Synthetic analogues of cyanobacterial alkaloid cylindrospermopsin and their toxicological activity

Article abstract of DOI:10.1016/j.tiv.2017.07.007

Cylindrospermopsin (CYN) is a naturally occurring alkaloid produced by a variety of cyanobacteria and known to induce oxidative stress-mediated toxicity in eukaryotic cells. Despite extensive research on the mechanism of CYN toxicity, an understanding of the structural features responsible for this toxicity and the mechanism by which it can enter the cell are still not clear. It was established that the presence of both the uracil and guanidine groups is essential in biological activity of CYN whilst not much is known in this regard on the role of tether that separates them and the attached hydroxyl group. Therefore, in the present study we have prepared three synthetic analogues possessing uracil and guanidine groups separated by a variable length tether (4–6 carbons) and containing a hydroxyl function in a position orientation to CYN, together with a tetracyclic analogue of CYN lacking the hydroxyl group at C-7. The toxicity of these compounds was then compared with CYN and guanidinoacetate (GAA; the primary substrate in CYN biosynthesis) in an in vitro model using human neutrophils isolated from healthy subjects. The lowest activity measured by means of reactive oxygen species generation, lipid peroxidation and cell death was observed for GAA and the tetracyclic analogue. The greatest toxicity was found in an analogue with a 6-carbon tether, but all three analogues and CYN caused rapid onset of redox imbalance. These results add to the general understanding of CYN toxicity and preliminary findings suggest that the –OH group at C-7 may be significant for the cellular transport of CYN and/or be involved in its toxic activity inside the cell, a hypothesis which requires further testing.

Full text of DOI:10.1016/j.tiv.2017.07.007

Accepted Manuscript
Synthetic
analogues
of
cyanobacterial
alkaloid
cylindrospermopsin and their toxicological activity
Christopher Cartmell, Daniel M. Evans, Jessica M.L. Elwood,
Hisham S. Fituri, Patrick J. Murphy, Thomas Caspari, Barbara
Poniedziałek, Piotr Rzymski
PII:
S0887-2333(17)30193-5
doi: 10.1016/j.tiv.2017.07.007
TIV 4056
DOI:
Reference:
To appear in:
Toxicology in Vitro
Received date:
Revised date:
Accepted date:
18 April 2017
19 June 2017
7 July 2017
Please cite this article as: Christopher Cartmell, Daniel M. Evans, Jessica M.L. Elwood,
Hisham S. Fituri, Patrick J. Murphy, Thomas Caspari, Barbara Poniedziałek, Piotr
Rzymski , Synthetic analogues of cyanobacterial alkaloid cylindrospermopsin and their
toxicological activity, Toxicology in Vitro (2017), doi: 10.1016/j.tiv.2017.07.007
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
1
Synthetic analogues of cyanobacterial alkaloid cylindrospermopsin and their
toxicological activity
Christopher Cartmell, Daniel M. Evans, Jessica M. L. Elwood, Hisham S. Fituri, Patrick J.
Murphy*
chs027@bangor.ac.uk
School of Chemistry, Bangor University, Bangor, Gwynedd, LL57 2UW, UK
Thomas Caspari
t.caspari@bangor.ac.uk
School of Medical Sciences, Bangor University, Bangor, Gwynedd, LL57 2UW, UK
Barbara Poniedziałek, Piotr Rzymski
Department of Environmental Medicine, Poznan University of Medical Sciences, Poznań,
Poland
bpon@ump.edu.pl
Abstract: Cylindrospermopsin (CYN) is a naturally occurring alkaloid produced by a variety of
cyanobacteria and known to induce oxidative stress-mediated toxicity in eukaryotic cells.
Despite extensive research on the mechanism of CYN toxicity, an understanding of the
structural features responsible for this toxicity and the mechanism by which it can enter the cell
are still not clear. It was established that the presence of both the uracil and guanidine groups is
essential in biological activity of CYN while not much is known in this regard on the role of tether
that separates them and the attached hydroxyl group. Therefore, in the present study we have
prepared three synthetic analogues possessing uracil and guanidine groups separated by a
variable length tether (4-6 carbons) and containing a hydroxyl function in a position orientation
to CYN, together with a tetracyclic analogue of CYN lacking the hydroxyl group at C-7. The
toxicity of these compounds was then compared with CYN and guanidinoacetate (GAA; the
primary substrate in CYN biosynthesis) in an in vitro model using human neutrophils isolated
from healthy subjects. The lowest activity measured by means of reactive oxygen species
generation, lipid peroxidation and cell death was observed for GAA and the tetracyclic
analogue. The greatest toxicity was found in an analogue with a 6-carbon tether, but all three
analogues and CYN caused rapid onset of redox imbalance. These results add to the general
understanding of CYN toxicity and preliminary suggest that the -OH group at C-7 may be
significant for the cellular transport of CYN and/or be involved in its toxic activity inside the cell,
a hypothesis which requires further testing.

ACCEPTED MANUSCRIPT
2
Keywords Cylindrospermopsin; guanidine synthesis, neutrophils; oxidative stress; reactive
oxygen species (ROS), superoxide dismutase (SOD), catalase (CAT). glutathione peroxidase
(GPx).
Introduction
Cylindrospermopsin (CYN) 1, is a naturally occurring cyanotoxin originating from the
1
cyanobacteria Cylindrospermopsis raciborskii. This was the first member of a small group of
related alkaloids to be isolated, along with 7-epi-cylindrospermopsin 2, 7-deoxy-
cylindrospermopsin 3, 7-deoxy-desulfo-cylindrospermopsin
4
and 7-deoxy-desulfo-12-
2-4
acetylcylindrospermopsin 5 from this and other cyanobacteria species. In 1979 a bloom of C.
raciborskii occurred and contaminated a local water supply and was thought to be responsible
5
for an outbreak of hepatoenteritis in 1979. The outbreak, coined the “Palm Island mystery
5
disease” was responsible for the hospitalization of 148 people, the vast majority children. The
highly toxic potencies of CYN in different types of mammalian cells were later extensively
6
demonstrated. Over ten freshwater species are currently known to be able to produce 1
including Aphanizomenon ovalisporum, Raphidiopsis curvata and terrestrial strain of
7,8,9,10
Hormoscilla pringsheimii (Figure 1).
The reasons behind its production are yet to be fully
elucidated though the toxin is known to be actively released from intact cells, to up-regulate
11
alkaline phosphatase in sympatric phytoplankton and to contribute to allelopathic interactions.
F
igure 1: The cylindrospermopsin alkaloids
Previous studies have attempted to link the activity of CYN 1 to specific functional groups
within the molecule, most notably a study by Sukenik who proposed that the uracil group was
partially responsible for its potent toxic activity, reasoning that there might be competitive or
12
inhibitory binding to a catalytic site. Since this study, there have been only a small number of

ACCEPTED MANUSCRIPT
3
reports on the synthesis of analogues of the cylindrospermopsin family, most probably due to
13
their structural complexity. However, Runnegar et al.
conducted a structure activity
relationship investigation (SAR) utilising a series of simplified synthetic analogues of the natural
products. Within this work the unsulfonated analogues ±6 and 7 were studied along with tricyclic
analogue ±8. The simplified uracil analogue ±9 was also investigated, as was the bicyclic
13
sulfonate analogue ±10, which lacks the uracil group (Figure 2).
Figure 2: Cylindrospermopsin analogues
Initially, the toxicity of synthetic racemic CYN ±1 was compared to that of the
stereoisomer 7-epi-cylindrospermopsin 2. The results obtained provided evidence to state that
the exact stereochemistry of the hydroxyl group located at C-7 had little effect on the biological
activity or transport of the toxin. Furthermore, a comparison of the activity of ±1 and 2 alongside
the corresponding diols ±6 and 7 indicated that ±1 and ±6 both inhibit protein synthesis with
almost identical IC values of 0.20 and 0.21 μM, whilst 7-epi-cylindrospermopsin 2 and diol ±6
50
each depleted cell antioxidant glutathione by similar amounts. These results indicate that the
sulfate group present within majority of the CYN alkaloids plays no appreciable role in either
their biological activity or cellular uptake. Of the other compounds studied, the tricyclic analogue
±
8 demonstrated an inhibitory effect on protein synthesis but only at concentrations between
00-1000 times greater than those of 1. Similarly analogues ±9 which lacks the guanidine
5
functionality, and ±10 which lacks the uracil ring displayed no biological activity at
13
concentrations between 800 and 2000 μM. From these results it was concluded, that the
presence of both the uracil and guanidine groups are essential and that the guanidine must be
contained in a lipophilic environment to enable any biological activity. Shaw also investigated
the role of the hydroxyl group located at C-7. This was done by analysing the effect of 7-deoxy-
cylindrospermopsin 3 in four different mammalian cell lines, the results demonstrated that 3

ACCEPTED MANUSCRIPT
4
displays the same levels of toxicity as CYN 1. The levels of protein synthesis inhibition were
14
also similar, with IC values of 340 and 220 nM for 3 and 1 respectively. Williams et al. also
50
determined that both toxins inhibit protein synthesis within one order of magnitude of each other
15
and that both have a similar inhibitory effect on cell glutathione. However, in vivo comparison
of both compounds revealed that 3 did not produce any toxic effects at levels exceeding the
3
lethal concentrations established for 1. Recently it was suggested that exposure to CYN at
environmentally relevant concentrations, has the potential to lead to a reduced ability to fight
16,17
pathogens and therefore increasing susceptibility to potential infections.
Further work in this
area has sought to elucidate the sequence of events underlying CYN induced oxidative stress in
human cells. The toxin has been shown to significantly increase levels of intracellular reactive
oxygen species (ROS) and significantly suppresses the activation of superoxide dismutase
18
(SOD) and catalase (CAT) whilst up-regulating glutathione peroxidase (GPx).
It is apparent that it is still unclear as to role of the uracil group in these alkaloids,
particularly as to its role in transportation into the cell or if it also plays a key role in the in vitro
activity. It is however apparent the sulfate group does not play a significant role in either the
biological activity or the transport of the CYN alkaloids, but that the hydroxyl group does play a
13,14
vital role, but the hydroxyl-stereochemistry does not seem to be a key factor in this activity.
With these observations in mind, we were motivated to prepare a series of analogues of 1 which
contain the key guanidine, uracil and hydroxyl groups found in 1 separated by a tether. Our plan
was to simplify the metabolites by removing the tricyclic ring system and obtaining the
structures 11a-c which are separated by a variable length tether (n = 1-3) (Figure 3). We
eventually compared the bioactive properties of 11a-c with CYN 1, a tetracyclic analogue 22
and guanidinoacetate (GAA; a primary substrate of CYN in cyanobacterial biosynthesis
19
pathway ). We used an in vitro human cell model to elucidate the potential role of tether length
in compound toxicity.
Figure 3: Proposal of the structure of analogues 11.
Results and discussion
Synthesis

