Synthesis and evaluation of 9-aminoacridines derived from benzyne click chemistry

Article abstract of DOI:10.1016/j.bmcl.2009.08.070

A small set of 9-aminoacridine-3- and 4-carboxamides were synthesized efficiently using the benzyne/azide click chemistry. The products bind to duplex DNA but have different antitumour activity in the HL60 cell line.

Full text of DOI:10.1016/j.bmcl.2009.08.070

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5880–5883
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of 9-aminoacridines derived from benzyne
click chemistry
*
Lesley A. Howell, Aaron Howman, Maria A. O’Connell, Anja Mueller, Mark Searcey
School of Pharmacy, University of East Anglia, Norwich NR4 7TJ, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
A small set of 9-aminoacridine-3- and 4-carboxamides were synthesized efficiently using the benzyne/
azide click chemistry. The products bind to duplex DNA but have different antitumour activity in the
HL60 cell line.
Received 5 August 2009
Revised 19 August 2009
Accepted 20 August 2009
Available online 23 August 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
DNA
Threading intercalators
Acridines
Benzyne click chemistry
DNA intercalators have found great utility as antitumour agents,
from natural products such as doxorubicin and actinomycin D to
synthetic agents such as mitoxantrone and amsacrine.1 Amsacrine
(1) is an acridine that binds to DNA and inhibits the enzyme topo-
isomerase II.1 It has formed the prototype for a series of new agents
that have had varying success in clinical development. Possibly of
most interest are the threading agents.2 These are intercalators in
which one substituent is located in one groove of the DNA and an-
other is threaded through the helix and resides in the other groove.
Denny and Wakelin have carried out extensive studies of these
agents, particularly modelled on the 9-aminoacridine-4-carboxam-
ides, which have consistently been shown to have potent antitu-
mour activity.3 Their mode of action has been suggested to
involve topoisomerase inhibition.4 Crystallographic studies5,6 have
recently demonstrated that the acridines bind with the 9-amino
substituent in the minor groove and the carboxamide in the major
groove. Our own studies7 of a bisintercalator that crosslinks du-
plexes showed that the 4-carboxamide makes H-bonding interac-
tions with the O6 or N7 of G and that the 9-amino substituent
protrudes from the minor groove.
activity.3 The more recent work showed that the antitumour activ-
ity is cell-line specific and that acridine-3-carboxamides such as 2
have topoisomerase inhibitory activity in pancreatic cell lines.9
This work demonstrates that the acridine-carboxamides still have
much to offer as potential therapeutic agents and herein, we de-
scribe the rapid synthesis of a small set of acridine compounds
(Fig 1, compounds 3–6) via a novel route involving benzyne click
chemistry.
At its most basic, click chemistry involves simple routes to no-
vel compounds using effective reactions with little or no chroma-
tography.10 Larock recently described the use of
a benzyne
generated in situ as one of the coupling partners in a click reaction
with an azide.11 This generates the benzotriazole structure, a fairly
common motif in medicinal chemistry. This reaction has also been
exploited by others,12 most recently by Moses and co-workers in a
further development of their aromatic azide-based click methodol-
ogy.13 In this case, the benzyne is generated in a similar manner to
the original benzotriazole route, first described in 1964,14 whereas
in the Larock chemistry the same molecule is generated via CsF-
mediated TMS-elimination.
Simple models of these interactions would suggest that a 9-
aminoacridine-3-carboxamide would not interact so well with
DNA, with the 3-substituent less favourably positioned within
the groove. However, studies by Ferguson and co-workers have
shown that acridine-3-carboxamides also have potential as antitu-
mour agents.8 Previous studies on these compounds suggested that
although they bound to duplex DNA, they had no antitumour
Prior to investigation of the click reaction in the context of acri-
dine chemistry, the synthesis of the set of substrates would require
formation of the 9-(2-azidoethyl)aminoacridine-4-carboxamide
from the 9-chloroacridine and azidoethylamine. 2-Azidoethyl-
amine was originally accessed in the literature via heating of 2-
bromoethylamine with sodium azide15 so it seemed likely that
the molecule was relatively stable (although care should be taken
when handling low molecular weight azides and this led us to seek
an alternative, milder approach to the synthesis). Commercially
available Boc-ethanolamine 7 was converted to the mesylate 8,
* Corresponding author.
