Deletion of a xenobiotic metabolizing gene in mice affects folate metabolism.

The mouse arylamine N-acetyltransferase 2 (Nat2) and its homologue (NAT1) in humans are known to detoxify xenobiotic arylamines and are also thought to play a role in endogenous metabolism. Human NAT1 is highly over-expressed in estrogen receptor positive breast tumours and is implicated in susceptibility to neural tube defects. In vitro assays have suggested an endogenous role for human NAT1 in folate metabolism, but in vivo evidence to support this hypothesis has been lacking. Mouse Nat2 provides a good model to study human NAT1 as it shows similar expression profiles and substrate specificities. We have generated transgenic mice lacking a functional Nat2 gene and compared the urinary levels of acetylated folate metabolite para-aminobenzoylglutamate in Nat2 knockout and Nat2 wild-type mice. These results support an in vivo role for mouse Nat2/human NAT1 in folate metabolism. In addition, effects of the Nat2 deletion on sex ratios and neural tube development are described.

Arylamine N-acetyltransferases have well-defined roles in detoxification of xenobiotic arylamines and hydrazines. There are two NAT genes in humans, showing distinct substrate specificities [1] and expression profiles [2], with human NAT2 expressed mainly in the liver and intestine, and widely held to have a role in xenobiotic metabolism. Human NAT1 and its homologue in mouse (mouse Nat2) is strongly expressed in the placenta, early in embryonic development [3] in the developing neural tube [4] and in folate-sensitive cells contributing to the neuroendocrine system [5], as well as in the adult, in erythrocytes and a range of epithelia [2,6].

Both human NAT1 and human NAT2 genes are polymorphic and there are suggestions that the polymorphic status may influence susceptibility to cancers in epithelia exposed to xenobiotic toxins [7]. Recent studies linking levels of NAT1 expression with those of the estrogen receptor in breast tumour tissue [8] indicate that the role of NAT1 in carcinogenesis and tumour growth in some tissues relates to endogenous as opposed to xenobiotic metabolism. In vitro studies have shown that human NAT1 is able to acetylate the folate catabolite para-aminobenzoylglutamate (pABAglu) [9,10], however, to date, there has been no confirmation of this activity in vivo. We have used a well-characterized transgenic mouse model to address the relationship in vivo between human NAT1/mouse Nat2 and folate metabolism, and describe phenotypic effects of deleting the mouse Nat2 gene in mice.

Methods

Nat2 transgenic mouse maintenance and breeding

All work involving animals was carried out according to the UK Animals (Scientific Procedures) Act of 1986 under license from the UK Home Office.

The generation of a stable Nat2 knockout line of mice by targeted insertion of a lacZ-containing cassette, has been described [11]. The Nat2 allele was bred from a129/Ola background onto two different genetic backgrounds (A/J and C57Bl/6, supplied by Harlan, UK) by backcrossing over ten generations. Nat2−/−, Nat2+/− and Nat2+/+ animals used for analysis were generated by intercrossing, (mating Nat2+/− male and Nat2+/− females). On weaning, offspring were sexed on the basis of the external morphology of the genitalia. Ear biopsies were taken for genotyping and sexing and DNA isolated using Sigma GenElute mammalian genomic DNA miniprep kit. Offspring sex was confirmed by screening for the Sry gene using Sry and MyoG primers and PCR conditions detailed in [12]. Primers used for Nat2 genotyping were as follows: Neo-T (forward, 5′CATCGCCTTCTATCGCCTTCT3′) and mNat2-910 (reverse, 5′TTCCAAGTACATGGAAGGACACC3′) were used to detect the Nat2 allele, with mNat2-1 (forward) 5′ATGGACATCGAAGCGTACTTTG3′, and mNat2-910 (reverse) used to detect the wild-type Nat2 allele as described [11]. Genotypic and phenotypic data pertaining to each animal were stored and analyzed using Microsoft Access databases.

Adult (8- to 10-week-old) male C57Bl/6 Nat2+/+ and Nat2−/− mice (n = 8 animals per genotype) were used for urine analysis in two independent experiments. All mice were fed on standard mouse chow given ad libitum. Animals given dietary folate supplement were fed ad libitum on mash made up with folic acid (20 μM, pH adjusted to pH 7.0) added to standard mouse chow at 20 mg folic acid/kg dry chow [13] over four to five days. Urine samples were collected during routine cage transfer on days 3–5 and samples from four animals per genotype were pooled for analysis.

To obtain embryos, timed matings were established, with noon on the day of the vaginal plug designated as embryonic day 0.5 (e0.5). Pregnant dams were killed by cervical dislocation. Embryos were dissected from the uterus into ice-cold 10 mM Potassium phosphate pH 7.5, 145 mM NaCl (phosphate buffered saline) containing 4% paraformaldehyde.

