Large Amounts of Picolinic Acid Are Lethal but Small Amounts Increase the Conversion of Tryptophan-Nicotinamide in Rats

Home , Kainic acid, Picolinic acid, Xanthurenic acid

J Nutr Sci Vitaminol, 60, 334–339, 2014

Katsumi Shibata and Tsutomu Fukuwatari

Department of Nutrition, School of Human Cultures, The University of Shiga Prefecture, Shiga 522–8533, Japan (Received May 13, 2014)

Summary Picolinic acid (PiA) is an endogenous metabolite of tryptophan that has been reported to possess a wide range of physiological actions. We investigated the effects of dietary PiA on the metabolism of tryptophan to nicotinamide in growing rats. Feeding an ordinary diet containing 1% PiA to growing rats (6 wk) caused death within a few days. Toxicity of PiA was higher than that of analogs such as nicotinic acid and quinolinic acid. Feeding an ordinary diet containing 0.05% and 0.1% PiA did not elicit decreased intake of food or loss in body weight. PiA did not affect the in vitro liver activities of quinolinic acid phosphoribosyltransferase or a-amino-b-carboxymuconate-e-semialdehyde decarboxylase (ACMSDase, a Zn-dependent enzyme). Concentrations of NAD and NADP in the liver and blood were not affected by PiA. PiA administration did not affect tryptophan metabolites such as anthranilic acid, kynurenic acid, and xanthurenic acid. However, quinolinic acid and subsequent metabolites such as nicotinamide and its catabolites were increased by administration of a diet containing 0.05% PiA but not by a 0.1% PiA diet. These results suggest that the in vivo activity of ACMSDase is controlled by the Zn level. Therefore, a small amount of PiA has a beneficial effect for conversion of tryptophan to nicotinamide, but an excessive amount of PiA can be very toxic. Key Words picolinic acid, tryptophan, nicotinamide, toxicity, rat

Picolinic acid (PiA) is a metabolite of the complete As in the case of QA, excessive amounts of nicotinic degradation pathway of tryptophan (Trp) (see, Fig. 1), acid, Nam, N1-methylnicotinic acid, and N1-methylnic- which has been reported to possess a wide range of otinamide (MNA) affect food intake and body-weight physiological actions (e.g., neuroprotective (1, 2) and gain only slightly (10). Administration of nicotinic acid immunological (3) effects) and zinc (Zn) metabolism (4). and Nam elicit a much greater increase in Nam catabo- PiA formation is controlled by the activities of two lites such as MNA, N1-methyl-2-pyridone-5-carbox- enzymes: a-amino-b-carboxymuconate-e-semialdehyde amide (2-Py) and N1-methyl-4-pyridone-3-carboxamide decarboxylase (ACMSDase) and a-aminomuconate-e- (4-Py). However, administration of MNA and N1-meth- semialdehyde dehydrogenase (AMSDHase). ACMSDase ylnicotinic acid do not produce increases in Nam catab- catalyzes the reaction of a-amino-b-carboxymuconate- olites (10). Whether dietary PiA affects the metabolism e-semialdehyde (ACMS) to a-aminomuconate-e-semial- of Trp to Nam is unknown. We investigated the effects dehyde (AMS) and AMSDHase the reaction of AMS to of dietary PiA on the growth, food intake, and metabo- 2-aminoadipic acid in the presence of NAD(P)H. lism of Trp to Nam in growing rats. PiA is created by non-enzymatic means from AMS. MATERIALS AND METHODS Quinolinic acid (QA) is produced non-enzymatically from ACMS, which is metabolized to nicotinic acid Chemicals. Vitamin-free milk casein, sucrose, and mononucleotide by QA phosphoribosyltransferase l-methionine were obtained from Wako Pure Chemical (QPRTase) in the presence of 5-phosphoribosyl 1-pyro- Industries, Ltd. (Osaka, Japan). Corn oil was purchased phosphate (PRPP). QA is the direct precursor of NAD in from Ajinomoto (Tokyo, Japan). Gelatinized cornstarch, the tryptophan (Trp)-nicotinamide (Nam) biosynthetic a mineral mixture (11), and a vitamin mixture (11) pathway, but dietary QA is not incorporated efficiently were purchased from Oriental Yeast Co., Ltd. (Tokyo, into NAD molecules (5). The low replacement efficiency Japan). PiA, anthranilic acid (AnA), QA and Nam were of QA for NAD is considered to result from inefficient obtained from Wako Pure Chemical Industries. MNA penetration of QA into cells (6–9). chloride, xanthurenic acid (XA), kynurenic acid (KA), The toxicity of QA is reported to be low. For exam- and 3-hydroxyanthranilic acid (3-HA) were purchased ple, feeding a diet containing 0.5% QA to weaning rats from Tokyo Chemical Industries (Tokyo, Japan). Com- affects food intake and body-weight gain slightly com- pounds 2-Py and 4-Py were synthesized according to pared with those of rats fed a non-QA control diet (10). the method of Pullman and Colowick (12) and that of Shibata et al. (13), respectively. All other chemicals were E-mail: [email protected] of the highest purity available from commercial sources.

