Synthesis of novel proxyphylline derivatives with dual Anti-Candida albicans and anticancer activity

Article abstract of DOI:10.1016/j.ejmech.2018.02.077

Three out of 16 newly synthesized 1,3-dimethylxanthine derivatives (proxyphylline analogues) exhibited consistencies between antifungal and anticancer properties. Proxyphylline possessing 1-(10H-phenothiazin-10-yl)propan-2-yl (6) and polybrominated benzimidazole (41) or benzotriazole moiety (42) remained selectively cidal against Candida albicans (lg R ≥ 3 at conc. of 31, 36 and 20 μM, respectively) however not against normal mammalian Vero cell line in vitro (IC₅₀ ≥ 280 μM) and Galleria mellonella in vivo. These compounds also displayed moderate antineoplastic activity against human breast adenocarcinoma (MCF-7) cell line (EC₅₀ = 80 μM) and high against peripheral blood T lymphoblast (CCRF-CEM) (EC₅₀ = 6.3–6.5 μM). In addition, 6 and 42 exerted: (1) dual activity against fungal adhesion and damage mature biofilm; (2) necrosis of planktonic cells due to loss of membrane function and of structural integrity; (3) biochemical (inhibition of sessile cell respiration) and morphological changes in cell wall polysaccharide contents. Therefore, leading proxyphylline derivatives can be employed to prevent cancer-associated biofilm Candida infections.

Full text of DOI:10.1016/j.ejmech.2018.02.077

Accepted Manuscript
Synthesis of novel proxyphylline derivatives with dual Anti-Candida albicans and
anticancer activity
Paweł Borowiecki, Patrycja Wińska, Maria Bretner, Małgorzata Gizińska, Mirosława
Koronkiewicz, Monika Staniszewska
PII:
S0223-5234(18)30214-9
10.1016/j.ejmech.2018.02.077
EJMECH 10254
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 21 December 2017
Revised Date: 10 February 2018
Accepted Date: 23 February 2018
Please cite this article as: Paweł. Borowiecki, P. Wińska, M. Bretner, Mał. Gizińska, Mirosł.
Koronkiewicz, M. Staniszewska, Synthesis of novel proxyphylline derivatives with dual Anti-Candida
albicans and anticancer activity, European Journal of Medicinal Chemistry (2018), doi: 10.1016/
j.ejmech.2018.02.077.
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.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Synthesis of Novel Proxyphylline Derivatives with Dual Anti-Candida albicans and
Anticancer Activity
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
a
a
a
b
Paweł Borowiecki *, Patrycja Wińska , Maria Bretner , Małgorzata Gizińska , Mirosława
c
b
Koronkiewicz , Monika Staniszewska *
a
Warsaw University of Technology, Faculty of Chemistry, Noakowskiego 3, 00-664 Warsaw,
Poland.
b
National Institute of Public Health-National Institute of Hygiene, Chocimska 24, 00-791
Warsaw, Poland.
c
National Medicines Institute, Chełmska 30/34, 00-725 Warsaw, Poland.
1
1
1
1
1
1
1
1
1
1
2
2
*Correspondence:
Paweł Borowiecki, Warsaw University of Technology, Faculty of Chemistry, Noakowskiego
3, 00-664 Warsaw, Poland
Phone: +48 22 660 53 42. Fax: + 48 22 628 27 41
E-mail: pawel_borowiecki@onet.eu
Monika Staniszewska, Independent Laboratory of Streptomyces and Fungi Imperfecti,
National Institute of Public Health – National Institute of Hygiene, 24 Chocimska, 00-791
Warsaw, Poland
Phone: +48 22 54 21 228. Fax: + 48 22 849 74 84
E-mail: mstaniszewska@pzh.gov.pl
2
2
2
2
2
2
2
2
3
3
3
3
2
3
4
5
6
7
8
9
0
1
2
3
1

ACCEPTED MANUSCRIPT
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Abstract: Three out of 16 newly synthesized 1,3-dimethylxanthine derivatives (proxyphylline
analogues) exhibited consistencies between antifungal and anticancer properties.
Proxyphylline possessing 1-(10H-phenothiazin-10-yl)propan-2-yl (6) and polybrominated
benzimidazole (41) or benzotriazole moiety (42) remained selectively cidal against Candida
albicans (lg R≥3 at conc. of 31, 36 and 20 µM, respectively) however not against normal
mammalian Vero cell line in vitro (IC ≥280 µM) and Galleria mellonella in vivo. These
5
0
compounds also displayed moderate antineoplastic activity against human breast
adenocarcinoma (MCF-7) cell line (EC =80 µM) and high against peripheral blood T
5
0
lymphoblast (CCRF-CEM) (EC =6.3–6.5 µM). In addition, 6 and 42 exerted: (1) dual
5
0
activity against fungal adhesion and damage mature biofilm; (2) necrosis of planktonic cells
due to loss of membrane function and of structural integrity; (3) biochemical (inhibition of
sessile cell respiration) and morphological changes in cell wall polysaccharide contents.
Therefore, leading proxyphylline derivatives can be employed to prevent cancer-associated
biofilm Candida infections.
Keywords: 1,3-Dimethylxanthines, Proxyphylline analogues, Fungicidal activity, Antibiofilm
activity, Selective toxicity, Anticancer activity.
2

ACCEPTED MANUSCRIPT
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
Introduction:
Candida albicans is the most prevalent opportunistic pathogen of the Candida infection
(CI) development in immunocompromised patients [1,2]. Frequently, many types of cancers
and hemathologic malignances as well as the therapy related to them constitute predisposing
risk factors for CI [3,4]. Candidiasis is an important problem in hospitals with an ever-
increasing mortality rate and economic costs due to the antimycotic resistance lowering the
efficiency of anti-infection therapies [3-7]. Candida albicans possesses unique fitness
attributes that give it the ability to colonize epithelial cells, adhere, develop biofilm, invade
deeper layers of the host epithelium, and to stress response [8]. Thus, inactivation of the
virulence factors can help attenuate the pathogenicity of C. albicans [9]. Nowadays, if the
candidiasis and cancer interdependence is an increasing concern [1], the design of effective
agents targeted at the induction of programmed cell death (PCD) of both proliferative
mammalian and fungal cells needs to be investigated.
One reason for choosing proxyphylline as a structural core motif for novel potential
biologically active compounds is its unique physicochemical properties, which make
proxyphylline highly soluble, permeable, and thus bioavailabile. Recent studies [10,11]
revealed that 8-substituted purines with five-membered heterocyclic rings display anticancer,
anti-HIV-1, and antibacterial activity via inhibiting nucleic acid synthesis. Moreover,
Voynikov et al. [12] showed that the theophylline-7-acetamides containing amido acid
moieties were active in vitro against Mycobacterium tuberculosis at the MIC values from 0.26
to 0.46 µM. However, research works devoted to the anticancer activity of 1,3-
dimethyxanthines are still rather scarce and definitely need more attention. Moreover, the
antifungal properties of 1,3-dimethylxanthine-based compounds, to the best of our
knowledge, have never been characterized. To achieve the desired antiproliferative effect of
proxyphylline (1), an appropriate modification of the alkyl side chain at the hydroxyl group
was performed. This included incorporation of various aromatic moieties in the xanthine at
some arbitrary distances. Moreover, we investigated the influence of the type of aromatic
substituent as well as different spacers (substituted acetamide, oxyacetamide, ester, and ether
moieties). Thus three variants of the proxyphylline (1) lead-structure were obtained in order to
gain insight into the structure–activity relationships (SAR). For a better understanding of the
structural variation groups see respective Schemes 1, 5 and 7 in the following sections.
Since breast cancer and hematologic malignancies exhibit a high risk of patient mortality
due to CI development [13], in this study novel proxyphylline analogues’ cytotoxicity against
3

