125
Biochem. J. (1992) 287, 125-129 (Printed in Great Britain)
Substrate specificity of the purine-2'-deoxyribonucleosidase of Crithidia luciliae Daniel J. STEENKAMP and Torsten J. F. HALBICH Department of Chemical Pathology, University of Cape Town Medical School, Observatory 7925, Cape Town, South Africa
The purine-2'-deoxyribonucleosidase of Crithidia luciliae catalyses an efficient deoxyribosyl transfer between a variety of purine bases, benzimidazole and 5,6-dimethylbenzimidazole. Since the deoxyriboside of a deoxyribosyl acceptor is necessarily also a substrate, the trans-N-deoxyribosylase activity of the enzyme allows a study of its specificity to be extended to a large number of purines and purine analogues. Amongst 27 different deoxyribosyl acceptors, only hypoxanthine gave rise to isomeric products. The introduction of methyl groups at appropriate positions in either purine or benzimidazole lowered the Michaelis constant, KB, for deoxyribosyl acceptors: by about 10-fold for 6-methylpurine (KB 351 + 87 uM) compared with purine (KB 3.91 +0.8 mM) and by about 103-fold for 5,6-dimethylbenzimidazole 7.8 + 2.4 mM). The maximal rates of deoxyribosyl transfer to (KB 7.0 ± 0.79 /tM) compared with benzimidazole (Km different acceptors, on the other hand, varied by only 4.5-fold, and can be ascribed to decreases in the rate of release of the newly formed purine deoxyriboside from the enzyme. app.
INTRODUCTION The catabolism of purine deoxyribonucleosides in Leishmania and the closely related Crithidia [1,2], protozoal parasites of the order Kinetoplastida, differ substantially from the corresponding process in humans [3]. In man the degradation of purine ribonucleotides and deoxyribonucleotides seems to be catalysed by the same enzymes [4]. The functional importance of these catabolic enzymes became apparent when it was recognized that certain immunological disorders can be attributed to deficiencies of either adenosine deaminase or purine nucleoside phosphorylase. Adenosine deaminase or purine nucleoside phosphorylase deficiency is thought to result in an accumulation of dATP or dGTP respectively, both of which are negative effectors of ribonucleotide reductase. An impairment in the synthesis of deoxyribonucleotides in the optimal proportions necessary for DNA synthesis is, therefore, likely to be an important contributing factor in the development of immunological dysfunction. Leishmania and Crithidia lack adenosine or deoxyadenosine deaminases, but produce adenase which converts adenine into hypoxanthine. The hydrolysis of deoxyguanosine, deoxyadenosine and deoxyinosine is catalysed by a single enzyme, the substrate specificity of which is limited to purine deoxyribonucleosides [5,6]. Nucleosidases which are specific for purine deoxyribonucleosides have a very limited distribution in Nature [5,7], and it is not known whether such enzymes are essential for growth. It is, however, of interest that deoxyadenosine was found to be considerably more inhibitory to the growth of Crithidia fasciculata than was adenosine [8]. It therefore seems possible that inhibition of an enzyme which cleaves all three naturally occurring purine deoxyribonucleosides could have growth-retarding, if not cytocidal, effects. Unfortunately, no specific inhibitors of the purine-2'-deoxyribonucleosidase which could be used to study its physiological relevance are presently available. With this in mind the substrate specificity of the enzyme is of interest. The enzyme from Crithidia luciliae lends itself particularly well to such a study, in that it also catalyses deoxyribosyl transfer between purine bases, an activity previously thought to be present only in the lactobacilli [9]. Enzymic catalysis of glycosidic bond cleavage is thought to proceed with the formation of an enzyme-bound carboxonium ion of the sugar moiety as a reaction intermediate [10,11]. Deoxyribosyl transfer involves a reversal of N-glycosidic bond
Vol. 287
cleavage, with a different nucleic acid base acting as the deoxyribosyl acceptor. It is most unlikely that the deoxyriboside of a purine base which is an efficient deoxyribosyl acceptor will not also be a substrate. Consequently the deoxyribosyl transferase activity of the purine-2'-deoxyribonucleosidase of C. luciliae allows a study of the substrate specificity of these enzymes to be extended to a large number of purine analogues, most of which are not commercially available as deoxyribonucleosides. Previous reports on the substrate specificity of the purine-2'deoxyribonucleosidases merely served to establish that the enzyme was specific for purine-2'-deoxyribonucleosides, and only three of these were examined [5,6]. EXPERIMENTAL Materials Deoxyribonucleosides and purine bases were obtained either from Sigma Chemical Co., St. Louis, MO, U.S.A., or from Fluka Chemie AG, Buchs, Switzerland, and were all chromatographically homogeneous on h.p.l.c. Growth of C. luciliae and enzyme purification C. luciliae was a gift from the South African Institute for Medical Research. The cultivation of the organism and isolation of purine-2'-deoxyribonucleosidase were as previously described [6]. Determination of the deoxyribosyl acceptor activity and Km,app. of various purines Deoxyribosyl transfer by the purine-2'-deoxyribonucleosidase can be approximated by the reactions of Scheme 1. It can be shown by application of the procedure of King & Altman [12] that the initial rate of hydrolysis of a donor purine deoxyribonucleoside (A) [deoxyinosine (HxdR) in Scheme 1] in the presence of an acceptor purine (B) is given by: K'
V-A+(VAB/KB)A-B +A +(KA/KB)B+A *B/KB+(B/KiB) [K'A +(KA/KB)B] (1)
where A and B are concentrations of A and B respectively, K'A is the Michaelis constant and V the maximal velocity for the hydrolysis of A in the absence of B, K1B is an inhibition constant
126
D. J. Steenkamp and T. J. F. Halbich
version 4.5 (DAPA Scientific Software, P.O. Box 58, Kalamunda, Western Australia 6076). Because the assay method relies on the accumulation of significant amounts of products, and also because it was not experimentally feasible to extrapolate to infinite concentrations of deoxyinosine, eqn. (3) only applies as an approximation. A Km,.app. rather than KB, was obtained by fitting the peak areas for the newly synthesized purine deoxyriboside, BdR, to eqn. (3) using Enzfitter (R. J. Leatherbarrow, Elsevier-Biosoft, Cambridge, U.K.). For those purines which are not xanthine oxidase substrates or inhibitors, the kinetic constants for deoxyribosyl transfer were also determined spectrophotometrically in a coupled assay using xanthine oxidase. If the possibility of dead-end inhibition of the purine-2'deoxyribonucleosidase by free purines is ignored, it can be shown that:
EHx
HxdR EB
k_7
Kk,
-E
,
EFl dxcdR
B
BdR
k-2 k5
Hx
k_5
\
\7
k_3
EBdR ,%A
k3
N\
EEdFR
B
Scheme 1. Reactions involved in deoxyribosyl transfer as an alternative reaction pathway additional to the nucleosidase activity of a
(
purine-2'-deoxyribonucleosidase
vV
K'A+ A AB
Hx, hypoxanthine; HxdR, deoxyinosine; B, deoxyribosyl acceptor; dR, deoxyribose.
