- Email: [email protected]
Overexpression, purification, and some properties of the AdoCbl-dependent ethanolamine ammonia-lyase from Salmonella typhimurium
ARCHIVES
OF BIOCHEMISTRY
AND BIOPHYSICS
Vol. 294, No. 1, April, pp. 50-54, 1992
Overexpression, Purification, and Some Properties of the AdoCbl-Dependent Ethanolamine Ammonia-Lyase from Salmonella typhimurium’ LaRosa P. Faust and Bernard M. Babior Department of Molecular and Experimental Medicine, The Scripps Research Institute, La Jolla, California 92037 Received September 3,199l
Recombinant ethanolamine ammonia-lyase from S. typhimurium has been overexpressed and purified in large quantities by a simple procedure. The molecular weight of the native enzyme is about 480 kDa, and it contains two active sites/molecule as shown by kinetic studies and by titration with CNCbl. Analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis confirms earlier cloning studies indicating that it is composed of two kinds of subunits, one of MW 3 1 kDa and the other of MW 50 kDa. These subunits, inactive by themselves, combine to produce an active enzyme whose composition is most likely &la. The IL, for AdoCbl is 0.6 fiM, and the turnover number is 65 s-l per active site at 22°C. o 1~02 Academic Press, Inc.
Ethanolamine ammonia-lyase is an enzyme found in several bacterial species that catalyzes the AdoCbl-dependent deamination of vicinal amino alcohols to 0x0 compounds and NH: (l-4). A great deal of work, carried out principally with the enzyme from an unclassified Clostridium, has shown that the reaction proceeds by a free radical mechanism in which the cofactor serves as a protected source of unpaired electrons (5). For further mechanistic studies, however, it would be desirable to have a recombinant enzyme that can be mutated at will. For this purpose, we have recently cloned and sequenced the genes encoding the ethanolamine ammonia-lyase from Salmonella typhimurium (6). We now report the overexpression and purification of this enzyme, and describe some of its properties. MATERIALS
AND
METHODS
The expression vector pKQV4 and Escherichie coli CAG626, the strain in which the recombinant plasmids were expressed, were the generous gifts of Dr. James A. Hoch. Restriction enzymes, T4 ligase, Klenow
1 Supported in part by USPHS 50
Grants GM-38050 and AI-24227.
fragment, Sl nuclease, isopropyl thiogalactoside (IPTG’), and other molecular biology supplies were purchased from Stratagene (La Jolla, CA), BRL Laboratories (Bethesda, MD), or Pharmacia (Piscataway, NJ). Agar and media were purchased from Difco (Detroit); AdoCbl, CNCbl, and ampicillin from Sigma (St. Louis); gel electrophoresis supplies and Bradford reagent from Bio-Rad (Richmond, VA); and [1,2-“Clethanolamine (2.4 mCi/mmol) from NEN (Boston, MA). Other reagents were of the best quality commercially available and were used without further purification.
Construction of Vectors That Overexpress the S. typhimurium Ethunolamine Ammoniu-Lyase and Its Subunits The genes for the AdoCbl-dependent ethanolamine ammonia-lyase from S. typhimurium occur as elements of an insert in the recombinant plasmid pBSE4.5, a derivative of pBR325 (6). The insert was excised from pBSE4.5 with Hind111 and ligated into the corresponding site of the expression vector pKQV4 (7). The new construct was transformed into E. coli XL-l Blue, and transformed bacteria were grown as individual clones on 1.5% YT agar (8) containing ampicillin (75 pg/ml). By mapping plasmids from several clones with HindI plus AocI, a plasmid containing the ethanolamine ammonia-lyase genes inserted in the sense direction (the “precursor plasmid”) was identified. Expression vectors for the complete enzyme and the individual subunits were prepared from the precursor plasmid as follows. A vector expressing only the small subunit, designated pEAL31, was constructed by digesting the precursor plasmid with EcoRI and NcoI, filling in the resulting overhangs with Klenow fragment and then religating with T4 ligase. The plasmid pEAL31/50, expressing the complete enzyme, was constructed from the precursor plasmid by shortening to 17 bp the distance between the Shine Delgarno box and the start of translation of the 50 K subunit; this was accomplished by digestion with Hind111 plus NcoI, removal of overhangs with Sl nuclease, and religation. The plasmid pEAL50, expressing only the large subunit, was constructed by cutting pEAL31/50 with EcoRI and AocI and then religating after filling in with Klenow fragment. All molecular genetic procedures were carried out according to standard protocols (8).
