Development of Potent Fluorescent Polyamine Toxins and Application

Jul 9, 2013 - Department of Drug Design and Pharmacology, University of Copenhagen, Universitetsparken 2, DK-2100 Copenhagen, Denmark. ‡ Department ...
4 downloads 12 Views 3MB Size
Articles pubs.acs.org/acschemicalbiology

Development of Potent Fluorescent Polyamine Toxins and Application in Labeling of Ionotropic Glutamate Receptors in Hippocampal Neurons Niels G. Nørager,† Christel B. Jensen,† Mette Rathje,‡ Jacob Andersen,† Kenneth L. Madsen,‡ Anders S. Kristensen,† and Kristian Strømgaard†,* †

Department of Drug Design and Pharmacology, University of Copenhagen, Universitetsparken 2, DK-2100 Copenhagen, Denmark Department of Neuroscience and Pharmacology, University of Copenhagen, Blegdamsvej 3B, DK-2200 Copenhagen, Denmark



S Supporting Information *

ABSTRACT: The natural product argiotoxin-636 (ArgTX636) found in the venom of the Argiope lobata spider is a potent open-channel blocker of ionotropic glutamate (iGlu) receptors, and recently, two analogues, ArgTX-75 and ArgTX48, were identified with increased potency and selectivity for iGlu receptor subtypes. Here, we have exploited these analogues as templates in the development of fluorescent iGlu receptor ligands to be employed as unique tools for dynamic studies. Eighteen fluorescent analogues were designed and synthesized, and subsequently pharmacologically evaluated at three iGlu receptor subtypes, which resulted in the discovery of highly potent fluorescent iGlu receptor antagonists with IC50 values as low as 11 nM. The most promising ligands were further characterized showing retention of their mechanism of action, as open-channel blockers of iGlu receptors, as well as preservation of the photophysical properties of the incorporated fluorophores. Finally, we demonstrate the applicability of the developed probes for imaging of iGlu receptors in hippocampal neurons.

F

psychiatric diseases, and iGlu receptors are therefore considered important drug targets for diseases such as stroke, Alzheimer’s disease, and depression and for the treatment of pain.14 The current understanding of the neurophysiological role of iGlu receptors is, to a large extent, based on pharmacological studies, and to address the specific function of the many iGlu receptor subtypes, development of subtype selective ligands has been a key challenge.15 The current understanding of the iGlu receptor structure and function has been greatly facilitated by seminal structural studies of ligand-binding and amino terminal domains,16 and in one case, an X-ray crystal structure of a full-length iGlu receptor (Figure 1A).17 Studies employing fluorescent labeling of iGlu receptors have generally focused on elucidation of localization, trafficking, subunit arrangements and assembly, as well as receptor conformational changes.18−20 In such studies, fluorescent labeling of iGlu receptors has mainly been achieved using genetically encoded fluorophores in the form of fluorescent proteins, covalent fluorescent labeling, or fluorophore-conjugated antibodies. In contrast, only very few fluorescent iGlu receptor ligands have been reported, which generally suffer

luorescence imaging is a powerful methodology for monitoring biomolecules in living systems, facilitating a plethora of studies examining pertinent biological questions.1−3 Benefitting from the high sensitivity of light monitoring, fluorescent imaging enables selective and specific detection of molecules at low concentrations. This is particularly true for studies of fundamental properties of receptor function and dynamics, where a number of methodologies such as fluorescence microscopy,4 fluorescence polarization,5 fluorescence resonance energy transfer (FRET),6 and fluorescence correlation spectroscopy (FCS)7 are being used. This includes, in particular, receptor expression, cellular localization8 and trafficking,9 internalization and recycling,10 protein−protein interactions,11 as well as studies of signal transduction and receptor−ligand interactions.3,7 In the latter case, fluorescent ligand are often considered superior to radioligands, as fluorescent ligands generally have higher sensitivity of detection, allow time-resolved studies, and are comparatively easier to generate and safer to handle.12 The ionotropic glutamate (iGlu) receptors constitute an important class of neurotransmitter receptors in the central nervous system, which mediate the majority of the excitatory synaptic transmission in the vertebrate brain and are crucial for normal brain function.13 Abnormal activity of iGlu receptors is believed to play an essential role in a range of neurological and © 2013 American Chemical Society

Received: April 22, 2013 Accepted: July 9, 2013 Published: July 9, 2013 2033

dx.doi.org/10.1021/cb400272m | ACS Chem. Biol. 2013, 8, 2033−2041

ACS Chemical Biology

Articles

Figure 1. (A) X-ray crystal structure of homomeric GluA2 AMPA receptor (PDB ID 3KG2) with agonist and polyamine toxin binding sites indicated. (B) Top-view of receptor showing opening of channel pore by agonist activation and subsequent pore blocking by polyamine toxins. (C) Modeling of polyamine toxin binding in the ion channel,30 showing the headgroup positioned in the vestibule and the polyamine moiety penetrating the narrow region of the ion channel. (D) General structure of polyamine toxins exemplified by ArgTX-636 (1) and the two analogues ArgTX-75 (2) and ArgTX-48 (3).

