*2.3. Presence of (S)-(*−*)-[18F]Fluspidine Binding Sites in Human GBM Tissue*

To initially assess the suitability of sig1R as specific target for molecular characterisation of human GBM, we performed in vitro autoradiography with (*S*)-(−)-[18F]fluspidine using cryosections of tissue samples obtained from 3 patients diagnosed with Glioblastoma multiforme IV. Total and nonspecific binding of the PET ligand was determined by incubation with only (*S*)-(−)-[18F]fluspidine or with co-administration of a high concentration of haloperidol to block the sig1R, followed by histological staining of the respective cryosections. As shown in Figure 6, the autoradiographic images indicate a heterogeneous pattern of binding sites of (*S*)-(−)-[18F]fluspidine in all three GBM samples with the highest density in regions histologically characterised by a high density of cells which we assume might be related to the highly proliferating tumour cells. Accordingly, although it was not possible within this preliminary study to confirm by immunohistochemistry the distribution pattern of sig1R in the cryosections or to identify the type of cells possessing high specific binding of (*S*)-(−)-[18F]fluspidine, these preliminary data motivate us to design a complementary study on the investigation of sig1R protein in a larger number of GBM samples by means of specific radioligands and antibodies.

**Figure 6.** Sig1R autoradiography with the sig1R-specific PET ligand (*S*)-(−)-[18F]fluspidine in human GBM in vitro. Binding of (*S*)-(−)-[18F]fluspidine at 4.5 nM in cryosections (12 <sup>μ</sup>m) of tumour tissue obtained from three patients (**A**,**B**,**C**) demonstrated heterogeneous distribution throughout the slices. By co-incubation with 1 μM haloperidol (**D,E,F**), a substantial reduction in activity accumulation was obtained. Histochemical analysis of corresponding sections was performed by Nissl staining (**G**,**H**,**I**). Analysis of one sample at higher magnification (red square in **C**) demonstrated correlation of the activity accumulation (**K**) with highly cell dense regions (H&E staining: (**L**); Nissl staining: (**M**). Length of the biopsies samples ~1 cm.

#### **3. Discussion**

In this study, we evaluated the availability of sig1R in an orthotopic mouse model of human GBM. High expression of sig1R in different cancer cell lines derived from prostate, breast, colon, melanoma, small and non-small cell lung cancer, brain tumours including GBM, neuroblastoma, and meningioma have already been reported [18,19,26,27,43–48]. The involvement of sig1R in many but selective protein interactions, the antiproliferative effect of putative antagonists, as well as their nonpleiotropic effects make sig1R a potent drug target prone to overcome adaptative drug resistance alone or in combination with other drugs [49,50]. It is also known that the upregulation of sig1R on both mRNA and protein level in the same cancer subtype differs from one cell line to another and from one patient to another, probably reflecting a context-dependent expression of sig1R [51–53]. Consequently, an improved understanding of how such patterns evolve and change during pathogenesis, by the use of noninvasive PET imaging, would promote the development of sig1R-based therapies. In this context, we chose to evaluate by PET the suitability of the clinically approved imaging agent (*S*)-(−)-[18F]fluspidine for the analysis of the expression of sig1R in GBM.

We first investigated in vitro the level of expression of sig1R protein in U87-MG cells, a human GBM cell line widely applied for orthotopic brain cancer mouse models. As consistently reported, sig1R is located at the endoplasmic reticulum-mitochondria interface and redistributes ligand-mediated and under conditions of cellular stress dynamically to the plasma membrane and the nucleus envelope [54,55]. The examination of sig1R expression in U87-MG cells by using (+)-[3H]pentazocine, a selective sig1R ligand widely applied in radioligand binding assays [56,57], demonstrated high affinity binding towards a single binding site expressed in U87-MG cells grown in 2D cell culture. The density of sig1R in this cell line, *B*max = 129 fmol/mg protein, is in the range of values determined by

(+)-[3H]pentazocine in other cancer cell lines such as 42 fmol/mg protein in the C6 murine glioblastoma cells, 76.5 fmol/mg protein in the NB41A3 neuroblastoma cells or 1115 fmol/mg protein in the U-138-MG cells [19]. We identified the U87-MG cells as suitable for the orthotopic GBM model applied in this study.

