DMPK147_proof ■ 23 November 2016 ■ 1/6

Drug Metabolism and Pharmacokinetics xxx (2016) 1e6

55
Contents lists available at ScienceDirect 56
57
58
Drug Metabolism and Pharmacokinetics 59
journal homepage: http://www.journals.elsevier.com/drug-metabolism-and- 60
pharmacokinetics 61
62
63
Regular Article 64
65
1 In vitro ocular metabolism and bioactivation of ketoconazole in rat, 66
2 67
3 rabbit and human 68
4 69
5 Q9 Amanda L. Cirello a, Jennifer L. Dumouchel a, Mithat Gunduz a, Christine E. Dunne b, 70
6
Upendra A. Argikar a, * 71
7 72
a
8 Novartis Institutes for BioMedical Research, Analytical Sciences and Imaging, Cambridge, MA 02139, USA
b 73
Colorado State University, Department of Chemistry, Fort Collins, CO 80523, USA
9 74
10 75
11 76
12 a r t i c l e i n f o a b s t r a c t
77
13 78
Article history: Oral ketoconazole is clinically administered for treatment of severe cases for fungal keratitis. Pharma-
14 Received 5 August 2016 79
codynamics and efficacy of oral and topical (ocular) ketoconazole is explored in rabbit. However,
15 Received in revised form 80
metabolism of ketoconazole in the eye in any species is not well explored in any preclinical species or
16 9 September 2016 81
human. An understanding of ocular drug metabolism in the eye is crucial for ocular therapeutics to
17 Accepted 8 November 2016
facilitate the risk assessment and development of potential drug candidates for the clinic. We aimed to 82
Available online xxx
18 investigate the metabolism of ketoconazole in rat, rabbit and human ocular S9 fractions. Metabolism in 83
19 liver S9 fractions was also studied for a direct comparison. Eleven putative metabolites were identified in 84
Keywords:
20 Ocular metabolism
the in vitro incubations. Of these metabolites, six were rat ocular whereas eight were present in rabbit 85
21 Ocular S9 fractions and human ocular matrices each. Metabolic pathways in rabbit and human ocular fractions suggested the 86
22 Liver S9 fractions formation of reactive intermediates in rabbit and human liver and ocular S9 incubations, which was
87
23 Bioactivation confirmed with trapping studies. Herein, we report eight human ocular metabolites of ketoconazole for
88
24 Potassium cyanide the first time. To the best of our knowledge, this is the first report of ocular metabolic pathways and
89
ocular bioactivation of ketoconazole in preclinical species and human.
25 90
© 2016 The Japanese Society for the Study of Xenobiotics. Published by Elsevier Ltd. All rights reserved.
26 91
27 92
28 93
29 94
30 95
31 1. Introduction Aspergillus keratitis [6]. In conjunction with topical antibiotics, KT is 96
32 also administered orally to treat Fusarium keratitis and prevent 97
33 The imidazole antifungal, ketoconazole (KT), is a medication progression to endophthalmitis [7]. Furthermore, KT has been 98
34 that is used to treat a variety of superficial and deep tissue fungal documented to be effective in preventing recurrent fungal keratitis 99
35 infections since its first approval in the United States in 1981 [1]. Its in patients previously on an intensive topical antifungal regimen 100
36 mode of action is to prevent production of fungal cell membranes [8]. Topical ocular administration of KT for treatment of fungal 101
37 [2]. Mycotic keratitis is an ocular condition that is manifested by keratitis has been well tolerated in humans [4] and experimentally, 102
38 inflammation, redness, tearing of cornea and can rapidly worsen if in rabbits [9e12]. Yet, ocular metabolism of KT has never been 103
39 not treated. It is believed to be caused by poor upkeep of contact documented. 104
40 lenses or prolonged exposure to fungi in an outdoor setting [3]. KT Orally administered KT has been linked with hepatotoxicity. 105
41 has been prescribed orally for the treatment of different types of Hepatic KT metabolism has been previously studied in vitro in 106
42 keratitis [4] alone or in combination with steroids and other anti- rabbit and human [13] and in vivo in mouse [14]. Bioactivation of KT 107
43 biotics. KT, 200 mg (once or twice daily) is clinically utilized for arises due to multiple mechanisms, including but not limited to the 108
44 treatment of severe cases of Acanthamoebic keratitis [5] and formation of the N-deacetyl-ketoconazole metabolite and further 109
45 metabolism by Flavin monooxygenases (FMO) enzymes [15]. 110
46 Therefore, we decided to examine the ocular metabolism and po- 111
47 tential ocular bioactivation of KT with the help of a unique in vitro 112
Abbreviations: CID, collision induced dissociation; [MþH]þ, protonated
48 ocular metabolism model recently established in our laboratory 113
molecular ion; FTMS, Fourier transformation mass spectrometry; KT, ketoconazole;
49 RT, retention time; DAK, N-deacetylated ketoconazole; FMO, Flavin [16,17]. Our objective was to document ocular metabolism of KT in 114
50 Q1 monooxygenase. S9 fractions obtained from rat, rabbit and human in comparison 115
51 * Corresponding author. with liver S9 fractions to contrast our finding from ocular studies 116
E-mail address: upendra.argikar@novartis.com (U.A. Argikar).
52 117
53 118
http://dx.doi.org/10.1016/j.dmpk.2016.11.001
54 1347-4367/© 2016 The Japanese Society for the Study of Xenobiotics. Published by Elsevier Ltd. All rights reserved. 119