ACCEPTED MANUSCRIPT
8
8
8
8
8
8
9
9
9
9
9
9
9
9
9
9
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
the breast cancer (MCF-7) and human T-cell leukemia (CCRF-CEM) was examined in vitro.
Following the current opinion that yeasts can constitute model for the development of both
anticancer and antifungal agents due to the high degree of conservation of basic cellular
processes between yeast and higher eucariota [14], we extended the PCD studies in C.
albicans under proxyphylline derivatives. The first aim of the antifungal studies includes an in
vitro screen of 16 proxyphylline analogues’ inhibitory and biocidal potential against C.
albicans. Based on the MIC and MFC values determined using the M27-A3 method, a few
proxyphylline derivatives which exhibit distinct antifungal properties were selected for further
investigation i.e., cytotoxicity against mammalian cells using the XTT method. Subsequently,
we extended the studies to elucidate the action mode of two proxyphylline derivatives: 6
(Schemes 1) and 42 (see Schemes 1 and 7), and finally one of these leading compounds
exhibiting the best fungicidal activity, 42 (see Scheme 7). Nuclear and cell wall
polysaccharides staining techniques using the confocal laser scanning microscope (including
mature biofilm) and flow cytometry (planktonic cells) were performed to define the cell
surface barrier damage and programmed cellular events under these compounds. Further
investigation of these compounds’ antifungal properties included: (1) effect on the formation
(adhesion to mammal cell line) and mature biofilm cell mitochondrial dysfunction using the
XTT assay; (2) chitinolytic activity using the colorimetric assay; (3) cell membrane damage
(phosphatidylserine externalization, increased permeabilization, and ergosterol synthesis
inhibition) with flow cytometry and microplate spectrophotometry analyses. With the
possibility of future clinical use in mind, the cytotoxic and antifungal effects (containing host-
pathogen study) of these compounds were investigated in vivo using an invertebrate Galleria
mellonella as a mini-host.
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
Results
Synthetic Chemistry
Synthesis of Proxyphylline Carbamates 3-5, Carbonate 6 and Thioester 7
The synthetic route of heterocyclic carbamates 3-5, carbonate 6 and thioester 7 based on
proxyphylline (1) scaffold is outlined in Scheme 1. The starting material 1 was prepared as
previously reported by us [15] via a regioselective propylene oxide ring-opening by
commercially available theophylline (1,3-dimethylxanthine) promoted by catalytic amount of
triethylamine (Et N) in boiling methanol. Next, the obtained racemic proxyphylline (1) was
3
treated with phosgene used in a 1.1-fold molar excess relative to the substrate in the presence
of 1.1 equiv of Et N as a base, thus leading to desired crude chlorocarbonate 2 in fair yield of
3
4

ACCEPTED MANUSCRIPT
117
118
119
120
121
122
123
124
125
126
127
128
129
130
ca. 70%. At this point, it is important to mention that performing the reaction with less
harmful triphosgene under the same conditions failed. In this instance, even if the base (Et N)
3
was changed on pyridine or Na CO , and the solvent (CH Cl ) on THF the crude 2 was
2
3
2
2
significantly contaminated with dimer of proxyphylline carbonate as well as the by-product of
elimination [1,3-dimethyl-7-[(1E)-prop-1-en-1-yl]-3,7-dihydro-1H-purine-2,6-dione]
1
according to H-NMR and HRMS. Nonetheless, with chloroformate intermediate 2 in hand,
we were able to carry out reaction with several aryl compounds, such as: aniline, 5H-
dibenzo[b,f]azepine, phenothiazine, 1-(10H-phenothiazin-10-yl)propan-2-ol and thiophenol,
respectively. The treatment of the above-mentioned compounds in a convenient manner by
using 1.2 equiv of 2 and 1.3 equiv of Et N in the presence of catalytic amount of 4-
3
dimethylaminopyridine (DMAP) in CH Cl and stirring at room temperature for 12 h resulted
2
2
in desired products 3-7 in moderate (25–44% for 3, 5, and 6) to good (69% for 7) and very
good (81% for 4) yields, accordingly.
1
31
32
1
Scheme 1. Synthesis of proxyphylline carbamates 3–5, carbonate 6 and thioester 7. Reagents
133
134
135
136
137
and conditions: (i) phosgene (1.1 equiv., as 15 wt.% in PhCH ), Et N (1.1 equiv.), dry
3
3
CH Cl , 0.5 h at -15 °C, then slowly warmed up to RT for 12 h; (ii) aniline or 5H-
2
2
dibenzo[b,f]azepine or phenothiazine or 1-(10H-phenothiazin-10-yl)propan-2-ol or
thiophenol, 2 (1.2 equiv.), Et N (1.3 equiv.), DMAP (cat.), dry CH Cl , 0–5 °C, then 12 h at
3
2
2
RT.
5

ACCEPTED MANUSCRIPT
138
139
140
141
142
143
144
145
146
147
148
149
150
151
Synthesis of Proxyphylline Esters 21-28
To identify novel proxyphylline analogues as dual anticancer-antifungal agents and to get
further SARs data, in the second series of experiments, we planned to introduce 1, 2 and 3
carbon atoms-long alkyl spacer between the xanthine moiety and various aromatic rings based
on the ester bond connection. Moreover, our major goal at this stage was the preparation of a
structurally more diverse library of the tested compounds than would have been possible
through the simple replacement of just the single-type substituents on a given proxyphylline
backbone. For these purposes, we decided to prepare three series of carboxylic acid
derivatives based on cheap starting materials, such as: acetophenones, arenecarbaldehydes
and aryl acetic acids.
The first series of derivatives posses three carbon atom alkyl spacer with 1,3-dicarbonyl
moiety. The method used for synthesis of 3-phenyl-3-oxo-propionic acids 10a-c is shown in
Scheme 2.
1
52
53
1
Scheme 2. Synthesis of β-keto acids 10a-c. Reagents and conditions: (iii) dimethyl carbonate
154
155
156
157
(3 equiv.), NaH (4 equiv., as 60% mineral oil dispersion), dry PhCH , 1 h at RT, then 12 h at
3
110 °C; (iv) NaOH (1.2 equiv.), H O/1,4-dioxane (2:1, v/v), 10 min at 0–5 ºC, then 12 h at
2
RT.
158
159
160
161
162
163
164
165
166
The two-step synthetic approach was initiated with the reaction between commercially
available acetophenones 8a-c and 3-fold molar excess of dimethyl carbonate under NaH-
mediated conditions in toluene at reflux for 12 h, which afforded aryl β-ketoesters 9a-c with
yields 90-93%. In the next step, to obtain 10a-c, mild reaction conditions had to used as 3-
oxo-3-phenylpropionic acids are thermodynamically unstable due to spontaneous
decarboxylation in solution [16,17]. To circumvent this instability of 1,3-dicarbonyl
compounds, and thus minimize degradation of β-keto acids 10a-c, methyl propionates 9a-c
were treated with cold H O/1,4-dioxane solution of NaOH used in 1.2-fold molar excess, and
2
left with stirring for further 12 h at ambient temperature under TLC-control. There are few
6

