- Email: [email protected]
Identification of amino-acid residues linked to different properties of phosphoribosylpyrophosphate synthetase isoforms I and II
BB
i!,J ELSEVIER
Biochimica et BiophysicaActa 1207 (1994) 126-133
Biochi~ic~a et BiophysicaA~ta
Identification of amino-acid residues linked to different properties of phosphoribosylpyrophosphate synthetase isoforms I and II Imtiaz Abroad, Sumio Ishijima, Kazuko Kita, Masamiti Tatibana * Department of Biochemistry, Chiba UniversitySchool of Medicine, Inohana 1-8-1, Chuo-ku, Chiba 260, Japan Received 17 December 1993; revised 1 April 1994
Abstract The catalytic subunit of rat liver phosphoribosylpyrophosphate synthetase is composed of two isoforms, PRS I and PRS II. The amino-acid sequences differ only by 13 residues, out of which two Lys residues of PRS I at positions 4 and 152 give net additional positive charges to PRS I. Previous work has shown that PRS I is more sensitive to inhibition by ADP and GDP and more stable to heat treatment than is PRS II. To identify amino-acid residues responsible for the different properties, five chimeric enzymes between rat PRS I and PRS II and two mutated enzymes with a single point mutation at position 152 were constructed; these enzymes were produced in Escherichia coli. Changing Lys-4 of PRS I to Val, together with Ile-5 to Leu, completely abolished sensitivity to GDP inhibition of PRS I, indicating that Lys-4 in PRS I is critical for GDP inhibition. The substitutions at position 152 had little effect on GDP inhibition. Characterization of the chimeric enzymes revealed that residues between residues 54-110 and 229-317, namely, Val-55 and/or Ala-81, and Arg-242 and/or Cys-264 of PRS I also contribute to the strong GDP inhibition. Lys-4 was also important for the strong ADP inhibition of PRS I. Regarding the physical properties, chimeric enzymes bearing residues 12-53 of PRS I were stable at 49°C and with digestion with papain and proteinase K. Our observations suggest that Lys-17, Ile-18, and/or Cys-40 of PRS I contribute to stability of the enzyme. Key words: Phosphoribosylpyrophosphate synthetase; Isoform; Chimeric enzyme; Nucleotide inhibition
I. Introduction P h o s p h o r i b o s y l p y r o p h o s p h a t e (PRPP) synthetase (ATP:D-ribose-5-phosphate pyrophosphotransferase, EC 2.7.6.1) catalyzes the formation of PRPP from ATP and ribose 5-phosphate. PRPP provides an important substrate for synthesis of almost all nucleotides [1,2], and is a critical control factor for synthesis de novo of purines [3-5] and pyrimidines [6-8]. PRPP synthetase has been purified from Salmonella typhimurium [9,10], Escherichia coli [11], Bacillus subtilis [12], human erythrocyte [13], and rat liver [14,15]. The rat
Abbreviations: PRPP, 5-phosphoribosyl 1-pyrophosphate;PRS I and PRS II, phosphoribosylpyrophosphatesynthetase 34-kDa subunits I and II, respectively; kb, kilobase pairs; HPLC, high-performanceliquid chromatography; PAGE, polyacrylamidegel electrophoresis; IPTG, isopropylfl-D-thiogalactopyranoside; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid. * Corresponding author. Fax: + 81 43 2262041. 0167-4838/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0167-4838(94)00065-O
liver enzyme exists as complex aggregates of 34-, 39- and 41-kDa components [16] (39- and 41-kDa components were formerly reported to be 38 and 40 kDa, respectively [15]), with the 34-kDa species being the catalytic subunit [15]. The cloning of rat c D N A and amino-acid sequencing of the catalytic subunit purified from the rat liver [17] revealed that the 34-kDa component is a mixture of two highly homologous isoforms, PRS I and PRS II, The deduced amino-acid sequences of both sets of 317 residues differ only by 13 residues, out of which 6 are non-conservative substitutions [17]. PRS I and PRS II are encoded by separate genes which have been mapped on distinct regions of the X chromosome. Rat mRNAs of the genes are expressed in all tissues but their relative expression is variable [18]. Characterization of the recombinant rat PRS I and PRS II expressed in E. coli [19] showed different properties between the two isoforms. PRS I is more sensitive to nucleotide inhibition than is PRS II. Inhibition of PRS I and PRS II by 0.3 mM A D P was 87% and 54%, respectively, and inhibition by 1 mM GDP was 93% and 24%, respectively. PRS II was 180-fold more sensitive to
L Ahmadet aL/Biochimica et BiophysicaActa 1207 (1994) 126-133 heat inactivation at 49°C than was PRS I. Regarding the structure-function relationship, the two substitutions of Lys for Val and Lys for Gin at positions 4 and 152 warrant attention. These two substitutions give net additional positive charges to PRS I. It is possible that the different sensitivity to nucleotide inhibition is due to either or both of the substitutions. To better comprehend the structure-function relationship between the two isoforms, we constructed five chimeric enzymes and two mutated enzymes with a single point mutation at position 152. We describe here the purification and characterization of the chimeric and mutated enzymes expressed in E. coli cells. On the basis of the results obtained, we propose that Lys-4 of PRS I is critical for GDP inhibition but at least two other residues also contribute to the strong GDP inhibition of PRS I. Regarding the physical properties, Lys-17, Ile-18 a n d / o r Cys-40 of PRS I play an important role in the thermal stability and resistance to proteolytic inactivation of the enzyme.
(A)
Sacl
eas, I
and
Accl Hindlll
I
I
I
'
"
I
Kpnl
I I
I
PRS III
(B)
"K'I
PRS I
I
"K'81~G~Vs'A~'2~'95"~'T ''D ='ZR " C
1
VL
I Rk, S I
I S QT
I I
E
] K
A
PRS II 1.228.11
2 2 8 VIlli/Ill/IlIA 11 o
1.110.11 1.53.11
1.11.11 2. Materials
127
F/////////////////////////////////////A 1
methods
2.1. Bacterial strains and plasmids E. coli strains MV1304 and AD8 were used for cloning and expression experiments. MV1304 contains the lacI q repressor and AD8 lacks the Ion proteinase. The expression vector pG1KHB [19] is derived from pKK233-2 (Pharmacia) and was used to express the chimeric and mutated enzymes.
I1.11.1 ~ / A l l [ Fig. 1. Construction of chimeric enzymes between rat PRS I and PRS II. (A) Common restriction enzyme sites of PRS I and PRS II cDNAs. Dotted lines indicate the newly created sites. (B) The upper two open bar and hatched box show the structures of PRS I and PRS II, respectively, with the 13 different amino-acid residues between PRS I and PRS II and their positions. The lower five open and hatched boxes show structures of the chimeric enzymes constructed between PRS I and PRS II. Open bars and hatched boxes show regions derived from PRS I and PRS II, respectively. For example, in 1.228.II, amino-acid residues 1-228 are derived from PRS I and the rest C-terminal residues are derived from PRS II.
