STUDEES ON A HWSEFLY P—HOSPHODEESYM {4.611% GN VAREOUS BEGLOGICAL GLYCEROPHOSPHODIESTERS Thesis f0? the Degree f0 Ph. D. MECHE'GAN STATE UNEVERSiTY GEORGE R. HILDENBRANDT 1971 ”wt-Gil. This is to certify that the thesis entitled STUDIES ON A HOUSEFLY PHOSPHODIESTERASE ACTING ON VARIOUS BIOLOGICAL GLYCEROPHOS PHOD I EST ERS presented by GEORGE R. HILDENBRANDT has been accepted towards fulfillment 20m requiremen? I fl/ 1 d e 'n r ,A/Z'Lf’gwfl'w' M /’ egr e 1 fl fl/éflg/K) Major professor Date y/y‘lj/y/ 0-7639 ABSTRACT STUDIES ON A HOUSEFLY PHOSPHODIESTBRASE ACTING ON VARIOUS BIOLOGICAL GLYCEROPHOSPHODIESTERS By George R. Hildenbrandt Housefly larvae metabolize phosphatidylinositol, phosphatidylserine, and phosphatidylglycerol by a micro- somal phospholipase A to the respective monoacylglycero- phosphatides. The monoacylglycerOphosphatides are dea- cylated by a lySOphospholipase to the resPective glycero- phosphoryl derivatives. An 88,000 x g soluble enzyme hydrolyzed glycerophosphoryl (GP)-ethanolamine (E), -choline (C), -B-methylcholine (BMC), -inositol (I), -serine (S), and -glycerol (G) to i-a-glycerolphosphate and the respective free base. The phosphodiesterase is stable to heating at 50° for 15 min and is not inhibited by sulfhydryl reagents. It is stable to treatment with some proteases and the reaction products are not inhibi- tory. The pH optimum is 7.2-7.” for GPC and GPE. Added Mg++ greatly stimulates activity. EDTA completely inhibits activity, but the inhibition is reversed by Mg++. Co++ is as stimulatory and Mn++ and Ni++ are about 50% as George R. Hildenbrandt stimulatory as Mg++. Other divalent cations tested show no stimulation. Mutual inhibition of hydrolysis by pairs of substrates incubated together suggests a single enzyme acting on all of the substrates. CPC is the preferential substrate on the basis of most favorable competition for hydrolysis, a 70% higher rate of hydrolysis than any other substrate (90 vs 50:§ nmoles/min/mg protein), and the lowest Km. The Km's for GPC, GPE, and GPI are 2 x 10'”, u x 10-”, and 2 x 10-3 M respectively. A tightly bound enzyme-glycerolphosphate intermediate is indicated by the following results. Labeled choline is exchanged into GPC and GPI but l-a-glycerolphosphate is not exchanged into GPC. Net reversal of the reaction is not observed. STUDIES ON A HOUSEFLY PHOSPHODIESTERASE ACTING ON VARIOUS BIOLOGICAL GLYCEROPHOSPHODIESTERS By George R. Hildenbrandt A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1971 , ,1. flex C C5 DEDICATION This thesis is dedicated to the beginning of the reassumption by pe0ple of the power unreasonably or ineffectually assumed by institutions. All power is in the people--may we grow to exercise it well. ii ACKNOWLEDGMENTS Without substantial.tolerance and good faith on the part of Loran Bieber, this thesis would not have been possible. My deepest gratitude is extended to the beautiful people with whom I have associated this past year through the Lansing Area Peace Council. Without the joy, love, and strength gained from these relationships, the absurdity of the past year would certainly have driven me from this doctoral program, iii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF ABBREVIATIONS INTRODUCTION Statement of problem Introduction to glycerophosphatide catabolism Insect glycerophosphatide metabolism Review of glycerophosphodiesterase literature EXPERIMENTAL PROCEDURES AND RESULTS DISCUSSION OF RESULTS Diacylglycerophosphatide catabolism in housefly larvae PrOperties of housefly larval glycero- phosphodiesterase CONCLUSIONS LIST OF REFERENCES APPENDIX A Metabolism of Ceramide Phosphorylethanol- amine, Phosphatidylinositol, Phosphatidyl- serine, and Phosphatidylglycerol by Housefly Larvae, G. R. Hildenbrandt, T. Abraham, and L. L. Bieber, Lipids E, 508 (1971). iv Page vi vii ix IO 13 15 15 16 19 23 508 TABLE OF CONTENTS--Continued Page APPENDIX B Characterization of GlycerOphosphoryl- Choline, -Ethanolamine, -Serine, -Inositol, and -Glycerol Hydrolytic Activity in Housefly Larvae, G. R. Hildenbrandt and L. L. Bieber, (manuscript for publication). 25 Table II. III. IV. VI. VII. LIST OF TABLES APPENDIX A Distribution of the chloroform-soluble radioactivity obtained from the enzymatic hydrolysis of 3H-ceramide phOSphoryl- ethanolamine. APPENDIX B Isolation of glycerophosphodiesterase from housefly larvae. Effects of heating and EDTA on glycero- phosphodiesterase activity. Effects of divalent cations on glycero- phosphodiesterase activity. Identification and quantitation of pro- ducts from hydrolysis of glycerophospho- diesters by larval phosphodiesterase. Effects of potential inhibitors of glycerophosphodiesterase activity. Effects of potential enzyme inhibitors on glycerophosphodiesterase activity. Reversal of enzymatic hydrolysis of GPC and exchange of Me-luc-choline or i-a- glycerol32P into GPC or GPI. vi Page 513 35 36 L42 1+6 51 52 56 Figure LIST OF FIGURES Diacylglycerophosphatide catabolic pathway in housefly larvae. APPENDIX A Tracings of the radiochromatogram scans of chromatograms of the chloroform- soluble and water-soluble extracts obtained from larval microsomes incubated with 2P-labeied phosphatidylglycerol. Tracings of the radiochromatogram scans of the chloroform-soluble and water- soluble extracts obtained from larval micro- somes incubated with 32P-labeled phospha- tidylserine. Tracings of radiochromatogram scans of chromatograms of the chloroform-soluble and water-soluble extracts obtained from larval microsomes incubated with 32P-labeled phosphatidylinositol. Time course for deacylation of 32P-labeled phOSphatidylinositol by housefly larvae microsomes. PH optimum and time course for hydrolysis of 3 P-labeled ceramide phosphoryl- ethanolamine. Identification of 32P-labeled phosphoryl- ethanolamine as a product of the cleavage of ceramide phosphorylethanolamine. APPENDIX B PH dependence of GPC and GPE hydrolysis by glycerophosphodiesterase. vii Page 20 509 510 510 511 511 512 A0 H. viii LIST OF FIGURES--Continued Effect of Mg"+ concentration on GPC and GPE enzymatic hydrolysis. Determination of Km values for GPC and GPI hydrolysis by glycerophosphodiesterase. Mutual inhibition of enzymatic hydrolysis by pairs of glycerophosphodiesters. Page w H9 55 EDTA GP GPBMC GPC GPE GPG GPI GPS GP-X PC PE PG PI PS LIST OF ABBREVIATIONS ethylene diaminetetra-acetic acid glycer0phosphate glycerophosphoryl-B-methylcholine glycerophosphorylcholine glycerOphosphorylethanolamine glycerophosphorylglycerol glycerophosphorylinositol glycerophosphorylserine R-a—glycerophosphoryl derivative of X phosphatidylcholine phosphatidylethanolamine phosphatidylglycerol phosphatidylinositol phOSphatidylserine ethanolamine, choline, B-methylcholine, inositol, glycerol, or serine ix INTRODUCTION INTRODUCTION Statement 9£_problem The work reported in this thesis was undertaken as part of an ongoing study of housefly (Musca domestica L.) phospholipid metabolism. It has been established (1 and 2) in our laboratory that housefly larvae catabolize phos- phatidyl-B-methylcholine, phosphatidylcholine, and phos- phatidylethanolamine by microsome bound phospholipase A1 and lysophospholipase to the respective glycerolphosphoryl compounds (GP-X). Glycero-32P-phosphoryl-B-methylcholine (G32PBMC) was found to be hydrolyzed by a 40,000 x g soluble larval enzyme predominately to glycerol-32P- phosphate (G32P). My task became the characterization of the glycerophosphodiesterase responsible for this hydrolysis. The minor glycerOphOSPhatides of the housefly, phosphatidyl-inositol, -serine, and -glycerol, were studied to determine whether they were deacylated by the microsomal phospholipase Al-B system to yield glycero- phosphoryl derivatives. In view of the alternative pathways of catabolism for phosphatidylinositol (PI) in some animals (3), it was of interest to establish the l 2 pathway(s) existing in housefly larvae. Having determined what GP-X compounds are produced by the larvae, I under- took to demonstrate the presence of glycerophOSphodies- terase activity capable of hydrolyzing all of these compounds to reutilizable components. Because of the sub- strate selectivity and particulate nature of the rat kidney enzyme (4), I attempted to characterize the larval activity and determine whether one or several enzymes are involved. Introduction to glycerophosphatide catabolism It is beyond the scope and intent of this thesis to review phospholipid biosynthesis and function, but a general introduction to diacylglycerophosphatide catabo- lism and a comprehensive review of glyceroPhosphodiester- ases will be given to orient the reader. Glycerophospha- tides are well established as being integral parts of virtually all biological membranes and are implicated in a rapidly growing list of biochemical reactions, especially those associated with membrane systems. Some peculiari- ties of phospholipid metabolism in houseflies making them particularly interesting to study will be discussed later. For more extensive background on phospholipids and their metabolism the reader is referred to one of the following reviews from which the subsequent discussion is taken: Ansell and Hawthorne, 196% (5); van Deenen and de Haas, l966 (6); and Rossiter, 1968 (3). The four major hydrolytic activities that can catabolize a l,2-diacyl-g27glycerol-3-ph03phatide are designated phospholipase A, B, C, and D. Phospholipase D is documented only in plants and catalyzes the exchange or cleavage of the moiety (e.g., choline, inositol, etc.) esterified to the phosphate of diacylglycerophosphate. Phospholipase C has been studied mainly in bacteria but has been reported to occur in some mammalian tissues. It hydrolyzes the phosphodiester linkage between the glycerol and phosphate moieties yielding, 1,2-diacylglycerol and phosphorylcholine or other phosphoryl derivative. The major catabolic pathway for diacylglycerophosphatides in animals is the hydrolysis of the two acyl groups, one at a time, to yield monoacylglycerophosphatide and finally GP-X. PhosPholipases Al and A2 are the enzymes catalyzing the hydrolysis of the first fatty acid at the l or 2 position respectively. The resulting 2- or l-acyl-gnf glycero-B-phosphatides are deacylated by phospholipase B, also designated lysophospholipase. One still encounters in the literature the term phospholipase B in reference to activity which removes both fatty acids without the demon- strable accumulation of any lyso derivative. Such activity is generally considered to represent sequential removal of the two fatty acids by one or several enzymes since to date no substantial evidence for the simultaneous cleavage of both fatty acids has been obtained. The phospholipase A and B pathway seems to hydrolyze phOSphatidyl-choline (PC), -ethanolamine (PE), -serine (PS), and -glycerol (PG) as well as cardiolipin in most systems that have been thoroughly studied. Phosphatidyl- inositol (P1) is a major exception to this scheme in many systems and will be discussed later. As indicated above, there are different phospholi- pase A's which can cleave either the l or 2 fatty acyl group. These enzymes are generally associated with a mem- braneous fraction in animal cells and are difficult to purify. Phospholipase A also occurs in snake and insect venoms as a soluble enzyme and is secreted as a zymogen by the pancreas. Early work with phospholipase A was done using the readily obtained and purified Ca++ dependent snake venom enzyme which was shown in 1963 to hydrolyze exclusively the 2 (sometimes designated 8) fatty acid. In 196“, using substrates labeled with 3H or luC specifi- cally in the l or 2 fatty acid, phospholipase A activity cleaving the l (or a) fatty acid of PC was found in rat liver and other tissues. These tissues also contain the enzyme hydrolyzing the 2 fatty acid since both lyso com- pounds were demonstrated to accumulate in rat liver. Studies (7, 8, and 9) on the subcellular distribution of phospholipases Al and A2 in rat liver have shown that mitochondria contain only A2 and microsomes contain only Al while lysosomes contain both Al and A2. The lysosomal activities (9) occur at acid pH and are distinct from the pH 8 A2 of mitochondria and the neutral to alkaline A1 of microsomes. The lysosomal pH ”.5 activity is substantially inhibited by Ca++ and activated by EDTA as is the microso- mal Al activity at pH 7.4. The A2 activity at pH u.5 may or may not be inhibited by Ca++, but the lysosomal A2 activity at pH 7 is stimulated by Ca++. Mitochondria con- tain small amounts of pH 7 A2 but contain mainly pH 8.8 A2 which is substantially stimulated by Ca++. These results and similar findings in other systems support a more common occurrence of Ca++ independent phospholipase A1 in microsomes and Ca++ dependent A2 in mitochondria. Lyso- somes appear to have the capacity to completely hydrolyze diacylglyceroPhosphatides by the combined action of phos- pholipases Al and A2 since A2 will also hydrolyze the 2- lyso compound resulting from Al activity. Lysophospholi- pase activity may also occur in lysosomes as well as in the other particulate fractions. It has also been re- ported as a soluble enzyme in some tissues. The questions of distinguishing lysophospholipase activity from phospholipase A2 activity which is known to act on the 2-lyso compound and of distinguishing between one and several enzymes where complete deacylation occurs without production of a detectable lyso intermediate are still inadequately answered. Except for soluble venom and pancreatic phospholipase Az's, purification has been nearly impossible due to the membraneous nature of the enzymes. Differences in heat stability and detergent effects have been the main tools used to assay and separate the different proposed activities. Another problem in the study of phospholipases has been getting the hydrophobic lipid substrates into a form that is accessible to the enzyme and into a form that approaches the situation found i§_zizg. Hanahan found in early work that ether enhanced snake venom phOSpholipase A2 activity on PC presumably by loosening the micelle structure to make the substrate more accessible. Detergents are fre- quently used to disperse the substrates. This often