Development of pH-responsive fluorescent false neurotransmitters

Article abstract of DOI:10.1021/ja101740k

We introduce pH-responsive fluorescent false neurotransmitters (pH-responsive FFNs) as novel probes that act as vesicular monoamine transporter (VMAT) substrates and ratiometric fluorescent pH sensors. The development of these agents was achieved by systematic molecular design that integrated several structural elements, including the aminoethyl group (VMAT recognition), halogenated hydroxy-coumarin core (ratiometric optical pH sensing in the desired pH range), and N- or C-alkylation (modulation of lipophilicity). Of 14 compounds that were synthesized, the probe Mini202 was selected based on the highest uptake in VMAT2-transfected HEK cells and desirable optical properties. Using Mini202, we measured the pH of catecholamine secretory vesicles in PC-12 cells (pH -5.9) via two-photon fluorescence microscopy. Incubation with methamphetamine led to an increase in vesicular pH (pH - 6.4), consistent with a proposed mechanism of action of this psychostimulant, and eventually to redistribution of vesicular content (including Mini202) from vesicles to cytoplasm. Mini202 is sufficiently bright, photostable, and suitable for two-photon microscopy. This probe will enable fundamental neuroscience and neuroendocrine research as well as drug screening efforts.

Full text of DOI:10.1021/ja101740k

Published on Web 06/11/2010
Development of pH-Responsive Fluorescent False Neurotransmitters
Minhee Lee,† Niko G. Gubernator,‡ David Sulzer,*,§ and Dalibor Sames*,†
Department of Chemistry, Columbia UniVersity, 3000 Broadway, New York, New York 10027, eMolecules,
San Diego, California 92014, and Departments of Neurology, Psychiatry, and Pharmacology, Columbia UniVersity
Medical Center, New York, New York 10032
Received March 12, 2010; E-mail: ds43@columbia.edu; sames@chem.columbia.edu
As part of a broad program aimed at optical imaging of metabolic
and signaling enzymes in cells and tissues,1,2 we became interested
in visualization of neurotransmission. Recently we have introduced
fluorescent false neurotransmitters (FFNs), probes that act as optical
tracers that provide the first means to image neurotransmitter release
from individual presynaptic terminals in the brain.3 We here report
the development of pH-responsive FFNs and demonstrate optical
in situ measurement of pH and its changes in catecholamine
secretory vesicles of intact PC-12 cells.4
Monoamine neurotransmitters are accumulated in synaptic
vesicles by vesicular monoamine transporter 2 (VMAT2), which
translocates the monoamines (e.g., dopamine) from cytosol to the
lumen of synaptic vesicles.5 Similarly, in chromaffin cells of adrenal
medulla, epinephrine and norepinephrine are accumulated in
secretory vesicles by a closely related protein, VMAT1 (Figure 1).
The vesicular lumen is acidic (pH 5-6) due to the action of
vacuolar-H+ ATPase, which imports H+ at the expense of ATP
hydrolysis. The pH gradient between the cytoplasm and the
Figure 1. Schematic illustration of (A) a dopaminergic presynaptic terminal
that modulates the activity of the postsynaptic neuron (DA ) dopamine,
vesicular lumen in turn provides the driving force for the accumula-
tion of transmitters in the vesicles. Thus the pH gradient is a key
D2 ) dopamine receptor 2). (B) A chromaffin cell in adrenal medulla
parameter regulating synaptic plasticity as it controls the vesicular
transmitter content and amount of transmitter released during vesicle
fusion. The pH gradient between the cytosol and the vesicular lumen
is ATP-dependent and thus closely coupled to the metabolic state
of the presynaptic terminals. Despite the importance of this
parameter and great efforts focused on development of fluorescent
pH indicators,6 there are currently no small molecule probes
available for selective measurement of pH in synaptic or secretory
vesicles. The pH-sensitive synapto-pHluorin protein can be used
to estimate pH in synaptic vesicles of cultured neurons.7 Alterna-
tively, construction of avidin-chimera proteins allows for anchoring
a pH-sensitive fluorescent dye, linked to biotin, to specific organelles
including secretory vesicles.8 These approaches, however, require
transfection of the cell culture prior to measurement or generation
of transgenic animals for studies with tissue.
secreting epinephrine (EPI). (C) Vesicular uptake of endogenous substrates
(dopamine or epinephrine) and exogenous fluorescent substrate (e.g.,
Mini102 or 202) by vesicular monoamine transporter (VMAT).
