Using in vitro/in silico data for consumer safety assessment of feed
ﬂavoring additives e A feasibility study using piperine
, S. Ethevea
, E. Fabianb
, W.R. Leemanc
, J.R. Plautza aDSM Nutritional Products AG, Wurmisweg 576, 4303 Kaiseraugst, Switzerland
bBASF SE, Experimental Toxicology and Ecology, 67056 Ludwigshafen, Germany cTNO, Utrechtseweg 48, 3704 HE Zeist, The Netherlands
a r t i c l e i n f o
Received 8 December 2014 Received in revised form 8 June 2015
Accepted 9 June 2015 Available online 21 June 2015
Keywords: Residue estimation Transfer database Piperine Metabolism Risk assessment
a b s t r a c t
Consumer health risk assessment for feed additives is based on the estimated human exposure to the additive that may occur in livestock edible tissues compared to its hazard. We present an approach using alternative methods for consumer health risk assessment. The aim was to use the fewest possible number of animals to estimate its hazard and human exposure without jeopardizing the safety upon use. As an example we selected the feed ﬂavoring substance piperine and applied in silico modeling for residue estimation, results from literature surveys, and Read-Across to assess metabolism in different species. Results were compared to experimental in vitro metabolism data in rat and chicken, and to quantitative analysis of residues' levels from the in vivo situation in livestock. In silico residue modeling showed to be a worst case: the modeled residual levels were considerably higher than the measured residual levels. The in vitro evaluation of livestock versus rodent metabolism revealed no major differ-ences in metabolism between the species. We successfully performed a consumer health risk assessment without performing additional animal experiments. As shown, the use and combination of different alternative methods supports animal welfare consideration and provides future perspective to reducing the number of animals.
© 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Feed additives, including feed ﬂavorings, require a thorough evaluation of consumer safety (Fig. 1). As presented by the Euro-pean Food Safety Authority (EFSA, 2012a), such an evaluation can be seen as a four step process from hazard characterization in laboratory animals or available literature through to determination of additive residues in the consumable target species. Step 1 is a hazard characterization usually done in rats, which will also include the kinetic and metabolic proﬁling of the additive. Step 2 is the comparison of these rat data with the kinetics and metabolism of the feed additive in the target livestock species to determine the identity and amount of the major metabolites. In case the target animal would form species-speciﬁc metabolites that were not present in the laboratory rat, gathering additional toxicological information about this metabolite would be warranted. Step 3 consists of determination of relevant residue levels in the edible
tissues of livestock. In Step 4, the risk characterization for the feed additive is derived based on the exposure of the consumer to the substance and its metabolites' via edible tissues that is then compared to the Health Based Guidance Values such as the Acceptable Daily Intake (ADI) derived from toxicological studies.
Brieﬂy, step 1 requires gathering the relevant toxicological (hazard) information either from literature (original literature and/ or reviews from competent bodies such as EFSA or the Joint FAO/ WHO Expert Committee on Food Additives (JECFA)) or from spe-ciﬁcally designed toxicological studies, step 2 requires metabolism studies in laboratory rat and target species using radiolabelled material. Step 3 requires the feeding of livestock animals followed by collection of their edible tissues for analytical determination of the residues' concentrations. Following the EFSA guidance docu-ment (EFSA, 2012a) such an evaluation will require a rather extensive use of both laboratory animals and livestock animals for meeting the experimental objectives. Under normal farming con-ditions, the target species animals would be marketed; however, because of the nature of the experimental conditions, the remains of the animals must be discarded.
With the current increased emphasis on reducing the use of
* Corresponding author.
E-mail address:firstname.lastname@example.org(A. Thiel).
Contents lists available atScienceDirect
Regulatory Toxicology and Pharmacologyj o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / y r t p h
animals for experimental purposes, weﬁnd it timely to present our case study of a modiﬁed approach to the safety assessment of feed additives thereby using and combining alternative methods. While the overall focus must remain on safety to the consumer, by providing alternative approaches for each step in the hazard char-acterization process, the current evaluation demonstrated that it is possible to reduce or avoid animal studies.
Information for hazard characterization can be obtained from literature from entries in databases such as that available from the European Chemicals Agency (ECHA). However, it may be that the
required information is not or not fully available. For such situations one can consider using the Threshold of Toxicological Concern (TTC) approach (Kroes et al., 2004; Munro et al., 2008) in case exposure is sufﬁciently low. It is also possible to use in silico ap-proaches to determine possible relevant endpoints,ﬁll data gaps and calculate/predict toxicological endpoints such as the No Adverse Effect Level (NAEL), which is key for risk assessment. Recently,Schilter et al. (2014)developed a decision tree approach on how and when to use in silico toxicological/hazard data in risk assessment.
The in vitro assessment of metabolism is possible and is used frequently. Usually, livers from the relevant species are used to prepare the in vitro functional metabolic components as found in the S9-fraction, microsomes, or primary hepatocytes; additionally, liver slices or liver cell lines are also used. Although such ap-proaches do require animal tissues, when compared to designed whole animal studies, these in vitro approaches yield information from fewer animals and avoid exposure of the animals. The use of comparative in vitro metabolism studies was recently recom-mended by the European Medicines Agency (EMA) as an alternative to in vivo metabolism studies (EMA, 2011). Likewise also other Regulatory Agencies clearly recommend a reduction in animal use and replacement by in vitro approaches (EFSA, 2009; ECHA, 2014). And, as noted above, predictions on metabolism in mammals can be done via in silico models (Schroeder et al., 2011). However, in silico metabolism predictions tools such as Meteor focus on biotrans-formation in mammalian species.
Further to the assessment process where residue levels in edible tissue data are needed, a non-animal approach to addressing these endpoints could be taken by using physiologically based pharma-cokinetic (PBPK) modeling (Cortright et al., 2009). It is also relevant to use databases that allow estimating the transfer of foreign sub-stances from the livestock diet into edible tissues (Leeman et al., 2007; MacLachlan, 2011), a concept we will demonstrate in our approach.
In this paper, we will present a case study of using a combina-tion of in vitro/in silico methods to perform a safety assessment of the feedﬂavoring piperine (EC, 2011,Fig. 2). Because it is a widely used feedﬂavoring ingredient, we have limited our focus to the 0.5 ppm piperine as added to poultry feed.
2. Materials and methods
In our approach we have addressed each of the risk assessment Abbreviations
ADI acceptable daily intake AI active incubations
AROD alkoxyresuroﬁn-O-dealkylase BROD benzyloxyresuroﬁn-O-debenzylase EROD ethoxyresuroﬁn-O-deethylase ECHA European Chemicals Agency EFSA European Food Safety Authority EMA European Medicines Agency
EPA Environmental Protection Agency (USA) HPV High Production Volume Chemicals Program HDC heat deactivated control
HPLC high performance liquid chromatography IPCS International Program on Chemical Safety JECFA Joint FAO/WHO Expert Committee Report on Food
LC-MS liquid chromatography-mass spectrometry L(O)AEL lowest (observed) adverse effect level LOQ limit of quantiﬁcation
MoE margin of exposure: N(L)OAEL/exposure MRTD maximum recommended therapeutic dose NTP National Toxicology Program
N(O)AEL no (observed) adverse effect level n.a. not applicable
PROD pentoxyresoruﬁn-O-depentylase PBPK physiologically based pharmacokinetic (Q)SAR (quantitative) structure activity relationship RP reverse phase
TTC threshold of toxicological concern TdB transfer database
Fig. 1. Graphical outline of the risk assessment approach used in this feasibility study: For hazard evaluation (step 1), use all available information such as literature or in silico computational tools to derive a NOAEL/Health Based Guidance Value. If necessary consider using the TTC concept or perform dedicated toxicology studies. For evaluation of comparative metabolism (step 2), three steps can be used: Part A) literature eval-uation to identify metabolites of piperine in mammals and chicken. Part B) check whether the known xenobiotic metabolism pathways relevant for piperine in mam-mals are reported to occur in chicken thereby including substances with comparable structure (Read-Across). Part C) if necessary perform comparative in vitro study. The residual concentration in edible tissues (step 3) can be assessed using the TdB or experimental analytical data. The resulting residue levels are used to calculate con-sumer exposure. In step 4, perform safety evaluation by comparing the Health Based Guidance Value for piperine with the calculated maximum intake and conclude on the risk for consumer.
steps (Fig. 1) using alternate considerations and methods.
