Effect of arylformamidase (kynurenine formamidase) gene inactivation in mice on enzymatic activity, kynurenine pathway metabolites and phenotype

Biochimica et Biophysica Acta 1724 (2005) 163 – 172 http://www.elsevier.com/locate/bba

Effect of arylformamidase (kynurenine formamidase) gene inactivation in mice on enzymatic activity, kynurenine pathway metabolites and phenotype Vasily N. Dobrovolskya,*, John F. Bowyerb, Michael K. Pabarcusc, Robert H. Heflicha, Lee D. Williamsd, Daniel R. Doerged, Bjo¨rn Arvidssone, Jonas Bergquiste, John E. Casidac a

Division of Genetic and Reproductive Toxicology, National Center for Toxicological Research, 3900 NCTR Rd., HFT-120, Jefferson, AR 72079, USA b Division of Neurotoxicology, National Center for Toxicological Research, Jefferson, AR, USA c Department of Environmental Science, Policy and Management, University of California, Berkeley, CA, USA d Division of Biochemical Toxicology, National Center for Toxicological Research, Jefferson, AR, USA e Department of Analytical Chemistry, Institute of Chemistry, Uppsala University, Uppsala, Sweden Received 21 December 2004; received in revised form 15 March 2005; accepted 16 March 2005 Available online 7 April 2005

Abstract The gene coding for arylformamidase (Afmid, also known as kynurenine formamidase) was inactivated in mice through the removal of a shared bidirectional promoter region regulating expression of the Afmid and thymidine kinase (Tk) genes. Afmid/Tk -deficient mice are known to develop sclerosis of glomeruli and to have an abnormal immune system. Afmid-catalyzed hydrolysis of N-formyl-kynurenine is a key step in tryptophan metabolism and biosynthesis of kynurenine-derived products including kynurenic acid, quinolinic acid, nicotinamide, NAD, and NADP. A disruption of these pathways is implicated in neurotoxicity and immunotoxicity. In wild-type (WT) mice, Afmidspecific activity (as measured by formyl-kynurenine hydrolysis) was 2-fold higher in the liver than in the kidney. Formyl-kynurenine hydrolysis was reduced by ¨50% in mice heterozygous (HZ) for Afmid/Tk and almost completely eliminated in Afmid/Tk knockout (KO) mice. However, there was 13% residual formyl-kynurenine hydrolysis in the kidney of KO mice, suggesting the existence of a formamidase other than Afmid. Liver and kidney levels of nicotinamide plus NAD/NADP remained the same in WT, HZ and KO mice. Plasma concentrations of formyl-kynurenine, kynurenine, and kynurenic acid were elevated in KO mice (but not HZ mice) relative to WT mice, further suggesting that there must be enzymes other than Afmid (possibly in the kidney) capable of metabolizing formyl-kynurenine into kynurenine. Gradual kidney deterioration and subsequent failure in KO mice is consistent with high levels of tissue-specific Afmid expression in the kidney of WT but not KO mice. On this basis, the most significant function of the kynurenine pathway and Afmid in mice may be in eliminating toxic metabolites and to a lesser extent in providing intermediates for other processes. Published by Elsevier B.V. Keywords: Arylformamidase; Formyl-kynurenine; Kynurenine; Kynurenic acid; Kynurenine formamidase; Tryptophan

1. Introduction A mouse model deficient for the endogenous cytosolic thymidine kinase (Tk) gene [1] is also deficient for the gene coding for arylformamidase (Afmid) (known additionally as kynurenine formamidase; EC 3.5.1.9) [2]. Afmid/Tk knock* Corresponding author. Tel.: +1 870 543 7549; fax: +1 870 543 7393. E-mail address: [email protected] (V.N. Dobrovolsky). 0304-4165/$ - see front matter. Published by Elsevier B.V. doi:10.1016/j.bbagen.2005.03.010

out (KO) mice rarely survive beyond 6 months of age. The omnipresent pathological changes in these mice are sclerosis of glomeruli, a deregulated immune system as evidenced by significantly reduced T-lymphocyte cloning efficiencies, alteration of the secretory function of the sublingual salivary gland from exclusively mucous to predominantly serous, and elevated levels of thymidine in the blood [1]. The phenotype of Afmid/Tk KO animals cannot be explained solely by the lack of Tk if its only function is pyrimidine

164

V.N. Dobrovolsky et al. / Biochimica et Biophysica Acta 1724 (2005) 163 – 172

nucleotide salvage. Mouse Afmid cDNA and protein structure have been defined [3]. The gene coding for Afmid is located on mouse chromosome 11 in close proximity to the Tk gene [2]. The way in which the Tk allele was disrupted in Tk KO mice, by insertion of a neo sequence, would inevitably inactivate the Afmid gene on the same chromosome (Fig. 1). This makes Tk KO animals doubleKOs for both the Tk and the Afmid genes. Afmid plays a key role in the kynurenine pathway of tryptophan degradation to kynurenic and quinolinic acids and ultimately to NAD/NADP (Fig. 2). The disruption of this pathway is implicated in neurotoxicity and immunotoxicity. Kynurenic acid is an antagonist for a subset of neuronal glutamate receptors sensitive to N-methyl-daspartate (NMDA), and quinolinic acid is an agonist for the same subset of receptors. Quinolinic acid exhibits neurotoxic properties through neuronal over-excitation while kynurenic acid may be a neuroprotector by blocking glutamate NMDA receptors and preventing their activation by either glutamate or quinolinic acid [4,5]. Tryptophan degradation is also important for the modulation of the immune response. The initial step of the degradation, opening the indole ring, is activated by ginterferon in the cells of the immune system, as well as in other tissues (reviewed in [6,7]). In addition, mitogen activation leads to increased production of quinolinic and kynurenic acids through the kynurenine pathway [8]. The