Purification
of Ethanolamine
Ammonia-Lyase
E. coli CAG626 transformed with pEAL31/50 were grown at 37°C with vigorous aeration in 90 liters of YT broth containing ampicillin
* Abbreviations used: IPTG, isopropyl thiogalactoside; PMSF, phenylmethylsulfonyl fluoride; SDS, sodium dodecyl sulfate; PAGE, plyacrylamide gel electrophoresis; FPLC, fast protein liquid chromatography. 0003-9861/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
AdoCbl-DEPENDENT
ETHANOLAMINE
AMMONIA-LYASE
(75 mg/liter) as selecting agent. When the culture reached an absorbance of 0.8 at 650 nm, solid IPTG was added to a final concentration of 1.3 mM. After four more hours of growth, the bacteria (yield ~1 kg) were harvested by centrifugation in a Sharples centrifuge, washed once with 0.02 M potassium phosphate buffer, pH 7.5, and frozen in liquid Ns. Step 1: Disruption of bacteria. Ninety-six grams of frozen bacteria were suspended in 192 ml 0.02 M potassium phosphate buffer, pH 7.5, containing 0.2 mM PMSF and disrupted by passage through a French press. Cell walls and unbroken bacteria were removed by centrifugation at 12,OOOgfor 20 min. The supernatant from this centrifugation constituted the starting material for the purification of the enzyme. Step 2: Removal of DNA and ammonium sulfate precipitation. To the supernatant (total volume, 205 ml) was added 20 ml streptomycin sulfate solution (10% [w/v] in water). The mixture was stirred for 10 min and then centrifuged (12,OOOg,20 min, 4°C) to remove precipitated DNA. Solid ammonium sulfate (16.4 g/100 ml) was added to the supernatant, which was stirred for 30 min to allow complete precipitation. The enzyme-containing precipitate was isolated by centrifugation, suspended in 100 ml of Buffer A (10 mM potassium phosphate buffer, pH 7.4, containing 10 mM ethanolamine * HCI, 10 mM KCl, 5 mM dithiothreitol, and 10% [v/v] glycerol), and dialyzed overnight against the same buffer. Step 3: Concentration. A dialysis bag was fitted tightly to the stem of a funnel inserted into a l-hole rubber stopper that was placed in the neck of a vacuum flask. The dialysis tubing was filled through the funnel with the solution obtained in step 2, and vacuum was applied to the flask, refilling the dialysis tubing as necessary until the solution obtained in the first step had been reduced to a total volume of 15 ml. On centrifuging the concentrated solution in a Beckman GPR refrigerated centrifuge at 3000 rpm for 20 min at 4”C, a translucent white precipitate of highly purified ethanolamine ammonia-lyase was obtained. This precipitate was washed once with Buffer B (Buffer A supplemented with 10 mM Na,EDTA and 1% (w/v) NaNa) and then suspended in the same buffer and dissolved by warming to 37°C. (Complete dissolution required enough buffer to achieve a final protein concentration of 1.5-2.0 mg/ ml). The entire purification was conducted at 4”C, and the solution of enzyme was stored at the same temperature.
Over-expression and Immunoblotting of the Individual Subunits of the S. typhimurium Ethanolumine Ammonia-Lyase To express the individual subunits, the expression plasmids described above were transformed into E. coli CAG626. The transformed bacteria were grown in 200-ml volumes of ampicillin-containing YT broth to an A,, of 0.9 and then induced with IPTG (final concentration, 1.3 mM) and grown for four more hours as described above. Bacteria were harvested by centrifugation, resuspended in 3 ml 0.02 M potassium phosphate buffer, pH 7.5, and disrupted by sonication (full power, five l-min bursts at O”C), followed by centrifugation (12,OOOgfor 20 min at 4°C) to remove membranes and unbroken cells. Controls included bacteria transformed with pEAL31/50 and pKQV4 (the original expression vector without an insert), grown and extracted in the same way. Protein concentrations in the extracts were as follows (mg/ml): pEAL31, 8.0; pEAL50, 4.6; pEAL31/50,5.4; and pKQV4,8.0. For immunoblotting of the pEAL31, pEAL50, and pEAL31/50 extracts, portions containing 150 pg protein were diluted with sample buffer and then subjected to SDS-PAGE as described below, except using Laemmli conditions (9). The proteins were transferred to a nitrocellulose membrane and detected with a 1:2000 dilution of a rabbit antiserum that had been raised against an extensively purified preparation of E. coli ethanolamine ammonia-lyase (6), visualizing the bound immunoglobulin with alkaline phosphatase using a kit purchased from Bio-Rad (10).
Other Methods SDS-PAGE was carried out using a 10% running gel prepared according to Laemmli and co-workers (9) and polymerized overnight, sur-
FROM
Salmonella
TABLE Purification
51
typhimurium I
of Ethanolamine
Ammonia-Lyase Yield
Step Bacterial lysate Ammonium sulfate precipitate Concentrate
Units
%
10,530 8,020 551
100 75 5.2
Specific activity (~mol/min/mg) 1.1 6.1 25.3
mounted by a stacking gel composed of 1.5% agarose in 0.75 M Tris, pH 6.8/10 mM dithiothreitol/lO mM Na thioglycolate/O.l% SDS.3 (An agarose stack was used because the 31 K subunit was often partly destroyed in the Laemmli stack.) Gels were stained with Coomassie blue and scanned with a Zeineh soft laser scanning densitometer. Protein was measured by the method of Bradford (11). Ethanolamine ammonialyase activity was assayed by a previously described radiochemical method (12). All experiments involving AdoCbl were carried out under Wratten 2A red safelights.
RESULTS Purification of S. typhimurium Ammonia-Lyase
Ethanolamine
In the course of working with the ethanolamine ammonia-lyase from S. typhimurium, it became apparent that the solubility of the enzyme was unusually low. Advantage was taken of this observation to develop a simple purification method that furnished pure enzyme in large quantities, albeit at relatively low yield. A typical purification is summarized in Table I, and SDS-PAGE analysis of materials obtained in the course of the purification is presented in Fig. 1. As expected from earlier experiments, the pure enzyme contained two subunits (M, -35 K and x50 K), both of which are clearly evident in the unfractionated bacterial extract. From the relative specific activities of the unfractionated extract and the pure enzyme, it can be calculated that about 4% of the protein in the extract consisted of ethanolamine ammonia-lyase.
Size and Subunit Composition of S. typhimurium Ethanolamine Ammonia-Lyme Gel filtration showed ethanolamine ammonia-lyase to be quite a large enzyme. When the purified enzyme was subjected to FPLC chromatography over a Superose 6 column calibrated with a group of standard proteins, it emerged at M, 480 K as determined both optically and by catalytic activity (Fig. 2). The ratio of the two subunits in the purified enzyme was estimated by densitometry of the Coomassie blue-stained gel. In scans of two different tracks, the areas corresponding to the 31 K peak were 84.5 and 85.0% of the areas corresponding to the 50 K peak. Taken in combination with the M, of 480 K deter3 Leslie E. Walker,
personal communication.
