Immunocytochemical study on the cytoplasmic side of cell membranes infected with vesicular stomatitis virus by quick-freezing and deep-etching replica method

Summary Synthesized N protein of vesicular stomatitis virus (VSV) is associated with replicated viral genomes in the infected cells. The cytoplasmic side of cell membranes was examined by quick-freezing and deep-etching replica method, in order to clarify the localization of VSV genomes. Control or infected monolayer Vero cells were fixed in 2% paraformaldehyde, scraped and centrifuged to make pellets. A drop of the cell pellet was put between two glass coverslips, which were coated with 3-aminopropyl triethoxy silane and glutaraldehyde. The cells were consequently split open and postfixed in the mixture of glutaraldehyde and paraformaldehyde. Some inside-out cell membranes on the coverslips were immunostained with anti-N monoclonal antibody directly coupled to gold particles. Others were immunostained with anti-N monoclonal antibody and rabbit anti-mouse IgG coupled to peroxidase and fixed again in glutaraldehyde. They were incubated in diaminobenzidine and hydrogen peroxide solution for 1 min. All of them were infiltrated with 10% methanol in distilled water and quickly frozen in a mixture of isopentane and propane cooled by liquid nitrogen. Such preparations were deep-etched and shadowed by platinum and carbon. Although many cell organelles were found to be associated with the cytoplasmic side of cell membranes in the normal Vero cells, few cell organelles were attached to it in the infected cells. On the contrary, special strand structures were identified, which could be immunostained with anti-N monoclonal antibody. It is concluded that platinum replicas have sufficient resolution to identify the VSV genomes coated with N protein and that these nucleocapsids can be associated with the cytoplasmic side of cell membranes in the infected cells.S. Ohno was a visiting fellow from the Fogarty International Center at the National Institutes of Health, USA, while this work was in progress     

Immunocytochemical localization of vesicular stomatitis virus proteins N and NS with monoclonal antibodies

The purpose of this paper is to describe the immunocytochemical-localization of N and NS nucleocapsid proteins of vesicular stomatitis virus in the cells throughout the infectious cycle. N protein was detected in the cytoplasm at 2 h after infection and formed small cytoplasmic clusters which progressively increased in size and number. At 5-6 h, it formed large cytoplasmic inclusions. NS protein was detected in the cytoplasm a little later than N protein and showed almost the same immunostaining pattern. However, diffuse background staining of NS protein was identified throughout the cytoplasm by double immunostaining methods. At electron microscopic level, N protein was mostly granular and occasionally organized in strands at 2-3 h. At 5-6 h, numerous immunostained reaction products were organized in strands. The reaction products of NS protein were almost the same as those of N protein with the exception that diffuse background staining was observed. Cos cells, transfected with SV40 vector containing N gene obtained by recombinant DNA technique, showed clusters of N protein, but virtually no strand at electron microscopic levels. The rapid-freezing and deep-etching replica method demonstrated that loosely coiled VSV genome coated with N protein was localized on cytoplasmic sides of cell membranes in the infected cells. These results showed that complete virus genome replication was needed for strand formation of N and NS proteins and suggested that they were bound to VSV genomes in the infected cells.     

Immunocytochemical study of the intracellular localization of M protein of vesicular stomatitis virus

Summary The purpose of this paper is to describe the immunocytochemical localization of M protein of vesicular stomatitis virus (VSV) in infected cells. Vero cells, MDBK cells, Swiss 3T3 cells, and BHK cells were examined at various times after infection. For immunofluorescent staining, the cells were fixed with PLP fixative and then treated with 0.05% Triton X-100 before incubation with antibodies. Three hours after infection, M protein exhibited diffuse immunostaining throughout the cytoplasm and later accumulated along the cell membrane. The localization of M protein differed from the granular localization of the nucleocapsid N protein of VSV in the cytoplasm. For electron microscopy, the cells were fixed first in a mixture of 2% paraformaldehyde and 0.05% glutaraldehyde and then with PLP fixative, this being followed by treatment with 0.05% saponin. They were then immunostained using the immunoperoxidase method. The M protein was found to be distributed throughout the cytoplasm and later under the cell membrane, especially at virus budding sites. We also used postembedding immunostaining and freeze-fracture immunostaining to avoid the translocation of M protein caused by the detergent treatment. These techniques confirmed our previous results. Our findings are consistent with the view that the M protein of VSV is synthesized on free ribosomes and is then associated with the cell membrane where viral assembly may occur.S. Ohno was a visiting fellow from the Fogarty International Center at the National Institutes of Health, USA, from September 1981 to August 1983, while some parts of this work were in progress.     

