Evaluation of [Cys(ATTO 488)8]Dermorphin-NH2 as a novel tool for the study of μ-opioid peptide receptors

The μ-opioid peptide (MOP) receptor is a member of the opioid receptor family and an important clinical target for analgesia. Measuring MOP receptor location and tracking its turnover traditionally used radiolabels or antibodies with attendant problems of utility of radiolabels in whole cells and poor antibody selectivity. To address these issues we have synthesized and characterised a novel ATTO488 based fluorescent Dermorphin analogue; [Cys(ATTO 488)8]Dermorphin-NH2 (DermATTO488). We initially assessed the binding profile of DermATTO488 in HEK cells expressing human MOP and CHO cells expressing human MOP, δ-opioid peptide (DOP), κ-opioid peptide (KOP) and Nociceptin/Orphanin FQ peptide (NOP) receptors using radioligand binding. Functional activity of the conjugated peptide was assessed by measuring (i) the ability of the ligand to engage G-protein by measuring the ability to stimulate GTPγ[35S] binding and (ii) the ability to stimulate phosphorylation of ERK1/2. Receptor location was visualised using confocal scanning laser microscopy. Dermorphin and DermATTO488 bound to HEKMOP (pKi: 8.29 and 7.00; p<0.05), CHOMOP (pKi: 9.26 and 8.12; p<0.05) and CHODOP (pKi: 7.03 and 7.16; p>0.05). Both ligands were inactive at KOP and NOP. Dermorphin and DermATTO488 stimulated the binding of GTPγ[35S] with similar pEC50 (7.84 and 7.62; p>0.05) and Emax (1.52 and 1.34fold p>0.05) values. Moreover, Dermorphin and DermATTO488 produced a monophasic stimulation of ERK1/2 phosphorylation peaking at 5mins (6.98 and 7.64-fold; p>0.05). Finally, in confocal microscopy DermATTO488 bound to recombinant MOP receptors on CHO and HEK cells in a concentration dependent manner that could be blocked by pre-incubation with unlabelled Dermorphin or Naloxone. Collectively, addition to ATTO488 to Dermorphin produced a ligand not dissimilar to Dermorphin; with ~10fold selectivity over DOP. This new ligand DermATTO488 retained functional activity and could be used to visualise MOP receptor location.


Introduction
Opioid receptors are members of the seven transmembrane-spanning G protein-coupled receptor (GPCR) superfamily. The μ (MOP), δ (DOP) and κ (KOP) receptors are classical or naloxone sensitive and the Nociceptin/Orphanin FQ (N/OFQ) receptor (NOP) is naloxone insensitive. Whilst all opioid receptors are capable of the production of analgesia, the main target in the clinic is the MOP receptor. MOP receptors couple to G i /G o G-proteins to enhance an outward potassium conductance to hyperpolarize, close voltage-sensitive calcium channels and inhibit adenylyl cyclase leading to the reduction of cAMP formation. In neurones this ultimately leads to reduced firing and neurotransmitter release [1][2][3][4][5].
The MOP receptor is widely distributed throughout the central nervous system and in neural and non-neural peripheral tissues [6]. Current methods to detect MOP receptor expression have several shortcomings. Use of radiolabels to study opioid receptors in native cells or tissue, where receptor densities are low is difficult due to the generally inadequate quantity of the sample that can be collected along with a relatively low specific activity of available radiolabels. Commercially available opioid receptor antibodies show poor selectivity and detection of mRNA does not necessarily indicate a functional protein [7][8][9]. There are a number of studies looking at turnover of tagged receptors. Typically these use receptors tagged with HA and FLAG but ultimately they require fixation and incubation with anti-HA or anti-FLAG antibodies [10,11].
Dermorphin is a MOP receptor agonist isolated from the skin of the Amazon frog Phyllomedusa sauvagei in the early 1980s [12,13]. Dermorphin binds to MOP with high affinity and an order of magnitude selectivity over DOP [14]. This relatively short (seven amino acids) peptide is easy to manipulate so we have used it as an acceptor for the fluorescent ATTO dye (488nm) to produce [Cys(ATTO 488) 8 ]Dermorphin-NH 2 (Derm ATTO488 ). The use of ATTO dyes leads to extended visualisation when compared to the more commonly used ALEXA dyes and linkage to bioactive peptides provides an increase in sensitivity when compared to use of antibodies or radioligand binding, particularly in low expression systems. Derm ATTO488 will have potential uses for tracking MOP receptors and when used in conjunction with other probes, for example N/OFQ ATTO594 [15] to examine opioid receptor interaction(s).
In this study we determine the binding properties of Dermorphin and Dermorphin ATTO488 along with functional activity in GTPγ[ 35 S] binding and ERK1/2 phosphorylation at recombinant human opioid receptors expressed in HEK and CHO cells. Importantly we use Derm ATTO488 to visualise MOP expression in live CHO and HEK cells using confocal microscopy.

