A novel bioassay for quantification of surface Cannabinoid receptor 1 expression

The cannabinoid receptor type 1 (CB1) plays critical roles in multiple physiological processes such as pain perception, brain development and body temperature regulation. Mutations on this gene (CNR1), results in altered functionality and/or biosynthesis such as reduced membrane expression, changes in mRNA stability or changes in downstream signaling that act as triggers for diseases such as obesity, Parkinson’s, Huntington’s, among others; thus, it is considered as a potential pharmacological target. To date, multiple quantification methods have been employed to determine how these mutations affect receptor expression and localization; however, they present serious disadvantages that may arise quantifying errors. Here, we describe a sensitive bioassay to quantify receptor surface expression; in this bioassay the Gaussia Luciferase (GLuc) was fused to the extracellular portion of the CB1. The GLuc activity was assessed by coelenterazine addition to the medium followed by immediate readout. Based on GLuc activity assay, we show that the GLuc signals corelate with CB1 localization, besides, we showed the assay’s functionality and reliability by comparing its results with those generated by previously reported mutations on the CNR1 gene and by using flow cytometry to determine the cell surface receptor expression. Detection of membrane-bound CB1, and potentially other GPCRs, is able to quickly screen for receptor levels and help to understand the effect of clinically relevant mutations or polymorphisms.

Scientific Reports | (2020) 10:18191 | https://doi.org/10.1038/s41598-020-75331-y www.nature.com/scientificreports/ the change in CB1 localization is point mutation F237L that reduces the membrane expression of this receptor 15 . Mutation T210I is characterized not only by reduced membrane expression but also by several times greater affinity for ligands, making it essentially constitutively active 16 , while mutation T210A results in a completely inactive CB1 despite membrane localization, although it seems to lack internalization capabilities 17 . Many techniques have been developed and employed for the quantification of GPCRs cell surface expression; the most common include flow cytometry using fluorescent probes 18,19 , ELISA 20 , surface biotinylation 21 , among others, which require long incubations and multiple washing steps that increase quantitative determination variability per assay, or involve complex imaging equipment. In order to solve these problems, here we report an inexpensive and easy bioassay for the relative quantification of the CB1 receptor expression at the cell surface; it is based on the genetic fusion of the CB1 and Gaussia luciferase (GLuc), a widely used enzymatic reporter in eukaryotes. As a proof of principle, we analyzed three previously reported point mutations (CB1 F237L , CB1 T210I and CB1 T210A ) and compared them to the wild-type (WT) counterpart 22 .

Materials and methods
Materials. KOD

Plasmid design.
A plasmid containing the CNR1 gene fused to the GLuc gene (GLuc-CNR1) was generated by Gibson Assembly by combining products from two PCR reactions previously carried out to amplify the CNR1 gene (F 1 : AAG TCG ATC CTA GAT GGC CTTG, R 1 : TCT TAA GGA GGG ATG GGG TGA F 2 : AAA AAT ACT GAC TCC AAC CAT TCA A R 2 : TTT GCC ATC AGA CTG TGA AAT AAG G). The CRN1 gene was cloned from genomic DNA isolated from a mouth-swab of the main author (IR-R), using the previously mentioned primers to amplify the coding region and the long 3′-UTR (5.4 Kb) fragments. Linearized vector was digested with DpnI to remove methylated DNA and purified using purification columns (DNA Clean and Concentrator Kits, ZYMO RESEARCH). The generated plasmid sequence was designed to provide both proteins (CB1 and GLuc) with a high rotational flexibility and enabling immunodetection by adding the sequence FLAG tag between them. The CRN1-FLAG-GLuc plasmid was then used for generating the CNR1 mutants (CB1 F237L , CB1 T210I and CB1 T210A ).

Construct description.
The architecture of the constructs consisted of the GLuc gene, which is comprised of 558 bp, cloned directly at the N-terminal of the CNR1 gene. To ensure proper folding and functioning they were linked by either a flexible domain (GLEG) or the FLAG sequence (DYKDDDDK), the latter for immunodetection purposes. The full-length CNR1 (consisting of 1416 bp) followed by the naïve long 3′-UTR of the coding sequence (3992 bp) was as well cloned into the vector. The 3′UTR of CNR1 has been shown to influence the expression levels of this receptor 23,24 . The signal peptide used to localize both proteins at the membrane is contained within the GLuc coding sequence (1 -17 amino acids). Schematic plasmid vector and protein construct is shown on Fig. 1.
Transformation and plasmid purification. Previously prepared electrocompetent E. coli bacteria 25 were electroporated with the plasmid generated by Gibson Assembly; resulting colonies (after selection with ampicillin 100 mg/mL) were checked by colony PCR. Positive colonies were then digested with EcoRI, those showing the expected length (10,146 bp) were sequenced. At least two sequence-verified clones for each construct were used for the experiments.
Plasmid DNA (pDNA) was isolated by minipreps (Zyppy Plasmid Miniprep Kit, ZYMO RESEARCH). Isolated plasmid's integrity was observed in 1% Agarose gel; enough plasmid concentration was isolated and stored at − 20 °C for until further use. The GLuc-linker-CNR1 plasmids purity and concentration were determined by UV spectroscopy, a concentration of 540.2 ng/µL was obtained with a A 260 /A 280 ratio of 1.8 indicating a high purity. Plasmids carrying mutations of CNR1 showed similar concentrations and A 260 /A 280 ratios.

