Loss of Hydrogen from Dihydroxyacetone Phosphate during Glyceryl Ether Synthesis*

mechanism of ether bond formation


SUMMARY
In order to obtain information on the mechanism of ether bond formation in the synthesis of 0-alkyl lipids, a study was carried out to determine if there is a loss of hydrogen from carbon 3 of dihydroxyacetone phosphate (DHAP) (nonphosphorus-linked carbon) in the course of its incorporation into glyceryl ethers. Accordingly, a mixture of [1,3-3H2] DHAP and [I ,  was prepared. In a microsomal enzyme system from Tetrahymena jyriformis which synthesizes 0-alkyl lipids, it was found that there was a loss of hydrogen from carbon 3 of glyceryl ethers synthesized from double-labeled DHAP.
The loss was quantitatively consistent with the Iabilization of one of the hydrogens linked to carbon 3 of DHAP and was not due to isomerase activity.
Previous work has shown that the alkyl side chain of glyceryl ethers is derived from long chain alcohols (l-3).
Subsequently, we found that glycerol used in glyceryl ether synthesis loses hydrogen from the carbon which forms the ether bond (4). This loss of hydrogen did not occur when glycerol was used in glyceride synthesis.
It was concluded that either S-phosphoglyceraldehyde or dihydroxyacetone phosphate was the source of glycerol in glyceryl ether synthesis. This view was supported by work in cell-free systems in other laboratories (5-8).
Since the rapid equilibrium reaction between these triose phosphates would produce rapid loss of some of the hydrogen from carbon 3,l the obligatory intermediate for the glycerol source was uncertain. Subsequently, studies by Hajra (9) pointed to dihydroxyacetone phosphate as an obligatory intermediate.
There is now good evidence that glyceryl ethers are synthesized from dihydroxyacetone phosphate and long chain alcohols and that there is * This work was supported by Research Grant 5 ROl-HE 12020 from the National Heart Institute, United States Public Health Service.
$ Recipient of Public Health Service Career Development Award 5 K04-HE 08306. 1 The unphosphorylated primary carbinol carbon is designated as carbon 3 of dihydroxyacetone l-phosphate. Carbon 3 of dihydroxyacetone l-phosphate corresponds to carbon 1 of the precursor sn-glyceryl-3-phosphate and carbon 1 of S-phospho-nglyceraldehyde. a requirement for ATP, coenzyme A, magnesium, and NADPH (10-12).
It has also been shown that the oxygen of l*O-labeled hexadecanol is incorporated into O-alkyl lipids (13). In addition, evidence has been presented which indicates that cetyl alcohol and acyl dihydroxyacetone phosphate may form glyceryl ethers in the absence of coenzyme A (14).
Because it has been suggested that glyceryl ether synthesis might occur from an enolized form of dihydroxyacetone phosphate or from a hemiacetal, it would be important to know if the hydrogens of carbon 3 of DHAP2 are labilized in the process of 0-alkyl ether bond formation.
The present study indicates that there is in fact a loss of hydrogen and that this loss is not the result of triose phosphate isomerase activity.
Hydrazine hydrate was obtained from Mallinckrodt Chemical Works. CAP, an inhibitor of triose phosphate isomerase, was prepared as described by Hartman (15). CAP completely inhibited both triose phosphate isomerase, which was obt#ained commercially and which was present in microsomes ( Fig. 1

temperature.
A solution of 0.2 M glycine (5 ml) containing 2.06 g of hydrazine hydrate was adjusted to pH 9.8 with concentrated HCI and added to the glycerol, together with ATP, 60 pmoles; MgCL, 12 pmoles; NAD, 22 pmoles; glycerophosphate dehydrogenase, 0.3 mg; and glycerokinase, 0.04 mg. Incubation was carried out for 2 hours at 37" with shaking.
The mixture was then treated with two 0.5-g portions of charcoal. The charcoal was washed with a small amount of water. The incubation mixture and water were combined and filtered. The hydrazine was removed by extracting the incubation mixture three times with benzaldehyde and then three times with ethyl ether. The extracts were combined and re-extracted once with a small amount of water.
