Characterization of novel kainic acid analogs as inhibitors of select microglial functions

Article abstract of DOI:10.1016/j.ejphar.2019.02.025

Alzheimer's disease (AD) is characterized by abnormal accumulation of extracellular amyloid beta protein (Aβ) plaques and intracellular neurofibrillary tangles, as well as by a state of chronic inflammation in the central nervous system (CNS). Adverse activation of microglia, the brain immune cells, is believed to contribute to AD pathology including excessive neuronal death. Thus, normalizing immune functions of microglia could slow neurodegeneration, and identification of novel compounds capable of modifying microglial functions is an important goal. Since kainic acid (KA) has been shown to modulate microglial morphology and immune functions, we synthesized six new KA analogs (KAAs) and tested their effects on select microglial functions by using three different cell types as microglia models. Four of the KAAs at low micromolar concentrations inhibited secretion of cytotoxins, monocyte chemoattractant protein (MCP)?1, reactive oxygen species and nitric oxide (NO) by immune-stimulated microglia-like cells. We hypothesize that the effects of the novel KAAs on microglia-like cells are not mediated by KA receptors since their biological activity was distinct from that of KA in all assays performed. A structural similarity search identified aldose reductase (AR) as a potential target for the novel KAAs. This hypothesis was supported by use of AR inhibitor zopolrestat, which abolished the inhibitory effects of two KAAs on microglial secretion of NO. Since the newly developed KAAs inhibited pro-inflammatory and cytotoxic functions of microglia, they should be further investigated for their potential beneficial effect on neuroinflammation and neurodegeneration in AD animal models.

Full text of DOI:10.1016/j.ejphar.2019.02.025

Author’s Accepted Manuscript
Characterization of novel kainic acid analogs as
inhibitors of select microglial functions
Morgan A. Alford, Zhenlin Tian, Frederic Menard,
Andis Klegeris
www.elsevier.com/locate/ejphar
PII:
S0014-2999(19)30122-0
https://doi.org/10.1016/j.ejphar.2019.02.025
EJP72238
DOI:
Reference:
To appear in:
European Journal of Pharmacology
Received date: 24 October 2018
Revised date: 13 February 2019
Accepted date: 14 February 2019
Cite this article as: Morgan A. Alford, Zhenlin Tian, Frederic Menard and Andis
Klegeris, Characterization of novel kainic acid analogs as inhibitors of select
microglial
functions, European
Journal
of
Pharmacology,
https://doi.org/10.1016/j.ejphar.2019.02.025
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 galley proof before it is published in its final citable 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.

Characterization of novel kainic acid analogs as inhibitors of select microglial functions
Morgan A. Alforda, Zhenlin Tianb, Frederic Menardb, Andis Klegerisa,*
aDepartment of Biology, University of British Columbia Okanagan Campus, Kelowna, BC,
Canada.
bDepartment of Chemistry, University of British Columbia Okanagan Campus, Kelowna, BC,
Canada.
*Correspondence to: Department of Biology, University of British Columbia Okanagan Campus,
3187 University Way, Kelowna, BC V1V 1V7, Canada; andis.klegeris@ubc.ca
Abstract
Alzheimer’s disease (AD) is characterized by abnormal accumulation of extracellular
amyloid beta protein (A) plaques and intracellular neurofibrillary tangles, as well as by a state
of chronic inflammation in the central nervous system (CNS). Adverse activation of microglia,
the brain immune cells, is believed to contribute to AD pathology including excessive neuronal
death. Thus, normalizing immune functions of microglia could slow neurodegeneration, and
identification of novel compounds capable of modifying microglial functions is an important
goal. Since kainic acid (KA) has been shown to modulate microglial morphology and immune
functions, we synthesized six new KA analogs (KAAs) and tested their effects on select
microglial functions by using three different cell types as microglia models. Four of the KAAs at
low micromolar concentrations inhibited secretion of cytotoxins, monocyte chemoattractant
protein (MCP)-1, reactive oxygen species and nitric oxide (NO) by immune-stimulated
microglia-like cells. We hypothesize that the effects of the novel KAAs on microglia-like cells
are not mediated by KA receptors since their biological activity was distinct from that of KA in
all assays performed. A structural similarity search identified aldose reductase (AR) as a
potential target for the novel KAAs. This hypothesis was supported by use of AR inhibitor
zopolrestat, which abolished the inhibitory effects of two KAAs on microglial secretion of NO.
Since the newly developed KAAs inhibited pro-inflammatory and cytotoxic functions of
1

microglia, they should be further investigated for their potential beneficial effect on
neuroinflammation and neurodegeneration in AD animal models.
Keywords: aldose reductase; anti-inflammatory drugs; neurodegenerative disease;
neuroinflammation; neuroprotection; zopolrestat
1. Introduction
Alzheimer’s disease (AD) is characterized by abnormal accumulation of amyloid beta
protein (A) plaques and neurofibrillary tangles; however, other pathological mechanisms, such
as instability of neurotransmission and synapses, could play an early role (Frere and Slutsky,
2018). AD is characterized by a state of chronic inflammation in the central nervous system
(CNS) including adverse activation of microglia, the immune cells of the brain (McKenzie et al.,
2017; Takahashi et al., 2017). Over-activated microglia secrete excessive amounts of
inflammatory cytokines and reactive oxygen species, which can disturb glutamatergic
neurotransmission and lead to excitotoxicity (Clark and Vissel, 2017; Wang et al., 2014b). As
neurons degenerate, they release factors that further activate microglia, and a cycle of chronic
neuroinflammation is established.
The neuroinflammation hypothesis proposes that inflammation triggered by AD-specific
molecular mechanisms drives the progression of neurodegeneration; therefore, anti-inflammatory
therapies are being considered for prevention and treatment of AD (McKenzie et al., 2017,
Morris et al., 2014). Numerous epidemiological studies have linked long-term use of non-
steroidal anti-inflammatory drugs (NSAIDs) that inhibit cyclooxygenases to a reduced risk of
developing AD (Wang et al., 2014a). However, NSAIDs have not been effective in most clinical
trials (Mullane and Williams, 2012). These mixed results warrant further research to identify
novel anti-inflammatory agents, with mechanisms of action different from NSAIDs, as
therapeutic options for AD (Drachman et al., 2014).
Kainic acid (KA) is a naturally occurring analog of glutamate, the principal excitatory
neurotransmitter in the CNS, which activates kainate receptors in neurons (Zheng et al., 2011). In
microglia, KA was found to modify cell morphology (Christensen et al., 2006), as well as their
expression and secretion of both pro- and anti-inflammatory cytokines and markers (Banerjee et
2

