New ratiometric fluorescent calcium indicators with moderately attenuated binding affinities

Article abstract of DOI:10.1016/S0960-894X(00)00280-8

Mono-halogenated derivatives of the calcium indicators fura-2 and indo-1 were synthesized and their spectroscopic properties evaluated. Halogenation ortho or para to the bridging oxygen in the BAPTA nucleus had a more pronounced weakening effect on bindin

Full text of DOI:10.1016/S0960-894X(00)00280-8

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1515±1518
New Ratiometric Fluorescent Calcium Indicators with Moderately
Attenuated Binding Anities
Kyle R. Gee,* Eric A. Archer, Laura A. Lapham, Mary E. Leonard, Zhang-Lin Zhou,
John Bingham and Zhenjun Diwu
Molecular Probes, Inc., 4849 Pitchford Ave., Eugene, OR 97402, USA
Received 8 March 2000; accepted 24 April 2000
AbstractÐMono-halogenated derivatives of the calcium indicators fura-2 and indo-1 were synthesized and their spectroscopic
properties evaluated. Halogenation ortho or para to the bridging oxygen in the BAPTA nucleus had a more pronounced weakening
e€ect on binding anity than in the meta position in the fura derivatives. Two new excitation ratioable ¯uorescent calcium indi-
cators, benzothiaza-1 and 2, were also synthesized. Kd values of 400 nM to 5.3 mM [Ca2+] were observed in these families of new
probes. # 2000 Elsevier Science Ltd. All rights reserved.
Fluorescent indicators of divalent metal ion concentra-
tions such as Ca2+, which were developed by Tsien and
colleagues in the 1980's, have been of tremendous
importance to researchers in a wide variety of biological
science disciplines. Two of the most popular ¯uorescent
calcium indicators are fura-2 and indo-1.1 Both combine
a ¯uorophore with a BAPTA (1,2-bis(2-aminophenoxy)
ethane-N,N,N,N0-tetraacetic acid)2 chelator and are
optimally excited with UV light, and both exhibit ratio-
metric wavelength dependencies as a function of calcium
concentration. The excitation maximum of fura-2 shifts
from 362 to 335 nm as [Ca2+] increases, while the
emission wavelength of indo-1 shifts from 485 to 410
nm. This ratiometric property is very important, as it
allows for internal calibration of [Ca2+] and obviates
the need for experimental corrections for photobleaching,
sample thickness variability, dye concentration, etc. in
intracellular systems.3,4
quickly and also partially act as bu€ers, resulting in
inaccurate measurements.5 For example, fura-2 has very
limited sensitivity to [Ca2+] above 1 mM.6 Molecules
like mag-fura-2,7 in which the BAPTA portion is trun-
cated into a lower anity chelator, have an attenuated
anity for calcium but are quite sensitive to magne-
sium. London et al. began to address this selectivity
problem by synthesizing fura-FF and indo-FF.8,9 These
di¯uorinated BAPTA-based low anity probes have
found some utility, for example being used to quantify
high calcium concentrations in endoplasmic reticulum10,11
and cytoplasm.12,13 However, these probes' high calcium
Kd values (30±50 mM, two orders of magnitude higher
than the parent probes) make them relatively insensitive
to [Ca2+] in the 1±5 mM range. Our aim in the present
study was to ®nd fura-2 and indo-1 analogues with
binding anities only about one order of magnitude less
than the parent probes, ®lling a needed gap in the bio-
scientist's toolbox for optical calcium concentration
measurements. We achieved this aim by exploring the
e€ect of single halogen atom attachment to various
positions on the BAPTA chelator part of the ¯uorescent
probe molecules, and also by synthesizing a new ¯uor-
escent calcium indicator. The syntheses and spectro-
scopic properties are reported herein.
The binding anity of ¯uorescent calcium probes such
as fura-2 and indo-1 are most often expressed in terms
of the dissociation constant Kd. Kd values for fura-2 and
indo-1 are in the 140±230 nM range, which allows for
accurate measurement of calcium concentrations close
to these concentrations. However, when calcium con-
centrations are higher or are involved in fast transients,
such as in cell excitatory states (intracellular) or when
secreted (extracellular), these probes can saturate
The synthetic strategy for fura-4F, fura-5F, fura-5Cl,
and fura-6F (5a±d) derives from that originally reported
by Grynkeiwicz et al.,1 beginning with an appropriately
halogenated 2-nitrophenol (2); this was also the strategy
employed in the synthesis of fura-FF8,9 (Scheme 1).
*Corresponding author. Tel.: +1-541-465-8300; fax: +1-541-344-
6504; e-mail: melodyouellet@probes.com
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00280-8