stimulates activity on a substrate not hydrolyzed in the absence of detergents. Mixtures of phospholipids also have been found to have a similar stimulatory effect on activity by altering the charge on the lipid micelle. When studying phospholipases in membranes, particularly of intact organelles, endogenous phospholipids may be cleaved at different rates than exogenous lipids due to poor pene- tration of substrate unless detergents or other disruptive measures are employed. As mentioned previously, PI is metabolized in many systems by a pathway other than phospholipase A and B deacylation. This is a phospholipase C generally re- quiring Ca++ and acid pH which is specific for PI. The products are diglyceride and inositolphosphate. See Appendix A for references and discussion of the distri- bution of the two PI catabolic pathways. In a recent report (10), White et al. have shown that guinea pig pancreas contains a microsomal phospholipase Al having high activity with PI. This phospholipase A1 is inhibited by Ca++. A lysophospholipase is also present yielding GPI as the final product. In contrast to this system, Dawson and co-workers (11 and 12) reported that thyroid gland of the pig metabolizes PI by the diglyceride pathway yielding cyclic inositolmonOphosphate. This system is active at acid pH and is stimulated by Ca++. The pancreatic phos- pholipase A1 of White (10) is also active on PC but is more active on PI suggesting involvement in the enhanced PI turnover occurring during secretion. Insect glycerophosphatide metabolism The following discussion is taken largely from a review by Past, 1970 (13) on insect lipids. Phospholipase D activity has been proposed in certain dipterous insects but has not been well established. The exchange of X in diacyl GP-X probably occurs suggesting an activity that is similar to phospholipase D but is unable to cause accumu- lation of phosphatidic acid. A phospholipase A1 not requiring Ca++ has been reported in three species of mos- quitoes. Rao and Subrahmanyam have studied this mosquito phospholipase A1 (1n) and have partially purified a phos- pholipase B activity away from the Al activity (15). The mosquito enzymes are microsomal, have Optimum activity at basic pH, and the A activity is inhibited by Ca**. The 1 housefly has been found by Kumar et a1. (2) to have a very similar phospholipase Al-B system. The mosquito system is reported (14) to be inactive on PI and phosphatidic acid but active on PE, PC, and cardiolipin. The housefly sys- tem was shown (2) to be active on PE, PC, and phospha- tidyl-B-methylcholine. Phospholipase A activity has been reported by Khan and Hodgson (16) in housefly and blowfly mitochondria as well as microsomes. To the extent phospholipid catabolism is known in insects it is similar to that in other animals. However, there are some interesting peculiarities in the total phospholipid story in dipterous insects, particularly houseflies. The predominate phospholipid is PE rather than PC which predominates in most insects and animals. It is known that houseflies cannot synthesize choline but require it as a dietary component for growth. Evidence is accumu- lating in this laboratory that houseflies have a normal one carbon metabolism utilizing Me-lQC-methionine but have a lesion in the system that methylates ethanolamine to form choline. If dietary choline is substituted for by choline analogs such as B-methylocholine, these bases will be incorporated into phospholipids in place of choline. Many of these analogs (17) do not provide for completely normal development and reproduction in houseflies. Findings of this nature suggest that choline plays an im- portant role in normal insect development and functioning. Essentially all the choline present in the insect is in the form of PC making it very likely that the critical role of choline is some structural or functional role of phosphatidylcholine. There may be some step in the metabolism of PC that fails to function adequately when an analog is substituted for choline. The matter of metabolic handling of choline is of particular interest in the insect. When houseflies meta- morphose from larvae to adults, they pass through a pupal stage during which they can ingest and excrete nothing with the exception of some gas exchange. In this closed system the tissues of the larva are broken down exten- sively and the components are reformed into the cells and tissues of the adult. Since choline cannot be synthe- sized, it seems crucial that the organism be able to efficiently reutilize the choline from the PC of the larva to synthesize the new membrane systems in the cells of the adult tissues. Since it is known that there is extensive phospholipid turnover, presumably by phospholipase A and B deacylation, in the late larval stage and in the pupal stage, it seems likely that there would be an enzyme such as glycerophosphodiesterase to release choline from glycer0phosphorylcholine (GPC) for reutilization in PC synthesis. Reacylation of GPC has not been detected and its accumulation has not been observed during PC break- down. LysoPC does accumulate during pupation but not during the turnover in the late larval stage. It may not be a satisfactory candidate for storage of choline and reacylation to PC in newly formed adult tissues and other lyso derivatives do not accumulate. 10 Review of glycerophosphodiesterase literature Hayaishi and Kornberg (18) isolated a strain Serratia plymuthicum capable of growing on glycerOphos- phorycholine (GPC), yeast extract and mineral salts. They partially purified a glycerophosphodiesterase from this bacterium that hydrolyzes GPC to L-a-GP and choline. The enzyme was stable to heat for 5 min. at 50°, had a pH opti- mum of from 8 to 9, and was inhibited 90% by 1 mM Mg++ or Mn++ or 0.1 mM EDTA. EDTA inhibition was not reversed by metal ions. GPC and glycerophosphorylethanolamine (GPE) gave Km's of 1.2 x 10’3M and 2.5 x 10‘3M respectively, and they each showed competitive inhibition of the hydrolysis of the other. In a brief communication, Prasad and Benson (19) reported that the S. plymuthicum enzyme hydro- lyzes glycerOphosphorylserine (GPS), glycerophosphorylino- sitol (GPI), and glycerophosphorylglucose as well as GPC and GPE. GlycerOphosphodiesterase activity has also been studied in the rat. Dawson (20) reported an enzyme in an acetone powder of rat liver that hydrolyzes GPC to GP and choline. GPE is also hydrolyzed and is a competitor of GPC cleavage. The pH optimum is 7.5 and 1 mM EDTA or Zn++ causes complete inhibition. 1 mM Mg++ stimulates activity, but higher concentrations abolish activity. Fowler and de Duve (21) have confirmed Dawson's findings and established that the GPC diesterase is not lysosomal or mitochondrial but primarily soluble. Webster et a1. ll (22) studied the glycerOphosphodiesterase activity of rat brain and other nervous tissues. The dialyzed rat brain homogenate cleaves GPC to choline and GP optimally at pH 9.5 without added divalent cations. EDTA inhibits com- pletely and the inhibition can be overcome with Mg++ or Mn++. Zn++ causes no inhibition or reversal of EDTA inhibition. A rat kidney microsome bound enzyme has been reported by Baldwin and Cornatzer (A). This enzyme hydrolyzes GPC and GPE to L-a-GP and free base. GPI and GPS apparently are not hydrolyzed by this kidney micro- somal preparation. The pH optimum is 9.5 and the Km's for GPE and GPC are 11.5 x 10'3 M and 2.2 x 10‘3M respec- tively. The enzyme was stable to heating for 5 min. at 50° and was not inhibited by either N-ethylmaleimide or iodoacetamide. At concentrations from 1 to 10 mM ethanolamine and its mono-, di-, and tri-N-methylated analogs show inhibition of GPC hydrolysis. Inhibition increased with concentration over this range and increases with degree of methylation. Choline, 10 mM, gives over 60% inhibition. The kidney enzyme shows no divalent cation stimulation even after dialysis. How- ever, 2 mM EDTA completely inhibits activity. It has been reported (23) that EDTA inhibition can be reversed by several cations including Zn++; however, these cations are inhibitory to the enzyme prior to dialysis against EDTA. The enzyme is thought to contain a metal ion, 12 possibly Zn++, which is lost only when a chelator such as EDTA is present. EXPERIMENTAL PROCEDURES AND RESULTS The results of my work on this thesis problem have been drawn up as two manuscripts for journal publication. A reprint of the published manuscript and the final draft of the other manuscript are included in this thesis as Appendices. The reader is referred to these Appendices for the presentation of methodology and experimental results discussed in the following section. Appendix A is a reprint of a journal article, Lipids E, 508 (1971), consisting of work I did independ- ently on the metabolism of phosphatidyl-inositol, -serine, and -g1ycerol. T. Abraham and L. L. Bieber investigated the metabolism of ceramide phosphorylethanolamine. Only my work will be discussed in the following section. The manuscript presented in Appendix B, describing the glycero- phosphodiesterase of housefly larvae, is entirely my work. This work will be submitted for publication. A brief outline of the methods used in Appendices A and B follows. Phospholipase A and B activity were assayed using 32P labeled phosPholipids prepared from housefly larvae reared on 32P-inorganic phosphate. The water soluble products were extracted for paper chromato- graphic identification. The lySOphosphatides and unreacted l3 19 diacylglycerophosphatides were separated by thin layer chromatography. The fractions were scraped from the plate and quantitated by counting 32F Cerenkov radiation. For assaying glycerophosphodiesterase activity G32P-X sub- strates were prepared by deacylation of 32P-phOSpholipids with methanolic KOH. The products of the incubations were separated by anion exchange chromatography. The 32P radioactivity was quantitated for the unreacted substrate fraction and the glycerolphosphate product fraction. DISCUSSION OF RESULTS DISCUSSION OF RESULTS Diacylglycerophosphatide catabolism in_housef1y larvae Appendix A presents data indicating that phospha- tidylglycerol (PG) (Appendix A, Figure l), phosphatidyl- serine (PS) (Appendix A, Figure 2), and phosphatidylino- sitol (PI) (Appendix A, Figure 3) are metabolized to monoacylglycerophosphatides and to GP-X derivatives by larval microsomes. It had been established (2) in our laboratory that phosphatidyl-ethanolamine, -choline, and -B-methylcholine are metabolized by larval microsomes to yield the lysoglycerophosphatides and GP-X compounds. It was also shown that the lysoglycerophosphatides were 2- acyl-glycerophosphatides and that the lyso compounds are precursors of the GP-X derivatives. It was not verified that the lysoPI, lysoPG, and lysoPS are 2-monoacy1glycero- phosphatides. It was shown that lysoPI is a precursor of GPI by following the course of PI breakdown with time (Appendix A, Figure B). The catabolism of PI in housefly larvae was found to be exclusively by phospholipases A and B to yield GPI. No significant amounts of breakdown by other pathways were detected in subcellular fractions other than microsomes at acidic or basic pH. 15 16 Properties of housefly larval glycerophosphodiesterase Appendix B presents results indicating the pres- ence of a glycerophosphodiesterase in housefly larvae that is not bound to any subcelluar particle (Appendix B, Table I). This hydrolytic activity is stable to heating at 50° for 15 min (Appendix B, Table II). The optimum pH for hydrolysis of GPC and GPE is 7.2-7.” (Appendix B, Figure 1). As indicated in Appendix B, Tables II and III and Figure 2, an ammonium sulfate fractionated enzyme prepa- ration has little activity without added divalent cations and attains optimal activity with u-io mM Mg++. EDTA completely inhibits the enzyme. This inhibition is re- versed by adding Mg++ during incubation. There is no significant inhibition of GPC or GPE hydrolysis by Mg++ up to 20 mM. Co++ is nearly as stimulatory as Mg++. Mn++ and Ni++ showed only half as much stimulation and other divalent cations showed none. Appendix B, Table VI shows that the enzyme is not inhibited by agents reactive with sulfhydryl groups and is not highly sensitive to protec- 1ytic activity. The major difference between the above findings and results of similar studies on the bacterial and rat enzymes is that the larval enzyme requires added Mg++ for activity and is not inhibited by high Mg++ con- centrations. Appendix B, Table IV shows that the larval enzyme cleaves GPC, GPE, GPS, GPI, and GPG to yield 90% of the phosphate released as Z-d-GP (ggfglycero-B-phosphate). 