Cholecystokinin 1 Receptor – A Unique G Protein-Coupled Receptor Activated by Singlet Oxygen (GPCR-ABSO)

Plasma membrane-delimited generation of singlet oxygen by photodynamic action with photosensitizer sulfonated aluminum phthalocyanine (SALPC) activates cholecystokinin 1 receptor (CCK1R) in pancreatic acini. Whether CCK1R retains such photooxidative singlet oxygen activation properties in other environments is not known. Genetically encoded protein photosensitizers KillerRed or mini singlet oxygen generator (miniSOG) were expressed in pancreatic acinar tumor cell line AR4-2J, CCK1R, KillerRed or miniSOG were expressed in HEK293 or CHO-K1 cells. Cold light irradiation (87 mW⋅cm-2) was applied to photosensitizer-expressing cells to examine photodynamic activation of CCK1R by Fura-2 fluorescent calcium imaging. When CCK1R was transduced into HEK293 cells which lack endogenous CCK1R, photodynamic action with SALPC was found to activate CCK1R in CCK1R-HEK293 cells. When KillerRed or miniSOG were transduced into AR4-2J which expresses endogenous CCK1R, KillerRed or miniSOG photodynamic action at the plasma membrane also activated CCK1R. When fused KillerRed-CCK1R was transduced into CHO-K1 cells, light irradiation activated the fused CCK1R leading to calcium oscillations. Therefore KillerRed either expressed independently, or fused with CCK1R can both activate CCK1R photodynamically. It is concluded that photodynamic singlet oxygen activation is an intrinsic property of CCK1R, independent of photosensitizer used, or CCK1R-expressing cell types. Photodynamic singlet oxygen CCK1R activation after transduction of genetically encoded photosensitizer in situ may provide a convenient way to verify intrinsic physiological functions of CCK1R in multiple CCK1R-expressing cells and tissues, or to actuate CCK1R function in CCK1R-expressing and non-expressing cell types after transduction with fused KillerRed-CCK1R.

We have unambiguously confirmed permanent photodynamic activation of CCK1 receptor (CCK1R) and the associated amylase secretion in pancreatic acinar cells Cui, 1989, 1990a,b;al-Laith et al., 1993;Cui et al., , 2012An et al., 2003). In such experiments, freshly isolated pancreatic acini were perifused, briefly exposed (10 min) to photosensitizer sulphonated aluminum phthalocyanine (SALPC), before washing out unbound SALPC. Subsequent light illumination (1 min) of rat pancreatic acini with plasma membrane-bound SALPC then triggered persistent calcium oscillations, in sharp contrast with oscillations induced by physiological CCK concentrations, which disappeared immediately after CCK wash-out . Photodynamically induced calcium oscillations were blocked by CCK1R antagonist FK480; after FK480 blockade of photodynamic calcium oscillations, the muscarinic agonist bethanechol triggered new calcium oscillations, indicating that after permanent photodynamic CCK1R activation and subsequent CCK1R blockade with FK480, pancreatic acinar cells remained perfectly healthy (An et al., 2003;Cui et al., 2012). 1 O 2 in the cellular milieu has a short lifetime (µs; Cui and Matthews, 1998;Bovis et al., 2012;Kim et al., 2014), and therefore has a limited effective diffusion or reactive distance (<10 nm; Moan and Berg, 1991;Cui and Matthews, 1998;Dougherty et al., 1998;Nowis et al., 2005;Cui et al., 2012). 1 O 2 generated in photodynamic action is therefore effective only at the site of generation, i.e., at the site of photosensitizer. Although in our previous experiments photosensitizer SALPC was incubated with pancreatic acini briefly to limit SALPCbinding to plasma membrane, therefore limiting 1 O 2 generation to plasma membrane, 1 O 2 localization would be much improved if the photosensitizer could be targeted with higher specificity and precision. Genetically encoded protein photosensitizers, due to the possibility of fusion with signal sequences for specific subcellular targeting, would be perfect for such precise localization.
The pancreatic acinar cells are typical CCK1R-expressing cells vital for digestive enzyme secretion An et al., 2003;Liang et al., 2013Liang et al., , 2017. But it has not been demonstrated whether CCK1R expressed in other cell types would be equally susceptible to photodynamic 1 O 2 activation. Therefore the aims of the present work were: (i) to examine whether CCK1R ectopically expressed in cells other than pancreatic acinar cells could be activated by SALPC photodynamic action, and (ii) to examine whether photodynamic action with KillerRed or miniSOG target-expressed to the plasma membrane could activate photodynamically the CCK1R in CCK1R-expressing pancreatic acinar tumor cell line AR4-2J, and in cell lines ectopically expressing CCK1R. The present work confirmed that CCK1R is photodynamically activated irrespective of the photosensitizers used or cell types where CCK1R is expressed. Importantly, KillerRed fused with CCK1R retains its photodynamic effect to activate the fused CCK1R after light irradiation. This important new finding may immediately open up new avenues to elucidate CCK1R physiology and extend CCK1R pharmacology in vivo both in the central nervous system and in peripheral organs.
Solid E. coli medium LB/Kana was sterilized and culture plates made. Liquid E. coli medium LB/Kana had the same composition but without agar.

