Caffeine and Methylliberine: A Human Pharmacokinetic Interaction Study

Introduction : Methylliberine and theacrine are methylurates found in the leaves of various Coffea species and Camellia assamica var. kucha , respectively. We previously demonstrated that the methylxanthine caffeine increased theacrine’s oral bioavailability in humans. Methods: Consequently, we conducted a double-blind, placebo-controlled pharmacokinetic study in humans administered methylliberine, theacrine, and caffeine to determine methylliberine’s pharmacokinetic interaction potential with either caffeine or theacrine. Subjects received an oral dose of either methylliberine, caffeine, methylliberine plus caffeine, or methylliberine plus theacrine using a randomized, double-blind, crossover design. Blood samples were analyzed using UPLC-MS/MS. Results : Methylliberine exhibited linear pharmacokinetics that were unaffected by coadministration of either caffeine or theacrine. However, methylliberine coadministration resulted in decreased oral clearance (41.9 ± 19.5 vs. 17.1 ± 7.80 L/hr) and increased half-life (7.2 ± 5.6 versus 15 ± 5.8 hrs) of caffeine. Methylliberine had no impact on caffeine’s maximum concentration (440 ± 140 vs. 458 ± 93.5 ng/mL) or oral volume of distribution (351 ± 148 vs. 316 ± 76.4 L). Conclusions : We previously demonstrated theacrine bioavailability was enhanced by caffeine, however, caffeine pharmacokinetics were unaffected by theacrine. Herein, we found that methylliberine altered caffeine pharmacokinetics without a reciprocal interaction, which suggests caffeine may interact uniquely with different methylurates.

Methylliberine has recently been granted new dietary ingredient (NDI) status following completion of a 90-day repeated-dose oral toxicity study 5 . In addition, human adverse event potential studies using methylliberine alone, and in combination with theacrine, found no effect of methylliberine on heart rhythm (electrocardiogram; ECG), resting heart rate, or blood pressure 6 . Anecdotal benefits of methylliberine suggesting reduced onset of action of EMF activity without anxiety has led to methylliberine being "stacked" (i.e., combined) with caffeine and/or theacrine. Because we previously demonstrated the interaction potential between the methylxanthine caffeine and the methylurate theacrine, we hypothesized that caffeine would interact with the methylurate methylliberine. Therefore, the purpose of this study was to determine methylliberine pharmacokinetics and its pharmacokinetic interaction potential with theacrine and caffeine following oral administration to humans.

Scientific Methods
The study protocol and informed consent were approved by the University of Memphis Institutional Review Board (Proposal # FY2018-490). Study participants were informed of all procedures, potential benefits, and risks associated with the study and provided informed consent prior to any study related procedures.
Eight hours post-dose, participants stayed in a conference room and were allowed to study, watch TV, or browse the internet. Meal replacement bars and shakes were provided 2-and 6-hours post-ingestion. Water intake was matched across conditions. Participants then went home with standardized food consisting of meal replacement bars and shakes that was kept consistent between visits. Participants returned the following day 24 hours after supplement ingestion. Participants were expected to arrive in a 10-hour fasted state at the same time for each visit between 6am-8am. Participants were asked to refrain from exercise the day prior to their lab visit. Upon arrival in the lab, participants rested for 10-15 minutes before blood pressure readings and catheter insertion. After the baseline sample was drawn, participants ingested the supplement, i.e. time of ingestion was matched across conditions. During screening, participants completed a publicly available caffeine consumption questionnaire. Participants were excluded for consuming greater than 400 mg daily. Participants kept diet logs 2 days prior to each lab visit. Meal replacement bars and shakes, as well as water intake, was matched across conditions

Pharmacokinetic Data Analysis
Caffeine, methylliberine, and theacrine oral pharmacokinetic parameters were estimated from plasma concentrationtime data, adjusted for lag time (tlag), using noncompartmental methods in Phoenix WinNonlin (version 8.2, Certara USA, Inc., Princeton, NJ) as previously described 8 . Maximum concentration (Cmax) and time corresponding to Cmax (Tmax) were determined from the plasma concentration versus time data. Area under the plasma concentration-time curve from time 0 to infinity (AUC0-∞) was calculated using the linear trapezoidal rule. The terminal half-life (t1/2), was evaluated using ln 2/kel, with kel as the terminal rate elimination constant estimated from the slope of the linear regression of the log plasma concentration versus time curve during the terminal phase. The oral clearance (CL/F) was calculated by dividing the administered oral dose by AUC0-. The apparent oral volume of distribution during the terminal elimination phase (Vz/F) was calculated by dividing CL/F by kel.

Statistical Analysis
To determine the probability of interaction magnitude between methylliberine and caffeine and/or theacrine, pharmacokinetic parameters were first logarithmically transformed. For each parameter, mean differences of the transformed values were obtained by taking the average of the difference of the transformed values, and upper and lower confident level with a 90% confidence interval (CI) were obtained using the paired t test function in Excel. The results of this analysis were exponentiated which corresponded to 90% confidence intervals around the geometric mean ratios for any observed pharmacokinetic parameters 9 .

Subject characteristics
Twelve healthy men (n=6; aged 24±4 years; 77±6 kg) and women (n=6; aged 22±3 years; 58±4 kg) participated in this study. Men and women ingested a daily amount of caffeine 155±85 mg and 230±103 mg, respectively. All subjects were well tolerated all treatments and no adverse effect was recorded. Diet intake was not changed across all treatment conditions for total kilocalorie, macro-and micro-nutrient composition.

