Pharmacological characterization of high‐affinity σ1 receptor ligands with spirocyclic thienopyran and thienofuran scaffold

In this study, the pharmacological properties of six spirocyclic piperidines 1–6 showing very high σ1 receptor affinity (Ki = 0.2–16 nm) were investigated.


Introduction
Initially, r receptors were assigned to the class of opioid receptors due to the high affinity for (+)-benzomorphans (such as (+)-pentazocine and (+)-SKF10,047). [1] However, this classification was dismissed later on since the opioid receptor antagonists naltrexone and naloxone were not able to compensate the (+)-benzomorphan mediated effects. [2,3] Two r receptor subtypes were pharmacologically identified and subsequently termed as r 1 and r 2 receptor. [4] The r 1 receptor shows a widespread distribution within the central nervous system (CNS) as well as in peripheral tissues such as kidney, liver, lung and heart. [5][6][7][8] The corresponding r 1 gene was cloned from several animal species and humans displaying a 95% similarity of the amino acid sequence. The human r 1 gene encodes for a protein with 223 amino acids with a mass of approximately 23.5 kDa. [9] The r 1 receptor is a unique protein without any homology to other known mammalian receptors or proteins. Although the r 1 receptor does not have any enzymatic sterol isomerase activity, the amino acid sequence shows a 30% homology to the fungal sterol D 8/7 -isomerase, an enzyme which catalyses a key step in the biosynthesis of ergosterol. [10,11] In 2002, Aydar et al. [12] postulated a r 1 receptor model with two transmembrane domains and both C-and N-terminus located intracellularly. Very recently, the crystal structure of the r 1 receptor together with various ligands was published. [13,14] The r 2 receptor has not been cloned yet and is even less explored than the r 1 receptor. Its molecular weight was estimated to be approximately 21.5 kDa. [15] Endogenous ligands for both r receptor subtypes have not been identified yet. Neuro(active)steroids such as pregnenolone and its sulfate, progesterone, dehydroepiandrosterone (DHEA) and the hallucinogenic N,Ndimethyltryptamine (DMT) are discussed as potential endogenous r ligands since they show moderate affinity towards r 1 receptors. [16][17][18] Ligands interacting with r 1 receptors are of particular interest due to their potential for the treatment of acute and chronic neurological disorders (e.g. epilepsy, schizophrenia, depression, Alzheimer's disease, Parkinson's disease), pain especially neuropathic pain besides alleviating the symptoms of alcohol and cocaine withdrawal. [7,[19][20][21][22] Also, a variety of human tumour cells expresses r 1 and/ or r 2 receptors in high density. [23] Labelled selective r 1 ligands might therefore be applicable for tumour imaging. As r 1 receptors are suggested to be involved in tumour proliferation, specific ligands may represent an interesting option for cancer therapy. [24] In 2010, Meunier and Hayashi [25] demonstrated that the knockdown of r 1 receptors in Chinese hamster ovarian cells (CHO) resulted in a potentiation of H 2 O 2 -induced apoptosis. For diabetes, it has also been reported that the disease is associated with oxidative stress. In addition, an imbalance in the endoplasmic reticulum (ER) of pancreatic b-cells was described. [26][27][28] The imbalance leads to an accumulation of misfolded and unfolded proteins in the ER lumen, often referred to as ER stress.
Since the r 1 receptor is considered to be an ER chaperone protein [29] and is involved in signalling between ER and mitochondria of cells, a linkage between the r 1 receptor and oxidative stress in pancreatic b-cells might exist.
Different methods are described to discriminate between r 1 receptor agonists and antagonists. It was shown that the r 1 receptor is involved in the regulation of intracellular Ca 2+ homoeostasis in retinal ganglion cells by influencing the activity of L-type voltage-regulated Ca 2+ channels. [30] r 1 agonists like (+)SKF-10,047 (alazocine) and opipramol inhibit the voltage-induced Ca 2+ influx, while antagonists (e.g. BD1047) are able to reverse this effect.
In the capsaicin assay, r 1 receptor antagonists inhibit mechanical hypersensitivity in mice induced by injection of capsaicin in the hind paw. The antiallodynic effect of r 1 antagonists is reversed by r 1 agonists. [31,32] With PC12 cells, which represent an in vitro model of neuronal differentiation, [33] it could be demonstrated that stimulation with r 1 receptor agonists potentiates neurite outgrowth and elongation induced by growth factors, such as nerve growth factor (NGF) or epidermal growth factor (EGF). [34] A large number of structurally diverse molecules interact with r 1 receptors. With the aim to find selective and potent r 1 receptor ligands and to gain insight into possible interactions within the binding site, a broad collection of substituted spirocyclic compounds with an annulated thiophene ring was synthesized in our laboratory and the r 1 affinity was evaluated. [35][36][37][38] Particular high r 1 affinity was found for the spirocyclic piperidines 1-6 displayed in Figure 1. While in many published articles the affinity of r ligands towards r and related receptors, such as the PCP binding site of the NMDA receptors, was studied, an in depth investigation of affinity of 1-6 to other targets and the activity in different functional assays to discriminate between agonistic and antagonistic activity or elucidate the therapeutic potential has not been performed.
Usually, particular features associated with the r 1 receptor were investigated with a particular compound or class. Herein, various pharmacological aspects of r 1 receptors were investigated with a coherent set of potent spirocyclic r 1 ligands 1-6 for the first time. First, the selectivity of the compounds was investigated in a variety of different binding assays, including opioid receptors, 5-HT receptors, serotonin transporter protein (SERT) and two different binding sites of the NMDA receptor. After determination of binding affinity, the ligands were studied in different cell-based assays as well as in an animal model of pain.
With the aim to classify the r ligands as agonists or antagonists, the functional activity was studied in the in vitro retinal ganglion cell assay and in the in vivo capsaicin assay.
Since r 1 receptors are also expressed on various tumour cell lines, the effect of the r 1 receptor ligands on the proliferation of A427 (non-small cell lung cancer), LCLC-103H (large cell lung cancer), 5637 (bladder cancer) and DAN-G (pancreatic cancer) tumour cell lines was studied.
Stimulated by the structural similarity between the r 1 receptor and fungal sterol D 8/7 -isomerase, the effects of the ligands on the biosynthesis of ergosterol in fungi and cholesterol in human cells were investigated.
Finally, the possible connection between the r 1 receptor and ER mediated oxidative stress prompted us to study the influence of one of our r 1 receptor ligands on the rate of apoptosis of murine pancreatic b-cells. If the r 1 receptor ligand is able to modulate the cell death rate of b-cells, the r 1 receptor could emerge as a new target in diabetes research.
The broad pharmacological evaluation of a coherent set of potent r 1 receptor ligands considering so many different aspects, including receptor affinity, selectivity, influence on ion channels, ergosterol and cholesterol biosynthesis, ER stress in mouse pancreatic b-cells, tumour cell growth and proliferation, and, finally, analgesic activity in vivo, represents the unique feature of this study. Moreover, the aspect of r 1 antagonistic activity was approached from various sides for the first time.

