In Vitro Rat Hepatic Metabolism of n-Alkanes: Nonane,Decane,and Tetradecane

Abstract

Jet propellant 8 (JP-8) jet fuel is a complex mixture of aromatic and aliphatic hydrocarbons. The aim of this study was to determine in vitro metabolic rate constants for semivolatile n-alkanes, nonane (C9), decane (C10), and tetradecane (C14), by rat liver microsomal oxidation. The metabolism was assessed by measuring the disappearance of parent compound by gas chromatography. Various concentrations of n-alkanes were incubated with liver microsomes from adult male F-344 rats. Nonlinear kinetic constants for nonane and decane were V _max (nmol/mg protein/min) = 7.26 ± 0.20 and 2.80 ± 0.35, respectively, and K _M (μM) = 294.83 ± 68.67 and 398.70 ± 42.70, respectively. Metabolic capacity as assessed by intrinsic clearance (V _max/K _M) was ~four-fold higher for nonane (0.03 ± 0.005) than for decane (0.007 ± 0.001). There was no appreciable metabolism of tetradecane even with higher microsomal protein concentration and longer incubation time. These results show a negative correlation between metabolic clearance and chain length of n-alkanes. These metabolic rate constants will be used to update existing physiologically based pharmacokinetic (PBPK) models for nonane and decane as part of developing a PBPK model for JP-8.

Jet propellant-8 (JP-8) fuel is the standard jet fuel used for military aircraft and ground vehicles in NATO (North Atlantic Treaty Organization) nations, with an annual use of approximately 5 billion gallons (Perleberg, Keys, and Fisher 2004; TIEHH 2001). JP-8 is a complex mixture including straight-chain alkanes, branched-chain alkanes, cycloalkanes, and napthalenes. Because of its widespread use, occupational exposure of military personnel to JP-8 has become a health concern to the military. In animal studies, JP-8 is reported to affect the immune, hepatic, central nervous, and respiratory systems (NRC 2003).

In assessing the risk of complex mixtures such as JP-8, it is not feasible to study the effects and disposition for each of the components. Instead, gaining knowledge on representative chemicals and using physiologically based pharmacokinetic (PBPK) modeling is likely to provide insight into JP-8 exposure and potential health risks (Robinson 2000). PBPK models are mathematical representations of the pharmacokinetic behavior of chemicals in the body. These models are increasing in use for describing the dosimetry of chemicals and estimating health risks. Experimentally derived metabolic rate constants are important for the development of PBPK models because metabolism may control the systemic clearance of the chemical or describe the rate of formation of a toxic metabolite.

Although several PBPK models have been developed for aromatic hydrocarbons found in jet fuel, PBPK models for semivolatile n-alkanes, an important class of hydrocarbons found in neat JP-8 fuel (Zeiger and Smith 1998) and aerosolized jet fuel inhalation studies (Dietzel et al. 2005), remain to be developed. The most prominent n-alkanes found in aerosolized JP-8 are n-nonane (C9), n-decane (C10), n-undecane (C11), n-dodecane (C12), n-tridecane (C13), and n-tetradecane (C14). As part of a larger study to develop PBPK models for these n-alkanes, we elected to conduct studies to determine rat hepatic metabolic constants for two n-alkanes (C9 and C10), which are known to be metabolized, and one n-alkane (C14) for which no metabolism information was found and is the least volatile of this fraction. Also, nonane (C9) and decane (C10) are considered representative chemicals of the C9–C14 class (Perleberg, Keys, and Fisher 2004; Robinson 2000).

Several urinary metabolites of nonane have been identified in the rat orally dosed with nonane (Serve et al. 1985). Using male mice microsomes, decane has been shown to be metabolized to three metabolites (Ichihara, Kusunose, and Kusunose 1969). These authors report a Michaelis-Menten constant (K _M) of 500 μM for hydroxylation of decane. Recently, Edwards, Rose, and Hodgson (2005) reported metabolic constants for formation of 2-nonanol from nonane using human hepatic microsomes (K _M = 62.5 μM and V _max = 429 nmol/min/mg protein) and identified the involvement of cytochrome P450 (CYP) isoforms 1A2, 2B6, and 2E1 in the formation of 2-nonanol. Mortensen et al. (2000) reported metabolic clearance of 25 petroleum hydrocarbons, including n-alkanes of C6 to C10, using rat liver slices by headspace analysis. For some unknown reason, the use of very high metabolic rates of Mortensen et al. (2000) in nonane and decane PBPK models were inconsistent with describing the systemic clearance of C9 and C10 from rats. As a result, the existing rat PBPK models have been developed without using independently determined metabolic rate constant values (C9; Robinson 2000) or not accounting for metabolism (C10; Perleberg, Keys, and Fisher 2004). Thus, the objectives of this research were to conduct metabolism studies with C9 and C10 to determine metabolic constant values for eventual use in updated PBPK models for nonane and decane and to evaluate the potential for tetradecane to be metabolized by rat microsomes in vitro.