d[BdR] dt
AB
(VAB/KB)A-B denominator
(2)
At saturation with A, eqn. (2) simplifies to the Michaelis-Menten relationship: d[BdR] VAB'B (3) dt
KB+ B
The Michaelis constant for B, KB, can therefore also in this case be approximated by using high concentrations of deoxyinosine as the deoxyribosyl donor. Two different experimental approaches were used to determine KB for purines which gave rise to significant amounts of deoxyribosides. The first of these entailed separation of the products by h.p.l.c., which allowed an estimation of KB from the integrated peak areas for the deoxyribosides. However, because the necessary purine deoxyribosides were in several cases not available for calibration purposes, the actual reaction rates required for an estimation of Vmax were not determined from the results of such analyses. Reaction mixtures contained 50 ,umol of Na Hepes, pH 7.8, 1.0 ,umol of deoxyinosine, acceptor purine in the range 22.5-150 /tmol, and 0.035 unit of the enzyme in a final volume of 0.5 ml. The mixtures were incubated for 0, 20 and 40 min and the reaction was then terminated by heating to 100 'C for 5 min. The different time points were required to ascertain that the reaction rate was approximately linear. The formation of purine deoxyribosides was detected by h.p.l.c. using C- 18 reversed-phase chromatography. Most of the reaction mixtures could be satisfactorily resolved on a Beckman C- 18 Ultrasphere ODS column (4.6 mm x 250 mm). The column was eluted for 2 min with solvent A (50 mM-ammonium acetate, pH 5.5) followed by an 18 min linear gradient to 80 solvent B (solvent B is 80 % methanol) and subsequently a 10 min plateau at 80 % solvent B. The flow rate was 0.8 ml/min. Data capture was by means of Beckman System Gold software or using DAPA
VAB*A . B KA+AJ
(4)
KB(K 'A+A)+B (KA+A)
When A is large, for the dissociation of B from the dead-end complex, EB, and VA B KA and KB are the constants usually associated with a double-displacement mechanism. The rate equation for formation of the new purine deoxyriboside, BdR (where B is the deoxyribosyl acceptor and dR is deoxyribose), lacks the first term in the numerator:
V-A
eqn.
(4) reduces to: (VV~AB=
+B B
(5)
KB+B
Provided that V VAB, the dependence of the rate of hypoxanthine formation from deoxyinosine on the concentration of the deoxyribosyl acceptor purine allows, in principle, a determination of VAB and KB under conditions of minimal product accumulation. Assay mixtures contained 40,umol of Na-Hepes, pH 7.8, 2 ,tmol of deoxyinosine, 0.05 unit of xanthine oxidase and variable amounts of deoxyribosyl acceptor in a volume of 1.0 ml. Reactions were started by the addition of purine-2'-deoxyribonucleosidase, and the absorbance change of 293 nm, due to the formation of uric acid (c 10.7 mm-v' cm-'), was monitored in a Hitachi 220S spectrophotometer thermostatically controlled at 30 'C. The concentration of deoxyinosine used in these assays exceeded its Km value, which was independently estimated as 50.8 + 2.8 /,M at pH 7.8, by about 40-fold. Utilization of 12'-3Hldeoxyadenosine by C. luciliae C. luciliae was cultured in 2 litres of the medium of Shames et al. [13] to an attenuance of 1.3 at 600 nm. The organisms were harvested by centrifugation, washed once with 200 ml of 50 mmNa-Hepes, pH 6.8, containing 0.15 M-KCl, and then resuspended in 20 ml of ice-cold HCl04. After stirring for 30 min on ice, the mixture was centrifuged at 30000 g for 20 min and the supernatant was neutralized to pH 8.0 with 4 M-KOH containing 0.4 M-Hepes. The mixture was left for 1 h on ice before removing the KCl04 precipitate by centrifugation. The supernatant was treated with 10 units of alkaline phosphatase. The progress of dephosphorylation was monitored by h.p.l.c. of 200 ,ul aliquots of the reaction mixture. H.p.l.c. was performed on two Waters C- 18 Novapak Radial-Pak cartridges which were placed in series. The elution conditions were: 5 min at 00% solvent B; 20 min linear gradient to 10 % B at a flow rate of 1.5 ml/min; 18 min linear gradient to 80 % B at a flow rate of 1.2 ml/min; 18 min plateau at 80 % B; and 10 min linear gradient to return to initial conditions (solvent A, 50 mM-ammonium acetate, pH 5.5; solvent B, 800% methanol). When no further changes in the elution profile could be detected, 15 ml of the mixture was incubated for 1 h with 1.5 units of adenosine deaminase. To ensure that all contaminating enzyme activities were destroyed, the mixture
was
then treated with 9.2 ml of
1
M-HClO4,
1992
Specificity of purine-2-deoxyribonuclease
127
neutralized with conc. KOH and the supernatant lyophilized. The freeze-dried material was redissolved in 2 ml of 50 mMammonium acetate, pH 5.5. The effect of purine-2'-deoxyribonucleosidase on this preparation was examined in order to identify purine-2'-deoxyribonucleosides. The effect of purine-2'deoxyribonucleosidase on the h.p.l.c. elution pattern was also examined after destroying the ribonucleosides as described by Garrett & Santi [14]. The conversion of [2'-3H]deoxyadenosine into deoxyguanosine in the DNA of C. luciliae was investigated by labelling a 20 ml culture, which had been grown to an attenuance of 0.9-1.1 at 600 nm, with 50 ,Ci of [2'-3H]deoxyadenosine. The cells were harvested after 2 h, washed twice with 25 mM-Hepes, pH 7.8, containing 0.15 M-KCl, and the DNA was isolated using the simultaneous method of Krieg et al. [15]. A 2.2 mg aliquot was incubated overnight with 0.25 unit of 5'-nucleotidase, 0.015 unit of phosphodiesterase 1, and 145 units of deoxyribonucleosidase in 0.1 M-Tris/HCl, pH 8.0, containing 10 mM-MgCl2. The deoxyribonucleosides were separated on a Beckman Ultrasphere 5,u ODS column (250 mm x 4.6 mm) at a flow rate of 0.8 ml/min. The column was eluted for 5 min at 100 % solvent A, for 25 min with a linear gradient to 10 % solvent B, followed by 25 min with a linear gradient to 80% B. The radioactivity of the eluate fractions was determined by liquid scintillation counting using Instagel II (Packard). RESULTS AND DISCUSSION It was previously found that the rate of deoxyribosyl transfer from deoxyinosine to guanine, catalysed by the purine-2'-deoxyribonucleosidase, converged with the rate of hydrolysis of deoxyinosine at high guanine concentrations [6], as predicted by eqn. (1). Similar results were obtained when deoxyribosyl transfer to a relatively poor deoxyribosyl acceptor, benzimidazole (see Table 2), was quantified. H.p.l.c. analysis indicated that the yield of benzimidazole deoxyriboside was approx. 88% of the hypoxanthine generated by hydrolysis of deoxyinosine (Fig. 1). Interesting similarities between the purine-2'-deoxyribonucleosidase of C. luciliae and the trans-N-deoxyribosylase of L.
M 0
2.5-
i
dlno
0.5
0
1*..
1 BldR
0
2.0
1.5
A 0
a
1.0
3050
0
9~0 1013
70
0.5
0 -10
10
~30
50
10
Time (min)
90U
110
130
Fig. 1. Formation of benzimidazole deoxyriboside with deoxyinosine deoxyribosyl donor
,umol
as
the
,umol of of benzimidazole, 2.5 deoxyinosine, 40 umol of Na-Hepes, pH 8.0, and 0.057 unit of purine-2'-deoxyribonucleosidase was incubated at 30 'C. Aliquots were removed at various times, heated at 100 'C for 5 min and then analysed by reversed-phase h.p.l.c. The concentrations of stock solutions used for calibration were determined spectrophotometrically assuming 6246 10.7mm-W cmr1 for hypoxanthine (Hx), 6249 12.3 mM-' cm-' for deoxyinosine (dIno) and 6279 2900 mm-' cm-' for benzimidazole deoxyriboside (BIdR) in water. Benzimidazole deoxyriboside was obtained from a largerscale preparation using the enzyme and was purified by h.p.l.c.
