Toxicokinetic and mass balance of morpholine in rats

Abstract Morpholine (MOR) has a broad spectrum of use and represents high risk of human exposure. Ingested MOR can undergo endogenous N-nitrosation in the presence of nitrosating agents forming N-nitrosomorpholine (NMOR), classified as possible human carcinogen by the International Agency for Research on Cancer. In this study, we evaluated the MOR toxicokinetics in six groups of male Sprague-Dawley rats orally exposed to 14C-radiolabelled MOR and NaNO2. The major urinary metabolite of MOR, N-nitrosohydroxyethylglycine (NHEG), was measured through HPLC as an index of endogenous N-nitrosation. Mass balance and toxicokinetic profile of MOR were determined by measuring radioactivity in blood/plasma and excreta. MOR reached maximum blood concentration 30 minutes after administration. Elimination rate was rapid (70% in 8h). Most of the radioactivity was excreted in the urine (80.9 ± 0.5%) and unchanged 14C-MOR was the main compound excreted in the urine (84% of the dose recovered). 5.8% of MOR is not absorbed and/or was not recovered. Endogenous nitrosation of MOR was demonstrated by the detection of NHEG. The maximum conversion rate found was 13.3 ± 1.2% and seems to be impacted by the MOR/NaNO2 ratio. These results help refining our knowledge of the endogenous production of NMOR, a possible human carcinogen.