17 Stoichiometric amounts of serine and inositol are also re- leased from GPS and GPI respectively. One enzyme apparent- ly hydrolyzes the above substrates and GPBMC. Appendix B, Figure A shows that when pairs of substrates are incubated together they are both hydrolyzed but at rates lower than when incubated separately. The total hydrolysis of both substrates does not exceed the hydrolysis of the most active substrate incubated alone. The two cases (Appendix B, Figure A, A and B) where combined hydrolysis slightly exceeds maximum hydrolysis of either single substrate may be accounted for by low rates for cleavage of G32PE (A) and GPC (B) when incubated alone. This may be due to some apparent inhibitor in these particular substrate prepara- tions that does not affect the other substrate in coincu- bations. GPC, which is hydrolyzed 70% faster than any of the other substrates (90 y§_5015 n moles/min/mg protein), competes much more favorably than GPE or GPI when coincu- bated with either of these substrates. Appendix B, Table V indicates that there is no extensive inhibition of GPC hydrolysis by any of the products of GP-X cleavage. In particular, choline shows no inhibition in sharp contrast to its 60% inhibition (at similar concentrations) of GPC cleavage by the rat kidney enzyme (M). Appendix B, Fig- nM for ure 3 gives an apparent Km determination of 2 x 10' GPC and 2 x 10'3M for GPI, GPE and GPBMC have Km's near that of GPC but Km's for GPS and GPG appear to be at least as high as the Km for GPI. 18 The findings outlined in the above paragraph suggest that housefly larvae have a single glycerophosPho- diesterase enzyme with broad substrate Specificity. The synthetic phosphodiesterase substrate bis-p-nitrophenyl- phOSphate does not inhibit GPC cleavage but is cleaved by other phosphodiesterases in the preparation. Appendix B, Table VIII shows that GPC cleavage is virtually irrevers- ible starting with choline and l-a-GP and that l-a-G32P cannot be exchanged into GPC. Since Me-luc-choline does exchange into GPC or GPI, it is suggested that the l-a-GP moiety of GP-X is firmly bound to the enzyme. After hydrolysis, the base moiety (e.g., choline or inositol) is free to exchange. The rate of release of the GP moiety is less than the rate at which the base can exchange and re- formed GP-X can be released. The failure of E-a-GP to exchange suggests that the formation of an enzyme-GP inter- mediate from free enzyme and GP is very unfavorable. The exchange of Me-luC-choline into GPI and GPC is additional support for the proposed single enzyme active on all GP-X compounds. CON CLUS IONS CONCLUSIONS From the results presented in this thesis and the available literature, the following conclusions are drawn: The microsomal phospholipase Al-B system of housefly larvae deacylates PI, PS, and PG, as well as PC, PE, and phosphatidyl-B-methylcholine (2) to the respective GP-X derivatives (see Figure 1). This system differs from the mosquito microsomal system in being active with PI. The catabolism of PI by the microsomal phospholipase Al-B system is very similar to the PI metabolizing pancreatic microsomal system of White et a1. (10). It remains to be established what the fate of GPI is in the pancreas. It may be metabolized by a glycerophosphodiesterase activity located in some fraction other than the microsomes. This would be in contrast to the microsomal enzyme found in rat tissues (n). It is possible that houseflies contain minor amounts of a specialized PI metabolic pathway such as the diglyceride-cyclic inositolphosphate reaction of thyroid gland (12). Any such activity could have been masked by the activity observed which is probably respon- sible for the extensive turnover of phospholipids in the late larval stage. 19 20 t {catty OH and 9 l .0 Of—O-X —I5-o-X 0" 0’ fatty % 2 acid f OH H V'— H X'OH 4" 9 3 9 -P-O' .-o-X .. 0" X-OH ' CHOLI NE, ETHANOLAMI NE, B-METHYLCHOLI NE, INOSITOL, SERINE or GLYCEROL Figure 1. Diacylglycerophosphatide catabolic pathway in housefly larvae. 21 In considering the extensive turnover of phOSpho- lipids occurring in late larval stage and particularly during pupation, it would be interesting to know what phos- pholipases are present in housefly lysosomes. If there is a system of phOSpholipases in housefly lysosomes similar to that in rat lysosomes (9 and 21), this system is probably very active during pupation when other lysosomal activity is causing extensive histolysis prior to formation of the adult fly. The soluble glycerophosphodiesterase reported on in Appendix B may be the only such activity functioning during pupation. There is no similar enzyme found in lyso- somes from rat liver (21) and I found none in larval lyso- some fractions. If this enzyme is functional during the histolysis of early pupation, as preliminary results suggest, a single enzyme active with many substrates and stable toward proteolysis would seem to be of physiologi- cal importance. GPC was found to be the best substrate of those tested with the glycer0phosphodiesterase. It was cleaved at a higher rate, had a lower Km, and competed more favorably for cleavage when incubated with another substrate. These observations may be of significance when one considers the necessity of efficient reutilization of choline in the formation of PC for adult tissues in late pupation. Figure 1 summarizes the findings of my work and previous work (2) in this laboratory on 22 diacylglycerophosphatide metabolism in housefly larvae. Reaction 1 is reversible in the sense that 2-1yso com- pounds can be reacylated, as can l-lyso compounds, to yield diacylglycerophosPhatides. LIST OF REFERENCES ll. 12. 13. 1”. LIST OF REFERENCES L. L. Bieber, L. G. Sellers, and S. S. Kumar. J., Biol. Chem. 2M”, 630 (1969). S. S. Kumar, R. H. Millay, and L. L. Bieber, Biochem- istry 9, 75% (1970). R. J. Rossiter, Metabolic Pathways II 3rd Edition, 69 (1968), D. M. Greenberg (Editor), AEEdemic Press, New York. Jerry J. Baldwin and W. E. Cornatzer, Biochim. Bio- phys. Acta 16%, 195 (1968). G. B. Ansell and J. N. Hawthorne, PhOSpholipids, B. B. A. Library vol. 3, 152 (196%), Elsevier, Amsterdam. L. L. M. van Deenen and G. H. de Haas, Ann. Rev. Bio- chem. 35, 157 (1966), P. D. Boyer (Editor), Annual Reviews Inc., Palo Alto. G. L. Scherphof et al., Biochim. BiOphys. Acta 125, #09 (1966). M. Waite and L. L. M. van Deenen, Biochim. Biophys. Acta 137, M98 (1967). M. Waite et al., J. Lipid Res. 10, All (1969). D. A. White, D. J. Founder, and J. N. Hawthorne, Bio- chim. Biophys. Acta 2H2, 99 (1971). R. M. C. Dawson, N. Freinkel, F. B. Jungalwala, and N. Clarke, Biochem. J. 122, 605 (1971). F. B. Jungalwala, N. Freinkel, and R. M. C. Dawson, Biochem. J. 123, 19 (1971). P. G. Fast, Progress in the Chemistry of Fats and other Lipids II part 2, 181 (1970), R. T. Holman (Editor), Pergamom Press. R. H. Rao and D. Subrahmanyam, Arch. Biochem. BiOphys. luo, nus (1970). 23 15. 16. 17. 18. 19. 20. 21. 22. 23. 2a R. H. Rao and D. Subrahmanyam, J. Lipid Res. 10, 636 (1970). _— M. A. Q. Kahn and E. Hodgson, Comp. Biochem. Physiol. 33, 899 (1967). R. G. Bridges and J. Ricketts, J. Insect Physiol. 16, 579 (1970). Osamu Hayaishi and Arthur Kornberg, J. Biol. Chem. 206, an? (1953). R. Prasad and A. A. Benson, Biochim. BiOphys. Acta 187, 269 (1969). R. M. C. Dawson, Biochem. J. 62, 689 (1956). S. Fowler and C. de Duve, J. Biol. Chem. 24%, H71 (1969). G. R. Webster, Elizabeth A. Marples, and R. H. S. Thompson, Biochem. J. 65, 37k (1957). Jerry J. Baldwin, Phyllis Lanes, and W. E. Cornatzer, Arch. Biochem. Biophys, 133, 22” (1969). APPENDICES APPENDIX A Reprinted from lems, Vol. 6, N0. 7, Pages: 5118511; (1971) Metabolism of Ceramide Phosphorylethanolamine, Phosphatidylinositol, Phosphatidylserine and Phosphatidylglycerol by Housefly Larvae1 G.R. HI LDENBRANDT, T. ABRAHAM and L.L. BIEBER, Department of Biochemistry, Michigan State University, East Lansing, Michigan 48823 ABSTRACT Microso me preparations (40,000-90,000 g sediment) from Musca domestica, housefly, larvae convert exo- genous 32P—labeled phosphatidylinositol, phosphatidylserine and phosphatidylgly— cerol to the respective lysoglycerophos- phatides and, ultimately, to the glycero- phosphoryl derivatives. These data, com- bined with previous results, demonstrate that housefly larvae can convert their nor- mal diacylglycerophosphatides to the res- pective glycerophosphoryl derivatives. Experiments utilizing exogenous 3H- labeled, 32P—labeled and MC-labeled cera- mide phosphorylethanolamine demon- strate that particulate preparations from housefly larvae convert ceramide phos- phorylethanolamine to ceramide, phos- phorylethanolamine, sphingosine and fat- ty acid. The presence of ceramide phos- phorylethanolamine phosphohydrolase and ceramidase activity in housefly larvae is consistent with the conclusion that ceramide phosphorylethanolamine is metabolized to ceramide and phosphoryl- ethanolamine and the ceramide is then hydrolyzed to sphingosine and fatty acid. Thus, metabolism of ceramide phos- phorylethanolamine by these insects is analogous to the metabolism of sphingo- myelin by mammalian systems. INTRODUCTION Phosphatidylcholine and phosphatidyl- ethanolamine are the principal phospholipids of many insects (l). Housefly (2) and blowfly (3) larvae also contain minor phospholipids such as phosphatidylserine and phosphatidylinositol as well as ceramide phosphorylethanolamine-type sphingolipids (2,4-6). Particulate preparations, containing microsomes, from larvae of the housefly (Musca domestica) (7,8), the blowfly (Phormia regina) (7) and the mosquito (Papiens fatigans) (9,10), contain phospholipase A1 and 1Paper No. 5250 from the Michigan Experiment Station. A2 activity towards phosphatidylcholine and phosphatidylethanolamine. Little is known con- cerning the metabolism of the minor lipids of insects. We will demonstrate that particulate preparations containing microsomes from housefly larvae deacylate phosphatidylserine, phosphatidylinositol and phosphatidylglycerol to the respective lysophosphatides and, ulti- mately, to the water soluble glycerophosphoryl derivatives. Ceramide phosphorylethanolamine is cleaved to ceramide and phosphorylethano- lamine, and the ceramide is cleaved to fatty acids and sphingosines. METHoos Isolation of Phospholipids and ‘Coramide Derivatives 3 2 P- labeled phosphatidylinositol, phos- phatidylglycerol, phosphatidylserine and cera- mide phosphorylethanolamine were isolated from housefly larvae that had been reared on a 32Pi-containing diet as described previously (6,8). The lipid extract containing phosphati- dylinositol, phosphatidylserine, phosphatidyl- glycerol and other phospholipids was chromato- graphed on silicic acid to remove neutral lipids and lecithin as described elsewhere (8,1 l). The remaining phospholipid classes were resolved by column chromatography using DEAE cellulose essentially as described by Rouser et al. (12). Column eluates containing acetic acid or formic acid were dripped directly into NH4 HCO3 solu- tions to neutralize the acid. These preparations were then partitioned between water and chlo- roform and the chloroform phase evaporated to dryness. The dried lipid samples were dissolved in warm tertiary butyl alcohol, lyophilized and stored dry at -20 C. After chromatography on DEAE cellulose, phosphatidylglycerol and phosphatidylinositol were separated by prepara- tive thin layer chromatography (TLC) using Brinkman preparative plates (Silica Gel F-254) and the solvent system, chloroform-methanol- H20 (65:35:4). Occasionally, the other phos- pholipids were further purified by preparative TLC using the solvent system, chloroform- methanol-H20 (6S:35:4) or chloroform- methanol-conc NH4OH (65:3SI4). The phos- pholipids were at least 95% pure before use, as determined by monitoring the 32F distribution 508 HOUSEFLY PHOSPHOLIPID METABOLISM 509 with a Packard Model 7201 radiochromatogram scanner after TLC in one or more solvent sys- tems. In experiments (unpublished data), G.R. Hildenbrandt has shown that the phosphorus- inositol ratio and the phosphorus-serine ratio of the glycerophosphoryl derivatives prepared from these preparations of phosphatidylinositol and phosphatidylserine are 1.0 :l: 0.1. 32F in column eluates was monitored with a recording Geiger Muller apparatus. Tritiated ceramide phosphorylethanolamine was prepared according to the procedure des- cribed above for preparation of 32P-labeled li- pids except that larvae were reared on a diet containing 1 curie of tritiated water. 14013- beled ceramide phosphorylethanolamine was prepared similarly except that larvae were reared on diets containing uniformly labeled L-serine from New England Nuclear. Ceramide was prepared from bovine heart sphingomyelin by treatment with 2-5 mg of phospholipase C from Worthington, Closm'dium Welchii, as fol- lows. The sphingomyelin (400 mg) in chloro- form was adsorbed onto 1 g of Celite using a rotary evaporator and the reaction was run overnight with stirring at room temperature in 20 ml of 0.025 M Hepes (N-2-hydroxypipera- zine-N'-2-ethane sulfuric acid), pH 7.3, con- taining 3 mM CaClz. The chloroform soluble- products were resolved on silicic acid (8). Ceramide was detected on thin layer chromato- grams utilizing a Chloroxbenzidine spray reagent (l3). Sphingosine standards were pre- pared by hydrolyzing bovine brain sphin- gomyelin in l N methanolic HCl for 18 hr at 105 C. Phospholipids were deacylated by add- ing 1.0 ml of 0.5 N KOH in 95% methanol to 2 ml of phospholipid in chloroform. After stirring for 10 min at room temperature, excess Dowex-SO resin, H+ form, was added and the samples were partitioned between equal vol- umes of chloroform and water. The resin was removed and the aqueous phase neutralized immediately with NaOH. Glycerophosphate and Pi were removed by column chromato- graphy on Dowex-l, formate form and the water soluble glycerophosphoryl derivatives were then chromatographed on Dowex-SO H+ as described below. Enzyme Preparations and Assays Microsomes were prepared from housefly larvae and phospholipase assayed as described by Kumar et al. (8). 32P-labeled diacylglycero- phosphatides were separated from the respec- tive monoacylglycerophosphatides by TLC as described in the figure legends. The water solu- ble portion of the reaction products, which contained the glycerophosphoryl derivatives, 863 FIG. 1. Tracings of the radiochromatogram scans of chromatograms of the chloroform soluble and water soluble extracts obtained from larval micro- somes incubated with 32P—Iabeled phosphatidylgly- cerol. Larval microsomes 0.6 mg protein, were incu- bated with 1 urnole of 32P-labeled phosphatidylgly- cerol (44,000 CPM) for 1 hr at 31 C in 1 ml of 50 mM Tris HCl, pH 7.2. Assays were performed as described in Methods. A contained 2.5 mM lauryl sulfate. Lauryl sulfate and microsomes were not added in B. A and B are radiochromatogram scans of the chloroform phases that were chromatographed on Silica Gel F-254 Brinkrmn thin layer plates in the solvent system, chloroform-methanol-H20 (65:35:4). 73% of total counts were in the chloroform phase of A and 100% of total counts were in the chloroform phase of B. C is a radiochromatogram seen of the water soluble frac- tion from A (27% of total counts) that were chroma- tographed on paper in the solvent system, methanol- conc. NH4OH-H20 (12:2:3). OR, origin; SF, solvent front; PG, phosphatidylglycerol; GP, a-glycerophoa- phate; Pi, inorganic phosphate. was passed through columns of Dowex-SO resin, H+ form, 100-200 mesh, 0.8 x 10 cm to remove cations. The column eluates (containing the glycerophosphorylinositol, serine, or glycerol) were lyophilized, dissolved in a small quantity of water (<0.5 ml), neutralized, and then chromatographed on paper as described in the legends of Figures 1-3. 