Isolation of Polysaccharides Sulfated during Early Embryogenesis in Fucus

Beginning 10 hours after fertilization, zygotes of Fucus distichus L. Powell incorporate (35)S into polysaccharides as a sulfate ester of fucose. These sulfated polysaccharides are sequestered in only the rhizoid cell of the two-celled embryo and can serve as a marker of cellular differentiation. Zygotes were pulsed at different times after fertilization with Na(2) (35)SO(4) to identify and isolate the fucans localized within the region of cytoplasm destined to become the rhizoid cell. Low molecular weight pools of (35)S were saturated within 60 minutes, with the greatest incorporation into ethanol-soluble and insoluble fractions occurring with 0.1 mm Na(2)SO(4) in the artificial sea water medium. At the time of rhizoid formation, four fucose-containing polysaccharide fractions incorporated (35)S. When each fraction was subjected to diethylaminoethyl chromatography, two components were eluted with KCl that contained over 84% of the fucose and 93% of the (35)S of the particular fraction. Highvoltage paper electrophoresis of each fraction also resulted in the separation of these two major components. Both components from each of the four fractions behaved identically when separated by diethylaminoethyl chromatography and paper electrophoresis. By comparing the incorporation of (35)S into the polysaccharide fractions at 4 and 16 hours after fertilization, the fucan-sulfate components that are localized in the cytoplasm at the time of rhizoid formation were isolated. Although sulfated polysaccharides in brown algae are reported to be very heterogeneous in terms of their sugar composition and complexes with other heteropolymers, we propose that there are two major components that are sulfated during early embryogenesis in Fucus. The location of these two sulfated polysaccharides in different chemical fractions may reflect their subcellular localization (e.g., cytoplasmic vesicles or cell walls), or their association with other heteropolymers.     

The relationship between changes in cell wall composition and the establishment of polarity in Fucus embryos

Changes in the appearance and location of fucoidin in the cell walls of Fucus embryos were related to embryo development. Fucoidin was not present in the cell wall until 10–14 hr after fertilization, when the embryos began to incorporate fucoidin preferentially into a localized area of the wall. At this time the site of rhizoid initiation was determined; that is, the embryos had undergone axis commitment. Germination of the single-celled embryo occurred between 12 and 16 hr, after fertilization, with all cell walls from germinated embryos showing fucoidin localization at the rhizoid end of the cell. The percentage of embryos with localized fucoidin at the time of axis fixation equaled the percentage of embryos that subsequently germinated. Culturing the embryos in sea water plus 0.8 M sucrose prevented the outgrowth of the rhizoid, but not the localization of fucoidin in the wall or axis commitment. Cycloheximide, nitroprusside, cytochalasin B, sulfate-free sea water, high levels of Ca²⁺, and a breakdown product of TIBA all prevented rhizoid growth and the specific localization of fucoidin. In addition, axis commitment could not be demonstrated in the presence of these inhibitors. DTNB, PCMBS, TIBA, HgCl₂, Mg²⁺ were ineffective as reversible inhibitors of rhizoid initiation. The authors propose that the fixation of axis commitment is accompanied by localized changes in the cell wall involving the incorporation of fucoidin as a structural component of the wall.     

In vivo sulfation of the contact site A glycoprotein of Dictyostelium discoideum

During the development of Dictyostelium discoideum from the growth phase to the aggregation stage, a glycoprotein with an apparent mol. wt. of 80 kd is known to be expressed on the cell surface. This glycoprotein, referred to as contact site A, has been implicated in the formation of species-specific, EDTA-stable contacts of aggregating cells. When developing cells were labeled in vivo with [³⁵S]sulfate, the 80-kd glycoprotein was found to be the most prominently sulfated protein. Another strongly sulfated protein had an apparent mol. wt. of 130 kd and was, like the 80-kd glycoprotein, developmentally regulated and associated with the particulate fraction of the cells. The [³⁵S]sulfate incorporated into the 80-kd and 130-kd proteins was not present as tyrosine-O-sulfate, a modified amino acid found in many proteins of mammalian cells. D. discoideum cells incubated with [³⁵S]sulfate in the presence of tunicamycin, an inhibitor of N-glycosylation, produced a 66-kd protein that reacted with monoclonal antibodies raised against the 80-kd glycoprotein, but no longer contained [³⁵S]sulfate. These results suggest that sulfation of the 80-kd glycoprotein occurred on carbohydrate residues. The possible importance of sulfation for a role of the 80-kd glycoprotein in cell adhesion is discussed.     

