- Email: [email protected]
Serine hydroxymethyltransferase in the central nervous system: regional and subcellular distribution studies
Brain Research, 54 (1973) 149-156
149
© ElsevierScientificPublishingCompany,Amsterdam- Printedin The Netherlands
SERINE HYDROXYMETHYLTRANSFERASE IN THE CENTRAL NERVOUS SYSTEM: REGIONAL AND SUBCELLULAR DISTRIBUTION STUDIES
L. P. DAVIES ANDG. A. R. JOHNSTON Department of Physiology, Australian National University, Canberra (Australia)
(AcceptedOctober20th, 1972)
SUMMARY
No significant differences were found in the serine hydroxymethyltransferase activity of extracts from different regions of the adult rat central nervous system. Within the cat spinal cord, some degree of correlation between glycine levels and hydroxymethyltransferase activity was apparent, with higher levels of both in ventral as compared with dorsal grey matter. Subcellular distribution studies in the rat cerebral cortex and spinal cord suggested that this enzymic activity is present in both the cytoplasm and the mitochondria. A number of centrally-active drugs and structural analogues of glycine and serine were found to inhibit hydroxymethyltransferase activity, the most effective being those which are known to inhibit other pyridoxal-dependent enzymes.
INTRODUCTION
The evidence that glycine is a major inhibitory transmitter in certain areas of the mammalian central nervous system, including the spinal cord and medulla oblongata, has been well documentedl, 5. The major pathways whereby glycine is metabolised in nervous tissue are, however, still unknown. The enzyme serine hydroxymethyltransferase (L-serine-tetrahydrofolate 5,10hydroxymethyltransferase, E.C. 2.1.2.1), which catalyses the interconversion of glycine and serine, has been recently purified from bovine brain 3 and there is some evidence to suggest that this enzyme may play an important role in glycine metabolism16. The observation2,16,17 that glycine levels vary from one CNS region to another suggested that the regional distribution of serine hydroxymethyltransferaseactivity might reveal important information about the possible role of this activity in glycine metab-
I 50
L. P. DAVIES A N D G. A. R. JOHNSTON
olism. We have studied the regional and subcellular distribution of this enzymic activity in extracts of rat and cat central nervous tissue, and investigated some structural analogues of glycine and serine, and various centrally-active drugs, as potential inhibitors of serine hydroxymethyltransferaseactivity. METHODS
Assay of hydroxymethyltransferase activity The rate of production of radioactive formaldehyde from radioactive serine was assayed by the method of Taylor and Weissbach 18. The reaction mixtures contained 0.5 mM [3-14C]serine, 0.5 mM pyridoxal phosphate, 2 mM tetrahydrofolate and 4 mM dithiothreitol in 100 mM potassium phosphate buffer (pH 7.4) in a final volume of 2 ml. Reactions carried out at 37 °C in a shaking water bath were initiated by the addition of 0.5 ml of extract. At various times, 0.4 ml aliquots were removed, and reaction terminated by addition to a mixture of 0.3 ml of 1 M sodium acetate solution (pH 4.5) and 0.2 ml of 0.1 M formaldehyde. The [laC]formaldehyde, formed during the reaction, was trapped with dimedone, extracted into toluene and counted in a toluene-based scintillator in a Beckman LSI00 counter, correcting for any quenching by means of an external standard. The enzymic activities measured in this way were linear for at least 40 rain. Each assay provided 4 time aliquots and the rate of production of formaldehyde was caIculated from the slope of the line of best fit through a plot of radioactivity versus time, by a least squares analysis using a PDP-8L computer. Protein concentrations in the tissue extracts were measured by the biuret method 9, or the method of Lowry et al. 14, removing lipid by ether extraction immediately prior to reading in the spectrophotometer. Regional distribution of serine hydroxymethyltransferase activity in extracts of rat central nervous tissue Adult Wistar rats (150-200 g) were killed by decapitation, the brains and spinal cords rapidly removed and chilled, and the various regions quickly separated. The tissue samples were then weighed and homogenised in 0.1 M potassium phosphate buffer (pH 7.4) containing 2 mM EDTA, 0.1 mM dithiothreitol and 0.1 ~ v/v Triton X-100, to give a 10~ w/v homogenate. The homogenates were dialysed for 5 h against several changes of the phosphate buffer containing EDTA and dithiothreitol, to remove endogenous glycine and serine. No loss of activity was observed after 16 h of dialysis. Each brain was divided into the following regions: cerebral cortex, midbrain (including the diencephalon and mesencephalon), cerebellum, and pons-medulla. The spinal cords were freed from spinal roots and the dura removed before homogenisation. Regional distribution of serine hydroxymethyltransferase activity in extracts of cat spinal cord Lumbosacral spinal cord was removed from cats under pentobarbitone anaes-
SERINEHYDROXYMETHYLTRANSFERASEIN CNS
151
