An "aND"-logic-gate-based fluorescent probe with dual reactive sites for monitoring extracellular methylglyoxal level changes of activated macrophages

Article abstract of DOI:10.1039/d1cc01859k

An "AND"-logic-gate-based fluorescent probe NAP-DCP-4 with dual reactive sites is reported, which has improved selectivity for methylglyoxal over glyoxal, featuring formaldehyde-enhanced methylglyoxal detection and irreversible and reversible turn-on fluo

Full text of DOI:10.1039/d1cc01859k

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
An ‘‘AND’’-logic-gate-based fluorescent probe
with dual reactive sites for monitoring
extracellular methylglyoxal level changes
of activated macrophages†
Cite this: Chem. Commun., 2021,
57, 8166
Received 7th April 2021,
Accepted 16th July 2021
Wenli Wang,‡a Junwei Chen,‡a Huijuan Ma,‡a Wanjin Xing,a Nan Lv,a
DOI: 10.1039/d1cc01859k
ab
a
Baoning Zhang,a Huan Xu,*a Wei Wang
*
and Kaiyan Lou
*
rsc.li/chemcomm
An ‘‘AND’’-logic-gate-based fluorescent probe NAP-DCP-4 with dual phosphorylation to glycolysis to satisfy the elevated energy
reactive sites is reported, which has improved selectivity for methyl- consumption.17 The increased glycolysis level leads to upregulated
glyoxal over glyoxal, featuring formaldehyde-enhanced methylglyoxal intracellular MGO levels resulting from both enhanced MGO
detection and irreversible and reversible turn-on fluorescence production and decreased glyoxalase I (GLO1) expression, the
responses at different excitation wavelengths. Its cell-impermeability key enzyme for MGO removal.18,19 While there were a few studies
enables facile monitoring of extracellular methylglyoxal level changes reported that intracellular MGO levels increase upon macrophage
in the supernatant of activated macrophages.
activation,18,19 extracellular MGO level changes and their potential
contributions to inflammatory responses are not clear. It is
Methylglyoxal (MGO) is a highly reactive carbonyl species mainly possible that activated macrophages that excrete MGO would
formed in glycolysis from the degradation of glyceraldehyde-3- induce an extracellular MGO level increase in their microenviron-
phophate and dihydroxyacetone phosphate.1,2 MGO forms ment, which regulates physiological functions of the surrounding
advanced glycation end products (AGEs) by glycating DNA, lipids, cells.20 Although quite a few fluorescent probes have been
and proteins,3 causing protein dysfunctions,4 activation of RAGE reported for MGO detection (Fig. S1, ESI†), so far, there is no
receptor,2 pro-inflammatory signaling and inflammation,5–8 and cell-impermeable probe available for monitoring the extracellular
oxidative stress.9,10 Elevated MGO and AGE levels are observed in MGO level changes.
many diseases including diabetes,11–13 obesity,14 cancer,15 and
Alzheimer’s disease.16 Therefore, development of MGO probes for impermeable fluorescent probe NAP-DCP-4 containing two
selective monitoring MGO levels is highly desirable. polar reactive sites: an o-phenylenediamino (OPD) group and
Herein, we would like to report the rational design of a cell-
In our lab, we are interested in MGO level changes of a guanidino (GND) group, forming an ‘‘AND’’ logic gate for the
macrophages and their surrounding microenvironments upon selective sensing of the extracellular MGO levels (Fig. 1 and
activation. Macrophages are critical innate immune cells for Fig. S2, ESI†). Both reactive groups’ reaction and sensing
host defence, immune surveillance of cancer cells and main- mechanism are well established in the literature. OPD is a
taining homeostasis.17 It is well known that macrophages can fluorescence quenching group via a photo-induced electron
be activated by heterologous stimuli, such as lipopolysacchar- transfer (PeT) mechanism.21–24 OPD-based fluorescent probes
ide (LPS), through multiple intracellular pro-inflammatory such as PND-1 and NP react with MGO irreversibly to form
signalling pathways, which shifts their metabolism from oxidative methylquinoxalines, which interrupts the PeT quenching
effect, resulting in a fluorescence turn-on response.22,23 Struc-
turally similar glyoxal (GO) normally does not interfere with
a State Key Laboratory of Bioengineering Reactor, Shanghai Key Laboratory of New
MGO detection in OPD-based probes because of its low
reactivity.21,22,24 However, formaldehyde (FA) and nitric oxide
Drug Design, and Shanghai Key Laboratory of Chemical Biology,
School of Pharmacy, East China University of Science & Technology,
(NO) may react with the OPD reactive group and interfere with
130 Meilong Road, Shanghai 200237, China. E-mail: xuhuan@ecust.edu.cn,
MGO detection.21,25 In contrast, GND-based fluorescent probes
kylou@ecust.edu.cn
b Department of Pharmacology and Toxicology and BIO5 Institute,
University of Arizona, Tucson, AZ 85721-0207, USA.
