Biochimica et Biophysica Acta, 1122 (1992) 85-92 © 1992 Elsevier Science Publishers B.V. All rights reserved 0167-4838/92/$05.0(I
85
BBAPRO 34237
Purification, characterization and preliminary X-ray study of fumarase from Saccharomyces cerevisiae Jan S. Keruchenko a, Irina D. Keruchenko ~, Kirill L. Gladilin a, Vyacheslav N. Zaitsev b and Nickolai Y. Chirgadze b A.N. l]ach htstitute of Biochemistry of the Russian Academy o f Sciences, Moscow (Russia) and t, Institute of Crystallography o f the Russian Academy of Sciences, Moscow (Russia) (Received 25 October 1991) (Revised manuscript received 6 January 1992)
Key words: Fumarase; Enzyme purification; Enz3,me characterization; (S. cerei'isiae) Fumarase (fumarate hydratase, EC 4.2.1.2) from Saccharomyces cerez'isiae has been purified to homogeneity by a method including acetone fractionation, D E A E ion-exchange and dye-sorbent affinity chromatography. The suggested method allows fumarase purification with a yield higher than 60% and may be used to obtain large enzyme quantities. The native protein consists o f four subunits with a --- 50 kDa molecular mass each and has an isoelectric point at p H 6.5 + 0.3. The equilibrium constant for fumarate hydration is about 4.3 (25°C, p H 7.5), the Michaelis constants for fumarate and l-malate are = 30 ~ M and ---250 /zM, respectively. The enzyme is activated by substrates and multivalent anions, the activation seems to be of a non-competitive type. The fumarase complex with mesodartaric acid has been crysta!lized by the vapor diffusion method. The unit cell parameters are a = 93.30, b = 94.05 and c = 106.07 ,~,, space group P212~2 ~. The unit cell contains 2 protein molecules. The crystals diffract to at least 2.6 A resolution and are suitable for X-ray structure analysis.
Introduction
Fumarase (fumarate hydratase, EC 4.2.1.2) catalyzes the reversible hydration reaction of fumarate to lmalate. Since this reaction is involved in Krebs cycle and some other metabolic pathways, fumarase is widely spread in the living cells. The recent data obtained from the genetic studies of fumarase [1-3] and the analysis of properties of the enzyme isolated from various species [4-6] suggest the existence of two genetically distinct classes of fumarases. The so-called "class II fumarase" is apparently the only class available in higher eukaryote cells; it has been found also in yeasts and some bacteria [7]. An intriguing feature of this fumarase is the complex kinetic behavior induced by anions and some nucleotides [8-11]. The mechanism of this phenomenon is still uncertain; apparently, it cannot be explained in terms
Abbreviations: SDS, sodium dodecyl sulphate: PEG, poly(ethylene glycol). Correspondence: J.S. Keruchenko, A.N. Bach Institute of Biochemistry of the Russian Academy of Sciences, 33 Leninsky Prospect, 117071 Moscow, Russia.
of the classic allosteric enzyme scheme [11]. Obviously, crystallographic studies of the enzyme structure may help in the understanding of this mechanism. The ew~lutionary aspect of the structural studies is also very important in view of the recently obtained data about evolutionary similarity of class II fumarase with aspartase and argininosuccinase that catalyse analogous trans-elimination reactions involving fumarate [7]. The best characterized member of class II fumarase is the enzyme from pig heart [12]. This enzyme has extensively been investigated in 1950's and 1960's. Early studies of pig heart fumarase have shown that its molecule consists of four polypeptide chains, each with a molecular weight of 48500 [12]. The presence of several species of subunit has also been indicated [13]. In 1986, the first X-ray diffraction studies were repo:ted [14]. The enzyme from Saccharomyces cerevisiae studied here, is evolutionarily very similar to pig heart fumarase [7]. However, the yeast fumarase has not yet been well characterized compared with pig enzyme. F-amarase from S. cerevisiae has been partially purified in 1958 [15]; isoenzymes in analytical quantities have been recently separated [16]. The amino acid composition based on gene sequence data [3] gives valuable information for further crystallographic investigation.
86 This paper describes a convenient method for fumarase purification that provides the enzyme in quantities sufficient for X-ray structural analysis. The main characteristics of the enzyme are studied in detail. Crystallization conditions and preliminary x-ray data are also reported. Materials
and Methods
Materials. DEAE-Sepharose Fast-Flow and Sepharose CL-6B were products of Pharmacia (Uppsala, Sweden). Chemicals were purchased from the following sources: E. Merck, L(-)malic acid, pyromellitic acid, meso-tartaric acid; Serva, Mes, Mops, Hepes, Cibacron blue F3GA; Ferak, PEG (M r 6000). Procion blue MX-R, Procion red H-E3B, Procion scarlet MX-G and Procion yellow H-A were kindly provided by Dr. S.F. Sadova (Textile Institute, Moscow, Russia). Acetone was of analytical grade. Pig heart fumarase was purified by the Kanarek and Hill method [17]. Baker's yeasts were obtained commercially. Enzyme assay. Fumarase activity was determined spectrophotometrically by the modified Racker method [18], with fumarate as a substrate (0.03 M fumarate in 0.05 M potassium phosphate buffer, pH 6.8) at 25°C. The reaction was initiated by the addition of enzyme and a decrease of absorbance at 300 nm was measured as a function of time. Activities were measured in /zmole of product per min. Protein determination. Protein concentrations were measured by the microbiuret method [19]. The measured extinction coefficient tg'280nm "-~ I m g / m l i"~ for pure yeast fumarase is about 0.26, i.e., the value is considerably lower than that of pig heart enzyme (0.51, Ref. 17), this value was used for the determination of fumarase concentration at the final purification step; it may also be used to estimate enzyme purity. Chromatographic methods. For analytical purposes, the FPLC system (Pharmacia) was used with the next columns: Superose 12 H R 1 0 / 3 0 (gel filtration), Mono-P H R 5/20 (chromatofocusing), Mono-Q H R 5 / 5 (anion exchange) and Phenyl Superose H R 5 / 5 (hydrophobic interaction chromatography).
