^{1}

^{1}

^{2}

^{2}

^{2}

^{1}

^{*}

^{1}

^{2}

Edited by: Chonlaphat Sukasem, Mahidol University, Thailand

Reviewed by: Volker Martin Lauschke, Karolinska Institute (KI), Sweden; Mirko Manchia, Dalhousie University, Canada

*Correspondence: Christian Maltecca

This article was submitted to Pharmacogenetics and Pharmacogenomics, a section of the journal Frontiers in Genetics

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

In livestock, the regulation of drugs used to treat livestock has received increased attention and it is currently unknown how much of the phenotypic variation in drug metabolism is due to the genetics of an animal. Therefore, the objective of the study was to determine the amount of phenotypic variation in fenbendazole and flunixin meglumine drug metabolism due to genetics. The population consisted of crossbred female and castrated male nursery pigs (

The regulation of drugs used to treat livestock has received increased attention due to animal welfare concerns, food safety, and implications of antibiotic resistance on human health (Landers et al.,

The resource population initially investigated four different drugs that involve both phase 1 and 2 drug reactions and are commonly used to treat livestock. A preliminary investigation that is described in Howard et al. (

This study was approved by the NCSU Institutional Animal Care and Use Committee (IACUC) and all experiments were performed in accordance with relevant guidelines and regulations.

The resource population that was utilized to estimate genetic parameters was developed to investigate the phenotypic and genetic variability related to drug metabolism in swine (Howard et al.,

The sires of the pigs were registered National Swine Registry sires mated to sows (Yorkshire X Landrace) at the North Carolina State University Swine Education Unit. A subset of the female and castrated male offspring produced by each mating pair was selected for the study. The sires utilized were either purebred Duroc (

The experimental design for batches 1–4 was a random crossover design and is described in detail by Howard et al. (

Descriptive statistics across batches.

^{a} |
||||
---|---|---|---|---|

1 | 7 | 34.15 ± 4.67 | 6 | 2 |

2 | 12 | 54.85 ± 11.13 | 9 | 10 |

3 | 15 | 31.66 ± 5.18 | 13 | 12 |

4 | 20 | 40.66 ± 5.88 | 17 | 18 |

5 | 29 | 26.03 ± 4.14 | 16 | 13 |

6 | 29 | 26.82 ± 3.78 | 13 | 16 |

7 | 27 | 33.67 ± 4.82 | 13 | 14 |

8 | 28 | 33.13 ± 5.29 | 15 | 13 |

9 | 31 | 34.06 ± 5.11 | 12 | 19 |

Fenbendazole and flunixin meglumine were administered intravenously (IV). The IV administration was utilized in order to remove inter- and intra-individual variability observed with extravascular administration routes (Petersen and Friis,

Heritability estimates for drug metabolism parameters were obtained based on traditional PK parameters or actual parent drug and metabolite concentrations across time. The latter heritability represents the heritability of the concentration of drug in the plasma across time, while the former represents the heritability of parameters that describe the concentration of drug in the plasma across time. The PK parameters are the traditional metric utilized in veterinary medicine to describe the rate at which the parent drug and metabolite are circulating in the plasma and are the ones that have been traditionally employed in pharmacogenetics analyses. Given the relatively large resource population in the current study, more complex random regression models were utilized to better describe the concentration of the parent drug or metabolite across time.

The PK parameters were calculated utilizing a traditional non-compartmental analysis of drug plasma concentration vs. time profiles. The analysis was conducted using the pharmacokinetic modeling Phoenix software (version 1.1; Pharsight, Cary, NC, USA). The PK parameters calculated for the parent drug included: area under the plasma concentration-time curve from time zero to infinity (AUC_{0 → ∞}; h^{*}μg/mL), clearance (Cl; L/h/kg), half-life (T_{1/2}; h), mean residence time (MRT; h), and volume of distribution at steady state (Vd_{ss}; L/kg). The PK parameters calculated for the metabolite included: AUC_{0 → ∞}, peak concentration (C_{max}; μg/mL), and time at which maximum concentration occurs (T_{max}; h). Summary statistics on the PK parameters across both drugs for the parent drug and metabolite are outlined in Table

where _{ijklm} represented the PK parameter, μ was the average PK parameter, _{i} was the fixed effect of ^{th} sex, _{j} was the fixed effect of the ^{th} breed of the sire, _{k} was the random effect of the ^{th} batch, sire_{l} was the random effect of the ^{th} sire and _{ijklm} was the random residual with a homogenous variance structure. The residuals were weighted according to the R-squared value of the model for each individual animal. The batch and sire effects were assumed to follow a normal distribution that was identically and independently distributed. The genetic relationships among sires was investigated by tracing back a pedigree for all sires and was found to be near zero in all cases. Therefore, the sires were assumed to be unrelated. Lastly, the impact of body weight, which was used to determine the dose, was investigated and found to be insignificant (^{2}) was estimated according to:

