J. Biochem. 115, 381-386 (1994)

Reaction Mechanism Underlying CMP-N-Acetylneuraminic Acid

Hydroxylation in Mouse Liver: Formation of a Ternary Complex of

Cytochrome b5, CMP-N-Acetylneuraminic Acid, and a Hydroxylation

Enzyme'

Hiromu Takematsu,* Takehiro Kawano,** Susumu Koyama,* Yasunori Kozutsumi,* Akemi Suzuki,** and Toshisuke Kawasaki*

*Department of Biological Chemistry, Faculty of Pharmaceutical Sciences, Kyoto University, Sakyo-ku, Kyoto 606;

and **Department of Membrane Biochemistry, The Tokyo Metropolitan Institute of Medical Science, Bunkyo-ku, Tokyo 113

Received for publication, September 10, 1993

We have proposed that CMP-N-acetylneuraminic acid (CMP-NeuAc) hydroxylation is mediated by an electron transport system consisting of cytochrome b5 (b5), b5 reducing factor(s), and CMP-NeuAc hydroxylase, all of which have been detected in the cytosolic fraction of mouse liver [Kozutsumi, Y., Kawano, T., Yamakawa, T., & Suzuki, A. (1990) J. Biochem. 108, 704-706]. In order to elucidate the reaction mechanism underlying CMPNeuAc hydroxylation, the interaction between b5 and the hydroxylase was studied using a b5-immobilized affinity column. The enzyme activity was retarded on the b5 column in the presence of the substrate, CMP-NeuAc, but not in the presence of the reaction product, CMP-N-glycolylneuraminic acid (CMP-NeuGc). These findings suggest that the binding of CMP-NeuAc to CMP-NeuAc hydroxylase changes the conformation of the enzyme so as to construct a recognition site for b5, followed by the formation of a ternary complex through this domain. Then the transport of electrons from NAD(P)H to the enzyme through b5 takes place, CMP-NeuAc is converted to CMP-NeuGc, and finally the ternary complex dissociates into its components to release CMP-NeuGc. It is known that a soluble form of b5 is abundant in erythrocytes and is synthesized from a mRNA different from that for the microsomal form of bs. In order to determine the origin of b5 detected in the cytosolic fraction of mouse liver, the molecular forms of b5 mRNA expressed in mouse liver were analyzed. The polymerase chain reaction, with mouse liver cDNA and specific primers for the respective forms of b5 mRNA, detected only the mRNA encoding the microsomal form in mouse liver. These results suggest the possibility that b5 involved in the ternary complex formation in the cytosolic fraction of mouse liver originates from the microsomal form, which was solubilized after its synthesis. Possible participation of the microsomal b5 on the endoplasmic reticulum in the hydroxylation reaction was also suggested by the results of an in vitro reconstitution assay involving isolated microsomes.

Key words: CMP-N-acetylneuraminic acid hydroxylase, cytochrome b5, N-glycolylneuraminic acid, Hanganutziu-Deicher antigen.

Sialic acid is an important component of glycoconjugates which participate in various biological functions. Namely, sialyl Le' is the determinant recognized by P-selectin and E-selectin, which mediate the interaction between platelets or endothelial cells and leukocytes (1, 2). Sialic acid is also suggested to be involved in the ligand-receptor recognition of L-selectin, a homing receptor of lymphocytes (3-5). Furthermore, sialylated glycolipids, gangliosides, induce neurite outgrowth of human neuroblastoma cell lines (6, 7) and differentiation of human myeloid cell lines (8-11).

Despite these lines of evidence, the biological significance of the occurrence of multiforms of sialic acids is not clear yet. However, there are several interesting lines of evidence concerning modification of the C-5 amino groups of sialic acids, i.e., N-acetylneuraminic acid (NeuAc) and N-glycolylneuraminic acid (NeuGc). The level of expression of NeuGc-containing glycoconjugates changes during the development of rat and piglet small intestine (12-14) and bovine fetal tissues (15), is associated with altered

growth of mouse lymphoma cells (16), and affects allotransplantation of mouse mammary tumor cells (17). Polymorphic expression of NeuGc-containing gangliosides was reported in erythrocytes of dog (18, 19) and cat (20). In mouse erythrocytes, ganglioside GM4 contains NeuAc but not NeuGc, while GM2 and GM1 contain NeuGc, suggesting the presence of mechanisms which determine ganglioside