Immunocytochemistry
Dispersed cells were attached to Cell-Tak-coated cover-slips before being fixed in paraformaldehyde 4% (10 min). Cells were permeabilized in 0.2% Triton X. Non-specific binding was blocked in 3% BSA (PBS) before incubation with primary antibodies in a humid chamber at 4 • C overnight, followed by incubation with 2nd ab (30 min). Cover-slips were placed on a slide, sealed and stored at 4 • C. For double staining, incubation with 1st Ab was repeated, and incubation with 2nd Ab was performed with mixed 2nd Abs. Imaging was done in a confocal microscope (Zeiss LSM 510 META, objective 63 × /1.40 oil).

Vector Constructs
Plasmids pKillerRed mem and pKillerRed dmito were bought from Evrogen (Moscow, Russia). Competent E. coli were infected with plasmid, cultured on solid LB/Kana. Bacteria colonies were picked and further cultured in liquid LB/Kana with shaking overnight. Proliferated plasmid was extracted and sequence verified.
pKillerRed lyso was constructed fusing pKillerRed mem with lysosomal localization sequence KGQGSMDEGTADERAPLIRT via in-fusion cloning. Forward primer 5 -CGCGGGCCCGGGA TCCATGAAAGGACAGGGATCCATGGATGAGGGAACAGC GGATGAAAGAGCACCCCTCATTCGAACCTCCGAGGGCG GCCCCG-3 was annealed with reverse primer 5 -CGGGGCC GCCCTCGGAGGTTCGAATGAGGGGTGCTCTTTCATCCG CTGTTCCCTCATCCATGGATCCCTGTCCTTTCATGGATC CCGGGCCCGCG-3 to obtain a double stranded complementary DNA. pKillerRed mem was linearized after digestion with BamH1-specific endonuclease. The above double stranded complementary DNA was fused with linearized pKillerRed mem by a fusion HD enzyme, to obtain vector pKillerRed lyso . pKillerRed lyso was transformed into DH5α, harvested and sequenced for verification.
To construct vector CCK1R-pKillerRed plasmid pKiller-Red mem was used as template to amplify KillerRed by PCR. The forward primer was 5 -GGATCCATGCTGTGCTGTATG AGAAGAA-3 , downstream primer was 3 -GAATTCATCCTC GTCGCTACCGATG-5 . The forward primer contained a BamHI site, the reverse an EcoRI site. These two restrictive sites were used to cut the PCR products and pcDNA3.1-CCK1R and target fragments were cloned into pcDNA3.1-CCK1R by transforming DH5α. The recombinant plasmids were extracted and the sequence verified and named CCK1R-pKillerRed. The transduced cells were named CCK1R-pKillerRed-CHO-K1.
To load transduced cells with fluorescent probes for mitochondrial or lysosomal tracking, cells were incubated with MitoTracker Green (0.05 mM) or LysoTracker Green (0.075 mM) for 30 min before imaging.

Photodynamic Action
To trigger photodynamic action, KillerRed-or miniSOGexpressing cells were irradiated with white light (87 mW·cm −2 ), SALPC-bound cells were irradiated with red light (>580 nm, 36.7 mW·cm −2 ), from a halogen cold light source (MegaLight 100, Hoya-Schott, Japan). Illuminance and irradiance were measured with a power meter (IL1700, International Light Inc., Newburyport, MA, United States). The fluorescence emission lightpath (for calcium measurements) need not be deflected off the detector during red light irradiation since Fura-2 emission was cut off at 550 nm (emitter D510/40 nm, see below).