A mixture containing 2.5
Vol. 287
helveticus are apparent in the utilization of benzimidazoles as substrates [16] and in the isomerization of at least one purine to the 7/3-isomer [6,17]. Moreover, the molecular masses of the two purine-2'-deoxyribonucleosidases [5,6] are quite similar to the subunit molecular mass of the trans-N-deoxyribosylase from L. helveticus [18]. The broad substrate specificity of the trans-N-deoxyribosylase from Lactobacillus spp. has been amply documented [18-20]. In the present study, reversed-phase h.p.l.c. was used to screen 27 different purines for the formation of the corresponding deoxyribosides. Compared with the enzymes from the Lactobacilli, the purine-2'-deoxyribonucleosidase from C. luciliae has a narrower specificity for purines as deoxyribosyl acceptors (Table 1). Compounds with modifications in the imidazole ring, as in allopurinol, aminopurinol, 8-azaguanine and 7-deazaguanine, were inactive as substrates. More substantial modifications of the pyrimidine ring were tolerated better by the transN-deoxyribosylase from L. helveticus than by the purine-2'deoxyribonucleosidase of C. luciliae. Thus a compound with an incomplete pyrimidine ring, such as 4-amino-5-carboxamidoimidazole, or with excessively bulky substituents, such as 6-[nhexylamino]purine, were not utilized by the C. luciliae enzyme, while both these compounds were substrates of the L. helveticus [16]. Unfortunately, the
enzyme
assay system for deoxyribosyl transfer lends itself poorly to the determination of the kinetic parameters
for various purines, and little quantitative information is available in the literature. It was, however, reported that 5,6dimethylbenzimidazole is a particularly efficient deoxyribosyl acceptor for the L. helveticus trans-N-deoxyribosylase [16]. In order to obtain a quantitative estimate of the efficiency of various purines as deoxyribosyl acceptors, Km,app. values were determined for a number of purines by product analysis using h.p.l.c. Because the method is time-consuming and suffers from a number of limitations which were outlined in the Experimental section, the results are mainly of interest in those cases where the spectrophotometric method could not be applied. The effects of three different deoxyribosyl acceptors on the spectrophotometrically observed rate of hypoxanthine formation from deoxyinosine are shown in Fig. 2. While these deoxyribosyl acceptors vary widely in their affinity for the enzyme, inhibition of the rate of hypoxanthine formation is in all three cases partial, indicating a transition from hydrolytic cleavage of deoxyinosine to an alternative reaction pathway involving deoxyribosyl transfer, as expected from Scheme 1. The partial nature of the inhibition indicates that binding of the deoxyribosyl acceptors to the free enzyme to form the dead-end complex, EB, does not contribute significantly, although such interactions at higher deoxyribosyl acceptor concentrations cannot be ruled out. Under certain conditions, as outlined in the Experimental section, it is therefore possible to estimate VAB and KB from the effect of deoxyribosyl acceptors on the rate of hypoxanthine formation, measured as uric acid by including xanthine oxidase in the assay. Detailed studies by Krenitsky et al. [21] on the specificity of xanthine oxidase, however, excluded compounds such as purine, adenine and 6-methylaminopurine, while for benzimidazole the VAR was not sufficiently different from V to allow a spectrophotometric determination of KB. For those deoxyribosyl acceptors with low values of KB, such as
5,6-dimethylbenzimidazole, considerable product
accumu-
lation occurred in the samples used for h.p.l.c. analysis, and this is reflected in a significant discrepancy between the apparent Km values obtained by h.p.l.c. analysis and the spectrophotometric result (Table 1). Considering the differences between the two methods, Km,app values obtained from h.p.l.c. analysis were, however, in reasonable agreement with KB values obtained from
128
D. J. Steenkamp and T. J. F. Halbich
Table 1. Purines and related compounds tested as deoxyribosyl acceptors of the purine-2'-deoxyribonucleosidase of C. luciliae Km app. refers to the Michaelis constant obtained by product analysis using h.p.l.c., while KB and VAB are the Michaelis constant and maximal velocity of deoxyribosyl transfer respectively obtained from initial rates of reaction measured spectrophotometrically.
(a)
A
RAB(Imol min-' mg-')
Purine
KB (uM)
Adenine Guanine 75.7+ 5.2 39.6+4.4 Hypoxanthine* Purine 6-Methylpurine 90.5 +6.0 351 +88 6-Chloropurine 43.4+ 15.8 259 + 77 6-Methylaminopurine 6-Methoxypurine 101 +5 122+24 6-Ethoxypurine 56.2 +4.4 53.7+ 9.0 Benzimidazole 158 5,6-Dimethyl34.9 + 2.8 7.0+0.8 benzimidazole * Deoxyadenosine was the deoxyribosyl donor.
Km.app. (uM) 115+27 74+9 146+ 12 3910+800 309+65
20
30
40
50
60
70
40
50
60
70
(b)
-
246+ 35 105+ 19 69.6+ 18.4 7840 + 2440 38.8 + 8.9
A 0
I
(.
'0
E 20
_- 160-
30
Retention time (min)
E Z 120E
Fig. 3. Identification of purine deoxynucleosides in an HCI04 extract of C. Ilcifae
(a) The traces represent elution patterns from reversed-phase h.p.l.c. of a dephosphorylated HCl04 extract of C. Iuciliae, which had also been incubated with adenosine deaminase before (bottom trace) and after (upper trace) treatment with purine-2'-deoxyribonucleosidase. (b) Effect of purine-2'-deoxyribonucleosidase on the same sample as was used in (a), but in which the ribonucleosides had been destroyed by treatment with periodate in the presence of methylamine.
.o 800
040 XC 40._ x ° I
°4--50
200
450 700 950 [Deoxyribosyl acceptor] (#iM)
1200
Fig. 2. Effect of deoxyribosyl acceptors on the production of hypoxanthine from deoxyinosine by purine-2'-deoxyribonucleosidase The initial rates of hypoxanthine formation were measured spectrophotometrically as uric acid in the presence of 6-methylpurine (@), 6-ethoxypurine (A) and 5,6-dimethylbenzimidazole (0). The reaction conditions were as described in the Experimental section.