The triazine dyes have been attached to Sepharose CL-6B by the method suggested by Atkinson et al. [20]. Dye concentration in Cibacron blue F3GA-Sepharose determined spectrometrically [21] was found to be 2.4 mg m l - i. Enzyme purification. A typical procedure of yeast fumarase purification is given below and summarized in Table I. All operations, unless otherwise specified, were carried out at room temperature. The buffers consisted of 0.2 M Mops-KOH (pH 7 . 5 ) ( A ) or 0.08 M Mes-KOH (pH 6.0) (B) were used. The A I - A 5 buffers were prepared from buffer A by dilution to Mops concentrations of 30, 40, 150, 150 and 20 mM, respectively; the A 4 and A s buffers contained ethylene glycol (25 vol.%).
Step 1, crude extract preparation and acetone fractionation. Yeast cells (900 g, net weight) were suspended in 900 ml cold medium containing 40 mM citrate, 3 mM EDTA-KOH (pH 7.4) and disrupted in a ML-1 bead mill disintegrator (Russia) at 4000 rpm; the temperature in the disintegration camera was maintained lower than 8°C. The homogenate (1600 ml) was diluted to 3200 ml with the same medium and then 1600 ml of cooled acetone (4°C) was added quickly under continuous stirring. The mixture was kept for 30 min in a cold room, then the resulting precipitate was removed by 15 min centrifugation at 3500 × g at 0°C. Acetone (one-half volume) was added to the clear supernatant. The mixture was occasionally stirred for 30 min and then was centrifuged as was indicated earlier. The supernatant solution was discarded and the pellet was dissolved in 600 ml of buffer A2.
Step 2, DEAE-Sepharose chromatography (Fig. 1). The protein solution obtained at step 1 was applied to a DEAE-Sepharose Fast-Flow column (2.5 × 75 era), pre-equilibrated with buffer A~. The column was washed with 500 ml of buffer A 1. The proteins were eluted with buffer B. In this procedure a flow rate of 12 ml min -~ was maintained. The fractions with fumarase activity above 30 U m l - l were pooled. Step 3, affinity chromatography (Fig. 2). The protein solution from the previous step (120 ml) was diluted to 200 ml with buffer A~, and then pH was adjusted to 7.1
TABLE i
Purificationof fumarasefrom S. cerecisiae Fraction
0. Extract 1. Acetone fractionation 2. DEAE-Sepharose Fast-Flow 3. First Cibacron F3GA Sepharose 4. Second Cibacron F3GA Sepharose
Volume (ml) -
7OO 120 24 7
Protein concn. (mg/ml) 21 23 10 1.7 5.2
Total protein (mg)
Total activity (U)
16000
59000
1200 42 36
56000 54000 46000 42000
Spec. activity (U/mg) ! .3
3.5 45 1090 1150
Purification
Yield (%)
-
-
2.7 35 840 885
95 91 78 71
87
using 1 M KOH. After l0 min centrifugation at 13000 x g, the supernatant was applied to a Cibacron blue F3GA Sepharose CL-6B column (1.6 × 40 cm) preequilibrated in buffer A 3. Then the column was washed with buffers A 3 (500 mi), A 4 (500 ml) and A 5 (100 ml) at a flow rate of 4 ml min-i. Fumarase was then selectively eluted with buffer A s containing 4 mM meso-tartarate. Active fractions were pooled and dialyzed overnight at 4°C against 5 l 5 mM Mops, 5 mM 2-mercaptoethanoI (pH 7.1). Step 4, rechromatography. The dialyzed sample was rechromatographied on a small-volume Cibacron blue Sepharose column (1 x 15 cm) in a similar manner. After protein loading, the column was washed with 100 ml of buffer A 4, then the flow rate was decreased from 2 ml min-t to 0.4 ml min-i and fumarase was eluted with buffer A s containing 4 mM meso-tartarate (pH 7.1). The active fractions were stored at - 4 ° C . Crystallization and X-ray methods. Crystals suitable for X-ray structure analysis were grown in a drop by the vapor diffusion method from a solution of 1.0% protein (w/v), 9% PEG-6000 (w/v) in the presence of 25% (v/v) ethylene glycol and 4 mM meso-tartaric acid, in 20 mM Mops, pH 7.5 at 20°C. The reservoirs were 14% (w/v) PEG-6000, 20 mM Mops (pH 7.5) containing 25% (v/v) ethylene glycol. The crystals grew for about 1 week. A preliminary X-ray study was carried out on an Elliot GX-20 rotating anode generator with 0.25 × 2.5 mm focal spot operated at 40 kV and 40 mA. The precession photographs of the hkO, Okl, hO1 and hhl zones were obtained at precession angles of 10-13 °.
A"'~ 1''~
A2
~
A4 A5 2M KCI
2000
3.0
J
2.0
1000
I
1.0
0
1000
o
Volume (ml)
Fig. 2. Elution profile of fumarase from a Cibacron blue F3GA Sepharose CL-6B column. The pH of protein solution from step 2 was adjusted to 7.1 and after centrifugation, the supernatant was applied to a Cibacron blue-Sepharose column 0 . 6 × 4 0 cm). Fumarase was selectively eluted with buffer A 5 containing 4 mM meso-tartarate. ( ), protein and (o). fumarase activity.
