Amphilphilic nature of spiralin, the major protein of the Spiroplasma citri cell membrane.

Spiralin could not be solubilized in the absence of detergents, and it was shown by charge-shift crossed immunoelectrophoresis that this protein was capable of binding detergents under nondenaturing conditions. These properties indicate the amphiphilic nature of spiralin, which therefore should be regarded as an intrinsic membrane protein. The efficiency of mild (ionic and neutral) detergents to solubilize spiralin was as follows: deoxycholate greater than lauroyl sarcosinate, cholate, taurocholate, taurodeoxycholate greater than Triton X-100 greater than Brij 58 greater than Tween 20, indicating that mild ionic detergents were more effective than neutral ones. Solubilization of spiralin was quantitative with sodium deoxycholate. It was also shown that although a membrane protein is not extractable by a given detergent from the membrane, this does not necessarily mean that the protein is not soluble in this detergent.     

Proteins of the kidney microvillar membrane. Aspartate aminopeptidase: purification by immunoadsorbent chromatography and properties of the detergent- and proteinase-solubilized forms.

Aminopeptidase A (aspartate aminopeptidase, EC 3.4.11.7) was purified 2000-fold from pig kidney cortex. The essential step in the purification was chromatography on an immunoadsorbent column prepared from a rabbit antiserum raised against pig intestinal aminopeptidase A. Glutamyl and aspartyl substrate were attacked most rapidly and their hydrolyses were stimulated by Ca2+. The 2-naphthylamide derivatives of neutral and basic amino acids were also hydrolysed by aminopeptidase A, but at rates about two orders of magnitude lower, and Ca2+ was inhibitory. The possibility that these atypical substrates were hydrolysed by traces of aminopeptidase M (EC 3.4.11.2) contaminating the preparation could be excluded on several grounds. Aminopeptidase A was sensitive to inhibition by chelating agents and the inactive enzyme could be reactivated by Ca2+ or Mn2+. Atomic absorption spectrophotometry revealed 1 g-atom of Ca/143000 g of protein. Two forms of the enzyme were purified: an amphipathic form solubilized from the membrane by Triton X-100 (detergent form) and a hydrophilic form released by incubation with trypsin (proteinase form). The detergent form exhibited charge-shift in crossed immunoelectrophoresis when anionic or cationic detergents were present. On gel filtration, mol.wts. of 350000--400000 and 270000 were calculated for the detergent and proteinase forms. Electron microscopy after negative staining of the proteinase form revealed a dimeric structure. Electrophoresis of either form in the presence of sodium dodecyl sulphate revealed four polypeptides with mobilities corresponding to apparent mol.wts. of 155000, 110000, 90000 and 45000. All four bands stained positively for carbohydrate. Pig serum possesses weak aminopeptidase A activity; immunological experiments showed it to be a similar protein.     

Horse kidney neutral alpha-D-glucosidase: purification of the detergent-solubilized enzyme; comparison with the proteinase-solubilized forms

Neutral alpha-D-glucosidase (alpha-D-glucoside glucohydrolase, EC 3.2.1.20) from horse kidney brush-border membranes was solubilized using Emulphogene BC 720 and purified by an affinity chromatography technique. The enzyme preparation (390-fold purified), which was free of other known microvillus hydrolases, exhibited one precipitate line in crossed immunoelectrophoresis and migrated as a single band on sodium dodecyl sulfate polyacrylamide gel electrophoresis. Several criteria (charge-shift crossed immunoelectrophoresis and hydrophobic chromatography) revealed the purified detergent form of the enzyme to be an amphipathic molecule. The papain treatment of either brush-border membrane vesicles or the purified detergent form of neutral alpha-D-glucosidase released an enzymatic form devoid of these amphipathic properties. Conversely, after trypsin treatment of the "d' form of the enzyme, two enzymatic forms were obtained: the first and major form retained these amphipathic properties; the second form exhibiting the same properties as the papain-released form. Furthermore, only a very small amount of neutral alpha-D-glucosidase can be released after trypsin solubilization of brush-border membrane vesicles, and the released enzyme did not exhibit amphipathic properties. These results were interpreted as meaning that the trypsin attack site on the detergent form of the enzyme had either poor affinity for, or obstructed access to, the proteinase when the enzyme was integrated in native membrane or in Triton X-100 micelles, whereas the proteolytic site of the papain was always accessible.     