' This work was supported by a Grant-in-Aid for Scientific Research

from the Ministry of Education, Science and Culture of Japan. Abbreviations: b5, cytochrome b5; ER, endoplasmic reticulum; GM1, Gal,81-3Ga1NAcj31-4[Siaa2-3]Galftl-4G1cf31-1'Cer; GM2, GalNAcQ14[~-,z~a2-3]Galf31-4G1cf31-1'Cer; GM4, Siaa2-3Galll-1'Cer; Neu N-acetylneuraminic acid; NeuGc, N-glycolylneuraminic acid.

ol. 115, No. 3, 1994 381

382 H . Takematsu et al.

species-specific sialic acids (21). In man, the expression of NeuGc is hardly detectable under normal conditions. When humans are treated with animal anti-sera, the Hanganutziu-Deicher (HD) antibody, that recognizes NeuGc in

glycoconjugates, will be produced (22, 23). NeuGc-containing glycoconjugates are also detected in colon cancer tissues

(24), melanoma tissues (25), retinoblastoma cells (26), breast cancer tissues (27), other malignant tumors (28, 29), and fetal tissues (24). Therefore, NeuGc-containing

glycoconjugates are oncofetal antigens in man. NeuGc is produced from NeuAc through the hydroxylation of CMP-NeuAc by a monooxygenase (30, 31). We have reported that CMP-NeuAc hydroxylation is carried out by an electron transport system consisting of cytochrome bs (b5), b5 reducing factor(s) and a hydroxylase as the terminal enzyme (32, 33), as shown below. All the components are detected in the cytosolic fraction of mouse liver, in which a large amount of NeuGc-containing glycoconjugates was detected (34).

Here, X and Y are putative NADH and NADPH-dependent b5 reductases, respectively. It has also been shown that the activity of the hydroxylase in several tissues is well correlated with the NeuGc content, indicating that the hydroxylase plays a key role in the regulation of NeuGc expression (35).

In CMP-NeuAc hydroxylation, the hydroxylase accepts electrons from bs and converts CMP-NeuAc to CMPNeuGc, implying the formation of an intermediate complex consisting of the hydroxylase, b5, and the substrate.

In this paper, we report evidence suggesting that the formation of a ternary complex consisting of the hydroxylase, b5, and CMP-NeuAc occurs during the hydroxylation reaction. In addition, the membrane form of bs may be involved in the hydroxylation of CMP-NeuAc in tissues such as mouse liver in which the soluble form of b5 is virtually absent.

EXPERIMENTAL PROCEDURES

Materials-Sepharose 4B and DEAE-Sepharose CL-4B were purchased from Pharmacia (Uppsala, Sweden). CMP-N-acetylneuraminic acid (CMP-NeuAc) was obtained from Kanto Chemicals (Tokyo), and CMP-N-glycolylneuraminic acid (CMP-NeuGc) was a gift from Drs. M. Ito and K. Tomita (MECT, Tokyo). Preparation of the cytosolic fraction and microsomal fraction from mouse liver was described previously (32). As the b5 reducing factor(s), named factor X, the cellulose phosphate P11 (P11) column bound fraction from the mouse liver cytosolic fraction was prepared as described previously (35). The P11 column unbound fraction from the mouse liver cytosolic fraction, which contains the hydroxylase activity and is free from b5 reducing factor(s) and b5 (35), was used for the further purification of the CMP-NeuAc hydroxylase on a DEAE-Sepharose CL-4B column, as described previously (33). The specific activity of the partially purified hydrox

ylase was 15 milliunits/mg protein. The CMP-NeuAc hydroxylase at this stage of purification was used in the following studies. The soluble form of b5 was purified from horse erythrocytes (33). Cytochrome b5 Column Chromatography bs-Sepharose was prepared by mixing the soluble form of horse b5 (5.25 mg protein) with 4.4 ml of BrCN-activated Sepharose 4B (36) in a total volume of 12 ml. After overnight shaking at 4°C, more than 90% of the b5 was attached to the resin. The b5 column (1 ml) was equilibrated with 10 mM Tris-HCI buffer, pH 7.5, containing 0.1 mM DTT. The partially purified CMP-NeuAc hydroxylase (0.6 milliunit in 100,u 1) was applied onto the b5 column at the flow rate of 0.1 ml/ min in the presence of various compounds at 4°C. Fractions (0.2 ml each) were collected and the enzyme activity was measured, after desalting if required. The CMP-N-Acetylneuraminic Acid Hydroxylation