32P in water, in organic solvents and on silicic acid was quantitated uti- lizing Cerenkov radiation as described by Havi- land and Bieber (l4). Ceramide phosphoryl- ethanolamine phosphohydrolase was prepared and measured as described in the legend of Figure 5. 32Pi was determined by the method of Lind- berg and Ernster (15). Pi was precipitated from LIPIDS, VOL. 6, NO. 7 B'MIN A / 5 < U (I) _J _l E B L‘— .4 O .. P PS OR 80- FF g; 3 ME 5° :- 3 05's 4o~ ' ~ I. ARKER I o FIG. 2. Tracings of the radiochromatogram scans of chromatograms of the chloroform soluble and water soluble extracts obtained from larval micro- somes incubated with 32P—labeled phosphatidylserine. Microsomes, 2.9 mg protein, were incubated for 8 min in 2.5 ml of a solution containing 0.05 M imidazole buffer, pH 7.4, 2.5 mM lauryl sulfate, 1 mM HgCl , and 2.75 umoles of phosphatidylserine (64,400 CPMI. The reaction was terminated and assayed as described in the Methods. The aqueous phase contained 7,762 CPM. SF, solvent front; OR, origin. For A, the chloroform soluble extract was chromatographed as described for Figures 1A and B. LPS, lysophosphati- dylserine; PS, phosphatidylserine. For B, the water soluble fraction was chromatographed on paper in the solvent system, picric acid-tertiary butyl alcohol-water (4 g:80:20). GP, a-glycerophosphate; GPS, glycero- phosphorylserine; PS, phosphorylserine; Pi, inorganic phosphate. aqueous solutions as the triethylamine phos- phomolybdate complex by the method of Sugino and Miyoshi (l6). Serine was quanti- tated using an automated amino acid analyzer as described by D.C. Robertson, H.B. Brock- man, W.I. Wood and W.A. Wood (personal com- munication). Protein was determined by the method of Ipwry et a1. (17). lnositol was quantitated by gas chromato- graphy of ,the TMSi derivatives as described by Wells et al. (18). R ESULTS Microsomel Deacylation of Phosphatidylinositol, Phosphatidylaorine, and Phosphatidylglyoorol When microsomes from housefly larvae were incubated with 32P-labeled phosphatidylserine and phosphatidylglycerol, TLC of the chloro- form soluble fraction demonstrated the pre- sence of the respective 32P-labeled lysophos- pholipids and paper chromatography of the LIPIDS, VOL. 6, NO. ‘7 G.R. HILDENBRANDT, T. ABRAHAM AND L.L. BIEBER so- , _. % OF FULL SCALE FIG. 3. Tracings of radiochromato am scans of chromatograms of the chloroform solu 1e and water soluble extracts obtained from larval microsomes incubated with 32P—labeled phosphatidylinositol. Incu- bations and assays were identical to those described for Figure 1 including quantity and specific radioacti- vity of substrate. For A, the lipid extract was chromatographed as in Figures 1A and B. The dashed curve (— —- — —) is a tracing of the radiochromatogram scan obtained with the lipid extract (100% of total counts) from an incubation without microsomes, and the solid curve (-——) is a tracing of the radiochroma- togram scan of the lipid extract (66% of total counts) from microsomes that were incubated 60 min with 2.5 mM lauryl sulfate. Pl, phosphatidylinositol; 0.7 mg protein per sample was used. For B, an aliquot of the water soluble portion (53% of total counts) of a 90 min incubation, as described for Part A, was chroma- tographed on paper in the solvent system described for Figure 1C. B is a tracing of the radiochromatogram scan. GPI, glycerophosphorylinositol; GP, a-glycero- phosphate; Pl, phosphorylinositol; and Pi, inorganic phosphate. In A and B, SF, solvent front; 0R, origin. water soluble fractions demonstrated produc- tion of glycerophosphoryl derivatives (Fig 1 and 2). The thin layer chromatograms for Figures IA and B were run in the solvent at different times. Thus, the difference in the Rf of phosphatidylglycerol in Fig. 1A and B is probably due to slightly different solvent com- position or different activation of the plates. Some inorganic phosphate was present in the water soluble fraction when phosphatidylserine was the substrate (Fig. 28). The inorganic phos- phate was probably produced by the phos- phatase(s) which are present in the enzyme preparations. Glycerophosphate, a phosphatase substrate, could be produced enzymatically HOUSEFLY PHOSPHOLIPID METABOLISM 51 ] B s 1 q SOLUBLE .1. g i o g 1 ._I_ 11.1 -YILL AIflljj l 2 3 4 HOURS FIG. 4. Time course for deacylation of 32P—Iabeled phosphatidylinositol by housefly larvae microsomes. Each sample of 1 ml, taken from an 8 ml incubation mixture at the indicated times, contained: 1 umole 32P-labeled phosphatidylinositol at zero time; 50 pmoles Tris Cl’, pH 8.0 and 2.5 umoles lauryl sulfate. Incubation was at 31 C, and assays were performed as described in Methods. PI, phosphatidylinositol, H; LPI, lysophosphatidylinositol, A—A; water soluble, 32P—labeled glycerophosphorylinositol formed, x—-x. The protein concentration was 0.7 mg/ ml. from glycerophosphorylserine by traces of phosphodiesterase. A large amount of this phos- phodiesterase activity is present in the 90,000 g supernatant fluid from larvae and occasionally small amounts remain with the microsome preparations. Attempts to detect production of lysophosphatidylserine were unsuccessful until lauryl sulfate and HgC12 were added to the reaction mixture. It was shown previously (8) that lauryl sulfate greatly stimulates monoacyl- glycerophosphatide production from the diacyl- glycerophosphatides. Apparently, in the ab- sence of lauryl sulfate and HgClz, housefly microsomes deacylate lysophosphatidylserine at a much greater rate than phosphatidylserine, preventing accumulation of the lyso-inter- mediate. When 32P-labeled phosphatidylino- sitol was the substrate, after 1 hr, microsome preparations had converted much of the 32P to a phospholipid which had the Rf on thin layer chromatograms of lysophosphatidylinositol (Fig. 3A). A water soluble compound was also formed. It had an Rf corresponding, to that of glycerophosphorylinositol rather than glycero- phosphate, inorganic phosphate or phosphoryli- inositol (Fig. 38). Since phosphatidylinositol is cleaved to phosphorylinositol and diglyceride in some systems, its cleavage by the insect system was further investigated. The microsomal system from housefly larvae is capable of deacylating most of the added phosphatidylinositol, as shown by the time course in Figure 4. At the end of the experiment, 4 hr, the concentration of exogenous phosphatidylinositol had de- creased from 1 mM to 0.04 mM. The amount of lOO ()1 O WATER - SOLUBLE 3 2p CPMX 10'2 MINUTES FIG. 5. pH Optimum and Time course for hydro- lysis of 32P-1abeled ceramide phosphoryl-ethanola- mine. In A, 0.2 ml of the 10,000-40,000 g fraction (06 mg protein) in 0.1 M KC1 was added to 0.8 ml of 0.05 M buffer, and 0.1 ml of 32P—labeled ceramide phos- phorylethanolamine in 5% Triton X-100. The final concentration of Triton X-100 was 1% and ceramide phos horylethanolamine was 0.25 mM = 32,140 CPM of 3 P. The samples were incubated at 31 C for 30 min, and the reaction was terminated by adding 6 ml of chloroform-methanol (1:2). After mixing or 15 min, 5 ml of chloroform and 5 ml of 0.02 M MgC12 were added. The samples were mixed and the layers separated by centrifugation. The water soluble frac- tion was counted and used as a measure of the extent of reaction. The water soluble CPM are plotted in A and B. The buffers were formate at pH 3.1 and 4.1- acetate at pH 4.5 and 5.5; citrate at pH 5.0; imidazole at pH 6.0, 6.5 and 7.0; Tris Cl‘ at pH 7.5, 8.5 and 9.5; phosphate at pH 8.0, and glycine at pH 9.0 and 10.5. In B, 1 ml of the 10,000-40,000 g pellet in 0.05 M KC1 was added to 4 ml of 0.05 M Tris HCl, pH 9.5, buffer containing sufficient Triton X-100 and ceramide phos- phorylethanolamine to make a final concentration of 1% Triton X-100 and 0.3 mM 32P—labeled ceramide phosphorylethanolamine. The sample was incubated at 31 C. 1.0 ml Aliquots were taken at the times indicated and assayed as described for part A. Each assay contained 0.6 mg protein. The l0,000-40,000 g fraction was prepared by homogenizing larvae in 0.15 M Tris, pH 7.5 (5ml/g larvae) in a micro Waring blendor for 30 sec at maximum speed. The homoge- nate was squeezed through several layers of cheese- cloth and a 10,000-40,000 g fraction was collected and suspended in 0.1 M KCl or the appropriate buffer. lysophosphatidylinositol increased rapidly and then decreased with time, while the glycero- phosphorylinositol only increased with time, in- dicating a product-precursor relationship. Alternate pathways of phosphatidylinositol LIPIDS, VOL. 6. NO. ‘7 m.— “- -— -_ _ -—_.-— ——____ —— _—___ —— ——__"—_——— __‘ —— __—’\———_ H .—-——-_J— —‘ _—‘ ‘0 512 G.R. HILDENBRANDT, T. ABRAHAM AND L.L. BIEBER FIG. 6. Identification of 32P-labeled phosphorylethanolamine as a product of the cleavage of ceramide phos horylethanolamine. To a 10,000-40,000 g fraction, 0.6 mg protein in 0.5 ml of 0.05 M KCl was ad ed 0.5 ml of 0.5 M Tris CI, pH 9.0 and 0.2 ml of 35P-labeled ceramide phosphorylethanolamine (0.43 umole) in 1% Triton X-100. The sample was incubated at 31 C for 30 min. The reaction was terminated by adding 3 ml of Ii uid phenol; see reference 39 for description of this method. The phenol was washed two times withg ml water, and the combined aqueous phases were applied to a Dowex-l, HCO ' column. The column was eluted with water, then 0.25 M tnethylammonium bicarbonate, pH .5, followed by l M triethylammonium bicarbonate, pH 7.5. Two peaks were obtained as described in the text. The first radioactive peak was chromatographed on paper using the solvent system, methanol-cone NH40H-H20 (12:2:3). Figure 6 is a tracing of a portion of a radiochromatogram scan of the paper chromatogram. One radioactive peak was detected; it had an Rf identical to phosphorylethanolamine. PEA, phosphorylethanolamine; Pi, inorganic phosphate. metabolism were not detected in housefly lar- vae. Attempts to detect production of phos- phorylinositol using whole homogenates as well as particulate fractions containing microsomes plus lysosomes and mitochondria were unsuc- cessful. Assays were performed at pH 5.0 and 8.0. Such results indicate that cleavage of phos- phatidylinositol to diglyceride and phosphoryl- inositol is not a major pathway of metabolism in housefly larvae. It should be noted that de- acylation of phosphatidylinositol was detected at pH 5.0 with particulate preparations that should contain lysosomes. The water soluble fraction from the pH 5.0 incubation contained inorganic 32P and glycerophosphorylinositol- 32P. The precursor of ”H was not investigated and could have been phosphorylinositol. In these experiments, the 32P-labeled water soluble products were not rigorously charac- terized; however, paper and column chromato- graphy demonstrated that the principal water soluble 32P was not inorganic phosphate, gly- cerophosphate, phosphorylinositol or phos- phorylserine. The water soluble 32F had paper chromatographic properties expected for the glycerophosphoryl derivatives of inositol, serine and glycerol. Authentic glycerophosphoryl LIPIDS, VOL. 6, NO. 7 derivatives of inositol, serine and glycerol were prepared from the respective diacylglycerophos- phatides by mild alkaline hydrolysis. Portions of these derivatives were used as standards for paper chromatography and as substrates for the phosphodiesterase that cleaves these com- pounds to glycerophosphate and free hydroxyl compound. Claavap of Ceramide Phosphorylethanolamine by Particulate Preparations Obtained From Housefly Larvae When whole homogenates as well as the 0-800 g, BOO-9,000 g, 9,00040,000 g, 40,000-90,000 g and the 90,000 g soluble frac- tions from housefly larvae were incubated with 32P—labeled ceramide phosphorylethanolamine at pH 7.5 and 9.5, water soluble 32P was re- leased. These experiments demonstrated that these fractions cleaved cerarnide phosphoryl- ethanolamine. The soluble fraction and the 0-800 g fraction contained the lowest specific activities. The ratio of the specific activities of the 800-9,000 g, 9,000-40,000 g and 40,000-90,000 g fractions varied from prepara- tion to preparation. The specific activity dif- ferences appeared to be related to the stage of HOUSEFLY PHOSPHOLIPID METABOLISM TABLE I Total Distribution of Fraction CPM CHCl3-soluble material, % Free fatty acid 13,320 10.2 Ceramide 2,500 1.9 Sphingosines 18,600 14.3 Ceramide phosphorylethanolamine 95,520 73.5 8Amounts of 0.5 mg of 3H-ceramide phosphorylethanolamine plus 4 mg of ceramide were dissolved in 1 ml of a solution of 1% Triton X-100 and 0.05 M glycine, pH 9.0. The 8,000-40,000 g particulate fraction in 2 ml of 0.05 M glycine buffer, pH 9.5, from 21 g of larvae was added to the substrate solution and incubated at 31 C for 15 min. The reaction was terminated by adding 8 ml of chloroform-methanol, 1:2, and stirring for 15 min. Then 5 ml of 0.02 M MgC12 and 5 ml of chloroform were added. The chloroform layer was washed three times with 5 ml of 0.02 M MgClz. The aqueous washes were combined and counted. The aqueous layer contained 8,430 CPM. The chloroform fraction was evaporated to dryness, and the fatty acids, sphingosine, ceramide and ceramide phosphorylethanola- mine separated by TLC as follows. The chloroform soluble material was streaked onto a Silica Gel F-254 (Brinkman) plate and developed in chloroform-glacial acetic acid (90:10). The plate was dried at approximately 50 C and placed in iodine vapors. The zones with Rf’s of ceramide and oleic acid were marked. The plate was then placed into the solvent system, chloroform-methanol cone NH4OH (65:35:4) and developed until the solvent front reached the ceramide zone. This solvent system separated ceramide phosphorylethanolamine from the sphingosines. The zones having Rf’s of sphingosines. fatty acids and ceramide were scraped from the plates. The scrapings were put into small columns and eluted with methanol. Aliquots of the methanol solution were transferred to scintillation vials and the samples counted for 3H. The remainder of the methanol was evaporated to dryness and the residue was dissolved in chloroform and applied to thin layer plates. The plates were developed in the solvent systems A, chloroformglacial acetic acid (96:4), and B. chloroformomethanol-conc, NH4OH (6513524). For each, the zones with Rf’s of sphingo- sine, fatty acid and ceramide were scraped from plates, eluted and counted as described above. The rechromatography was essential because the ceramide and sphingosine overlapped slightly on the initial chromatography. In solvent A, the Rf‘s of ceramide phosphorylethanolamine, sphingosine, ceramide and free fatty acids were 0.0, < 0.02, 0.26, and 0.59, respectively. In solvent B. the Rf’s of ceramide phosphorylethanolamine and sphingosine were 0.42 and 0.88, respectively. The tritiated samples were dissolved in 1 ml of methanol or 1 ml of water. Then 0.4 ml of Triton X-100 and 10 ml of toluene scintillation solution (4 g PPO and 100 mg POPOP per liter toluene) were added and the samples were counted. 