The water was added to the incubation mixture and extracted once more with ethyl ether. One drop of 5% human serum albumin was added, and the proteins were precipitated with 0.5 ml of 50% trichloroacetic acid and removed by centrifugation.
The mixture was then adjusted to pH 6 with the use of phenol red as an indicator, and 0.25 ml of 1 M CaClt was added.
A small amount of precipitate which formed immediately was removed by centrifugation. Ten volumes of acetone-ethanol (1: 1) were added, and the mixture was kept at -20" overnight.
The precipitate was removed by centrifugation, dried with nitrogen, and mixed with 1 ml of a slurry of Dowex AG 5OW-X8 (100-200 mesh, hydrogen form). This was added to a l-ml column of the same resin in the sodium form and eluted with water. The eluate was lyophilized and dissolved in 5 ml of water.
The yield was approximately 40%. The composition of the mixture was examined by thin layer chromatography on cellulose layers 0.1 mm thick, developed in a mixture of tert-butyl alcohol, 60 ml; water, 30 ml; and p-toluene sulfonic acid, 2 g. In this system, there was poor separation of cr-glycerophosphate from DHAP (17). There was, however, good separation of a-glycerophosphate and DHAP from 3-phosphoglyceraldehyde.
High voltage electrophoreris at pH 1.5 with either 0.1 M oxalate buffer (18) or 8% aqueous formic acid gave good separation of a-glycerophosphate and DHAP, but 3phosphoglyceraldehyde migrated at the same rate as oc-glycerophosphate.
With the use of these procedures, it was found that the double-labeled DHAP contained about 25% cr-glycerophosphate. Further isolation and purification of labeled DHAP were accomplished by high voltage electrophoresis in 8% formic acid. As shown in Fig. 2 (top), the purified material gave a single peak on thin layer chromatography. A single peak was also obtained on high voltage electrophoresis at pH 1.5 in 0.1 M oxalate buffer and at pH 6 in 0.1 M sodium bisulfite.
It was also observed that the 3H:14C ratio of the purified DHAP was 7.6, whereas the starting glycerol had a ratio of 9.6. This suggested that there was a loss of tritium from DHAP during its synthesis.
The problem was investigated by incubating a mixture of [I, glycerol and [l , glycerol in the glytine hydrazine buffer used to synthesize DHAP. A tritium loss was indicated by a decrease in 3H :r4C ratio in an aliquot of the dried mixture after incubation.
The tritium which was lost was recovered in water distilled from the incubation mixture. A smaller loss of tritium was noted in the formation of a-glycerophosphate in triethanolamine in the absence of hydrazine hydrate. 3. High voltage electrophoresis of labeled DHAP and derivatives.
A. f3H. 14ClDHAP and 13H. l%la:-glvceronhosnhate were separated by 'high voltage electrophoresis-ih 0.1 A~ oxalate buffer, pH 1.5. The radioactive peaks were located with a strip scanner and with appropriate standards.
The radioactive areas were then cut into 4-mm strips and counted in 2: 1 phosphor-Triton. B, [3H, l%]DHAP and [3H, 14C]or-glycerophosphate were treated with a-glycerophosphate dehydrogenase and NADH in triethanolamine buffer 0.3 M, pH 7.6. The reaction was stopped with trichloroacetic acid and the products were separated as in A. C, the procedure which was used in B was repeated.
After incubation with a-glycerophosphate dehydrogenase and precipitation of enzyme with trichloroacetic acid, an aliquot was separated by electrophoresis with added [3H,14C]DHAP. dehydrogenase and NADH to convert the DHAP to L-cy-glycerophosphate.
The products of the incubation were separated by high voltage electrophoresis in 0.1 M oxalate buffer. The disappearance of the DHAP peak and the appearance of an ar-glycerophosphate peak established with certainty the identity of the former (Fig. 3).