Introduction
N-nitrosamines (or N-nitrosoamines: NAs) constitute a subclass of the superfamily of N-nitroso compounds, characterised by the general structure R1R2-N-N ¼ O, where R1 and R2 may be substituted or unsubstituted aryl, alkyl or heterocyclic groups.Their formation requires a nitrosating agent (e.g.nitrates, nitrites) and a nitrosatable substrate (e.g.primary amine, secondary amine, quaternary ammonium salt or amino acid) (Mirvish 1975;Rostkowska et al. 1988;Park et al. 2015).NAs are produced as unintended by-products in various industrial and natural processes such as for example water disinfection, food processing and/or cooking, and drugs metabolism (Mitch and Sedlak 2004;Herrmann 2014;Zeng and Mitch 2015;Park et al. 2015Park et al. , 2018)).This is a topic of concern because of the high carcinogenic potency of some of them.Accumulated data indicate that many NAs induce tumours in animal species, including higher primates (Bogovski and Bogovski 1981).NAs are pro-carcinogens which require metabolic activation by cytochrome P-450 enzymes to exert carcinogenic and mutagenic effects (Bartsch et al. 1989).
N-nitrosodimethylamine and N-nitrosodiethylamine were classified as 2 A (probably carcinogenic to humans) in 1987, and six others (N-Nitrosomethylethylamine, N-nitrosodibutylamine, N-nitrosopiperidine, N-nitrosopyrrolidine, N-nitrosomorpholine and N-nitrososarcosine) as 2B (possibly carcinogenic to humans) by the International Cancer Research Centre (IARC).To date, there is no known industrial use of NAs.Human exposure to NAs and/or their precursors occurs through exogenous and endogenous sources (Jakszyn and Gonz alez 2006;Loh et al. 2011).Exogenous sources refer to preformed N-nitrosamines presents in diverse consumer products such as food, beverages, tobacco smoke, cosmetics and drugs as well as environmental sources (raw water) (ANSES, 2012;Clapp et al. 2012;Qiu et al. 2017;Gushgari and Halden 2018;Sluggett et al. 2018;Lim et al. 2018).The second group covers NAs formed in vivo from ingested precursors (both nitrosating agents and amines) (Tannenbaum, 1978;Winter et al. 2007).Endogenous nitrosation may account for up to 97% of the total human exposure to N-nitroso compounds (Tricker 1997;Jakszyn and Gonz ale, 2006).
Morpholine (MOR: 1,4-tetrahydro-oxazine; CAS N 110-91-8), as shown in Figure 1, is a heterocyclic secondary amine with great industrial applications.It is used as an anticorrosive agent in water boiling systems, as a key intermediate in the synthesis of a wide range of chemicals (e.g.pharmaceuticals), as a surfactant and emulsifier in cosmetic products and as a food additive (Mcguire and Hagenmaier, 1996;Poupin et al. 1998;de Vocht et al. 2007;Flick et al. 2016).Due to its solubility in water and its broad spectrum of use, MOR can reach the natural aquatic environment via the discharge of wastewater and industrial effluents.The concentrations reported in some inland surface waters and runoffs around the world vary from 0.15 to 9.25 mg/L (Pietsch et al. 2001;Aky€ uz and Ata 2006;Wang et al. 2011, Ma et al. 2012;Zhao et al. 2022).Human exposure to morpholine is well documented and known, but only a little information is available concerning the nitrosation of morpholine.The formation of N-nitrosomorpholine (NMOR) has been observed in vitro in human saliva (Tannenbaum et al. 1978) and in an epithelium equivalent of the gastrointestinal tract (Iijima et al. 2003).This highlights the possibility of in vivo formation of NMOR via consummation of MOR-contaminated water.
Human exposure to MOR, therefore, requires an assessment of the risks associated with the potential endogenous formation of NMOR.This risk assessment is commonly based on the conversion rate of MOR to NMOR, as determined by Hecht and Morrison (1984a).In this study, male F-344 rats were exposed by gavage to MOR and NaO 2 .The concentrations of NMOR formed were estimated by measuring N-nitrosohydroxyethylglycine (NHEG), the major urinary metabolite of NMOR (see Figure 1).The conversion rates obtained ranged between 0.6% ± 0.5 and 12% ± 3. The 12% level corresponds to doses of 19.2 mmol (equivalent 5.6 mg/kg bw) and 95.7 mmol (equivalent 22 mg/kg bw) of MOR and NaNO 2 , respectively.This study, while being informative, has some limitations.The only measurement is performed 24 h post-exposure, which does not allow the observation of potential early excretion.Moreover, the NHEG measurement was performed using GLC (gas-liquid chromatography) coupled with TEA (thermal energy analyser).This method is valid but was technically refined several times since 1984.In another study, the mutagenic potential of MOR was assessed in the intrahepatic host-mediated bacterial mutagenicity assay using young adult female CD-1 mice (Edwards et al. 1979).The obtained conversion rate was between 9 and 15%.This study presents a major weakness in evaluating MOR-NMOR conversion rate: this is not a toxicokinetic study and the NMOR is detected through its mutagenicity.Mice were orally exposed to MOR, and bacteria were injected into their tail vein.
A mutagenicity study was performed on these bacteria with an AMES -like genotoxicity study without following good laboratory practices or OECD guidelines for this type of study (OECD 471).
In conclusion, the existing literature characterising MOR-MOR conversion rate is quite dated and presents limitations.Significant quantities of MOR reach the aquatic environment, where nitrates can often be found.MOR-reaching organisms which are also exposed to nitrate can be transformed into NMOR, which is known as a possible carcinogen.Hence, refining our knowledge of the conversion rate from MOR to NMOR contributes to the assessment of human health risk since the MOR -NMOR conversion rate is an important parameter for MOR risk assessment.The experiments reported here were designed to examine the conversion rate of MOR to NMOR under relevant exposure conditions.Male healthy rats were exposed by gavage to a single dose of MOR, concomitantly with NaNO 2 .At the same time, the mass balance and toxicokinetic of MOR were evaluated by administering carbon-14 labelled morpholine (14 C-MOR).NHEG was measured in urine samples at each condition to confirm the ability of MOR to undergo N-nitrosation, as well as to evaluate its conversion rate.

Animals and husbandry
Sprague-Dawley rats [Crl CD (SD) IGS BR, Caesarian Obtained Barrier Sustained-Virus Antibody free (COBS-VAF)] were selected historical data was available.On arrival, the animals were given a clinical examination to ensure that they were in good condition.During the acclimation period, the required number of animals (36 males) were selected according to body weight and clinical condition.Each animal was identified by an individual ear tattoo.The animals were housed in environmentally controlled animal rooms at 18-26 C, 30-70% relative humidity and kept in a 12-hour light/dark periodic cycle.Rodent food and tap water filtered with a 0.22-micron filter were provided ad libitum.The rats were fasted overnight before treatment (food only) for a period of at least 14 h; the food was given about 4 h after dosing (after blood sampling when applicable).At the beginning of the treatment period, the animals were approximately 6 weeks old and had a mean body weight of 275 g (262 g to 299 g).The study was conducted in compliance with the CRL standard operating procedures and Good Laboratory Practice guidelines.The study was entirely performed under the supervision of the CRL Ethics Committee.