2.2. Enzymes and other materials Restriction and D N A modifying enzymes were obtained from Takara Shuzo, Toyobo and New England Biolabs and were used according to the manufacturer's directions. Site-directed mutagenesis was carried out using the Amersham oligonucleotide-directed in vitro mutagenesis kits. TSK G3000SW and G4000SW columns were purchased from Tosoh Manufacturing. Sources of all other reagents were as described elsewhere [15]. 2.3. Plasmid construction of chimeric PRS I / PRS H cDNAs for chimeric enzymes were constructed by connecting fragments of the cDNAs for PRS I and PRS II. A restriction enzyme site common to the two cDNAs at the same locus was used for splicing c D N A fragments. For this purpose, new sites were introduced into the cDNAs by oligonucleotide-directed mutagenesis (Fig. 1A). An antisense synthetic 19-mer 5 ' - A T G A C A A A T G G T A C C A C A A-3' was used to create a site for KpnI in PRS I c D N A and antisense 19-mers 5 ' - G T G A G A G C T C C C G C T G A A G-3' and 5 ' - C T G G A T G A T G T A G A C A T C T - 3 ' were used
to create unique SacI and AccI sites, respectively, in PRS II cDNA. The new restriction enzyme sites created were confirmed by mapping with restriction enzymes. We constructed five chimeric cDNAs by interchanging appropriate fragments of the two cDNAs, as shown in Fig. lB. Junctional sites between PRS I and PRS II in the chimeras were identified by numbers referring to the last amino-acid residue of the N-terminal region. For example, in the chimeric enzyme 1.228.11, amino-acid residues from 1 to 228 are derived from PRS I and remaining residues are derived from PRS II. After ligation of the appropriate c D N A fragments, the products were cloned into the vector pG1KHB. The resulting expression plasmids contained the complete coding region of 954 bp and 3'-noncoding region (0.24 kb for PRS I c D N A or 1.1 kb for PRS II cDNA), under control of the trc promoter. The plasmids were introduced into E. coli MV1304 cells, except for the plasmid for the chimeric enzyme 1.11.II, which was introduced into AD8 cells. To introduce a single point mutation of Lys-152 to Gin in PRS I, an antisense 19-mer synthetic oligonucleotide 5 ' - G A A T C C A C T G C A G G A C A G C - 3 ' was
128
I. Ahmadet al. /Biochimica et BiophysicaActa 1207 (1994) 126-133
used. This mutagenesis created a new site for PstI. Another antisense 19-mer synthetic oligonucleotide 5'-GAATCCACITAAGAACTGC-3' was used to introduce Gln-152 to Lys mutation in PRS II. This mutagenesis created a new site for AfllI. New restriction sites created were verified by restriction mapping. These plasmid constructs were introduced into cells of E. coli strain MV1304. The mutant PRS I bearing a single replacement of Lys-152 to Gin, is referred to as IK152Q and the mutant PRS II bearing a single replacement of Gln-152 to Lys, is referred to as IIQ152K. 2.4. Cell growth The transformed cells were grown at 30°C for 15 h in 1.5-3 1 of M9 medium [20] containing 1 mM thiamine and ampicillin (50 /zg/ml). After adding 1 mM IPTG, the cells were grown for an additional 3 h at 30°C and then harvested by centrifugation. 2.5. Purification of chimeric and mutated enzymes All procedures were carried out at 4°C and all buffers used contained 0.3 mM ATP, 6 mM MgC12, 0.1 mM EDTA and 2.5 mM 2-mercaptoethanol as enzyme stabilizing agents [13,14] unless otherwise stated. The pH values indicated were measured at 4°C. The E. coli cells were suspended in a minimum volume of a solution containing 50 mM potassium phosphate (pH 7.4) and 0.1 mM phenylmethylsulfonyl fluoride (buffer A) and lysed by treatment with egg white lysozyme (1 m g / m l ) for 20 min. The mixture was then sonicated with Sonifier Cell Disrupter 185 (Branson Sonic Power, USA), using a standard microtip. The output control of the sonifier was placed at a setting of 4 and the cells were disrupted by sonic treatment, 5-times each for 15 s. The cell debris was sedimented at 10 000 X g for 15 min. The protein concentration of the cellular extracts ranged from 5 to 10 mg/ml. All subsequent steps were as described previously [15] but with the following modifications: (a) The IK152Q and IIQ152K proteins in the cellular extracts were precipitated with 5% ( w / v ) polyethylene glycol 6000. The precipitates were dissolved in buffer A and the protein concentration was adjusted to 3.3 m g / m l and the enzymes were then precipitated at pH 6.1. The acid precipitates of the mutated enzymes were dissolved in buffer A and pH was adjusted to 7.4. (b) Chimeric enzymes 1.11.II, 1.228.1I, 1.110.II were precipitated with 6.6% ( w / v ) of polyethylene glycol 6000 and the precipitates were dissolved in buffer A. The protein concentration was adjusted to 3.3 mg/ml. 1.11.II and 1.228.1I were then precipitated at pH 5.7 while 1.110.II was precipitated at pH 6.2. The acid precipitates were dissolved in buffer A and pH was adjusted to 7.4. (c) For purification of II.11.I, the cellular extract was treated with 0.8% streptomycin sulfate, then the enzyme was precipitated with 45%-saturated ammonium sulfate. The precipitate
was dissolved in buffer A and the protein concentration was adjusted to 3.3 mg/ml. The enzyme was precipitated at pH 4.9, dissolved in buffer A and the pH was adjusted to 7.4. (d) For purification of 1.53.I1, the cellular extract was treated with 0.8% streptomycin sulfate, and then with 90%-saturated ammonium sulfate. The precipitate was then dissolved in buffer A, and the solution was applied to a TSK Gel 3000SW column (0.75 X 60 cm) equilibrated with 50 mM potassium phosphate (pH 7.4). The column was eluted with the same buffer at a flow rate of 0.8 ml/min, and active fractions were pooled. The above preparations of the chimeric and mutated enzymes were stored in small portions at - 8 0 ° C until their properties were examined. 2.6. Enzyme assay The enzyme activity was assayed using the two methods described elsewhere [15]. Briefly, method 1 for a crude enzyme preparation, we used a reaction mixture (0.5 ml) containing 50 mM Hepes-KOH (pH 7.4 at 37°C), 10 mM potassium phosphate, 4 mM MgC12, 1 mM EDTA, 0.8 mM ATP, 0.2 mM ribose 5-phosphate, 1 mM dithiothreitol, 1.2 mM phosphoenolpyruvate, pyruvate kinase (9 units), and the enzyme, unless otherwise stated. The amount of PRPP synthesized was measured by a modification [21] of the enzymatic method of Kornberg et al. using labeled orotic acid [22]. Method 2, used for preparations after the acid precipitation step, measured ribose 5-phosphate dependent [14C]AMPproduction from [14C]ATP. The composition of the assay mixture (0.1 ml) was identical with that of the mixture for method 1, except that the unlabeled ATP was replaced with 0.4 mM [8-14C]ATP, and phosphoenolpyruvate and pyruvate kinase were omitted. One unit of enzyme activity was defined as the amount catalyzing formation of 1 /zmol of PRPP/min, under standard conditions. 2.7. Protein concentration Protein concentration was determined by the method of Bradford [23], using bovine serum albumin as a standard. For determination of specific activities of the chimeric and mutated enzymes, each purified preparation was applied to SDS-PAGE and the amount of the 34-kDa protein was determined based on densitometric analysis of the gel stained with Coomassie brilliant blue R-250. 2.8. Molecular weight determination by gel filtration The molecular weight of the chimeric and mutant enzymes was determined by gel filtration using TSK G4000SW HPLC column (0.75 X 60 cm). The running buffer was 50 mM potassium phosphate (pH 7.4) containing the enzyme stabilizing agents.
I. Ahmad et al. / Biochimica et BiophysicaActa 1207 (1994) 126-133 2.9. Other methods
SDS-PAGE was performed by the method of Laemmli [24]. In western blot analysis, proteins were separated by SDS-PAGE and transferred electrophoretically to a polyvinylidine difluoride membrane [25]. Protein was immunodetected using rabbit antiserum raised against rat liver PRPP synthetase and anti-rabbit IgG goat antibody conjugated with horseradish peroxidase (E.Y. Labs, USA). Color was developed with 0.8 m g / m l of diaminobenzidine tetrahydrochloride and 0.009% H 2 0 2 .
3. Results and discussion 3.1. Expression of chimeric PRS I / enzymes in E. coli
PRS II and mutated
We constructed five chimeric cDNAs using convenient restriction sites (Fig. 1B) and two cDNAs for the mutated enzymes at position 152. The plasmids carrying the cDNAs for chimeric and mutated enzymes were introduced into the host cells of E. coli. As the chimeric enzyme 1.11.II from the MV1304 transformants was not obtained in a sufficient amount, the strain AD8 which lacks the Ion proteinase was used for this chimeric enzyme. The cells were grown at 30°C in M9 medium to minimize expression of the constitutive E. coli enzyme [19]. Expression of the chimeric and mutated enzymes was induced by IPTG. The enzyme activity in soluble fractions of the crude cellular extracts increased after the addition of IPTG and reached maximum in 3 - 4 h. At maximal levels, the enzyme activities of the extracts of the cells containing the chimeric and mutated plasmids were from 4- to 60-fold higher than that from the control MV1304 or AD8 cells containing the vector alone: 1.228.11, 10.5-fold; 1.110.II, 24-fold; 1.53.I1, 4.5-fold; 1.11.II, 23-fold; II.11.I, 5.6-fold; IK152Q, 48-fold; IIQ152K, 60-fold. The elevated enzyme activity in cells carrying the cDNAs suggested that the chimeric and mutated enzymes had catalytic activity, which was expressed. 3.2. Purification of chimeric and mutated enzymes
Since homology of the deduced amino-acid sequences of the E. coli and the recombinant rat enzymes is high (47%), separation could be difficult. Therefore, E. coli cells containing the plasmids were grown under conditions of culture that would repress the synthesis of bacterial PRPP synthetase [19]. The chimeric and mutated enzymes were purified from E. coli MV1304 and AD8 transformants. The procedure for chimeric enzymes 1.228.1I, 1.110.II, 1.11.II and mutated IK152Q and IIQ152K included fractionation with polyethylene glycol and acid precipitation. As the other two chimeric enzymes, 1.53.11 and I1.11.I were not precipitated with polyethylene glycol,