Characterization of an active hydrophilic erythrocyte membrane acetylcholinesterase obtained by limited proteolysis of the purified enzyme

Purified human erythrocyte membrane acetylcholinesterase was subjected to limited proteolysis with papain. This treatment generated a hydrophilic form of the enzyme as determined by charge-shift crossed immunoelectrophoresis and by binding to phenyl-Sepharose. The hydrophilic enzyme was stable and its activity was independent of the presence of amphiphiles. Electroimmunochemical analysis showed no antigenic difference between the two enzyme forms. Although the proteolytic treatment only brought about a small change in molecular weight, marked differences in the hydrodynamic properties were encountered. The Stokes radius decreased from 8.2 to 5.9 nm and the sedimentation coefficient increased from 6.3 to 7.0 S. The results are consistent with the view that a short hydrophobic peptide responsible for the amphipatic character of acetylcholinesterase is removed by the treatment with papain.     

Proteins of the kidney microvillar membrane. The amphipathic forms of endopeptidase purified from pig kidneys.

The purification of detergent-solubilized kidney microvillar endopeptidase (EC 3.4.24.11) by immuno-adsorbent chromatography is described. The product (the d-form) was 270-fold purified compared with the homogenate of kidney cortex and was obtained in a yield of 5%. It was free of other peptidase activities and homogeneous by electrophoretic analyses. It contained about 15% carbohydrate and one Zn atom/subunit. Two trypsin-treated forms were also characterized. One (dt-form) was obtained by treatment of the d-form. The other (tt-form) was the result of solubilizing the membrane by treatment with toluene and trypsin. All three forms had apparent subunit Mr values of approx. 89 000, but the d-form appeared to be slightly larger than the other two. Estimates of Mr by gel filtration showed that of the tt-form to be 216 000 whereas those of the other forms were 320 000. An estimate of the detergent (Triton X-100) bound to the d- and dt-forms accounted for this difference. By several criteria, including charge-shift crossed immunoelectrophoresis and hydrophobic chromatography, the d- and dt-forms were shown to be amphipathic molecules. In contrast, the tt-form was hydrophilic in its properties. Differences in ionic properties were also noted, consistent with the loss, in the case of the dt-form, of a positively charged peptide. The results indicate that the native endopeptidase is a dimeric molecule, each subunit being anchored in the membrane by a relatively small region of the polypeptide close to one or other terminus. The d- and dt-forms had similar enzyme activity when assayed by the hydrolysis of 125I-insulin B-chain. Chelating agents and phosphoramidon inhibited the endopeptidase. The kinetic constants were determined by a new two-stage fluorimetric assay using glutarylglycylglycylphenylalanine 2-naphthylamide as substrate and aminopeptidase N (EC 3.4.11.2) to hydrolyse phenylalanine 2-naphthylamide. The Km was 68 microM and Vmax. 484nmol X min-1 X (mg of protein)-1.     

Membrane anchors of alkaline phosphatase and trehalase associated with the plasma membrane of larval midgut epithelial cells of the silkworm, Bombyx mori

The larval midgut epithelial cell of the silkworm, Bombyx mori, has two forms of alkaline phosphatase and trehalase, soluble and membrane-bound. Alkaline phosphatase and trehalase of the latter form are found in the brush border membrane and the basolateral membrane, respectively. In this work we studied the membrane anchors of these membrane-bound enzymes. Alkaline phosphatase was solubilized by phosphatidyl-inositol-specific phospholipase C, but not by papain. Conversely, trehalase was released from the membrane by papain, but not by phosphatidylinositol-specific phospholipase C. Both enzymes were solubilized in an amphiphilic form with 0.5% Triton X-100 plus 0.5% sodium deoxycholate (pH 7.0). The detergent-solubilized alkaline phosphatase and trehalase were converted to hydrophilic form on incubation with phosphatidylinositol-specific phospholipase C and papain, respectively. The effects of papain on solubilization and conversion of trehalase were completely inhibited by leupeptin. These results suggest that, in the silkworm larvae, alkaline phosphatase is anchored in the brush-border membrane via a glycosyl-phosphatidylinositol, while trehalase is associated with the basolateral membrane through a hydrophobic segment of the polypeptide.     