Assay-The activity of CMP-NeuAc hydroxylase was measured in the presence of the soluble form of b, factor X, and NADH, as described in the previous paper (35). For reconstitution assays to determine the contribution of the microsomal b5, the microsomal fraction containing 20 pmol b5 protein and the soluble form of b5 (20 pmol protein) were used. The cytosolic fraction (56.3 pg protein), containing the hydroxylase, b5, and b5 reducing factor(s), was mixed with other factors and the hydroxylation activity was measured using NADH as an electron donor, in a final volume of 50 pl (33). To study the effect of phospholipids on the hydroxylation, the partially purified CMP-NeuAc hydroxylase (14.5 punts), the soluble form of b5 (40 pmol protein), factor X (5.5 pg protein), and the microsomal fraction containing 40 pmol b5 protein were used. Various phospholipids were dissolved in chloroform : methanol (2 : 1) and then added to the reaction tubes (1 mg/tube). After the solvent had been evaporated off, other components were added, followed by sonication for 15 s on ice. To determine the affinity of the soluble form of b5 to the hydroxylation system, various amounts of the soluble form of b5 were added to the mouse liver cytosolic fraction (99.0 pg protein) and then the hydroxylase activity was measured using NADH. The activity, which had been corrected for the control values observed in the absence of the soluble form of bs, were used for Lineweaver-Burk plot analysis. Preparation of RNA from Mouse Tissues-Blood and liver were obtained from 5-week-old BALB/c mice. The blood was collected in heparinized plastic tubes and applied to a cellulose column (37). The flow-through fraction (erythrocyte fraction) contained erythrocytes and reticulocytes, but was free from platelets and leukocytes. RNA was isolated from the tissues by acid guanidium thiocyanatephenol-chloroform extraction (38). Amplification of the Coding Region of Cytochrome b5

mRNAs-For cDNA synthesis, RNA (20 ,u g) was incubated for 60 min at 37°C in a reaction mixture (20 pl) containing 300 units of Molony murine leukemia virus reverse transcriptase (GIBCO-Bethesda Research Laboratories), 15 units of human placenta RNase inhibitor (Wako Pure Chemicals, Osaka), and 50 pmol of oligo(dT). To amplify the coding region of the cDNA of b, the polymerase chain reaction (PCR) was carried out for 30 cycles in a reaction mixture (50 p l) containing 1 p 1 of the above cDNA solution, 2.5 units of Taq DNA polymerise (Promega), and 20 pmol each of sense and antisense primers, which corresponded to

J. Biochem.

Reaction Mechanism of CMP-NeuAc Hydroxylase 383

the nucleotide sequences of mouse b5 cDNA. The primers used for the PCR were 5'-ACTCAGAATTCCGAGATGGCCGGGCAGT-3' (primer 1), 5'-CAGGAATTCCACCAACTGGAATTAGACTCC-3' (primer 2), and 5'-CTGGACACCTTTAAGATTC-3' (primer 3), in which the underlined sequences corresponded to those of mouse b5 cDNA. The reaction products were fractionated by electrophoresis on a 2% agarose gel. The molecular weight markers used were the 100 Base-Pair Ladder (Pharmacia). Determination of the Nucleotide Sequence of the Ampli

fied DNA-The amplified DNA, which had been fractionated by agarose gel electrophoresis, was isolated using a Geneclean-II kit (Bio-101 Inc.). The nucleotide sequences of the isolated fragments were determined by the chain termination method (39) using a Cycle sequencing kit (ABI).

Cytochrome b5 Determination-The amount of the purified soluble form of b5 was determined as described (40). The concentration of the microsomal b5 in the microsomal fraction of mouse liver was determined as described (41). Protein Determination-Protein concentrations were determined by the BCA method of Smith et al. (42) with bovine serum albumin as a standard.

