CHARACTERIZATION OF THE HEXOKINASE MEMBRANE _ ' ' INTERACTION: RAPID PHOTOLYSIS 0F DANSYL TTROSINE: ‘ PARTIAL CHARACTERIZATION OF BINDABLE ANO NON - : BINDABLE HEXOKINASE ISOZYMES ‘ Thesis for the. Degree Of M. S. MICHIGAN STATE UNIVERSITY PHILIP L. FELGNER 1975 d\' ABSTRACT CHARACTERIZATION OF THE HEXOKINASE MEMBRANE INTERACTION: RAPID PHOTOLYSIS OF DANSYL TYROSINE; PARTIAL CHARAC- TERIZATION 0F BINDABLE AND NON-BINDABLE HEXOKINASE ISOZYMES By Philip L. FeTgner Two of the three dansyT chloride derivatives of tyrosine have been identified and characterized by three independent methods. These two products are the mono- and di-dansyTated forms of tyro- sine. The mono-dansyIated species has the dansyT group attached to the primary amine and the di-dansyIated product is TabeTed at the phenoTic oxygen and at the primary amine. N-ethyI morphoIine, the standard solvent for protein dansyTation as suggested by Gray (Gray, N. R., Methods in EnzymoTogy, 25, 121 [1972]) was found to give an unexpectedTy high yier of mono-dansyIated tyrosine; therefore sodium bicarbonate was used as buffer since this was found to favor compTete conversion to the di-dansyI product. Under these circumstances, the N-terminaT amino acid of rat brain hexo- kinase (ATP:D-hexose-G-phosphotransferase, E.C.2.7.T.I) was found to be tyrosine, not gTycine, as previousTy reported (Chou, A. C., WiTson, J. E., Arch. Biochem. Biophys., 151, 48 [1972]). Chymo- trypsin treatment of this enzyme exposed two new N-terminaTs, phenyTaTanine and Tysine, completer removing tyrosine. And, finaTTy, it has been shown that di-dansyT tyrosine has an unusuaTTy Philip L. Felgner high photolysis rate under U.V. Tight, a property which can easily lead to complications in interpretation of experimental results unless measures are taken to restrict photolytic degradation. Type I rat brain hexokinase purified by the method of Chou and Wilson (Chou, A. C., Wilson, J. E., Arch. Biochem. Biophys., lgl, 48 [1972]) has been shown to consist of at least two isozymes. One of these isozymes (type 1b) binds to mitochondrial membranes and the other (type In) does not. Recently discovered peculiarities in the binding assay have obscured this discovery for some time. (i) Hexokinase binds to the polypropylene microcentrifuge tubes routinely used for the binding assay. (ii) There is an unexpected inhibition of binding by relatively low concentrations of salt. (iii) The routinely used rat liver mitochondria bind less hexokinase and are more susceptible to the salt effect than rat brain mitochondria. A peptide(s) can be removed from pure hexokinase, at least part of which comes from the N-terminal end and may be involved in the binding process. This peptide(s) awaits further characterization. The phosphate-induced reversal of solubilization first demon- strated by Rose and Narms (Rose, I. A., and Harms, J. V. B., J. Biol. Chem., 242, l635 [l967]) can also be done with potassium chloride. This salt induced rebinding suggests a hypothesis for the binding mechanism. Further experiments will be required to ascertain whether or not the salt-induced rebinding is due to identical factors necessary for phosphate-induced rebinding. CHARACTERIZATION OF THE HEXOKINASE MEMBRANE INTERACTION: RAPID PHOTOLYSIS OF DANSYL TYROSINE; PARTIAL CHARAC- TERIZATION OF BINDABLE AND NON-BINDABLE HEXOKINASE ISOZYMES By Philip L. Felgner A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Biochemistry 1975 TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES ABBREVIATIONS BACKGROUND AND PURPOSE OF THE PROJECT Background . Purpose . Chapter I. RAPID PHOTOLYSIS 0F DANSYL TYROSINE Introduction . Materials and Methods Chemicals Dansylation . . Photolysis of Di -Dansyl Tyrosine on Thin Layer Plates . . . . . . Results . Characterization of Two Dansylated Forms of Tyrosine . . Effect of N- -Ethyl Morpholine as Buffer in the Dansylation Reaction . . Photolysis of Tyrosine on Thin Layer Plates Hexokinase N- Terminal Determination . Discussion . Swmmy. II. PARTIAL CHARACTERIZATION OF BINDABLE AND NON-BINDABLE HEXOKINASE ISOZYMES Materials and Methods Chemicals Hexokinase Assay . Preparation of Mitochondria . . Glucose- 6- -Phosphate Solubilized Enzyme . Purification of Rat Brain Hexokinase ii Page iv vii REFERENCES 1% Sodium Dodecyl Sulfate- Polyacrylamide Disc Gel Electrophoresis . . Mitochondrial Binding Assay . Sephadex Column Chromatography of Chymotrypsin Treated Pure Hexokinase . Results . Binding of Hexokinase to Polypropylene Centrifuge Tubes . The Effect of Salt on MgClz- -Induced Rebinding. of Hexokinase to Mitochondria . - Binding Assay With Pure Hexokinase Hexokinase Isozymes . Chymotrypsin Treatment of the Pure Type Ib Hexokinase . Potassium Chloride- Induced Rebinding of. Hexokinase to Mitochondria . Discussion . Hexokinase Binding to Polypropylene . . . Inhibition of Rebinding by Ionic Strength . Differences Between Liver and Brain Mitochondria . . Purified Bindable and Non- Bindable Isozymes Pure Hexokinase and the "Binding Peptide" Salt-Induced Reversal of Solubilization and the Putative Binding Mechanism . . Summary . Page Table LIST OF TABLES The relative amounts of the three dansyl tyrosine derivatives produced at various pHs . The determination of the ratio of dansyl groups to tyrosine on compounds I and II . . . . . The relative amounts of the three dansyl tyrosine derivatives at various concentrations of N-ethyl morpholine . Comparison of Na-bicarbonate and N-ethyl morpholine as buffers for N-terminal determination of hexokinase Assay of the activity that remains bound to polypropylene tubes . iv Page l3 l4 15 15 29 LIST OF FIGURES Figure 1. Relative activity of soluble and bound forms of brain hexokinase 2. Difference between the catalytic activities of soluble and bound forms of brain hexokinase 3. Thin layer chromatography of dansylated tyrosines and dansyl glycine, before and after photolysis 4. Time course for the photolysis of di-dansyl tyrosine . . . . . . . 5. Increased binding to polypropylene with increasing purity of hexokinase . . . . . . 6. Assaying successive aliquots from a polypropylene tube . . . . . . . . 7. Hexokinase binding to successive polypropylene tubes . . . . . . . 8. The Inhibition of rebinding by increasing ionic ' strength . . . . . . . 9. Salt-induced inhibition of binding; comparison between liver and brain mitochondria . lO. Bindability of pure hexokinase ll. DEAE-cellulose column chromatograph of rat brain hexokinase . . . . . . . . 12. DEAE-cellulose column chromatography of rat brain hexokinase; elution with a shallower gradient . l3. l% SDS-polyacrylamide gel electrophoresis of purified rat brain hexokinase . . . l4. Inhibition of binding by chymotrypsin treatment of crude and pure hexokinases . Page ll l7 27 28 30 32 33 34 35 36 38 4O Figure Page 15. Sephadex G- 50 column chromatography of chymotrypsin. treated hexokinase . . . . 42 T6. Reversibility of solubilization by salt . . . . . 45 vi ATP BSA Dansyl-Cl DEAE-cellulose DTNB EDTA G-6-P HEPES HK NADP NADPH SDS Tris ABBREVIATIONS Adenosine triphosphate Bovine serum albumin l-dimethylaminonaphthalene-S-sulfonyl chloride Diethylaminoethyl-cellulose 5,5l-dithiobis-(2-nitrobenzoic acid) 'Ethylenediamine tetraacetate Glucose-6-phosphate N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid Hexokinase Nicotinamide adenine dinucleotide phOSphate Reduced nicotinamide adenine dinucleotide phosphate Sodium dodecyl sulfate Tris-(hydroxymethyl) aminomethane vii BACKGROUND AND PURPOSE OF THE PROJECT Background In l953 Crane and 8015 (1) observed that a major portion of the hexokinase activity in mammalian brain was associated with a particulate fraction. Johnson's report that the particulate enzyme was bound to mitochondria (2) was subsequently confirmed by other workers (3-8). It was later reported by Kropp and Wilson (9) for liver and by Craven and Basford (6) for brain that the enzyme was located on the outer-mitochondrial membrane. Wilson and others found that in the presence of glucose-6- phosphate (1 mM), ATP (l0 mM), or high ionic strength (200 mM) the enzyme could be solubilized (l, 8, 10). Neither ADP, other 6-phosphosugars, nor low salt could solubilize the enzyme. Because of this specificity and because glucose-6-phosphate and ATP are end products of hexokinase and glycolysis, respectively, it was suggested the ATP/G6P-induced soluble/particulate equilibrium played a regulatory role in_viyg_(8, 10). In support of this view, it was determined that the bound enzyme was less susceptible to inhibition by glucose-6-phosphate and had a greater affinity for ATP (ll). Calculations based on the Michaelis-Menten equation (see Figure l) indicated that a shift from soluble to bound states under I'physi- ological" substrate concentrations (l2, 13) would result in a 3-4 fold increase in the hexokinase catalytic rate. This compared Figure l.--Relative activity of soluble and bound forms of brain hexokinase. The ordinate gives the velocity of the reaction catalyzed by the bound (mitochondrial) form relative to the velocity for the soluble form of hexokinase as a function of concentration of the substrate ATP (indicated on the abscissa) and of the inhibi- tory product, glucose-6-phosphate. Values for Km (ATP) and Ki (G6P) were obtained from Tuttle and Wilson (ll). v bound v soluble I; Soluble: Km .-. 0.25 mM ‘l I'— K;(G-6-P) = 0.007mM Km: OJ mM Ki(G-6-P) = 0.035mM I Ia ATP (mM) favorably with the estimated 3-7 fold increase in the glycolytic flux during ischemia (12) or convulsions (14). It was also calcu- lated that it is precisely under "physiological" substrate concen- trations where the difference between the velocities of the two forms is greatest (see Figure 2). Furthermore, Knull et al. (15, 16) have shown that disturbances of energy metabolism in chick brain (caused by-ischemia, insulin shock, or high blood galactose levels) result in altered glucose-6-phosphate levels and soluble/particulate hexokinase distribution in accord with the hypothesis (10). All these data taken together provide considerable support for the pro- posed regulatory role of the ATP/G6P-induced soluble/particulate equilibrium. Purpose The objective of this project has been to describe in greater detail the nature of the interaction between rat brain hexokinase and mitochondria. While progressing toward this goal, two new hexokinase isozymes were isolated in pure form; type Ib binds to mitochondria and type In does not. This topic is covered in Chapter II. Chapter I is concerned with technical difficulties in N-terminal determination by the dansyl chloride method. .a»w>wpum mmm:_xoxm: mo Focucou com Emwcmgums m>vgwmcmm m mcwuw>oca nova Ianwxwmwu mumpauerma\m_a:_om cw cowgmwcm> gum: ucmpmwmcou an upzoz sows: .Amm_ .N_H zpm>wwomnmmc .25 op.o use :5 m.mv mace; pmuwmopownga may cw mcowumgwcmocou mumzamocaioimmoua_m new ah< um caouo op :mmm mew mace; cczon use mpnz—om wgp :mwzpmn mmucmcwmmwu Pmswxmz .mm>c:u mg“ co umumu Iwu:_ .mcowpmgucmucou wpmgamosaioimmouspm pcmcmwwwu mugs“ so» czmcu mew mm>cau mg» ucm .mmmwumam mg» :0 cm>wm mw cowumgpcmocou ah< .Aogm~ ma upzoz >< as» sues; any mh< we mpm>mp mcwumcaumm um um>smmno up upzoz gown: .Ah>v zuwuopm> mpamcwmppm Esewxme msu mo magma cw ummmmcnxm my mucwcmmwwu mesh .m> I m> u >< ..m.w .mmmcwxoxm; cwmcn we macaw Amv mpnzpom new Amv ucaon ecu xa vaAFmpmu m:o_pummg map mo mm_pwuo_m> mg» :mwzumn mucmcmmwwu may mm>wm mpmcwuco mmh .mmmcmxoxw; :wmcn mo maxed uczon use m~22P0m mo mwwpw>wuum umu»_m¢mu mzu :mmzpmn mucmcmemwoii.m mczmwm CHAPTER I RAPID PHOTOLYSIS OF DANSYL TYROSINE Introduction The problems covered in this chapter concern some peripheral discoveries regarding technical difficulties in N-terminal determina- tion by the dansyl chloride method. This problem was initiated to determine from which end of hexokinase a peptide was clipped by the action of chymotrypsin. (See Figure 8, Chapter II, for the effect of chymotrypsin on bindable hexokinase.) Since published data on the identity of the N-terminal of pure hexokinase was being contra- dicted (25), an exhaustive study was necessary. Materials and Methods Chemicals Dansyl chloride, L-tyrosine and standard dansyl amino acids were obtained from Sigma Chemical Company. Sodium dodecyl sulfate (sequenation grade), N-ethyl morpholine (sequanal grade), and dimethylformamide (silylation grade) were purchased from Pierce Chemical Company (Rockford, Illinois). U-[14CJ-L-tyrosine and G-[3H]dansyl chloride were products of Amersham/Searle Corporation. l-nitroso-Z—naphthol came from Mallinckrodt Chemical (St. Louis, Missouri) and Cheng-Ching polyamide layer sheets were obtained from Ballard-Schlesinger Chemical Mfg. Corp. (Carle Place, New York). All chemicals were reagent grade if not otherwise specified. 7 Dansylation Free tyrosine was dansylated essentially according to the method of Gros and Labouesse (17) with some minor modifications. 10 ul of l mM_tyrosine (raised to pH 10 to promote solubilization) was added to 20 ul of 100 mM_sodium bicarbonate, pH 9.5. The reac— tion was started by adding 30 ul of 10 mM_dansyl chloride in ace- tone. After one hour at room temperature the reaction was stopped with 50 ul 0.1 N NaOH and the solution acidified with 20 ul 6 N HCl. After removal of acetone under a stream of nitrogen, the samples were completely dried on a rotary evaporator at 35°C, suspended in 10 pl 40% acetic acid-acetone and then chromatographed on 5 x 5 cm polyamide layer plates according to Weiner et al. (18). In some experiments, 100 mM_sodium borate or 100 mM_sodium phosphate at various pHs replaced the bicarbonate. The method outlined by Gray (19) for dansylation of proteins was routinely followed with three modifications: N-ethyl morpholine was replaced by 500 mM_sodium bicarbonate, pH 9.5; protein samples were not performic acid oxidized, the dansylated protein sample was washed more extensively. The procedure was as follows: 50-250 pg of lyophilized protein was placed in a 6 x 50 mm culture tube and dissolved in 50 ul of 1% SDS by heating in a boiling water bath for about five minutes. If the protein sample contained buffer or salts it was dialysed overnight against 2 liters of water before lyophilization. After cooling, 50 ul of 500 mM_sodium bicarbonate, pH 9.5, followed by 75 ul of a 25 mg/ml dansyl chloride solution in dimethyl formamide were added. The test tube was vortexed, covered with parafilm and incubated at room temperature for at least one hour. Since the dansyl chloride was not entirely soluble in this aqueous solution the reaction mixture had a cloudy appearance that cleared as the reaction went to completion, but this did not affect the reaction. One volume of 20% TCA (w/v) was added to stop the reac- tion and precipitate the protein. The precipitated protein was washed once with 0.2 ml 1N HCl and once with 80% acetone:water.' To the dried precipitate was added 50 ul constant boiling HCl, and the culture tube was sealed in_vagug_and placed in a boiling tolu- ene bath at 110°C for 18 hours. The dried hydrolysate was dis- solved in 10 ul 2:3, acetic acidzacetone, spotted on 5 x 5 cm polyamide layer sheets and chromatographed according to Weiner et al. (4). Photolysis of Di-Dansyl Tyrosine on Thin Layer Plates Radioactive di-dansyl tyrosine was prepared from commercial tyrosine and radiOactive dansyl chloride as described above. The dansylated tyrosine was chromatographed in two dimensions and the well resolved di-dansyl tyrosine spot was irradiated with an ultra- violet light source (100 watt mercury spot lamp from General Electric with a long wave length filter) for various time intervals. The plates were placed exactly 20 cm from the lamp, and centered in the most intense portion of the beam. At this distance from the source, the temperature was 35°C which was 16°C above ambient temperature. 10 Results Characterization of Two Dansylated Forms of_1yrosine From Figure' 3, it is clear that commercially obtained di-dansyl tyrosine (top two chromatograms) can contain numerous chromatographically distinct fluorescent species. Further charac- terization of some of these was considered necessary because pub- lished data on their identity was either scarce or absent. Based on the literature, compound 11 (Figure 3) was di-dansyl tyrosine and compound I was N-dansyl tyrosine (20). To confirm this, three experiments were performed. (i) Since dansyl chloride reacts with unprotonated amines and with unprotonated phenols, the reactivity of either group depends both on the pH of the reaction mixture and on the pK of the reactive group (17, 20). Since the pK of the phenolic hydroxyl on tyrosine is 10.1 and the pK of the amino group is 9.1, the amino group should be more reactive at lower pHs than the phenolic hydroxyl. Table 1 indicates that as the pH of the reaction was decreased, relatively more compound I was formed than compound II. This was consistent with the proposal that compound II was di-dansyl tyrosine and compound I was N-dansyl tyrosine. (ii) l-nitroso-Z-naphthO1 in the presence of nitric acid has been shown to react Specifically with the phenolic group on tyrosine to give a red colored product (21). When they were sprayed with this reagent, compound I reacted to give a red color and com- pound II did not. This strongly indicated that the phenolic group on compound II was blocked and on compound I was not. 11 Figure 3.¥-Thin layer chromatography of dansylated tyrosines and dansyl glycine, before and after photolysis. Approximately 0.5 pl of commercial di-dansyl tyrosine (250 ug/ml in 40% acetic acid: acetone) or 0.5 ul of dansyl-glycine was spotted on a thin layer plate and chromatographed in three dimensions as described by Weiner et a1. (18). The top left hand chromatogram is didansyl- tyrosine (A) and the bottom left is dansyl glycine (C). The chromatograms on the right represent identical amounts of each amino acid derivative, photolysed for 20 min after the second dimension as described in the methods section and then chromato- graphed in the third solvent in order to resolve the amino acid derivative from its degradation products. Plate 8 is photolysed tyrosine; plate 0 is photolysed glycine. The drawing that accom- .panies the photograph identifies all the spots on plates A and B. Di-dansyl tyrosine (compound II) N-dansyl tyrosine (compound I) 0-dansyl tyrosine (not visible on these plates) Dansyl amine Dansyl sulfonic acid Photolysis products OSU'IDde The identities of O-dansyl tyrosine, dansyl amine and dansyl sul- fonic acid are well supported in the literature (18, 19, 20). 12 13 TABLE 1. --The relative amounts of the three dansy1_tyrosine deriva- tives produced at various pHs. Dansylation was carried 3out as described in the methods section except that l uCi of [ 3H]dansyl chloride was mixed with 0. 5 ml cold dansyl chloride before adding it to buffered tyrosine. Samples were chromatographed as in Figure l and the apprOpriate Spots scraped off, placed into scintillation solution (4 gm PPO; 0.05 gm POPOP; 1 liter toluene), mixed and counted in a Packard Tri-Carb liquid scintillation counter. (n.d. = not detected.) Buffer % di-dansyl % N-dansyl % 0-dansy1 tyrosine tyrosine tyrosine 33 mM_Na-phosphate pH 7.2 50 50 n.d. 33 mM_Na-phosphate pH 8.0 62 38 n.d. 33 mM Na-bicarbonate pH 8.3 70 30 n.d. 33 mM Na-bicarbonate pH 9.5 99 l n.d. 33 mM_Na—bicarbonate pH 10.5 99 1 n.d. 33 mM_Na-Borate pH 8.5 99 1 n.d. 33 mM_Na-Borate pH 9.5 99 l n.d. (iii) An experiment was performed to determine whether compound II in fact contained twice as many dansyl groups as compound I. [140]Tyrosine was dansylated with [3H]dansy1 chloride at a low pH so that both dansylated products were generated. Under these circumstances, the tritium to carbon-14 ratio should be twice as large for the di-dansyl tyrosine as for N-dansyl tyrosine. The data in Table 2 indicate that this was nearly the case. The data from all three experiments indicate that the identity of derivates I and II is as given in Figure 3. 14 TABLE 2.--The determination of the ratio of dansyl groups to tyrosine on compounds I and II. Dansylation was car§ied out as described in the methodS section except that 10 mM [ H]dansyl chloride (1 uCi/O. 2 m1) and 1 mM [ 4CJtyrosine (l uCi/O. 1 ml) were used. Samples were chromatographed as in Figure 3 and the appro- priate spots scraped off, placed into scintillation solution (3300 m1; 200 gm naphthalene; 20 gm PPO; 1.6 gm dimethyl POPOP; 2000 ml xylenes; 1,100 Triton X-ll4), mixed and counted in a Beckman LS 230 liquid scintillation counter. Appropriate cor- rections were made from standard curves (provided by Dr. A. J. Morris) to account for counting efficiency and carbon-l4 spill into the tritium window. Two additional experiments were performed with essentially similar results. DPM Compound 3H 3H 14C /Cl4 Di-dansyl tyrosine 2566 3569 0.72 Amino-dansyl tyrosine 1964 4672 0.42 Dansyl amine 315 12 41 Dansyl sulfonic acid 16032 390 26 Random scrapings 21 9 -- Effect of N-Ethyl Morpholine as Buf- fer in the Dansylation Reaction To buffer the dansylation reaction mixture, Gray (19) used 2.3 fl N-ethyl morpholine. This buffer has detergent-like Charac- teristics to help solubilize protein and it keeps the H20 concen- tration low. Table 3 indicates, however, that the use of N-ethyl morpholine during dansylation of free tyrosine can lead to very high levels of N-dansyl tyrosine because as the buffer concentration is increased more N-dansyl tyrosine is produced. Table 4 illustrates that this same phenomena, though not as pronounced, is observed 15 TABLE 3.--The relative amounts of the three dansyl tyrosine deriva- tives at various concentrations of N-ethyl morpholine. The experi- ment was performed exactly as described under Table 1 with the indi- cated concentrations of N-ethyl morpholine. (n.d. = none detected.) Buffer % di-dansyl % N-dansyl % 0-dansy1 tyrosine tyrosine tyrosine 33 mM_N—ethy1 morgnolgng 44 55 n.d. 133 mM N-e-thyl morgflolgng 42 58 n.d. 533 mM_N-ethyl morgnolgng 29 7] n.d. 2230 mMDN-ethyl morgnolgng 1 99 n.d. TABLE 4.--Comparison of Na-bicarbonate and N-ethyl morpholine as buffers for N-terminal determination of hexokinase. Dansylation was performgd as described in the methods section except that 10 uCi of [ H]dansy1 chloride was added to the reaction mixture. Spots were cut out and counted as described under Table l. Buffer % di-dansyl % N-dansyl % 0-dansyl tyrosine tyrosine tyrosine 215 mM_Bicarbonate pH 9.5 99 1 -- 230 mM_ N-ethyl morpholine 37 63 __ pH 10.3 16 with hexokinase and demonstrates conditions under which 99% di-dansyl tyrosine can be generated. Photolysis of Tyrosine on Thin Layer Plates Figure 4 shows the rapid U.V. light catalysed photo- degradation of di-dansyl tyrosine on thin layer plates. The half time for photolysis under these conditions was 3—4 minutes. Figure 3 indicates that di-dansyl tyrosine is photolysed more rapidly than either dansyl glycine or N-dansyl tyrosine, suggesting that dansyl groups attached to amines are more stable than those attached to phenols. Hexokinase N-Terminal Determination The reliability of the revised method for N-terminal deter- mination was evaluated by checking N-terminals on several commercial standard proteins. Lysozyme, cytochrome-c, RNAse, hemoglobin, aldolase, and KDPG aldolase (from W. A. Wood) all unambiguously gave the expected N-terminal amino acid. Using this method, it was determined that hexokinase, puri- fied by the method of Chou and Wilson (25), contained mostly tyrosine, not glycine, as reported previously (25). Quantitation of the dansylated amino acids obtained using tritiated dansyl chloride gave 85% tyrosine, 10% phenylalanine and 5% leucine. The leucine and phenylalanine spots have consistently appeared in many preparations but never exceed 5% and 10%, respectively. This result was verified using the Edman procedure (kindly performed by 17 .Amueswusov mswswesms wswmoszu meseuiwu & use «pepsupeu op umm: we; Am mesmwu mmmv mausuoss mwmxpouoss mew op mswswesms ms_mos»u Fxmseuiwu mo ovges as» .sowpumm muosume esp s_ umswsummu me m~e>smusw use“ msowse> so; ummmpoposs use F mpneh smus: umswsummu me appuexm umusuoss we: mswmosxp Fzmseu m>wuueowuem .mswmosau meseuawu so mwmxHOHoss esp sow mmssou mewhui.u «seams 33:32 00 .\ om ON 2 I . I. u .//. CI o\° o L!) lAsunp-gq waned l aunsoM osoo— 18 Dr. Dean Ersfeld, working in the laboratory of Dr. W. A. Wood). By this method the N-terminal was predominantly tyrosine, with faintly observable traces of phenylalanine and leucine. The N-terminal of chymotrypsin treated pure hexokinase (see Figure 15, Chapter II of this thesis) was found to be about 50% phenylalanine and 50% lysine. The tyrosine was completely removed. Discussion In order to avoid hydrolysis of dansyl chloride and to help solubilize the protein, Gray (19) suggests N-ethyl morpholine to buffer the dansylation reaction. Table 3, however, indicates that such high concentrations (greater than 2 molar) of N-ethyl morpholine interact with tyrosine to inhibit the reaction of dansyl chloride with the phenolic hydroxyl. Because of its low pKa (7.70), low concentrations of N-ethyl morpholine cannot successfully buffer the reaction around the optimum pH of 9.5, also leading to multiple tyrosine derivatives (Table 1). The N-ethyl morpholine effect is seen both with the free amino acid and protein (Tables 3 and 4), so it seems advisable to use pH 9.5 sodium bicarbonate which gives only the di-dansylated product. The increased water content of the reaction mixture did not give any obvious problems with the technique. All the relevant dansylated amino acids except N-dansyl tyrosine can be purchased as pure standards from various chemical companies (e.g., Sigma Chemical Company), suggesting that a method is needed to synthesize N-dansyl tyrosine. These data indicate 19 that N-ethyl morpholine could be used to devise a simple preparation of pure N-dansyl tyrosine in excellent yields. Pouchan and Passeron (22) found that dansyl glycine and dansyl proline (both mono-dansylated in the N-position) underwent photodegradation on thin layer plates at identical rates. The rela- tively mild photolysis conditions in Figure 4 (22, 23) were suffi- cient to degrade most of the di-dansyl tyrosine in 10 minutes while leaving dansyl glycine largely intact (Figure 3). These data indicate that in a dansyl protein hydrolysate exposed to excessive U.V. light, an N-terminal amino acid such as glycine, present in relatively low amounts, could be erroneously assigned the N-terminal in preference to tyrosine. Or in a similar sample, an N-terminal tyrosine might be entirely lost due to excessive exposure to U.V. light, leaving one to conclude that the N-terminal is blocked. In addition, the photolysis products of tyrosine degradation lead to numerous spots that complicate interpretation of any chromatogram. Figure 3 shows that this problem is more obvious with tyrosine than glycine. All of the previously mentioned difficulties have been experienced with rat brain hexokinase and the N-terminal was found to be tyrosine, not glycine as reported by Chou and Wilson (25). A complete explanation for their result is not available, but I have found that the Woods and Wang (24) method that was used does not remove some trace amino acid contaminants from the dansylation reaction that show up in the dansyl protein hydrolysate. Among these contaminants is a prominent glycine which could have been 20 confused for the N-terminal. And in view of all the difficulties I have uncovered regarding the rapid photolysis and consequent loss of dansyl tyrosine (Figures 3 and 4), it is not hard to imagine why this amino acid was not observed.' Summary Two of the three dansyl chloride derivatives of tyrosine have been identified and characterized by three independent methods. These two products are the mono- and di-dansylated forms of tyrosine. The mono-dansylated species has the dansyl group attached to the primary amine and the di-dansylated product is labeled at the phenolic oxygen and at the primary amine. N-ethyl morpholine, the standard solvent for protein dansylation as suggested by Gray (19), was found to give an unexpectedly high yield of mono-dansylated tyrosine; therefore, sodium bicarbonate was used as buffer since this was found to favor complete conversion to the di-dansyl product. Under these circumstances, the N-terminal amino acid of rat brain hexokinase was found to be tyrosine, not glycine, as previously reported (25). Chymotrypsin treatment of this enzyme exposed two new N-terminals, phenylalanine and lysine, completely removing tyrosine. And finally, it has been shown that di-dansyl tyrosine has an unusually high photolysis rate under U.V. light, a property which can easily lead to complications in interpretation of experimental results unless measures are taken to restrict photolytic degradation. CHAPTER II PARTIAL CHARACTERIZATION OF BINDABLE AND NON-BINDABLE HEXOKINASE ISOZYMES Introduction The main thrust of this research topic has been to describe the nature of the interaction between rat brain hexokinase and mito- chondria. To completely satisfy this query, it will be necessary to purify both membrane and enzyme components required in the binding process. This chapter deals with the purification of enzyme that can bind to mitochondria and the partial characterization of some requisite binding components on the enzyme. Materials and Methods Chemicals Biochemicals and HEPES buffer were obtained from Sigma Chemical Company. DEAE-cellulose was purchased from Gallard- Schlesinger. Polypropylene microcentrifuge tubes were the product of Brinkmann Instruments. All other chemicals were reagent grade, obtained from commercial sources. Adult male and female rats (ranging from 150-500 gm) of the Sprague-Dawley type were obtained from Spartan Research (Haslett, Michigan) and maintained on a common laboratory diet and water a_d_ libitum. 21 22 Hexokinase Assay_ Hexokinase activity was determined spectrophotometrically at 25°C i 0.5 in an assay mixture containing 3.3 mM_91ucose, 6.7 mM_ATP, 6.7 mM Mg C12, 40 mM_HEPES, 10 mM_l-thioglycerol, 0.64 mM_NADP, and 2 units of 91ucose-6-phosphate dehydrogenase in a total volume of 1.0 m1 (pH 7.5). The enzyme sample was added to an assay mixture containing all components except ATP, and oxidation of any endogenous glucose-6-phosphate was permitted to go to comple- tion if necessary (up to approximately 3 min). Subsequently, the hexokinase assay was initiated by addition of 0.030 ml of 220 mM_ATP solution (pH 7.3). NADPH formation was followed at 340 nm with a Turner Model 330 Spectrophotometer connected to a Sargent SRL Recorder. One unit is defined as the amount of enzyme which cata- lyzes the formation of l pmole of glucose-6-phosphate per minute. Preparation of Mitochondria Rat liver mitochondria were prepared by homogenizing the liver from a starved rat (15-17 hr, with water 29 libitum) in 10 volumes (10 ml/gm tissue) cold 0.25 M_sucrose with a Teflon-glass homogenizer. The homogenate was centrifuged at 6009 for 10 min and the pellet discarded. The supernatant was centrifuged at 6,5009 for 15 min and the supernatant discarded. The 6,5009 pellet was resu5pended in 10 volumes sucrose and centrifuged. The final pellet was resu5pended in 2 volumes of 0.25 M_sucrose and 1 ml aliquots stored at -20°C. 23 Crude rat brain mitochondria were prepared by homogenizing frozen brains (stored in liquid N2) in 10 volumes of 0.25 M sucrose with a Teflon-glass homogenizer. The homogenate was centrifuged at 40,0009 for 15 min and the supernatant discarded. The pellet was washed three times by rehomogenization in 10 volumes 0.25 M_sucrose and centrifugation. The final pellet was suspended in 4 volumes 0.25 M_sucrose and frozen at —20°C in 1 ml aliquots. Glucose-6-Phosphate Solubilized Enzyme . Rat brains, frozen in liquid N2, were thawed and homogenized in 0.25 M_sucrose (10 ml/gm). The homogenate was centrifuged at 10009 for 10 min, and the pellet discarded. The 10009 supernatant was centrifuged at 40,0009 for 10 min and the resulting pellet washed several times, depending on the experiment, by rehomogeniza- tion in 10 volumes 0.25 M_sucrose and centrifugation. The washed pellet was resuspended in 10 volumes 0.25 M_sucrose containing 1.2 mM_glucose-6-phosphate and incubated for 30 min at 25°C. The solubilized enzyme was obtained in the supernatant after centri- fugation at 40,0009 for 10 min. Purification of Rat Brain Hexokinase The three times washed particulate fraction from 195 gms rat brain was prepared exactly according to Chou and Wilson (25). The resuspended particles were incubated with l mM_glucose-6- phosphate at 25°C for 1 hour and centrifuged at 40,0009 for 40 min. The supernatant, which contained solubilized hexokinase, was 24 decanted and potassium phosphate buffer (0.5 M, pH 7.0), glucose (0.5 M), NaZEDTA (25 mM), and thioglycerol (12 M) were added to final concentrations of 0.01 M, 0.01 M, 0.5 mM, and 10 mM, respec- tively. The supernatant fluid was then concentrated by use of an Amicon ultrafiltration device with a PM-10 membrane, to about 300 ml and centrifuged a final time at 105,0009 for 1 hour. This super- natant was applied to a DEAE-cellulose column and washed exactly according to Chou and Wilson (25). The enzyme was eluted from the column with a 600 ml linear gradient from 0 to 0.2 M KCl in column buffer, collecting 3.8 m1 fractions. Overall recovery of glucose- 6-phosphate solutilized hexokinase was 83%. 1% Sodium Dodecyl Sulfate- Pblyacrylamide Disc Gel Electrophoresis The method of Fairbanks et al. (26) was only slightly modi- fied. Samples (1-5 mg/ml) were prepared in 1% sodium dodecyl sulfate, 5-10% sucrose, 10 mM Tris-HCl (pH 8.0), 1 mM_EDTA and 2% mercaptoethanol. They were then heated at 100°C for 15 min, 0.2% pyronin B tracking dye was added, and the samples were layered on 5.6% polyacrylamide gels (5 mm x 100 mm) prepared in tubes which had been coated with dimethyl dichlorosilane. Electrophoresis was performed at constant current of 4 ma/gel with a running time of about 4 hours. Gels were fixed and washed by agitating each gel in a 30 m1 capacity test tube with 10% TCA/33% isopropanol. The washing solution was changed every two hours for three washes. The dehydrated gels were placed into 10% TCA until they regained their 25 original size and then were stained with xylene brilliant cyanin-G (K + K Laboratories, Inc., Plainview, N.Y.) according to published procedures (27, 28). The rather extensive gel wash removes $05 that interferes with the stain. Mitochondrial Binding Assay All assays were done in polypropylene microcentrifuge tubes coated with 2% BSA. The coating was done by dipping each tube in 2% BSA and drying in an oven at about 60°C. If the tubes were left uncoated, artifaCts were introduced as explained below. For the assay, aliquots of hexokinase and either liver or brain mitochondria were mixed in a microcentrifuge tube and 3 mM MgCl2 was added. Tubes were incubated at 0°C for 15 min and Spun at room temperature in an Eppendorf 3200 microcentrifuge for 2 min. Hexokinase activity in the supernatant was measured directly. Pellets were assayed after suspending them in a known volume of 0.5% Triton X-100 - 0.25 M_sucrose by vortexing in the presence of glass beads (Sargent, No. S-61740, size A-7) until homogeneous. With each new mitochondrial preparation the number of binding sites per aliquot was approximated by titrating in a fixed volume of mito- chondria with increasing hexokinase. In this way, it was always possible to determine when the binding sites were in excess for a given amount of hexokinase. When pbrified enzyme, containing column buffer, was assayed for bindability to mitochondria, a l to 20 dilution into a 0.25 M sucrose-50 mM glucose was first made in order to lower the salt concentration. 26 Sephadex Column Chromatography of Chymotrypsin Treated Pure Hexokinase In order to remove any small molecular weight fragments, .. before use pure hexokinase and alpha-chymotrypsin (Worthington iochemical Corporation) were individually passed over a Sephadex G-25 column that was equilibrated with DEAE column buffer (10 mM potassium phosphate pH 7.0, 10 mM glucose, 0.5 mM_NaZEDTA, 10 mM. thioglycerol). Chymotrypsin (4 mg) was incubated at 25°C for 10 min with 1 mg hexokinase in a total volume of 15 mls. The sample was frozen and lyophilized down to 1.5 mls and 70% of the hexo- kinase activity was recovered. This mixture was sephadexed on a G-50 column that was equilibrated with DEAE column buffer. From this column, 100% of the remaining activity was recovered (total recovery equals 70%). Results Binding of Hexokinase to Poly- propylene Centrifuge Tubes Figure 5 shows that adding G6P-solubi1ized hexokinase to polypropylene centrifuge tubes decreases the activity. As the G6P-solubilized enzyme becomes more pure,1 exposure to polypropylene leads to a greater activity loss. Figure 6 indicates that succes- sive assays taken out of the same tube contain less hexokinase activity. Both of these problems are eliminated by coating the 1Albert Chou has shown that as mitochondria are washed more times the solubilized enzume obtained from them has a higher Specific activity. (Personal communication.) 27 100 \ BSA coated ty rema' Ivi /uncoated percent act 500 1 2 3 4 number of washes Figure 5.--Increased binding togpolypropylene with increas- ing purity of hexokinase. G6P-solubilized hexokinase was obtained from brain particles that were washed various times as described under methods. Activities ranged from 0.56 units/m1 (0 times washed) u)0.37 units/ml (5 times washed). As mitochondria are washed more often, the solubilized enzyme becomes more pure (see footnote 1). About 0.3 ml of the enzyme was added to and vortexed in BSA coated and uncoated tubes. Freshly prepared (0), 18 hours refrigerated (c>), and 42 hours refrigerated (C3) enzymes all responded the same. 28 NBC); H BSA coated o~ _ <3 uncoated percent activity remaining 2c>i o 100 200 pl removed Figure 6.--Assaying successive aliquots from a polypropylene tube. 0.2 m1 of hexokinase (0.43 units/ml) was put in a polypropo- lene tube and vortexed. 50 ul was removed for the first assay, 50 ul for the second point, and so on. 29 tubes with 2% BSA. Washing uncoated tubes with detergent, 1:1 chloroformzmethanol, 6N HCl or 1% dichlorodimethyl silane in benzene, has no effect. The activity loss was too fast to deter- mine any time dependence (faster than 1 min). Figure 7 demonstrates that an aliquot vortexed in succes- sive tubes progressively loses activity, suggesting that hexo- kinase binds to polypropylene. BSA coated tubes showed a markedly diminished effect. It was demonstrated that hexokinase activity remains tightly bound in spite of washing. A tube containing hexokinase was rinsed several times with 0.25 M sucrose and hexokinase assay mix was added to the tube. Table 5 shows that less glucose-6- phosphate accumulated in the BSA coated tubes than in the uncoated tubes. These data prove that hexokinase can bind tightly to poly- propylene and that BSA substantially reduces the effect. TABLE 5.--Assay of the activity that remains bound to polypropylene tubes. G6P-solubilized hexokinase from twice washed particles (0.52 units/ml) was added to BSA coated or uncoated tubes, drawn out and the tubes then washed with 0.25 M sucrose. Hexokinase assay mix containing everything but glucose-6-phosphate dehydrogenase was added to the tubes and incubated at 25°C for 15 min. The "control" was an uncoated tube that did not have any hexokinase added. uMole G6P Formed in 15 min Control 0.0 Uncoated tube 0.080 BSA coated 0.035 30 Too 4 BSA coated -e uncoated percent activity remaining on O 2 4 6 tube number Figure 7.--Hexokinase binding to successive polypropylene tubes. G6P-solubilized hexokinase from twice washed particles (0.52 units/m1) was added to BSA coated (e) or uncoated (o) polypropylene tubes, drawn out and added to other tubes for the indicated number of times, and finally assayed. I‘ll 5". ll llI lllllll .Ill! 31 The Effect of Salt on MgClz-Induced Rebinding ofTHexoanase_to Mitochondria From Figure 8, it is clear that low ionic strength severely inhibits binding of G6P-solubilized hexokinase to liver mitochondria. Fifty percent inhibition occurs at 20-30 mM_ionic strength. Figure 9 shows that the salt effect is not as pronounced with rat brain mitochondria. Binding Assgy With Pure Hexokinase Figure 10 Shows the results of a binding assay on pure rat brain hexokinase using low salt, BSA coated centrifuge tubes and rat brain mitochondria. In this experiment, the enzyme was more than 50% bindable. This is the first time that purified type I mammalian hexokinase has been reported to bind to mitochondria. Hexokinase Isozymes It was noticed that pure hexokinase as eluted from a DEAE- cellulose column, according to Chou and Wilson (25), sometimes yielded a slightly skewed curve. When a shallower salt gradient was used to elute the enzyme, the results in Figure 11 were obtained. The elution pattern clearly indicates two enzyme types represented by the shoulder eluting at 0.065 M KCl and the peak at 0.075 M KCl. An even shallower gradient resolves two distinct peaks (Figure 12). The relative amounts of the two peaks have varied in several experiments from about 75%:25% to 25%:75%; the reasons for this remain under investigation. 1% SDS gel 32 percent recovery 0 . .050 .100 , ionic strength Figure 8.--The inhibition of rebinding by increasing ionic strength. Binding assays were performed as describedTTn the text with G6P-solubilized hexokinase from 1X washed mitochondria (0.58 units/m1), with liver mitochondrial binding sites in excess. Appropriate salt concentrations were added to the tubes before adding MgClz. Open dots (c), A ) represent KCl and closed dots (e , A ) potassium phosphate. 33 Too" percent recovery . .050 .Too ionic strength Figure 9.--Salt-induced inhibition of binding; comparison between liver and brain mitochondria. This figure compares the difference in the salt effects between rat liver and brain mito- chondria. Experimental details were the same as under Figure 8. 34 100 supernatants 50 percent recovery o . 25 so pl mitochondria Figure 10.--Bindability of pure hexokinase. The binding assay was done as described in the methods section for pure enzyme using rat brain mitochondria, low salt and BSA coated tubes. 100% recovery (ordinate) in this experiment was equivalent to 0.26 units/m1. 35 .so_uoom muospoe use s? uoswsomou me umssmeos use: uszos “smosoo use Aev 33.5.8 mmesono: .msoBoe: use. so 3:953.»er SEBoousoo .3 uosEsopou we: on soEespsoosoo 8V. 9:. .5593 muosuos. 2: E umstomou me uosCooCoo mes xsoesmopesosso .mmeswxoxms smesn peg wo soesmopeEosso sE:Poo mmo_:FFoqusomso speme moso mp xppppse mspusps use .msoppoese FE w.m msppomppou .smwmzs caepoo sp pox.a mp.o op o.o Eose psopuesm Leospp FE coo e spwz umpzpm we: oEsto use .eou paose xpso we: >so>ooos mesto osp .Asowpem Ismpspsoo sow uopsp we: sopos pesoN ev uopssmppe mspms use: mesopssoop 3o: oEom oospm pas .pp ossmps sop we osem osp xppeppsommo osoz psospsooxo mesp so mppepou use .psopuesm sozoppesm e sppz soppspo momespxoxos specs pes so asmesmopeeosso caspoo omo—oppooim eoueqaosqe 4O .su>s_ so ueupmss ueme use: essusossoopss ssess pem .eos so mssppsso use asupv ppoe umessxoxes esp pousse pos usu pseEpeusp ssmssspoexsu .spsssse mssusss xospmuu pos upu msospsu Isoo peoppsuus suuse uupeusp essusosooppz .useueooss useusepm esp he ueEsosses use; meemme assussm .sospmumsu umeeposs sopm op ueuue we: Aeussoesspesossemsxspuesxsessv sts.mE m. use eos so uusssso use: musep sse .uossus sospeseoss esp so use usp p< .ssmsssposxso so mpseose uupeopuss esp spsz comm pe sss op sos uepeseoss me: npe\mpsse pu.ov umessxoxus so use m.o .umessxoxus uueso esp sos .mumessxoxus uses use uueso so psespeusp sswmxspoExso stmssusss so sospssssspii.up usemss 41 On \ E _\:_mn>spoE>:o m1 ON 0— 'VI'-p. .-.-,--, r . JD on BUIPUIq £0 uomqmu: % 42 .supesosoessopossospousm seezom e so uuspssupuu we; uosuomusoess e>spepus .msospoespmss m.susepoesesee esp op mssusoooe A.osp .esooseeissesssozv esesesss assme esseosspueosoeps uessssupuu we; sseposs .Auseepop sepss — ”sesos Em mo.o moss Em es ssepxooo uemes esuesop e ss supseoo sospessspssom useese usexoes e ss uepseoo me: xps>spoeosues .Aoev museuuooss uesmssses op assusoooe .msueu.a s.o so uosumuss esp ss Asupme szspe ussmoszpieiseoesusizv muss mssme oomm pe uesemees me: ssmsxspoexsu .muospus.:x:=euusssomuu me umessxoxes sos uuxemme use .uussesmopeEosso .uupeusp me: ussEem ess .emessxoxes uepeesp ssmsxspoexso so ssQMsmopeEosso sEesoo omiu xuuessemii.ms usemss 43 PROTEIN relative fluorescence no ..... no. 1000 0 “c-cwcose 100,000 (f ..... . C / ....... IW'ul‘Vaug 'G'OV NISdAlIlOWAHD .. ' lW/ssgun JSVNIXOXil-l 20 Gene 2" d .g L FRACTION NUMBER 44 Potassium Chloride-Induced Rebinding of Hexoanase to Mitochondria Figure 16 demonstrates the phosphate-induced rebinding phenomena observed in several laboratories (29, 47, 48). Since phosphate had been Shown to have the specific effect on some hexokinases of overcoming glucose-6—phosphate inhibition (29, 30), it was presumed that phosphate-induced rebinding of hexokinase was also Specific. Figure 10 suggests that this is not the case because reversal of solubilization can be induced equally well with potassium chloride. Discussion Hexokinase Binding to Polypropylene It has long been known that certain enzymes such as RNAse bind to glass test tubes (31), but no reference has been made regarding binding to polypropylene tubes. The forces responsible for polypropylene binding are probably analogous to those in hydrophobic affinity chromatography (32, 33). Supporting this idea, J. E. Wilson has found that hexokinase can stick to an affinity column prepared by coupling glucosamine to agarose via a long hydrophobic arm. The enzyme was eluted by high salt, but not by glucose; this would be consistent with hydrophobic interactions being required rather than binding to the glucosamine moiety II (personal communication). Polypropylene ( -C-C- )n is extremely HCH3 hydrophobic and should make good affinity material for hexokinase. 45 U1 0 units ml Ivity _ 25 phosphate supernatant act 0‘ .020 .040 ionic strength Figure 16.--Reversibi1ity of solubilization py salt. Before removal of the particles from 2X washed G6P-solubilized hexokinase, an appropriate aliquot of 0.1 M potassium phosphate or 1.0 M KCl was added. These tubes were incubated at 25°C for 30 min, spun in an Eppendorf micrOcentrifuge and the supernatants assayed. 46 It has already been shown that serum albumins can be purified by hydrophobic affinity methods (33), so it iS not unreasonable that BSA could be binding to the available Sites on the polypropylene tubes in place of hexokinase. It would be interesting to try a polypropylene fiber affinity column to purify hexokinase. Inhibition of Rebinding by_ Ionic Strength Salt effects on solubilization of hexokinase have been thoroughly studied (10), but salt effects on MgCl2 mediated rebind- ing have not. Since salt-induced solubilization was observed to be 50% maximum at 0.120 p ionic strength, it was always assumed that low ionic strength (less than 0.05 u) would not greatly affect rebinding. However, the data on Figure 8 indicate that 50% inhibition of rebinding occurs at 0.025u clearly indicating that the ionic strength in the binding assay must be kept very low (less than 0.005u). Differences Between Liver and Brain Mitochondria A considerably diminished salt effect was observed for the MgCl2 mediated rebinding of hexokinase to rat brain particles. This marks the first time that a difference between the binding sites in mitochondrial preparations from different tissues has been observed. After this discovery, J. E. Wilson found that there is a much higher energy of activation for the binding of hexokinase to brain particles than liver mitochondria (personal communication). 47 Investigation into the nature of these differences is war- ranted. Since the brain preparation is much cruder than liver, the brain particles must first be purified to see if the differences between rat and brain depend on the purity of the organelles rather than an actual difference in the membrane binding sites. To carry out a further purification, outer mitochondrial membranes from both tissues could be isolated (9, 34) and checked for differences in binding. Purified Bindable and Non-Bindable Isozymes With the increased knowledge concerning rebinding and an appropriately modified assay, it has been possible to Show for the first time that pure hexokinase can bind to mitochondria (Figure 10). In addition, two new isozymes have been obtained in a puri- fied form. Type Ib is the bindable form and type In is non-bindable. Investigation into the differences between these two types is obviously required. To aid in this type of study, a quick assay method for the different isozymes is necessary. For this purpose, isoelectric focusing on gels is proposed. Preliminary evi- dence has already indicated microheterOgeneity in different prepara- tions of purified hexokinase with at least three isozymes. This heterogeneity was not observed previously (25) because the sucrose density gradient technique employed by Chou and Wilson is inherently less sensitive (35, 36). There are numerous possible explanations for the hetero- geneity (37). Genetic heterogeneity and errors made during 48 synthesis of the messenger RNA or protein strand are ruled out because these effects would be constant in all preparations. In fact, different preparationsIrfpurehexokinase have different ratios of type Ib to type In. The enzyme does not contain carbohydrate (38) and has only one subunit (25, 39), removing these variables from consideration. N-terminal cleavage has been tentatively ruled out because both forms contain the same N-terminal amino acid, tyrosine. The possibility that a second tyrosine is generated after proteolysis cannot be ignored, however. C-terminal cleavage is also possible. Three promising possibilities for the heterogeneity are sulfhydryl modification, deamination and/or stable conformational isomers. Of these possibilities, the first two are documented with examples (39, 40, 41, 42). There is no proven example of the lat- ter; however, the possibility has been raised (43, 44, 45). The sulfhydryl modification idea is easily tested by treat- ing the protein with DTNB and stopping the reaction at various times with a stabilizing substrate. It has already been shown by Chou (39) that different enzymes whose sulfhydryls have been variously modified can be obtained in this way. It may be fbund that certain sulfhydryls are required for binding to mitochondria and certain are not. Pure Hexokinase and the "BindingiPeptide" The ability of chymotrypsin to convert hexokinase to a non- bindable form was first shown by Rose and Warms with crude ascites 49 tumor enzyme (8). Following the precedence of cytochrome—b5 reductase, their result suggests that the enzyme contains a bind- ing peptide that "holds on to" the membrane like a prehensile tail (46). Figure 14 demonstrates the chymotrypsin-induced loss in bindability of a pure preparation and Figure 15 shows that the cleaved fragment(s) can be easily isolated. N-terminal analysis of the chymotrypsin treated enzyme indicates that this fragment(s) comes, at least partially, from the N-terminus because N-terminal tyrosine is completely replaced by phenylalanine and lysine (Chapter I of this thesis). The amino acid composition of the fragment obviously needs to be determined. An alternative explanation, of course, would be that chymotrypsin may not be removing a ''binding peptide" but instead simply altering the protein conformation to make it non-bindable. In this case, the amino acid composition of the fragment(s) would not be expected to be particularly notable. Salt-Induced Reversal of Solubiliza- tion and the Putative Binding_ Mechanism Phosphate reversal of solubilization has been demonstrated in several laboratories (29, 47, 48), but in none of their work was a control run to see if the reversal was an ionic strength effect. In Figure 10 is the first demonstration that potassium chloride works as well in the binding assay as potassium phosphate. In light of this new evidence, a hypothesis for the mechanism of binding can be proposed. This hypothesis is somewhat 50 analogous to a membrane reconstitution hypothesis already in the literature (49). Because of phospholipids that are in membranes, there is a negative charge envelope around each mitochondria at neutral pH (50). Since the pI of hexokinase is 6.35 (25) it, too, has a negative charge at neutral pH and consequently the two negatively charged particles repel each other. Binding of cations, especially divalent, serves to decrease the negative charge envelope around the mitochondria, allowing hexokinase to interact with the membrane binding site. Glucose-6-phosphate affects solubilization by causing a conformational change in the enzyme that makes it more easily repelled by the membrane. This hypothesis can be readily tested with a number of obvious experiments. (i) The mitochondrial charge barrier can be gradually removed by phoSpholipases which Should lead to a diminishing salt requirement for binding. (ii) Lowering the pH Should lead to a diminished salt requirement as the enzyme and membrane become less negatively charged. (iii) The cationic por- tion of the salt should show some specificity; i.e., some cations should work better than others. Summar Type I rat brain hexokinase purified by the method of Chou and Wilson (25) has been shown to consist of at least two isozymes. One of these isozymes (type Ib) binds to mitochondrial membranes and the other (type In) does not. Recently discovered peculiarities in the binding assay have obscured this discovery 51 for some time. (i) Hexokinase binds to the polypropylene micro- centrifuge tubes routinely used for the binding assay. (ii) There is an unexpected inhibition of binding by relatively low concentra- tions of salt. (iii) The routinely used rat liver mitochondria bind less hexokinase and are more susceptible to the salt effect than rat brain mitochondria. A peptide(s) can be removed from pure hexokinase, at least part of which comes from the N-terminal end and may be involved in the binding process. This peptide(s) awaits further character- ization. The phosphate-induced reversal of solubilization first demonstrated by Rose and Warms (8) and confirmed by other workers (47, 48) can also be done with potassium chloride. This salt- induced rebinding suggests a hypothesis for the binding mechanism. Further experiments will be required to ascertain whether or not the salt-induced rebinding is due to identifical factors necessary for phosphate-induced rebinding. REFERENCES 52 10. 11. 12. 13. 14. 15. 16. REFERENCES Crane, R. R., and $015, A., J. Biol. Chem., 203, 273 (1953). Johnson, M. K., Biochem. J., 22, 610 (1960). Beattie, D. S., Sloan, H. R., and Basford, R. E., J. Cell ' Biol., 19, 309 (1963). Bachelard, H. S., Biochem. J., 104, 289 (1966). Craveg, P. A., and Basford, R. E., Biochemistry, 8, 3520 1969 . Craven, P. A., Goldblatt, P. J., and Basford, R. E., 819: chemistry, 8, 3525 (1969). Rose, I. A., and Warms, J. V. 8., Fed. Proc., 24, 297 (1965). Rose, I. A., and Warms, J. V. 8., J. Biol. Chem., 242, 1635 (1967). Kropp, E. S., and Wilson, J. E., Biochem. Biophys. Res. Commun., 28, 74 (1970). Wilson, J. E., J. Biol. Chem., 243, 3640 (1968). Tuttle, J. P., Wilson, J. E., Biochim. Biophys. Acta, 212, 185 1970 . Lowry, 0. H. Passonneau, J. V., Hasselberger, F. X., and Schultz, D. W., J. Biol. Chem., 239, 18 (1964). Veech, R. L., Harris, R. L., Veloso, 0., and Veech, E. H., J. Neurochem., 29, 183 (1973). Sacktor, 8., Wilson, J. E., and Tiekert, C. G., J. Biol. Chem., 24.1. 5071 (1966). Knull, H. R., Taylor, W. F., and Wells, W.IN..LJ.Biol. Chem., QB. 5414 (1973). Knull, H. R., Taylor, W. F., and Wells, W.1N.,.J.Biol. Chem., 249, 6930 (1974). 53 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 54 Gros, C., and Labouesse, 8., European J. Biochem., Z, 463 (1969). Weiner, A. M., Platt, T., and Weber, K., J. Biol. Chem., 247, 3242 (1972). Gray, W. R., Methods in Enzymology, 28, 121 (1972). Hartley, 8. S., Biochem. J., 119, 805 (1970). Greenstein, J. P., and Winita, M., Chemistry of the Amino Acids, Wiley and Sons, 8, 2353 (196111 Pouchan, M. I., and Passeron, E. J., Analytical Biochem., 88, 585 (1975). D'Souza, L., Bhatt, K., Madaiah, M., and Day, R. A., Arch. Biochem. Biophy§,, 141, 690 (1970). Woods, K. R., and Wang, K. T., Biochim. Bipphys. Acta, 133, 369 (1967). Chou, A. C., Wilson, J. E., Arch. Biochem. Biophys., 151, 48 (1972). Fairbanks, C., Steck, T. L., and Wallack, D. F. H., Biochemistry, 18, 2607 (1971). Diezel, N., et al., Anal. Biochem., 48, 617 (1972). Molick, N., and Berzie, A., Anal. Biochem., 48, 173 (1973). Kosow, D. P., Oski, F. A., Warms, J. V. B., and Rose, 1. A., Arch. Biochem. ijphys., 157, 114 (1973). Rose, I. A., Warms, J. V. B., and O'Connell, E. L., Biochem. BiOphys. Res. Commun., 18, 33 (1964). Parish, J. H., Principles and Practice of Experiments with Nucleic Acids, Longman Group Limited (1972). Hofstee, B. H. J., Anal. Biochem., 82, 430 (1973). Peters, T., Jr., Taniuchi, H., and Anfinsen, C. 8., Jr., J. Biol. Chem., 248, 2447 (1973). Sottocasa, C. L., Kuylenstierna, B., Ernster, L., and Berg- strand, A., J. Cell Biol., 82, 415 (1967). Righetti, P. 6., and Drysdale, J. W., J. Chromat., 88, 271 (1974). 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 55 Haglund, H., Meth. BioChem. Anal., 18, l (1971). Williamson, A. R., Salaman, M. R., and Kreth, H. W., N.Y. Acad. Sci. Annals, 209, 210 (1973). ’ Craven, P. A., and Basford, R. E., Biochem. Biophys. Acta, 338. 619 (T974). Chou, A. C., Thesis, Doctor of PhiloSOphy, Michigan State Uni- versity (1973). Edmunson, A. 8., Nature (London), 205, 883 (1965). Elatmerk, T., Vesterberg, 0., Acta Chem. Scand., 29, T947 1966 . Chou, A. C., and Wilson, J. E., Arch. Biochem. Biophys., 163, 191 (1974). Kaplan, N. 0., Ann. N.Y. Acad. Sci., 151, 382 (1968). Kitto, B. G., Wasserman, P. M., and Kaplan, N. 0., Proc. Nat. Acad. Sci., 88, 578 (1966). Kitto, B. G., Solzenbach, F. E., and Kaplan, N. 0., Biochem. BiOphyS. Res. Commun., 88, 31 (1970). Spatz, L., and Strittmatter, P., J. Biol. Chem., 248, 293 (1973). Purich, D. L., and Fromm, H. J., J. Biol. Chem., 246, 3456 (1971). - Wilson, J. E., Arch. Biochem. Biophys,, 188, 543 (1973). Rottem, S., Stein, 0., and Razin, S., Arch. Biochem. Biophys., L25. 46 (1968). Racker, E. F., Membranes of Mitochondria and Chloroplasts, Van Vostrant Reinhold Company, New York (1970). Perlmann, G. E., and Lorand, L., Meth. in Enzymol., 18, 38 (1970). "1111111111IIII'IIIII