induction of the immunosuppressive effect of tryptophan degradation may help tumor cells avoid destruction by cytotoxic lymphocytes [9]. In vivo models for the alteration of the kynurenine pathway of tryptophan metabolism involved, until recently, inhibiting key enzymatic reactions by exogenous compounds. For instance, 1-methyl-tryptophan administered systemically appears to inhibit indoleamine-2,3-dioxygenase (IDO), which catalyzes the first step of the pathway. This results in a higher rate of spontaneous abortion in mice [10]. Some organophosphorous compounds inhibit Afmid and cause marked developmental abnormalities in avian embryos but not in mammalian systems [11– 13]. The present investigation extends this chemically-induced Afmid inhibition in an avian system to gene inactivation in mammals. It examines Afmid/Tk KO mice for alterations in Afmid activity and the concentrations of kynurenine pathway metabolites in relation to kidney failure and immune cell proliferation.

2. Materials and methods 2.1. Mice The animal procedures were approved by the National Center for Toxicological Research Institutional Animal Care

Fig. 1. Genomic organization of the Afmid and Tk genes. (A) Relative location and orientation of the Afmid and Tk genes in mouse chromosome 11 (78 cM). Directions of transcription are shown by arrows for respective genes. (B) In-scale exon/intron structure of the Afmid and the Tk genes. Solid black line represents introns and other non-coding sequences; open boxes represent exons. Directions of transcription from the initiation sites of major transcripts for the genes are shown by angled arrows originating from the shared promoter region. The features within the gray box are enlarged in panel C. Panels A and B were adapted from the NCBI map viewer web site (http://www.ncbi.nlm.nih.gov/mapview/). (C) Organization of the wild-type and inactivated alleles of the Afmid/ Tk locus. Locations of XhoI and ApaI restriction sites on the wild-type allele employed for inserting the neo cassette are shown by 4. Exons are shown by numbered open boxes. The vertical ATGs within the first exons of each gene mark translation initiation sites, which are separated by 170 bases on the genomic DNA. Primers TK14, TK16 and NEO4 (shown by horizontal arrows) were used for genotyping of animals. In the targeted allele both genes were inactivated by replacing the shared promoter region and the first exons of the Tk and Afmid genes with the neo sequence.

V.N. Dobrovolsky et al. / Biochimica et Biophysica Acta 1724 (2005) 163 – 172

165

Fig. 2. Key role of Afmid in tryptophan degradation via the kynurenine pathway. Solid arrows represent single-step reactions (direct conversions); dashed arrows represent conversions through multiple steps (catalyzed by more than one enzyme, with intermediate metabolites that are not shown).

and Use Committee. KO mice were produced by conventional breeding of heterozygous (HZ) parents, which were at least the 5th generation descendent from breeding the original 129/C57Bl6 chimeras to C57Bl6 mice. HZ mice are available from the Mutant Mouse Regional Resource Center, http://www.mmrrc.org/. The three possible genotypes were identified by a three-primer (TK14, TK16 and NEO4 (Fig. 1C)) PCR as described earlier [1]. 2.2. Northern blot Total RNA was prepared from mouse tissues using the RNeasy kit (Qiagen, Valencia, CA). About 5 Ag of RNA was resolved on an agarose gel and transferred onto a nylon membrane using standard protocols. The 400 bp Afmid probe was PCR-synthesized from a plasmid containing Afmidexpressed sequence tag AA245789 (gift from Dr. C. Seiser [2]), using primers Afmid1 (5V-GCTGGATGTCCCCTATGGAGAT; annealing to exon 3 of the Afmid gene) and Afmid2 (5V-TAGATCCCACTCACCAGGAGAAAG; annealing to exon 8 of the Afmid gene), and the HotStarTaq kit (Qiagen); temperature profile: 95 -C  15 min + (95 -C  1 min + 56 -C  1 min + 72 -C  3 min)  30. The complete Afmid cDNA sequence (accession number NM_027827) is available on the NCBI web site (http://www.ncbi.nlm.nih.gov/ entrez). The accession number for the DNA sequence of mouse chromosome 11 comprising the Afmid gene is NT_039521. The h-actin probe was a ¨570 bp DNA fragment produced by RT-PCR of total RNA from mouse lymphocytes using primers m-actin-F (5V-TGGGTCAGAAGGACTCCTATG) and m-actin-R (5V-CAGGCAGCT-

CATAGCTCTTCT). The probes were 32P-labeled using the Rediprime II random priming kit (Amersham, Piscataway, NJ) and hybridized according to standard protocols. The blot was hybridized with the Afmid probe, then washed and re-hybridized with the h-actin probe for normalization. 2.3. Enzymatic activity Frozen liver and kidney were homogenized at 20% w/v in 50 mM sodium phosphate (pH 7.4) and 1 mM EDTA at 4 -C in a glass homogenizer. The homogenate was centrifuged at 10,000  g for 10 min and the supernatant immediately used for assay. Afmid specific activity was determined for 6 –200 Ag protein with 0.45 mM formyl-kynurenine in 200 Al of 50 mM sodium phosphate buffer (pH 7.4), monitoring the absorbance increase at 365 nm for 90 s resulting from the release of l-kynurenine. Flat-bottomed 96-well plates were used with a Versamax Kinetic Microplate Reader (Molecular Devices, Sunnyvale, CA) and linear regression with SOFTmax v. 4.0 data analysis software. N-formyl-lkynurenine was synthesized as previously described [3,14]. Protein was determined by the Bradford procedure [15] using a Microplate Protein Assay (Bio-Rad, Hercules, CA) with bovine serum albumin as the standard. Activity was expressed as nmol formyl-kynurenine hydrolyzed (i.e., kynurenine liberated)/min/mg protein. 2.4. Analysis of blood plasma Blood was collected by cardiac puncture into heparincoated Microtainer gel tubes (Beckton-Dickinson, Franklin

