[ 11 C]glyburide PET imaging for quantitative determination of the importance of Organic Anion-Transporting Polypeptide transporter function in the human liver and whole-body

Organic Anion-Transporting Polypeptides (OATPs) are known to control the liver uptake of many drugs. Non-hepatic expression of OATPs has been reported although functional importance for whole-body pharmacokinetics (WBPK) remains unknown. Glyburide is a well described substrate of several hepatic and non-hepatic OATPs. Dynamic whole-body positron emission tomography (DWB-PET) with [ 11 C]glyburide was performed in humans for determination of the importance of OATPs for liver uptake and WBPK. Seven healthy male subjects (24.7 ± 3.2 years) underwent [ 11 C]glyburide PET scan with concomitant blood sampling. All subjects underwent Abbreviations: ABC, ATP-binding cassette; AUC, area under the curve; AUCR, ratio of AUC tissue to AUC blood ; BBB, blood-brain barrier; BCRP, breast cancer resistance protein; C blood,t , blood concentration at time t; C liver,t , liver concentration at time t; COVID-19, coronavirus disease-19; CT, computed tomography; DDI, drug-drug interaction; DILI, drug-induced liver injury; DWB, dynamic whole-body; f b , fraction bound to plasma protein; HPLC, high-performance liquid chromatography; IDIF, image-derived input function; k uptake , uptake transfer constant; MR, magnetic resonance; MRAC, MR based PET attenuation correction; OATP, organic anion-transporting polypeptide; OAT, organic anion transporter; OCT, organic cation transporter; P-gp, P-glycoprotein; PET, positron emission tomography; PK, pharmacokinetics;