Commercially available pH-sensitive dyes (e.g., Lysotracker and
Lysosensor)9 are not suitable for this task as they label other acidic
organelles including endosomes and lysosomes,10 and in the brain
they do not label presynaptic terminals. A fixable weak base, known
as DAMP (3-(2,4-dinitroanilino)-3′-amino-N-methyldipropylamine),
can be used to estimate vesicular pH but requires cell fixation,
immunostaining with gold particles, and electron microscopy and
thus is unsuitable for intact cells and tissues.10
Aware of these limitations, we set out to develop pH-responsive
FFNs that would act as VMAT substrates with a built-in ratiometric
fluorescent pH sensor. The design of pH-responsive FFNs was
modeled on probe FFN511 (Figure 2A);3 the aminoethyl group was
Figure 2. Design of pH-responsive FFNs based on the coumarin nucleus.
(A) Structure of FFN511, the first example of FFN probes. (B) Structure
of umbelliferone. (C) Chemical structure of Mini101. (D) The key structural
elements for optimization of photophysical, physical, and functional
properties of pH-responsive FFNs.
maintained as the key VMAT recognition element, while the amine
in the 7-position of the coumarin nucleus was replaced with the
phenol. It is known that phenols exhibit pH-dependent photophysical
properties related to the equilibrium between the protonated phenol
and the deprotonated phenolate forms. Specifically, 7-hydroxycou-
marin (umbelliferone, Figure 2B) exhibits a pH-responsive, ratio-
† Columbia University.
‡ eMolecules.
§ Columbia University Medical Center.
9
8828 J. AM. CHEM. SOC. 2010, 132, 8828–8830
10.1021/ja101740k  2010 American Chemical Society

C O M M U N I C A T I O N S
Figure 3. (A) Structure of Mini102 and Mini202. (B) Excitation spectra
of Mini202 (λem ) 458 nm) at different pH values in phosphate buffer. (C)
Emission spectra of Mini202 in pH 7.43 phosphate buffer (λex ) 335 and
370 nm, [Mini202] ) 2 µM).
metric absorption/excitation profile.11 Accordingly, we synthesized
probe Mini101 (Figure 2C), and the pKa value of 7.8 was obtained
from the absorption measurements. Since the pKa of Mini101 is
too high for an accurate measurement in the relevant pH range of
synaptic or secretory vesicles (5-6), electron-withdrawing groups
(Cl and F) were introduced at the 6-position of coumarin to decrease
the pKa of the phenolic hydroxyl group.
Figure 4. Fluorescence microscopy images of Mini202 (20 µM, 30 min
incubation, all images were taken after rinsing the cells with dye-free media)
in (A) nontransfected HEK cells, (B) VMAT2-HEK cells, and (C) VMAT2-
HEK cells preincubated with 1 µM TBZ or (D) 1 µM reserpine. (E)
VMAT2-HEK cells labeled with Mini202. (F) The lipophilic base chloro-
quine (300 µM, 3 min) led to redistribution of Mini202 from the acidic
vesicles to cytosol. λex ) 350 ( 25 nm, λem ) 460 ( 25 nm.
In addition to the photophysical properties, we also aimed to
examine the lipophilicity of FFNs as an additional key design
parameter that affects membrane permeability by passive diffusion
and determines selectivity between VMAT-expressing and nonex-
pressing cells. Lipophilicity can be fine-tuned by adding alkyl
groups to either the coumarin ring (at position C-8) or the amino
group (Figure 2D).
According to these design directives, a series of 14 compounds
was prepared (see Supporting Information for structure and
synthesis). For all compounds, the photophysical properties, pH
responsiveness, and lipophilicity were investigated (see Supporting
Information). Placing the chlorine or fluorine adjacent to the
hydroxy group resulted in a substantial decrease of pKa, as
exemplified by probes Mini102 and Mini202 (Figure 3A, pKa )
6.2 and 6.4, respectively). The excitation spectra of the probe
Mini202 are shown in Figure 3B; they are strongly dependent on
pH, exhibiting two fully resolved maxima at 335 and 370 nm, where
the former corresponds to the protonated phenolic form and the
latter to the deprotonated phenolate form. Mini202 is sufficiently
fluorescent, and the emission intensity (λmax ) 458 nm) is dependent
on the excitation wavelength and pH (Figure 3C). Therefore, the
fluorescence intensity ratio at 458 nm obtained by dual excitation
at two different wavelengths yields the pH of the solution. In
addition, Mini202 is highly polar (log D ) -1.5, obtained by
partitioning between pH 7.4 phosphate buffer and n-octanol), which
is presumably responsible for slow cell uptake and low background
fluorescence (see below).
Figure 5. (A) PC-12 cells incubated with 20 µM Mini202 for 1 h. (B)
PC-12 cells preincubated with VMAT1 inhibitor reserpine (1 µM, 1 h)
followed by 20 µM Mini202 treatment for 1 h. The uptake of Mini202 was
blocked by reserpine. λex ) 350 ( 25 nm, λem ) 460 ( 25 nm.