13.9 x 104 DPM were used in the experiment. 513 larval development and were not investigated further. Since all three particulate fractions contained ceramide phosphorylethanolamine phosphohydrolase activity, a l0,000-40,000 g particulate fraction was used for most of the following investigations. In all of the prepara- tions, 32P-labeled phosphorylethanolamine as well as 32Pi were detected when 32P-labeled ceramide phosphorylethanolamine was the sub- strate. pH Optimum and Time Course Four pH optimum curves were run on l0,000-40,000 g preparations. In all four deter- minations, the greatest activity was obtained near pH 9.0; however, in two of the runs, con- siderably more activity at pH 4-5 and at 7-8 was obtained than is shown in Figure 5A. At pH 9.5, hydrolysis of ceramide phos- phorylethanolamine was nearly linear for over an hour, as shown in Figure SB. Identification of Phosphorylathanolamine as a Reaction Product ‘ When 32P-labeled ceramide phosphoryl- ethanolamine was incubated with the 0-800 3, 800-9,000 g, 9,000-40,000 g, 40,000-90,000 g, or the 90,000 g supernatant fraction, water soluble 32P was released. From 10% to 45% of the water soluble radioactivity from each frac- tion was inorganic phosphate as determined by partitioning the phosphomolybdate complex between sulfuric acid and isotubanol-benzene (15). When the water soluble fraction from a 10,000-40,000 g fraction was exchanged onto Dowex-l in the bicarbonate form and the col- umn eluted successively with 0.25 M and l M triethylammonium bicarbonate, pH 7.5, two radioactive peaks wer obtained. The first peak had chromatographic properties on Dowex—l of authentic phosphorylethanolamine. It also had paper chromatographic properties of phos- phorylethanolamine as shown in Figure 6. The second peak obtained from the Dowex-l col- umn was Pi as determined by paper chromato- graphy and by partitioning radioactivity as the molybdate complex into isobutanol-benzene. Inorganic phosphate would be expected in these preparations from the action of phos- phatase on phosphorylethanolamine. Each of LIPIDS, VOL. 6, N0. 7 514 the fractions contained phosphatase activity as determined by release of Pi from glucose 6- phosphate and release of p-nitrophenol from p-nitrophenyl phosphate. When the water-soluble 32P was partitioned between water and chloroform as described by Hirschberg et al. (19), negligible radioactivity was detected in the chloroform layer. Since the partition coefficient in this system for phos- phoryl sphingosines is approximately 0.73 (19), some radioactivity should have partitioned into the organic phase if phosphoryl sphingosines were present. Identification of Ceramides as a Product of Ceramide Phosphorylethanolamine Cleavage by Larvae Particulate Preparations When l4C-labeled ceramide phosphoryl- ethanolamine was incubated with a 10,000-40,000 g particulate fraction, l‘iC-la- beled sphingosine and l‘iC-labeled phosphoryl- ethanolamine were detected; however, ceramide was not detected in the organic soluble frac- tion, nor was l4C detected on thin layer chro- matograms at the Rf corresponding to ceramide standards. 14C was anticipated in ethanolamine and sphingosine because the l‘iC-labeled cera- mide phosphorylethanolamine was isolated from larvae that were reared in the presence of uniformly labeled L-serine. The data mentioned above indicated that either ceramide was not a reaction product or the preparations contained considerable ceramidase activity. To distinguish between these possibilities, 0.5 mg of tritiated ceramide phosphorylethanolamine—both the sphingosine and fatty acid portion of the sub- strate contained tritium— was combined with 4 mg of ceramide and incubated with an 8,000-40,000 g fraction. Carrier ceramide was added to dilute any radioactive ceramide that would be produced. The organic soluble reac- tion products were separated by TLC in three solvent systems, as described in the legend of Table 1. Very weak iodine-absorbing spots were detected in the ceramide regions of the thin layer chromatograms although samples equiva- lent to 1 mg of initial ceramide were applied to the thin layer chromatograms indicating that the particulate preparations contained cerami- dase activity. Compounds with Rf’s of sphingo- sine and fatty acids were detected. Although very little, if any, ceramide was detected by iodine vapors on the thin layer chromatograms, some tritium was detected in the ceramide area with the three solvent systems. used. When the tritiated material in the ceramide region of the thin layer chromatograms was eluted and re- chromatographed, as described in the legend of Table I, the tritium again migrated with cera- LIPIDS, VOL. 6, NO. 7 G.R. HILDENBRANDT, T. ABRAHAM AND L.L. BIEBER mide. Thus, small amounts, approximately 2% (Table I) of the chloroform soluble tritiated material obtained from the reaction media had chromatographic properties of ceramide. Large amounts of sphingosine and fatty acids were produced, presumably from ceramide via the ceramidase activity. The appearance of 2% of the radioactivity of tritiated ceramide phosphorylethanolamine in the ceramide fraction when carrier ceramide was added to the incubation media is signifi- cant, as indicated by the following results. When l‘iC-labeled ceramide phosphorylethano- lamine was incubated with the 10,000-40,000 g particulate fraction in the absence of carrier ceramide, only 0.02% of the initial 14C was detected in ceramide. This amount is 1% of the radioactivity that accumulated in the presence of a large pool of cold ceramide. Ceramide phosphorylethanolamine might be initially deacylated to sphingosine phosphoryl- ethanolamine and fatty acids, but the data do not indicate such cleavage. 32P-labeled material with thin layer chromatographic properties ex- pected for sphingosine phosphorylethanolamine was not detected in the experiments described for Table I. DISCUSSION The data show that a microsome-enriched fraction from housefly larvae converts phos- phatidylinositol, phosphatidylserine and phos- phatidylglycerol to lysophospholipids and to the respective glycerophosphoryl derivatives. A typical product-precursor relationship was ob- served for lysophosphatidylinositol (Fig. 4). Previous investigations (8) demonstrated that phosphatidylcholine, phosphatidylethanola- mine and phosphatidyl-{i-methylcholine are de— acylated similarly by microsome-containing pre- parations. Thus, housefly larvae have the capa- city for completely deacylating their principal and minor diacylglycerophosphatides; cardioli- pin and phosphatidic acid, both occurring in larvae, have not been investigated. Presumably, this deacylation system can also remove the fat- ty acids from the numerous abnormal glycero- phosphatides that are formed by housefly lar- vae (20-24). Formation of glycerOphosphorylinositol from phosphatidylinositol occurs in mammalian systems such as ram seminal fluid (25), ox pan- creas (26), rat liver (27) and rat prostate (29), and in the microorganism, penicillium notarum (26). Phosphatidylinositol is also cleaved by phosphoinositide inositolphosphohydrolase to diglyceride and phosphorylinositol by ram sper- matozoa (25), ox pancreas (26), rat liver (27), HOUSEFLY PHOSPHOLIPID METABOLISM guinea pig intestine (28) and guinea-pig brain (40). Our studies indicate that housefly larvae contain little, if any, phosphatidylinositol phos- phohydrolase activity. When 32P-labeled phos- phatidylinositol was incubated with larval mi- crosomes, no phosphorylinositol was detected in the reaction mixture (Fig. 38). Some gly- cerophosphate was found (Fig. 3B), but this un- doubtedly was caused by a larval phosphodies- terase that cleaves glycerophosphoryl deriva- tives (unpublished results) to glycerophosphate and the hydroxyl compound. Phosphorylinosi- tol was not detected when the soluble or parti- culate fractions were used as the enzyme source. It should be noted that small amounts of ”H were detected in some of the reactions. Such results indicate that third instar M. domes- tica larvae contain little, if any, phospholipase C type activity towards phosphatidylinositol. Ceramide phosphorylethanolamine is a phos- pholipid in several species of flies (46, 30,31) as well as in honey bees (32), scorpions (32), fresh water mollusks (33) and pond snails (34). Our data are consistent with the conclusion that housefly larvae can metabolize ceramide phosphorylethanolamine, as follows: I. Ceramide phosphoryléthanolamine + HOH -> ceramide + phosphorylethanolamine 2. Ceramide + HOH -> fatty acids + sphingo- sines The above-mentioned pathway is supported by the following results: 1. Phosphorylethanolamine was the principal water soluble product when 32P-labeled cera- mide phosphorylethanolamine was the sub- strate. ”H was also detected, but it most likely was cleaved from phosphorylethanolamine by a phosphomonoesterase. 2. Tritiated ceramide was detected in the re- action mixture when tritiated ceramide phos- phorylethanolamine was the substrate. Tritiated sphingosine and fatty acid were also detected, but these products could be caused by the ac- tion of ceramidase on the ceramide. 3. The particulate enzyme preparations con- tained ceramidase activity. The ceramidase acti- vity was greater than the ceramide phosphoryl- ethanolamine phosphohydrolase activity. Thus, ceramide phosphorylethanolamine is metabolized by housefly larvae similar to the metabolism of ceramide phosphorylcholine, sphingomyelin, in mammals. In mammals, sphingomyelin is hydrolyzed to ceramide and phosphorylcholine (35,36) and ceramide is cleaved to fatty acids and sphingosines (37,38). Although phosphatidase C and D activity 515 was not detected with any of the 32P—labeled diacylglycerophospholipids, the metabolism of diacylglycerophosphatides via pathways other than those described above has not been rigo- rously excluded. Similarly, pathways involving conversion of ceramide phosphorylethanola- mine to sphingosine-phosphorylethanolamine or to phosphorylceramide, especially by non- microsome fractions, were not investigated and could occur in housefly larvae. ACKNOWLEDGMENT Supported in part by Grant No. 15450 from the National Science Foundation. REFERENCES 1. Fast, P.G., Memoirs Entomol. Soc. Can. 37:1964. 2. Crone, H.D., and R.G. Bridges, Biochem. J. 89:11 (1963). 3. Bieber, L.L., E. Hodgson, V.H. Cheldelin, VJ. Brookes and R.W. Newburgh, J. Biol. Chem. 236:2590 (1961). 4. Dawson, R.M.C., and P. 106:3l9(l968). 5. Bieber, L.L., V.H. Cheldelin and R.W. Newburgh, J. Biol. Chem. 238:1262 (1963). 6. Bieber, L.L., J.D. O’Connor and C.C. Sweeley, Biochim. Biophys. Acta 187:157 (1969). 7. Khan, M.A.Q., and E. Hodgson, Comp. Biochem. Physiol. 23:899 (1967). 8. Kumar, S.S., R.H. Millay and L.L. Bieber, Bio- chemistry 9:754 (1970). 9. Rao, R.H., and D. Subrahmanyam, J. Physiol. 15:149 (1969). 10. Rao, R.H., and D. Subrahmanyam, J. Lipid Res. 10:636 (1969). 11. Bieber, L.L., L.G. Sellers and 8.8. Kumar, J. Biol. Chem. 244:630 (1969). 12. Rouser, G., G. Kritchevsky, A. Yamamoto, G. Simon, C. Galli and AJ. Bauman, in “Methods in Enzymology XIV,” Edited by J. Lowenstein, Academic Press, New York, 1969, p. 272. 13. Skipski, V.P., A.F. Smolow and M. Barclay, J. Lipid Res. 8:295 (1967). 14. Haviland, R.T., and L.L. Bieber, Anal. Biochem. 33:323 (1970). 15. Lindberg, 0., and L. Ernster, Method. Biochem. Anal. 3:1 (1955). 16. Sugino, Y., and Y. Miyoshi, J. Biol. Chem. 239:2360 (1964). 17. Lowry, 0.1-1., N.J. Rosebrough, A.L. Farr and R.J. Randall, Ibid. 193:265 (1951). 18. Wells, W.W., T.A. Pittman and I-I.J. Wells, Anal. Biochem. 10:450 (1965). 19. Hirschberg, C.B., A. Kisic and G.J. Schroepfer, J. Biol. Chem. 245:3084 (1970). 20. Bridges, R.G., and J.S. Holden, J. Insect Physiol. 15:779 (1969). ' 21. Bridges, R.G., and J. Ricketts, Ibid. (1970) 22. Bridges, R.G., J. Ricketts and LT. Cox, Ibid. 11:225 (1965). 23. Hodgson, E., and W.C. Dauterman, Ibid. 10:1005 (1964). 24. Hodgson, E., W.C. Dauterman, H.M. Mehendale, E. Smith and M.A.Q. Khan, Comp. Biochem. Physiol. 29:343 (1969). Kemp, Biochem. J. Insect 16:579 LIPIDS, VOL. 6, NO. 7 I 516 25. 26. 27. 2s. 29. 30. 31. 32. 33. G.R. HILDENBRANDT, T. ABRAHAM AND L.L. BIEBER' Scott, T.W., and R.M.C. Dawson, Biochem. J. 108:457 (1968). Dawson, R.M.C., Biochem. Biophys. Acta 33:68 (1959). Kemp, P., G. Hubscher and LN. Hawthorne, Biochem. J. 79:193 (1961). Atherton, R.S., P. Kemp and LN. Hawthorne, Biochim. Biophys. Acta 125:409 (1966). Seamark, R.F., M.E. Tate and T.C. Smeaton, J. Biol. Chem. 243:2424 (1968). Dawson, R.M.C., and P. Kemp, 106:319(1968). Hori, T., O. Itasaka, M. Sugita and I. Arakawa, J. Biochem. (Japan) 64:123 (1968). O’Connor, J.D., A.J. Polito, R.E. Monroe, C.C. Sweeley and L.L. Bieber, Biochim. Biophys. Acta 202:195 (1970). Hort, T., 0. Itasaka, T. Hoshimoto and H. Inoue, Bioche m. J. 34. 35. 36. 37. 38. 39. 