E-mail address: m.searcey@uea.ac.uk (M. Searcey).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.08.070

L. A. Howell et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5880–5883
5881
Figure 1. Compound 2 has been shown to have activity against topoisomerase I.
Compounds 3–6 were synthesised in this study.
which then smoothly underwent direct substitution with sodium
azide to give the Boc-protected azide 9. Removal of the amine pro-
tecting group under standard anhydrous conditions gave the azi-
doethylamine as the hydrochloride salt 10 (Scheme 1).
9-Aminoacridine carboxamides substituted in the 3- and 4-po-
sition can be readily accessed via reported methods (Scheme 2).3,16
For example, aniline reacted with 2-bromoterephthalic acid 11 un-
der Jourdan–Ullmann conditions (Cu, CuI, Py, 100 °C, 53%) to give
2-phenylamino terephthalic acid 12 which was cyclised in essen-
tially quantitative yield to give 9(10H)-acridone-3-carboxylic acid
13 (PPA, 120 °C, 99%). Conversion to the 9-chloroacridine-3-car-
boxamides 14 and 15 by treatment with thionyl chloride under re-
flux followed by rapid addition of an excess of the required amine
in dichloromethane yielded the desired products (Scheme 2). Com-
pound 14 was purified by column chromatography whereas 15 has
been reported in the literature to be unstable, a property confirmed
in our hands and as such was used in the next step without further
purification.16 For the 4-carboxamides, cyclisation of the commer-
cially available 2,20-iminodibenzoic acid 16 with sulfuric acid gave
acridone-4-carboxylic acid 17 which was then converted to 9-chlo-
roacridine-4-carbonyl chloride using thionyl chloride with cata-
lytic DMF. Selective reaction of the acid chloride involving rapid
addition of the required amine along with triethylamine in cold
CH2Cl2 gives the target carboxamides 18 and 19. Compound 19
again proved to be unstable and was used without further purifica-
tion. The amines to form the carboxamide were selected on the ba-
sis of previous work. Thus the 2-(dimethylamino)ethylamine side
chain has previously been used extensively by us and others and
Scheme 2. Synthesis of (i) 9-chloroacridine-4-carboxamides and (ii) 3-carbox-
amides.
consistently generates compounds with good cytotoxicity and
DNA binding characteristics. The piperazine sidechain was used
by Ferguson and co-workers and had good activity in a cell line
selective fashion.
The substrates for the click reaction were generated through an
in situ reaction with the 9-phenoxyacridine, generated by, for
example, heating 14 to 110 °C in phenol for 15 min (Scheme 3).
Cooling to 55 °C was followed by addition of 10 and stirring at this
temperature for 24 h. Compound 20 was obtained in excellent yield
as the hydrochloride salt following purification and the structure
Scheme 1. Synthesis of 2-azidoethylamine 10.
Scheme 3. Synthesis of 9-(2-azidoethyl)aminoacridine carboxamides.

5882
L. A. Howell et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5880–5883
was verified by 1H NMR, IR and high resolution mass spectrometry.
Compounds 21–23 were prepared in a similar fashion from the
respective phenoxyacridines in good to excellent yield.17
Initial studies focussed on the Larock click chemistry involving
reaction of o-(trimethylsilyl)phenyltriflate with CsF to generate the
benzyne in situ, which then reacts directly with the azide. The opti-
mum conditions described by the original authors involved stirring
at room temperature in CH3CN and reaction times of 18–24 h.
Gratifyingly, all of the azide substrates reacted under these
conditions to give the benzotriazole products 3–6 in good yield
(Scheme 4). All of the compounds generated were extremely polar
and purification by column chromatography under normal phase
conditions was not possible. However, it was found that basifica-
tion of the reaction mixture with NaHCO3, followed by extraction
into CH2Cl2 and removal of solvent gave an oil that could then be
converted into the HCl salt (1.25 M HCl/CH3OH), which was found
to be >90% pure as assessed by proton NMR. This purification of the
target compounds led us to discount the similar Feringa methodol-
ogy12 for this reaction, which uses an excess of one of the sub-
strates and would probably require further purification, although
this later proved not to be a problem with the Moses chemistry
(see below).