Study design and experiment procedure
A total of 36 male rats were divided into six groups: three groups of nine animals for toxicokinetic (TK) investigations and three groups of three animals for mass balance (MB) investigations.Allocation was made using a computerised stratification procedure to obtain a similar mean body weight per group.The overall study design is illustrated in Figure 2. Information on the concentrations tested, the collection of samples as well as the various physicochemical analyses carried out are detailed in the following sections.

Dose and justification
The oral route was used since it is the intended mode of exposure in humans.Isotopic mixture (MOR and 14 C-MOR) and sodium nitrite were administered by oral route under a constant dose volume of 5 mL/kg and dose-radioactivity of 2.2 MBq/kg.As mentioned above, the MOR gavage solution was prepared by dilution 14 C-MOR with the corresponding cold compound (MOR) so that the animals received a commonly used radioactivity level of approximately 2.2 MBq/kg by gavage.The doses of nitrites (96 mmol and 4 mmol), as well as the highest dose of MOR (19.2 mmol), were selected considering literature using the same exposure protocol (Hecht and Morrison, 1984a).The lowest nitrite value is also in the same range as WHO's drinking water guidelines (3 mg/L) (WHO, 2017).The lowest dose level of 0.14 mg/kg of morpholine was based on the minimum feasible dose, considering the level of radioactivity (2.2 MBq/kg) usually used in this type of study and the route of administration, and the specific activity of the radiolabelled test compound.Overall, these concentrations were not expected to induce toxicity.These doses are described in Table 1.The animals were first exposed to MOR, and immediately afterwards to NaNO2, both administered by oral gavage.Oral administration was performed using a plastic syringe fitted with a gavage tube.There was a good agreement between the theoretical and actual dose levels achieved as all deviations were within [-5.8%; 1.2%] of the target values.

Sample collection
As shown in Figure 2, blood samples were taken from the animals used for the toxicokinetic test at the following time points: before administration and then 0.25, 0.5, 2, 4, 8, 24, 48 and 96 h post-administration.For mass-balance, urine, faeces and cage wash were collected from dedicated animals at the following collection times: 24 h before administration and then over the periods 0-8, 8-24, 24 48, 48, 72-96, 96-120, 120-144 and 144-168 h post-administration.
2.3.3.Sample analysis 2.3.3.1 Radioactivity measurement by liquid scintillation counting.The radioactivity of biological samples as well as cage wash waters was measured by liquid scintillation counting (LSC) using a LS 6000 TA system (Beckman Instruments, Fullerton, USA).That of plasma, urine and cage wash (all duplicate samples) was determined directly while that for blood and faeces samples was measured after combustion in a Packard Sample Oxidiser (Model 307 oxidiser, Packard Instrument Company, Meriden, CT).The dpm (disintegrations per minute) values of each sample type were obtained from the corresponding cpm (counts per minute) values using a quench standard curve.An external standard source (Cesium 137) was used to monitor sample quenching.The mean dpm value from the replicate analysis was used to calculate the corresponding radioactivity concentration (as Bq/g or ng-eq/g), according to the calculation method detailed in Supporting Information.Values were corrected by subtracting radioactivity from samples taken prior to MOR/14C-MOR administration (blank).

Analysis of MOR-NMOR metabolite by UV-Radio-HPLC method.
For NHEG analysis in urine samples, 0.5 mL ali- quots coming from each rat were diluted 2-fold with mobile phase (eluent A), vigorously mixed (10 s) and transferred to appropriate vials for HPLC analysis.Analyses were performed with a Perkin Elmer Series 200 liquid chromatograph equipped with an auto-sampler.The injection volume of the sample was 20 mL.Chromatographic separations were achieved on a Spherisorb ODS2 C18 column (5 mm particle size, 250 Â 4.6 mm, Waters) at a flow rate of 1 mL/min using a gradient elution program.The mobile phase consisted of eluent A (water) and eluent B (water/acetonitrile, 25:75 (v/v), both containing trifluoroacetic acid (0.01 M) and heptane sulphonic acid (0.05 M).The gradient program starts at 100% A, gradually decreases to 0% in 20 min, and remains in an isocratic mode for 5 min before returning to 100% in 1 min.The system was re-equilibrated in 4 min before the next injection.The HPLC flow was split between a Radiomatic 150TR flow scintillation analyser (Perkin Elmer) and a Waters 2487 dual absorbance detector, operating at 245 nm.Data processing was performed using Empower 2 software from Waters.The number of mmol of NMOR formed is estimated by dividing the quantity of NHEG in mmol by 0.52 as calculated by Hecht and Morrison (1984a).