Immunochemically identical hydrophilic and amphiphilic forms of the bovine adrenomedullary dopamine beta-hydroxylase.

By means of a monospecific antibody, dopamine beta-hydroxylase was monitored immunoelectrophoretically in various extracts of chromaffin granules. Approximately one-third of the dopamine beta-hydroxylase present was located in the membrane fraction and could only be liberated with detergent. The dopamine beta-hydroxylases of the buffer and membrane fractions were antigenically identical, but differed in their amphiphilicity, as demonstrated by the change in precipitation patterns on removal of Triton X-100 from the gel, on charge-shift crossed immunoelectrophoresis and on crossed hydrophobic interaction immunoelectrophoresis with phenyl-Sepharose. Furthermore, immunoelectrophoretic analysis in the presence of Triton X-100 plus the cationic detergent cetyltrimethylammonium bromide indicates additional heterogeneity of the membrane-bound dopamine-beta-hydroxylase. By limited proteolysis with chymotrypsin and thermolysin the amphiphilic form could be convered into its hydrophilic counterpart.     

Proteins of the kidney microvillar membrane. Reconstitution of endopeptidase in liposomes shows that it is a short-stalked protein.

Pig kidney microvillar proteins were extracted with octyl beta-glucoside and reconstituted in liposomes prepared from microvillar lipids of known composition. Four peptidases, namely endopeptidase (EC 3.4.24.11), aminopeptidases N (EC 3.4.11.2) and A (EC 3.4.11.7) and dipeptidyl peptidase IV (EC 3.4.14.5), were shown to be reconstituted. At lipid/protein ratios greater than 4:1, about half the detergent-solubilized protein and nearly all of the activity of the four peptidases were reconstituted. Dissolution of the liposomes with Triton X-100 did not increase the activity of any of these peptidases, a result consistent with an asymmetric, 'right-side-out', orientation of these enzymes. When purified, endopeptidase was subjected to the same procedure; the two amphipathic forms of the enzyme (the detergent form and the trypsin-treated detergent form) were fully reconstituted. The amphiphilic form, purified after toluene/trypsin treatment, failed to reconstitute. Electron microscopy of microvilli showed that the appearance of the surface particles was profoundly altered by treatment with papain. Before treatment, the microvilli were coated with particles of stalk lengths ranging from 2.5 to 9 nm. After papain treatment nearly all the particles had stalks of 2-3 nm. Reconstituted microvillar proteins in liposomes showed the same heterogeneity of stalk length. In contrast, liposomes containing reconstituted endopeptidase revealed a very homogeneous population of particles of stalk length 2 nm. Since the smallest dimension of a papain molecule is 3.7 nm, the ability of papain, and other proteinases of similar molecular size, to release microvillar enzymes is crucially affected by the length of the junctional peptide that constitutes the stalk of this type of membrane protein.     

Crossed immunoelectrophoresis, in the presence of tween 20 or sodium deoxycholate, of purified membrane proteins from Acholeplasma laidlawii.

Five membrane proteins from Acholeplasma laidlawii have been previously purified on a large scale. These proteins have been used to establish the relationship between the precipitation lines obtained by crossed immunoelectrophoresis of solubilized cell membrane proteins from A. laidlawii in the presence of the neutral detergent Tween 20 or those obtained in the presence of the anionic detergent sodium deoxycholate. This relationship, which was unambiguously established for four of the five proteins, was determined by tandem or "parallel" crossed immunoelectrophoresis of the sodium deoxycholate-solubilized membrane together with the purified proteins. Membranes from strain A of A. laidlawii were composed of proteins, which were immunologically related to and probably identical to membrane proteins from strain B of this organism.     

Proteins of the kidney microvillar membrane. The amphipathic form of dipeptidyl peptidase IV.