166

V.N. Dobrovolsky et al. / Biochimica et Biophysica Acta 1724 (2005) 163 – 172

Lakes, NJ). Plasma was separated by a 2 min centrifugation at 15,000  g, transferred to fresh tubes, and stored at 80 -C until ready to use. Solid-phase extraction of plasma samples was carried out using Isolute ENV+ 96-well array cartridges (50 mg, 1 cc; Argonaut Technologies, Foster City, CA) under reduced pressure. The activation of the cartridges was achieved with two 1-ml washes of methanol, followed by two 1-ml washes of aqueous 0.1% formic acid. Plasma samples (20 Al) were diluted with 200 Al of 0.1% formic acid and applied to the cartridges which were then washed twice with 1 ml of water followed by elution with three 400 Al volumes of 5% ammonium hydroxide in methanol. The eluate was reduced to dryness using a centrifugal vacuum concentrator and reconstituted into 200 Al of 0.1% formic acid. Liquid chromatography (LC) was performed using a Waters 2695 liquid handling system (Waters Assoc., Milford, MA). Chromatographic separation was achieved on a Betamax base analytical column (2  100 mm  5 Am particles; Thermo Hypersil-Keystone, Bellefonte, PA) equipped with a Betamax base guard cartridge (2 mm; Thermo Hypersil-Keystone). An isocratic mobile phase of 95% 0.1% formic acid and 5% acetonitrile was delivered at a flow rate of 0.25 ml/min. Injection volumes were 50 Al and all separations were performed at ambient temperature. The entire column effluent from the LC step was directed into a Quattro Micro triple quadrupole mass spectrometer (Waters Assoc., Manchester, UK) equipped with an electrospray interface. Positive precursor and product ions were acquired in the multiple reaction monitoring (MRM) mode using a desolvation temperature of 400 -C, a source temperature of 120 -C, a dwell time of 0.1 s, and a collision cell pressure of argon at 3.2  10 3 mBar. Three MRM transitions were acquired for each of the four compounds analyzed for additional specificity, although only the strongest transition for each was used for quantitation. The mass spectrometry (MS) conditions are shown in Table 1. Individual analytes were quantified by the method of standard addition (i.e., the samples were processed twice,

Table 1 Conditions for mass spectrometry analysis of tryptophan degradation metabolites in mouse plasma Compound (cone voltage, V)

Transition (collision energy, eV)

Tryptophan (15)

205 205 205 237 237 237 209 209 209 190 190 190

N-formyl-kynurenine (15)

Kynurenine (15)

Kynurenic acid (25)

a

Y Y Y Y Y Y Y Y Y Y Y Y

118 (26) 146 (17) 188a (10) 136 (17) 192 (12) 220a (9) 146 (18) 174 (17) 192a (9) 144a (18) 162 (17) 172 (12)

Major transitions used for the quantitation of metabolites.

once with a known amount of authentic standard added and the second time without). The difference in peak areas was attributed to the standard addition; hence, average response factors were calculated for kynurenine, kynurenic acid, and tryptophan using standards purchased from Sigma (St. Louis, MO). N-formyl-kynurenine was arbitrarily assigned a response factor equal to that for kynurenine since a standard sample of adequate purity was not available at the time of these determinations. For determining the levels of kynurenine metabolites in the blood of 4-week-old mice, a similar LC-MS procedure was employed using alternative hardware, an UltiMatei capillary HPLC system (LC Packings, Amsterdam, Netherlands) for chromatography and an API 365 triple quadrupole mass spectrometer (PE-Sciex, Concord, Canada) for the detection of analytes. Previously described conditions were used for the analysis [16]. 2.5. Analysis of liver and kidney The LC procedure of Shibata et al. [17] was used to analyze nicotinamide and its derivatives in the liver and kidney. The tissue homogenate was diluted with water and placed in a boiling water bath for 10 min. This procedure stoichiometrically converted NAD and NADP to nicotinamide, so the method determined total nicotinamide equivalents. The cleanup procedure involved the addition of potassium carbonate to the aqueous sample and then extraction with ether, giving high recoveries of nicotinamide (98% reported by Shibata et al. [17] and confirmed in the present study) with little or no interfering peaks. More specifically, the liver and kidney were homogenized in water at 20% w/v. A 200-Al aliquot was added to 800 Al of water in a 1.5-ml Eppendorf tube. The samples were held at 100 -C for 10 min and chilled on ice for 20 min before centrifugation at 10,000  g for 10 min, retaining the supernatant fraction. The pellet was washed with 1 ml of water, centrifuged, and the supernatants combined. Extraction used 1 ml of the recovered supernatants after the addition of 1.2 g of potassium carbonate and cooling on ice. Diethyl ether (5 ml) was added at 25 -C, mixed well by intermittent vortexing over a period of 5 min, and centrifuged at 600  g for 1 min. The ether layer was retained and the extraction was repeated two more times. The combined ether layers were evaporated to dryness in a Speed-Vac Concentrator (Savant, Farmingdale, NY) and resuspended in 500 Al of water. Samples were syringe filtered through a 4 mm  0.45 Am Supelco IsoDisk PFTE-4-4 (Supelco, Bellefonte, PA) and an aliquot of 50– 75 Al was injected onto the column for analysis. The LC conditions were: column, Phenomenex Ultremex 5 C18 (4.6 mm  25 cm); mobile phase, 10 mM potassium phosphate pH 4.5/acetonitrile (94/6 v/v); flow, 1 ml/min; detection, OD 260 nm; Waters HPLC system with 600E controller and 994 detector (Waters Assoc., Milford, MA); and nicotinamide retention time, 8.2 min. The analyzed samples were quantitated against a calibration curve for nicotinamide.