thesia and separated into various regions as described previously12. The tissue samples were weighed and extracted as described above for the samples of rat nervous tissue. Subcellular distribution of serine hydroxymethyltransferase activity in homogenates of rat cerebral cortex and spinal cord Tissue was removed as described above and homogenised in 0.32 M sucrose at a concentration of 10 ~o w/v using a stainless steel mortar and Perspex pestle (rotating at 870 rev./min, clearance 0.25 mm difference in diameter). Subcellular fractions were prepared by differential and density gradient centrifugation as described by Johnston et al. 13, with the exception that the A1 myelin layer and the upper $2 layer were not separated prior to the next centrifugation. Pellets were resuspended in phosphate buffer containing EDTA and dithiothreitol and dialysed overnight against several changes of this buffer. The soluble fraction was concentrated prior to assay in an ultrafiltration apparatus (Amicon, UM-10 Diaflo ultrafilters). Inhibition Potential inhibitors of the hydroxymethyltransferase activity in extracts of rat spinal cord were tested at a final concentration of 0.5 mM, using the same concentration of reactants as above. For each inhibitor, triplicate determinations were made of the amount of formaldehyde formed after 40 min incubation. Materials 5-Azaindole was the gift of Prof. A. Albert (Canberra), and O-glycyl-L-serine the gift of Dr. J. R. Dice (Parke Davis, Detroit). Tetrahydrofolate was prepared from folic acid by catalytic hydrogenation as described by Hatefi et al. 11. Hydrazinoacetic acid and DL-serine-O-sulphate were prepared by published procedures 4,8. L-[3-14C]Serine, specific activity 48 mCi/mmole, was purchased from the Radiochemical Centre, Amersham. Other chemicals were purchased as follows: aminooxyacetic acid, 3-mercaptopropionic acid and thiocarbohydrazide (Eastman); DL-Callylglycine, D-cycloserine and pyridoxal phosphate (Sigma); 5-aminotetrazole, dimedone, kojic acid, and L-serine (Fluka); thiosemicarbazide and L-threonine (B.D.H.); glycine and sucrose (Mann); aminopterin and DL-2-methylserine (Calbiochem); Nacetylglycine (Light's); DL-3-phenylserine (Nutritional Biochemical Corp.) and strychnine (T. and M. Smith). RESULTS
The regional distribution of serine hydroxymethyltransferase activity in the extracts of rat central nervous tissue is presented in Table I, together with the concentrations of glycine and serine in the various regions as determined by Shank and Aprison is. No significant differences in enzymic activity were observed between the 5 regions examined, whereas the concentrations of both glycine and serine show appreciable variation. The distribution of serine hydroxymethyltransferase activity in the 4 regions of
152
L , P . D A V I E S A N D G . A. R . J O H N S T O N
TABLE I REGIONAL DISTRIBUTION OF SERINE HYDROXYMETI-IYLTRANSFERASE ACTIVITY IN EXTRACTS OF ADULT RAT CENTRAL NERVOUS TISSUE
Serine hydroxymethyltransferase activity (/~moles formaldehyde produced/h/g wet tissue or /ms protein) is expressed as the means 4- S.E.M. of values from 4 experiments. The glycine and serine concentrations (#moles/g wet tissue) are those of Shank and Aprison16. Region
Serine hydroxymethyltransferase activity /g wet tissue
~ragprotein
Cerebral cortex Cerebellum Pons-medulla
2.36 4- 0.19 2.42 4- 0.18 2.64 4- 0.32
0.0374- 0.004 0.040 4- 0.005 0.0394- 0.006
Diencephalon-mesencephalon
2.32 ± 0.37
0.0344- 0.005
Spinal cord
2.53 i 0.23
0.043i 0.004
Glycine concentration
Serine concentration
0.67 4- 0.02 0.63 4- 0.03 2.93 ± 0.15 4.02 4- 0.11 0.93 ± 0.02 1.63 4- 0.04 4.36*
1.03 4- 0.02 0.60 4- 0.02 0.40 4- 0.01 0.44 ± 0.01 0.61 4- 0.01 0.42 _-%0.01 0.46*
* Average of values for grey and white matter.
TABLE II REGIONAL DISTRIBUTION OF SERINE HYDROXYMETHYLTRANSFERASE
ACTIVITY 1N EXTRACTS
OF CAT
SPINAL CORD
Hydroxymethyltransferase activities (#moles formaldehyde produced/h/g wet tissue o r / m s protein) are expressed as the means 4- S.E. of values from each of 3 cats. The concentrations of glycine and serine (#moles/g wet tissue) are those reported previouslylL Region
Serine hydroxymethyltransferase activity Glycine concentration /g wet tissue /mg protein
Serine concentration
Dorsal white Dorsal grey Ventral grey Ventral white
2.46 4.29 4.97 2.47
0.58 0.92 0.92 0.61
4- 0.06 4- 0.20 4- 0.07 4- 0.12
0.066 ± 0.002 0.095 ± 0.004 0.113 4- 0.003 0.066 4- 0.002
2.92 4- 0.10 5.42 4- 0.14 6.51 ± 0.17 4.55 4- 0.09
4- 0.16 ± 0.26 i 0.27 4- 0.27
cat spinal cord is given in Table | I , together with the d i s t r i b u t i o n of glycine a n d serine as d e t e r m i n e d by J o h n s t o n l L The values for activity in the extracts of grey matter are almost double those in the extracts of white matter. Activity in the ventral grey matter extracts is significantly higher t h a n in the dorsal grey matter extracts at P < 0.05 based o n wet weight of tissue, a n d at P < 0.025 based o n protein, using Student's t-test. The subcellular d i s t r i b u t i o n of serine hydroxymethyltransferase activity in h o m o g e n a t e s of rat cerebral cortex a n d spinal cord was very similar (Table III). Activity was concentrated i n the mitochondrial, soluble a n d nuclear fractions. The activity observed in the nuclear fraction m a y be due to c o n t a m i n a t i o n with m i t o c h o n d r i a a n d
SERINE HYDROXYMETHYLTRANSFERASE IN C N S
153
TABLE III SUBCELLULAR DISTRIBUTION OF SERINE HYDROXYMETHYLTRANSFERASEACTIVITY IN HOMOGENATES OF ADULT RAT CEREBRALCORTEXAND SPINALCORD Results (given in #moles formaldehyde produced/h/g wet tissue) are means 5: S.E. of values from each of 3 rats. Values in brackets refer to the percentage of the total recovered activity.