The unit cell parameters were refined on a Syntex P2~ diffractometer. Other methods. The electrophoresis and isoelectrophocusing was carried out on a PhastSystem apparatus or on other Pharmacia standard equipment according to the recommended techniques [22,23]. The kinetic constants were calculated by the non-linear least-squares method. Results and Discussion
Fumarase purification A~
Buffer B
:D
8.0
6.0 1000
4.0
%
2.0 J
0
500
I
2000 Volume (ml)
Fig. 1. Elution profile of fumarase from a DEAE-Sepharose FastFlow column. The protein fraction obtained after treatment of homogenate by acetone was dissolved in 600 ml of buffer A 2, then applied to a DEAE-Sepharose column (2.5x75 cm). Elution was performed with buffer 13. ( ), protein as absorbance at 280 nm; (o), fumarase activity.
The developed procedure of yeast fumarase purification is rapid enough and provides large quantities of very pure enzyme (Fig. 3). The method was successfully used in our laboratory. Below we discuss some important details of the procedure. (a) Acetone fractionation appears to be very convenient for the preliminary fumarase purification. Upon addition of the first portion of acetone a substantial amount of protein precipitated and the cells debris readily aggregated so that it could be removed completely by short centrifugation. Mixing an organic solvent with aqueous solution immediately resulted in a temperature rise up to 16-18°C, but there was no need to cool the mixture intensively as fumarase seemed to be relatively stable under these conditions. Additional quantities of proteins precipitated after slow cooling in the cold room due to denaturation. Buffers containing diisopropyl fluorophosphate and pepstatin at concentrations of 0.2 ml l-~ and 0.3 mg l-l, respectively, have been used in some preparations.
88
O
D ~ D
12345678 Fig. 3. SDS-PAGE of fumarase preparations. SDS-PAGE was'performed on PhastGel Gradient 10-15, the gel was stained with Coomassie R-350. From left to right: lanes 1 and 6, standard marker with molecular mass 94, 67, 43, 30, 20 and 14 kDa (each of 0.I /~g); lane 2, protein fraction after DEAE-Sepharose step (2.6/zg); lane 3, fumarase fraction after Ist Cibacron blue Sepharose step (0.4 /~g); lanes 4 and 5, the final preparation (2.1 #g); lane 7, pig heart fumarase (0.3 ~zg); lane 8, the final preparation (2.0/zg).
However, since no difference either in activity recovery or in properties of the purified enzyme was observed, we believe that such additions are unnecessary. (b) Chromatography on DEAE-Sepharose was used to remove proteins that would bind very tightly with the affinity sorbent in the next column thereby reducing its capacity. The typical elution profile is shown in Fig. 1. This intermediate step, although simple, leads to significant purification. Similar results may be obtained with other weak anion-exchangers, and yet DEAE-Sepharose Fast-Flow seems to be optimal due to its high protein capacity and flow rates. Strong anion-exchangers (Q-Sepharose, QAE-Sephadex, etc.) tend to yield poor fumarase recovery and therefore should be avoided. Thin and long columns are recommended for c~.~antitative fumarase binding. (c) Thus far fumarases from different sources have been isolated by the affinity technique developed by Beeckmans and Kanarek [24]. This technique appears to be powerful. However, it requires the preparation of a special sorbent: pyromellitic acid-agarose. The enzyme obtained usually requires further purification. We suggest an alternative procedure in which the general-ligand chromatography with a dye sorbent is used as the central step of fumarase isolation. A similar approach has been employed earlier in the partial purification of fumarase from yeast mitochondria.
However, as reported in Ref. 25, the enzyme yields were low and, once purified, the enzyme was unstable. During the series of our experiments, designed to find optimum conditions, the recovery has been significantly improved and very good purification has been achieved. The fumarase sorption on dye-agarose columns seems to be connected with the enzyme ability to bind ATP and NAD [9,12]. The efficiency of elution with substrates and strong competitive inhibitors implies that fumarase, similar to nucleotide dependent enzymes, may be bound with such sorbents by direct 'active site-dye' interactions. Since hydrophobic interactions also take place, the presence of ethylene glycol or an analogous agent is necessary to ensure good fumarase recovery in this procedure. In preliminary experiments, different Sepharose CL6B-attached triazine dyes (Cibacron blue F3GA, Procion blue MX-R, Procion red H-E3B, Procion scarlet MX-G, Procion yellow H-A) were tested. We have observed that all these matrices can bind fumarase to a certain extent, but, surprisingly, the most common sorbent, Cibacron blue F3GA Sepharose, yields the maximum recovery and reliable results and therefore it was used for further work. The elution profile of fumarase from Cibacron blue-Sepharose is shown in Fig. 2. The ability of Cibacron blue F3GA-agarose to bind fumarase is affected by many factors. The optimum pH region was found as 7.4-7.6; binding is incomplete at alkaline pH; the substrate elution is non-effective at pH below 6.7. The ionic composition of the media is also important. Any multivalent anions that may be considered as potential enzyme inhibitors preventbinding and cause partial or full fumarase de~orption. Moreover, similar 'specific' interactions have also been observed, although to a lesser degree, for some monovalent anions (chloride, acetate). For instance, in the presence of 0.08 M chloride a remarkable fumarase activity was detected in the eluate, although enzyme could be eluted completely only for 0.6 M KCI. Therefore the ionic strength required for column washing was maintained by the use of 0.15 M Mops.
Stability The purified enzyme was found to be quite stable. The final fractions from dye-ligand chromatography (containing 25% (v/v) ethylene glycol) may be ~tored at -4°C for at least 6 months without any loss of activity. Yeast fumarase is stable at pH values from 4.5 to 8.2; for more alkaline pH, inactivation proceeds rapidly, but it may be slightly reduced by the addition of substrates and dithiothreitol. In aqueous solution, containing 0.05 M Mops, 0.05 M fumarate (pH 7.5) the rate of inactivation at 20°C is negligible, if the enzyme concentration is comparatively high (above 50 .v.g ml- ~). However, with dilution the loss of activity increases
89 and reaches 30% per h at a fumarase concentration of 3 / z g ml-I. At higher temperatures, fumarase stability decreases drastically, and at 50°C complete inactivation takes place within several seconds.