Please cite this article in press as: Cirello AL, et al., In vitro ocular metabolism and bioactivation of ketoconazole in rat, rabbit and human, Drug
Metabolism and Pharmacokinetics (2016), http://dx.doi.org/10.1016/j.dmpk.2016.11.001
 DMPK147_proof ■ 23 November 2016 ■ 2/6

2 A.L. Cirello et al. / Drug Metabolism and Pharmacokinetics xxx (2016) 1e6

1 with those from hepatic metabolism studies documented in the The sample aliquots were eluted at a flow rate set at 0.25 mL/min. 66
2 literature. Bioactivation of KT was studied in rabbit and human, The sample gradient started at 10% B for 2 min and was linearly 67
3 specifically because of their pharmacological and clinical relevance. increased to 90% over 27 min, and held for 3 min. The column was 68
4 returned to 10% B and held for 3 min before the next injection. LTQ- 69
5 2. Materials and methods Orbitrap XL was calibrated in positive electrospray ionization as 70
6 recommended by the vendor. Detailed MS tune parameters have 71
7 2.1. Chemicals and reagents been described elsewhere [18]. Samples were analyzed with a 35 V 72
8 capillary voltage, 100 V tube lens, sheath gas of 35 (arbitrary units), 73
9 KT was obtained from the compound inventory at Novartis In- auxiliary gas of 5 (arbitrary units), and capillary temperature of 74
10 stitutes for Biomedical Research (Basel, Switzerland). Human 325  C. Samples were analyzed under data dependent scanning 75
11 (mixed gender pool, 11 donors), rabbit (male New Zealand white, methods detailed in Bushee and Argikar, 2011 [19]. In short, full 76
12 75 donors) and rat (male SpragueeDawley, 200 donors) ocular S9 scans were acquired at 30,000 resolution and CID MS/MS and MS3 77
13 (5 mg/mL protein content) were purchased from XenoTech (Lenexa, scans were obtained at 7500 resolution. Analysis of the metabolite 78
14 KS). Rabbit (male New Zealand white, 11 donors) and rat (male identification and reactive intermediate trapping samples were 79
15 SpragueeDawley, 400 donors) liver S9 (20 mg/mL protein content) manually analyzed with Xcalibur software (version 2.0.7 SP1, 80
16 were purchased from XenoTech (Lenexa, KS) whereas, human Thermo Scientific). 81
17 (mixed gender, 50 donors) liver S9 (20 mg/mL protein content) was 82
18 obtained from BD Bioscience (San Jose, CA). Solvents were HPLC or 83
3. Results
19 MS grade and were purchased from Sigma Aldrich (St. Louis, MO, 84
20 USA) or Mallinckrodt Baker (Phillipsburg, NJ, USA). NADPH and 85
3.1. CID product ion spectrum of KT
21 alamethicin from Trichoderma viride were purchased from MP 86
22 Biomedicals (Solon, OH). All remaining chemicals were purchased 87
KT showed a protonated molecular ion [MþH]þ at m/z 531. The
23 from Sigma Aldrich. 88
LC retention time was approximately 17.0 min. The CID product ion
24 89
spectrum (Fig. 1) showed fragment ions 489, 446, 421, 311 and 255.
25 2.2. In vitro rat, rabbit and human ocular and liver S9 incubations 90
Fragment ion at m/z 489 was formed from the loss of acetyl moiety.
26 91
Fragment ion 446 was formed by loss of ethyleneamine moiety,
27 With the following adjustments, KT ocular and liver S9 in- 92
whereas, fragment ion 421 was formed by loss of pyrazole moiety
28 cubations were conducted at 1.25 mg/mL protein concentration 93
from m/z 489. Fragment ions 311 and 255 correspond to carboca-
29 (1 mL incubation volume) according to commonly accepted pro- 94
tions containing the pyrazolo and dichlorophenyl structural motifs.
30 tocols, as previously established in our laboratory [16]. All cofactors 95