Liquid scintillation counting of aqueous solutions was done by the method of Patterson and Greene (19) with 2 parts of toluene phosphor and 1 part of Triton X-100.
In order to insure reproducibility, the amount of aqueous material to be counted was maintained at 0.5% by volume of the total counting solution. Tritiated and 14C-labeled toluene were used as internal standards to determine absolute activities when necessary.
Preparation of Microsomes-Tetrahymena pyrijormis, strain HSM, was grown in 6 liters of medium with aeration at room temperature.
After 4 days of growth, the cells were harvested by centrifugation at 1,000 to 1,500 rpm for 2 min at 10" in a a-liter rotor. The cells were washed once with distilled water and concentrated again. The volume of the cells was measured after centrifugation in graduated centrifuge tubes and 5 volumes of ice-cold buffer, containing 0.10 M phosphate buffer, pH 7.2; 0.04 M sodium fluoride, 0.25 M sucrose, and 0.005 M mercaptoethanol were added. The cells were then disrupted by sonication at 2" in an ice bath by three bursts of 15-set duration each, applied to about 50 ml of solution at a time. Between bursts the mixture was allowed to cool and the extent of cell disruption was examined microscopically.
The material was then centrifuged at 8,000 rpm at 2" for 20 min. The resulting supernatant was centrifuged at 100,000 X g for 1 hour to sediment microsomes.
The microsomes were washed three times in the same buffer, made up to a final volume of one-half of the volume of the original sonicated mixture, and stored at -70".
At the end of 6 weeks of storage there was little loss of activity.
Phospholipids were developed in chloroform-methanolacetic acid-water (50 : 25 : 7 : 3). Triose phosphates were separated on cellulose layers (Brinkmann) 0.1 mm thick, developed in a mixture of tert-butyl alcohol, 60 ml; water, 30 ml; and p-toluene sulfonic acid, 2 g. Silicic acid column chromatography was done on a column (1 cm in diameter) containing 1 g of silicic acid. Neutral lipids were eluted with ethyl ether and phospholipids with methanol.
Chemical Treatments and Other Methods-Lipids were treated with LiA1H4 at room temperature as described by Thompson (22), and glyceryl ethers produced were converted to their isopropylidine derivatives by the method of Hanahan, Ekholm, and Jackson (23). Glyceryl ethers were cleaved at the ether bond with hydrogen iodide, and the alkyl iodides produced were separated by thin layer chromatography as previously described (1). Reduction with sodium borohydride was carried out as described by Hajra and Agranoff (24). Less than 1 pmole of lipid was dissolved in 0.1 ml of ethanol and 0.02 ml of 0.1 M NaBH4 was added. After 15 min at room temperature an additional 0.01 ml of 0.1 M NaBH4 was added. After an additional 15 min 4 ml of chloroform-methanol-concentrated HCl (20 : 10 : 0.1,

Issue of
X. J. Friedberg, A. Heifetz, and R. C. Greene 5825 v/v) were added followed by 1 ml of water. The upper phase was removed and the lower phase was washed with simulated upper phase. The lower phase was then dried and stored in benzene.
Phosphorus was determined by the method of Bartlett (25) and proteins by the method of Lowry et al. (26).
Periodate cleavage of glyceryl ethers between vicinal hydroxyl groups (27) or of DHAP between the carbon atoms 1 and 2 was performed as follows.
Glyceryl ethers in 1 ml of ethanol or DHAP in 1 ml of water were mixed with a a-fold molar excess of 0.05 N sodium metaperiodate in 25% acetic acid and left in the dark for 30 min with occasional shaking.
At times, 18 pmoles (2 ~1) of carrier formaldehyde were added before the periodate to minimize oxidation of labeled formaldehyde.
The reaction was stopped by addition of a IO-fold excess of 0.5 N sodium arsenite and the long chain aldehydes were extracted with ethyl ether. The water phase was then brought to pH 6 with 0.1 N NaOH and 15 ml of saturated aqueous dimedon were added (approximately 0.47,).

Incorporation
of [I ,~Hz, 1 ,S-W22]Dihydroxyacetone Phosphate into 0-Alkyl Lipids-The criterion for the formation of 0-alkyl lipids in this study was the chromatographic isolation of labeled glyceryl ethers and their conversion to isopropylidene derivatives. Thus, lipids extracted after incubation were treated with LiAlH( and separated by thin layer chromatography. Two peaks are apparent (Fig. 4): one coincides with glyceryl ether, and another is an unknown material remaining at the origin. The identity of the glyceryl ethers was established by conversion to their isopropylidene derivatives followed by thin layer chromatography (Fig. 4).
It was also considered of importance to examine the products of [l , 3-3H2, 1, 3-i4Cs]DHAP incubation as compared with the products obtained after incubation of [1-i4C]hexadecanol prior to further treatment with other agents. Accordingly, phospholipids and neutral lipids were first separated by silicic acid column chromatography.
Material obtained after the incubation of labeled DHAP and labeled hexadecanol was then compared by thin layer chromatography.
In the neutral lipid fraction, a peak corresponding to the 0-alkyl dihydroxyacetone described by Snyder, Wykle, and Malone (5) was found. This material gave a peak with the mobility of glyceryl ether after t,reatment with sodium borohydride and rechromatography in the same system. In the phospholipid fraction, a number of peaks were found which we have not identified.
It was of interest to note, however, that the several radioactive peaks derived from labeled DHAP and labeled hexadecanol corresponded, although the relative amounts differed.
In order to obtain further evidence that the isotope incorporated into glyceryl ethers came from DHAP and not from an unidentified radioactive contaminant, the effect of added unlabeled DHAP was evaluated.
The addition of increasing concentrations of DHAP from 0.0133 to 1.2 mM yielded increasing amounts of glyceryl ethers. A plot of the data according to the method of Lineweaver and Burk suggested typical first order enzyme kinetics.
The results are shown in Fig. 5. The addition of increasing amounts of unlabeled DHAP caused a progressive decrease in the amount of radioactivity incorporated. This is mm FIG. 4. Thin layer chromatography of 0-alkyl lipid derivatives obtained from microsomal enzyme system. Bottom, [3H ,"C]DHAP (6,950,OOO dpm 3H and 909,000 dpm I%, 0.02 rmole) was incubated with 1 ml of microsomes free of isomerase activity (2 mg of protein), 10 mM ATP, 2.7 mM Mg++, 0.67 mM CoA, and 0.067 mM cetyl alcohol in 1% Tween 80, in a total volume of 1.5 ml. The lipids were extracted as described in the text, treated with LiAlH4, and separated by thin layer chromatography on Silica1 Gel G and developed with chloroform-methanol-water (90:5: 0.5). The bar indicates the position of the standard.
Top, glyceryl ethers purified by thin layer chromatography were converted to their isopropylidene derivatives and identified by thin layer chromatography on Silica Gel G (developed in ligroine-ethyl-acetic acid (90: 10: 1)). The lipids were extracted and glyceryl ethers were isolated by thin layer chromatography as described in Fig. 4. The results were plotted in terms of aH activity and amount of glyceryl ethers formed.
characteristic of dilution of the labeled precursor and substantiates our previous characterization of the substance as DHAP. Analysis of Tritium Activity and Its Distribution on 0-Alkyl Lipids-In accordance with the principal objective of this study, a series of experiments was carried out to determine whether or Glyceryl Ether Synthesis Vol. 246,No. 18 not any of the hydrogens of the ether-linked carbon of DHAP are labilized in the process of 0-alkyl lipid formation.
Incubat,ion was carried out for 3 hours at 30". The lipids were extracted by the method of Bligh and Dyer (20) and stored in benzene.
The aqueous phase containing residual DHAP and other materials was filtered, dried on a rotary evaporator, and dissolved in 5 ml of water.
An aliquot of the lipid sample was treated with LiAlH4 and purified by thin layer chromatography after addition of 5 mg of carrier tetradecyl glyceryl ether. Rechromatography of this material gave a single peak of radioactivity.