Sulfation of fucoidin in focus embryos: III. Required for localization in the rhizoid wall

Zygotes of the brown alga Fucus distichus L. Powell accumulate a sulfated polysaccharide (fucoidin) in the cell wall at the site of rhizoid formation. Previous work indicated that zygotes grown in seawater minus sulfate do not sulfate the preformed fucan (an unsulfated fucoidin) but form rhizoids. Under these conditions, we determined whether sulfation of the fucan is required for its localization in the rhizoid wall. This was accomplished by developing a specific stain for both the fucan and fucoidin. Using a precipitin assay, we demonstrated in vitro that the lectin ricin (RCA(I)) specifically complexes with both the sulfated and desulfated polysaccharide. No precipitate is observed when either is incubated in 0.1 M D-galactose or when RCA(I) is mixed with laminarin or alginic acid, the other major polysaccharides in Fucus. RCA(I) conjugated with fluorescein isothiocyanate (FITC) is also shown to bind specifically to fucoidin using a filter paper (DE81) assay. When added to zygotes, RCA(I)-FITC binds only to the site of fucoidin localization, i.e., the rhizoid cell wall. However, RCA(I)-FITC is not observed in the rhizoid wall of zygotes grown in the absence of sulfate. This observation is not due to inability of RCA(I)-FITC to bind to the fucan in vivo. Chemically desulfated cell walls that contained fucoidin in the rhizoid wall bind RCA(I)-FITC only in the rhizoid region. Also, the concentration of fucose-containing polymers and polysaccharides that form precipitates with RCA(I) is the same in embryos grown in the presence or absence of sulfate. If sulfate is added back to cultures of zygotes grown without sulfate, fucoidin is detected at the rhizoid tip by RCA(I)-FITC several hours later. These results support the conclusion that the enzymatic sulfation of the fucan is a modification of the polysaccharide required for its localization and/or assembly into a specific region of the cell wall.     

Intracellular transport and tyrosine sulfation of procollagens V

Several tyrosine residues of the extracellular p-collagens V and collagens V are sulfated [Fessler, L. I., Brosh, S., Chapin, S. and Fessler, J. H. (1986) J. Biol. Chem. 261, 5034-5040]. Here, the sulfation of their intracellular precursors, the procollagens V, was studied. A Golgi-enriched subcellular fraction of chick embryo tendon catalyzed the sulfation of tyrosine residues in both endogenous and added, unsulfated procollagens V with the sulfate donor 3'-phosphoadenosine 5'-[35S]phosphosulfate. Intracellular tyrosine sulfation of procollagen V occurred at a point distal to the cis Golgi compartment as judged by change of the N-linked carbohydrate of procollagen V from being endoglycosidase-H-sensitive to being resistant. The time course of the intracellular modifications of procollagen V was determined by incubating tendons with 3H-labeled amino acids and with [35S]sulfate. The pro alpha(V) chains were synthesised in about 10 min and then assembled into unsulfated procollagen V molecules. Tyrosine sulfation occurred 50 min after completion of polypeptide synthesis and the molecules were successively sulfated in the order in which they had been synthesized. The antimicrotubular drug Nocodazole, which disrupts the spatial organization of the Golgi, decreased the time interval between synthesis of procollagens V and sulfation. The sulfated procollagens V were soon secreted and cut to sulfated p-collagens V. Sulfated pro alpha 1(V) chains were cleaved faster than sulfated pro alpha 1'(V) chains. The relationship of sequential protein modification to spatial cellular organization is discussed.     

Fucoidan in Padina sanctae-crucis spores

The presence and distribution of fucoidan in the vegetative fronds and spores of the brown alga Padina sanctae-crucis Børg. was studied. Autoradiography using ³⁵S showed that fucoidan is localized in the walls of the meristematic cells of the rhizoid of the developing spore. It is suggested that fucoidan is structurally involved in the formation of the new cell walls.     