The product ion spectrum of KT has been described in depth pre-
31 (SAM, PAPS, Acetyl CoA, etc) including NADPH (1 mM) and UDPGA 96
viously by Whitehouse et al. and Fitch et al. [20,21].
32 (5 mM) were added after the two pre-incubations and the reaction 97
33 was started upon the addition of substrate, KT (10 mM). Aliquots 98
34 (0.3 mL) were quenched with equal amounts of ice-cold acetonitrile 3.2. Structural elucidation of M1 and M6 99
35 with 0.1% formic acid to stop the reaction at 0 and 1 h. The 100
36 quenched reactions were vortexed and centrifuged for 5 min at M1 showed a protonated molecular ion [MþH]þ at m/z 545, 14 101
37 4630 g. The supernatants were transferred to a DW-96 plate, amu higher than the parent. The LC retention time was approxi- 102
38 dried down under gentle stream of N2(g) to a pellet, and recon- mately 19.5 min and was observed in rabbit and human ocular and 103
39 stituted (15 concentration) in the initial mobile phase conditions. liver S9 fractions (Supp. Fig. 1). The retention time for M6 was 104
40 The samples were subjected to analysis by LCeMSn. approximately 15.5 min and was observed across all matrices 105
41 (Supp. Fig. 6). The addition of 14 amu to parent fragment ion m/z 106
42 2.3. In vitro human and rabbit ocular and liver S9 fractions for 489 indicates oxidation. M1 and M6 were identified as oxidative 107
43 iminium ion trapping metabolites of ketoconazole. 108
44 109
45 Human and rabbit ocular and liver S9 fractions (1 mg/mL protein 110
3.3. Structural elucidation of M2
46 concentration) were pre-warmed for 3 min in 0.1 M potassium 111
47 phosphate buffer (pH 7.4) with magnesium chloride, KT (10 mM) 112
M2 showed a protonated molecular ion [MþH]þ at m/z 529, 2
48 and KCN (1 mM KCN:K13C15N 70:30 (v/v)) at 37  C. NADPH (1 mM) 113
amu less than the parent. The LC retention time was approximately
49 was then added to start the reaction. Reaction mixture was 114
18.6 min and was observed in rat, rabbit and human ocular and liver
50 quenched at 0 and 1 h and processed as previously described. The 115
S9 fractions. The CID MS/MS fragment ion used to elucidate the
51 supernatants were transferred to a DW-96 plate and dried down 116
metabolite was m/z 487 and 255 (Supp. Fig. 2). Based on the ion
52 under N2(g) to half the original volume. The samples were then 117
product spectra the site of metabolism was most likely on the
53 analyzed by LCeMSn. 118
piperazine moiety. M2 was identified as hydroxylation followed by
54 119
dehydration.
55 2.4. LCeMSn for metabolism identification 120
56 121
57 All samples were analyzed on a Thermo Orbitrap-XL mass 3.4. Structural elucidation of M3 122
58 spectrometer (Thermo Fisher Scientific, Waltham, MA) equipped 123
59 with CTC PAL autosampler, a 3 Ti high-performance LC pump M3 showed a protonated molecular ion [MþH]þ at m/z 547, 16 124
60 (LEAP Technologies, Carrboro, NC) amu more than the parent. The LC retention time was approxi- 125
61 The analytes were injected and separated on a Waters Sym- mately 16.7 min and was observed in rat, rabbit and human ocular 126
62 metry C18 analytical column (5 mm 2.1  150 mm; Milford, MA) and liver S9 fractions (Supp. Fig. 3). The addition of 16 amu to 127
63 with a 35 min gradient elution method. Mobile phase A consisted of parent fragment ion m/z 489 to afford m/z 505 indicates hydrox- 128
64 10 mM ammonium formate in MS-grade water with 0.1% formic ylation. M3 was identified as a hydroxylated metabolite of 129
65 acid. Mobile phase B consisted of acetonitrile with 0.1% formic acid. ketoconazole. 130

