Theophylline (TP) is a bronchodilator that has been commonly used for the treatment of asthma and chronic obstructive pulmonary disease (Pesce et al. 1998; Navid et al. 2016). However, its use has declined due to the narrow therapeutic range and the emergence of more potent steroid inhalants (Ward et al. 1993; Sullivan et al. 1994; Boswell-Smith et al. 2006; Zhou et al. 2006). Despite plasma concentration monitoring of TP to prevent adverse effects, unexpected adverse effects such as seizures, arrhythmias, and neurologic toxicities have frequently occurred (Hurwitz 1969; Carney 1982; Jonkman et al. 1991; Lieberman and Nelson 1993; Benowitz et al. 1995). Concurrent use of foods or drugs that are substrates, inhibitors, and/or inducers of CYP1A2 can result in unwanted changes in plasma concentration of TP, as has been observed in many cases (Conrad and Nyman 1980; Jackson et al. 1981; Tse et al. 1981; Jonkman and Upton 1984; Sarkar et al. 1992; Goasduff et al. 1996; Navid et al. 2016; Britz et al. 2019). Shown in Fig. 1, the extensive hepatic metabolism of TP by cytochrome P450 (CYP)1A2, 2E1, and 3A4 is the elimination pathway that determines TP concentrations in the body (Jonkman et al. 1991; Sarkar et al. 1992; Tjia et al. 1996; Ginsberg et al. 2004).

Figure 1. Possible metabolic pathways of TP and CA.

Caffeine (CA) is a widely used compound in food and drugs, often taken in combination with other medications (Robson et al. 1988; Navid et al. 2016). Once completely absorbed from the gastrointestinal tract, CA is metabolized primarily by CYP1A2 to produce paraxanthine (PA), theobromine (TB), and TP, which are then further metabolized into secondary metabolites (Tassaneeyakul et al. 1994; Chen et al. 2017; Nehlig 2018; Aranda and Beharry 2020; Dai et al. 2022) (Fig. 1). These metabolites have pharmacological effects, with PA and TB having similar stimulating effects to CA, and TP acting as a bronchodilator (Carney 1982; Benowitz et al. 1995). The ratio of PA/CA in plasma, also known as the AUC_PA/AUC_CA ratio, has been validated as a phenotypic indicator of CYP1A2 activity since the conversion of PA and its further metabolites is exclusively mediated by CYP1A2 in the primary metabolic pathway of CA (Fuhr and Rost 1994; Perera et al. 2010; Navid et al. 2016; Yu et al. 2016).

When using CA with TP, precautions should be taken, as these compounds share overlapping metabolic pathways via CYP1A2. Although numerous evaluations of the pharmacokinetic interaction of CA or TP with other drugs have been conducted, reports on measuring the plasma concentrations of CA and TP together, as well as their metabolites in vivo systems, are scarce. Therefore, changes in plasma concentrations of CA, including its metabolites, and TP were investigated from the perspective of CYP1A2-mediated pharmacokinetic interactions.

Materials

Caffeine (CA), theobromine (TB), theophylline (TP), paraxanthine (PA), and acetaminophen [internal standard (IS) of ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS)] were purchased from Sigma-Aldrich (St. Louis, MO, USA). All other chemicals and reagents used were of analytical grade.

Animals

Animal studies were conducted in accordance with the protocols approved by the Institute of Laboratory Animal Resources of Dongguk University_Seoul, Republic of Korea (No. IACUC-2022-011, approved on June 10th, 2022). Male Institute of Cancer Research (ICR) mice (6 weeks old and weighing 25^-30 g) were purchased from the Charles River Company Korea (Seoul, Republic of Korea). The mice were acclimated for a week before starting the pretreat process. Upon arrival, the mice were housed randomly in groups of five under strictly controlled environmental conditions, including a temperature of 20^-25℃ and 48^-52% relative humidity, with a 12 h light/dark cycle and light intensity between 150 to 300 Lux. The mice were given ad libitum freely access to food and water. They fasted for 12 h before drug administration.

The mice were divided into three groups: CA, TP, and CA+TP. The CA group received oral administration of CA, the TP group received oral administration of TP, and the CA+TP group received oral administration of CA and TP simultaneously.