V.N. Dobrovolsky et al. / Biochimica et Biophysica Acta 1724 (2005) 163 – 172

167

2.6. Lymphocyte cloning assay Cloning efficiencies for mouse spleen lymphocytes were determined by limiting-dilution culture in 96-well plates as described earlier [18]. l-Kynurenine was added to the growth medium at 5 mg/l where appropriate. 2.7. Evaluation of neurodegeneration For the histological evaluation of neurodegeneration, the mice were anesthetized with 80 mg/kg of pentobarbital and perfused with 50 ml of saline followed by 250 ml of 4% formaldehyde in 0.1 M sodium phosphate buffer, pH 7.4. Brains were postfixed for 2 days, and then 40-Am-thick coronal sections were cut and collected in 2% formaldehyde with 0.1 M phosphate buffer, pH 7.4, and stored at 4 -C until processing. Continuous sectioning, with all sections collected, was performed in the forebrain from 0.5 mm anterior to the genu of the corpus collosum through the substantia nigra (midbrain). The Fluoro-Jade\B labeling procedure was performed according to Schmued and Hopkins [19]. Any fluorescently labeled cells present in the sections were detected with an epifluorescence microscope with a filter system designed for visualizing fluorescein. The isolectin B4 labeling of microglia for detecting microgliosis was performed as described previously [20]. Brain sections were incubated overnight at 4 -C in a solution of isolectin B4 from Griffonia simplicifolia (10 Ag/ ml; Sigma) coupled to horseradish peroxidase and the binding sites were visualized with 3,3V-diaminobenzidine (Sigma) and H2O2. Immunohistochemical staining for glial fibrillary acidic protein (GFAP) was used to detect astrocytic hypertrophy. A rabbit antiserum to GFAP (1:1000) from Dako (Carpinteria, CA) was incubated with the sections overnight at 4 -C; then the slides were hybridized with swine anti-rabbit secondary antibody (Dako) and rabbit peroxidase-anti-peroxidase complex (Dako). GFAP-positive astrocytes within the sections were visualized by staining with 3,3V-diaminobenzidine. 2.8. Statistical analysis Statistical analysis was performed using SigmaStat for Windows Version 3.0 (SPSS, Chicago, IL). Group comparisons (WT vs. KO, HZ vs. KO) were made using the t-test or rank sum test (when normality or equal variance tests failed). The difference between two groups was considered statistically significant when P < 0.05.

3. Results 3.1. Simultaneous disruption of Afmid and Tk genes The gene coding for Afmid was mapped to chromosome 11, next to the endogenous cytosolic Tk gene in a

Fig. 3. Identification and characterization of Afmid/Tk -deficient mice. (A) Typical PCR genotyping of offspring from two heterozygous (HZ) parents using three primers, TK14, TK16 and NEO4 (Fig. 1). (B) Detection of the Afmid transcript in the liver, kidney and salivary gland of mice of three different genotypes. h-Actin probe was used for estimating relative amount of total RNA in each sample. (C) Afmid enzymatic activity in the liver and the kidney of mice of three different genotypes. The data are presented as average values for three animals of each genotype, with error bars indicating the standard deviation.

168

V.N. Dobrovolsky et al. / Biochimica et Biophysica Acta 1724 (2005) 163 – 172

head-to-head array [2] (Fig. 1A, B). On the genomic DNA, 170 bp separate their respective translation initiation codons. In the targeted allele, both genes were inactivated by replacing the shared promoter region and the first exons of the Afmid and Tk genes with a neo sequence (Fig. 1C). PCR genotyping of the offspring produced by two HZ parents confirmed the wild-type (WT) (Afmid +/+/Tk +/+), HZ (Afmid +/ /Tk +/ ), and inactivated KO (Afmid / /Tk / ) alleles (Fig. 3A). Afmid mRNA was present in the liver and the kidney of WT and HZ animals, but not in the salivary gland. KO

animals lacked detectable Afmid transcript in the liver, kidney and salivary gland (Fig. 3B). 3.2. Effect of gene inactivation on Afmid enzymatic activity In WT mice, Afmid specific activity was 2.2- to 2.3-fold higher in liver than in the kidney (Fig. 3C). In HZ animals, the activity was reduced to 53% in the liver and 57% in the kidney of that in the WT mice. The KO mice were essentially devoid of liver Afmid activity but 13% of the kidney activity was retained (Fig. 3C).

Fig. 4. Metabolites of the kynurenine pathway in the plasma of mice of three different genotypes. The increase in the concentration of formyl-kynurenine, kynurenine and kynurenic acid in 20-week-old KO mice was significant ( P < 0.05) when compared to either WT or HZ animals. One 20-week-old KO animal produced an extremely high level of kynurenic acid (¨9 AM or ¨1.7 ng/Al) and was excluded from the statistical analysis as an outlier. Formyl-kynurenine concentrations were determined using the response factor for kynurenine and are, therefore, approximate (see Materials and methods). The data are presented as average concentrations, with error bars indicating standard deviation. Replication for each genotype at 9 weeks involved 4 mice and at 20 weeks involved 3, 6, and 4 or 5 mice for WT, HZ, and KO genotypes, respectively.