129
the cellular extracts were treated with streptomycin, after which the enzymes were precipitated with ammonium sulfate. 1.53.II was further subjected to gel filtration on TSK Gel 3000SW. II.11.I was further purified by acid precipitation at pH 4.9. The E. coli enzyme expressed in a small amount was separated in these purification steps; in a separate purification of the E. coli enzyme, the bacterial enzyme showed behaviour which differed from that of the chimeric and mutated enzymes seen at steps of polyethylene glycol precipitation and acid precipitation. While we can not completely rule out the possibility of contamination by the E. coli enzyme, the amount of the contaminated E. coli enzyme should be limited. The expression levels of the chimeric enzymes 1.53.II and II.11.I in the cells were relatively low, but the physical properties of the partially purified enzymes were quite different from that of the E. coli enzyme. The partially purified bacterial enzyme was unstable at 49°C at a protein concentration of 28 ~ g / m l . After 10 min of incubation, below 10% of the initial activity remained. Except for the chimeric 1.11.II, all other enzymes were much more stable (see Fig. 2A). Specific activities of the purified chimeric and mutated enzymes, calculated as milliunit/mg of 34-kDa protein were as follows: 1.228.II, 12500; 1.110.II, 13000; 1.53.11, 12800; 1.11.II, 16900; II.11.I, 8500; IK152Q, 10400; IIQ152K, 14 400. These values are significantly lower than values of the specific activities of PRS I (25 700) and PRS II (34 500) [19]. 3.3. Characterization of chimeric and mutated enzymes
Physical and kinetic properties of the chimeric and mutated enzymes were examined and compared. The mammalian PRPP synthetases exist in multiple aggregated forms [26]. To examine whether the purified chimeric and mutated enzymes also existed in aggregated forms, each enzyme was applied to HPLC gel filtration on TSK 4000SW. All these enzymes did elute in aggregated forms at the position of a molecular mass ranging from 300 to 1100 kDa. We determined the apparent Michaelis constants for ATP and ribose 5-phosphate of the partially purified chimeric and mutated enzymes. The K m values for ATP of the enzymes are as follows (in /~M): 1.228.II, 17; 1.110.II, 25; 1.53.1I, 51; 1.11.II, 45; II.11.I, 50; IK152Q, 23; IIQ152K, 42, and the V / K m values ranged from 6-25 min-1/xM -1. The values of K m and V / K m for ribose 5-phosphate of these enzymes were within the range of 25-50 /zM and 9-29 min-1/zM -1, respectively. These values indicated that the chimeric and mutated enzymes retained functioning ATP and ribose 5-phosphate binding sites and catalytic efficiencies of these enzymes, as represented by V//Km, w e r e 30-125% of those of native PRS I and PRS II. Regarding physical properties of the isoforms, PRS I is far more stable than PRS II to heat inactivation [19]. To
I. Ahmad et aL / Biochimica et Biophysica Acta 1207 (1994) 126-133
130
ascertain which amino-acid residues contribute to the stability of PRS I, heat inactivation of the chimeric and mutated enzymes was examined (Fig. 2A). 1.228.1I, 1.110.II and 1.53.11, which retain a considerable part of the Nterminal region of PRS I, were fairly stable when incubated for 60 min at 49°C. The residual activities were 65%, 61% and 77%, respectively. In contrast, 1.11.II showed residual activity of only 0.6% when treated similarly. This enzyme has three substitutions compared with 1.53.II, i.e., Lys-17 to Arg, Ile-18 to Val and Cys-40 to Ser. II.11.I, which retains Lys-17, Ile-18 and Cys-40, showed a stability similar to those of 1.228.1I, 1.110.II and 1.53.II. These results indicate that at least one of the three residues, Lys-17, Ile-18 and Cys-40, is important for the thermostability of PRS I.
1.228.11, 1.110.II and 1.53.II. Therefore, at least one of the three residues, Lys-17, Ile-18 and Cys-40, contributes to resistance to the proteinase inactivation of PRS I. The finding that 1.11.II was readily inactivated by proteinase is consistent with the result that the yield of 1.11.II production was poor in the E. coli strain MV1304, but was increased in the strain AD8 which lacks the Ion proteinase. Different sensitivities to the proteinase inactivation were also observed between PRS II and 1.11.II. PRS II was more susceptible to inactivation by papain than was I. 11.II, but was more resistant to inactivation by proteinase K. The results shown in Fig. 2 show a correlation between thermostability and resistance to proteolytic inactivation. A remarkable difference was evident between chimeric 1.53.II and 1.11.II regarding the response to heat and proteinase treatment. Among the three residues involved in the difference (positions 17, 18 and 40) the substitution of Cys vs. Ser at the position 40 is notable, as substitutions at the positions 17 and 18, namely, Lys vs. Arg and Ile vs. Val, are conservative. Cys-40 might play an important role in the stability of PRS I. It is interesting that Cys-40 is replaced by Ser in the E. coli PRPP synthetase [11], which was susceptible to heat inactivation. Further studies with site directed mutagenesis will elucidate a definitive role of Cys-40.
3.4. Proteinase digestion A remarkable difference was found between the sensitivity to inactivation by proteinase digestion of PRS I and PRS II. Activity of PRS II was lost more rapidly than was that of PRS I in case of treatment with papain and proteinase K at 25°C for 4 h (Fig. 2B and 2C). To identify which residues of PRS I contribute to the resistance to proteinase inactivation, all the chimeras were treated with papain and proteinase K; similar responses were observed with this treatment. 1.228.II, 1.110.II and 1.53.II were resistant to proteinase inactivation and sensitivities were similar to that of PRS I. In contrast, 1.11.II showed a much lower residual activity when treated similarly. II.11.I was resistant to proteinase inactivation to a similar degree to
3.5. Inhibition by nucleotides ADP and GDP are the most potent inhibitors of PRPP synthetases purified from various sources [12]. We reported different sensitivities of PRS I and PRS II to the
A Heat stability
Enzyme
B Stability to papain inactivation
C Stability to proteinase K inactivation
il
m
1.228.11 1
1.110.111
~///////~
m []
1.53.,I I. 11., I
,. l.IH
I
m
m ,
,
20
40
,, 60
n
. . . . 0 100
Residual Activity (%)
0
20
, 40
60
. . . . . . .
0 100
Residual Activity (%)
20 40 80 80 100120
Residual Activity (%)
Fig. 2. Inactivationby heat and protease treatment of the native and chimeric enzymes. Structures of the enzymes are shown on the left. Open and hatched boxes show regions derived from PRS I and PRS II, respectively. (A) Heat stability of the enzymes. Residual enzyme activity was assayed after incubating at 49°C for 1 h at a protein concentrationof 28 /xg/ml in 50 mM potassium phosphate buffer (pH 7.4) containing the enzyme stabilizing agents. (B) and (C) Resistance to inactivation by papain and proteinase K, respectively. Residual enzyme activity was assayed after the enzymes were treated with the respective proteinase at a ratio of 50:1 (w/w) for 4 h at 25°C.
131
L Ahmad et al./Biochimica et Biophysica Acta 1207 (1994) 126-133
A
B i
i
i
i
i
C i
\
i
i
i
~
,
i
,
,
,
" ql
~ '~ ° 40I
"~
1~40I
, 0 o0.2 0.4 0.6 0.8 .0 GDP (mM)
~X,,xx~
o
l~
, , , , ,0 0.2 0.4 0.6 0.81.0 GDP (mM)
40I
o,
, , , , , 0 0.2 0.4 0.6 0.8 1.0 GDP (raM)
Fig. 3. GDP inhibition of chimeric and mutated enzymes. Standard assay conditions (0.4 mM ATP, 0.2 mM ribose 5-phosphate, 4 mM MgC12, 10 mM potassium phosphate buffer, pH 7.4) were used. ( O ) PRS I, ( O ) PRS II, (zx) II.11.I, ( • ) 1.11.II, ( [ ] ) IK152Q, ( • ) IIQ152K, ( [ ] ) 1.228.II, (half-filled circle) 1.110.II, and ( ~ ) 1.53.II.
inhibition by ADP and GDP [19]. To identify the aminoacid residues contributing to the difference, inhibition of the chimeric and mutated enzymes by ADP and GDP were examined (Figs. 3 and 4). The most notable finding was that the chimeric enzyme II.11.I was inhibited only 1% by 1 mM GDP, whereas PRS I was inhibited by 91% (Fig. 3A). Only 2 residues are different between II.11.I and PRS I, Val vs. Lys at the position 4 and Leu vs. lie at the position 5. Since substitution at position 5 is conservative, its effect on function of the enzyme is probably limited. Thus, the finding strongly suggests that Lys-4 in PRS I is critical for the sensitivity to GDP inhibition. However, the
A
extent of inhibition of PRS 1I and the chimeric 1.11.II is small, thereby showing that merely changing Val-4 to Lys in PRS II will not induce a strong GDP inhibition. Regarding the function of Lys-152 in PRS I, the inhibition of IK152Q was only slightly weaker than that of PRS I (Fig. 3B). The extent of GDP inhibition of IIQ152K was similar to that of PRS II. These results indicate that Lys-152 is not a critical factor in the PRS I sensitivity to GDP inhibition. In analyses of other chimeric enzymes, the extent of inhibition of the chimeric enzyme 1.228.1I was significantly smaller than that of PRS I (Fig. 3C). As the sequence of this chimeric enzyme differs from that of PRS
B ,
,
,
i
C
f
100;
00
)0
._~ 80 "5
80
30
60
60
60
4o
40
40
2O
20
20
0
0
~e
~E
8 ~ 13. n" 0_
' 0 0.2 0.4 0.6 0.8 1.0 ADP (mM)
0 0.2 0.4 0.6 0.81.0 ADP (mM)
0
0 0.2 0.4 0.6 0.81.0 ADP (raM)
Fig. 4. ADP inhibition of chimeric and mutated enzymes. Standard assay conditions (0.4 mM ATP, 0.2 mM ribose 5-phosphate, 4 mM MgCI2, 10 mM potassium phosphate buffer (pH 7.4)) were used. ( O ) PRS I, ( O ) PRS II, (zx) II.11.I, ( • ) 1.11.II, ( [ ] ) IK152Q, ( • ) IIQ152K, ( [ ] ) 1.228.II, (half-filled circle) 1.110.II, and ( ~ ) 1.53.II.