1516
K. R. Gee et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1515±1518
Indo-5F (9) was obtained beginning with 5-¯uoro-2-
nitrophenol and 2-nitrophenol. Construction of the
BAPTA nucleus went as described above, followed by
formylation, to give key intermediate 8a. Wittig reaction
and reductive cyclization, as described for indo-1,1 gave
the desired indicator after saponi®cation of the esters
(Scheme 2). Condensation of 5-formyl-50-methyl-BAPTA
methyl ester1 (8b) and 8a with 5-tri¯uoromethyl-2-amino-
thiophenol in DMSO gave the novel dyes benzothiaza-1
(10a) and-2 (10b), respectively, after ester saponi®cation
(Scheme 2).
All of the probes were puri®ed and characterized in
their potassium salt forms.14 From this, they were con-
vertable into their cell permeant acetoxymethyl (AM)
ester15 forms (6) by conversion to the free acids, fol-
lowed by alkylation with bromomethyl acetate in DMF,
mediated by diisopropylethylamine.
In the monohalogenated fura series (5a±d), the halogen
atom exerted a larger e€ect on calcium binding anity
when located ortho or para to the bridging oxygen atom
of the BAPTA nucleus, as opposed to when para to the
Scheme 1. Reagents: (a) K2CO3, DMF; (b) H2, Pd/C; (c) BrCH2CO2CH3, DMF; (d) POCl3, DMF; (e) H2, Pt/C; (f) 5-chloromethyloxazole-2-car-
boxylic acid, ethyl ester, K2CO3, DMF; (g) KOH(aq), MeOH/dioxane; (h) HCl; (i) BrCH2OC(O)CH3, DIEA, DMF.
Scheme 2. Reagents: (a) K2CO3, DMF; (b) H2, Pd/C; (c) BrCH2CO2CH3, DMF; (d) POCl3 DMF; (e) 4-Methoxycarbonyl-2-nitrobenzyl-a-tri-
phenylphosphonium bromide, K2CO3, DMF; (f) P(OEt)3; (g) KOH(aq), MeOH/dioxane; (h) DMSO.