Materials
Dermorphin and Naloxone were purchased from Sigma-Aldrich Co. (Dorset, U.K.). Naltrindole and Dynorphin A were from Tocris (Abingdon, U.K.). Tritiated Diprenorphine ([ 3 H]-DPN) and tritiated N/OFQ ([ 3 H]-N/OFQ) was from Perkin Elmer. Tissue culture media and supplements were from Sigma-Aldrich and Gibco. All other materials and reagents were of the highest purity available.

Synthesis of [Cys(ATTO 488)8]Dermorphin-NH2 (DermATTO488)
The conjugation of [Cys 8 ]Dermorphin-NH 2 to ATTO 488 maleimide (purchased from ATTO-TEC GmbH Am Eichenhang 50 D-57076 Siegen Germany) was achieved using the classic thiol-Michael reaction. A solution of maleimide derivative fluorescent probe (1 mg, 1 equiv.) in CH 3 CN (250 μL) was added to a stirred solution of [Cys 8 ]Dermorphin-NH 2 (1.1 equiv.) in 250 μL of H 2 O, followed by the addition of 25 μL of NaHCO 3 5%. The mixture was stirred in the dark under a nitrogen atmosphere and at room temperature for 15 minutes. The reaction was monitored by analytical HPLC analysis and after completion, preparative HPLC purification of the reaction mixture. Quantitative yield is depicted in the supplement. S1

Cell culture
Human Embryonic Kidney (HEK) or Chinese Hamster Ovary (CHO) cells expressing recombinant human opioid receptors were cultured in MEM or HAM F-12 media respectively until >90% confluence. HEK hMOP cells were prepared in house (plasmid coding for the MOP receptor was from https://www.cdna.org/files/data/OPR0M10000.pdf). CHO hMOP and CHO hKOP cells were provided by T Costa (Instituto Superiore di Sanita, Rome, Italy). CHO hDOP were supplied by E Varga (University of Arizona, USA). CHO hNOP were from GSK (Stevenage, Herts, UK). Media was supplemented with 10% foetal bovine serum, 100μg/ml penicillin, 100μg/ml streptomycin and 2.5μg/ml fungizone. Selection media used for stock cells expressing human recombinant classical opioid receptors (HEK hMOP , CHO hMOP , CHO hKOP and CHO hDOP ) was supplemented with 200μg/ml geneticin (G418). The selection media for cells expressing human recombinant NOP (CHO hNOP ) also contained 200μg/ml hygromycin B. For the live-cell imaging experiments, cells were plated on ethanol-sterilised 25mm coverslips for 48h before use.

Membrane protein preparation
HEK hMOP , CHOP hMOP , CHO hNOP , CHO hKOP and CHO hDOP cells were harvested in harvest buffer consisting of 154mM NaCl, 10mM HEPES and 1.7mM EDTA, pH 7.4. Cells were homogenised (Ultra-Turrax T25 cell homogeniser) and membrane fragments were collected by centrifugation at 13,500rpm for 10min at 4 0 C. Membranes were resupended in fresh harvest buffer and the process was repeated once more. Finally, membranes were suspended in the required volume of assay buffer (0.5% BSA in 50mM Tris, pH 7.4). Protein concentration was determined using the Lowry method. For the GTPγ[ 35 S] binding assay, after harvesting the cells were suspended in homogenisation buffer consisting of 50mM Tris-HCl and 0.2mM EGTA, pH 7.4 and at the final step, membrane fragments were suspended in assay buffer consisting 50mM Tris, 0.2mM EGTA, 100mM NaCl and 1mM MgCl 2 , pH 7.4.