Dipeptidyl peptidase IV was solubilized from the microvillar membrane of pig kidney by Triton X-100. The purified enzyme was homogeneous on polyacrylamide-gel electrophoresis and ultracentrifugation, although immunoelectrophoresis indicated that amino-peptidase M was a minor contaminant. A comparison of the detergent-solubilized and proteinase (autolysis)-solubilized forms of the enzyme was undertaken to elucidate the structure and function of the hydrophobic domain that serves to anchor the protein to the membrane. No differences in catalytic properties, nor in sensitivity to inhibition by di-isopropyl phosphorofluoridate were found. On the other hand, several structural differences could be demonstrated. Both forms were about 130,000 subunit mol.wt., but the detergent form appeared to be larger by no more than about 4,000. Electron microscopy showed both forms to be dimers, and gel filtration revealed a difference in the dimeric mol.wt. of about 38 000, mainly attributable to detergent molecules bound to the hydrophobic domain. Papain converted the detergent form into a hydrophilic form that could not be distinguished in properties from the autolysis form. A hydrophobic peptide of about 3500 mol.wt. was identified as a product of papain treatment. The detergent and proteinase forms differed in primary structure. Partial N-terminal amino acid sequences were shown to be different, and the pattern of release of amino acids from the C-terminus by carboxypeptidase Y was essentially similar. The results are consistent with a model in which the protein is anchored to the microvillar membrane by a small hydrophobic domain located within the N-terminal amino acid sequence of the polypeptide chain. The significance of these results in relation to biosynthesis of the enzyme and assembly in the membrane is discussed.     

Studies on the enzymology of purified preparations of brush border from rabbit kidney

1. A method for the preparation of brush border from rabbit kidneys is described. Contamination by other organelles was checked by electron microscopy and by the assay of marker enzymes and was low. 2. Seven enzymes, all hydrolases, were substantially enriched in the brush-border preparation and are considered to be primarily located in this structure. They are: alkaline phosphatase, maltase, trehalase, aminopeptidase A, aminopeptidase M, gamma-glutamyl transpeptidase and a neutral peptidase assayed by its ability to hydrolyse [(125)I]iodoinsulin B chain. 3. Adenosine triphosphatases were also present in the preparation, but showed lower enrichments. 4. Alkaline phosphatase was the most active phosphatase present in the preparation. The weak hydrolysis of AMP may well have been due to this enzyme rather than a specific 5'-nucleotidase. 5. The two disaccharidases in brush border were distinguished by the relative heat-stability of trehalase compared with that of maltase. 6. The individuality of the four peptidases was established by several means. The neutral peptidase and aminopeptidase M, both of which can attack insulin B chain, differed not only in response to inhibitors and activators but also in the inhibitory effect of a guinea-pig antiserum raised to rabbit aminopeptidase M. This antiserum inhibited both the purified and the brush-border activities of aminopeptidase M. The neutral peptidase and gamma-glutamyl transpeptidase were unaffected but aminopeptidase A was weakly inhibited. The characteristic responses to Ca(2+) and serine with borate served to distinguish aminopeptidase A and gamma-glutamyl transpeptidase from other peptidases. 7. No dipeptidases, tripeptidases or carboxypeptidases were identified as brush-border enzymes. 8. Incubation of brush border with papain released almost all the aminopeptidase M activity but only about half the activities of maltase, gamma-glutamyl transpeptidase and aminopeptidase A. No release of alkaline phosphatase, trehalase or the neutral peptidase was observed.     

Proteins of the kidney microvillar membrane. Asymmetric labelling of the membrane by lactoperoxidase-catalysed radioiodination and by photolysis of 3,5-di[125I]iodo-4-azidobenzenesulphonate.

