Ribose. An oxidation product of glucose 6-phosphate in microsomal fraction.

An oxidative metabolism of glucose 6-phosphate was studied in rat liver microsomal fraction. Although radioactive 14CO2 was formed from [1-14C]glucose 6-phosphate in the microsomal fraction (Hino, Y., and Minakami, S. (1982) J. Biochem. (Tokyo) 92, 547-557), the formation was negligible when [2-14C]glucose 6-phosphate was used as a starting substrate. These results indicated an inability of the microsomal fraction to rearrange [2-14C]glucose 6-phosphate to form [1-14C] glucose 6-phosphate, and it was expected that a certain compound derived from glucose 6-phosphate accumulated as an end-product of the reaction. We, therefore, have tried to identify the product by high performance liquid chromatography, and found that ribose accumulated as the end-product. The formation of ribose was inhibited in the same manner as that of 14CO2 by antibodies against rat liver microsomal hexose-6-phosphate dehydrogenase, and the ratios of ribose to 14CO2 formed in the reaction were 0.5-0.8 on a molar basis. The finding of ribose formation further suggested the involvement of ribose phosphate isomerase and phosphatase activities in the reaction.

concerned with further metabolism of G6P in the microsomal fraction. In this communication, we present results which suggest that the microsomal fraction was unable to rearrange [2-I4C]G6P to form [1-14C] G6P, and that ribose accumulated as an end-product of the reaction. This is the first report to demonstrate the formation of ribose during the G6P oxidation in microsomes.
EXPERIMENTAL PROCEDURES Subcellular Fractions-Microsomal and cytosol fractions were prepared as described before (3,4), except that 55%' sucrose layer was placed at the bottom of the centrifuge tube to remove glycogen particles from the microsomal fraction.
Analytical and Assay Methods-Protein concentrations were determined by the method of Lowry et al. (6) with bovine serum albumin as a reference. The reaction mixture for determining the ribose formation was essentially similar to that for the "CO, formation (5). It contained (final volume, 0.2 ml) 50 mM HEPES buffer (pH 7.5), 25 mM nicotinamide, 0.6%. Emulgen 913, 2 mM ATP, 2 units of hexokinase, 2.5 mM NADP+, 10 mM glucose, and an appropriate amount of an enzyme source. When radioactive "Cor was to be determined, trace amount of [l-14C]glucose, [2-"C]glucose, or [6-"C] glucose was included in the reaction mixture.
For identification of the dead-end product of G6P oxidation in the microsomes, the reaction was conducted as described above using 16-"Clglucose (4 pCi), stopped by adding 5% trichloroacetic acid, and an aliquot of the clear supernatant fraction obtained by centrifugation was added with an equal volume of ethanol solution of dansyl hydrazine (10 mg/ml) ( 7 , 8 ) . The dansylation was carried out by incubating the mixture for 2-3 h at 30 "C. Portions of the mixtures (2 pl) were analyzed for the dansylated derivatives by HPLC (Hitachi 635A liquid chromatography), which was performed on a Merck LiChrosorb KP-18 column (particle size, 5 pm) at room temperature (20-25 "C) by increasing the concentration of acetonitrile from 5 to 30%' (v/v) linearly at a flow rate of 1.0 ml/min (column pressure, 120-150 kg/ cm"). The gradient was made with the solvent programmer by mixing two stock solutions of 0.2 mM KHrP04 and 50% (v/v) acetonitrile in 0.2 mM KH2POI. Addition of low concentration (0.1-0.4 mM) of KH2P0, to the aqueous acetonitrile solution was essential to separate both phosphorylated and nonphosphorylated sugars on the column. The elution profiles were monitored by absorbance changes at 254 nm and by radioactivity. When measured by absorbance changes, the peak heights made by the dansylated derivatives of phosphorylated sugars were smaller than those made by the derivatives of the corresponding nonphosphorylated sugars; e.g. the peak heights for ribulose 5-phosphate and ribose 5-phosphate were about b and VI of those of ribulose and ribose, respectively. For determination of ribose formed, 1-4 mM of an authentic o-ribose was included in the incubated mixture as an internal standard.
Effects ofAntibodies-The antibodies against rat liver microsomal hexose-6-phosphate dehydrogenase raised in a rabbit were shown to be specific for the antigen (3).
A mixture containing microsomes, HEPES buffer (pH 7.5), Emulgen 913, and antibodies (or unimmunized globulin fraction) was preincubated for 30 min at 30 "c. For determination of the remaining activities of "COS and ribose formation, ATP, hexokinase, glucose (in case of I4COy formation, [~-'~c ] glucose (20 d p m b m o l of glucose) was included), nicotinamide, and NADP' were added to make the constituents of the mixture the Same as those for the assays as described above.

