Infrared-emitting, peptidase-resistant fluorescent ligands of the bradykinin B2 receptor: application to cytofluorometry and imaging

Background We have previously reported the design, pharmacological properties and imaging application of bradykinin (BK) B2 receptor (B2R) ligands conjugated with fluorophores such as fluorescein derivatives at their N-terminus. To take advantage of the high penetration of infrared light into living tissues and their low autofluorescence in this region of the spectrum, additional probes conjugated with cyanine dye 7 (Cy7) were synthesized and characterized. Results The antagonist B-9430 (D-Arg-[Hyp3,Igl5,D-Igl7,Oic8]-BK) and the agonist B-9972 (D-Arg-[Hyp3,Igl5,Oic7,Igl8]-BK) were N-terminally extended with the infrared fluorophore Cy7, producing the peptides B-10665 and B-10666, respectively. Pharmacological studies indicated that the agonist B-10666 lost much affinity for the B2R vs. the parent peptide, whereas the antagonist B-10665 better retained its potency vs. B-9430 (competition of [3H]BK binding to human B2R, contractility of the human isolated umbilical vein for which potency losses were more important in each case). Both probes stained HEK 293 cells that expressed the B2R-green fluorescent protein (GFP) construction in a specific manner (confocal microscopy) and with very extensive co-localization of the green and infrared fluorescence in either case. The agonist B-10666 at 100 nM promoted the endocytosis of B2R-GFP in live cells, but not the antagonist version at 10–25 nM. The Cy7-labeled peptides did not label cells expressing the β2-adrenoceptor-GFP construction. B-10665 at low nanomolar concentrations was an effective probe for the recombinant B2Rs in cytofluorometry and macroscopic imaging of cell wells (IVIS imaging system operated for infrared fluorescence detection). Conclusions Despite a propensity for non-specific binding when used at high concentrations and limited sensitivity, Cy7-conjugated peptidase-resistant B2R ligands support original imaging and cytofluorometric applications. Electronic supplementary material The online version of this article (doi:10.1186/s13104-016-2258-1) contains supplementary material, which is available to authorized users.


Background
We have previously reported the design, pharmacological properties and imaging application of bradykinin (BK) B 2 receptor (B 2 R) ligands conjugated with fluorophores such as fluorescein derivatives or AlexaFluor-350 at their N-terminus, with application to microscopy in cells that expressed recombinant receptors and their molecular partners such as arrestins, angiotensin converting enzyme and Rab small GTPases [1][2][3][4].
To take advantage of the high penetration of infrared light into living tissues and their low autofluorescence in this region of the spectrum, we wished to produce and characterize additional probes conjugated with a suitable fluorophore, cyanine dye 7 (Cy7). The parent peptides for the antagonist and the agonist versions are B-9430 (D-Arg-[Hyp 3 ,Igl 5 ,D-Igl 7 ,Oic 8 ]-BK) and B-9972 (D-Arg-[Hyp 3 ,Igl 5 ,Oic 7 ,Igl 8 ]-BK), respectively. They are well characterized pharmacologically [5]. The major kinininactivating ectopeptidases are angiotensin converting enzyme and aminopeptidase P [6,7] and both parent peptides are intrinsically resistant to them due to extensive substitutions with non-natural amino acid residues (Table 1). Thus, the Cy7-conjugated versions, termed B-10665 and B-10666, respectively, were produced, pharmacologically characterized and exploited in original imaging experiments.

Radioligand binding assay
The binding of [ 3 H]BK ([2,3-prolyl-3,4-3 H(N)]-bradykinin, 85.4 Ci/mmol, PerkinElmer, Boston, MA, USA) to adherent intact cells expressing a form of the B 2 R was evaluated precisely as described [8]. HEK 293a cells stably expressing myc-tagged human B 2 Rs [8] were used to estimate receptor affinity via the binding competition of a fixed concentration of the radioligand (3 nM) with several unlabeled peptides applied in a wide concentration range.

Contractility assays
Local ethics review boards approved the anonymous use of segments of human umbilical cords obtained following elective cesarean sections; informed consent was obtained from the mothers. Novel Cy-7 conjugated B 2 R ligands were assayed via the contraction of the human isolated umbilical vein [9,10]. Cumulative concentration effect curves were constructed for the agonist version and its potency compared to that of BK. The antagonist version was introduced 30 min before the construction of the concentration-effect curves of the reference agonist BK and the shift to the right was used to determine their potencies using the Schild regression and pA 2 scale [1].

Cytofluorometry
A subclone of HEK 293 cells, called HEK 293a, originally obtained from Sigma-Aldrich was used in these experiments, either as a non-transfected cell type or as one stably expressing myc-tagged human B 2 Rs [8]. These cells were grown in Dulbecco's modified Eagle's medium supplemented with 10 % fetal bovine serum, 1 % l-glutamine, and 1 % penicillin-streptomycin stock solutions (100×). Cells were detached using the protease-free Cell Dissociation Buffer (Invitrogen), incubated in D-MEM without serum at 37 °C for 30 min under agitation in the presence of a Cy7-conjugated ligands and other drugs, rapidly centrifuged (30 s, 11,000g) and resuspended in phosphate buffered saline. Then, the fluorescence of the cell suspensions was assessed using the BD SORP LSR II cell analyzer for the uptake of the as a function of stimulation and transgene expression. The cytofluorometry results were analyzed using the BD FACS DIVA software.