ACCEPTED MANUSCRIPT
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
reasons for using dioxane-water mixture as the reaction medium. An addition of 1,4-dioxane
is required to enhance poor aqueous solubility of aromatic β-keto esters 9a-c, and as it was
reported by Straub et al. [18] benzoylacetic acids show a minor decarboxylation rate in the
presence of 50% aq. 1,4-dioxane solution when compared with mixtures of i.e. isopropanol-
water, acetonitrile-water or pure water. Finally, with the employed reaction conditions aryl
acids were obtained with the following yields: 78% for benzoylacetic acid (10a), 56% for
para-fluorobenzoylacetic acid (10b), and 41% for para-methylbenzoylacetic acid (10c). It
should be pointed out that using acid-catalyzed (1M HCl) or lipase-mediated [Novozym
435/Tris-HCl buffer (pH 7.5)] hydrolytic conditions failed as reactions proceeded with
significant decomposition of the formed labile products 10a-c.
The result concerning hydrolysis based on enzymatic catalysis was in accordance with
already reported data [19,20]. However, shortly after successfully performed base-catalyzed
hydrolysis of 9a-c, we found the report on efficient enzymatic hydrolysis with lipase from
Candida rugosa Type VII in the presence of 10% DMSO in Tris-HCl buffer (100 mM of pH
8.0) [21]. Nevertheless, at this stage of our investigations we did not need alternatives, and
thus abandoned to follow this new procedure. Moreover, it is worth noting that the obtained
β-keto acids 10a-c turned out to be greatly prone to decarboxylation especially in chloroform
solution. We observed this feature trying to record NMR spectra of 10a-c in CDCl . It turned
3
out that shortly (0.5-1 h) after preparing the samples of 10a-c, the spectrum showed only the
presence of corresponding acetophenons 8a-c. Interestingly, this is in contradiction to
previously reported observations concerning the independence of the decarboxylation rate of
β-keto acids on both ring substituent and solvent effects [22]. The clear differences regarding
the stability was then observed when spectra were recorded again in deuterated chloroform
(CDCl ), acetone [(CD ) CO] or dimethylsulfoxide [(CD ) SO], respectively (see Supporting
3
3 2
3 2
Information).
In order to investigate the role and importance of two carbon atoms-long linkers connected
with various other aryl moieties for cytotoxic activity against human cancer cells and C.
albicans yeast, the acid chlorides of O-methylmandelic acids 16a-c were obtained in a five-
step reaction sequence starting from arenecarbaldehydes 11a-c (Scheme 3). The detailed
synthetic procedures including preparation of α-methoxyphenylacetic acids 15a-c from the
corresponding aryl carboxaldehydes 11a-c were reported by our group previously [23]. In this
work, we have extended the reaction pathway of one step to obtain reactive acid chlorides
16a-c useful for further proxyphylline (1) structural modifications. This task was
7

ACCEPTED MANUSCRIPT
200
201
202
203
accomplished by reaction of the appropriate acid 15a-c with 10 equiv of SOCl in refluxing
2
benzene for 2 h, and subsequent stirring the mixtures at ambient temperature for 4 h. The
crude compounds 16a-c were further used without purification.
2
04
05
2
Scheme 3. Synthesis of O-methylmandelic acid chlorides 16a-c. Reagents and conditions: (v)
206
207
208
209
210
211
CHBr (1 equiv.), LiCl (1.9 equiv.), KOH (3.9 equiv.), finely crushed ice cubes, 1,4-dioxane,
3
36 h at 0–5 °C, 24 h at 35 °C; (vi) SOCl (1.1 equiv.), dry MeOH, 0.5 h at -30 °C, then 4 h at
2
RT; (vii) Ag O (1.18 equiv.), roasted CaSO (1.18 equiv.), CH I (24.91 equiv.), low-light
2
4
3
conditions, 36 h at reflux; (viii) LiOH (4 equiv.), MeOH, 24 h at RT, then 36 % HCl, 0–5 °C;
(ix) SOCl (10 equiv.), benzene, 2 h at reflux (under N -atmosphere), then 4 h at RT.
2
2
212
213
214
215
216
217
218
219
220
221
222
The next task was to prepare aryl compounds that would provide framework consisting of
1-2 carbon atom-long linkers and bulky aromatic ring systems of high intercalative potential
to DNA strands, and thus an improved inducement of cellular death. Moreover, it was
worthwhile to investigate how structural flexibility/rigidity among aromatic pattern could
impact on biological activity of the designed compounds. In this regard, we synthesized two
closely related precursors, which main motif would posses two benzene rings connected to
carbons with fully rotatable bonds [like in diphenylacetic acid (17)] and with blocked flexible
rotation [like in 9H-fluorene-9-carboxylic acid (19)]. The respective acid chlorides 18 and 20
were synthesized from commercially available acids 17 and 19 in accordance with the above-
mentioned procedure concerning chlorination of mandelic acid derivatives 15a-c by using
SOCl (Scheme 4).
2
2
23
8

ACCEPTED MANUSCRIPT
2
24
25
2
Scheme 4. Synthesis of diphenylacetic acid chloride 18 and 9H-fluorene-9-carbonyl chloride
20. Reagents and conditions: (x) SOCl (15 equiv.), benzene, 24 h at reflux (under N -
226
227
228
229
2
2
atmosphere), then 24 h at RT; (xi) SOCl (5.94 equiv.), benzene, 6 h at reflux (under N -
2
2
atmosphere).
230
231
232
233
234
235
236
237
238
239
240
241
242
243
Although both starting materials 17 and 19 are similar to each other in terms of chemical
structure, the experiments revealed that they differ significantly in reactivity. In order to
obtain 18 and 20 with reasonable timescale and high yields a slightly modified stoichiometry
had to be adjusted. In the case of diphenylacetic acid (17) huge 15-fold molar excess of SOCl2
and elongated reaction time (up to 48 h) had to be applied, while in the case of 9H-fluorene-9-
carboxylic acid (19) ca. 6 equiv of SOCl was sufficient to obtain 20 in quantative yield after
2
conducting the reaction for 6 h in refluxing benzene.
Next, we focused our efforts on coupling all the previously obtained acids 10a-c or acid
chlorides 16a-c, 18, and 20 with proxyphylline (1) (Scheme 5). The syntheses of first set of
proxyphylline esters 21-25 were accomplished by using basic conditions (Et N) in the
3
presence of DMAP as the catalyst in dry CH Cl at room temperature. The acylation of
2
2
proxyphylline (1) was achieved in fair (45% for 22), moderate (62% for 23), good (70% for
25), very good (87% for 24), and excellent (98% for 21) yields.
9