Subcellularfraction
Nuclear pellet Myelin layer Synaptosomes, vesicles, etc. Mitochondrial pellet Soluble fraction Microsomal pellet
Serine hydroxymethyltransferase activity Cerebral cortex
Spinal cord
0.18 4- 0.07 (23)
0.20 0.02 0.03 0.30 0.12 0.02
0.04 + 0.02 (5) 0.05 0.33 0.16 0.04
4444-
0.01 (6) 0.13 (41) 0.01 (20)
0.02 (5)
4- 0.09 4- 0.01 i 0.02 4- 0.11 -4- 0.04 4- 0.01
(29) (3) (4) (44) (18) (2)
TABLE IV INHIBITORS OF SERINEHYDROXYMETHYLTRANSFERASEACTIVITYIN EXTRACTSOF ADULTRAT SPINALCORD Values are means 4- S.E.M. of 3 determinations.
Inhibitor (0.5 mM)
% Inhibition
Hydrazinoacetic acid Thiocarbohydrazide Aminooxyacetic acid D-Cycloserine Thiosemicarbazide O-Glycyl-L-serine Aminopterin N-Acetylglycine DL-Serine-O-sulphate Kojic acid 3-Mercaptopropionic acid Threonine DL-2-Methylserine Nikethamide
94 93 83 76 49 37 30 22 21 20 17 15 13 5
4- 8 i 4 ± 1 ± 2 4- 2 i 2 + 4 ± 2 ± 3 4- 2 4- 2 4- 1 4- 1 4- 0.5
No significant inhibition: DL-C-allylglycine, 5-aminotetrazole, 5-azaindole, DL-3-phenylserine, strychnine
unbroken cells and thus no definite conclusion can be drawn as to the occurrence of the enzyme in or associated with the nuclei. Under the experimental conditions used, 14 substances significantly inhibited the serine hydroxymethyltransferase activity in extracts of rat spinal cord (Table IV), the most effective being hydrazinoacetic acid, thiocarbohydrazide, aminooxyacetic acid and cycloserine. DISCUSSION
No.significant differences were found in the serine hydroxymethyltransferase
154
L.P.
DAVIES A N D G . A. R. J O H N S T O N
activity in extracts of various regions of the rat central nervous system. As there are regional differences in serine and glycine content, it seems likely that serine hydroxymethyltransferase is not rate determining with respect to either glycine or serine metabolism. The rate controlling step for the formation of glycine from glucose may be the conversion of D-glycerate to hydroxypyruvate, catalysed by D-glycerate dehydrogenase, since this enzyme is inhibited by glycine in a non-competitive manner suggestive of end-product inhibition19 and its activity correlates with the regional distribution of glycine 2°. Serine hydroxymethyltransferase, like other amino acid metabolising enzymes including D-amino acid oxidasev, glutamine synthetase, glutaminase, GABA-2-oxoglutarate aminotransferase, aspartate-2-oxoglutarate aminotransferase, glutamate decarboxylase, glutamate dehydrogenase10, and glycine-2-oxoglutarate aminotransferase la, was found to be more concentrated in the grey matter than in the white matter of cat spinal cord. Serine hydroxymethyltransferase and glutamate decarboxylase1° are the only ones to show dorsoventral differences within the spinal grey matter. Serine hydroxymethyltransferase and glycine are more concentrated in the ventral grey matter than in the dorsal grey matter, while the reverse is true for glutamate decarboxylase, glutamate and GABA. Dorsoventral differences in enzymic activity were not apparent in the white matter, whereas glycine is more concentrated in the ventral than in the dorsal regions of both the grey and white matter 2,12. Under the assay conditions employed in this and an earlier study13, the rate of glycine production from serine by the hydroxymethyltransferase activity in the cat spinal cord extracts was approximately one-half the rate of glycine production from glyoxylate by the glycine-2-oxoglutarate aminotransferase activity in these extracts. Shank and Aprison 16 have commented that this aminotransferase activity is 'very weak in comparison' to the activity of other amino acid metabolising enzymes. The activity of both these glycine metabolising enzymes may be comparatively weak, but, in as much as enzymic activities measured in vitro can be related to those in vivo, they appear to be more than adequate to account for the estimated rates of glycine and serine formation from glucose in vivo 16. The subcellular distribution of serine hydroxymethyltransferase activity was at least bimodal in the homogenates of rat cerebral cortex and spinal cord, being concentrated in the mitochondrial and soluble fractions. A similar subcellular distribution has been found in homogenates of rat liver 15. No differences were observed in the subcellular distribution of hydroxymethyltransferase activity in the cerebral cortex, where glycine is unlikely to be a synaptic transmitter, and in the spinal cord, where a transmitter role is highly probable. These findings provide no evidence to suggest that this enzyme is specifically associated with synaptosomes produced on homogenisation of nerve endings which release glycine as a transmitter in the spinal cord. Most of the substances found to inhibit serine hydroxymethyltransferase activity in extracts of rat spinal cord are known to inhibit other pyridoxal-dependent enzymes (hydrazinoacetic acid, thiosemicarbazide, thiocarbohydrazide, aminooxyacetic acid, D-cycloserine, 3-mercaptopropionic acid) and/or are structural analogues of glycine and serine (hydrazinoacetic acid, aminooxyacetic acid, o-cycloserine, O-glycyl-
SERINE HYDROXYMETHYLTRANSFERASE IN C N S
155
L-serine, N-acetylglycine, DL-serine-O-sulphate, L-threonine, DL-2-methylserine). The inhibitors include certain convulsants (thiocarbohydrazide, thiosemicarbazide, kojic acid, 3-mercaptopropionic acid), whereas other convulsants (nL-C-allylglycine, 5azaindole and strychnine) had no effect on hydroxymethyltransferase activity. Glycine levels in rat cerebellum are known to be reduced during convulsions induced by 3mercaptopropionic acid 0: inhibition of serine hydroxymethyltransferase activity may contribute to this reduction in glycine level. ACKNOWLEDGEMENTS
We are grateful to Dr. M. L. Uhr for valuable advice and to Miss H. M. Alexander for some initial studies on serine hydroxymethyltransferase activity in CNS extracts.