al., 2015; Hong et al., 2010; Zheng et al., 2010). Accordingly, we aimed to design analogs of KA
that prevent adverse activation of immune-stimulated microglia. Six novel kainic acid analogs
(KAAs) were synthesized and assessed for their ability to modulate select microglial functions in
vitro. These novel KAAs differ from KA: their substituents are less polar, much larger, or both
(Fig. 1). A molecular similarity search returned hits for aldose reductase (AR) inhibitors,
suggesting that the KAAs reported here may share functional group complementarity to the AR
active site (Mylari, 1992; Chatzopolou et al., 2014). AR is an enzyme that mediates
hyperglycemia-induced oxidative stress when reducing glucose to sorbitol by producing
metabolites, such as glutanionyl-1,4-dihydroxynonanol (GS-DHN), which enhance inflammation
through activation of the nuclear factor (NF)-κB (Baba et al., 2013; Tang et al., 2012; Verdile et
al., 2015). In the brain, AR is expressed constitutively by microglia, and is upregulated under
oxidative stress. AR also detoxifies reactive aldehydes such as 3,4-
dihydroxyphenylacetaldehyde, and its deficiency has been linked to Parkinson’s disease (Yeung
et al., 2017; Zeng et al., 2013). In the present study KAA1-4 are shown to downregulate
microglial secretion of cytotoxins, monocyte chemoattractant protein (MCP)-1, reactive oxygen
species and nitric oxide (NO) at low micromolar concentrations. We hypothesize that these
KAAs inhibit select microglial functions by interacting with AR, and not KA receptors.
2. Materials and methods
2.1 Reagents
The following reagents were obtained from Sigma Aldrich (Oakville, ON, Canada): 3-
(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT), dimethyl sulfoxide
(DMSO), extravidin alkaline phosphatase (#E2636), lipopolysaccharide (LPS; from Escherichia
coli O55:B5, #L2880), luminol, N-(1-naphthyl)ethylenediamine dihydrochloride, N-formyl-Met-
Leu-Phe (fMLP), phosphate buffered saline (PBS), phosphatase substrate tablets, sulfanilamide,
triton X-100 and the AR inhibitor zopolrestat. KA was obtained from Tocris/Bio-Techne
(Oakville, ON, Canada). Human interferon gamma (IFN-) and the MCP-1 enzyme-linked
immunosorbent assay (ELISA) kit were purchased from Peprotech (Rocky Hill, NJ, USA). All
other reagents were obtained from ThermoFisher Scientific (Ottawa, ON, Canada) unless
otherwise stated.
3

2.2 Cell culture
THP-1 and HL-60 cell lines were purchased from the American Type Culture Collection
(ATCC, Manassas, VA, USA), and the BV-2 cell line was a gift from Dr. G. Garden (University
of Washington, Seattle, WA, USA). The human neuroblastoma SH-SY5Y cell line, used as a
neuron model, was generously donated by Dr. R. Ross (Fordham University, Bronx, NY, USA).
All cells were grown in Dulbecco’s modified Eagle medium nutrient mixture F-12 Ham
(DMEM-F12) supplemented with 10% calf bovine serum (CBS) and penicillin/streptomycin
x100 (PSA) (F10 medium). The cells were cultured in T-75 flasks and incubated at 37 ºC in a
humidified 5% CO2 and 95% air atmosphere. For experiments, cells were cultured in DMEM-
F12 supplemented with PSA and 5% CBS (F5 medium) unless stated otherwise. All cell lines
were used without differentiation, except HL-60 cells, which were treated with DMSO as
described in 2.7.
2.3 Synthesis of novel kainic acid analogs (KAAs)
For general techniques, detailed experimental procedures, and complete characterization
of KAA1, KAA2, KAA5, KAA6, 9-hydroxyphenylfluorenyl (PfOH), and 9-phenylfluorene
(PfH) see Tian and Menard (2018). The synthesis of KAA3 and KAA4 is outlined in Fig. 1.
Briefly: KAA1 (101 mg, 0.203 mmol) was hydrogenolysed using a solid-phase Pd/C catalyst (5
mg) in methanol (10 ml). The solution was degassed and charged with a hydrogen atmosphere.
After 24 h, the reaction solution was filtered through a celite pad. The filtrate was concentrated
to dryness under reduced pressure. The residue was triturated with methanol (15 ml) and
afforded a white precipitate. The precipitate was collected by filtration to yield hydrogenated
phenylfluorenyl (PfH) (47 mg, 96%) as a white solid. The filtrate was concentrated to yield
KAA3 (48 mg, 92%) as a pale-yellow liquid. KAA3: 1H NMR (400 MHz, chloroform-d) δ 3.98
(d, J = 9.7 Hz, 1H), 3.79 (s, 3H), 3.71 (s, 1H), 3.59 (d, J = 18.0 Hz, 1H), 3.40 (d, J = 18.0 Hz,
1H), 2.77–2.70 (m, 2H), 2.57-2.53 (m, 1H), 1.44 (s, 9H). KAA4 was prepared according to the
procedure described for KAA3 by using KAA2 as the starting material (Fig. 1C). Using KAA2
1
(62 mg, 0.13 mmol) afforded KAA4 as a pale-yellow liquid (28 mg, 93%). KAA4: H NMR
(400 MHz, chloroform-d) δ 3.91 (d, J = 10.1 Hz, 1H), 3.87 (t, J = 5.4 Hz, 2H), 3.78 (s, 3H), 3.48
(d, J = 7.0 Hz, 1H), 3.34 (d, J = 7.4 Hz, 1H), 2.99 (dd, J = 4.7, 2.6 Hz, 1H), 2.71 (t, J = 5.4 Hz,
2H), 2.64 (dt, J = 10.6, 5.4 Hz, 2H).
4

2.4 THP-1 cell toxicity toward neuronal SH-SY5Y cells
Human THP-1 cells were selected as they have been shown to be good models for
studying microglial cytotoxic responses and cytokine secretion (Gouveia et al., 2017; Klegeris et
al., 2005). THP-1 cells were seeded into 24-well plates at a concentration of 5 x 105 cells/ml in
one ml F5 medium. THP-1 cells were then treated with the synthetic compounds KAA1-4, KA,
and PfOH (1-20 M), or their vehicle (0.2% DMSO), 15 min prior to activation with LPS (0.2
ng/ml) plus IFN- (150 U/ml), or their PBS vehicle. LPS and IFN- were used as potent immune
stimuli to model microglial activation and detect the biological activity of KAAs. At the
concentration used, DMSO alone did not affect any of the parameters studied across the different
cell types. Following 48 h incubation, the viability of THP-1 cells was measured using the MTT
assay. Additionally, 0.5 ml of THP-1 cell supernatant were transferred onto SH-SY5Y cells,
which had been plated 24 h earlier at a concentration of 2 x 105 cells/ml in 0.5 ml F5 medium.
Following a 72 h incubation period in the THP-1 supernatants, the viability of the SH-SY5Y
cells was measured by the MTT assay.
2.5 Cell viability assay
The cell viability assay measures the conversion of MTT to purple formazan by viable
cells. It was performed as described previously by Hansen et al. (1989). Briefly, MTT (0.5
mg/ml) was added to cells in a 24-well plate for one h at 37 ºC under a 5% CO2 atmosphere.
Dark formazan crystals that had formed were dissolved by adding an equal volume of SDS/DMF
(20% sodium dodecyl sulfate, 50% N,N-dimethyl formamide solution) for three h. 100 l
aliquots were transferred to 96-well plates and optical densities measured at 570 nm with a
FLUOstar Omega microplate reader (BMG Labtech, Nepean, ON, Canada). The background
absorbance of MTT added to cell culture media was corrected for, and data were normalized to
the absorbance values from wells containing viable unstimulated control cells not exposed to any
of the compounds.
2.6 Measurement of MCP-1 secretion by ELISA
ELISA was used to monitor the secretion of MCP-1, a prototypical pro-inflammatory
chemotactic cytokine that is both secreted and recognized by microglia (Yang et al., 2011). THP-
5