Immunocytochemical localization of vesicular stomatitis virus proteins N and NS with monoclonal antibodies

Summary The purpose of this paper is to describe the immunocytochemical localization of N and NS nucleocapsid proteins of vesicular stomatitis virus in the cells throughout the infectious cycle. N protein was detected in the cytoplasm at 2h after infection and formed small cytoplasmic clusters which progressively increased in size and number. At 5–6 h, it formed large cytoplasmic inclusions. NS protein was detected in the cytoplasm a little later than N protein and showed almost the same immunostaining pattern. However, diffuse background staining of NS protein was identified throughout the cytoplasm by double immunostaining methods. At electron microscopic level, N protein was mostly granular and occasionally organized in strands at 2–3 h. At 5–6 h, numerous immunostained reaction products were organized in strands. The reaction products of NS protein were almost the same as those of N protein with the exception that diffuse background staining was observed. Cos cells, transfected with SV40 vector containing N gene obtained by recombinant DNA technique, showed clusters of N protein, but virtually no strand at electron microscopic levels. The rapid-freezing and deep-etching replica method demonstrated that loosely coiled VSV genome coated with N protein was localized on cytoplasmic sides of cell membranes in the infected cells. These results showed that complete virus genome replication was needed for strand formation of N and NS proteins and suggested that they were bound to VSV genomes in the infected cells.S. Ohno was a visiting fellow from the Fogarty International Center at the National Institutes of Health, while this work was in progress     

Immunocytochemical localization of basic protein in major dense line regions of central and peripheral myelin

To localize basic protein (BP) in the lamellar structure of central and peripheral myelin, we perfused newborn and 7-11-day rat pups with a phosphate-buffered fixative that contained 4% paraformaldehyde and 0.05 or 0.2% glutaraldehyde. Teased, longitudinally split or "brush" preparations of optic and trigeminal nerves were made by gently teasing apart groups of myelinated fibers with fine forceps or needles. Some of these preparations were immunostained without pretreatment in phosphate-buffered antiserum to BP according to the peroxidase-antiperoxidase method. Others were pretreated in ethanol before immunostaining. Then, all of them were dehydrated, embedded in Epon, and sectioned for electron microscopic study. In optic and trigeminal nerves that were not pretreated, myelin, glial cells, and their organelles were well preserved. BP immunostaining was present on cytoplasmic faces of oligodendroglial and Schwann cell membranes that formed mesaxons and loose myelin spirals. In compact central and peripheral myelin, reaction product was located in major dense line regions, and the myelin periodicity was the same as that observed in unstained control myelin that had been treated with preimmune serum. In ethanol-pretreated tissue, the myelin periodicity was reduced but dense line staining still was present. Our immunocytochemical demonstration of dense line localization of BP in both CNS and PNS myelin that was not disrupted or pretreated with solvents is important because of conflicting evidence in earlier immunostaining studies. Our results also support biochemical and histochemical evidence suggesting that BP exists in vivo as a membrane protein interacting with lipids on the cytoplasmic side of the bilayer in the spirally wrapped compact myelin membrane.     

The herpes simplex virus UL20 protein compensates for the differential disruption of exocytosis of virions and viral membrane glycoproteins associated with fragmentation of the Golgi apparatus.

The Golgi apparatus is fragmented and dispersed in Vero cells but not in human 143TK- cells infected with wild-type herpes simplex virus 1. Moreover, a recombinant virus lacking the gene encoding the membrane protein UL20 (UL20- virus) accumulates in the space between the inner and outer nuclear membranes of Vero cells but is exported and spreads from cell to cell in 143TK- cell cultures. Here we report that in Vero cells infected with UL20- virus, the virion envelope glycoproteins were of the immature type, whereas the viral glycoproteins associated with cell membranes were fully processed up to the addition of sialic acid, a trans-Golgi function. Moreover, the amounts of viral glycoproteins accumulating in the plasma membranes were considerably smaller than those detected on the surface of Vero cells infected with wild-type virus. In contrast, the amounts of viral glycoproteins present on the plasma membranes of 143TK- cells infected with wild-type or UL20- virus were nearly identical. We conclude that (i) in Vero cells infected with UL20- virus the block in the export of virions is at the entry into the exocytic pathway, and a second block in the exocytosis of viral glycoproteins associated with cytoplasmic membranes is due to an impairment of transport beyond Golgi fragments containing trans-Golgi enzymes and not to a failure of the Golgi oligosaccharide-processing functions; (ii) these defects are manifested in cells in which the Golgi apparatus is fragmented; and (iii) the UL20 protein compensates for these defects by enabling transport to and from the fragmented Golgi apparatus.     