E-mail: wwang@pharmacy.arizona.edu
NAP-DCP-1 and NAP-DCP-3 react with MGO or GO reversibly
and show fluorescence turn-on responses via an intramolecular
change transfer (ICT) mechanism and excitation wavelength
modulation to selectively excite the deprotonated adducts.26
They show high selectivity over FA and NO, but cannot differ-
entiate between MGO and GO with even higher fluorescence
† Electronic supplementary information (ESI) available: Experimental details,
additional spectra, and 1H NMR, 13C NMR and HRMS data. See DOI: 10.1039/
d1cc01859k
‡ These authors contributed equally to this work.
8166 | Chem. Commun., 2021, 57, 8166–8169
This journal is © The Royal Society of Chemistry 2021

View Article Online
Communication
ChemComm
methylquinoxaline formation (labelled in blue, Fig. 1b) and the
‘‘ICT On’’ of the dihydroxyimidazolidine formation (labelled in
red, Fig. 1b) resulted in the fluorescence turn-on response for
MGO (Fig. 2a). As expected, incorporation of the OPD group
significantly reduced the probe’s turn-on response for GO
(Fig. 2b). To estimate the contribution of the OPD group to
the improvement of MGO selectivity over GO, the probe was
excited at the isobestic absorption wavelength at 365 nm
(see Fig. S3, ESI† for the isobestic point) to minimize the effect
of the GND group’s reaction with MGO towards the fluores-
cence turn-on response. Under this excitation condition, both
unreacted and reacted GND-substituted fluorophores (labelled
in red in Fig. 1) have the same absorbance and are all emissive
from either excited state proton transfer (ESPT) excited states
(‘‘ESPT, ICT on’’ for the protonated species in Fig. 1b) or excited
state (‘‘ICT on’’ for 6-I in Fig. 1b) where there is no PeT
quenching group attached.26,28 The fluorescence turn-on ratios
Fig. 1 (a) Probe design of NAP-DCP-4: an ‘‘AND’’-logic-gate-based fluor-
escent probe with both OPD and GND reactive groups for improved MGO
selectivity; (b) illustration of irreversible turn-on responses for MGO at
lex = 365 nm and reversible turn-on responses for MGO at lex = 425 nm
(only major species in equilibrium are shown).
turn-on ratio for GO.26 We envisioned that incorporation of
dual reactive groups, OPD and GND, into an ‘‘AND’’-logic-gate-
based probe would improve the GND-based probe’s selectivity
for MGO over GO, while retaining high selectivity over FA and
NO (Fig. 1a). We also envisioned that the incorporation of both
reversible GND and irreversible OPD reactive groups may render the
probe with both features of reversible (lex = 425 nm, Off–On–Off)
and irreversible (lex = 365, Off–On–On) fluorescence responses for
reaction with MGO and then addition of N-acetyl cysteine (NAC), a
known MGO quencher (Fig. 1b). Additionally, the combined use of
the two polar reactive groups could possibly afford a cell-
impermeable probe for monitoring extracellular MGO levels,
as the calculated log P values from the molinspiration website27
for NP,23 NAP-DCP-1,26 and NAP-DCP-4 are 3.69, 1.69, and 1.19,
respectively.