1 cells were plated, treated and stimulated for 48 h as described in 2.4. MCP-1 ELISA was
performed according to the manufacturer’s (Peprotech) protocol.
2.7 Measurement of reactive oxygen species secretion by chemiluminescence (CHL)
Differentiated HL-60 cells were chosen to model a microglial respiratory burst response;
these cells have been shown to express a high level of functional NADPH oxidase (Collins et al.,
1978). A previously described chemiluminescence (CHL) assay was carried out with minor
modifications (Gouveia et al., 2017). Prior to experimentation, HL-60 cells were differentiated
by 1.3% DMSO in a six cm dish for five days. HL-60 cells were washed and seeded into 96-well
plates at a concentration of 1 x 106 cells/ml in 0.25 ml phenol red-free DMEM-F12
supplemented with 2% CBS (F2 medium). HL-60 cells were treated with the synthetic KAA1-4
compounds, KA, and PfOH (1-20 M), or their vehicle (0.2% DMSO), for 15 min prior to
priming their respiratory burst response with LPS (0.5 g/ml in PBS) for 24 h. 165 l of the
medium were removed without disturbing cells and the plates were placed in a FLUOstar Omega
microplate reader. CHL intensity in each well was measured after sequential injections of
luminol (10 mg/ml dissolved in PBS) and fMLP (20 M dissolved in PBS). fMLP is an
established inducer of respiratory burst response in phagocytes. Viability of HL-60 cells was
assessed by the MTT assay in parallel wells as described in 2.5. CHL was quantified as described
previously and the data were normalized to the signal obtained from LPS-primed and fMLP-
stimulated control cells not exposed to the compounds.
2.8 Measurement of NO secretion
Murine BV-2 microglia were used to study NO secretion because this cell type has been
shown to robustly respond to LPS and other inflammatory stimuli with the upregulated
expression of inducible NO synthase (iNOS). BV-2 cells were seeded into 24-well plates at a
concentration of 2 x 105 cells/ml in 0.5 ml of F5 medium and incubated for 24 h to allow for
their adherence. F5 medium was replaced with a fresh aliquot and cells were treated with
compounds KAA1-6, KA, and PfOH (1-20 M), or their vehicle (0.2% DMSO), 15 min prior to
activation with LPS (0.2 g/ml), or its PBS vehicle solution. After 24 h incubation, 50 l
aliquots of cell-free BV-2 cell supernatant were transferred to 96-well plates and an equal
volume of Griess reagent was added (1% sulfanilamide, 0.1% N-(1-napthyl)ethylenediamine
6

dihydrochloride, 2.5% phosphoric acid). In this assay, the Griess reagent reacts with nitrite ions,
which are the stable primary breakdown products of NO, thereby causing a colorimetric change
that can be measured at 550 nm (Yoshida and Kasama, 1987). Viability of the BV-2 cells was
assessed by the MTT assay in parallel wells as described in 2.5.
2.9 Aldose reductase inhibition
BV-2 microglial cells were plated as described in 2.8 and incubated with the competitive
AR inhibitor zopolrestat (100 M), or its vehicle (0.05% DMSO), for 15 min. Cells were then
treated with KAA1 or KAA2 (1-20 M) for 15 min prior to activation with LPS (0.2 g/ml), or
its PBS vehicle. After 24 h incubation, 50 l aliquots of cell-free BV-2 supernatants were
transferred to 96-well plates and the Griess assay was performed. Viability of the BV-2 cells was
assessed by the MTT assay in parallel wells as described in 2.5.
2.10 Statistical analysis
GraphPad Prism software (version 7.0, GraphPad Software Inc., La Jolla, CA, USA) was
used to conduct statistical analyses. The randomized-block design one-way analysis of variance
(ANOVA) followed by Dunnett’s post-hoc test was performed to detect differences between
treatment groups. Significance was established at P < 0.05.
3. Results
3.1 Effect of compounds on THP-1 cytotoxicity toward SH-SY5Y neuronal cells
The ability of the novel KAAs to reduce the secretion of cytotoxins by THP-1 monocytic
cells was studied first. Fig. 2 shows that exposure of neuronal SH-SY5Y cells to the supernatants
from THP-1 cells stimulated with LPS plus IFN- without KAA (0 M) causes significant
reduction in viability of neuronal cells when compared to cells exposed to supernatants from
unstimulated THP-1 cells (vehicle). Treatment of THP-1 cells with KAA1-4 prior to cell
stimulation with LPS and IFN- reduces their cytotoxicity toward SH-SY5Y cells (Fig. 2A-D).
At 20 M, KAA1-4 reduce THP-1 cytotoxicity toward SH-SY5Y cells by 83%, 72%, 64% and
73%, respectively. Fig. 2E and F demonstrate that KA and PfOH do not have significant effects
on the cytotoxicity of stimulated THP-1 cells toward SH-SY5Y cells. For all concentrations
studied, none of KAA1-4, KA, or PfOH affected the viability of stimulated THP-1 cells (Fig.
7

3A-F). KAA1-4 were also applied directly to SH-SY5Y cells to investigate whether they affect
neuronal cell viability. At the concentrations studied (up to 20 M), no significant effects on SH-
SY5Y cell viability were observed according to the MTT assay (data not shown).
3.2 Effect of compounds on secretion of MCP-1 by THP-1 cells
Fig. 4A-D show that treatment of THP-1 cells with KAA1-4 prior to cell stimulation with
LPS and IFN- significantly reduces secretion of MCP-1, when compared to stimulated THP-1
cells treated with the vehicle only (0 M). The inhibitory effects of KAA1 and KAA3 are
significant at 1 M and higher, whereas KAA2 and KAA4 are inhibitory at 5 M and 20 M. At
20 M, KAA1-4 reduce the concentration of MCP-1 in supernatants of stimulated THP-1 cells
by 17%, 40%, 54% and 51%, respectively. Fig. 4E and 4F show that at all concentrations
studied, KA and PfOH have no significant effects on the concentration of MCP-1 in supernatants
of stimulated THP-1 cells. At the concentrations studied, KAA1-4, KA and PfOH do not affect
the viability of stimulated THP-1 cells (Fig. 3A-F).
3.3 Effect of compounds on secretion of reactive oxygen species by primed HL-60 cells
Fig. 5A-D show that treatment of HL-60 cells with KAA1-4 prior to their priming with
LPS and their stimulation with fMLP inhibits the CHL response of cells, when compared to the
CHL response of LPS-primed and fMLP-stimulated HL-60 cells that have been treated with
vehicle only (0 M). KAA1 inhibits the CHL response of primed and stimulated cells at 1 M
and higher, whereas KAA2-4 are inhibitory at 5 M and 20 M. At 20 M, KAA1-4 reduce the
CHL response of primed and stimulated cells by 50%, 53%, 42% and 34%, respectively. At the
concentrations studied, KA and PfOH have no significant effects on the CHL response of primed
and stimulated cells (Fig. 5E, F). The addition of KAA1-4, KA and PfOH to LPS-primed HL-60
cells just prior to stimulation with fMLP has no significant effects on the CHL response of cells
(data not shown). At the concentrations studied, KAA1-4, KA and PfOH do not affect the
viability of primed and stimulated HL-60 cells (Fig. 6A-F).
3.4 Effect of compounds on secretion of NO by BV-2 microglia
Fig. 7A-D show that treatment of BV-2 microglia with KAA1-4 prior to LPS stimulation
significantly inhibits the secretion of NO, when compared to the LPS-stimulated cells treated
8