Materials and Methods
Receptor binding studies (1) Details on the assay to determine the affinity towards the ifenprodil binding site of GluN2B subunit containing NMDA receptors are given in Ref. [39] . (2) PCP binding assay see Refs [40,41] .

Inhibition of ergosterol and cholesterol biosynthesis
These assays were performed following the methods described by us previously Appendix S1. [11,[48][49][50][51] Ca 2+ influx assay, fluorescence measurements with fura-2-AM Concentrations c(Ca 2þ i ) were measured in single PC12 cells using the fluorescent indicator fura-2-AM in combination with a monochromator-based imaging system (FEI today Thermo Fisher Scientific, Langenselbold, Germany, SCR_008452) attached to a fluid immersion objective. Cells were loaded with 0.5 µM fura-2-AM and 0.01% Pluronic F-127 for 30 min at 37°C in a standard solution composed of 138 mM NaCl, 6 mM KCl, 1 mM MgCl 2 , 2 mM CaCl 2 , 5.5 mM glucose and 10 mM HEPES (adjusted to pH 7.4 with NaOH at 37°C). Cover slips were then washed in fresh buffer for 30 min and mounted in a perfusion chamber on the stage of the microscope (Olympus EX51WI, Hamburg, Germany). For measurements of c(Ca 2þ i ), cells were excited at 340 and 380 nm and emission was measured at 510 nm. After correction for background fluorescence, the fluorescence ratio F340/F380 of the emission was calculated. Fura-2-signals were calibrated according to the method of Grynkiewicz et al., [52] using a K D value of 224 nM. 10-20 cells were measured on slide and at least two replicates/independent experiment were conducted. At least five independent experiments were measured.