52
FAUST
AND
BABIOR
o.40 I I2
0.32
E 2
0.24
0; 5E& 0.16
46-
310.00 0.00
0.40
0.80
1.20
1.60
2.00
211 /[AdoCbl]
2.40
(PM)-’
o’40 4
14FIG. 1. SDS-PAGE analysis of ethanolamine ammonia-lyase at various stages of purification. (1) Extract (150 pg protein); (2) ammonium sulfate precipitate (50 pg protein); (3) concentrate (20 pg protein). The specific activity of the sample used for each track is indicated in Table I.
I -?
6.00
-
0.16
-
0.08
-
l
E D 0; El!. r-
mined by size exclusion chromatography, these data are most compatible with an enzyme composition of (Y,& (MW 486 kDa) or (Y&& (MW 548 kDa), where (Y and /I represent the 31 K and 50 K subunits, respectively. Because of the very close agreement between the observed M, and the molecular weight calculated from the a& structure and because polypeptides vary considerably in their ability to take up Coomassie blue, it seems most
0.32
> :
t 0.00 ’ 0
J 2
4
1 /[Enzyme]
6
8
(pg/ml)-’
FIG. 3. Kinetics. (top) Enzyme limiting. Reaction mixtures contained 5 pg (10.3 pmol) ethanolamine ammonia-lyase, 1 Nmol (40,000 cpm) [“C]ethanolamine . HCl, 50 ~1 Buffer B, and AdoCbl as indicated, in a total volume of 0.1 ml. (bottom) AdoCbl limiting. Reaction mixtures contained 15 pmol AdoCbl, 1 rmol(40,OOO cpm) [“C]ethanolamine * HCl, 50 pl Buffer B, and enzyme as indicated, in a total volume of 0.1 ml. Incubations were conducted at 22°C for 2 min, starting the reactions with AdoCbl. Reactions were terminated and the product was assayed as described under Materials and Methods.
likely that (Y,& represents the true structure of the enzyme. This structure will therefore be assumed in subsequent calculations of the molarity of solutions of the enzyme. 9
10
Retention
11
12
13
(mln)
FIG. 2. Molecular weight of the native enzyme. Puritied ethanolamine ammonia-lyase (0.65 mg in 0.5 ml Buffer B) was analyzed by FPLC over a Superose 12 column, eluting with Buffer A/150 mM KC1 at 0.15 ml/min and assaying for protein (Am) and activity. The column was standardized with 25 pl of a mixture of alcohol dehydrogenase (150 kDa), j3-amylase (200 kDa), apoferritin (443 kDa), and thyroglobulm (669 kDa), each dissolved in Buffer A/150 mM KC1 at the concentration recommended by the manufacturer, and eluted with the same buffer at 0.15 ml/min.
Kinetics and Active Sites Two approaches were used to determine the number of active sites in ethanolamine ammonia-lyase. In the first approach, saturation kinetics were conducted according to standard methods, except that two sets of measurements were made: one set with limiting enzyme (the usual conditions) and one set with limiting cofactor. From these results it was possible to calculate Michaelis constants and V,,, values for both ethanolamine and AdoCbl. The experimental results are shown in Fig. 3, and the kinetic
AdoCbl-DEPENDENT TABLE Kinetic
Component AdoCbl Enzyme
ETHANOLAMINE
AMMONIA-LYASE
FROM
Salmonella
53
typhimurium
II
Constants
(2, 0.51 + 0.05 0.31 If: 0.08’
V, (nmol/min/pmol) 3.31 f 0.17 6.78 + 0.35
Note. The experiments were conducted as described in Fig. 3. Results are expressed as the mean f 1 SD. a Expressed as active sites, assuming two active sites/molecule. 0
constants are presented in Table II. The V, values show that the rate of consumption of ethanolamine/mol of enzyme was twice the rate of consumption/mol AdoCbl, suggesting that a molecule of ethanolamine ammonialyase contains two active sites. Corroborating this interpretation is the similarity between the K,,, of the enzyme (expressed in terms of the concentration of active sites) and that of AdoCbl. In the second approach, the number of active sites in ethanolamine ammonia-lyase was determined by titration with CNCbl, a potent inhibitor of the enzyme. When this titration was carried out, the activity of the enzyme was found to decline in direct proportion to the amount of CNCbl added to the reaction mixture (Fig. 4). Full inhibition was achieved at a ratio of 1.97 mol of CNCbl/mol of enzyme. This result adds further support to the idea that ethanolamine ammonia-lyase contains two active sites/molecule of enzyme. Both Subunits of Ethunolamine Ammonia-Lyase Required for Catalytic Activity
Are
Purified ethanolamine ammonia-lyase preparations from each of three different bacterial species have been shown to contain two polypeptides, one of M, = 50 K and the other of M, ~35 K (2, 6, 13). The finding that ethanolamine ammonia-lyase activity in S. typhimurium is abolished by a mutation in the gene encoding either polypeptide suggests that both are required for catalytic activity (14). By using plasmids pEAL31 and pEAL50, which express the small and the large polypeptide of ethanolamine ammonia-lyase, respectively, we have been able to study this question more directly. We have found that E. coli transformed with these plasmids are able to produce the encoded polypeptides, but not in equivalent amounts (Fig. 5). The amount of large subunit in bacteria expressing pEAL50 is similar to the amount in bacteria expressing pEAL31/50 (i.e., the plasmid containing both ethanolamine ammonia-lyase genes), but in bacteria expressing pEAL31, the quantity of small subunit is at least an order of magnitude lower than the quantity in bacteria expressing pEAL31/50. These results suggest that in the E. coli host, the 31 K subunit of S. typhimurium
20
15
10
5
CNCbl
25
30
35
(pmoles)
FIG. 4. Titration of active sites with CNCM. CNCbl in the quantities indicated was preincubated for 10 min at 22°C with 4.2 pg (8.6 pmol) ethanolamine ammonia-lyase in 65 rl Buffer B. Assay mixtures were then made up to a final volume of 0.1 ml with 1 rmol(40,OOO dpm) [U“Clethanolamine . HCl and 0.25 nmol AdoCbl, adding the AdoCbl last. Incubations were carried out for 2 min at 22’C. Reactions were terminated and the product was assayed as described under Materials and Methods.