ACCEPTED MANUSCRIPT
5
20
The synthesis of the analogues 11a-c began with the known aldehyde 12 which was
treated with the corresponding acetylinic Grignard reagents 13a-c (n = 1-3) in THF to give the
corresponding alcohols 14a-c in 55-75 % yield. Hydrogenation of these gave the alcohols 15a-c
(79-100 % yield), which were acetylated using acetic anhydride in pyridine to give acetates 16a-
c (68-88 % yields) and then desilylated with TBAF to give the alcohols 17a-c (71-80 % yield).
Alcohols 17a-c were reacted with N,N’,N’’-tri-Boc-guanidine 18 in the presence of DIAD and
triphenylphosphine to give guanidines 19a-c. These were deprotected by refluxing in 0.5 M HCl
to give the desired guanidinium salts 11a-c. Unfortunately, on deprotection, the guanidine
groups underwent partial hydrolysis to the amines 20a-c which were inseparable from
guanidines 11a-c. However, treatment of this mixture with 1H-pyrazole-1-carboxamidine
hydrochloride 21 and DIPEA, followed by purification on Amberlite© CG-50 gave the desired
guanidines 11a-c in 12-54% yields over 3 steps (Scheme 1).
Scheme 1: Synthesis of the analogues 11a-c. (i) 13a-c, EtMgBr in THF (1M), -78 °C, 2 h, 0 °C – rt, 16 h. then
NH Cl aq. n = 1, 75 %; n = 2, 71 % n = 3, 55 %. (ii) Pd/C (10%), EA, H , 2 h. n = 1, 79 %, n = 2, 100 %, n = 3, 86
4
2
%
2
. (iii) Ac O, pyridine, DMAP, CH Cl rt, 1 h. n = 1, 68 %, n = 2, 74 %, n = 3, 88 %. (iv) TBAF (1M), THF, 0 °C - rt,
2 2 2,
h, n = 1, 80 %; n = 2, 74 %, n = 3, 71 %. (v) DIAD, PPh N,N’,N”-tri-Boc-guanidine 18, THF, rt 16 h. (vi) HCl (0.5
3,
M), 100 °C, 16 h. (vii) 1H-Pyrazole-1-carboxamidine hydrochloride 21, DIPEA, DMF, 24 h, then Amberlite© CG-50l;
1a; n = 1, 54 %; 11b; n = 2, 12 %; 11c; n = 3, 23 % (over three steps).
1

ACCEPTED MANUSCRIPT
6
Concomitant to the preparation of the tethered analogues described above we prepared
a tetracyclic analogue 22 of 7-deoxy-cylindrospermopsin 3. The synthesis of 22 began with our
previously prepared aldehyde 23 which was subjected to a tethered Biginelli condensation with
21,22
the β-keto ester 24 using our previously reported conditions.
Under these conditions it was
possible to isolate the tetracyclic guanidine 25 in 19% yield, the relative stereochemistry of
which was determined by 2D NOESY NMR. Deallylation/decarboxylation of 25 was achieved by
treatment with Pd(PPh ) , followed by reduction with NaBH CN to give guanidine 26 in 75%
3
4
3
yield. Again 2D NOESY NMR was used to determine the orientation of the newly introduced
methine proton. Deprotection of the N-benzyl groups was achieved by treatment with BBr3
which furnished the 7-deoxy-cylindrospermopsin analogue 22 in 93% yield (Scheme 2).
Scheme 2: Preparation of analogue 22. reagents and conditions: (a) AcOH, 24 h; (b) Morpholine acetate, 24,
Na SO , CF CH OH, 100 °C, 12 d, 19%; (c) Pd(PPh ) , pyrollidine, THF-MeOH (1:1), 4 h; (d) NaBH CN, AcOH-
2
4
3
2
3 4
3
MeOH (1:1), 0 °C to rt, 16 h, 75% (2 steps); (e) i) BBr , Xylenes, 130 °C, 16h; ii) MeOH, rt, 16 h, 93%.
3
The β-keto ester 24 required for the synthesis of 22 was prepared by condensation of
23
N,N'-dibenzylurea 27 with diethyl 1,3-acetonedicarboxylate 28 in the presence of p-TsOH, to
give the benzyl protected pyrimidine ester 29 in 67 % yield. Saponification of 29 by treatment
with ethanolic NaOH, gave the corresponding carboxylic acid 30 in 78 % yield. The desired β-
24
keto ester functionality was incorporated using CDI mediated coupling of 30 with Meldrum’s
acid 31 followed by heating the intermediate adduct in neat allyl alcohol to give 24 in 75% yield.

ACCEPTED MANUSCRIPT
7
Scheme 2: preparation of analogue 24. (a) p-TsOH, 28, 4 Å MS, benzene, 4 d, 67%; (b) NaOH, EtOH, 0 °C to rt,
5
d, 78%; (c), CDI, 31, THF, 50 °C, 16 h; (d) Allyl alcohol, 110 °C, 16 h, 75%.
Toxicity studies
The present study compared the toxicity of the synthetic CYN analogues (11a, 11b, 11c
and 22), 1 and GAA in human-derived neutrophils in vitro. Neutrophils are the most abundant
25
and chemo-sensitive type of leukocytes with a circulating half-life of only 6-8 h, previously
26-28
shown to be a suitable in vitro model to evaluate the toxicity of cyanobacterial metabolites.
The employed model is rapid in contrast to those employing cell lines that usually require
culturing and a longer assay time to allow for cell responses. Moreover, in contrast to numerous
toxicological in vitro studies involving human cell lines, our experimental model used cells
isolated from healthy donors, thereby taking into account individual variability and susceptibility
and for this reason the observed effects may be more relevant in human risk assessment. This
study is amongst those that highlight that toxicology in vitro can be used as a convenient tool to
investigate the biological activity of synthetic toxin analogues, prepared in order to elucidate
which structures may play a pivotal in the toxicity of the parent compound.
The first assay monitored the intracellular levels of ROS and was employed to directly
evaluate the potential of tested compounds in altering the cellular redox balance. The assay is
based on a fluorogenic cell permeant dye which is oxidized to fluorescent compound
(dichlorofluorescein) in the presence of different ROS. The vast majority of cyanobacterial
metabolites and other toxic compounds (e.g. tBHP used in our study as positive control) are
29-31
known to trigger ROS generation above physiological levels.
Increased ROS can lead to
oxidative stress which overwhelms the ability of biological system to readily detoxify the reactive
32
intermediates resulting in damage to cellular structures and the genetic information.
-1
As shown, all the compounds tested, at concentration of 2.0 µg mL , had the ability to
temporarily increase the intracellular ROS levels to different extents (Figure 4A), which is in
contrast to tBHP which caused a gradual increase in ROS over the exposure time. All the CYN

ACCEPTED MANUSCRIPT
8
analogues tested showed a peak-pattern effect on the redox balance with the greatest levels of
ROS noted after 15 minutes after the exposure was initiated. The most significant increase was
observed in the presence of 11a which amounted to 121.6% of the control level. A slightly lower
effect was found for 1 (120.5% of control level) and 11c (119.1% of control level). Importantly,
1
1a was the only compound that caused a statistically significant increase in ROS level after
just 5 min of exposure, indicating the highest bioavailability of all tested compounds. After 1 h of
exposure the ROS level remained significantly increased after exposure to 1, 11a and 11c. The
lowest effect on neutrophils redox balance was observed for 22 and GAA (Figure 4A).
The kinetics of the increase in ROS levels may be determined by the toxic properties of
the compound but potentially also by how rapidly the toxic compound is being taken up and
metabolized by cells. The potential detrimental outcomes of elevated generation of ROS levels
include DNA damage, protein modifications, lipid peroxidation, and necrotic or apoptotic cell
32
death. In the present study the two latter effects were also assessed to determine whether
potential change in ROS levels produced adverse effects or whether neutrophils had the ability
to cope with them through adaptive responses.
Peroxidation of lipids was measured by means of the malondialdehyde (MDA) content.
-1
MDA levels were significantly increased following the 1 h exposure to 2 µg mL of 1, 11a, 11b
and 11c. The 11c compound caused its elevation by as much as 33.3% when compared to the
negative control, at level similar to positive control (Figure 4B). Peroxidation of lipids is a chain
reaction which is initiated by the hydrogen abstraction or addition of an oxygen radical, resulting
in the oxidative damage of polyunsaturated fatty acids. Increased lipid oxidation can decrease
the membrane fluidity and the final products of peroxidation (predominantly MDA and 4-
33,34
hydroxy-2-nonenal) can induce genotoxicity, and cell death.
increase ROS and MDA levels, 1, 11a, 11b and 11c also reduced cell viability accordingly
Figure 4C). In summary, the ability of compounds 11a, 11b and 11c to elevate ROS levels
Consistent with their ability to
(
correlated closely with an increase in lipid peroxidation and cell death. If one considers the
short-term exposure model applied in this study, the magnitude of observed effects highlights
the potentially fast bioavailability of the tested compounds and onset of effects on cellular level.
It should however be stressed that the employed concentration is much above the maximal
-1
observed for 1 in aquatic environment (usually found in the range of 1–10 μg L , maximal
-1
35
reported concentration exceeded 500 μg L )
and concentrations lower by order(s) of
magnitude may not produce such adverse effects. Nevertheless, in the present study, a
-1
concentration of 2.0 µg mL was used intentionally to maximize the potential differences in
action of particular compounds.