REFERENCES 1 APRISON, M. H., DAVIDOFF,R. A., AND WERMAN, R., Glycine: its metabolic and possible transmitter roles in nervous tissue. In A. LAJTHA(Ed.), Handbook ofNeurochemistry, Vol. 3, Metabolic Reactions in the Nervous System, Plenum Press, New York, 1970, pp. 381-397. 2 APRISON, M. H., AND WERMAN, R., The distribution of glycine in cat spinal cord and roots, Life Sci., 4 (1965) 2075-2083. 3 BRODERICK,D. S., CANDLAND,K. L., NORTH, J. A., AND MANGUM,J. H., The isolation of serine transhydroxymethylase from bovine brain, Arch. Biochem., 148 (1972) 196-198. 4 CARMI,A., POLLAK,G., AND YELLIN, H., ct-Hydrazino acids. I. a-Hydrazino aliphatic acids and a-(1-methylhydrazino) aliphatic acids, J. org. Chem., 25 (1960) 44 46. 5 CURTIS,D. R., AND JOHNSTON,G. A. R., Amino acid transmitters. In A. LmTI-IA(Ed.), Handbook of Neurochemistry, VoL 4, Central Mechanisms in the Nervous System, Plenum Press, New York, 1970, pp. 115-134. 6 DE LORESARNAIZ, G. R., ALBERICI, i . , AND DE ROBERTIS,E., Alteration of GABA system and Purkinje cells in the rat cerebellum by the convulsant 3-mercaptopropionic acid, J. Neurochem., 19 (1972) 1379-1385. 7 DE MARCHI, W. J., AND JOHNSTON,G. A. R., The oxidation of glycine by D-amino acid oxidase in extracts of mammalian central nervous tissue, J. Neurochem., 16 (1969) 355-361. 8 DODGSON, K. S., LLOYD, A. G., AND TUDBALL,N.,O-Sulphate esters of L-serine, L-threonine and L-hydroxyproline, Biochem. J., 79 (1961) 111-117. 9 GORNALL, A. G., BARDAWILL,C. J., AND DAVID, i . M., Determination of serum proteins by means oftbe biuret reaction, J. biol. Chem., 177 (1949) 751-766. 10 GRAHAM, JR., L. T., AND APRISON, i . H., Distribution of some enzymes associated with the metabolism of glutamate, aspartate, 7-aminobutyrate and glutamine in cat spinal cord, J. Neurochem., 16 (1969) 559-566. 11 HATEFI, Y., TALBERT, P. T., OSBORN, M. J., AND HUENNEKENS,F. M., Tetrahydrofolic acid. In H. A. LARDY(Ed.), BiochemicalPreparations, Vol. 7, Wiley, New York, 1960, pp. 89-92. 12 JOHNSTON,G. A. R., The intraspinal distribution of some depressant amino acids, J. Neurochem., 15 (1968) 1013-1017. 13 JOHNSTON,G. A. R., VITALI, i . V., AND ALEXANDER,H. i . , Regional and subcellular distribution studies on glycine-2-oxoglutarate transaminase activity in cat spinal cord, Brain Research, 20 (1970) 361-367. 14 LOWRY, O. H., ROSENBROUGH,N. H., FARR, A. L., AND RANDALL, R. J., Protein measurement with the Folin phenol reagent, J. biol. Chem., 193 (1951) 265-275. 15 NAKANO,Y., FUJIOKA, M., AND WADA, H., Studies on serine hydroxymethylase isoenzymes from rat liver, Biochim. biophys. ,4cta (,4rest.), 159 (1968) 19-26. 16 SHANK,R. P., AND APRISON, i . H., The metabolism in vivo of glycine and serine in eight areas of the rat central nervous system, J. Neurochem., 17 (1970) 1461-1475.
156
L.P. DAVIESAND G. A. R. JOHNSTON
17 SHAW,R. K., ANDHEINE,J. D., Ninhydrin positive substances present in different areas of normal rat brain, J. Neurochem., 12 (1965) 151-155. 18 TAYLOR,R. T., AND WEISSBACH,H., Radioactive assay for serine transhydroxymethylase, Analyt. Bioehem., 13 (1965) 80-84. 19 UHR, M. L., AND SNEDDON,M. K., Glycine and serine inhibition of D-glycerate dehydrogenase and 3-phosphoglycerate dehydrogenase of rat brain, FEBSLetters, 17 (1971) 137-140. 20 UHR, M. L., AND SNEOOON,M. K., The regional distribution of o-glycerate dehydrogenase and 3-phosphoglycerate dehydrogenase in the cat central nervous system: correlation with glycine levels, J. Neurochem., 19 (1972) 1495-1500.