studies of the conformational behavior of tetrameric voltagegated potassium channels.28 A prominent class of iGlu receptor ion channel blockers is polyamine toxins, which constitute a group of natural products from spiders and wasps.29 Argiotoxin-636 (ArgTX-636, 1, Figure 1D) isolated from the venom of the orb weaver spider Argiope lobata, is a prototypical example of polyamine toxins. ArgTX-636 contains an aromatic headgroup, an optional amino acid linker, a polyamine backbone, and an optional amino acid tail29 and inhibits iGlu receptors in use-and voltage-dependent manner. We have recently explored structure−activity relationship (SAR) studies of ArgTX-63630 and have identified two analogues of ArgTX-636 as potent and subtype selective iGlu receptor antagonists.31 These analogues, ArgTX-75 (2) and ArgTX-48 (3), showed IC50 values as low as 17 and 19 nM for the N-methyl-D-aspartate (NMDA) and AMPA receptor, respectively. Thus, we envisioned that ArgTX-75 and ArgTX48 would be excellent templates for developing fluorescent ligands, as probes to visualize iGlu receptors in living tissue.

from low potencies, lack of selectivity, and/or employ fluorophores with limited applicability.21−24 Improved fluorescent ligands targeting iGlu receptors are therefore of interest as these offer a number of advantages as tools for fluorescent visualization of native receptors, in particular, in studies of native receptors in cells or tissue. Specifically, advantages include minimum or no perturbation of native receptors, reversible labeling as well as distinctive monolabeling of tetrameric iGlu receptors. This is even further pertinent in studies of Ca2+-permeable α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, which have important physiological functions,25,26 but cannot be exclusively examined in fluorescent studies using genetically encoded tags or through indirectly labeling with antibodies.13 Thus, development of potent, selective fluorescent probes targeting iGlu receptors would enable insight into dynamic ligand−receptor interactions in living cells. As templates for design of fluorescent iGlu receptor ligands, ion channel blockers are particularly attractive (Figure 1A), as such ligands can offer excellent selectivity, high-affinity, and slow off-rates. Furthermore, several classes of iGlu receptor blockers bind in the ion channel of the activated iGlu receptor (Figure 1),27 thus being use-dependent blockers that can potentially be used to selectively visualize active iGlu receptors. Moreover, as the ion channel constitute a spatially well-defined singular binding site in the tetrameric iGlu receptors (Figure 1B), fluorescent channel blockers can potentially be of use for resonance energy transfer (RET) studies, as demonstrated in



RESULTS AND DISCUSSION Design of Fluorescent iGlu Receptor Ligands. Fluorescent ligands are typically generated either by appending a fluorophore to a ligand via a linker or by substituting part of the ligand with a fluorophore.21,32 The former strategy is intended to leave the binding interactions of the ligand unaltered; however, linker type, length, and attachment point can affect 2034

dx.doi.org/10.1021/cb400272m | ACS Chem. Biol. 2013, 8, 2033−2041

ACS Chemical Biology

Articles

Scheme 1. Solid-Phase Synthesis of Compounds 10a−b, 11a−b, 12a−b, and 18a−ba

a Reagents and conditions: (a) BAL polystyrene resin (loading: 0.87 mmol/g), NaBH(OAc)3, DMF/AcOH (9:1), RT; (b) Boc-L-Arg(Pbf)−OH, HATU, DIPEA, CH2Cl2/DMF (9:1), RT; (c) N-Teoc 7-aminoheptan-1-ol or N-Teoc 4-aminobutan-1-ol, Bu3P, ADDP, CH2Cl2/THF (1:1), RT; (d) TBAF, THF, 55 °C; (e) Fmoc-L-Asn(Trt)−OH, HATU, DIPEA, CH2Cl2/DMF (9:1), RT; (f) piperidine, DMF, RT; (g) 7-amino-4-methyl-3coumarinylacetic acid, 5(6)-carboxy-2′,7′-difluorofluorescein or 5-carboxyfluorescein, HATU, DIPEA, CH2Cl2/DMF (9:1), RT; (h) DBU, βmercaptoethanol, DMF, RT; (i) TFA/CH2Cl2/TIPS/H2O (75:20:2.5:2.5), RT.

fluorophore attached to the headgroup via a polyethylene glycol (PEG) linker were designed. Synthesis. Inspired by the previously reported solid-phase synthesis of 2 and 3, we developed a series of synthetic procedures for the designed fluorescent analogues, and applied these in parallel for polyamine moieties of both compounds 2 and 3.31 To enable late-stage diversification and incorporation of sensitive fluorophores, a revised resin attachment point and direction of synthesis was required. The syntheses commenced with the loading of two mononosyl (Ns) protected diamines, 4, onto a polystyrene resin through reductive amination onto a backbone amide linker (BAL) resin, 5 (Scheme 1).40 A protected arginine was subsequently coupled to the resinbound secondary amine providing 6, and the polyamine chain was elongated with N-2-(trimethylsilyl)ethyloxycarbonyl (Teoc) protected amino alcohols in a Fukuyama−Mitsunobu alkylation.41 Subsequent removal of the Teoc group resulted in intermediate 7, which is a common intermediate for all synthetic strategies. For the synthesis of fluorescent toxins comprising carboxyl fluorophores AMC (10a−b), Oregon Green (11a−b), and fluorescein (12a−b), intermediate 7 was coupled to a protected asparagine, followed by removal of the Fmoc group, to give intermediate 8. The three commercially available carboxyl fluorophores AMC, Oregon Green, and fluorescein were readily coupled to 8 and removal of the Ns group, cleavage from the resin, and concomitant deprotection provided the desired fluorescent analogues 10a−b, 11a−b, and 12a−b in good yields (11−20%, 9 steps, 78−84% average yield per step). For BODIPY containing toxins, the procedure used for compounds 10−12 did not provide the desired compounds, as the BODIPY moiety was unstable to the acidic conditions required for cleavage from the BAL resin. Instead, we considered that an off-resin procedure would be feasible.