Calcium Measurements
Dispersed cells were loaded with Fura-2 AM (10 µM, 30 min), attached to Cell-Tak-coated cover-slip bottom of Sykes-Moore chambers for 30 min before perfusion. Cytosolic calcium was measured in an inverted fluorescent microscope (Olympus IX 70 or Nikon NE3000) coupled to a calcium measurement system (PTI, New Jersey, United States) with alternating excitations at 340 nm/380 nm (DeltaRam V or X). Emission (dichroic mirror 400DCLP, emitter D510/40 nm) was detected with a PMT (pmt814, PTI) or a CCD (NEO-5.5-CL-3, Andor). Calcium concentration was expressed as F 340 /F 380 and plotted against time with SigmaPlot as reported before (Jia and Cui, 2011;Liang et al., 2013;Li et al., 2015).

Statistical Analysis
To analyze the statistical significance of the differences of peak values before, during, and after devazepide (1 nM) in Figure 3C, all calcium peaks were normalized to the mean of peaks before perfusion of devazepide. Student's t-test was used, and P < 0.05 was taken as statistically significant, indicated with an asterisk ( * ).

SALPC Photodynamic Activation of CCK1R Expressed in HEK293 Cells
CCK1R was transiently transduced into HEK293 cells. Immunocytochemistry done 48 h after transfection showed plasma membrane CCK1R localization; note the complete lack of CCK1 receptor in un-transfected cells in the same field ( Figure 1A). CCK up to 500 pM had no effect on cytosolic calcium in un-transfected cells (Figure 1Ba), in these untransfected cells no changes in basal calcium were ever found after photodynamic action (SALPC 2 µM, light λ > 580 nm, 36.7 mW·cm −2 ; Figure 1Bb). CCK (20 pM) triggered regular calcium oscillations in CCK1R-HEK293 cells, CCK wash-out led to immediate cessation of induced calcium spikes (Figure 1Bc). Exposure to the chemical photosensitizer SALPC (2 µM) in the dark had no effect on basal calcium in the same cell (Figure 1Bc), but subsequent light irradiation (λ > 580 nm, 36.7 mW·cm −2 ) triggered regular and persistent calcium oscillations, which continued long after cessation of light illumination (Figure 1Bc). Note the difference of CCK-and photodynamically induced calcium oscillations: the former disappeared immediate after CCK wash-out, the latter continued long after completion of light irradiation.
The pKillerRed mem -AR4-2J cells showed a stable cytosolic calcium baseline in the dark (Figure 2D), but white light irradiation (87 mW·cm −2 ) induced oscillatory increases in cytosolic calcium, which persisted after completion of light irradiation ( Figure 2E). In non-transfected AR4-2J, basal calcium remained stable, chemical photosensitizer SALPC (0.5 µM) was perfused briefly, subsequent red light irradiation (> 580 nm, 36.7 mW·cm −2 ) induced calcium oscillations ( Figure 2F) Experimental data as shown in (C) were analyzed, with all calcium peaks normalized to the mean of peaks before devazepide (taken as 1.00, indicated with a horizontal dashed blue line). The asterisk ( * ) indicates statistical significance at P < 0.05. Note that in (C) the ordinate was different, due to the use of the CCD-based calcium measurement system. One only needs to note the dynamic changes.
which were similar to white light irradiation-induced calcium oscillations in pKillerRed mem -AR4-2J cells ( Figure 2E). These data indicate that KillerRed mem photodynamic action fully duplicates SALPC photodynamic action ;  An et al., 2003), both KillerRed and SALPC photodynamically activate the plasma membrane CCK1 receptor.