334 Picolinic Acid and Tryptophan Metabolism 335

Fig. 1. Metabolic pathway of Trp. ACMS, a-amino-b-carboxymuconate-e-semialdehyde; AMS, a-aminomuconate-e-semialdehyde; NMN, nicotinamide mononucleotide.

Table 1. Compositions of diets.

20% Casein diet 10.05% PiA 10.10% PiA 11.0% PiA

g/kg diet Vitamin-free casein 200 200 200 200 l-Methionine 2 2 2 2 Gelatinized-cornstarch 464 463.5 463 454 Sucrose 224 224 224 224 Corn oil 50 50 50 50 Mineral mixture1 50 50 50 50 Vitamin mixture (NiA-free)1 10 10 10 10 PiA 0 0.5 1 10

1 Shibata K, Matsuo H. 1989. Effect of supplementing low protein diets with the limiting amino acids on the excretion of N1-methylnicotinamide and its pyridones in rats. J Nutr 119, 896–901.

Animals and treatment. The care and treatment of and allowed to drink tap water ad libitum for 21 d; this experimental animals conformed to guidelines set by group was the control group (Table 1). The other groups The University of Shiga Prefecture (Shiga, Japan). The (n55) were fed the same diets with PiA addition (Table animal room was maintained at temperature of 22˚C 1) and allowed to drink tap water ad libitum for 21 d. and with 60% humidity. A 12-h light–dark cycle was The concentration of dietary PiA was selected from the also in operation (18:00–06:00). data about the toxicity of quinolinic acid (10) and nico- Six-week-old male Wistar rats obtained from CLEA tinic acid (10, 14). Urine samples (24 h; 09:00–09:00) Japan, Inc. (Tokyo, Japan) were housed individually in were collected at day 21 in amber bottles containing metabolic cages (CT-10; CLEA Japan). Rats were divided 1 mL of 1 mol/L HCl. Acidified urine samples were into four groups at 09:00, which was designated as stored at 225˚C until needed. After the last urine sam- the “zero time” of the experiment elapsed time. One ples had been collected, the rats were killed. Blood from group (n55) was fed a conventional 20% casein diet the carotid artery and the livers were removed, and were 336 Shibata K and Fukuwatari T

Fig. 2. Effects of dietary PiA on body-weight changes (A) and food intake (B) in rats. Values are mean6SE for 5 rats.

Table 2. Effects of PiA on body weight, food intake, liver weight, NAD concentration and NADP concentration, and liver activities of QPRTase and ACMSDase in rats on day 21.

20% Casein diet 10.05% PiA 10.10% PiA

Initial body weight, g 133.861.33 133.760.15 133.460.90 Final body weight, g 273.966.0 274.966.5 268.067.6 Body-weight gain, g/21 d 140.165.9 141.264.2 134.665.7 Total food intake, g/21 d 399610.8 392610.5 397613.6 Liver weight, g 12.2960.34 12.0760.48 11.9360.51 Liver NAD, nmol/g 826642.4 810633.7 824650.1 Liver NADP, nmol/g 43267.8 448637.5 437632.2 Blood NAD, nmol/mL 82.861.7 79.261.4 84.963.0 Blood NADP, nmol/mL 13.560.5 13.860.5 14.760.4 Liver QPRTase, nmol/h/g 1.1860.05 1.1560.04 1.1960.04 Liver ACMSDase, nmol/h/g 0.7460.11 0.9760.16 0.8660.13

Values are means6SE for 5 rats. Significant differences were not observed between values in rows.