149
© ElsevierScientificPublishingCompany,Amsterdam- Printedin The Netherlands
SERINE HYDROXYMETHYLTRANSFERASE IN THE CENTRAL NERVOUS SYSTEM: REGIONAL AND SUBCELLULAR DISTRIBUTION STUDIES
L. P. DAVIES ANDG. A. R. JOHNSTON Department of Physiology, Australian National University, Canberra (Australia)
(AcceptedOctober20th, 1972)
SUMMARY
No significant differences were found in the serine hydroxymethyltransferase activity of extracts from different regions of the adult rat central nervous system. Within the cat spinal cord, some degree of correlation between glycine levels and hydroxymethyltransferase activity was apparent, with higher levels of both in ventral as compared with dorsal grey matter. Subcellular distribution studies in the rat cerebral cortex and spinal cord suggested that this enzymic activity is present in both the cytoplasm and the mitochondria. A number of centrally-active drugs and structural analogues of glycine and serine were found to inhibit hydroxymethyltransferase activity, the most effective being those which are known to inhibit other pyridoxal-dependent enzymes.
INTRODUCTION
The evidence that glycine is a major inhibitory transmitter in certain areas of the mammalian central nervous system, including the spinal cord and medulla oblongata, has been well documentedl, 5. The major pathways whereby glycine is metabolised in nervous tissue are, however, still unknown. The enzyme serine hydroxymethyltransferase (L-serine-tetrahydrofolate 5,10hydroxymethyltransferase, E.C. 2.1.2.1), which catalyses the interconversion of glycine and serine, has been recently purified from bovine brain 3 and there is some evidence to suggest that this enzyme may play an important role in glycine metabolism16. The observation2,16,17 that glycine levels vary from one CNS region to another suggested that the regional distribution of serine hydroxymethyltransferaseactivity might reveal important information about the possible role of this activity in glycine metab-
I 50
L. P. DAVIES A N D G. A. R. JOHNSTON
olism. We have studied the regional and subcellular distribution of this enzymic activity in extracts of rat and cat central nervous tissue, and investigated some structural analogues of glycine and serine, and various centrally-active drugs, as potential inhibitors of serine hydroxymethyltransferaseactivity. METHODS
Assay of hydroxymethyltransferase activity The rate of production of radioactive formaldehyde from radioactive serine was assayed by the method of Taylor and Weissbach 18. The reaction mixtures contained 0.5 mM [3-14C]serine, 0.5 mM pyridoxal phosphate, 2 mM tetrahydrofolate and 4 mM dithiothreitol in 100 mM potassium phosphate buffer (pH 7.4) in a final volume of 2 ml. Reactions carried out at 37 °C in a shaking water bath were initiated by the addition of 0.5 ml of extract. At various times, 0.4 ml aliquots were removed, and reaction terminated by addition to a mixture of 0.3 ml of 1 M sodium acetate solution (pH 4.5) and 0.2 ml of 0.1 M formaldehyde. The [laC]formaldehyde, formed during the reaction, was trapped with dimedone, extracted into toluene and counted in a toluene-based scintillator in a Beckman LSI00 counter, correcting for any quenching by means of an external standard. The enzymic activities measured in this way were linear for at least 40 rain. Each assay provided 4 time aliquots and the rate of production of formaldehyde was caIculated from the slope of the line of best fit through a plot of radioactivity versus time, by a least squares analysis using a PDP-8L computer. Protein concentrations in the tissue extracts were measured by the biuret method 9, or the method of Lowry et al. 14, removing lipid by ether extraction immediately prior to reading in the spectrophotometer. Regional distribution of serine hydroxymethyltransferase activity in extracts of rat central nervous tissue Adult Wistar rats (150-200 g) were killed by decapitation, the brains and spinal cords rapidly removed and chilled, and the various regions quickly separated. The tissue samples were then weighed and homogenised in 0.1 M potassium phosphate buffer (pH 7.4) containing 2 mM EDTA, 0.1 mM dithiothreitol and 0.1 ~ v/v Triton X-100, to give a 10~ w/v homogenate. The homogenates were dialysed for 5 h against several changes of the phosphate buffer containing EDTA and dithiothreitol, to remove endogenous glycine and serine. No loss of activity was observed after 16 h of dialysis. Each brain was divided into the following regions: cerebral cortex, midbrain (including the diencephalon and mesencephalon), cerebellum, and pons-medulla. The spinal cords were freed from spinal roots and the dura removed before homogenisation. Regional distribution of serine hydroxymethyltransferase activity in extracts of cat spinal cord Lumbosacral spinal cord was removed from cats under pentobarbitone anaes-
SERINEHYDROXYMETHYLTRANSFERASEIN CNS
151