RESULTS

Formation of a Ternary Complex of Cytochrome b5, the Hydroxylase, and CMP-N-Acetylneuraminic Acid in the CMP-N-Acetylneuraminic Acid Hydroxylation ReactionThe interaction between b5 and CMP-NeuAc hydroxylase was analyzed on a b5 immobilized affinity resin. The partially purified hydroxylase, freed from b5 and b5 reducing factors, was loaded on the b5 column. As shown in Fig. 1, elution of the enzyme activity from the column was retarded in the presence of CMP-NeuAc but not in the presence of CMP-NeuGc. The CMP-NeuAc dependent retardation was abolished in the presence of 0.1 M NaCl, which totally inhibits the CMP-NeuAc hydroxylation reaction (33). NaN3 is another inhibitor of CMP-NeuAc hydroxylation and has been considered to act on the hydroxylase itself (33). However, NaN3 did not affect the retardation (data not shown), suggesting that NaN3 does not affect at least the binding of the enzyme to bs, but does affect other

processes. In order to measure the Km for b5, various amounts of the soluble b5 were added to the cytosolic fraction and then the hydroxylation activity was measured using NADH as an electron donor. As shown in Fig. 2, the apparent Km was estimated to be 0.15,uM, showing high affinity of the soluble b5 to the hydroxylation system. On the basis of these findings we propose a scheme for the reaction mechanism, i.e., that a ternary complex consisting of the soluble b5i CMP-NeuAc hydroxylase and CMPNeuAc is formed in the hydroxylation pathway (Fig. 3 and see "DISCUSSION"). Molecular Species of Cytochrome b5 Expressed in Mouse Liver-b5 occurs in two forms, a microsomal form and a soluble form. The microsomal form is ubiquitous in various tissues and the soluble form is only abundant in erythrocytes. It has also been shown that the two forms are encoded by different mRNAs in mammals (43-45). In oral to elucidate the mechanism of the biosynthesis of b5, t i is detected in the cytosolic fraction of mouse liver

involved in the CMP-NeuAc hydroxylation, the

amounts of mRNA for the microsomal and soluble forms were measured. According to our recent results (Takematsu, H., Kozutsumi, Y., and Kawasaki, T., unpublished results), the nucleotide sequence of the mouse soluble b5 cDNA is identical to that of the microsomal form, except for the insertion of a 19 base-pair (bp) sequence in the soluble form. This insertion is located between the catalytic domain and the membrane-bound domain of the microsomal form (Fig. 4A). This insertion introduces a new termination codon following the coding sequence of the catalytic domain, resulting in the expression of a truncated b5, which has lost the membrane domain and therefore has become the soluble form of b5 (Fig. 4A). The inserted sequence is very short and contains a sequence homologous to another part of the coding sequence. Therefore, Northern blot analysis is not practical for distinguishing these two mRNAs. An alternative approach for the detection of each of these mRNAs is polymerise chain reaction (PCR) analysis using cDNAs synthesized from mRNAs. PCR primers were designed, as shown in Fig. 4A. The PCR products with primers 1 and 2 are expected to be 372 by in length for the soluble form and 353 by for the microsomal form. PCR with primers 1 and 3 is expected to produce only a soluble form DNA segment of 322 bp. As shown in Fig. 4B, the only PCR product with the liver cDNA was the microsomal form (lanes 1 and 2). On the other hand, when the cDNA of mouse erythrocytes was used as the template, soluble form segments could be detected as well as the microsomal form segment (Fig. 4B, lanes 3 and 4). All DNA segments amplified were confirmed to be derived from the mouse soluble or microsomal b5 cDNA by DNA sequencing analysis (data not shown). These data demonstrated that mouse erythrocytes contain both the soluble and microsomal b5 mRNAs, while mouse liver contains only the microsomal form of b5. Mouse Liver Microsomal Fraction as a Source of Cytochrome b5 Involved in CMP-N-Acetylneuraminic Acid

Fig. 1. Cytochrome b5 Sepharose chromatography of the CMPNeuAc hydroxylase. The CMP-NeuAc hydroxylase was applied to a cytochrome b5-Sepharose column (1 ml) in the presence of 100 pM CMP-NeuAc (0), 100 pM CMP-NeuGc (C)), and 100 uM CMP-NeuAc and 100 mM NaCI (x). Fractions of 0.2 ml were collected and aliquots thereof were used for monitoring the enzyme activity. The arrow indicates the position of the flow-through fraction on elution of the column.

,1. 115, No. 3, 1994

384 H Takematsu et al.

Fig. 2. Dependence of the CMP-NeuAc hydroxylase activity on the amount of cytochrome b5 added. The mouse liver cytosolic fraction (99.0 ug), containing the hydroxylase, cytochrome bs (b5), and b5 reducing factor(s), was incubated with various amounts of the soluble form of bs. (Inset) Lineweaver-Burk plot of the reaction. The activity increase on the addition of b5 is plotted,

Fig. 3. The proposed reaction mechanism for CMP-NeuAc hydroxylation. Hydroxylase, CMP-NeuAc hydroxylase; b5i cytochrome b5. Open arrows denote the possible sites where electrons are transferred to b5.