these interactions, and the physiochemical properties of the ligand are often compromised by a substantial increase in size.33 Introduction of a fluorescent moiety through replacement of part of the ligand will inevitably alter the ligand−receptor interactions. However, this strategy has the potential of increasing spatial resolution, relative to a ligand with a flexible linker-bound fluorophore, thus improving feasibility of bioimaging experiments measuring molecular distances such as RET studies. The two templates, compounds 2 and 3 (Figure 1D), comprise a long linear polyamine chain and a bulky aromatic headgroup, and modeling studies propose that the headgroup is positioned in the vestibule of the ion channel, while the polyamine moiety penetrates the narrow region of the ion channel (Figure 1C).30,34 Since fluorophores are generally composed of large conjugated systems,35,36 we envisioned that the bulky headgroup could be replaced by a fluorescent moiety. In the selection of fluorescent moieties, we focused on employing small fluorophores with high brightness and absorption and emission wavelength shifted toward the red end of the spectrum. The first choice was the group of 4,4difluoro-4-borata-3a-azonia-4a-aza-s-indacene (BODIPY) fluorophores,37 and second the less bright but smaller group of coumarins, 38 specifically the 7-amino-4-methylcoumarin (AMC) analogue.36 Moreover, we selected the widely used but bulky xanthene dyes, fluorescein36 and Oregon Green,39 to explore the toleration of bulkiness. We have recently demonstrated that the asparagine amino acid linker of compounds 2 and 3 (Figure 1D) can be substituted with bulky, aromatic amino acids (His, Trp), while retaining or even improving potency.31 Thus, we designed analogues where BODIPY and fluorescein fluorophores replaced both headgroup and asparagine linker of 2 and 3. To explore the linker strategy, analogues carrying a fluorescein 2035

dx.doi.org/10.1021/cb400272m | ACS Chem. Biol. 2013, 8, 2033−2041

ACS Chemical Biology

Articles

Scheme 2. Synthesis of Compounds 15a−b, 16a-b, 21a−b, 26a−b, and 27a−ba

Reagents and conditions: (a) N3-L-Asn-OH, HATU, DIPEA, CH2Cl2/DMF (9:1), RT; (b) DBU, β-mercaptoethanol, DMF, RT; (c) TFA/ CH2Cl2/TIPS/H2O (75:20:2.5:2.5), RT; (d) H-BODIPY acetylene or Me-BODIPY acetylene, CuSO4, sodium ascorbate, DMF/H2O (1:1), RT. (e) Imidazole-1-sulfonyl azide hydrogen sulfate, ZnCl2, DIPEA, CH2Cl2, RT; (f) N3−PEG(7)−C6H4CH2CO2H, HATU, DIPEA, CH2Cl2/DMF (9:1), RT. a

tosylation and subsequent sodium azide substitution47 of the same intermediate provided a complex mixture (Supporting Information Scheme S1). Instead, diazo transfer with imidazole1-sulfonyl azide on intermediate 7 was attempted.48 Use of CuSO4 and K2CO3 as catalysts gave no conversion, probably due to the very low solubility of the catalysts in solvents required for swelling of the resin. However, use of ZnCl2 and DIPEA resulted in near quantitative conversion (Scheme 2). Following standard procedures for deprotection and cleavage from the resin provided 20, and the acetylene functionalized BODIPY fluorophores were readily reacted with 20 to furnish compounds 21a−b. Finally, four analogues containing a monodisperse PEG linker between the headgroup and fluorophore were synthesized, with and without the asparagine linker. We envisioned that the PEG linker could be readily attached via the phenol moiety of the headgroup and exploited that a p-hydroxyl group is sufficient for biological activity,31 thereby circumventing challenges of regioselectivity for the linker attachment. A headgroup with a PEG linker containing a terminal azido group was prepared (see the Supporting Information (SI)) and subsequently coupled to intermediates 7 and 8 (Scheme 2). The standard procedures for deprotection and cleavage from the resin provided azido derivatives 24 and 25, which were reacted with the Me-BODIPY acetylene fluorophore in a CuAAC reaction yielding the four analogues 26a−b and 27a−b (Scheme 2). Pharmacology. To address the potencies of the compounds for iGlu receptors, their inhibitory potency was evaluated at recombinantly expressed NMDA, AMPA, and kainate (KA) receptors using two-electrode voltage-clamp electrophysiology in Xenopus laevis oocytes. Specifically, we

Attachment of the BODIPY moiety to the polyamine chain would require a reaction orthogonal to the nucleophilic functionalities of the polyamine, which we envisioned could be achieved by the copper-catalyzed azide−alkyne 1,3-dipolar cycloaddition (CuAAC) reaction.42 This would replace an amide bond with a 1,2,3-triazole, which has previously been used as an amide bond bioisoster.43 For the synthesis of BODIPY analogues 15a−b and 16a−b, commercially available azido-asparagine was coupled to intermediate 7, followed by removal of Ns, cleavage from resin, and deprotection to furnish azido polyamine 14 (Scheme 2). Two acetylene functionalized BODIPY fluorophores, readily synthesized in a condensation reaction between 4-pentynoyl chloride and 1H-pyrrole or 2,4dimethyl-1H-pyrrole (see Supporting Information),44 were reacted with 14 in a CuAAC reaction, yielding the target compounds 15a−b and 16a−b. For the analogues where fluorescein and BODIPY fluorophores are replacing both a headgroup and an asparagine linker, corresponding to compounds 12a−b and 15a−b with an excised asparagine linker, two synthetic strategies were employed. The fluorescein analogues were readily synthesized by omission of the asparagine coupling step, providing analogues 18a−b (Scheme 1). In the synthesis of the corresponding BODIPY analogues, the Fukuyama−Mitsunobu reaction of azido alcohol building blocks to give intermediate 6 was not feasible, due to the explosive nature of these building blocks.45 Instead, we anticipated that the azido group could be introduced by either substitution of an alcohol or diazotransfer on a resin-bound intermediate.45 An alcohol derivate of 7 was reacted with diphenylphosphoryl azide (DPPA) under Mitsunobu conditions, but no conversion was observed,46 and similarly, 2036