Pharmacokinetics
Mean plasma concentration (± standard deviation) time profiles for methylliberine, caffeine, and theacrine are shown in Figures 1A-C. Methylliberine pharmacokinetic parameters for each cohort are shown in Table 1. We found the methylliberine was rapidly absorbed from the oral administration, with Cmax reached on average at 0.6 and 0.9 hours after following low (25 mg) and high (100 mg) doses of methylliberine, respectively ( Figure 1A, Table 1). Thereafter, methylliberine was eliminated with a half-life averaging 1 to 1.4 hours. Dose-normalized Cmax and AUC were significantly higher, oral clearance and oral volume of distribution were significantly lower, following 100 mg dose of methylliberine compared to 25 mg of methylliberine.    The study was not designed to determine the effect of caffeine and/or methylliberine co-administration on theacrine pharmacokinetics, viz., there was not an arm where subjects received only theacrine. However, based on our previous pharmacokinetic studies with theacrine and caffeine, it appears that methylliberine increased the half-life of theacrine by approximately two-fold 8 .

Discussion
Variation in caffeine sensitivity has spurred the discovery of wider therapeutic index natural stimulant platforms that include a unique class of purine alkaloids known as methylurates, e.g., theacrine, also exert their pharmacologic effects via adenosine receptor modulation 2,4,10,11,12,13 . Intriguingly, however, the pharmacologic profile of theacrine appears distinct from caffeine in that it does not alter cardiovascular parameters (e.g., heart rate) [14][15][16] . For this reason, theacrine is frequently combined ("stacked") with caffeine in energy, mood, and focus dietary supplements; unfortunately, with little regard for pharmacokinetic and pharmacodynamic interaction potential. We previously demonstrated that when combined, caffeine diminished theacrine's oral clearance (CL/F) without altering its half-life (t1/2 ~Vd/CL), which suggested that the most likely mechanism for the observed interaction was that caffeine increased theacrine's oral bioavailability (F) 8 . In the present study, we expanded our investigation of the interaction potential between caffeine and methylurates by examining the impact of methylliberine on caffeine pharmacokinetics. Similar to our previous study, we found that caffeine co-administration led to a modest, but significant, decrease in CL/F and Vd/F, as well as an increase in plasma area under the curve (AUC = (F*Dose)/CL) of methylliberine. However, methylliberine halflife, and by extension Vd and CL, were unaltered by caffeine.
In our previous study investigating the pharmacokinetic interaction potential between theacrine and caffeine, theacrine was found to have essentially no effect on caffeine bioavailability or clearance 8 . The inability of theacrine to increase caffeine bioavailability is not surprising as caffeine is a low extraction drug with an oral bioavailability approaching unity. However, the lack of an effect of theacrine on caffeine clearance is informative since it implied that theacrine, while it may be a CYP1A2 substrate, is not a clinically significant CYP1A2 inhibitor. In the current study, however, concomitant administration of caffeine and methylliberine led to significantly increased caffeine exposure (AUC), which was accompanied by commensurate decreases in half-life and oral clearance (CL/F). Interestingly, data from our previous study, although not designed to evaluate pharmacokinetic interaction potential, clearly showed that caffeine oral clearance (CL/F) was substantially lower than literature reports when co-administered as a cocktail also containing both theacrine and methylliberine 7 . The mechanism by which methylliberine reduced oral clearance (CL/F) of caffeine is unlikely related to increased bioavailability (F) considering the fact that caffeine's bioavailability is complete and that caffeine's oral volume of distribution (Vd/F) was unchanged.
A clue as to the potential mechanism by which methylliberine decreases the oral clearance (CL/F) of caffeine is provided by the fact that caffeine is a low hepatic extraction drug that is extensively metabolized (>90%) by CYP1A2 to the N3-demethylated metabolite paraxanthine 17 . Hepatic clearance of low extraction drugs is approximated by multiplying the fraction of unbound drug (fup) and intrinsic clearance (CLint) 18 . Thus, a reduction in caffeine's hepatic clearance is likely attributable to a reduction in intrinsic clearance, which reflects CYP1A2 activity. Thus, our data support the notion that methylliberine decreases the intrinsic clearance of caffeine through mechanisms likely including inhibition of CYP1A2. However, we cannot discount many other potential factors such as gender, race, genetic variation, disease, and exposure to inducers, which contribute to large interindividual variability in CYP1A2 activity and thus caffeine clearance 17,19,20 . For example, caffeine's plasma clearance is reduced in patients with liver cirrhosis, hepatitis B, and hepatitis C 21,22 . Moreover, smoking stimulates caffeine clearance via CYP1A2 induction, whereas cessation of smoking decreases caffeine clearance 20,[23][24][25] . It is also puzzling that theacrine, which was administered at doses similar to methylliberine doses in this study, did not affect caffeine clearance in our previous study 8 .

Conclusions
In conclusion, methylliberine, a methylurate analog of caffeine, increased plasma exposure and half-life of caffeine following concomitant oral administration. The mechanism underlying this pharmacokinetic interaction is likely attributable to methylliberine inhibition of CYP1A2, which is a major determinant of intrinsic clearance, and thus hepatic clearance, of caffeine. Several important consequences, with regard to herb drug interaction potential, can be inferred from the data assuming reproducibility in larger more diverse populations. First, caffeine is commonly used as a probe drug to examine CYP1A2-mediated drug interactions 26 . Consequently, our data demonstrate that methylliberine has the potential to interact with other drugs whose elimination depends on CYP1A2. Secondly, methylurate pharmacology is still in its infancy, but early studies imply that methylxanthine and methylxanthine ligands differ in their affinity and selectivity for the adenosine A1 and A2A receptors 12 , as well as, the sirtuin 3 receptor 27 . Ergo, additional pharmacology studies are needed to provide insight into the pharmacodynamic interaction potential between methylxanthines and methylurates.