2.1. Non-laboratory aspects
2.1.1. Literature search approach, TTC and Cramer class determination
For hazard evaluation (step 1), it is possible to use existing data either from original literature or reviews by competent bodies such as EFSA or JECFA to identify the appropriate NOAEL or ADI. In case public information is not sufﬁcient for determination of a NOAEL or ADI, the respective threshold of toxicological concern (TTC) (Kroes et al., 2004; Munro et al., 2008; EFSA, 2012c) may be considered using the TTC decision tree. For most substances, the substance has to be assigned according to its structure into the Cramer Classes I, II, or III (Cramer et al., 1978). This can be done using online web-tools such asToxTree. The TTC rejects high potency substances such as steroids or anorganic substances whereas speciﬁc thresholds for potential genotoxic substances and organophosphates/carbamates are present.
The literature search was performed using the CAS Registry and RTECS ﬁles, the STN/Toxcenter online database as well as the websites of the following organizations: International Programme on Chemical Safety (IPCS), OECD, Environmental Protection Agency High Production Volume Chemicals Program (EPA HPV), National Toxicology Program (NTP), and EFSA. The search terms were piperine and its CAS number (94-62-2). From the retrieved litera-ture search hits, the relevant information was extracted with focus on toxicology/pre-clinical safety and metabolism/tissue concen-trations in both rats and poultry.
2.1.2. In silico predictions of toxicological endpoints and identiﬁcation of structural analogs
For toxicological endpoints where literature information for piperine was not available, in silico predictions were performed. Carcinogenicity was predicted using the models available from the Lazar website (Helma, 2006). Quantitative toxicological endpoint (i.e. the NAEL) was predicted using the Maximum Recommended Therapeutic Dose (MRTD) prediction model also available from the Lazar website (Maunz and Helma, 2008). In addition, theToxPredict similarity search function with a threshold of 0.7 was used to identify structural analogs. The literature was searched for toxico-logical data of the identiﬁed structural analogs.
2.1.3. Comparative metabolism
For evaluation of comparative metabolism (step 2), we followed three approaches. Part A) consisted of a literature evaluation to identify metabolites of piperine in mammals. In Part B) we evalu-ated the literature on existing metabolism studies for piperine in chickens thereby we included the identiﬁed analogs (see section 2.1.2). Part C) was added and consisted of comparative in vitro metabolism studies in liver preparations of both species (see sec-tion2.2.2), focused on phase I metabolism as conjugation reactions (phase II metabolism) is highly related to detoxiﬁcation reactions.
2.1.4. In silico modeling for calculation of residual concentrations in edible tissues
The TNO transfer factor database (TdB) is described byLeeman et al., 2007. Brieﬂy, the database was constructed from a meta-analysis of public literature on the transfer of chemicals such as pesticides, dioxins, furans, biphenyls, heavy metals, mycotoxins, hormones, veterinary drugs, and nitrosamines from feed into edible animal commodities. Data gathered and recorded in the database included the species (e.g. poultry, swine, ruminants), a chemical's concentration in the feed, feeding periods (animal age and/or duration), residue levels in the animal's edible products, and a number of other parameters. The key information derivable from this database is the transfer factors, which are deﬁned as the ratio of the concentration of the chemical in the edible animal com-modity (mg/kg wet weight) to the concentration of the chemical in the animal feed (mg/kg dry weight). Thus, from the concentration of a chemical added to feed, a prediction is made on the concen-tration of the chemical that could be expected to be found in edible commodities. The compounds included in the transfer database cover a log Po/w range of9.4 to 9.5 (personal communication) by which both hydrophilic and lipophilic compounds are about equally present. Moreover, it was demonstrated that transfer is related to the log Po/w, showing accumulation of compounds in the log Po/w range between 3 and 8 only. Therefore the transfer database can be used for a worst case estimation for transfer of hydrophilic and lipophilic compounds from feed to edible com-modities. Speciﬁcally, one transfer factor is assigned to a substance and edible animal commodity. It should be noted that the transfer database does not predict metabolism of chemicals, as such the transfer from feed to food is considered for piperine only.
The transfer factors speciﬁc for piperine are not present in the database and had to be derived from values available for substances with a similar log Po/w (±0.5) within the database. The log Po/w of 3.69 was used for piperine for the in silico modeling. This log Po/w is an estimated value retrieved fromChemIDPlusand conﬁrmed by EPA EPI Suite software (V4.00). It should be noted that this additive is intended for use only in feed for broilers so a reﬁnement in the selection of data was made to cover poultry data only. Piperine is added to the target species' feed at a concentration of 0.5 ppm with a maximum feeding period of 42 days in broiler chicken. Therefore, a further reﬁnement was made by selecting data covering a limited feeding period of 1e50 days. Considering the log Po/w of the con-cerned substances, animal species and relevant exposure periods, transfer factors were derived from the database for eggs, meat, fat and the edible organs liver and kidney.
2.1.5. Exposure assessment and risk characterization
From the in silico modeling of worst case residual piperine levels in edible tissues (section 2.1.4) and from the experimentally determined residual concentrations (section2.2.3), we calculated consumer exposure by using the guidance for exposure assessment of feed additives brought forward by EFSA (EFSA, 2012a). As the ﬁnal steps we have performed the risk evaluation and compared the estimated Health Based Guidance Value for piperine with the calculated maximum intake.
2.2. Laboratory experiments 2.2.1. Chemicals
Piperine (CAS 94-62-2) was obtained from Acros Organics (pu-rity: 98.7%). Capsaicin was obtained from Sigma (product number 360,376); testosterone (purity:>99.9%) was from SigmaeAldrich. For the poultry feeding studies (see section2.2.3), CRINA®Poultry Plus, containing 0.1% piperine, was added in amounts up to 4500 ppm in the feed.
2.2.2. In vitro experiments
184.108.40.206. Animals used for preparation of liver fractions. For the in vitro experiments livers from male Crl:WI (Han) Wistar rats (Charles River, Sulzfeld, Germany) and female Leghorn Braun chickens (obtained from a local farmer) were used. The animals were not pre-treated with enzyme inducers before sacriﬁce. 220.127.116.11. Preparation of liver S9-fraction and liver microsomes for in vitro metabolism experiments. Liver samples from hen or rat were cut into small pieces, mixed with homogenization buffer (saccha-rose (250 mM), EDTA-Na2(1 mM)), and homogenized with a Potter
Elvehjem homogenizer. Raw cell compartments were separated by centrifugation at 9,000 g (15 min/4 C), which yielded the S9-fraction in the supernatant. The microsomes were collected via sedimentation from the S9-faction by centrifugation at 100,000 g (60 min/0e4C), washed with washing solution (KCl, 150 mM) and
re-centrifuged at 100,000 g (60 min/0e4C). Re-suspension buffer (glutathion 1 mM; EDTA, 1 mM; MgCl2 6H2O, 4 mM; KH2PO4,
0.1 M; glycerol, 20% (v/v), adjusted to pH 7.5) was added to the pellet, which was then re-homogenized.
The rat or hen liver preparations were characterized using the following parameters: protein content according to Bradford, to-tal cytochrome P450 content according to Omura and Sato (1964a,b), and AROD-activities according to Lubet et al. (1990). Since liver microsomes are integral constituents of the S9-fraction and contribute essentially to their metabolic activities, the char-acterization of a liver S9-fraction is simultaneously addressed by the characterization of the associated microsomes prepared from that S9-fraction. This characterization was done to guarantee valid microsome preparations compared to internal, historical controls.