Scheme 5. Moses click chemistry to generate the compounds 3–6.
The use of anthranilic acid as a starting material for the synthesis
was also attractive, as the number of substituted analogues of these
compounds that are commercially available or easily accessed is
greater than that for the o-(trimethylsilyl)phenyltriflates. With this
in mind, the Moses methodology for synthesis of similar com-
pounds was also briefly investigated (Scheme 5). The optimised
conditions for this reaction involved heating to reflux in
acetonitrile for 15 h with 2 equiv of anthranilic acid and 2 equiv
of t-BuONO.13 In fact, under these conditions with our substrates,
excessive decomposition was observed. The reaction failed to
progress at room temperature. Ultimately, we found that the use
of 2 equiv of the anthranilic acid and four of t-BuONO gave the
required products in fair to good yields, although in the case of 4
there was a substantial decrease in product obtained. Work-up
was exactly as for the Larock conditions, basification and extraction
into CH2Cl2 and gave products of similar purity on acidification.
Compounds 3–6 were assessed initially for their antitumour
activity against the human leukaemia cell line HL60 and the results
are shown in Table 1. None of the click chemistry products had
antitumour activity higher than the parent azide 20–23 and the
piperidine side chain (compounds 4, 6, 21, 23) clearly leads to a
Table 1
Cytotoxicity and DNA binding constants for compounds 3–6 and 16–19
Compounds
Cytotoxicity
IC50 (lM)
C50
(
l
M)b
Log Kapp
a
3
4
5
262.4
na
23.4
na
12.1
26.2
30.0
93.2
17.6
>200
13.6
>200
6.54
6.20
6.14
5.65
6.37
—
6
20
21
22
23
5.4
45.7
0.72
52.0
6.49
—
a
Carried out in HL60 cell line (na = not active).
Competitive ethidium bromide displacement assay.
b
further decrease in antitumour activity. This decrease in antitu-
mour activity could be a consequence of a decrease in DNA binding
affinity. Initially, we assessed the relative binding affinity of the
various agents via an ethidium displacement assay at a single con-
centration (2 lM) for each agent. None of the click compounds 3–6
showed any appreciable displacement at this concentration,
whereas the 9-(2-azidoethyl)aminoacridine-4-carboxamides 20–
23 all had significant effects. Titration of the agents against
ethidium bromide gave C50 values and Kapp values as shown in
Table 1.19
The antitumour activity of these compounds against HL60 cells
is clearly not directly related to their DNA binding affinity, as com-
pounds with similar log Kapp values varied from low micromolar
activity to inactive. Interestingly, it is clear from these results that
a change in carboxamide substitution pattern from the 3- to the 4-
position has little effect on DNA binding affinity and that the struc-
ture of the side chain has a much greater impact. For three of the
four piperidine compounds (6, 21, 23) there is a significant de-
crease in DNA binding affinity, with no binding detected for the
azide derivatives 21 and 23. However, these same compounds
maintain antitumour activity, whereas 4 and 6, which have mea-
surable binding affinity, have no antitumour effect in HL60 cells.
This suggests that either DNA binding is not required for biological
activity of acridine carboxamides or that the compounds with dif-
ferent side chains are acting through different mechanisms.
As the acridines are intrinsically fluorescent, we briefly and
qualitatively investigated their uptake in HL60 cells to see if this
Scheme 4. Larock click chemistry to generate compounds 3–6.18

L. A. Howell et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5880–5883
13. Zhang, F.; Moses, J. E. Org. Lett. 2009, 11, 1587.
5883
may have affected their antitumour activity (data not shown). All
compounds were taken up by the cells and this is unlikely to ex-
plain the total lack of activity of the 3-carboxamides. If the acri-
dine-4-carboxamides are inhibitors of topoisomerase II, as has
previously been reported, then it seems likely that the piperidine
substituted analogues are unable to exert an effect on this enzyme,
at least in these cell lines.