Two methods were used to label pig kidney microvillar membrane proteins from the luminal and cytoplasmic surfaces of closed membrane vesicles. The first method was lactoperoxidase-catalysed radioiodination. The enzyme reagents, lactoperoxidase and glucose oxidase, were positioned inside the vesicles before sealing or externally after sealing, iodination being initiated by the subsequent addition of glucose and 125I-. After resolution of the labelled proteins by electrophoresis in the presence of dodecyl sulphate, asymmetric labelling patterns on radioautographs were observed. However, the major disadvantage of this method is the high degree of intramembrane labelling of the fatty acid chains of membrane lipids, a reaction that undermines any conclusions about the location of the label in that region of the protein supposedly exposed at the surface of the membrane. The second method overcame this disadvantage. A new hydrophilic photoreagent, 3,5-di[125I]iodo-4-azidobenzesulphonate, was synthesized via the intermediate, diazotized 3,5-di[125I]iodosulphanilic acid. It was transported by a Na+-dependent system into microvillar vesicles, thus permitting labelling from either side of the membrane when the vesicles were photolysed. The labelling of membrane lipids was less than with the first method and was essentially confined to the polar headgroups. The activity of several microvillar peptidases survived the labelling reaction and they could be identified in the immunoprecipitates after resolution of the detergent-solubilized membrane proteins by crossed-immunoelectrophoresis. Treatment with papain converted the detergent-solubilized form of susceptible enzymes into the proteinase-solubilized form, which lacked the intramembrane domain and any portion exposed at the cytoplasmic surface. Radioautography established that aminopeptidases M and A, dipeptidyl peptidase IV and neutral endopeptidase were transmembrane proteins. This novel approach to the investigation of membrane topology may be applicable to other complex membranes.     

Topological studies on the hydrolases bound to the intestinal brush border membrane. I. Solubilization by papain and triton X-100

Papain digestion of closed, right side out vesicles from pig, rat and rabbit jejunum brush border induces the release of the hydrolases bound to the membrane without grossly affecting the lipid bilayer limiting the vesicles. This observation definitely proves that intestinal hydrolases are surface components attached to the external side of the membrane. All proteins released by papain could be identified by electrophoresis and immunoelectrophoresis to already known intestinal hydrolases, with the exception of an unidentified substance strongly stained by the Schiff's reagent.The early observation that the aminopeptidase form released from pig brush border by Triton X-100 is different from that released by papain was extended to other hydrolases from pig, rat and rabbit. In some cases, the Triton-released form could be converted by further proteolytic digestion into a new form similar to that liberated by papain. These facts may be related to the existence of hydrophobic anchors retaining the intestinal hydrolases to the membrane surface.     

Enzymic solubilization of the human intestinal brush border membrane enzymes.

The releases of proteins, maltase, lactase, sucrase, trehalase, alkaline phosphatase, gamma-glutamyltransferase and leucylnaphthylamide-hydrolyzing activity from human intestinal brush bborder membrane vesicles by various enzymes (especially pancreatic proteases) have been studied. The brush border membrane enzymes are not solubilized by digestion with trypsin and chymotrypsin but are largely released after treatment with papain or elastase. Most of the enzymes are fully active after the proteolytic treatment. All proteins released by papain and elastase have been identified by electrophoresis to already known intestinal hydrolases. Electron microscopy of brush border membrane vesicles demonstrates "knob-like" structures (particles) attached to the external side of the membrane. During papain treatment, enzyme removal runs parallel with the disappearance of the particles. During elastase treatment it is not possible to correlate the release of the enzymic activities with the removal of the particles. The results indicate that most of the intestinal hydrolases are surface components attached to the external side of the membrane. They are in accord with the concept that the brush border membrane enzymes are organized within the membrane in a mosaic-like pattern.     

Enzymic solubilization of the human intestinal brush border membrane enzymes

The releases of proteins, maltase, lactase, sucrase, trehalase, alkaline phosphatase, γ-glutamyltransferase and leucylnaphthylamide-hydrolyzing activity from human intestinal brush border membrane vesicles by various enzymes (especially pancreatic proteases) have been studied.The brush border membrane enzymes are not solubilized by digestion with trypsin and chymotrypsin but are largely released after treatment with papain or elastase. Most of the enzymes are fully active after the proteolytic treatment. All proteins released by papain and elastase have been identified by electrophoresis to already known intestinal hydrolases.Electron microscopy of brush border membrane vesicles demonstrates “knob-like” structures (particles) attached to the external side of the membrane. During papain treatment, enzyme removal runs parallel with the disappearance of the particles. During elastase treatment it is not possible to correlate the release of th enzymic activities with the removal of the particles.The results indicate that most of the intestinal hydrolases are surface components attached to the external side of the membrane. They are in accord with the concept that the brush border membrane enzymes are organized within the membrane in a mosaic-like pattern.     

