In vitro test systems to determine tetracycline residue binding to human feces

Abstract

The use of antimicrobials, such as tetracycline, in food-producing animals may result in antimicrobial drug residues (ADR) in edible tissues from treated animals and contribute to the emergence of antibiotic resistant bacteria. The Veterinary International Conference on Harmonization (VICH) document (VICH GL36(R)/FDA-CVM Guidance for Industry#159) provides guidance on evaluating the safety of veterinary ADR in the human foods as related to effects on the human intestinal microbiome. One recognized research gap is a need for additional data and testing requirements to determine the fraction of an oral dose of ADR available to intestinal microorganisms. In the present study, we address this need by examining the binding of tetracycline to human feces using chemical and microbiological assays. High-performance liquid chromatography and liquid chromatography mass spectrometry assays showed that 25% (w/v) diluted steam sterilized feces dosed with 0.15 and 1.5 μg/ml tetracycline had binding of 58.2 ± 10.8% and 56.9 ± 9.1%, respectively. Tetracycline binding to fecal slurries gave similar results. Microbiological assays with two reference bacterial strains validated the results of the chemical assays. Based on data from chemical and microbiological assays methods, the fraction of dose available to microorganisms was 0.418 and 0.431 of the 0.15 and 1.5 μg/ml tetracycline treatments, respectively. This study also proposes factors to be considered when designing and conducting experiments to determine the percent of an antimicrobial agents that is available to microorganisms in the gastrointestinal tract.

1. Introduction

The tetracyclines, oxytetracycline (OTC), chlortetracycline (CTC) and tetracycline (TET), belong to a family of broad-spectrum polyketide antimicrobials used in human and veterinary medicine, and are prescribed for many infections caused by aerobic and anaerobic Grampositive and Gram-negative bacteria (Chopra and Roberts, 2001; Michalova et al., 2004). They have been extensively used in humans as therapeutic agents, and in the livestock as additives to feeds to enhance growth of animals and to treat clinical disease (Backhed et al., 2005; Chopra and Roberts, 2001; Daghrir and Drogui, 2013; Ley et al., 2006; Speer et al., 1992; Wagner et al., 2008). Under FDA Guidance for Industry (GFI) # 213, the growth promotion uses of medically important antibiotics are not considered judicious use, thus the tetracyclines growth promotion uses in food producing animals are no longer legal approved uses (FDA, 2017). The use of veterinary antimicrobial agents in food-producing animals may result in antimicrobial drug residues (ADR) in edible tissues and meat products from the treated animals (Martinez et al., 2014). The VICH GL36(R) provides guidance for assessing the human food safety of residues from veterinary antimicrobial drugs regarding effects on the human intestinal microbial community (FDA, 2013). The objectives of this guidance were (1) to outline the recommended steps in determining the need for establishing a microbiological acceptable daily intake (mADI); (2) to recommend test systems and methods for determining no-observable adverse effect concentrations (NOAECs) and no-observable adverse effect levels (NOAELs) for the endpoints of health concern; and (3) to recommend procedures to derive a microbiological ADI (Cerniglia and Kotarski, 1999; Cerniglia and Kotarski, 2005; Cerniglia et al., 2016; FDA, 2013). To provide guidance on the safety of veterinary ADR in the human food supply, tetracyclines has been evaluated by the United States-Food and Drug Administration (US-FDA), the Joint [Food and Agriculture Organization/World Health Organization (FAO/WHO)] Expert Committee on Food Additives (JECFA), and the European Medicines Agency (EMA) to assess its safety with respect to residues in animal-derived foods (EMEA, 1995; FDA, 1996; JECFA, 1999). JECFA determined an acceptable daily intake (ADI) value for tetracycline at 30 μg/kg bw/day; however, US-FDA established a codified ADI value for all tetracyclines at 25 μg/kg bw/day. This is about ten times higher than the ADI set by the EMA (0–3 μg/kg bw/day) (Cerniglia and Kotarski, 2005; EMEA, 1995). All these determined ADIs values preceded the implementation of the harmonized VICH guideline GL#36.

Appendix D of VICH GL36(R) provided strategies for in vivo and in vitro testing methods to determine the fraction of oral dose available to intestinal microbiota (Cerniglia et al., 2016; FDA, 2013; FDA, 2017). Research to determine drug inactivation and binding to fecal matter has been reported earlier (Ahn et al., 2012a; Cerniglia et al., 2016; Edlund et al., 1988, 1994; Hazenberg et al., 1983, 1984; Jansen et al., 1992; Vansaene et al., 1985; Veringa and Vanderwaaij, 1984; Wagner et al., 2008). These investigations mostly used therapeutic human doses of antimicrobial agents rather than residual levels in the study design. Previously, we examined the low residual concentration levels of enrofloxacin bound to fecal content by microbiological and analytical chemistry assay methods (Ahn et al., 2012a). We found that 10% and 25% diluted steam sterilized fecal samples dosed with enrofloxacin showed binding of 50 ± 4.6% and 54 ± 6.5% of the enrofloxacin, respectively. Our earlier results concluded that the degree of binding of enrofloxacin to fecal material and inactivation of the drug varied depending upon chemical structure, concentration of the drug added, and physicochemical properties of the antimicrobial class used in a sorption/desorption of fecal slurry analysis. However, little is known about binding of tetracyclines to human fecal contents.