Receptor relative quantification.
To quantify the expression of the CB1 receptor, the GLuc detection assay was performed 48 h after transfection. The culture medium was transferred to another well and each well containing transfected HEK293T cells was washed multiple times with PBS. A solution of 20 µM coelenterazine was diluted in 5 mL PBS and incubated for 30 min at room temperature under dark to allow its stabilization. Subsequently, the final 10 µM concentration of cell-added coelenterazine was obtained by mixing 50 µL of the stable coelenterazine with 50 µL of the cells medium. The GLuc activity was measured immediately by triplicate using a plate luminometer (TECAN Infinite 200 PRO); results were then statistically analyzed.

Flow cytometry analyses.
GLuc-CB1 detection in the cells was assessed by measurement of the fluorescence intensity from binding the primary anti-FLAG mouse antibody (Cell Signaling Technology, Cat. Nr: #8146) and secondary anti-mouse antibody conjugated with Alexa Fluor 555 dye (Invitrogen Cat. Nr: A-31570). The HEK293T cells were seeded onto 6-well plates 24 h before transfection. Transfection procedure was described at Transfection and receptor quantification section. After 48 h, the cells were washed (Ca 2+ and Mg 2+ -free PBS) and harvested (5 mM solution of EDTA in Ca 2+ and Mg 2+ -free PBS). Both, free floating as well as adherent cells were gathered. Next, after centrifugation (500×g, 5 min), the appropriate cells pellets were separated into two main parts. The part of cells was used for total (intracellular and surface membrane) CB1 detection. Cells were fixed by resuspending in fixation and permeabilization buffer (BD Pharmingen, Cytofix/Cytoperm solution, cat. numb. 554722) and incubated for 20 min on ice. Next, cells were washed (BD Perm/Wash buffer, cat. numb. 554723) and centrifuged (500×g, 5 min). Mocked (UT), CB1 F237L , CB1 T210I and CB1 T210A cells were incubated with primary anti-flag antibody (1 h, 37 °C and 5% CO 2 ) and subsequently after wash step, labeled with secondary Alexa Fluor 555-conjugated antibody (1 h, 37 °C and 5% CO 2 ). Part of the cells were incubated only with Alexa Fluor 555 conjugated Ig (described as control). For total CB1 quantification all Ig dilutions and wash steps were made in BD Perm/Wash buffer. www.nature.com/scientificreports/ The second part of harvested cells were used for surface membrane CB1 detection. The cell fixation and permeabilization steps were omitted. Next, cells were prepared as described above. All Ig dilutions and wash steps for surface membrane CB1 detection were made in Ca 2+ and Mg 2+ -free PBS.
All stainings were performed directly before the flow cytometric analysis (BD FACSCalibur, CellQuest Pro Version 6.0. software for the Macintosh operating system). The fluorescence Alexa Fluor 555 dye intensity of individual cell was determined and at least 10,000 events were measured within an acquisition rate of 100-300 events/s. Statistical analyses. Significance among luminescence readings was assessed by one-way ANOVA followed by Tukey's post-hoc test. Statistical analyses of the flow cytometry were performed using GraphPad Prism 8.0 (GraphPad Software Inc., California, U.S.A). ANOVA with Tukey post hoc test and column statistics were used for comparisons (*, p < 0.05; **, p < 0.01; ***, p < 0.001 was considered statistically significant). All tests were performed in the triplicates, at least.