Pharmacokinetics of TP, CA, PA, and/or TB after oral administration of TP, CA, or CA plus TP in mice

The procedure for drug administration and blood sampling was modified based on a previous report (You et al. 2021). Before the experiment, the mice were fasted overnight but were allowed access to water. The drugs were administered to the mice by gavage: a single dose of 30 mg (5 mL)/kg of TP (dissolved in 0.9% sodium chloride solution) for the TP group, a single dose of 10 mg (5 mL)/kg of CA (dissolved in distilled water) for the CA group, and a simultaneous administration of 30 mg (5 mL)/kg of TP and 10 mg (5 mL)/kg of CA (dissolved in 0.9% sodium chloride solution for TP and distilled water for CA) for the CA+TP group. At 5, 15, 30, 60, 120, 240, 360, 480, 600, or 720 min after the oral administration of the drugs, blood samples (150 µL) were collected via heart puncture using a heparinized insulin syringe. Considering the permissible volume of blood sample collection from one moue, blood sampling was conducted by parallel blood sampling method (Kumar et al. 2017). The blood was then centrifuged at 13,000 rpm for 1 min, and a 50 µL aliquot of the supernatant (plasma) was collected. The collected samples were stored at −20℃ for LC-MS/MS analysis of TP, CA, TB, and/or PA.

LC-MS/MS analysis of TP, CA, PA, and TP in mouse plasma

All analyses were conducted using a Waters XEVO TQ-S/UPLC system from Waters Corporation. The chromatographic separation was achieved using a reversed-phase C₁₈ column (ACQUITY UPLC BEH, 2.1 mm × 100 mm, 1.7 µm; Waters Corporation, St. Milford, MA, USA) at a flow rate of 0.3 mL/min. The mobile phase consisted of 0.1% formic acid in water (A) and methanol (B), and gradient elution was performed with A:B ratios of 88:12 (v/v) from 0 to 4.5 min, 5:95 (v/v) at 4.7 min for 5.5 min, and returned to the initial composition at 5.7 min, which was then maintained for 8 min. The total run time was 8 min.

For positive ions ([M+H]⁺), the multiple reaction monitoring (MRM) mode with an electrospray ionization (ESI) interface was used at a capillary voltage of 3.0 kV, a desolvation gas temperature of 350℃, and a gas flow of 650 L/h. The m/z values, along with their corresponding cone voltage (CV) and collision energy (CE) values, were as follows: 151.93→109.97 (20 eV of CV and 15 eV of CE) for IS, 181.02→138.07 (34 eV and 18 eV) for TB, 181.00→123.90 (20 eV and 20 eV) for PA and TP, and 195.04→138.02 (35 eV and 20 eV) for CA. The data were analyzed using MassLynx software (Version 4.1; Waters Corporation, St. Milford, MA, USA).

To prepare working solutions of CA, PA, TB, and TP, a 20 mg/mL stock solution of each compound was dissolved in dimethyl sulfoxide and mixed and serially diluted by methanol. To prepare plasma standard samples, the 100-fold concentration of the working solution containing CA, PA, TB, and TP was added to drug-free plasma samples, which were then diluted to achieve plasma standard samples with final concentrations of 0.01, 0.05, 0.1, 0.5, 1, 5, 10, or 50 µg/mL.

To deproteinize plasma samples, 100 µL of methanol containing 2.5 µg/mL of IS was added to 50 µL of plasma sample. After vortexing and centrifugation for 10 min at 12,000 rpm, the supernatant was evaporated under a stream of N₂ gas at 30℃ and reconstituted by 300 µL of distilled water with 0.1% formic acid. After vortexing and centrifugation for 5 min, 5 µL of the supernatant was injected into the column.

The typical chromatograms of drug-free mouse plasma with IS, plasma standard samples at 0.5 µg/mL TB, PA, TP, and CA, and plasma samples obtained at 360 min after simultaneous oral administration of 10 mg/kg CA plus 30 mg/kg TP (i.e., in CA+TP group) are shown in Fig. 2. The peaks of TB, IS, PA, TP, and CA appeared at 2.29, 2.35, 3.01, 3.44, and 6.41 min, respectively. The lower limits of quantification (LLOQ) values of TB, PA, TP, and CA in plasma samples were 10 ng/mL for each. The calibration curves for TB, PA, TP, and CA in the plasma samples were constructed via linear regression analysis of the peak area ratios relative to those of the IS within the range of 0.01 to 50 µg/mL.

Figure 2. Representative chromatograms of TB, IS, PA, TP, and CA. (A) Drug-free mouse plasma with IS; (B) plasma standard samples at 0.5 µg/mL of TB, IS, PA, and TP; (C) plasma samples obtained at 360 min after simultaneously oral administration of 10 mg/kg CA with 30 mg/kg TP in mice.