Therefore, in this study we investigated the binding of tetracycline to human fecal slurries using in vitro test systems to determine the fraction of oral dose available for intestinal microorganisms following three steps as illustrated in Fig. 2. First, we optimized the experimental conditions regarding the drug concentration to be used in the study and the fecal parameters. Secondly, we calculated the fraction of oral dose available to microorganisms based on in vitro chemical assay. Thirdly, we confirmed the fraction of oral dose of tetracycline available to microorganisms based on in vitro microbiological methods. The approach used in this study is similar to one of the approaches recommended in Appendix D of the VICH GL 36 (R) guideline. Our intention of this investigation, besides presenting data on tetracycline binding to fecal slurries, was to evaluate the testing scheme listed as Approach 2 of Appendix D VICH GL 36 (R) and provide additional guidance on conducting in vitro testing methods to determine the fraction of oral doses available to microorganisms.

Fig. 2.

Overview of the in vitro test systems approach to determine the fraction of oral dose available to microorganisms. This process is composed of three steps: Experimental conditions and setup (Step 1), in vitro chemical assay (Step 2) and in vitro microbiological assay (Step 3). In this study, the fraction of the dose available for colonic microorganisms was calculated by an in vitro chemical assay (Step 2) and confirmed by an in vitro microbiological assay (Step 3).

2. Materials and methods

2.1. Human fecal specimens

The use of human fecal samples was approved by the FDA Research Involving Human Subjects Committee (RIHSC #14–061T). The human fecal samples were collected (self-sampled) from three healthy subjects, who had not been treated with any antibiotics for a minimum of 6 months and samples were transferred to the anaerobic chamber hood for microbiological processing. The collected samples were diluted with Brain Heart Infusion (BHI; Difco Lab., Detroit, MI. USA) broth (Ahn et al., 2012a) to a final concentration of 50% (w/v). The samples were mixed by magnetic stirring for 30 min to produce homogeneous slurries. The fecal slurries were further diluted with BHI to produce 3, 10 and 25% (w/v) dilutions. The fecal specimens were typically used on the day of collection.

2.2. Establishment of experimental conditions for tetracycline concentration and fecal parameters used in the in vitro tetracycline binding studies

2.2.1. Designed test study of tetracycline concentration

US FDA has set a codified ADI value for tetracyclines at 25 μg/kg bw/day (equivalent to 1.5 mg/60 kg person/day) (Cerniglia et al., 2016). Earlier, van Marwyck (1958) showed that 1.5 mg/day would produce a fecal level of 0.15 μg/g (equivalent to 0.15 μg/ml) (Carman et al., 2005; van Marwyck, 1958). This dose level of 0.15 μg/ml corresponds to the United States ADI value of 25 μg/kg bw/day. Therefore, we tested 0, 0.15, 1.5, 15 and 150 μg/ml equivalent to doses of 0, 1.5, 15, 150 and 1500 mg/60 kg person/day. Tetracycline stock solution (1 mg/ml in water; Sigma-Aldrich, St. Louis, MO. USA) was added to sterile serum bottles (in triplicate) to final concentrations from 0.15 to 150 μg/ml.

2.2.2. Tetracycline stability test in fecal supernatant

The 25% (w/v) fecal samples were mixed by magnetic stirring to produce homogeneous slurries (Ahn et al., 2012a). The fecal slurries were sterilized by autoclaving at 121 °C for 15 min on three consecutive days. The sterile fecal slurries were centrifuged at 10,000 g for 20 min and the supernatants were filtered (0.2 μm, 25 mm, Millipore, Billerica, MA, USA). The filtered supernatants were transferred to new tubes for further use in experiments. As the 25% (w/v) fecal extract is opaque and its OD is higher than the limit of detection by spectrophotometer, it was diluted in BHI to a final concentration of 3% (w/v), suitable for monitoring its effect on the growth of bacteria (Ahn et al., 2012b). Triplicate bottles containing a 3% (w/v) sterile fecal supernatant or 0.1% formic acid in BHI were established in volumes of 25 ml in 50 ml amber serum bottles. Each amber serum bottle, containing 0.1% formic acid or 3% fecal supernatant was dosed with 30 μg/ml of tetracycline. The 0.1% formic acid served as an abiotic transformation controls. Serum bottles were incubated quiescently in the dark at 37 °C for 9 days. For tetracycline sampling, the bottles were shaken thoroughly, and 1 ml of sample was withdrawn at 0, 1, 2, 3, 7, 8 and 9 days. The samples were filtered through a 13-mm nylon syringe filter (0.45 μm pore size) before high-performance liquid chromatography (HPLC, Agilent 1200 series, Wilmington, DE). Tetracycline and its metabolites in supernatants were determined using HPLC without extraction as described below.