ACCEPTED MANUSCRIPT
2
44
45
2
Scheme 5. Synthesis of proxyphylline esters 21-28. Reagents and conditions: (xii) 16a-c or 18
or 20 (1.1 equiv.), Et N (1.3 equiv.) and DMAP (cat./0.06 equiv.), CH Cl , 0.5 h at 0–5 °C,
246
247
248
249
3
2
2
then 12 h at RT; (xiii) 10a-c (1.0 equiv.), EDCI·HCl (1.1 equiv.), DMAP (cat./0.4 equiv.),
CH Cl , 0.5 h at 0–5 °C, then 12 h at RT.
2
2
250
251
252
253
254
255
256
257
258
259
260
261
262
263
In turn, esters of benzoylacetic acid derivatives and proxyphylline 26-28 were not obtained
by the above-described procedure due to the fact that β-keto acids enolize extensively in
benzene, so that side reactions of non-regioselective chlorination of enol form might occur
with thionyl chloride, and what is even more harmful for the reaction outcome, enolization of
1,3-dicarbonyl compounds in benzene is a driving force of carbon-carbon bond cleavage via
rapid decarboxylation [16, 24]. However, replacement of benzene with chloroform, which is
other common solvent in SOCl -based chlorinations, would also be unfeasible since, as
2
mentioned previously, these compounds decompose very rapidly in CHCl solution. In
3
addition, thionyl chloride employed in excess generates considerable amounts of hydrogen
chloride, which also strongly promotes elimination of carbon dioxide from benzoylacetic
acids. To overcome this drawback, we turned our attention toward alternative esterification
methods. In this regard, simple esterification based on carbodiimide activators [i.e. N,N′-
dicyclohexylcarbodiimide
(DCC),
N,N′-diisopropylcarbodiimide
(DIC),
N-(3-
dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDCI·HCl) etc.] seemed to be
1
0

ACCEPTED MANUSCRIPT
264
265
266
267
268
269
270
271
272
the methods of choice. Therefore, at first we used DCC coupling reagent as a promoter and
classical conditions for so-called Steglich esterification. Unfortunately, treatment of 1 with the
respective acids 10a-c failed to afford the desired pure esters 26-28 as it turned out that the
removal of the dicyclohexylurea (DCU) by-product was very problematic. A highly stable
possibly-formed complex between DCU and the product molecules (Fig. 1) could not be
resolved either by column chromatography or selective precipitation. However, with the use
of EDCI hydrochloride which generates an urea by-product easily removable from the
reaction medium by water extraction, the respective substrates underwent esterification to
give 26-28 in 26-57% yield, respectively.
2
73
2
74
75
2
Fig 1. Probable stable DCU-(26-28) complexes that prevent preparation of pure esters 26-28.
2
76
277
278
279
280
281
282
283
284
285
286
287
288
289
290
Synthesis of Proxyphylline 1,2,3-Triazolic Ethers 40-42
In order to synthesize new proxyphylline analogues that would exhibit improved dual
anticancer-antifungal activity in vitro, and thus to expand SAR studies, we decided to prepare
compounds with increased enzymatic stability. In this regard, we tried to introduce non-
hydrolyzable linkages into the designed hybrid molecules that would become resistant against
cellular proteolytic degradation when compared to derivatives with ester/amide linkers.
Unfortunately, all the attempts toward coupling the desired molecules using direct C-C, C-N
or C-S bond ended up without success since the nucleophilic substitution (S 2) of the
N
prepared proxyphylline halogen-, tosyl- and triflate- derivatives with the different carbon-,
amine-, and thiol-like nucleophiles revealed to be unobtainable. Under various basic
conditions [(i) Et N/CH Cl , (ii) DIEA/CH Cl , (iii) K CO /DMF, (iv) t-BuOK/DMF, (v)
3
2
2
2
2
2
3
NaH/THF, (vi) NaH/PhCH , (vii) BuLi/THF, (viii) KOH/18-Crown-6/DMSO-H O (15:1),
3
2
and (ix) 50% aq. NaOH/DMSO] all the alkylation reactions failed, to give only product of
elimination [1,3-dimethyl-7-[(1E)-prop-1-en-1-yl]-3,7-dihydro-1H-purine-2,6-dione] or the
1
1

ACCEPTED MANUSCRIPT
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
unreacted, almost quantitatively recovered starting material 1. In addition, the direct
amination of proxyphylline (1) hydroxyl moiety with the nucleophilic amines (including
aniline, naphthalen-1-amine, anthracen-1-amine and morpholine) under Mitsunobu reaction
conditions (Ph P/DEAD/THF) proceeded unsuccessfully as well. This prompted us to connect
3
proxyphylline (1) to aromatic units via stable linkers based on carbon−heteroatom−carbon
bonds, and we decided to use copper(I)-catalyzed azide−alkyne 1,3-dipolar cycloaddition
(CuAAC) reaction. The Cu(I)-mediated Meldal-Sharpless version [25,26] of the Huisgen 1,3-
dipolar cycloaddition methodology serves as powerful, efficient, broad-in-scope, highly
chemo- and regio-selective tool extensively applied for the preparation of valuable
compounds with broad applications [27-34], including medicinal chemistry [35]. Moreover,
the application of 1,2,3-triazole moiety, which is known to be bioisostere of the amide bond,
can be useful from the view-point of biological activity of the obtained products since it is
well-documented that heteroaromatic azole functionalities play a fundamental role in
development of antifungal agents [36,37]. The synthesis of such triazole-linked molecular
hybrids employing CuAAC-promoted ‘click reaction’ requires preparation of building blocks
possessing suitable terminal functional groups, in this case: −N and –alkyn.
3
The azide-like reagent 30 was synthesized by employing a commercially available 1-
iodonaphthalene (29) and 6-fold molar excess of NaN in DMSO/H O (10:1, v/v) mixture,
3
2
catalyzed by CuI/sodium ascorbate (NaAsc)/NaOH (Scheme 6). The introduction of an azide
functional group into naphthalene ring proceeded with 57% yield after the reaction at 70 °C
for 24 h.
1
2

ACCEPTED MANUSCRIPT
3
13
14
3
Scheme 6. Synthesis of azide derivatives 30, 37, 38. Reagents and conditions: (xiv) sodium
315
316
317
318
319
320
ascorbate (0.6 equiv.), NaOH (0.6 equiv.), NaN (6.0 equiv.), DMSO/H O (10:1, v/v), 20 min
3
2
at RT (degassing under Ar-atmosphere), then CuI (0.6 equiv.), 24 h at 70 °C; (xv) 69%
HNO /100% HNO (17:1, v/v), Br (6 equiv.), 1 h at 80 °C, then 48 h at 60 °C; (xvi)
3
3
2
CH CN/MeOH (3:1, v/v), DBU (2.9 equiv.), 5 min at RT, then 1,3-dibromopropane (12.9
3
equiv.), 7 d at RT; (xvii) NaN (1.55 equiv.), DMF, 48 h at RT.
3
321
322
323
324
325
326
327
328
329
330
331
332
333
The subsequent azido compounds 37 and 38 were designed on the basis of our previous
research experience and well-documented anticancer properties of derivatives prepared from
selective protein kinase 2 (CK2) inhibitors, that are: 4,5,6,7-tetrabromo-1H-benzimidazole
(TBBi, 33) and 4,5,6,7-tetrabromo-1H-benzotriazole (TBBt, 34) [38,39]. Moreover, Świder
and co-workers have reported that (TBBt)-derivatives functionalized with 1,2,3-triazole
exhibit very promising multisite-directed CK2 inhibitory activity [40] as well as proapoptotic
activity towards human Jurkat T-leukemia and MCF-7 breast adenocarcinoma cells [41]. In
addition, we have recently published the results of studies on antiproliferative activity in
tumor cells and kinase profile of novel TBBi and TBBt derivatives also possessing N-
substituted 1,2,3-triazole ring [42]. The above-mentioned results encouraged us to continue
the evaluation of 1,2,3-triazolic TBBi/TBBt motif in our anticancer agents study, this time by
using it in the structural modifications of proxyphylline (1). It was motivated not only by the
desire to enlarge the structural diversity of 1 analogues, but also to find if the proxyphylline
1
3