ACCEPTED MANUSCRIPT
9
Figure 4. Generation of intracellular ROS (A), lipid peroxidation (B) and cell survival (C) after 1
h exposure to different CYN analogues (1, 11a-11c and 22), GAA and 10 µM tBHP (positive
control) expressed as percentage of control. Bars represent mean ± SD from three independent
experiments corresponding to three different donors (n = 3). Asterisks represent statistically
significant difference to the control (* - p < 0.05; ** - p < 0.01; *** - p < 0.001; Wilcoxon signed-
rank test).
The general toxicity decreased in the following order: 11c > 11a > 1 > 11b > 22 > GAA.
Since all of the CYN analogues tested contained guanidine and uracil, either group could be
involved in generating the observed in vivo effects. The uracil group was demonstrated to be a
12
pharmacophore of CYN, likely facilitating direct binding to nucleic acids and its further effects
36
such as DNA strand breaks and other genotoxic effects , all events usually accompanied by an
37
imbalance in cellular ROS levels. It is however likely that other structures of CYN are also
involved in its bioactivity. It remains so far unclear how CYN moves across the cellular
6
membrane and which structures are crucial in this process . It was however, demonstrated that
13
the sulfate group has no importance for this import process . The role of the hydroxyl group in
this matter is unknown. Interestingly, the naturally occurring 7-deoxy-cylindrospermopsin 3, that
lacks the hydroxyl group, did not produce any toxic effects in an in vivo rodent model at levels
3
previously shown to exceed the lethal concentrations established for 1 . Consistent with the

ACCEPTED MANUSCRIPT
1
0
conclusion that the hydroxyl group is important for cytotoxicity all active compounds in this
study, 11a-11c and 1, contained this group. The 22 analogue, which is most closely related to 1
in terms of its general structure and molecular weight, but lacks the hydroxyl group, did also not
cause toxic outcomes. These findings support the view that the hydroxyl group may play an
important role in the bioactivity of CYN and highlight the need for further studies focusing in
detail on the mechanism of cellular uptake, intracellular localization and intracellular
concentration of tested compounds. It should also be noted that exposure to 11a-11c analogues
and 1 led to a more rapid onset of ROS production, indirectly indicating that their transport
across the plasma membrane might be more efficient. Other compounds, such as cholesterol,
are known to enter cellular membrane by inserting their hydroxyl group into hydrophilic
38
phospholipid head group . Therefore one hypothesis is that the -OH group at C-7 may be
involved in the cellular transport of CYN possibly resembling the uptake of cholesterol.
Moreover, it was established that CYN can be, at least partially, transported across the cellular
membrane via passive diffusion owing to the low molecular weight and high hydrophilicity. The
presence of –OH group in 1 and 11a-11c supports the hydrophilicity of those analogues due to
the polarity of this functional group and likely increases the facilitated diffusion over this
presented by 22 and GAA that lack it. As proposed, hydroxylation of C-7 is the last step of
cyanobacterial biosynthesis of CYN and is catalyzed by a product encoded by cyrI gene of cyr
19
cluster. This might also support the significance of –OH group in CYN bioactivity. Further
studies are however necessary to elucidate the exact sequence of events in which hydroxyl
group is involved in triggering the toxic potencies of CYN in eukaryotic cells. The present study
did not find a linear association between toxicity and the length of tether. Since the greatest
toxicity was found for the analogue 11c, which from the point of view of the proximity of the
uracil, hydroxyl and guanidine groups, might be best able to adopt a similar conformation to that
found in the naturally occurring compound 1.
Conclusions
This study describes the synthesis of four CYN analogues differing in length of tether
between guanidine and uracil groups, and the presence or absence of a hydroxyl group. The
comparison of bioactivity of the synthetic analogues revealed that only those having a hydroxyl
group at C-7 produced significant in vitro toxicity in human cells. The largest toxic effect was
found in compound 11c, which has a 6-carbon tether separating the guanidine and uracil
groups. This tether length might allow the molecule to adopt a similar conformation to that of
CYN as might a 4-carbon tether as 11a had similar activity. The tether length of 5-carbons gave
an appreciably lower activity. Overall the present study adds to the general understanding that

ACCEPTED MANUSCRIPT
1
1
CYN damages human tissue by increasing oxidative stress levels which in turn damages
membrane lipids and can trigger cell death. The results also highlight the potentially important
role of the hydroxyl group at C-7 in cellular uptake and toxicity. Further studies on detailed
mechanism of cellular uptake, intracellular localization and intracellular concentration are
required to fully elucidate these points.
Acknowledgements
Thanks are given to the EPSRC for a fellowship (DE, EP/J01821X/1), to the BEACON
(ERDF) program for support (PJM, DE), the Libyan Authority for Research, Science &
Technology (HSF) and the Foundation for Polish Science (PR "Start" Program: 091.2016). We
would also like to thank the EPSRC National Mass Spectrometry Facility at Swansea.
Experimental
All glassware used was washed with acetone and dried with N or in a vacuum oven. Reagents
2
and starting materials were from commercial suppliers and used as supplied. Diethyl ether and
tetrahydrofuran were distilled from benzophenone and sodium wire, whilst dichloromethane was
dried over CaH and freshly distilled. Dry DMF and methanol were purchased from Aldrich.
Petroleum ether (PE) refers to the fraction distilled between boiling range of 40-60 ºC. Column
chromatography and TLC chromatography was carried out using silica and silica plates
purchased from Fluorochem Ltd (particle size 35-70 ) using HPLC grade acetic acid (AA)
chloroform (CFM), dichloromethane (DCM), diethyl ether (DE), ethyl acetate (EA), hexane (HX)
and methanol (ME). Compounds were visualized using UV, iodine or stained using
polyphosphomolybdic acid (PMA) in EtOH or vanillin in EtOH/H SO4, with heating. IR spectra
2
were recorded on a Perkin-Elmer 1600 series FTIR instrument as KBr discs, thin films, Nujol
mulls, or chloroform solution on NaCl plates with absorption frequencies reported in
-1
wavenumbers (cm ). Routine NMR samples were performed on a Bruker AC250, AC400 or on
a Bruker Avance-500 spectrometer. Chemical shifts for spectra are all reported in  values
(ppm) relative to the residual solvent peak in each case. Electron Ionisation (EI) and Chemical
Ionisation (CI) mass spectra were recorded on an Agilent Tech. 6890N spectrometer or an
XCalibur MAT900 XLT spectrometer and ESI were recorded on a LTQ Orbitrap XL at the
EPSRC Mass Spectrometry Service (Swansea).
General method for the Grignard reaction.
°
A solution of EtMgBr in THF (1M, 1.9 eqv.) was added to a stirred and cooled (0 C)
solution of the alkyne 13 (2 eqv.) in dry THF (2.5 mL per mmol of 13) and the resulting mixture

ACCEPTED MANUSCRIPT
1
2
°
was stirred for 1 h. This solution was then added via cannula to a cooled (-78 C) solution of
aldehyde 12 (1.0 eqv) in dry THF (4 mL per mmol of 12) and the mixture stirred for 2 h. After
warming to rt overnight NH Cl solution (aqueous, saturated, 20 mL) was added and the mixture
4
extracted with EA (3 × 50 mL) and the combined organic extracts washed with brine (25 mL)
and water (25 mL), dried (MgSO ) and evaporated under reduced pressure. The crude residue
4|
was purified by flash chromatography on silica gel (EA/PE 5-30%) to give the acetylinic alcohols
1
4a-c as unstable oils.
4
-((tert-Butyldimethylsilyl)oxy)-1-(2,6-dimethoxypyrimidin-4-yl)but-2-yn-1-ol 14a: 75 %; Rf
0
.32 (25% EA/PE);  0.09 (6H, s, 2 × CH ), 0.87 (9H, s, 3 × CH ), 1.95 (1H, br s, OH), 3.97
H
3
3
(
3H, s, CH ), 3.99 (3H, s, CH ), 4.35 (2H, d J 1.1 Hz, CH ), 5.31 (1H, t, J 1.1 Hz, CH), 6.54 (1H,
3 3 2
s, CH);  -5.1, 18.4, 25.9, 51.8, 54.3, 55.1, 63.3, 82.8, 85.1, 98.2, 165.2, 168.7, 172.5; vmax
C
3
8
392, 2955, 2930, 2885, 2858, 1597, 1572, 1484, 1463, 1382, 1358, 1255, 1203, 1129, 1059,
+
+
15, 779; m/z (CI) 339.2 (100%, [M+H] ), 577.2 (90%, [2M+H] ); HRMS (CI) C H N O Si
16
27
2
4
+
(
[M+H] ) requires 339.1735, found 339.1735.
-((tert-Butyldimethylsilyl)oxy)-1-(2,6-dimethoxypyrimidin-4-yl)pent-2-yn-1-ol 14b: 71 %;
R 0.32 (25% EA/PE);  0.04 (6H, s, 2 × CH ), 0.87 (9H, s, 3 × CH ), 2.45 (2H, dt, J 2.0, 7.1 Hz,
5
f
H
3
3
CH ) 3.72 (2H, t, J 7.1 Hz, CH ), 3.90 (1H, br s, OH) 3.99 (3H, s, CH ), 4.01 (1H, s, CH), 5.27
2
2
3
(
1H, br t, J 2.0 Hz, CH), 6.56 (1H, s, CH);  -5.2, 18.4, 23.3, 26.0, 54.3, 55.2, 61.7, 63.3, 79.2,
C
8
1
4.5, 98.2, 165.1, 169.1, 172.6; vmax 3384, 2953, 2929, 2883, 2857, 1596, 1572, 1482, 1462,
+
381, 1356, 1253, 1203, 1102, 1056, 836, 777; m/z (CI) 339.2 (100%, [M+H] ), 577.2 (90%,
+
+
[
2M+H] ) HRMS (CI) C H N O Si ([M+H] ) requires 353.1890, found 353.1891.
17 28 2 4
6
-((tert-Butyldimethylsilyl)oxy)-1-(2,6-dimethoxypyrimidin-4-yl)hex-2-yn-1-ol 14c: 55 %; Rf
0
.34 (25% EA/PE);  0.00 (6H, s, 2 × CH ), 0.84 (9H, s, 3 × CH ), 1.69 (2H, tt, J 6.0, 7.1 Hz,
H
3
3
CH ), 2.29 (2H, dt, J 2.0 Hz, CH ), 3.63 (2H, t, J 6.0 Hz, CH ), 3.90 (1H Br s OH), 3.95 (3H, s,
2
2
2
CH ), 3.98 (3H, s CH ), 5.25 (1H, br t, J 2.0 Hz, CH), 6.54 (1H, s, CH);  -5.3, 15.3, 26.0, 31.5,
3
3
C
5
2
4.2, 55.0, 61.6, 63.5, 78.3, 87.0, 98.0, 165.1, 169.5, 172.5; vmax 3434, 2955, 2930, 2897, 2858,
216, 1590, 1570, 1483, 1463, 1386, 1360, 1253, 1203, 1100, 1056, 836, 758; m/z (CI) not
determined (unstable to oxidation).
General method for hydrogenation.
Palladium on activated carbon (1:1 by mass with 14) was added to a solution of alkyne
1
4 (1.0 eqv.) in EA (4 mL per mmol) under an argon blanket. The flask was purged with
hydrogen gas (balloon) and the mixture stirred vigorously under a hydrogen atmosphere for 2 h.
©
The mixture was then filtered through a Celite pad which was washed with further portions of
EA and the filtrate concentrated by evaporation under reduced pressure. The crude product was