ethanolamine ammonia-lyase is relatively unstable unless it is combined with the 50 K subunit. Ethanolamine ammonia-lyase activity was measured in extracts from bacteria expressing the 50 K subunit alone, the 31 K subunit alone, or the complete enzyme. Before assay, the extract containing the complete enzyme was diluted 1:lOO with an extract containing no ethanolamine ammonia-lyase components (i.e., an extract from bacteria transformed with pKQV4); extracts containing the individual subunits were assayed without dilution. The results (Table III) showed that although considerable ethanolamine ammonia-lyase activity was present in the
Enzyme 50K Subunit
30K Subunit
AA&
----^
..-__
4002001005040020010050
30
-.. _I___ 10 3 1
l/Dilution FIG. 5. SDS-PAGE analysis of extracts from bacteria transformed with plasmids expressing various ethanolamine ammonia-lyase components. Extracts from bacteria transformed with pEAL31/50, pEAL31, or pEAL50 were diluted with sample buffer as indicated in the figure and immunoblotted as described under Materials and Methods.
54
FAUST TABLE
Ethanolamine
Ammonia-Lyase
Enzyme component Complete enzyme’ Large subunit Small subunit Control extract Complete enzymea Complete enzyme’
Plasmid pEAL31/50 pEAL50 pEAL31
PKQV~
AND
III
Activity
of Isolated Subunits
Diluent Control
extract -
Large subunit Small subunit
Product (nmol) 6.69 + 0.20
Large subunit + small subunitb
Note. Reaction mixtures contained 50 pl of enzyme component (diluted or mixed as indicated), 0.1 pmol(40,060 cpm) [“C]ethanolamine * HCl, and 7 nmol AdoCbl in a total volume of 0.1 ml. Incubations were started with AdoCbl and carried out for 2 min. Reactions were terminated and the product was assayed as described under Materials and Methods. Results are expressed as the mean + 1 SD. ’ Enzyme/diluent = l/100 (v/v). b 25 pl of each.
extract containing the complete enzyme, no activity was detected in extracts containing the isolated subunits, either before or after mixing. Activity was only slightly altered when the complete enzyme was diluted into the extracts containing the isolated subunits, indicating that the lack of activity of the isolated subunits was not due to an inhibitor, but was an intrinsic property of the subunits themselves. These findings indicate that both subunits are required for catalysis by the S. typhimurium ethanolamine ammonia-lyase. DISCUSSION
It is difficult to determine by enzymological methods the exact subunit composition of a molecule as large and complex as ethanolamine ammonia-lyase. The scatter of the measurements, the assumptions made in determining protein concentration, and the variation in dye binding by different polypeptides all introduce uncertainties into the conclusions. From the results presented here it is possible to state with a reasonable degree of confidence that the enzyme contains two active sites/molecule, but the subunit structure of (Y&, postulated according to what_ to us is the best interpretation of the present data, could
BABIOR
be subject to revision when studies by more definitive techniques (e.g., X-ray crystallography) become available. Our studies showed that the activity of the isolated subunits amounted to no more than 0.01% that of the complete enzyme, as estimated from the results presented in Table III and the low concentration of complete enzyme used in the experiments described in that table. Although this low activity could be attributed to denaturation, it is more likely to represent an intrinsic property of the subunits, since denaturation would have been expected to lead to the destruction of the subunits in the microorganism, but the immunoblot showed that they survived. Even though neither subunit is catalytically active, it will be of interest to see if one of them can mediate a mechanistically significant partial reaction (e.g., the cleavage of the C-Co bond of the cofactor) or if either can bind a cobalamin. ACKNOWLEDGMENTS We thank Drs. Earl R. Stadtman and J. Michael Poston for indispensable help with the growth of the transformed E. co& We are indebted to Julie M. Ruedi for valuable assistance and to Judith A. Connor for advice.
REFERENCES 1. Bradbeer, C. (1965) J. Bid. Chem. 240, 4675-4681. 2. Blackwell, C. M., and Turner, J. M. (1978) Biochem. J. 176, 555-563. 3. Jones, P. W., and Turner, J. M. (1984) J. Gen. Microbial. 130, 299-308. 4. Roof, D. M., and Roth, J. R. (1988) J. Bacterial. 170,3855. 5. Babior, B. M. (1988) Bidactors 1,21-26. 6. Faust, L. P., Connor, J. A., Roof, D. M., Hoch, J. A., and Babior, B. M. (1990) J. Biol. Chem. 266,12,462-12,466. 7. Strauch, M. A., Spiegelman, G. B., Perego, M., Johnson, W. C., and Hoch, J. A. (1989) EMBO J. 8, 1615-1621. 8. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) in Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 9. Cleveland, D. W., Fischer, S. G., Kirschner, M. W., and Laemmli, U. K. (1977) J. Bid. Chem. 262,1102-1106. 10. Gaastra, W. (1984) in Methods in Molecular Biology (Walker, J. M., Ed.), pp. 350-352, Humana Press, Clifton, NJ. 11. Bradford, M. M. (1976) And. Biochem. 72,248-254. 12. Babior, B. M., and Li, T. K. (1969) Biochemistry 8,154-160. 13. Wallis, 0. C., Johnson, A. W., and Lappert, M. F. (1979) FEBS L&t. 97,196-199. 14. Roof, D. M., and Roth, J. R. (1989) J. Bacterial. 171, 3316-3323.