Mass spectroscopy

Mouse urine analysis by mass spectroscopy. Urine samples (∼200 μl) were diluted 1:1 with ammonium hydroxide solution (40 μl of concentrated ammonium hydroxide in 1 ml water, pH 10) and centrifuged at 8000g for 10 min. The supernatant (∼400 μl) was applied to an anion-exchange cartridge (Oasis MAX, Waters, UK) previously washed with 1 ml of methanol and conditioned with 1 ml of water prior to sample application. Analytes were eluted from the cartridge with 5% ammonium hydroxide in water, followed by 1 ml of methanol and finally 1 ml of 5% formic acid in methanol.

All fractions were analysed by both negative- and positive-ion electrospray (ES) mass spectrometry (MS) and tandem mass spectrometry (MS/MS) on a Q-TOF Global (Waters, UK) instrument. The instrument was operated in the “V-mode” at a resolution of ∼7500 (FWHM). MS/MS spectra were recorded on selected precursor ions using a collision-energy of 20 eV.

Results and discussion

Metabolic effect of deleting mouse Nat2

The ability of adult Nat2 and Nat2 mice to acetylate the folate catabolite para-aminobenzoylglutamate (pABAglu) was tested by mass spectroscopic analysis of urine samples. Negative-ion ES-MS analysis of purified acetylated pABAglu gave a [M-H]− ion at m/z 307.06 (theoretical m/z 307.09). This compound eluted in the formic acid fraction from the anion-exchange column. The ion of m/z 307.06 was also found in ES-MS spectra of the formic acid anion-exchange fraction from the urine of Nat2 mice given a dietary folate supplement, indicating that acetyl-pABAglu is present in their urine (Fig. 1). However, this ion was absent from the ES-MS spectra of all anion-exchange fractions from Nat2−/− mouse urine. To confirm the identity of components having a mass of 307.06 in Nat2 mouse urine, the ion at m/z 307 found in the formic acid fraction from the anion-exchange column was further analysed by MS/MS, and the resultant spectrum compared to that of the [M-H]− ion of purified acetyl-pABAglu (Fig. 1, Table 1). The MS/MS spectra obtained from the ion of m/z 307 in urine of the Nat2 mice were essentially identical to that of standard acetyl-pABAglu, however the corresponding spectra from Nat2 mice urine did not correspond to the acetyl-pABAglu standard (Table 1). These data indicate that acetyl-pABAglu is present in the urine of Nat2 mice given dietary folate supplements, but undetectable in urine from Nat2 mice.

Fig. 1

ES-MS and -MS/MS analysis of authentic acetyl-pABAglu. (A) Negative ion ES-MS spectrum of authentic acetyl-pABAglu, giving a [M-H]− ion of m/z 307.06. Inset; chemical structure of the [M-H]− ion of acetyl-pABAglu showing exact mass. (B) MS/MS of the ion of [M-H]− ion (m/z 307) from authentic acetyl-pABAglu. (C) Negative ion ES-MS spectrum of urine from Nat2+/+ mice given dietary folate supplement. (D) MS/MS of the ion of [M-H]− ion (m/z 307) from urine of Nat2+/+ mice given dietary folate supplement. See Table 1 for interpretation of fragment ions.

Table 1

MS/MS analysis of the ion of m/z 307 in the formic acid fraction from the anion-exchange column from urine of Nat2+/+ and Nat2−/− mice

MS/MS Fragment ion m/z	Proposed fragment ion elemental formula	Authentic acetyl-pABAglu	Urinary 307.06 peaka		Acetyl-pABAGlu peak intensitiesb	Nat2+/+ peak intensitiesb
			Nat2+/+	Nat2−/−
134.06	C8H8NO	+	+	−	1.0	1.0
178.04	C9H8NO3	+	+	−	0.66	0.65
219.09	C12H15N2O2	+	+	−	0.29	0.25
245.07	C13H13N2O3	+	+	−	0.18	0.18
263.07	C13H15N2O4	+	+	−	0.30	0.25

Urine samples collected from adult male mice given a dietary supplement of folic acid were fractionated by anion-exchange chromatography and ions of m/z 307 further analysed by MS/MS.

For both authentic acetyl-pABAglu and Nat2+/+ derived samples, peak intensities are expressed relative to the peak of mass 134.06. +, peak present; −, peak absent. Experimental m/z values agree (<±0.03 Da) with theoretical m/z values given in the Table.