Pain behavioural studies
To evaluate the effect of drugs on mechanical allodynia induced by capsaicin, a previously described experimental procedure was used Appendix S1. [31] Animal care, husbandry and research were conducted in a sound-attenuated, air-regulated, experimental room of the Animal Facility of the Parc Cient ıfic de Barcelona (PCB), in accordance with institutional protocols approved by the local Committee of Animal Use and Care of our Institution (Animal Research Ethics Committee) and by Generalitat de Catalunya (Registry number: 9015-368574/ 2016) and international standards, including the Care and Use of Laboratory Animals Guidelines of the European Community (European Directive 2010/63/EU).

ER stress
Pancreatic islets were isolated from C57BL/6N mice by collagenase digestion. Single islet cells were cultured in RPMI 1614 medium with 10% fetal calf serum and 1% penicillin/ streptomycin (in 10 mM glucose). ER stress was induced by addition of 10 µM cyclopiazonic acid (CPA) for 24 h. Apoptosis was determined by TUNEL staining (in situ cell death detection kit, fluorescein, Roche Diagnostics, Mannheim, Germany) according to the manufacturer's protocol. For determination of the total cell number, nuclei were stained with bisbenzimide (Hoechst-33258). Fluorescence of apoptotic cells was excited at 480 nm. Nuclear staining was excited at 380 nm. A minimum of 200 cells for each condition was analysed per preparation to determine the rate of apoptosis.