2.2.3. Comparing sterile fecal slurries and fecal slurries

Using sterile fecal slurries in the tetracycline binding assay provided the experimental milieu, representative of in vivo conditions devoid of survived/active replicating bacterial population and to minimize growth of bacteria on plating media used in the tetracycline bioassay. To compare the difference in percent binding between sterilized fecal slurries versus fecal slurries, 25% (w/v) diluted fecal slurries from three human volunteers were used. Twenty-four bottles containing fecal slurry were established under anaerobic conditions. Among them, fecal slurries were sterilized in twelve amber serum bottles by autoclaving at 121 °C for 15 min on three consecutive days. Sterilized fecal slurries and fecal slurries were treated in triplicate with four different concentrations of tetracycline. Tetracycline was added from deoxygenated stock solutions in methanol to a final concentrations of 15, 20, 30, and 50 μg/ml. Serum bottles were incubated quiescently in the dark at 37 °C for 24 h. Aliquots of these incubations were removed at 2 h, and the fecal slurries were separated into supernatants and pellets by centrifugation at 10,000 × g for 20 min. The supernatants were filtered through a 13-mm nylon syringe filter (0.45-μm pore size) and evaluated for tetracycline concentration by HPLC and microbiological bioassay.

2.2.4. Comparing fecal concentration

The dilute fecal samples (25%, w/v) were homogenized using a magnetic stirrer. The fecal slurries were diluted with CO₂ saturated BHI diluent to produce 10% and 3% (w/v) dilutions of fecal slurries. The fecal slurries (3%, 10% and 25%, w/v) in amber serum bottles were sterilized by autoclaving at 121 °C for 15 min on three consecutive days. Then, each serum bottle with the fecal slurries was dosed with 0.15, 1.5, 15, or 150 μg/ml of deoxygenated tetracycline. After centrifugation at 10,000 × g for 20 min, the resulting supernatants were evaluated for higher tetracycline concentration by HPLC (15 and 150 μg/ml) and lower tetracycline concentrations by LC-MS/MS (0.15 and 1.5 μg/ml) analyses.

2.2.5. Comparing incubation time

The effect of culture incubation time on drug binding was quantitatively assessed by incubating the 25% (w/v) steam sterilized fecal slurries with CO₂ saturated BHI at 37 °C for 1, 2 and 6 h with either 15 or 150 μg/ml of deoxygenated tetracycline added. Aliquots were removed at 1, 2, and 6 h, and the supernatants were evaluated for tetracycline concentration by HPLC.

2.2.6. Comparing tetracycline concentration

To find the maximum tetracycline concentration on drug binding to 25% (w/v) steam sterilized fecal slurries, the fecal slurries with CO₂ saturated BHI were dosed with either 15, 20, 30, 50, 70, 100, or 150 μg/ml of deoxygenated tetracycline. Aliquots were removed at 2 h and the supernatants were evaluated for tetracycline concentration by HPLC.

2.3. In Vitro chemical assays

2.3.1. HPLC and LC-MS/MS analysis for tetracycline quantification

The supernatants were evaluated for tetracycline concentration by HPLC (15, 20, 30, 50, 70, 100, and 150 μg/ml) and LC-MS/MS (0.15 and 1.5 μg/ml). Tetracycline was analyzed in aqueous phase by HPLC with a C-18 Gemini column (4.6 × 250 mm; 5 μm, 110 Å) (Phenomenex, Torrance, CA) with ultraviolet detection at 365 nm. The initial mobile phase composition was 100% mobile phase A (99.9:0.1, water:formic acid) at a flow rate of 0.5 ml/min. This was changed to 100% A–0% B (99.9:0.1, acetonitrile:formic acid) over 40 min with a linear gradient. The lower limit of quantification for tetracycline-added fecal slurries samples was 6.9 μg per ml.

For LC-MS/MS analysis, tetracycline-d₆ (Santa Cruz Biotechnology, Dallas TX) was spiked into the fecal supernatant as an internal standard (final concentration: 10 ng/ml). Tetracycline was assayed by LC-MS/MS using an Agilent 1260 HPLC system coupled to an AB Sciex QTRAP 5500 mass spectrometer using electrospray ionization (ESI). Samples (5 μl) were injected onto a C-18 Kinetex Minibore column (2.1 × 50 mm; 2.6 μm, 100 Å) (Phenomenex, Torrance, CA) and eluted from the column with a gradient flow of 500 μl/min of 0.1% formic acid in water (mobile phase A) and 0.1% formic acid in acetonitrile (mobile phase B). Mobile phase A was held at an initial time of 0.5 min at 95%, then the proportion of mobile phase B was increased linearly to 50% in the following 6 min. At 6.1 min the proportion of mobile phase B was increased to 99%, followed by a hold of 2 min at 99%. At 8.2 min mobile phase B was returned to 5%, and the column was re-equilibrated for 2 min prior to the subsequent injection. The mass spectrometer data were collected in the positive ESI mode using multiple responses monitoring (MRM) (precursor and MS/MS fragment ions obtained are listed in Table 1). The spray voltage was optimized at 5500 V, curtain gas at 10 psi, source heater 650 °C, declustering potential at 70 V, ion source gas 1 at 40 psi, and ion source gas 2 at 30 psi. The autosampler temperature was set to 5 °C to prevent sample degradation during LC-MS/MS analysis.

Table 1.

MS/MS acquisition parameters for tetracycline analysis.

Analyte	Precursor ion, m/z	Product ions, m/z	CE, volts	Retention time, min
Tetracycline	445.1	a154.1, 410.3	35, 19	2.4
Tetracycline d6	451.1	160.2	19	2.4

^aProduct ion (m/z) in bold indicate the ion was used for quantification.