Pharmacokinetic analysis

Non-compartmental analysis (PK solver, version 2.1; Scientific) was used to calculate several pharmacokinetic parameters, including the total area under the plasma concentration-time curve from time zero to the last measured time in plasma (AUC_0-last) or to infinity (AUC), using the trapezoidal rule method. The area from the last datum point to time infinity was estimated by dividing the last measured plasma concentration by the terminal-phase rate constant, according to standard methods (Gibaldi and Perrier 1982).

Additionally, the following parameters were calculated: terminal half-life (t_1/2), and apparent oral clearance (CL/F). The early peak plasma concentration (C_max) and time to reach C_max (T_max) were obtained directly from the plasma concentration-time data. The data are presented as mean ± standard deviations, except for T_max, which is presented as median (ranges).

Statistical analysis

The statistical comparison of pharmacokinetic parameters was conducted using a Students’ t-test. The pharmacokinetic parameters of CA and its metabolites were compared between the CA and CA+TP groups, while those of TP were compared between the TP and CA+TP groups. All data are presented as mean ± standard deviation, except T_max, which is expressed as median (ranges).

Pharmacokinetics of TP after oral administration of TP alone and in combination with CA in TP and CA+TP groups, respectively

The mean plasma concentration-time profiles of TP after oral administration of TP alone and in combination with CA for the two groups are shown in Fig. 3, and the relevant pharmacokinetic parameters of TP are listed in Table 1. The observed and corrected TP concentrations and their pharmacokinetic parameters in the CA+TP group are shown in Table 1 and Supplementary Fig. 1. The corrected TP concentrations in the CA+TP group were calculated from the observed TP concentrations minus the TP concentrations derived from CA metabolism (i.e., from the CA group), which was used to calculate the pharmacokinetic parameters of the corrected TP.

Figure 3. Plasma concentrations of (A) TP, (B) CA, (C) PA, and (D) TB after oral administration of CA (○), TP (●), and simultaneous CA plus TP (; CA+TP) to mice, respectively. The doses of CA and TP were 10 and 30 mg/kg, respectively.

Table 1 . Mean values (± standard deviation except T_max) of pharmacokinetic parameters of TP, CA, PA and TB after oral administration of TP, CA and simultaneous CA plus TP (CA+TP) to mice, respectively. The doses of CA and TP were 10 and 30 mg/kg, respectively.

Parameters	TP (n = 7)	CA (n = 7)	CA+TP (n = 6)
Body weight (g)	32.2 ± 1.93	29.2 ± 3.17	31.9 ± 0.724
TP			Observed TP	Corrected TP
t1/2 (min)	57.5 ± 12.1	69.2 ± 14.1	75.1 ± 33.7	82.1
Tmax (min)*,a	15.0 (5-30)	60.0 (30-120)	15.0 (5-30)	15.0
Cmax (μg/mL)a	30.9 ± 2.28	0.447 ± 0.0954	29.8 ± 2.43	28.9
AUC (μg min/mL)a,b	2,606 ± 480	85.4 ± 29.6	4,090 ± 768	4,192
Vz/F (mL/kg)b	981 ± 258	13,744 ± 7,667	788 ± 270	848
CL/F (mL/min/kg)a,b	11.8 ± 1.86	133 ± 54.1	7.59 ± 1.62	7.16
AUCTP/AUCCA (%)a		9.30 ± 3.00	277 ± 13.9	285
CA
t1/2 (min)		60.4 ± 15.7	89.2 ± 28.5
Tmax (min)*		30.0 (15-60)	22.5 (5-30)
Cmax (μg/mL)		6.48 ± 0.720	7.85 ± 0.717
AUC (μg min/mL)a		914 ± 98.9	1,472 ± 250
Vz/F (mL/kg)		960 ± 253	862 ± 156
CL/F (mL/min/kg)a		11.0 ± 1.15	6.97 ± 1.27
PA
t1/2 (min)		144 ± 47.7	117 ± 22.3
Tmax (min)*,a		120 (60-120)	240 (120-360)
Cmax (μg/mL)a		1.01 ± 0.111	0.756 ± 0.107
AUC (μg min/mL)		337 ± 54.2	313 ± 30.6
Vz/F (mL/kg)		6,476 ± 1,957	5,398 ± 1,012
CL/F (mL/min/kg)		30.5 ± 4.48	32.2 ± 3.13
AUCPA/AUCCA (%)a		37.2 ± 7.32	21.6 ± 2.96
TB
t1/2 (min)		133 ± 38.7	189 ± 52.6
Tmax (min)*,a		240 (120-360)	300 (240-360)
Cmax (μg/mL)a		0.838 ± 0.178	0.560 ± 0.0695
AUC (μg min/mL)		343 ± 67.7	302 ± 40.4
Vz/F (mL/kg)		6,430 ± 2,872	8,972 ± 2,149
CL/F (mL/min/kg)		29.7 ± 5.84	33.6 ± 4.62
AUCTB/AUCCA (%)a		37.8 ± 7.74	20.7 ± 2.17

*T_max values are median values (ranges)..