( 25 nm to afford punctate fluorescence patterns, consistent with
VMAT2 expression in acidic organelles (Figure 4B). No uptake of
Mini202 was observed in nontransfected HEK cells (Figure 4A),
and the uptake of Mini202 was strongly diminished by preincu-
bating VMAT2-HEK cells with the VMAT2 inhibitors, tetrabena-
zine (TBZ) or reserpine (Figure 4C, 4D). These results indicate
that selective accumulation of Mini202 in intracellular organelles
is mediated by VMAT2. Furthermore, addition of the lipophilic
base chloroquine to the VMAT2-HEK cells labeled with Mini202
resulted in loss of the punctate fluorescent pattern, which is
consistent with redistribution of the probes from the vesicles to
the cytoplasm caused by collapse of the pH gradient (Figure 4F).
Next, we tested the uptake of Mini202 in PC-12 cells, a cell
line derived from a rat pheochromocytoma which is widely used
as a model of chromaffin cells and neuronal presynaptic terminals.4
PC-12 cells express mostly VMAT1 on their secretory large dense
core vesicles (LDCVs) where catecholamines accumulate (mostly
dopamine and norepinephrine). PC-12 cells were incubated with
The compound series was profiled by fluorescence microscopy
using human embryonic kidney (HEK) cells stably transfected with
VMAT2 (VMAT2-HEK). Six probes showed VMAT2-dependent
uptake, with Mini202 displaying the highest uptake at the selected
incubation time (Figure 4). VMAT2-HEK cells were incubated with
Mini202 (final concn ) 20 µM) for 30 min and imaged by
fluorescence microscopy with λex ) 350 ( 25 nm and λem ) 460
9
J. AM. CHEM. SOC. VOL. 132, NO. 26, 2010 8829

C O M M U N I C A T I O N S
Figure 6. Measuring pH of catecholamine secretory vesicles in PC-12 cells with Mini202 via two-photon fluorescence microscopy. (A) In situ calibration
curve of fluorescence intensity ratio from 760 nm irradiation and 692 nm irradiation (λem ) 470 ( 30 nm, pKa ) 5.93 ( 0.04, n ) 3) in PC-12 cells as a
function of vesicle pH. (B) Two-photon image of PC-12 cells incubated with 20 µM Mini202 for 1 h at λex ) 760 nm and (C) λex ) 692 nm. (D) Pseudocolor
image of I760/I692 and corresponding pH values.
Mini202 (20 µM) for 1 h and displayed fluorescent puncta consistent
with the distribution of LDCVs (Figure 5A). Preincubation of PC-
12 cells with VMAT1 inhibitor reserpine (1 µM) resulted in no
detectable labeling by Mini202 (Figure 5B), indicating that Mini202
labels LDCVs and that Mini202 is a VMAT1 substrate.
pH of secretory vesicles (e.g., transmitter releasing activity or
toxicity screens) will be possible. The pH measurement of individual
presynaptic terminals in the brain may also be feasible, which is a
focus of current studies in our laboratories.
Acknowledgment. The authors thank the G. Harold & Leila
Y. Mathers Charitable Foundation, the McKnight Foundation, the
Picower and Parkinson’s Disease Foundations, and NIDA for
financial support. We thank Dr. Robert Edwards (UCSF) for
providing HEK cells stably transfected with VMAT2 and Dr. Mark
Sonders for technical assistance with two-photon microscopy and
valuable discussions.
Finally, we pursued in situ pH measurement of LDCVs in PC-
12 cells using a two-photon fluorescence microscope. An in situ
calibration curve was generated by dual excitation (760 and 692
nm) ratiometric imaging of Mini202 in PC-12 cells, incubated in a
series of buffers of known pH in the presence of 5 µM nigericin
(K+/H+ ionophore) and monensin (Na+/H+ ionophore), which act
to equilibrate the vesicular pH with the surrounding media (Figure
6A).12 Using this calibration curve, the mean pH of LDCVs in PC-
12 cells was determined to be 5.88 ( 0.08 by converting the ratio
(0.75 ( 0.08) of the two fluorescence intensities obtained from the
vesicles after excitation at 760 nm (Figure 6B) and 692 nm (Figure
6C). Although the pH of secretory vesicles has not previously been
measured in PC-12 cells, this value is in general agreement with
measurements in related cells via in situ and other methods (5.4-5.7
in primary chromaffin cells,10 5.5 in AtT-20 cells,8 5.6-5.7 in
synaptic vesicles of hippocampal and dopaminergic neuronal
culture7).