Based on HPLC and LC-MS/MS analysis, tetracycline binding percent was calculated by the equation as defined below:

$$\begin{gathered} \text{Binding}\mathit{percent}(\mathrm{\%})=\frac{\mathit{total}\mathit{added}\mathit{concentration}-\mathit{detected}\mathit{concentration}}{\mathit{total}\mathit{added}\mathit{concentration}} \\ \times 100 \end{gathered}$$

2.3.2. Tetracycline binding percent

The effect of fecal binding on tetracycline availability was quantitatively assessed by incubating tetracycline at concentrations of 0, 0.15, and 1.5 μg/ml with steam sterilized 25% (w/v) diluted steam sterilized fecal slurries at 37 °C for 2 h. The fecal slurries were separated into supernatants and pellets by centrifugation at 10,000 × g for 20 min. The supernatants were evaluated for tetracycline concentration by LC-MS/MS. Based on LC-MS/MS analysis, tetracycline binding percent (%) was calculated and the fraction of the dose available for colonic microorganisms was estimated by the equation below.

$$\begin{gathered} Fractionoforaldoseavailableforcolonicmicroorganisms \\ =1-\left(\frac{\mathit{Binding}\left(\mathrm{\%}\right)}{100}\right) \end{gathered}$$

2.4. In vitro microbiological assays

Escherichia coli ATCC 25922 and Salmonella enterica subsp. Enterica Serovar Gaminara SEA2575 (S. Gaminara SEA2575) were grown on Blood Agar (TSA with 5% sheep blood) (Thermo Fisher Scientific, Lenexa, KS) at 37 °C for 20 h, and then a colony was picked and transferred into 5 ml steam sterilized Mueller-Hinton broth (MHB) (Sigma, St. Louis, MO. USA) (final inoculum of approximately 1.5 × 10⁸ CFU/ml (optical density, OD600 = 0.08–0.1). Tetracycline susceptibility testing was determined by broth microdilution method following the Clinical Laboratory Standards Institute (CLSI) guidelines (CLSI, 2012). The diluent was then used to inoculate 10 ml of fresh MHB at a dilution of 1 in 10 to achieve final cell numbers of approximately 10⁷ CFU/ml and 10⁶ CFU/ml. Twenty μl of tetracycline stock solution and 20 μl suspended cells was added to each well in a 96-well plate with 160 μl of MHB. To determine the effects of media supplemented with sterilized fecal supernatants on the MIC of E. coli ATCC 25922 and S. Gaminara SEA2575 in MHB, 20 μl of tetracycline stock solutions, 20 μl of suspended cells and 20 μl of 25% sterilized fecal supernatants was added to each well in a 96-well plate with 140 μl of MHB. The final inoculum dose became approximately 1.5 × 10⁵ CFU/ml and the final chemical concentration became 0.25, 0.5, 1, 2 or 4 μg/ml for tetracycline. The 96-well plates were incubated at 37 °C for 48 h under anaerobic conditions and growth was measured by optical density at 600 nm with a Synergy MX spectrophotometer (BioTek Instruments, Winooski, VT) (Kim et al., 2015; Rose et al., 2009). After 48 h incubation at 37 °C, the number of wells in which growth had occurred (the negative control was OD₆₀₀ < 0.045) was recorded. The MIC was calculated as the minimum concentration at which the optical density did not exceed this breakpoint.

For the microbiological assay, the effect of fecal binding on tetracycline activity was assessed by incubating selected tetracycline concentrations (0, 0.15, 0.5, 1, 1.5, 2, 4, 8, and 16 μg/ml) with 25% (w/v) steam sterilized human fecal slurry in MHB for 48 hat 37 °C (Table 3). Aliquots were removed at 2 h and centrifuged for 20 min at 10,000 × g, and the steam sterilized human fecal supernatants then were analyzed for tetracycline activity. Tetracycline activity of the steam sterilized human fecal supernatant was determined using E. coli ATCC 25922 and S. Gaminara SEA2575. Twenty μl of MHB and suspended cells (final inoculum dose: approximately 1.5 × 10⁵ CFU/ml) was added to each well of a 96-well plate with 160 μl of steam sterilized human fecal supernatants. Duplicate control wells, containing steam sterilized human fecal supernatant without suspended cells as well as with or without MHB, were also incubated at 37 °C for 48 h under anaerobic conditions. Antibiotic susceptibility testing was conducted as described above.

Table 3.

Comparative MIC test using MHB and steam sterilized 2.5% (w/v) fecal supernatant for 48 h for the microbiological assays. The experiment was performed in triplicate on a 96-well plate.

Strains	Supplement	Added concentration of tetracycline (μg/ml).
		0	0.15	0.5	1	1.5	2	4
E. coli ATCC 25922	MHB only	+	+	+	–	–	–	–
	MHB +2.5% (w/v) fecal supernatant	+	+	+	+	+	–	–
S. enterica subsp. Enterica Serovar Gaminara SEA2575	MHB only	+	+	+	–	–	–	–
	MHB +2.5% (w/v) fecal supernatant	+	+	+	+	–	–	–

+ = growth, - = no growth.

3. Results

3.1. Experimental scheme

The experimental strategy to investigate the in vitro tetracycline binding on human fecal slurries is outlined in Fig. 2. This strategy is constituted of three steps,

• Step 1: determination of the experimental conditions, drug concentration and fecal parameters, including abiotic transformation of tetracycline (Fig. 3), and the comparison of sterilized fecal slurries and fecal slurries (Fig. 4), incubation time (Fig. 6), as well as fecal slurry (Fig. 5), and tetracycline concentrations (Fig. 7);

• Step 2: in vitro chemical assay (Fig. 8, Table 4); and

• Step 3: in vitro microbiological assay (Table 3, Table 4).