Toxicokinetic in blood
Profiles of total radioactivity in blood are shown in Figure 3.Following oral gavage at the dose-level of 5.6 (groups 1 and 3) or 0.14 (group 5) mg/kg to Sprague-Dawley rats, the maximum mean blood total radioactivity levels (Cmax) were 1862, 1606 and 40 ng-eq/g, respectively.Peak concentrations were reached within 30 min.post-gavage for all groups, indicating rapid absorption of MOR.Thereafter, levels decreased until the last quantifiable time-point at 48 h post-gavage for group 1 (71.3 ng-eq/g) or 8 h postgavage for groups 3 (466 ng-eq/g) and 5 (15.3 ng-eq/g).The blood radioactivity levels were characterised by a low inter-animal variability.The mean exposures (as AUC0-t) were 15,101, 8178 and 229 ng-eq.h/gfor groups 1, 3 and 5, respectively.Table S1 in Supporting Information summarises the main calculated toxicokinetic parameters for blood and plasma radioactivity.It shows that the plasma radioactivity profiles are very close to those observed in the blood.Figure S1 in the Supporting Information represents a typical total radioactivity concentration-time curve in blood and plasma.The whole blood-to-plasma ratio of total radioactivity approaches unity, for Cmax as well as AUC0-t, indicating that there was no presence of MOR in blood cells.

Excretion balance
For all three administered doses, the concentrations of radioactivity in biological samples and cage rinses, over the 168 h period are presented in Table 2.The mean total radioactivity recovered from excreta (urine and faeces) and cage washes was 94.2% (range, 93.5-94.8%)within 168 h after dose administration, indicating almost complete recovery of administered radioactivity and good absorption of MOR.No measurable radioactivity was detected after 168 h.The mean cumulative excretion in urine was 80.2 ± 4.8, 81.4 ± 1.0 and 81.0 ± 2.9% for groups 2, 4 and 6, respectively, whereas only a slight amount of radioactivity was recovered in the faeces (2.8 ± 0.4, 3.0 ± 0.4 and 3.2 ± 0.5% of the administered radioactive dose, respectively).A moderate amount of radioactivity was recovered in the cage washes (10.0 ± 1.8, 10.5 ± 0.7 and 10.7 ± 2.5%, respectively).Analysis of all the results shows that 80.9 ± 0.5% of the administered radioactive dose was excreted in the urine and 3.1 ± 0.1% in the faeces, indicating that renal excretion is the major route of elimination of MOR.The cumulative amount of morpholine-associated radioactivity excreted is shown in Figure 4. Of radioactivity recovered from urine and faeces, most were recovered within 8 h (mean, 71.1%) after dose administration.No difference was observed between the rate of elimination in faeces and urine.