- - i
0.02 ,-5
Molecular properties Using polyacrylamide-SDS gel electrophoresis, we have estimated the molecular weight of the enzyme subunit as = 50000 (Fig. 3). This value may be expected from yeast fumarase gene sequence data [3] and is in agreement with the previous observations [16]. According to our results for gel filtration and electrophoresis under native conditions, one molecule of yeast fumarase consists of four identical subunits and hence resembles the well-characterized pig heart enzyme. Since both enzymes display significant homology in amino acid sequences [7], this resemblance is obvious. As has been demonstrated by other laboratories, fumarase purified from various sources may consist of two forms, corresponding to the mitochondrial and cytosolic fractions [16,26,27]. Using the method of chromatographic differentiation on hydroxyapatite sorbents [28], Boonyarat and Doonan [16] managed to separate and partially characterize yeast fumarase isoenzymes. However, N-terminal residue analysis failed, probably owing to N-terminus blocking [16]. The marked difference between molecular weights of mitochondrial and cytosolic isoenzymes has also been reported [16]. In our preliminary experiments on hydroxyapatite with the yeast fumarase purified by our method we have clearly observed the separation of fumarase activity on several peaks, however, the analysis of eluted fractions by SDS-electrophoresis on the gradient polyacrylamide gel and by the Laemmli method did not reveal any differences in the molecular weights of these forms within the accuracy of these methods. We suppose that the forms obtained have very close or even identical kinetic properties; our attempts to separate them by alternative methods (electrophoresis in 6 M urea, chromatofocusing, ion exchange and hydrophobic interaction chromatography) were unsuccessful. Further studies are now in progress. Proceeding from the ion-exchange chromatography and chromatofocusing data, the isoelectric point of yeast fumarase lies in the neutral area (6.1-6.8). The p l values, obtained by isoelectric f~cusing belong to the same region, but they depend on the ampholyte composition. Besides, in some cases the band corresponding to fumarase was splitted in 4 - 7 minor components. We suggest that this phenomenon follows from the formation of an enzyme-ampholyte complex and is not associated with the presence of fumarase multiple forms.
oo of"
°°'t
r------.--.-
~o
1/[s]
...g._
0
100
v/[S], (pmol/min per mg)/mM
F i g . 4. E a d i e - H o f s t e e p l o t f o r t h e m a l a t e d e h y d r a t i o n r e a c t i o n . T h e
reaction rates were measured at 25°C in 20 mM Hepes (pH 7.5). Insert: Lineweaver-Burkplot of the same data.
Kinetic properties The equilibrium constant for fumarate-malate interconversion measured at pH 7.2 was found as 4.3 + 0.5. The rate of malate dehydration reaches the maximum values at pH 7.6; for the reverse reaction the constantrate 'plateau' is observed at pH 5.3-6.8. The reaction rates determined over the broad substrate concentration range (10-5-10 -~ M) display an unusual kinetic behavior of fumarase (Fig. 4). For both the hydration and the reverse reactions the considerable stimulation of acfivi%, is clearly observed at substrate concentrations above 1 mM. The experimental data seem to be in good agreement with the equation:
+"Ka) tsl
' = (,+lSllfi+lSl
/
which may be derived from the simple kinetic scheme: E
~
ES'
~
S \ ~ K~
Ks
ES
ESS'
k
, P+...
~k , P +
"'"
The calculated values of K s, K,, k and a are 0.030 +_ 0.020 mM, 45 + 20 mM, 105 + 40 s- i, 9.2 + 3.0 for the hydration reaction and 0.25 + 0.09 mM, 33 + 10 mM, 120 + 25 s-i, 6.2 + 3.0 for the reverse reaction. Unfortunately, in spite of a large set of experimental points (up to 75), the standard errors remain to be rather high as a consequence of four unknown quantities in Eqn. 1. The pronounced stimulating effect of substrates and other multivalent anions on the fumarase-catalyzed reaction is a well known fact for the pig heart enzyme
90 [12]. The action of multivalent ions rather than substrates is, however, known to be more complex due to potential structural similarity of these compounds with natural substrates. As a result, in dependence on the reaction conditions, concentrations and particular effector, either stimulation or inhibition occurs. Since almost all multieharged anions can activate the enzyme to some extent, the same mechanism for substrate and anion activation may be suggested. It has been shown, that the activation effect cannot be explained by an increase in ionic strength [29]. Apparently, some low specific interactions between enzyme and activator take place, but the details of this mechanism are still uncertain. In comparison with the pig heart enzyme, yeast fumarase is characterized by similar but more pronounced dependences. Although its affinity to substrates is 3-6-fold lower, the parameter a is 2-4-fold higher as it follows from the interpolation of data published for the pig heart enzyme [8,29] and f r o m o u r experiments. The effect of phosphate on yeast fumarase activity is illustrated by Figs. 5 and 6. To explain the data obtained we suggest the logical extension of the previous scheme. Including an additional effector A, it may be rewritten as: /3Ki
EAA'
' EA' ~ #K~ ESA'
~
EA
E
>
EAS'
t~k
~
> p+...
,
I
2
,
1/[S], mM "1
Fig. 6. Double-reciprocal plot for the fumarate hydration reaction. The reaction rates were measured at 25°C in 20 mM Hepes (pH 7.5) in the absence of phosphate ( [S]/K a) the equation (1) reduces to the Michaelis form (see Fig. 6):
, P+...
ES
ES'
t
1
, p+...
ESS'
:1
%
v~ r =
Km I+--
(3)
IS]
where [At
l+ot-K, I'm = kEa [A] I+-Kx
>='li i
50
100
tx [
[Phosphate], mM
Fig. 5. The effect of phosphate on the activity of yeast fumarase. The reaction rates were measured at 25°C in 30 mM Hepes (pH 7.5) at famarate concentrations 30 mM (o), 3.75 mM ( [ ] ) and 0.50 mM
(A).