Radioligand displacement assays
40μg of freshly prepared membrane protein was suspended in 0.5ml of buffer containing 50mM Tris, 0.5% BSA and~0.8nM [ 3 H]-Diprenorphine (DPN) (for HEK hMOP , CHO hMOP , CHO hKOP and CHO hDOP ) or~0.8nM [ 3 H]-N/OFQ (for CHO hNOP ) and varying concentrations (1pM-1μM) of unlabelled Dermorphin or Derm ATTO488. Non-specific binding was evaluated with 10μM naloxone for cells expressing the classical opioid receptors and 1μM N/OFQ for CHO hNOP . Reference ligands were added as indicated in the results. After 1h incubation at room temperature reactions were terminated by vacuum filtration onto a PEI-soaked Whatman GF/B filter paper using a Brandel harvester [14,16].

Live-cell imaging
Coverslip cultures of CHO or HEK cells were placed on a Harvard Peltier plate and perfused with Krebs buffer, pH 7.4 at 4˚C (low temperature selected because Derm ATTO488 is an agonist). Derm ATTO488 was injected onto coverslips at four different cumulative concentrations; 1nM, 10nM, 100nM or 1μM. Cells were incubated with the lowest Derm ATTO488 concentration for 3mins, washed with ice-cold Krebs buffer for 2min then imaged. After a further wash of 1mins in HEK cells or 3min in CHO cells (to produce z-stack) the next concentration was added and imaged as described. To define non-specific label binding cells were incubated with 25μM of unlabelled Dermorphin for 5min and then with 1μM of Derm ATTO488 added as above. Additional experiments were also performed using 15μM Naloxone to determine NSB. Images were taken using an oil immersed 60x objective in a Nikon C1Si microscope. Both HEK and CHO cells were imaged using the 488nm wavelength laser with a 20% power setting and a 7.15 gain-green channel. The laser and gain settings were maintained constant in both cell lines. Images were collected by the Nikon C1Si software. For the CHO cells only, due to the low signal observed, images were captured using the 'z-stack project'. All images were analysed using FIJI. Four random regions of interest (ROI) were selected and averaged for background fluorescence. For concentration fluorescence relationship analysis total corrected cell fluorescence was calculated according to the equation: Integrated density-(Area of selected cell X Mean fluorescence of background Readings) [15].

Data analysis
Data are expressed as mean±SEM for (n) individual experiments. In confocal experiments representative images are depicted from 5 independent experiments. All curve fitting was performed using Graphpad Prism 7. In radioligand experiments, ligand affinity (pK i ) was estimated from the concentration producing 50% displacement corrected for the competing concentration of label according to Cheng and Prusoff [17]. In GTPγ[ 35 S] assays, data are presented as a stimulation factor, that is the fold change in GTPγ[ 35 S] binding relative to the basal [14,16]. The concentration of drug producing 50% of the maximum response (pEC 50 ) and the maximum response (E max ) are shown. Statistical analysis (as noted in text and Display legends) was via ANOVA with post-hoc correction and/or t-test as appropriate with P values <0.05 considered significant.

Results
We initially assessed binding affinity and selectivity before confirming ATTO conjugation retained biological activity in GTPγ[ 35 S] binding assay and ERK1/2 stimulation. Once this basic profile was confirmed live cell imaging was used to visualise cell surface binding.

