig. 2). At a concentration of 30
µM, compound 7 led to ~11-fold repression, which is about 10-fold stronger compared to the lead
structure 1, compound 9 even led to a 15.3-fold repression. However, at lower concentrations (30
µM) the antiandrogenic potential of compounds 2 to 6 converged to those of the lead structure (Fig.
1A and B).
Adding oxygen atoms to the ester chain, to receive alkoxy esters (Fig. 1C), led also to an increase of
the AR antagonism. At 300 µM, compounds 10 to 12 showed a strong repression of the R1881-
mediated AR activation, although, when compared to similar structures without oxygen atoms (Fig.
1A and B), the repression effect was decreased and nearly vanished using 30 µM of the compounds.
Flexible six-carbon ring structures in the ester group (Fig. 1D) increased the repression of the effect
of R1881 (compound 13 and 14) at 300 µM compared to the leading structure 1, but the additional
methylene group in compound 14 led to a sharp decline when compared to compound 13. The
addition of a rigid benzene ring structure, such as phenyl (compound 15) or cinnamoyl (compound
16, Supplemental Fig. 2) led to a strong decrease of repression being tantamount to the AR
inhibitory effect.
Changing the amino group in the leading structure 1 to a secondary amine as N-methyl derivative
(compound 17, Fig. 1E), seemed to increase the AR inhibitory potency slightly at 30 µM
concentration, however, an elongation of the alkyl chain to ethyl or propyl (compound 18 and 19,
Supplemental Fig. 2) lowered the effect. A change to a tertiary amine (compound 20) strongly
reduced the antiandrogenic effect, but integrating the amino group as tertiary amine into a ring
structure such as pyrrole (compound 22) or morpholine (compound 21, Supplemental Fig. 2) led to
increased AR inhibition.
The introduction of amido groups into the lead structure 1 (Fig. 1F) resulted in a decrease of AR
inhibition (compound 23 and 24). Moreover, adding a rigid space-consuming benzene ring with
substituted carboxy group resulted in a total loss of the AR inhibitory potential (compound 25).
22

Halogen substitution at the benzene ring (-Cl or -I) of the lead structure 1 (Fig. 1G) strongly
increased the AR inhibitory effect. Therefore, we focused on these derivatives since halogenated
compounds are also used as clinically used as antiandrogens, such as OH-F. By comparison of
mono chlorine substituted derivatives at concentrations of 300 and 30 µM (compound 26 and 27), it
became apparent that a halogen substitution in meta position to the carbonyl group is leading to a
stronger repression of the androgenic effect of R1881 than substitution at the para position. A
derivative with halogens at both meta positions to the carbonyl group showed a stronger effect
(compound 28) than the derivate with only one halogen at the meta position (compound 26).
Substitutions at the benzene ring with two methoxy groups or a trifluoromethyl group (compound
29 and 30, Supplemental Fig. 2) only slightly increased AR inhibition. Extending the hydrophobic
side chain, compound 30a, enhanced AR antagonism.
Exchanging the benzene ring in the lead structure 1 (Fig. 1H) to a thiophene (compound 32 and
compound 31, Supplemental Fig. 2) decreased the AR inhibition compared to the lead compound. A
change to a pyridine ring (compound 33) almost led to a loss of the AR inhibitory potential. Adding
an annulated saturated ring system to give a naphthalene derivative (compound 34), showed a strong
increase in repression of the R1881-mediated AR transactivation.
To investigate the influence of the amino group in the lead structure, derivatives with different
substitution position of the amino group and an exchange of the amino group to a hydroxyl group
were tested (Fig. 1I). Shifting the amino group from ortho in meta position (compound 35) resulted
in a total loss of the AR inhibitory effect, a shift to the para position (compound 36) led to a strong
decrease. Also, the replacement of the amino group with a hydroxyl group resulted in a total loss of
AR inhibitory potential (compound 37).
3.3. Halogen-substituted compounds inhibit the androgen-activated AR mutants that mediate
resistance to current clinically used AR antagonists.
Halogen-substituted anthranilic acid esters (compounds 26, 27, 28 (Fig. 1G) including compounds
30, 30a, Supplemental Fig. 2) revealed throughout and consistently high levels of AR antagonistic
activity as shown above. Therefore, we chose this group for further mechanistical investigation at
23