132
1. Ahmad et al. / Biochimica et Biophysica Acta 1207 (1994) 126-133
I only by 2 residues in the C-terminal region, the result shows that changes of Arg-242 to Lys a n d / o r of Cys-264 to Ala are responsible in reducing the sensitivity to inhibition. The inhibition of 1.110.II was similar to that of 1.228.II, but the inhibition of 1.53.II was much weaker. The findings indicate that the distinctive residues between 5 4 110 of PRS I, namely, Val-55 a n d / o r Ala-81, also contribute to the inhibition of GDP. For the inhibition by ADP (Fig. 4), all the enzymes were strongly inhibited at 1 mM, but at 0.3 mM, the pattern of inhibition differed with each enzyme. The inhibition of PRS I and PRS II by 0.3 mM ADP was 95% and 55 %, respectively, while the inhibition of II. 11.I and I. 11.II was 31% and 43%, respectively (Fig. 4A). The lower sensitivity of II. 11.I to ADP inhibition, compared with that of PRS I, indicates that two residues at positions 4 and 5 of PRS I contribute to the strong ADP inhibition. The extent of inhibition of mutant IK152Q was 73%, compared with the 95% inhibition of PRS I (Fig. 4B). Decrease in the extent indicates the contribution of Lys-152 of PRS I to the strong ADP inhibition, however, the contribution is less than that seen with Lys-4. The extent of inhibition of PRS II and IIQ152K was similar (55% and 59%, respectively). As for the chimeric enzymes, the extent of inhibition of the 1.228.I1, 1.110.II, 1.53.I1 was 80%, 58% and 61%, respectively and the extent of inhibition of PRS II, and chimeric enzymes 1.110.II and 1.53.I1 was much the same. These findings indicate that amino-acid residues between 111-228 and 229-317 of PRS I also contribute to the higher sensitivity to ADP inhibition. ADP inhibition of mammalian PRPP synthetases was reported to be competitive with respect to ATP [27,28] and GDP inhibits noncompetitively. ADP also inhibits bacterial PRPP synthetases, and evidence indicated that ADP binds to an allosteric site as well as competitively with ATP to the active site [29,30]. Although the ADP binding site has been examined carefully only with the PRPP synthetase from S. typhimurium, all PRPP synthetases appear to contain an allosteric inhibitor site which binds ADP. No structural information about the allosteric sites for ADP and GDP had been available. Recently, Roessler et al. reported that mutant human PRS I with a single replacemnt of A s n - l l 3 by Ser or Asp-182 by His is resistant to inhibition by both ADP and GDP [31]. In our studies, the patterns of ADP inhibition for the chimeric and mutated enzymes were similar to those for GDP inhibition; only a small distinction between the ADP and GDP inhibition was observed with 1.110.II; in the GDP inhibition, the extent of inhibition of 1.110.II Was similar to that of 1.228.II, while the ADP inhibition of 1.110.II was weaker than that of 1.228.II. This similarity in the ADP and GDP inhibition patterns supports the hypothesis that ADP and GDP bind to the same site of the enzyme [12]. Furthermore, the extent of ADP inhibition of II.11.I was significantly smaller than that of PRS I, while the K m value for ATP of II.11.I was similar to that of PRS I (50 vs. 44/xM
[19], respectively). The findings are not consistent with the view that ADP inhibits the enzyme by simple steric interference with ATP binding, but rather suggest that it acts at an allosteric site. The mechanism by which ADP inhibits the mammalian PRPP synthetases remains to be elucidated. Regarding the structure-function relationships of PRPP synthetase, Bower et al. characterized the E. coli mutant enzyme to identify a divalent cation-nucleotide binding site [32], and Harlow and Switzer used the active site-directed reagent 5'-(p-fluorosulfonylbenzoyl)adenosine to identify an active-site His residue of the S. typhimurium enzyme [33]. We have now constructed the chimeric enzymes between rat PRS I and PRS II; construction of chimeric enzymes is useful for structure-function analyses of similar proteins. The amino-acid sequences of rat PRS I and PRS II are 96% identical and this high sequence homology allows for the generation of catalytically active chimeras that maintain conformational integrity. The resuits obtained with several chimeric enzymes and the mutated enzymes with a single point mutation identified amino-acid residues contributing to different properties of the enzyme isoforms.
Acknowledgments We thank M. Ohara for critical comments. This study was supported in part by grants from the Ministries of Education, Science, and Culture and of Health and Welfare of Japan.
References [1] Kornberg, A., Lieberman, I. and Simms, E.S. (1955) J. Biol. Chem. 215, 417-427. [2] Flaks, J.G., Erwin, M.J. and Buchanan, J.M. (1957) J. Biol. Chem. 228, 201-213. [3] Fox, I.H. and Kelly, W.N. (1971) Ann. Intern. Med. 74, 424-433. [4] Holmes, E.W.(1978) in Handbook of Experimental Pharmacology (Kelly, W.N. and Weiner, I.M., eds.), Vol. 51, pp. 21-41, Springer, Heidelberg. [5] Bagnara, A.S. and Finch, L.R. (1974) Eur. J. Biochem. 41, 421-430. [6] Tatibana, M. and Shigesada, K. (1972) J. Biochem. (Tokyo) 72, 549-560. [7] Tatibana, M. (1978) in Handbook of Experimental Pharmacology (Kelly, W.N. and Weiner, I.M., eds.), Vol. 51, pp. 125-154, Springer, Heidelberg. [8] Keppler, D. and Holstege, A. (1982) in Metabolic Compartmentation (Sies, H., ed.), pp. 147-203, Academic Press, London. [9] Schubert, K.R., Switzer, R.L. and Shelton, A. (1975) J. Biol. Chem. 250, 7492-7500. [10] Switzer, R.L. and Gibson, K.J. (1978) Methods Enzymol. 51, 3-11 [11] Hove-Jensen, B., Harlow, K.W., King, C.J. and Switzer, R.L. (1986) J. Biol. Chem. 261, 6765-6771. [12] Arnvig, K., Hove-Jensen, B. and Switzer R.L. (1990) Eur. J. Biochem. 192, 195-200.
L Ahmad et al./Biochimica et Biophysica Acta 1207 (1994) 126-133
[13] Fox, I.H. and Kelly, W.N. (1971) J. Biol. Chem. 246, 5739-5748. [14] Roth, D.G., Shelton, E. and Deuel, T.F. (1974) J. Biol. Chem. 249, 291-296. [15] Kita, K., Otsnki, T., Ishizuka, T. and Tatibana, M. (1989) J. Biochem. (Tokyo) 105, 736-741. [16] Kita, K., Ishizuka, T., Ishijima, S., Sonoda, T. and Tatibana, M. (1994) J. Biol. Chem. 269, 8334-8340. [17] Taira, M., Ishijima, S., Kita, K., Yamada,K., Iizasa,T. and Tatibana, M. (1987) J. Biol. Chem. 262, 14867-14870. [18] Taira, M., lizasa, T., Yamada, K., Shimada, H. and Tatibana, M. (1989) Biochim. Biophys. Acta 1007, 203-208. [19] Ishijima, S., Kita, K., Ahmad, I., Ishizuka, T., Taira, M. and Tatibana, M. (1991) J. Biol. Chem. 266, 15693-15697. [20] Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular cloning: A laboratory Manual, 2nd Edn., Cold Spring Harbor Press, Cold Spring Harbor, New York. [21] Gliemann, J. (1967) Diabetologia 3, 382-388. [22] Kornberg, A., Liebermann, I. and Simms, E.S. (1955) J. Biol. Chem. 215, 389-402.