Fig. 3.

The abiotic transformation of tetracycline in 0.1% formic acid and 3% (w/v) human fecal supernatant for 9 days.

Fig. 4.

Comparison of the percent of tetracycline binding to steam sterilized fecal slurries and non-sterilized fecal slurries after 2 h incubation in BHI medium with selected tetracycline concentrations (15, 20, 30, and 50 μg/ml). Graphs represent averages of triplicate samples; selected tetracycline concentrations and error bars represent the standard deviations.

Fig. 6.

Time course for determining the amount of tetracycline remaining in the supernatant in steam sterilized 25% fecal slurries at 15 μg/ml (A) and 150 μg/ml (B) tetracycline concentrations. Graphs represent averages of triplicate samples; selected tetracycline concentrations and error bars represent the standard deviations.

Fig. 5.

The percent of tetracycline binding under different fecal slurry concentrations over 2 h incubation in BHI medium with selected tetracycline concentrations (0.15, 1.5, 15 and 150 μg/ml). Fecal samples were collected from three healthy volunteers. Graphs represent averages of triplicate tests from three human fecal slurries; selected tetracycline concentrations and error bars represent the standard deviations.

Fig. 7.

Comparison of the tetracycline concentration to binding under steam sterilized 25% fecal slurries after 2 h incubation in BHI medium with selected tetracycline concentrations (15, 20, 30, 50, 70, 100 and 150 μg/ml). Graphs represent averages of triplicate samples and error bars represent the standard deviations.

Fig. 8.

In Vitro fecal slurry test system-chemical assays. The percent of tetracycline binding to 0.15 μg/ml (A) and 1.5 μg/ml (B) tetracycline over 2 h incubation in steam sterilized 25% (w/v) fecal slurries. Graphs represent averages of triplicate samples and error bars represent the standard deviations.

Table 4.

Fraction of the dose available for colonic microorganisms from the microbiological and chemical assays.

Added concentration of tetracycline (μg/ml)		0	0.15	0.5	1	1.5	2	4
aChemical assay		0	0.087 ± 0.016			0.853 ± 0.136
bMicrobiological assay	E. coli ATCC 25922	+	+	+	+	+	+	-
	S. enterica subsp. Enterica Serovar Gaminara SEA2575	+	+	+	+	+	+	-
cPercentage of free drug in the supernatant.		0	58.2 ± 10.8			56.9 ± 9.1
dFraction of the dose available for colonic microorganisms		0	0.418			0.431

^aChemical assay (0.15 and 1.5 μg/ml by LC-MS/MS), measured concentration in the supernatant (μg/ml).

^bMicrobiological assay (+ = growth, - = no-growth) measure the OD₆₀₀ after 48 h in steam sterilized 25% (w/v) fecal slurries. Each sample was tested in triplicate on a 96-well plate.

^cPercentage of free drug in the supernatant equal to binding percent.

^dFraction of the dose available for colonic microorganisms calculated by 1-binding%/100.

The rationale for the experimental approach is based on the guidance given in Appendix D, VICH GL 36 (R).

3.2. Step 1: determination of the experimental conditions, drug concentration and fecal parameters

3.2.1. Abiotic transformation of tetracycline

Abiotic transformation of tetracycline was observed within 1 day in the presence of 3% (w/v) fecal supernatant. After 7 days of incubation, the remaining proportion of the 30 μg/ml tetracycline concentration reached 84% and 67% in 0.1% formic acid and 3% (w/v) fecal supernatant, respectively (Fig. 3). These percentages remained relatively stable for up to 9 days of the tested incubation period. This result suggested that fecal supernatant affected the abiotic transformation but did not result in more than 50% transformation within 9 days.

3.2.2. Comparing sterilized fecal slurries and fecal slurries

A comparative study was conducted using sterilized fecal slurries and fecal slurries for drug binding (Fig. 4). Fecal slurries (25%, w/v) prepared from three individual fecal specimens were used in this experiment and tested with 15, 20, 30, and 50 μg/ml tetracycline concentrations for the binding experiment. Tetracycline binding ranges of 57.9–77.1%, and 59.6–77.7% were obtained in the selected tetracycline concentrations after 2 h of incubation with 25% (w/v) fecal slurries and steam sterilized fecal slurries, respectively (Fig. 4). The highest percentage of added tetracycline, detected in the supernatants of 25% (w/v) fecal slurry preparations of person C, was 77.1 ± 7.2% and 77.7 ± 7.8% in fecal samples and steam-sterilized fecal samples, respectively. The tetracycline binding measured in the steam sterilized fecal slurries after 2 h of incubation was approximately the same as the concentrations measured in the fecal slurries.

3.2.3. Comparing fecal slurry concentrations

The effect of fecal slurry concentration on tetracycline binding was assessed using steam sterilized fecal slurries after 2 h incubation in BHI medium with selected tetracycline concentrations (0.15, 1.5, 15 and 150 μg/ml). Fecal samples from three healthy volunteers were used. Tetracycline binding ranges of 24–55%, 38–67%, and 60–73% were obtained after 2 h of incubation, in selected tetracycline concentrations, with 3%, 10% and 25% (w/v) steam sterilized fecal slurries, respectively (Fig. 5). A 25% (w/v) suspension of fecal slurry showed more constant values of binding percent at different tetracycline concentrations than the 3 and 10% (w/v) slurries.