thesia and separated into various regions as described previously12. The tissue samples were weighed and extracted as described above for the samples of rat nervous tissue. Subcellular distribution of serine hydroxymethyltransferase activity in homogenates of rat cerebral cortex and spinal cord Tissue was removed as described above and homogenised in 0.32 M sucrose at a concentration of 10 ~o w/v using a stainless steel mortar and Perspex pestle (rotating at 870 rev./min, clearance 0.25 mm difference in diameter). Subcellular fractions were prepared by differential and density gradient centrifugation as described by Johnston et al. 13, with the exception that the A1 myelin layer and the upper $2 layer were not separated prior to the next centrifugation. Pellets were resuspended in phosphate buffer containing EDTA and dithiothreitol and dialysed overnight against several changes of this buffer. The soluble fraction was concentrated prior to assay in an ultrafiltration apparatus (Amicon, UM-10 Diaflo ultrafilters). Inhibition Potential inhibitors of the hydroxymethyltransferase activity in extracts of rat spinal cord were tested at a final concentration of 0.5 mM, using the same concentration of reactants as above. For each inhibitor, triplicate determinations were made of the amount of formaldehyde formed after 40 min incubation. Materials 5-Azaindole was the gift of Prof. A. Albert (Canberra), and O-glycyl-L-serine the gift of Dr. J. R. Dice (Parke Davis, Detroit). Tetrahydrofolate was prepared from folic acid by catalytic hydrogenation as described by Hatefi et al. 11. Hydrazinoacetic acid and DL-serine-O-sulphate were prepared by published procedures 4,8. L-[3-14C]Serine, specific activity 48 mCi/mmole, was purchased from the Radiochemical Centre, Amersham. Other chemicals were purchased as follows: aminooxyacetic acid, 3-mercaptopropionic acid and thiocarbohydrazide (Eastman); DL-Callylglycine, D-cycloserine and pyridoxal phosphate (Sigma); 5-aminotetrazole, dimedone, kojic acid, and L-serine (Fluka); thiosemicarbazide and L-threonine (B.D.H.); glycine and sucrose (Mann); aminopterin and DL-2-methylserine (Calbiochem); Nacetylglycine (Light's); DL-3-phenylserine (Nutritional Biochemical Corp.) and strychnine (T. and M. Smith). RESULTS
The regional distribution of serine hydroxymethyltransferase activity in the extracts of rat central nervous tissue is presented in Table I, together with the concentrations of glycine and serine in the various regions as determined by Shank and Aprison is. No significant differences in enzymic activity were observed between the 5 regions examined, whereas the concentrations of both glycine and serine show appreciable variation. The distribution of serine hydroxymethyltransferase activity in the 4 regions of
152
L , P . D A V I E S A N D G . A. R . J O H N S T O N
TABLE I REGIONAL DISTRIBUTION OF SERINE HYDROXYMETI-IYLTRANSFERASE ACTIVITY IN EXTRACTS OF ADULT RAT CENTRAL NERVOUS TISSUE
Serine hydroxymethyltransferase activity (/~moles formaldehyde produced/h/g wet tissue or /ms protein) is expressed as the means 4- S.E.M. of values from 4 experiments. The glycine and serine concentrations (#moles/g wet tissue) are those of Shank and Aprison16. Region
Serine hydroxymethyltransferase activity /g wet tissue
~ragprotein
Cerebral cortex Cerebellum Pons-medulla
2.36 4- 0.19 2.42 4- 0.18 2.64 4- 0.32
0.0374- 0.004 0.040 4- 0.005 0.0394- 0.006
Diencephalon-mesencephalon
2.32 ± 0.37
0.0344- 0.005
Spinal cord
2.53 i 0.23
0.043i 0.004
Glycine concentration
Serine concentration
0.67 4- 0.02 0.63 4- 0.03 2.93 ± 0.15 4.02 4- 0.11 0.93 ± 0.02 1.63 4- 0.04 4.36*
1.03 4- 0.02 0.60 4- 0.02 0.40 4- 0.01 0.44 ± 0.01 0.61 4- 0.01 0.42 _-%0.01 0.46*
* Average of values for grey and white matter.
TABLE II REGIONAL DISTRIBUTION OF SERINE HYDROXYMETHYLTRANSFERASE
ACTIVITY 1N EXTRACTS
OF CAT
SPINAL CORD
Hydroxymethyltransferase activities (#moles formaldehyde produced/h/g wet tissue o r / m s protein) are expressed as the means 4- S.E. of values from each of 3 cats. The concentrations of glycine and serine (#moles/g wet tissue) are those reported previouslylL Region
Serine hydroxymethyltransferase activity Glycine concentration /g wet tissue /mg protein
Serine concentration
Dorsal white Dorsal grey Ventral grey Ventral white
2.46 4.29 4.97 2.47
0.58 0.92 0.92 0.61
4- 0.06 4- 0.20 4- 0.07 4- 0.12
0.066 ± 0.002 0.095 ± 0.004 0.113 4- 0.003 0.066 4- 0.002
2.92 4- 0.10 5.42 4- 0.14 6.51 ± 0.17 4.55 4- 0.09
4- 0.16 ± 0.26 i 0.27 4- 0.27
cat spinal cord is given in Table | I , together with the d i s t r i b u t i o n of glycine a n d serine as d e t e r m i n e d by J o h n s t o n l L The values for activity in the extracts of grey matter are almost double those in the extracts of white matter. Activity in the ventral grey matter extracts is significantly higher t h a n in the dorsal grey matter extracts at P < 0.05 based o n wet weight of tissue, a n d at P < 0.025 based o n protein, using Student's t-test. The subcellular d i s t r i b u t i o n of serine hydroxymethyltransferase activity in h o m o g e n a t e s of rat cerebral cortex a n d spinal cord was very similar (Table III). Activity was concentrated i n the mitochondrial, soluble a n d nuclear fractions. The activity observed in the nuclear fraction m a y be due to c o n t a m i n a t i o n with m i t o c h o n d r i a a n d
SERINE HYDROXYMETHYLTRANSFERASE IN C N S
153
TABLE III SUBCELLULAR DISTRIBUTION OF SERINE HYDROXYMETHYLTRANSFERASEACTIVITY IN HOMOGENATES OF ADULT RAT CEREBRALCORTEXAND SPINALCORD Results (given in #moles formaldehyde produced/h/g wet tissue) are means 5: S.E. of values from each of 3 rats. Values in brackets refer to the percentage of the total recovered activity.