(4)
and [A] O- 1 K..=K~ ,+ r--~-+ K. +------~] [A]
(5)
91
F
L
u
In the cases where K i
85
BBAPRO 34237
Purification, characterization and preliminary X-ray study of fumarase from Saccharomyces cerevisiae Jan S. Keruchenko a, Irina D. Keruchenko ~, Kirill L. Gladilin a, Vyacheslav N. Zaitsev b and Nickolai Y. Chirgadze b A.N. l]ach htstitute of Biochemistry of the Russian Academy o f Sciences, Moscow (Russia) and t, Institute of Crystallography o f the Russian Academy of Sciences, Moscow (Russia) (Received 25 October 1991) (Revised manuscript received 6 January 1992)
Key words: Fumarase; Enzyme purification; Enz3,me characterization; (S. cerei'isiae) Fumarase (fumarate hydratase, EC 4.2.1.2) from Saccharomyces cerez'isiae has been purified to homogeneity by a method including acetone fractionation, D E A E ion-exchange and dye-sorbent affinity chromatography. The suggested method allows fumarase purification with a yield higher than 60% and may be used to obtain large enzyme quantities. The native protein consists o f four subunits with a --- 50 kDa molecular mass each and has an isoelectric point at p H 6.5 + 0.3. The equilibrium constant for fumarate hydration is about 4.3 (25°C, p H 7.5), the Michaelis constants for fumarate and l-malate are = 30 ~ M and ---250 /zM, respectively. The enzyme is activated by substrates and multivalent anions, the activation seems to be of a non-competitive type. The fumarase complex with mesodartaric acid has been crysta!lized by the vapor diffusion method. The unit cell parameters are a = 93.30, b = 94.05 and c = 106.07 ,~,, space group P212~2 ~. The unit cell contains 2 protein molecules. The crystals diffract to at least 2.6 A resolution and are suitable for X-ray structure analysis.
Introduction
Fumarase (fumarate hydratase, EC 4.2.1.2) catalyzes the reversible hydration reaction of fumarate to lmalate. Since this reaction is involved in Krebs cycle and some other metabolic pathways, fumarase is widely spread in the living cells. The recent data obtained from the genetic studies of fumarase [1-3] and the analysis of properties of the enzyme isolated from various species [4-6] suggest the existence of two genetically distinct classes of fumarases. The so-called "class II fumarase" is apparently the only class available in higher eukaryote cells; it has been found also in yeasts and some bacteria [7]. An intriguing feature of this fumarase is the complex kinetic behavior induced by anions and some nucleotides [8-11]. The mechanism of this phenomenon is still uncertain; apparently, it cannot be explained in terms
Abbreviations: SDS, sodium dodecyl sulphate: PEG, poly(ethylene glycol). Correspondence: J.S. Keruchenko, A.N. Bach Institute of Biochemistry of the Russian Academy of Sciences, 33 Leninsky Prospect, 117071 Moscow, Russia.
of the classic allosteric enzyme scheme [11]. Obviously, crystallographic studies of the enzyme structure may help in the understanding of this mechanism. The ew~lutionary aspect of the structural studies is also very important in view of the recently obtained data about evolutionary similarity of class II fumarase with aspartase and argininosuccinase that catalyse analogous trans-elimination reactions involving fumarate [7]. The best characterized member of class II fumarase is the enzyme from pig heart [12]. This enzyme has extensively been investigated in 1950's and 1960's. Early studies of pig heart fumarase have shown that its molecule consists of four polypeptide chains, each with a molecular weight of 48500 [12]. The presence of several species of subunit has also been indicated [13]. In 1986, the first X-ray diffraction studies were repo:ted [14]. The enzyme from Saccharomyces cerevisiae studied here, is evolutionarily very similar to pig heart fumarase [7]. However, the yeast fumarase has not yet been well characterized compared with pig enzyme. F-amarase from S. cerevisiae has been partially purified in 1958 [15]; isoenzymes in analytical quantities have been recently separated [16]. The amino acid composition based on gene sequence data [3] gives valuable information for further crystallographic investigation.
86 This paper describes a convenient method for fumarase purification that provides the enzyme in quantities sufficient for X-ray structural analysis. The main characteristics of the enzyme are studied in detail. Crystallization conditions and preliminary x-ray data are also reported. Materials
and Methods
Materials. DEAE-Sepharose Fast-Flow and Sepharose CL-6B were products of Pharmacia (Uppsala, Sweden). Chemicals were purchased from the following sources: E. Merck, L(-)malic acid, pyromellitic acid, meso-tartaric acid; Serva, Mes, Mops, Hepes, Cibacron blue F3GA; Ferak, PEG (M r 6000). Procion blue MX-R, Procion red H-E3B, Procion scarlet MX-G and Procion yellow H-A were kindly provided by Dr. S.F. Sadova (Textile Institute, Moscow, Russia). Acetone was of analytical grade. Pig heart fumarase was purified by the Kanarek and Hill method [17]. Baker's yeasts were obtained commercially. Enzyme assay. Fumarase activity was determined spectrophotometrically by the modified Racker method [18], with fumarate as a substrate (0.03 M fumarate in 0.05 M potassium phosphate buffer, pH 6.8) at 25°C. The reaction was initiated by the addition of enzyme and a decrease of absorbance at 300 nm was measured as a function of time. Activities were measured in /zmole of product per min. Protein determination. Protein concentrations were measured by the microbiuret method [19]. The measured extinction coefficient tg'280nm "-~ I m g / m l i"~ for pure yeast fumarase is about 0.26, i.e., the value is considerably lower than that of pig heart enzyme (0.51, Ref. 17), this value was used for the determination of fumarase concentration at the final purification step; it may also be used to estimate enzyme purity. Chromatographic methods. For analytical purposes, the FPLC system (Pharmacia) was used with the next columns: Superose 12 H R 1 0 / 3 0 (gel filtration), Mono-P H R 5/20 (chromatofocusing), Mono-Q H R 5 / 5 (anion exchange) and Phenyl Superose H R 5 / 5 (hydrophobic interaction chromatography).