133
[23] Bradford, M.M. (1976) Anal. Biochem. 72, 248-254. [24] Laemmli, U.K.(1970) Nature 227, 680-687. [25] Towbin, H., Staehelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. 76, 4350-4354. [26] Becker, M.A., Raivio, K.O. and Seegmiller, J.E. (1979) Adv. Enzytool. Relat. Areas Mol. Biol. 49, 281-306. [27] Fox, I.J. and Kelly, W.N. (1972) J. Biol. Chem. 247, 2126-2131 [28] Becker, M.A., Kostel, P.J. and Meyer, L.J. (1975) J. Biol. Chem. 250, 6822-6830 [29] Switzer, R.L. and Sogin, D.C. (1973) J. Biol. Chem. 248, 1063-1073 [30] Gibson, K.J., Schubert, K.R. and Switzer, R.L. (1982) J. Biol. Chem. 257, 2391-2396 [31] Roessler, B.J., Nosal, J.M., Smith, P.R., Heidler, S.A., Palella, T.D., Switzer, R.L. and Becker, M.A. (1993) J. Biol. Chem. 268, 2647626481 [32] Bower, G.S., Harlow, K.W., Switzer, R.L. and Hove-Jensen, B. (1989) J. Biol. Chem. 264, 10287-10291 [33] Harlow, K.W. and Switzer, R.L. (1990) J. Biol. Chem. 265, 54875493.
i!,J ELSEVIER
Biochimica et BiophysicaActa 1207 (1994) 126-133
Biochi~ic~a et BiophysicaA~ta
Identification of amino-acid residues linked to different properties of phosphoribosylpyrophosphate synthetase isoforms I and II Imtiaz Abroad, Sumio Ishijima, Kazuko Kita, Masamiti Tatibana * Department of Biochemistry, Chiba UniversitySchool of Medicine, Inohana 1-8-1, Chuo-ku, Chiba 260, Japan Received 17 December 1993; revised 1 April 1994
Abstract The catalytic subunit of rat liver phosphoribosylpyrophosphate synthetase is composed of two isoforms, PRS I and PRS II. The amino-acid sequences differ only by 13 residues, out of which two Lys residues of PRS I at positions 4 and 152 give net additional positive charges to PRS I. Previous work has shown that PRS I is more sensitive to inhibition by ADP and GDP and more stable to heat treatment than is PRS II. To identify amino-acid residues responsible for the different properties, five chimeric enzymes between rat PRS I and PRS II and two mutated enzymes with a single point mutation at position 152 were constructed; these enzymes were produced in Escherichia coli. Changing Lys-4 of PRS I to Val, together with Ile-5 to Leu, completely abolished sensitivity to GDP inhibition of PRS I, indicating that Lys-4 in PRS I is critical for GDP inhibition. The substitutions at position 152 had little effect on GDP inhibition. Characterization of the chimeric enzymes revealed that residues between residues 54-110 and 229-317, namely, Val-55 and/or Ala-81, and Arg-242 and/or Cys-264 of PRS I also contribute to the strong GDP inhibition. Lys-4 was also important for the strong ADP inhibition of PRS I. Regarding the physical properties, chimeric enzymes bearing residues 12-53 of PRS I were stable at 49°C and with digestion with papain and proteinase K. Our observations suggest that Lys-17, Ile-18, and/or Cys-40 of PRS I contribute to stability of the enzyme. Key words: Phosphoribosylpyrophosphate synthetase; Isoform; Chimeric enzyme; Nucleotide inhibition
I. Introduction P h o s p h o r i b o s y l p y r o p h o s p h a t e (PRPP) synthetase (ATP:D-ribose-5-phosphate pyrophosphotransferase, EC 2.7.6.1) catalyzes the formation of PRPP from ATP and ribose 5-phosphate. PRPP provides an important substrate for synthesis of almost all nucleotides [1,2], and is a critical control factor for synthesis de novo of purines [3-5] and pyrimidines [6-8]. PRPP synthetase has been purified from Salmonella typhimurium [9,10], Escherichia coli [11], Bacillus subtilis [12], human erythrocyte [13], and rat liver [14,15]. The rat
Abbreviations: PRPP, 5-phosphoribosyl 1-pyrophosphate;PRS I and PRS II, phosphoribosylpyrophosphatesynthetase 34-kDa subunits I and II, respectively; kb, kilobase pairs; HPLC, high-performanceliquid chromatography; PAGE, polyacrylamidegel electrophoresis; IPTG, isopropylfl-D-thiogalactopyranoside; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid. * Corresponding author. Fax: + 81 43 2262041. 0167-4838/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0167-4838(94)00065-O
liver enzyme exists as complex aggregates of 34-, 39- and 41-kDa components [16] (39- and 41-kDa components were formerly reported to be 38 and 40 kDa, respectively [15]), with the 34-kDa species being the catalytic subunit [15]. The cloning of rat c D N A and amino-acid sequencing of the catalytic subunit purified from the rat liver [17] revealed that the 34-kDa component is a mixture of two highly homologous isoforms, PRS I and PRS II, The deduced amino-acid sequences of both sets of 317 residues differ only by 13 residues, out of which 6 are non-conservative substitutions [17]. PRS I and PRS II are encoded by separate genes which have been mapped on distinct regions of the X chromosome. Rat mRNAs of the genes are expressed in all tissues but their relative expression is variable [18]. Characterization of the recombinant rat PRS I and PRS II expressed in E. coli [19] showed different properties between the two isoforms. PRS I is more sensitive to nucleotide inhibition than is PRS II. Inhibition of PRS I and PRS II by 0.3 mM A D P was 87% and 54%, respectively, and inhibition by 1 mM GDP was 93% and 24%, respectively. PRS II was 180-fold more sensitive to
L Ahmadet aL/Biochimica et BiophysicaActa 1207 (1994) 126-133 heat inactivation at 49°C than was PRS I. Regarding the structure-function relationship, the two substitutions of Lys for Val and Lys for Gin at positions 4 and 152 warrant attention. These two substitutions give net additional positive charges to PRS I. It is possible that the different sensitivity to nucleotide inhibition is due to either or both of the substitutions. To better comprehend the structure-function relationship between the two isoforms, we constructed five chimeric enzymes and two mutated enzymes with a single point mutation at position 152. We describe here the purification and characterization of the chimeric and mutated enzymes expressed in E. coli cells. On the basis of the results obtained, we propose that Lys-4 of PRS I is critical for GDP inhibition but at least two other residues also contribute to the strong GDP inhibition of PRS I. Regarding the physical properties, Lys-17, Ile-18 a n d / o r Cys-40 of PRS I play an important role in the thermal stability and resistance to proteolytic inactivation of the enzyme.
(A)
Sacl
eas, I
and
Accl Hindlll
I
I
I
'
"
I
Kpnl
I I
I
PRS III
(B)
"K'I
PRS I
I
"K'81~G~Vs'A~'2~'95"~'T ''D ='ZR " C
1
VL
I Rk, S I
I S QT
I I
E
] K
A
PRS II 1.228.11
2 2 8 VIlli/Ill/IlIA 11 o
1.110.11 1.53.11
1.11.11 2. Materials
127
F/////////////////////////////////////A 1
methods
2.1. Bacterial strains and plasmids E. coli strains MV1304 and AD8 were used for cloning and expression experiments. MV1304 contains the lacI q repressor and AD8 lacks the Ion proteinase. The expression vector pG1KHB [19] is derived from pKK233-2 (Pharmacia) and was used to express the chimeric and mutated enzymes.
I1.11.1 ~ / A l l [ Fig. 1. Construction of chimeric enzymes between rat PRS I and PRS II. (A) Common restriction enzyme sites of PRS I and PRS II cDNAs. Dotted lines indicate the newly created sites. (B) The upper two open bar and hatched box show the structures of PRS I and PRS II, respectively, with the 13 different amino-acid residues between PRS I and PRS II and their positions. The lower five open and hatched boxes show structures of the chimeric enzymes constructed between PRS I and PRS II. Open bars and hatched boxes show regions derived from PRS I and PRS II, respectively. For example, in 1.228.II, amino-acid residues 1-228 are derived from PRS I and the rest C-terminal residues are derived from PRS II.