The rat liver microsomes were characterized as follows: total protein content 18.4 mg/mL; cytochrome b5-content 0.5 nmol/mg protein; total cytochrome P450 content 0.6 nmol/mg protein; and EROD, PROD, and BROD-activities were 29.1, 22.4, 63.3 pmol/min/ mg protein, respectively. The hen liver microsomes were charac-terized as follows: total protein content 18.4 mg/mL; cytochrome b5-content 0.2 nmol/mg protein; total cytochrome P450 content <0.01 nmol/mg protein; the EROD and PROD-activities were 0.4 and 0.3 pmol/min/mg protein, respectively. BROD activities were not detectable.
18.104.22.168. Incubation of S9-fraction and microsomes with piperine. Piperine at a nominal concentration of 100
mM and 10
mM were each incubated with the liver microsomes and the S9-fractions obtained from rat or hen. Incubations with microsomes were car-ried out with 0.5 mg microsomal protein/mL incubate. For in-cubations with S9-fraction, 250
mL/mL incubate were used. Piperine in DMSO was added at aﬁnal DMSO concentration of 2.5% (v/v). The incubations were performed at 37C for 2 h including 30
mL/mL incubate of a NADPH-regeneration system. The NADPH-regeneration system contained glucose-6-phosphate mono-sodium salt 131 mg/mL, NADPH 33.3 mg/mL, MgCl21 M (143
mL), and glucose-6-phosphate-dehydrogenase 200
mL/mL (¼ 140 units).
In addition, appropriate controls were processed under identical incubation conditions: (i) testosterone as positive control at 200
mM without addition of piperine, (ii) heat-deactivated (100C, about 10 min) controls with piperine, and (iii) buffer controls with piperine but without microsomes or S9-mix. Further, an extraction efﬁciency control was used that was not incubated but immediately stopped using acetone. After incubation, the proteins were precipitated by adding 1 volume of acetone and centrifuged. From the resulting supernatant, an aliquot was taken for HPLC-UV-analysis. All incubations were carried out in two replicates. Testosterone was incubated in parallel and used as a positive con-trol to show the validity of the test system.
22.214.171.124. Analysis of metabolites in incubates of liver S9-fraction or microsomes. Piperine and its metabolites were separated by HPLC (HP 1200) with UV detection (240 nm) using a Gemini 5
mC18, 150 3 mm column. Gradient elution was a mixture of acetonitrile/ 0.1% HCOOH: v/v (A) and deionized water/0.1% HCOOH: v/v (B) with a constantﬂow rate in all steps of 0.6 mL/min. The gradient was as follows: 0e1 min: kept at 95% B; 1e35 min: gradient change to 10% B; 35e40 min: kept at 10% B; 40e45 min: gradient change to 95% B; 45e55 min: kept at 95% B.
The positive control testosterone and its metabolites were separated by HPLC (Waters Aliance) with Purospher RP18, 150 4 mm column and UV-detection (240 nm). Gradient elution was a mixture of deionized water/25% methanol/0.1% HCOOH: v/v/ v (A) and deionized water/63.5% methanol/1.5% acetonitrile/0.1% HCOOH: v/v/v (B) with a constantﬂow rate of 0.7 mL/min in all steps. The gradient was as follows: 0e10 min: kept at 25% B; 10e35 min: gradient change to 100% B; 35e50 min: kept at 100% B; 50e60 min: gradient change to 25% B; 60e70 min: kept at 25% B. 2.2.3. Residue measurements in in vivo target animal safety studies
Livers were obtained from Ross 308 broiler chicken of both sexes that from hatch until 35 days of age were treated with CRINA® Poultry Plus at dietary concentrations of 0, 450 and 4500 ppm in a target animal safety/tolerance study. CRINA® Poultry Plus is a preparation ofﬂavoring substances used for animal feed and con-tains amongst others piperine (0.1%). Thus, the piperine concen-trations fed to the animals were 0, 0.45, and 4.5 ppm. The birds were housed separated by sex and had unrestricted access to a pelletized broiler diet and drinking water. Each dose group con-sisted of 8 replicate pens (4 pens per sex) with each replicate pen containing 20 birds. On day 0,ﬁve birds per pen were randomly pre-selected. On day 35, the pre-selected birds were sacriﬁced. Their livers were collected, put on ice and subsequently stored at20C until analysis of piperine content. In total 20 liver per sex
were analyzed in the control and 40 animals in the low dose group for the presence of piperine. In the high dose group in total 13 birds were analyzed with a minimum of 1 bird per pen. It was not considered necessary to analyze the remaining livers because all samples were below the LOQ.
Before analysis, the liver samples were thawed at room tem-perature in the dark. An amount of 50
mL of internal standard so-lution (capsaicin) was added to about 200 mg of liver tissue, followed by 1.5 mL of ethyl acetate. The mixture was homogenized using Precellys system and centrifuged. Afterwards 1 mL of the supernatant was evaporated to dryness and then dissolved with 150
mL of acetonitrile. An amount of 10
mL of the mixture was injected into the LC-MS/MS system.
LC analysis was performed using HPLC system (Agilent, USA) with a Sunﬁre C18 (50 4.6 mm, dp ¼ 2.5
mm, Waters) column. Gradient elution was a mixture of 20 mM ammonium acetate (A) and acetonitrile (B) with a constantﬂow rate of 0.8 mL/min in all steps. The gradient was as follows: 0.00e0.50 min: kept at 20% B;
0.50e20.00: gradient change to 90% B; 20.0e25.0: kept at 90% B; 25.0e30.0: gradient change to 20% B and 30.0e35.0: kept at 20% B. A triple quadrupole API 4000 mass spectrometer (AB Sciex, Tor-onto, Canada) equipped with a Turbo Ionspray (ESI) in positive ion mode was used. Quantiﬁcation was done by applying external calibration using internal standard. The criteria for identiﬁcation and quantiﬁcation were the retention time and MRM transition m/z 286/201 for piperine and m/z 306/137 for the internal standard (capsaicin).
3.1. Hazard evaluation (step 1)
Using the ToxTree tool, piperinee a non-mutagenic substance was determined to be a Cramer Class III compound.
Literatureﬁndings: Piperine was reported to be not mutagenic in the Ames Test (EFSA, 2011), negative in two in vitro micronucleus (MNT) assays (Singh et al., 1994; Thiel et al., 2014), negative in several in vivo MNT studies (EFSA, 2011) and in a recently con-ducted in vivo MNT (Thiel et al., 2014). Kinetic and metabolism data were identiﬁed for mammalian species including laboratory rat (Bajad et al., 2002, 2003a; Suresh and Srinivasan, 2010) and human (Sethi et al., 2009; Hoelzel and Spiteller, 1984; Kakarala et al., 2010). However, metabolism data in target species were not available. Piperine was tested in rats for 8 weeks as dietary mixture at con-centrations approximately equivalent to doses of 10e20 mg/kg bw/ d (Bhat and Chandrasekhara, 1986). In this study, no adverse effects on hematological and clinical chemistry parameters were seen. In other experiments focusing on the male reproductive tract, doses ranging from 1 to 100 mg/kg bw/d given by gavage to male rats for 30 days (Malini et al., 1999; D'Cruz et al., 2005, 2008). The lowest dose administered, 1 mg/kg bw/d, produced no adverse effects on sexual organs and sperm. Doses equivalent to 5 mg/kg bw/d or higher produced reduced weights of male sexual organs including testis, cauda epididymides, vas deferens, seminal vesicle and ventral prostate. Sperm counts, motility and viability were reduced. Histopathologically, desquamation of spermatocytes, round and elongated spermatids and their accumulation in the lumen of seminiferous tubules was seen. In a further experiment with a study design focusing on liver (Gagini et al., 2010), a dose level of 1.12 mg/kg bw/d was given orally for 23 days to rats. The livers of the animals were examined histopathologically without signiﬁcant ﬁndings. Recently, EFSA summarized the results of a 90-day study with piperine. EFSA concluded that the NOAEL of this study was 5 mg/kg bw/d based on a dose-dependent increase in plasma cholesterol levels in males at the next higher doses (EFSA, 2015).