40. J. Biochem. (Japan) 55:545 (1964). Hori, T., O. Itasaka, H. Inoue, M. Gamo and I. Arakawa, Japan J. Exp. Med. 36:85 (1966). Kanfer, J.N.. and R.G. Brady, in “Methods in Enzymology XIV," Edited by J. Lowenstein, Academic Press, New York, 1969, p. 131. Gatt, S., Ibid. p. 134. Gatt, S., and E. Yavin, Ibid. p. 139. Nilsson, A., Biochim. Biophys. (1969). Boyer, P.D., and L.L. Bieber, in “Methods in Enzymology X,” Edited by R.W. Estabrook and M.E. Pullman, Academic Press, New York, 1967, p. 768. Friedel, R.O., J.D. Brown and J. Durell, Biochim. Biophys. Acta 144:684 (1967). [Received December 24, 1970] AC“ 1 76:339 LIPIDS. VOL. 6, NO. 7 APPENDIX B CHARACTERIZATION OF GLYCEROPHOSPHORYL-CHOLINE, -ETHANOLAMINE, -SERINE, -INOSITOL, AND -GLYCEROL HYDROLYTIC ACTIVITY IN HOUSE FLY LARVAE G. R. Hildenbrandt* and L. L. Bieber** Department of Biochemistry Michigan State University East Lansing, Michigan #8823 *Current address: Department of Biochemistry The University of Arizona Tucson, Arizona 85721 **To whom inquiries should be sent. 25 SUMMARY Homogehates of Musca domestica (housefly) larvae contain glycerophosphodiesterase activity which fraction- ates into the 88,000 x g supernatant fluid. The phospho- diesterase is inhibited by EDTA and is stimulated by Mg++, Ni++, Co++, and Mn++. The pH optimum is 7.2. The enzyme is stable to heating at 50° for 15 minutes and is insensi- tive to sulfhydryl inhibitors. Glycerophosphoryldiesters of choline, ethanolamine, inositol, serine, glycerol, and B-methylcholine are hydrolyzed to the common product, l-d- glycerOphOSphate, and the appropriate free alcohol. The rate of glycerophosphorylcholine hydrolysis is 70% greater than the rate of hydrolysis of the other glycerOphospho- diesters. Apparent Km values for glycerophosphorylcholine, glycerophosphorylethanolamine, and glycerophOSphoryl-B- methylcholine are 2 - 4 x 10'“ M and for glycerophosphoryl- inositol, 2 x 10"3 M. Competitive studies using various pairs of substrates as well as the exchange of free choline into both glycerophosphorylcholine and glycerophosphoryl- inositol suggest that a single enzyme cleaves all substrates. Product inhibition and reversal of the reaction were not detected. Choline, but not fl-a-glycerophosphate exchanges into glycerophosphorylcholine and glycerophosphorylinosi- tol. 26 27 Introduction Catabolism of glycerophosphatides in animals is generally thought to follow the pattern of phospholipase A and B hydrolysis of the two fatty acyl groups yielding free fatty acids and glycerOphosphoryl-alcohols (GP-Xl compounds); see review by Rossiter (l). Phosphatidyl- inositol is an exception since it is also catabolized by an additional or alternative pathway to diglyceride and inositol phosphate in several mammalian tissues (1). In- vestigations previously reported from this laboratory (2- 4) showed that in housefly larvae, the glycerophosphatides of ethanolamine, choline, B-methylcholine, serine, glycerol, and inositol can be catabolized by microsomal phospholipase A1 and lysophospholipase to yield GP-X compounds. It was also reported (3) that glycerophospho- ryl-B-methylcholine is hydrolyzed by a 90,000 x g super- natant fraction from housefly larvae to yield glycerol-P and B-methylcholine. The limited literature on glycerophosphodiesterase activity in animals does not clearly document the meta- bolic fate of GPS, GPG, and GPI. Baldwin and Cornatzer (5) reported that rat kidney microsomal glycerOphospho- diesterase does not cleave GPI or GPS. In contrast, lGP-X = sn-glycero-3-phosphoryl-X X = Efioline (C) ethanolamine (E), inositol (I), serine (8), glycerol (G), or B-methylcholine. 28 Serratia plymuthicum apparently can hydrolyze all of its GP-X compounds (6). The work reported herein was under- taken to determine whether housefly larvae can hydrolyze all GP—X compounds and whether a single or multiple enzymes are involved. Materials and Methods Enzyme Preparation: Musca domestica larvae were grown aseptically to the third instar (mature, wandering) stage on the diet described by Monroe (7). Larvae free of medium were weighed and homogenized at full Speed in 2 - 3 volumes of 0.025 M Tris, pH 8.0, for l min in a Waring blender. The homogenate was filtered through eight layers of cheesecloth and centrifuged at 40,000 x g for 20 min using a Sorvall SS-34 rotor. The 40,000 x g super- natant fluid was fractionated with solid ammonium sulfate to yield a 35-50% saturation insoluble fraction. This fraction, after resuspension in homogenizing buffer, was lyophilized and stored at -20°. All of the above steps were done at 0-4°. Assays were done using enzyme freshly dissolved in distilled H20. Most of the investigations reported herein were done with a single enzyme preparation. Where indicated, a preparation with about one-third the specific activity of the normal preparation was used. This preparation was redissolved in 0.05 M imidazole, pH 7.2, rather than H20. 29 32 Substrate Preparation: P-labeled substrates were used for the routine assay of GP-X phosphodiesterase activ- ity. Substrates were prepared by rearing housefly larvae as described above except that 32Pi was added to the diets. The phospholipids were extracted and purified as previously described (2-4). Glycerophosphoryl-derivatives were prepared by alkaline deacylation (4) of the various pure phospholipids. Glycerophosphoryl-B-methylcholine was prepared from phosphatidyl-B-methylcholine, which was iso— lated from larvae reared on choline-free, carnitine- supplemented diets (3). The same procedures were followed for preparation of unlabeled GPS from bovine brain lipids and GPI from a commercial preparation enriched in phosPha- tidylinositol. Ion Exchange Separations: Glycerophosphoryldies- ters were separated from the phosphorus-containing hydrolysis products on columns of Dowex l - 8x, 100-200 mesh, formate form. Unless otherwise indicated, columns 6 x 0.5 cm were used. Paper chromatography (4) of enzyme reaction products and other criteria (see Results) demon— strated that GP-X hydrolysis yielded no phosphoryl-X compounds. A batch elution procedure was used to elute unhydrolyzed substrate into fraction A, and glycerophos- phate into fraction B. Fraction B also contained any Pi that was released during the incubation. Fraction A was collected during application of a 2.5 ml aqueous load, a 30 1.5 ml H20 rinse, and a 5 ml elution with 0.1 M ammonium formate. Fraction B was collected immediately following A by application of 10 ml of 0.4 M ammonium formate-0.2 M formic acid. Serine from GPS incubations was separated from phosphate-containing compounds on the above described columns by elution with 5 ml of H20 after loading the incubation mixture at pH 7. Inositol from GPI incu- bations was collected the same as serine and then run over Dowex 50 X 8. H+ form. The eluate was lyophilized and used for gas chromatographic analysis. When total phos- phorus assays were done on the substrate and product fractions or z-a-glycerophosphate analysis was done on fraction B, the fractions were passed through Dowex 50, H+ form, columns to convert formate salts to formic acid and to remove cations. The acidic column eluates were immediately frozen and lyophilized to remove formic acid. In the exchange and reversal experiments larger columns were used and the incubations were diluted to 10 ml with H20 prior to loading onto the columns. When GPC and choline in methyl-lQC-choline labeling experiments were separated, the anion exchange columns were washed with H20 as for serine and inositol, and this fraction was loaded onto Dowex 50, H+ form, columns. GPC was eluted using several volumes of H20 and choline was recovered by elution with several column volumes of 4 N HCl. 31 Assay Procedures: All incubations were done in 12 m1 conical glass centrifuge tubes. Incubations were initiated by adding enzyme to the temperature equilibrated reaction mixtures. The reactions were terminated by heating for four minutes in a boiling water bath. The 0.5 ml reaction volume was diluted with 2 ml of H20 and vortexed prior to loading onto anion exchange columns. When t-d-glycerophosphate was assayed directly (Table IV, exp. no. 3, 8, and 9), the boiled incubation mixtures were lyophilized and resuspended in a known volume of distilled water and aliquots were taken for the 1-9- glycerophosphate assay. The remainder of the mixtures was put onto anion exchange columns. Controls were run for each experiment and each substrate preparation. Iden- tical incubation mixtures and conditions as for experi- mental tubes were used for the controls except that the enzyme was inactivated by boiling prior to the addition of substrate. The per cent of total radioactivity recovered off anion exchange columns in the 8 fraction (phosphate- containing hydrolysis products of GP-X) for control incubations was always less than 2%. This blank value largely represents breakdown products from the substrate, as similar values were obtained when samples of sub- strate were applied to columns without incubation. The precision of per cent distribution of radioactivity from anion exchange columns is i 0.2%. In all results reported 32 32P were incubated and recovered a minimum of 2,000 Cpm of from the column fractionation. The assays routinely yielded 10-40% (above control values) of total radio- activity as product (fraction B). 32P Cerenkov radiation was quantitated in a Packard Tricarb scintillation counter for 10 ml aqueous samples in polyethylene vials according to Haviland and Bieber (8). For monitoring anion exchange columns, the 10-ml substrate and product fractions were collected di- rectly into polyethylene vials and counted. For quanti- tating ll‘C, 0.1 ml aqueous aliquots in 10 ml of a 1:1 mixture of Triton X-100ttoluene scintillator2 were counted. l”C-labeled choline were lyo- GPC samples containing methyl- philized and redissolved to 0.5 m1 from which aliquots were counted. lnositol enzymatically hydrolyzed from GPI was quantitated by gas chromatography (9) of the trimethyl- silyl derivative. Aliquots of an internal standard, a- methylmannoside, were added to incubation mixtures prior to ion exchange purification, lyophilization, and deri- vatization. Ratios of inositol to phosphate in GPI prep- arations were determined to be 110.1. Free inositol in 2Toluene scintillation mixture: 4 g 2,5-diphenyloxazole (PPO) 100 mg l,4-bis-2-(4-methyl-5-phenyloxazolyl)-benzene (POPOP) l 2 toluene 33 boiled enzyme control incubations was less than 0.5% of GPI-bound inositol. Serine hydrolyzed from GPS was quantitated with an automated amino acid analyzer.3 The ratio of phosphate to serine in a GPS acid hydrolyzate was found to be 1.1/0.85. This low value for serine is due to the manipulations in- volved in the serine preparation. The phOSphatidylserine source was greater than 90% pure, and routine ion exchange purification of its deacylation product removes essen- tially all non-GPS phosphate. Boiled enzyme control incubations showed negligible amounts of free serine in the GPS preparation. fi-G-glycerophOSphate was determined fluorometri- cally or spectrophotometrically following NAD+ reduction by t-a-glycerophosphate dehydrogenase (Sigma). The protocol for this assay was essentially that described for glycerol analysis (10) with the omission of glycerol kinase, Mg++, and ATP. Protein was determined according to Lowry et a1. (11). Total phosphate was determined as described by Bartlett (12). Results Distribution, Purification,gand Stability of the Glycerophosphodiesterase: Homogenization of housefly 3D. C. Robertson, H. B. Brockman, W. I. Wood, and W. A. Wood, personal communication. 34 larvae in 0.025 M Tris, pH 8, and subsequent fraction- ation by differential centrifugation yielded fractions containing nuclei and mitochondria (250-9,000 x g), lyso- somes and microsomes (9,000-40,000 x g), microsomes (40,000-88,000 x g), and soluble components (88,000 x g supernatant). No significant glycerophosphorylcholine phosphodiesterase activity was found in any of the partic- ulate fractions. All of the activity detectable in the 250 x g supernatant fluid (Table I) is recovered in the 40,000 x g supernatant fraction. The greater activity in the 250 x g supernatant fraction than in the homogenate was not always observed. Greater than 70% of the activity can be recovered in an 88,000 x g supernatant fluid, but the limited purification obtained did not warrant this high-speed centrifugation. The phosphodiesterase activity in the 40,000 x g supernatant fluid is 75-80% stable to heating for 15 min at 50° (Table II), and is stable to freezing and thawing. Fractionation of the 40,000 x g supernatant fraction with ammonium sulfate, 35-50% saturation, routinely yielded 90% of the total activity in the precipitate with a 2- to 5-fold purification. This 35-50% fraction was dissolved in 0.05 M imidazole, pH 7.2, buffer, lyophilized, and stored at -20° for up to 2 years with less than 15% loss of activity. 35 TABLE I Isolation of glycerophosPhodiesterase fromihousefly larvae Total Specific Activity -fold Fraction Protein Activity Recovered Purified nmoles/min/ mg mg protein % Homogenate 536 4.4 100 1.0 250 x g Supernatant 345 7.8 115 1.8 u0,000 x g Supernatant 226 12.0 115 2.7 35-50% (NHQZSO1+ Precipitate 72 35.0 106 8.0 12.5 g of larvae were homogenized for l min in 30 ml of 0.025 M Tris buffer, pH 8.0. All manipulations were carried out at 044° as described in Methods. 32p-labeled glycerophosphorylcholine was the substrate. Assays for 2P-labeled glycerophosphate were performed as described in the Methods section. 36 TABLE II Effects of heating and EDTA on glycero- phosphodiesterase actIvity a-Glycero- Enzyme phosphate Experiment Additions formed nmoles Unheated 15 mM MgCl2 330 A Unheated 2.5 mM EDTA 10 Heated (15 min at 50°) 15 mM MgCl2 260 Heated None 150 B Heated 15 mM MgCl2 210 — Heated 2.5 mM EDTA 110 Heated 2.5 mM EDTAE 210 15 mM MgCl2 2The sample was incubated with EDTA for 10 min prior to addition of MgCl2 and then the incubation was continued for 20 minutes. All incubation mixtures contained approximately 2 mg of 40,000 x g supernatant protein unheated or heated for 15 min at 50°, 0.15 M sucrose, and 15 mM Hepes, pH 7.4, in a final volume of 2 ml. Incubation was at 30° for 20 min. The glycero-32P-phosphoryl-B-methylcholine sub- strate was 0.56 mM in experiment A and 0.28 mM in experi- ment B. Assays were performed as_described in the Methods section. 