2.2. Enzymes and other materials Restriction and D N A modifying enzymes were obtained from Takara Shuzo, Toyobo and New England Biolabs and were used according to the manufacturer's directions. Site-directed mutagenesis was carried out using the Amersham oligonucleotide-directed in vitro mutagenesis kits. TSK G3000SW and G4000SW columns were purchased from Tosoh Manufacturing. Sources of all other reagents were as described elsewhere [15]. 2.3. Plasmid construction of chimeric PRS I / PRS H cDNAs for chimeric enzymes were constructed by connecting fragments of the cDNAs for PRS I and PRS II. A restriction enzyme site common to the two cDNAs at the same locus was used for splicing c D N A fragments. For this purpose, new sites were introduced into the cDNAs by oligonucleotide-directed mutagenesis (Fig. 1A). An antisense synthetic 19-mer 5 ' - A T G A C A A A T G G T A C C A C A A-3' was used to create a site for KpnI in PRS I c D N A and antisense 19-mers 5 ' - G T G A G A G C T C C C G C T G A A G-3' and 5 ' - C T G G A T G A T G T A G A C A T C T - 3 ' were used
to create unique SacI and AccI sites, respectively, in PRS II cDNA. The new restriction enzyme sites created were confirmed by mapping with restriction enzymes. We constructed five chimeric cDNAs by interchanging appropriate fragments of the two cDNAs, as shown in Fig. lB. Junctional sites between PRS I and PRS II in the chimeras were identified by numbers referring to the last amino-acid residue of the N-terminal region. For example, in the chimeric enzyme 1.228.11, amino-acid residues from 1 to 228 are derived from PRS I and remaining residues are derived from PRS II. After ligation of the appropriate c D N A fragments, the products were cloned into the vector pG1KHB. The resulting expression plasmids contained the complete coding region of 954 bp and 3'-noncoding region (0.24 kb for PRS I c D N A or 1.1 kb for PRS II cDNA), under control of the trc promoter. The plasmids were introduced into E. coli MV1304 cells, except for the plasmid for the chimeric enzyme 1.11.II, which was introduced into AD8 cells. To introduce a single point mutation of Lys-152 to Gin in PRS I, an antisense 19-mer synthetic oligonucleotide 5 ' - G A A T C C A C T G C A G G A C A G C - 3 ' was
128
I. Ahmadet al. /Biochimica et BiophysicaActa 1207 (1994) 126-133
used. This mutagenesis created a new site for PstI. Another antisense 19-mer synthetic oligonucleotide 5'-GAATCCACITAAGAACTGC-3' was used to introduce Gln-152 to Lys mutation in PRS II. This mutagenesis created a new site for AfllI. New restriction sites created were verified by restriction mapping. These plasmid constructs were introduced into cells of E. coli strain MV1304. The mutant PRS I bearing a single replacement of Lys-152 to Gin, is referred to as IK152Q and the mutant PRS II bearing a single replacement of Gln-152 to Lys, is referred to as IIQ152K. 2.4. Cell growth The transformed cells were grown at 30°C for 15 h in 1.5-3 1 of M9 medium [20] containing 1 mM thiamine and ampicillin (50 /zg/ml). After adding 1 mM IPTG, the cells were grown for an additional 3 h at 30°C and then harvested by centrifugation. 2.5. Purification of chimeric and mutated enzymes All procedures were carried out at 4°C and all buffers used contained 0.3 mM ATP, 6 mM MgC12, 0.1 mM EDTA and 2.5 mM 2-mercaptoethanol as enzyme stabilizing agents [13,14] unless otherwise stated. The pH values indicated were measured at 4°C. The E. coli cells were suspended in a minimum volume of a solution containing 50 mM potassium phosphate (pH 7.4) and 0.1 mM phenylmethylsulfonyl fluoride (buffer A) and lysed by treatment with egg white lysozyme (1 m g / m l ) for 20 min. The mixture was then sonicated with Sonifier Cell Disrupter 185 (Branson Sonic Power, USA), using a standard microtip. The output control of the sonifier was placed at a setting of 4 and the cells were disrupted by sonic treatment, 5-times each for 15 s. The cell debris was sedimented at 10 000 X g for 15 min. The protein concentration of the cellular extracts ranged from 5 to 10 mg/ml. All subsequent steps were as described previously [15] but with the following modifications: (a) The IK152Q and IIQ152K proteins in the cellular extracts were precipitated with 5% ( w / v ) polyethylene glycol 6000. The precipitates were dissolved in buffer A and the protein concentration was adjusted to 3.3 m g / m l and the enzymes were then precipitated at pH 6.1. The acid precipitates of the mutated enzymes were dissolved in buffer A and pH was adjusted to 7.4. (b) Chimeric enzymes 1.11.II, 1.228.1I, 1.110.II were precipitated with 6.6% ( w / v ) of polyethylene glycol 6000 and the precipitates were dissolved in buffer A. The protein concentration was adjusted to 3.3 mg/ml. 1.11.II and 1.228.1I were then precipitated at pH 5.7 while 1.110.II was precipitated at pH 6.2. The acid precipitates were dissolved in buffer A and pH was adjusted to 7.4. (c) For purification of II.11.I, the cellular extract was treated with 0.8% streptomycin sulfate, then the enzyme was precipitated with 45%-saturated ammonium sulfate. The precipitate
was dissolved in buffer A and the protein concentration was adjusted to 3.3 mg/ml. The enzyme was precipitated at pH 4.9, dissolved in buffer A and the pH was adjusted to 7.4. (d) For purification of 1.53.I1, the cellular extract was treated with 0.8% streptomycin sulfate, and then with 90%-saturated ammonium sulfate. The precipitate was then dissolved in buffer A, and the solution was applied to a TSK Gel 3000SW column (0.75 X 60 cm) equilibrated with 50 mM potassium phosphate (pH 7.4). The column was eluted with the same buffer at a flow rate of 0.8 ml/min, and active fractions were pooled. The above preparations of the chimeric and mutated enzymes were stored in small portions at - 8 0 ° C until their properties were examined. 2.6. Enzyme assay The enzyme activity was assayed using the two methods described elsewhere [15]. Briefly, method 1 for a crude enzyme preparation, we used a reaction mixture (0.5 ml) containing 50 mM Hepes-KOH (pH 7.4 at 37°C), 10 mM potassium phosphate, 4 mM MgC12, 1 mM EDTA, 0.8 mM ATP, 0.2 mM ribose 5-phosphate, 1 mM dithiothreitol, 1.2 mM phosphoenolpyruvate, pyruvate kinase (9 units), and the enzyme, unless otherwise stated. The amount of PRPP synthesized was measured by a modification [21] of the enzymatic method of Kornberg et al. using labeled orotic acid [22]. Method 2, used for preparations after the acid precipitation step, measured ribose 5-phosphate dependent [14C]AMPproduction from [14C]ATP. The composition of the assay mixture (0.1 ml) was identical with that of the mixture for method 1, except that the unlabeled ATP was replaced with 0.4 mM [8-14C]ATP, and phosphoenolpyruvate and pyruvate kinase were omitted. One unit of enzyme activity was defined as the amount catalyzing formation of 1 /zmol of PRPP/min, under standard conditions. 2.7. Protein concentration Protein concentration was determined by the method of Bradford [23], using bovine serum albumin as a standard. For determination of specific activities of the chimeric and mutated enzymes, each purified preparation was applied to SDS-PAGE and the amount of the 34-kDa protein was determined based on densitometric analysis of the gel stained with Coomassie brilliant blue R-250. 2.8. Molecular weight determination by gel filtration The molecular weight of the chimeric and mutant enzymes was determined by gel filtration using TSK G4000SW HPLC column (0.75 X 60 cm). The running buffer was 50 mM potassium phosphate (pH 7.4) containing the enzyme stabilizing agents.
I. Ahmad et al. / Biochimica et BiophysicaActa 1207 (1994) 126-133 2.9. Other methods
SDS-PAGE was performed by the method of Laemmli [24]. In western blot analysis, proteins were separated by SDS-PAGE and transferred electrophoretically to a polyvinylidine difluoride membrane [25]. Protein was immunodetected using rabbit antiserum raised against rat liver PRPP synthetase and anti-rabbit IgG goat antibody conjugated with horseradish peroxidase (E.Y. Labs, USA). Color was developed with 0.8 m g / m l of diaminobenzidine tetrahydrochloride and 0.009% H 2 0 2 .
3. Results and discussion 3.1. Expression of chimeric PRS I / enzymes in E. coli
PRS II and mutated
We constructed five chimeric cDNAs using convenient restriction sites (Fig. 1B) and two cDNAs for the mutated enzymes at position 152. The plasmids carrying the cDNAs for chimeric and mutated enzymes were introduced into the host cells of E. coli. As the chimeric enzyme 1.11.II from the MV1304 transformants was not obtained in a sufficient amount, the strain AD8 which lacks the Ion proteinase was used for this chimeric enzyme. The cells were grown at 30°C in M9 medium to minimize expression of the constitutive E. coli enzyme [19]. Expression of the chimeric and mutated enzymes was induced by IPTG. The enzyme activity in soluble fractions of the crude cellular extracts increased after the addition of IPTG and reached maximum in 3 - 4 h. At maximal levels, the enzyme activities of the extracts of the cells containing the chimeric and mutated plasmids were from 4- to 60-fold higher than that from the control MV1304 or AD8 cells containing the vector alone: 1.228.11, 10.5-fold; 1.110.II, 24-fold; 1.53.I1, 4.5-fold; 1.11.II, 23-fold; II.11.I, 5.6-fold; IK152Q, 48-fold; IIQ152K, 60-fold. The elevated enzyme activity in cells carrying the cDNAs suggested that the chimeric and mutated enzymes had catalytic activity, which was expressed. 3.2. Purification of chimeric and mutated enzymes
Since homology of the deduced amino-acid sequences of the E. coli and the recombinant rat enzymes is high (47%), separation could be difficult. Therefore, E. coli cells containing the plasmids were grown under conditions of culture that would repress the synthesis of bacterial PRPP synthetase [19]. The chimeric and mutated enzymes were purified from E. coli MV1304 and AD8 transformants. The procedure for chimeric enzymes 1.228.1I, 1.110.II, 1.11.II and mutated IK152Q and IIQ152K included fractionation with polyethylene glycol and acid precipitation. As the other two chimeric enzymes, 1.53.11 and I1.11.I were not precipitated with polyethylene glycol,