Standard experimental carcinogenicity studies are not available for piperine and therefore in silico predictions were performed using the in silico models available from theLazar website(Helma, 2006). Piperine was predicted to be a non-carcinogen by those in silico models. The conﬁdence for the predictions was above 0.025 for 4 out of the 6 models: The DSSTox Carcinogenicity Potency DBS MulitCellCall and the DSSTox Carcinogenic Potency DBS Mouse model showed conﬁdence values below 0.025.
Interestingly, several published studies suggest an onco-preventative activity rather than a carcinogenic activitye a result which supports the predictions of the absence of carcinogenicity (Wbra et al., 1992; Selvendiran et al., 2003, 2004, 2005a, 2005b, 2006; Manoharan et al., 2009; Krishnakumar et al., 2009; Vellaichamy et al., 2009). In these experimental studies, piperine was administered subcutaneously or orally together with known carcinogenic substances using intermittent or continuous piperine exposures at dose levels up to 100 mg/kg bw/d. In none of the studies did the co-administration of piperine at the doses used
increase the incidence in benign or malignant tumors in the investigated organs. In contrast, piperine co-administration reduced the incidence of tumors, reduced the tumor volumes, and the tumor burden in animals in most of the studies.Wiseman et al. (1987), reported an absence of carcinogenic activity in pre-weanling mice that received 4 intraperitoneal injections and were then observed for 18 months. The overall weight of evidence therefore suggests that piperine does not have a carcinogenic potential.
Conﬂicting results are available in terms of immunomodulation: The administration of piperine at dose levels of 0, 1.12, 2.25, or 4.5 mg/kg bw for 5 consecutive days to 6-week old male Swiss mice resulted in a toxicological proﬁle for immunotoxicity as indicated by reduced relative weights of spleen, thymus, and mesenteric lymph node, as well as the lower number of white blood cells in the high dose group. Several functional tests for humoral and cellular immunity were also performed and indicated changes in immune function. The authors concluded that the high dose had a consistent immunosuppressive effect and that the NOAEL from this study was 1.12 mg/kg bw/d (Dogra et al., 2004). In contrast, in mice a pro-tective effect of piperine on cadmium-induced toxicity to the im-mune system at comparable dose levels was reported byPathak and Khandelwal (2009).
Further information related to pharmacological effects both in vitro and in vivo are available in the public domain (as reviewed e.g. by Perez Gutierrez et al., 2013 or Meghwal and Goswami (2013)). However, with this information a NOAEL for risk assess-ment cannot be deduced.
Judicious choice of the in vitro dose concentrations relative to the in vitro biological component's size, collection of culture me-dium for quantiﬁcation of parent and metabolite free concentra-tions at different time points during the assays, and their association to the assay's endpoint, would be relevant for the use of in vitro information in (quantitative) risk assessment (Schroeder et al., 2011; Leist et al. 2012,Blaauboer et al., 2012; Bessems et al., 2014), however, such details were not available.
In conclusion, the most comprehensive study (90-day study in rats summarized byEFSA (2015)) results in a NOAEL of 5 mg/kg bw/ d whereas adverse effects on male sexual organs were reported at the same dose by other authors. A dose of 1 mg/kg bw/d was not associated with toxicity to male sexual organs (Malini et al., 1999; D'Cruz et al., 2005, 2008). It is not explicitly mentioned in the EFSA (2015) report whether male reproductive organs were examined, however, the reported study is a regulatory toxicological study and therefore the male reproductive organs were most likely examined. Nevertheless, the available literature data make the deduction of an overall NOAEL from experimental data difﬁcult mainly due to conﬂicting and missing information (especially on the toxicological endpoint reprotoxicity). For the purpose of this manuscript, we used 1 mg/kg bw/d as a NOAEL. Applying the EFSAs standard uncertainty factors of 10 10 2 for inter-and intra-species variations and for extrapolation from subchronic to chronic exposure (EFSA, 2012b), a Health Based Guidance Value of 5
mg/kg bw/d can be calculated.
Considering the uncertainty in terms of the NOAEL described above and because piperine is not genotoxic, we decided to consider as well the TTC value for Cramer Class III compounds. The original Cramer Class III TTC threshold was found to be 90
mg/d or 1.5
mg/kg bw/d (EFSA, 2012; Kroes et al., 2004). Recently, the orig-inal dataset was re-evaluated with focus on organophosphates including carbamates, organohalogens and remaining Cramer Class III compounds. This allowed the deﬁnition of a reﬁned Cramer Class III threshold being 4
mg/kg bw/d (Leeman et al., 2014).
A further possibility is to predict quantitative toxicological endpoints by using the MRTD model which is available from the
Lazar website (Maunz and Helma, 2008): The predicted MRTD was 0.014 mmol/d which is considered a human LAEL (Schilter et al., 2014). The predicted MRTD is equivalent to 60
mg/kg bw/ d thereby considering piperines' molecular weight of 285.34 g/mol and a 70 kg adult person (EFSA, 2012b). Applying a conservative uncertainty factor of 3 to convert the LAEL into a human NAEL (ECHA, 2012), this results in 20
mg/kg bw/d which is at the same order of magnitude as the derived Health Based Guidance Value and the TTC thresholds for a Cramer Class III compound as dis-cussed above.
The similarity search using ToxPredict software and a similarity search threshold of 0.7 revealed two structural related substances: antiepilepsirine (CAS 23434-86-8) and piperonic acid (CAS 5285-18-7). Piperonic acid lacks the piperidine group of piperine (Fig. 2). For both substances comprehensive toxicological data were not identiﬁed in peer reviewed scientiﬁc literature. Antiepilepsirine has anticonvulsant properties and has been used for decades in the treatment of epilepsy in China with therapeutic doses of 10 mg/kg bw/d. At doses above 200 mg/kg bw/d clinical side effects con-sisting of mild vertigo, drowsiness, anorexia and/or nausea can occur (Pei 1999).
3.2. Comparative metabolism from literature (step 2) 3.2.1. Piperine metabolism in rat (Fig. 1, step 2, part A)
In rat urine, the metabolites piperonylic acid, piperonal, vanillic acid, and piperonyl alcohol were found; in the bile, piperic acid was found, suggesting that the amide eCO-N-bond was cleaved (Bhat and Chandrasekhara, 1987). However, the presence of piperic acid in bile was not conﬁrmed (Bajad et al., 2002) and the urinary metabolites previously described could not be conﬁrmed (Bajad et al., 2003b). Instead, in rat plasma and urine a major metabolite was characterized by MS and 1H NMR: 5-(2,4-methylenedioxy phenyl)-2E,4E-pentadienoic acid-N-(3-yl propi-onic acid)-amide resulting from oxidation of the piperidine ring in the 4-position followed by loss of an ethylene group (seeFig. 5). In addition, a minor metabolite was proposed to be E,E-1-[5-(3-methoxy-4-hydroxyphenyl)-1-oxo-2,4-pentadienyl]piperidine (Bajad et al., 2003a).
In human urine four metabolites were identiﬁed: 5-(3,4-dihydroxyphenyl)valeric acid piperidide, its 4-hydroxylated deriv-ative, and 5-(3,4-dihydroxyphenyl)-2,4-pentaidenoic acid piper-idide as well as its 4-hydroxylated derivative (Hoelzel and Spiteller, 1984). The apparent differences between the metabolites observed in rats vs. humans may be explained by the stringent conditions (pH2 adjusted with 4 N HCl) used during metabolite isolation re-ported byBhat and Chandrasekhara (1987). This suggests that the cleavage of the amide bond may have been produced artiﬁcially.
Overall, metabolites observed in rats and humans are consid-ered to be derived from identical metabolism pathways. The steps for mammalian metabolism of piperine are: cleavage of an amide-group (A), hydroxylation of the piperidine ring in position 4 (D), opening of the benzodioxole group, and reduction of the double-bonds (Table 1andFig. 5).
Metabolism of the structural related antiepilepsirine, studied in isolated perfused rat liver, resulted in two metabolites: 3,4-methylene dioxycinnamyl hydroxypiperidine and 4-hydroxy-3-methoxycinnamyl piperidine (Dong et al., 1989). Information on the metabolism of piperonic acid was not found; however, it is most likely the initial metabolite of piperine once the amide-function is cleaved (Fig. 5, reaction A).