The triazine dyes have been attached to Sepharose CL-6B by the method suggested by Atkinson et al. [20]. Dye concentration in Cibacron blue F3GA-Sepharose determined spectrometrically [21] was found to be 2.4 mg m l - i. Enzyme purification. A typical procedure of yeast fumarase purification is given below and summarized in Table I. All operations, unless otherwise specified, were carried out at room temperature. The buffers consisted of 0.2 M Mops-KOH (pH 7 . 5 ) ( A ) or 0.08 M Mes-KOH (pH 6.0) (B) were used. The A I - A 5 buffers were prepared from buffer A by dilution to Mops concentrations of 30, 40, 150, 150 and 20 mM, respectively; the A 4 and A s buffers contained ethylene glycol (25 vol.%).
Step 1, crude extract preparation and acetone fractionation. Yeast cells (900 g, net weight) were suspended in 900 ml cold medium containing 40 mM citrate, 3 mM EDTA-KOH (pH 7.4) and disrupted in a ML-1 bead mill disintegrator (Russia) at 4000 rpm; the temperature in the disintegration camera was maintained lower than 8°C. The homogenate (1600 ml) was diluted to 3200 ml with the same medium and then 1600 ml of cooled acetone (4°C) was added quickly under continuous stirring. The mixture was kept for 30 min in a cold room, then the resulting precipitate was removed by 15 min centrifugation at 3500 × g at 0°C. Acetone (one-half volume) was added to the clear supernatant. The mixture was occasionally stirred for 30 min and then was centrifuged as was indicated earlier. The supernatant solution was discarded and the pellet was dissolved in 600 ml of buffer A2.
Step 2, DEAE-Sepharose chromatography (Fig. 1). The protein solution obtained at step 1 was applied to a DEAE-Sepharose Fast-Flow column (2.5 × 75 era), pre-equilibrated with buffer A~. The column was washed with 500 ml of buffer A 1. The proteins were eluted with buffer B. In this procedure a flow rate of 12 ml min -~ was maintained. The fractions with fumarase activity above 30 U m l - l were pooled. Step 3, affinity chromatography (Fig. 2). The protein solution from the previous step (120 ml) was diluted to 200 ml with buffer A~, and then pH was adjusted to 7.1
TABLE i
Purificationof fumarasefrom S. cerecisiae Fraction
0. Extract 1. Acetone fractionation 2. DEAE-Sepharose Fast-Flow 3. First Cibacron F3GA Sepharose 4. Second Cibacron F3GA Sepharose
Volume (ml) -
7OO 120 24 7
Protein concn. (mg/ml) 21 23 10 1.7 5.2
Total protein (mg)
Total activity (U)
16000
59000
1200 42 36
56000 54000 46000 42000
Spec. activity (U/mg) ! .3
3.5 45 1090 1150
Purification
Yield (%)
-
-
2.7 35 840 885
95 91 78 71
87
using 1 M KOH. After l0 min centrifugation at 13000 x g, the supernatant was applied to a Cibacron blue F3GA Sepharose CL-6B column (1.6 × 40 cm) preequilibrated in buffer A 3. Then the column was washed with buffers A 3 (500 mi), A 4 (500 ml) and A 5 (100 ml) at a flow rate of 4 ml min-i. Fumarase was then selectively eluted with buffer A s containing 4 mM meso-tartarate. Active fractions were pooled and dialyzed overnight at 4°C against 5 l 5 mM Mops, 5 mM 2-mercaptoethanoI (pH 7.1). Step 4, rechromatography. The dialyzed sample was rechromatographied on a small-volume Cibacron blue Sepharose column (1 x 15 cm) in a similar manner. After protein loading, the column was washed with 100 ml of buffer A 4, then the flow rate was decreased from 2 ml min-t to 0.4 ml min-i and fumarase was eluted with buffer A s containing 4 mM meso-tartarate (pH 7.1). The active fractions were stored at - 4 ° C . Crystallization and X-ray methods. Crystals suitable for X-ray structure analysis were grown in a drop by the vapor diffusion method from a solution of 1.0% protein (w/v), 9% PEG-6000 (w/v) in the presence of 25% (v/v) ethylene glycol and 4 mM meso-tartaric acid, in 20 mM Mops, pH 7.5 at 20°C. The reservoirs were 14% (w/v) PEG-6000, 20 mM Mops (pH 7.5) containing 25% (v/v) ethylene glycol. The crystals grew for about 1 week. A preliminary X-ray study was carried out on an Elliot GX-20 rotating anode generator with 0.25 × 2.5 mm focal spot operated at 40 kV and 40 mA. The precession photographs of the hkO, Okl, hO1 and hhl zones were obtained at precession angles of 10-13 °.
A"'~ 1''~
A2
~
A4 A5 2M KCI
2000
3.0
J
2.0
1000
I
1.0
0
1000
o
Volume (ml)
Fig. 2. Elution profile of fumarase from a Cibacron blue F3GA Sepharose CL-6B column. The pH of protein solution from step 2 was adjusted to 7.1 and after centrifugation, the supernatant was applied to a Cibacron blue-Sepharose column 0 . 6 × 4 0 cm). Fumarase was selectively eluted with buffer A 5 containing 4 mM meso-tartarate. ( ), protein and (o). fumarase activity.
The unit cell parameters were refined on a Syntex P2~ diffractometer. Other methods. The electrophoresis and isoelectrophocusing was carried out on a PhastSystem apparatus or on other Pharmacia standard equipment according to the recommended techniques [22,23]. The kinetic constants were calculated by the non-linear least-squares method. Results and Discussion
Fumarase purification A~
Buffer B
:D
8.0
6.0 1000
4.0
%
2.0 J
0
500
I
2000 Volume (ml)
Fig. 1. Elution profile of fumarase from a DEAE-Sepharose FastFlow column. The protein fraction obtained after treatment of homogenate by acetone was dissolved in 600 ml of buffer A 2, then applied to a DEAE-Sepharose column (2.5x75 cm). Elution was performed with buffer 13. ( ), protein as absorbance at 280 nm; (o), fumarase activity.