Fig. 2 (a and b) Fluorescence intensities of NAP-DCP-4 (2 mM) upon
addition of increased concentrations (0, 10, 20, 40, 60, 100, 150, 200, 300,
400, 500, 600, 800, 1000, and 1200 mM) of MGO (a) and GO (b) for 1.5 h;
(c) fluorescence emission spectra of the probe NAP-DCP-4 (2 mM) upon
incubation with 100 equiv. MGO for 1.5 h versus further addition of
The probe NAP-DCP-4 was conveniently synthesized from
commercially available 4-bromo-1,8-naphthalic anhydride (1) 1000 equiv. NAC and then incubation for another 3 h; (d) fluorescence
in two steps in a 7.7% total yield and characterized by 1H NMR,
intensity at 559 nm of the probe NAP-DCP-4 (2 mM) upon incubation with
100 equiv. of various other species. (1): probe only, (2): MGO, (3): MGO and
13C NMR and HRMS (see ESI† Part II for more details).
1000 equiv. FA, (4): FA, (5): acetaldehyde, (6): GO, (7): GO and 1000 equiv.
We then studied the probe NAP-DCP-4’s in vitro fluorescence
FA, (8): o-phthalaldehyde, (9): glyoxylic acid, (10): benzaldehyde, (11) GSH,
response for MGO in PBS buffer at pH 7.4. Similar to the
(12): glucose, (13): cysteine, (14): proline, (15): isoleucine, (16): H2O2, (17):
GND-based probe NAP-DCP-1,26 the UV-Vis spectra showed a K+, (18): Na+, (19): Cu2+, (20): Zn2+, (21): Al3+, (22): Ca2+, (23): Fe2+, (24):
Fe3+, (25) NOC-18; (e) pH-dependent fluorescence intensities at 559 nm
red-shifted peak at 391 nm (Fig. S3, ESI†) due to the formation
of the deprotonated dihydroxyimidazolidine adduct 6-II
with increased electron-donation at the 4-position of the 1,
8-naphathalimide (ICT On, Fig. 1b). When 6-II was selectively
of the probe NAP-DCP-4 in response to 100 equiv. MGO or GO; (f) time-
dependent fluorescence intensity increase of the probe NAP-DCP-4
(2 mM) at 559 nm upon addition of 100 equiv. MGO with or without
1000 equiv. FA (all measurements were performed in 10 mM PBS buffer at
excited at 425 nm, the ‘‘AND’’ effect of the ‘‘PeT Off’’ of the 37 1C, lex = 425 nm).
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 8166–8169 | 8167

View Article Online
ChemComm
Communication
were 24.7-fold for MGO versus 9.8-fold for GO (Fig. S4c and d,
From the above studies, the reaction and sensing mechnism
ESI†) under typical reaction conditions (2 mM probe incubated was proposed in Fig. 1b based on the well-established reaction
with 200 equiv. MGO or GO at pH 7.4 and 37 1C for 1.5 h), and sensing mechanism of each reactive group and formation
indicating that OPD indeed reacts much faster with MGO than of an ‘‘AND’’ logic gate (Fig. S2, ESI†). The proposed reaction
with GO, which provides the basis for the improved selectivity for mechanism was supported by HRMS studies that confirmed
MGO over GO in NAP-DCP-4. Upon excitation with 425 nm light, the formation of di-MGO adduct 6 (Fig. S11, ESI†) between the
the ‘‘AND’’-logic-gate-based probe showed 21.2 and 11.1-fold for probe and excess amount of MGO and formation of the mono
MGO and GO, respectively (Fig. S3a and b, ESI†). In contrast, the adduct 7 (Fig. S12, ESI†) when the reaction mixture was further
probe NAP-DCP-1 bearing only the GND group had fluorescence treated with NAC. 3D fluorescence spectra also confirmed that
turn-on ratios 11- and 19-fold for MGO and GO, respectively.26
the probe gives an irreversible turn-on response (Off–On–On,
When excited at 425 nm, the probe’s fluorescence turn-on labelled as the red triangle) at lex/em = 365/559 nm and
response for MGO was reversible upon addition of NAC reversible turn-on response (Off–On–Off, lablled as the green
(Fig. 2c). The reversibility for GO was not as good as that of square) at lex/em = 425/559 nm (Fig. S10, ESI†).