As the microenvironment is known to influence gene expression, e.g., the hypoxia-stimulated HIF-1α expression in glioma or the culture mode (2D vs. 3D)-dependent differential gene expression of colorectal cell lines [58–61], we further investigated the expression of sig1R in the intracerebral U87-MG tumour. The comparable immunofluorescence staining on 2D-cultivated U87-MG cells and on the orthotopically grown U87-MG tumour along with the similarity of the KD values of the PET tracer (S)-(−)-[18F]fluspidine, indicate the conservation of sig1R expression and conformation over the translation from in vitro culture to in vivo implantation. Furthermore, the cytoplasmic localization of the sig1R fluorescence signal which was observed in vitro and in the explanted brain tumours matches with the cellular localization of the receptor found in rat astrocytes and mouse neurons [62,63].

Subsequently, we quantified the number of sig1R expressed in the U87-MG tumours implanted in the right striatum as well as in the internal control region, the left striatum. The equivalent *K*<sup>D</sup> values obtained for (*S*)-(−)-[18F]fluspidine in both regions indicate that the PET radiotracer binds to the same target, i.e., the sig1R, in both compartments. The analysis of the binding parameter *B*max excludes conformational differences between sig1R in cancer and normal cells, as discussed by Kim et al., as a possible reason for the higher accumulation of sig1R-targeting radioligands in tumour tissue, but clearly indicates an about 2-fold higher density of sig1R in the U87-MG tumour in comparison to the healthy brain [51]. Thus, the herein exploited orthotopic U87-MG GBM mouse model is appropriate for the following imaging studies. As an add-on to the extensive and validated data on the selectivity of the clinically applied PET radioligand (*S*)-(−)-[18F]fluspidine obtained mainly by pharmacological intervention studies [64,65], we made use of access to a sig1R-knockout mouse model to measure the actual contribution of the off-target binding of the radiotracer to the uptake of activity in the brain in imaging studies in mouse [66]. In accordance with the fast washout kinetics observed in the knockout model, we supposed only a weak background signal in imaging studies with (*S*)-(−)-[18F]fluspidine in the orthotopic brain cancer model and no relevant interaction with off-target binding sites in vivo.

Eventually, the results of the fundamental characterisation of the components of the experimental setting, i.e., the mouse model and the PET radioligand, with respect to availability of and selectivity to sig1R, prompted us to proceed with dynamic PET studies in the orthotopic U87-MG glioblastoma mouse model. Only few studies have explored the use of PET radiotracers for sig1R imaging of tumours, and even less have addressed brain tumours in particular [36,39,67–70] such as the investigation of sig1R in an ectopic glioma rat model as well as in spontaneous pituitary tumours in rats using [11C]SA4503 by the group of van Waarde [39,69,70]. To the best of our knowledge, we are the first exploring the sig1R availability of human glioblastoma in an orthotopic tumour mouse model. The in vivo imaging studies revealed a tumour-to-background ratio (TBR) of only slightly higher than 1, detectable from the late PET images. Even though we observed a continuous washout of activity from the tumour, this process was slower than in contralateral tissue. Accordingly, the activity concentration in the tumour surpassed that in the contralateral striatum over time.

Despite this, the TBR value determined in our study is in fact notably lower than the values reported for the [11C]SA4503 PET studies mentioned above. However, we assume that this discrepancy is related mainly to the characteristics of the background region, in particular the physiological expression of sig1R in the different grafting sites. An ectopic tumour obtained by e.g., implantation of C6 glioma in the shoulder in the soft tissue [71], close to the muscle, benefit of an ideal background tissue with low expression of sig1R [20], leading to a TBR values > 4. Such values are not comparable to orthotopically transplanted brain tumours due to the comparatively high expression of sig1R in the surrounding nondiseased brain, as indicated by e.g., in the herein performed PET studies with (*S*)-(−)-[18F]fluspidine in healthy mice [72,73].