spectrophotometric measurement of initial velocities for most of the purines. Major differences between the various deoxyribosyl acceptors are reflected in changes in the values of KR rather than VAB. Thus the K. and Km, app. values span a range of 103, while VAR values range over only 4.53-fold (Table 1). VAR can be expressed in terms of the microscopic rate constants (Scheme 1): Vk3+k5
(6)
while V= k3k4
(7)
The decrease in the rate of hypoxanthine production as a result of deoxyribosyl transfer to an acceptor purine can, therefore, be ascribed to a slower dissociation of the newly formed purine deoxyriboside from the enzyme than is the case for deoxyribose as a product, since V- VAB > O for k4 > k5. For the limited range of purines for which both VA. and KB
were obtained, decreases in K, which reflect a tighter association of the purine, also resulted in a lowering of VAR, as might be
expected. A significant trend apparent from the values given in Table 1 is evident in a lowering of Km app for hypoxanthine, 6-methoxypurine and 6-ethoxypurine upon introduction of the short alkyl chains. This effect of increased hydrophobicity is also apparent in a decrease in the Km,app. of more than 10-fold upon the introduction of a methyl group in the 6-position of purine, and by a factor of about 103 upon introducing two methyl groups at the 5- and 6-positions of benzimidazole. Among the compounds listed in Table 1, 5,6-dimethylbenzimidazole had the lowest KB. A similar effect of increased hydrophobicity on the Km app. values was not apparent in the series adenine, 6-methylaminopurine and 6-(dimethylamino)purine. Only trace amounts of a deoxyriboside of 6-(dimethylamino)purine were observed, which suggests that steric hindrance at the 6-position of the purine may be important. The efficiency of deoxyribosyl transfer catalysed by the enzyme in vitro raises a question concerning the physiological relevance of this reaction. Efficient deoxyribosyl transfer necessarily depends on restricted access of water or phosphate to the active site of a glycosidase. The decreased KB values observed with more hydrophobic substrates could be a mere reflection of this requirement. It is, however, not easy to exclude the possibility that 5,6-dimethylbenzimidazole deoxyriboside or 7,/-deoxyribofuranosylhypoxanthine may be naturally occurring substrates of the purine-2'-deoxyribonucleosidase, since the concentrations of deoxyribosides in Nature are very low. It is of interest to note that hypoxanthine was the only purine which gave rise to 1992
129
Specificity of purine-2-deoxyribonuclease Table 2. Purines tested as deoxyribosyl acceptors, but for which kinetic constants were not determined
Purine
Allopurinol Aminopurinol 8-Azaguanine 8-Azahypoxanthine 7-Deazaguanine 4-Amino-5-carboxamidoimidazole 2,6-Dichloropurine 1-Methyladenine 2-Methyladenine 3-Methyladenine 2-Aminopurine Isoguanine 6-Dimethylaminopurine 6-(n-hexylamino)purine
6-Methylmercaptopurine 2-Amino-6-methylmercaptopurine
Deoxyriboside No No No No No No Yes No Yes No Yes Yes Yes No Yes Yes
isomeric deoxyribosides [6], indicating a specificity which could be physiologically relevant. It was, however, not possible to demonstrate the presence of deoxyribonucleosides other than deoxyadenosine, deoxyinosine and deoxyguanosine in perchlorate extracts of C. luciliae after dephosphorylation. The results of an experiment which examines the effect of purine-2'-deoxyribonucleosidase on a perchlorate extract of C. luciliae which had been dephosphorylated and deaminated are shown in Fig. 3. The only detectable changes were the disappearance of peaks representing deoxyinosine and deoxyguanosine. Treatment of extracts with purine-2'-deoxyribonucleosidase, after first destroying the ribonucleosides with periodate, gave similar results (Fig. 3b). Since it was quite possible that the presence of low concentrations of novel deoxyribonucleosides would not have been detected by these procedures, the distribution of radiolabel from [2'-3H]deoxyadenosine between deoxyguanosine and deoxyadenosine in the DNA of C. luciliae is a more readily quantifiable measure of deoxyribosyl transfer in vivo. The amount of label from [2'-2H]deoxyadenosine present in DNA as deoxyguanosine and deoxyadenosine was 7.3 % and 92.7% of the total respectively. The appearance of a small proportion of the label in DNA as deoxyguanosine indicates that deoxyribosyl transfer does occur in vivo, but to a rather limited extent. The data reported in Tables 1 and 2 could serve as a useful Received 2 March 1992; accepted 9 April 1992
Vol. 287
guide for an examination of the substrate and inhibitor specificities of the purine-2'-deoxyribonucleosidase of Leishmania spp. It seems reasonable to suggest that methylated benzimidazole deoxyribosides, deactivated with respect to Nglycosidic bond cleavage by fluorination at the 2'-carbon [22] or as carbocyclic derivatives, may be potent slow substrate inhibitors of the enzyme which could be used to probe the functional significance of the purine-2'-deoxyribonucleosidase and its associated trans-N-deoxyribosylase activity. We thank the South African Medical Research Council and the University of Cape Town for financial support, and Mrs. B. Chodalski for able technical assistance.
REFERENCES 1. Coburn, C. M., Otteman, K. M., McNeely, T., Turco, S. J. & Beverly, S. M. (1991) Mol. Biochem. Parasitol. 46, 169-180 2. Looker, D., Miller, L. A., Elwood, H. J., Stickel, S. & Sagin, M. L. (1988) Nucleic Acids Res. 16, 7198 3. Elion, G. B. (1985) Adv. Enzyme Regul. 24, 323-334 4. Kredich, N. M. & Herschfield, M. S. (1989) in The Metabolic Basis of Inherited Disease (Scriver, C. R., Beaudet, A. L., Sly, W. S. & Valle, D., eds.), vol. 1, pp. 1045-1075, McGraw Hill, New York 5. Koszalka, G. W. & Krenitsky, T. A. (1979) J. Biol. Chem. 254, 8185-8193 6. Steenkamp, D. J. (1991) Eur. J. Biochem. 197, 431-439 7. Miller, R. L., Sabourin, C. L. K., Krenitsky, T. A., Berens, R. L. & Marr, J. J. (1984) J. Biol. Chem. 259, 5073-5077 8. Dewey, V. C., Kidder, G. W. & Nolan, L. L. (1978) Biochem. Pharmacol. 24, 1479-1485 9. Beck, W. S. & Levine, M. (1963) J. Biol. Chem. 238, 702-709 10. Cherian, X. M., van Arman, S. A. & Czarink, A. W. (1990) J. Am. Chem. Soc. 112, 4490-4498 11. Parkin, D. W. & Schramm, V. L. (1984) J. Biol. Chem. 259, 9418-9425 12. King, E. L. & Altman, C. (1956) J. Phys. Chem. 60, 1375-1378 13. Shames, S. L., Fairlamb, A. H., Cerami, A. & Walsh, C. T. (1986) Biochemistry 25, 3519-3526 14. Garrett, C. & Santi, D. V. (1979) Anal. Biochem. 99, 268-273 15. Krieg, P., Amtmann, E. & Sauer, G. (1983) Anal. Biochem. 134, 288-294 16. Holguin, J., Cardinaud, R. & Samelink, C. A. (1975) Eur. J. Biochem. 54, 515-520 17. Holguin-Hueso, J. & Cardinaud, R. (1972) FEBS Lett. 20, 171-173 18. Holguin, J. & Cardinaud, R. (1975) Eur. J. Biochem. 54, 505-514 19. Huang, M.-C., Hatfield, K., Roetker, A. W., Montgomery, J. A. & Blakley, R. L. (1981) Biochem. Pharmacol. 30, 2663-2671 20. Huang, M.-C., Montgomery, J. A., Thorpe, M. C., Stewart, E. L., Secrist, J. A. & Blakley, R. L. (1983) Arch. Biochem. Biophys. 222, 133-144 21. Krenitsky, T. A., Neil, S. M., Elion, G. B. & Hitchings, G. H. (1972) Arch. Biochem. Biophys. 150, 585-599 22. Abeles, R. H. & Alston, T. A. (1990) J. Biol. Chem. 265,16705-16708
Biochem. J. (1992) 287, 125-129 (Printed in Great Britain)
Substrate specificity of the purine-2'-deoxyribonucleosidase of Crithidia luciliae Daniel J. STEENKAMP and Torsten J. F. HALBICH Department of Chemical Pathology, University of Cape Town Medical School, Observatory 7925, Cape Town, South Africa
The purine-2'-deoxyribonucleosidase of Crithidia luciliae catalyses an efficient deoxyribosyl transfer between a variety of purine bases, benzimidazole and 5,6-dimethylbenzimidazole. Since the deoxyriboside of a deoxyribosyl acceptor is necessarily also a substrate, the trans-N-deoxyribosylase activity of the enzyme allows a study of its specificity to be extended to a large number of purines and purine analogues. Amongst 27 different deoxyribosyl acceptors, only hypoxanthine gave rise to isomeric products. The introduction of methyl groups at appropriate positions in either purine or benzimidazole lowered the Michaelis constant, KB, for deoxyribosyl acceptors: by about 10-fold for 6-methylpurine (KB 351 + 87 uM) compared with purine (KB 3.91 +0.8 mM) and by about 103-fold for 5,6-dimethylbenzimidazole 7.8 + 2.4 mM). The maximal rates of deoxyribosyl transfer to (KB 7.0 ± 0.79 /tM) compared with benzimidazole (Km different acceptors, on the other hand, varied by only 4.5-fold, and can be ascribed to decreases in the rate of release of the newly formed purine deoxyriboside from the enzyme. app.