Hydroxylation-It has been believed that the electron transport system for CMP-NeuAc hydroxylation consists entirely of soluble components (12, 13, 15, 30-33, 35). However, Shaw et al. reported that the purified microsomal form of b5 can be utilized by CMP-NeuAc hydroxylase in the presence of a detergent (46). In order to determine whether or not the microsomal fraction isolated from mouse liver can function as active b5 for the hydroxylation, the fraction was subjected to in vitro reconstitution assay. The isolated cytosolic fraction contained the hydroxylase, b5 reducing factor(s), and b5. However, the level of bs was very low, and increased activity was detected when b5 was exogenously added (Fig. 2 and Table I). On the other hand, there was no significant increase in activity when the hydroxylase or factor X [ b5 reducing factor(s) ] was added (Table I). When the microsomal fraction was added, the CMP-NeuAc hydroxylase activity slightly increased, but

Fig. 4. (A) The orientation of oligonucleotide primers used for the amplification of mouse cytochrome b5 cDNA. m-b5, the microsomal form of cytochrome b5 (b5); s.b5, the soluble form of b5; white bars, the coding region for the catalytic domain; black bar, the region specific for s-b5 cDNA; striped bars, the coding region for the membrane-binding domain; solid lines, the 5' and 3'-noncoding regions of b5 cDNA; ", termination codon. (B) Agarose gel electrophoresis of DNA segments of mouse cytochrome b5 amplified by means of the polymerase chain reaction. The cDNA from liver (lanes 1 and 2), and erythrocytes (lanes 3 and 4) was amplified with primers 1 and 2 (lanes I and 3), and primers 1 and 3 (lanes 2 and 4), and the products were analyzed by agarose gel electrophoresis. M, the molecular weight markers.

TABLE I. Effects of cofactors and the microsomal fraction of mouse liver on the CMP-NeuAc hydroxylation using the cytosolic fraction as the hydroxylase source. The hydroxylation activity of the cytosolic fraction (56.3 ug protein) was measured under the various conditions listed. Cytosol, the cytosolic fraction; Hyd, the partially purified CMP-NeuAc hydroxylase (14.5 uunits); X, a fraction containing b5 reducing factor(s) (5.5 ug protein); s-bs, the soluble form of cytochrome b5 (20 pmol protein) ; Microsome, the microsomal fraction containing 20 pmol b5 protein; Triton, Triton X-100 (0.5%). The data are presented as relative values, the endogenous activity of the cytosolic fraction being taken as 1.00.

the level of the activity was much lower than that with the equivalent amount of the soluble form of b5 (Table I). However, in the presence of a detergent, Triton X 100, the activity was markedly elevated, reaching almost the same level as detected in the presence of soluble b5 (Table I). It has been shown that liver microsomes contain b5 reducing factor(s) as well as b5 (47, 48). In order to determine whether or not the b5 reducing factor(s) of the microsomes can also be coupled with the hydroxylase through the b5 of the microsomes, the membrane fraction was mixed with the partially purified hydroxylase, which was free from b5 and b5 reducing factors, and the hydroxylation activity was measured. As shown in Table II, the hydroxylase activity was successfully reconstituted without the addition of b5 and b5 reducing factors, if Triton X-100 was added to the

J. Biochem.

Reaction Mechanism of CMP-NeuAc Hydroxylase 385

TABLE II. Effects of phospholipids on reconstitution of the CMP-NeuAc hydroxylation activity with the partially purified hydroxylase and the microsomal fraction. The hydroxylation activity was measured using the partially purified hydroxylase (14.5 punts) under the various conditions listed. s-b5, the soluble form of cytochrome b5 (40 pmol protein); X, a fraction containing b5 reducing factor(s) (5.5 pg protein); Microsome, the microsomal fraction containing 40 pmol b5 protein; Triton, Triton X-100 (0.5%); PS, phosphatidylserine; PE, phosphatidylethanolamine; PI, phosphatidylinositol; PC, phosphatidylcholine. Activity is expressed as a percentage of that with the hydroxylase, s-b5 and X (100%).

incubation mixture. As an endogenous factor in place of Triton X-100, phosphatidylserine slightly increased the hydroxylation activity in this system (Table II). These data indicate that the b5 reducing factor(s) as well as b5 in the microsomal fraction are coupled in vitro with the CMPNeuAc hydroxylase of the mouse liver cytosolic fraction in the presence of the detergent. Taking the virtually null level of b5 in the cytosolic fraction of mouse liver into account, these data raise the possibility of the predominant participation of the microsomal b5 and b5 reductase(s) of the endoplasmic reticulum (ER) in the CMP-NeuAc hydroxylation in mouse liver.