dx.doi.org/10.1021/cb400272m | ACS Chem. Biol. 2013, 8, 2033−2041

ACS Chemical Biology

Articles

relative to the parent template. Compounds 10a−b carrying the relatively small AMC fluorophore produced significantly more inhibition than the larger Oregon Green (11a−b) and fluorescein (12a−b) analogues at both AMPA and NMDA receptors (Figure 2). This suggests that the size of the fluorophore is important for affinity, although the negative charge of the fluorophores in 11a−b and 12a−b might also affect affinity. The correlation between size of the fluorophore and inhibition is further substantiated by compounds 15a−b and 16a−b carrying the medium-sized BODIPY fluorophores, which generally showed lower inhibition than compounds 10a−b but higher than 11a−b and 12a−b (Figure 2). In addition, these results demonstrate that the bioisosteric replacement of an amide bond with a triazole in compounds 15a−b and 16a−b was tolerated. Interestingly, the ArgTX-48 analogues 15b and 16b show significantly higher AMPA receptor affinity compared to ArgTX-75 analogues 15a and 16a at the NMDA receptor, as measured by their inhibition at 1 μM (Figure 2). This could be rationalized by differences in size of the vestibule above the ion channel (Figure 1C), which is supposed to be narrower in the NMDA receptor than in the AMPA receptor,49,50 or by differences in binding conformations in the two receptors. The substitution pattern of the BODIPY fluorophore, being either unsubstituted (15a−b) or tetramethylated (16a−b), did not show difference in inhibition (Figure 2). Excision of the asparagine linker generally improved affinity at the NMDA receptor, as both the fluorescein (18a) and BODIPY (21a) compounds showed increased inhibition compared to their asparagine containing counterparts (12a and 15a, respectively). In contrast, similar modifications of 12b and 15b did not result in improved inhibition at the AMPA

selected the NMDA receptor subtype GluN1/2A, the AMPA receptor subtype GluA1 and the KA receptor subtype GluK1 as representative subtypes for the three major iGlu receptor subclasses. Compound affinities were initially assessed by measuring their degree of inhibition at a concentration of 1 μM (Figure 2 and SI Table S1), followed by determination of IC50

Figure 2. Screening of percentage inhibition (mean ± standard deviation; N > 4) by 1 μM fluorescent ligand at NMDA (black bars) and AMPA (gray bars) receptors in oocytes held at a membrane potential of −80 mV.

values from full dose−response curves for the most potent compounds (Table 1 and Figure 3). Since ligand occupation of the channel result in complete inhibition of the ion flow, the inhibitory potency accurately reflects the ligand binding affinity of the compounds examined. The initial assessment of compound inhibition at a single concentration of 1 μM allowed us to compare the impact of incorporation of the different fluorophores on ligand affinity

Table 1. IC50 Values of Fluorescent Compounds at NMDA and AMPA Receptors

Values determined from nonlinear regression fitting to a logistic equation of composite dose−response data obtained from 4 to 8 oocytes. Numbers in brackets denote the 95% confidence interval for IC50. bInhibition of the current elicited by 100 μM L-glutamate and 100 μM glycine by simultaneous coapplication of the antagonist in oocytes injected with a 1:10 ratio of GluN1/N2A. cInhibition of the current elicited by 300 μM Lglutamate by simultaneous coapplication of the antagonist in oocytes injected with GluA1. dData from ref 31. a

2037

dx.doi.org/10.1021/cb400272m | ACS Chem. Biol. 2013, 8, 2033−2041

ACS Chemical Biology

Articles

Figure 3. Inhibitory potency of compounds at recombinant iGlu receptor subtypes. (A) Representative two-electrode voltage-clamp current recording illustrating the standard testing protocol. Oocytes expressing iGlu receptor were exposed to maximally effective concentrations of agonist, followed by increasing concentrations of the test toxin plus agonist (B, C). Composite concentration−response curves for compounds 10a (B) and 21a (C) at GluN1/2A (■) and GluA1 (▲) receptors at membrane potentials of −80 mV. Error bars are SEM (standard error of the mean) and are shown when larger than symbol size.

Figure 4. (A) Absorption spectra and (B) fluorescence emission spectra for 10a−b, 15b, 16b, and 21a−b. 10a−b was excited at 360 nm and 15b, 16b and 21a−b at 498 nm. (C) Photophysical data of the most potent fluorescent analogues. bAbsorption maximum. cEmission maximum. d Maximum molar extinction coefficient. eAbsolute quantum yield. f Determined by multiplication of extinction coefficient and quantum yield.