Microscopy
HEK 293 cells, either not transfected or stably expressing a fully functional rabbit B 2 R construction fused to green fluorescent protein (B 2 R-GFP) [2,11] were grown as described and plated on poly-l-lysine precoated glass surfaces. B 2 R-GFP is a high affinity, fully functional receptor that supports subcellular localization in intact cells; further, the fusion protein is not degraded by short treatments with various ligands (immunoblot evidence) [5] and is not expected to participate to any energy transfer with Cy7-labeled ligands. As a specificity control, other HEK 293 cells were transiently transfected with a vector coding for the β 2 -adrenoceptor-GFP Topaz fusion protein [12] (the vector is a generous gift from Prof. Michel Bouvier, Université de Montréal) using polyethylenimine as described [13]. Cells were generally treated for 30 min with Cy7-conjugated ligands (incubation carried out at 37 °C in humidified atmosphere containing 5 % CO 2 ), rinsed 4-5 times with phosphate buffered saline, observed in microscopy for green and infrared epifluorescence and differential interference contrast microscopy (DIC) using a Quorum Wave FX confocal spinning disk microscope coupled to a Hamamatsu ImageEM digital camera (filters for GFP: excitation 491 nm, emission 536 nm; for Cy7: excitation 642 nm, emission 810). The objective was Leica HCX PL APO 63X/1,30 glycerol immersion lens.

Imaging of infrared fluorescence in macroscopic objects
The integrated imaging station IVIS Lumina LT Series III (PerkinElmer), along with its Living Image software, was exploited to evaluate the distribution of B 2 Rs in macroscopic object using the intrinsic infrared fluorescence of the antagonist B-10665. The apparatus was set in fluorescence mode with filters at 675 nm (excitation) and 831 nm (emission), medium binning and 60 s exposure time. This experiment exploited HEK 293a cells grown in 24-well plates that stably expressed either human or rat B 2 R sequences N-terminally tagged with the myc epitope [8]. Expression of these receptors as a function of pharmacologic treatment was controlled in separate plates of cells that were not fixed or permeabilized using the binding of anti-myc monoclonal antibody (clone 4A6, Millipore, dilution 1:1000) added along goat horseradish peroxidase-conjugated anti-mouse IgG antibodies (Santa Cruz, dilution 1:1000). After 15 min of incubation at 37 °C and 4 washing with phosphate buffered saline, the reaction was revealed by adding the TrueBlue substrate for immunohistochemistry (Kirkegaard & Perry Lab, Inc., Gaithersburg, MD, USA) used as directed. The resulting blue precipitate was photographed at the macroscopic scale.

Data analysis
Results are presented as mean ± SEM. Radioligand binding data were fitted by nonlinear regression to a one-site competition equation using a least-square method (Prism 5.0, GraphPad Software Inc., San Diego, CA, USA) and IC 50 values with their 95 % confidence limits were calculated from this procedure. The same computer program was used to draw concentration-effect curves (least square fitting of sigmoidal dose-response equation with variable slope) and to derive contractile EC 50 values.

Pharmacology of Cy7-conjugated bradykinin analogs
A radioligand binding assay was conducted to evaluate the affinity of the novel fluorescent putative B 2 R ligands ( Fig. 1 [8]. However, the Cy7-conjugated form of the antagonist B-9430, termed B-10665, is about equipotent to its parent peptide ( Fig. 1; Table 1). The human umbilical vein is a contractile bioassay for the endogenous B 2 R, and was exploited to further study the pharmacology of Cy7-conjugated analogs (Fig. 2). The fluorescent peptide B-10666 was confirmed a B 2 R agonist, but with a massive 4240-fold loss of potency vs. BK based on an extrapolated concentration-effect for the  Table 1 former peptide (Fig. 2a). B-10665 has no direct effect on the isolated umbilical vein, but was a surmountable and rather potent antagonist of BK (Fig. 2b) with an estimated pA 2 of 6.83, but significantly less potent than the parent peptide B-9430 in this preparation (pA 2 7.7) [5]. Thus, in this assay, the potency of both Cy7-conjugated peptides vs. their respective parent peptides was inferior to what could have been expected from the radioligand binding competition assay.

Applications based on the fluorescence of Cy7-labeled B 2 R ligands: cytofluorometry
Binding of the antagonist B-10665 was detected using its intrinsic infrared fluorescence in detached, intact HEK 293a cells that stably expressed human myc-tagged B 2 Rs (Fig. 3). The low autofluorescence of cells in the infrared range allowed detecting the binding of the probe at subnanomolar concentrations. The binding was specific as shown by its very low intensity in non-transfected cells and its extensive competition by the non-fluorescent B 2 R antagonist icatibant in co-treated cells (Fig. 3). Higher concentrations of B-10665 produced a specific binding that plateaued in the low nanomolar range, as expected with cells expressing a finite number of receptors, but non-specific fluorescence (measured in the presence of icatibant) increased considerably more, possibly related to the lipophilic nature of the fluorophore (Additional file 1: Figure S4).