The precision of the heritability estimate, indicated by its standard error, was evaluated in order to determine if the estimation error surrounding the heritability (i.e., h^{2} ±

Observed average (± ^{a}

^{b} |
_{1/2} (h) |
_{0 → ∞} (h^{*}μg/mL) |
_{ss} (L/kg) |
_{max} (h) |
_{max} (μg/mL) |
|||
---|---|---|---|---|---|---|---|---|

FBZ | Parent | 15.14 ± 12.43 | 5.62 ± 4.67 | 15.33 ± 14.7 | 0.27 ± 0.22 | 2.93 ± 1.85 | – | – |

Metabolite | – | 7.02 ± 1.97 | – | – | – | 3.28 ± 0.69 | 0.47 ± 0.09 | |

FLU | Parent | 6.62 ± 2.32 | 28.04 ± 8.27 | 3.85 ± 1.46 | 0.12 ± 0.04 | 0.43 ± 0.18 | – | – |

Metabolite | – | 0.83 ± 0.48 | – | – | – | 0.51 ± 0.08 | 0.21 ± 0.10 |

_{1/2}; h), clearance (Cl; L/h/kg), area under the plasma concentration-time curve from time zero to to infinity (AUC_{0 → ∞}; h^{*}μg/mL), mean residence time (MRT; h), volume of distribution at steady state (Vd_{ss}; L/kg), peak concentration (C_{max}; μg/mL), and time at which maximum concentration occurs (T_{max}; h)

The drug concentration curve heritability for the parent drug and metabolite within each drug was estimated utilizing plasma concentrations across time for each individual. The raw drug concentration across time for the parent drug and metabolite is summarized in Figures

where _{ijklmnt} represents the concentration of the parent drug or metabolite, ϕ_{mtn} is Legendre polynomial_{n} for the ^{th} animal at the ^{th} hour since the drug was administered, β_{n} are the fixed Legendre polynomial regression coefficients, _{i} was the fixed effect of ^{th} sex, _{j} was the fixed effect of the ^{th} breed of the sire, _{k} was the random effect of the ^{th} batch, _{ln} are the random Legendre polynomial regressions for the ^{th} sire and _{ijklmnt} was the random residual with a heterogeneous variance structure by ^{th} hour since the drug was administered. The order of the Legendre polynomial for the fixed effect (nf) was 6 for fenbendazole across both compounds and 7 for flunixin meglumine across both compounds. The Legendre polynomial for the random sire regression effect (nr) included only the intercept (i.e., order = 0) for all models except for the fenbendazole metabolite, which included an intercept and slope (i.e., order = 1). Lastly, across both compounds, the residual covariance structure was independent across time points within an individual and across animals for fenbendazole and had an autoregressive 1 (AR1) structure within an animal and independent across animals for flunixin meglumine.

Observed concentration of the parent drug and metabolite across time for fenbendazole. Light gray lines represent individual animals and the solid black line represents the mean across all animals.

Observed concentration of the parent drug and metabolite across time for flunixin meglumine. Light gray lines represent individual animals and the solid black line represents the mean across all animals.

Similar to the PK heritability analysis, the sire variance estimate was transformed algebraically into direct additive genetic effects _{t}

where _{t} is equal to the t^{th} row vector of ϕ and

Lastly, the change in heritability for the fenbendazole metabolite model across time points is a function of not only the change in sire variance across time points, but also the change in residual variance.

The heritability (± SE) estimates for PK parameters for the parent drug and metabolite for fenbendazole and flunixin meglumine are outlined in Table ^{2} ± _{max} for the metabolite for flunixin meglumine was outside the bounds and therefore was not shown. Least-squares means for fenbendazole and flunixin meglumine and its associated metabolite by breed and sex are outlined in Tables _{0 → ∞} for the parent drug of flunixin meglumine. Similarly, the impact of breed on PK parameters never reached the significance level for fenbendazole, but was trending toward significance for MRT for the parent drug flunixin meglumine.

Heritability (± SE) across pharmacokinetic (PK) parameters by drug.