An important property for the applicability of fluorescent probes in a biological system is that the probes remain associated with the target of interest during the period of observation. For channel blockers, this property is inversely proportional to the loss of inhibition for the compounds at the receptor. We therefore estimated this by measuring loss of inhibition after 1.5 and 5 min at NMDA and AMPA receptors, respectively, for the most potent compounds (SI Table S2). Although the fluorescent analogues showed a slightly increased loss of inhibition compared to 2 and 3, most compounds retained 40−60% of the inhibition in the period of observation. The introduction of changes to the parent argiotoxins could potentially affect the voltage-dependent inhibition and uncompetitive antagonism that are hallmarks of the pore blocking mechanism of polyamine toxins.29 We measured voltage dependency by determining compound inhibition at membrane potentials of −40, −60, and −80 mV for the most potent compounds (SI Figure S1). As expected, the inhibition decreased at less negative membrane potentials. Additionally, for all compounds, no changes in inhibition were observed when glutamate concentration was increased from 300 to 3000 μM, thus indicating that all compounds retain their uncompetitive mechanism of action (SI Figure S2).

receptor (18b and 21b, respectively). Compounds 26a−b and 27a−b, having the fluorophore attached to the headgroup via a PEG linker, showed a remarkable reduction in inhibition at NMDA receptors compared to 2, while all four compounds showed approximately 50% inhibition of the AMPA receptor at a 1 μM concentration (Table 1). Finally, all compounds were also tested at the kainate receptor subtype GluK1, but none showed inhibition >50% (SI Table S1), thus showing that the fluorescent polyamine toxin analogues have preference for NMDA and AMPA receptors. For the compounds 10a−b, 15b, 16b, 21a, and 21b that had shown the highest inhibition at 1 μM concentration, full dose− response curves were generated at AMPA, NMDA, or both receptor subtypes to determine their IC50, which reflects Kd values (Table 1, Figure 3). The ArgTX-48 analogues (10b, 15b, 16b, and 21b) all showed IC50 values on the AMPA receptor in the nM range, with 10b and 15b being the most potent inhibitors with IC50 values of 61 nM for both compounds. The ArgTX-75 analogues 10a and 21a had IC50 values of 86 and 124 nM, respectively at the NMDA receptor. Both compounds also showed potent inhibition of the AMPA receptor, with 10a being more potent than ArgTX-48 (3) with an IC50 value as low as 11 nM. 2038

dx.doi.org/10.1021/cb400272m | ACS Chem. Biol. 2013, 8, 2033−2041

ACS Chemical Biology

Articles

Figure 5. Confocal laser scanning images of cultured hippocampal neurons expressing mCherry-tagged NMDA receptors. Cultures were subjected to 2 min incubation with 50 nM 21a (left) or 50 nM 26a (right) during stimulation with 100 μM glycine in the absence of Mg2+. Live-cell images of mCherry and 21a or 26a at the beginning of wash-out of excess ligand (T0) and following five minutes wash (T5). Lower panels show overlay of mCherry and 21a or 26a fluorescence in representative dendritic regions in the culture (dashed boxes). Note clusters of mCherry-NMDA receptors along the dendrite that overlap with intense 21a fluorescence, whereas no fluorescence is observed for 26a. The shown cultures also contain untransfected neurons expressing endogenous NMDA receptors, which are also targeted by 21a. Scale bars: 50 μm.

Spectroscopic Properties. The fluorophores were selected based on their favorable size, brightness, and absorption at longer wavelength toward the red end of the spectrum. However, these properties might be affected when introduced into the polyamine scaffold. Thus, we examined absorption and emission properties for all compounds (Figure 4 and SI Table S3), and found that these properties had been retained for all fluorophores, except for the BODIPY containing analogues. For these, the extinction coefficient was found to be sensitive to the position of the fluorophore. When the fluorophore was attached via the PEG-linker (26a−b and 27a−b), no significant change was observed (SI Table S3). However, when attached directly onto the polyamine chain (21a−b) or via the asparagine linker (15a−b and 16a−b) slightly lower extinction coefficients were observed. For the six most potent compounds, those containing a BODIPY moiety showed most favorable spectroscopic properties, with absorption and emission at longer wavelengths, as well as high brightness (Figure 4). Thus, in particular, compounds 21a and 21b are highly promising candidates for labeling iGlu receptors. Imaging of Native iGlu Receptors in Hippocampal Neurons. The primary purpose of the development of fluorescent channel blockers is their potential use for visualization of iGlu receptors in live cells. To examine this, we tested the ability of 21a to label NMDA receptors in hippocampal neurons dissociated from E19 rats (see the Supporting Information). From the two potent NMDA receptor inhibitors identified, 10a and 21a, we selected 21a because of its emission at longer wavelengths relative to 10a. We also selected 26a as a representative ligand with low affinity for the NMDA receptor to assess nonspecific binding. To enable comparison of subcellular labeling patterns with NMDA receptor expression patterns, we transfected neurons with cDNA encoding GluN1 and GluN2A subunits genetically tagged with the red fluorescent protein mCherry to allow simultaneous visualization of NMDA receptor with green ligand fluorescence. We imaged neurons after application of a 50 nM concentration of 21a or 26a in the presence of 100 μM glycine and absence of Mg2+ to stimulate synaptic NMDA receptor activity (Figure 5). mCherry-tagged NMDA receptors were