Metabolic pattern in urine
In this work, a method based on the online coupling of highperformance liquid chromatography (HPLC) with flow scintillation analysis (FSA) and UV absorbance detection was developed for the separation and detection of analytes of interest in the different study matrices.The optimisation of the chromatographic conditions as well as the verification of the purity of the standards were carried out using UV detection, while radioactivity detection was used to measure the residual MOR and its metabolites in urine samples.A chromatographic profile of blank urine samples is provided in Supplementary Information Figure S5. Figure S2 in Supporting Information show typical chromatograms of the urine samples supplemented with 14 C-MOR and NHEG obtained under the separation conditions described in Section 2.3.3.It shows that MOR and NHEG are well separated, with average retention times of 2.95 min and 9.14 min, respectively.To accurately determine the expected retention time of radiolabelled NHEG, the lag time between UV detection and radioactivity detection was measured by injection of 14 C-Palmitic acid as tracer, since 14 C-NHEG standard was not commercially available.The lag time determined was 0.25 min.
The radio-chromatograms obtained vary according to the group of rats and the collection period.Figure 5 shows a representative chromatogram of a urine sample from group 2 (treated at a dose of 5.6 mg/kg), collected within the interval of 0-8 h.This chromatogram reveals the presence of five metabolites: a first group (three peaks, between 3.2 and 4.4 min) and a second group (two peaks, at 6.7 and 7.2 min) before the 14 C-MOR peak (9.1 min).Peaks observed at approximately 3.2 min matched the expected retention time of 14 C-NHEG.In urine samples collected over the 8-24h period, only 14 C-MOR was detected with traces of the second group of metabolites.The chromatograms for other groups can be found in Supporting Information S4.For groups 4 and 6, only the second group of metabolites (6.7 and 7.2 min) as well as residual 14 C-MOR were detected in the urine collected over the period 0-8 h.In contrast, in the urines collected over the period 8-24 h, one peak was detected in all animals of group 4 (at 3.42-3.43min), and one to two peaks were detected in 2/3 animals of group 6 (between 2.93-3.50min).These peaks were also considered to represent the entire NHEG and derivative ratio.No peaks were detected in urine samples collected over the 24-48 h and 48-72 h periods for all groups.Unchanged 14 C-MOR was the main compound excreted in urine with more than 84% of the dose recovered (Table S2 in Supporting Information).

Conversion rate of MOR to NMOR
To assess the amount of metabolite, peak areas were measured for all 14 C-compound detected and the results were expressed as a percentage of the total area (see Table S2 in Supporting Information).Hecht and Morrison (1984a), who studied the urinary metabolism of NMOR to NHEG in F344 rats, found that more than 95% of the NHEG formed was excreted within 24 h of NMOR administration.They established that the mean conversion of NMOR to urinary NHEG is 52%.Since urinary excretion of NHEG is rapid and in significant concentrations, it's quantitative monitoring in urine has been suggested as a procedure to estimate the amount of endogenously formed NMOR.Thus, the amount of NMOR formed in our study following rats' exposure to MOR was calculated by dividing NHEG levels by 0.52.For group 2, the percentage conversion of MOR-to-NMOR was estimated at 10.5 ± 1.6%, while those of groups 4 and 6 were estimated in the ranges of 7.1-12.0%and 12.1-14.6%,respectively.