MGO, providing a further differentiation between MGO and GO
We further examined the probe’s cell permeability and
(Fig. S6, ESI†). The probe has good selectivity for MGO (Fig. 2d) potential utility in sensing MGO level changes upon macro-
over many other reactive carbonyl species (FA, acetaldehyde, phage activation. RAW264.7 cells were selected as the model
o-phthalaldehyde, glyoxylic acid, and benzaldehyde), GSH, glu- cell line, which were activated by lipopolysaccharide (LPS)
cose, amino acids (cysteine, proline, and isoleucine), H2O2, stimulation (see ESI† part V for more details). In the cytotoxicity
metal ions (K+, Na+, Cu2+, Zn2+, Al3+, Ca2+, Fe2+, and Fe3+), studies evaluated by MTS assay, the probe showed surprising
and NO (from NOC-18, an NO donor). Further pH-dependence nontoxicity up to 100 mM for both non-activated and activated
studies indicated that the probe is suitable for detection of RAW264.7 cells (Fig. S13, ESI†). Microplate readings of fluores-
MGO at pH higher than 6.0 (Fig. 2e).
cence intensity for both non-activated and activated RAW264.7
Since FA may exist in biological systems at much higher cells incubated with the probe (20 mM) did not show a statis-
concentration than MGO (up to 500 mM for FA29 versus typical tically significant intensity increase compared with the control
1–4 mM for MGO25) and often presents as an interfering species cells (Fig. S14, ESI†), indicating that the probe is cell imperme-
for MGO detection, we examined the potential effect of FA able. Otherwise, the probe-incubated cells should have shown
towards MGO detection by the probe NAP-DCP-4. When excited enhanced fluorescence intensity compared with the control
at 365 nm, no significant fluorescence intensity increase was cells after the probe reacted with the intracellular MGO increas-
observed for FA (Fig. S8, ESI,† column 4) compared with MGO ingly produced upon LPS activation. Further fluorescence ima-
(Fig. S8, ESI,† column 2), suggesting that FA reacts much slower ging studies using cell-permeable NAP-DCP-1 as the control
with the OPD group than MGO under the same conditions. also confirmed cell impermeability of the probe (Fig. S15, ESI†).
However, the coexistence of 10 times concentration of FA We then took the cell culture supernatant for fluorescence
enhanced the probe’s fluorescence turn-on response for MGO analysis. When excited at 425 nm, RAW264.7 cells stimulated
(Fig. 2d, columns 2 and 3), indicating that FA can compete with with LPS for 4 h had an apparent fluorescence intensity
MGO to react with the OPD group and help the removal of its increase, which decreased upon addition of 5 mM NAC, con-
fluorescence quenching effect. When excited at 425 nm, time- firming that the observed fluorescence increase was mostly due
dependent fluorescence intensity studies also confirmed FA- to increased MGO levels (Fig. 3a). Extended stimulation with
enhanced turn-on detection of MGO as a result of the ‘‘AND’’ LPS for 24 h resulted in further increase of fluorescence
logic gate (Fig. 2f). Accordingly, the limit of detection (LOD) for intensity, suggesting further elevated supernatant MGO level
MGO improved from 2.2 mM to 1.8 mM in the presence of (Fig. 3b). A similar fluorescence intensity increase was observed
biologically relevant 400 mM FA (Fig. S7, ESI†). As a proof-of-
concept study, the ‘‘AND’’-logic-gate-based probe NAP-DCP-4
turns an interfering specie (FA) into a promoting specie for the
detection of the analyte of interest (MGO).
The potential interference of NO for MGO detection was also
studied. UV-Vis studies showed that the presence of 1 mM
NOC-18 seemed not to affect the formation of GND-MGO
adducts at around 390 nm (Fig. S9a, ESI†). However, the
fluorescence turn-on ratio for MGO in the presence of 1 mM
NOC-18 was only 3-fold (Fig. S9b, ESI†). We propose that the
formed benzotriazole between OPD and NO may deprotonate at
pH 7.4, and function as a fluorescence quenching group via a
PeT mechanism (Scheme S2, ESI†). Since the NO concentration
is relatively low compared with those of FA and MGO in
Fig. 3 Fluorescence intensity at 564 nm of NAP-DCP-4 in the super-
biological systems, its interference to MGO detection can be
neglected.
natant of RAW264.7 cells stimulated with LPS for 4 h (a and c) and 24 h
(b and d): (a and b) lex = 425 nm; (c and d) lex = 365 nm.