ACCEPTED MANUSCRIPT
1
3
purified by column chromatography on silica gel (10-40% EA/PE) to give the alkanes 15a-c as
colourless oils.
4
-((tert-Butyldimethylsilyl)oxy)-1-(2,6-dimethoxypyrimidin-4-yl)butan-1-ol 15a: 79 %; Rf
0
1
.25 (20% EA/PE);  0.05 (6H, s, 2 × CH ), 0.89 (9H, s, 3 × CH ), 1.60-1.78 (3H, m, CH, CH ),
H
3
3
2
.93-2.03 (1H, m, CH), 3.65 (2H, t, J 6.5 Hz, CH ), 3.95 (3H, s, Me), 3.97 (3H, s, Me), 4.13 (1H,
2
br s, OH), 4.56 (1H, m, CH), 6.44 (1H, s, CH);  5.3, 18.4, 26.0, 28.6, 34.7, 54.0, 54.8, 63.4,
C
7
1
2.7, 97.5, 165.1, 172.3, 174.2; vmax 3402, 2955, 2930, 2885, 2857, 1598, 1570, 1481, 1462,
+
382, 1355, 1255, 1203, 1153, 1097, 1006, 836, 776; m/z (CI) 343.2 (100%, [M+H] ), 365.2
+
+
+
(
15%, [M+Na] ), 707.4 (30%, [2M+Na] ); HRMS (CI) C H N O Si ([M+H] ) requires 343.2048,
16 31 2 4
found 343.2049.
5
-((tert-butyldimethylsilyl)oxy)-1-(2,6-dimethoxypyrimidin-4-yl)pentan-1-ol 15b: 100 %; Rf
0
1
3
1
2
8
.27 (20% EA/PE);  0.00 (6H, s, 2 × CH ), 0.85 (9H, s, 3 × CH ), 1.39-1.58 (4H, m, 2 x CH ),
H
3
3
2
.61-1.70 (1H, m, CH), 1.74-1.83 (1H, m, CH), 3.37 (1H, br s, OH), 3.52 (2H, t, J 6.3 Hz, CH2),
.94 (3H, s, Me), 3.97 (3H, s, Me), 4.53 (1H, dd, J 4.6, 7.6 Hz, CH ) 6.32 (1H, s, CH);  -5.2,
2
C
8.4, 21.6, 26.0, 32.6, 37.4, 54.1, 54.9, 63.1, 72.5, 97.5, 165.0, 172.2, 173.5; vmax 3422, 3018,
953, 2931, 2898, 2858, 1598, 1570, 1482, 1462, 1382, 1357, 1255, 1215, 1100, 1056, 1006,
+
+
37, 758; m/z (CI) 357.2 (100%, [M+H] ), 735.4 (20%, [2M+Na] ); HRMS (CI) C H N O Si
17
33
2
4
+
(
[M+H] ) requires 357.2204, found 357.2202.
-((tert-Butyldimethylsilyl)oxy)-1-(2,6-dimethoxypyrimidin-4-yl)hexan-1-ol 15c: 86 %; Rf
.34 (25% EA/PE);  0.00 (6H, s, 2 × CH ), 0.86 (9H, s, 3 × CH ), 1.31-1.43 (4H, m, 2 x CH ),
6
0
1
H
3
3
2
.46-1.53 (2H, m, CH ), 1.60-1.69 (1H, m, CH), 1.73-1.81 (1H, m, CH), 3.47 (1H, br s, OH), 3.57
2
(
2H, t, CH J 6.3 Hz), 3.95 (3H, s, Me), 3.98 (3H, s, Me), 4.52 (1H, dd, J 4.6, 7.6 Hz, CH), 6.32
2
(
1H, s, CH);  -5.2, 18.5 (C), 25.0, 25.8, 32.8, 37.7, 54.1, 54.9, 63.2, 72.5, 97.4, 165.0, 172.2,
C
1
73.5; vmax 3418, 2951, 2932, 2858, 1598, 1570, 1480, 1461, 1381, 1356, 1253, 1203, 1099,
+
+
1
056, 1006, 837, 758; m/z (CI) 371.2 (100%, [M+H] ); HRMS (CI) C H N O Si ([M+H] )
18 35 2 4
requires 371.2361, found 371.2360.
General method for acetylation.
Acetic anhydride (3.7 eqv.) was added drop wise to a cooled (0˚C) and stirred mixture of
the alcohol 15 (1.0 eqv.), pyridine (3.9 eqv.) and DMAP (0.1 eqv.) in dry DCM (10 mL per mmol
of 15). The mixture was warmed to rt and stirred for 1 h, whereupon HCl (aqueous. 1 M, 10
eqv.) was added and the mixture stirred for 30 min. The mixture was then extracted with DCM
(3 x 25 mL) and the combined extracts washed with HCl (aqueous, 1M, 3 × 10 mL) and water
(
20 mL). After drying (MgSO ) and evaporated under reduced pressure, column
4
chromatography on silica gel (2.5-17.5% DE in PE) gave the acetates 16a-c as oils.

ACCEPTED MANUSCRIPT
1
4
1
-(2,6-Dimethoxypyrimidin-4-yl)-4-((tert-butydimethylsilyl)oxy)butyl acetate 16a: 68 %; Rf
0
1
3
1
2
1
.28 (20% DE/PE);  0.02 (6H, s, 2 × CH ), 0.87 (9H, s, 3 × CH ), 1.51-1.62 (2H, m, CH ),
H
3
2
2
.84-1.94 (1H, m, CH), 1.96-2.04 (1H, m, CH), 2.13 (3H, s, CH ), 3.60 (2H, t, J 6.3 Hz, CH ),
3
2
.94 (3H, s, Me), 3.96 (3H, s, Me), 5.61 (1H, dd, J 4.8, 7.8 Hz, CH), 6.30 (1H, s, CH);  -5.2,
C
8.4, 21.1, 26.0, 28.4, 30.6, 54.0, 54.9, 62.7, 75.2, 97.7, 165.4, 170.3, 170.6, 172.3; vmax 2956,
930, 2898, 2858, 2253, 1744, 1598, 1572, 1482, 1463, 1388, 1358, 1242, 1206, 1162, 1104,
+
+
059, 964, 910, 836; m/z (CI) 385.2 (100%, [M+H] ), 791.4 (15%, [2M+H] ), HRMS (CI)
+
C H N O Si ([M+H] ) requires 385.2153, found 385.2146.
18
33
2
5
1
-(2,6-Dimethoxypyrimidin-4-yl)-4-((tert-butydimethylsilyl)oxy)butyl acetate 16b: 74 %; Rf
0
1
3
0
.64 (20% EA/PE);  0.01 (6H, s, 2 × CH ), 0.86 (9H, s, 3 × CH ), 1.34-1.42 (2H, m, CH ),
H
3
2
2
.46-1.57 (2H, m, CH ), 1.80-1.98 (2H, m, CH ), 2.13 (3H, s, CH ), 3.57 (2H, t, J 6.3 Hz, CH ),
2
2
3
2
.94 (3H, s, Me), 3.96 (3H, s, Me), 5.58 (1H, dd, J 4.8, 7.8 Hz, CH), 6.30 (1H, s, CH);  -5.2,
C
.1, 18.5, 21.2, 21.6, 26.1, 32.6, 33.9, 54.0, 54.9, 62.9, 75.5, 97.7, 165.4, 170.4, 170.7, 172.3;
vmax 2956, 2929, 2857, 1747, 1597, 1573, 1461, 1386, 1357, 1233, 1205, 1101, 837; m/z (CI)
+
+
+
3
3
99.2 (100%, [M+H] ), 819.4 (10%, [2M+H] ); HRMS (CI) C H N O Si ([M+H] ) requires
19 35 2 4
99.2310, found 399.2299.
-(2,6-Dimethoxypyrimidin-4-yl)-4-((tert-butydimethylsilyl)oxy)butyl acetate 16c: 88 %; Rf
.44 (20% EA/PE);  -0.02 (6H, s, 2 × CH ), 0.87 (9H, s, 3 × CH ), 1.30-1.40 (4H, m, 2 x CH ),
1
0
1
3
1
2
H
3
2
2
.45-1.52 (2H, m, CH ), 1.79-1.97 (2H, m, CH ), 2.13 (3H, s, CH ), 3.57 (2H, t, J 6.3 Hz, CH ),
2
2
3
2
.95 (3H, s, Me), 3.97 (3H, s, Me), 5.58 (1H, dd, J 4.8, 7.8 Hz, CH), 6.30 (1H, s, CH);  -5.2,
C
8.5, 21.2, 25.0, 25.7, 32.7, 34.2, 54.0, 54.9, 63.2, 75.4, 97.6, 165.4, 170.4, 170.7, 172.3; vmax
952, 2932, 2858, 1747, 1597, 1574, 1481, 1462, 1386, 1357, 1232, 1205, 1101, 1045, 835;
+
+
m/z (CI) 413.2 (100%, [M+H] ); HRMS (CI) C H N O Si ([M+H] ) requires 413.2466, found
20
37
2
5
4
13.2458.
General method for silyl deprotection.
A solution of TBAF (1M, 1.5 eqv.) in THF was added to a cooled (0˚C) solution of the
silane 16 (1.0 eqv.) in THF (10 mL per mmol) and the resulting mixture stirred for 2 h. The
reaction mixture was evaporated onto silica gel (ca 1-2 g per mmol of 16) under reduced
pressure, then purified by flash chromatography on silica gel, (gradient elution; 30-40% EA/PE)
to give 17a-c as colorless oils.
1
-(2,6-Dimethoxypyrimidin-4-yl)-4-hydroxybutyl acetate 17a: 80 %; R 0.26 (40% EA in
f
PE);  1.64-1.89 (4H, m, 2 × CH ), 2.00 (3H, s, CH ), 3.59 (1H, br s, OH), 3.94 (3H, s, CH ),
H
2
3
3
3
2
.97 (3H, s, CH ), 4.04-4.11 (2H, m, CH ), 4.55 (1H, dd, J 2.6, 6.3 Hz, CH), 6.33 (1H, s, CH); 
3
2
C
1.0, 24.8, 34.0, 54.1, 54.9, 64.3, 72.0, 97.4, 165.1, 171.3, 172.3, 173.1; vmax 3440, 3428, 2955,