The developed procedure of yeast fumarase purification is rapid enough and provides large quantities of very pure enzyme (Fig. 3). The method was successfully used in our laboratory. Below we discuss some important details of the procedure. (a) Acetone fractionation appears to be very convenient for the preliminary fumarase purification. Upon addition of the first portion of acetone a substantial amount of protein precipitated and the cells debris readily aggregated so that it could be removed completely by short centrifugation. Mixing an organic solvent with aqueous solution immediately resulted in a temperature rise up to 16-18°C, but there was no need to cool the mixture intensively as fumarase seemed to be relatively stable under these conditions. Additional quantities of proteins precipitated after slow cooling in the cold room due to denaturation. Buffers containing diisopropyl fluorophosphate and pepstatin at concentrations of 0.2 ml l-~ and 0.3 mg l-l, respectively, have been used in some preparations.
88
O
D ~ D
12345678 Fig. 3. SDS-PAGE of fumarase preparations. SDS-PAGE was'performed on PhastGel Gradient 10-15, the gel was stained with Coomassie R-350. From left to right: lanes 1 and 6, standard marker with molecular mass 94, 67, 43, 30, 20 and 14 kDa (each of 0.I /~g); lane 2, protein fraction after DEAE-Sepharose step (2.6/zg); lane 3, fumarase fraction after Ist Cibacron blue Sepharose step (0.4 /~g); lanes 4 and 5, the final preparation (2.1 #g); lane 7, pig heart fumarase (0.3 ~zg); lane 8, the final preparation (2.0/zg).
However, since no difference either in activity recovery or in properties of the purified enzyme was observed, we believe that such additions are unnecessary. (b) Chromatography on DEAE-Sepharose was used to remove proteins that would bind very tightly with the affinity sorbent in the next column thereby reducing its capacity. The typical elution profile is shown in Fig. 1. This intermediate step, although simple, leads to significant purification. Similar results may be obtained with other weak anion-exchangers, and yet DEAE-Sepharose Fast-Flow seems to be optimal due to its high protein capacity and flow rates. Strong anion-exchangers (Q-Sepharose, QAE-Sephadex, etc.) tend to yield poor fumarase recovery and therefore should be avoided. Thin and long columns are recommended for c~.~antitative fumarase binding. (c) Thus far fumarases from different sources have been isolated by the affinity technique developed by Beeckmans and Kanarek [24]. This technique appears to be powerful. However, it requires the preparation of a special sorbent: pyromellitic acid-agarose. The enzyme obtained usually requires further purification. We suggest an alternative procedure in which the general-ligand chromatography with a dye sorbent is used as the central step of fumarase isolation. A similar approach has been employed earlier in the partial purification of fumarase from yeast mitochondria.
However, as reported in Ref. 25, the enzyme yields were low and, once purified, the enzyme was unstable. During the series of our experiments, designed to find optimum conditions, the recovery has been significantly improved and very good purification has been achieved. The fumarase sorption on dye-agarose columns seems to be connected with the enzyme ability to bind ATP and NAD [9,12]. The efficiency of elution with substrates and strong competitive inhibitors implies that fumarase, similar to nucleotide dependent enzymes, may be bound with such sorbents by direct 'active site-dye' interactions. Since hydrophobic interactions also take place, the presence of ethylene glycol or an analogous agent is necessary to ensure good fumarase recovery in this procedure. In preliminary experiments, different Sepharose CL6B-attached triazine dyes (Cibacron blue F3GA, Procion blue MX-R, Procion red H-E3B, Procion scarlet MX-G, Procion yellow H-A) were tested. We have observed that all these matrices can bind fumarase to a certain extent, but, surprisingly, the most common sorbent, Cibacron blue F3GA Sepharose, yields the maximum recovery and reliable results and therefore it was used for further work. The elution profile of fumarase from Cibacron blue-Sepharose is shown in Fig. 2. The ability of Cibacron blue F3GA-agarose to bind fumarase is affected by many factors. The optimum pH region was found as 7.4-7.6; binding is incomplete at alkaline pH; the substrate elution is non-effective at pH below 6.7. The ionic composition of the media is also important. Any multivalent anions that may be considered as potential enzyme inhibitors preventbinding and cause partial or full fumarase de~orption. Moreover, similar 'specific' interactions have also been observed, although to a lesser degree, for some monovalent anions (chloride, acetate). For instance, in the presence of 0.08 M chloride a remarkable fumarase activity was detected in the eluate, although enzyme could be eluted completely only for 0.6 M KCI. Therefore the ionic strength required for column washing was maintained by the use of 0.15 M Mops.
Stability The purified enzyme was found to be quite stable. The final fractions from dye-ligand chromatography (containing 25% (v/v) ethylene glycol) may be ~tored at -4°C for at least 6 months without any loss of activity. Yeast fumarase is stable at pH values from 4.5 to 8.2; for more alkaline pH, inactivation proceeds rapidly, but it may be slightly reduced by the addition of substrates and dithiothreitol. In aqueous solution, containing 0.05 M Mops, 0.05 M fumarate (pH 7.5) the rate of inactivation at 20°C is negligible, if the enzyme concentration is comparatively high (above 50 .v.g ml- ~). However, with dilution the loss of activity increases
89 and reaches 30% per h at a fumarase concentration of 3 / z g ml-I. At higher temperatures, fumarase stability decreases drastically, and at 50°C complete inactivation takes place within several seconds.