The reasons for the discrepant results obtained in the present paper regarding the in vitro and in vivo quantification of sig1R in the U87-MG tumours are not clear at the moment. We assume, that factors such as microenvironment, vascularisation, or interstitial fluid pressure affect the binding parameters of (*S*)-(−)-[18F]fluspidine in vivo. The U87-MG tumour is known to be highly vascularised and presenting necrotic foci [62], suggesting a first uptake in the vascularised periphery and a later accumulation by diffusion in the core of the tumour tissue. However, a systematic investigation of these processes was beyond the scope of this study. Notwithstanding this limitation, a detailed investigation of the PET images of the intracranial U87-MG tumours revealed that the heterogeneous pattern of activity accumulation discovered already in vitro could be detectable by the in vivo imaging approach as well. Interestingly, a similar distribution of [11C]SA4503 in the tumour outer rim was reported in the already mentioned PET study of the ectopic C6 glioma model as well as in a patient with non-small cell lung cancer in the tumour tissue [74,75]; noteworthy is the discrepancy between the distribution patterns of [11C]SA4503 and [18F]FDG [69,74]. Since [18F]FDG PET images may be misleading due to an increased glucose metabolism in noncancerous but inflammatory tissues, the authors suggested the use of sig1R PET imaging to discriminate between tumour and inflammation [69].

A final aspect addressed in this study on the suitability of PET imaging of sig1R in glioblastoma was the investigation of the expression of sig1R in human GBM tissue. The accordingly performed receptor autoradiography with (*S*)-(−)-[18F]fluspidine on cryosections of human glioblastoma obtained from three patients consistently showed a heterogeneous distribution of binding sites of the sig1R-targeting radioligand with high-density binding in cell-dense regions as suggested by the subsequent histological analysis. However, although sig1R appears to play a role in proliferation, this preliminary examination does not allow to speculate about a correlation between receptor expression and tumour proliferation but nevertheless suggests to design a respective large-scale study [28].

#### **4. Materials and Methods**

All experimental work including animals has been conducted in accordance with the national legislation on the use of animals for research (Tierschutzgesetz (TierSchG), Tierschutz-Versuchstierverordnung (TierSchVersV)) and has been approved by the responsible research ethics committee (TVV 30/17; TVV 18/18 Landesdirektion Sachsen).

#### *4.1. Radiochemistry*

Enantiomerically pure (*S*)-(−)-[18F]fluspidine was prepared on a TRACERlab FXN synthesizer (GE Healthcare, Waukesha, WI, USA) as described in previous publications [65]. The radiochemical purity of (S)-(−)-[18F]fluspidine was >99%, and the molar activity (Am) at the end of the synthesis (EOS) was 89–180 GBq/μmol (*n* = 2).

#### *4.2. Cell Culture*

U87-MG cells (obtained from Jens Pietzsch/Birgit Belter, Department Radiopharmaceutical and Chemical Biology, Helmholtz-Zentrum Dresden-Rossendorf, Rossendorf, Germany) and human hsig1R-transfected Human Embryonic Kidney (HEK) cells (obtained from Olivier Soriani, Institut de Biologie Valrose—University Côte d'Azur, Sophia Antipolis, France) were maintained in monolayer culture (37 ◦C, 5% CO2, 95% O2) in Dulbecco's Modified Eagle Medium (DMEM, Gibco, Invitrogen, Dun Laoghaire, Ireland) supplemented with 10% heat inactivated fetal bovine serum (Gibco, Invitrogen, Dun Laoghaire, Ireland), 5% penicillin and streptomycin, 1.25% sodium pyruvate, 1% l-glutamine (Gibco, Invitrogen, Ireland) and 1 μg/mL puromycin (Gibco, Invitrogen, Dun Laoghaire, Ireland) only for the transfected cells.