ACCEPTED MANUSCRIPT
1
5
+
2
927, 1738, 1596, 1569, 1482, 1381, 1351; m/z (CI) 271.1 (100%, [M+H] ), 293.1 (40%,
+
+
+
[
M+Na] ), 563.2 (20%, [2M+Na] ); HRMS (CI) C H N O ([M+H] ) requires 271.1288, found
12 19 2 5
2
71.1287.
-(2,6-Dimethoxypyrimidin-4-yl)-4-hydroxybutyl acetate 17b: 74. %; R 0.26 (40% EA in
1
f
PE);  1.35-1.43 (2H, m, CH ), 1.49-1.62 (2H, m, CH ), 1.79-1.97 (2H, m, CH ) 2.11 (3H, s,
H
2
2
2
CH ), 2.12 (1H, br s, OH), 3.57 (2H, t, J 6.5 Hz, CH ), 3.92 (3H, s, CH ), 3.94 (3H, s, CH ), 5.56
3
2
3
3
(
1H, dd, J 4.8, 7.5 Hz, CH), 6.28 (1H, s, CH);  21.1, 21.4, 32.3, 33.8, 54.0, 54.9, 62.5, 75.2,
C
9
7.5, 165.3, 170.4, 170.5, 172.2; vmax 3408, 2948, 2870, 1744, 1597, 1571, 1483, 1386, 1359;
+
+
m/z (CI) 285.1 (100%, [M+H] ), HRMS (CI); C H N O ([M+H] ) requires 285.1443, found
13
20
2
5
2
85.1445.
-(2,6-Dimethoxypyrimidin-4-yl)-6-hydroxyhexyl acetate 17c: 71 %; R 0.26 (40% EA in
1
f
PE);  1.27-1.39 (4H, m, 2 x CH ), 1.49-1.56 (2H, m, CH ), 1.78-1.95 (3H, m, CH , OH), 2.12
H
2
2
2
(
3H, s, CH ), 3.58 (2H, t, J 6.6 Hz, CH ), 3.93 (3H, s, CH ), 3.95 (3H, s, CH ), 5.56 (1H, dd, J
3 2 3 3
4
1
.9, 7.8 Hz, CH), 6.28 (1H, s, CH);  21.1, 21.9, 25.5, 32.5, 34.0, 54.0, 54.9, 62.7, 75.3, 97.5,
C
65.3, 170.4, 170.6, 172.3; vmax 3414, 2934, 2862, 1746, 1598, 1573, 1483, 1386, 1358; m/z
+
+
(
CI) 299.2 (100%, [M+H] ), HRMS (CI) C H N O ([M+H] ) requires 299.1601, found
14 23 2 5
2
99.1599.
General method for the guanylation, deprotection, reguanylation and purification.
Guanylation: DIAD (1.5 eqv.) was added in a drop wise manner to a cooled (0 ˚C) and
stirred solution of the alcohol 17 (1.0 eqv.), PPh (1.53 eqv.) and N,N′,N′′-tri-Boc-guanidine 18 (2
3
eqv.) in dry THF (8 mL per mmol of 17). After stirring to rt overnight, water (3-4 eqv.) was added
and the mixture stirred for 30 min after which the solution was evaporated under reduced
pressure. The crude product was purified by flash chromatography on silica gel (5-20% EA/PE)
to give 19a-c as gums, which were contaminated with quantities of DIAD byproducts and were
used in the next step without further purification. Deprotection: Hydrochloric acid (6 mL per 0.1
mmol, 0.5 M) was added to 19a-c (1.0 eqv.) and the mixture heated (100°C) at reflux for 16h.
After cooling and evaporation under reduced pressure the residue was dried in a desiccator
(
P O ) overnight, dissolved in ME (ca 2 mL) and DE (3-5 ml) was added to the cloud point
2 5
whereupon an oily precipitate on cooling (-20˚C) overnight. This was found to be a mixture of
1a-c and 20a-c. Reguanylation and ion exchange purification: The crude mixture from the
1
previous step was dissolved in DMF (2-3 mL per 0.1 mmol) and 1H-pyrazole-1-carboxamidine
hydrochloride (0.5-0.7 eqv) was added together with DIPEA (1.5-2.0 eqv) and the mixture
stirred at rt for 24 hrs. DE (5-10 ml) was added which resulted in the formation of an oily
precipitate formed, whereupon the supernatant liquid was decanted. The precipitate was

ACCEPTED MANUSCRIPT
1
6
dissolved in ME (1-2 ml) and DE (3-5 ml) was added to the cloud point, and an oily precipitate
formed on cooling (-20˚C) overnight. After decanting the supernatant liquid, the precipitate was
dissolved in water (5 mL) then passed several times through a short column (3 g) of activated
Amberlite© CG-50 (the resin was pre-washed with MeOH, then HCl (2 M), then water to pH =
7
). The column was then eluted with water (20 mL), ammonia (1 M (30 mL) and 2M (30 mL)),
evaporation of the 2M ammonia washings gave on evaporation the crude guanidines 11a-c (the
water fraction was found to contain residual product, which could be purified by repeating the
Amberlite© CG-50 purification steps). The crude guanidine obtained was dissolved in ME (1-2
ml) containing a drop of HCl (12 M) and DE (3-5 ml) was added to the cloud point. After cooling
(
-20˚C) overnight, the supernatant liquid was decanted to give 11a-c as gums after drying.
-(2,6-dimethoxypyrimidin-4-yl)-4-(1,2,3-tris(tertbutoxycarbonyl)guanidino)butyl acetate
9a: R 0.26 (20% EA/PE);  1.49 (18H, s, 6 × CH ), 1.50 (9H, s, 3 × CH ), 1.71-1.91 (2H, m,
1
1
f
H
3
3
CH ), 2.04 (3H, s, CH ), 2.17-2.38 (2H, m, CH ), 3.94 (3H, s, CH ), 3.96 (3H, s, CH ), 4.11 (2H,
2
3
2
3
3
t, J 6.4 Hz, CH ), 5.40-5.68 (1H, m, CH), 6.50 (1H, s, CH), 10.73 (1H, s, NH);  (partial data)
2
C
2
1
3
1.1, 25.7, 27.9, 27.9, 28.1, 28.2, 54.0, 54.9, 61.1, 72.3, 99.1, 149.7, 152.5, 155.0, 170.8,
+
+
72.2, 172.3 (5 x C not observed); m/z (CI) 612.3 (100%, [M+H] ), 1245.6 (90%, [2M+Na] ),
+
+
12 (96%, [M-2Boc+H] ); HRMS (CI) C H N O ([M+H] ) requires 612.3239, found 612.3225.
28
46
5
10
1
-(2,6-Dimethoxypyrimidin-4-yl)-5-(1,2,3-tris(tert-butoxycarbonyl)guanidino)pentyl acetate
1
9b: R 0.26 (20% EA/PE);  1.46 (9H, s, 3 × CH ), 1.49 (18H, s, 6 × CH ), 1.35-1.42 (2H, m,
f
H
3
3
CH ), 1.59-1.73 (2H, m, CH ), 1.79-2.02 (2H, m, CH ) 2.14 (3H, s, CH ), 3.75 (2H, t, J 7.4 Hz,
2
2
2
3
CH ), 3.95 (3H, s, CH ), 3.97 (3H, s, CH ), 5.58 (1H, dd, J 4.6, 7.7 Hz, CH), 6.30 (1H, s, CH),
2
3
3
1
7
0.61 (1H, s, NH);  (partial data) 21.2, 22.5, 28.1, 28.1, 28.2, 28.6, 33.8, 47.7, 54.0, 54.9,
C
5.4, 97.6, 149.6, 152.8, 153.6, 170.4, 170.6, 172.3 (5 x C not observed); m/z (CI) 626.3 (7%,
+
+
+
[
M+H] ), 648.3 (100%, [M+Na] ); HRMS (CI) C H N O Na ([M+Na] ) requires 648.3215,
29 47 5 10
found 648.3221.
1
-(2,6-Dimethoxypyrimidin-4-yl)-4-(1,2,3-tris(tertbutoxycarbonyl)guanidino)hexyl acetate
1
9c: R 0.26 (20% EA/PE);  1.26-1.36 (4H, m, 2 x CH ), 1.45 (9H, s, 3 × CH ), 1.47 (18H, s, 6
f
H
2
3
×
CH ), 1.55-1.66 (2H, m, CH ), 1.77-1.92 (2H, m, CH ), 2.13 (3H, s, CH ), 3.72 (2H, t, J 7.5 Hz,
3 2 2 3
CH ), 3.95 (3H, s, CH ), 3.96 (3H, s, CH ), 5.55 (1H, dd, J 4.8, 7.7 Hz, CH), 6.27 (1H, s, CH),
2
3
3
1
5
0.59 (1H, s, NH);  (partial data) 21.1, 24.9, 26.6, 28.1, 28.1, 28.2, 28.7, 34.1, 47.9, 54.0,
C
4.9, 75.4, 97.6, 149.1, 152.7, 153.6, 170.4, 170.6, 172.3 (5 x C not observed).
Analogues 11a-c
Amino((4-(2,6-dioxo-1,2,3,6-tetrahydropyrimidin-4-yl)-4-
hydroxybutyl)amino)methaniminium chloride 11a: 54 % (over three steps);  (CD OD) 1.59-
H
3
1
.74 (2H, m, CH ), 1.76-1.86 (1H, m, CH ), 2.00-2.08 (1H, m, CH), 3.63 (2H, t, J 6.0 Hz, CH ),
2 2 3