INTRODUCTION The catabolism of purine deoxyribonucleosides in Leishmania and the closely related Crithidia [1,2], protozoal parasites of the order Kinetoplastida, differ substantially from the corresponding process in humans [3]. In man the degradation of purine ribonucleotides and deoxyribonucleotides seems to be catalysed by the same enzymes [4]. The functional importance of these catabolic enzymes became apparent when it was recognized that certain immunological disorders can be attributed to deficiencies of either adenosine deaminase or purine nucleoside phosphorylase. Adenosine deaminase or purine nucleoside phosphorylase deficiency is thought to result in an accumulation of dATP or dGTP respectively, both of which are negative effectors of ribonucleotide reductase. An impairment in the synthesis of deoxyribonucleotides in the optimal proportions necessary for DNA synthesis is, therefore, likely to be an important contributing factor in the development of immunological dysfunction. Leishmania and Crithidia lack adenosine or deoxyadenosine deaminases, but produce adenase which converts adenine into hypoxanthine. The hydrolysis of deoxyguanosine, deoxyadenosine and deoxyinosine is catalysed by a single enzyme, the substrate specificity of which is limited to purine deoxyribonucleosides [5,6]. Nucleosidases which are specific for purine deoxyribonucleosides have a very limited distribution in Nature [5,7], and it is not known whether such enzymes are essential for growth. It is, however, of interest that deoxyadenosine was found to be considerably more inhibitory to the growth of Crithidia fasciculata than was adenosine [8]. It therefore seems possible that inhibition of an enzyme which cleaves all three naturally occurring purine deoxyribonucleosides could have growth-retarding, if not cytocidal, effects. Unfortunately, no specific inhibitors of the purine-2'-deoxyribonucleosidase which could be used to study its physiological relevance are presently available. With this in mind the substrate specificity of the enzyme is of interest. The enzyme from Crithidia luciliae lends itself particularly well to such a study, in that it also catalyses deoxyribosyl transfer between purine bases, an activity previously thought to be present only in the lactobacilli [9]. Enzymic catalysis of glycosidic bond cleavage is thought to proceed with the formation of an enzyme-bound carboxonium ion of the sugar moiety as a reaction intermediate [10,11]. Deoxyribosyl transfer involves a reversal of N-glycosidic bond
Vol. 287
cleavage, with a different nucleic acid base acting as the deoxyribosyl acceptor. It is most unlikely that the deoxyriboside of a purine base which is an efficient deoxyribosyl acceptor will not also be a substrate. Consequently the deoxyribosyl transferase activity of the purine-2'-deoxyribonucleosidase of C. luciliae allows a study of the substrate specificity of these enzymes to be extended to a large number of purine analogues, most of which are not commercially available as deoxyribonucleosides. Previous reports on the substrate specificity of the purine-2'deoxyribonucleosidases merely served to establish that the enzyme was specific for purine-2'-deoxyribonucleosides, and only three of these were examined [5,6]. EXPERIMENTAL Materials Deoxyribonucleosides and purine bases were obtained either from Sigma Chemical Co., St. Louis, MO, U.S.A., or from Fluka Chemie AG, Buchs, Switzerland, and were all chromatographically homogeneous on h.p.l.c. Growth of C. luciliae and enzyme purification C. luciliae was a gift from the South African Institute for Medical Research. The cultivation of the organism and isolation of purine-2'-deoxyribonucleosidase were as previously described [6]. Determination of the deoxyribosyl acceptor activity and Km,app. of various purines Deoxyribosyl transfer by the purine-2'-deoxyribonucleosidase can be approximated by the reactions of Scheme 1. It can be shown by application of the procedure of King & Altman [12] that the initial rate of hydrolysis of a donor purine deoxyribonucleoside (A) [deoxyinosine (HxdR) in Scheme 1] in the presence of an acceptor purine (B) is given by: K'
V-A+(VAB/KB)A-B +A +(KA/KB)B+A *B/KB+(B/KiB) [K'A +(KA/KB)B] (1)
where A and B are concentrations of A and B respectively, K'A is the Michaelis constant and V the maximal velocity for the hydrolysis of A in the absence of B, K1B is an inhibition constant
126
D. J. Steenkamp and T. J. F. Halbich
version 4.5 (DAPA Scientific Software, P.O. Box 58, Kalamunda, Western Australia 6076). Because the assay method relies on the accumulation of significant amounts of products, and also because it was not experimentally feasible to extrapolate to infinite concentrations of deoxyinosine, eqn. (3) only applies as an approximation. A Km,.app. rather than KB, was obtained by fitting the peak areas for the newly synthesized purine deoxyriboside, BdR, to eqn. (3) using Enzfitter (R. J. Leatherbarrow, Elsevier-Biosoft, Cambridge, U.K.). For those purines which are not xanthine oxidase substrates or inhibitors, the kinetic constants for deoxyribosyl transfer were also determined spectrophotometrically in a coupled assay using xanthine oxidase. If the possibility of dead-end inhibition of the purine-2'deoxyribonucleosidase by free purines is ignored, it can be shown that:
EHx
HxdR EB
k_7
Kk,
-E
,
EFl dxcdR
B
BdR
k-2 k5
Hx
k_5
\
\7
k_3
EBdR ,%A
k3
N\
EEdFR
B
Scheme 1. Reactions involved in deoxyribosyl transfer as an alternative reaction pathway additional to the nucleosidase activity of a
(
purine-2'-deoxyribonucleosidase
vV
K'A+ A AB
Hx, hypoxanthine; HxdR, deoxyinosine; B, deoxyribosyl acceptor; dR, deoxyribose.