In summary, benzyne click chemistry is an important addition
to the arsenal of reactions available to the medicinal chemist. We
have shown that it can be applied to the synthesis of threading
intercalators that bind to DNA with reasonable affinity and have
up to low micromolar antitumour activity. The reasons for differ-
ences in the activity of the different compounds are still under
investigation but do not seem to correlate with DNA binding affin-
ity or cellular uptake. This may suggest that these compounds vary
in their ability to inhibit topoisomerases, as described by Ferguson
and co-workers, and we are currently investigating this potential
mode of action. This is also the first description of the azide-substi-
tuted threading intercalators and these compounds are currently
being studied for their ability to undergo click reactions under
the usual Sharpless conditions.
14. Reynolds, G. A. J. Org. Chem. 1964, 29, 3733.
15. Forster, M. O.; Newman, S. H. J. Chem. Soc. 1911, 99, 1277.
16. Goodell, J. R.; Madhok, A. A.; Hiasa, H.; Ferguson, D. M. Bioorg. Med. Chem. 2006,
14, 5467.
17. Compound 12:
(N3), 1643 (CO). dH (300 MHz; CDCl,) 8.32 (1H, d, J = 1.5 Hz, H-4), 8.13 (2H, t,
J = 9.3 Hz, H-1 + H-8), 7.93 (1H, d, J = 8.7 Hz, H-5), 7.78 (1H, dd, J1 = 1.8 Hz, J2
m
max/cmꢀ1 3275 (NH), 2936 (CH2/CH3), 2819 (CH2/CH3), 2094
=
9.0 Hz, H-2), 7.69–7.63 (1H, m, H-6), 7.12–7.36 (2H, m, H-7 + NH), 3.95 (2H, t,
J = 5.4 Hz, CH2N3), 3.65–3.55 (4H, m, CONHCH2 + NHCH2), 2.56 (2H, t, J = 5.9 Hz,
CH2N(CH3)2), 2.29 (6H, s, N(CH3)2). m/z (ES+) 378.2035 [M+H]+ C20H24N7O
requires 378.2037; compound 13: m
max/cmꢀ1 3273 (NH), 2935 (CH2/CH3), 2790
(CH2/CH3), 2095 (N3), 1628 (CO). dH (300 MHz; CDCl3) 8.13 (1H, d, J = 9.0 Hz, H-
8), 8.09 (1H, d, J = 9.0 Hz, H-1), 7.89 (2H, br s, H-4 + H-5), 7.62 (1H, t, J = 7.5 Hz,
H-6), 7.37–7.30 (2H, m, H-7 + H-2), 3.92 (2H, t, J = 5.6 Hz, CH2N3), 3.83 (2H, br s,
CONCH2), 3.59 (2H, t, J = 5.6 Hz, NHCH2), 3.51 (2H, br s, CONCH2), 2.49 (2H, br s,
CH2NCH3), 2.35 (2H, br s, CH2NCH3), 2.30 (3H, s, NCH3). m/z (ES+) 390.2035
[M+H]+ C21H24N7O requires 390.2037; compound 14: max/cmꢀ1 3313 (NH),
m
2935 (CH2/CH3), 2856 (CH2/CH3), 2095 (N3), 1638 (CO). dH (300 MHz; MeOD)
8.54 (1H, d, J = 6.6 Hz, H-1), 8.36 (1H, dd, J1 = 1.4 Hz, J2 = 8.7 Hz, H-3), 8.21 (1H,
d, J = 8.7 Hz, H-8), 7.87 (1H, d, J = 8.1 Hz, H-5), 7.66 (1H, t, J = 8.4 Hz, H-6), 7.37–
7.32 (2H, m, H-2 + H-7), 3.92 (2H, t, J = 6.3 Hz, CH2N3), 3.69 (2H, t J = 6.3 Hz,
CONHCH2), 3.60 (2H, t, J = 5.7 Hz, NHCH2), 2.70 (2H, t, J = 6.5 Hz, CH2N(CH3)2)
2.39 (6H, s, N(CH3)2). m/z (ES+) 378.2039 [M+H]+ C20H24N7O requires
378.2037; compound 15: m
max/cmꢀ1 3323 (NH), 2936 (CH2/CH3), 2799 (CH2/
CH3), 2096 (N3), 1608 (CO). dH (500 MHz; CH3OD) 8.34 (1H, dd, J1 = 1.0 Hz,
J2 = 8.5 Hz, H-1), 8.24 (1H, d, J = 8.5 Hz, H-8), 7.94 (1H, d, J = 8.5 Hz, H-5), 7.65–
7.61 (2H, m, H-3 + H-6), 7.37–7.32 (2H, m, H-2 + H-7), 3.96 (2H, br s, CONCH2),
3.92 (2H, t, J = 5.5 Hz, CH2N3), 3.58 (2H, t, J = 5.5, NHCH2), 3.20 (2H, br, s,
CONCH2), 2.65 (2H, br s, CH2NH), 2.29 (5H, s, NCH3 + CH2NH). m/z (ES+)
390.2032 [M+H]+ C21H24N7O requires 390.2037.