with the vehicle only (0 M). KAA1 and KAA3 significantly inhibit the secretion of NO by cells
at 20 M, whereas KAA2 and KAA4 are inhibitory at 5 M and 20 M. At 20 M, KAA1-4
inhibit secretion of NO by 31%, 53%, 27% and 14%, respectively. At the concentrations studied,
PfOH, KAA5 and KAA6 have no significant effects on the secretion of NO by BV-2 cells (Fig.
7F-H). Fig. 7E shows that treatment of BV-2 cells with 5 or 20 M KA prior to LPS stimulation
enhances NO secretion, when compared to LPS-stimulated cells treated with the vehicle only (0
M). At the concentrations studied, KAA1-6, KA and PfOH do not affect the viability of LPS-
stimulated BV-2 microglia (Fig. 8A-H). Nitrite was not detected in the supernatants of
unstimulated BV-2 cells (data not shown). Detectable secretion of NO was not induced when
KAA1-6, KA and PfOH were applied to BV-2 cells on their own (1 - 20 M, data not shown).
3.5 Effect of compounds on secretion of NO by BV-2 microglia in the presence of zopolrestat
Fig. 9 illustrates that, similar to the results shown on Fig. 7, treatment of BV-2 cells with
KAA1 or KAA2, prior to cell stimulation with LPS, inhibits their secretion of NO, when
compared to LPS-stimulated cells treated with the vehicle only (0 M). However, when the cells
are sequentially pre-incubated with the competitive AR inhibitor zopolrestat (100 M), then
treated with KAA and stimulated with LPS, the inhibitory effects of KAA1 and KAA2 on NO
secretion are abolished (Fig. 9A, B). Cell viability of LPS-stimulated BV-2 microglia was not
affected by pre-incubation of cells with 100 M zopolrestat followed by LPS-stimulation, nor by
KAA1, KAA2, or zopolrestat at the studied concentrations and combinations (Fig. 10A, B).
4. Discussion
Natural kainoids that bind to KA receptors are usually neurotoxic and are therefore
undesirable candidates to develop AD therapies. However, in the current study, the modifications
made to the kainoid core molecule should prevent KAA1-6 from acting through KA receptors,
thereby possibly eluding neurotoxic effects (Majik and Mandrekar, 2016). Indeed, the key
interactions between KA and its receptors are well defined (Protein Data Bank identifier: PDB
4E0X; Krogsgaard-Larsen et al., 2015). Compounds KAA1-4 were synthesized as part of a
different research program (Tian and Menard, 2018), yet their low molecular weight and
desirable lipophilic characteristics suggest they may cross the blood-brain barrier. The only
9

conserved structural elements between the natural kainoids and the KAAs reported here are: the
presence of a carboxyl group at the C2 position of the pyrrolidine ring, and a trans relationship
between the C2-C3 substituents. The C4 alkene substituent of KA is replaced by a simple ketone
(KAA1-5), or its stereochemistry is inverted (KAA6). Two of the novel analogs (KAA2 and
KAA4) also differ from kainoids at the C3-side chain: the usual carboxymethyl group is replaced
by a 2’-ketobutan-4’-ol sidechain. The nitrogen of the pyrrolidine core is alkylated with a large
polyaromatic group (9-phenylfluorene) in four compounds (KAA1/2/5/6), making these KAAs
much more hydrophobic than KA (cLogPs range from 3.4 to 6.8 in contrast to –0.6 for KA). In
KAA3 and KAA4, the nitrogen can participate in hydrogen bonds and/or be protonated. Finally,
compounds PfOH and PfH were tested to ensure that the observed effects are not due to
inadvertent release of the polyaromatic portion of the molecules under physiological conditions;
both were found to be inactive.
Based on the above structural disparity between the novel KAAs and typical kainoids, we
hypothesize that the newly synthesized KAAs do not act through KA receptors. Consistent with
this hypothesis, we observed that KAA1-4 are not toxic to SH-SY5Y cells at the concentrations
studied. In contrast, KA is toxic to SH-SY5Y cells (Cannarsa et al., 2008). In addition, KAA1-4
compounds exhibit a biological activity different from that of KA in each of the assays
performed, further supporting our hypothesis that KA receptors are not the primary target of
these KAAs.
The maximal inhibition exhibited by KAA2 is superior to that of KAA1, which suggests
that a relatively polar hydroxylic substituent at C3 of the pyrrolidine ring enhances the affinity of
the compounds for their target. Moreover, the presence of a functional group at the C3 position
and a ketone at the C4 position of the pyrrolidine ring may be essential to the biological activity
of KAAs since KAA5 and KAA6 have no effect on the secretion of NO. Our data suggest that
the mechanism of action may differ slightly between the compounds tested, since addition of a
N-phenylfluorenyl group on the pyrrolidine ring of KAAs differentially modulates the inhibitory
activity of KAAs on microglial secretions. For example, removal of the large, lipophilic N-
phenylfluorenyl protecting group from KAAs enhances the inhibitory effect exhibited by the
compound with an ester-containing functional group at C3 (KAA1 versus KAA3), but does not
impact the effect exhibited by the compound with a hydroxylic substituent at the same position
(KAA2 versus KAA4).
10

We hypothesized that the inhibitory activity of KAA1-4 on microglial cells is at least
partially due to AR inhibition because of the structural analogy between KAA1-4 and ethyl 1-
(9H-fluoren-9-yl)-4,5-dioxopyrrolidine-3-carboxylate, which is an established AR inhibitor
(Chemical Abstracts Service identifier: 142774-45-6; Mylari, 1992). This hypothesis is
supported by our data showing that the selective AR inhibitor zopolrestat abolishes the inhibitory
effects of KAA1 and KAA2 on microglial NO secretion. Since AR mediates hyperglycemia-
induced oxidative stress, AR inhibitors have been used clinically for treatment of complications
associated with type II diabetes mellitus (Chatzopoulou et al., 2013). More recently, however,
AR inhibitors have entered the research spotlight as possible therapeutics for inflammatory
disorders such as rheumatoid arthritis, asthma and AD, due to findings that AR produces pro-
inflammatory GS-DHN (Frohnert et al., 2014; Song et al., 2017). Consistent with our data
showing the anti-inflammatory potential of KAAs due to inhibition of MCP-1 secretion, previous
studies have demonstrated that AR inhibitors reduce the secretion of two other inflammatory
cytokines, tumor necrosis factor (TNF)-α and interleukin (IL)-1β, by BV-2 cells, RAW264.7
macrophages or retinal microglia (Chang et al., 2014; Song et al., 2017; Zeng et al., 2013). AR
activity has been linked to activation of NF-B as well as upregulated microglial secretion of
MCP-1, reactive oxygen species and NO (Ramana et al., 2004; Song et al., 2017; Zeng et al.,
2013), indicating that the inhibitory effects of KAAs may include downregulation of NF-B
activity in microglia.
KAA1-4 described in this study have the potential to slow down neuroinflammation in
AD by inhibiting microglial cytotoxicity, secretion of MCP-1, reactive oxygen species and NO.
MCP-1 can be secreted by microglia in response to tissue lesions to promote the chemotaxis of
monocytes and other glial cells toward the site of injury for mounting an immune response
(Azizi et al., 2015; Clark and Vissel, 2017). In diseased brains, however, over-activated
microglia secrete MCP-1 at high concentrations that augment the deposition of A plaques and
worsen AD pathology (Porcellini et al., 2013). Reactive oxygen species production is increased
in AD due to age-related mitochondrial deterioration and microglial activation (Wang et al.,
2014b). High concentrations of reactive oxygen species hinder a broad spectrum of cellular
processes, which may cause further cellular damage and exacerbation of AD-associated
neurodegeneration. NO is yet another important CNS messenger that has been implicated in the
progression of AD when produced at high concentrations (Azizi et al., 2015; Balez and Ooim,
11