Introduction
It is often hypothesized that passive diffusion is the main mechanism mediating drug diffusion from circulation to organs.However, identification of membrane transporters at the blood-tissue interface suggests that delivery of drugs to tissues may be controlled by carrier-mediated processes [1][2][3][4].Influx transporters of the Solute Carrier (SLC) superfamily enable the blood-to-tissue transfer of several endogenous and exogenous compounds.Transporters of the Organic Anion-Transporting Polypeptide (OATP) family are well known to mediate the hepatic uptake of many drugs [5].Liver OATPs include OATP1B1, OATP1B3, and OATP2B1.OATP1B1/1B3 are assumed to be liver-specific although non-hepatic expression of OATP1B3 has been reported [6].Expression of OATP2B1 is predominant in the liver but was also detected in non-clearance organs including human myocardium, blood-brain barrier (BBB), epithelial lung cells, ovary, mammary duct, and muscles.Other non-hepatic OATPs have been reported, including OATP1A2, whose expression has been detected in brain, liver, lung, kidney, and testis [7].The importance for non-hepatic OATPs in controlling drug delivery to these potential target/vulnerable tissues relative to hepatic OATP function remains to be assessed in vivo [8].
So far, pharmacokinetic (PK) studies are mainly based on the timeconcentration profile of drugs in plasma and body fluids [9].This approach is commonly performed to investigate pharmacokinetic drug-drug interactions (DDIs) involving transporters including OATPs [3].However, from a perspective of drug efficacy or toxicity, it is the tissue concentration rather than plasma concentration that is the most relevant.Efforts are currently being made to consider PK as a global and whole-body approach (WBPK) [10,11].In this framework, dynamic whole-body (DWB) positron emission tomography (PET) imaging using radiolabeled analogs of drugs is emerging as an innovative method to unveil the delivery of drugs in the whole-body [12,13] and explore molecular mechanisms controlling the WBPK of drugs in animals and humans [14].
Efforts have been made to develop dedicated PET probes to investigate the impact of efflux transporters of the ATP-binding cassette (ABC) family on drug delivery to selected organs such as the brain [15,16] or the lungs [17].The role of SLC transporters did not receive as much attention although their importance for WBPK remains unclear [2,18].Several probes have been proposed for quantitative PET imaging of liver OATP function [19].So far, clinically validated PET probes used in humans include (15 R)-[ 11   [20][21][22][23][24].The substrate profile of these probes corresponds to hepatic OATPs (OATP1B1/1B3 and/or OATP2B1) and imaging acquisitions consisted of dynamic acquisitions focused on the liver.A major limitation of current probes is the formation of radiometabolites causing the PET signal associated with radiometabolites in tissues to be indistinguishable from that associated with the parent (unmetabolized) radioligand.Metabolic stability is therefore essential for correct interpretation of PET data, especially in the liver which is the main site for drug metabolism [14].
Another important feature of PET probes for influx transporters is the sensitivity to detect changes in transporter function [14].Sensitivity of investigated substrate probes to inhibition is usually estimated by the magnitude of the decreased liver uptake precipitated by inhibition/depletion of transporter function.This can be easily achieved for liver OATPs in humans using a single dose of rifampicin, a safe and potent OATP inhibitor [3].Finally, the liver uptake of an ideal PET probe for OATPs should be negligible when transporters are inhibited, thus ensuring the specificity of the carrier-mediated transport over passive diffusion or other transporter systems.
Glyburide (5-chloro-N-[2-[4-(cyclohexylcarbamoylsulfamoyl) phenyl]ethyl]-2-methoxybenzamide, a.k.a glibenclamide) is a hypoglycemic agent commonly used in the treatment of noninsulin-dependent diabetes.The type 1 surfonylurea receptors (SUR1) subunits are the binding site of glyburide which induces insulin secretion by pancreatic cells through inhibitory effect on the K ATP channel [25].Recent in vitro and simulation studies suggest that the carrier-mediated uptake of glyburide by the liver is predominantly mediated by OATP1B1 [26].This is consistent with clinical studies showing significant DDIs with decreased plasma clearance of glyburide following administration of rifampicin in humans [27,28].However, previous in vitro studies suggest that glyburide may also be substrate for non-hepatic OATP2B1 and OATP1A2, which may impact WPBK [29,30].
Isotopic radiolabeling of glyburide with carbon-11 ([ 11 C]glyburide) is feasible [31].[ 11 C]glyburide PET imaging was therefore evaluated in healthy volunteers as a probe for OATP function.An optimized DWB-PET acquisition protocol using a PET-MR (magnetic resonance) system was performed to simultaneously compare [ 11 C]glyburide distribution in the liver and other organs in order to assess the importance of OATP-mediated transport on the WBPK of glyburide in humans.

Radiosynthesis of [ 11 C]glyburide
[ 11 C]glyburide was synthesized using a previously described automated two-step method [31].Detailed information are provided as Supplementary data.

Subjects and study design
The study protocol was approved by an Ethic Committee (CPP IDF5: 17041) and the experiments were conducted with respect of the Declaration of Helsinki.All subjects had a first medical examination where the inclusion criteria were checked by the clinical investigator.A written informed consent was signed after the subjects received the information about the study.
Eleven healthy male volunteers (< 30 years-old) were included in the study.Four of them did not participate to imaging sessions due to technical issues regarding radiosynthesis or the PET-MR system.Seven subjects (24.7 ± 3.2 years-old, body weight was 75 ± 10 kg) S. Marie et al. participated to two subsequent imaging sessions which took place during a single day to avoid inter-day variability.An intravenous (i.v.) bolus of [ 11 C]glyburide was first injected.After 3 h decay, total PET signal represented < 0.22 % of injected radioactivity due to the very short radioactive half-life of carbon-11 (20.4 min).The second i.v.bolus of [ 11 C]glyburide was therefore injected at least 3 h later, immediately after infusion of rifampicin (9 mg/kg diluted in glucose 5 % perfused, within 45 min of [ 11 C]glyburide injection, Rifadin®, Sanofi-Aventis, Gentilly, France).All volunteers received the first injection of [ 11 C] glyburide (147 ± 63 MBq i.v.).Five of them received the second injection of [ 11 C]glyburide (161 ± 69 MBq i.v.) after infusion of rifampicin.The mean specific radioactivity at the time of injection was 7988 ± 4250 MBq/µmol.The pharmacokinetic half-life of i.v injected glyburide (1 mg) in human plasma is ~ 1.5 h [32].Injected mass of glyburide associated with the first injection of [ 11 C]glyburide was probably not fully eliminated from the body at the time of the second injection.However, microdoses of glyburide (11.2 ± 5.8 µg) injected for PET imaging were very low compared with pharmacological dose (1.25-5 mg).Cumulative saturation of carrier-mediated transport is therefore very unlikely.