OF BIOCHEMISTRY
AND BIOPHYSICS
Vol. 294, No. 1, April, pp. 50-54, 1992
Overexpression, Purification, and Some Properties of the AdoCbl-Dependent Ethanolamine Ammonia-Lyase from Salmonella typhimurium’ LaRosa P. Faust and Bernard M. Babior Department of Molecular and Experimental Medicine, The Scripps Research Institute, La Jolla, California 92037 Received September 3,199l
Recombinant ethanolamine ammonia-lyase from S. typhimurium has been overexpressed and purified in large quantities by a simple procedure. The molecular weight of the native enzyme is about 480 kDa, and it contains two active sites/molecule as shown by kinetic studies and by titration with CNCbl. Analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis confirms earlier cloning studies indicating that it is composed of two kinds of subunits, one of MW 3 1 kDa and the other of MW 50 kDa. These subunits, inactive by themselves, combine to produce an active enzyme whose composition is most likely &la. The IL, for AdoCbl is 0.6 fiM, and the turnover number is 65 s-l per active site at 22°C. o 1~02 Academic Press, Inc.
Ethanolamine ammonia-lyase is an enzyme found in several bacterial species that catalyzes the AdoCbl-dependent deamination of vicinal amino alcohols to 0x0 compounds and NH: (l-4). A great deal of work, carried out principally with the enzyme from an unclassified Clostridium, has shown that the reaction proceeds by a free radical mechanism in which the cofactor serves as a protected source of unpaired electrons (5). For further mechanistic studies, however, it would be desirable to have a recombinant enzyme that can be mutated at will. For this purpose, we have recently cloned and sequenced the genes encoding the ethanolamine ammonia-lyase from Salmonella typhimurium (6). We now report the overexpression and purification of this enzyme, and describe some of its properties. MATERIALS
AND
METHODS
The expression vector pKQV4 and Escherichie coli CAG626, the strain in which the recombinant plasmids were expressed, were the generous gifts of Dr. James A. Hoch. Restriction enzymes, T4 ligase, Klenow
1 Supported in part by USPHS 50
Grants GM-38050 and AI-24227.
fragment, Sl nuclease, isopropyl thiogalactoside (IPTG’), and other molecular biology supplies were purchased from Stratagene (La Jolla, CA), BRL Laboratories (Bethesda, MD), or Pharmacia (Piscataway, NJ). Agar and media were purchased from Difco (Detroit); AdoCbl, CNCbl, and ampicillin from Sigma (St. Louis); gel electrophoresis supplies and Bradford reagent from Bio-Rad (Richmond, VA); and [1,2-“Clethanolamine (2.4 mCi/mmol) from NEN (Boston, MA). Other reagents were of the best quality commercially available and were used without further purification.
Construction of Vectors That Overexpress the S. typhimurium Ethunolamine Ammoniu-Lyase and Its Subunits The genes for the AdoCbl-dependent ethanolamine ammonia-lyase from S. typhimurium occur as elements of an insert in the recombinant plasmid pBSE4.5, a derivative of pBR325 (6). The insert was excised from pBSE4.5 with Hind111 and ligated into the corresponding site of the expression vector pKQV4 (7). The new construct was transformed into E. coli XL-l Blue, and transformed bacteria were grown as individual clones on 1.5% YT agar (8) containing ampicillin (75 pg/ml). By mapping plasmids from several clones with HindI plus AocI, a plasmid containing the ethanolamine ammonia-lyase genes inserted in the sense direction (the “precursor plasmid”) was identified. Expression vectors for the complete enzyme and the individual subunits were prepared from the precursor plasmid as follows. A vector expressing only the small subunit, designated pEAL31, was constructed by digesting the precursor plasmid with EcoRI and NcoI, filling in the resulting overhangs with Klenow fragment and then religating with T4 ligase. The plasmid pEAL31/50, expressing the complete enzyme, was constructed from the precursor plasmid by shortening to 17 bp the distance between the Shine Delgarno box and the start of translation of the 50 K subunit; this was accomplished by digestion with Hind111 plus NcoI, removal of overhangs with Sl nuclease, and religation. The plasmid pEAL50, expressing only the large subunit, was constructed by cutting pEAL31/50 with EcoRI and AocI and then religating after filling in with Klenow fragment. All molecular genetic procedures were carried out according to standard protocols (8).