Early work on human NAT1 enzyme demonstrated that the acetylation of NAT1-specific substrate para-aminobenzoic acid by NAT1 in human erythrocyte cytosols is inhibited in vitro by folic acid, and folate levels in these cells appears to show an inverse correlation with acetylation activity [14]. Subsequently, recombinant human NAT1 enzyme was shown to acetylate the folate catabolite pABAglu [9,10]. The role of folate in cancer biology is complex, with folate supplements used to prevent tumour formation in normal tissues, although these supplements may accelerate the growth of established tumours, and antifolate agents have some efficacy as cancer treatments [15,16]. This duality shown by folate mirrors the effects of human NAT1 on cell growth. High level of expression of human NAT1 has been shown to enhance cell growth under certain conditions [17], but is associated with improved clinical outcome in ER positive breast cancer [8].

Nat2 deletion affects offspring sex ratios in A/J strain

Nat2−/− mice are described as overtly aphenotypic [11]. However, on an A/J genetic background, within homozygous null litters there is a gender bias with a 1.5 fold excess of males (89 males; 59 females) that differs significantly from the predicted 1:1 ratio (p = 0.01) suggesting that homozygosity of the Nat2 allele may be deleterious to females, or beneficial in males. A reversed gender bias is seen in offspring from intercross matings (Nat2 M × Nat2 F), in which both parents are heterozygous (Table 2), with a ratio of males to females of 0.7:1 (184 males to 254 females p = 0.001). In these intercross litters, the female excess is made up of Nat2+/+ and Nat2+/− animals (Table 2), again suggesting that in females, the wild-type allele has selective advantage. Together these results indicate that mutations at the Nat2 locus have a gender-dependent effect on offspring survival. By screening for the Sry gene [12], we have confirmed that the skewed sex ratios do not result from sex reversal. (Supplementary Figure S1).

Table 2

Analysis of sex and genotype in A/J intercross offspring

Nat2 genotype	Offspring numbers
	Males	(Predicted)	Females
+/+	50	(55)	77
+/−	77	(110)	122
−/−	62	(55)	55
χ2	p < 0.01a		p < 0.01a

Inheritance of the Nat2null allele was determined in offspring of intercross mating using PCR. Numbers of offspring predicted assuming independent segregation and Mendelian inheritance of Nat2 (located on chromosome 8) and Sry gene (located on Y chromosome) are given in parentheses.

Significant deviation from predicted values.

Effect of Nat2 deletion on neural tube closure

We have tested whether Nat2 genotype influences susceptibility to neural tube defects, using the C57Bl/6 strain, which is not prone to neural tube defects, and scoring embryos for neural tube defects at e10.5 and e11.5. Amongst Nat2−/− offspring derived by incrossing, out of 64 embryos examined, only one embryo was observed to have a neural tube defect. To investigate the effects of an imbalance in parental Nat2 phenotype, crosses were set up using Nat2+/+ and Nat2−/− parents, generating Nat2+/− offspring. The incidence of neural tube defects amongst Nat2+/− offspring derived in this way was 14% (4/28). In three out of the four cases described in here, associated with the Nat2 null allele, the position of the defect is caudal to the hindbrain, at the cervical level (Fig. 2).

Fig. 2

Wholemount views of embryos showing neural tube defects. Embryos derived by crossing C57Bl/6 Nat2 and Nat2 mice, and harvested at e10.5–e11.5. (A) Normal embryo (e11.5) viewed from right hand side. (B) Embryo (e11.5) with neural tube defect, showing incomplete closure of the anterior neuropore, viewed from right hand side in main panel; inset, dorsal view. (C) Embryo (e10.5) with neural tube defect at cervical level, viewed from right in main panel; inset, dorsal view. Arrowheads point to regions of incomplete closure of the neural tube. The orientation of the embryos is indicated in dorsal views; C, caudal; R, rostral.

An allelic association between the orthologous human NAT1 and orofacial clefting has been recorded [18] and a link between NAT1 imbalance and developmental defects has been described for the human NAT1 alleles with respect to the incidence of spina bifida [19]. In this human study, natural variants with little acetylating activity were found to have a protective effect amongst offspring, and both maternal and offspring genotypes affected the outcome. Gender bias towards females amongst exencephalic embryos both in mice and in men is well documented, for review [20]. The A/J strain is known to be susceptible to neural tube defects, and it is possible that the male bias seen amongst A/J Nat2 offspring is due to developmental abnormalities and lower survival rates amongst female embryos, although the numbers presented here are too small at present for statistical analysis.

Folate supplements have been used widely to prevent neural tube defects and to reduce the risk of colon cancer, reviewed in [15]. The data presented here indicate that deleting mouse Nat2 reduces the acetylation of the folate catabolite pABAglu. Our results suggest that the complex relationship between human NAT1 polymorphism and epithelial cancers [7] including breast cancer [8,17], and the influence of NAT1 genotype on neural tube defects [19] may result from a functional link in vivo between NAT1 acetylation activity and folate metabolism.