Stimulation of ERK1/2 phosphorylation
A time-response curve in HEK MOP showed that stimulation by 1μM of Derm ATTO488 led to a monophasic pattern of ERK1/2 phosphorylation peaking at 5min and returning towards baseline by 20mins. The peak activity was statistically significant compared to t = 0 and was 7.64 ±1.06 fold greater. Similar data were obtained for Dermorphin (6.98±1.22 fold). These maximum stimulation values were not significantly different from each other (p>0.05). The phosphorylation pattern for Dermorphin and Derm ATTO488 were essentially superimposable, Fig 3  (S3 Fig in S1 File shows the uncropped and unadjusted images of the blots).

Live-cell imaging
Four different concentrations of Derm ATTO488 (1nM, 10nM, 100nM and 1μM) were added to live HEK MOP and CHO hMOP cells and binding was measured using confocal microscopy and a 488nm laser, Figs 4 and 5. In all images nuclei are labelled with DAPI (blue) and the binding of Derm ATTO488 is shown in green. In both cell lines the binding was concentration dependent with the highest cell surface binding measured at the highest concentration used (1μM). Typical images are shown with average corrected total fluorescence for n = 5 independent experiments. Specificity of Derm ATTO488 (1μM) binding in single transfected lines (note; only MOP) was assessed by 5 min pre-incubation with an excess of unlabelled Dermorphin (25μM). As can be seen in Figs 4A and 5A there was demonstrable non-specific binding and this amounted to 16% and 19% in HEK hMOP and CHO hMOP respectively. Pre-incubation with unlabelled Naloxone (15μM) also defined non-specific binding in both HEK hMOP (16%) and CHO hMOP (20%) (S4 and S5 Figs in S1 File). In untransfected wild type HEK and CHO cells Derm ATTO488 (1μM) bound at ultra-low levels, comparable to that defined by unlabelled Dermorphin and Naloxone (Figs 4A-inset and 5A-inset).