Inhibition of human tumour cell proliferation
The cytotoxic effects of the test compounds were investigated in four cell lines as previously described Appendix S1. [53] Effects on human tumour cell line A427 Cultivation of A427 cells. For the preparation of the medium for the A427 cells 450 ml of RPMI 1640, 50 ml of FBS Superior and 5 ml of penicillin/streptomycin (10 000 µg/ ml) were warmed to 37°C in the water bath and mixed under sterile conditions in the biological safety cabinet. The content of one 2.0 ml cryovial (approx. 1-2 million cells) was added to 12 ml of the prepared medium in a 75 ml culture flask (VWR â Tissue Culture Flasks, VWR, Langenselbold, Germany). This flask was then stored in the incubator at 37°C and a water saturated atmosphere with 5% CO 2 . The unused medium was stored at 5°C. The medium in the flask was changed when the indicator changed its colour.
Subcultivation of A427 cells. At a confluence of 80-90%, the A427 cells of one flask were split over three flasks (VWR). The medium, PBS Dulbecco and trypsin were warmed to 37°C in the water bath. The medium was removed from the flask, and the cells were rinsed with 12 ml of PBS. After the removal of the PBS, 4 ml of trypsin were added. The cells were incubated in the incubator at 37°C for 5-10 min until approx. 80% of the cells were detached from the surface. Then, 15 ml of medium was added. The mixture was centrifuged at 800 9 g for 10 min at 4°C. The supernatant was discarded, and the pellet was resuspended in 12 ml of medium by manual vortexing. The medium with the cells was equally distributed over three culture flasks (VWR) and filled up with medium to a total volume of 12 ml. The cultivation conditions were the same as described above.
Live cell imaging using the IncuCyte â S3 live cell analysis system General remarks. The measurement of the cell confluence and the determination of the IC 50 values were performed with the IncuCyte â S3 Live Cell Analysis System.
Preparation of 96-well plates. For the preparation of the 96-well plates for the Live Cell Imager, A427 cells from the flask were detached and centrifuged. The pellet was resuspended in 20 ml of the medium using the vortexer. 1 ml was used to determine the number of cells, and the volume required for 8000-12 000 cells was calculated. This volume was then added to each well of a sterile 96-well plate (Tissue Culture Testplates 96F 92696, TPP, Trasadingen, Switzerland), and the well was filled up to a total volume of 150 µl with medium.
At a confluence of 30-40%, the test compounds were added. Dilutions of concentrations of 200, 100, 40, 20, 10 and 4 µM were prepared in medium with 2% DMSO. 50 µl of each dilution was added to the wells, resulting in a final assay concentration of 50, 25, 10, 5, 2.5 and 1 µM, respectively. Each concentration was measured in triplicates on each plate and on three different days.
Competitive confluence assay. The plates were prepared as described above with the exception that the cells were grown in 100 µl instead of 150 µl of medium. After the cell confluence reached 30-40%, the cells were treated with the respective test compounds and the agonist (+)-pentazocine (final concentration: 10 µM).
Measurement of cell confluence. The cell confluence was measured using the IncuCyte â S3. For the scanning with the company's own programme (IncuCyte S3 2017A), the HD phase imaging mode was chosen at a 10-fold magnification. Four pictures were taken from each well every 2 h until the growth curve of the cells reached a plateau (approx. 5 days). The mask for the analysis of the confluence in the pictures had a segmentation adjustment of 1.1, a cleanup adjustment size of À1 pixel and an area filter of minimum 200 µm 2 . After the analysis, a graph showing the development of the confluence over time was produced.
Calculation of IC 50 values. The cell confluence (in %) was exported to Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) to calculate the inhibition of the cell growth. The last time point from the measurement of each concentration was taken as the maximal confluence was then reached. The calculation was done as described in equation 1.
The inhibition value was then transferred to Origin 2018b (OriginLab, Northampton, MA, USA) to generate a graph showing the inhibition in dependence of the common logarithm of the concentration. First, a scatter diagram was generated. The fitting of the curve was done by non-linear, sigmoidal, dose/response regression. The non-linear regression was calculated with equation 2.
with y being the inhibition in %, A1 being the horizontal approach to the lowest value in inhibition (bottom (left) asymptote), A2 being the horizontal approach to the highest value in inhibition (top (right) asymptote), A1 < A2, log x0 being the centre of the curve, x being the common logarithm of the concentration and p being the hill slope. At an inhibition of 50%, the x value is equivalent to the log 10 (IC 50 ) which then had to be transformed to the IC 50 value. From the obtained values, the mean value and the corresponding standard deviation (SD) were calculated.
Cytotoxicity assay. The A427 cells were grown on a sterile 96-well plate (TPP) at 37°C until a confluence of approx. 90% was reached. The medium was then removed. Afterwards, the plate was blocked with 200 µl of medium containing 1% BSA and washed with 200 µl of medium. 150 µl of medium and 50 µl of the diluted test compounds were added to each well. As a positive control for a maximum cytotoxicity, 0.1% Triton X was used. The negative control was the medium without a test compound. This was incubated for 3 and 24 h at 37°C. The final concentrations of the test compounds in the assay were 50 and 5 µM, respectively.
After the incubation, 25 µl of supernatant and 50 µl of LDH-assay reagent (1 U/ml diaphorase, 1% sodium lactate, 0.1% NAD + , 0.08% BSA and 0.4% iodonitrotetrazolium chloride in 75 mM PBS) were transferred to a clear 96-well plate (Diagonal GmbH & Co. KG, M€ unster, Germany), previously blocked with BSA (see above). The plate was incubated at 37°C for 45 min. After the incubation, the absorption was measured at a wavelength of 490 nm in the plate reader. The obtained results were then put into relation to the controls.

Results and Discussion
Receptor selectivity r 1 affinity and selectivity over r 2 receptor The r affinities of 1-6 were determined in competition experiments with [ 3 H]-(+)-pentazocine and [ 3 H]DTG as radioligands. [42][43][44] Homogenates from guinea pig brain (r 1 ) and rat liver (r 2 ) were used as receptor material. All spirocyclic thienopyrans included in this study show a very high selectivity for the r 1 receptor (Table 1). Only for the thienofuranone 3, a moderate r 2 affinity (K i = 379 nM) was observed. The combination of the slightly lower r 1 affinity (K i = 16 nM) with moderate r 2 affinity resulted in a reduced selectivity of this compound. The increased r 2 affinity could be attributed to the unique carbonyl moiety serving as H-bond acceptor or the different structure of the furan ring together with the sp 2 hybridization in 3-position.