3.2.2. Piperine metabolism in chicken (part B)
Relevant information on chicken metabolism was not available in the published literature. From the existing primary literature or
from literature reviews (Pan and Fouts, 1978a and b) on existing xenobiotic metabolism pathways in poultry, it remained unclear whether the identiﬁed steps on mammalian piperine metabolism exist in chicken (Table 1). Metabolism information in target animal species for the two structurally related substances was not found. Therefore no prediction of metabolism of piperine in chicken could be made and subsequently a comparative in vitro metabolism study was performed.
3.2.3. Comparative in vitro metabolism study using piperine (part C)
A comparative in vitro metabolism study using male rat liver fraction and microsomes and female Brown Leghorn hen liver S9-fraction and microsomes was conducted. Piperine was metabo-lized using liver microsomes and liver S9-fraction of rats and hens under the chosen incubation conditions of the study (Figs. 3 and 4). In each microsomal and S9-fraction incubate of rats and hens, 10e13 metabolites could be separated by HPLC. All metabolites had shorter retention times in the RP-chromatogram than the parent compound (retention time for piperine, isomer 1 about 22.6 min and isomer 2 about 23.0 min) and are therewith more polar than piperine. The E/Z-isomerization of piperine is a typical behavior and is also known from other matrices (Ternes and Krause, 2002). The most polar metabolites have retention times of about 9e11 min. Characteristic patterns of metabolites that occur in both in vitro metabolism systems of both species elute at about 16e17, 17e18, and 19e20 min (Figs. 3 and 4). A metabolite with a retention time of 14.9 min was detected in liver microsomes of hens and in liver S9-fraction of hens and rats. A minor metabolite with a retention time of 14.3 min was detected in S9-fraction of hens only. However, taken together, it can be concluded from the comparison of the performed in vitro experiments that the observed metabolic proﬁles of piperine in liver systems of rats and hens are similar. 3.3. Residue levels (step 3)
3.3.1. Residue estimation using the TNO transfer factor database Piperine was evaluated for its transfer from feed to edible commodities using the transfer database. No data for piperine were present in the database. For the case under consideration, the transfer factor was derived based on the physical chemical prop-erties of piperine, more speciﬁcally on the log of the octanol:water partition coefﬁcient (log Po/w). For the log Po/w, a very good cor-relation with the respective transfer factors is known (Leeman et al., 2007).
The transfer factors of piperine from feed to edible commodities were based on 1) the concentration of piperine in the feed, 2) the transfer derived using the 95th percentile of substances with a log Po/w in the same range as piperine (range used of± 0.5), 3) the relevant animal species, and 4) the relevant exposure duration. Although the initial approach was for meat, fat, liver and kidney of broilers only, transfer factors were also calculated for eggs and are provided as additional information. The results are presented in Table 2. Based on these transfer factors the piperine concentrations were calculated for each of the various edible commodities (see Table 3).
The transfer factors derived from the TdB are all based on 10 or more data points and are therefore considered of relevance for piperine. Because the selected dataset of transfer factors are based on the 95th percentile of the distribution, the calculated residual piperine concentrations are considered to represent a reasonable worst case for subsequent consumer risk assessment.
3.3.2. Residue determination by analytical means
estimated with the TdB, predicted that the liver would have the highest amount. However, these amounts were only in the upper ppb range, which was used to guide the sensitivity of an analytical method.
The analytical method developed for residue level measure-ments in liver had a limit of quantiﬁcation (LOQ) of 0.1
mg/kg (0.1 ppb). 73 samples from the control (0 ppm), 450 ppm and 4500 ppm CRINA®Poultry Plus groups, corresponding to 0, 0.45 and 4.5 ppm piperine, respectively, were analyzed for piperine. In each of the analyzed liver samples, the determined concentrations were below the LOQ.
3.4. Risk assessment (step 4)
A consumer exposure estimate for piperine was performed using the calculated worst case residual piperine levels and the actual measured piperine residual levels based on the in-life exposure situation. The ﬁrst exposure estimation was designed in a
conservative way and therefore we used only the high consuming persons acute consumptionﬁgures for adult and toddlers as given in the respective EFSA guidance (EFSA, 2012a) (Table 2). Acute intake estimates usually result in higher exposure and were therefore used also to assess chronic exposures. The acute intake levels for piperine were calculated to be 1.3
mg/kg bw/d for adults (78
mg/d) and 1.1
mg/ kg bw/d for toddlers (13.2
mg/d) based on the transfer database estimation, and 0.002
mg/kg bw/d for adults and 0.001
mg/kg bw/ d for toddlers based on the measured concentrations in the broiler liver (seeTable 4). Comparison with available intake data coming from use of piperine in food, reveals that intakes of piperine by the consumer is about 0.3
mg/kg bw/d (20
mg/person/d) (EFSA, 2011) which is in a comparable range and below the deduced Health Based Guidance Value of 5
mg/kg bw/d, below the original and reﬁned TTC threshold level of 1.5
mg/kg bw/d and 4
mg/kg bw/d for Cramer Class III compounds, respectively (Kroes et al., 2004; Leeman et al., 2014). The intake is also below the predicted human NAEL of 20
mg/kg bw/ d. The highest calculated acute exposure is 1.3
mg/kg bw/d, which
Fig. 3. Metabolite patterns (HPLC-UV chromatogram) of piperine; 100mM in liver microsomes from male Wistar-Han Rats and female Leghorn Braun hens: from top to bottom: Heat deactivated controls (HDC) from rat and hen, Active incubations (AI) from rat and hen, piperine isomers elute after 22.5 and 23 min.
can be considered as a worst case exposure estimate, and the resulting margin of exposure (MoE) is 15.4(= 20
mg/kg bw/d / 1.3
mg/ kg bw/d) for the predicted human NAEL. The MoE is high enough to account for an uncertainty factor of 10 for intraspecies variability (EFSA, 2012b). The use of further uncertainty factors such as extrapolation of duration or for interspecies extrapolation is not considered relevant because we consider comparable exposure durations. Based on these considerations, the use of piperine as a feed ﬂavoring at a concentration of 0.5 ppm in chicken feed is considered to pose no unacceptable risk for consumer.
In this paper we describe combination of non-animal alternative methods for hazard characterization to fulﬁll the feed additives guidance for consumer safety assessment as detailed by theEFSA (2012a). Instead of performing in vivo toxicological, kinetics, metabolism, and residue studies, we used alternative, non-animal approaches. These included information from public literature, in silico predictions of qualitative and quantitative toxicological end-points, Read-Across techniques to assess metabolism and toxico-logical endpoints, an in vitro comparative metabolism study and an in silico-based estimation of residue levels predicted for the edible tissues.
As a case study we selected the feed additive piperine, which was added in a feeding study to poultry feed at a concentration of 0.5 ppm and a treatment duration of 42 days, the maximum feeding period for fattening chickens in the broiler stage.
Determination of a Health Based Guidance Value to use for hazard characterization (Step 1) was attempted by doing a review of existing data in the published literature which clearly shows the absence of mutagenicity or genotoxicity. The available literature consisted of a summary of a regulatory toxicological study, of re-ports of dedicated studies that investigated speciﬁc endpoints such as testicular toxicity, chemo-preventive activity or interaction with the immune system. However, the data were conﬂicting because the NOAEL from the regulatory 90-day toxicity study (5 mg/kg bw/ d) was reported in the literature to produce adverse effects on male reproductive organs (EFSA, 2015; Malini et al., 1999; D'Cruz et al., 2005, 2008). We therefore used as the NOAEL, the dose which produced no effects on male sexual organs which was 1 mg/kg bw/ d to deduce a Health Based Guidance Value of 5
mg/kg bw/ d considering the guidance fromEFSA (2012b). To account for the above mentioned uncertainties, we used the TTC concept as described byKroes et al. (2004)in addition. Piperine is a Cramer Class III substance for which a threshold of 1.5
mg/kg bw/d is considered to be the systemic exposure threshold below which toxicity concerns are minimal. In addition, we estimated a Health
Fig. 4. Metabolite patterns (HPLC-UV chromatogram) of piperine; 100mM in liver S9 fraction from male Wistar-Han Rats and female Leghorn Braun hens: from top to bottom: Heat deactivated controls (HDC) from rat and hen, Active incubations (AI) from rat and hen, piperine isomers elute after 22.5 and 23 min.