Sulfation of thyroglobulin: A ubiquitous modification in vertebrates

Summary Mammalian thyroglobulin is released by thyroid follicle cells as a sulfated glycoprotein; the sulfate residues are mostly linked to tyrosine, but they are also attached to the high-mannose carbohydrate side-chains. To decide whether sulfation of thyroglobulin is confined to mammals, representatives of other vertebrate classes were analyzed for the presence of sulfated thyroglobulin: fish (trout), amphibians (clawed toad) and birds (chicken). Mini-organs were prepared from thyroid tissue and suspended in a ³⁵SO ₄ ^-- -containing culture medium. Light- and electron-microscope autoradiographs prepared from the mini-organs showed that thyroid follicle cells from all species examined incorporate ³⁵SO ₄ ^-- and synthesize a sulfated secretory product which accumulates in the follicle lumen. The Golgi complex was detected as the primary intracellular site of sulfate organification. The ³⁵SO ₄ ^-- -radiolabeled secretory product of all species was shown by polyacrylamide-gel-electrophoretic analyses to consist of thyroglobulin, identified by comparison with biosynthetically ¹²⁵I-labeled thyroglobulin. The results indicate that the sulfation of thyroglobulin is a ubiquitous post-translational modification observed already in the thyroglobulin of lower vertebrates. Our observations suggest that sulfation of thyroglobulin was acquired in the early stages of thyroid evolution.     

Cell-free sulfation of the contact site A glycoprotein of Dictyostelium discoideum and of a partially glycosylated precursor

An 80-kDa glycoprotein of Dictyostelium discoideum, designated contact site A, has been implicated in EDTA-stable cell adhesion. This protein is known to be the major sulfated protein of aggregation-competent cells and has been shown to contain two types of carbohydrate, sulfated type 1 and unsulfated type 2 carbohydrate moieties. Here we investigate the cell-free sulfation of this protein. In the homogenate of developing cells, [35S]sulfate was transferred by endogenous sulfotransferase from [35S]3'-phosphoadenosine-5'-phosphosulfate to the contact site A glycoprotein and to various other endogenous proteins. The sulfate was transferred to carbohydrate rather than to tyrosine residues. After differential centrifugation of the homogenate, the capacity for sulfation of the contact site A glycoprotein was barely detected in the plasma membrane-enriched 10,000 X g pellet fraction which contained the bulk of this glycoprotein, but was largely recovered in the 100,000 X g pellet fraction which contained only a small portion of this glycoprotein. After sucrose gradient centrifugation, the membranes containing the sulfation capacity were found to have a density characteristic for Golgi membranes. In immunoblots, monoclonal antibodies raised against the contact site A glycoprotein recognized not only this 80-kDa protein, but also a sulfatable 68-kDa protein found in the 100,000 X g pellet fraction. The 68-kDa protein did not react with monoclonal antibodies against type 2 carbohydrate but was converted by endoglycosidases F and H into a 53-kDa protein, indicating that it was a partially glycosylated form of the 80-kDa glycoprotein containing only type 1 carbohydrate. Isoelectric focusing showed that a substantial portion of the 68-kDa glycoprotein was unsulfated, even after cell-free sulfation. The 68-kDa glycoprotein was not found in the plasma membrane-enriched 10,000 X g pellet fraction and did not accumulate in parallel with the 80-kDa contact site A glycoprotein during cell development. We conclude that the 68-kDa glycoprotein is a precursor that is converted by attachment of type 2 carbohydrate and sulfation of type 1 carbohydrate into the mature 80-kDa glycoprotein. The precursor nature of the 68-kDa glycoprotein was supported by results obtained with mutant HL220 which is defective in glycosylation (Murray, B. A., Wheeler, S., Jongens, T., and Loomis, W. F. (1984) Mol. Cell. Biol. 4, 514-519). This mutant specifically lacks type 2 carbohydrate and produces a 68-Kda glycoprotein instead of the 80-kDa contact site A glycoprotein (Yoshida, M., Stadler, J., Bertholdt, G., and Gerisch, G. (1984) EMBO J. 3, 2663-2670).(ABSTRACT TRUNCATED AT 400 WORDS)     