Selectivity over opioid and NMDA receptors
To elucidate the selectivity of the spirocyclic thiophenes 1-6, a comprehensive screening of the compounds at opioid receptors and the NMDA receptor was performed. Although the r 1 receptor is not structurally related to the NMDA receptor, many compounds binding to the GluN2B subunit of the NMDA receptor show high affinity to both r receptor subtypes, including the lead compound ifenprodil. [30] Therefore, the NMDA receptor was included in this study. The affinities to the µ-, dand j-opioid receptor subtypes (MOR, DOR, KOR) were determined with Table 1 r Receptor affinity and selectivity of spirocyclic thienopyrans, thienofuranone 3 and reference compounds (+)-pentazocine and haloperidol Compd.
homogenates from guinea pig (MOR and KOR) and rat brain (DOR). For the determination of affinity to the PCP binding site, a membrane preparation of pig brain cortex was used, [40,41] while membrane fragments from recombinant mouse fibroblasts stably expressing the GluN1a/ GluN2B subunits of the NMDA receptor were used in the assay for the ifenprodil binding site of the GluN2B subunit. [39] In Table 2, the radioligands employed in each binding assay and affinity data for reference compounds for each assay are listed. For all compounds, a very low affinity to MOR was observed. Surprisingly, a low-to-moderate affinity to the KOR and DOR was observed for the spirocyclic thiophenes 1, 2 and 6. The highest affinity was recorded for the thieno [3,4- The interaction of all investigated compounds with the PCP binding site of the NMDA receptor is also negligible. At a concentration of 1 µM of each test compound, only a small amount of the radioligand [ 3 H]MK-801 was displaced in the competition assay. A very interesting result was found for the spirocyclic compound 2 bearing the pcyanophenyl residue in the 1 0 -position of the thiophene ring. While this compound does not show any interaction with the opioid receptors and the PCP binding site, an unexpected affinity (K i = 91 nM) to the GluN2B binding site of the NMDA receptor was determined. In contrast, the structurally closely related arylated compound 5 did not display measurable affinity to the GluN2B binding site rendering 5 the compound with the highest selectivity for the r 1 receptor in this series.
Selectivity over 5-HT 1A , a 1A , a 2A receptors and 5-HT transporter In order to gain additional information regarding the selectivity over other pharmacological targets, a screening of the spirocyclic thiophenes 1, 3, 4 and 6 was performed. An affinity to 5-HT 1A receptors, the serotonin transporter (SERT) and the human a 1A adrenergic receptor were not observed for any of the compounds. For the non-arylated thiophenes 1 and 4, a low affinity to the human a 2A adrenergic receptor was observed. At a concentration of 1 µM in the binding assay, an inhibition of 52% and 46% was recorded, respectively, indicating IC 50 values of approximately 1 µM.
In conclusion, high-affinity spirocyclic r 1 ligands 1-6 show high selectivity for the r 1 receptor over the investigated related receptor systems.

Inhibition of the biosynthesis of ergosterol and cholesterol
To analyse potential inhibitory effects on the ergosterol and cholesterol biosynthesis, the non-arylated spirocyclic thiophene derivatives 1, 3, 4 and 6 were studied in cellular assays. [11,[48][49][50] In the ergosterol biosynthesis assay, three different fungi species (Candida glabrata, Saccharomyces cerevisiae and Yarrowia lipolytica) were incubated with the spirocyclic thiophenes. Cell lysis and extraction yielded a crude fraction of fungal sterols, which were transformed into the corresponding trimethylsilyl ethers and analysed by GLC/MS. In neither of the three species, inhibitory effects of the spirocyclic thiophenes on the post-squalene part of ergosterol biosynthesis were found, including the fungal sterol D 8/7 -isomerase.
The tritiated radioligands used in the assays are mentioned, respectively. a In case of high affinity (K i < 100 nM), the experiments were performed as triplicate (n = 3) and SEM values are given. b Values in % reflect the inhibition of the radioligand binding at a test compound concentration of 1 lM.
In the cholesterol biosynthesis inhibition assay, human promyelocytic leukaemia cells (HL60) were incubated with the test compounds. Isolation and silylation of the sterol fraction were performed as described above, followed by GLC/MS analysis. For the thienopyrans 1, 4 and 6, analysis of the sterol pattern clearly indicated an inhibitory effect on the human sterol C14-reductase at high concentrations (50 µM), evident from an accumulation of the intermediate 4,4-dimethyl-5a-cholesta-8,14-dien-3b-ol in the biological samples. For the thienofuranone 3, accumulation of zymostenol (cholest-8-en-3b-ol) clearly indicated an inhibition of the mammalian sterol D 8/7 -isomerase. These results were unexpected since the related human enzymes do not even show a structural similarity to the r 1 receptor. Similar results were obtained by us before for aminomethyl-substituted spiroketals. [51] It is noteworthy that the inhibition of the mammalian sterol C14-reductase was detected for all three isomeric thienopyrans, regardless of the position of the S atom in the thiophene ring. However, since the sterol C14-reductase is located in the ER membranes, a modulatory effect of the r 1 receptor acting as a chaperone might provide an explanation for the inhibitory effect of the spirocyclic r 1 ligands on this enzyme.