Microscopy
Adherent and intact cells that stably expressed the rabbit B 2 R-GFP construction exhibited the typical membraneassociated green fluorescence corresponding to the resting receptor that was obviously absent in non-transfected HEK 293 cells (Fig. 4). There was no infrared fluorescence in resting cells, expressing or not B 2 R-GFP. The antagonist B-10665 (10 or 25 nM) labelled essentially plasma membranes in the infrared if and only if B 2 R-GFP was expressed; both signals co-localized. The plasma membrane fluorescence of B-10665 (10 nM) was prevented in B 2 R-GFP-expressing cells that had been pretreated with the B 2 R antagonist icatibant (Fig. 4), supporting the competition of the fluorescent antagonist receptor binding by the non-fluorescent antagonist.
The agonist B-10666 (100 nM) determined the endocytosis of a fraction of the plasma membrane receptors and colocalized, in part, with B 2 R-GFP in cells (Fig. 5). At the used concentration, the agonist did not label untransfected cells or icatibant-pretreated receptor-expressing cells, supporting a high specificity in cells that were vigorously rinsed before observation.
The comparable β 2 -adrenoceptor-GFP Topaz construction was used in additional control experiments. This fusion protein, when transiently expressed in HEK 293 cells, is mainly expressed at the plasma membrane level but with scattered intracellular labeling more intense than that of stably expressed B 2 R-GFP (Additional file 1: Figure S5). β 2 -adrenoceptor-GFP Topaz is pharmacologically reactive, as shown in isoproterenol-stimulated cells: the receptors are then translocated into endosomes that are better differentiated from the cytosolic background fluorescence at high magnification (Additional file 1: Figure  S5). The agonist-induced endocytosis was prevented by a co-treatment with propranolol (Additional file 1: Figure  S5), supporting the receptor identity. Both the C7-conjugated B 2 R ligands failed to label HEK 293 cells expressing β 2 -adrenoceptor-GFP Topaz (Additional file 1: Figure S6).

Imaging of infrared fluorescence in macroscopic objects
The fluorescence imaging system was used to probe the properties of the antagonist B-10665 in controlled conditions were cell wells containing confluent HEK 293a cells expressing human or rat B 2 Rs were photographed  (Fig. 6a). B-10665 stained the cell wells (from a 24-well plate) in a concentration-dependent manner (1-10 nM), more so for the rat construction. Non-transfected cells, used as controls, showed some non-specific signal when treated with 10 nM of the probe, but not at the 1 nM concentration level. The fluorescence essentially faded after a single 60-sec acquisition period (Fig. 6a, bottom). The expression of the myc-tagged receptors was verified by an immunohistochemistry technique based on antimyc antibodies applied to the same type of macroscopic object, a cell well plate (Fig. 6b). The blue labeling of receptor-expressing cell populations was more intense than the background staining, and the rat construction apparently more densely expressed than the human one. The pharmacological treatments had no obvious effect on the density of receptors. B-10665 (up to 250 nM) could not detect endogenous B 2 R presence in slices of human umbilical cord using this imaging system (data not shown), probably related to the low physiological density of the receptors.

Discussion
The pair of Cy7-conjugated agonist/antagonist ligands of the B 2 R, B-10666 and B-10665, respectively, shared the same design as a previously reported pair based on the same parent peptides, fluorescein-5-thiocarbamoyl (FTC)-B-9972 and FTC-B-9430 [4]. While the conjugation to a fluorophore produced a loss of receptor affinity for both sets of peptides, the Cy7-conjugated agonist B-10666 was particularly poor, whereas FTC-B-9972 fared better (EC 50 108 nM in the umbilical vein contractility assay [4] . 4) [4].
The original properties of the studied Cy7-conjugated ligands included: (1) superior performance in co-localization experiments with the B 2 R-GFP fusion protein (Figs. 4, 5), which was not possible with fluoresceinlabeled ligands, supporting the very extensive co-localization of ligands with the receptor at both the cell surface and, for the agonist version, in endosomes; (2) high signal/noise ratio of nanomolar levels of B-10665 in cytofluorometry, owing to minimal cell autofluorescence in  Fig. 6 Imaging of the B 2 R distribution in a plate of wells containing HEK 293a cells that stably express myc-tagged recombinant human or rat receptors (hu myc-B 2 R, rt myc-B 2 R, respectively) or non-transfected cells (nt). a Infrared fluorescence following treatment with B-10665 (30 min, 37 °C, followed by ample rinsing). The color signal represents infrared fluorescence intensity, as quantified on the scale at the right hand side, and is overlayed on a black and white photograph of the object. Top fluorescence recorded during the first exposure (60 s); bottom results from the second exposure of the same object (60 s). b Detection of the receptors via immunohistochemistry of their myc tag (blue precipitate). The cell wells were submitted to the same treatments as in a. In both a and b, results are representative of 2 separate experiments