^aCA group was statistically different (p < 0.05) from CA+TP groups..

^bTP group was statistically different (p < 0.05) from CA+TP groups..

Compared to the TP group, the AUC of TP was significantly greater (by 56.9%) in the CA+TP group due to the significantly slower CL/F of TP (by 35.7%). However, other pharmacokinetic parameters of TP between the two groups were not significantly different. The same pattern was observed in the corrected TP concentrations between the two groups.

Pharmacokinetics of CA and its metabolites (i.e., TP, TB, and PA) after oral administration of CA alone and in combination with TP in CA and CA+TP groups, respectively

The mean plasma concentration-time profiles of CA, TP, TB, and PA after oral administration of CA without and with TP in CA and CA+TP groups are shown in Fig. 3, and their relevant pharmacokinetic parameters are listed in Table 1.

In the pharmacokinetic profile of CA, it was detected in plasma based on early blood sampling time points (5 min) and the time to reach the peak concentration (T_max) was 30 min. This pattern was similarly observed in the CA+TP group. However, the AUC and C_max of CA in the CA+TP group were significantly greater (by 61.1%) and higher (by 21.1%) than that in the CA group. The t_1/2 of CA in the CA+TP group was also longer (by 47.6%) than that in the CA group. The significantly slower CL/F (by 36.9%) of CA in the CA+TP group was observed compared to that in the CA group.

In the pharmacokinetic profile of TP, the observed TP concentrations represent the TP from CA metabolism in the CA group, whereas the TP concentrations in the CA+TP group are the sum of TP from CA metabolism and orally administered TP. In the CA group, TP was detected in plasma-based at an early blood sampling time point (5 min), with a T_max of 60 min. The ratio of AUC_TP/AUC_CA was 9.30% in the CA group, suggesting that the contribution of TP concentrations derived from CA metabolism was minor in the observed TP concentrations (i.e., the sum concentrations of orally administered TP and formed TP) in the CA+TP group. Compared to those in the CA group, the AUC and CL/F of TP were significantly greater (by 4,692%) and slower (by 94.3%) in the CA+TP group.

In the pharmacokinetic profile of PA, PA in the CA group was detected in plasma based on early blood sampling time point (5 min), but it was detected from 15 min in the CA+TP group. The C_max and T_max of PA in the CA+TP group significantly decreased and increased (by 25.1 and 100%, respectively) compared to those in the CA group. Although there was no change in AUC values of PA between the CA and CA+TP groups, the ratio of AUC_PA/AUC_CA in the CA+TP group was significantly decreased (by 41.9%) compared to that in the CA group.

In the pharmacokinetic profiles of TB, TB in the CA group was detected in plasma based on an early blood sampling time point (5 min), but the detection of TB in the CA+TP group was delayed to 30 min. The C_max and T_max of TB were significantly reduced and increased (by 33.2 and 25.0%, respectively) in the CA+TP group compared to those in the CA group. Although there was no change in the AUC values of TB between the CA and CA+TP groups, the ratio of AUC_TB/AUC_CA was significantly decreased (by 45.2%) in the CA+TP group compared to that in the CA group. Other parameters of TB were comparable between CA and CA+TP groups.

Comparing the pharmacokinetic profiles of TP between TP and CA+TP groups, the observed TP concentrations represent the disposition of orally administered TP in the TP group, whereas the TP concentrations in the CA+TP group are the sum of orally administered TP (i.e., the corrected TP concentrations) and TP from CA metabolism (i.e., the TP concentrations in CA group). The corrected TP concentrations were similar to the observed TP concentrations in the CA+TP group, proposing that the contribution of TP derived from CA metabolism is negligible to the observed concentrations of TP. Thus, a 1.61-fold increase in the AUC of TP was due to a 0.63-fold decrease in CL/F in the CA+TP group compared to the TP group. This might be due to the inhibition of CYP1A2-mediated TP metabolism by CA considering that both CA and TP were substrates of CYP1A2 (Ginsberg et al. 2004; Yu et al. 2016; Britz et al. 2019; Dai et al. 2022). In other words, CA inhibited CYP1A2-mediated TP metabolism in the CA+TP group.