distributed throughout soma and dendrites with some synaptic localization (Figure 5). The subcellular expression pattern was consistent with previous reports on expression of GFP-tagged GluN1/2A NMDA receptors in hippocampal neurons.51 Immediately following switch into wash-out of excess ligand (Figure 5, T0), uniform green fluorescence was present at all cell surfaces, indicating substantial unspecific binding to the cell membrane. As the polyamine tail is most likely positively charged at the physiological pH used during the labeling experiments, this observed nonspecific binding to the phospholipid bilayer is expected. However, green fluorescence diminished completely within minutes of continued perfusion and following a 5-min period of wash, the culture with reference compound 26a showed no detectable signs of membrane localized fluorescence (Figure 5; T5). In contrast, following a similar period of wash-out, the culture treated with 21a displayed punctuate fluorescence along dendrites with overlap with punctuate mCherry fluorescence in the dendritic regions of the transfected neurons (Figure 5, lower panels), which is consistent with these regions harboring excitatory synaptic sites. The fluorescence from 21a did not overlap with mCherry fluorescence in the soma region, where the strong mCherry signal likely is due to the intracellular receptor pools that cannot be reached by 21a. Note that the culture also contains dendrites from untransfected neurons, which express endogenous NMDA receptors that also are targets for the fluorescent ligands. These results demonstrate that fluorescent polyamine toxins have the potential as markers of NMDA receptors in live neuronal cells. Notably, in addition to their general use as labels, potential advantages of fluorescent polyamine toxins over other types of fluorescent probes, such as fluorophore-conjugated antibodies, include their potential use for activity-dependent labeling by taking advantage of the use-dependent binding mode of the toxins. Thus, these probes can be used to selectively identify and visualize active iGlu receptor populations, which is relevant in several neurobiological settings. In summary, we have designed and synthesized fluorescent analogues of ArgTX-75 and ArgTX-48, which were all prepared from a common intermediate using a revised solid-phase synthesis procedure. Both on- and off-resin labeling strategies, 2039

dx.doi.org/10.1021/cb400272m | ACS Chem. Biol. 2013, 8, 2033−2041

ACS Chemical Biology

Articles

determine the properties of ligand−receptor complexes. Trends Pharmacol. Sci. 28, 637−645. (8) Wu, B., Piatkevich, K. D., Lionnet, T., Singer, R. H., and Verkhusha, V. V. (2011) Modern fluorescent proteins and imaging technologies to study gene expression, nuclear localization, and dynamics. Curr. Opin. Cell Biol. 23, 310−317. (9) Pick, H., Preuss, A. K., Mayer, M., Wohland, T., Hovius, R., and Vogel, H. (2003) Monitoring expression and clustering of the ionotropic 5HT3 receptor in plasma membranes of live biological cells. Biochemistry 42, 877−884. (10) Daigle, T. L., Kwok, M. L., and Mackie, K. (2008) Regulation of CB1 cannabinoid receptor internalization by a promiscuous phosphorylation-dependent mechanism. J. Neurochem. 106, 70−82. (11) Maurel, D., Comps-Agrar, L., Brock, C., Rives, M.-L., Bourrier, E., Ayoub, M. A., Bazin, H., Tinel, N., Durroux, T., Prezeau, L., Trinquet, E., and Pin, J.-P. (2008) Cell-surface protein−protein interaction analysis with time-resolved FRET and snap-tag technologies: Application to GPCR oligomerization. Nat. Methods 5, 561−567. (12) Bohme, I., and Beck-Sickinger, A. (2009) Illuminating the life of GPCRs. Cell Commun. Signal. 7, 16. (13) Traynelis, S., Wollmuth, L., McBain, C., Menniti, F., Vance, K., Ogden, K., Hansen, K., Yuan, H., Myers, S., and Dingledine, R. (2010) Glutamate receptor ion channels: Structure, regulation, and function. Pharmacol. Rev. 62, 405−496. (14) Pinheiro, P. S., and Mulle, C. (2008) Presynaptic glutamate receptors: Physiological functions and mechanisms of action. Nat. Rev. Neurosci. 9, 423−436. (15) Ogden, K., and Traynelis, S. (2011) New advances in NMDA receptor pharmacology. Trends Pharmacol. Sci. 32, 726−733. (16) Stawski, P., Janovjak, H., and Trauner, D. (2010) Pharmacology of ionotropic glutamate receptors: A structural perspective. Bioorg. Med. Chem. 18, 7759−7772. (17) Sobolevsky, A., Rosconi, M., and Gouaux, E. (2009) X-ray structure, symmetry and mechanism of an AMPA-subtype glutamate receptor. Nature 462, 745−756. (18) Ramanoudjame, G., Du, M., Mankiewicz, K. A., and Jayaraman, V. (2006) Allosteric mechanism in AMPA receptors: A FRET-based investigation of conformational changes. Proc. Natl. Acad. Sci. U.S.A. 103, 10473−10478. (19) Rambhadran, A., Gonzalez, J., and Jayaraman, V. (2010) Subunit arrangement in N-methyl-D-aspartate (NMDA) receptors. J. Biol. Chem. 285, 15296−15301. (20) Henley, J. M., Barker, E. A., and Glebov, O. O. (2011) Routes, destinations and delays: Recent advances in AMPA receptor trafficking. Trends Neurosci. 34, 258−268. (21) Marchand, P., Becerril-Ortega, J., Mony, L., Bouteiller, C., Paoletti, P., Nicole, O., Barré, L., Buisson, A., and Perrio, C. (2011) Confocal microscopy imaging of NR2B-containing NMDA receptors based on fluorescent Ifenprodil-like conjugates. Bioconjugate Chem. 23, 21−26. (22) Joubert, J., Dyk, S. v., Green, I. R., and Malan, S. F. (2011) Synthesis and evaluation of fluorescent heterocyclic aminoadamantanes as multifunctional neuroprotective agents. Bioorg. Med. Chem. 19, 3935−3944. (23) Nishimaru, T., Sano, M., Yamaguchi, Y., and Wakamiya, T. (2009) Syntheses and biological activities of fluorescent-labeled analogs of acylpolyamine toxin NPTX-594 isolated from the venom of Madagascar Joro spider. Bioorg. Med. Chem. 17, 57−63. (24) Vytla, D., Combs-Bachmann, R. E., Hussey, A. M., Hafez, I., and Chambers, J. J. (2011) Silent, fluorescent labeling of native neuronal receptors. Org. Biomol. Chem. 9, 7151−7161. (25) Kwak, S., and Weiss, J. (2006) Calcium-permeable AMPA channels in neurodegenerative disease and ischemia. Curr. Opin. Neurobiol. 16, 281−287. (26) Man, H.-Y. (2011) GluA2-lacking, calcium-permeable AMPA receptors-inducers of plasticity. Curr. Opin. Neurobiol. 21, 291−298. (27) Green, A., Nakanishi, K., and Usherwood, P. (1996) Polyamine amides are neuroprotective in cerebellar granule cell cultures challenged with excitatory amino acids. Brain Res. 717, 135−146.