Acknowledgements
18. Compound 3: m
max /cmꢀ1 3253 (NH), 3034 (CH), 2935 (CH2/CH3), 1637 (CO). dH
This work was funded by a UEA graduate studentship (L.H.) and
a summer studentship (A.H.). Mass spectra were obtained from the
National Mass Spectrometry service at Swansea.
(500 MHz; MeOD) 8.57 (1H, d, J = 9.0 Hz, H-1), 8.54 (1H, d, J = 8.5 Hz, H-5), 8.15
(1H, d, J = 1.0 Hz, H-4), 8.01 (1H, t, J = 7.5 Hz, H-7), 7.96 (2H, d, J = 8.5 Hz,
benzyl), 7.87 (1H, d, J = 9.0 Hz, H-8), 7.73–7.71 (1H, m, H-2), 7.64–7.59 (2H, m,
H-6 + benzyl), 7.49 (1H, app q, J = 7.5 Hz, benzyl), 4.36 (2H, t, J = 5.5 Hz, CH2N3),
4.31 (2H, t, J = 6.3 Hz, CH2N(CH3)2), 3.96 (2H, t, J = 5.5 Hz, NHCH2), 3.78 (6H, s,
N(CH3)2), 3.69 (2H, t, J = 6.3 Hz, CONHCH2). m/z (ES+) 454.2345 [M+H]+
References and notes
C26H28N7O requires 454.2350; compound 4:
m
max/cmꢀ1 3349 (NH), 3024
(CH), 2937 (CH2/CH3), 2915 (CH2/CH3), 1635 (CO). dH (500 MHz; MeOD) 8.66
(1H, d, J = 8.5 Hz, H-1), 8.56 (1H, d, J = 8.5 Hz, H-5), 8.04–8.01 (1H, m, H-7), 8.00
(1H, d, J = 1.0 Hz, H-4), 7.96 (2H, d, J = 8.0 Hz, benzyl), 7.89 (1H, d, J = 8.5 Hz, H-
8), 7.74 (1H, t, J = 7.8 Hz, benzyl), 7.68 (1H, d, J = 7.0 Hz, benzyl), 7.66–7.63 (2H,
m, H-2 + H-6), 4.68 (2H, br s, CH2), 4.58 (2H, br s, CH2), 4.38 (2H, t, J = 5.5 Hz,
CH2N3), 4.22 (2H, br s, CH2), 3.97 (2H, t, J = 5.5 Hz, NHCH2), 3.74 (2H, br s, CH2),
3.63 (3H, s, NCH3). m/z (ES+) 466.2347 [M+H]+ C27H28N7O requires 466.2350;
1. Brana, M. F.; Cacho, M.; Gradillas, A.; de Pasxual-Teresa, B.; Ramos, A. Curr.
Pharm. Des. 2001, 7, 1745.
2. Wakelin, L. P. G.; Chetcuti, P.; Denny, W. A. J. Med. Chem. 1990, 33, 2039.
3. Atwell, G. J.; Cain, B. F.; Baguley, B. C.; Finlay, G. J.; Denny, W. A. J. Med. Chem.
1984, 27, 1481.