Discussion and conclusion
To perform a well-founded and updated assessment of human exposure to endogenously produced NMOR, updated in vivo data on the kinetics of metabolism of MOR and on its conversion rate to NMOR were required.For that purpose, we designed a study to investigate toxicokinetic, excretion and mass balance of 14 C-MOR in the presence of NaNO 2 in male Sprague-Dawley rats, both orally administered.
After a single oral administration, the radioactivity rapidly increased in the blood, with a peak measured at 30 min, and then gradually decreases to become undetectable between 8h and 24h post-exposure.This indicates that the metabolism of 14 C-MOR as well as its eventual 14 C-metabolits is rapid.The whole blood-to-plasma ratio of radioactivity approaches unity, suggesting a low uptake into blood cells.
Mass balance showed that the recovery of the cumulative radioactivity was very high, with a mean recovery of 94.2%, indicating almost complete recovery of administered doses.The most radioactivity was recovered within 8 h (mean of 71.1%) after dose administration.The mean recovery in urine was 80.9 ± 0.5%, demonstrating that renal excretion is the main route of elimination of MOR.
Urine radio chromatograms showed the presence of a total of six peaks, observed in 0-8h and 8-24h collection periods, including those of 14 C-MOR and its main known urinary metabolite: NHEG.NHEG is also a urinary metabolite of other reactions, such as the formation of N,N-dinitrosopiperazine after administrating rats piperazine and sodium nitrate (Hecht et al. 1984b).However, in our study, rats were only administered morpholine and nitrites and were not exposed to piperazine.In our experimental protocol, we can then consider that the NHEG measured is a specific metabolite of the MOR-NMOR conversion.It can also be measured with a LOQ of 1 ng/analysis, our used HPLC-FSA method appears more sensitive than that reported by Hecht and Morrison (1984a) based on gas chromatography utilising chemical derivatization.Unchanged 14 C-MOR was the main compound excreted in the urine, accounting for more than 84% of the total peak area in HPLC-FSA chromatograms.It is important to emphasise that all detectable chromatographic peaks correspond to urinary metabolites of 14 C-MOR since only radioactive compounds can be revealed by HPLC-FSA.In another study, a metabolic mechanism was proposed by Hecht and Young (1981), highlighting the formation of several NMOR metabolites, including NHEG (See Figure S3 in Supporting Information).In another study, Sohn et al. (1982) identified other metabolites and intermediates in rat, hamster, and guinea pig urines.
Since urinary excretion of NHEG is rapid and in significant concentrations, it's quantitative monitoring in urine has been suggested as a procedure to estimate the amount of endogenously formed NMOR.Thus, the amount of NMOR was calculated by dividing the NHEG concentrations by 0.52, based on the results of Hecht and Morrison (1984a).The percentage conversion of MOR to NMOR was estimated between 9.8 and 13.3%, with a mean of 11.2%, whatever NaNO2/MOR ratios and rat groups.These values are very close to those previously reported by Hecht and Morrison (1984a).The results obtained seem to show that the rate of conversion increases as NaNO2 initial concentration increases.However, no clear relationship between the NaNO2/MOR ratio and this conversion rate could be established.
For regulatory studies, the MOR carcinogenic risk assessment can include its conversion rate into NMOR in the body, from which an equivalent exposure to NMOR can be inferred.To date, the intracorporeal conversion rate of MOR to NMOR is estimated at 12% based on the toxicokinetic study by Hecht and Morrison (1984a) and at 12% using the Host Mediated test by Edwards et al. (1979).In this work, a method based on the online coupling of HPLC with FSA was developed.This method is more sensitive than that reported by Hecht and Morrison (1984a) which is based on gas chromatography equipped (GC) with the specific thermal energy analyser detector (TEA).Moreover, HPLC-FSA method does not require time-consuming and tedious extraction and chemical derivatization steps, unlike the GC-TEA approach.Table 3 below summarises the data available on the conversion rate for comparison.
As shown in Table 3, the results obtained in this study are in good agreement with those of Hecht and Edwards.In the two toxicokinetic studies, the same dose levels of MOR and NaNO 2 substantially give a similar conversion rate: 10% in the current study vs. 12% in Hecht and Morrison's study.However, similar doses (0.48 mmol MOR and 4 mmol of NaNO 2 in the present study vs. 0.92 mmol MOR and 4.8 mmol of Nitrites for Hecht and Morrison's study) with a different NaNO 2 /MOR ratio (8.3 in the present study versus 5.2 at Hecht and Morrison) gave a very different conversion rate: 13.3% in present study vs. 0.6% at in Hecht and Morrison's study.
The findings of this study allow a clearer overview of the toxicokinetic of morpholine in rats and enable to refinement of the MOR-NMOR conversion rate.Morpholine is not classifiable as to its carcinogenicity to humans whereas N-Nitrosomorpholine is possibly carcinogenic to humans (2B).Hence, having a better knowledge of the expected endogenous N-nitrosation is critical for morpholine's risk assessment.Although we performed our study on rats, these conclusions can be used in the context of human risk assessment.Indeed, the NMOR metabolism mostly involves aand b-hydroxylation pathways (Li and Hecht, 2022) which can be found in both rats (Hecht and Young, 1981) and humans (Hecht et al. 1979).
The data of this study contributes to improving the assessment of the risk associated with the consumption of water and/or food containing morpholine and nitrites.However, this study has some limitations, such as the limited number of MOR to NaNO 2 conditions tested and the lack of structural elucidation of the detected metabolites.Moreover, we did not directly detect NMOR; we estimated its concentration through NHEG.This method was previously published (Hecht and Morrison, 1984a) with experimental conditions very close to ours (the most important difference being the rat strain: we worked on Sprague Dawley while the conversion rate was originally measured on F344).Measuring the conversion rate between NMOR and urinary NHEG under our specific experimental conditions may strengthen this estimation.

Figure 4 .
Figure 4. Arithmetic mean cumulative excretion of radioactivity (as % of dose) vs. time following a single oral of 14 C-MOR to rats: group 2 (A), group 4 (B) and group 6 (C).

Figure 5 .
Figure 5. Radio-chromatogram profile of a urine sample from group 2 collected over the 0-8 h period.