DISCUSSION

The CMP-NeuAc hydroxylation step is mediated by an electron transport system comprising b5 reducing factors, b5j and CMP-NeuAc hydroxylase (36). The data obtained in the present study led us to propose the hypothesis illustrated in Fig. 3. First of all, the hydroxylase binds to the substrate, CMP-NeuAc, and then the enzyme-substrate complex interacts with b5, resulting in the formation of a ternary complex, i.e., b5-enzyme-substrate. The enzyme does not bind to b5 without the substrate. The interaction, therefore, seems to occur after conformational changes of the enzyme, which may be triggered by the binding of the substrate to the enzyme. After the conversion of CMPNeuAc to CMP-NeuGc, the ternary complex may dissociate into the components. Although we do not have evidence demonstrating that only the oxidized form of b5 interacts with the enzyme-substrate complex, the complex can interact with b5 without the addition of a b5 reducing factor and NAD(P)H. Therefore, it is quite reasonable to assume that the oxidized form of b5 can interact with the complex.

Judging from the Km value (0.15 pM) obtained in the soluble assay system, b5 has a high affinity to the hydroxylation system. However, CMP-NeuAc hydroxylase did not bind tightly to the immobilized b5 but was eluted from the b5 column with retardation. The Km value estimated in the soluble system could be that of the reduced b5 and the ternary complex might contain the reduced b5 which may form a more stable complex than that containing the oxidized form. In this case, electrons should be transferred to b5 before the ternary complex formation.

In this paper, horse b5 was used to explore the interaction with mouse CMP-NeuAc hydroxylase for practical reasons. However, mouse liver b5 transferred electrons to mouse hydroxylase to the same extent as horse b5 (Table I), suggesting that mouse b5 also participates in the complex formation with mouse hydroxylase.

In the previous paper, we demonstrated that CMPNeuAc hydroxylation is inhibited by salt and azide, which affect the electron transport from NAD(P)H to b5 and from b5 to oxygen molecules, respectively (33). In the presence of 100 mM KCI, the electron transfer from NAD(P)H to b5 was partially inhibited, though the hydroxylation reaction was almost completely abolished, suggesting the occurrence of other salt-sensitive pathways in the reaction (33). The present data indicated that the ternary complex formation was totally inhibited in the presence of 100 mM KC1. Taking these data together, it may be concluded that salt inhibits both the reduction of b5 and the formation of the ternary complex. Azide did not inhibit the complex formation, indicating that it may directly affect the final stage of the conversion of CMP-NeuAc to CMP-NeuGc, but the details of the action mechanism remain unclear.

It has been shown that the soluble form of b5 is abundant in mammalian erythrocytes and is synthesized from mRNA different from that for the microsomal form of b5 (43-45). In the case of mouse liver, however, a soluble b5-specific mRNA could not be detected on PCR analysis (Fig. 4B), even though CMP-NeuAc hydroxylation activity was detected in the cytosolic fraction of mouse liver (32, 33). These results suggest that the b5 activity detected in the cytosolic fraction may be ascribed to b5 which is synthesized as the microsomal form and then solubilized in the cytosol in vivo. The microsomal b5 is amphipathic and bound to the ER membrane through its C-terminal portion (49). Therefore, the solubilization might result from the proteolytic cleavage of the membrane-bound domain. Alternatively, b5 detected in the cytosolic fraction might be the microsomal form of b5 itself, which is temporarily localized in the cytosol en route to the ER membranes. The microsomal b5 is known to be synthesized on free polysomes and then attached to the ER membranes without the signal recognition particle receptor system (50, 51).