3.2.4. Comparing incubation times

The effect of culture incubation time on drug binding was quantitatively assessed by incubating 25% (w/v) steam sterilized fecal slurries with 15 or 150 μg/ml of tetracycline for 1, 2, and 6 h (Fig. 6). Within 1 h of incubation, fecal slurries incubated with 15 or 150 μg/ml of tetracycline had 46.4 ± 8.2% and 63.7 ± 1.7% tetracycline binding, respectively. After 2 h of incubation, the tetracycline binding measured in the 15 μg/ml (60.3 ± 8.0%) and 150 μg/ml tetracycline (73.2 ± 1.9%) was not much different to the concentrations measured in these slurries after 6 h (67.3 ± 6.9% and 80.5 ± 11.3%, respectively).

3.2.5. Comparing tetracycline concentrations

Tetracycline concentrations of 15, 20, 30, 50, 70, 100 and 150 μg/ml were tested for fecal binding in samples of steam sterilized 25% (w/v) fecal slurries after 2 h of incubation. During the incubation with these tetracycline concentrations, the average binding in sterilized fecal slurries was 48.3 ± 10.1%, 56.9 ± 12.7%, 65.8 ± 15.6%, 70.5 ± 7.8%, 64.7 ± 8.9%, 65.6 ± 7.4%, and 65.2 ± 4.7%, respectively (Fig. 7). These results show that incubation of fecal slurries with 15–50 μg/ml of tetracycline resulted in an increase of the percentage bound. However, there was essentially no difference in the binding patterns of steam sterilized fecal slurries with 50–150 μg/ml tetracycline concentrations.

3.3. Step 2: In Vitro fecal slurry test system-chemical assays

The 0.15 and 1.5 μg/ml tetracycline were intended to simulate exposure levels equivalent to and above the corresponding codified ADI for the United States (25 μg/kg bw/day, as mentioned earlier) (Carman et al., 2005). The three individual steam sterilized fecal slurries (25%, w/v) with 0.15 μg/ml tetracycline had 49.5 ± 4.8%, 54.9 ± 21.7%, and 70.2 ± 0.3% binding of tetracycline at 37 °C after 2 h of incubation, respectively (Fig. 8A). The fecal slurries with 1.5 μg/ml tetracycline had 47.3 ± 4.7%, 57.8 ± 4.8%, and 65.5 ± 7.1% tetracycline binding (Fig. 8B). Individual to individual variation of tetracycline binding occurred in fecal slurries. The average binding of 0.15 and 1.5 μg/ml tetracycline with steam sterilized 25% (w/v) fecal slurries was 58.2 ± 10.8% and 56.9 ± 9.1% at 37 °C after 2 h of incubation, respectively. Based on percent binding results from chemical assay methods, the fractions of the oral dose available to microorganisms were 0.418 and 0.431 in the 0.15 μg/ml and 1.5 μg/ml of tetracycline, respectively (Table 4).

3.4. Step 3: In Vitro fecal slurry test system-microbiological assays

An in vitro drug binding assay by microbiological methods was conducted based on results from chemical assays. For the microbiological assay, the MICs of two reference strains (E. coli ATCC 25922 and S. Gaminara SEA2575) were determined. The E. coli ATCC 25922 and S. Gaminara SEA2575 reference strains did not display growth in sample wells containing tetracycline concentrations other than 0, 0.15, and 0.5 μg/ml (Table 3). Media supplemented with sterilized fecal supernatants better supported the growth of E. coli ATCC 25922 and S. Gaminara SEA2575 with 1 and 1.5 μg/ml of tetracycline (Table 3). The growth of E. coli ATCC 25922 and S. Gaminara SEA2575 were substantially higher in the medium supplemented with 2.5% sterilized fecal supernatant than in MHB media only.

The effect of fecal binding on tetracycline activity was assessed by microbiological methods incubating selected tetracycline concentrations in 25% (w/v) steam sterilized human fecal supernatant. The 25% (w/v) steam sterilized human fecal slurries supported the growth of E. coli ATCC 25922 and S. Gaminara SEA2575 in the added concentration of 2 μg/ml tetracycline, which is equal to two times the MIC of the drug, respectively, for both strains (Table 4). These results suggest that the active concentration may be lower than 1 μg/ml due to binding of the drug to the fecal slurries since E. coli 25922 and S. Gaminara SEA2575 are susceptible to 1 μg/ml of tetracycline.

4. Discussion

In this study, we investigated in vitro tetracycline binding on human fecal slurries to determine the fraction of oral dose available to microorganisms following the VICH mADI Expert Working Group (mADI EWG) deliberations that resulted in Appendix D, VICH GL 36 (R) (Cerniglia et al., 2016; FDA, 2013). Approach 2 consisted of Phases A and B using in vitro test systems. Phase A is an initial experiment including both chemical and microbiological assays. Phase B, based on results from Phase A, uses microbiological assays. The data for all six donors recommended in the approach were used for the final determination of the fraction of oral dose available to microorganisms (FDA, 2013). In this study, we determined the fraction of oral dose available for intestinal microorganisms following three steps. As shown in Fig. 2, the fraction of oral dose available to microorganisms based on the experimental conditions of drug concentration and fecal parameters (Step 1) was calculated by an in vitro chemical assay (Step 2) and confirmed by an in vitro microbiological assay (Step 3).