2015). The expression of iNOS, which is the enzyme that produces NO, is upregulated in
microglia exposed to harmful stimuli such as A. NO causes direct neurotoxicity, similarly to
reactive oxygen species, by disrupting mitochondrial integrity (Azizi et al., 2015; Wang et al.,
2014b).
Elevated levels of NO (Yuste et al., 2015), reactive oxygen species (Cruz-Haces et al.,
2017) and inflammatory cytokines (Viviani et al., 2014) have also been linked to enhanced
excitotoxicity, thereby increasing neuronal death indirectly. Excitotoxicity involves excessive
activation of the neuronal N-methyl-D-aspartate (NMDA) receptors by glutamate leading to a
large calcium influx, which causes loss of mitochondrial and nuclear functions as well as
activation of calcium-dependent proteases (Meldrum and Garthwaite, 1990). Excitotoxicity is
facilitated by microglial NO and reactive oxygen species blocking the reuptake of glutamate,
which increases its concentration in the synaptic cleft (Cruz-Haces et al., 2017; Yuste et al.,
2015). High levels of inflammatory cytokines facilitate glutamate release from glial cells and
enhance excitotoxic neuronal death by modulating the expression, phosphorylation and activity
of NMDA receptors (Viviani et al., 2014).
By inhibiting secretion of cytotoxins and several different microglial inflammatory
mediators, the newly synthesized KAA1-4 have the potential to reduce neuronal death in
neurodegenerative diseases, including AD, which have a strong neuroinflammatory component;
however, further preclinical studies will be required to establish this activity in AD animal
models. Further in vivo testing of the novel KAAs is especially important, as microglial
populations are heterogeneous and they display functional diversity along spatial and temporal
axes (Hanisch, 2013). This could lead to potentially competing physiological effects: the
activation of certain microglial subtypes may be detrimental, while activation of other subtypes
could be beneficial for CNS homeostasis. Current in vitro models of microglial activation cannot
reliably predict clinically beneficial modulators since the gene expression patterns in these
different microglia subtypes are not well characterized at the single-cell level (Silvin and
Ginhoux, 2018). Recent experiments using PLX5622, an inhibitor of colony-stimulating factor 1
receptor that depletes microglia, further illustrate the opposite roles that microglia sometimes
play in neuropathologies. For example, depletion of microglia can avert plaque deposition and
cognitive decline in AD animal models (Dagher et al., 2015; El-Gamal et al., 2018), or prevent
disruption of the blood-retina barrier during chronic inflammation induced by LPS (Kokona et
12

al., 2018), but it is detrimental in viral encephalitis (Waltl et al., 2018) and prion disease (Carroll
et al., 2018). Pharmacological modulators of microglia functions may have drastically different
clinical effects in neuropathologies depending on the overall role neuroinflammation and
microglia play in each specific case. We posit that the compounds reported herein provide new
tools to investigate the complex functions of microglia in neurodegenerative pathologies.
5. Conclusions
Dysregulated microgliosis contributes to neuroinflammation and neurotoxicity; therefore,
therapeutics that attenuate the adverse activation of microglia could be used to slow down the
progression of AD. In a search for small molecules that can regulate microglial immune
functions, we synthesized a series of compounds based on the core structure of KA. We studied
their effects in three different microglia cell models: THP-1, HL-60 and BV-2 cells, which were
chosen due to their robust responses in four different functional assays measuring secretion of
cytotoxins, the chemokine MCP-1, reactive oxygen species and NO. We show that four of the
eight compounds tested inhibit the above microglial functions at low micromolar concentrations.
By using a selective AR inhibitor we show that the inhibitory action of bioactive KAAs is at
least partially mediated by this enzyme. Such inhibitory effects on microglia can have an indirect
beneficial outcome on neuron survival. Thus, KAA1-4 identified herein may represent novel
candidate molecules for future preclinical studies aimed at developing anti-AD therapeutics.
Funding support
This work was supported by grants from the Natural Sciences and Engineering Research
Council of Canada (NSERC), the Jack Brown and Family Alzheimer’s Disease Research
Foundation and the University of British Columbia (UBC) Okanagan campus.
Conflict of interest
The authors have declared that there is no conflict of interest.
13

References
Azizi, G., Navabi, S., Al-Shukaili, A. Seyedzadeh, M.H., Yazdani, R., Mirshafiey A. 2015. The
role of inflammatory mediators in the pathogenesis of Alzheimer’s disease. Sultan
Qaboos Univ. Med. J. 15, 305-316. doi: 10.18295/squmj.2015.15.03.002
Baba, S.P., Hoetker, J.D., Merchant, M., Klein, J.B., Cai, J., Barski, O.A., Conklin, D.J.,
Bhatnagar, A. 2013. Role of aldose reductase in the metabolism and detoxification of
carosine-acrolein conjugates. J. Biol. Chem. 288, 28163-28179. doi:
10.1074/jbc.M113.504753
Balez, R., Ooim, L. 2015. Getting to NO Alzheimer’s disease: neuroprotection versus
neurotoxicity mediated by nitric oxide. Oxid. Med. Cell. Longev. 1, 1-8. doi:
10.1155/2016/3806157
Banerjee, M., Sasse, V.A., Wang, Y., Maulik, M., Kar, S. 2015. Increased levels and activity of
cathepsins B and D in kainate-induced toxicity. Neuroscience 22, 360-373. doi:
10.1016/j.neuroscience.2014.10.003
Cannarsa, R., Landuzzi, D., Cavina, C., Candeletti, S., Romualdi, P. 2008. Kainic acid
downregulated NOP receptor density and gene expression in human neuroblastoma SH-
SY5Y cells. J. Mol. Neurosci. 35, 171-177. doi: 10.1016/j.neuint.2012.02.015
Carroll, J.A., Race, B., Williams, K., Striebel, J., Chesebro, B. 2018. Microglia are critical in
host defence against prion disease. J. Virol. Epub 29769333 doi: 10.1128/JVI.00549-18
Chang, K.C., Ponder, J., LaBarbera, D.V., Petrash, J.M. 2014. Aldose reductase inhibition
prevents endotoxin-induced inflammatory responses in retinal microglia. Invest.
Ophthalmol. Vis. Sci. 55, 2853-2861. doi: 10.1167/iovs.13-13487
Chatzopoulou, M., Pegklidou, K., Papastavrou, N., Demopoulos, V.J. 2013. Development of
aldose reductase inhibitors for the treatment of inflammatory disorders. Exp. Opin. Drug
Discov. 8, 1365-1380. doi: 10.1517/17460441.2013.843524
Chatzopoulou, M., Parsilinakos, A., Vallianatou, T., Prnova, M.S., Zakelj, S., Ragno, R., Stefek,
M., Kristl, A., Tsantili-Kakoulidou, A., Demopoulos, V.J. 2014. Decreasing acidity in a
series of aldose reductase inhibitors: 2-fluoro-4-(1H-pyrrol-1-yl)phenol as a scaffold for
improved membrane permeation. Bioorganic. Med. Chem. 22, 2194-2207. doi:
10.1016/j.bmc.2014.02.016
14