It should be noted that although the b5 reducing factors, b5 and hydroxylase of the electron transport system of mouse liver are all present in the cytosolic fraction (33, 35), the concentration of b5 is rate-limiting for the hydroxylation in vitro. In fact, in the presence of an excess amount of the soluble b5, the activity of the cytosolic fraction increased to more than 60-fold (Fig. 2). It is thus possible that the amount of b5 is also rate-limiting in vivo. The

present data on in vitro reconstitution using the cytosolic and microsomal fractions suggested the possibility of the participation of the microsomal b5 of the ER in CMP-NeuAc hydroxylation. However, the in vitro system does not completely mirror the in vivo system, as is evident from the requirement of a detergent in vitro. It is possible that a conformational change of the membrane during isolation prevents the access of the hydroxylase to the microsomal form of b5i or another endogenous factor(s) in place of the detergent may be involved in the in vivo reaction for the utilization of b5 on the ER. Screening of the endogenous factor(s) might be possible using the in vitro reconstitution system. In fact, of the phospholipids tested, only phos

Vol. 115, No. 3, 1994

386 H . Takematsu et al.

phatidylserine slightly increased the hydroxylation activity (Table II). Currently, it is not clear which b5, that localized in the

cytosol or that in the ER, is involved in the CMP-NeuAc hydroxylation in vivo. In order to determine which one is involved, further studies are required. In the case of participation of b5 of the cytosol in the hydroxylation, the ternary complex consisting of b5, the hydroxylase and CMP-NeuAc must be soluble. On the other hand, the participation of the microsomal b5 of the ER would result in the formation of a membrane-associated ternary complex. Determination of the in vivo localization of the hydroxylase using an antibody against the hydroxylase may solve this

problem.

We would like to thank Drs. Tomita and Itoh (MECT Co., Tokyo) for the gift of CMP-NeuGc.

REFERENCES

1. Polley, M.J., Phillips, M.L., Wayner, E., Nudelman, E., Singhal, A.K., Hakornori, S.-I., & Paulson, J.C. (1991) Proc. Natl. Acad. Sci. USA 88, 6224-6228 2. Larsen, G.R., Sako, D., Ahern, T.J., Shaffer, M., Erban, J., Sajer, S.A., Gibson, M.R., Wagner, D.D., Furie, B.C., & Furie, B.

(1992) J. Biol. Chem. 267,11104-11110 3. True, D.D., Singer, M.S., Lasky, L.A., & Rosen, S.D. (1990) J. Cell Biol. 111, 2757-2764 4. Imai, Y., Singer, M.S., Fennie, C., Lasky, L.A., & Rosen, S.D.

(1991) J. Cell Biol. 113,1213-1221 5. Lasky, L.A., Singer, M.S., Dowbenko, D., Imai, Y., Henzel,

W.J., Grimley, C., Fennie, C., Gillett, N., Watson, S.R., & Rosen, S.D. (1992) Cell 69, 927-938 6. Tsuji, S., Yamashita, T., &Nagai, Y. (1983) J. Biochem. 94,303

306 7. Tsuji, S., Yamashita, T., & Nagai, Y. (1988) J. Biochem. 104, 498-503

8. Saito, M., Terui, H., & Nojiri, H. (1985) Biochem. Biophys. Res. Commun. 132, 223-231 9. Nojiri, H., Takaku, F., Terui, H., Miura, Y., & Saito, M. (1986)

Proc. Natl. Acad. Sci. USA 83, 782-786 10. Nojiri, H., Kitagawa, S., Nakamura, S., Kirito, K., Enomoto, Y.,

& Saito, M. (1983) J. Biol. Chem. 263, 7443-7446 11. Nakamura, M., Kirito, K., Yamanoi, J., Wainai, T., Nojiri, H., & Saito, M. (1991) Cancer Res. 51, 1940-1945

12. Bouhours, D. & Bouhours, J.-F. (1988) J. Biol. Chem. 263, 15440-15545

13. Bouhours, J.-F. & Bouhours, D. (1989) J. Biol. Chem. 264, 16992-16999

14. Yuyama, Y., Yashimatsu, K., Ono, E., Saito, M., & Naiki, M. (1993) J. Biochem. 113, 488-491

15. Schauer, R., Stoll, S., & Reuter, G. (1991) Carbohydr. Res. 213, 353-359

16. Shaw, L., Yousefi, S., Denis, J.W., & Schauer, R. (1991) Glycoconjugate J. 8, 433-441 17. Sherbiom, A.P. & Dahlin, C. E. (1985) J. Biol. Chem. 260, 1484