37 In order to determine whether the phosphodies- terase activity occurred only in housefly larvae entering the pupal stage of development or was present in other develOpmental stages, eggs and l- to 2-day old pupae were homogenized and assayed for activity. Results (not shown) demonstrated that both of these developmental stages had the same activity on a live weight basis as third instar larvae. Because the preparation was not pure, it was essential to know if contaminating activities were present that could interfere with glycerophosphodiester assays. All of the phosphomonoester released was fi-a-glycerphos- phate as determined by t-d-glycerophosphate dehydro- genase assays. Phosphatases active on fi-a-glycerophos- phate were not detected by paper chromatographic analysis of products of GP-X hydrolysis. The enzyme preparation did not liberate 32Pi from 32P-1abeled glycerophosphate. This was determined by partitioning the reaction products into isobutanol-benzene (13). Traces of phOSphatase activity were detectable using p-nitrophenylphosphate (14) as substrate. This activity at pH 7.2 was less than 2% that of the phosphodiesterase. Slightly higher p-nitro- phenylphosphate hydrolysis was detected at pH 4.5 and pH 9.0. Biologically occurring phosphodiesters that do not contain glycerol were not tested as competitive in- hibitors of GP-X hydrolysis, but the synthetic 38 phosphodiesterase substrate, bis-p-nitrOphenylphosphate, was assayed (15). The synthetic substrate was cleaved at about 25% the rate observed with G32PE. This activity was apparently different than the G32PE hydrolytic activity because cleavage of the synthetic substrate was not inhibited by G32PE, and G32PE cleavage was not inhibited by the synthetic substrate. Effect ofng on Glycerophosphodiesterase Activity: The glycerOphosphodiesterase has a pH optimum of 7.2 that falls markedly within 1 0.5 pH units when G32PC is the substrate, as shown in Figure 1. With G32PE as substrate, the pH optimum is 7.4, but the curve is nearly flat from pH 7.0 to pH 7.8. The rates of G32PC cleavage above 40 nmoles/min/mg protein may be slightly low due to the sub- strate becoming limiting during the incubation period. The maximal rates for G32PE cleavage shown in Figure l are lower than sometimes observed with the same enzyme prepa- ration. This was undoubtedly due to the presence of an impurity in the substrate for this determination. In data not shown, the pH optimum for G32PS was similar to G32PE. Divalent Cation Requirement of the Glycerophos- phodiesterase: Mg++ stimulates the phosphodiesterase activity. Small amounts of glycerophosphodiesterase activity were obtained in the absence of added Mg++, but 39 .mcflamao:mnvoamponmmonmimmmioboomam ma mammw mmafiaonoambonamoamimmmrosoomam ma ommmw .mpoepmz one as confinemep we meow ones m>Mmm< .nvaHo>Hw can . Carey .0 eHoNMpHEfl .e eyebrow 9.33 mmmmw pom mbmmmsm . Conflomam pad .0 mflfi. . Omaowmfiodfi . Cowmvmom 6903 um me now mbmmmsm .oom we CHE om how we: COAHMQsocH .cflopmbm we is can .mmmmw so ommmw as H {when as 2. seems as .a. pmdflmpcoo mobdyxfls :owvmnsucH .ommbmpweaponamonmosmomaw mzu >9 mflmmaogpzm mew can 0mm mo mocmpcomma mm a obzmflm 40 I I l l O O (I) <1' ugamrd bw/ugw/sepwu Figure 1 111 this low activity varied from enzyme preparation to prepa- ration and was probably due to differences in residual divalent cations. As shown in Table IIA, EDTA inhibits the phosphodiesterase activity. The inhibition by EDTA was usually greater than 95%. Addition of excess MgC12 over- comes the EDTA inhibition and restores the activity to that of the control (see Table IIB). Table III shows that the divalent cation require- ment is quite specific. The highest activity was obtained with Mg++ and Co++. Mn++ and Ni++ gave approximately 50% as much activity as Mg++. The variation in salt concen- tration in Table III suggests that the major portion of cation stimulation occurs at concentrations below 0.5 mM. As shown in Figure 2, submillimolar Mg++ concentrations give a definite stimulation. The activity leveled off 32 above 4 mM with G PE. The optimum Mg++ concentration for G32PC cleavage was approximately 10 mM. Quantitation and Identification of GlycerOphospho- diesterase Reaction Products: When either GPS, GPE, GPI, GPC, or GPG were used as substrate, the only phosPhomono- ester detected was t-a-glycerOphosphate. PhOSphoryl monoesters of serine, ethanolamine,'choline, and inositol were not detected using column and paper chromatographic techniques. For all of the above-mentioned substrates, the amounts of z-a-glycerOphosphate determined enzymati- cally were very similar to the values obtained using the 112 TABLE III Effects of divalent cations on glycero- phosphodiesterase activity 32 , 32 Substrate G PL G PC Salt mM of the salt 0.5 2.5 10 10 nmoles diglycerophosphate/min/mg protein None 3.6 4.8 MgC12 50 52 58 80 MgSOu 54 51 61 97 Mg(NO3)2 41 51 52 84 CoC12 55 57 48 94 MnSOq 29 31 30 30 NiCl2 18 22 29 55 Zn(C2H302)2 3.5 0.7 0 0 Ca(N03)2 3.9 0 0 1.4 CdC12 3.7 1.9 0.3 1.9 CuC12 0 0.8 0.3 0 The incubation mixEure contained the indicated salt, and 2. 8 mM G3 2PC or G3 PE, 40 mM imidazole buffer, pH 7.2, and 0.17 mg of protein. Samples were incubated at 30° for 30 min and the assays were performed as described in Methods. G32PE = glycero-32P-phosphorylethanolamine; G32PC = g1ycero32P-phosphorylcholine. 43 Figure 2 Effect of Mg++ Concentration on the Enzymatic Hydrolysis of GPE and GPC. The incubation mixtures con- tained MgCl as indicated as well as 40 mM imidazole, pH 7.2, 0.1 mg protein, and G32PE or G32PC in a final volume of 0.5 m1. G32PE (2.8 mM) was incubated at 30° for 30 min and G32PC (3.0 mM) was incubated at 30° for 15 min. The assays were performed as described in the Methods. G32PE is glycero-32P-phosphorylethanolamine; G32PC is glycero- 2P-phosphorylcholine. v.11 1251— 5295 oE\c_E\ 8.06: 1 l2 20 Figure 2 45 column assay procedure; compare the nmoles of a l-d- glycerophosphate to the nmoles of 32P—labeled product in experiments 3, 4, 5, 8, and 9 of Table IV. The molar ratio of fi-a-glycerophosphate determined enzymatically to the glycerophosphate determined by the column assay was be- tween 0.89 to 0.95. When GPS was the substrate, the amount of serine liberated was similar to the amount of l-G-glycerophosphate released. Compare 238 nmoles serine to 260 nmoles l-d- glycerOphOSphate in experiment 1 and 340 nmoles serine to 388 nmoles 2-d-glycerophosphate in experiment 2 of Table IV. Quantitative analysis of inositol hydrolyzed from G32PI compared very closely to the amount of 32? product formed; i.e., in experiment 6 of Table IV, 350 nmoles 32P- labeled anglycerophosphate and 356 nmoles inositol were released. In addition, experiments 6 and 7 show that the values are quite reproducible for identical incubations. Compare 350 to 333 nmoles 32 P product and 356 to 320 nmoles inositol. Experiment 9, Table IV, shows that G32PG is, with- in experimental error, completely hydrolyzed to the z-a- glycerOphosphate, indicating that the d isomer is not formed. The control contained less than 2% as much l-a- glycer0phosphate as the experimental product fraction. No attempt was made to assay free glycerol released from G32PG. 46 TABLE IV Identification and quantitation of_products from hydrolysis of—glycerophosphodiesters’by larval phosphodiesterase EXP° , 32P-labeled No. Substrate 2-a-GP* product X-OH V nmoles £-u-GP*/ min/mg nmoles nmoles nmoles ,protein 1 GPS (5.5 mM) 260 238 38 2 GPS (11 mM) 390 388 99 3 C32Ps (8.0 mM) 490 ueu 7.33 u G32PE (5.0 mM) 910 1020 172 5 G32PC (4.2 mM) 800 900 80 6 G32PI (1.7 mM) 350 356 412 7 G32PI (1.7 mM) 333 320 402 8 G32PI (8.0 mM) 955 1020 ’ 153 9 G32PG (8.0 mM) 800 890 8.92 a Low Specific activity enzyme preparation was used. 2 v = nmoles 32F product/min/mg protein. X-OH = serine (experiments 1 and 2) or inositol (experi- ments 6 and 7). * t-a-GP = R-a-glycerophosphate. All incubations were done at 30° with a final volume of 0.5 ml. The buffer was imidazole, pH 7.2. Composition of mixtures and incubation times were as follows: 1 and 2, 40 mM buffer, 2 mM MgC12, 0.23 mg protein, and 30 min; 3, 60 mM buffer, 4 mM MgC12, 1 mg protein, and 63 min; 4, 40 mM buffer, 4 mM MgC12, 1 mg protein, and 60 min; 5, 40 mM buffer, 2 mM MgC12, 0.5 mg protein, and 20 min; 6 and 7, 40 mM buffer, 2 mM MgC12, 0.14 mg protein, and 60 min; 8, 50 mM buffer, 4 mM MgC12, 1 mg protein, and 63 min; 9, 50 mM buffer, 4 mM MgC12, 1 mg protein, and 90 min. i-a-glycero- phosphate was assayed enzymatically, and the P product was determined by the column fractionation assay. X-OH was assayed by amino acid analyzer for serine (experiments 1 and 2) and by gas chromatography for inositol ( experiments 6 and 7). See Methods for details of the assays. The values in parentheses in the substrate column indicate the concentration of substrate used. 47 Glycer0phosphoryldiesterase Sgbstrate Affinities and Vmax's: The affinity of the enzyme(S) for two of the substrates is indicated by the data in Figure 3. The Km for G32PC is an order of magnitude lower than that for GPI. The apparent Km for GPC was 2 x lO-“M and for GPI was 2 x 10"3 M. From data not shown, values for GPE and gly- cerophosphoryl-B-methylcholine were similar to the Km for GPC; the variation was between 2 and 4 x 10'” M. Hydroly- sis of glycerophosphoryl-B-methylcholine by a 40,000 x g supernatant fraction from housefly larvae has been previously reported by this laboratory (3). Complete characterization of the products has not been done, but kinetically, glycerophosphoryl-B-methlycholine behaves as a competitive inhibitor of G32PC hydrolysis with a Ki of 7 x 10'” M. Complete data is not available for GPG or GPS Km's, but they appear to be on the order of GPI rather than GPC. Experiments 1 and 2 of Table IV indicate that the Km for GPS is at least as large as the Km for GPI. Values for Vmax with various enzyme and substrate preparations varied, but the Vmax for GPC was consistently 70% higher than that for GPI (see Figure 3). With the enzyme preparation used for most of these studies, the Vmax (nmoles/min/mg prOtein) for GPC was near 90 and for GPI was 50-55. GPE, GPS, GPG, and glycerophosphoryl-B- methylcholine all had apparent Vmax values of 50 1,5. 48 Figure 3 Determination of Km Values for GPC and GPI. Samples containing G32PC contained 2 mM MgC12, 00 mM imidazole, pH 7.2, and 0.084 mg protein in 0.5 ml. The samples were incubated at 30° for 10 min. Mixtures containing GPI contained 4 mM MgC12, 40 mM imidazole, pH 7.2, and 0.23 mg protein/ml in a final volume of l or 3 ml. The incubations were done at 30° for 15 or 30 min. Hydrolysis of GPI was quantitated by measuring the inositol released. Inositol was quantitated using gas chromatography as described in the Methods. G32PC was assayed as described in the Methods. G32PC is glycero-32P-phosphorylcholine; GPI is glycerophosphorylinositol. 49 Vmox=92 nmoles/min mg Km = 0.2 mM 0.02 LIB tic its Vmox = 55 nmoles/ min mg 9 ‘T__'_' JR 3 u no 3 Z O.|5 l/v (nmoles" x min x mg protein) GPI 4 8 l/ [5] (mM") Figure 3 50 Effects of Potential Inhibitors and Reaction Products on Glycerophosphodiesterase Activity: The phos- phodiesterase activity was not significantly affected by the reaction products, choline, ethanolamine, inositol, and glycerophosphate at concentrations near that of the substrates (see Table V). It should be noted that the hydrolysis in the control is about 10% below values ob- tained in other experiments. This may partially account for the apparent stimulation at 1 mM by most of the com- pounds tested. In data not shown, 25 mM concentrations were also used. These data are in close agreement with the 5 mM values except for choline and ethanolamine which give 75 and 20% of the control rate, respectively. The data in Table VI indicate that the phosphodi- esterase activity is not dependent on catalytic sulfhydryl groups since N-ethylmaleimide, iodoacetamide, and p- chloromercuribenzoate were not inhibitory. The failure of trypsin (Nutritional Biochemical Co.) and pronase (Cal Biochem) to inhibit the activity after 50 min of incu- bation (Table VI) or even after 80 min (not shown) indi- cates that the phosphodiesterase enzyme(s) is not highly sensitive to these types of protease activity. Mutual Inhibition Between Pairs of Glycerophos- phodiester Substrates: To obtain some indication as to whether or not a Single or multiple enzymes are involved Effects of potential inhibitors and reaction 51 TABLE V products on the glycerophosphodiesterase activity 1 mM compound nmoles l-a-GP/ 5 mM compound nmoles l-a-GP/ Compound added min/mg protein % min/mg protein .3 None 72 100 72 100 di-a-GP 80 111 60 83 B-GP 76 106 67 93 Choline 81 113 72 100 Ethanolamine 83 115 67 93 Inositol 82 114 88 122 PhOSphoryl- choline 83 115 76 106 Incubation mixtures containing 3.1 mM glycero-32P- phosphorylcholine, 4 mM MgC12, 40 mM imidazole, pH 7.2, and 0.17 mg of protein in 0.5 ml final volume were incu- bated at 30° for 30 min. GP = glycerophosphate. See Methods for assay procedures. 52 TABLE VI Effects of potential enzyme inhibitors on glycerOphosphodiesterase activity Inhibitor nmoles 2-a-glycerophosphate/min/mg protein None 21.3 NEME, 2.0 mM 20.3 IAAE, 2.0 mM 22.7 PCMBE: 0.04 mM 27.6 Trypsin, 0.025 mg 20.9 Pronase, 0.025 mg 22.9 ENEM, N-ethylmaleimide. ~ EIAA, iodoacetamide. c h —PCMB, p-chloromercurlbenzoate. Incubation mixtures containing inhibitor, 4 mM MgC12, 50 mM imidazole, pH 7.2, and 1 mg of protein of low specific activity in 0.5 m1 final volume were preincubated at 30° for 10 min (20 min with pronase and trypsin). 5 mM glycero-32P-ph08phorylcholine was added and incubation was continued at 30° for 30 min. See Methods for assay procedure. 