as well as a linker approach, were employed providing the 18 target compounds in good yields. Evaluation at subtypes of iGlu receptors led to the discovery of a group of potent, fluorescent iGlu receptor antagonists, with the most potent compound, 10a, having IC50 values as low as 11 nM and 86 nM on AMPA and NMDA receptors, respectively. It was also demonstrated that these compounds have preserved their mechanism of action, that is, binding in the ion channel of iGlu receptors as voltage-dependent and uncompetitive antagonists. Moreover, one of the key parameters for the applicability, the timedependent loss of inhibition, was shown to be comparable to that of their parent compounds. Finally, we have demonstrated that the fluorescent probe 21a can be used to visualize NMDA receptors in hippocampal neurons. However, the applicability of the developed fluorescent ligands is broader, as the compounds also target Ca2+-permeable AMPA receptors and can potentially be unique tools for visualizing this important subgroup of iGlu receptors.

■ ■

METHODS A complete methods section can be found in the Supporting Information. ASSOCIATED CONTENT

* Supporting Information S

Experimental section, supporting schemes, tables and figures, and detailed compound characterization. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We are grateful for support from the Carlsberg Foundation, the Lundbeck Foundation, and the Drug Research Academy (University of Copenhagen).



REFERENCES

(1) Lichtman, J., and Conchello, J.-A. (2005) Fluorescence microscopy. Nat. Methods 2, 910−919. (2) Qian, X., Xiao, Y., Xu, Y., Guo, X., Qian, J., and Zhu, W. (2010) ″Alive″ dyes as fluorescent sensors: Fluorophore, mechanism, receptor, and images in living cells. Chem. Commun. 46, 6418−6436. (3) Leopoldo, M., Lacivita, E., Berardi, F., and Perrone, R. (2009) Developments in fluorescent probes for receptor research. Drug Discov. Today 14, 706−712. (4) Hern, J. A., Baig, A. H., Mashanov, G. I., Birdsall, B., Corrie, J. E. T., Lazareno, S., Molloy, J. E., and Birdsall, N. J. M. (2010) Formation and dissociation of M1 muscarinic receptor dimers seen by total internal reflection fluorescence imaging of single molecules. Proc. Natl. Acad. Sci. U.S.A. 107, 2693−2698. (5) Bach, A., Clausen, B. H., Møller, M., Vestergaard, B., Chi, C. N., Round, A., Sørensen, P. L., Nissen, K. B., Kastrup, J. S., Gajhede, M., Jemth, P., Kristensen, A. S., Lundström, P., Lambertsen, K. L., and Strømgaard, K. (2012) A high-affinity, dimeric inhibitor of PSD-95 bivalently interacts with PDZ1-2 and protects against ischemic brain damage. Proc. Natl. Acad. Sci. U.S.A. 109, 3317−3322. (6) Roy, R., Hohng, S., and Ha, T. (2008) A practical guide to singlemolecule FRET. Nat. Methods 5, 507−516. (7) Briddon, S. J., and Hill, S. J. (2007) Pharmacology under the microscope: the use of fluorescence correlation spectroscopy to 2040