Based Guidance Value from the available in silico tools and pre-dicted the MRTD that was converted into a human NAEL of 20
For step 2, which is the comparison of metabolites formed in rats versus those seen in poultry, based only on available literature we could not reach what we considered to be a reliable conclusion. From the review ofPan and Fouts (1978a and b)it is evident that many of the metabolism steps occurring in rats, such as conjugation of phenolic compounds with sulfate or glucuronic acid, conjugation of carboxylic acids, or ester hydrolysis, are also well documented in poultry. However, from the literature we could not conﬁrm that poultry would use the major steps for piperine metabolism found in mammals, including amide-bond cleavage, hydroxylation of a piperidine-ring or opening of a benzodioxole group (seeTable 1and Fig. 4). Likewise, no metabolism data of structural analogs in
Fig. 5. Metabolites of piperine described for mammals.
Comparison of principal metabolic pathways of piperine.
Major metabolism step in mammals Occurrence in chicken for the following substances
A: Cleavage of amide-group No data
B: Opening of the1,3-benzodioxole group No data C: Reduction of double-bonds No data D: Hydroxylation of piperidine ring in position 4 with/without loss of ethylene group No data
Calculated transfer factors for chicken products using the TNO transfer factor database.
Animal product GM GSD Median P90 P95 P99 N Max Eggs 0.03 0.04 0.01 0.11 0.14 0.16 29 0.16 Meat 0.02 0.04 0.00 0.03 0.19 0.19 17 0.19 Fat 0.01 0.01 0.01 0.03 0.04 0.04 14 0.04 Edible offals 0.09 0.20 0.02 0.18 0.29 0.99 25 0.99 GM: geometric mean.
GSD: geometric standard deviation. Pxx: xxthpercentile.
N: amount of data underlying the statistical data.
poultry was found in the public domain. Therefore, we performed an in vitro metabolism comparative study that clearly showed similar metabolites proﬁles for rat and chicken (Fig. 3). The results of this in vitro metabolism study together with the available liter-ature information increased the level of conﬁdence that the rat metabolism study was a reliable predictor of poultry metabolism of piperine.
Overall, from our experience with the present study, the scope of our approach would be more widely applicable if the metabolic pathways for a broader number of species were readily available and accessible. Expansion of existing databases beyond poultry to other consumable livestock such asﬁsh or ruminants would pro-vide a useful platform for further development of a reduced animal approach to food or feed additives risk assessments.
For step 3, we successfully used the TdB to estimate the worst case piperine residual amounts in edible tissues. The use of the TdB allowed us during determination of residues under normal in-life situations to focus on the tissue with the expected highest resid-ual concentration. Another helpful information which was obtained from the TdB was the required sensitivity of the method for analysis because clear guidance is given on the required sensitivity of the analytical method. The use of the 95th percentile of the database distribution on residual piperine levels in edible tissues conﬁrmed the assumed conservative (worst case) residue estimations by the TdB, being considerably higher than the actual measured concen-tration in tissues from exposed target animals under normal in-life conditions. Thus, using the transfer database from a food safety perspective was also shown to be suitable.
In step 4, we demonstrated, using the guidance from EFSA (2012a), that the consumer's calculated intakes of piperine were considerably lower than the estimated Health Based Guidance Value, the TTC value and the predicted NAEL. Therefore, we concluded that piperine could be safely used in poultry nutrition at the routinely applied dietary concentrations.
In this case study with the feed additive piperine, we have
successfully used a combination of in vitro/in silico based alternative methods for the consumer risk assessment. Our approach is based on (i) read-across techniques to assess metabolism in target animal complemented with in vitro comparative metabolism work, (ii) calculated residual concentrations derived from a database thereby using worst case i.e. highest distributions within the database, and (iii) comparison of the determined exposure with the derived Health Based Guidance Value, the respective threshold of the TTC concept and/or predicted quantitative toxicological endpoints. The in silico residue modeling used for piperine showed to be a worst case approach: the modeled residual levels were considerably higher as compared to the measured residual levels. This strategy provides an alternative to animal testing without jeopardizing consumer safety. The use of in vitro/in silico methods may therefore also serve in future for risk assessment for other feed additives and was shown that the alternative methods results in conservative outcomes.
Conﬂicts of interest
A Thiel, S Etheve, and JR Plautz are employees of DSM Nutri-tional Products AG. E Fabian is an employee of BASF SE. WR Leeman is an employee of TNO.
None of the authors are aware of a conﬂict of interest except employment.
We would like to thank Mrs. Manuela Baur, Mrs. Elodie Chenal and Mrs. Christiane Grunenwald (DSM Nutritional Products, Analytical Research Center) who developed the method and con-ducted the measurements of piperine in liver samples.
We would like to thank Dr. Paul Beilstein for thorough review of the manuscript.
Transparency document related to this article can be found
Calculated versus measured residue levels of piperine in liver.
Concentration in feed Calculated concentration in edible commodity using TNO transfer model Measured concentration in liver Egg Meat Fat Kidney and liver
(ppm) (mg/kg wet weight)
0.5 0.07 0.1 0.02 0.15 <0.0001
The concentrations in edible commodity (mg/kg wet weight) were calculated as follows: transfer factor (seeTable 2) x concentration of the substance in the animal feed (mg/ kg dry weight (ppm)).
Acute adult and toddlers exposure estimates based on calculated residue levels and comparison with TTC level. Population Reference level for risk
Consumer exposurecresulting from consumption of edible chicken tissues based on
TTC ADI Calculated worst case residue levelsa Measured residue levelsb
Adults 1.5 n.a. 1.3 0.002
Toddlers 1.5 n.a. 1.1 0.001
aUsing calculated residue levels speciﬁed inTable 3.
bUsing measured residue levels speciﬁed inTable 3thereby assuming identical amounts in each tissue. From the TdB data it can be argued that other edible commodities
contain less residue than liver.
c Consumer exposure was calculated for high consuming adult and toddlers thereby assuming acute exposure to reﬂect worst case. According to EFSA guidance (EFSA, 2012a) the following consumption data are applicable for this exposure scenario: toddlers weighing 12 kg consume 135 g meat/day; adults weighing 60 kg consume 390 g meat/day, 170 g liver/day, 100 g kidney/day, 40 g fat/day, eggs not given.
online athttp://dx.doi.org/10.1016/j.yrtph.2015.06.006. References
Bajad, S., Singla, A.K., Bedi, K.L., 2002. Liquid chromatographic method for deter-mination of piperine in rat plasma: application of pharmacokinetics. J. Chromatogr. B 776, 245e249.
Bajad, S., Khajuria, R.K., Suri, O.P., Bedi, K.L., 2003a. Characterisation of a new minor urinary metabolite of piperine, an omnipresent food component, by LC-MS/MS. J. Sep. Sci. 26, 943e946.
Bajad, S., Coumar, M., Khajuria, R., Suri, O.P., Bedi, K.L., 2003b. Characterisation of a new rat urinary metabolite by LC/NMR/MS studies. Eur. J. Pharm. Sci. 19, 413e421.
Bessems, J.G., Loizou, G., Krishnan, K., Clewell, H.J., Bernasconi, C., Bois, F., Coecke, S., Collnot, E.-M., Diembeck, W., Farcal, L.R., Geraets, L., Gundert-Remy, U., Kramer, N., Küsters, G., Leite, S.B., Pelkonen, O.R., Schr€oder, K., Testai, E., Wilk-Zasadna, I., Zaldívar-Comenges, J.-M., 2014. PBTK modelling platforms and parameter estimation tools to enable animal-free risk assessment recommen-dations from a joint EPAAe EURL ECVAM ADME workshop. Regul. Toxicol. Pharmacol. 68, 119e139.