PET imaging
Whole-body dynamic acquisitions were performed using a SIGNA® PET/MR scanner (GE Healthcare, Waukesha, WI, USA).The three first subjects had a catheter inserted in one arm for arterial blood sampling and a venous catheter in the other arm for [ 11 C]glyburide injection.Due to the COVID-19 pandemic, access to anesthesiologist staff for insertion of the arterial catheter was no longer possible and a venous catheter was inserted instead for subsequent subjects to estimate the proportion of radiometabolites.The subjects were positioned on the patient bed and MR localization acquisitions were performed.A first blood sample was collected before [ 11 C]glyburide injection for determination of the binding to plasma proteins.PET acquisition began with the injection of [ 11 C]glyburide (1 min bolus).The first 3-minute dynamic mono-bed acquisition was positioned on the abdomen (16 frames of 10 s), to capture uptake in the liver and some abdominal organs including spleen, pancreas, kidneys, and myocardium.This was immediately followed by repeated multi-bed whole-body acquisitions performed over the next 27 min (5 bed positions, 11 frames of 2.5 min).Arterial blood samples were collected every 25 s in the first 4 min and at 5, 7, 10, 15, 20, and 30 min post-injection.Venous blood samples were collected at 1, 3, 5, 7, 10, 15, 20, and 30 min post-injection.

PET data analysis
PET images were reconstructed using a 3D-iterative reconstruction algorithm, including resolution modeling and time-of-flight.Data were corrected for radioactive decay, attenuation, and random and scattered coincidences.Attenuation correction was based on a 2-point Dixon MR sequence (MRAC).Volumes of interest (VOIs) were delineated for tissues (liver, kidneys, spleen, myocardium, pancreas, brain, testis, muscle, and eyes) which were delineated from PET and/or co-registered MRAC images using the PMOD® software (version 3.9, PMOD Technologies LLC, Zurich, Switzerland).An additional VOI was delineated on a multi-bed volume which includes the left ventricle and the aorta to derive an image-derived input function (IDIF).Time-activity curves (TACs) were generated by calculating the mean radioactivity in these VOI (kBq/cc).Standardized uptake values (SUV) were also generated by correcting the radioactivity in the VOI for injected dose and body weight.Areas under the TACs (AUC, expressed in SUV.min) were calculated from 0 to 30 min for each TAC to describe tissue exposure.
The transfer constant (k uptake ) of [ 11 C]glyburide from blood to the liver was estimated using the integration plot method [24,33] from 5 to 30 min after [ 11 C]glyburide injection using the equation , where C liver,t and C blood,t represent the concentrations of [ 11 C]glyburide in the liver and blood (IDIF) respectively at time t.AUC blood,0-t represents the area under the IDIF TAC from time 0 to t. k uptake corresponds to the slope of the linear regression plot.V E represents the initial distribution volume in the liver at time 0.
The [ 11 C]glyburide fraction bound to plasma protein (f b ) with and without rifampicin infusion was estimated on the first sample (obtained before injection of [ 11 C]glyburide).Plasma sample was added ~4 MBq of [ 11 C]glyburide and f b was determined as previously described using a validated ultrafiltration method (MPS Micropartition Microcon®-YM-10 membrane, Millipore, Molsheim, France) [34].

Data and statistical analysis
Data were calculated as the mean ± SD (standard deviation).Linear regression was used to test the correlation between activity in arterial blood samples and IDIF.Paired t-test was used to identify differences before and after rifampicin-treatment on [ 11 C]glyburide kinetics (p < 0.05).Statistical analysis was performed using GraphPad Prism® software (version 8.4, San Diego, CA, USA).