#### *4.3. In Vivo Competitive Radioligand Binding Assay*

Cell membrane homogenates of U87-MG cells were obtained by gentle scraping the cells grown to confluency in one 175 cm<sup>2</sup> flask, followed by sedimentation of the cells suspended in cell culture medium by centrifugation at 800 rpm for 3 min at room temperature, re-suspension of the cells in 1 mL 50 mM TRIS-HCl, pH 7.4/4 ◦C and incubation on ice for 20 min, centrifugation of the suspension at 15,000 rpm for 15 min at 4 ◦C, and finally re-suspension of the pellet in 200 μL 50 mM TRIS-HCl, pH 7.4/4 ◦C and storage at −25 ◦C. The radioligand binding assay was performed by incubating the U87-MG cell membrane homogenate (226 μg protein/mL) with the Sig1R agonist (+)-[3H] pentazocine (working concentration = 3.25 nM; Am = 995 GBq/mmol; PerkinElmer LAS GmbH, Rodgau, Germany) in incubation buffer (50 mM TRIS-HCl, pH 7.4, 120 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2) without (total binding, TB; *n* = 3) or with co-incubation of 1 μM haloperidol (nonspecific binding, NB; *n* = 3) at room temperature for 60 min. The incubation was terminated by filtration via a Whatman® glass microfibre filter (Grade GF/B, pre-incubated in freshly prepared polyethyleneimine (3%) at room temperature for 90 min), followed by quadruplicate washing with 50 mM TRIS-HCl, pH 7.4/4 ◦C using a semi-automated cell harvester (48-samples; Brandel, Gaithersburg, MD, USA). Filter-bound radioactivity was detected in terms of DPM/vial by liquid scintillation counting (Beckman LS 6500; Beckman Coulter Inc., Fullerton, CA, USA) of the isolated filters immersed for two hours in liquid scintillation cocktail (Ultima Gold; PerkinElmer LAS GmbH, Rodgau, Germany). Specific binding (SB) was calculated by SB (DPM/vial) = TB (DPM/vial) − NB (DPM/vial). The Bmax and the KD values were estimated by a nonlinear regression model (equation: one-site binding (hyperbola)) using GraphPad Prism, Version 4.1 (GraphPad Inc., La Jolla, CA, USA).

#### *4.4. In Vitro Autoradiography on Human Glioblastoma Tissue*

Cryosections of brain tumour tissue from three patients (Glioblastoma multiforme IV) were obtained using a microtome (MICROM HM560, Fisher Scientific GmbH, Schwerte, Germany), mounted on microscopy slides (SuperFrost, Thermo Scientific Menzel, Fisher Scientific GmbH, Schwerte, Germany), dried for ~2 h at room temperature, and stored at −25 ◦C until the autoradiography study. For the experiment, the slides were taken out from the freezer, the cryosections dried under a stream of cold air, and pre-incubated with incubation buffer (50 mM TRIS-HCl, pH 7.4, 120 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2) at room temperature for 15 min. The pre-incubation solution was decanted, the slices dried again under a stream of cold air, and covered afterwards with the incubation solution ((*S*)-(−)-[18F]fluspidine, 197 kBq/mL incubation buffer <sup>=</sup> 4.5 nM at the time of incubation, without (total binding) or with co-incubation with 1 μM haloperidol to assess nonspecific binding). Incubation at room temperature was terminated after 60 min, the slides were washed two times in 50 mM TRIS-HCl, pH 7.4 at 4 ◦C, on ice for two minutes each followed by dipping in ice-cold demineralized water for 5 s and rapid drying under a stream of cold air. Afterwards, the slides were exposed to a phosphor imager plate (BAS-IP TR 2025, FujiFilm Corporation, Tokyo, Japan) along with standards obtained by pipetting and drying 1 μL of each concentration of a serial dilution of the radioligand solution on to a microscopic slide. The exposed phosphor-imaging plates were scanned using a high resolution scanner (HD-CR 35 Bio; Dürr NDT GmbH & Co. KG, Bietigheim-Bissingen, Germany) at a laser spot size of 12.5 μm (pixel size: 12.5 μm2) followed by two-dimensional analysis of the digitized images (AIDA 4.27; Elysia-raytest GmbH, Straubenhardt, Germany). The tracer distribution in the autoradiographic images obtained for total and nonspecific binding was compared by visual inspection and correlated with the histochemical staining (Nissl- and Hematoxiline-eosin staining) of the corresponding tissue sections.

#### *4.5. In Vitro Autoradiography on Mice Brain-Bearing Glioblastoma*

Cryosections of brains obtained from female athymic nude mice (Rj:NMRI-Foxn1 nu/nu) (10–12 weeks old, 25–38 g), were obtained as described above. The same protocol as in Section 4.3 was used. The incubation step was performed with 0.1–0.2 MBq/mL (*S*)-(−)-[18F]fluspidine in buffer for 60 min at room temperature. Nonspecific binding was determined in the presence of 10 μM of SA4503 (Tocris, Bio-Techne GmbH, Wiesbaden-Nordenstadt, Germany) or 100 μM to 10 nM of (*S*)-(−)-fluspidine, respectively. Developed autoradiographs were analysed in a phosphor imager (HD-CR 35; Dürr NDT GmbH & Co. KG, Bietigheim-Bissingen, Germany). The quantification was performed by using 2D-densitometric analysis (AIDA 2.31 software; raytest Isotopenmessgeräte GmbH, Straubenhardt, Germany). The Bmax and the KD values were estimated by a linear regression model

(equation: one-site binding (hyperbola)) using GraphPad Prism, Version 4.1 (GraphPad Inc., La Jolla, CA, USA).