V.N. Dobrovolsky et al. / Biochimica et Biophysica Acta 1724 (2005) 163 – 172

169

3.3. Effect of gene inactivation on the levels of kynurenine pathway metabolites The levels of tryptophan and tryptophan metabolites were compared in plasma from WT, HZ and KO mice at 9 and 20 weeks of age (Fig. 4); kynurenine concentrations were also determined at 4 weeks (data not shown). Although Afmid enzymatic activity in WT mice was twice as high as in HZ animals, concentrations of all measured metabolites in WT and HZ animals were not significantly different. The tryptophan concentration remained unchanged (within the variability of the experiment) in animals having different Afmid/Tk genotypes. Formyl-kynurenine was increased in KO animals at 9 and 20 weeks of age, with more pronounced effects in older animals. The products downstream of formyl-kynurenine were also increased in KO animals. Kynurenine was higher in 4-, 9- and 20-week-old KO animals than in age-matched WT or HZ animals. Kynurenic acid was also significantly increased (more than 5-fold) in 9- and 20-week-old KO animals. Despite the high variability of the samples derived from KO animals, the increase in the concentration of formyl-kynurenine, kynurenine and kynurenic acid in 20-week-old KO mice was significant ( P < 0.05) when compared to either WT or HZ animals. 3.4. Effect of gene inactivation on liver and kidney nicotinamide equivalents There was no significant difference in the levels of total nicotinamide and NAD-type compounds in the livers from WT, HZ and KO mice (Fig. 5). This was also true for kidney levels of nicotinamide equivalents. 3.5. Effect of L -kynurenine on reduced cloning efficiencies of spleen lymphocytes from mice deficient in Afmid and Tk activities

Fig. 5. Nicotinamide equivalents in liver and kidney of mice of three different genotypes. The data are presented as average values for four WT animals, four HZ animals and three KO animals, with error bars representing the standard deviation.

section [19]. Examination of the GFAP-positive astrocytes in the forebrain of the KO mice, compared to either the WT or HZ mice, indicated there also was no ongoing reactive astrocytosis. In addition, there were no changes in the number of either the phagocytic or activated microglia which indicated there was no active microgliosis in the forebrain of KO mice (data not shown).

4. Discussion There was poor ex vivo activation/propagation of lymphocytes from Afmid/Tk KO mice (Table 2, and reported previously [1]). This was conceivably due to a deficiency in kynurenine or one of its derivatives in primary cultures. However, supplementation of the growth medium with 5 mg/l kynurenine, an amount equal to the tryptophan concentration in the RPMI1640 medium (5 mg/l), failed to reverse the reduced cloning efficiencies of lymphocytes derived from KO mice (Table 2). Moreover, added kynurenine decreased the cloning efficiency of lymphocytes from both Afmid-deficient and Afmid-proficient mice. 3.6. Neurological consequences of gene inactivation There was no neurodegeneration in KO mice as evidenced by the absence of neuronal staining with Fluoro-Jade\B (not shown), which has sufficient sensitivity to detect the degeneration of one or two neurons per brain

4.1. Possible unidentified formamidase(s) in formyl-kynurenine metabolism Afmid/Tk KO mice still have 13% of their original kidney formyl-kynurenine-hydrolyzing activity suggesting that there may be a kidney enzyme other than Afmid with formamidase activity. It is also conceivable that in order to eliminate excess formyl-kynurenine nonspecific formamidases may be upregulated in KO mice which, in turn, may be toxic for a specific cell type (e.g., the cells of the glomeruli). The simultaneous accumulation of the substrate, formyl-kynurenine, and the product, kynurenine, in animals deficient for Afmid may be due to the activity of unidentified formamidases, which are less efficient than Afmid. These unidentified formamidases may be regulated in a different way than Afmid and have different tissue/cell-type specificity. As an analogous example, in the absence of cytosolic tyrosine amino-

170

V.N. Dobrovolsky et al. / Biochimica et Biophysica Acta 1724 (2005) 163 – 172

Table 2 Cloning efficiencies (CEs) of spleen lymphocytes from mice deficient in Afmid and Tk activities in regular growth medium and in medium supplemented with 5 mg/l of l-kynurenine Mouse

Genotypea

Growth medium

CE (%)

1 2 3

HZ HZ HZ

4

KO

5

KO

Regular Regular Regular Regular + kynurenine Regular Regular + kynurenine Regular Regular + kynurenine

13.7 8.9 11.6 8.5 2.28 0.88 1.67 1.02

a HZ—heterozygous, Afmid/Tk +/ ; KO—knockout, Afmid/Tk total of 3 HZ and 2 KO animals was used in this experiment.