^{a} |
|||
---|---|---|---|

Drug | T_{1/2} |
0.16 ± 0.36 | 0.20 ± 0.37 |

Cl | 0.01 ± 0.26 | 0.42 ± 0.39 | |

AUC_{0 → ∞} |
0.13 ± 0.38 | 0.32 ± 0.35 | |

MRT | 0.16 ± 0.38 | 0.04 ± 0.28 | |

Vd_{ss} |
0.28 ± 0.44 | 0.58 ± 0.44 | |

Metabolite | AUC_{0 → ∞} |
0.36 ± 0.41 | 0.61 ± 0.42 |

C_{max} |
0.13 ± 0.25 | 0.18 ± 0.25 | |

T_{max} |
0.06 ± 0.36 | – |

_{0 → ∞}; h^{*}μg/mL), mean residence time (MRT; h), volume of distribution at steady state (Vdss; L/kg), peak concentration (Cmax; μg/mL), and time at which maximum concentration occurs (Tmax; h)

The heritability (± SE) across time points for the parent drug and metabolite for fenbendazole and flunixin meglumine is outlined in Table

Heritability (± SE) across time since the drug was administered for fenbendazole and flunixin meglumine parent drug and metabolite concentrations.

^{a} |
|||||
---|---|---|---|---|---|

^{b} |
|||||

0.5 | 0.451 ± 0.20 | 0.732 | 0.5 | 0.193 ± 0.10 | 0.312 ± 0.14 |

1 | 0.195 ± 0.09 | 0.611 | 1 | 0.118 ± 0.06 | 0.470 ± 0.21 |

2 | 0.097± 0.05 | 0.675 | 2 | 0.167 ± 0.09 | 0.443 ± 0.20 |

3 | 0.110± 0.05 | 0.549 | 3 | 0.162 ± 0.09 | 0.249 ± 0.12 |

4 | 0.311± 0.14 | 0.479 | 4 | 0.122 ± 0.06 | 0.466 ± 0.21 |

6 | 0.181± 0.08 | 0.599 | 8 | 0.125 ± 0.07 | 0.525 ± 0.23 |

12 | 0.333± 0.15 | 0.355 | 12 | 0.081 ± 0.04 | 0.468 ± 0.21 |

– | – | – | 16 | 0.142 ± 0.08 | 0.487 ± 0.22 |

24 | 0.225± 0.10 | 0.783 | 24 | 0.106± 0.06 | 0.550 ± 0.24 |

48 | 0.527± 0.23 | – | 48 | 0.200 ± 0.11 | 0.420± 0.19 |

The current study has utilized PK parameters or plasma concentrations of the drug or metabolite across time to determine the amount of phenotypic variation explained by genetics. To our knowledge heritability estimates on drug metabolism parameters have not been estimated for any drug in livestock populations, although multiple twin—(Matthaei et al.,

In humans, the reported proportion of phenotypic variation for a PK parameter explained by genetics across multiple drugs ranges from minimal (Kalow et al.,

Given the large sample size of the current study, models that characterize the drug concentration curve across time were also investigated. The average heritability estimate for the parent drug across time for the drug concentration curve was similar (i.e., 0.20) to the point estimate derived from the PK model. The average heritability estimate for the metabolite across time was 0.51, which was higher than the point estimate derived from the PK model. More importantly, the heritability estimation error (i.e., h^{2} ±

The initial study by Howard et al. (_{ss}) and oxfendazole (_{0 → ∞}). Given the larger sample size and a different objective in the current study, an additional sire effect was included in the model compared to that used in Howard et al. (

As discussed in Howard et al. (

The current study estimated the proportion of phenotypic variation in metabolizing fenbendazole and flunixin meglumine that is due to additive genetics. Models to estimate heritability based on PK parameters or the observed plasma drug concentration across time were utilized. Across both types of models a moderate heritability was estimated. Furthermore, the model that utilized the plasma drug concentration across time vs. the PK parameters resulted in more precise estimates.

JH performed the statistical analysis. RB, JB, and JY performed the pharmacokinetic analysis. CM, RB, and MA designed and conceived the experiment. JH wrote the first draft of the paper. All authors read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The authors wish to thank the many undergraduate students that helped with animal care and sample collection. Furthermore, the authors would like to thank Patty Roth, Audrey O'Nan, Maria Stone, and Krista Browning. The research was supported by the North Carolina Biotechnology Center (grant #2012-MRG-1108) and Food Animal Residue Avoidance Data- bank (USDA #2013-41480-21002).

The Supplementary Material for this article can be found online at: