Target Size of D-P-Hydroxybutyrate Dehydrogenase FUNCTIONAL AND STRUCTURAL MOLECULAR WEIGHT BASED ON RADIATION INACTIVATION*

D-P-Hydroxybutyrate dehydrogenase is a lipid-re- quiring enzyme which is localized on the inner face of the mitochondrial inner membrane. The apoenzyme has been purified to homogeneity from beef heart; it is devoid of lipid and inactive. It can be functionally re- constituted with lecithin or phospholipid mixtures containing lecithin. The active form of the enzyme is the enzyme-phospholipid complex. Classical target analysis of radiation-inactivation data has now been used to determine the molecular size of the enzyme both in the native membrane (submitochondrial vesicles) and in the reconstituted enzyme inserted into phospholipid vesicles containing lecithin. For both forms of the en- zyme, we find the same molecular size, -110,000 daltons. This size is consistent with a tetramer. Radiation results in fragmentation of the polypeptide and the destruction of the polypeptide correlates with loss of enzymic function. A similar size is obtained when purified D-p-hydroxybutyrate dehydrogenase is inserted into a nonactivating mixture of phospholipid (ie. in the absence of lecithin). We conclude that: 1) the native enzyme in submitochondrial vesicles and the purified active enzyme in phospholipid vesicles are the same size, approximating a tetramer; 2) radiation of D-P-hydroxybutyrate

D-P-Hydroxybutyrate dehydrogenase is a lipid-requiring enzyme which is localized on the inner face of the mitochondrial inner membrane. The apoenzyme has been purified to homogeneity from beef heart; it is devoid of lipid and inactive. It can be functionally reconstituted with lecithin or phospholipid mixtures containing lecithin. The active form of the enzyme is the enzyme-phospholipid complex. Classical target analysis of radiation-inactivation data has now been used to determine the molecular size of the enzyme both in the native membrane (submitochondrial vesicles) and in the reconstituted enzyme inserted into phospholipid vesicles containing lecithin. For both forms of the enzyme, we find the same molecular size, -110,000 daltons. This size is consistent with a tetramer. Radiation results in fragmentation of the polypeptide and the destruction of the polypeptide correlates with loss of enzymic function. A similar size is obtained when purified D-p-hydroxybutyrate dehydrogenase is inserted into a nonactivating mixture of phospholipid ( i e . in the absence of lecithin). We conclude that: 1) the native enzyme in submitochondrial vesicles and the purified active enzyme in phospholipid vesicles are the same size, approximating a tetramer; 2) radiation of D-Phydroxybutyrate dehydrogenase results in loss of activity and fragmentation of the polypeptide; and 3) the role of lecithin in activation of D-p-hydroxybutyrate dehydrogenase is unrelated to determining oligomeric size of the enzyme since both active and nonactive forms exhibit the same structural size. D-P-Hydroxybutyrate dehydrogenase (D-3-hydroxybutyrate:NAD' oxidoreductase, EC 1.1.1.30) is a lipid-requiring enzyme with a specific requirement of lecithin for enzymic activity. It has been found to be localized on the inner face of the mitochondrial inner membrane (McIntyre et al., 1978a). The enzyme has been purified from bovine heart mitochondria to homogeneity as the soluble, phospholipid-free apodehydrogenase which is inactive Fleischer, 1974, 1975). When D-P-hydroxybutyrate apodehydrogenase is admixed with phospholipid vesicles, it appears to insert unidirectionally into the outer leaflet of the bilayer. When lecithin is present Health Grants AM14632 and AM13376 and by a Biomedical Research * These studies were supported in part by National Institutes of Support Grant from Vanderbilt University. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 3 Supported swiss National Science Foundation Fellowship in the vesicles, an active enzyme-phospholipid complex is formed (Fleischer et al., 1974). The enzyme can also be activated with soluble lecithins, below the critical micellar concentration (Gazzotti et al., 1975). Analysis of the activation of the apodehydrogenase with soluble lecithins indicates that two lecithins per functional unit are required for activation (Cortese et al., 1982). The kinetic mechanism of D-/3-hydroxybutyrate dehydrogenase is consistent with an ordered sequential reaction mechanism with kinetic parameters being similar for the enzyme in the native membrane and for the purified enzyme reactivated with MPL' (Nielsen et al., 1973;Brenner et al.,).* Therefore, the isolation procedure does not appear to have significantly altered the enzyme and there does not appear to be substantial influence of other protein components of the mitochondrial membrane on enzymic catalysis (Nielsen et ul., 1973).
The subunit size of D-P-hydroxybutyrate dehydrogenase is 31,000 daltons as determined by polyacrylamide gel electrophoresis in sodium dodecyl sulfate . The hydrodynamic characteristics of both the inactive soluble apodehydrogenase and of the enzyme reactivated with short chain soluble lecithin (i.e. below the critical micellar concentration), indicated that the protein undergoes concentrationdependent self-association (McIntyre et al., 1978b). Active enzyme sedimentation indicates that the active form of the enzyme in the presence of soluble dioctanoyl lecithin is a dimer. These studies, together with NADH binding studies (1 mol of NADH binds per 2 mol of monomer, Gazzotti et al., 1974), indicated that the minimum functional size is a dimer of identical subunits. However, hydrodynamic techniques cannot be used to study either the size of the native enzyme in the mitochondrial membrane or the size of the purified enzyme inserted into natural phospholipid vesicles.
In the present work, we have used the radiation-inactivation technique (Pollard, 1953) to measure the molecular size of D-P-hydroxybutyrate dehydrogenase in the membrane (submitochondrial vesicles) as well as for the purified enzyme reconstituted in phospholipid vesicles. Further, we have compared the size of the reactivated enzyme with that of the enzyme inserted into a nonactivating mixture of phospholipids, i.e. in the absence of lecithin.3 ' The abbreviations used are: MPL, the mixture of mitochondrial phospholipids extracted from beef heart mitochondria ; Hepes, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid; PAGE, polyacrylamide gel electrophoresis.

953
This is an Open Access article under the CC BY license.

EXPERIMENTAL PROCEDURES
Sucrose, density gradient grade, was obtained from Scbwartz/ Mann. Glucose-6-phosphate dehydrogenase (EC 1.1.1.49) from Leuconostoc mesenteroides (Type XXIV), DL-P-hydroxybutyric acid (sodium salt), bovine serum albumin (Fraction V., powder), antimycin A, and Tris were obtained from Sigma. Bovine plasma albumin solution, from Armour, was used as the protein standard. NAD' and dithiothreitol were obtained from Chemical Dynamics Corp. (South Plainfield, NJ). NAD' was also obtained from P-L Biochemicals, and Hepes from Calbiochem. All other chemicals were reagent grade. All solutions were prepared in deionized water.
Assays-Protein was measured by the procedure of Lowry et al. (1951) wi.th bovine plasma albumin as protein standard. When dithiothreitol was present in the sample, the assay was carried out as described by Ross and Schatz (1973) using iodoacetate to carboxymethylate the dithiothreitol which would otherwise interfere with the assay for protein. Phosphorus was measured using a modification  of the method of Chen et ul. (1956). D-P-Hydroxybutyrate dehydrogenase activity was measured spectrophotometrically as the rate of reduction of NAD' with D-,L-hydroxybutyrate (sodium salt) as substrate as described previously . The activity was measured using the complex assay method in which a complex of the apodehydrogenase with MPL vesicles was preformed by incubation of the apoenzyme with the lipid vesicles as described below. An aliquot of the complex was added to the cuvette which was preincubated to 37 "C and contained 1 ml (final volume) of assay medium (10 mM potassium phosphate (pH 7.351, 0.5 mM EDTA, 0.4 mg/ml of bovine serum albumin, 1.27% (v/v) ethanol, 0.3 mM dithiothreitol and 5 mM NAD'). After 1 min incubation the reaction was started by addition of substrate DL-P-hydroxybutyrate, to a final concentration of 20 mM. The initial rate of NADH production was measured a t 340 nm and the D-P-hydroxybutyrate dehydrogenase specific activity of the sample was calculated taking the extinction coefficient of NADH to be 6.22 mM" cm". When submitochondrial vesicles were assayed, the method was modified. The sample (30 pI of 2 mg/ml) was added to the cuvette containing assay medium without substrate and then a 5pl aliquot of antimycin A (0.1 mg/ml of ethanol) was added to block reoxidation of NADH by electron transport. After 2 min of incubation, the enzymic reaction was initiated by the addition of 0.1 ml of 0.2 M DL-,L-hydroxybutyrate. Glucose-6-phosphate dehydrogenase was assayed at 32 "C in 50 mM Tris-HCI (pH 7.8 at 25 "C), 3 mM MgCI2, 3 mM NAD', and 3.3 mM glucose-6-phosphate (Olive and Levy, 1971).
Preparation of Suhrnitochondrial Vesicles, D -@-Hydroxybutyrate Dehydrogenase, and Phospholipid Vesicles-The heavy fraction of bovine heart mitochondria was prepared on a large scale (Blair, 1967) with modifications described previously (Bock and Fleischer, 1974) and were stored frozen a t -80 "C. Submitochondrial vesicles were prepared from these mitochondria by nitrogen compression- deet al. (1974b).
compression and shear, using the Parr bomb, as described by Fleischer Mitochondrial lipids were prepared by extraction of bovine heart mitochondria with chloroform/methanol (2:l) and then back extracted to remove non-lipid materials (Folch as described in . Mitochondrial phospholipids were prepared from the total lipid extract by separation of the neutral lipid using silicic acid chromatography . Liposomes of MPL were prepared by dialysis of a butanol/cholate solution of MPL versus 20 mM Tris-HCI, 1 mM EDTA, pH 8.1, as described previously . Single molecular species of phospholipids (dioleoylphosphatidyldimethylethanolamine, dioleoylphosphatidylethanolamine, and according to the methods of Eibl (1978Eibl ( , 1980Eibl ( , 1981aEibl ( , and 1981b).
Enzymic activity was measured following 1 h preincubation at 25 "C to obtain optimal activity, after which time activity remained constant (Isaacson et al., 1979). The specific activity of the purified enzyme preparations used in these studies ranged between 115 and 135 pmol of NAD' reduced/min/mg of protein, a t 37 "C, when optimally activated with MPL. The enzyme, added to phospholipids in aqueous buffer, appears to insert unidirectionally into the outer face of the phospholipid vesicles (McIntyre et al., 1979).
Preparation of Samples for Radiation-The effect of irradiation on the native enzyme in the mitochondrial inner membrane was determined using submitochondrial vesicles. In intact mitochondria, the mitochondrial inner membrane presents a permeability barrier for NAD(H) to reach the enzyme, which is localized on the inner face, thus complicating the measurement of enzymic activity (McIntyre et al., 1978a). For this reason, submitochondrial vesicles were used since enzymic activity is readily measured in such inside-out vesicles which have their matrix face exposed to the medium (Fleischer, et al. 1974b). Submitochondrial vesicles were diluted to 2 mg of protein/ml in 0.25 M sucrose, 5 mM Hepes (pH 7.3), 1 mM EDTA, and 5 mM dithiothreitol. D-P-Hydroxybutyrate dehydrogenase (0.96 mg) was activated with MPL vesicles (58 or 174 pg of P) which had previously been centrifuged a t 100,OOO X g for 60 min and filtered through a 0.22-p Triton-free filter (GSTF filter, Millipore Corp., Bedford, MA) in a total volume of 9.6 ml of 0.25 M sucrose, 5 mM Hepes (pH 7.3), 1 mM EDTA, and 5 mM dithiothreitol. After incubation for 1 h at 25 "C, the active enzyme-phospholipid complex was then diluted with the same buffer to 24 ml. Formation of an inactive enzyme-phospholipid complex was achieved in like manner except that, instead of MPL, a nonactivating ternary mixture of phospholipids was used (see above).
This mixture of lipids without lecithin did not activate the enzyme but forms a nonactive complex (Gazzotti, et al., 1975). Glucose-6phosphate dehydrogenase, used as an internal standard was diluted into 0.3 M sucrose, 0.1 M KC1, and 5 mM Hepes (pH 7.0) to a concentration of 5 units/ml. For each preparation, 2-ml aliquots were plated out to a depth of 1.14 mm in open aluminum trays (surface area -17.6 c d ) . T h e trays were each frozen by mounting them level over an insulated channel which was subsequently filled with liquid nitrogen. Freezing time was approximately 10-15 s per tray. The trays were shipped to Buffalo and maintained in a liquid nitrogen refrigerator for 24-48 b prior to irradiation. Samples were then shipped back in like manner to Nashville for analysis.
Radiation Inactivation Procedure-The sample in trays was itradiated in the frozen state in a Van de Graff generator, producing a 0.5-mA beam of 1.5 MeV electrons. Radiation dose was measured using the transmittance change of blue cellophane calibrated against a chemical dosimeter (Fricke and Hart, 1966). The dose for each sample was controlled by varying the number of passes through the electron beam from zero to 16 with a measured dose of between 0.55 and 0.65 Mrad/pass. The radiation chamber was cooled with a stream of liquid nitrogen and the sample temperature (-40 to -50 "C) was monitored as described previously (Saccomani et al., 1981). Samples of either submitochondrial vesicles or enzyme-MPL or enzyme-nonactivating phospholipid complexes or glucose-6-phosphate dehydrogenase standard, were irradiated in pairs, so that reliable comparisons could be made between different sample preparations. After radiation, the trays were stored in a liquid nitrogen refrigerator for 24 h. The samples were then thawed and enzymic activities were promptly measured. Within experimental error, the samples were quantitatively recovered from the plates after thawing, as estimated by measurement of phosphorus concentration of samples with various radiation doses. Thus, loss of activity after radiation was not due to loss of sample. Storage of the samples in the Liquid nitrogen refrigerator for up to 4 days did not affect the results. Measured enzymic activities were stable for samples stored on ice, for up to 6 h after thawing. Identical results could be obtained after quick-freezing and storage of the samples in liquid nitrogen for up to 1 month. Control samples either simply frozen and thawed using the same protocol or manipulated through the Van de Graff generator with the radiation beam turned off ("sham" irradiated), showed no loss of activity compared with the initial activity prior to freezing.
Data Presentation and Calculations-The data were analyzed with a single target, single hit model of radiation inactivation (Pollard, 1953). Plots of the logarithm of the percentage of surviving enzymic activity were linear over one order of magnitude with correlation coefficients >0.99 (>0.98 for PAGE data). The data were fitted by least-squares with the value for the control sample weighted as twice the other values since the error in this point is smaller. Apparent molecular weights were calculated using the formula of Kepner and Macey (19681, i.e. molecular weight = 6.4 x IO"/D:E (rads), where DJj is the dose of absorbed radiation required to reduce the enzyme activity or PAGE hand intensity to 37% of the original.
Quantitation by Polyacrylamide Gel ELectrophoresis-The amount of purified D-8-hydroxybutyrate dehydrogenase remaining in samples after irradiation was quantitated on gels after silver staining. Polyacrylamide gel electrophoresis in sodium dodecyl sulfate was carried out according to Lammli (1970) using 159 acrylamide slab gels. Samples of D-8-hydroxybutyrate dehydrogenase-phospholipid complexes, which had previously been irradiated, were prepared for PAGE as follows: Each sample (75 p1 of 40 pg of D-8-hydroxybutyrate dehydrogenase/ml) was diluted with 25 p1 of sample buffer (2 M Tris-HCI, pH 6.8, 8.0% sodium dodecyl sulfate, 22% (v/v) P-mercaptoethanol, 22% (v/v) glycerol and 0.22 mg bromophenol blue/ml) and incubated for 30 min at room temperature; 15 pI (450 ng enzyme) of each sample was applied to the gel. Sample denaturation appeared to he complete after this treatment since similar results were obtained with samples denatured by boiling in sample buffer, although there was greater error introduced by variations in the sample volume after boiling. Electrophoresis (1.5 milliwatts/gel) was carried out using a Hoefer slab gel apparatus (Hoefer Scientific Instruments, San Francisco, CA) with cooling. Electrophoresis was terminated when the tracking dye (bromophenol blue) reached the end of the gel (-4 h). Gels were subsequently stained with silver according to Merril et al. (1982). The gels were then scanned at 622 nm with an E-C densitometer (E-C Apparatus Corporation, St. Petersburg, FL). Each band was scanned at 3 different, positions and the areas of the peaks were measured with an Apple graphic tablet (Apple Computer, Inc., Cupertino, CA). Peak areas were normalized for variations in the measured width of each band. For each set of samples, 5 gels were prepared and the average values of the peak areas were used for calculations. A series of calibration samples was prepared with different amounts of enzyme-MPL complex diluted with appropriate amounts of MPL (to keep the amount of MPL constant and vary the amount of enzyme) which were then electrophoresed and analyzed in cornparable manner. It was found that the absorbance of the enzyme hand was linear for between 45 and 450 ng of D-8-hydroxybutyrate dehydrogenase loaded onto the gel (cf. Fig. 3). For some experiments, 46 ng of cytochrome c (Sigma) per sample was added to the sample buffer to control for any variability in the loading of the gels.

RESULTS
Effect of Radiation on Enzymic Activity-Inactivation of D-P-hydroxybutyrate dehydrogenase, in both submitochondrial vesicles and the purified enzyme reconstituted into MPL vesicles, occurred as a simple exponential function of radiation dose (Fig. 1). The lines drawn in Fig. 1 were determined by least-squares fit of the weighted data (cfi "Experimental Procedures") to the theoretical exponential decay of activity with dose. From the slope of these lines, D37 values were determined from which target sizes were calculated (Fig. 1). From several such studies, average values for molecular sizes of 106,000 k 8,000 and 119,000 k 8,000 were determined for D-P-hydroxybutyrate dehydrogenase activity in submitochondrial vesicles and enzyme-MPL complex, respectively (Table I). These values are the same within experimental error and were found not to be significantly different using Student's t test ( p < 0.05). T h e oligomeric structure of purified and reactivated D-P-hydroxybutyrate dehydrogenase therefore appears to be the same as in the membrane of submitochondrial vesicles.
Loss in enzymic activity by irradiation might have resulted from the destruction of the phospholipid, which is essential for function, or from the generation of potent inhibitors ofthe enzyme. This does not appear to be the case. The lipid of t h e irradiated enzyme-MPL complexes served to activate freshly added D-P-hydroxybutyrate apodehydrogenase similarly to the nonirradiated sample (Fig. 2). Two parameters of the activation can be evaluated by titrating the apodehydrogenase The enzymic activities of the initial samples before freezing were 130 and 0.84 pmol of NAD' reduced/min/mg of protein at 37 "C for BDH-MPL and SMV, respectively. "Sham" irradiated controls (electron beam turned of0 retained full activity. The slope of each inactivation profiie was determined by least-squares fit of the data. The molecular weight (M,) was calculated using the established relationship M , = 6.4 X 10"/D.,j (Kepner and Macey, 1968) where D3: is the amount of radiation in rads required to reduce activity to 37% of the original. The values obtained from these data were 97,000 & 8,000 (D:,7 = 6.60 Mrad) and 114,000 8,000 ( D , 3~ = 5.63 Mrad) daltons for the enzyme in SMV and purified BDH-MPL, respectively. Theoretical inactivation lines (solid) are shown for either monomer ( a ) , dimer ( b ) , or tetramer (c) of BDH with target sizes of 30, 60, or 120 kiodaltons, respectively. Experimental data from a number of experiments are averaged in Table I. with phospholipid vesicles or membranes (Gazzotti et al., 1975;McIntyre et al., 1979). Maximal specific activity, obtained in excess phospholipid, was the same with the irradiated and nonirradiated enzyme-MPL samples as with MPL alone. Efficiency of activation, the amount of lipid required to obtain half-maximal activation, is an index of the accessible lipid in the membrane (McIntyre et al., 1979). For this parameter, a higher number denotes poorer efficiency. The values obtained for activation of newly titrated apodehydrogenase by control and two irradiated samples were the same (23 mol of phospholipid/mol of added apodehydrogenase). This was somewhat less efficient than that obtained with MPL alone (19 mol of phospholipid/mol of added apodehydrogenase, Fig.   2). This lower efficiency of activation by the enzyme-MPL complexes versus MPL merely reflects that the enzyme, in the liposome, diminishes in part the amount of accessible phospholipid. If t h e loss in activity of the enzyme-MPL complexes after irradiation treatment were due to either the generation of inhibitors or phospholipid destruction, a decrease in the maximal specific activity and/or efficiency of activation would be obtained with freshly added apodehydrogenase. Neither of these effects is observed, suggesting that a process other than lipid damage or generation of inhibitors, results in loss in enzymic activity. Further, the loss of activity could be directly correlated with destruction of the enzyme.
Correlation of Loss of Activity with Polypeptide Fragmentation-The amount of enzyme in the purified enzyme-phospholipid complex, after exposure to various doses of radiation up to 8.8 megarads, was measured by densitometry of PAGE of D-@-Hydroxybutyrate Dehydrogenase pg of P/mg of BDH) which had previously been exposed to different doses of irradiation. They were: no irradiation control (A), 3.3 megarads radiation (0, residual activity 57% of control), or 8.8 megarads radiation (0, residual activity 22% of control). BDH (3.2 pg) was added to increasing amounts of either MPL or previously irradiated BDH-MPL complexes, as indicated, in a total volume of 200 pl of 0.25 M sucrose, 5 mM Hepes (pH 7.3), 1 mM EDTA, 10 mM dithiothreitol, and 5 mM NAD'. After incubation for 2 h at 25 "C, to ensure optimal activation, enzymic activity was measured at 37 "C (cf. "Experimental Procedures"). Control samples of BDH-MPL, to which fresh BDH was not added, were likewise assayed to determine, bv difference, the activity of the freshly inserted BDH. The activity of the original BDH (in BDH-MPL complex) contributed up to a maximum of 40% of the total BDH activity at the highest amount of complex added (120 pg of P/mg of newly added BDH for activation of BDH by the nonirradiated BDH-MPL complex) and lower values (28 and 148) for the two previously irradiated BDH-MPL samples in sodium dodecyl sulfate using silver staining (Fig. 3). There was a progressive decrease in silver stain intensity of the enzyme band with increasing radiation dose (Fig. 3, A and B ) . The gels with decreased enzyme band intensity generally showed some lower molecular weight material as a broadened smear, which stained only faintly and variably at this level of loading. The fragments could be observed routinely at higher protein loading either by the silver stain or with Coomassie brilliant blue (not shown). Irradiation did not give rise to higher molecular weight material in either the stacking or running gel or at the interface between the two. Therefore, we conclude that irradiation of D-/3-hydroxybutyrate dehydrogenase results in polypeptide fragmentation with production of a distribution of lower molecular weight peptides. Calibration of the enzyme band intensity versus the amount of applied enzyme is linear in the concentration range from 45 to 450 ng (Fig. 3 0 . The plot of the logarithm of the integrated peak area of the PAGE band as a function of radiation dose for the enzyme-MPL complex gave a straight line (Fig. 4). Target size analysis of radiation-induced fragmentation of the polypeptide, from two separate experiments, yields a structural size of about 104,000 daltons, the same within experimental error as that obtained by enzymic activity measurements of these same samples ( Fig. 4 and Table I). Thus, irradiation leads to concomitant loss of enzymic activity and fragmentation of the protein.
newly inserted BDH (120 * 5 pmol of min/mg of NAD' reduced/ min/mg of BDH) was obtained with each of the samples. The efticiency of activation (amount of lipid which gives half-maximal activation, see text) was the same (23 pg of P/mg of BDH) for the three BDH-MPL complexes which had received different dosages of irradiation. MPL, to which BDH had not previously been added, had somewhat better efficiency of activation (19 pg of P/mg of BDH). The reactivation characteristics of freshly introduced BDH by the lipid in BDH-MPL complexes was, within experimental error, independent of the irradiation exposure.  doses of 0, 1.1, 1.65,2.2,2.75,3.3,4.4,5.5, 7.15, and 8.8 Mrads, for tracks I to IO, respectively. Gels were loaded with 450 ng of BDH as the BDH-MPL complex, electrophoresed (1.5 milliwatts/gel) and then silver stained. For details see "Experimental Procedures." B, representative scans of the gel (top of gel to the left) used to quantitate the band intensity for various doses of radiation. The scans shown labeled A to E, are for the BDH bands of samples in tracks 1, 3, 6, 8, and 10, respectively, of the gel shown in A. A few scans (e.g. that shown in D) exhibited baseline shifts which arise from variable staining intensity of lower molecular weight breakdown products visible in some samples. In such cases, the peak area was estimated with baseline correction as indicated by the dashed line in d. Each gel band was scanned at three different positions and the averaged peak areas were each normalized for slight variations in the measured width of each band. The integrated peak areas as a function of radiation dose are shown in Fig. 4. C, calibration of silver staining density of BDH (integrated peak areas) as a function of the amount of BDH, 45 to 450 ng, loaded onto the gel. Samples were prepared as described under "Experimental Procedures" then electrophoresed and band intensity quantitated as described in A and B, above.   Table I. BDH was also inserted into vesicles of a nonactivating phospholipid mixture of dioleoylphosphatidyldimethylethanolamine/dioleoylphosphatidylethanolamine/l-palmitoyl-2-oleoylphosphatidylpropan-1,3-diol (5:4:1 ratio by phosphorus). In these lipid vesicles, the phospholipid to enzyme ratio was 180 p g of P/mg of BDH. The molecular size (Mr = 110,000 & 7,000) of the nonactivated enzyme was determined from the decrease in PAGE band intensity of BDH (A---A) as a function of irradiation dose (average of five separate gel analyses of one set of irradiated samples).

of ~-P-Hydroxybutyrute Dehydrogenase
The size of the inactive enzyme, inserted into phospholipid vesicles devoid of lecithin was also measured. The decrease in intensity of the enzyme band on polyacrylamide gels with increasing radiation dose was similar to that of the active enzyme-MPL complex (Fig. 4). Target size analysis yielded a molecular size of approximately 110,000 daltons. Thus, the nonactive enzyme in phospholipid has a size comparable to that of the active enzyme-MPL complex and as well as the enzyme in the membrane of submitochondrial vesicles (cf- Table I).

D-P-Hydroxybutyrate dehydrogenase is one
of the most extensively studied lipid-requiring enzymes. This is the fist study in which the radiation-inactivation technique has been used to compare the target size of a purified and reconstituted membrane enzyme with that of the enzyme in the native membrane. Within experimental error, the size of D-P-hydroxybutyrate dehydrogenase, either native or reconstituted, is the same and approximates a tetramer. Inactivation analysis has also been used to evaluate the influence of lecithin on the oligomeric size of this lipid-requiring enzyme. We find the same oligomeric size for the enzyme in activating mixtures of phospholipid containing lecithin as with a nonactivating phospholipid mixture. Therefore, the lecithin does not activate the enzyme by way of regulating oligomeric size. We find that inactivation of enzymic activity upon irradiation is not attended by destruction of the phospholipid which is required for enzymic function, but arises from fragmentation of the ' ' All samples, prepared in 0.25 M sucrose, 5 mM Hepes (pH 7.3), 1 mM EDTA, and 5 mM dithiothreitol, were frozen over liquid nitrogen and irradiated at -40 to -50°C. Submitochondrial vesicles were irradiated at 2 mg/ml (cf. Fig. 1). BDH-MPL is the active complex of purified D-P-hydroxybutyrate dehydrogenase (BDH) reactivated with mitochondrial phospholipids (MPL) to obtain optimally activated enzyme. The samples were irradiated at 40 pg of BDH/ml with either 60 or 180 pg of P/mg of BDH (60 or 180 mol of phospholipid/mol of BDH) (cf. Fig. 1 and "Experimental Procedures"). Results obtained at the two lipid/protein ratios were the same within experimental error and are averaged. BDH-DiMePE is inactive complex of the enzyme inserted into a nonactivating ternary mixture of phospholipids similar to MPL but containing phosphatidyldimethylethanolamine (DiMePE) instead of lecithin (cf. Fig. 4); BDH to lipid ratio was 180 pg of P/mg of BDH (180 mol of phospholipidimol of BDH).
The effect of irradiation on BDH was analyzed either by loss in enzymic activity (cf. Fig. 1) or by decrease in intensity of the BDH band separated by PAGE in sodium dodecyl sulfate (SDS) and visualized by silver staining (cf. Figs. 3 and 4).
'' The data were analyzed with a single target, single hit model of radiation inactivation (Pollard, 1953). Apparent molecular weights were calculated using the formula of Kepner and Macey (19681, i.e. M , = 6.4 x 10"/D:3j (rads), where D:1: is the dose of absorbed radiation required to reduce the enzymic activity to 37% of the original. The errors given are the standard deviation of separate experiments except the last value which is the standard deviation of the slope of the linear least-square fit of the data (Fig. 4). Glucose-6-phosphate dehydrogenase from Leuconostoc mesenteroides was incorporated in some samples as a standard and was found to have a target size of 90,000 5,000 daltons in the presence or absence of membranes. This size is 87% of the size (103,700 daltons) of this enzyme measured by sedimentation equilibrium at 20 "C (Olive and Levy, 1971).
polypeptide. Further freeze-fracture electron microscopy studies indicate that loss of D-P-hydroxybutyrate dehydrogenase activity by irradiation of submitochondrial vesicles is not attended by gross morphological changes in membrane structure.s T h e size of D-P-hydroxybutyrate dehydrogenase obtained by classical target theory analysis of radiation inactivation data is approximately 110,000 daltons both in submitochondrial vesicles and for the purified enzyme reactivated by insertion into vesicles of MPL. Therefore, the oligomeric structure of D-P-hydroxybutyrate dehydrogenase does not appear to be altered following isolation, purification, and reactivation with MPL. This is consistent with the unaltered kinetic parameters of this preparation, reactivated with MPL, compared with the native enzyme (Nielsen et al., 1973). T h e subunit size by PAGE in sodium dodecyl sulfate is 31,000 daltons ), but appears slightly smaller (28,500 daltons) by equilibrium sedimentation under denaturing conditions.' The radiation inactivation data analyzed by the single target, single hit model (Pollard, 1953) obtained in this study are consistent with a tetramer.
BDH is an amphipathic molecule and as such is highly associating in the absence of lipid. It is only by means of using active enzyme sedimentation that the hydrodynamic characteristics of the active BDH-soluble lecithin complex could be ~ -' A. Maurer and S. Fleischer, unpublished studies. studied at low concentration of enzyme (30 pg of protein/ml or lower). The phospholipid-free apodehydrogenase in aqueous buffer was studied by conventional sedimentation analysis at higher protein concentrations (60-1200 p g of protein/ml) and was found to undergo concentration-dependent self-association. Under these conditions, the enzyme associates to form a heterogeneous population of components, some much larger than a tetramer (McIntyre et al., 1979b). Therefore, the presence of phospholipid serves to limit self-association to a tetramer. Active enzyme sedimentation studies of the enzyme, activated with soluble lecithins, show a dimeric functional unit (McIntyre et al., 1978b). The molecular size of the enzyme, activated with soluble lecithin, could not be confirmed by the radiation-inactivation method due to the The apparent molecular size of purified enzyme-MPL complex was not significantly different at two different lipid/ protein ratios used in this study. The MPL vesicles, prepared as described here, have an average diameter of 650 8, by quasielastic light scattering measurements6 so that there are approximately 530 or 175 enzyme monomers/phospholipid vesicle for 60 to 180 mol of phospholipid/mol of enzyme monomer, respectively. Thus, the apparent molecular size of the D-P-hydroxybutyrate dehydrogenase in MPL vesicles is not influenced by the number of polypeptides per phospholipid vesicle within the range of this study.
The loss of enzymic activity upon irradiation of the purified D-P-hydroxybutyrate dehydrogenase-MPL complex appears to result from radiation-induced fragmentation of the 31,000dalton band  observed by PAGE in sodium dodecyl sulfate.
Fragmentation of the polypeptide exhibited the same radiation dose dependency as the loss of enzymic activity, within experimental error, with a target size equivalent to a tetramer.
Since this enzyme has a specific requirement for lecithin for enzymic activity, it was important to evaluate whether destruction of the phospholipid contributes to the observed loss of activity. Target theory predicts that destruction of the lipid in the samples by direct hit would not be significant at the dose levels used. However, the phospholipid which surrounds the protein could constitute a portion of the observed target. Further, destruction of a small number of phospholipid molecules might have generated inhibitors of the enzyme. We evaluated this possibility by testing the competence of the phospholipid present in the irradiated samples to activate freshly inserted enzyme. The activation characteristics of D-P-hydroxybutyrate dehydrogenase by enzyme-MPL complexes were the same for nonirradiated and maximally irradiated samples in which the original enzyme activity had been reduced to less than 25% of the control. Activation of D-phydroxybutyrate dehydrogenase by phospholipid present in membrane preparations is a sensitive test for the phospholipid integrity since the activation characteristics are critically dependent on the phospholipid structure and composition  (Isaacson et al., 1979). Thus, the bulk of the phospholipid after irradiation can activate D-/3-hydroxybutyrate dehydrogenase suggesting that phospholipid destruction does not occur and there is no inactivation by this mechanism. It should be noted that MPL alone, activates the freshly added enzyme more efficiently than the previously irradiated enzyme-phospholipid complexes even after greater than 75% of the original enzyme present had been fragmented and inactivated. Thus, the phospholipid that is associated with the enzyme in the complex ( i e . not accessible to freshly added enzyme) does not become accessible after radiation-inactivation and fragmentation of the enzyme. This would imply that either the inaccessible phospholipid remains inaccessible when the polypeptide is degraded or that some phospholipid constitutes a portion of the target. However, this inaccessible phospholipid in reconstituted enzyme-MPL complex constitutes only a small portion of the total phospholipid; the bulk of the phospholipid in the irradiated enzyme-MPL complexes, behaves similarly to that in nonirradiated complex for the activation of D-/3-hydroxybutyrate dehydrogenase. D-P-hydroxybutyrate dehydrogenase, inserted into a nonactivating mixture of phospholipid vesicles, exhibits a tetrameric structure, the same as that of the enzyme in an activating phospholipid mixture. Hence, the unique role of lecithin in activating the enzyme is not related to modulating the oligomeric size. The difference in oligomeric size of the enzyme reactivated with soluble lecithins (see above) as compared with the enzyme inserted into phospholipid vesicles is an important one. It points out that enzymic activity is more critically dependent on the presence of lecithin rather than whether the enzyme exists as a dimer or tetramer.
Tne temperature of irradiation influences the apparent target size of some enzymes Schlegel et al., 1979). In studies of the (H' + K')-ATPase, a 20% higher molecular size was obtained for the enzyme as a lyophilized fiim at 20 "C as compared with the apparent target size in frozen suspensions a t -50 "C (Saccomani et al., 1981).
The temperature for radiation used in the study reported here (-50 "C) and that reported by Saccomani et al., (1981) in the frozen state were the same. In our studies, glucose-6-phosphate dehydrogenase, has been used as an internal standard, to correct for temperature and other variables. The target size for the internal standard, irradiated under conditions similar to those used for D-P-hydroxybutyrate dehydrogenase, was 90,000 k 5,000 (cf. Table I) or 13% lower than the size as measured by sedimentation equilibrium at 20 "C (Olive and Levy, 1971). This correction factor is, within experimental error, the same as the temperature correction factor obtained by Saccomani et al. (1981). Thus, the -110,000-dalton target size that we obtain for D-P-hydroxybutyrate dehydrogenase appears to be underestimated by 10-20% due to the conditions for the low temperature of irradiation. When a temperature correction factor of 13% is applied to our results with D-phydroxybutyrate dehydrogenase, a target size of -124,000 daltons is obtained, which is close to a tetramer.
The usefulness of target size analysis lies in the ability to determine the functional molecular weight of a protein regardless of the state of purification as long as the size is unaltered by sample preparation. In the study reported here, we have obtained the size of D-P-hydroxybutyrate dehydrogenase both in the membrane and as the purified enzymephospholipid complex. The functional sizes of a number of transport proteins have previously been reported. The (Na' + K')-ATPases of human erythrocyte ghosts, guinea pig kidney microsomes, and plasma membrane preparations have been shown to have sizes of 250,000 daltons (Kepner and Macey, 1968). The calcium pimp protein of sarcoplasmic reticulum vesicles prepared from rabbit muscle appears to be a dimer (Vegh et al., 1968).' The functional size of the human erythrocyte D-glUCOSe carrier has been determined to be -200,000 daltons based both on the inactivation of cytochalasin B binding (Jung et al., 1980) and the inactivation of Dglucose flux . The target size of the purified gastric (H' + K+)-ATPase has been found, by both activity and PAGE analyses, to be 270,000 daltons suggesting that this transport protein is a trimeric assembly of the 100,000-dalton polypeptide (Saccomani et al., 1981). Schlegel et al. (1979) found that activators alter the target size of adenylate cyclase, thus providing insight into structural changes within the membrane that accompany activation. Radiation inactivation analysis studies are not always clearcut. It has been reported that the molecular weight of acetylcholine esterase in lyophilized samples is dependent on the choice of buffer salts (Parkinson and Callingham, 1982). This dependency might be related to the influence of salt on protein aggregation. Recent studies of cytochrome oxidase have suggested that the cytochrome c oxidase activity may reside in a 70,000-dalton unit which is smaller than the minimum size (125,000 daltons) required for two hemes, postulated to be essential for this function (Thompson et al., 1982). Similarly, in brush-border membranes, the target size molecular weight of alkaline phosphatase, sucrast and leucine aminopeptidase, each determined by inactivation of enzymic activities, was less than the molecular weight of the polypeptides separated by PAGE in sodium dodecyl sulfate (Stevens et al., 1982). Such results could arise from uncertainty in the absolute values of the radiation incident within the experimental sample (Lo et al., 1982) or from errors in determination of polypeptide molecular weights by PAGE in sodium dodecyl sulfate or from incomplete inactivation of the polypeptide by a single hit. The latter result could be explained by retention of enzymic function of a large polypeptide that has been hit at a point remote from the catalytic site so that the integrity and function of the catalytic domain of the protein is maintained.