Specific incorporation of host cell surface proteins into budding vesicular stomatitis virus particles

The specific incorporation of cell surface proteins into budding Vesicular Stomatitis Virus (VSV) particles was shown by two approaches. In the first, monolayer cultures of Vero or L cells were labeled by lactoperoxidase-catalyzed iodination and the cells were then infected with VSV. Approximately 2% of the cell surface ¹²⁵1 radioactivity was incorporated into particles which co-purify with normal, infectious virions by both velocity and equilibrium gradient centrifugation and which are precipitated by antiserum specific for the VSV glycoprotein. Control experiments establish that these ¹²⁵I-labeled particles are not cell debris or cellular material which aggregate with or adhere to VSV virions. VSV virions contain only a subset of the 10–15 normal ¹²⁵1-labeled cell surface polypeptides resolved by SDS gel electrophoresis; VSV grown in L cells and Vero cells incorporate different host polypeptides. In a second approach, Vero cells were labeled with ³⁵S-methione, then infected with VSV. Two predominant host polypeptides (molecular weights 110,000 and 20,000) were incorporated into VSV virions. These proteins, like VSV G protein, are exposed to the surface of the virion. They co-migrate with the major incorporated ¹²⁵1 host polypeptides. These host proteins are present in approximately 10 and 80 copies, respectively, per virion. Specific incorporation of host polypeptides into VSV virions does not require the presence of viral glycoprotein. This was shown by use of a ts VSV mutant defective in maturation of VSV G protein to the cell surface. Budding from infected cells are noninfectious particles which contain all the viral proteins except for G; these particles contain the same proportion and spectrum of ¹²⁵1-labeled host surface polypeptides as do wild-type virions. These results extend previous conclusions implicating the submembrane viral matrix protein, or the viral nucleocapsid, as being of primary importance in selecting cell surface proteins for incorporation into budding VSV virions.     

Loss of antibody binding to prefixed cells: Fixation parameters for immunocytochemistry

Summary The denaturing effects of various types of fixative solutions on 5 cell surface antigens on mouse T-lymphocytes (Thy-1, T-200, Lyt-1, Lyt-2, and Th-B) were studied. For this purpose, cells were fixed with paraformaldehyde, glutaraldehyde, acrolein and osmium tetroxide at various concentrations. Fixed cells were then incubated with monoclonal antibodies and appropriate second stage antibodies or conjugates. The degree of antibody binding to these cells was determined quantitatively using flow-cytometry with a fluorescence-activated cell sorter or with a semi-automatic micro-ELISA system. The data obtained indicate that paraformaldehyde and glutaraldehyde preserve all five tested antigen molecules, whereas antibody binding to cells fixed in acrolein and osmium tetroxide is rapidly reduced at increasing concentrations of the fixative. The optimal concentration of paraformaldehyde is in the range 0.5–1%, whereas glutaraldehyde should be used at concentrations between. 0.05 and 0.1%. Cells fixed with 0.5% paraformaldehyde or with 0.05% glutaraldehyde are stable and can be stored for at least one week prior to incubation with antibodies.     

Assembly of viral membranes: nature of the association of vesicular stomatitis virus proteins to membranes.

To explore the interaction of vesicular stomatitis virus (VSV) proteins with cellular membranes, we have isolated membranes from infected cells that have been radioactively pulse-labeled. We have found conditions of isolation that result in membrane preparation which contain primarily the VSV membrane protein (M) and glycoprotein (G). Both of these proteins are very firmly attached to membranes: conditions known to release peripherally associated membrane proteins from membranes (S. Razin, Biochim, Biophys. Acta 265:241-246, 1972; S. J. Singer, Annu. Rev. Biochem. 43:805-826, 1974; S. J. Singer and G. L. Nicholson, Science 175:720-731, 1972) are ineffective in detaching either the G or the M protein. The results of trypsin digestion of these membrane fractions suggest that the M protein resides primarily on one side, the cytoplasmic side of cellular membranes, whereas the glycoprotein has been transported to the lumen of the membrane vesicle. However, we present evidence that the glycoprotein is transmembranal and that approximately 3,000 daltons of one end of the molecule is on the cytoplasmic side of the membrane. We have also found that undenatured VSV M protein contains a trypsin-resistant core with a molecular weight of 22,000. This region of the M protein is trypsin-resistant regardless of its association with membranes.     