8168 | Chem. Commun., 2021, 57, 8166–8169
This journal is © The Royal Society of Chemistry 2021

View Article Online
Communication
ChemComm
Notes and references
1 M. P. Kalapos, Toxicol. Lett., 1999, 110, 145–175.
2 N. Rabbani and P. J. Thornalley, Amino Acids, 2012, 42,
1133–1142.
3 P. J. Thornalley, Gen. Pharmacol., 1996, 27, 565–573.
4 R. Singh, A. Barden, T. Mori and L. Beilin, Diabetologia, 2001, 44,
129–146.
5 D. Yao, T. Taguchi, T. Matsumura, R. Pestell, D. Edelstein,
I. Giardino, G. Suske, N. Rabbani, P. J. Thornalley, V. P. Sarthy,
H.-P. Hammes and M. Brownlee, J. Biol. Chem., 2007, 282,
31038–31045.
6 H. Wang, Q. H. Meng, J. R. Gordon, H. Khandwala and L. Wu,
Clin. Biochem., 2007, 40, 1232–1239.
Fig. 4 Fluorescence intensities at 564 nm of NAP-DCP-4 in the super-
natant of RAW264.7 cells in normal or hypoxia microenvironments stimu-
lated with LPS for 0, 2, 4, 8, 12, and 24 h: (a) lex = 425 nm; (b) lex = 365 nm.
7 B. Vulesevic, B. McNeill, F. Giacco, K. Maeda, N. J. R. Blackburn,
M. Brownlee, R. W. Milne and E. J. Suuronen, Diabetes, 2016, 65,
1699–1713.
8 C. Tikellis, R. J. Pickering, D. Tsorotes, O. Huet, M. E. Cooper,
K. Jandeleit-Dahm and M. C. Thomas, Diabetes, 2014, 63,
3915–3925.
9 L. Wu and B. H. J. Juurlink, Hypertension, 2002, 39, 809–814.
10 N. Shangari, W. R. Bruce, R. Poon and P. J. O’Brien, Biochem. Soc.
Trans., 2003, 31, 1390–1393.
upon LPS activation when excited at 365 nm, although it was not
reversible upon addition of NAC (Fig. 3c and d). The above studies
confirmed that the probe can be used to monitor the increase of
supernatant MGO levels of macrophages upon activation.
We continued to study the probe’s use for the monitoring of
the supernatant MGO level changes of activated RAW264.7 cells
under hypoxia conditions induced by CoCl2.30 Time-dependent
fluorescence intensity increases upon LPS stimulation were
observed when excited with either 425 nm (Fig. 4a) or 365 nm
(Fig. 4b) light. Upon addition of NAC, the fluorescence intensity
increase remained almost unchanged when excited at 365 nm
(Fig. 4b) in contrast to the fluorescence intensity dropping to low
background values when excited at 425 nm (the blue curve in
Fig. 4a), which confirmed MGO detection. It was also found that
the fluorescence intensities measured were higher under hypoxia
conditions at each time point, suggesting higher excreted MGO
levels at hypoxia conditions due to further increased glycolysis for
energy consumption. It was possible that the fluorescence inten-
sity increase observed may partially come from increased FA
concentrations. However, the increasing differences between the
green curve under hypoxia conditions and the red curve under
normal conditions when excited at 425 nm (Fig. 4a) compared
with the rather unchanged fluorescence intensity differences
when excited at 365 nm (Fig. 4b) still suggested more reacted
GND group and higher MGO levels under hypoxia conditions.
In conclusion, we have reported an ‘‘AND’’-logic-gate-based
fluorescent probe NAP-DCP-4 adopting two reactive sites: OPD
and GND. The probe has improved selectivity for MGO over GO,
featuring FA-enhanced MGO detection and unique irreversible
and reversible fluorescence turn-on responses at two different
excitation wavelengths. The potential utility of the probe for
extracellular MGO detection was demonstrated in monitoring
the MGO level changes in the supernatant of RAW264.7 cells
upon LPS stimulation under either normal or hypoxia conditions.
The financial support by Shanghai Natural Science Fund
(Grant No. 20ZR1414700), Shanghai Sailing Program (Grant No.
19YF1412500), the National Natural Science Foundation of China
(Grant No. 21577037, 21738002, and 21906057) and the State Key
Laboratory of Bioreactor Engineering is greatly appreciated.