Please cite this article in press as: Cirello AL, et al., In vitro ocular metabolism and bioactivation of ketoconazole in rat, rabbit and human, Drug
Metabolism and Pharmacokinetics (2016), http://dx.doi.org/10.1016/j.dmpk.2016.11.001
 DMPK147_proof ■ 23 November 2016 ■ 3/6

A.L. Cirello et al. / Drug Metabolism and Pharmacokinetics xxx (2016) 1e6 3

1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 80
16 81
17 82
18 83
19 84
20 85
21 86
22 87
23 88
24 89
25 90
26 91
27 92
28 93
29 94
30 95
31 96
32 97
33 98
34 99
35 100
36 101
37 102
38 103
39 104
40 105
41 106
42 107
43 108
44 109
45 110
46 111
47 112
48 113
49 114
50 115
51 116
52 117
53 118
54 119
55 120
56 121
57 122
58 123
59 124
60 125
61 126
62 Fig. 1. (A) CID of KT was obtained at normalized collision energy of 35. Protonated molecular ion [MþH]þ is represented by m/z 531. The diagnostic ions include 489 and 255 127
63 representing the loss of the acetyl group and ketene, respectively (B) LCeMS chromatogram of ketoconazole, (C) LCeMS chromatogram of the trapped cyano adduct metabolite 128
64 (M11) (D) a full scan accurate mass measurement of M11 and the stable labeled cyano adduct (13C15N). 129
65 130

Please cite this article in press as: Cirello AL, et al., In vitro ocular metabolism and bioactivation of ketoconazole in rat, rabbit and human, Drug
Metabolism and Pharmacokinetics (2016), http://dx.doi.org/10.1016/j.dmpk.2016.11.001
 DMPK147_proof ■ 23 November 2016 ■ 4/6

4 A.L. Cirello et al. / Drug Metabolism and Pharmacokinetics xxx (2016) 1e6

1 3.5. Structural elucidation of M4 suggesting N-dealkylation. M9 was identified as an aniline 66