4. Woynarowski, J. M.; McCarthy, K.; Reynolds, B.; Beerman, T. A.; Denny, W. A.
Anticancer Drug Des. 1994, 9, 9.
5. Adams, A.; Guss, J. M.; Collyer, C. A.; Denny, W. A.; Wakelin, L. P. G. Biochemistry
1999, 38, 9221.
6. Adams, A.; Guss, J. M.; Denny, W. A.; Wakelin, L. P. G. Nucleic Acids Res. 2002, 30,
719.
7. Hopcroft, N. H.; Brogden, A. L.; Searcey, M.; Cardin, C. J. Nucleic Acids Res. 2006,
34, 6663.
8. Goodell, J. R.; Ougolkov, A. V.; Hiasa, H.; Kaur, H.; Remmel, R.; Billadeau, D. D.;
Ferguson, D. M. J. Med. Chem. 2008, 51, 179.
compound 5: m
max/cmꢀ1 3359 (NH), 3039 (CH), 2935 (CH2/CH3), 1621 (CO). dH
(500 MHz; MeOD) 8.71 (1H, d, J = 8.0 Hz, H-1), 8.55 (1H, d, J = 8.5 Hz, H-5), 8.19
(1H, d, J = 7.5 Hz, H-3), 8.04 (1H, t, J = 7.8 Hz, H-7), 7.99 (2H, d, J = 8.5 Hz,
benzyl), 7.94 (1H, d, J = 8.5 Hz, H-8), 7.64 (1H, t, J = 7.5 Hz, H-6)), 7.58–7.53 (2H,
m, H-2 + benzyl), 7.41 (1H, t, J = 7.5 Hz, benzyl), 4.39–4.34 (4H, m,
CH2N(CH3)2 + CH2N3), 3.97 (2H, t, J = 5.0 Hz, NHCH2), 3.80 (6H, s, N(CH3)2),
3.75 (2H, t, J = 6.0 Hz, CONHCH2). m/z (ES+) 454.2354 [M+H]+ C26H28N7O
requires 454.2350; compound 6: m
max/cmꢀ1 3358 (NH), 3020 (CH), 2987 (CH2/
CH3), 2945 (CH2/CH3), 1630 (CO). dH (500 MHz; MeOD) 8.63 (1H, d, J = 8.0 Hz,
H-1), 8.51 (1H, d, J = 8.5 Hz, H-5), 8.04 (2H, br d, J = 8.0 Hz, H-8 + H-3), 7.98–
7.93 (3H, m, H-7 + benzyl), 7.74 (1H, t, J = 7.5 Hz, benzyl), 7.66 (1H, t, J = 7.0 Hz,
benzyl), 7.61–7.56 (2H, m, H-6 + H-2), 4.67 (2H, br s, CH2), 4.54 (2H, br s, CH2),
4.35 (2H, s, CH2N3), 4.23 (2H, br s, CH2), 4.08 (2H, br s, CH2), 3.97 (2H, s,
NHCH2), 3.65 (3H, s, NCH3). m/z (ES+) 466.2353 [M+H]+ C27H28N7O requires
466.2350.
9. Oppegard, L. M.; Ougolkov, A. V.; Luchini, D. N.; Schoon, R. A.; Goodell, J. R.;
Kaur, H.; Billadeau, D. D.; Ferguson, D. M.; Hiasa, H. Eur. J. Pharmacol. 2009, 602,
223.
10. Moses, J. E.; Moorhouse, A. D. Chem. Soc. Rev. 2007, 36, 1249.
11. Shi, F.; Waldo, J. P.; Chen, Y.; Larock, R. C. Org. Lett. 2008, 10, 2409.
12. Campbell-Verduyn, L.; Elsinga, P. H.; Mirfeizi, L.; Dierckx, R. A.; Feringa, B. L.
Org. Biomol. Chem. 2008, 6, 3461; Chandrasekhar, S.; Seenaiah, M.; Rao, C. L.;
Reddy, C. R. Tetrahedron 2008, 64, 11325.
19. Jenkins, T. C. Methods Mol. Biol. 1997, 90, 195–218.