Before comparing the pharmacokinetic profiles of TP between CA and CA+TP groups, it is necessary to compare the pharmacokinetic profiles of CA as a parent drug of TP after CA administration. The absorption of CA from the gastrointestinal tract was rapid in both of CA and CA+TP groups, and co-administered TP resulted in a 1.57-fold increase in AUC and a 0.64-fold decrease in CL/F of CA in the CA+TP group compared to the CA group. These results suggest that co-administered TP did not affect the absorption rate of CA, but the elimination of CA (mainly through the metabolism of CA) was inhibited by TP, as evidenced by the significantly slower CL/F of CA in the CA+TP group compared to that in the CA group. Especially, the inhibitory effect of TP on CA metabolism was caused by the CYP1A2-mediated metabolism of CA in the liver because it is the main elimination pathway of CA after the complete absorption of CA (Ginsberg et al. 2004; Yu et al. 2016; Britz et al. 2019; Dai et al. 2022).

In the pharmacokinetic profile of TP between CA and CA+TP groups, it is natural that there was a statistically significant difference in the pharmacokinetic parameters of TP between the two groups because both the orally administered TP and TP derived from co-administered CA were added, which is reflected in the pharmacokinetic profile of TP in CA+TP group. However, the similar TP concentrations in the TP and CA+TP groups mentioned above (i.e., litter contribution of TP derived from CA to the disposition of TP) make it clear that the contribution of TP derived from CA was negligible to produce the observed concentration of TP in the CA+TP group. Practically, the elimination pattern of TP between CA and CA+TP groups was comparable supported by the comparable t_1/2 values. The C_max and T_max values (30.9 µg/mL at 15 min) in the CA+TP group were also comparable to those (29.8 µg/mL at 15 min) in the TP group, indicating that the TP concentrations in the CA+TP group mainly originated from the orally administered TP concentrations, not from the TP derived from CA. Although the contribution of TP on CA metabolism is a little and the degree of the inhibitory effect of TP on CA metabolism in the CA+TP group, the inhibitory effect of TP on CA metabolism might exist as previously reported (Yu et al. 2016; Chen et al. 2017): a 47.9-fold increase in AUC despite a 0.05-fold decrease in CL/F of TP in CA+TP group as well as the increased ratio of AUC_TP/AUC_CA from 9.30% in CA group to 277% in CA+TP group were observed. These results suggest that a decrease in the AUC of CA was greater than a decrease in the AUC of TP, which might be because of the inhibition of TP metabolism to form further metabolites, as a secondary metabolic pathway of CA, occurring through CYP1A2, as well as inhibiting CA metabolism to TP (Yu et al. 2016; Chen et al. 2017).

In the pharmacokinetic profile of PA, the formation of PA from CA metabolism started immediately after CA absorption, which might have been slowed down by co-administered TP in the CA+TP group. This could be supported by the delayed time to appear peak of PA on early blood sampling time point in the CA+TP group compared to the CA group. In addition, the decreased ratio of AUC_PA/AUC_CA in the CA+TP group indicated the metabolism of CA to PA was significantly reduced by co-administered TP compared to the TP group. In the pharmacokinetic profiles of TB, co-administered TP slowed down the formation of TB derived from CA metabolism, and the decreased ratio of AUC_PA/AUC_CA was observed in the CA+TP group compared to the CA group. These similarly changed patterns appeared in the formation of PA and TB from CA metabolism by TP in the CA+TP group suggested that TP might inhibit CA metabolism to form PA and CA. Especially, the decreased ratios of AUC_PA/AUC_CA, an indicator of CYP1A2 inhibition (Ginsberg et al. 2004; Yu et al. 2016; Britz et al. 2019; Dai et al. 2022), critically proposed that co-administered TP inhibited CYP1A2-mediated metabolism of CA.

In conclusion, the potential of pharmacokinetic interaction of CA and TP is proposed because of their overlapping CYP1A2-mediated metabolism being inhibited by their co-administration. Thus, the pharmacological interaction between CA and TP is required to be investigated for the safe co-administration of CA and TP.

This research was funded by the National Research Foundation of Korea in the Korean government (MSIT) [NRF-2016R1C1B2010849, NRF-2018R1A5A2023127, and NRF-2021R1A2C1094462 (Y.H.C.)].

Supplementary Materials can be found via https://doi.org/10.58502/DTT.23.0024.

The authors declare that they have no conflict of interest.