11 S.-Y. Goh and M. E. Cooper, J. Clin. Endocrinol. Metab., 2008, 93,
1143–1152.
12 A. Prasad, P. Bekker and S. Tsimikas, Curr. Cardiol. Rev., 2012, 20,
177–183.
13 A. K. Berner, O. Brouwers, R. Pringle, I. Klaassen, L. Colhoun,
C. McVicar, S. Brockbank, J. W. Curry, T. Miyata, M. Brownlee,
R. O. Schlingemann, C. Schalkwijk and A. W. Stitt, Diabetologia,
2012, 55, 845–854.
14 J. Masania, M. Malczewska-Malec, U. Razny, J. Goralska,
A. Zdzienicka, B. Kiec-Wilk, A. Gruca, J. Stancel-Mozwillo,
A. Dembinska-Kiec, N. Rabbani and P. J. Thornalley, Glycoconjugate
J., 2016, 33, 581–589.
15 D. Schroeter and A. Hoehn, Curr. Pharm. Des., 2018, 24,
5245–5251.
16 C. Angeloni, L. Zambonin and S. Hrelia, BioMed Res. Int., 2014,
238485.
´
´
17 A. Viola, F. Munari, R. Sanchez-Rodrıguez, T. Scolaro and
A. Castegna, Front. Immunol., 2019, 10, 1462.
18 T. Aki, T. Funakoshi, K. Noritake, K. Unuma and K. Uemura,
Sci. Rep., 2020, 10, 10581.
19 D. Prantner, S. Nallar, K. Richard, D. Spiegel, K. D. Collins and
S. N. Vogel, J. Leukocyte Biol., 2021, 109, 605–619.
20 T. Baumann, A. Dunkel, C. Schmid, S. Schmitt, M. Hiltensperger,
K. Lohr, V. Laketa, S. Donakonda, U. Ahting, B. Lorenz-Depiereux,
¨
J. E. Heil, J. Schredelseker, L. Simeoni, C. Fecher, N. Korber,
T. Bauer, N. Hu¨ser, D. Hartmann, M. Laschinger, K. Eyerich,
S. Eyerich, M. Anton, M. Streeter, T. Wang, B. Schraven,
D. Spiegel, F. Assaad, T. Misgeld, H. Zischka, P. J. Murray,
¨
A. Heine, M. Heikenwalder, T. Korn, C. Dawid, T. Hofmann,
¨
P. A. Knolle and B. Hochst, Nat. Immunol., 2020, 21, 555–566.
21 T. Wang, E. F. Douglass, Jr., K. J. Fitzgerald and D. A. Spiegel, J. Am.
Chem. Soc., 2013, 135, 12429–12433.
22 T. Tang, Y. Zhou, Y. Chen, M. Li, Y. Feng, C. Wang, S. Wang and
X. Zhou, Anal. Methods, 2015, 7, 2386–2390.
23 S. Gao, Y. Tang and W. Lin, J. Fluoresc., 2019, 29, 155–163.
24 M. Yang, J. Fan, J. Zhang, J. Du and X. Peng, Chem. Sci., 2018, 9,
6758–6764.
25 F. Shaheen, A. Shmygol, N. Rabbani and P. J. Thornalley,
Biochem. Soc. Trans., 2014, 42, 548–555.
26 H. Xu, Q. Liu, X. Song, C. Wang, X. Wang, S. Ma, X. Wang, Y. Feng,
X. Meng, X. Liu, W. Wang and K. Lou, Anal. Chem., 2020, 92,
13829–13838.
27 https://www.molinspiration.com/cgi-bin/properties, 2021.
28 J. Zhou, H. Liu, B. Jin, X. Liu, H. Fu and D. Shangguan, J. Mater.
Chem. C, 2013, 1, 4427–4436.
29 K. J. Bruemmer, T. F. Brewer and C. J. Chang, Curr. Opin. Chem.
Biol., 2017, 39, 17–23.
Conflicts of interest
˜
´
´
´
30 J. Munoz-Sanchez and M. E. Chanez-Cardenas, J. Appl. Toxicol.,
2019, 39, 556–570.
There are no conflicts to declare.
This journal is © The Royal Society of Chemistry 2021
Chem. Commun., 2021, 57, 8166–8169 | 8169