used for measurements of NAD and NADP as well as the Statistical analyses. Values are the mean6SE. Sig- enzyme activities of QPRTase and ACMSDase. nificance was determined by one-way ANOVA, followed Enzyme assays. The liver was dissected, and one by Tukey’s multiple-comparison test; p,0.05 was con- portion homogenized immediately with a Teflon-glass sidered significant. Prism version 5.0 (Graph Pad, San homogenizer in 5 volumes of cold 50 mmol/L KH2PO4– Diego, CA) was used for all analyses. K HPO buffer at pH 7.0. The resulting homogenate 2 4 RESULTS was used as an enzyme source. The methods of measuring of QPRTase (nicotinate Effects of a high-PiA diet on body weight and food intake nucleotide: pyrophosphate phosphoribosyltransferase, Body weights of rats fed the 1% PiA diet decreased by EC 2.4.2.19) (15) and ACMSDase (EC 4.1.1.45) (16) 1.9 g on day 1, and by 20 g on day 3. Three of the five have been reported previously. rats died on day 6, one rat on day 8, and one rat on day Measurement of Trp metabolites. Amounts of Trp 9. Rats ate 10 g of the 1% PiA diet on day 1, and 3 g/d metabolites such as AnA (17), KA (18), XA (19), 3-HA during days 2 to 5, but ate little after day 5. (19), QA (20), Nam (13), MNA (21), 2-Py (13), and Effects of low PiA on the metabolism of Trp to Nam 4-Py (13) in urine were measured by high-performance Feeding a diet containing low amounts of PiA (0.05% liquid chromatography. and 0.1%) to rats did not affect food intake or body- Measurement of levels of NAD and NADP. Concentra- weight gain (Fig. 2 and Table 2). Liver weights were also tions of NAD (22) and NADP (23) in the livers and blood not affected by administration of low amounts of PiA were measured by enzyme cycling methods. (Table 2). Concentrations of NAD and NADP in the liv- Picolinic Acid and Tryptophan Metabolism 337

Table 3. Effects of PiA on urinary excretion of Trp metabolites including Nam and its catabolites on day 21.

20% Casein diet 10.05% PiA 10.10% PiA

AnA, nmol/d 6767 6366 6769 KA, nmol/d 1,5486111 1,5726110 1,414692 XA, nmol/d 2,0826122 2,142665 1,9466104 3-HA, nmol/d 180620 180610 194614 QA, nmol/d 793639b 1,412674a 9336160b Nam, nmol/d 5266b 167637a 5663b MNA, nmol/d 121611b 203628a 13568b 2-Py, nmol/d 160615b 253624a 186610b 4-Py, nmol/d 2,4756186b 4,1706634a 2,4426122b SUM1, nmol/d 2,8086209b 4,7936801a 2,8196100b Conversion percentage2, % 1.2460.10b 2.1560.17a 1.2160.04b

1 SUM5Nam1MNA12-Py14-Py. Nam and its catabolites originated only from dietary Trp because the diets did not contain preformed niacin. 2 (SUM, nmol/d)/(Trp intake, nmol/d)3100. Values are the means6SE for 5 rats. Mean values in a row without a common superscripted letter differ by p,0.05 as determined by one-way ANOVA followed by Tukey-Kramer multiple comparison tests.

ers and blood were not affected by PiA administration that decreased synthesis of PiA leads to Zn deficiency (Table 2). Diets were free of preformed niacin, so NAD and then to decreased levels of urocanic acid (a natural and NADP originated only from dietary Trp in protein. sunscreen compound in skin (25) and increased levels The in vitro activities of liver QPRTase and ACMSDase of the hem precursor 5-aminolevulinic acid and pho- were not affected by PiA administration (Table 2). We toreactive porphyrins. It is probable that PiA affects the could not measure the activity of AMSDHase. Available metabolism of divalent minerals because PiA can che- substrate could not be obtained because the substrate late with divalent cations. However, we did not inves- (ACM) is very unstable, and cycles automatically to form tigate the effects of PiA on the metabolism of divalent PiA in vitro. cations. We investigated the scale of toxicity of dietary Twenty-four-hour urine samples were collected on PiA as well as the effects on the metabolism of the Trp day 21. Table 3 shows the effects of low levels of dietary to Nam pathway, which is very important for supplying PiA on urinary excretion of AnA, KA, 3-HA, and QA, as Nam (26). well as the sum of Nam, MNA, 2-Py, and 4-Py. Diets were The toxicity of dietary PiA is very high compared free of preformed niacin, so all of Nam and its metabo- with that of QA (10). Administration of a diet contain- lites originated only from dietary Trp in protein. Urinary ing 1% PiA to rats will kill them within a few days. Nico- levels of AnA, KA, XA, and 3-HA were not affected by tinic acid has been reported to be non-toxic (10). It is administration of a diet low in PiA (Table 3). However, very unlikely that mammals (including humans) would urinary levels of QA and its metabolites (sum of Nam, consume large amounts of PiA, but the present study MNA, 2-Py, and 4-Py) were increased by PiA admin- is important because some people take chromium(III) istration (Table 3). Therefore, the conversion of Trp to picolinate for health reasons. We have no data on the Nam was increased by PiA administration (Table 3). extent of endogenous formation of PiA in rats. Forma- tion of QA is 6% of Trp intake in mice (27). DISCUSSION Administration of a small amount of PiA (0.05%) In the Trp catabolic pathway, two types of pyridine strengthened QA formation, and subsequent formation carboxylic acids are created: QA and PiA. ACMS is of Nam and its metabolites. Thus, PiA increased the formed from 3-HA by catalysis of 3-HADOase, and the conversion percentage of Trp to Nam. The reaction of ACMS that “escapes” from the catalysis of ACMSDase PiA to QA has not been reported previously. Increased cycles spontaneously to become QA. QA is metabolized formation of QA by PiA administration cannot be to nicotinic acid mononucleotide in the presence of explained clearly, but PiA might inhibit ACMSDase PRPP by the catalysis of QPRTase. Dietary QA is not activity because the enzyme is a Zn-dependent amido- used efficiently for a NAD precursor (5) and the toxic- hydrolase (28). Therefore, a moderate concentration of ity is low (10). AMS “escaping” from the catalysis of PiA, for example a diet containing 0.05% PiA, increases AMSDHase cycles spontaneously to form PiA. PiA is a the efficiency of Zn availability, whereas an excessive metabolite of the kynurenine pathway in Trp catabo- concentration of PiA, for example a diet containing over lism that has been reported to possess neuroprotective 0.1% PiA, might decrease the availability of Zn. A lower (1, 2) and immunological (3) effects, and to metabolize activity of ACMSDase leads to accumulation of ACMS Zn (4). Nutritionally, the association between PiA and (which is produced from 3-HA by 3-HADOase). QA is Zn metabolism is very important. Badawy (24) proposed made spontaneously from ACMS. 338 Shibata K and Fukuwatari T