K. R. Gee et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1515±1518
Table 1. Binding anity and wavelengths
1517
Compound
Kd (Ca2+)
Exc l max (low Ca2+)
Exc l max (high Ca2+)
Em l max
Ref
1
Fura-2
140 nM
770 nM
400 nM
400 nM
5.3 mM
660 nM
1.4 mM
230 nM
470 nM
364 nm
371 nm
367 nm
366 nm
367 nm
368 nm
367 nm
349 nm
346 nm
338 nm
339 nm
339 nm
337 nm
337 nm
333 nm
331 nm
331 nm
330 nm
512 nm
510 nm
511 nm
511 nm
510 nm
470 nm
470 nm
485/410 nma
478/401 nma
Fura-4F (5a)
Fura-5F (5b)
Fura-5Cl (5c)
Fura-6F (5d)
Benzothiaza-1 (10a)
Benzothiaza-2 (10b)
Indo-1
1
Indo-5F (9)
aEmission maxima in low/high [Ca2+]; excitation at 334 nm.
Figure 1. Fluorescence excitation spectra of 5a.
Figure 2. Fluorescence excitation spectra of 10a.
BAPTA nitrogen atom (Table 1). In the 50-halo-fura
molecules 5b±c, chlorine or ¯uorine had essentially the
same e€ect. All of the monohalogenated fura derivatives
had ¯uorescence quantum yields similar to fura-2 in
bu€er, with little change in wavelength maxima.
®broblasts, RBL mast cells and BC3H-1 myocytes using
incubation at 37 ꢀC for 30 min with 1±5 mM (10a) as its
tetra-AM ester, imaged via ¯uorescence microscopy
using standard fura-2 optics. We also observed the
expected shift in excitation maxima from 370 to 330 nm
upon addition of the calcium ionophore ionomycin to
the extracellular medium (data not shown). Also, BHK
cells have been loaded with 10a as its tetra-AM ester for
[Ca2+] measurement by two-photon laser scanning
microscopy.4
The modest e€ect of 50-halogenation was also observed
for indo-5F (9), which has a 2-fold drop in binding a-
nity relative to the 50-methyl analogue, aka indo-1.
The benzothiaza probes 10a±b exhibit the same type of
excitation ratiometric behavior with changing [Ca2+] as
fura-2 does, at the same wavelengths (Fig. 2). Their
straightforward synthesis allows for easy structural
modi®cation. However, the emission wavelengths of the
benzothiaza dyes are about 40 nm shorter than fura-2.
This property may allow for dual imaging of ben-
zothiaza and ¯uorescein-based probes, since the ¯uo-
rescence emissions can be measured distinct from each
other. Such an experiment is not possible with fura
derivatives, because the emission (510 nm) coincides
with that of ¯uoresceins (510±520 nm). Benzothiaza-2
(10b) achieved the goal of about 10-fold lower binding
anity, compared with fura-2. However, the quantum
yields of both benzothioaza-1 and-2 are somewhat
lower, i.e. about 25% that of fura-based probes.
Nevertheless, we obtained successful loading of 3T3
Acknowledgements
We thank Nabi Malekzadeh and Dr. Richard Haugland
for helpful discussions. We thank Diane Ryan and Cailan
Zhang for spectroscopic measurements.
References and Notes
1. Grynkiewicz, G.; Poenie, M.; Tsien, R. Y. J. Biol. Chem.
1985, 260, 3440.
2. Tsien, R. Y. Biochemistry 1980, 19, 2396.
3. Kao, J. P. Y. Meth. Cell Biol. 1994, 40, 155.
4. Takahashi, A.; Camacho, P.; Lechleiter, J. D.; Herman, B.
Phys. Rev. 1999, 79, 1090.

1518
K. R. Gee et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1515±1518
5. Regehr, W. G.; Atluri, P. P. Biophys. J. 1995, 68, 2156 and
references cited therein.
6. Stout, A. K.; Reynolds, I. J. Neuroscience 1999, 89, 91.
7. Raju, B.; Murphy, E.; Levy, L. A.; Hall, R. D.; London, R.
E. Am. J. Physiol. 1989, 256, C540.
8. London, R. E.; Levy, L. A.; Murphy, E. US Patent
5,516,911 (1996); Chem. Abstr. 1996, 125, 81296.
9. London, R. E.; Rhee, C. K.; Murphy, E.; Gabel, S.; Levy,
L. A. Am. J. Physiol. 1994, 266, C1313.
10. Hajnoczky, G.; Hager, R.; Thomas, A. P. J. Biol. Chem.
1999, 274, 14157.
11. Golovina, V. A.; Blaustein, M. P. Science 1997, 275, 1643.
12. Tse, A.; Tse, F. W. J. Physiol. 1998, 512, 385.
13. Hanitkevich, V. Ya. Cell Calcium 1998, 23, 213.
1
14. For example, for 5d: H NMR (300 MHz, D2O) d 7.65 (s,
1H), 7.55 (s, 1H), 7.36 (s, 1H), 7.08 (s, 1H), 6.97 (dd, J=14, 6.6
Hz, 1H), 6.67 (t, J=9.6 Hz), 6.50 (d, J=8.4 Hz, 1H), 4.35 (br s,
4H), 3.99 (s, 4H), 3.93 (s, 4H); 19F NMR (D2O) ꢀ 67.0 (referenced
to CF3±toluene); HPLC 99% pure (254 nm, C18 4.6Â250 mm
column, 5±95 CH3CN/0.1% TFA for 30 min, retention time 19.9
min); E 25,000 cm 1 M 1 (10 mM EGTA/10 mM MOPS, pH 7.2).
15. Tsien, R. Y. Nature 1981, 290, 2396.