Replication of mitochondrial DNA in the early development of Misgurnus fossilis

Mitochondria isolated from Misgurnus fossilis embryos at various developmental stages were incubated with ³H-dTTP in vitro and the incorporation into mtDNA was determined. It has been found that the rate of mtDNA labeling increases exponentially with a doubling time of 7 hr from 0.01 pmole of ³H-dTMP/mg protein/hr in mitochondria from unfertilized eggs to 0.4 pmoles of ³H-dTMP/mg protein/hr in mitochondria of 35 hr embryos. The pool of intramitochondrial dTTP decreases 2.5 times during the first 10 hr after fertilization, then remains practically constant up to 35 hr of development. The rate of exogenous ³H-dTTP incorporation into the acid soluble pool of isolated mitochondria at two stages is approximately proportional to the pool size. Thus identical specific activities of ³H-dTTP inside mitochondria would be obtained even with pools of different sizes. We conclude that the increase of ³H-dTMP incorporation into mtDNA in development reflects genuine activation of mtDNA synthesis. As early as 6 hr after fertilization the bulk of the label incorporated into mtDNA is found in the fraction associated with covalently closed molecules. This pattern of labeling characteristic for replicating mtDNA is maintained throughout early development. In contrast such preferential label incorporation into the closed circular fraction was not found with mitochondria of unfertilized eggs. Closed mtDNA from unfertilized eggs contains not more than 1% of molecules with D-loops. In 35 hr embryos the corresponding value is equal to about 4%. Activation of mtDNA replication in embryogenesis is probably due to the activation of mechanisms responsible for the generation of primers for replication. DNA polymerase activity solubilized from mitochondria remains unchanged in the course of embryogenesis.     

Nonsulfated cholecystokinins in the small intestine of pigs and rats

Cholecystokinin (CCK) is a gut hormone that acts via two receptors. The CCK_A-receptor requires the tyrosyl residue in the C-terminal bioactive site of CCK to be O-sulfated, whereas, the CCK_B-receptor binds irrespective of sulfation. Consequently, unsulfated CCK peptides – if present – may constitute a hormone system that acts only through the CCK_B-receptor. Therefore, we have now examined whether, CCK peptides occur in nonsulfated form in the small intestine of pigs and rats. The concentrations of sulfated and nonsulfated CCK were measured by RIAs, one specific for sulfated CCKs and a new two-step assay specific for nonsulfated CCK. For further characterization, the intestinal extracts were subjected to size- and ion exchange-chromatography.The intestinal concentrations of sulfated and nonsulfated CCK were highest in the duodenum and the proximal part of jejunum both in the pig and the rat. The porcine duodenal mucosa contained 193 ± 84 pmol/g sulfated CCK and 31 ± 10 pmol/g nonsulfated CCK, and the upper rat intestine 70 ± 19 pmol/g and 8 ± 2 pmol/g, respectively. The degree of sulfation correlated with the endoproteolytic proCCK processing. Thus, 38% of porcine CCK-58 was unsulfated, whereas, only 12% of CCK-8 was unsulfated.The results show that a substantial part of intestinal CCK peptides in rats and pigs are not sulfated, and that the longer peptides (CCK-58 and CCK-33) are less sulfated than the shorter (CCK-22 and CCK-8). Hence, the results demonstrate that proCCK in the gut is processed both to sulfated and unsulfated α-amidated peptides of which the latter are assumed to act via the CCK_B-receptor.     

Synthesis of sulfated glycoproteins by glial precursor cells

Populations of enriched glial precursor cells and astrocytes isolated from primary cultures of newborn rat brain were used to study the synthesis of sulfated glycoproteins. Both cell types incorporated [³H]glucosamine and [³⁵S]sulfate into carbohydrate side chains of proteoglycans and glycoproteins. The rate of incorporation of [³H]glucosamine into the oligosaccharides and the pattern of distribution of the label into high mannose and complex glycopeptides recovered from the glycoproteins appeared to be similar for the two glial cell types. However, clear differences were noted in the rate of oligosaccharide sulfation activities. Thus the cultures of precursor glia were about four times more active than cultures enriched in astroglia in their ability to incorporate [³⁵S]sulfate into glycoproteins.     

The role of the Golgi complex in sulfate metabolism

This investigation was designed to determine if sulfate metabolism is the function of a particular cell organelle, or whether the site of sulfation varies, depending upon the type of cell and the class of sulfated compound. Rats and mice were injected intravenously with inorganic sulfate labeled with ³⁵S (H₂ ³⁵SO₄), and were then killed by vascular perfusion of fixative 5–30 min later. Several tissues were prepared for electron microscope autoradiography. 14 different types of specialized cells which incorporated the labeled sulfate were analyzed. In every case, the sulfate was initially detected in the smooth membranes and vesicles of the Golgi complex. Available evidence indicates that these cells were engaged in the synthesis of several different sulfated compounds, including mucopolysaccharides, glycoproteins, lipids, and steroids. These results lead to the generalization that the enzymes required for the transfer of inorganic sulfate to a variety of acceptor molecules are located in the Golgi complex.     