10) Shibata K, Tanaka K. 1986. Effect of supplementation of CONCLUSION excessive nicotinic acid, nicotinamide, quinolinic acid, 1 A high concentration of dietary PiA was lethal, but a trigonelline or N -methylnicotinamide on the metabo- low concentration of dietary PiA increased the conver- lism of niacin in rats. Bull Teikoku-Gakuen 12: 1–9 (in Japanese). sion of Trp to Nam by changing the bioavailability of 11) Reeves PG. 1997. Components of the AIN-93 diets as Zn, which leads to control of Zn-dependent ACMSDase improvements in the AIN-76A diet. J Nutr 127(Suppl): activity in vivo. 838S–841S. 12) Pullman ME, Colowick SP. 1954. Preparation of 2- Author contributions and 6-pyridones of N1-methylNam. J Biol Chem 206: K.S. designed the study. K.S. drafted the manuscript. 121–127. T.F. assisted in these tasks. 13) Shibata K, Kawada T, Iwai K. 1988. Simultaneous micro-determination of Nam and its major metabolites, 1 1 Funding N -methyl-2-pyridone-5-carboxamide and N -methyl- This study was part of the project “Studies on the 4-pyridone-3-carboxamide, by high-performance liquid dietary reference intakes for Japanese people” (princi- chromatography. J Chromatogr 424: 23–28. 14) Fukuwatari T, Kurata K, Shibata K. 2009. Effects of pal investigator, Katsumi Shibata) and “Studies on the excess nicotinic acid on growth and the urinary excre- construction of evidence to revise the dietary reference tion of B-group vitamins and the metabolism of trypto- intake for Japanese people–elucidation of the balance of phan in weaning rats. Food Hyg Safe Sci 50: 80–84 (in micronutrients and major elements” (principal inves- Japanese). tigator, Katsumi Shibata), which were supported by a 15) Shibata K, Fukuwatari T, Sugimoto E. 2000. Reversed- research grant for Comprehensive Related Diseases from phase high-performance liquid chromatography of nic- the Ministry of Health, Labour and Welfare of Japan. otinic acid mononucleotide for measurement of quino- This work is a publication of the Department of Nutri- linate phosphoribosyltransferase. J Chromatogr 749: tion, School of Human Cultures, The University of Shiga 281–285. Prefecture. 16) Ikeda M, Tsuji H, Nakamura S, Ichiyama A, Nishizuka Y, Hayaishi O. 1965. Studies on the biosynthesis of nicotinamide adenine dinucleotide. II. A role of picolinic Author disclosure carboxylase in the biosynthesis of nicotinamide adenine K. Shibata and T. Fukuwatari have no conflicts of dinucleotide from tryptophan in mammals. J Biol Chem interest. 240: 1395–1401. REFERENCES 17) Shibata K, Onodera M. 1991. Measurement of 3-hydroxyanthranilic acid and anthranilic acid in urine 1) Vrooman L, Jhamandas K, Boegman RJ, Beninger RJ. by high-performance liquid chromatography. Agric Biol 1993. PiA modulates kainic acid-evoked glutamate Chem 55: 143–148. release from the striatum in vitro. Brain Res 627: 18) Shibata K. 1988. Fluorimetric micro-determination of 193–198. kynurenic acid, an endogenous blocker of neurotoxic- 2) Blasi E, Mazzolla R, Pitzurra L, Barluzzi R, Bistoni F. ity, by high-performance liquid chromatography. J Chro- 1993. Protective effect of PiA on mice intracerebrally matogr 430: 376–380. infected with lethal doses of Candida albicans. Antimi- 19) Shibata K, Onodera M. 1992. Simultaneous high-perfor- crob Agents Chemother 37: 2422–2426. mance liquid chromatographic measurement of xanth- 3) Varesio L, Clayton M, Blasi E, Ruffman R, Radzioch D. urenic acid and 3-hydroxyanthranilic acid in urine. 