[ 11 C]glyburide blood data
[ 11 C]glyburide was barely metabolized in both the baseline and rifampicin conditions (Supplemental Fig. 1).Parent (unmetabolized) [ 11 C]glyburide in either arterial (n = 6) or venous (n = 4) plasma samples accounted for more than 90 % of the plasma radioactivity over PET acquisition (Fig. 1A).A small proportion of radiometabolites could be detected in baseline condition at each time point (maximum 6.8 ± 3.6 % at 10 min).In the presence of rifampicin, radiometabolites could only be detected at 5 min after injection (6.0 ± 2.4 %) but not at later time points.It was therefore decided not to correct the arterial input function from radiometabolites given their limited impact on PET quantification.
[ 11 C]glyburide is known to be highly bound to plasmatic proteins [35].Plasma protein binding was 99.45 ± 0.01 % in baseline experiments (n = 7).Rifampicin did not impact binding to plasma proteins which was 99.33 ± 0.01 % in rifampicin-treated subjects who underwent arterial blood sampling (n = 3).Rifampicin did not impact the plasma/whole-blood ratio which was 1.65 ± 0.17 in baseline and 1.62 ± 0.14 in rifampicin-treated subjects at the end of PET acquisition (n = 4, p > 0.05).
Correlation of [ 11 C]glyburide concentrations measured from wholeblood samples and IDIF obtained in the 3 subjects who underwent arterial blood sampling is presented in Fig. 1B.Linear regression was calculated for both baseline and rifampicin situations and confirmed excellent correlation (R 2 = 98.6 % in baseline condition, R 2 = 96.8% after treatment by rifampicin and R 2 = 97.9% with both combined).This enabled the use of IDIF for kinetic modeling of [ 11 C]glyburide PET data, with no need for arterial catheterization.

[ 11 C]glyburide PET data
Representative PET images of the whole-body distribution of [ 11 C] glyburide with and without rifampicin are shown in Fig. 2.Under baseline conditions, [ 11 C]glyburide showed a slow but predominant uptake by the liver (SUV max = 18.6 ± 3.9 at 30 min, Fig. 3) but also substantial distribution to the kidneys.Lower uptake by the myocardium and spleen was observed (Fig. 2).Uptake by other tissues was extremely low and associated with high variability, probably due to the poor signal counting.Strikingly, radioactivity was obvious in the circulation, reflecting high retention of [ 11 C]glyburide in the plasma, probably due to high binding to plasmatic proteins and/or limited diffusion of [ 11 C]glyburide from blood to most tissues.
Treatment by rifampicin drastically decreased [ 11 C]glyburide uptake by the liver (Fig. 2) with a 48.5 ± 6.5 % decrease in liver exposure (AUC liver , Fig. 3 and Table 1).Consistently, kinetic modeling confirmed a 77.3 ± 7.3 % decrease in the transfer constant of [ 11 C]glyburide from blood to the liver (k uptake ) which ranged from − 67.4 % to − 87.9 % among subjects (Fig. 4).The magnitude of the decrease in k uptake in response to inhibition was strongly correlated with baseline k uptake (Fig. 4C, R 2 = 0.98, p < 0.01).
As a consequence to decreased liver uptake, OATP inhibition using rifampicin significantly increased concentrations of [ 11 C]glyburide in the vasculature (Figs. 2 and 3 and Table 1).AUC blood was significantly higher (p = 0.003, n = 5) after treatment with rifampicin (mean increase of + 25.0 ± 10.3 %, n = 5) (Table 1).Urinary bladder was more obvious on PET images after rifampicin infusion (Fig. 2).Urinary clearance of radioactivity was not significantly modified at 30 min after treatment by rifampicin (Supplemental Fig. 2A).Radioactivity detected in urine almost exclusively corresponded to radiometabolites (n = 10 samples in both baseline and rifampicin-treated conditions) (Supplemental Fig. 2B).
Pretreatment by rifampicin increased tissue exposure (AUC 0-30 min ) for kidneys (+ 21.5 ± 12.1, p = 0.0043) and myocardium (+ 43.3 ± 25.1, p = 0.034) (Table 1).Transfer constants from blood to the tissues (k uptake ) could not be accurately estimated in organs other than the liver.Importance of OATP function for distribution of [ 11 C]glyburide to tissues was therefore estimated as change in AUCR.No significant change in AUCR could be observed in organs other than the liver suggesting limited importance of OATP in mediating the tissue distribution of [ 11 C]glyburide in other tissues (Table 1).