Results and discussion
We took advantage of the natural properties of GLuc which does not require cofactors and it is normally secreted-it contains a secretion signal peptide 26 . This allows us to quantify the CB1 relative expression on the surface of HEK293T cells. The relative quantification is easily done by measuring the luminescence generated by the GLuc enzyme linked to the CB1 protein, since this luminescence depends on the amount of receptor, changes on its expression will be quantifiably noted. A schematic representation of this method is shown in Fig. 1C.
As described in the Construct Description section, the signal peptide (SP in Fig. 1) used to localize the receptor at the membrane is encoded in the GLuc gene, in agreement with previous research works that have employed artificial signal sequences at the N-terminus to allow membrane localization 27 since the majority of the GPCR receptors including CB1 do not possess a signal peptide to be trafficked to the cell surface 28 . Foreign sequences at the N-terminal of GPCRs, are known to affect the addressing of the receptor to the cell surface due to a low efficiency of translocation through the endoplasmic reticulum (ER). For this reason, the addition of a signal sequence (signal peptide, SP) at the N-terminus of a fusion construct is necessary 27,29 . In this case, GLuc, which already contains a SP, was inserted in front of CB1 separated by a short linker to avoid interfering with the CB1 folding and topology; in some other cases the fusion of such chimeric proteins has been done between residues 25 and 26 of the extracellular N terminus following the same principle 27 .
At first, the initial architecture containing the GLuc linked to the CB1 by a flexible domain consisting of 4 amino acids (GLEG), resulted in that most of GLuc was found in the medium and not attached to the cells ( Fig. 2A), what seems to be caused by the cleavage of the GLuc-CB1 by a protease. To address this, we incubated cells expressing this GLuc-CB1 with several protease inhibitors and found the Batimastat, a pan-metalloprotease inhibitor, diminished the amount of free GLuc (Fig. 2B). We then analyzed a short sequence that comprises the GLEG flexible domain and the proteins GLuc and CB1 (specifically DKIKGAGGDGLEGKSILD) for metalloprotease sites using a prediction software "SitePrediction" (https ://www.dmbr.ugent .be/prx/bioit 2-publi c/SiteP redic tion/) and found a potential cleavage site that could disrupt the generated chimeric protein causing GLuc enzyme to be released ( Supplementary Fig. S1). This sequence was partially removed by inserting a FLAG Tag between the GLuc and the CB1, which plays the role of a linker and a spacer for proper folding Fig. 1. Subsequent experiments showed that the GLuc activity, after cells were transfected with the GLuc-FLAG-CNR1 plasmid, was virtually only from cells (Cells) and not in the culture medium (Medium) (Fig. 2C). Mock transfected cells and their medium showed baseline luminescence values, as shown in Fig. 2C. Based on the results, the generated luminescence specifically comes from the transfected cells, clearly demonstrating that the enzyme is not being secreted and that it is only readable in cells expressing the GLuc-FLAG-CB1. Full sequences of both architectures (GLuc-FLAG-CNR1 and GLuc-GLEG-CNR1) are provided in the Supplementary material, both as nucleic acids and amino acid sequences, as well, the plasmid maps are shown in Supplementary Figs. S2 and S3.
The above results show that adding a FLAG tag between the GLuc and the CB1 let the chimeric protein be expressed at the cell surface and being properly folded. This also suggests that any intracellular GLuc signal is minimal e.g. during biosynthesis and localization, as proteolytically processed GLuc would show activity within the cells while transporting, which is virtually unperceived.
The one-way ANOVA showed that there is a statistically significant difference among the means of the four factors (cells and medium from transfected GLuc-FLAG-CB1 and mock transfected cells) (p ≤ 0.001); the Tukey's test revealed that the transfected cells (Cells) mean is significantly different from those of the other three factors, confirming the source of luminescence.
Having established that the receptor is at the cell membrane and can be quantified through this bioassay, we next generated a series of CNR1 mutants (CB1 F237L , CB1 T210I , CB1 T210A ), in order to demonstrate changes in the cell surface expression caused by these point mutations are in line with previous observations. Indeed, the mutants behaved as expected, where the CB1 T210A exhibited similar expression to CB1 WT , while CB1 F237L and CB1 T210I showed considerable reduced expression on membrane (Fig. 3).
The GLuc-CB1 reporter showed that there is a statistically significant difference among the means of the cells transfected by GLuc-FLAG-CB1 WT and GLuc-FLAG-CB1 F237L , or GLuc-FLAG-CB1 T210I (Fig. 3), however, no difference was observed between those transfected by GLuc-FLAG-CB1 WT and GLuc-FLAG-CB1 T210A , indicating similar cell surface receptor expression. Figures 2C and 3  www.nature.com/scientificreports/ considered as a 100-fold reduction in the expression but are in fact inter assay variations. Comparisons can only be done between samples of the same run, as we have done for WT vs mutants in Fig. 3. Through this bioassay we corroborated previous data in that the CB1 F237L and CB1 T210I mutants cause a reduced CB1 receptor expression level on the cell surface 15,16 . Our data also supports that CB1 T210A mutation results in unaffected expression 17 .
Next, we tested whether luciferase readouts agreed with membrane-anchored GLuc-FLAG-CB1 expression using an antibody against the FLAG tag by flow cytometry. The analyses are shown in Fig. 4. The CB1 WT and CB1 T210A mutant showed similar GLuc-CB1 expression both in permeabilized as well as in not permeabilized cells, while CB1 F237L and CB1 T210I have significantly lower (p ≤ 0.001) membrane expression (Fig. 4C) as expected from the GLuc readouts. GLuc-CB1 expression in permeabilized (total expression) and not permeabilized (membrane-anchored only) cells is shown in Fig. 4A,B, respectively. See Supplementary Table S1 for a statistical analysis of the collected data.
According to the flow-cytometry results, the total expression for all the mutants was apparently the same as the wild type, however the results demonstrate how the point mutations CB1 F237L and CB1 T210I directly diminish the membrane-anchored receptor quantity. The membrane expression of the wild-type and CB1 T210A clearly indicate that the addition of the signal peptide aided the membrane localization of the receptor, and that the lower levels of expression seen for the mutants CB1 F237L and CB1 T210I are caused by different pathways, as alterations in biosynthesis, internalization, a phenomenon previously observed for the mutated receptor F237L 15 , or by a constitutive endocytosis. Either way, the bioassay is able to detect such changes which is essential for screening of hitherto mutations.
The results obtained by cell cytometry and GLuc-generated relative luminescence units were both in accordance with each other and those in literature, clearly indicating that the readouts generated by both experiments quantify correctly the receptor expressed at the cell membrane for all the analyzed cases. This is also supported www.nature.com/scientificreports/ by the GLuc-GLEG-CB1 test where virtually no intracellular readout was detected ( Fig. 2A) and instead almost all luminescence is found in the culture media.
There are some limitations that should be considered; in this study we have used the CB1 to demonstrate that the luminescence comes only from those transfected cells; however, it has been reported that the expression of this receptor occurs at other intracellular structures such as the mitochondria 30 . In addition, coelenterazine has a lipophilic nature and its able cross the cell membrane; a non-permeable substrate is available 31 , although this is easily detected by flow cytometry or confocal microscopy if the mutant is suspected to be mostly intracellularly located. Another limitation is that our assay cannot assess CB1 activity, for which the system could be coupled to some of the others signaling assays such as TANGO 32 , GloSensor 33 , or other assays that measure cAMP, inositol phosphate or calcium accumulation methods 34 and two measurements could be accounted for, expression vs activity.