Interaction with voltage-gated Ca 2+ channels
With the aim to study the functional activity of the spirocyclic r 1 ligands, one non-arylated (1) and one arylated compound (2) of the set were chosen to assess the influence on voltage dependent intracellular Ca 2+ levels in retinal ganglion cells. [52] In short, the cells were preincubated with the Ca 2+ sensitive fluorescent dye fura-2-AM and the test compound. Subsequently, the cells were depolarized with KCl to initiate influx of Ca 2+ ions. Both test compounds were not able to inhibit the KClinduced Ca 2+ influx, suggesting that they are not acting as r 1 agonists. To study a possible antagonistic effect, the test compounds were co-incubated with the r 1 agonist opipramol. The inhibitory effect of opipramol was reversed by the p-cyanophenyl substituted thiophene 2 in a dose-dependent manner (Figure 2). Therefore, 2 can be classified as an r 1 antagonist in this assay system. For the non-arylated thiophene 1, an increase of the KCl-induced Ca 2+ influx was not observed.

Analgesic activity in the capsaicin assay
The analgesic activity of the spirocyclic thiophenes 1, 4, 5 and 6 was investigated in a capsaicin assay. [31] In this assay, mice were injected with 1 µg capsaicin in the right hind paw to induce hypersensitivity. After 15 min, a punctuate mechanical stimulation was applied with a nonflexible filament to the hind paw. When a paw withdrawal response of the animal occurred, the stimulation was automatically terminated and the response latency time recorded. The test compounds were injected intraperitoneally 5 min before the capsaicin treatment. For compounds 1, 4 and 5, a significant analgesic effect with a prolonged response latency time was observed (Figure 3). In addition to the pharmacological effect, the assay also demonstrates that the compounds are acting as r 1 antagonists, which are in good accordance with the results of the retinal ganglion Ca 2+ influx assay. For the thieno[3,2-c]pyran 6, only weak analgesic activity was found. However, in contrast to the other test compounds 6 was administered 30 min before the capsaicin treatment so a metabolic degradation might explain the reduced activity.

Inhibition of ER stress in mouse pancreatic b-cells
The effect of the spirocyclic piperidine 1 on murine pancreatic b-cells was investigated by an apoptosis assay. Briefly, the cells were incubated with the test compound in standard medium before ER stress was induced by treatment with a high concentration of cyclopiazonic acid (CPA, 10 µM). CPA is an inhibitor of SERCA, a Ca 2+ ATPase that transfers Ca 2+ from the cytosol of the cell to the lumen of the ER. The Ca 2+ depletion of the ER causes an increased rate of apoptosis. The apoptotic rate was determined by counting the number of total and apoptotic cells, respectively. [54] Islet cells were preincubated overnight with the r 1 ligand before addition of CPA for another 24 h. Compound 1 slightly increased apoptotic cell death (TUNEL-positive cells, control: 0.16 AE 0.08%, + compound 1, 1 µM: 1.8 AE 0.2%, n = 3 independent preparations, P < 0.05). An effect of the r 1 antagonist 1 on the CPA-induced apoptosis was not observed (TUNEL-positive cells, CPA 10 µM: 3.6 AE 0.7%, + compound 1, 1 µM: 3.02 AE 0.82%, n = 3, n.s.). Obviously, the test compound neither increased nor prevented the release of Ca 2+ ions from the ER into the cytoplasm under stress conditions.