Purification
of Ethanolamine
Ammonia-Lyase
E. coli CAG626 transformed with pEAL31/50 were grown at 37°C with vigorous aeration in 90 liters of YT broth containing ampicillin
* Abbreviations used: IPTG, isopropyl thiogalactoside; PMSF, phenylmethylsulfonyl fluoride; SDS, sodium dodecyl sulfate; PAGE, plyacrylamide gel electrophoresis; FPLC, fast protein liquid chromatography. 0003-9861/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
AdoCbl-DEPENDENT
ETHANOLAMINE
AMMONIA-LYASE
(75 mg/liter) as selecting agent. When the culture reached an absorbance of 0.8 at 650 nm, solid IPTG was added to a final concentration of 1.3 mM. After four more hours of growth, the bacteria (yield ~1 kg) were harvested by centrifugation in a Sharples centrifuge, washed once with 0.02 M potassium phosphate buffer, pH 7.5, and frozen in liquid Ns. Step 1: Disruption of bacteria. Ninety-six grams of frozen bacteria were suspended in 192 ml 0.02 M potassium phosphate buffer, pH 7.5, containing 0.2 mM PMSF and disrupted by passage through a French press. Cell walls and unbroken bacteria were removed by centrifugation at 12,OOOgfor 20 min. The supernatant from this centrifugation constituted the starting material for the purification of the enzyme. Step 2: Removal of DNA and ammonium sulfate precipitation. To the supernatant (total volume, 205 ml) was added 20 ml streptomycin sulfate solution (10% [w/v] in water). The mixture was stirred for 10 min and then centrifuged (12,OOOg,20 min, 4°C) to remove precipitated DNA. Solid ammonium sulfate (16.4 g/100 ml) was added to the supernatant, which was stirred for 30 min to allow complete precipitation. The enzyme-containing precipitate was isolated by centrifugation, suspended in 100 ml of Buffer A (10 mM potassium phosphate buffer, pH 7.4, containing 10 mM ethanolamine * HCI, 10 mM KCl, 5 mM dithiothreitol, and 10% [v/v] glycerol), and dialyzed overnight against the same buffer. Step 3: Concentration. A dialysis bag was fitted tightly to the stem of a funnel inserted into a l-hole rubber stopper that was placed in the neck of a vacuum flask. The dialysis tubing was filled through the funnel with the solution obtained in step 2, and vacuum was applied to the flask, refilling the dialysis tubing as necessary until the solution obtained in the first step had been reduced to a total volume of 15 ml. On centrifuging the concentrated solution in a Beckman GPR refrigerated centrifuge at 3000 rpm for 20 min at 4”C, a translucent white precipitate of highly purified ethanolamine ammonia-lyase was obtained. This precipitate was washed once with Buffer B (Buffer A supplemented with 10 mM Na,EDTA and 1% (w/v) NaNa) and then suspended in the same buffer and dissolved by warming to 37°C. (Complete dissolution required enough buffer to achieve a final protein concentration of 1.5-2.0 mg/ ml). The entire purification was conducted at 4”C, and the solution of enzyme was stored at the same temperature.
Over-expression and Immunoblotting of the Individual Subunits of the S. typhimurium Ethanolumine Ammonia-Lyase To express the individual subunits, the expression plasmids described above were transformed into E. coli CAG626. The transformed bacteria were grown in 200-ml volumes of ampicillin-containing YT broth to an A,, of 0.9 and then induced with IPTG (final concentration, 1.3 mM) and grown for four more hours as described above. Bacteria were harvested by centrifugation, resuspended in 3 ml 0.02 M potassium phosphate buffer, pH 7.5, and disrupted by sonication (full power, five l-min bursts at O”C), followed by centrifugation (12,OOOgfor 20 min at 4°C) to remove membranes and unbroken cells. Controls included bacteria transformed with pEAL31/50 and pKQV4 (the original expression vector without an insert), grown and extracted in the same way. Protein concentrations in the extracts were as follows (mg/ml): pEAL31, 8.0; pEAL50, 4.6; pEAL31/50,5.4; and pKQV4,8.0. For immunoblotting of the pEAL31, pEAL50, and pEAL31/50 extracts, portions containing 150 pg protein were diluted with sample buffer and then subjected to SDS-PAGE as described below, except using Laemmli conditions (9). The proteins were transferred to a nitrocellulose membrane and detected with a 1:2000 dilution of a rabbit antiserum that had been raised against an extensively purified preparation of E. coli ethanolamine ammonia-lyase (6), visualizing the bound immunoglobulin with alkaline phosphatase using a kit purchased from Bio-Rad (10).
Other Methods SDS-PAGE was carried out using a 10% running gel prepared according to Laemmli and co-workers (9) and polymerized overnight, sur-
FROM
Salmonella
TABLE Purification
51
typhimurium I
of Ethanolamine
Ammonia-Lyase Yield
Step Bacterial lysate Ammonium sulfate precipitate Concentrate
Units
%
10,530 8,020 551
100 75 5.2
Specific activity (~mol/min/mg) 1.1 6.1 25.3
mounted by a stacking gel composed of 1.5% agarose in 0.75 M Tris, pH 6.8/10 mM dithiothreitol/lO mM Na thioglycolate/O.l% SDS.3 (An agarose stack was used because the 31 K subunit was often partly destroyed in the Laemmli stack.) Gels were stained with Coomassie blue and scanned with a Zeineh soft laser scanning densitometer. Protein was measured by the method of Bradford (11). Ethanolamine ammonialyase activity was assayed by a previously described radiochemical method (12). All experiments involving AdoCbl were carried out under Wratten 2A red safelights.
RESULTS Purification of S. typhimurium Ammonia-Lyase
Ethanolamine
In the course of working with the ethanolamine ammonia-lyase from S. typhimurium, it became apparent that the solubility of the enzyme was unusually low. Advantage was taken of this observation to develop a simple purification method that furnished pure enzyme in large quantities, albeit at relatively low yield. A typical purification is summarized in Table I, and SDS-PAGE analysis of materials obtained in the course of the purification is presented in Fig. 1. As expected from earlier experiments, the pure enzyme contained two subunits (M, -35 K and x50 K), both of which are clearly evident in the unfractionated bacterial extract. From the relative specific activities of the unfractionated extract and the pure enzyme, it can be calculated that about 4% of the protein in the extract consisted of ethanolamine ammonia-lyase.
Size and Subunit Composition of S. typhimurium Ethanolamine Ammonia-Lyme Gel filtration showed ethanolamine ammonia-lyase to be quite a large enzyme. When the purified enzyme was subjected to FPLC chromatography over a Superose 6 column calibrated with a group of standard proteins, it emerged at M, 480 K as determined both optically and by catalytic activity (Fig. 2). The ratio of the two subunits in the purified enzyme was estimated by densitometry of the Coomassie blue-stained gel. In scans of two different tracks, the areas corresponding to the 31 K peak were 84.5 and 85.0% of the areas corresponding to the 50 K peak. Taken in combination with the M, of 480 K deter3 Leslie E. Walker,
personal communication.