Bhat, B.G., Chandrasekhara, N., 1986. Lack of adverse inﬂuence of black pepper, its oleoresin and piperine in the weanling rat. J. Food Saf. 7, 215e233.
Bhat, B.G., Chandrasekhara, N., 1987. Metabolic disposition of piperine in the rat. Toxicology 44, 99e106.
Blaauboer, B.J., Boekelheide, K., Clewell, H.J., Daneshian, M., Dingemans, M.M.L., Goldberg, A.M., Heneweer, M., Jaworska, J., Kramer, N.I., Leist, M., Seibert, H., Testai, E., Vandebriel, R.J., Yager, J.D., Zurlo, J., 2012. The use of biomarkers of toxicity for integrating in vitro hazard estimates into risk assessment for humans. ALTEX 29 (4), 411e425.
Chem IDPlus, Available at: http://chem.sis.nlm.nih.gov/chemidplus/ (accessed 28.02.14).
Cramer, G.M., Ford, R.A., Hall, R.L., 1978. Estimation of toxic hazarde a decision tree approach. Food Cosmet. Toxicol. 16, 255e276.
Cortright, K.A., Wetzlich, S.E., Craigmill, A.L., 2009. A PBPK model for midazolam in four avian species. J. Veterinary Pharmacol. Ther. 32 (6), 552e565.
D'Cruz, S.C., Mathus, P.P., 2005. Effect of piperine on the epididymides of adult rats. Asian J. Androl. 7 (4), 363e368.
D'Cruz, S.C., Vaithinathan, S., Saradha, B., Mathur, P.P., 2008. Piperine activates testicular apoptosis in adult rats. J. Biochem. Mol. Toxicol. 22 (6), 382e388.
Dogra, R.K.S., Khanna, S., Shanker, R., 2004. Immunotoxicological effects of piperine in mice. Toxicology 196, 229e236.
Dong, S.N., Bai, F., Yang, H.J., Lou, Y.Q., Liang, W.S., 1989. Studies on the metabolism of antiepilepserine in isolated perfused rat liver. Acta Pharm. Sin. 24 (4), 241e245.
EC, 2011. European Union Register of Feed Additives Pursuant to Regulation (EC) No 1831/2003. Appendixes 3b& 4, Annex : List of additives (Status: Released 28 February 2011), Available at:http://www.platform-fefana.org/Website/DOCS/ comm_register_feed_additives_1831-03.pdf. accessed on 28-February-2014. ECHA, 2012. Guidance on Information Requirements and Chemical Safety
Assess-ment Chapter R.8: Characterisation of Dose (Concentration)-response for Hu-man Health, Version 2.1. November 2012, Available at:http://echa.europa.eu/ documents/10162/13632/information_requirements_r8_en.pdf (accessed 05.11.14).
ECHA, 2014. The Use of Alternatives to Testing on Animals for REACH. ECHA-14-A-08-EN, Available at: http://echa.europa.eu/documents/10162/13639/ alternatives_test_animals_2014_summary_en.pdf(accessed 21.11.14). EFSA, 2009. Opinion of the Scientiﬁc Committee on a request from EFSA on existing
approaches incorporating replacement, reduction and reﬁnement of animal testing: applicability in food and feed risk assessment. EFSA J. (2009) 1052, 1-77, Available at:http://www.efsa.europa.eu/en/efsajournal/doc/1052.pdf(accessed 21.11.14).
EFSA, 2011. Scientiﬁc opinion on ﬂavouring group evaluation 86, revision 1 (FGE.86Rev1): consideration of aliphatic and aromatic amines and amides evaluated by JECFA (65th meeting), EFSA panel on food contact materials, en-zymes, ﬂavourings and processing aids (CEF). EFSA J. 9 (4), 1926. http:// dx.doi.org/10.2903/j.efsa.2011.1926.
EFSA, 2012a. EFSA Panel on additives and products or substances used in animal feed (FEEDAP): guidance for establishing the safety of additives for consumer. EFSA J. 10 (1), 2537.http://dx.doi.org/10.2903/j.efsa.2012.2537.
EFSA, 2012b. EFSA Scientiﬁc Committee; guidance on selected default values to be used by the EFSA Scientiﬁc Committee, Scientiﬁc Panels and Units in the absence of actual measured data. EFSA J. 10 (3), 2579.http://dx.doi.org/10.2903/ j.efsa.2012.2579.
EFSA, 2012c. Scientiﬁc opinion on exploring options for providing advice about possible human health risks based on the concept of threshold of toxicological concern (TTC). EFSA J. 10 (7), 2750.http://dx.doi.org/10.2903/j.efsa.2012.2750. EFSA, 2015. EFSA CEF Panel (EFSA panel on food contact materials, enzymes,
ﬂa-vourings and processing aids), 2015. Scientiﬁc opinion on ﬂavouring group evaluation 86, revision 2 (FGE.86Rev2): consideration of aliphatic and arylalkyl amines and amides evaluated by JECFA (65th meeting). EFSA J. 13 (1)http:// dx.doi.org/10.2903/j.efsa.2015.3998, 3998, 49.
EMA, 2011. European Medicines Agency, Committee for Medicinal Products for Veterinary Use (CVMP) VICH GL47: Studies to Evaluate the Metabolism and Residue Kinetics of Veterinary Drugs in Food-producing Animals: Laboratory
Animal Comparative Metabolism Studies. 14-March-2011 EMA/CVMP/VICH/ 463104/2009.
Gagini, T.B., Silva, R.E., Castro, I.S., Soares, B.A., Lima, M.E.F., Brito, M.F., Mazur, C., Direito, G.M., Danelli, M., 2010. Oral administration of piperine for the control of aﬂatoxin intoxication in rats. Braz. J. Microbiol. 41, 345e348.
Helma, C., 2006. Lazy structure-activity relationships (lazar) for the prediction of rodent carcinogenicity and Salmonella mutagenicity. Mol. Divers. 10, 147e158.
Hoelzel, C., Spiteller, G., 1984. Piperinee an example of individually different (polymorph) metabolism of an omnipresent nutrition component. Liebigs Ann. Chem. 1319e1331.
Kakarala, M., Dubey, S.K., Tarnowski, M., Cheng, C., Liyanage, S., Strawder, T., Tazi, K., Sen, A., Djuric, Z., Brenner, D.E., 2010. Ultra-lowﬂow liquid chromatography assay with ultraviolet (UV) detection for piperine quantitiation in human plasma. J. Agric. Food Chem. 58, 2594e6599.
Krishnakumar, N., Manoharan, S., Palaniappan, L.R.M., Venkatachalam, P., Manohar, M.G.A., 2009. Chemopreventive efﬁcacy of piperine in 7,12-dimehtly benz (a) anthracene (DMBA)-induced hamster buccal pouch carcinogenesis: an FT-IR study. Food Chem. Tox. 47, 2813e2820.
Kroes, R., Renwick, A.G., Cheeseman, M., Kleiner, J., Mangelsdorf, I., Piersma, A., Schilter, B., Schlatter, J., van Schothorst, F., Vos, J.G., Würtzen, G., 2004. Struc-ture-based thresholds of toxicological concern (TTC): guidance for application to substances present at low levels in the diet. Food Chem. Toxicol. 42, 65e83. Lazar website:http://lazar.in-silico.ch/predict.
Leeman, W.R., van den Berg, K.J., Houben, G.F., 2007. Transfer of chemicals from feed to animal products: the use of transfer factors in risk assessment. Food Addit. Contam. 24 (1), 1e13.
Leeman, W.R., Krul, L., Houben, G.F., 2014. Reevaluation of the munro dataset to derive more speciﬁc TTC thresholds. Regul. Toxicol. Pharmacol. 69, 273e278.
Leist, M., Lidbury, B.A., Yang, C., Hayden, P.J., Kelm, J.M., Ringeissen, S., Detroyer, A., Meunier, J.R., Rathman, J.F., Jackson Jr., G.R., Stolper, G., Hasiwa, N., 2012. Novel technologies and an overall strategy to allow hazard assessment and risk pre-diction of chemicals, cosmetics, and drugs with animal-free methods. Altex 29, 373e388.