4.1. The experimental conditions of in vitro chemical assay

Before beginning an in vitro chemical assay, questions to be addressed included: a) what is the physicochemical characteristics of the tetracycline in fecal slurries?; b) what is the difference in tetracycline binding between the sterilized and fecal slurries for further chemical assays?; c) what fecal concentrations should be tested?; d) what are appropriate incubation times to determine kinetics of binding?, and e) what range of antibiotic concentrations could be assessed successfully using the optimized experimental conditions?

The physicochemical characteristics, as well as the pharmacological and pharmacokinetic parameters of the tetracycline HCl have been well investigated (Liang et al., 1998; Moreno-Cerezo et al., 2001; Wu and Fassihi, 2005). Tetracycline easily degrades under conditions of weak acids, strong bases, ultraviolet radiation, and temperatures above 37 °C, forming more than fourteen different degradation products (Hsieh et al., 2011; Kuhne et al., 2000; Liang et al., 1998; Mason et al., 2011; Moreno-Cerezo et al., 2001; Pena et al., 1998; Wu and Fassihi, 2005). We detected trace amounts of 4-epitetracycline in the fecal supernatants and sterile controls (data not shown). Epimerization occurs in acid solutions (pH 2.0–6.0) and leads to 4-epitetracycline, which has less antimicrobial activity (Pena et al., 1998). Although abiotic transformations have been shown, the amount of tetracycline in 3% (w/v) fecal supernatant was approximately 94% of the original concentration during the 2 h of incubation (data not shown). Also, tetracycline binding measured approximately 60% in the 15 μg/ml tetracycline after 2 h of incubation, which was not much different from the concentrations measured in these slurries after 6 h (67.3%). Thus, it appears that incubation time does not affect binding after 2 h, as indicated by no change in drug distribution between the 2 and 6 h samplings.

Sterilized fecal slurries are not representative of in vivo conditions. Appendix D, VICH GL36(R) (FDA, 2013) recommended that feces without sterilization should be used, where possible, when conducting in vitro drug-binding/inactivation studies. However, small differences between binding to fecal slurries and sterilized fecal slurries may allow further studies to be based on sterilized feces only (FDA, 2013). In this study, sterilized fecal slurries and fecal slurries gave approximately the same binding during the 2 h incubation periods. Sterilized fecal slurries may allow for further chemical and microbiological assay studies.

The difference in the binding concentration observed with the tetracycline could be caused by the fecal dilution (Fig. 5). The weight of the pellets from the 25% (w/v) steam sterilized fecal slurries was more consistent than that observed for the pellets from the 3 and 10% (w/v) steam sterilized fecal slurries (data not shown), which is consistent with the differences in observed tetracycline binding percent. Since the 50% (w/v) fecal slurry was highly viscous and difficult to manipulate (Ahn et al., 2012a), the 25% (w/v) suspension of fecal slurry was found to be more practically manageable for determining binding of the drug to fecal slurries. Appendix D, VICH GL36(R) (FDA, 2013) recommended 25% (w/v) fecal slurries as representative of the colon content.

In previous studies, the binding capacity was noted to depend on the concentration of antibiotic added to the fecal slurry (Ahn et al., 2012a; Hazenberg et al., 1983; Veringa and Vanderwaaij, 1984; Wagner et al., 2008). Also, Veringa and van der Waaij (Veringa and Vanderwaaij, 1984) observed that the percentage of remaining antibiotic in the supernatant declined with increasing concentrations of antibiotic. Consistent with previous studies, there was increased binding concentrations of steam sterilized fecal slurries up to 50 μg/ml tetracycline slurry sample. High concentrations of tetracycline (> 50 μg/ml) may be saturated on fecal slurries, indicating that every possible binding site is filled with tetracycline.

4.2. In vitro chemical assay

We attempted to address the methodological questions raised by the VICH GL #36 Step 3, “do residues entering the human colon remain microbiologically active?” by an in vitro chemical assay using HPLC and LC-MS/MS (Ahn et al., 2012a; Cerniglia et al., 2016). The active (intact) form and the concentration of drug able to reach the gut from an oral dose should be considered when estimating the value of the fraction of the oral dose. However, due to unavailability of data, the value is assumed to be one, which means the antimicrobial activity of drug reaching the gut is equal to the original oral dose (Fig. 1) (Cerniglia and Kotarski, 1999, 2005; Cerniglia et al., 2016). However, drug binding to fecal material is a major consideration; thus, an in vitro drug binding test systems for the drugs was developed. This system can determine the fraction of the oral dose bioavailable to microorganisms in the human intestinal tract (Ahn et al., 2012a; Edlund et al., 1988; Hazenberg et al., 1983, 1984; Jansen et al., 1992; Veringa and Vanderwaaij, 1984; Wagman et al., 1974).

Fig. 1.

VICH Guideline 36 formula approach equation for determining a microbiological ADI.