ACCEPTED MANUSCRIPT
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
fragment could enhance the aqueous solubility of polybrominated counterparts in biological
media, and thus increase their in vitro cellular activity. It is worth to point out that i.e.
although TBBi induces apoptosis in cancer cells, it has low solubility. Thereby, to obtain the
azides 37 and 38, the procedure reported by us previously [39, 42] was used with some
modifications involving the reaction work-up (see Experimental section) (Scheme 6). It is
worth mentioning that when TBBi (33) was alkylated by 1,3-dibromopropane a significant
amount of an elimination by-product, namely 4,5,6,7-tetrabromo-1-(prop-2-en-1-yl)-1H-
benzimidazole, was formed. Furthermore, a negligible difference in polarity between 35 and
the above-mentioned contamination excluded efficient purification of the final product.
Fortunately, this problem was successfully resolved by using the crude reaction mixture as a
starting material for the later two steps, and performing the purification procedure only after
obtaining the desired cycloaddition product 41. In turn, under the same alkylation reaction
conditions TBBt (34) was transformed with excellent regioselectivity by using 1,3-
dibromopropane into exclusively N-2-substituted derivative 36 as the major isomer, and
without formation of by-product of an elimination reaction what gave opportunity of minimal
work-up and purification. Afterwards, both compounds 37 and 38 were prepared by using
1.55 equiv of NaN as azide anion donor in DMF under mild reaction conditions at room
3
temperature over 2 days. After routine work-up an azido derivatives 37 and 38 were further
used without additional purification.
The second component that possesses terminal alkyne moiety required for the CuAAC
reaction with the respective azides 30, 37, 38 was prepared as shown in Scheme 7.
1
4

ACCEPTED MANUSCRIPT
3
56
57
3
Scheme 7. Synthesis of proxyphylline 1,2,3-triazolic ethers 40-42. Reagents and conditions:
(xviii) NaH (2 equiv., as 60% w/w dispersion in mineral oil), dry THF, 10 min. at 0–5 °C,
then 30 min at RT (under Ar-atmosphere), then TBAI (0.1 equiv.), propargyl bromide (2
358
359
360
361
362
363
364
365
equiv., as 80 % w/w solution in PhCH ), 20 min. at 0–5 °C, then 2 h at RT; (xix) 30 (1.1
3
equiv.), CuSO ·5H O (0.01 equiv.), sodium ascorbate (0.02 equiv.), CH CO H (0.1 equiv.),
4
2
3
2
tert-BuOH/H O (1:2, v/v), 10 min. at RT; (xx) 37 or 38 (1.1 equiv.), CuSO ·5H O (0.1
2
4
2
equiv.), sodium ascorbate (0.2 equiv.), CH CO H (0.2 equiv.), DMF/tert-BuOH/H O (5:1:2,
3
2
2
v/v), 30 min. at 50 °C.
366
367
368
369
370
371
372
373
374
375
376
377
The propargylation of 1 was accomplished according to procedure reported by Martín-
Rapún et al. [43]. The (NaH/TBAI)-promoted conditions provided alkyne-like proxyphylline
derivative 39 in very good yield (75 %) after almost 3 h. Afterwards we turned our attention
toward connection of both synthons into 1,2,3-triazolic ethers 40-42. From a huge repertoire
of published copper(I)-catalyzed azide-alkyne cycloaddition procedures, we decided to chose
highly efficient modified Straub protocol [44] proposed by Shao and co-workers [45], which
assumes utilization of standard catalytic system composed of CuSO ·5H O/NaAsc in neutral
4
2
aqueous tert-BuOH solution and an additive of various carboxylic acids (i.e. HOAc, tert-
BuCO H, and PhCO H) as ultimately potent CuAAC-promoters. Initially, we have coupled
2
2
39 with 1-azidonaphthalene 30. Although Shao and co-workers’ reported that 0.2 equiv of
PhCO H as a proton source gave the best result of acceleration of the substrates’ conversion,
2
in our case, a remarkably high reactivity was successfully achieved by using 0.1 equiv of
1
5

ACCEPTED MANUSCRIPT
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
cheaper HOAc. The 1,2,3-triazolic cycloaddition product 40 was obtained in an excellent
yield (81%) within less than 10 min at room temperature.
Unfortunately, for substrates 37 and 38 the employed conditions for Cu-catalyzed clicking
gave significantly worst reaction rates and yields. Therefore, we have studied other Shao’s
acid-base (CuI/DIPEA/HOAc) promoted CuAAC conditions [46], in which instead of tert-
BuOH/H O solvent system CH Cl was used. The conducted experiments revealed that low
2
2
2
reactivity was mostly related with very poor solubility of polybrominated azido compounds
37 and 38 in both reaction media [tert-BuOH/H O (1:2, v/v) and CH Cl ]. In order to
2
2
2
overcome the encountered drawbacks, an addition of DMF to a reaction mixture as well as an
increase of the amounts of CuSO ·5H O and sodium ascorbate by one order of magnitude,
4
2
and HOAc up to 0.2 equiv had to be applied. Moreover, to boost the formation of the desired
products, we have also changed the reaction temperature from an ambient to 50°C. Using this
improved method, hybrid molecules 41 and 42 were synthesized in a relative short time (30
min.) in 73% and 69% yield, respectively.
Antitumor Activity
The in vitro cytotoxic activity of 16 newly synthesized proxyphylline derivatives was
determined by the MTT assay after 72 h of exposure of CCRF-CEM (human acute
lymphoblastic leukemia) and MCF-7 (human breast adenocarcinoma) cell lines to each analog
at 25, 50, 100 and 200 µM concentration, respectively (Table 1). Interesting results and SAR
data emerged from this screening.
Table 1. Effects of proxyphylline analogues on viability of the human leukemia cancer
(CCRF-CEM) and breast cancer (MCF-7) cells after 72 h treatment (by MTT assay).
a
Compound
Cell viability% ± SD
CCRF-CEM
MCF-7
2
5 µM
78.9 ± 10.7 73.1 ± 4.3
52.4 ± 0.4 39.3 ± 3.4
105.2 ± 5.8 77.4 ± 10.5
22.1 ± 3.1 7.9 ± 2.0
53.9 ± 14.1 37.1 ± 9.0
50 µM
100 µM
200 µM
25 µM
50 µM
100 µM
200 µM
3
4
5
6
7
53.5 ± 5.7
23.9 ± 0.8
38.5 ± 2.7
23.4 ± 1.1
91.9 ± 6.7
93.4 ± 7.3
90.5 ± 7.2
58.8 ± 5.2
86.8 ± 2.7
89.5 ± 6.5
83.6 ± 2.1
98.8 ± 3.9
89.9 ± 6.3
38.3 ± 4.3
81.9 ± 2.1
75.2 ± 3.8
84.7 ± 4.1
82.5 ± 0.9
73.4 ± 4.6
27.1 ± 0.9
69.9 ± 5.1
67.3 ± 6.3
63.5 ± 4.2
81.0 ± 6.1
76.5 ± 1.3
37.0 ± 6.4
20.8 ± 4.3
43.1 ± 5.3
43.5 ± 2.1
28.3 ± 3.6
33.8 ± 19.0 47.1 ± 4.8
0.0 ± 0.0
14.5 ± 1.4
30.5 ± 9.5
0.1 ± 0.0
0.0 ± 0.0
0.5 ± 1.3
2.5 ± 0.6
2.3 ± 1.2
2
2
1
2
84.8 ± 0.6
41.5 ± 3.8
33.1 ± 2.6
2.8 ± 0.5
101.2 ± 12.3 85.1 ± 6.5
1
6

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
4
4
4
3
3
75.8 ± 1.2
84.7 ± 41.4 28.2 ± 0.6
96.6 ± 4.4 87.9 ± 8.3
85.1 ± 15.3 88.1 ± 5.4
55.3 ± 6.3
28.6 ± 4.0
27.6 ± 2.7
0.0 ± 0.0
82.2 ± 0.2
93.5 ± 3.7
92.6 ± 2.7
91.3 ± 6.0
88.5 ± 5.4
89.0 ± 10.0
93.5 ± 9.4
74.5 ± 6.7
58.3 ± 3.3
19.6 ± 4.3
79.5 ± 2.9
86.3 ± 3.4
87.7 ± 2.9
79.8 ± 9.4
77.9 ± 8.6
82.2 ± 2.9
91.9 ± 3.2
29.7 ± 5.4
59.8 ± 2.1
24.6 ± 2.0
63.5 ± 1.3
83.7 ± 4.5
83.3 ± 5.0
79.3 ± 3.9
83.8 ± 2.5
86.6 ± 2.1
93.3 ± 4.9
2.0 ± 1.8
35.0 ± 1.0
68.8 ± 6.3
76.9 ± 2.1
81.8 ± 5.1
76.8 ± 6.6
82.7 ± 5.2
77.0 ± 7.0
4.1 ± 1.6
4
15.0 ± 1.4
5
51.4 ± 10.5 34.6 ± 6.9
65.6 ± 11.1 38.5 ± 10.9
6
7
83.5 ± 3.2
89.5 ± 2.8
90.0 ± 4.6
0.0 ± 0.0
0.0 ± 0.0
6.3 ± 4.5
51.1 ± 21.4
74.2 ± 13.7
67.7 ± 5.0
0.0 ± 0.0
43.8 ± 7.2
67.3 ± 4.7
34.2 ± 5.2
0.0 ± 0.0
0.0 ± 0.0
0.0 ± 0.0
20.1 ± 6.9
31.6 ± 11.6
4.1 ± 1.9
0.0 ± 0.0
0.0 ± 0.0
0.0 ± 0.0
8
0
1
2
0.0 ± 0.0
59.1 ± 2.6
9.0 ± 1.4
53.7 ± 6.4
1.7 ± 0.4
3 (TBBi)
0.0 ± 0.0
a
4
4
4
02
03
04
Standard deviation; the results are expressed in percentage of cell death relative to control (cells without
inhibitor in 0.5% DMSO). Experiments were performed in triplicate and the % of dead cells are mean values of
three independent experiments.
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
Shortly, the MTT cell viability test indicated that most of the synthesized compounds
displayed moderate-to-high cytotoxic activity against CCRF-CEM cells, among which
compounds 41 and 42 exhibited the highest antineoplastic activity causing a complete loss of
the cell viability at all of four examined dose concentrations. With regard to those derivatives
it seems clear that the presence of heterocyclic polybrominated fragment is fundamental for
improvement of anticancer potency since in the series of proxyphylline-1,2,3-triazole-
aromatic hybrids 40-42, derivative with the naphthalene substituent 40 was found to be
significantly less active. It has to be pointed out that compounds 41 and 42 exhibited higher
cytotoxicity than parent TBBi (33) and TBBt (34) molecules (data not shown) [39, 42]. In
turn, compound 6 bearing a phenothiazine moiety with alkyl-carbonate linker also showed
good activity in cellular assay, leading to complete death of the leukemia cancer cells at the
100 µM concentration treatment. In the case of CCRF-CEM cell line, also worth attention are
compounds 4 and 22, both containing bulky tricyclic aromatic substituent, that effectively
inhibited cell growth at the 25 µM concentration with nearly similar values to their EC . The
50
cytotoxic effect of compounds 4, 6 and 22 on leukemia cells seems to be not related with
hypothetical products of cellular hydrolysis since the related components, including 5H-
dibenzo[b,f]azepine, 1-(10H-phenothiazin-10-yl)propan-2-ol and fluorene were completely
inactive. Cytotoxic activity against CCRF-CEM cells was also demonstrated by thioester
derivative 7 (ca. 86% cell death at 100 µM). In this case, the biological data indicated that
simple replacement of the sulfur atom with nitrogen cause a loss of activity in respective
amide 3. Similarly, the introduction of free rotation along the bonds engaged in the formation
of rigid fluorene ring was found detrimental for antitumor effectiveness of compound 21 as
1
7

ACCEPTED MANUSCRIPT
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
its’ antiproliferative activity drop to a half of the value represented by 22. On the other hand,
MCF-7 cell line reveals higher resistance for all the new proxyphylline analogues. Beside
compounds 6, 41 and 42 which exerted a moderate effect on MCF-7 cells viability (58.3-
74.5% at 25 µM concentration), the rest of the examined derivatives showed negligible
cytotoxic activity, and thus higher compound concentrations were necessary to observe
significant decrease of the cell growth. It is somehow interesting how presence of TBBi-
substituent highly potentiates the anticancer activity of 41 when compared to TBBt moiety in
42 in the concentration range of 50–200 µM. It is in contradiction to the results obtained for
MTT assay conducted in CCRF-CEM cell culture, where there was only a little or no
difference in activity between the TBBt and TBBi substitution in the synthesized
proxyphylline derivatives. The other unsettling results concerned linear cytotoxic activity of
compound 42, regardless of its’ concentration. The lack of any difference between the
cytotoxic behaviour of the tested derivative 42 indicates that it might be related within
different transmembrane active transport mechanism in MCF-7 cells. In this regard, the
maximum level of receptors’ occupation (saturation) causes the limitations of overloaded
compound activity as it is likely not further transferred across cellular membranes.
Nevertheless, more in-depth cytotoxicity evaluations against MCF-7 cells were omitted from
further biological evaluation as the results were unsatisfactory for the examined proxyphylline
derivatives.
After first swift screening of the compounds, the MTT test was assessed again, this time for
detail evaluation of half-maximal effective concentration values (EC ) of the most potent
5
0
compounds 41 and 42. The EC index was used to express the concentration of compound
5
0
which kills 50% of the tumor cells in vitro in comparison to control culture, that is cells
incubated without inhibitor in 0.5% DMSO. The demonstrated negative effect on CCRF-CEM
cells viability at low micromolar levels was confirmed with an EC value of 6.3 µM for
5
0
compound 41 and 6.5 µM for compound 42. In turn, for MCF-7 cell culture the best EC50
value reaching approximately 80 µM was obtained for both compounds 6 and 42.
Candidacidal Activity of Proxyphylline Derivatives
Our initial screen of 16 proxyphylline derivatives using M27-A3 [47] found the most
effective C. albicans inhibitors as presented in extensive Table S1 (Supporting Information).
Sixteen-content assays identified the leading fungicidal proxyphylline derivatives (Table 2) to
be used in a series of follow-up studies to establish their action mode and to examine their in
1
8