RESULTS AND DISCUSSION
14C02 Was Not Formed with (2-I4C/G6P as Substrate-We have determined, by measuring the formation of radioactive 14C02, whether the microsomes have an activity to rear-

Ribose Formation in Microsomes
range [2-'4C]G6P to form [1-14C]G6P. The rationale was that, once [1-14C]G6P was formed, the l-I4C would be liberated as 14C02 by the actions of hexose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase present in the microsomes. The results shown in Fig. 1 indicate that, though significantly large amount of I4CO2 was formed with [2-14C] G6P in the cytosol fraction, it was only negligible when the microsomal fraction was employed as an enzyme source. The molar ratios of '*CO2 formed with [2-I4C]G6P to that formed with [1-14C]G6P, which could be taken as a conventional measure of the rearranging activity, were about 0.23 for the cytosol fraction and 0.01 for the microsomal fraction (determined after 10-h incubations). It could be concluded that the microsomes were unable to form [ Identification of Ribose-Considering the possibility that the microsomes might be deficient in the activities necessary for converting [2-'4C]G6P to [1-I4C]G6P, we could imagine that a certain compound derived from G6P might accumulate as a dead-end product during the reaction. In order to identify the product, the reactions were conducted with [6-I4C]G6P and the dansylated products were analyzed by the HPLC as shown in Fig. 2.
Although several absorption peaks appeared, only three major peaks (designated a, b, and c ) could be detected when radioactivity was used for the detection. Peak c decreased below the detectable level when either NADP' or microsomes was omitted from the complete system. Zero time control did not give rise to peak c. As shown in Fig. 3, where absorbance changes were used for the detection, a peak corresponding to peak c shown in Fig. 2 increased significantly as a function of incubation time. Furthermore, peak c increased in proportion to the amount of microsomal protein and the increase was inhibited by adding 2.5 mM p-chloromercuribenzoate or antihexose-6-phosphate dehydrogenase antibodies (see Fig. 5) or by heating the microsomes at 95 "C for 3 min. Taken together with other control experiments (not shown), we could conclude that peak c was the dansylated derivative of the deadend product. Peak c was not formed when one of the components necessary for generating G6P was omitted from the   [6-'4C]glucose, NADP+, ATP, and hexokinase), and the reactions were carried out for 13.5 h at 30 "C. The HPLC elution profdes of the dansylated samples were monitored by absorbance changes (broken line, arbitrary unit) and radioactivity (solid line, arbitrary unit). There appeared three major radioactive peaks designated a, 6, and c, of which peak c could be seen only in the complete system. Peaks b and c corresponded to the dansylated derivatives of glucose and ribose, respectively, and peak a, in this particular case, mostly of G6P. reaction mixture, so G6P formed by the hexokinase reaction might be the substrate for the reaction. Several sugars mentioned in Fig. 3 could be distinguished on the chromatogram, and we found that peak c had a similar elution time with an authentic dansyl ribose. Further, the dansylated reaction mixture and the authentic dansyl ribose were combined together and the resulting mixture was subjected to HPLC, thus making it highly likely that the unknown peak e corresponded to dansyl ribose. The derivatives of phosphorylated sugars were eluted at retention times between 6 and 9 min, and those of nonphosphorylated sugars between 14 and 20 min (Fig. 3). The radioactive peaks observed in the front of the chromatograms might represent polar compounds not having the dansyl group.
Relationship between l4 COZ a n d Ribose Formed-We have attempted to determine the amount of ribose with respect to l4CO2 formed in the reaction. For this purpose we have used