the molecular level in comparison to the lead structure compound 1. In addition, these compounds
contain some structural similarities to clinically applied antiandrogens such as hydroxyflutamide
(OH-F).
Even though the synthetic R1881 is preferentially used in cell culture experiments, since it is less
metabolized compared to the natural androgen dihydrotestosterone (DHT), it had to be clarified
whether the compounds can also inhibit the DHT-activated AR. This is important since the AR
conformational change is dependent on the bound ligand and might therefore alter possible
subsequent antiandrogenic effects. Similar results compared to the R1881-activated AR were
obtained in the presence of DHT (supplementary Fig. 4A). Further, we tested whether the halogen-
substituted anthranilic acid esters can also inhibit the AR-T877A mutant. This AR mutant is often
expressed in resistant PCa tumors [5] and confers resistance to OH-F treatment. Comparing the
inhibitory potential of the compounds towards the AR-wt and the AR-T877A suggests a distinct
pattern of inhibition. While compound 28 inhibits most potently the AR-wt, the compound 30a is
more efficiently in inhibiting AR-T877A (Fig. 2A). Interestingly this effect is independent of the
used AR-agonist, since we obtained similar results with the DHT-activated AR-T877A mutant
(supplementary Fig. 4B). The AR-F876L mutant confers resistance to the second generation AR
antagonist enzalutamide [6]. Analyzing this AR mutant in the same experimental setup, the
halogenated compounds inhibited the AR-F878L mediated transactivation, whereas enzalutamide
activated this receptor mutant using each 30 µM of compounds (Fig. 2B). Compound 28 appeared
to be the most effective inhibitor of AR-F876L. To include also bicalutamide resistance AR mutant
(AR-W741C) and to compare the activity of compound 28 with enzalutamide and bicalutamide with
AR mutants 10 µM of each compound was used (Fig. 2C and supplementary Fig. 5). The data
suggest that compound 28 inhibits not only wild-type but also the AR-mutants AR-T877A, AR-
F876L, and AR-W741C. This indicates that compound 28 exhibits a different molecular mechanism
of AR antagonism compared to enzalutamide and bicalutamide.
24

Fig. 2. Halogen-substituted anthranilic acid esters reduce the transactivation of the AR-T877A mutant. Similar
experimental setup of reporter assays as described in Fig. 1 using the expression vectors for the AR mutants or
AR-T877A, AR-F876L, and AR-W741C in the presence of R1881 (30 pM). Compounds were applied in
30 µM (A) and (B) or 10 µM (C) final concentration. Results represent luciferase activity normalized to β-
galactosidase activity derived from the co-transfected pCMV-lacZ vector. Error bars show standard deviation
of the mean. Enz, enzalutamide; Bic bicalutamide.
3.4. Lack of inhibition of the GR and PR by halogen-substituted MA-derivatives
To determine receptor specificity, we tested whether compound 26, 27, 28, 30 or 30a effect the
closely AR-related nuclear receptors GR and PR. Similar experimental setups as for AR were used
including the MMTV-luc reporter. Transfected cells with expression vectors for the GR or PR were
treated with the glucocorticoid dexamethasone or progesterone, respectively. Interestingly, we did
neither observe antagonism nor agonism by the compounds for either GR or PR suggesting that the
compounds are AR-specific antagonists (Fig. 3).
25

Fig. 3. Halogen-substituted anthranilic acid esters do not affect glucocorticoid- (GR) or progesterone- (PR)
mediated transactivation. Similar experimental setup of reporter assays as described in Figure 1 using (A) the
transcriptional active human GR or (B) the PR-B expression vector and the designated agonists
dexamethasone (10 nM) or progesterone (10 nM), respectively. Compounds were applied in 30 µM final
concentration. Error bars show standard deviation of the mean.
3.5. Inhibition of endogenous AR target gene expression
To investigate whether the chosen compounds affect the expression of endogenous AR target genes,
RT-qPCR was performed. We detected mRNA levels of the androgen-induced target genes PSA and
FKBP5 and the androgen repressed gene maspin (SERPINB5) in human PCa LNCaP cells, which
endogenously express the AR-T877A mutant. As expected, both PSA and FKBP5 mRNA levels
were increased upon treatment with androgen (Fig. 4A and B) whereas, the maspin mRNA levels
26

were reduced (Fig. 4C). The tested compounds inhibited the androgen-induced PSA and FKBP5
mRNA levels while compound 1 and compounds 27, 28 and 30 abolished the androgen-mediated
repression of maspin mRNA levels.
27