Christensen, R.N., Ha, B.K., Bresnahan, J.C., Beattie, M.S. 2006. Kainate induces rapid
redistribution of the actin cytoskeleton in ameboid microglia. J. Neurosci. Res. 84, 170-
181. doi: 10.1002/jnr.20865
Clark, I. A., Vissel, B. 2017. The meteorology of cytokine storms, and the clinical usefulness of
this knowledge. Sem. Immunopathology 39, 505-516. doi: 10.1007/s00281-017-0628-y
Collins, S.J., Ruscetti, F.W., Gallagher, R.E., Gallo, R.C. 1978. Terminal differentiation of
human promyelocytic leukemia cells induced by dimethyl sulfoxide and other polar
compounds. Proc. Natl. Acad. Sci. U.S.A. 75, 2458-2462
Cruz-Haces, M., Tang, J., Acosta, G., Fernandez, J., Shi, R. 2017. Pathological correlations
between traumatic brain injury and chronic neurodegenerative diseases. Transl.
Neurodegener. 6, 20. doi: 10.1186/s40035-017-0088-2
Dagher, N.N., Najafi, A.R., Kayala, K.M., Elmore, M.R., White, T.E., Medeiros, R., West, B.L.,
Green, K.N. 2015. Colony-stimulating factor 1 receptor inhibition prevents microglial
plaque association and improves cognition in 3xTg-AD mice. J. Neuroinflam. 12: 139, 1-
14. doi: 10.1186/s12974-015-0366-9
Drachman, D.A. 2014. The amyloid hypothesis, time to move on: amyloid is the downstream
result, not cause, of Alzheimer’s disease. Alz. Dementia 10, 372-380. doi: 10.1016/j.jalz.
2013.11.003
El-Gamal, M.I., Al-Ameen, S.K., Al-Koumi, D.M., Hamad, M.G., Jalal, N.A., Oh, C.H. 2018.
Recent advances of colony-stimulating factor-1 receptor (CSF-1R) kinase and its
inhibitors. J. Med. Chem. 61, 5450-5466. doi: 10.1021/acs.jmedchem.7b00873
Frere, S., Slutsky, I. 2018. Alzheimer’s disease: From firing instability to homeostasis network
collapse. Neuron 97, 32–58. doi: 10.1016/j.neuron.2017.11.028
Frohnert, B.I., Long, E.K., Hahn, W.S., Bernlohr, D.A. 2014. Glutathionated lipid aldehydes are
products of adipocyte oxidative stress and activators of macrophage inflammation.
Diabetes 63, 89-100. doi: 10.2337/db13-0777
Gouveia, A., Bajwa, E., Klegeris, A. 2017. Extracellular cytochrome C as an intercellular
signaling molecule regulating microglial functions. Biochim. Biophys. Acta 1861(9):
2274-2281. doi: 10.1016/j.bbagen.2017.06.017
Hanisch, U.K. 2013. Functional diversity of microglia – how heterogeneous are they to begin
with? Front. Cell. Neurosci. 7: 65, 1-18. doi: 10.3389/fncel.2013.00065
15

Hansen, M.B., Nielsen, S.E., Berg, K. 1989. Re-examination and further development of a
precise and rapid dye method for measuring cell growth/cell kill. J. Immunol. Methods
119, 203-210.
Hong, J., Cho, I.H., Kwak, K. I., Suh, E. C., Seo, J., Min, H. J., Choi, S. Y., Kim, C. H., Park, S.
H., Jo, E. K., Lee, S., Lee, K. E., Lee, S. J. 2010. Microglial toll-like receptor 2
contributes to kainic acid-induced glial activation and hippocampal neuronal death. J.
Biol. Chem. 285, 39447-39457. doi: 10.1074/jbc.M110.132522
Klegeris, A., Bissonnette, C. J., McGeer, P.L. 2005. Modulation of human microglia and THP-1
cell toxicity by cytokines endogenous to the nervous system. Neurobiol. Aging 26, 673-
682. doi: 10.1016/j.neurobiolaging.2004.06.012
Kokona, D., Ebneter, A., Escher, P., Zinkernagel, M.S. 2018. Colony-stimulating factor 1
receptor inhibition prevents disruption of the blood-retina barrier during chronic
inflammation. J. Neuroinflam. 25: 340. doi: 10.1186/s12974-018-1373-4
Krogsgaard-Larsen, N., Storgaard, M., Møller, C., Demmer, C.S., Hansen, J., Han, L., Monrad
R.N., Nielsen, B., Tapken, D., Pickering, D.S., Kastrup, J.S., Frydenvang, K., Bunch, L.
2015. Structure-activity relationship study of ionotropic glutamate receptor antagonist
(2S,3R)-3-(3-carboxyphenyl)pyrrolidine-2-carboxylic acid. J. Med. Chem. 58, 6131-
6150. doi: 10.1021/acs.jmedchem.5b00750
Majik, M., Madrekar, K. 2016. Recent progress in syntheses and biological activities of kainic
acid and its derivatives. Curr. Org. Chem. 20, 939-951. doi: 10.2174/13852728
19666150804000728
McKenzie, J.A., Spielman, L.J., Pointer, C.B., Lowry, J.R., Bajwa, E., Lee, C.W., Klegeris, A.
2017. Neuroinflammation as a common mechanism associated with the modifiable risk
factors for Alzheimer’s and Parkinson’s diseases. Curr. Aging Sci. 10, 158-176. doi:
10.2174/1874609810666170315113244.
Meldrum, B., Garthwaite, J. 1990. .Excitatory amino acid neurotoxicity and neurodegenerative
disease. Trends Pharmacol. Sci. 11, 379−387. doi: 10.1016/0165-6147(90)90184-A
Morris, G.P., Clark, I.A., Vissel, B. 2014. Inconsistencies and controversies surrounding the
amyloid hypothesis of Alzheimer’s disease. Acta Neuropathol. Commun. 2, 1-21. doi:
10.1186/s40478-014-0135-5
16