MATERIALS AND METHODS

Chemicals

Nonane, decane, and tetradecane of 99%+ pure were obtained from Sigma Chemical (St. Louis, MO). Methanol and methylene chloride were purchased from Fisher Scientific (Pittsburgh, PA).

Animals

Male Fischer 344 rats (70 days old) were obtained from Charles River (Raleigh, NC). Rats (two/cage) were housed in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)-approved animal care facility maintained at 72°F ± 2°F and 50% ± 10% humidity with a 12-h light/dark cycle. Food (5001 Rodent Diet; PMI Nutrition International, MO) and tap water were provided ad libitum. Rats were used when they were 90 days old. The experimental protocol was reviewed and approved by the University of Georgia Animal Use and Care Committee.

Preparation of Liver Microsomes

Liver samples were collected from four untreated rats following decapitation. To prepare hepatic microsomes by the procedure of Omura and Sato (1964), 5 g of fresh liver were homogenized in 25 ml of cold 0.02 M Tris-KCl buffer, pH 7.4, using a Teflon/glass tissue homogenizer. The homogenized liver was centrifuged using a Sorvall RC5C centrifuge (DuPont, Newton, CT) for 30 min at 4°C at 12000 ×g. The supernatant was collected and centrifuged using a Beckman Coulter Optima XL-100K ultracentrifuge (Fullerton, CA) at 4°C at 105,000 ×g for 60 min. The pellet containing the microsomes was resuspended in 5 ml Tris-KCl, pH 7.4, buffer and frozen at −80°C. Protein content of the microsomes was measured by the method of Lowry et al. (1951).

Metabolism of n-Alkanes

The kinetics of in vitro metabolism of n-alkanes was evaluated by measuring the disappearance of parent compound from four individual liver microsomes (n = 4) collected from different rats. Initial experiments were conducted for each compound to establish the optimal incubation conditions. Disappearance of n-alkanes was linear with respect to time as well as to microsomal protein over concentration ranges relevant to this study. Note that separate incubations were performed for each n-alkane. The velocity (V _o ) of the reaction was expressed as nmol n-alkane disappearance/min/mg protein. Affinity Km was expressed as μM.

Various concentrations of nonane (20 to 4000 μM final concentration), decane (10 to 3500 μM final concentration), and tetradecane (13 to 1300 μM final concentration) were incubated with microsomal protein with and without NADPH (1.2 mM, final) at 37°C. Whereas nonane and decane were incubated with 1 mg protein/940 μl (diluted using Tris-HCl buffer, pH 7.4) for 30 and 60 min, respectively, tetradecane was incubated with 1 to 5 mg protein/940 μl for 30 to 120 min. The reaction was initiated by adding microsomes (940 μl) to mixture of 50 μl NADPH or buffer and 10 μl of various substrate concentrations (in methanol, never more than 1% of total volume). After the incubation time, 2 ml of methylene chloride were added to stop the reaction and to extract n-alkanes. Incubations were performed in sealed tubes to minimize any loss of parent compound. Buffer and heat-inactivated microsome controls were conducted to account for background loss and nonspecific binding.

Extraction and GC-FID Analyses

After the addition of methylene chloride, the tubes were vortexed for 5 min and centrifuged for 10 min at 2500 ×g. Two microliters of supernatant were injected onto the gas chromatography (GC) column.

The concentrations of n-alkanes were measured according to the method of Smith et al. (2005) with some modifications. The GC-Flame Ionization Detector (FID) system used was Agilent 6890 Series II (Milford, MA) with a HP-5 column (30 m × .32 mm × 0.25 μm) and an autosampler. The injector and detector temperatures were 200°C and 260°C, respectively for nonane, decane, and tetradecane. For nonane and decane, initial oven temperature was 35°C for 3 min and then increased to 65°C using 5°C/min increments followed by 30°C/min increments until 300°C. For tetradecane, initial oven temperature was 45°C for 3 min and then increased to 160°C using 15°C/min increments followed by 40°C/min increments until 300°C. The nitrogen, hydrogen, and airflows for all three compounds were 7.5, 37, and 370 ml/min. The limit of detection for these n-alkanes was 0.1 μg/ml.

Data Analysis

Data are presented as mean and standard error (SE) of four independent microsomal samples from 4 different animals. The metabolic rate constants (V _max and K _M) for parent disappearance were calculated by nonlinear regression using Prism (3.03). Student t test was used to compare the statistical significance (p < .05) in kinetic parameters between C9 and C10 using Prism (3.03).

RESULTS

There was no marked difference in n-alkane levels between buffer controls and corresponding standards, indicating that loss of these compounds to headspace due to volatilization at 37°C is negligible. Similar degree of loss of n-alkanes (~10%) was observed between heat-inactivated controls and without NADPH, suggestive of nonspecific binding.