129
the cellular extracts were treated with streptomycin, after which the enzymes were precipitated with ammonium sulfate. 1.53.II was further subjected to gel filtration on TSK Gel 3000SW. II.11.I was further purified by acid precipitation at pH 4.9. The E. coli enzyme expressed in a small amount was separated in these purification steps; in a separate purification of the E. coli enzyme, the bacterial enzyme showed behaviour which differed from that of the chimeric and mutated enzymes seen at steps of polyethylene glycol precipitation and acid precipitation. While we can not completely rule out the possibility of contamination by the E. coli enzyme, the amount of the contaminated E. coli enzyme should be limited. The expression levels of the chimeric enzymes 1.53.II and II.11.I in the cells were relatively low, but the physical properties of the partially purified enzymes were quite different from that of the E. coli enzyme. The partially purified bacterial enzyme was unstable at 49°C at a protein concentration of 28 ~ g / m l . After 10 min of incubation, below 10% of the initial activity remained. Except for the chimeric 1.11.II, all other enzymes were much more stable (see Fig. 2A). Specific activities of the purified chimeric and mutated enzymes, calculated as milliunit/mg of 34-kDa protein were as follows: 1.228.II, 12500; 1.110.II, 13000; 1.53.11, 12800; 1.11.II, 16900; II.11.I, 8500; IK152Q, 10400; IIQ152K, 14 400. These values are significantly lower than values of the specific activities of PRS I (25 700) and PRS II (34 500) [19]. 3.3. Characterization of chimeric and mutated enzymes
Physical and kinetic properties of the chimeric and mutated enzymes were examined and compared. The mammalian PRPP synthetases exist in multiple aggregated forms [26]. To examine whether the purified chimeric and mutated enzymes also existed in aggregated forms, each enzyme was applied to HPLC gel filtration on TSK 4000SW. All these enzymes did elute in aggregated forms at the position of a molecular mass ranging from 300 to 1100 kDa. We determined the apparent Michaelis constants for ATP and ribose 5-phosphate of the partially purified chimeric and mutated enzymes. The K m values for ATP of the enzymes are as follows (in /~M): 1.228.II, 17; 1.110.II, 25; 1.53.1I, 51; 1.11.II, 45; II.11.I, 50; IK152Q, 23; IIQ152K, 42, and the V / K m values ranged from 6-25 min-1/xM -1. The values of K m and V / K m for ribose 5-phosphate of these enzymes were within the range of 25-50 /zM and 9-29 min-1/zM -1, respectively. These values indicated that the chimeric and mutated enzymes retained functioning ATP and ribose 5-phosphate binding sites and catalytic efficiencies of these enzymes, as represented by V//Km, w e r e 30-125% of those of native PRS I and PRS II. Regarding physical properties of the isoforms, PRS I is far more stable than PRS II to heat inactivation [19]. To
I. Ahmad et aL / Biochimica et Biophysica Acta 1207 (1994) 126-133
130
ascertain which amino-acid residues contribute to the stability of PRS I, heat inactivation of the chimeric and mutated enzymes was examined (Fig. 2A). 1.228.1I, 1.110.II and 1.53.11, which retain a considerable part of the Nterminal region of PRS I, were fairly stable when incubated for 60 min at 49°C. The residual activities were 65%, 61% and 77%, respectively. In contrast, 1.11.II showed residual activity of only 0.6% when treated similarly. This enzyme has three substitutions compared with 1.53.II, i.e., Lys-17 to Arg, Ile-18 to Val and Cys-40 to Ser. II.11.I, which retains Lys-17, Ile-18 and Cys-40, showed a stability similar to those of 1.228.1I, 1.110.II and 1.53.II. These results indicate that at least one of the three residues, Lys-17, Ile-18 and Cys-40, is important for the thermostability of PRS I.
1.228.11, 1.110.II and 1.53.II. Therefore, at least one of the three residues, Lys-17, Ile-18 and Cys-40, contributes to resistance to the proteinase inactivation of PRS I. The finding that 1.11.II was readily inactivated by proteinase is consistent with the result that the yield of 1.11.II production was poor in the E. coli strain MV1304, but was increased in the strain AD8 which lacks the Ion proteinase. Different sensitivities to the proteinase inactivation were also observed between PRS II and 1.11.II. PRS II was more susceptible to inactivation by papain than was I. 11.II, but was more resistant to inactivation by proteinase K. The results shown in Fig. 2 show a correlation between thermostability and resistance to proteolytic inactivation. A remarkable difference was evident between chimeric 1.53.II and 1.11.II regarding the response to heat and proteinase treatment. Among the three residues involved in the difference (positions 17, 18 and 40) the substitution of Cys vs. Ser at the position 40 is notable, as substitutions at the positions 17 and 18, namely, Lys vs. Arg and Ile vs. Val, are conservative. Cys-40 might play an important role in the stability of PRS I. It is interesting that Cys-40 is replaced by Ser in the E. coli PRPP synthetase [11], which was susceptible to heat inactivation. Further studies with site directed mutagenesis will elucidate a definitive role of Cys-40.
3.4. Proteinase digestion A remarkable difference was found between the sensitivity to inactivation by proteinase digestion of PRS I and PRS II. Activity of PRS II was lost more rapidly than was that of PRS I in case of treatment with papain and proteinase K at 25°C for 4 h (Fig. 2B and 2C). To identify which residues of PRS I contribute to the resistance to proteinase inactivation, all the chimeras were treated with papain and proteinase K; similar responses were observed with this treatment. 1.228.II, 1.110.II and 1.53.II were resistant to proteinase inactivation and sensitivities were similar to that of PRS I. In contrast, 1.11.II showed a much lower residual activity when treated similarly. II.11.I was resistant to proteinase inactivation to a similar degree to
3.5. Inhibition by nucleotides ADP and GDP are the most potent inhibitors of PRPP synthetases purified from various sources [12]. We reported different sensitivities of PRS I and PRS II to the
A Heat stability
Enzyme
B Stability to papain inactivation
C Stability to proteinase K inactivation
il
m
1.228.11 1
1.110.111
~///////~
m []
1.53.,I I. 11., I
,. l.IH
I
m
m ,
,
20
40
,, 60
n
. . . . 0 100
Residual Activity (%)
0
20
, 40
60
. . . . . . .
0 100
Residual Activity (%)
20 40 80 80 100120
Residual Activity (%)
Fig. 2. Inactivationby heat and protease treatment of the native and chimeric enzymes. Structures of the enzymes are shown on the left. Open and hatched boxes show regions derived from PRS I and PRS II, respectively. (A) Heat stability of the enzymes. Residual enzyme activity was assayed after incubating at 49°C for 1 h at a protein concentrationof 28 /xg/ml in 50 mM potassium phosphate buffer (pH 7.4) containing the enzyme stabilizing agents. (B) and (C) Resistance to inactivation by papain and proteinase K, respectively. Residual enzyme activity was assayed after the enzymes were treated with the respective proteinase at a ratio of 50:1 (w/w) for 4 h at 25°C.
131
L Ahmad et al./Biochimica et Biophysica Acta 1207 (1994) 126-133
A
B i
i
i
i
i
C i
\
i
i
i
~
,
i
,
,
,
" ql
~ '~ ° 40I
"~
1~40I
, 0 o0.2 0.4 0.6 0.8 .0 GDP (mM)
~X,,xx~
o
l~
, , , , ,0 0.2 0.4 0.6 0.81.0 GDP (mM)
40I
o,
, , , , , 0 0.2 0.4 0.6 0.8 1.0 GDP (raM)
Fig. 3. GDP inhibition of chimeric and mutated enzymes. Standard assay conditions (0.4 mM ATP, 0.2 mM ribose 5-phosphate, 4 mM MgC12, 10 mM potassium phosphate buffer, pH 7.4) were used. ( O ) PRS I, ( O ) PRS II, (zx) II.11.I, ( • ) 1.11.II, ( [ ] ) IK152Q, ( • ) IIQ152K, ( [ ] ) 1.228.II, (half-filled circle) 1.110.II, and ( ~ ) 1.53.II.
inhibition by ADP and GDP [19]. To identify the aminoacid residues contributing to the difference, inhibition of the chimeric and mutated enzymes by ADP and GDP were examined (Figs. 3 and 4). The most notable finding was that the chimeric enzyme II.11.I was inhibited only 1% by 1 mM GDP, whereas PRS I was inhibited by 91% (Fig. 3A). Only 2 residues are different between II.11.I and PRS I, Val vs. Lys at the position 4 and Leu vs. lie at the position 5. Since substitution at position 5 is conservative, its effect on function of the enzyme is probably limited. Thus, the finding strongly suggests that Lys-4 in PRS I is critical for the sensitivity to GDP inhibition. However, the
A
extent of inhibition of PRS 1I and the chimeric 1.11.II is small, thereby showing that merely changing Val-4 to Lys in PRS II will not induce a strong GDP inhibition. Regarding the function of Lys-152 in PRS I, the inhibition of IK152Q was only slightly weaker than that of PRS I (Fig. 3B). The extent of GDP inhibition of IIQ152K was similar to that of PRS II. These results indicate that Lys-152 is not a critical factor in the PRS I sensitivity to GDP inhibition. In analyses of other chimeric enzymes, the extent of inhibition of the chimeric enzyme 1.228.1I was significantly smaller than that of PRS I (Fig. 3C). As the sequence of this chimeric enzyme differs from that of PRS
B ,
,
,
i
C
f
100;
00
)0
._~ 80 "5
80
30
60
60
60
4o
40
40
2O
20
20
0
0
~e
~E
8 ~ 13. n" 0_
' 0 0.2 0.4 0.6 0.8 1.0 ADP (mM)
0 0.2 0.4 0.6 0.81.0 ADP (mM)
0
0 0.2 0.4 0.6 0.81.0 ADP (raM)
Fig. 4. ADP inhibition of chimeric and mutated enzymes. Standard assay conditions (0.4 mM ATP, 0.2 mM ribose 5-phosphate, 4 mM MgCI2, 10 mM potassium phosphate buffer (pH 7.4)) were used. ( O ) PRS I, ( O ) PRS II, (zx) II.11.I, ( • ) 1.11.II, ( [ ] ) IK152Q, ( • ) IIQ152K, ( [ ] ) 1.228.II, (half-filled circle) 1.110.II, and ( ~ ) 1.53.II.