Mullane, K., Williams, M. 2012. Alzheimer’s therapeutics: continued clinical failures question
the validity of the amyloid hypothesis- but what lies beyond? Biochem. Pharmacol. 85,
289-305. doi: 10.1016/j.bcp.2012.11.014
Mylari, B.L. 1992. Preparation of pyrrolidinedionecarboxylates as aldose reductase inhibitors.
PCT Int. Appl. Patent no: WO 9206954 A2 19920430
Porcellini, E., Ianni, M., Carbone, I., Francheschi, M., Licastro, F. 2013. Monocyte
chemoattractant protein-1 promoter polymorphism and plasma levels in Alzheimer’s
disease. Immun. Ageing 10, 1-6. doi: 10.1186/1742-4933-10-6
Ramana, K.V., Friedrich, B., Srivastava, S., Chatnagar, A., Srivastava, S.K. 2004. Activation of
nuclear factor-kappaB by hyperglycemia in vascular smooth muscle cells is regulated by
aldose reductase. Diabetes 53, 2910-2920. doi: 10.2337/diabetes.53.11.2910
Silvin, A., Ginhoux, F. 2018. Microglia heterogeneity along a spatio-temporal axis: more
questions than answers. Glia 66, 2045-2057. doi: 10.1002/glia.23458
Song, X.M., Yu, Q., Dong, X., Yang, H.O., Zeng, K.W., Li, J.,Tu, P.F. 2017. Aldose reductase
inhibitors attenuate -amyloid induced TNF- production in microglia via ROS-PKC-
mediated NF-B and MAPK pathways. Int. Immunopharmacol. 50, 30-37. doi:
10.1016/j.intimp.2017.06.005
Takahashi, R.H., Nagao, T., Gouras, G.K. 2017. Plaque formation and the intraneuronal
accumulation of -amyloid in Alzheimer’s disease. Pathol. Int. 67, 185-193. doi:
10.1111/pin.12520
Tang, W.H., Martin, K.A., Hwa, J. 2012. Aldose reductase, oxidative stress, and diabetic
mellitus. Front. Pharmacol. 3, 87-95. doi: 10.3389/fphar.2012.00087
Tian, Z., Menard, F. 2018. Synthesis of kainoids and C4 derivatives. J. Org. Chem. 83, 6162-
6170. doi: 10.1021/acs.joc.8b00179
Verdile, G., Keane, K.N., Cruzat, V.F., Medic, S., Sabale, M., Rowles, J., Wijesekara, N.,
Martins, R.N., Fraser, P.E., Newsholme, P. 2015. Inflammation and oxidative stress: the
molecular connectivity between insulin resistance, obesity and Alzheimer’s disease.
Mediators Inflamm. 1, 1-17. doi: 10.1155/2015/105828
Viviani, B., Boraso, M., Marchetti, N., Marinovich, M. 2014. Perspectives on neuroinflammation
and excitotoxicity: a neurotoxic conspiracy? Neurotoxicology 43, 10-20. doi:
10.1016/j.neuro.2014.03.004.
17

Waltl, I., Kaufer, C., Gerhauser, I., Chhatbar, C., Ghita, L., Klinke, U., Loscher, W. 2018.
Microglia have a protective role in viral encephalitis-induced seizure development and
hippocampal damage. Brain Behav. Immun. 74, 186-204. doi: 10.1016/j.bbi.2018.09.006
Wang, J., Tan, L., Wang, H.F., Tan, C.C., Meng, X.F., Wang, C., Tang, S.W., Yu, J.T. 2014a.
Anti-inflammatory drugs and risk of Alzheimer’s disease: an updated systematic review
and meta-analysis. J. Alzheimer Dis. 44, 385-396. doi: 10.3233/jad-141506
Wang, X., Wang, W., Li, L., Perry, G., Lee, H. G., Zhu, X. 2014b. Oxidative stress and
mitochondrial dysfunction in Alzheimer’s disease. Biochim. Biophys. Acta 1842, 1240-
1247. doi: 10.1016/jbbadis.2013.10.015
Yang, G., Meng, Y., Li, W., Yong, Y., Fan, Z., Ding, H., Wei, Y., Luo, J., Ke, Z.J. 2011.
Neuronal MCP-1 mediates microglia recruitment and neurodegeneration induced by the
mild impairment of oxidative metabolism. Brain Pathol. 21, 279-297. doi:
10.1111/j.1750-3639.2010.00445.x
Yeung, P.K.K., Lai, A.K.W., Son, H.J., Zhange, X., Hwang, O., Chung, S.S.M., Chung, S.K.
2017. Aldose reductase deficiency leads to oxidative stress-induced dopaminergic
neuronal loss and autophagic abnormality in an animal model of Parkinson's disease.
Neurobiol. Aging 50, 119-133. doi: 10.1016/j.neurobiolaging.2016.11.008
Yoshida, K., Kasama, K. 1987. Biotransformation of nitric oxide. Environ. Health Perspect. 73,
201-206.
Yuste J.E., Tarragon, E., Campuzano, C.M., Ros-Bernal, F. 2015. Implications of glial nitric oide
in neurodegenerative diseases. Front. Cell. Neurosci. 9, 322. doi:
10.3389/fncel.2015.00322
Zeng, K.W., Li, J., Dong, X., Wang, Y.H., Ma, Z.Z., Jiang, Y., Jin, H.W., Tu, P.F. 2013. Anti-
neuroinflammatory efficacy of the aldose reductase inhibitor FMHM via phospholipase
C/ protein kinase C-dependent NF-B and MAPK pathways. Toxicol. Appl. Pharmacol.
273(1), 159-171. doi: 10.1016/j.taap.2013.08.028
Zheng, H., Zhu, W., Zhao, H., Wang, X., Wang, W., Li, Z. 2010. Kainic acid-activated microglia
mediate increased excitability of rat hippocampal neurons in vitro and in vivo: crucial
role of interleukin-1 beta. Neuroimmunomodulation 17, 31-38. doi: 10.1159/000243083
18

Zheng, X.Y., Zhang, H.L., Luo, Q., Zhu, J. 2011. Kainic acid-induced neurodegenerative model:
potentials and limitations. J. Biomed. Biotechnol. 457079, 1-10. doi:
10.1155/2011/457079
19

Figure 1. Structures of the compounds tested and synthesis of KAA3 and KAA4. (A)
Compounds KAA1-6, KA, PfOH and PfH. KAA1/3/5/6, PfH and PfOH were synthesized as
previously reported (Tian and Menard, 2018). (B) KAA3 was synthesized by removing the Pf
group from KAA1. (C) KAA4 was synthesized by removing the Pf group from KAA2.
Abbreviations used: Dess-Martin periodinane (DMP), dichloromethane (DCM),
hexamethylphosphoramide (HMPA), hydrogenated protecting group (PfH), hydroxylated
protecting group (PfOH), kainic acid (KA), kainic acid analog (KAA), meta-
chloroperoxybenzoic acid (mCPBA), n-butyllithium (nBuLi), palladium/carbon catalyst (Pd/C);
protecting group (Pf), tert-butyl (tBu), tert-butyldimethylsilyl (TBS), tetrahydrofuran (THF).
Figure 2. KAA1-4 inhibit the toxicity of stimulated human monocytic THP-1 cells toward
SH-SY5Y cells. THP-1 cells were treated with KAA1-4 (A-D), KA (E), PfOH (F) or their
vehicle (0.2% DMSO) for 15 min. Cells were then stimulated with LPS (0.2 ng/ml) & IFN-γ
(150 U/ml) as indicated on the abscissa. Following 48 h incubation, THP-1 cell supernatant was
transferred to SH-SY5Y cells. Following 72 h exposure to the THP-1 cell supernatants, SH-
SY5Y cell viability was assessed by the MTT assay. Data from four independent experiments are
presented as mean % of the control cell viability ± standard error of the mean (SEM). Control
cells were treated with fresh F5 media instead of THP-1 cell supernatant (data not shown). * P <
0.05, ** P < 0.01 different from vehicle-treated, stimulated cells (0 μM) according to the
randomized block one-way ANOVA followed by Dunnett’s post-hoc test. P and F values for the
main effects of the ANOVA are shown on the graphs.
Figure 3. At the concentrations studied, compounds do not affect the viability of stimulated
human monocytic THP-1 cells. Cells were treated with KAA1-4 (A-D), KA (E), PfOH (F) or
their vehicle (0.2% DMSO) for 15 min. THP-1 cells were then stimulated with LPS (0.2 ng/ml)
& IFN-γ (150 U/ml) as shown on the abscissa. Following 48 h incubation, THP-1 cell viability
was assessed by the MTT assay. Data from four independent experiments are presented as mean
% of the unstimulated control cell viability ± SEM. Unstimulated control cells were seeded in the
absence of KAAs and solvent (data not shown). * P < 0.05 different from vehicle-treated,
stimulated cells (0 μM) according to the randomized block one-way ANOVA followed by
20