The binding of the antibiotic to feces was calculated to investigate the difference between the amount of added antibiotic and the amount remaining in the supernatant solution after centrifugation. Based on preliminary results for the experimental conditions of drug concentration and fecal parameters, chemical assay methods should be quantitatively assessed by incubating the drugs at a range of residual low level and therapeutic concentrations. In this study, tetracycline concentrations were intended to simulate the exposure levels reported by Carman et al. (2005). For example, 0.15 μg/ml tetracycline in feces would be formed by ingestion of 1.5 mg/day by an adult (60 kg person)(Carman et al., 2005; Cerniglia et al., 2016), which corresponds to 25 μg/kg bw/day of acceptable daily intake (ADI). In this study, the fraction of oral dose available to microorganisms was 0.418 in the 25% (w/v) steam sterilized fecal slurries with 0.15 μg/ml tetracycline after 2 h of incubation (Table 4).

The value of the fraction of oral dose available to microorganisms was based on the value of chemical binding percent to fecal slurries. There was no validation of the recovery of antimicrobial activity and drug amounts in the pellet fractions that was a concern expressed by the VICH mADI EWG. We conducted studies to determine the amount of drug in the pellet. Total recoveries of tetracycline with acetonitrile in the pellet were less than 30% (data not shown), but were neither reliable nor consistent with the chemical assay to detect tetracycline concentration. However, the fraction of oral dose available to microorganisms from the pellet and supernatant should be taken into account in the mADI formula.

4.3. In vitro microbiological assay

The microbiological bioassay is based on the presence or absence of bacterial growth after incubation of the drug residue-containing supernatant with a chosen indicator test organism (Cerniglia et al., 2016). The screening test for antibiotic residues (STAR) was used for the detection of tetracyclines (Lainscek et al., 2014; Tumini et al., 2016). It was not sensitive enough to detect tetracycline residues in milk and required prolonged incubation time (18–24 h). Also, the STAR method was not sensitive enough to satisfy the requirements for determining tetracycline at residue levels in fecal binding studies. Thus, we chose E. coli ATCC 25922 and S. Gaminara SEA2575 as indicator test microorganisms for our investigation. The 25% (w/v) steam sterilized human fecal slurries supported the growth of E. coli ATCC 25922 and S. Gaminara SEA2575 at the added concentrations of 1.5 and 2 μg/ml tetracycline; however, the concentration detected in the supernatant in the 25% (w/v) steam sterilized fecal slurries was about 0.87 μg/ml (58.2% of 1.5 μg/ml). Decreased sensitivity of indicator microorganisms in the presence of fecal supernatants used in the bioassay may indicate either decreased potency of antibiotics or reduced cell penetration. This result is consistent with our previous enrofloxacin studies (Ahn et al., 2012a, 2012b). Microbiological assays did not provide a quantitative fraction of oral dose available to microorganisms, even though it offered a cost-effective method to quantitatively monitor antibiotic residues. The microbiological assays method would be a useful and effective method for preliminary and confirmatory screening after chemical analysis.

5. Conclusions

The experimental procedures to better understand in vitro test methods to determine the fraction of oral doses available to microorganisms are outlined in Fig. 9. Concerns regarding fecal parameters include fecal concentrations, incubation time, and use of non-sterile or sterile fecal samples (Table 2). We confirmed that 25% (w/v) fecal slurries, 2 h incubation, and sterilized feces allowed a better understanding of antibiotic binding parameters. Tetracycline had 56.9 ± 9.1% and 58.2 ± 10.8% binding with 25% (w/v) fecal slurries in 0.15 and 1.5 μg/ml of tetracycline concentrations, respectively. Tetracycline in media supplemented with 2.5% (w/v) sterilized human fecal supernatant reduced sensitivity of E. coli ATCC 25922 and S. Gaminara SEA2575 to a concentration of tetracycline two or three times higher than the MIC. The 25% (w/v) steam sterilized human fecal slurries supported the growth of E. coli ATCC 25922 and S. Gaminara SEA2575 in the added concentration of 1.5 and 2 μg/ml tetracycline; however, the real concentration may be lower than 1 μg/ml due to binding of the drug to the fecal slurries. Based on results from chemical assays and microbiological methods, the fraction of oral doses available to microorganisms was 0.418 in the 0.15 μg/ml tetracycline concentration. There was consistent correlation between the chemical and microbiological assays.

Fig. 9.

Procedure of appropriate methods to determine fraction of oral dose available to microorganisms (N = test number).

Table 2.

The experimental conditions of tetracycline concentration and fecal parameters for in vitro tetracycline binding studies using human fecal slurries.

	Sample set	Fecal concentration (%)	Tetracycline concentration (μg/ml)	Tetracycline detection methods	Optimal conditions
Stability	Designed tetracycline concentration		0, 0.15, 1.5, 15, 150	HPLC, LC-MS/MS	0.15 μg/ml (equivalent of 25 μg/kg bw/day or 1.5 mg/60 kg person/day)
	Abiotic transformation of tetracycline	3	15, 150	HPLC	Abiotic transformation
Fecal parameters	Use of non-sterile or sterile fecal slurries	25	15, 20, 30, 50	HPLC	Sterile fecal slurries
	Fecal concentrations	3, 10, 25	0, 0.15, 1.5, 15, 150	HPLC, LC-MS/MS	25%
	Fecal incubation time	25	15, 150	HPLC	2 h
	Maximum tetracycline concentration	25	15, 20, 30, 50, 70, 100, 150	HPLC	Under 50 μg/ml