Acid mucopolysaccharide metabolism, the cell surface, and primary mesenchyme cell activity in the sea urchin embryo

The relationship between ³⁵SO₄ incorporation into acid mucopolysaccharides and the appearance and activity of the primary mesenchyme cells has been studied in the sea urchin, Lytechinus pictus. The ratio of the uptake of ³⁵SO₄ to its incorporation into cetylpyridinium chloride precipitable material varies over a wide range during early development, with the smallest ratio, therefore the greatest sulfation activity, being found at the early mesenchyme blastula stage. The types of mucopolysaccharides produced have not been identified, but are heterogeneous. At the mesenchyme blastula stage nearly 90% of the polysaccharides produced become sulfated. When embryos develop in sulfate-free sea water to the mesenchyme blastula stage there is a 70% decrease in the incorporation of ³H-acetate into polysaccharides and a 13-fold decrease in the ratio of sulfated to nonsulfated polysaccharides produced. Embryos raised in sulfate-free sea water develop normally to the mesenchyme blastula stage at which time there is an accumulation in the blastocoel of primary mesenchyme cells that do not migrate. The surface of the primary mesenchyme cells of sulfate-deficient embryos has a smooth appearance in the scanning electron microscope, while the surface of these cells in control embryos is rough, possibly reflecting the presence of an extracellular coat. It is suggested that there is a correlation between sulfated polysaccharide synthesis, cell surface morphology and cell movement.     

The degree of polymerization and sulfation patterns in heparan sulfate are critical determinants of cytomegalovirus entry into host cells

Several enveloped viruses, including herpesviruses attach to host cells by initially interacting with cell surface heparan sulfate (HS) proteoglycans followed by specific coreceptor engagement which culminates in virus-host membrane fusion and virus entry. Interfering with HS-herpesvirus interactions has long been known to result in significant reduction in virus infectivity indicating that HS play important roles in initiating virus entry. In this study, we provide a series of evidence to prove that specific sulfations as well as the degree of polymerization (dp) of HS govern human cytomegalovirus (CMV) binding and infection. First, purified CMV extracellular virions preferentially bind to sulfated longer chain HS on a glycoarray compared to a variety of unsulfated glycosaminoglycans including unsulfated shorter chain HS. Second, the fraction of glycosaminoglycans (GAG) displaying higher dp and sulfation has a larger impact on CMV titers compared to other fractions. Third, cell lines deficient in specific glucosaminyl sulfotransferases produce significantly reduced CMV titers compared to wild-type cells and virus entry is compromised in these mutant cells. Finally, purified glycoprotein B shows strong binding to heparin, and desulfated heparin analogs compete poorly with heparin for gB binding. Taken together, these results highlight the significance of HS chain length and sulfation patterns in CMV attachment and infectivity.     

Structure and anti-dengue virus activity of sulfated polysaccharide from a marine alga

A sulfated polysaccharide, named fucoidan, from the marine alga Cladosiphon okamuranus is comprised of carbohydrate units containing glucuronic acid and sulfated fucose residues. Here we found this compound potently inhibits dengue virus type 2 (DEN2) infection. Viral infection was inhibited when DEN2, but not other serotypes, was pretreated with fucoidan. A carboxy-reduced fucoidan derivative in which glucuronic acid was converted to glucose did not inhibit viral infection. Elimination of the sulfated function group from fucoidan significantly attenuated the inhibitory activity on DEN2 infection with <1% fucoidan. DEN2 particles bound exclusively to fucoidan, indicating that fucoidan interacts directly with envelope glycoprotein (EGP) on DEN2. Structure-based analysis suggested that Arg323 of DEN2 EGP, which is conformationally proximal to one of the putative heparin binding residues, Lys310, is critical for the interaction with fucoidan. In conclusion, both the sulfated group and glucuronic acid of fucoidan account for the inhibition of DEN2 infection.     

Cholinephosphotransferase activity during development of the sea urchin, Arbacia punctulata

Cholinephosphotransferase activity was examined during early development of Arbacia punctulata embryos. CMP was rapidly incorporated by enzyme preparation from Arbacia embryos into a compound identified as CDP-choline. Activity was dependent upon the presence of Mg²⁺, Mn²⁺, or Co²⁺, and was maximal in phosphate buffer, pH 7.0, containing 3 × 10^?2M MgCl₂ and 2 × 10^?5M CMP. Addition of ATP or egg lecithin had no effect on enzyme activity. The activity was localized in the 1000 g and 10,000 g pellets, with little or no activity in the microsomal fraction.Upon fertilization, cholinephosphotransferase activity decreased rapidly, detectable changes in activity being observed within 2 min after fertilization. After reaching a minimum activity 2 hr after fertilization, the enzyme activity increased up to the gastrula stage, then decreased once more. Possible interference by competing enzymes was examined and found to be negligible.     