Discussion
DWB-PET imaging using [ 11 C]glyburide was performed in humans to investigate the WBPK of glyburide and estimate the importance of OATP function on drug delivery to tissues.The main result of the study was the dramatic impact of OATP inhibition using rifampicin on the liver uptake of [ 11 C]glyburide in humans.Rifampicin decreased the liver AUCR by 58.7 ± 5.2 % and the k uptake by 77.3 ± 7.3 %, suggesting high importance of OATP in controlling the uptake of glyburide by human hepatocytes.A preclinical study using [ 11 C]glyburide PET in non-human primates demonstrated that the OATP inhibitors rifampicin (same dose than in the present study) or cyclosporin (15 mg/kg/h, i.v.) decreased the liver uptake of [ 11 C]glyburide in a higher extent (− 90 % in AUCR liver ) [36].This is consistent with the higher expression of OATP1B1, assessed using quantitative proteomics, in the liver of non-human primates compared with humans, suggesting species differences in liver OATP function [37].
Our clinical data show that [ 11 C]glyburide is poorly metabolized in humans, which is consistent with the known PK of glyburide (< 5 % proportion of 4-hydroxy-glyburide in plasma over 24 h) [28].This is a major advantage because it enables correct interpretation of PET data, assuming suitable radiochemical purity of the PET signal in tissues over the investigated timeframe.Metabolic stability of [ 11 C]glyburide is a major asset compared with previous OATP probes which are extensively metabolized.Radiometabolites of [ 11 C]dehydropravastatin were predominant in human blood at 30 min [22].(15 R)-[ 11 C]TIC-Me is a prodrug that is rapidly metabolized into (15R)-[ 11 C]TIC (10 min) which is the OATP substrate, and for which another radiometabolite becomes predominant after 20 min [22,24].
Liver uptake of [ 11 C]glyburide in humans was very sensitive to the inhibition by rifampicin as revealed by the ~ 77% decrease in k uptake .In comparison, only 25 % decrease in the uptake of [ 11 C]erlotinib was observed after treatment by rifampicin (600 mg i.v.), 45 % for (15R)-  [ 11 C]TIC and 58 % for [ 11 C]dehydropravastatin (both using the same dose of rifampicin 600 mg per os) [20,22,24].High response to OATP inhibition by rifampicin compared to previous radioligands suggests exquisite sensitivity of [ 11 C]glyburide to detect change in transporter function, in a limited number of subjects.[ 11 C]glyburide is a poorly permeable, highly lipophilic compound, with high binding to plasma proteins.It may be hypothesized that transport of [ 11 C]glyburide across the basolateral pole of hepatocytes may be predominantly governed by OATP function, with negligible concurrent passive diffusion.Consistently, k uptake of [ 11 C]glyburide during OATP inhibition was extremely low (0.0154 ± 0.074 min − 1 ) suggesting that almost complete OATP inhibition was achieved, leading to low uptake by hepatocytes.Li et al. used in vitro models to simulate the liver uptake clearance of glyburide and concluded that OATP-mediated uptake accounted for 98% of the total liver uptake of glyburide (5560 L/h for carrier-mediated transport vs 111 L/h for passive diffusion) [26].They also suggested that liver uptake of metabolites is ~10-fold lower compared with parent glyburide, thus supporting suitable radiochemical purity of the PET signal for [ 11 C]glyburide in the liver [26].A large and consistent body of research suggests that the liver uptake of glyburide is predominantly mediated by OATP transporter(s) [26,38,39].However, contribution of other transporters in the liver uptake of [ 11 C]glyburide cannot be excluded.To the best of our knowledge, glyburide has not been reported as a substrate of organic cation transporter 1 (OCT1), another important influx transporter expressed at the sinusoidal pole of hepatocytes [38,39].In our study, almost complete inhibition of the liver uptake of [ 11 C] glyburide was achieved using rifampicin which may inhibit multiple hepatocyte transporters including OATPs but does not inhibit OCT1 [40,41].Therefore, comparison of the transporter substrate/inhibitor profile of glyburide and rifampicin does not support a role for OCT1 in controlling the liver uptake of glyburide in humans.