1990. PiA, a catabolite of tryptophan, as the second Biosci Biotechnol Biochem 56: 974. signal in the activation of IFN-gamma-primed macro- 20) Mawatari K, Oshida K, Iinuma F, Watanabe M. 1996. phages. J Immunol 145: 4265–4271. Determination of QA by liquid chromatography with 4) Krieger I, Statter M. 1987. Tryptophan deficiency and fluorimetric detection. Adv Exp Med Biol 398: 697–701. PiA: effect on zinc metabolism and clinical manifesta- 21) Shibata K. 1987. Ultramicro-determination of N1-meth- tions of pellagra. Am J Clin Nutr 46: 511–517. ylNam in urine by high-performance liquid chromatog- 5) Shibata K, Tanaka K, Murata K. 1986. Efficiency of QA raphy. Vitamins 61: 599–604 (in Japanese). as niacin in rats. Agric Biol Chem 50: 2025–2032. 22) Shibata K, Murata K. 1986. Blood NAD as an index of 6) Ijichi H, Ichiyama A, Hayaishi O. 1966. Studies on the niacin nutrition. Nutr Int 2: 177–181. biosynthesis of Nam adenine dinucleotide. 3. Compara- 23) Shibata K, Tanaka K. 1986. Simple measurement of tive in vivo studies on nicotinic acid, Nam, and QA as blood NADP and blood levels of NAD and NADP in precursors of Nam adenine dinucleotide. J Biol Chem humans. Agric Biol Chem 50: 2941–2942. 241: 3701–3707. 24) Badawy AA. 2014. Pellagra and alcoholism: A bio- 7) Hagino Y, Lan SJ, Ng CY, Henderson LM. 1968. Metabo- chemical perspective. Alcohol Alcohol 49: 238–250. lism of pyridinium precursors of pyridine nucleotides in 25) Shibata K, Nishioka Y, Kawada T, Fushiki T, Sugimoto E. perfused rat liver. J Biol Chem 243: 4980–4986. 1997. High-performance liquid chromatographic mea- 8) Lin LF, Henderson LM. 1972. Pyridine nucleotide syn- surement of urocanic acid isomers and their ratios in thesis. Purification of Nam mononucleotide pyrophos- naturally light-exposed skin and naturally shielded skin. phorylase from rat erythrocytes. J Biol Chem 247: J Chromatogr 695: 434–438. 8023–8030. 26) Fukuwatari T, Murakami M, Ohta M, Kimura N, Jin-no 9) Henderson LM, Gross CJ. 1979. Transport of niacin Y, Sasaki R, Shibata K. 2004. Changes in the urinary and niacinamide in perfused rat intestine. J Nutr 109: excretion of the metabolites of the tryptophan-niacin 646–653. Picolinic Acid and Tryptophan Metabolism 339

pathway during pregnancy in Japanese women and rats. 28) Martynowski D, Eyobo Y, Li T, Yang K, Liu A, Zhang J Nutr Sci Vitaminol 50: 392–398. H. 2006. Crystal structure of alpha-amino-beta- 27) Terakata M, Fukuwatari T, Sano M, Nakao N, Sasaki R, carboxymuconate-epsilon-semialdehyde decarboxylase: Fukuoka S, Shibata K. 2012. Establishment of true nia- insight into the active site and catalytic mechanism cin deficiency in quinolinic acid phosphoribosyltransfer- of a novel decarboxylation reaction. Biochemistry 45: ase knockout mice. J Nutr 142: 2148–2153. 10412–10421.