Subcellularfraction
Nuclear pellet Myelin layer Synaptosomes, vesicles, etc. Mitochondrial pellet Soluble fraction Microsomal pellet
Serine hydroxymethyltransferase activity Cerebral cortex
Spinal cord
0.18 4- 0.07 (23)
0.20 0.02 0.03 0.30 0.12 0.02
0.04 + 0.02 (5) 0.05 0.33 0.16 0.04
4444-
0.01 (6) 0.13 (41) 0.01 (20)
0.02 (5)
4- 0.09 4- 0.01 i 0.02 4- 0.11 -4- 0.04 4- 0.01
(29) (3) (4) (44) (18) (2)
TABLE IV INHIBITORS OF SERINEHYDROXYMETHYLTRANSFERASEACTIVITYIN EXTRACTSOF ADULTRAT SPINALCORD Values are means 4- S.E.M. of 3 determinations.
Inhibitor (0.5 mM)
% Inhibition
Hydrazinoacetic acid Thiocarbohydrazide Aminooxyacetic acid D-Cycloserine Thiosemicarbazide O-Glycyl-L-serine Aminopterin N-Acetylglycine DL-Serine-O-sulphate Kojic acid 3-Mercaptopropionic acid Threonine DL-2-Methylserine Nikethamide
94 93 83 76 49 37 30 22 21 20 17 15 13 5
4- 8 i 4 ± 1 ± 2 4- 2 i 2 + 4 ± 2 ± 3 4- 2 4- 2 4- 1 4- 1 4- 0.5
No significant inhibition: DL-C-allylglycine, 5-aminotetrazole, 5-azaindole, DL-3-phenylserine, strychnine
unbroken cells and thus no definite conclusion can be drawn as to the occurrence of the enzyme in or associated with the nuclei. Under the experimental conditions used, 14 substances significantly inhibited the serine hydroxymethyltransferase activity in extracts of rat spinal cord (Table IV), the most effective being hydrazinoacetic acid, thiocarbohydrazide, aminooxyacetic acid and cycloserine. DISCUSSION
No.significant differences were found in the serine hydroxymethyltransferase
154
L.P.
DAVIES A N D G . A. R. J O H N S T O N
activity in extracts of various regions of the rat central nervous system. As there are regional differences in serine and glycine content, it seems likely that serine hydroxymethyltransferase is not rate determining with respect to either glycine or serine metabolism. The rate controlling step for the formation of glycine from glucose may be the conversion of D-glycerate to hydroxypyruvate, catalysed by D-glycerate dehydrogenase, since this enzyme is inhibited by glycine in a non-competitive manner suggestive of end-product inhibition19 and its activity correlates with the regional distribution of glycine 2°. Serine hydroxymethyltransferase, like other amino acid metabolising enzymes including D-amino acid oxidasev, glutamine synthetase, glutaminase, GABA-2-oxoglutarate aminotransferase, aspartate-2-oxoglutarate aminotransferase, glutamate decarboxylase, glutamate dehydrogenase10, and glycine-2-oxoglutarate aminotransferase la, was found to be more concentrated in the grey matter than in the white matter of cat spinal cord. Serine hydroxymethyltransferase and glutamate decarboxylase1° are the only ones to show dorsoventral differences within the spinal grey matter. Serine hydroxymethyltransferase and glycine are more concentrated in the ventral grey matter than in the dorsal grey matter, while the reverse is true for glutamate decarboxylase, glutamate and GABA. Dorsoventral differences in enzymic activity were not apparent in the white matter, whereas glycine is more concentrated in the ventral than in the dorsal regions of both the grey and white matter 2,12. Under the assay conditions employed in this and an earlier study13, the rate of glycine production from serine by the hydroxymethyltransferase activity in the cat spinal cord extracts was approximately one-half the rate of glycine production from glyoxylate by the glycine-2-oxoglutarate aminotransferase activity in these extracts. Shank and Aprison 16 have commented that this aminotransferase activity is 'very weak in comparison' to the activity of other amino acid metabolising enzymes. The activity of both these glycine metabolising enzymes may be comparatively weak, but, in as much as enzymic activities measured in vitro can be related to those in vivo, they appear to be more than adequate to account for the estimated rates of glycine and serine formation from glucose in vivo 16. The subcellular distribution of serine hydroxymethyltransferase activity was at least bimodal in the homogenates of rat cerebral cortex and spinal cord, being concentrated in the mitochondrial and soluble fractions. A similar subcellular distribution has been found in homogenates of rat liver 15. No differences were observed in the subcellular distribution of hydroxymethyltransferase activity in the cerebral cortex, where glycine is unlikely to be a synaptic transmitter, and in the spinal cord, where a transmitter role is highly probable. These findings provide no evidence to suggest that this enzyme is specifically associated with synaptosomes produced on homogenisation of nerve endings which release glycine as a transmitter in the spinal cord. Most of the substances found to inhibit serine hydroxymethyltransferase activity in extracts of rat spinal cord are known to inhibit other pyridoxal-dependent enzymes (hydrazinoacetic acid, thiosemicarbazide, thiocarbohydrazide, aminooxyacetic acid, D-cycloserine, 3-mercaptopropionic acid) and/or are structural analogues of glycine and serine (hydrazinoacetic acid, aminooxyacetic acid, o-cycloserine, O-glycyl-
SERINE HYDROXYMETHYLTRANSFERASE IN C N S
155
L-serine, N-acetylglycine, DL-serine-O-sulphate, L-threonine, DL-2-methylserine). The inhibitors include certain convulsants (thiocarbohydrazide, thiosemicarbazide, kojic acid, 3-mercaptopropionic acid), whereas other convulsants (nL-C-allylglycine, 5azaindole and strychnine) had no effect on hydroxymethyltransferase activity. Glycine levels in rat cerebellum are known to be reduced during convulsions induced by 3mercaptopropionic acid 0: inhibition of serine hydroxymethyltransferase activity may contribute to this reduction in glycine level. ACKNOWLEDGEMENTS
We are grateful to Dr. M. L. Uhr for valuable advice and to Miss H. M. Alexander for some initial studies on serine hydroxymethyltransferase activity in CNS extracts.