#### *4.6. Immunohistochemistry*

Tissues were cryopreserved by incubation in 2-Methylbutane at −25 ◦C (Merck, Germany). The brains were cut into coronal sections 10 μm thickness with cryostat (MICROM HM560, Fisher Scientific GmbH, Schwerte, Germany) and kept at −25 ◦C. Immunostaining was performed after fixation in PFA 4% for 20 min at 4 ◦C of the slides. Detection of the sig1R protein was performed by overnight incubation at 4 ◦C of the slides with the primary mouse monoclonal antibody (1:500 in blocking buffer 5% normal goat serum, B-5: sc-137075, Santa Cruz Biotechnology, Inc., Dallas, TX, USA). After washing with a solution of 1% BSA in PBS, the slides were incubated for 1 h at room temperature with the secondary polyclonal goat anti-mouse antibody (1:200 in dilution buffer 1% BSA, Alexa Fluor ® 488; ab150117, Abcam, Berlin, Germany). A Hoechst counterstaining, 10 min at room temperature, was performed to visualize the nuclei of the cells (1:1000 in PBS, Hoechst 33258, Life Technologies, Carlsbad, Ca, USA). After a step of washing and drying, slides were cover up with mounting medium. (Aquapolymount, Polysciences Europe GmbH, Hirschberg an der Bergstrasse, Germany). Visualization of the slides was performed by fluorescence microscopy (Leica, DMi8, software Leica LASX, Leica Mikrosysteme Vertrieb GmbH, Wetzlar, Germany).

#### *4.7. Animals and Orthotopic Brain Tumour Model*

Female athymic nude mice (Rj:NMRI-Foxn1 nu/nu) were chosen for this study (Janvier labs, France). The mice were used for tumour implantation at the age of 8 weeks (26–30 g). During microsurgery mice were anesthetized with a mixture of air and isoflurane concentrate (1.5–2% depending on the breathing) under sterile conditions. The mice were placed into a Stoelting stereotactic frame (just for mouseTM, Stoelting Europe, Dublin, Ireland). A midline incision was done and a burr hole was drilled 0.5 mm anterior and 2.5 mm lateral to the bregma. 5 <sup>×</sup> 104 U87-MG cells were suspended in 1 <sup>μ</sup>L Hank's Buffered Salt Solution (HBSS, 1X) and were injected 3.0 mm into the brain parenchyma with a flow of 0.1 μL/min using a 10 μL Hamilton syringe. After injection, the burr hole was filled with bonewax (Ethicon, US, LLC), the scalp incision sutured (Vicryl 6.0, Ethicon, US, LLC) and the surface antiseptically cleaned. Animal sacrifice was performed by induction of anesthesia with a mixture of air and isoflurane concentrate followed by cervical dislocation.

#### *4.8. Small Animal PET*/*MR Imaging*

For the time of the experiments, female CD-1 mice (*n* = 3; age: 10 weeks; weight: 30–35 g) or *nude* mice (*n* = 3; age: 10 weeks; weight: 25–30 g) (Janvier Labs, Le Genest-Saint-Isle, France) and one CD-1 sig1R-knockout mouse (*n* = 1; age: 10 weeks; weight: 27 g) (Envigo RMS, SARL, Bresso, Italy) were kept in a dedicated climatic chamber with free access to water and food under a 12:12h dark:light cycle at a constant temperature of 24/26 ◦C. The animals were anaesthetized (Anaesthesia Unit U-410, AgnTho's, Lidingö, Sweden) with isoflurane (1.8%, 0.35 L/min) delivered in a 60% oxygen/40% air mixture (Gas Blender 100 Series, MCQ instruments, Rome, Italy) and maintained at 37 ◦C with a thermal bed system. (*S*)-(−)-[18F]fluspidine was injected into the lateral tail vein (control group: 3.5 <sup>±</sup> 1.9 MBq, Am: 94 ± 7 GBq/μmol EOS; *h*sig1R-knockout mouse: 1.9 MBq; Am: 89 GBq/μmol EOS; tumour group: 5.7 <sup>±</sup> 3.7 MBq; Am: 119 <sup>±</sup> 41 GBq/μmol EOS) followed by a 60 min PET/MR scan (nanoScan®, Mediso, Hungary). Each PET image was corrected for random coincidences, dead time, scatter and attenuation (AC), based on a whole body (WB) MR scan. The list mode data were sorted into sonograms using a framing scheme of 12 × 10 s, 6 × 30 s, 5 × 300 s, 9 × 600 s. The reconstruction parameters for the list mode data are: 3D-ordered subset expectation maximization (OSEM), 4 iterations, 6 subsets, energy window: 400–600 keV, coincidence mode: 1–5, ring difference: 81. The mice were positioned prone in a special mouse bed (heated up to 37 ◦C), with the head fixed to a mouth piece for the anesthetic gas supply with isoflurane in 40% air and 60% oxygen. The animal head was positioned in the center