Dunnett’s post-hoc test. P and F values for the main effects of ANOVA are shown on the graphs.
Figure 4. KAA1-4 reduce the concentration of MCP-1 in the supernatants of stimulated
human monocytic THP-1 cells. Cells were treated with KAA1-4 (A-D), KA (E), PfOH (F) or
their vehicle (0.2% DMSO) for 15 min. THP-1 cells were then stimulated with LPS (0.2 µg/ml)
& IFN-γ (150 U/ml). MCP-1 (ng/ml) was measured in THP-1 cell supernatants following 48 h
incubation. Data from four independent experiments are presented as means ± SEM. * P < 0.05,
** P < 0.01 different from vehicle-treated, stimulated cells (0 μM) according to the randomized
block one-way ANOVA followed by Dunnett’s post-hoc test. P and F values for the main effects
of ANOVA are shown on the graphs. The limit of detection of the ELISA is displayed as a
dashed line at 1.1 ± 0.2 ng/ml.
Figure 5. KAA1-4 inhibit reactive oxygen species secretion by primed and stimulated
human HL-60 cells in a concentration-dependent manner. DMSO-differentiated cells were
treated with KAA1-4 (A-D), KA (E), PfOH (F) or their vehicle (0.2% DMSO) for 15 min and
then primed with LPS (0.5 µg/ml). Following 24 h incubation, the cells were stimulated with
fMLP (1 µM) and the luminol-dependent chemiluminescence (CHL) was measured. Data from
four independent experiments are presented as mean % CHL intensity of the primed and
stimulated control HL-60 cells ± SEM. Control HL-60 cells were seeded in the absence of
compounds and solvent (data not shown). * P < 0.05, ** P < 0.01 different from vehicle-treated,
stimulated cells (0 μM) according to the randomized block one-way ANOVA followed by
Dunnett’s post-hoc test. P and F values for the main effects of ANOVA are shown on the graphs.
Figure 6. At the concentrations studied, compounds do not affect the viability of primed
and stimulated human HL-60 cells. Cells were treated with KAA1-4 (A-D), KA (E), PfOH (F)
or their vehicle (0.2% DMSO) for 15 min and then primed with LPS (0.2 µg/ml). Following 24 h
incubation, HL-60 cell viability was assessed by the MTT assay. Data from four independent
experiments are presented as mean % of the primed unstimulated control cell viability ± SEM.
Control cells were primed with LPS but seeded in the absence of compounds and solvent (data
not shown). No treatment groups were different from vehicle-treated, stimulated cells (0 μM)
according to the randomized block one-way ANOVA followed by Dunnett’s post-hoc test. P and
21

F values for the main effects of ANOVA are shown on the graphs.
Figure 7. KAA1-4 reduce nitrite concentration in the supernatant of stimulated murine
BV-2 microglia. Cells were treated with KAA1-4 (A-D), KA (E), PfOH (F), KAA5 (G), KAA6
(H) or their vehicle (0.2% DMSO) for 15 min and then stimulated with LPS (0.2 µg/ml).
Following 24 h incubation, nitrite (µM) was measured in cell supernatants. Data from four
independent experiments are presented as means ± SEM. * P < 0.05, ** P < 0.01 different from
vehicle-treated, stimulated cells (0 μM) according to the randomized block one-way ANOVA
followed by Dunnett’s post-hoc test. P and F values for the main effects of ANOVA are shown
on the graphs. The limit of detection of the Griess assay is displayed as a dashed line at 1.2 ± 0.3
μM.
Figure 8. At the concentrations studied, compounds do not affect the viability of stimulated
murine BV-2 microglia. Cells were treated with KAA1-4 (A-D), KA (E), PfOH (F), KAA5
(G), KAA6 (H) or their vehicle (0.2% DMSO) for 15 min and then stimulated with LPS (0.2
µg/ml). Following 24 h incubation, BV-2 cell viability was assessed by the MTT assay. Data
from four independent experiments are presented as mean % of the unstimulated control cell
viability ± SEM. Unstimulated control cells were seeded in the absence of compounds and
solvent (data not shown). No statistically significant effects were observed according to the
randomized block one-way ANOVA. P and F values for the main effects of ANOVA are shown
on the graphs.
Figure 9. Zopolrestat abolishes the effects of KAA1 and KAA2 on nitrite secretion by
stimulated murine BV-2 microglia. BV-2 cells were preincubated with zopolrestat (100 μM) or
its vehicle (0.05% v/v DMSO) for 15 min. Cells were then treated with KAA1 (A), KAA2 (B) or
their vehicle (0.2% v/v DMSO) for 15 min and stimulated with LPS (0.2 µg/ml). Following 24 h
incubation, the concentration of nitrite (µM) was measured in cell supernatants by the Griess
assay. Data from five independent experiments are presented as means ± SEM. The randomized
block one-way ANOVA and Dunnett’s post-hoc test were used to determine the effects of KAA
treatment in the presence or absence of zopolrestat. * P < 0.05, ** P < 0.01 different from
vehicle-treated, stimulated cells (0 μM). P1 and F1 values for the main effect of KAA treatment
22

alone, and P2 and F2 values for the main effect of KAA treatment in the presence of zopolrestat..
The limit of detection of the Griess assay is displayed as a dashed line at 0.9 ± 0.7 μM.
Figure 10. At the concentrations studied, zopolrestat in combination with KAA1 or KAA2
does not affect the viability of stimulated murine BV-2 microglia. Cells were preincubated
with zopolrestat (100 μM) or its vehicle (0.05% v/v DMSO) for 15 min. Cells were then treated
with KAA1 (A), KAA2 (B) or their vehicle (0.2% v/v DMSO) for 15 min and stimulated with
LPS (0.2 µg/ml). Following 24 h incubation, BV-2 cell viability was assessed by the MTT assay.
Data from five independent experiments are presented as mean % of the unstimulated control
cell viability ± SEM. Control cells were seeded in the absence of compounds and solvent (data
not shown). According to the randomized block one-way ANOVA, there were no significant
effects of KAA in the presence or absence of zopolrestat. P1 and F1 values for the main effect of
KAA treatment alone, and P2 and F2 values for the main effect of KAA treatment in the presence
of zopolrestat.
23

Fig. 1.
24

Fig. 2.
25

Fig. 3.
26

Fig. 4.
27

Fig. 5.
28

Fig. 6.
29