1492 18. Yasue, S., Handa, S., Miyagawa, S., Inoue, J., Hasegawa, A., &

Yamakawa, T. (1978) J. Biochem. 83, 1101-1107

19. Hashimoto, Y., Yamakawa, T., & Tanabe, Y. (1984) J. Biochem. 96,1777-1782 20. Hamanaka, S., Handa, S., Inoue, J., Hasegawa, A., &Yamakawa, T. (1979) J. Biochem. 86, 695-698 21. Nakamura, K., Hashimoto, Y., Moriwaki, K., Yamakawa, T., & Suzuki, A. (1990) J. Biochem. 107, 3-7 22. Hanganuitziu, M. (1924) Soc. Biol. 91, 1457-1459 23. Deicher, H. (1926) Z. Hyg. 106, 561-579 24. Hirabayashi, Y., Kasakura, H., Matsumoto, M., Higashi, H., Kato, S., Kasai, N., & Naiki, M. (1987) Jpn. J. Cancer Res. 78, 251-260 25. Hirabayashi, Y., Higashi, H., Kato, S., Taniguchi, M., &

Matsumoto, M. (1987) Jpn. J. Cancer Res. 78, 614-620 26. Ohashi, Y., Sasabe, T., & Nishida, T. (1983) Am. J. Ophthalmol. 96,321-325 27. Devine, P.L., Clark, B.A., Birrell, G.W., Layton, G.T., Ward, B.G., Alewood, P.F., & McKenzie, F.C. (1991) Cancer Res. 51, 5826-5836 28. Kasukawa, R., Kano, K., & Milgron, F. (1976) Clin. Exp. Immunol. 25, 122-132 29. Nishimaki, T., Kano, K., & Milgron, F. (1979) J. Immunol. 122, 2314-2318 30. Shaw, L. & Schauer, R. (1988) Biol. Chem. Hoppe-Seyler 369, 477-486 31. Shaw, L. & Schauer, R. (1989) Biochem. J. 263, 355-363 32. Kozutsumi, Y., Kawano, T., Yamakawa, T., & Suzuki, A. (1990) J. Biochem. 108, 704-706 33. Kozutsumi, Y., Kawano, T., Kawasaki, H., Suzuki, K., Yama kawa, T., & Suzuki, A. (1991) J. Biochem. 110, 429-435 34. Nakamura, K., Hashimoto, Y., Yamakawa, T., & Suzuki, A.

(1988) J. Biochem. 103, 201-208 35. Kawano, T., Kozutsumi, Y., Takematsu, H., Kawasaki, T., & Suzuki, A. (1993) Glycoconjugate J. 10, 109-115 36. Kohn, J. & Wilchek, M. (1982) Biochem. Biophys. Res. Commun. 107,878-884 37. Beutler, E. & West, J. (1976) J. Lab. Clin. Med. 88, 328-333 38. Chomczynski, P. & Sacchi, N. (1989) Anal. Biochem. 162, 156

159 39. Sanger, F., Nicklen, S., & Coulson, A.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467 40. Omura, S. & Takesue, S. (1970) J. Biochem. 67, 249-257 41. Omura, T. & Sato, R. (1964) J. Biol. Chem. 239, 2379-2385 42. Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K.,

Gartner, F.H., Provenzano, M.D., Fujimoto, E.K., Goeke, N.M., Olson, B.J., & Klenk, D.C. (1985) Anal. Biochem. 150, 76-85 43. Giordano, S.J. & Steggles, A.W. (1991) Biochem. Biophys. Res. Commun. 178, 38-44 44. Takematsu, H., Kozutsumi, Y., Suzuki, A., & Kawasaki, T.

(1992) Biochem. Biophys. Res. Commun. 185, 845-851 45. Giordano, S.J. & Steggles, A.W. (1993) Biochim. Biophys. Acta 1172,95-100 46. Shaw, L., Schneckenburger, P., Carlsen, J., Christiansen, K., & Schauer, R. (1992) Eur. J. Biochem. 206, 269-277 47. Spatz, L. & Strittmatter, P. (1973) J. Biol. Chem. 248, 793-799 48. Mihara, K. & Sato, R. (1975) J. Biochem. 78, 1057-1073 49. Abe, K., Kimura, S., Kizama, R., Anan, F.K., & Sugita, Y.

(1985) J. Biochem. 97,1659-1668 50. Bendzko, P., Prehn, S., & Rapoport, T.A. (1982) Eur. J. Biochem. 123,121-126 51. Sakaguchi, M., Mihara, K., & Sato, R. (1984) Proc. Natl. Acad. Sci. USA 81, 3361-3364

J. Biochem.