Altogether, our PET data interpreted in the light of previous in vitro data, suggest that the uptake of [ 11 C]glyburide by the liver is predominantly governed by OATPs.Compared with previously reported OATP probes, [ 11 C]glyburide benefits from exquisite sensitivity and suitable radiochemical purity of the PET signal in both the blood and the liver.The negligible proportion of radiometabolites in the blood and liver allows for correct estimation of the OATP-mediated transport capacity of hepatocytes which was a major limitation of previous PET radioligands that undergo substantial liver metabolism [14,19].
The magnitude of response to inhibition, which requires two subsequent PET scans prior to and after OATP inhibition, was used to estimate the total transport capacity of OATP in the liver.Interestingly, baseline k uptake accurately predicted the magnitude of response to inhibition.Two subsequent scans will still be needed to investigate the risk of DDIs involving OATPs using [ 11 C]glyburide, as exemplified with rifampicin in the present study.However, our results suggest that a single [ 11 C]glyburide PET scan is sufficient for quantitative determination of the total transport capacity of OATP, given the limited importance of non-OATPmediated uptake of [ 11 C]glyburide by the liver.This allows for quantitative study of the regulation of OATP activity observed in various patho-physiological situations [42].Imaging liver OATP function using [ 11 C]glyburide PET may therefore be useful as a molecular determinant of liver function in animal model or patients with liver dysfunction.To the best of our knowledge, such property has not been demonstrated for previous PET probes for OATPs.
The WBPK approach used in this study illustrates the relevance of DWB-PET imaging to explore the impact of transporter-mediated DDI at both the blood and the tissue levels [13].Interestingly, the DDI between glyburide and rifampicin was initially described through the magnitude change in plasma concentration of glyburide in a clinical PK study [28].In a population similar to ours, a same i.v.infusion dose of rifampicin (600 mg) significantly increased the plasma AUC of orally administered glyburide by 125 ± 51.5 % compared with the baseline situation.Magnitude of this DDI was much higher than the change observed in our study (25.0 ± 10.3 %) [28] which may be explained by several methodological differences.First, glyburide was given orally at a dose of 1.25 mg while in our study [ 11 C]glyburide was given i.v. at microdoses (11.2 ± 5.8 µg).Above all, our PET data were obtained only the 30 min after injection, thus predominantly reflecting the distribution phase while the plasma PK study reported by Zheng et al. was carried out over 24 h, mainly reflecting the elimination phase.This suggests that early PET data may underestimate the overall impact of transporter-mediated DDI on drug elimination, especially when hepatic elimination is involved [28].Altogether, this suggests that WBPK imaging using short-life isotopes such as carbon-11 is useful to unveil the tissue distribution of drugs but may not accurately capture the entirety of the plasma PK of drugs with longer plasma half-life.PK PET data should be interpreted in the light of known plasma PK but cannot not replace it.
The DWB-PET protocol enabled direct visualization of the WBPK of glyburide.Our results show that dramatic decrease in liver uptake of [ 11 C]glyburide resulted in increased exposure in the blood and several organs including kidneys, myocardium, spleen, and brain.Increase in the pancreas, which is assumed to be the target tissue for glyburide, was not significant (p > 0.05).This may be due to poor quantification associated with low uptake and radioactive counting compared with surrounding abdominal tissues.From a therapeutic perspective, this suggests that modest uptake suffices to promote insulin release by pancreatic cells, although extra-pancreatic mechanism of action has been proposed for glyburide, with a central role of the liver [43].From an imaging perspective, this suggests that [ 11 C]glyburide is not appropriate as a probe for imaging SUR1 expression in the pancreas.