52
FAUST
AND
BABIOR
o.40 I I2
0.32
E 2
0.24
0; 5E& 0.16
46-
310.00 0.00
0.40
0.80
1.20
1.60
2.00
211 /[AdoCbl]
2.40
(PM)-’
o’40 4
14FIG. 1. SDS-PAGE analysis of ethanolamine ammonia-lyase at various stages of purification. (1) Extract (150 pg protein); (2) ammonium sulfate precipitate (50 pg protein); (3) concentrate (20 pg protein). The specific activity of the sample used for each track is indicated in Table I.
I -?
6.00
-
0.16
-
0.08
-
l
E D 0; El!. r-
mined by size exclusion chromatography, these data are most compatible with an enzyme composition of (Y,& (MW 486 kDa) or (Y&& (MW 548 kDa), where (Y and /I represent the 31 K and 50 K subunits, respectively. Because of the very close agreement between the observed M, and the molecular weight calculated from the a& structure and because polypeptides vary considerably in their ability to take up Coomassie blue, it seems most
0.32
> :
t 0.00 ’ 0
J 2
4
1 /[Enzyme]
6
8
(pg/ml)-’
FIG. 3. Kinetics. (top) Enzyme limiting. Reaction mixtures contained 5 pg (10.3 pmol) ethanolamine ammonia-lyase, 1 Nmol (40,000 cpm) [“C]ethanolamine . HCl, 50 ~1 Buffer B, and AdoCbl as indicated, in a total volume of 0.1 ml. (bottom) AdoCbl limiting. Reaction mixtures contained 15 pmol AdoCbl, 1 rmol(40,OOO cpm) [“C]ethanolamine * HCl, 50 pl Buffer B, and enzyme as indicated, in a total volume of 0.1 ml. Incubations were conducted at 22°C for 2 min, starting the reactions with AdoCbl. Reactions were terminated and the product was assayed as described under Materials and Methods.
likely that (Y,& represents the true structure of the enzyme. This structure will therefore be assumed in subsequent calculations of the molarity of solutions of the enzyme. 9
10
Retention
11
12
13
(mln)
FIG. 2. Molecular weight of the native enzyme. Puritied ethanolamine ammonia-lyase (0.65 mg in 0.5 ml Buffer B) was analyzed by FPLC over a Superose 12 column, eluting with Buffer A/150 mM KC1 at 0.15 ml/min and assaying for protein (Am) and activity. The column was standardized with 25 pl of a mixture of alcohol dehydrogenase (150 kDa), j3-amylase (200 kDa), apoferritin (443 kDa), and thyroglobulm (669 kDa), each dissolved in Buffer A/150 mM KC1 at the concentration recommended by the manufacturer, and eluted with the same buffer at 0.15 ml/min.
Kinetics and Active Sites Two approaches were used to determine the number of active sites in ethanolamine ammonia-lyase. In the first approach, saturation kinetics were conducted according to standard methods, except that two sets of measurements were made: one set with limiting enzyme (the usual conditions) and one set with limiting cofactor. From these results it was possible to calculate Michaelis constants and V,,, values for both ethanolamine and AdoCbl. The experimental results are shown in Fig. 3, and the kinetic
AdoCbl-DEPENDENT TABLE Kinetic
Component AdoCbl Enzyme
ETHANOLAMINE
AMMONIA-LYASE
FROM
Salmonella
53
typhimurium
II
Constants
(2, 0.51 + 0.05 0.31 If: 0.08’
V, (nmol/min/pmol) 3.31 f 0.17 6.78 + 0.35
Note. The experiments were conducted as described in Fig. 3. Results are expressed as the mean f 1 SD. a Expressed as active sites, assuming two active sites/molecule. 0
constants are presented in Table II. The V, values show that the rate of consumption of ethanolamine/mol of enzyme was twice the rate of consumption/mol AdoCbl, suggesting that a molecule of ethanolamine ammonialyase contains two active sites. Corroborating this interpretation is the similarity between the K,,, of the enzyme (expressed in terms of the concentration of active sites) and that of AdoCbl. In the second approach, the number of active sites in ethanolamine ammonia-lyase was determined by titration with CNCbl, a potent inhibitor of the enzyme. When this titration was carried out, the activity of the enzyme was found to decline in direct proportion to the amount of CNCbl added to the reaction mixture (Fig. 4). Full inhibition was achieved at a ratio of 1.97 mol of CNCbl/mol of enzyme. This result adds further support to the idea that ethanolamine ammonia-lyase contains two active sites/molecule of enzyme. Both Subunits of Ethunolamine Ammonia-Lyase Required for Catalytic Activity
Are
Purified ethanolamine ammonia-lyase preparations from each of three different bacterial species have been shown to contain two polypeptides, one of M, = 50 K and the other of M, ~35 K (2, 6, 13). The finding that ethanolamine ammonia-lyase activity in S. typhimurium is abolished by a mutation in the gene encoding either polypeptide suggests that both are required for catalytic activity (14). By using plasmids pEAL31 and pEAL50, which express the small and the large polypeptide of ethanolamine ammonia-lyase, respectively, we have been able to study this question more directly. We have found that E. coli transformed with these plasmids are able to produce the encoded polypeptides, but not in equivalent amounts (Fig. 5). The amount of large subunit in bacteria expressing pEAL50 is similar to the amount in bacteria expressing pEAL31/50 (i.e., the plasmid containing both ethanolamine ammonia-lyase genes), but in bacteria expressing pEAL31, the quantity of small subunit is at least an order of magnitude lower than the quantity in bacteria expressing pEAL31/50. These results suggest that in the E. coli host, the 31 K subunit of S. typhimurium
20
15
10
5
CNCbl
25
30
35
(pmoles)
FIG. 4. Titration of active sites with CNCM. CNCbl in the quantities indicated was preincubated for 10 min at 22°C with 4.2 pg (8.6 pmol) ethanolamine ammonia-lyase in 65 rl Buffer B. Assay mixtures were then made up to a final volume of 0.1 ml with 1 rmol(40,OOO dpm) [U“Clethanolamine . HCl and 0.25 nmol AdoCbl, adding the AdoCbl last. Incubations were carried out for 2 min at 22’C. Reactions were terminated and the product was assayed as described under Materials and Methods.