ACCEPTED MANUSCRIPT
1
7
4
1
2
.33 (1H, dd, J 3.8, 9.0 Hz, CH), 5.56 (1H, s, CH);  (CD OD) 29.4, 31.8, 54.0, 61.8, 98.0,
c
3
+
+
52.2, 157.1, 158.7, 166.8; m/z (CI) 242.1 (100%, [M] ); HRMS (CI) C H N O ([M] ) requires
9
16
5
3
42.1248, found 242.1250.
Amino((5-(2,6-dioxo-1,2,3,6-tetrahydropyrimidin-4-yl)-5-
hydrooxypentyl)amino)methaniminium chloride 11b: 12 % (over three steps);  (CD OD) 1.42-
H
3
1
.80 (6H, m, 3 x CH ), 3.19 (2H, t, J 7.0 Hz, CH ), 4.38 (1H, dd, J 4.5, 7.6 Hz, CH), 5.63 (1H, s,
2 2
CH);  (CD OD) 23.3, 29.6, 36.5, 42.3, 70.2, 97.3, 153.4, 158.6, 161.6, 167.4; m/z (CI) 256.1
c
3
+
+
(
100 %, [M] ); HRMS (CI) C H N O ([M] ) requires 256.1404, found 256.1411.
10 18 5 3
Amino((5-(2,6-dioxo-1,2,3,6-tetrahydropyrimidin-4-yl)-6-
hydrooxypentyl)amino)methaniminium chloride 11c: 23 % (over three steps);  (CD OD) 1.37-
H
3
1
.76 (8H, m, 4 x CH ), 3.17 (2H, t, J 7.0 Hz, CH ), 4.36 (1H, dd, J 4.8, 7.2 Hz, CH), 5.62 (1H, s,
2 2
CH);  (CD OD) 25.8, 27.4, 29.7, 36.9, 42.4, 70.4, 97.3, 153.4, 158.6, 161.8, 167.4; m/z (CI)
c
3
+
+
2
70.2 (100 %, [M] ); HRMS (CI) C H N O ([M] ) requires 270.1561, found 270.1557.
10 18 5 3
Allyl (5aR,8aS)-4-((1,3-dibenzyl-2,6-dioxo-1,2,3,6-tetrahydropyrimidin-4-yl)methyl)-
3
,5a,6,7,8,8a-hexahydro-1H-2,2a1,3-triazaacenaphthylene-5-carboxylate acetate 25.
16,17
A solution of aldehyde 23
(255.1 mg, 0.719 mmol) in AA (3 mL) was stirred at rt for
2
4 h. After evaporation under reduced pressure, residual AA was removed by co-evaporation
with CFM (3 x 15 mL) and the residue dried under high vacuum for 4 h. The resultant salts
where dissolved in trifluoroethanol (1.5 mL) and morpholine acetate (264.7 mg, 1.798 mmol), 24
(1.55 g, 3.595 mmol) and anhydrous sodium sulphate (1.0 g) were added and the resultant
mixture stirred at 70°C in a sealed Carius tube for 12 d. After cooling to rt and evaporation,
purification was achieved by flash column chromatography on silica gel using ME/CFM
(stepwise gradient of 1% increments 0:100 to 8:92) containing 1% AA. Fractions eluting in 5:95
ME/CFM gave 25 as a tan oil (82.5 mg, 0.135 mmol, 19%). R 0.53 (10% ME/CFM);  1.36-
f
H
1
3
3
.41 (2H, m, 2 x CH), 1.52-1.64 (3H, m, 3 x CH), 1.78-1.81 (2H, m, 2 x CH), 1.96 (3H, s, CH3),
.19 (1H, dd, J = 9.6, 1.3 Hz, CH ), 3.54-3.58 (1H, m, CH), 3.79 (1H, d, J = 16.6 Hz, CH ), 3.83-
2
2
.88 (1H, m, CH), 3.98 (1H, d, J = 16.6 Hz, CH ), 4.40 (1H, dd, J = 11.2, 2.4 Hz, CH), 4.46-4.48
2
(2H, m, CH ), 5.09-5.22 (6H, m, 3 x CH ), 5.82 (1H, ddt, J = 16.1, 10.8, 5.6 Hz, CH), 7.17-7.37
2 2
(
10H, m, 10 x CH);  (partial data) 21.8, 22.5, 30.2, 31.8, 39.1, 45.5, 47.3, 48.9 (HSQC), 54.2,
C
5
1
1
5
8.5, 65.5, 101.0, 103.0, 118.2, 127.4, 128.5, 128.6, 129.3, 129.4, 129.8, 133.9, 137.6, 138.3,
54.0, 156.5, 159.8, 164.6, 166.8 (1 x C not observed); vmax 3064, 3010, 2957, 2928, 2856,
+
700, 1659, 1617, 1562, 1496, 1449, 1395, 1361, 1314, 1248; m/z (ESI) 1103 (16%, [2M+H] ),
+
+
74 (14%, [M+Na] ), 552 (100%, [M+H] ); m/z (ESI negative) 121 (100), 75 (88), 59 (85%, [M-
-
+
H] ); HRMS (ESI) C H N O [M+H] requires 552.2605, found 552.2602.
32
34
5
4

ACCEPTED MANUSCRIPT
1
8
1
,3-Dibenzyl-6-(((4R,5aR,8aS)-3,4,5,5a,6,7,8,8a-octahydro-1H-2,2a1,3-triazaacenaphthylen-4-
yl)methyl)pyrimidine-2,4(1H,3H)-dione acetate 26.
Tetrakis(triphenylphosphine)palladium(0) (8.1 mg, 0.007 mmol) and pyrrolidine (0.027
mL, 0.337 mmol) were added to a stirred solution of 25 (206.0 mg, 0.337 mmol) in a 1:1 mixture
of anhydrous ME/THF (6 mL). The reaction was stirred at rt and progress monitored by TLC.
After 4 h the solvent was removed under reduced pressure and the residue dissolved in ME (3
mL) and AA (3 mL) and cooled (0 °C), whereupon sodium cyanoborohydride (105.8 mg, 1.685
mmol) was added and the mixture stirred to rt over 16 hours. After evaporation, the reaction
material was purified by flash column chromatography on silica gel using ME/CFM (0:100 to
1
7
5:85) containing 1% AA. Fractions eluting in 10:90 ME/CFM gave 26 (133.8 mg, 0.253 mmol,
5%) as a pale yellow oil. R 0.18 (10% ME/CFM);  0.95-1.10 (2H, m, 2 x CH), 1.17-1.26 (1H,
f
H
m, CH ), 1.32-1.42 (1H, m, CH ), 1.77-1.81 (1H, m, CH ), 1.85-1.89 (1H, m, CH ), 1.94 (3H, s,
2
2
2
2
CH ), 1.97-2.01 (1H, m, CH ), 2.03-2.07 (1H, m, CH ), 2.45 (1H, dd, J = 9.9, 2.7 Hz, CH ), 2.98-
3
2
2
2
3
3
.03 (1H, m, CH), 3.12-3.20 (2H, m, 2 x CH), 3.40-3.45 (1H, m, CH), 3.52-3.60 (1H, m, CH),
.73-3.78 (1H, m, CH ) 4.94-5.37 (4H, m, 2 x CH ), 5.64 (1H, s, CH), 7.16-7.49 (10H, m, 10 x
2
2
CH);  22.0, 23.3, 29.5, 30.9, 34.2, 36.9, 44.7, 47.7, 47.9, 48.5, 49.6, 56.7, 102.7, 126.4, 127.6,
C
1
3
27.8, 128.3, 129.0, 129.0, 136.0, 136.6, 150.7, 152.6, 155.6, 161.8, 178.0; vmax 3155, 3065,
034, 2947, 2931, 2862, 1704, 1668, 1620, 1586, 1557, 1497, 1451, 1398, 1340 and 1262; m/z
+
(
ESI) 470 (100%, [M+H] ); m/z (ESI negative) 157 (18), 141 (20), 121 (100), 75 (86), 73 (12),
-
+
5
9 (83%, [M-H] ); HRMS (ESI) C H N O [M+H] requires 470.2551, found 470.2546.
28 32 5 2
6
-(((4R,5aR,8aS)-3,4,5,5a,6,7,8,8a-Octahydro-1H-2,2a1,3-triazaacenaphthylen-4-
yl)methyl)pyrimidine-2,4(1H,3H)-dione hydrobromide 22.
Boron tribromide (106 µL, 1.120 mmol) was added to a stirred solution of 26 (118.4 mg,
0
.224 mmol) dissolved in xylenes (12 mL) and the mixture heated at reflux for 16 h. On cooling
to rt ME (5 mL) was slowly added and the mixture stirred for a further 16 h. After evaporation,
the crude product was purified by flash column chromatography on silica gel (ME:CFM (0:100 to
3
9
0:70 containing 1% AA). Fractions eluting in 12:88 ME/CFM gave 22 (77.9 mg, 0.210 mmol,
4%) as a waxy solid. R 0.11 (20% ME/CFM);  1.21-1.29 (1H, m, CH ), 1.33-1.41 (1H, m,
f
H
2
CH ), 1.47-1.56 (2H, m, 2 x CH), 1.93-1.99 (2H, m, 2 x CH), 2.06-2.09 (1H, m, CH ), 2.29 (1H,
2
2
dt, J = 13.5, 3.6 Hz, CH ), 2.69 (2H, d, J = 6.8 Hz, CH ), 3.18-3.25 (1H, m, CH ), 3.40 (1H, tt, J
2
2
2
=
4
2
10.9, 3.6 Hz, CH), 3.80-3.86 (3H, m, 3 x CH), 5.59 (1H, s, CH);  23.1, 30.4, 31.7, 35.1, 38.3,
C
9.3, 49.6, 51.2, 58.3, 101.8, 153.3, 153.4, 156.4, 166.8; vmax 3322, 3159, 3054, 2963, 2927,
+
859, 1713, 1659, 1607, 1445, 1342, 1331; m/z (ESI) 579 (4%, [2M+H] ), 394 (5), 372 (10), 312