Inhibition of the growth of tumour cells
The influence of the r 1 ligands on the proliferation of the tumour cell lines A427 (non-small-cell lung cancer), LCLC-103H (large cell lung cancer), 5637 (bladder cancer) and DAN-G (pancreatic cancer) was determined in vitro by a crystal violet staining assay. (+)-Pentazocine and haloperidol were included as reference compounds. The results are summarized in Table 3.
The cytotoxic effects of thienopyrans 1, 2 and 5 on the non-small-cell lung cancer line A427 and the bladder cancer cell line 5637 are highly remarkable. These compounds inhibit the proliferation of both cell lines with IC 50 values in the range between 2.6 and 9.1 µM. In the case of A427 cells, the same effect was observed for the r 1 antagonist haloperidol, while the r 1 agonist (+)-pentazocine did not show growth inhibiting effects. This observation is very interesting as the thienopyrans 1, 2 and 5 have also been classified as antagonists in the functional assays described above. Similar results have been reported for other classes of r ligands on this cell line.
For the 5637 cell line, it should be noted that cytotoxic effects were found for both (+)-pentazocine and haloperidol which indicates that the antiproliferative activity might not be related to the r 1 receptor in this case.
A cytotoxic activity of lactone 3 and thieno[2,3-c]pyran 4 was abundant on all investigated cell lines. At least for 4, this observation was surprising as not only the chemical structure but also the binding data and functional activity are very similar to the thieno[3,4-c]pyran 1.
The thienopyran 2 bearing a p-cyanophenyl moiety shows a rather unselective inhibition of tumour cell growth with very similar IC 50 values over all cell lines, indicating rather unspecific cytotoxicity than a r receptor-mediated effect.
Analysis of the effects of spirocyclic thiophenes 2 and 5 on proliferation and morphology of the human tumour cell line A427 In the screening of different human tumour cell lines, the spirocyclic thiophenes 2 and 5 showed strong inhibition of the proliferation of the A427 tumour cell line. Therefore, these compounds were selected to determine the effects on this cell line in more detail. The A427 cells were incubated in the Live Cell Imager IncuCyte â S3, which allows permanent observation of the morphology and behaviour of cells over a long period without interruption of the incubation process. Both spirocyclic compounds led to inhibition of the proliferation of the A427 cell line. The cyanophenyl substituted thiophene 2 possessing higher r 1 affinity revealed 3-fold stronger inhibition of cell proliferation than the phenyl substituted spirocyclic thiophene 5 (Table 4). These data obtained by determination of the confluence in the live cell imager correlate well with those from the crystal violet staining assay shown in Table 3.  Whereas (+)-pentazocine, which is regarded as r 1 agonist did not influence the proliferation of the A427 tumour cells, the r 1 antagonist haloperidol showed considerable inhibition of proliferation (IC 50 = 16 µM). (Table 4) Therefore, it was investigated whether the tumour cell proliferation inhibition of the spirocyclic thiophenes 2 and 5 could be antagonized by (+)-pentazocine (10 µM). The effect of the more potent cyanophenyl derivative 2 was reduced in the presence of (+)-pentazocine, but the effect of the less potent phenyl derivative 5 could not be inhibited.
In addition to recording IC 50 values, the morphology of the A427 cells upon incubation under various conditions was observed and analysed (Figure 4). At first the A427 cells appear vital (Figure 4a), but 48 and 96 h after addition of compound 2 (10 µM), cells were detached from the surface and cell fragments could be observed (Figure 4b). This cytotoxic effect of 2 could be antagonized at least partially by addition of (+)-pentazocine. Several cells were still attached to the surface, and the amount of cell fragments was reduced (Figure 4c). At a concentration of 50 µM of the spirocyclic thiophene 2, the rescue effect of (+)-pentazocine was lower (Figure 4d). The less potent phenyl substituted thiophene 5 shows qualitatively the same effects as the cyanophenyl substituted derivative 2, but to a lesser extent.
In order to analyse the cytotoxic effect of the spirocyclic thiophenes 2 and 5, the lactate dehydrogenase (LDH) assay was performed. Damage or destruction of cells leads to release of LDH, which can be measured by UV-spectroscopic determination of a coloured formazane dye. At a concentration of 50 µM, the cyanophenyl derivative 2 demonstrated strong cytotoxic activity (65%) in this assay, whereas the cytotoxicity of the phenyl substituted derivative 5 is considerably lower (28%). This result nicely correlates with the unspecific inhibition of cell growth of the four tumour cell lines presented in Table 3.