/

. A

transferase, a mitochondrial enzyme compartmentalizes the product of the reaction ( p-hydroxyphenyl pyruvate) in the mitochondria and shields the product from further processing by cytosolic enzymes. This results in increased urinary excretion of the substrate (tyrosine) and the oxidized product of the pathway [21]. Little is known about alternative kynurenine formamidases, except for the fact that they may exist in several species [22 –24]. As a possibility, kynurenine removal (presumably though the kidneys) may be coupled with a reaction catalyzed by Afmid (presumably in the kidney). Apparently, kynurenine (and its downstream metabolites) cannot be efficiently removed from the blood. 4.2. Afmid disruption relative to tryptophan as a source of nicotinamide and NAD/NADP Tryptophan is an essential amino acid and must be supplied in the diet. The standard NIH-31 diet for laboratory rodents contains adequate amounts of nicotinamide or precursors thereof. The question remains as to whether the kynurenine pathway is used primarily for the excretion of toxic metabolites or as a backup to de novo NAD synthesis. The viability of KO animals for only an average of about 170 days [1] and the specific kidney failure suggests that tryptophan degradation contributes minimally to NAD metabolism at the level of the whole animal. Cell culture media contain an adequate amount of niacin to minimize the in vitro effect of the kynurenine pathway in conversion of tryptophan to NAD. Avian species may be more dependent upon tryptophan degradation as the source of nicotinamide due to the closed nature of development in the egg. Chick malformations are common in systems where Afmid activity is even partially reduced by chemical inhibitors; in open mammalian systems, complete inhibition is impossible and partial inhibition of Afmid has less noticeable effects [25]. The present study is consistent with this view since almost complete loss of Afmid activity in KO mice did not significantly alter the liver or kidney level of nicotinamide derivatives.

4.3. Afmid gene disruption relative to kynurenine pathway metabolites Disruption of the Afmid gene causes a significant imbalance in several metabolites of the tryptophan degradation pathway consistent with gradual kidney deterioration. Thus, the observed changes resemble the biochemical condition in animals with surgically-induced renal failure, i.e., increased levels of kynurenine and the products of kynurenine downstream degradation in the blood [26]. Despite the increase of formyl-kynurenine in the blood, there was no increase in tryptophan concentration in KO mice, which may indicate upregulation of the initial step of tryptophan degradation by liver tryptophan 2,3-dioxygenase (TDO) (usually regulated by general mechanisms, such as insulin and corticosteroids) or by the ubiquitous IDO (usually expressed in response to g-interferon, viral infections and endotoxins [27 –29]). In WT animals, Afmid transcript is constitutively expressed at high levels in the liver, to a lesser extent in the kidney, and at low levels or not at all in many other tissues, such as heart, brain, spleen, lung, skeletal muscle, testis and salivary gland (this study and [2]). Afmid enzymatic activity in the liver and the kidney is much higher (20 – 500) than in other tissues; nevertheless, Afmid (or another formamidase(s)) is present in most tissues at levels capable of rapidly producing kynurenine upon the stimulation of IDO, so that tissue concentrations of formyl-kynurenine are low in WT mice [30]. Excess formyl-kynurenine in tissues also can enter the blood pool to be further converted by Afmid in the liver or in the kidney. In Afmid/Tk KO mice the degradation pathway is blocked, resulting in the accumulation of formyl-kynurenine in the blood plasma and, probably, in tissues. It is not known to what degree excess formyl-kynurenine may inhibit or promote other steps of the overall kynurenine pathway further contributing to altered tryptophan metabolism. The imbalance in kynurenine metabolites does not directly explain the mechanism of kidney failure in KO animals, but suggests that higher levels of formyl-kynurenine due to Afmid deficiency may be toxic for the cells of the glomerulus. 4.4. Kidney failure in Afmid/Tk KO mice Kidney pathology in Afmid/Tk KO mice is consistent with the current knowledge of the tissue-specific expression of the Afmid gene, with the kidney functioning as the organ for removal of kynurenine catabolites and the glomerulus as the interface between the blood and urine. Perhaps, Afmid participates in pathways other than tryptophan degradation that are required for proper vascular functioning. The exact mechanism for obliteration of the glomeruli remains to be defined. In the liver, where TDO is constitutively expressed, the imbalance in tryptophan metabolites does not appear to

V.N. Dobrovolsky et al. / Biochimica et Biophysica Acta 1724 (2005) 163 – 172

be toxic, as no obvious pathology has been found in the liver of KO animals.

171

mamidases in KO mice may not provide sufficient levels of products rapidly.

4.5. Neurological consequences of Afmid/Tk deficiency 5. Conclusions Although kynurenine metabolism is significantly disturbed in the blood of Afmid/Tk-deficient mice, it doesn’t appear to affect the brain. The blood levels of quinolinic acid in KO mice were not determined in this study. As evidenced by staining with Fluoro-Jade\B, however, there was no neurotoxicity in KO mice that could be associated with neuronal over-excitation (e.g., potentiated by quinolinic acid). However, if the levels of kynurenic acid are elevated to the same extent in the brain as in the blood, the signaling through NMDA receptors may be reduced in KO animals. This might be neuroprotective, and contribute to the sluggishness and reduced motor coordination (hunched posture) seen in KO mice by the end of their natural life span [1]. Combined with the absence of stress indication in microglia and astrocytes of the forebrain, the evidence suggests that the elevated levels of metabolites of the kynurenine pathway not only did not result in neurodegeneration but also did not induce either nerve terminal or dendrite degeneration. The inactivation of another key enzyme of the kynurenine pathway, kynurenine aminotransferase, does not result in significant neurotoxicity in mice [31]. This indicates that compensatory mechanisms, such as enzyme isoforms, may play a significant role in tryptophan degradation in the brain. In addition, the existence of a brain-specific formamidase has been proposed [22]. The levels of formyl-kynurenine, kynurenine, kynurenic acid and quinolinic acid within most regions of the brain are regulated by the selective transporters present in the vasculature and the choroid plexus of the brain. Thus, their levels in the brain might not be as drastically altered as in the blood, which might also explain the lack of neurotoxicity observed in the brain. 4.6. Immune system in Afmid/Tk KO mice The lowered lymphocyte count, poor lymphocyte cloning efficiency, occasional arteritis, restructured spleen [1], and increased micronucleus formation [32] in Afmid/Tk KO mice indicate that disruption of tryptophan metabolism is involved in disregulation of lymphoid tissue and erythropoiesis. Certain types of T-cells may be sensitive to kynurenine metabolites [33]. There also are indications that the pathway and its metabolites may be needed not only for fine-tuning immunosuppression [34], but also for lymphocyte maturation [35]. The details of how tryptophan degradation may be involved in the modulation of the immune response require further investigation. But it cannot be excluded that surges (e.g., induced by g-interferon) in demand for kynurenine metabolites may be required for proper response and that non-specific (unidentified) for-