ACCEPTED MANUSCRIPT
1
9
+
+
(
9%, [M+Na] ), 290 (100%, [M+H] ); m/z (ESI negative) 125 (53), 85 (27), 81 (95), 79 (100%,
-
+
[
M-H] ), 62 (20), 32 (21); HRMS (ESI) C H N O [M+H] requires 290.1612, found 290.1607.
14 20 5 2
18
1
,3-Dibenzylurea 27.
A vigorously stirred suspension of finely powdered carbohydrazide (9.0 g, 100.0 mmol) in
DMF (200 mL) and benzylamine (109 mL, 1000 mmol) was cooled to 0 °C, to which I (100.0 g,
2
0
.40 mol) was added in portions over 1 h until a brown colour persisted. Excess Iodine was
removed by the addition of Na S O (Sat, ca. 2-3 mL) and the desired compound was
2
2
3
precipitated from the resulting straw yellow solution by the addition of water (800 mL). After
filtration and drying under vacuum, 27 (15.3 g, 64 mmol, 64%) was obtained as a white solid
which was used without further purification. Recrystallisation from THF/toluene gave An
39
analytically pure of sample. R 0.20 (50% EA/PE); Mp 165-168 °C (lit. 168-169 °C);  4.23
f
H
(
2H, s, CH ), 4.24 (2H, s, CH ), 6.43 (2H, t, J = 5.8 Hz, 2 x NH), 7.20-7.33 (10H, m, 10 x CH); 
2
2
C
4
1
3.0, 43.0, 126.5, 127.0, 128.2, 140.9, 158.1; vmax 3337, 3086, 3062, 3030, 2919, 2874, 1627,
+
613, 1589, 1574, 1493, 1472, 1453, 1421; m/z (ESI) 503.2 (25%, [2M+Na] ), 481.3 (100%,
+
+
+
+
[
2M+H] ), 263.1 (8%, [M+Na] ), 241.1 (95%, [M+H] ); HRMS (ESI) C H N O [M+H] requires
15 17 2
2
41.1335, found 141.1333.
Ethyl 2-(1,3-dibenzyl-2,6-dioxo-1,2,3,6-tetrahydropyrimidin-4-yl)acetate 29.
Urea 27 (11.5 g, 47.7 mmol), diethyl 1,3-acetonedicarboxylate 28 (17.3 mL, 95.4 mmol)
and p-TsOH (907 mg, 4.8 mmol) were dissolved in benzene (250 mL) and the mixture heated to
reflux under a soxhlet extractor containing powdered 4 Å molecular sieves (ca 20 g). The
1
molecular sieves were replaced every 2 days and after 4 days H NMR analysis indicated
complete consumption of the starting material. The reaction mixture was cooled to rt, diluted
with DCM (800 mL) and washed with NaHCO (sat. 3 x 150 mL) and brine (100 mL). After
3
drying and evaporation the resulting viscous oil was stirred with PE (300 mL). After 1 hour the
solvent was decanted and the resultant precipitate triturated with DE (4 x 100 mL) to give ester
2
9 as a pale yellow solid (12.1 g, 32.0 mmol, 67%). R 0.41 (50% EA/PE); Mp 119-122 °C; H
f
1
.26 (3H, t, J 7.2 Hz, CH ), 3.4 (2H, s, CH ), 4.14 (2H, q, J 7.2 Hz, CH ), 5.13 (2H, s, CH ), 5.17
3
2
2
2
(
2H, s, CH ), 5.73 (1H, s, CH), 7.11-7.50 (10H, m, 10 x CH);  14.0, 38.7, 44.8, 48.1, 62.2,
2
C
1
3
1
04.3, 126.1, 127.6, 127.9, 128.4, 129.0, 129.1, 135.8, 136.7, 147.7, 152.5, 161.8, 167.4; vmax
090, 3061, 3023, 3008, 2979, 2935, 1738, 1702, 1693, 1654, 1662, 1617, 1584, 1495, 1451,
+
+
423, 1396, 1367, 1333; m/z (ESI) 779.3 (6%, [2M+Na] ), 757.3 (100%, [2M+H] ), 401.1 (5%,
+
+
+
[
M+Na] ), 379.2 (95%, [M+H] ); HRMS (ESI) C H N O [M+H] requires 379.1652, found
22 23 2 4
3
79.1650.

ACCEPTED MANUSCRIPT
2
0
2
-(1,3-Dibenzyl-2,6-dioxo-1,2,3,6-tetrahydropyrimidin-4-yl)acetic acid 30.
Sodium hydroxide (2.25 g, 56.1 mmol) was added to a cooled (0 °C), stirred suspension
of ester 29 (10.62 g, 28.1 mmol) in ethanol (500 mL) and the mixture stirred to rt over 5 d. The
mixture was quenched with H O (600 mL) and the pH adjusted to pH 1-2 by the slow addition of
2
1
M HCl, then extracted with DCM (3 x 100 mL) and the combined organic extracts dried over
magnesium sulfate. After evaporation the resulting pale brown solid was triturated with DE (3 x
0 mL) to give acid 30 as an off white solid (7.70 g, 22.0 mmol, 78%). R 0.26 (20% ME/DCM);
7
f
Mp 138-141 °C;  3.57 (2H, s, CH ), 5.02 (2H, s, CH ), 5.04 (2H, s, CH ), 5.88 (1H, s, CH),
H
2
2
2
7
1
3
1
.15-7.36 (10H, m, 10 x CH), 13.07 (1H, s, COOH);  38.0, 43.8, 47.5, 103.0, 125.9, 127.2,
C
27.4, 127.6, 128.4, 128.7, 136.4, 137.0, 149.6, 152.0, 161.3, 169.4; vmax 3442, 3096, 3063,
034, 2921, 2793, 2695, 2576, 1722, 1705, 1693, 1642, 1598, 1496, 1472, 1454, 1423, 1399,
+
+
367, 1351, 1315; m/z (ESI) 701.3 (100%, [2M+H] ), 373.1 (15%, [M+Na] ), 351.1 (100%,
+
+
[
M+H] ); HRMS (ESI) C H N O [M+H] requires 351.1399, found 351.1340.
20 19 2 4
Allyl 4-(1,3-dibenzyl-2,6-dioxo-1,2,3,6-tetrahydropyrimidin-4-yl)-3-oxobutanoate 24.
Pyrimidine 30 (1.50 g, 4.28 mmol), Meldrum's acid 31 (0.68 g, 4.71 mmol) and CDI (1.04
g, 6.42 mmol) were dissolved in THF (20 mL) and the mixture heated at 50 °C for 16 h. Upon
cooling to rt the solvent was removed under reduced pressure and the residue diluted with
DCM/H O (1:1, 160 mL) and the pH of the aqueous layer adjusted to pH 2 by the slow addition
2
of 1 M HCl. The organic layer was then separated and washed with H O (50 mL), dried over
2
magnesium sulfate and the solvent removed under vacuum. The resulting yellow solid was
dissolved in allyl alcohol (20 mL) and the mixture heated at 110 °C for 16 h. After cooing the
mixture was evaporated under reduced pressure and the resulting material purified by flash
column chromatography on silica gel using EA/PE (15:85 to 30:70). Fractions eluting in 25:75
EA/PE gave 24 as a viscous yellow oil (1.38 g, 3.19 mmol, 75%). R 0.08 (30% EA/PE);  3.77
f
H
(
2H, s, CH ), 3.98 (2H, s, CH ), 4.58 (2H, dt, J 5.5, 1.4 Hz, CH ), 4.69 (2H, s, CH ), 5.02 (2H, s,
2 2 2 2
CH ), 5.19-5.33 (2H, m, CH ), 5.81 (1H, s, CH), 5.83-5.92 (1H, m, CH), 7.14-7.36 (10H, m, 2 x
2
2
Ph);  43.8, 46.1, 47.6, 48.2, 65.2, 103.2, 118.3, 126.0, 127.3, 127.4, 127.5, 128.4, 128.7,
C
1
2
8
32.1, 136.6, 137.1, 149.3, 151.9, 161.2, 166.6, 198.4; vmax 3094, 3067, 3023, 3002, 2968,
939, 1746, 1727, 1705, 1702, 1663, 1628, 1586, 1497, 1453,1396, 1347 and 1328; m/z (ESI)
+
+
+
65.3 (50%, [2M+H] ), 455.2 (6%, [M+Na] ), 433.2 (100%, [M+H] ); HRMS (ESI) C H N O
5
25
25
2
+
[
M+H] requires 433.1758, found 433.1756.