Lubet, R.A., Syi, J.L., Nelson, J.O., Nims, R.W., 1990. Induction of hepatic cytochrome P-450 mediated alkoxyresoruﬁn O-dealkylase activities in different species by prototype P-450 inducers. Chem. Biol. Interact. 75, 325e339.
MacLachlan, D.J., 2011. Estimating the transfer of contaminants in animal feedstuffs to livestock tissues, milk and eggs: a review. Anim. Prod. Sci. 51, 1067e1078.
Malini, I., Manimaran, R.R., Arunakaran, J., Aruldhas, M.M., Govindarajulu, P., 1999. Effects of piperine on testis of albino rats. J. Ethnopharmacol. 64, 219e225.
Manoharan, S., Balakrishnan, S., Menon, V.P., Alias, L.M., Reena, A.R., 2009. Che-mopreventive efﬁcacy of curcumin and piperine during 7,12-dimethylbenz[a] anthracene-induced hamster buccal pouch carcinogenesis. Singap. Med. J. 50 (2), 139e146.
Maunz, A., Helma, C., 2008. Prediction of chemical toxicity with local support vector regression and activity-speciﬁc kernels. SAR QSAR Environ. Res. 19 (5e6), 413e431.
Meghwal, M., Goswami, T.K., 2013. Piper nigrum and piperine: an update. Phy-thotherapy Res. 27, 1121e1130.http://dx.doi.org/10.1002/ptr.4972.
Munro, I.C., Renwick, A.G., Danielewska-Nikiel, B., 2008. The threshold of toxico-logical concern (TTC) in risk assessment. Toxicol. Lett. 180, 151e156.
Omura, T., Sato, R., 1964a. The carbon monoxide-binding pigment of liver micro-somes. I. evidence for its hemoprotein nature. J. Biol. Chem. 239, 2370e2378.
Omura, T., Sato, R., 1964b. The carbon monoxide-binding pigment of liver micro-somes: II. Solubilization, puriﬁcation, and properties. J. Biol. Chem. 239, 2379e2385.
Pan, H.P., Fouts, J.R., 1978a. Drug metabolism in birds: part 1. Drug Metab. Rev. 7, 1e140.
Pan, H.P., Fouts, J.R., 1978b. Drug metabolism in birds: part 2. Drug Metab. Rev. 7, 141e253.
Pathak, N., Khandelwal, S., 2009. Immunomodulatory role of piperine in cadmium induced thymic atrophy and splenomegaly in mice. Environ. Toxicol. Pharma-col. 28, 52e60.
Pei, Y.Q., 1999. A review of pharmacological and clinical use of antiepilepsirine and its derivatives. Res. Commun. Pharmacol. Toxicol. 4 (3e4), II13eII23.
Perez Gutierrez, R.M., Neira Gonzalez, A.M., Hoyo-Vadillo, C., 2013. Alkaloids from piper: a review of its phytochemistry and pharmacology. Mini Rev Med. Chem. 13, 163e193.
Schilter, B., Benigini, R., Boobis, A., Chiodini, A., Cockburn, A., Cronin, M.T.D., Lo Piparo, E., Modi, S., Thiel, A., Worth, A., 2014. Estabilishing the level of safety concern for chemicals in food without the need for toxicity testing. Regul. Toxicol. Pharmacol. 68 (2), 275e296. http://dx.doi.org/10.1016/ j.yrtph.2013.08.018.
Schroeder, K., Bremm, K.D., Alepee, N., Bessems, J.G.M., Blaauboer, B., Boehn, S.N., Burek, C., Coecke, S., Gobmau, L., Heweitt, N.J., Heylings, J., Huqyler, J., Jaeger, M., Jagelavicius, M., Jarrett, N., Ketelslegers, H., Kocina, I., Koester, J., Kreysa, J., Note, R., Poth, A., Radtke, M., Rogiers, V., Scheel, J., Schulz, T., Steinkellner, H., Toeroek, M., Whelan, M., Winkler, P., Diembeck, W., 2011. Report from the EPAA workshop: In vitro ADME in safety testing used by EPAA industry sectors. Toxicol. In Vitro 25 (3), 589e604.http://dx.doi.org/10.1016/j.tiv.2010.12.005.
Selvendiran, K., Banu, S.M., Sakthisekaran, D., 2004. Protective effect of piperine on benzo(a)pyrene-induced lung carcinogenesis in Swiss albino mice. Clin. Chim. Acta 350, 73e78.
Selvendiran, K., Banu, S.M., Sakthisekaran, D., 2005a. Oral supplementation of piperine leads to altered phase II enzymes and reduced DNA damage and
DNA-protein cross links in Benzo(a)pyrene induced experimental lung carcinogen-esis. Mol. Cell. Biochem. 268, 141e147.
Selvendiran, K., Singh, J.P.V., Krishnan, K.B., Sakthisekaran, D., 2003. Cytoprotective effect of piperine against benzo(a)pyrene induced lung cancer with reference to lipid peroxidation and antioxidant system in Swiss albino mice. Fitoterapia 74, 109e115.
Selvendiran, K., Thirunavakkarasu, C., Singh, J.P.V., Padmavathi, R., Sakthisekaran, D., 2005b. Chemopreventive effect of piperine on mitochondrial TCA cycle and phase-I and glutathione-metabolizing enzymes in benzo(a)pyrene induced lung carcinogenesis in Swiss albino mice. Mol. Cell. Biochem. 271, 101e106.
Selvendiran, K., Prince Vijeya Singh, J., Sakthisekaran, D., 2006. In vivo effect of piperine on serum and tissue glycoprotein levels in benzo(a)pyrene induced lung carcinogenesis in Swiss albino mice. Pulm. Pharmacol. Ther. 19, 107e111.
Sethi, P., Dua, V.K., Mohanty, S., Mishra, S.K., Jain, R., Edwards, G., 2009. Develop-ment and validation of a reversed phase HPLC method for simultaneous determination of curcumin and piperine in human plasma for application in clinical pharmacological studies. J. Liq. Chromatogr. Relat. Technol. 32, 2961e2974.
Singh, J., Reen, R., Wiebel, F.J., 1994. Piperine, a major ingredient of black and long peppers, protects against AFB1- induced cytotoxicity and micronuclei formation in H4IIEC3 rat hepatoma cells. Cancer Lett. 86, 195e200.
Suresh, D., Srinivasan, K., 2010. Tissue distribution & elimination of capsaicin, piperine, and curcumin following oral intake in rats. Indian J. Med. Res. 131, 682e691.
Ternes, W., Krause, E.L., 2002. Characterization and determination of piperine and piperine isomers in eggs. Anal. Bioanal. Chem. 374, 155e160.
Thiel, A., Buskens, C., Woehrle, T., Etheve, S., Schoenmakers, A., Fehr, M., Beilstein, P., 2014. Black pepper constituent piperine: genotoxicity studies in vitro and in vivo. Food Chem. Toxicol. 66, 350e357.
ToxPredict, Available at:http://apps.ideaconsult.net:8080/ToxPredict/. ToxTree, Available at:http://toxtree.sourceforge.net/.
Vellaichamy, L., Balakrishnan, S., Panjamurthy, K., Manoharan, S., Alias, L.M., 2009. Chemopreventive potential of piperine in 7,12-dimethylbenz[a]anthracene induced skin carcinogenesis in Swiss albino mice. Environ. Toxicol. Pharmacol. 28, 11e18.
Wbra, H., El-Mofty, M.M., Schwaireb, M.H., Dutter, A., 1992. Carcinogenicity testing of some constituents of black pepper (piper nigrum). Exp. Toxic. Pathol. 44, 61e65.
Wiseman, R.W., Miller, E.C., Miller, J.A., Liem, A., 1987. Structure-activity studies on the hepatocarcinogenicities of alenylbenzene derivatives related to estragole and safrole on administration to preweanling male C57BL/6J x C3H/heJ F1 mice. Cancer Res. 47, 2275e2283.