Aryl sulfatase of unusual specificity from the liver of marine mollusk Littorina kurila

An aryl sulfatase of unusual specificity has been isolated from the liver of marine mollusk Littorina kurila. It hydrolyzes p-nitrophenyl sulfate, does not affect the natural fucoidan, and catalyzes splitting off the sulfate group in position C4 of xylose residues within the carbohydrate chains of holostane triterpene glycosides from sea cucumbers. The properties of the enzyme were studied at pH 5.4. The protein is homogeneous according to electrophoresis and has M 45 ± 1 kDa. The semiinactivation time of the enzyme at 60°C is 20 min, and its K _m value for the hydrolysis of p-nitrophenyl sulfate is 8.7 ± 1 mM. It was shown that natural sulfated polyhydroxysteroids inhibit activity of the sulfatase; their I ₅₀ values depend on their structures and are within the range from 10^?3 to 10^?5 M.     

Pituitary gastrins. Different processing in corticotrophs and melanotrophs

The anterior, intermediate, and neural lobes of porcine pituitaries contain different molecular forms of gastrin. In the anterior lobe, unsulfated large gastrins, component I, and gastrin-34 constitute the main forms. In contrast in the intermediate and neural lobes, gastrin-17, sulfated as well as unsulfated, predominates. Since component I and gastrin-34 are biosynthetic precursors of gastrin-17, the findings indicate that proteolytic processing and amino acid derivatization (sulfation) differs between the lobes. This difference, together with the localization of anterior lobe gastrin in corticotrophs and intermediate lobe gastrin in melanotrophs, emphasizes the close relation and parallelism of the biosynthesis of gastrin and corticotropin peptides.     

Tubulin:tyrosine ligase in oocytes and embryos of Xenopus laevis

Tubulin:tyrosine ligase (TTL), which catalyzes the post-translational addition of tyrosine to the α chain of tubulin, exists in a wide variety of embryonic and adult vertebrate tissues. In the present study, we report that TTL exists in amphibian oocytes at a time when tubulin is a poor substrate for tyrosination, and when, in immature oocytes, tubulin is not polymerizable. Ligase activity detected at several stages of oogenesis and embryogenesis in Xenopus is compatible with mammalian brain tubulin in the tyrosination reaction. Within 3–5 hr after fertilization, [³H] tyrosine incorporated/μg endogenous tubulin increases approximately 3.5-fold over that in extracts prepared from the largest oocytes obtained. This increase cannot be accounted for by increasing levels of TTL. Ligase activity remains fairly constant throughout oogenesis and early embryogensis and rises significantly (2-fold) only 35–50 hr after fertilization. The late rise in embryonic ligase activity is not accompanied by a change in apparent k_m for tubulin.     

In vivo and in vitro tyrosine sulfation of a membrane glycoprotein

A431 cells incorporate 35SO4 into a protein of Mr 61,000 (P61). We examined sulfation of P61 by cells (in vivo) and by a cell-free system (in vitro) which requires only addition of A431 cell membranes and a 3'-phosphoadenosine 5'-phospho[35S]sulfate-generating system prepared from Krebs ascites cells. Sulfate is found exclusively in the form of tyrosine SO4 by two-dimensional high voltage electrophoresis following Pronase digestion. Endoglycosidase F digestion reduces the Mr by 2,000 but does not release the sulfate, indicating that P61 is a glycoprotein but that sulfate is not incorporated into the carbohydrate. Sulfated P61 is not found in the medium from cultured cells and remains associated with the membrane fraction following cell lysis. Treatment of membranes with 0.4 M NaCl, 0.3 M KCl, 15 mM EDTA, or pH 11.0 does not release sulfated P61. P61 is solubilized by Triton X-114 treatment of membranes and partitions into the detergent phase upon warming. Based on these characteristics, we conclude that P61 is an integral membrane protein. Trypsin digestion experiments with intact cells suggest that sulfated P61 is predominantly located in the plasma membrane. This is the first example of an integral membrane protein which is sulfated on tyrosine. The properties of the sulfation reaction are distinct from those reported for secreted proteins and are consistent with the possibility that this modification occurs at the plasma membrane rather than in the Golgi.