ACCEPTED MANUSCRIPT
2
1
Toxicological studies: The toxicity of CYN variants was evaluated using human neutrophils in
an in vitro model. Heparinized blood samples (6.0 mL) were collected in lithium heparin tubes
from 3 healthy (screened by physical examination, medical history and initial blood tests), non-
smoking and normal weighted (BMI 18.5–24.9) male donors (aged 21–25 years old) at the
Regional Centre of Blood and Blood Treatment in Poznan, Poland, according to accepted
safeguard standards and legal requirements. Human neutrophils were isolated immediately
after collection using a one-step density-gradient centrifugation on Gradisol G of a specific
-1
gravity of 1.115 g mL (Polfa, Poland) at 400 g at room temperature for 30 min. The residual
erythrocytes were removed from the cell population by hypotonic lysis. The purity of the
neutrophils (>90%) was verified by counting under a light microscope after May-Grunwald-
Giemsa staining. Neutrophils were exposed for 1 h to 2.0 µg/mL of 11a-c, 22, 1 or GAA and
their toxicity was compared by monitoring the intracellular generation of reactive oxygen species
(ROS), peroxidation of lipids and cell survival. Each experimental assay was performed on three
independent neutrophil isolates with three technical replicates. The negative control was
constituted of unexposed cells. The positive control was constituted of cells exposed to 10 µM
tert-butyl hydroperoxide (tBHP), a well established inducer of oxidative stress and specifically,
lipid peroxidation and oxidative stress-induced death in eukaryotic cells.
Intracellular reactive oxygen species assay: Human neutrophils were loaded for 30 min at 37

C in darkness with 20 M of 2’,7’ –dichlorofluorescin diacetate (DCFDA; Abcam, UK), a
fluorogenic dye that measures hydroxyl, peroxyl and other reactive oxygen species (ROS)
activity within the cell. Cells were then washed, dissolved in PBS, seeded in a black clear
4
bottom 96-well plate at a density of 20 x 10 neutrophils per well (90 μL aliquots/well) and
exposed to tested compounds for 1h at 37C for 1 h. Fluorescence of DCFDA was measured
kinetically after 5, 15, 30, 45 and 60 min of incubation using a Synergy HTX multi-mode plate
reader (BioTek, USA) at an excitation of 495 nm and emission of 528 nm. The background
signal, measured in exposed neutrophils not loaded with DCFDA, was withdrawn from the
corresponding samples. The final results were presented as a percentage of the negative
control.
Lipid peroxidation assay: Lipid peroxidation was analyzed using a Lipid Peroxidation
Colorimetric/Fluorometric Assay Kit (BioVision, UK) by means of malondialdehyde (MDA)
4
content. Human neutrophils were seeded in a 96-well plate at a density of 20 x 10 neutrophils
per well (90 μL aliquots/well) and exposed to each tested compound for 1 h at 37 C. The
control was constituted of cells incubated with 10 μL of PBS. After the experiments, cells were

ACCEPTED MANUSCRIPT
2
2
harvested from each well and homogenized on ice in 300 μL of provided lysis buffer and
centrifuged to remove insoluble material. The resulting 200 μL of supernatants were transferred
to a microcentrifuge tube and supplemented with 600 μL of thiobarbituric acid (TBA) to generate
an MDA–TBA adduct. To accelerate the reaction, samples were incubated at 95 °C for 60 min
and the final product was measured colorimetrically at 532 nm. The calculated values were
compared to a calibration curve prepared using MDA standard (BioVision, UK). The coefficient
2
of variation (r ) for the calibration curve was 0.99. The final results were presented as a
percentage of the negative control.
Cell viability assay: Cell viability was determined using the colorimetric 3-(4,5-dimethylthiazol-
2
-yl)-2,5 diphenyltetrazolium bromide (MTT) metabolic activity assay (BioVision, UK). Human
4
neutrophils were seeded in a 96-well plate at a density of 20 x 10 neutrophils per well (90 μL
aliquots/well) and exposed to each tested compound for 1 h at 37 °C. Afterwards, cells were
washed, seeded again, and 10 µL of MTT was added to each well for 3 h. Neutrophils were
then treated with 10% sodium dodecyl sulfate in 0.01 M HCl and incubated for another 6 h in
the darkness to dissolve formazan crystals. The optical density (OD) of the final product (the
formazan crystals) was measured at 570 nm using a Synergy HTX microplate reader (BioTek,
USA). The final results were presented as a percentage of the negative control.
References
1
2
3
. I. Ohtani, R. E. Moore and M. T. C. Runnegar, J. Am. Chem. Soc., 1992, 114, 7941.
. R. Banker, B. Teltsch, A. Sukenik and S. Carmeli, J. Nat. Prod., 2000, 63, 387.
. R. L. Norris, G. K. Eaglesham, G. Pierens, G. R. Shaw, M. J. Smith, R. K. Chiswell, A. A.
Seawright and M. R. Moore, Environ. Toxicol., 1999, 14, 163.
4
5
6
. K. M. Wimmer, W. K. Strangman, J. L.C. Wright, Harmful Algae., 2014, 37, 203.
. S. Byth, Med. J. Aust., 1980, 2, 40.
. B. Poniedziałek, P. Rzymski, M. Kokociński, Environ. Toxicol. Pharmacol., 2012, 34,
6
51.
. R. Banker, S. Carmeli, O. Hadas, B. Teltsch, R. Porat, A. Sukenik, J. Phycol., 1997, 33,
13.
. R. Li, W. W. Carmichael, S. Brittain, G. K. Eaglesham, G. R. Shaw, Y. Liu and M. W.
Watanabe, J. Phycol., 2001, 37, 1121.
7
8
6
9
1
. P. Rzymski, B. Poniedziałek, B. J Phycol., 2015, 51, 1037.
0.M. Bohunicka, J. Mareš, P. Hrouzek, P. Urajová, M. Lukeš, J. Šmarda, J. Komárek, L.
Gaysina , O. Strunecky, O. J. Phycol. 2015, 51, 1040.
1
1
1
1.P. Rzymski, B. Poniedziałek, M. Kokociński, T. Jurczak, D. Lipski, K. Wiktorowicz K.
Harmful Algae 2014, 35, 1.
2.R. Banker, S. Carmeli, M. Werman, B. Teltsch, R. Porat and A. Sukenik, J. Toxicol.
Environ. Health, Part A, 2001, 62, 281.
3.M. T. Runnegar, C. Xie, B. B. Snider, G. A. Wallace, S. M. Weinreb, J. Kuhlenkamp,
Toxicol. Sci., 2002, 67, 81.

ACCEPTED MANUSCRIPT
2
3
1
1
1
1
1
1
4.C. Neumann, P. Bain and G. Shaw, J. Toxicol. Environ. Health, Part A., 2007, 70, 1679.
5.R. E. Looper, M. T. C. Runnegar, R. M. Williams, Angew. Chem. Int. Ed., 2005, 44, 3879.
6.B. Poniedziałek, P. Rzymski, K. Wiktorowicz, Toxicol. In Vitro., 2014, 28, 968.
7.B. Poniedziałek, P. Rzymski, J. Karczewski, Toxicon., 2014, 87, 113.
8.B. Poniedziałek, P. Rzymski, J. Karczewski, Toxicol. In Vitro., 2015, 29, 926.
9.T.K. Mihali, R. Kellmann, J. Muenchhoff, K.D. Barrow B.A. Neilan, Appl. Environ.
Microbiol., 2008, 74, 716.
2
2
2
2
2
0.B. B. Snider, C. Xie, Tetrahedron Lett., 2000, 41, 1849.
1.D. M. Evans, P. J. Murphy, Chem. Commun., 2011, 47, 3225.
2.D. M. Evans, P. N. Horton, M. B. Hursthouse, P. J. Murphy, RSC Adv., 2014, 4, 20744.
3.Y. Wolman, P. M. Gallop J. Org. Chem., 1962, 27, 1902.
4.V. Z. Shimkin, A. K. Shirinian, D. V. Mailian, D. V. V.Lonshakov, M. M. Gorokhov, M.
Krayushkin, Russ. Chem. Bull., 2011, 60, 139.
2
2
2
5.C. Nathan, Nat. Rev. Immunol., 2006, 3, 173.
6.B. Poniedziałek, P. Rzymski, M. Kokociński, J. Karczewski, Chemosphere., 120, 608.
7.P. Rzymski, P. Niedzielski, N. Kaczmarek, T. Jurczak, P. Klimaszyk, Harmful Algae,
2
015, 46, 34.
2
8.P. Rzymski, B. Poniedziałek, J. Mankiewicz-Boczek, E.J. Faassen, T. Jurczak, I. Gągała-
Borowska, A. Ballot, M. Lürling, M. Kokociński, Algal Res., 2017, 24, 72.
9.D. Lobner, P. M. Piana, A. K. Salous, R. W. Peoples, Neurobiol. Dis., 2007, 25, 360.
0.A. Rymuszka, A. Sieroslawska, Endocrinol. Lett. 2011, 32, Suppl. 1, 84.
1.B. Poniedziałek, P. Rzymski, J. Karczewski, Toxicol. In Vitro., 2015, 29, 926.
2.S. Halliwell, J. M. C. Gutteridge, Free Radicals in Biology and Medicine., Oxford
University Press, New York, 4th Ed., 2007.
2
3
3
3
3
3
3
3
3
3
3
3.P. C. Burcham, Mutagenesis, 1998, 13, 287.
4.A. Ayala, M. F. Muñoz, S. Argüelles, Oxid. Med. Cell. Longev., 2014, 360438.
5.P. Rzymski, B. Poniedziałek, Water Res., 2014, 66, 320.
6.X. Shen, P.K. Lam, G.R. Shaw, W. Wickramasinghe W, Toxicon 2002, 40, 1499.
7.B. Žegura, G. Gajski, A. Štraser, V, Garaj-Vrhovac V, Toxicon, 2011, 58, 471.
8.G.M. Cooper, 2002, Sunderland, Sinauer Associates.
9.L. Hong-Xi, C. Mei-Ling, W. He-Ming, Y. Xiao-Juan, R. Zhi-Gang, and L. Jian-Ping,
Organometallics., 2011, 30, 208.

ACCEPTED MANUSCRIPT
2
4

ACCEPTED MANUSCRIPT
2
5

ACCEPTED MANUSCRIPT
2
6

ACCEPTED MANUSCRIPT
2
7

ACCEPTED MANUSCRIPT
2
8

ACCEPTED MANUSCRIPT
2
9
TOC
Analogues 11a-c of the alkaloid cylindrospermopsin 1, containing the key guanidine, hydroxyl and uracil
motifs, have shown comparable cytotoxic activity to the natural material
Graphical abstract