of the field of view in order to benefit from the highest spatial resolution possible (spatial resolution center of the FOV: 900 μm). A dynamic PET scan of a duration of 60 min was performed followed by a T2 weighted sequence (Fast Spin Echo, TR/TE: 4377/88.5 ms, NEX: 4, FOV: 70 × 70 mm, matrix: 256 × 256, SI: 0.9 mm) and a T1 weighted sequence (Gradient Echo, TR/TE: 15/2.59 ms, NEX: 4, FOV: 60 × 60 mm, matrix: 160 × 160, slice thickness: 0.5 mm) for anatomical orientation and AC correction respectively. Image registration and evaluation of the region of interest (ROI) was done with PMOD (PMOD Technologies LLC, v. 3.9, Zurich, Switzerland). The respective brain regions were identified using the T2 weighted sequence and the tumour area and the contralateral area were delineated manually. The hypersignal due to the tumour in T2 weighted images was manually segmented and described as "tumour ROI", and due to the compression of the contralateral side a fixed circled shape ROI was used to delineate the striatum avoiding nearby structure (cortex, ventricles, hypothalamus). The image-derived input function (IDIF) was extracted from a voxel of interest (VOI) segmented on the inferior vena cava (IVC). The IVC was identified using the first time frames showing the first passage of (*S*)-(−)-[18F]fluspidine bolus. An automatic algorithm from PMOD was used to identify the IVC signal avoiding heart and kidney area [76]. The activity data are expressed as mean standardized uptake value (SUV) of the overall ROI or as SUV ratio of the striatum ROI over the IDIF (SUVR). Data are presented as mean ± standard deviation (SD). Microsoft Excel was used to perform statistical tests. A parametric student *t*-test preceded by a Fischer test for variance were used to compare the groups with *p* < 0.05.

#### **5. Conclusions**

To conclude, we showed for the first time in an orthotopic GBM model, the U87-MG mouse model of glioblastoma and the suitability of a sig1R-targeting PET radioligand, (*S*)-(−)-[18F]fluspidine, to investigate the tumour-specific expression pattern of sig1R by in vivo imaging. Whether the inferior outcome in vivo in comparison to in vitro is caused by the physiological expression of sig1R in the healthy brain or by certain pathophysiological characteristics of the orthotopic mouse model of GBM, remains to be elucidated. Nevertheless, this first evaluation of the sig1R availability in an orthotopic in vivo model of brain tumour contributes to a better understanding of this model and suggests an expression of sig1R in the tumour periphery as found in other studies, which may be related to proliferation and invasiveness. In conclusion, the data obtained in the U87-MG mouse model of GBM along with the detection of sig1R in human GBM tissue for the first time by a PET radioligand, indicate not only the relevance of this target but also the suitability of (*S*)-(−)-[18F]fluspidine for sig1R-targeted cancer research and drug development.

**Author Contributions:** Conceptualization, P.B., W.D.-C., G.S.; methodology, M.T., M.K., W.D.-C., S.F., F.-A.L., M.P., O.S., T.A.J., G.S.; formal analysis, M.T., W.D.-C.; investigation, M.T., M.K., W.D.-C.; tracer synthesis and analysis, O.S., M.P.; writing—original draft preparation, M.T.; writing—review and editing, M.T., W.D.-C., M.K., M.P., O.S., P.B, F.A.L., S.F, B.W. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** The technical assistance of Tina Spalholz is acknowledged. The sig1R knockout mice were kindly provided by Envigo RMS SARL.

**Conflicts of Interest:** The authors declare no conflicts of interest.

#### **References**


**Sample Availability:** Samples of the compounds are not available from the authors.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