d[BdR] dt
AB
(VAB/KB)A-B denominator
(2)
At saturation with A, eqn. (2) simplifies to the Michaelis-Menten relationship: d[BdR] VAB'B (3) dt
KB+ B
The Michaelis constant for B, KB, can therefore also in this case be approximated by using high concentrations of deoxyinosine as the deoxyribosyl donor. Two different experimental approaches were used to determine KB for purines which gave rise to significant amounts of deoxyribosides. The first of these entailed separation of the products by h.p.l.c., which allowed an estimation of KB from the integrated peak areas for the deoxyribosides. However, because the necessary purine deoxyribosides were in several cases not available for calibration purposes, the actual reaction rates required for an estimation of Vmax were not determined from the results of such analyses. Reaction mixtures contained 50 ,umol of Na Hepes, pH 7.8, 1.0 ,umol of deoxyinosine, acceptor purine in the range 22.5-150 /tmol, and 0.035 unit of the enzyme in a final volume of 0.5 ml. The mixtures were incubated for 0, 20 and 40 min and the reaction was then terminated by heating to 100 'C for 5 min. The different time points were required to ascertain that the reaction rate was approximately linear. The formation of purine deoxyribosides was detected by h.p.l.c. using C- 18 reversed-phase chromatography. Most of the reaction mixtures could be satisfactorily resolved on a Beckman C- 18 Ultrasphere ODS column (4.6 mm x 250 mm). The column was eluted for 2 min with solvent A (50 mM-ammonium acetate, pH 5.5) followed by an 18 min linear gradient to 80 solvent B (solvent B is 80 % methanol) and subsequently a 10 min plateau at 80 % solvent B. The flow rate was 0.8 ml/min. Data capture was by means of Beckman System Gold software or using DAPA
VAB*A . B KA+AJ
(4)
KB(K 'A+A)+B (KA+A)
When A is large, for the dissociation of B from the dead-end complex, EB, and VA B KA and KB are the constants usually associated with a double-displacement mechanism. The rate equation for formation of the new purine deoxyriboside, BdR (where B is the deoxyribosyl acceptor and dR is deoxyribose), lacks the first term in the numerator:
V-A
eqn.
(4) reduces to: (VV~AB=
+B B
(5)
KB+B
Provided that V VAB, the dependence of the rate of hypoxanthine formation from deoxyinosine on the concentration of the deoxyribosyl acceptor purine allows, in principle, a determination of VAB and KB under conditions of minimal product accumulation. Assay mixtures contained 40,umol of Na-Hepes, pH 7.8, 2 ,tmol of deoxyinosine, 0.05 unit of xanthine oxidase and variable amounts of deoxyribosyl acceptor in a volume of 1.0 ml. Reactions were started by the addition of purine-2'-deoxyribonucleosidase, and the absorbance change of 293 nm, due to the formation of uric acid (c 10.7 mm-v' cm-'), was monitored in a Hitachi 220S spectrophotometer thermostatically controlled at 30 'C. The concentration of deoxyinosine used in these assays exceeded its Km value, which was independently estimated as 50.8 + 2.8 /,M at pH 7.8, by about 40-fold. Utilization of 12'-3Hldeoxyadenosine by C. luciliae C. luciliae was cultured in 2 litres of the medium of Shames et al. [13] to an attenuance of 1.3 at 600 nm. The organisms were harvested by centrifugation, washed once with 200 ml of 50 mmNa-Hepes, pH 6.8, containing 0.15 M-KCl, and then resuspended in 20 ml of ice-cold HCl04. After stirring for 30 min on ice, the mixture was centrifuged at 30000 g for 20 min and the supernatant was neutralized to pH 8.0 with 4 M-KOH containing 0.4 M-Hepes. The mixture was left for 1 h on ice before removing the KCl04 precipitate by centrifugation. The supernatant was treated with 10 units of alkaline phosphatase. The progress of dephosphorylation was monitored by h.p.l.c. of 200 ,ul aliquots of the reaction mixture. H.p.l.c. was performed on two Waters C- 18 Novapak Radial-Pak cartridges which were placed in series. The elution conditions were: 5 min at 00% solvent B; 20 min linear gradient to 10 % B at a flow rate of 1.5 ml/min; 18 min linear gradient to 80 % B at a flow rate of 1.2 ml/min; 18 min plateau at 80 % B; and 10 min linear gradient to return to initial conditions (solvent A, 50 mM-ammonium acetate, pH 5.5; solvent B, 800% methanol). When no further changes in the elution profile could be detected, 15 ml of the mixture was incubated for 1 h with 1.5 units of adenosine deaminase. To ensure that all contaminating enzyme activities were destroyed, the mixture
was
then treated with 9.2 ml of
1
M-HClO4,
1992
Specificity of purine-2-deoxyribonuclease
127
neutralized with conc. KOH and the supernatant lyophilized. The freeze-dried material was redissolved in 2 ml of 50 mMammonium acetate, pH 5.5. The effect of purine-2'-deoxyribonucleosidase on this preparation was examined in order to identify purine-2'-deoxyribonucleosides. The effect of purine-2'deoxyribonucleosidase on the h.p.l.c. elution pattern was also examined after destroying the ribonucleosides as described by Garrett & Santi [14]. The conversion of [2'-3H]deoxyadenosine into deoxyguanosine in the DNA of C. luciliae was investigated by labelling a 20 ml culture, which had been grown to an attenuance of 0.9-1.1 at 600 nm, with 50 ,Ci of [2'-3H]deoxyadenosine. The cells were harvested after 2 h, washed twice with 25 mM-Hepes, pH 7.8, containing 0.15 M-KCl, and the DNA was isolated using the simultaneous method of Krieg et al. [15]. A 2.2 mg aliquot was incubated overnight with 0.25 unit of 5'-nucleotidase, 0.015 unit of phosphodiesterase 1, and 145 units of deoxyribonucleosidase in 0.1 M-Tris/HCl, pH 8.0, containing 10 mM-MgCl2. The deoxyribonucleosides were separated on a Beckman Ultrasphere 5,u ODS column (250 mm x 4.6 mm) at a flow rate of 0.8 ml/min. The column was eluted for 5 min at 100 % solvent A, for 25 min with a linear gradient to 10 % solvent B, followed by 25 min with a linear gradient to 80% B. The radioactivity of the eluate fractions was determined by liquid scintillation counting using Instagel II (Packard). RESULTS AND DISCUSSION It was previously found that the rate of deoxyribosyl transfer from deoxyinosine to guanine, catalysed by the purine-2'-deoxyribonucleosidase, converged with the rate of hydrolysis of deoxyinosine at high guanine concentrations [6], as predicted by eqn. (1). Similar results were obtained when deoxyribosyl transfer to a relatively poor deoxyribosyl acceptor, benzimidazole (see Table 2), was quantified. H.p.l.c. analysis indicated that the yield of benzimidazole deoxyriboside was approx. 88% of the hypoxanthine generated by hydrolysis of deoxyinosine (Fig. 1). Interesting similarities between the purine-2'-deoxyribonucleosidase of C. luciliae and the trans-N-deoxyribosylase of L.
M 0
2.5-
i
dlno
0.5
0
1*..
1 BldR
0
2.0
1.5
A 0
a
1.0
3050
0
9~0 1013
70
0.5
0 -10
10
~30
50
10
Time (min)
90U
110
130
Fig. 1. Formation of benzimidazole deoxyriboside with deoxyinosine deoxyribosyl donor
,umol
as
the
,umol of of benzimidazole, 2.5 deoxyinosine, 40 umol of Na-Hepes, pH 8.0, and 0.057 unit of purine-2'-deoxyribonucleosidase was incubated at 30 'C. Aliquots were removed at various times, heated at 100 'C for 5 min and then analysed by reversed-phase h.p.l.c. The concentrations of stock solutions used for calibration were determined spectrophotometrically assuming 6246 10.7mm-W cmr1 for hypoxanthine (Hx), 6249 12.3 mM-' cm-' for deoxyinosine (dIno) and 6279 2900 mm-' cm-' for benzimidazole deoxyriboside (BIdR) in water. Benzimidazole deoxyriboside was obtained from a largerscale preparation using the enzyme and was purified by h.p.l.c.