dx.doi.org/10.1021/cb400272m | ACS Chem. Biol. 2013, 8, 2033−2041

ACS Chemical Biology

Articles

(28) Posson, D. J., Ge, P., Miller, C., Bezanilla, F., and Selvin, P. R. (2005) Small vertical movement of a K+ channel voltage sensor measured with luminescence energy transfer. Nature 436, 848−851. (29) Olsen, C. A., Kristensen, A. S., and Strømgaard, K. (2011) Small molecules from spiders used as chemical probes. Angew. Chem., Int. Ed. 50, 11296−11311. (30) Nelson, J. K., Frølund, S. U., Tikhonov, D. B., Kristensen, A. S., and Strømgaard, K. (2009) Synthesis and biological activity of argiotoxin 636 and analogues: Selective antagonists for ionotropic glutamate receptors. Angew. Chem., Int. Ed. 48, 3087−3091. (31) Poulsen, M. H., Lucas, S., Bach, T. B., Barslund, A. F., Wenzler, C., Jensen, C. B., Kristensen, A. S., and Strømgaard, K. (2013) Structure−activity relationship studies of argiotoxins: Selective and potent inhibitors of ionotropic glutamate receptors. J. Med. Chem. 56, 1171−1181. (32) Briddon, S. J., Kellam, B., and Kill, S. J. (2011) In Methods in Molecular Biology: Receptor Signal Transduction Protocols (Willars, G. B., Challiss, J., Eds.) 3rd ed., Vol. 746, pp 211−236, Springer ScienceBusiness Media, New York. (33) Baker, J. G., Middleton, R., Adams, L., May, L. T., Briddon, S. J., Kellam, B., and Hill, S. J. (2010) Influence of fluorophore and linker composition on the pharmacology of fluorescent adenosine A1 receptor ligands. Br. J. Pharmacol. 159, 772−786. (34) Andersen, T. F., Tikhonov, D. B., Bølcho, U., Bolshakov, K., Nelson, J. K., Pluteanu, F., Mellor, I. R., Egebjerg, J., and Strømgaard, K. (2006) Uncompetitive antagonism of AMPA receptors: Mechanistic insights from studies of polyamine toxin derivatives. J. Med. Chem. 49, 5414−5423. (35) Wysocki, L. M., and Lavis, L. D. (2011) Advances in the chemistry of small molecule fluorescent probes. Curr. Opin. Chem. Biol. 15, 752−759. (36) Johnson, I. and Spence, M. T. Z. (Eds.) Molecular Probes Handbook, A Guide to Fluorescent Probes and Labeling Technologies (2010) 11th ed., Molecular Probes, Eugene, OR. (37) Loudet, A., and Burgess, K. (2007) BODIPY dyes and their derivatives: Syntheses and spectroscopic properties. Chem. Rev. 107, 4891−4932. (38) Bryantseva, N. G., Sokolova, I. V., Tsyrenzhapova, A. B., Selivanov, N. I., Khilya, V. P., and Garazd, Y. L. (2008) Fluorescent characteristics of coumarin photosensitizers. J. Appl. Spectrosc. 75, 700−705. (39) Sun, W.-C., Gee, K. R., Klaubert, D. H., and Haugland, R. P. (1997) Synthesis of fluorinated fluoresceins. J. Org. Chem. 62, 6469− 6475. (40) Jensen, K. J., Alsina, J., Songster, M. F., Vágner, J., Albericio, F., and Barany, G. (1998) Backbone amide linker (BAL) strategy for solid-phase synthesis of C-terminal-modified and cyclic peptides. J. Am. Chem. Soc. 120, 5441−5452. (41) Olsen, C. A., Witt, M., Hansen, S. H., Jaroszewski, J. W., and Franzyk, H. (2005) Fukuyama−Mitsunobu alkylation in amine synthesis on solid phase revisited: N-alkylation with secondary alcohols and synthesis of curtatoxins. Tetrahedron 61, 6046−6055. (42) Tornøe, C. W., Christensen, C., and Meldal, M. (2002) Peptidotriazoles on solid phase: [1,2,3]-triazoles by regiospecific copper(I)-catalyzed 1,3-dipolar cycloadditions of terminal alkynes to azides. J. Org. Chem. 67, 3057−3064. (43) Bock, V. D., Speijer, D., Hiemstra, H., and van Maarseveen, J. H. (2007) 1,2,3-Triazoles as peptide bond isosteres: Synthesis and biological evaluation of cyclotetrapeptide mimics. Org. Biomol. Chem. 5, 971−975. (44) Shah, M., Thangaraj, K., Soong, M.-L., Wolford, L. T., Boyer, J. H., Politzer, I. R., and Pavlopoulos, T. G. (1990) Pyrromethene−BF2 complexes as laser dyes: 1. Heteroat. Chem. 1, 389−399. (45) Bräse, S., Gil, C., Knepper, K., and Zimmermann, V. (2005) Organic azides: An exploding diversity of a unique class OFC. Angew. Chem., Int. Ed. 44, 5188−5240. (46) Nicolaou, K. C., Winssinger, N., Vourloumis, D., Ohshima, T., Kim, S., Pfefferkorn, J., Xu, J. Y., and Li, T. (1998) Solid and solution

phase synthesis and biological evaluation of combinatorial sarcodictyin libraries. J. Am. Chem. Soc. 120, 10814−10826. (47) Lee, C. E., Kick, E. K., and Ellman, J. A. (1998) General solidphase synthesis approach to prepare mechanism-based aspartyl protease inhibitor libraries. Identification of potent cathepsin D inhibitors. J. Am. Chem. Soc. 120, 9735−9747. (48) Goddard-Borger, E. D., and Stick, R. V. (2007) An efficient, inexpensive, and shelf-stable diazotransfer reagent: Imidazole-1sulfonyl azide hydrochloride. Org. Lett. 9, 3797−3800. (49) Burnashev, N., Villarroel, A., and Sakmann, B. (1996) Dimensions and ion selectivity of recombinant AMPA and kainate receptor channels and their dependence on Q/R site residues. The J. Physiol. 496, 165−173. (50) Villarroel, A., Burnashev, N., and Sakmann, B. (1995) Dimensions of the narrow portion of a recombinant NMDA receptor channel. Biophys. J. 68, 866−875. (51) Barria, A., and Malinow, R. (2002) Subunit-specific NMDA receptor trafficking to synapses. Neuron 35, 345−353.

2041

dx.doi.org/10.1021/cb400272m | ACS Chem. Biol. 2013, 8, 2033−2041