DISCUSSION
In the present work we have found that CCK1R ectopically expressed in HEK293 was activated by SALPC photodynamic action, duplicating SALPC photodynamic CCK1R activation in isolated rat pancreatic acini. KillerRed mem or miniSOG mem photodynamically activated CCK1R in AR4-2J cells. Fused protein CCK1R-KillerRed mem maintained both KillerRed and CCK1R activities, KillerRed mem photodynamically activated CCK1R in KillerRed-CCK1R-CHO-K1 cells. These data together suggest that photodynamic 1 O 2 activation is an intrinsic property of CCK1R, independent of the photosensitizer used or cell types where CCK1R is expressed.
KillerRed has a GFP-like structure with a central chromophore of Q65-Y66-G67 (Pletnev et al., 2009;Roy et al., 2010). Q65-Y66-G67 connects with an aqueous channel (formed by I142, L143, P144, I199, I200, T201) to ensure oxygen supply from the medium. The excited chromophore transfers its excitation energy to ground state molecular oxygen to generate reactive oxygen species (ROS) which also exit KillerRed via this aqueous channel (Carpentier et al., 2009;Pletnev et al., 2009;Serebrovskaya et al., 2009;Roy et al., 2010). The purified KillerRed may undergo type I photodynamic action to generate superoxide (Pletnev et al., 2009;Shu et al., 2011;Vegh et al., 2011;Kim et al., 2014), although evidence for 1 O 2 generation by a Type II photodynamic action in the cellular context is very strong (Roy et al., 2010;Petrova et al., 2016). The case for 1 O 2 generation is indicated in the present work, since KillerRed photodynamically activated CCK1R, rather similar to SALPC photodynamic CCK1R activation. It is known that neither superoxide nor H 2 O 2 had any effect on pancreatic acinar cell CCK1R (data not shown).
The flavin mononucleotide (FMN)-binding mini singlet oxygen generator (miniSOG) is composed of two α-helices interspersed in five β-sheets, with FMN located in between (Shu et al., 2011;Pietra, 2014). 1 O 2 probing with anthracene-9, 10dipropionic acid (ADPA) obtained a quantum yield of 0.47 (Shu et al., 2011). But direct measurement of 1 O 2 phosphorescence at 1275 nm, and the use of uric acid as an 1 O 2 probe obtained a quantum yield of 0.03 (Ruiz-González et al., 2012;Pimenta et al., 2013). Since miniSOG is less than half the size of KillerRed, miniSOG may have some advantages over KillerRed in sensitizer/target protein fusion experiments, taking into account the fact that miniSOG mem photodynamically activates CCK1R similarly.
The present work was carried out at the cellular level in vitro. Based on the present work, in vivo photodynamic 1 O 2 CCK1R activation is also possible. CCK1R activations in situ would have immediate physiological and pharmacological significance, either with peripheral or central CCK1R. CCK1R in nodose and dorsal root ganglia are known to play vital roles in satiety and other peripheral sensations (Broberger et al., 2001;Li et al., 2011;Kaczynska and Szereda-Przestaszewska, 2015). Highly localized CCK1R expression in the mouse hippocampus and defined extracortical sites are also well-recognized (Nishimura et al., 2015). Third ventricular ependymal cell CCK1R is known to be important for infant mouse satiety (Ozaki et al., 2013). A CCK-CCK receptor-like satiety-control system is commonly found in lower invertebrates such as Caenorhabditis elegans (Janssen et al., 2008;Bhattacharya et al., 2014). Our accumulated works suggest that CCK1R is unique among class A GPCR: it is activated permanently by type II photodynamic action (i.e., by 1 O 2 ). CCK1R is the only G protein-coupled receptor activated by singlet oxygen ( 1 O 2 ; GPCR-ABSO) identified so far, adding new arsenals alongside RASSL and DREADD (Alvarez-Curto et al., 2011;Gomez et al., 2017) for the elucidation of GPCR functions.
In conclusion, photodynamic activation is an intrinsic property of CCK1R, independent of photosensitizers used or CCK1R-expressing cell types. Photodynamic CCK1R activation by 1 O 2 after transduction of genetically encoded photosensitizer in situ would provide a convenient way to verify unambiguously intrinsic physiological functions of CCK1R in multiple CCK1R-expressing cells or tissues, or to actuate CCK1R function in expressing and non-expressing cell types after transduction with fused KillerRed-CCK1R, miniSOG-CCK1R or other similar constructs.

Physiological Relevance and Perspectives
The present work found that CCK1R is activated by type II photodynamic action (i.e., 1 O 2 ) irrespective of the photosensitizers used or the CCK1R-expressing cell types. Therefore CCK1R is a unique G protein coupled receptor activated by singlet oxygen (GPCR-ABSO). The 1 O 2 could in the future be provided in vivo by photodynamic action of knockedin expression of photosensitizers such as KillerRed, miniSOG, to use the GPCR-ABSO property to confirm unambiguously CCK1R functions by directing focused light to central or peripheral cells or tissues. 1 O 2 can be generated endogenously from photodynamic action in the skin. Skin photodynamic action is triggered after absorption of sunlight in the ultraviolet A region by endogenous photosensitizers. 1 O 2 is also generated in neutrophil respiratory burst in neutrophil-infiltrated/inflamed tissues. Therefore the GPCR-ABSO property of CCK1R is highly relevant in cellular physiology and is likely to play a significant role in future physiological research.

AUTHOR CONTRIBUTIONS
ZJC conceived the idea of the project, supervised all the experiments, and finalized the manuscript. HNJ performed the experiments with KillerRed, YL with miniSOG, and WYJ with SALPC. HJ, YL, and WYJ wrote up the respective sections and all authors checked and approved the final submitted version.