Ribose Formation in Microsomes 1417
product suggested the involvements of ribose phosphate isomerase (~-ribose-5-phosphate ketolisomerase, EC 5.3.1.6) and phosphatase activities in the reaction. When the microsomes were incubated with ribulose 5-phosphate, a peak corresponding to ribose appeared in addition to a ribulose peak (Fig. 6). The ribose peak was not detected, however, when ribulose was used in place of the phosphorylated compound (not shown). It could be explained as that ribulose 5-phosphate was first isomerized by the isomerase to form ribose 5-phosphate, which was then converted into ribose by a phosphatase present in the microsomes. The isomerase activity might be inhibited by p-chloromercuribenzoate, because the drug inhibited the appearance of the ribose peak, whereas the ribulose peak was less sensitive to the SH reactive drug. The contribution of cytosol enzymes to the observed activity ultraviolet absorption rather than radioactivity as a detection method, because relatively high background values observed in the latter method might render the results less reliable. Fig.  4 (left) shows that the height ofpeak c increased biphasically in the presence of NADP+, and the increase in the slow phase was similar to that observed in the control experiments carried out in the absence of NADP+. The NADP+-independent increase ofpeak c might not be due to the product of interest, because no appreciable radioactivity could be detected in this region when [6-14C]G6P was used as substrate and the radioactivity was employed for the detection (see Fig. 2, -NADP+). Therefore, we corrected the results by subtracting the NADP+-independent increases of the peak heights as blank from the NADP'-dependent ones (Fig. 4 (left), broken line). As shown in Fig. 4 (right), the formation of 14C02 occurred rapidly and linearly for the initial 60 min and reached the maximum at about 2 h. On the contrary, the ribose formation was slower than the 14C02 formation and reached the maximal level after a 6-h incubation. The maximal level of ribose formed was consistently lower than that of I4CO2 formed on a molar basis. We could estimate the molar ratios of ribose to I4CO2 accumulated at the maximum to be 0.5-0.8, though these ratios varied significantly from experiment to experiment. Fig. 5 shows that the 14C02 formation was inhibited in the same manner as the ribose formation by anti-hexose-6-phosphate dehydrogenase antibodies. The similarity of the inhibition patterns and the observation that ribose was formed much slower than 14C02 might support our proposal thatpeak c, that was ribose, was the end product of the reaction.
Involvements of Ribose Phosphate Isomerase and Phosphatase Activities-The formation of ribose as a final reaction  seemed to be negligible, because the microsomes used in these experiments were extensively washed and the specific activity of lactate dehydrogenase, a marker enzyme for the cytosol fraction, detected in the microsomal fraction was less than 0.2% of that detected in the cytosol fraction. Moreover, the results shown in Figs. 1-4 could hardly be explained by the cytosol contamination; rather they seemed to be an indication that the cytosol contamination was minimum.

In in Microsomes
General Comments-We have presented results which indicate the formation of ribose as an end-product of G6P oxidation in the microsomes, though we are not sure whether the results can be applied to the metabolism in situ. These results are in favor of our previous report (5) which indicated that hexose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase present in the microsomal vesicles catalyzed the formation of reduced NADP' through the oxidation of G6P, which was presumably transported from the cytoplasm uia the G6P specific transporter (9)(10)(11)(12). Ribose, the product of the reaction, seems to permeate the membrane barrier without any special transporting devices because of small size and having no charged groups on the molecule (13,  14), and so can leave the luminal space of the microsomal vesicles without much difficulty. No accumulation of the product within the vesicular compartment can make it possible for the enzyme system to continue to generate reduced NADP+ in response to demand for it.