53 in hydrolysis of the various substrates, eXperiments in- volving incubation of pairs of substrates separately and in combination were performed. Figure 4 shows that the hydrolysis of substrate pairs was not significantly greater than the most active substrate incubated alone. For each substrate pair, both substrates were hydrolyzed and the rates were less than the rates of hydrolysis for the indi- vidual substrates. For example, as shown in Figure 4A, G32PE hydrolysis in the presence of GPI was about half the control rate. This and the fact that the GPI hydrolysis was inhibited by one-third suggests that G32PE competes slightly less favorably than GPI for hydrolysis when co- incubated. Figure 4D indicates that GPI may be a slightly better competitor than the G32PS and is hydrolyzed at a higher rate than G32PS. In evaluating Figure 4B and C, it is noted that the GPC preparation used in B gave rates about 35% below normal (e.g., G32PC vmax. Figure 3). Even considering this limitation for GPC hydrolysis, it is apparent from Figure 4B and C that GPC competes much more favorably for hydrolysis than does GPI or GPE. All data obtained are consistent with a single enzyme hydrolyzing all of the substrates. Reversal of the Glycerophosphodiesterase Reaction: Attempts to demonstrate formation G32PC from 32P-labeled fi-a-glycerOphosphate and choline were unsuccessful (experi- ments I and II of Table VII), even at very high (0.1 M) 59 Figure 4 Mutual Inhibition of Enzymatic Hydrolysis by Pairs of Glycerophosphodiesters. All incubation mixtures contained 4 mM MgC12 and 65 mM imidazole, pH 7.2, in addition to enzyme and substrate in a final volume of 0.5 ml. Incubation temperature was 30°. The substrate concentrations were: for Experiment A, G32PE, 4.5 mM and GPI, 10 mM; for Experiment B, G32PE, 5 mM and GPC, 5 mM; for Experiment C, G32PC, 5 mM, and GPI, 10 mM; for Experiment D, G32PS, 9.6 mM, and GPI, 10 mM. Hydrolysis of the labeled substrates was measured by counting the anion exchange column fractions and hydrolysis of un- labeled substrate and combined substrates was measured using total phosphate assay of the substrate and product fractions. The open bar in the column for both substrates indicates total 2-a-g1ycer0phosphate produced. G32PE is glycero-32P-phosphorylethanolamine; GPI is glycerOphosphorylinositol° GPC is glycerophos- phorylcholine; G32PC is glycero-ézP-phosphorylcholine; G32PS is glycero-32P-phosphorylserine. 55 2.2.2.35 SE on acct—5:65 SE 0N. £9.05 9: mm.o\om>om_o 8.9:: unlabeled I both lncuboted labeled Substroleis) Figure 4 56 TABLE VII {‘ ‘ ; o ‘0 u- f o o o _i ‘0. 21¢ e-ch:_:e of Me- C-choline or 2-a-glycerol- P into GPC and GPI Incu- Radioactive Other bation Radioactivity re- Exp. compound used Reactants time covered in GPC min- w. 32.53.03. l-a-glycerol-B Egg mM 90 2.7 x 10”(2 6.3 1° 32p (:SOXmfi? 1.9 x 10” 0.0 cpm, V . B Choline— 90 3.0 x 10”3 7.0 100 mM 1.0 x 10“ 3.3 lgg'glycer°l‘6 CholineE 90 753 0.10 II. P(7.5 x 10 10 mM 100 0.13 cpm, 15 mM) Mel”c- GPC 60 1003 0.83 choline 6 10 mM 2060 17.0 III. (1.2 x 10 cpm, 10 90 2252 1.9 mM) GPI 30 630 5.3 10 mM 60 1020 8.5 90 1170 9.7 a . -801led enzyme control. 2 50 umoles of carrier GPC were added after stopping the reaction. 3 5 nmoles of carrier GPC were added after stOpping the reaction. All reaction mixtures contained, in addition to compo- nents indicated, 0 mM MgC12, 0.5 mg of low specific activity protein, and 65 mM imidazole, pH 7.2 (reactants were adjusted to pH 7.0) in a final volume of 0.5 ml. Assay for-32P-labeling of GPC was by anion exchange columns of a sufficient capacity to separate the larger than usual quantities of anionic material. The incorporation of Meluc- choline into GPC was determined by cation exchange separation of GPC and choline on Dowex 50 H+ columns. GPC = glycero- phosphorylcholine; GPI = glycerophosphorylinisitol. 57 concentrations of l-a-glycerophosphate and choline. Ex- change of 32P-labeled 2-a-g1ycerophosphate into GPC was also not detected as shown by the data given in experiment I of Table VII. The incubations employing 0.1 M reactants were adjusted to pH 7.0 with NaOH and the choline was made up in imidazole buffer, pH 7.2. Thus, the effective buffering capacity was higher than indicated in the legend of Table VII. ll"C-choline was incubated with either When methyl- GPC or GPI, a significant percentage (on the order of 0.1%) of the total activity was recovered as GPluC. In the ex- change 0f methYl‘luC-choline into GPI, the choline incor- poration appeared to be approaching a steady level after 90 min of incubation. This suggests that the rate of exchange of choline for inositol is being approached by the rate of hydrolysis of the spluc produced. Total GPluc con- centration approached 0.01 mM after 90 min of incubation. For the experiment involving exchange of methyl-luc- choline into GPC, after 60 min, 1/600 (0.l7%) of the label appeared in GPC which is 20 times the amount of luC in the blank. Approximately 20% of the initial GPC was hydrolyzed during the incubation. Thus, “0% of the total choline in the mixture was present in the GPC fraction at the end of the experiment. The data strongly indicate that choline was exchanged into GPC and GPI. 58 Discussion Subcellular Distribution: Our results show that housefly larvae contain glycerOphosphodiesterase activity which is soluble in a 88,000 x g supernatant fraction. The particulate fractions, including microsomes, contain little, if any, of the phosphodiesterase activity. This is in contrast to the report by Baldwin and Cornatzer (5) that rat kidney glycer0phosphodiesterase is microsome- bound. Divalent Cation Effects: The glycerophosphodies- terase is greatly stimulated by the divalent cation Mg++ and is also stimulated by Ni++, Co++, and Mn++. The opti- mum Mg++ concentration is M mM for GPE and “-10 mM for GPC (Figure 2). EDTA can completely inhibit the activity and the inhibition by EDTA can be overcome by excess Mg++. This is in contrast to the enzyme from Serratia plymuthicum (16) which is inhibited by mM concentrations of divalent cations and shows an EDTA inhibition that is irreversible with metal ions. EDTA also inhibits the rat kidney gly- cerophosphodiesterase (5), but this EDTA effect is reversed with equimolar Zn++ (l7) and partially inhibited by higher Zn++ concentrations. Like rat kidney microsomes rat brain homogenates (18) show no loss of activity on dialysis unless EDTA is added. This inhibition is reversed by several metal ions in both tissues. Rat liver 59 GPC diesterase (19) is stimulated slightly by Mg++ after dialysis but is inhibited at concentrations above 3 mM. The rat liver diesterase is also inhibited by Zn++ and Ca++. The larval enzyme appears to be quite different from all the rat tissue glycerophosphodiesterases and the §. plymuthicum enzyme. It is nearly inactive, even in undi- alyzed preparations, unless divalent cations are added. The specificity for Mg++ (Table III) and the failure of Mg”+ to inhibit at 10 mM concentration are in sharp con- trast to the bacterial and rat enzymes. pH Optima and Heat Stability: The larval enzyme is 80% stable to heating 50° for 15 min. This is very similar to the results reported for rat kidney (5) and g; plymuthi- cum (16). pH optima for the bacterial, rat kidney, and rat brain (18) enzyme are about pH 9. Larval activity is Optimal at pH 7.2 which is similar to the rat liver prepa- ration (pH 7.5) of Dawson (19). Substrate §pecificity: Housefly larval glycerophos- phodiesterase is active on gn-glycero-3-phosphoryl- ’ethanolamine, -choline, -serine, -inositol, -1'§nfglycerol, and -B-methy1choline. A similar spectrum of substrates has been reported (6) for the §; plymuthicum enzyme. The products of enzyme hydrolysis are without exception l-a- glycerophosphate (gn—glycerol-3-ph05phate) and free alcohol moiety. Our results for the hydrolysis of 60 Enfglycerol-3-phosphoryl-l'snfglycerol (Prepared by deacyl- ating housefly larval phosphatidylglycerol) show the product to be all snfglycerol-3-phosphate and not s2: glycerol-l-phosphate. This agrees with the results for é; plymuthicum (6). Km and Vmax: housefly larval glycerophOSphodies- terase has an apparent Km of 2.0 x 10'” M for GPC and 2 x 10‘3 M for GPI. GPE and glycerOphosphoryl-B-methyl- choline have Km's in the range of 2 - u x 10'” M. The limited data available suggests that GPG and GPS have Km's on the order of 10'3 M rather than 10‘“ M. The reported values for GPC and GPE Km's with E; plymuthicum enzyme (16) are 2'5 X 10-3 and 1-2 X 10'3 M, respectively. Baldwin and Cornatzer (5) determined the rat kidney Km's to be 2.2 x 10-3 M for GPC and 11.5 x 10-3 M for GPE. Vmax values for the larval enzyme preparation used for most of this work show that GPC is hydrolyzed (92 nmoles/min/mg protein) about 80% faster than any of the other substrates studied (GPG, GPE, GPI, GPS, and glycero- phOSphoryl-B-methylcholine in the range 501_5 hmoles/min/ mg protein). GPC is cleaved about 2.5 times as fast as GPE by the rat kidney enzyme (5) and about 30% faster by the §; plymuthicum enzyme (16). Other: Sulfhydryl groups apparently are not re- quired for activity since N-ethylmaleimide, 61 p-chloromercuribenzoate, and iodoacetamide are not inhibi- tory. Similar findings have been reported for the mammalian and bacterial enzymes. The larval enzyme shows no significant inhibition of GPC hydrolysis by the reaction products at l or 5 mM concentrations and only slight (25%) inhibition by 25 mM choline. This is in contrast to the rat kidney enzyme (5) which was inhibited up to 60% by choline at 1-10 mM con- centrations. This product inhibition and the apparent substrate specificity for GPC and GPE suggest that the rat kidney enzyme differs from the larval enzyme and probably the bacterial enzyme in that the larval enzyme is not as specific. It does not readily distinguish ethanolamine and choline from serine, inositol, and glycerol. Our results indicate, but do not prove, that all GP-X compounds are cleaved by one larval enzyme. When pairs of substrates were co-incubated, both were hydrol- yzed but at rates below noncompetitive values (Figure 0). GPI competes slightly more favorably for hydrolysis (i.e., it is less inhibited relative to noncompetitive rates) when co-incubated with GPE or GPS. GPC is a better com- petitor for hydrolysis than GPE or GPI. GPC cleavage is inhibited only 10-15% in the presence of GPI or GPS, but GPC inhibits hydrolysis of GPI and GPS 80-85%. Total cleavage of any pair of substrates does not exceed signif- icantly the cleavage of the single most active substrate 62 in the pair. The bacterial enzyme (16) and the rat liver enzyme (19) exhibit competitive kinetics for hydrolysis of GPC in the presence of GPE. In results not published, the larval enzyme deviates from competitive kinetics when GPE is used as the inhibitor of GPC hydrolysis but glycero- phosphoryl-B-methylcholine is a competitive inhibitor of GPC. Hydrolysis of GPC by the larval glycerophospho- diesterase is virtually irreversible; however, lL’C-methyl- labeled choline did exchange slowly into GPC and GPI. These observations suggest that the z-a-glycerophosphate is released from the enzyme slower than the alcohol moiety. Release of fi-a-glycerophosphate is nearly irreversible since 32P-labeled l-a-glycerophosphate was not exchanged into GPC. The exchange of choline into GPI and GPC indi- cates, but does not prove, that some type of enzyme- phosphoryl-glycerol intermediate is formed during the enzymatic conversion of glycerophosphodiesters to £-a— glycerOphosphate and XOH. A plausible reaction sequence that is consistent with our limited data is: l. GPX + enzyme I glycerOphosphorylenzyme + XOH 2. Glycerophosphoryl-enzyme + i-a-glycerophosphate + + HOH enzyme 63 The exchange of free choline into GPC and GPI, the only phosphodiesters tested, is strong evidence that a single enzyme hydrolyzes both GPC and GPI. From the above discussion, it is apparent that the housefly larval enzyme has several significant differences from both the bacterial and the rat kidney microsomal en- zymes. The properties observed for the larval enzyme indicate that it can hydrolyze all the GP-X compounds that are known (2 and H) to be major catabolic products of gly- cerOphOSphatides in housefly larvae. The amount of activity, about 200 nmoles/min/g of larvae, would be suf- ficient to turn over the 7 umoles of lipid phosphate/g of larvae1+ in less than an hour. Insects in the mature larval stage that are entering the pupal stage undergo extensive breakdown of phospholipids and changes in phos- pholipid composition (20). It is likely that glycero- phosphodiesterase activity is essential during this phase of metamorphosis when the organism, operating as a nearly closed system, extensively hydrolyzes its cellular constit- uents and reassembles the components into the tissues of the adult organism. The apparent insensitivity of the en- zyme to protease may be important during the autolysis of pupation. 1+L. L. Bieber, unpublished observation. 64 Acknowledgments Supported, in part, by Grant No. 15450 from the National Science Foundation. Journal Article No. - - - - from the Michigan Agricultural Experiment Station. The authors wish to thank D. C. Robertson for per- forming the determination of serine. 10. ll. 12. 13. l“. 15. 65 References R. J. Rossiter, Metabolic Pathways II 3rd Edition, 69 (1968), David M. Greenberg (Editor):TAcademic Press, New York. 8. S. Kumar, R. H. Millay, and L. L. Bieber, Biochem- istry g, 754 (1970). L. L. Bieber, L. G. Sellers, and S. S. Kumar, J. Biol. Chem. 29%, 630 (1969). G. R. Hildenbrandt, T. Abraham, and L. L. Bieber, Lipids g, 508 (1971). Jerry J. Baldwin and W. E. Cornatzer, Biochim. Bio- phys. Acta 169, 195 (1968). R. Prasad and A. A. Benson, Biochim. Biophys. Acta 187, 269 (1969). R. E. Monroe, Ann. Entomol. Soc. Amer. 55, 190 (1962). R. T. Haviland and L. L. Bieber, Anal. Biochem. 33, 323 (1970). W. W. Wells, T. A. Pittman, and H. J. Wells, Anal. Biochem. 12, 950 (1965). Sidney S. Chernick, Methods in Enzymology XIV, 627 (1969), J. Lowenstein (Editor), Academic Press, New York. 0. H. Lowry, N. J. Rosebrough, A. L. Farr, and R. J. Randall, J. Biol. Chem. 193, 265 (1951). G. R. Bartlett, J. Biol. Chem, 239, #66 (1959). O. Lindberg and L. Ernster, Methods of Biochemical Analysis 3, l (1955), David Glick (Editor), Inter- science Publishers, New York. Kurt Linhardt and Klaus Walter, Methods of Enzyme Analysis, 779 (1963), H. U. Bergmeyer (Editor), Academic Press, New York. W. E. Razzell and H. G. Khorana, J. Biol. Chem. 236, 1199 (1961). 16. 17. 18. 19. 20. 66 Osamu Hayaishi and Arthur Kornberg, J. Biol. Chem, 2 6, 697 (1953). Jerry J. Baldwin, Phyllis Lanes, and W. E. Cornatzer, Archives Biochem. Biophys. 133, 224 (1969). G. R. Webster, Elizabeth A. Marples, and R. H. S. Thompson, Biochem. J. 65, 374 (1957). R. M. C. Dawson, Biochem. J. 62, 689 (1956). Paul G. Fast, Progress in the Chemistry of Fats and Other Lipids II, part 2, 181 (1970), R. T. Holman (Editor), Pergamon Press.