2 metabolite of KT resulting from two successive N-dealkylations at 67
3 M4 showed a protonated molecular ion [MþH]þ at m/z 533, 2 the piperazine nitrogen atom (Supp. Fig. 9). 68
4 amu higher than the parent. The LC retention time was approxi- 69
5 mately 15.8 min and was observed in rat, rabbit and human ocular 3.10. Structural elucidation of M10 70
6 and liver S9 fractions. Fragment ion, m/z 489 from the parent 71
7 spectrum gained 16 amu to yield m/z 505 indicating hydroxylation M10 showed a molecular ion [MþH]þ at m/z 329, 276 amu less 72
8 on the piperazine (Supp. Fig. 4). M4 was identified as hydroxylation than the parent. The LC retention time was approximately 13.2 min 73
9 and formyl conjugation of ketoconazole. It is possible that M4 is and was observed in rabbit and human ocular and liver S9 fractions 74
10 Q2 metabonate of KT arising from M5 (Table 1). and rat liver S9 fractions. Fragment ion m/z 255, remaining un- 75
11 changed from the parent spectra, highly suggests that M10 is O- 76
12 3.6. Structural elucidation of M5 dealkylated KT (Supp. Fig. 10). 77
13 78
14 M5 showed a protonated molecular ion [MþH]þ at m/z 505, 26 3.11. Structural elucidation of M11
79
15 amu less than the parent. The LC retention time was approximately 80
16 15.2 min and was observed in rat, rabbit and human liver S9 frac- 81
M11 showed a molecular ion [MþH]þ at m/z 556 and 558 (cold
17 tions and human ocular S9 fractions. Fragment ion m/z 487 was 82
and stable label, respectively), 25 and 27 amu higher than the
18 formed from loss of water. M5 was identified as hydroxylated and 83
parent with retention time of 15.5 min. Protonated molecular ions
19 N-deacetyled ketoconazole metabolite (Supp. Fig. 5). 84
at m/z 556 and 558 were observed corresponding to regular and
20 85
stable labled cyano adducts, respectively. Fragment ion of m/z value
21 86
3.7. Structural elucidation of M7 531 was recorded upon neutral losses of 27 and 29 amu, corre-
22 87
sponding to cold and stable labeled cyano groups, respectively.
23 88
M7 showed a molecular ion [MþH]þ at m/z 707, 176 amu higher Thus, M11 was identified as a cyano adduct. M11 was observed in
24 89
than the parent. The LC retention time was approximately 14.5 min human and rabbit ocular and liver S9 fractions (Fig. 1). M11 was
25 90
and was observed in rat, rabbit and human liver S9 fractions. identified as the cyano adduct of KT, arising from the reaction with
26 91
Addition of 176 amu to ketoconazole indicates a direct glucur- an iminium intermediate.
27 92
onidation without any modification. Fragment ion m/z 531 (keto-
28 93
conazole) formed upon cleavage of the sugar moiety which is a
29 4. Discussion 94
typical fragmentation pattern of glucuronides. M7 was identified as
30 95
N-glucuronide of ketoconazole (Supp. Fig. 7).
31 Our work aimed to investigate the metabolic fate of KT in rat, 96
32 rabbit and human ocular S9 fractions. Our results indicated that the 97
33 3.8. Structural elucidation of M8 metabolic pathways were mostly consistent between ocular and 98
34 hepatic models across species. Metabolic pathways in the liver S9 99
35 M8 showed a molecular ion [MþH]þ at m/z 489, 42 amu less fractions were in agreement with the literature [20]. We observed 100
36 than the parent. The LC retention time was approximately 14.0 min that KT was mainly metabolized to oxidative products in both 101
37 and was observed in rat, rabbit and human ocular and liver S9 ocular and liver S9 fractions (Fig. 2). In the present study, we 102
38 fractions. Fragment ions m/z 446, 421, 311 and 255, remaining demonstrated that ocular drug metabolizing enzymes could also 103
39 intact highly suggest that M8 is deacetylated KT (Supp. Fig. 8). catalyze many of the metabolic reactions documented for KT by 104
40 liver drug metabolizing enzymes. Subtle species and organ differ- 105
41 3.9. Structural elucidation of M9 ence were observed in the overall metabolic profile. For example, 106
42 M1 was not observed in rat, whereas M9 was not observed in hu- 107
43 M9 showed a molecular ion [MþH]þ at m/z 420, 111 amu less man. Furthermore, M5 and M10 were liver specific metabolites in 108
44 than the parent, eluting at approximately 14.0 min. M9 was rat; M7 was identified to be liver specific in rabbit and human. In 109
45 observed in rabbit and rat ocular and liver S9 fractions. Fragment contrast M10 was only liver specific only in rat. Due to lack of 110
46 ions m/z 311 and 255 remained intact from the parent spectra, authentic reference standards, the metabolites reported here in are 111
47 112
48 113
49 Table 1 114
50 Rabbit, rat and human metabolites identified in ocular and liver S9 incubations (þ indicates presence of metabolites and  indicates that the metabolite was not observed). Q3 115
51 116
[MþH]þ RT (min) Key CID product ions Rat ocular S9 Rat liver S9 Rabbit ocular S9 Rabbit liver S9 Human ocular S9 Human liver S9
52 117
53 KT 531.1544 17.04 489.1429, 446.1139, 421.1059, þ þ þ þ þ þ 118
311.0335, 255.0079
54 M1 545.1344 19.5 503.1228   þ þ þ þ
119
55 M2 529.1380 18.6 487.1277, 255.0058 þ þ þ þ þ þ 120
56 M3 547.1485 16.7 529.1407, 505.1204 þ þ þ þ þ þ 121
57 M4 533.1329 15.8 515.1220, 505.1379, 473.1112 þ þ þ þ þ þ 122
M5 505.1384 15.2 487.1276, 420.0883  þ  þ þ þ
58 123
M6 545.1337 15.5 503.1221, 473.1115, 432.0905, þ þ þ þ þ þ
59 255.0070 124
60 M7 707.1852 14.5 531.1536  þ  þ e þ 125
61 M8 489.1431 14.0 446.1010, 421.1066, 311.0348, þ þ þ þ þ þ 126
62 255.0075 127
M9 420.0862 14.0 311.0354, 255.0081 þ þ þ þ e e
63 M10 329.0456 13.2 255.0081 e þ þ þ þ þ
128
64 M11 556.1513 15.5 NA NA þ þ þ þ 129
65 130

Please cite this article in press as: Cirello AL, et al., In vitro ocular metabolism and bioactivation of ketoconazole in rat, rabbit and human, Drug
Metabolism and Pharmacokinetics (2016), http://dx.doi.org/10.1016/j.dmpk.2016.11.001
 DMPK147_proof ■ 23 November 2016 ■ 5/6