Simultaneous inactivation of the Afmid and Tk genes makes it difficult to delineate the contribution of each gene individually to the KO phenotype. Single-gene KO animals may not be easily produced because of the close proximity of the two genes, overlapping transcription pattern, and possible shared regulatory sequences [2]. A conventional targeted inactivation of one gene may change the expression pattern for the other. Adding either a Tk or Afmid transgene to double-KO animals may not necessarily create a true single-KO animal, as the transgene may not follow the expression pattern of the endogenous genes. Afmid (but not Tk) is normally expressed in the kidney, suggesting that Afmid deficiency contributes to sclerosis of kidney glomeruli more than Tk deficiency in Afmid/Tk KO mice. The phenotype and known biochemical activities encoded by the Afmid gene suggest that Afmid deficiency may contribute to the kidney damage and deranged immune system of KO mice. The kynurenine pathway of tryptophan degradation remains functional although altered in vivo when its key enzyme, Afmid, is inactivated. The most significant function of the kynurenine pathway and Afmid in vivo may be in eliminating toxic metabolites and to a lesser extent in providing intermediates for other processes.

Acknowledgments This work was partially supported by Grant R01 ES08762 to John E. Casida from the National Institute of Environmental Health Sciences (NIEHS), NIH, and its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIEHS, NIH. It was also partially supported by Grants 621-20025261 and 629-2002-6821 from the Swedish Research Council. The views presented in this article do not necessarily reflect those of the US Food and Drug Administration.

References [1] V.N. Dobrovolsky, T. Bucci, R.H. Heflich, J. Desjardins, F.C. Richardson, Mice deficient for cytosolic thymidine kinase gene develop fatal kidney disease, Mol. Genet. Metab. 78 (2003) 1 – 10. [2] B. Schuettengruber, A. Doetzlhofer, K. Kroboth, E. Wintersberger, C. Seiser, Alternate activation of two divergently transcribed mouse genes from a bidirectional promoter is linked to changes in histone modification, J. Biol. Chem. 278 (2003) 1784 – 1793. [3] M.K. Pabarcus, J.E. Casida, Kynurenine formamidase: determination of primary structure and modeling-based prediction of tertiary

172

[4] [5]

[6] [7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

V.N. Dobrovolsky et al. / Biochimica et Biophysica Acta 1724 (2005) 163 – 172 structure and catalytic triad, Biochim. Biophys. Acta 1596 (2002) 201 – 211. T.W. Stone, Kynurenines in the CNS: from endogenous obscurity to therapeutic importance, Prog. Neurobiol. 64 (2001) 185 – 218. T.W. Stone, G.M. Mackay, C.M. Forrest, C.J. Clark, L.G. Darlington, Tryptophan metabolites and brain disorders, Clin. Chem. Lab. Med. 41 (2003) 852 – 859. J.R. Moffett, M.A. Namboodiri, Tryptophan and the immune response, Immunol. Cell Biol. 81 (2003) 247 – 265. B. Wirleitner, G. Neurauter, K. Schrocksnadel, B. Frick, D. Fuchs, Interferon-gamma-induced conversion of tryptophan: immunologic and neuropsychiatric aspects, Curr. Med. Chem. 10 (2003) 1581 – 1591. K. Saito, J.S. Crowley, S.P. Markey, M.P. Heyes, A mechanism for increased quinolinic acid formation following acute systemic immune stimulation, J. Biol. Chem. 268 (1993) 15496 – 15503. M. Friberg, R. Jennings, M. Alsarraj, S. Dessureault, A. Cantor, M. Extermann, A.L. Mellor, D.H. Munn, S.J. Antonia, Indoleamine 2,3dioxygenase contributes to tumor cell evasion of T cell-mediated rejection, Int. J. Cancer 101 (2002) 151 – 155. A.L. Mellor, P. Chandler, G.K. Lee, T. Johnson, D.B. Keskin, J. Lee, D.H. Munn, Indoleamine 2,3-dioxygenase, immunosuppression and pregnancy, J. Reprod. Immunol. 57 (2002) 143 – 150. M. Eto, J. Seifert, J.L. Engel, J.E. Casida, Organophosphorus and methylcarbamate teratogens: structural requirements for inducing embryonic abnormalities in chickens and kynurenine formamidase inhibition in mouse liver, Toxicol. Appl. Pharmacol. 54 (1980) 20 – 30. J. Seifert, J.E. Casida, Relation of yolk sac membrane kynurenine formamidase inhibition to certain teratogenic effects of organophosphorus insecticides and of carbaryl and eserine in chicken embryos, Biochem. Pharmacol. 27 (1978) 2611 – 2615. A.D. Moscioni, J.L. Engel, J.E. Casida, Kynurenine formamidase inhibition as a possible mechanism for certain teratogenic effects of organophosphorus and methylcarbamate insecticides in chicken embryos, Biochem. Pharmacol. 26 (1977) 2251 – 2258. C.E. Dalgliesh, The synthesis of N-formyl-dl-kynurenine, N a-acetyldl-kynurenine and related compounds, and observations on the synthesis of kynurenine, J. Chem. Soc. 1952 (1952) 137 – 141. M.M. Bradford, A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein – dye binding, Anal. Biochem. 72 (1976) 248 – 254. A. Amirkhani, E. Heldin, K.E. Markides, J. Bergquist, Quantitation of tryptophan, kynurenine and kynurenic acid in human plasma by capillary liquid chromatography-electrospray ionization tandem mass spectrometry, J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. 780 (2002) 381 – 387. K. Shibata, T. Kawada, K. Iwai, High-performance liquid chromatographic determination of nicotinamide in rat tissue samples and blood after extraction with diethyl ether, J. Chromatogr. 422 (1987) 257 – 262. V.N. Dobrovolsky, J.G. Shaddock, R.H. Heflich, 7,12-dimethylbenz[a]anthracene-induced mutation in the Tk gene of Tk +/ mice: automated scoring of lymphocyte clones using a fluorescent viability indicator, Environ. Mol. Mutagen. 36 (2000) 283 – 291. L.C. Schmued, K.J. Hopkins, Fluoro-Jade B: a high affinity fluorescent marker for the localization of neuronal degeneration, Brain Res. 874 (2000) 123 – 130.

