The niacin required for optimum growth can be synthesized from L-tryptophan in growing mice lacking tryptophan-2,3-dioxygenase.

In mammals, nicotinamide (Nam) is biosynthesized from l-tryptophan (l-Trp). The enzymes involved in the initial step of the l-Trp→Nam pathway are l-Trp-2,3-dioxygenase (TDO) and indoleamine-2,3-dioxygenase (IDO). We aimed to determine whether tdo-knockout (tdo(-/-)) mice fed a diet without preformed niacin can synthesize enough Nam to sustain optimum growth. Wild-type (WT) and tdo(-/-) mice were fed a chemically defined 20% casein diet with or without preformed niacin (30 mg nicotinic acid/kg) for 28 d. Body weight, food intake, and liver NAD concentrations did not differ among the groups. In the groups of mice fed the niacin-free diet, urinary concentrations of the upstream metabolites kynurenine (320% increase, P < 0.0001), kynurenic acid (270% increase, P < 0.0001), xanthurenic acid (770% increase, P < 0.0001), and 3-hydroxyanthranilic acid (3-HA; 450% increase, P < 0.0001) were higher in the tdo(-/-) mice than in the WT mice, while urinary concentrations of the downstream metabolite quinolinic acid (QA; 50% less, P = 0.0010) and the sum of Nam and its catabolites (10% less, P < 0.0001) were lower in the tdo(-/-) mice than in the WT mice. These findings show that the kynurenine formed in extrahepatic tissues by IDO and subsequent enzymes can be metabolized up to 3-HA, but not into QA. However, the tdo(-/-) mice sustained optimum growth even when fed the niacin-free diet for 1 mo, suggesting they can synthesize the minimum necessary amount of Nam from l-Trp, because the liver can import blood kynurenine formed in extrahepatic tissues and metabolize it into Nam via NAD and the resulting Nam is then distributed back into extrahepatic tissues.

Establishment of true niacin deficiency in quinolinic acid phosphoribosyltransferase knockout mice.

Pyridine nucleotide coenzymes are involved in >500 enzyme reactions and are biosynthesized from the amino acid L-tryptophan (L-Trp) as well as the vitamin niacin. Hence, "true" niacin-deficient animals cannot be "created" using nutritional techniques. We wanted to establish a truly niacin-deficient model animal using a protocol that did not involve manipulating dietary L-Trp. We generated mice that are missing the quinolinic acid (QA) phosphoribosyltransferase (QPRT) gene. QPRT activity was not detected in qprt(-/-)mice. The qprt(+/+), qprt(+/-), or qprt(-/-) mice (8 wk old) were fed a complete diet containing 30 mg nicotinic acid (NiA) and 2.3 g L-Trp/kg diet or an NiA-free diet containing 2.3 g L-Trp/kg diet for 23 d. When qprt(-/-)mice were fed a complete diet, food intake and body weight gain did not differ from those of the qprt(+/+) and qprt(+/-) mice. On the contrary, in the qprt(-/-) mice fed the NiA-free diet, food intake and body weight were reduced to 60% (P < 0.01) and 70% (P < 0.05) of the corresponding values for the qprt(-/-) mice fed the complete diet at d 23, respectively. The nutritional levels of niacin, such as blood and liver NAD concentrations, were also lower in the qprt(-/-) mice than in the qprt(+/+) and the qprt(+/-) mice. Urinary excretion of QA was greater in the qprt(-/-) mice than in the qprt(+/+) and qprt(+/-) mice (P < 0.01). These data suggest that we generated truly niacin-deficient mice.