A mixture containing 2.5
Vol. 287
helveticus are apparent in the utilization of benzimidazoles as substrates [16] and in the isomerization of at least one purine to the 7/3-isomer [6,17]. Moreover, the molecular masses of the two purine-2'-deoxyribonucleosidases [5,6] are quite similar to the subunit molecular mass of the trans-N-deoxyribosylase from L. helveticus [18]. The broad substrate specificity of the trans-N-deoxyribosylase from Lactobacillus spp. has been amply documented [18-20]. In the present study, reversed-phase h.p.l.c. was used to screen 27 different purines for the formation of the corresponding deoxyribosides. Compared with the enzymes from the Lactobacilli, the purine-2'-deoxyribonucleosidase from C. luciliae has a narrower specificity for purines as deoxyribosyl acceptors (Table 1). Compounds with modifications in the imidazole ring, as in allopurinol, aminopurinol, 8-azaguanine and 7-deazaguanine, were inactive as substrates. More substantial modifications of the pyrimidine ring were tolerated better by the transN-deoxyribosylase from L. helveticus than by the purine-2'deoxyribonucleosidase of C. luciliae. Thus a compound with an incomplete pyrimidine ring, such as 4-amino-5-carboxamidoimidazole, or with excessively bulky substituents, such as 6-[nhexylamino]purine, were not utilized by the C. luciliae enzyme, while both these compounds were substrates of the L. helveticus [16]. Unfortunately, the
enzyme
assay system for deoxyribosyl transfer lends itself poorly to the determination of the kinetic parameters
for various purines, and little quantitative information is available in the literature. It was, however, reported that 5,6dimethylbenzimidazole is a particularly efficient deoxyribosyl acceptor for the L. helveticus trans-N-deoxyribosylase [16]. In order to obtain a quantitative estimate of the efficiency of various purines as deoxyribosyl acceptors, Km,app. values were determined for a number of purines by product analysis using h.p.l.c. Because the method is time-consuming and suffers from a number of limitations which were outlined in the Experimental section, the results are mainly of interest in those cases where the spectrophotometric method could not be applied. The effects of three different deoxyribosyl acceptors on the spectrophotometrically observed rate of hypoxanthine formation from deoxyinosine are shown in Fig. 2. While these deoxyribosyl acceptors vary widely in their affinity for the enzyme, inhibition of the rate of hypoxanthine formation is in all three cases partial, indicating a transition from hydrolytic cleavage of deoxyinosine to an alternative reaction pathway involving deoxyribosyl transfer, as expected from Scheme 1. The partial nature of the inhibition indicates that binding of the deoxyribosyl acceptors to the free enzyme to form the dead-end complex, EB, does not contribute significantly, although such interactions at higher deoxyribosyl acceptor concentrations cannot be ruled out. Under certain conditions, as outlined in the Experimental section, it is therefore possible to estimate VAB and KB from the effect of deoxyribosyl acceptors on the rate of hypoxanthine formation, measured as uric acid by including xanthine oxidase in the assay. Detailed studies by Krenitsky et al. [21] on the specificity of xanthine oxidase, however, excluded compounds such as purine, adenine and 6-methylaminopurine, while for benzimidazole the VAR was not sufficiently different from V to allow a spectrophotometric determination of KB. For those deoxyribosyl acceptors with low values of KB, such as
5,6-dimethylbenzimidazole, considerable product
accumu-
lation occurred in the samples used for h.p.l.c. analysis, and this is reflected in a significant discrepancy between the apparent Km values obtained by h.p.l.c. analysis and the spectrophotometric result (Table 1). Considering the differences between the two methods, Km,app values obtained from h.p.l.c. analysis were, however, in reasonable agreement with KB values obtained from
128
D. J. Steenkamp and T. J. F. Halbich
Table 1. Purines and related compounds tested as deoxyribosyl acceptors of the purine-2'-deoxyribonucleosidase of C. luciliae Km app. refers to the Michaelis constant obtained by product analysis using h.p.l.c., while KB and VAB are the Michaelis constant and maximal velocity of deoxyribosyl transfer respectively obtained from initial rates of reaction measured spectrophotometrically.
(a)
A
RAB(Imol min-' mg-')
Purine
KB (uM)
Adenine Guanine 75.7+ 5.2 39.6+4.4 Hypoxanthine* Purine 6-Methylpurine 90.5 +6.0 351 +88 6-Chloropurine 43.4+ 15.8 259 + 77 6-Methylaminopurine 6-Methoxypurine 101 +5 122+24 6-Ethoxypurine 56.2 +4.4 53.7+ 9.0 Benzimidazole 158 5,6-Dimethyl34.9 + 2.8 7.0+0.8 benzimidazole * Deoxyadenosine was the deoxyribosyl donor.
Km.app. (uM) 115+27 74+9 146+ 12 3910+800 309+65
20
30
40
50
60
70
40
50
60
70
(b)
-
246+ 35 105+ 19 69.6+ 18.4 7840 + 2440 38.8 + 8.9
A 0
I
(.
'0
E 20
_- 160-
30
Retention time (min)
E Z 120E
Fig. 3. Identification of purine deoxynucleosides in an HCI04 extract of C. Ilcifae
(a) The traces represent elution patterns from reversed-phase h.p.l.c. of a dephosphorylated HCl04 extract of C. Iuciliae, which had also been incubated with adenosine deaminase before (bottom trace) and after (upper trace) treatment with purine-2'-deoxyribonucleosidase. (b) Effect of purine-2'-deoxyribonucleosidase on the same sample as was used in (a), but in which the ribonucleosides had been destroyed by treatment with periodate in the presence of methylamine.
.o 800
040 XC 40._ x ° I
°4--50
200
450 700 950 [Deoxyribosyl acceptor] (#iM)
1200
Fig. 2. Effect of deoxyribosyl acceptors on the production of hypoxanthine from deoxyinosine by purine-2'-deoxyribonucleosidase The initial rates of hypoxanthine formation were measured spectrophotometrically as uric acid in the presence of 6-methylpurine (@), 6-ethoxypurine (A) and 5,6-dimethylbenzimidazole (0). The reaction conditions were as described in the Experimental section.