In non-human primates, rifampicin-inhibitable uptake of [ 11 C]glyburide occurred in the myocardium and kidney cortex suggesting a functional role for non-hepatic OATPs in these organs [36].This was not observed in humans suggesting a limited role for non-hepatic OATPs in controlling the blood-to-tissue uptake of [ 11 C]glyburide [44].Non-hepatic OATPs of putative importance for PK are OATP2B1 and OATP1A2 which are ubiquitously expressed [7].However, non-hepatic tissues with reported OATP2B1, OATP1A2 or OATP1B3 expressions such as the brain, kidneys, myocardium, spleen, pancreas, testis, muscle, and eyes display very low baseline uptake of [ 11 C]glyburide and no or limited changes in AUCR after OATP inhibition using rifampicin [8].Altogether, this suggests that in human [ 11 C]glyburide behaves as a liver-specific OATP substrate in vivo, suggesting species differences in the importance of non-hepatic OATPs [45].
Several hypotheses can be drawn to explain this result with respect to investigated organs.First, recent in vitro and simulation studies, performed using MDCKII-cells transfected with the human genes, suggest that glyburide may specifically be transported by OATP1B1 which is a liver-specific member.In the brain, PET images showed the negligible uptake of [ 11 C]glyburide although OATP2B1 and OATP1A2 are expressed at the BBB (Fig. 2) [8,46].[ 11 C]glyburide is transported by efflux transporters from the ATP-Binding Cassette (ABC) family such as the P-glycoprotein (P-gp, ABCB1) and the breast cancer resistance protein (BCRP, ABCG2) which restrict the brain penetration of their substrates across the BBB [47][48][49][50]36,51].In this situation, a competition may occur between influx and efflux transporters at the same interface [14].The role of OATPs expressed in the kidneys such as OATP4C1 is poorly understood [52].Given the high proportion of radiometabolites in urine, the radio-HPLC analysis of urine samples was not sensitive enough to detect the excreted amount of parent (unmetabolized) [ 11 C] glyburide at the end of PET acquisition.It was therefore not possible to assess whether rifampicin altered CL urine of [ 11 C]glyburide from our PET data.In a clinical PK study, Zheng et al. consistently showed that the 0-12 h accumulated amount of glyburide in urine was 198-fold lower than that of metabolite (4-hydroxyglyburide) in baseline condition, and 117-fold after OATP inhibition using i.v rifampicin [28].Moreover, they reported no change in the urinary clearance of glyburide and 4-hydroxy-glyburide after rifampicin injection, suggesting a limited role for rifampicin-inhibitable transport in controlling the urinary clearance of glyburide in humans [28].Additionally, in vitro studies suggest that the metabolites of glyburide, but not glyburide itself, are secreted in the urine via the organic anion transporters 3 (OAT3) expressed on the basolateral membrane of the kidney proximal tubule cells [26].Further experiments are needed to untangle the importance of carrier-mediated processes, which may involve influx and efflux transporters, for the renal elimination of glyburide and its metabolites.
To the best of our knowledge, it is the first application of PET-MR DWB imaging for the study of dynamic processes such as PK and DDIs in humans.Compared with current static or whole-body acquisition routinely performed using PET-CT (computed tomography) scanners, DWB imaging PET-MR reduces volunteer exposure to radiation, allows for repeated scanning without additional CT for attenuation correction, and ensures accurate delineation in organs including soft tissues [53].This original method enables quantification of radiolabeled substrates of transporters in several organs in a parallel manner which in turn allows for direct comparison of the importance of transporter function between organs in each individual.This suggests that the development of DWB acquisition and total-body scanners may foster the use of PET imaging to elucidate the WBPK in early drug development as part of clinical pharmacology studies [54].Finally, efforts are being made to ensure suitable drug delivery to tissues while reducing liver exposure of drugs and/or metabolites and prevent drug-induced liver injury (DILI) [55,56].Our clinical results suggest that almost complete inhibition of liver OATP function can be achieved using a clinical dose of rifampicin.Given the limited role for non-hepatic OATPs in controlling the uptake of drugs by tissues, this strategy may be further used to block the liver uptake of OATP substrates while ensuring increased delivery to other organs, which can be monitored using DWB-PET.