132
1. Ahmad et al. / Biochimica et Biophysica Acta 1207 (1994) 126-133
I only by 2 residues in the C-terminal region, the result shows that changes of Arg-242 to Lys a n d / o r of Cys-264 to Ala are responsible in reducing the sensitivity to inhibition. The inhibition of 1.110.II was similar to that of 1.228.II, but the inhibition of 1.53.II was much weaker. The findings indicate that the distinctive residues between 5 4 110 of PRS I, namely, Val-55 a n d / o r Ala-81, also contribute to the inhibition of GDP. For the inhibition by ADP (Fig. 4), all the enzymes were strongly inhibited at 1 mM, but at 0.3 mM, the pattern of inhibition differed with each enzyme. The inhibition of PRS I and PRS II by 0.3 mM ADP was 95% and 55 %, respectively, while the inhibition of II. 11.I and I. 11.II was 31% and 43%, respectively (Fig. 4A). The lower sensitivity of II. 11.I to ADP inhibition, compared with that of PRS I, indicates that two residues at positions 4 and 5 of PRS I contribute to the strong ADP inhibition. The extent of inhibition of mutant IK152Q was 73%, compared with the 95% inhibition of PRS I (Fig. 4B). Decrease in the extent indicates the contribution of Lys-152 of PRS I to the strong ADP inhibition, however, the contribution is less than that seen with Lys-4. The extent of inhibition of PRS II and IIQ152K was similar (55% and 59%, respectively). As for the chimeric enzymes, the extent of inhibition of the 1.228.I1, 1.110.II, 1.53.I1 was 80%, 58% and 61%, respectively and the extent of inhibition of PRS II, and chimeric enzymes 1.110.II and 1.53.I1 was much the same. These findings indicate that amino-acid residues between 111-228 and 229-317 of PRS I also contribute to the higher sensitivity to ADP inhibition. ADP inhibition of mammalian PRPP synthetases was reported to be competitive with respect to ATP [27,28] and GDP inhibits noncompetitively. ADP also inhibits bacterial PRPP synthetases, and evidence indicated that ADP binds to an allosteric site as well as competitively with ATP to the active site [29,30]. Although the ADP binding site has been examined carefully only with the PRPP synthetase from S. typhimurium, all PRPP synthetases appear to contain an allosteric inhibitor site which binds ADP. No structural information about the allosteric sites for ADP and GDP had been available. Recently, Roessler et al. reported that mutant human PRS I with a single replacemnt of A s n - l l 3 by Ser or Asp-182 by His is resistant to inhibition by both ADP and GDP [31]. In our studies, the patterns of ADP inhibition for the chimeric and mutated enzymes were similar to those for GDP inhibition; only a small distinction between the ADP and GDP inhibition was observed with 1.110.II; in the GDP inhibition, the extent of inhibition of 1.110.II Was similar to that of 1.228.II, while the ADP inhibition of 1.110.II was weaker than that of 1.228.II. This similarity in the ADP and GDP inhibition patterns supports the hypothesis that ADP and GDP bind to the same site of the enzyme [12]. Furthermore, the extent of ADP inhibition of II.11.I was significantly smaller than that of PRS I, while the K m value for ATP of II.11.I was similar to that of PRS I (50 vs. 44/xM
[19], respectively). The findings are not consistent with the view that ADP inhibits the enzyme by simple steric interference with ATP binding, but rather suggest that it acts at an allosteric site. The mechanism by which ADP inhibits the mammalian PRPP synthetases remains to be elucidated. Regarding the structure-function relationships of PRPP synthetase, Bower et al. characterized the E. coli mutant enzyme to identify a divalent cation-nucleotide binding site [32], and Harlow and Switzer used the active site-directed reagent 5'-(p-fluorosulfonylbenzoyl)adenosine to identify an active-site His residue of the S. typhimurium enzyme [33]. We have now constructed the chimeric enzymes between rat PRS I and PRS II; construction of chimeric enzymes is useful for structure-function analyses of similar proteins. The amino-acid sequences of rat PRS I and PRS II are 96% identical and this high sequence homology allows for the generation of catalytically active chimeras that maintain conformational integrity. The resuits obtained with several chimeric enzymes and the mutated enzymes with a single point mutation identified amino-acid residues contributing to different properties of the enzyme isoforms.
Acknowledgments We thank M. Ohara for critical comments. This study was supported in part by grants from the Ministries of Education, Science, and Culture and of Health and Welfare of Japan.
References [1] Kornberg, A., Lieberman, I. and Simms, E.S. (1955) J. Biol. Chem. 215, 417-427. [2] Flaks, J.G., Erwin, M.J. and Buchanan, J.M. (1957) J. Biol. Chem. 228, 201-213. [3] Fox, I.H. and Kelly, W.N. (1971) Ann. Intern. Med. 74, 424-433. [4] Holmes, E.W.(1978) in Handbook of Experimental Pharmacology (Kelly, W.N. and Weiner, I.M., eds.), Vol. 51, pp. 21-41, Springer, Heidelberg. [5] Bagnara, A.S. and Finch, L.R. (1974) Eur. J. Biochem. 41, 421-430. [6] Tatibana, M. and Shigesada, K. (1972) J. Biochem. (Tokyo) 72, 549-560. [7] Tatibana, M. (1978) in Handbook of Experimental Pharmacology (Kelly, W.N. and Weiner, I.M., eds.), Vol. 51, pp. 125-154, Springer, Heidelberg. [8] Keppler, D. and Holstege, A. (1982) in Metabolic Compartmentation (Sies, H., ed.), pp. 147-203, Academic Press, London. [9] Schubert, K.R., Switzer, R.L. and Shelton, A. (1975) J. Biol. Chem. 250, 7492-7500. [10] Switzer, R.L. and Gibson, K.J. (1978) Methods Enzymol. 51, 3-11 [11] Hove-Jensen, B., Harlow, K.W., King, C.J. and Switzer, R.L. (1986) J. Biol. Chem. 261, 6765-6771. [12] Arnvig, K., Hove-Jensen, B. and Switzer R.L. (1990) Eur. J. Biochem. 192, 195-200.
L Ahmad et al./Biochimica et Biophysica Acta 1207 (1994) 126-133
[13] Fox, I.H. and Kelly, W.N. (1971) J. Biol. Chem. 246, 5739-5748. [14] Roth, D.G., Shelton, E. and Deuel, T.F. (1974) J. Biol. Chem. 249, 291-296. [15] Kita, K., Otsnki, T., Ishizuka, T. and Tatibana, M. (1989) J. Biochem. (Tokyo) 105, 736-741. [16] Kita, K., Ishizuka, T., Ishijima, S., Sonoda, T. and Tatibana, M. (1994) J. Biol. Chem. 269, 8334-8340. [17] Taira, M., Ishijima, S., Kita, K., Yamada,K., Iizasa,T. and Tatibana, M. (1987) J. Biol. Chem. 262, 14867-14870. [18] Taira, M., lizasa, T., Yamada, K., Shimada, H. and Tatibana, M. (1989) Biochim. Biophys. Acta 1007, 203-208. [19] Ishijima, S., Kita, K., Ahmad, I., Ishizuka, T., Taira, M. and Tatibana, M. (1991) J. Biol. Chem. 266, 15693-15697. [20] Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular cloning: A laboratory Manual, 2nd Edn., Cold Spring Harbor Press, Cold Spring Harbor, New York. [21] Gliemann, J. (1967) Diabetologia 3, 382-388. [22] Kornberg, A., Liebermann, I. and Simms, E.S. (1955) J. Biol. Chem. 215, 389-402.
133
[23] Bradford, M.M. (1976) Anal. Biochem. 72, 248-254. [24] Laemmli, U.K.(1970) Nature 227, 680-687. [25] Towbin, H., Staehelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. 76, 4350-4354. [26] Becker, M.A., Raivio, K.O. and Seegmiller, J.E. (1979) Adv. Enzytool. Relat. Areas Mol. Biol. 49, 281-306. [27] Fox, I.J. and Kelly, W.N. (1972) J. Biol. Chem. 247, 2126-2131 [28] Becker, M.A., Kostel, P.J. and Meyer, L.J. (1975) J. Biol. Chem. 250, 6822-6830 [29] Switzer, R.L. and Sogin, D.C. (1973) J. Biol. Chem. 248, 1063-1073 [30] Gibson, K.J., Schubert, K.R. and Switzer, R.L. (1982) J. Biol. Chem. 257, 2391-2396 [31] Roessler, B.J., Nosal, J.M., Smith, P.R., Heidler, S.A., Palella, T.D., Switzer, R.L. and Becker, M.A. (1993) J. Biol. Chem. 268, 2647626481 [32] Bower, G.S., Harlow, K.W., Switzer, R.L. and Hove-Jensen, B. (1989) J. Biol. Chem. 264, 10287-10291 [33] Harlow, K.W. and Switzer, R.L. (1990) J. Biol. Chem. 265, 54875493.