Immunocytochemical evidence for colocalization in neurite growth cones of actin and myosin and their relationship to cell-substratum adhesions

Sensory neurons from 8- to 11-day chick embryos were cultured on polyornithine-treated coverslips, fixed with glutaraldehyde, and stained for immunofluorescent localization of actin. Actin was distributed in a fibrous form in the growth cones, extending into filopodia and lamellipodial expansions of the growth cone margin. Often, these actin fibers were located at sites of linear adhesions to the glass substratum, as viewed by interference reflection optics. Our antisera to myosin did not recognize myosin in glutaraldehyde-fixed cells, and paraformaldehyde, which preserves the antigenicity of myosin, did not fix embryonic neurons well. Thus, myosin was localized in NGF-stimulated PC12 cells, whose morphology is better preserved by paraformaldehyde. Within the growth cones of PC12 neurites, actin and myosin are distributed into fibrous arrays which resemble the actin fibers seen in the growth cones of sensory neurons. Thus, actomyosin-like contractile forces may be exerted in neurite growth cones. These forces may act in concert with cell-substratum adhesive bonds to move the growth cone across the substratum or move organelles within the growth cone.     

Application of avidin-biotin-peroxidase complex method with a monoclonal antibody to human Ia-like antigen in immunoelectron microscopy

Monoclonal antibody (MAb) to human Ia-like (HLA-DR) antigen was applied with the avidin-biotin-peroxidase complex (ABC) immunostaining method to localize the Ia-like antigen at the electron microscopic level. Our results indicated that in human tonsils and adenoids fixed with 4-6% phosphate-buffered paraformaldehyde for 4-6 hr, sharply delineated electron-dense products of the antigen and antibody complex were detectable on the outer cell membranes of lymphoblasts, lymphocytes, reticular cells, and macrophages. In our study, the vibratome sections of the paraformaldehyde-fixed, pre-embedding immunostained tissues consistently showed more satisfactory morphology than frozen sections. The combined use of the anti-human Ia monoclonal antibody and the ABC procedure with paraformaldehyde fixation provides a simple and sensitive method to study at the ultrastructural level the Ia-like antigen-bearing cells, which are vital in the immune response.     

Ultrastructure and location of cytoplasmic fibrils inSpiroplasma floricola OBMG

The ultrastructure ofSpiroplasma floricola OBMG was investigated to identify subcellular structures that might be involved in motility and helicity. Optimal preservation for thin sectioning was achieved with either glutaraldehyde or a mixture of glutaraldehyde plus paraformaldehyde followed by OsO₄ and uranyl acetate. In thin sections, a 94-nm-wide band consisting of 4-nm-diameter fibrils was observed apposed to the cytoplasmic side of the plasma membrane. The band of fibrils extended axially the entire length of the cell. The addition of rethenium red to fixative solutions resulted in condensation of the fibrils. Freeze-substitution increased the apparent thickness of membranes but did not improve preservation of the fibrils. Freeze-fracturing revealed a 99-nm-wide zone containing few particles in fractured membrane surfaces. Treatment with deoxycholate or Triton X-100 to dissolve membranes yielded bands of fibrils comparable to those seen in thin sections. Based on these findings, a model indicating the intracellular location of the fibrils is proposed.     

Cell membrane coating with glutaraldehyde: application to a versatile solid-phase assay for thyroid membrane proteins and molecules interacting with thyroid membranes

In defined conditions, glutaraldehyde was shown to tightly bind cell membranes to flexible microtiter plates without significant alteration of the antigenic and functional properties of membrane proteins. In the presence of 0.06% glutaraldehyde, human thyroid membranes were bound to plastic firmly enough to resist numerous washing and flicking steps; the coated membranes remained almost unaltered with regard to monoclonal antibody and thyrotropin binding as well as adenylate cyclase and peroxidase activities. Based on the use of thyroid membrane-coated microtiter plates, a versatile solid-phase assay was developed which allowed screening of anti-membrane monoclonal antibodies, detection of thyrotropin-displacing activity in hormone and antibody preparations, and monitoring of fractionation experiments of solubilized membrane antigens and thyrotropin receptor. It was concluded that the use of glutaraldehyde for coating cell membranes to flexible microtiter plates enabled the establishment of simple, rapid, and reliable assays for detection and quantitation of membrane proteins and molecules interacting with membranes.     

High-resolution three-dimensional views of membrane-associated clathrin and cytoskeleton in critical-point-dried macrophages

We obtained high-resolution topographical information about the distribution of clathrin and cytoskeletal filaments on cytoplasmic membrane surfaces of macrophages spreading onto glass coverslips by both critical-point drying of broken-open cells and preparation of rotary platinum replicas. Irregular patches of the adherent ventral surface of the plasma membrane were exposed in these cells, and large areas of these exposed membranes were covered with clathrin-coated patches, pits, and vesicles. Various amounts of cytoskeleton were attached to the plasma membranes of these spreading cells, either as distinct starlike foci, or as individual filaments and bundles radiating out from the cytoskeletal meshwork. In newly adherent cells a well developed Golgi-GERL complex, characterized by smooth, dish-like cisternae associated with rough endoplasmic reticulum, was observed. There were many coated vesicles budding off from the Golgi cisternae, and these were predominantly of the large type (150 nm) usually associated with the plasma membrane. In critical-point-dried samples, both cytoskeleton and membranes were preserved in detail comparable to that of quick-frozen samples, after appropriate fixation. Rotary replication of critical-point-dried cells provides a rapid, easily controlled, and generally easy to perform method for obtaining samples of exposed membrane large enough to permit quantification of membrane- associated clathrin and cytoskeleton under various experimental conditions.     

Defective interfering particles modulate VSV infection of dissociated neuron cultures.

Infection of dissociated neuron cultures of mice with VSV and its defective particle DI-T was studied using fluorescent light microscopy as well as transmission and scanning electron microscopy. When cultures are infected with wild virus, VSV replicates selectively in neurons, producing cell death within 24-48 hr. Sensory and immature neurons express viral antigen most rapidly. Viral antigen and viral budding sites are detected along the neuron soma and dendrites. When large amounts of DI-T particles are added to the wild virus inoculum, viral growth is completely suppressed in mature neurons, the cell killing effects of VSV are considerably delayed and co-infected cultures survive for 5-16 days. Viral antigen accumulates in cytoplasmic inclusions and on the membrane of neuron cell somas and dendrites in the virtual absence of viral assembly. Identical modulation of VSV infection in mature neuron cultures is obtained when DI-T particles are added before or after the wild virus, but ultraviolet inactivation of DIs completely abolishes their protective effect. Immature neurons or Vero cells cannot be protected from acute cytopathic changes by an equivalent amount of DI particles. Thus DIs interfere with replication and assembly of the wild virus and attenuate cell killing effects in mature neurons in vitro.     

Cellular origin and ultrastructure of membranes induced during poliovirus infection.

Poliovirus RNA replicative complexes are associated with cytoplasmic membranous structures that accumulate during viral infection. These membranes were immunoisolated by using a monoclonal antibody against the viral nonstructural protein 2C. Biochemical analysis of the isolated membranes revealed that several organelles of the host cell (lysosomes, trans-Golgi stack and trans-Golgi network, and endoplasmic reticulum) contributed to the virus-induced membranous structures. Electron microscopy of infected cells preserved by high-pressure freezing revealed that the virus-induced membranes contain double lipid bilayers that surround apparently cytosolic material. Immunolabeling experiments showed that poliovirus proteins 2C and 3D were localized to the same membranes as the cellular markers tested. The morphological and biochemical data are consistent with the hypothesis that autophagy or a similar host process is involved in the formation of the poliovirus-induced membranes.     

Cross-reactivity of antibody against SARS-coronavirus nucleocapsid protein with IL-11

Infection of SARS-associated coronavirus (SARS-CoV) induced a strong anti-nucleocapsid (anti-N) antibody response. However, the pathophysiological significance of the anti-N antibodies in SARS pathogenesis is largely unknown. To profile the anti-N antibodies, a phage-displayed scFv library was prepared from mice immunized with heat-inactivated SARS-CoV-infected Vero E6 cell lysate. Specific anti-N scFvs were isolated by panning against a recombinant nucleocapsid protein and reactivity was confirmed with phage-ELISA. Sequence analysis indicated that two of the isolated anti-N scFv clones were identical and displayed a high homology with an scFv specific for interleukin 11 (IL-11), an anti-inflammatory cytokine derived from bone marrow stroma cells. In a neutralization assay, IL-11-induced STAT 3 phosphorylation in rat intestinal epithelial IEC-18 cells was completely suppressed by the anti-N scFv clone L9N01.     