Conclusion
[ 11 C]glyburide was validated as a PET probe for imaging OATP function in the human liver.[ 11 C]glyburide benefits from suitable PK properties including metabolic stability and high magnitude of response to OATP inhibition, suggesting suitable sensitivity and specificity.[ 11 C] glyburide provides a relevant tool to investigate OATP-mediated DDI at the liver level or the regulation of OATP function in health and disease.The use of a WBD PET-MR acquisition provided unique information on the impact of OATP-mediated DDIs on tissue exposure to drugs which brings new insights for WBPK and carrier-mediated drug delivery to the whole-body in humans.

CRediT authorship contribution statement
TOURNIER Nicolas contributed to the study conception and design.Material preparation, data collection and analysis were performed by MARIE Solène, CHALAMPALAKIS Zacharias, BREUIL Louise, BECQUE-MONT Laurent, LECOQ Anne-Lise, CAILL É Fabien, GERVAIS Philippe, COMTAT Claude, BOTTLAENDER Michel, and TOURNIER Nicolas.The first draft of the manuscript was written by MARIE Solène and all authors commented on previous versions of the manuscript.All authors read and approved the final manuscript.

Fig. 1 .
Fig. 1. [ 11 C]glyburide metabolism in healthy volunteers in absence and presence of rifampicin (9 mg/kg i.v.) (A).Correlation of whole-blood concentrations of 11 C]glyburide (kBq/cc) measured from gamma-counting of blood samples with PET quantification (image derived input function) obtained in the baseline and the rifampicin-treated groups (B).Metabolism data are presented as mean percentages ± SD of unchanged [ 11 C]glyburide in plasma vs time after i.v.injection of [ 11 C]glyburide in baseline conditions and after rifampicin infusion.

Fig. 2 .
Fig. 2. Whole-body PET images of [ 11 C]glyburide distribution in baseline condition and after treatment with rifampicin (9 mg/kg i.v.) in a healthy volunteer.Images are the maximum intensity projections of the frames summed from 3 to 30 min and are expressed in standardized uptake values (SUV).

Fig. 3 .
Fig. 3. Time-activity curves of [ 11 C]glyburide in liver (A, B) and blood (image-derived, C, D) in baseline conditions and after treatment with rifampicin (9 mg/kg i. v.) from to 30 min (B, D) and with a focus from 0 to 3 min (A, C).Data are mean ± SD.

Table 1
Areas under the curve (AUC) and area under the curve ratios (AUCR ¼ AUC tissue /AUC blood ) of the time-activity curves of [ 11 C]glyburide calculated for each subject in baseline condition and after treatment with rifampicin (9 mg/kg i.v.).Reported blood pool data come from the validated image-derived input function.Data are mean ± SD; *, p < 0.05, paired t-test compared to baseline).

Fig. 4 .
Fig. 4. Transfer of [ 11 C]glyburide from blood to the liver assessed by integration plot analysis from 5 to 30 min (A), change in transfer constant (k uptake ) (B), and magnitude of the decrease in k uptake (C) for each subject between baseline condition and after treatment with rifampicin (9 mg/kg i.v.).Blood data come from a validated image-derived input function.The k uptake was assessed from the integration plot analysis by determining the slope of the regression plot for each PET acquisition (*, p < 0.05, paired t-test).The magnitude of change was assessed by subtracting for each subject k uptake after treatment by rifampicin to k uptake in baseline condition.