A.L. Cirello et al. / Drug Metabolism and Pharmacokinetics xxx (2016) 1e6 5

1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 80
16 81
17 82
18 83
19 84
20 85
21 86
22 87
23 88
24 89
25 90
26 91
27 92
28 93
29 94
30 95
31 96
32 97
33 98
34 99
35 100
36 101
37 102
38 103
Fig. 2. Proposed metabolic pathways for KT in rabbit, rat and human ocular and liver S9 fractions.
39 104
40 105
41 qualitative in nature. In vivo, the formation of ocular metabolites Efficacy of topical and oral KT for treatment of fungal keratitis 106
42 may be limited by ketoconazole available in the eye due to melanin has been extensively documented in rabbits [8,10,11]. There have 107
43 binding, and transport across the blood-ocular barriers or by been many publications documenting administration of oral KT in 108
44 intrinsic metabolic capacity of the eye. Because ketoconazole is the form of a once a day dose of 200 mg for the treatment of ocular 109
45 dosed orally and not topically in the clinic, metabolites observed fungal infections. The treatment regimen mostly involves combi- 110
46 in vivo in the eye may either be ocular or may be hepatic/extra- nation of oral KT with other antibacterial and steroidal agents 111
47 hepatic metabolites that reach the eye via circulation. In vitro in- [6,7,12,24]. A report by Rosa et al., showed that oral KT in combi- 112
48 cubations of KT in rabbit and human ocular S9 fractions contained nation was used in at least 20% of fungal keratitis cases in Florida 113
49 KCN as a trapping agent for reactive, hard electrophiles. Herein, we over a decade [25]. In clinical therapy, KT administration at a dose of 114
50 demonstrate that ocular enzymes can also catalyze a reaction to 800 mg/day for 4 months led to papilledema, which was reversible 115
51 form an iminium ion intermediate of KT, thus triggering bio- after termination of therapy [26]. In a rabbit model, KT has been 116
52 activation of KT in the eye. KT bioactivation was previously studied shown to produce modest pathological changes in regeneration of 117
53 in vitro in rat and human liver microsomes fortified with KCN [21] corneal epithelium [9]. Interestingly, after administration of 2% KT 118
54 and clinically, KT administration has also been linked with hepa- topically to patients, Torres and coworkers noted that there was 119
55 totoxicity. Hepatotoxicity associated with KT is believed to be due little to no evidence for microscopic changes in the front of the eye 120
56 to the formation of the N-deacetylation of KT (DAK) followed by [4]. This data coupled with the bioactivation pathway identified in 121
57 FMO mediated metabolism [15]. It was also reported that KT forms this study provides indirect evidence for back of the eye bio- 122
58 the iminium ion intermediate trappable by KCN [21,22]. There are activation of KT. 123
59 many literature reports on the reactivity of iminium ions [23], thus Although KT has been studied extensively, our work is the first 124
60 providing circumstantial evidence for organ toxicity. The mecha- to demonstrate that KT formed a reactive iminium ion trappable by 125
61 nism of toxicity could arise from an iminium ion reacting with cyanide in ocular S9 fractions, indicating topically administered KT 126
62 proteins in its vicinity. The iminium ion is thought to cause adverse could potentially be bioactivated in the eye. However, the safety 127
63 drug reactions including DNA adducts. Our report is the first to data on topically administered KT may be limited. It is challenging 128
64 investigate the bioactivation potential of ocular S9 fractions. to investigate the fate of the drug when administered topically, 129
65 130

Please cite this article in press as: Cirello AL, et al., In vitro ocular metabolism and bioactivation of ketoconazole in rat, rabbit and human, Drug
Metabolism and Pharmacokinetics (2016), http://dx.doi.org/10.1016/j.dmpk.2016.11.001
 DMPK147_proof ■ 23 November 2016 ■ 6/6