The rate of nonane oxidation by liver microsomes was linear up to 400 μM, and plateaued at higher concentrations (Fig. 1). The Michaelis-Menten rate constants for the microsomal metabolism of nonane were calculated using nonlinear regression, V _max = 7.26 ± 0.20 nmol/mg protein/min and K _M = 294.83 ± 68.67 μM. The intrinsic clearance (V _max/K _M) was 0.03 ± 0.005 ml/min/mg protein.

Fig. 2 presents the rate of decane disappearance by liver microsomal oxidation. The metabolism of decane was linear up to 700 μM, followed by saturation at higher concentrations. The Michaelis-Menten rate constants for the microsomal metabolism of decane were calculated using nonlinear regression, V _max = 2.8 ± 0.35 nmol/mg protein/min and K _M = 398.7 ± 42.7 μM. The intrinsic clearance (V _max/K _M) was 0.007 ± 0.001 ml/min/mg protein.

There was no appreciable oxidative metabolism of tetradecane in liver microsomes over the concentration range of 13 to 1300 μM. Even with increased protein concentration from 1 to 5 mg protein or longer incubation time from 30 to 120 min, tetradecane metabolism was not apparent in liver microsomes (data not shown).

DISCUSSION

Aliphatic hydrocarbons of C9–C14 chain length make up a significant proportion of neat JP-8 (Zeiger and Smith 1998) and inhalation chamber atmospheres used to conduct aerosolized JP-8 toxicity studies (Dietzel et al. 2005). In the interest of constructing PBPK models, the present study reports comparative oxidative metabolism of semivolatile C9, C10, and C14 n-alkanes in rat liver microsomes and compares the results with the previously reported values of lower chain n-alkane, hexane.

The present results showed that the metabolism decreases with increasing carbon length. Tab. 1 compares the hepatic metabolic rate constants of C9, C10, and C14 n-alkanes. Whereas K _M values were not statistically different between nonane and decane, the V _max value of nonane was about threefold higher than decane. Correspondingly, intrinsic hepatic clearance rate of nonane was about four times greater than decane. There was no observed metabolism of tetradecane by liver microsomes under present experimental conditions.

To better show the relationship between metabolism and chain length, we attempted to compare our findings with published values for lower chain length n-alkanes (<C9). Interestingly, we only found metabolic constants for n-hexane (Ali and Tardif 1999; Dennison, Andersen, and Yang 2003; Perbellini et al. 1990). Intrinsic clearance of n-hexane (C6; Perbellini et al. 1990) was compared with the results of C9, C10, and C14 reported in this study. The metabolic rate constants reported by Perbellini et al. (1990) were V _max 2.8 nmoles/g/min and K _M 1.7 μM. The units for V _max and K _M in the present study were nmoles/min/mg protein and μM. For the purpose of comparison, the unit of V _max of this study was converted to nmoles/g/min. Because the liver microsomes in this study were resuspended in a volume (5 ml) of buffer same as the original volume (5 g) of the liver, enzyme activities measured in each milliliter of suspension could be expressed per gram liver. The average microsomal protein concentration was 20 mg/ml. Hence, the V _max values were multiplied by 20 to get nmoles/g/min. The intrinsic clearance values (ml/min/g) as calculated by V _max/K _M for C6, C9, and C10 are presented in Fig. 3. The intrinsic clearance of C6 (1.7) was about threefold higher than C9 (0.5), which is fivefold higher than C10 (0.1), followed by no observable metabolism with C14. These metabolic calculations suggest a reciprocal relationship between carbon number and metabolism.

Although difficult to compare directly,Itchihara, Kusunose, and Kusunose (1969) reported K _M of 500 μM for decane by mouse liver microsomes, which is similar to the present findings with rat microsomes (400 μM). Similar to the present study, decane was incubated with mouse liver microsomes and extracted by liquid-liquid extraction.Itchihara, Kusunose, and Kusunose (1969) did not report V _max values. On the other hand, the metabolic rate constants of Mortensen et al. (2000) for C9 and C10 using rat liver slices by headspace analysis were inconsistent with our findings with rat liver microsomes and liquid-liquid extraction. They reported intrinsic clearance (V _max/K _M) values for C9 and C10 many fold greater than values reported in the present study. They also observed a positive correlation between metabolism of C6–C10 n-alkanes and chain length as opposed to negative or inverse correlation seen in the present study. The exact reason(s) for this discrepancy is unknown. One reason could be the loss of n-alkanes to glass in the headspace vials occurred in a similar fashion as in gas-uptake chamber (Perleberg, Keys, and Fisher 2004), leading to an erroneously high estimate of metabolic capacity.

In summary, this study provides information on the ability of rat liver microsomes to metabolize three common n-alkanes found in jet fuel. These metabolic constant calculations will assist in the implementation of updated PBPK models for nonane and decane as part of a JP-8 PBPK model and provide evidence for negligible hepatic metabolic clearance of tetradecane in rats. It should be noted that extrahepatic metabolism and factors other than metabolism might play role in the clearance of n-alkanes from the system. Information on these processes can be incorporated in the models, as they are available. In fact, PBPK models can be useful in identifying key input parameters.