spectrophotometric measurement of initial velocities for most of the purines. Major differences between the various deoxyribosyl acceptors are reflected in changes in the values of KR rather than VAB. Thus the K. and Km, app. values span a range of 103, while VAR values range over only 4.53-fold (Table 1). VAR can be expressed in terms of the microscopic rate constants (Scheme 1): Vk3+k5
(6)
while V= k3k4
(7)
The decrease in the rate of hypoxanthine production as a result of deoxyribosyl transfer to an acceptor purine can, therefore, be ascribed to a slower dissociation of the newly formed purine deoxyriboside from the enzyme than is the case for deoxyribose as a product, since V- VAB > O for k4 > k5. For the limited range of purines for which both VA. and KB
were obtained, decreases in K, which reflect a tighter association of the purine, also resulted in a lowering of VAR, as might be
expected. A significant trend apparent from the values given in Table 1 is evident in a lowering of Km app for hypoxanthine, 6-methoxypurine and 6-ethoxypurine upon introduction of the short alkyl chains. This effect of increased hydrophobicity is also apparent in a decrease in the Km,app. of more than 10-fold upon the introduction of a methyl group in the 6-position of purine, and by a factor of about 103 upon introducing two methyl groups at the 5- and 6-positions of benzimidazole. Among the compounds listed in Table 1, 5,6-dimethylbenzimidazole had the lowest KB. A similar effect of increased hydrophobicity on the Km app. values was not apparent in the series adenine, 6-methylaminopurine and 6-(dimethylamino)purine. Only trace amounts of a deoxyriboside of 6-(dimethylamino)purine were observed, which suggests that steric hindrance at the 6-position of the purine may be important. The efficiency of deoxyribosyl transfer catalysed by the enzyme in vitro raises a question concerning the physiological relevance of this reaction. Efficient deoxyribosyl transfer necessarily depends on restricted access of water or phosphate to the active site of a glycosidase. The decreased KB values observed with more hydrophobic substrates could be a mere reflection of this requirement. It is, however, not easy to exclude the possibility that 5,6-dimethylbenzimidazole deoxyriboside or 7,/-deoxyribofuranosylhypoxanthine may be naturally occurring substrates of the purine-2'-deoxyribonucleosidase, since the concentrations of deoxyribosides in Nature are very low. It is of interest to note that hypoxanthine was the only purine which gave rise to 1992
129
Specificity of purine-2-deoxyribonuclease Table 2. Purines tested as deoxyribosyl acceptors, but for which kinetic constants were not determined
Purine
Allopurinol Aminopurinol 8-Azaguanine 8-Azahypoxanthine 7-Deazaguanine 4-Amino-5-carboxamidoimidazole 2,6-Dichloropurine 1-Methyladenine 2-Methyladenine 3-Methyladenine 2-Aminopurine Isoguanine 6-Dimethylaminopurine 6-(n-hexylamino)purine
6-Methylmercaptopurine 2-Amino-6-methylmercaptopurine
Deoxyriboside No No No No No No Yes No Yes No Yes Yes Yes No Yes Yes
isomeric deoxyribosides [6], indicating a specificity which could be physiologically relevant. It was, however, not possible to demonstrate the presence of deoxyribonucleosides other than deoxyadenosine, deoxyinosine and deoxyguanosine in perchlorate extracts of C. luciliae after dephosphorylation. The results of an experiment which examines the effect of purine-2'-deoxyribonucleosidase on a perchlorate extract of C. luciliae which had been dephosphorylated and deaminated are shown in Fig. 3. The only detectable changes were the disappearance of peaks representing deoxyinosine and deoxyguanosine. Treatment of extracts with purine-2'-deoxyribonucleosidase, after first destroying the ribonucleosides with periodate, gave similar results (Fig. 3b). Since it was quite possible that the presence of low concentrations of novel deoxyribonucleosides would not have been detected by these procedures, the distribution of radiolabel from [2'-3H]deoxyadenosine between deoxyguanosine and deoxyadenosine in the DNA of C. luciliae is a more readily quantifiable measure of deoxyribosyl transfer in vivo. The amount of label from [2'-2H]deoxyadenosine present in DNA as deoxyguanosine and deoxyadenosine was 7.3 % and 92.7% of the total respectively. The appearance of a small proportion of the label in DNA as deoxyguanosine indicates that deoxyribosyl transfer does occur in vivo, but to a rather limited extent. The data reported in Tables 1 and 2 could serve as a useful Received 2 March 1992; accepted 9 April 1992
Vol. 287
guide for an examination of the substrate and inhibitor specificities of the purine-2'-deoxyribonucleosidase of Leishmania spp. It seems reasonable to suggest that methylated benzimidazole deoxyribosides, deactivated with respect to Nglycosidic bond cleavage by fluorination at the 2'-carbon [22] or as carbocyclic derivatives, may be potent slow substrate inhibitors of the enzyme which could be used to probe the functional significance of the purine-2'-deoxyribonucleosidase and its associated trans-N-deoxyribosylase activity. We thank the South African Medical Research Council and the University of Cape Town for financial support, and Mrs. B. Chodalski for able technical assistance.
REFERENCES 1. Coburn, C. M., Otteman, K. M., McNeely, T., Turco, S. J. & Beverly, S. M. (1991) Mol. Biochem. Parasitol. 46, 169-180 2. Looker, D., Miller, L. A., Elwood, H. J., Stickel, S. & Sagin, M. L. (1988) Nucleic Acids Res. 16, 7198 3. Elion, G. B. (1985) Adv. Enzyme Regul. 24, 323-334 4. Kredich, N. M. & Herschfield, M. S. (1989) in The Metabolic Basis of Inherited Disease (Scriver, C. R., Beaudet, A. L., Sly, W. S. & Valle, D., eds.), vol. 1, pp. 1045-1075, McGraw Hill, New York 5. Koszalka, G. W. & Krenitsky, T. A. (1979) J. Biol. Chem. 254, 8185-8193 6. Steenkamp, D. J. (1991) Eur. J. Biochem. 197, 431-439 7. Miller, R. L., Sabourin, C. L. K., Krenitsky, T. A., Berens, R. L. & Marr, J. J. (1984) J. Biol. Chem. 259, 5073-5077 8. Dewey, V. C., Kidder, G. W. & Nolan, L. L. (1978) Biochem. Pharmacol. 24, 1479-1485 9. Beck, W. S. & Levine, M. (1963) J. Biol. Chem. 238, 702-709 10. Cherian, X. M., van Arman, S. A. & Czarink, A. W. (1990) J. Am. Chem. Soc. 112, 4490-4498 11. Parkin, D. W. & Schramm, V. L. (1984) J. Biol. Chem. 259, 9418-9425 12. King, E. L. & Altman, C. (1956) J. Phys. Chem. 60, 1375-1378 13. Shames, S. L., Fairlamb, A. H., Cerami, A. & Walsh, C. T. (1986) Biochemistry 25, 3519-3526 14. Garrett, C. & Santi, D. V. (1979) Anal. Biochem. 99, 268-273 15. Krieg, P., Amtmann, E. & Sauer, G. (1983) Anal. Biochem. 134, 288-294 16. Holguin, J., Cardinaud, R. & Samelink, C. A. (1975) Eur. J. Biochem. 54, 515-520 17. Holguin-Hueso, J. & Cardinaud, R. (1972) FEBS Lett. 20, 171-173 18. Holguin, J. & Cardinaud, R. (1975) Eur. J. Biochem. 54, 505-514 19. Huang, M.-C., Hatfield, K., Roetker, A. W., Montgomery, J. A. & Blakley, R. L. (1981) Biochem. Pharmacol. 30, 2663-2671 20. Huang, M.-C., Montgomery, J. A., Thorpe, M. C., Stewart, E. L., Secrist, J. A. & Blakley, R. L. (1983) Arch. Biochem. Biophys. 222, 133-144 21. Krenitsky, T. A., Neil, S. M., Elion, G. B. & Hitchings, G. H. (1972) Arch. Biochem. Biophys. 150, 585-599 22. Abeles, R. H. & Alston, T. A. (1990) J. Biol. Chem. 265,16705-16708