REFERENCES 1 APRISON, M. H., DAVIDOFF,R. A., AND WERMAN, R., Glycine: its metabolic and possible transmitter roles in nervous tissue. In A. LAJTHA(Ed.), Handbook ofNeurochemistry, Vol. 3, Metabolic Reactions in the Nervous System, Plenum Press, New York, 1970, pp. 381-397. 2 APRISON, M. H., AND WERMAN, R., The distribution of glycine in cat spinal cord and roots, Life Sci., 4 (1965) 2075-2083. 3 BRODERICK,D. S., CANDLAND,K. L., NORTH, J. A., AND MANGUM,J. H., The isolation of serine transhydroxymethylase from bovine brain, Arch. Biochem., 148 (1972) 196-198. 4 CARMI,A., POLLAK,G., AND YELLIN, H., ct-Hydrazino acids. I. a-Hydrazino aliphatic acids and a-(1-methylhydrazino) aliphatic acids, J. org. Chem., 25 (1960) 44 46. 5 CURTIS,D. R., AND JOHNSTON,G. A. R., Amino acid transmitters. In A. LmTI-IA(Ed.), Handbook of Neurochemistry, VoL 4, Central Mechanisms in the Nervous System, Plenum Press, New York, 1970, pp. 115-134. 6 DE LORESARNAIZ, G. R., ALBERICI, i . , AND DE ROBERTIS,E., Alteration of GABA system and Purkinje cells in the rat cerebellum by the convulsant 3-mercaptopropionic acid, J. Neurochem., 19 (1972) 1379-1385. 7 DE MARCHI, W. J., AND JOHNSTON,G. A. R., The oxidation of glycine by D-amino acid oxidase in extracts of mammalian central nervous tissue, J. Neurochem., 16 (1969) 355-361. 8 DODGSON, K. S., LLOYD, A. G., AND TUDBALL,N.,O-Sulphate esters of L-serine, L-threonine and L-hydroxyproline, Biochem. J., 79 (1961) 111-117. 9 GORNALL, A. G., BARDAWILL,C. J., AND DAVID, i . M., Determination of serum proteins by means oftbe biuret reaction, J. biol. Chem., 177 (1949) 751-766. 10 GRAHAM, JR., L. T., AND APRISON, i . H., Distribution of some enzymes associated with the metabolism of glutamate, aspartate, 7-aminobutyrate and glutamine in cat spinal cord, J. Neurochem., 16 (1969) 559-566. 11 HATEFI, Y., TALBERT, P. T., OSBORN, M. J., AND HUENNEKENS,F. M., Tetrahydrofolic acid. In H. A. LARDY(Ed.), BiochemicalPreparations, Vol. 7, Wiley, New York, 1960, pp. 89-92. 12 JOHNSTON,G. A. R., The intraspinal distribution of some depressant amino acids, J. Neurochem., 15 (1968) 1013-1017. 13 JOHNSTON,G. A. R., VITALI, i . V., AND ALEXANDER,H. i . , Regional and subcellular distribution studies on glycine-2-oxoglutarate transaminase activity in cat spinal cord, Brain Research, 20 (1970) 361-367. 14 LOWRY, O. H., ROSENBROUGH,N. H., FARR, A. L., AND RANDALL, R. J., Protein measurement with the Folin phenol reagent, J. biol. Chem., 193 (1951) 265-275. 15 NAKANO,Y., FUJIOKA, M., AND WADA, H., Studies on serine hydroxymethylase isoenzymes from rat liver, Biochim. biophys. ,4cta (,4rest.), 159 (1968) 19-26. 16 SHANK,R. P., AND APRISON, i . H., The metabolism in vivo of glycine and serine in eight areas of the rat central nervous system, J. Neurochem., 17 (1970) 1461-1475.
156
L.P. DAVIESAND G. A. R. JOHNSTON
17 SHAW,R. K., ANDHEINE,J. D., Ninhydrin positive substances present in different areas of normal rat brain, J. Neurochem., 12 (1965) 151-155. 18 TAYLOR,R. T., AND WEISSBACH,H., Radioactive assay for serine transhydroxymethylase, Analyt. Bioehem., 13 (1965) 80-84. 19 UHR, M. L., AND SNEDDON,M. K., Glycine and serine inhibition of D-glycerate dehydrogenase and 3-phosphoglycerate dehydrogenase of rat brain, FEBSLetters, 17 (1971) 137-140. 20 UHR, M. L., AND SNEOOON,M. K., The regional distribution of o-glycerate dehydrogenase and 3-phosphoglycerate dehydrogenase in the cat central nervous system: correlation with glycine levels, J. Neurochem., 19 (1972) 1495-1500.