Conclusion
CB1 plays key homeostatic role in a variety of physiological processes, new tools have been constantly being designed to provide a better understanding of its mechanisms and variability. With our quantification assay, determining how mutations affect the receptor expression in a much easier way than commonly used methods will ease research regarding the endocannabinoid system and its alterations upon previously reported pathological conditions. Moreover, GLuc has been used for BRET and therefore it can be used to detect ligand-receptor or receptor-receptor interactions on the plasma membrane of living cells.
The main novelty of our work is determined by the fact that it could be possible to detect the effect of various alterations in the chemical environment (such as the constant binding of ligands) or genetical factors (such as the presence of point mutations or polymorphisms) on the GPCR receptors' expression 14 .
The detection of membrane localization of particular GPCRs is fundamental to understand the roles of mutations and protein interactions. In fact, the altered expression of the CB1 receptor is in clear association with various mental-related issues such as major depression, schizophrenia, bipolar disorder, among others [35][36][37] . Our research could provide the foundation to better understanding how dysregulation of the membrane expression of CB1 plays a role in the molecular basis for the afore mentioned diseases.
Multiple mutations, polymorphisms, and alternative splice variants of the CB1 receptor have been identified and recognized as important pharmacological targets in multiple diseases, these genetic variants of the CB1 have been deeply described elsewhere 38 . Thus, screening how these genetic variants affect the level of expression of this and other GPCRs can speed up the diagnosis and treatment, with e.g. molecular chaperones 39 , of multiple genetic-based diseases.
As previously described, other quantifying methods including flow-cytometry or ELISA can be employed to determine surface expression of the CB1 and other GPCRs, however, through a simplified bioassay as the one described here, we offer a new simple methodology with results that can be compared with those obtained by flow-cytometry, this, we believe, will potentially aid research on the GPCR field and many other membrane expressed receptors.  (T210I) are shown, in order to generate a read-out, coelenterazine was added to the cells after medium was washed away. Similar expression levels can be observed for CB1 T210A and the CB1 WT , showing that this mutation does not affect the expression of the receptor, however, mutations CB1 F237L and CB1 T210I reduced the expression at the surface of HEK293T cells, as compared to the CB1 WT .

Data availability
The datasets generated during the current study are available from the corresponding author on request. The plasmid (GLuc-FLAG-CNR1) will be available from Addgene.