ethanolamine ammonia-lyase is relatively unstable unless it is combined with the 50 K subunit. Ethanolamine ammonia-lyase activity was measured in extracts from bacteria expressing the 50 K subunit alone, the 31 K subunit alone, or the complete enzyme. Before assay, the extract containing the complete enzyme was diluted 1:lOO with an extract containing no ethanolamine ammonia-lyase components (i.e., an extract from bacteria transformed with pKQV4); extracts containing the individual subunits were assayed without dilution. The results (Table III) showed that although considerable ethanolamine ammonia-lyase activity was present in the
Enzyme 50K Subunit
30K Subunit
AA&
----^
..-__
4002001005040020010050
30
-.. _I___ 10 3 1
l/Dilution FIG. 5. SDS-PAGE analysis of extracts from bacteria transformed with plasmids expressing various ethanolamine ammonia-lyase components. Extracts from bacteria transformed with pEAL31/50, pEAL31, or pEAL50 were diluted with sample buffer as indicated in the figure and immunoblotted as described under Materials and Methods.
54
FAUST TABLE
Ethanolamine
Ammonia-Lyase
Enzyme component Complete enzyme’ Large subunit Small subunit Control extract Complete enzymea Complete enzyme’
Plasmid pEAL31/50 pEAL50 pEAL31
PKQV~
AND
III
Activity
of Isolated Subunits
Diluent Control
extract -
Large subunit Small subunit
Product (nmol) 6.69 + 0.20
Large subunit + small subunitb
Note. Reaction mixtures contained 50 pl of enzyme component (diluted or mixed as indicated), 0.1 pmol(40,060 cpm) [“C]ethanolamine * HCl, and 7 nmol AdoCbl in a total volume of 0.1 ml. Incubations were started with AdoCbl and carried out for 2 min. Reactions were terminated and the product was assayed as described under Materials and Methods. Results are expressed as the mean + 1 SD. ’ Enzyme/diluent = l/100 (v/v). b 25 pl of each.
extract containing the complete enzyme, no activity was detected in extracts containing the isolated subunits, either before or after mixing. Activity was only slightly altered when the complete enzyme was diluted into the extracts containing the isolated subunits, indicating that the lack of activity of the isolated subunits was not due to an inhibitor, but was an intrinsic property of the subunits themselves. These findings indicate that both subunits are required for catalysis by the S. typhimurium ethanolamine ammonia-lyase. DISCUSSION
It is difficult to determine by enzymological methods the exact subunit composition of a molecule as large and complex as ethanolamine ammonia-lyase. The scatter of the measurements, the assumptions made in determining protein concentration, and the variation in dye binding by different polypeptides all introduce uncertainties into the conclusions. From the results presented here it is possible to state with a reasonable degree of confidence that the enzyme contains two active sites/molecule, but the subunit structure of (Y&, postulated according to what_ to us is the best interpretation of the present data, could
BABIOR
be subject to revision when studies by more definitive techniques (e.g., X-ray crystallography) become available. Our studies showed that the activity of the isolated subunits amounted to no more than 0.01% that of the complete enzyme, as estimated from the results presented in Table III and the low concentration of complete enzyme used in the experiments described in that table. Although this low activity could be attributed to denaturation, it is more likely to represent an intrinsic property of the subunits, since denaturation would have been expected to lead to the destruction of the subunits in the microorganism, but the immunoblot showed that they survived. Even though neither subunit is catalytically active, it will be of interest to see if one of them can mediate a mechanistically significant partial reaction (e.g., the cleavage of the C-Co bond of the cofactor) or if either can bind a cobalamin. ACKNOWLEDGMENTS We thank Drs. Earl R. Stadtman and J. Michael Poston for indispensable help with the growth of the transformed E. co& We are indebted to Julie M. Ruedi for valuable assistance and to Judith A. Connor for advice.
REFERENCES 1. Bradbeer, C. (1965) J. Bid. Chem. 240, 4675-4681. 2. Blackwell, C. M., and Turner, J. M. (1978) Biochem. J. 176, 555-563. 3. Jones, P. W., and Turner, J. M. (1984) J. Gen. Microbial. 130, 299-308. 4. Roof, D. M., and Roth, J. R. (1988) J. Bacterial. 170,3855. 5. Babior, B. M. (1988) Bidactors 1,21-26. 6. Faust, L. P., Connor, J. A., Roof, D. M., Hoch, J. A., and Babior, B. M. (1990) J. Biol. Chem. 266,12,462-12,466. 7. Strauch, M. A., Spiegelman, G. B., Perego, M., Johnson, W. C., and Hoch, J. A. (1989) EMBO J. 8, 1615-1621. 8. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) in Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 9. Cleveland, D. W., Fischer, S. G., Kirschner, M. W., and Laemmli, U. K. (1977) J. Bid. Chem. 262,1102-1106. 10. Gaastra, W. (1984) in Methods in Molecular Biology (Walker, J. M., Ed.), pp. 350-352, Humana Press, Clifton, NJ. 11. Bradford, M. M. (1976) And. Biochem. 72,248-254. 12. Babior, B. M., and Li, T. K. (1969) Biochemistry 8,154-160. 13. Wallis, 0. C., Johnson, A. W., and Lappert, M. F. (1979) FEBS L&t. 97,196-199. 14. Roof, D. M., and Roth, J. R. (1989) J. Bacterial. 171, 3316-3323.