6 A.L. Cirello et al. / Drug Metabolism and Pharmacokinetics xxx (2016) 1e6

1 especially to the eye. The drug could be absorbed by the blood [4] Torres MA, Mohamed J, Carvazos-Adame H, Martinez LA. Topical ketoconazole 61
for fungal keratitis. Am J Ophthalmol 1985;100:293e8.
2 vessels in the eye to the systemic circulation which then would be 62
[5] Asbell PA, Torres MA. Therapeutic dilemmas in external ocular diseases. Drugs
3 subject to all bioactivation pathways even if it escapes the eye. 1991;42:606e15. 63
4 Iminium ions are reported quite extensively on understanding [6] Kredics L, Varga J, Kocsube S, Rajaraman R, Raghavan A, Doczi I, et al. Infec- 64
5 toxicity observed in many drugs. These reactions could be catalyzed tious keratitis caused by Aspergillus tubingensis. Cornea 2009;28:951e4. 65
[7] Dursun D, Fernandez V, Miller D, Alfonso EC. Advanced fusarium keratitis
6 by CYP450, myeloperoxidases, hemoglobin/hydrogen peroxide and progressing to endophthalmitis. Cornea 2003;22:300e3. 66
7 CYP450/t-butylperoxide [27]. KT has been shown to be hydrolyzed [8] Komadina TG, Wilkes TD, Shock JP, Ulmer WC, Jackson J, Bradsher RW. 67
8 to DAK. In addition to KT, DAK has been shown to be bioactivated by Treatment of Aspergillus fumigatus keratitis in rabbits with oral and topical 68
ketoconazole. Am J Ophthalmol 1985;99:476e9.
9 FMO enzymes [28]. Our work with ocular S9 fractions demon- [9] Foster CS, Lass JH, Moran-Wallace K, Giovanoni R. Ocular toxicity of topical
69
10 strated the formation of the DAK. The DAK may be subject to a antifungal agents. Arch Ophthalmol 1981;99:1081e4. 70
11 similar bioactivation pathway in the eye and most certainly in the [10] Zhang J, Wang L, Gao C, Zhang L, Xia H. Ocular pharmacokinetics of topically- 71
applied ketoconazole solution containing hydroxypropyl beta-cyclodextrin to
12 liver, if it reaches to the systemic circulation. It has been shown that rabbits. J Ocul Pharmacol Ther 2008;24:501e6.
72
13 KT is extensively metabolized by CYP3A4 [21], an enzyme that KT [11] Oji EO. Study of ketoconazole toxicity in rabbit cornea and conjunctiva. Int 73
14 inactivates in a time-dependent manner. The importance of Ophthalmol 1982;5:169e74. 74
[12] Singh SM, Khan R, Sharma S, Chatterjee PK. Clinical and experimental mycotic
15 CYP3A5 in the metabolism of KT has recently been highlighted 75
corneal ulcer caused by Aspergillus fumigatus and the effect of oral ketoco-
16 [29,30]. Interestingly, little or no CYP3A4 has been detected in the nazole in the treatment. Mycopathologia 1989;106:133e41. 76
17 eye [31,32]. Thus our report indirectly points toward either the [13] Rodriguez RJ, Miranda CL. Isoform specificity of N-deacetyl ketoconazole by 77
human and rabbit flavin-containing monooxygenases. Drug Matab Dispos
18 presence of CYP3A5 in the eye or an unrelated ocular enzyme that 78
2000;28:1083e6.
19 metabolizes KT. As additional information on ocular drug metab- [14] Whitehouse LW, Menzies A, Dawson B, Cyr TD, By AW, Black DB, et al. Mouse 79
20 olizing enzymes comes to light in the upcoming years, it will be hepatic metabolites of ketoconazole: isolation and structure elucidation. 80
21 worthwhile investing the potential role of CYP3A5 and other en- J Pharm Biomed Anal 1994;12:1425e41. Q7 81
[15] Acosta Ra. Metabolism of ketoconazole and deacetylated ketoconazole by rat
22 zymes in the metabolism of KT. hepatic microsomes and flavin-containing monooxygenases. Drug Matab 82
23 In summary, although, KT has been extensively investigated in Dispos 1997;25:772e7. Q8 83
24 hepatic in vitro systems, this is the first report of ocular metabolism [16] Bushee JL, Dunne CE, Argikar UA. An in vitro approach to investigate ocular 84
metabolism of a topical, selective beta1-adrenergic blocking agent, betaxolol.
25 and ocular bioactivation of KT in eye. Bioactivation was found to Xenobiotica 2015;45:396e405.
85
26 occur via iminium ion in both human and rabbit ocular S9 fractions [17] Argikar UA, Dumouchel JL, Dunne CE, Saran C, Cirello AL, Gunduz M. Ocular 86
27 and the reactive intermediates were successfully trapped via po- metabolism of levobunolol: historic and emerging metabolic pathways. Drug 87
Matab Dispos 2016;44:1304e12.