[20] W.J. Streit, An improved staining method for rat microglial cells using the lectin from Griffonia simplicifolia (GSA I-B4), J. Histochem. Cytochem. 38 (1990) 1683 – 1686. [21] J.H. Fellman, P.J. Vanbellinghen, R.T. Jones, R.D. Koler, Soluble and mitochondrial forms of tyrosine aminotransferase. Relationship to human tyrosinemia, Biochemistry 8 (1969) 615 – 622. [22] R.B. Cumming, M.F. Walton, J.C. Fuscoe, B.A. Taylor, J.E. Womack, F.H. Gaertner, Genetics of formamidase-5 (brain formamidase) in the mouse: localization of the structural gene on chromosome 14, Biochem. Genet. 17 (1979) 415 – 431. [23] G.P. Moore, D.T. Sullivan, The characterization of multiple forms of kynurenine formidase in Drosophila melanogaster, Biochim. Biophys. Acta 397 (1975) 468 – 477. [24] R. Shinohara, I. Ishiguro, New formamidase having substrate specificity for o-formylaminoacetophenone in pig liver, Biochim. Biophys. Acta 483 (1977) 409 – 415. [25] J. Seifert, J.E. Casida, Mechanisms of teratogenesis induced by organophosphorous and methylcarbamate insecticides, in: D.H. Hudson, T.R. Roberts (Eds.), Progress in Pesticide Biochemistry, vol. 1, John Wiley and Sons, Ltd., 1981, pp. 219 – 246. [26] D. Pawlak, A. Tankiewicz, T. Matys, W. Buczko, Peripheral distribution of kynurenine metabolites and activity of kynurenine pathway enzymes in renal failure, J. Physiol. Pharmacol. 54 (2003) 175 – 189. [27] R. Yoshida, J. Imanishi, T. Oku, T. Kishida, O. Hayaishi, Induction of pulmonary indoleamine 2,3-dioxygenase by interferon, Proc. Natl. Acad. Sci. U. S. A. 78 (1981) 129 – 132. [28] R. Yoshida, O. Hayaishi, Induction of pulmonary indoleamine 2,3dioxygenase by intraperitoneal injection of bacterial lipopolysaccharide, Proc. Natl. Acad. Sci. U. S. A. 75 (1978) 3998 – 4000. [29] R. Yoshida, Y. Urade, M. Tokuda, O. Hayaishi, Induction of indoleamine 2,3-dioxygenase in mouse lung during virus infection, Proc. Natl. Acad. Sci. U. S. A. 76 (1979) 4084 – 4086. [30] O. Takikawa, R. Yoshida, R. Kido, O. Hayaishi, Tryptophan degradation in mice initiated by indoleamine 2,3-dioxygenase, J. Biol. Chem. 261 (1986) 3648 – 3653. [31] P. Yu, N.A. Di Prospero, M.T. Sapko, T. Cai, A. Chen, M. Melendez-Ferro, F. Du, W.O. Whetsell Jr., P. Guidetti, R. Schwarcz, D.A. Tagle, Biochemical and phenotypic abnormalities in kynurenine aminotransferase II-deficient mice, Mol. Cell. Biol. 24 (2004) 6919 – 6930. [32] V.N. Dobrovolsky, L.J. McGarrity, L.S. Vontungeln, R.A. Mittelstaedt, S.M. Morris, F.A. Beland, R.H. Heflich, Micronucleated erythrocyte frequency in control and azidothymidine-treated Tk +/+, Tk +/ and Tk / mice, Mutat. Res. 570 (2005) 227 – 235. [33] T. Morita, K. Saito, M. Takemura, N. Maekawa, S. Fujigaki, H. Fujii, H. Wada, S. Takeuchi, A. Noma, M. Seishima, 3-Hydroxyanthranilic acid, an l-tryptophan metabolite, induces apoptosis in monocytederived cells stimulated by interferon-gamma, Ann. Clin. Biochem. 38 (2001) 242 – 251. [34] F. Fallarino, U. Grohmann, C. Vacca, R. Bianchi, C. Orabona, A. Spreca, M.C. Fioretti, P. Puccetti, T cell apoptosis by tryptophan catabolism, Cell Death Differ. 9 (2002) 1069 – 1077. [35] J.R. Moffett, M.G. Espey, M.A. Namboodiri, Antibodies to quinolinic acid and the determination of its cellular distribution within the rat immune system, Cell Tissue Res. 278 (1994) 461 – 469.