ACCEPTED MANUSCRIPT
vitro and in vivo activity. Since 6 and 42 manifested the widest antifungal activity at the
concentrations of 15.5 and 5 µM respectively, they were selected to identify their action
mode.
460
461
462
463
464
465
Table 2. Proxyphylline derivatives at the highest concentration tested against C. albicans
90028 in the MFC assay [48].
a
b
c
Proxyphylline
derivatives
Concentrations
Proxyphylline
Growth
Logarithm
(µM)
treated cells
control cells
reduction (R)
5
7
(
CFU × 10 )
427
0
(CFU × 10 )
3
45
35
34
31
43
37
37
36
37
37
40
40
40
36
36
20
0.6
8
d
4
5
425
0
0.3
8
d
6
7
616
632
146
482
315
255
167
109
250
240
0
0.1
0.1
0.8
0.2
0.4
0.5
0.7
0.9
0.5
0.5
8
2
2
2
2
2
2
2
2
4
1
2
3
4
5
6
7
8
0
8
d
4
1
2
d
4
0
8
a
4
4
4
4
4
4
4
4
4
66
67
68
69
70
71
72
73
74
Legend: Each concentration is espressed as molar quantity (µM) corresponding to conc. of 16 µg/mL prepared
b
according to M27-A3 [47] as described in Supporting Information; CFU of C. albicans after treatment with
c
proxyphylline derivative at proper concentrations; CFU of untreated inoculum of C. albicans. Decrease is
expressed as decimal log reduction using the formula: lg R = lg CFU/mL control cells – lg CFU/mL compound
treated cells. Proxyphylline derivative reducing C. albicans cells at least for lg R≥3 was defined as fungicidal,
d
data represent mean from two independent experiments (nꢀ=ꢀ2), leading compounds displaying fungicidal
activity against 90028 at the highest conc. tested, with 4 and 41 at 2-fold lower conc. (i.e., 17 and 16 µM
resectively) showing R<3 (data not shown), whereas 6 and 42 displaying extended fungicidal activity to 15 and 5
µM (R>3), respectively (data not shown).
4
4
75
76
1
9

ACCEPTED MANUSCRIPT
477
478
479
480
481
482
483
484
485
486
487
488
Biofilm-Inhibitory Activity of the Leading Proxyphylline Derivatives 6 and 42 In Vitro
Figure 2 shows the detailed results for the concentration-dependent inhibitory effects on C.
albicans biofilm formation for the leading compounds 6 and 42. The latter was more effective
against the mature biofilm (almost 50%-inhibition of cell respiration). The potency of 6 and
42 was measured respectively, and IC was determined (Fig. 2). The results indicated that the
5
0
created biofilms were mostly recalcitrant to treatment with 42’s concentrations up to 20 µM
(IC ) as well as 6’s concentrations up to 248 µM (IC ). On the basis of these characteristics
5
0
50
and our prediction that the concentrations of 20 and 248 µM for 42 and 6 respectively can be
effective against candidiosis in vivo, we selected them as the most promising for further
characterization. This study was extended by microscopy with fluorescent detection of the cell
death under 6 and 42.
4
89
90
4
Fig. 2. In vitro characterization of the leading proxyphylline derivatives’ impact on C.
albicans biofilm. Ability of proxyphyline derivatives 6 and 42 to inhibit biofilm formation
was determined using the microtiter plate method described previously [49]; the experiments
were conducted in duplicate on different days; results are shown as mean percent biofilm
inhibition compared with untreated control biofilms using the XTT assay; bars show standard
deviations (SDs); % cytotoxicity was calculated according to formula: % = (Tested compound
– Untreated biofilm)/(Blank – Untreated biofilm) × 100, where: Blank means RPMI and
Untreated biofilm means 24-h biofilm grown at well bottom in RPMI; 6’s and 42’s
491
492
493
494
495
496
497
2
0

ACCEPTED MANUSCRIPT
498
499
500
501
concentration-dependent inhibitory effect on C. albicans biofilm formation was observed; *
42 at 20 µM or ** 6 at 248 µM inhibited biofilm respectively at 47% (IC not significant at
5
0
p≤0.05) or 61% (IC ; not significant at p≤0.05).
5
0
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
Mature Biofilms Prone to Treatment with 6 and 42
The confocal laser scanning microscopy results confirmed the C. albicans biofilm
metabolic decrease under 6 and 42 at IC = 248 µM and IC = 20 µM, respectively. It was
5
0
50
shown that the anti-biofilm activity of 6 and 42 is caused by the ability of these compounds to
disorganize and penetrate the cell membrane. Other cellular changes, such as cell wall
rearrangements under these compounds were observed as below-described. The various
fluorescence staining techniques of the mature biofilm cells treated with 6 or 42 at the
concentration of 2 × IC (data not shown) revealed comparable fluorescence distribution
5
0
within the biofilm cells treated with 248 or 20 µM, respectively. When treated with 42, a
negatively charged fluorochrome (CFW) showed a distinctly higher affinity to chitin localized
in septal junctions (true and breakable septa) and blastoconidial scares, corresponding to the
expected chitin distribution comparable to the untreated control biofilm (Fig. 3 A vs E). On
the other hand, the slight increase in chitin fluorescence around the sessile cells was observed
in the biofilm treated with 6 also at the concentration of 248 µM vs the untreated biofilm
growth (Fig. 3 C vs E). To further characterize the phenomenon of changes in the chitin
content under 6 and 42, their chitinolytic activity and staining index (SI) solely for the potent
cell wall inhibitor, i.e. 42 at two concentrations, were studied against C. albicans planktonic
cells whose results are described below.
2
1

ACCEPTED MANUSCRIPT
5
20
21
5
Fig. 3. Chitin distribution in C. albicans under 6 or 42 treatment. Confocal laser scanning
microscope; fluorescence (A, C, E) and overlaid fluorescence and differential interference
contrast (DIC) images (B, D, F) representing C. albicans biofilm cells stained with calcofluor
white (CFW) at 10 mg/mL after treatment with proxyphylline derivatives: 42 and 6. Cells
treated with 42 (A, B) and 6 (C, D) as well as untreated cells (E, F) displayed different chitin
distribution over cell wall, as follows: (A, B) 42-treated sessile cells at conc. of 20 µM stained
522
523
524
525
526
2
2

ACCEPTED MANUSCRIPT
527
528
529
530
531
with CFW, resulting in bright bluish labeled septal junction and blastoconidial poles
(expressing biofilm viability) comparable with (E, F) untreated control sessile cells. (C, D)
CFW staining revealed considerable changes in chitin content of 6-treated cells at 248 µM in
comparison with control cells (E, F). Scale bars represent 5 µm.
532
533
534
535
536
537
538
539
The Congo red fluorescence of the 42-treated biofilm showed β-glucan uniform distribution
compared with the regular cells (Fig. 4 A vs E). Strikingly, the 20 µM-treated cells showed
labelling around the cell wall, suggesting this polysaccharide synthesis to be a compensatory
cell mechanism (cell adaptation). In contrast, the 6-treated sessile cells similarly as the
untreated control biofilm (Fig. 4 C vs E), indicated the physiological polysaccharides
distribution (a weak signal was detected using this approach). In the control biofilm (intact
cells), β-glucan is localized in the cell wall’s inner layer inaccessible to the Congo red
staining.
2
3

ACCEPTED MANUSCRIPT
5
40
41
5
Fig. 4. Content of β-glucan in C. albicans under 6 or 42 treatment. Confocal laser scanning
microscope; fluorescence (A, C, E) and overlaid fluorescence and differential interference
contrast (DIC) images (B, D, F) depicted β-glucan’s exposure to staining with 2% Congo Red
of biofilm cells after treatment with 42 and 6. (A, B) Unmasked β-glucan in cells treated with
42 at 20 µM. Clear red labelling in 42-treated cells was visible vs a weak fluorescent signal
detected in untreated counterparts (E, F). (C, D) In cells treated with 6 at 248 µM, a staining
542
543
544
545
546
2
4