Fig. 4. Inhibition of endogenous AR induced target genes in LNCaP cells by halogen-substituted anthranilic
acid esters. Relative mRNA expression was determined by RT-qPCR. Cells were treated of indicated
compounds with or without the addition of R1881 (100 pM). Compounds were applied in 30 µM final
concentration except 1 had a final concentration of 300 µM. Whole RNA was isolated and analyzed by RT-
qPCR. Results show normalized fold expression of the mRNA of the androgen-regulated genes (A) FKBP5,
(B) PSA (KLK3) and (C) maspin (SERPINB2) normalized to β-actin and relative to DMSO control (solvent
control). Error bars show standard deviation of the mean.
3.6. Halogen-substituted anthranilic acid esters require the AR-LBD for AR-binding
Next, we tested whether halogen-substituted anthranilic acid esters require the AR-LBD for AR
inhibition applying an AR mutant lacking the LBD (AR∆LBD) (Fig. 5A). This mutant is known to
be transcriptionally active in the absence of androgens. The analyzed compounds did not inhibit the
AR-mediated transactivation, indicating that these compounds act through the LBD of the AR.
To further confirm binding of the compounds to the AR, whole-cell binding assays with the
compounds 26, 27, 28, 30 and 30a were performed (Fig. 5B). The obtained results suggest that 1
and the halogenated anthranilic acid ester derivatives compete for androgen binding and further
indicate that these compounds bind to the LBD of AR. Our results reveal that for the action of 1 and
the tested derivatives the AR LBD is crucial.
Fig. 5. Halogen-substituted anthranilic acid esters require the AR-LBD for antagonism and compete for
androgen binding to the AR. (A) The activity of the constitutively nuclear localized ARΔLBD, which lacks
the ligand-binding domain (LBD), was tested in reporter assays as described in Fig. 1. Error bars show
standard deviation of the mean. (B) Competitive whole-cell binding assays were performed using pSG5-hAR-
3
wt transfected COS-7 cells co-incubated with 1 nM (H)-mibolerone in the absence or presence of increasing
concentrations of either unlabeled mibolerone or the indicated compounds for 90 min. Competition for
28

binding is measured as disintegrations per unit time (dpm) normalized with an internal transfection control
using the β-galactosidase activity derived from co-transfected pCMV-lacZ vector and illustrated as percentage
of specific binding. Results are averages of triplicates with error bars as standard deviations of the mean.
Reduced dpm suggest competition for ligand binding.
3.7. Halogen-substituted MA-derivatives inhibit cell proliferation of human PCa cells.
To address whether the inhibition of AR-mediated transactivation coincides with inhibition of cell
proliferation, we treated human androgen-dependent growing LNCaP cells with the indicated
compounds (Fig. 6). Interestingly, the treatment with all tested compounds led to a reduced growth
of LNCaP cells (Fig. 6A).
To analyze the underlying mechanism of growth inhibition, we explored the activity of senescence-
associated β-galactosidase (SA-β Gal) activity, a specific marker for cellular senescence in which
cells irreversibly arrest in the G1/G0 cell cycle phase. We observed that the expression of SA-β Gal
activity was induced to various degrees in treated cells (Fig. 6B). The efficacy of inhibition of
LNCaP cell growth by compounds 26 and 28 correlated with a higher induction of SA-β Gal
activity. In line with this, the FACS data suggest a higher number of cells in the G1/G0-phase when
treated with 28 without an increase in the SubG1-phase, suggesting that apoptosis is not induced
(Fig. 6C). Our data indicate that the compounds inhibit cell proliferation, the cell cycle and are able
to induce cellular senescence.
Since compounds 26 and 28 had similar activities and acted AR specifically, we focused on
compound 28 for further investigations.
29