Localization and putative function of the UL20 membrane protein in cells infected with herpes simplex virus 1.

The UL20 protein of herpes simplex virus 1, an intrinsic membrane protein, is required in infected Vero cells in which the Golgi apparatus is fragmented for the transport of virions from the space between the inner and outer nuclear membranes and for the transport of fully processed cell membrane-associated glycoproteins from the trans-Golgi to the plasma membrane. It is not required in the human 143TK- cell line, in which the Golgi apparatus remains intact. We report the following. (i) The UL20 protein was detected in infected cells beginning at 6 h postinfection and was regulated as a gamma 1 gene. (ii) Pulse-chase experiments revealed no detectable alteration in the mobility of the UL20 protein in polyacrylamide gels. (iii) In both infected Vero and infected 143TK- cells, the UL20 protein was detected by immunofluorescence in association with nuclear membranes and in the cytoplasm. Some of the cytoplasmic fluorescence colocalized with beta-COP, a protein associated with Golgi-derived transport vesicles. UL20 protein was present in virions purified from the extracellular space but could not be detected in the plasma membrane. These results are consistent with the hypothesis that UL20 is a component of virion envelopes and membranes of virion transport vesicles and is selectively retained from the latter in a Golgi compartment.     

Delayed appearance of pseudotypes between vesicular stomatitis virus influenza virus during mixed infection of MDCK cells.

In intact Madin-Darby canine kidney (MDCK) cell monolayers, vesicular stomatitis virus (VSV) matures only at basolateral membranes beneath tight junctions, whereas influenza virus buds from apical cell surfaces. Early in the growth cycle, the viral glycoproteins are restricted to the membrane domain from which each virus buds. We report here that phenotypic mixing and formation of VSV pseudotypes occurred when influenza virus-infected MDCK cells were superinfected with VSV. Up to 75% of the infectious VSV particles from such experiments were neutralized by antiserum specific for influenza virus, and a smaller proportion (up to 3%) were resistant to neutralization with antiserum specific for VSV. The latter particles, which were neutralized by antiserum to influenza A/WSN virus, are designated as VSV(WSN) pseudotypes. During mixed infections, both wild-type viruses were detected 1 to 2 h before either phenotypically mixed VSV or VSV(WSN) pseudotypes. Coincident with the appearance of cytopathic effects in the monolayer, the yield of pseudotypes rose dramatically. In contrast, in doubly infected BHK-21 cells, which do not show polarity in virus maturation sites and are not connected by tight junctions, VSV(WSN) pseudotypes were detected as soon as VSV titers rose to the minimum levels which allowed detection of pseudotypes, and the proportion observed remained relatively constant at later times. Examination of thin sections of doubly infected MDCK monolayers revealed that polarity in maturation sites was preserved for both viruses until approximately 12 h after inoculation with influenza virus, when disruption of junctional complexes was evident. Even at later periods, the majority of each virus type was associated with its normal membrane domain, suggesting that the sorting mechanisms responsible for directing the glycoproteins of VSV and influenza virus to separate surface domains continue to operate in doubly infected MDCK cells. The time course of VSV(WSN) pseudotype formation and changes in virus maturation sites are compatible with progressive mixing of viral glycoproteins at either intracellular or plasma membranes of doubly infected cells.     

Ultrastructural Localization of P2 Protein in Actively Myelinating Rat Schwann Cells

Abstract: The myelin specific protein, P₂, was localized immunocytochemically in electron micrographs of 4-day-old rat peripheral nerve by a preembedding technique. P₂ staining was restricted to Schwann cells that had established a one-to-one relationship with an axon. P₂ antiserum produced a diffuse staining throughout the entire cytosol of myelinating Schwann cells. In addition, the cytoplasmic side of Schwann cell plasma membranes and the membranes of cytoplasmic organelles that were exposed to cytosol were stained by P₂ antiserum. This cytoplasmic localization of P₂ protein is similar to that described for soluble or peripheral membrane proteins that are synthesized on free ribosomes. P₂ antiserum stained the cytoplasmic side of Schwann cell membranes that formed single or multiple loose myelin spirals around an axon. In the region of the outer mesaxon, P₂ antiserum stained the major dense line of compact myelin. These results demonstrate that P₂ protein is located on the cytoplasmic side of compact myelin membranes and are consistent with biochemical studies demonstrating P₂ to be a peripheral membrane protein.