ACCEPTED MANUSCRIPT
pattern similar to control cells was noted (β-glucan inaccessible to Congo red dying). Scale
bars represent 5 µm.
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
The staining procedure employing the metachromatic acridine orange (AO) and the other
intercalating ethidium bromide (EB) dye in the unfixed biofilm cells indicated the metabolic
changed state of cells under the hypotonic treatment with 6 or 42. The differences noted
between the 6- or 42-treated biofilms compared to the untreated biofilm were due to lysis of
the cell membrane. The 6- or 42-treated cells at 248 or 20 µM respectively displayed early
apoptotic phenotypes and morphological changes characteristic of late apoptosis rarely
occurred. It was noted that the mature biofilms’ programmed cell death is varied in form
depending on the damage severity. Under 42, dual AO and EB staining identified an early cell
apoptosis being in the process of disintegrating (50%-reduction of dehydrogenases in the
process of sessile cell respiration according to XTT) compared with the negative control
biofilm (Fig. 5 A vs E). Treatment with 6 produced early- and late-apoptotic cells (61% of
sessile cells were metabolically impaired according to XTT), with uneven reddish-orange
fluorescence on their periphery. The fluorescence microscopic examination of the 6- or 42-
treated biofilm cells revealed the apoptotic cells at least twice as intense in their green
fluorescence vs the untreated biofilm with faint green cytoplasm and nuclei with bright green
fluorescence. A distinct population of the early apoptotic cells (Fig. 5 A and C; see open
arrows) displayed an extremely packed cellular mass at one side of the cells (polarized
distribution), and typical apoptotic events such as nuclear condensation, while their nuclei
were only weakly stained with EB compared with the late apoptotic cells with brightly EB-
stained nuclei (reddish-orange). Under the 42- or 6-treatment (Fig. 5 A-D, arrowhead)
comparable to the untreated control biofilm conditions (Fig. 5 E and F) a subpopulation of
viable cells showing faint green fluorescence (live cells with masked DNA) was observed
(merged images B, D, F exemplified unclear staining with AO; note arrowheads vs green
nucleus, see arrow). Contrariwise, the late apoptotic/necrotic cells tend to lose the proton
pomp function resulting in AO reddish-orange fluorescence. Moreover, the live cells with
cytoplasmic membrane integrity did not uptake EB (lack of red fluorescence, Fig. 5 B, D, F).
The above described features (in live and necrotic populations) were quantified by the XTT
assay, as described above.
2
5

ACCEPTED MANUSCRIPT
5
5
78
79
Fig. 5. Morphological and staining characteristics of unfixed biofilms made permeable
through the treatment with 6 and 42 prior to dual staining AO and EB in situ. Extrinsic
apoptotic morphological changes in 6- or 42-treated C. albicans cells after dual AO and EB
staining. Confocal laser scanning microscope; fluorescence (A, C, E) and overlaid
fluorescence and differential interference contrast (DIC) images (B, D, F) showing fungal
membrane permeabilization to EB. Dual staining with AO (penetrating intact membrane) and
with cell-membrane impermeable EB dye (each at 100 µg/mL) of unfixed biofilm cells in situ
5
5
5
5
5
5
80
81
82
83
84
85
2
6

ACCEPTED MANUSCRIPT
586
587
588
589
590
591
592
593
594
595
596
597
598
599
showed evidence of cell-membrane disruption under 42 (A, B) and 6 (C, D) vs untreated cells.
(A, B) Analysis of cellular staining pattern of 42-treated biofilm revealed early-stage
apoptotic cells marked by a crescent-shaped AO nuclear staining (chromatin aggregation; see
open arrow). Note higher light green fluorescence intensity compared with untreated biofilm.
(C, D) Early apoptotic cells with bright-green fluorescence of nuclei with AO (crescent or
dense green area of fragmented chromatin located on one side of cells; see open arrow). A
population of cells was seen at elevated reddish-orange fluorescence distribution suggesting
late apoptotic cells. These cells appearing reddish-orange on their periphery with unapparent
outline (near disintegration) were noted. Note increase in bright green or reddish-orange
fluorescence intensity vs untreated biofilm. (E, F) Negative control biofilm group – normal
cells with green fluorescence nuclei (fluorescent spots). (B, D, F) Merged fluorescence and
DIC exemplified viable EB (–) and AO (+) cell types with green nuclei (arrow) as well as
poorly green fluorescence nuclei (arrowhead). Scale bars represent 5 µm.
600
601
602
603
604
605
606
607
608
609
Flow Cytometry – Planktonic Cell Death
The biofilm cell-death assessed with fluorescent microscopy differed from the results
attained for planktonic phenotypes using flow cytometry under the compounds as described
below. Importantly, an essential majority of all cells that responded respectively to 6 or 42
without Annexin-V binding were propidium iodide (PI) positive at 95.9%±0.26 or
93.94%±1.11 (necrotic populations in Table 3). The dead cells unable to exclude PI indicated
the onset of secondary necrosis. The necrotic cells did not resuscitate after 48 h incubation on
the YEPD agar medium. PI negative cells (early apoptotic cells) were observed at 1.2%±0.17
under 6 or 2.87%±0.37 under 42.
6
6
10
11
Table 3. Detection of apoptosis-like programmed cell death under 6 and 42 by concurrent
staining with Annexin-V and PI.
b
Proxyphylline
Protoplasts (%)
Early apoptosis
a
derivatives
Live
Late apoptosis
Necrosis
or Control
6
4
2.80±0.30
2.87±0.99
96.38±2.05
1.2±0.17
2.87±0.33
3.00±1.65
0.1±0.1
95.9±0.26
2
0.22±0.14
0.3±0.30
93.94±2.05
0.33±0.20
Control
2
7

ACCEPTED MANUSCRIPT
a
b
6
6
6
6
6
6
6
12
13
14
15
16
17
18
Legend: 6 at IC = 248 µM or 42 at IC50 = 20 µM; Data are mean percentages ± SDs; results between
5
0
untreated control and 42-treated or 6-treated cells were respectively not significant or significant at p≤0.05; 6- or
42-treated cells as well as untreated cells (Control) were subsequently stained with Annexin-V-FITC conjugated
and PI, and their far-red or red fluorescence was measured with flow cytometry. Live cells were both Annexin-V
(–) and PI (–); at early stage of apoptosis cells bound Annexin-V and excluded PI giving the following staining
pattern: Annexin-V (+) and PI (–); at late stage of apoptosis they bound Annexin-V and were PI positive, i.e.,
Annexin-V (+) and PI (+); necrotic cells showed as follows: Annexin-V (–) and PI (+).
6
19
620
621
622
623
624
625
Effect of Proxyphylline Derivatives 6 and 42 Against the Fungal Cell Wall
As the specific fungal cell wall inhibitors showing a reversed effect in a medium containing
an osmotic protector i.e., sorbitol are their distinctive feature, we indicated that antifungal
properties of 42 are related to the biosynthetic pathways of the cell wall because MFC was
increased two-fold (10 µM with lg R = 6.74, Fig. 6) vs in-medium without sorbitol (MFC = 5
µM with lg R = 6.81, Fig. 6).
6
26
2
8