The 3H:14C ratio (dpm) of this material was 7.1, which is not significantly different from that of the starting DHAP.
The results of the degradation experiments described below are therefore unexpected.
The procedure was repeated several times with the same and different batches of labeled DHAP and different microsomes. The results were similar in each instance.
That is, there was a loss of tritium from the ether-linked carbon.
There was also a substantial increase in 3H:14C ratio of carbon 1. The results of five periodate cleavages are summarized in Table II. After treatment of glyceryl ethers with hydrogen iodide, it was found that the alkyl residue contained no activity.
An examination of the DHAP remaining after incubation was carried out to determine if an explanation for the increase in 3H :  I 14C ratio of carbon 1 in glycerolipids would become apparent. Thin layer chromatography on cellulose with tert-butyl alcoholwater-p-toluene sulfonic acid (60:30:2, v/v/w) revealed the appearance of several unidentified new tritium poor peaks and an increase in relative tritium activity of the remaining DHAP (Fig. 2). In addition, no activity in 3-phosphoglyceraldehyde was detected, indicating complete inhibition of triose phosphate isomerase by CAP. The peak corresponding to DHAP was eluted.
This material showed a relative tritium enrichment as compared with the starting DHAP (Table II).
A fraction of the water phase remaining after incubation was also treated with periodate, and the formaldehyde containing the carbon 3 of DHAP was precipitated with dimedon. The 3H:14C ratio of formaldehydodimedon was 10.3, indicating that there was no loss of tritium from the ether-linked carbon of residual DHAP during the course of the incubation.
The possibility that either LiAlH4 or NaBH4 reduction of Oalkyl lipids might have resulted in tritium loss was considered. This was ruled out by the fact that there was no change in 3H :r4C ratio when the dihydroxyacetone ether formed in the course of incubation was treated with these agents and converted to glyceryl ether. DISCUSSION We have previously indicated that we could visualize the possibility of a mechanism whereby glyceryl ether synthesis proceeds via the formation of an alkyl glycoside (4). This implies the synthesis of a hemiacetal as an intermediate.
Others have suggested similar mechanisms, such as a DHAP-CoA complex, an aldehydrogenic form, that leads to the formation of 0-alkyl dihydroxyacetone phosphate through a hemiacetal intermediate (5, 10). It has also been proposed that CoA might form a thioester with DHAP which could then react with fatty alcohol to form 0-alkyl dihydroxyacetone phosphate and that an enol form of DHAP might be involved in these mechanisms (10). The recent work of Hajra (9) indicates that acyl dihydroxyacetone phosphate may be an intermediate and points to some interesting possibilities.
The data reported here reveal the loss of 1 hydrogen from carbon 3 of DHAP in the course of glyceryl ether synthesis. This finding could be explained by the formation of an enediol, possibly of acyl dihydroxyacetone phosphate. However, several other mechanisms can be envisioned, and the data are not sufficient to draw firm conclusions at this time. Nevertheless, any explanation of the mechanism of 0-alkyl bond formation may have to take three facts into considerat'ion.
These are the retention of the oxygen of hexadecanol, the preferential utilization of acyl dihydroxyacetone phosphate, and the loss of hydrogen from the ether-linked carbon of DHAP.
The results indicate that somewhat more than half of the tritium of carbon 3 of DHAP is lost in the process of 0-alkyl lipid synthesis.
If 3-phosphoglyceraldehyde were an intermediate and had appeared as a result of isomerase activity, a tritium loss would also have occurred.
However, CAP completely inhibited the isomerase present in the microsomal system, and labeled 3phosphoglyceraldehyde did not appear in the course of incubation as determined by thin layer chromatography.
Although the loss of tritium from carbon 3 of the glycerol moiety of glyceryl ethers can be mechanistically explained, the reason for enrichment of trit.ium in carbon 1 of the glyceryl ethers and in the residual DHBP is not so clear cut. An artifact which