Discussion
In this study we have shown that conjugation of the MOP receptor agonist Dermorphin with the ATTO488 fluorophore produced a peptide ligand with slightly reduced binding affinity. Derm ATTO488 retained functional activity in GTPγ[ 35 S] binding and ERK1/2 phosphorylation with effects superimposable to those of the native peptide Dermorphin. The new ligand has 10fold selectivity over DOP and is inactive at KOP and NOP opioid receptors. This new ligand could be effectively used to visualise MOP receptors in live cells. This is a significant advantage over the use of radiolabels with their attendant problems, the generally poor selectivity of commercially available antibodies and the need to employ tagged receptors.
Physical-chemical characterization of ATTO488 is extensive and resulted in fluorophore characteristics for use in localization-based super-resolution imaging [18]. Moreover, ATTO dyes are particularly useful as they exhibit prolonged fluorescence and reduced photobleaching properties [19]. For a synthetic perspective, the fast and high yield reactivity of the ATTO488 maleimide derivative in very mild conditions (room temperature, acetonitrile/water as a solvent) with thiol groups (like the-SH of Cys side chain) make this dye an excellent candidate for easy conjugation with peptide sequences.
The binding affinity of Dermorphin at MOP was 8.29 (HEK) and 9.26 (CHO) representing just under a 10-fold difference. The difference between CHO and HEK (higher affinity in HEK) is difficult to explain but could result from a larger proportion of receptors in the G-protein coupled state as this has higher affinity for agonists [20]. In a previous paper [14] using Dermorphin in CHO cells we reported pK i values at MOP of 8.69 (4-fold weaker than the current data); pK i at DOP was essentially identical and there was no activity at KOP or NOP. In this paper we reported GTPγ[ 35 S] binding and ERK1/2 phosphorylation [14]. A further potential explanation for the discrepancy in binding affinities could be differential coupling of MOP to additional protein(s) in CHO and HEK cells. Opioid receptors are known undergo extensive protein-protein interactions [21]. One such protein that has gained interest with respect to opioid receptor trafficking is Filamin A, a protein that couples membrane proteins to actin. However with respect to our data, Onoprishvili et al [22] showed that the binding of [ 3 H] Diprenorphine and [ 3 H]DAMGO were identical in M2 melanoma cells that do not express Filamin A and its subclone A7 transfected with this protein. A key factor in using any receptor probe (fluorescent or radiolabelled) is to determine specificity of binding; or more specifically to define low non-specific binding (NSB). In these experiments we pre-incubated live cultures with the fluorophore parent; Dermorphin (an agonist) or the non-selective opioid receptor antagonist naloxone. In both cases Derm ATTO488 binding was displaced such that at 1μM labelling concentrations only 16-20% of it was nondisplaceable. NSB was at the higher end in CHO compared to HEK cellular background. To add a further layer of definition, untransfected HEK and CHO cells were incubated with 1μM Derm ATTO488 ; in the absence of MOP receptors, probe binding was low and comparable (raw fluorescence) to that observed in the traditional displacement based NSB definition.
We have also reported previously on the conjugation of ATTO594 (red) to the NOP ligand N/OFQ. Unlike conjugation of ATTO488 to Dermorphin this conjugation produced a fluorescent ligand with no loss of affinity. Moreover, N/OFQ ATTO594 displayed no activity at MOP, DOP or KOP and the ligand retained full biological activity in experiments to measure inhibition of cAMP formation and ERK1/2 phosphorylation. N/OFQ ATTO594 was used to visualise receptor location in recombinant and native (immune cells) NOP expression systems [15]. A further difference between the binding of Derm ATTO488 and N/OFQ ATTO594 relates to NSB. In the present study we report NSB for the Dermorphin ligand between 16 and 19%; for N/OFQ ligand the NSB was effectively undetectable in the conditions used.
There is extensive evidence for opioid dimers in both in vitro recombinant expression systems and in vivo [23][24][25]. Use of both N/OFQ ATTO594 (red) and DermA TTO488 (green) in double expression systems where there is a possibility to use FRET to visualize NOP-MOP dimers will an interesting further use of these new ligands. Studies with these new fluorescent ligands as agonists offer the possibility to examine receptor trafficking following activation of the receptor. There have been other studies using opioid peptides conjugated to other fluorescent 'probes'. Gaudriault et al (1997) synthesized BODIPY derivatives of Dermorphin and the DOP peptide agonist Deltorphin-I. These probes were used in COS cells transfected with rat MOP and DOP receptors. The authors used these ligands in internalisation studies of live cells and showed that both were capable of visualising receptors and that receptor (probe) internalisation was time and temperature dependent [26]. A range of fluorescent peptide agonists including Dermorphin-Bodipy Texas Red and Dermorphin Alexa488 were synthesized by Arttamangkul et al. In radioligand binding studies using CHO cells expressing rat MOP and DOP the authors reported good MOP selectivity; no data are presented for KOP or NOP. Functional activity was confirmed electrophysiologically in rat locus ceruleus. Importantly at low temperature (4-8˚C) binding measured by confocal microscopy in CHO cells was located to the membrane that internalized on warming (32-35˚C). Interestingly, Alexa488 conjugated to the DOP antagonist TIPP bound but failed to internalize [27]. Further use of these Dermorphin-Bodipy Texas Red and Dermorphin Alexa488 probes in primary cultures of mouse locus ceruleus showed that desensitization is not dependent on receptor internalization [28].
There are a number of drawbacks to Derm ATTO488 . First, Dermorphin is an agonist and as such it will produce receptor activation; we have clearly shown that in this paper. In imaging experiments where cell surface receptor expression is critical activation will lead to Arrestin recruitment, we have shown this previously [14] and ultimately internalization. As such confocal experiments need to be performed at 4˚C to limit this process. Whilst a clear disadvantage for studies examining internalisation, protocols to warm cells should enable the internalisation process to be tracked. The second major drawback (no issue in single recombinant expression systems) relates to selectivity of only~10 fold over DOP noted above. In native tissues aimed at exploring MOP receptor expression, DOP receptors are also often co-expressed so careful use of label concentration is needed to allow discrimination or pre-blocking with highly selective DOP antagonists such as naltrindole.
In summary Derm ATTO488 represents a useful addition to the ligand toolkit to study opioid receptor expression, function and turnover.