- - i
0.02 ,-5
Molecular properties Using polyacrylamide-SDS gel electrophoresis, we have estimated the molecular weight of the enzyme subunit as = 50000 (Fig. 3). This value may be expected from yeast fumarase gene sequence data [3] and is in agreement with the previous observations [16]. According to our results for gel filtration and electrophoresis under native conditions, one molecule of yeast fumarase consists of four identical subunits and hence resembles the well-characterized pig heart enzyme. Since both enzymes display significant homology in amino acid sequences [7], this resemblance is obvious. As has been demonstrated by other laboratories, fumarase purified from various sources may consist of two forms, corresponding to the mitochondrial and cytosolic fractions [16,26,27]. Using the method of chromatographic differentiation on hydroxyapatite sorbents [28], Boonyarat and Doonan [16] managed to separate and partially characterize yeast fumarase isoenzymes. However, N-terminal residue analysis failed, probably owing to N-terminus blocking [16]. The marked difference between molecular weights of mitochondrial and cytosolic isoenzymes has also been reported [16]. In our preliminary experiments on hydroxyapatite with the yeast fumarase purified by our method we have clearly observed the separation of fumarase activity on several peaks, however, the analysis of eluted fractions by SDS-electrophoresis on the gradient polyacrylamide gel and by the Laemmli method did not reveal any differences in the molecular weights of these forms within the accuracy of these methods. We suppose that the forms obtained have very close or even identical kinetic properties; our attempts to separate them by alternative methods (electrophoresis in 6 M urea, chromatofocusing, ion exchange and hydrophobic interaction chromatography) were unsuccessful. Further studies are now in progress. Proceeding from the ion-exchange chromatography and chromatofocusing data, the isoelectric point of yeast fumarase lies in the neutral area (6.1-6.8). The p l values, obtained by isoelectric f~cusing belong to the same region, but they depend on the ampholyte composition. Besides, in some cases the band corresponding to fumarase was splitted in 4 - 7 minor components. We suggest that this phenomenon follows from the formation of an enzyme-ampholyte complex and is not associated with the presence of fumarase multiple forms.
oo of"
°°'t
r------.--.-
~o
1/[s]
...g._
0
100
v/[S], (pmol/min per mg)/mM
F i g . 4. E a d i e - H o f s t e e p l o t f o r t h e m a l a t e d e h y d r a t i o n r e a c t i o n . T h e
reaction rates were measured at 25°C in 20 mM Hepes (pH 7.5). Insert: Lineweaver-Burkplot of the same data.
Kinetic properties The equilibrium constant for fumarate-malate interconversion measured at pH 7.2 was found as 4.3 + 0.5. The rate of malate dehydration reaches the maximum values at pH 7.6; for the reverse reaction the constantrate 'plateau' is observed at pH 5.3-6.8. The reaction rates determined over the broad substrate concentration range (10-5-10 -~ M) display an unusual kinetic behavior of fumarase (Fig. 4). For both the hydration and the reverse reactions the considerable stimulation of acfivi%, is clearly observed at substrate concentrations above 1 mM. The experimental data seem to be in good agreement with the equation:
+"Ka) tsl
' = (,+lSllfi+lSl
/
which may be derived from the simple kinetic scheme: E
~
ES'
~
S \ ~ K~
Ks
ES
ESS'
k
, P+...
~k , P +
"'"
The calculated values of K s, K,, k and a are 0.030 +_ 0.020 mM, 45 + 20 mM, 105 + 40 s- i, 9.2 + 3.0 for the hydration reaction and 0.25 + 0.09 mM, 33 + 10 mM, 120 + 25 s-i, 6.2 + 3.0 for the reverse reaction. Unfortunately, in spite of a large set of experimental points (up to 75), the standard errors remain to be rather high as a consequence of four unknown quantities in Eqn. 1. The pronounced stimulating effect of substrates and other multivalent anions on the fumarase-catalyzed reaction is a well known fact for the pig heart enzyme
90 [12]. The action of multivalent ions rather than substrates is, however, known to be more complex due to potential structural similarity of these compounds with natural substrates. As a result, in dependence on the reaction conditions, concentrations and particular effector, either stimulation or inhibition occurs. Since almost all multieharged anions can activate the enzyme to some extent, the same mechanism for substrate and anion activation may be suggested. It has been shown, that the activation effect cannot be explained by an increase in ionic strength [29]. Apparently, some low specific interactions between enzyme and activator take place, but the details of this mechanism are still uncertain. In comparison with the pig heart enzyme, yeast fumarase is characterized by similar but more pronounced dependences. Although its affinity to substrates is 3-6-fold lower, the parameter a is 2-4-fold higher as it follows from the interpolation of data published for the pig heart enzyme [8,29] and f r o m o u r experiments. The effect of phosphate on yeast fumarase activity is illustrated by Figs. 5 and 6. To explain the data obtained we suggest the logical extension of the previous scheme. Including an additional effector A, it may be rewritten as: /3Ki
EAA'
' EA' ~ #K~ ESA'
~
EA
E
>
EAS'
t~k
~
> p+...
,
I
2
,
1/[S], mM "1
Fig. 6. Double-reciprocal plot for the fumarate hydration reaction. The reaction rates were measured at 25°C in 20 mM Hepes (pH 7.5) in the absence of phosphate ( [S]/K a) the equation (1) reduces to the Michaelis form (see Fig. 6):
, P+...
ES
ES'
t
1
, p+...
ESS'
:1
%
v~ r =
Km I+--
(3)
IS]
where [At
l+ot-K, I'm = kEa [A] I+-Kx
>='li i
50
100
tx [
[Phosphate], mM
Fig. 5. The effect of phosphate on the activity of yeast fumarase. The reaction rates were measured at 25°C in 30 mM Hepes (pH 7.5) at famarate concentrations 30 mM (o), 3.75 mM ( [ ] ) and 0.50 mM
(A).
(4)
and [A] O- 1 K..=K~ ,+ r--~-+ K. +------~] [A]
(5)
91
F
L
u
In the cases where K i