Furthermore, we quantitatively examined a pharmacological
manipulation of vesicular acidity. It has previously been reported
that acute exposure of chromaffin cells to methamphetamine rapidly
diminishes the pH gradient.10,13 When PC-12 cells loaded with
Mini202 were exposed to 100 µM methamphetamine for 5 min,
the emission ratio increased to 1.19, which corresponds to pH 6.36
(Supporting Information, Figure S4). This result is consistent with
the proposed pharmacological mode of action of methamphetamine
as a transmitter releaser, which at high doses redistributes the
vesicular content to the cytoplasm by causing collapse of the pH
gradient.
Supporting Information Available: The chemical structures,
synthetic procedure, structural and photophysical characterization of
compounds, and cellular imaging protocols. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Chen, G.; Yee, D. J.; Gubernator, N. G.; Sames, D. J. Am. Chem. Soc.
2005, 127, 4544–4545. (b) Halim, M.; Tremblay, M. S.; Jockusch, S.; Turro,
N. J.; Sames, D. J. Am. Chem. Soc. 2007, 129, 7704–7705. (c) Froemming,
M. K.; Sames, D. J. Am. Chem. Soc. 2007, 129, 14518–14522. (d)
Tremblay, M. S.; Lee, M.; Sames, D. Org. Lett. 2008, 10, 5–8. (e) Halim,
M.; Yee, D. J.; Sames, D. J. Am. Chem. Soc. 2008, 130, 14123–14128.
(2) For examples from other groups: (a) Sharma, V.; Wang, Q.; Lawrence,
D. S. Biochim. Biophys. Acta 2008, 1784, 94–99. (b) Lukovic´, E.; Taylor,
E. V.; Imperiali, B. Angew. Chem., Int. Ed. 2009, 48, 6828–6831. (c) Albers,
A. E.; Rawls, K. A.; Chang, C. J. Chem. Commun. 2007, 44, 4647–4649.
(3) Gubernator, N. G.; Zhang, H.; Stall, R. G. W.; Mosharov, E. V.; Pereira,
D. B.; Yue, M.; Balsanek, V.; Vadola, P. A.; Mukherjee, B.; Edwards,
R. H.; Sulzer, D.; Sames, D. Science 2009, 324, 1441–1444.
(4) Greene, L. A.; Tischler, A. S. Proc. Natl. Acad. Sci. U.S.A. 1976, 73, 2424–
2428.
(5) Yelin, R.; Schuldiner, S. Neurotransmitter Transporters: Structure, Func-
tion, and Regulation, 2nd ed.; Humana Press: Totowa, NJ, 2002; pp 313-
354.
(6) Han, J.; Burgess, K. Chem. ReV. 2010, 110, 2709–2728.
(7) (a) Miesenbo¨ck, G.; Angelis, D. A. D.; Rothman, J. E. Nature 1998, 394,
192–195. (b) Mani, M.; Ryan, T. A. Front. Neural Circuits 2009, 3, 1-9.
(8) Wu, M. M.; Grabe, M.; Adams, S.; Tsien, R. Y.; Moore, H.-P. H.; Machen,
T. E. J. Biol. Chem. 2001, 276, 33027–33035.
(9) Haugland R. P. The Handbook: A Guide to Fluorescent Probes and Labeling
Technologies, 10th ed.; Invitrogen: Eugene, OR, 2005; pp 937-955.
(10) Markov, D.; Mosharov, E. V.; Setlik, W.; Gershon, M. D.; Sulzer, D.
J. Neurochem. 2008, 107, 1709–1721.
(11) Sun, W.-C.; Gee, K. R.; Haugland, R. P. Bioorg. Med. Chem. Lett. 1998,
8, 3107–3110.
(12) This is an established method for in situ pH calibration of fluorophores in
intracellular organelles. For examples: (a) Llopis, J.; McCaffery, J. M.;
Miyawaki, A.; Farquhar, M. G.; Tsien, R. Y. Proc. Natl. Acad. Sci. U.S.A.
1998, 95, 6803–6808. (b) Holopainen, J. M.; Saarikoski, J.; Kinnunen,
P. K. J.; Ja¨rvela¨, I. Eur. J. Biochem. 2001, 268, 5851–5856.
(13) Sulzer, D.; Rayport, S. Neuron 1990, 5, 797–808.
In summary, using rational molecular design, we were able to
integrate two molecular functions, the transport by VMAT and
ratiometric optical pH sensing, to develop ratiometric pH-responsive
FFN probes. Through a systematic effort, Mini202 emerged as the
most promising probe, enabling in situ pH measurement of
catecholamine secretory vesicles and methamphetamine-induced pH
changes in PC-12 cells. Mini202 is sufficiently bright, photostable,
and suitable for two-photon microscopy. This new agent comple-
ments the fluorescent protein tags and may enable the study of
mechanisms controlling the secretory pathways in neuroendocrine
cells. Also, screening of drugs and other agents for their effects on
JA101740K
9
8830 J. AM. CHEM. SOC. VOL. 132, NO. 26, 2010