28 tassium cyanide based trapping methodology. This is the first [18] Gunduz M, Argikar UA, Baeschlin D, Ferreira S, Hosagrahara V, Harriman S.
88
29 report of trapping an iminium ion intermediate arising from ocular Identification of a novel N-carbamoyl glucuronide: in vitro, in vivo, and 89
30 bioactivation of KT and may provide indirect evidence for ocular mechanistic studies. Drug Matab Dispos 2010;38:361e7. 90
[19] Bushee JL, Argikar UA. An experimental approach to enhance precursor ion
31 toxicity of KT in rabbit. In addition, in contrast to rabbit ocular 91
fragmentation for metabolite identification studies: application of dual colli-
32 toxicity, such clinical observations have not been documented in sion cells in an orbital trap. Rapid Commun Mass Spectrom 2011;25:1356e62. 92
33 patients, thus pointing toward potential favorable or compensatory [20] Whitehouse LW, Menzies A, Dawson B, Cyr AW, Black DB, Zamecnik J. Mouse 93
hepatic metabolites of ketoconazole: isolation and structure elucidation.
34 elimination mechanisms in humans. 94
J Pharm Biomed Anal 1994;12:1425e41.
35 [21] Fitch LW, Tran T, Young M, Liu L, Chen Y. Revisiting the metabolism of ke- 95
36 Conflict of interest toconazole using accurate mass. Drug Metab Lett 2009;3:191e8. 96
37 [22] Argoti D, Liang L, Conteh A, Chen L, Bershas D, Yu C, et al. Cyanide trapping of 97
iminium ion reactive intermediates followed by detection and structure
38 The authors have no conflict of interest. identification using liquid chromatography-tandem mass spectrometry (LC- 98
39 MS/MS). Chem Res Toxicol 2005;18:1537e44. 99
40 [23] Masic LP. Role of cyclic tertiary amine bioactivation to reactive iminium 100
Acknowledgments species: structure toxicity relationship. Curr Drug Metab 2011;12:35e50.
41 [24] Ishibashi Y. Oral ketoconazole therapy for keratomycosis. Am J Ophthalmol
101
42 1983;95:342e5. 102
The authors acknowledge helpful discussions with Reiner
43 [25] Rosa Jr RH, Miller D, Alfonso EC. The changing spectrum of fungal keratitis in 103
Aichholz and Bindi Sohal. south Florida. Ophthalmology 1994;101:1005e13.
44 [26] Novack GD. Ocular toxicology. Curr Opin Ophthalmol 1994;5:110e4.
104
45 [27] Obach RS, Dalvie DK. Metabolism of nomifensine to a dihydroisoquinolinium 105
Appendix A. Supplementary data
46 ion metabolite by human myeloperoxidase, hemoglobin, monoamine oxidase 106
A, and cytochrome P450 enzymes. Drug Matab Dispos 2006;34:1310e6.
47 107
Supplementary data related to this article can be found at http:// [28] Rodriguez RJ, Acosta Jr D. Metabolism of ketoconazole and deacetylated ke-
48 toconazole by rat hepatic microsomes and flavin-containing mono- 108
dx.doi.org/10.1016/j.dmpk.2016.11.001.
49 oxygenases. Drug Matab Dispos 1997;25:772e7. 109
[29] Niwa T, Murayama N, Emoto C, Yamazaki H. Comparison of kinetic parameters
50 110
for drug oxidation rates and substrate inhibition potential mediated by cy-
51 Q4,5 References 111
tochrome P450 3A4 and 3A5. Curr Drug Metab 2008;9:20e33.
52 [30] Soars MG, Grime K, Riley RJ. Comparative analysis of substrate and inhibitor 112
[1] Heel RC, Brogden RN, Carmine A, Morley PA, Speight TM, Avery GS. Ketoco- interactions with CYP3A4 and CYP3A5. Xenobiotica 2006;36:287e99.
53 nazole: a review of its therapeutic efficacy in superficial and systemic fungal 113
[31] Nakamura K, Fujiki T, Tamura HO. Age, gender and region-specific differences in
54 infections. Drug Eval 1982;23:1e36. drug metabolising enzymes in rat ocular tissues. Exp Eye Res 2005;81:710e5. 114
55 [2] Scheinfeld N. Ketoconazole: a review of a workhorse antifungal molecule with [32] Al-Ghananeem AM, Crooks PA. Phase I and phase II ocular metabolic activities 115
a focus on new foam and gel formulations. Drugs Today 2008;44:369e80. and the role of metabolism in ophthalmic prodrug and codrug design and
56 [3] Thomas PA, Kaliamurthgy J. Mycotic keratitis: epidemiology, diagnosis and
116
delivery. Molecules 2007;12:373e88.
57 Q6 management. J Clin Microbiol Infect Dis 2013;19:210e20. 117
58 118
59 119
60 120

Please cite this article in press as: Cirello AL, et al., In vitro ocular metabolism and bioactivation of ketoconazole in rat, rabbit and human, Drug
Metabolism and Pharmacokinetics (2016), http://dx.doi.org/10.1016/j.dmpk.2016.11.001