Immunoelectrophoretic characterization of the ADP/ATP carrier from heart, kidney, and liver

Earlier studies gave an indication for an organ specificity of the ADP/ATP carrier. We used a modified charge-shift crossed immunoelectrophoresis for a more precise immunochemical characterization of this detergent-solubilized hydrophobic membrane protein. Immunological differences between the carrier protein from heart, kidney, and liver were demonstrated by a different electrophoretic migration of the three ligand-protein complexes in the first dimension and a distinct staining intensity, sharpness, and shape of the precipitates in the second dimension. However, the antibodies against the heart and kidney protein showed a cross-reactivity between the three antigens. The results are consistent with the view that the ADP/ATP carrier has organ-specific antigenic determinants although there is a partial identity between the carrier proteins from heart, kidney, and liver.     

Angiotensin III: A Potent Inhibitor of Enkephalin-Degrading Enzymes and an Analgesic Agent

Various angiotensins, bradykinins, and related peptides were examined for their inhibitory activity against several enkephalin-degrading enzymes, including an aminopeptidase and a dipeptidyl aminopeptidase, purified from a membrane-bound fraction of monkey brain, and an endopeptidase, purified from the rabbit kidney membrane fraction. Angiotensin derivatives having a basic or neutral amino acid at the N-terminus showed strong inhibition of the aminopeptidase. Dipeptidyl aminopeptidase was inhibited by angiotensins II and III and their derivatives, whereas the endopeptidase was inhibited by angiotensin I and its derivatives. The most potent inhibitor of aminopeptidase and dipeptidyl aminopeptidase was angiotensin III, which completely inhibited the degradation of enkephalin by enzymes in monkey brain or human CSF. The Ki values for angiotensin III against aminopeptidase, dipeptidyl aminopeptidase, endopeptidase, and angiotensin-converting enzyme, which degraded enkephalin, were 0.66 X 10(-6), 1.03 X 10(-6), 2.3 X 10(-4), and 1.65 X 10(-6) M, respectively. Angiotensin III potentiated the analgesic activity of Met-enkephalin after intracerebroventricular coadministration to mice in the hot plate test. Angiotensin III itself also displayed analgesic activity in that test. These actions were blocked by the specific opiate antagonist naloxone.     

Separation of the proteins of bovine milk-fat-globule membrane by electrofocusing with retention of enzymatic and immunological activity

1. Proteins of fat-globule membrane from bovine milk were solubilized with the non-ionic detergent Triton X-100 in the presence of protease inhibitors. Approximately 25% of the total membrane protein was solubilized and the extracts were shown to contain a sample of most of the major membrane proteins and glycoproteins. 2. The solubilized proteins were separated in flat-beds of Ultrodex by electrofocusing and the pI values for the major proteins, glycoproteins and certain enzymes determined. Several of the proteins displayed marked heterogeneity indicating the existence of protein variants and isoenzymes. Principal pI values for the enzymes assayed were as follows: xanthine oxidase, 7.35--7.55; NADH2: iodonitrotetrazolium reductase, less than 4.5; 5'-nucleotidase, 7.15--7.4; alkaline phosphatase, 5.4--5.7; phosphodiesterase, 4.6--4.8; gamma-glutamyl transpeptidase, 4.4--4.55. 3. Fractions after electrofocusing were analyzed by 'fused rocket' immunoelectrophoresis and crossed immunoelectrophoresis after separation in polyacrylamide gels containing sodium dodecyl sulphate. Major antigens of the membrane include xanthine oxidase and glycoproteins of apparent molecular weights 67 000, 49 500 and 46 000. The latter two components share common antigenic determinants and could not be separated by gel filtration, ion-exchange chromatography, lectin-affinity chromatography or preparative electrofocusing.     

Microvillar membrane neutral endopeptidases

Recent developments on neutral endopeptidase (NEP, EC 3.4.24.11) are described. These include (1) the development of a novel colorimetric assay with a chromogenic substrate (Glutaryl-Gly-Gly-Phe-2-naphthylamide) coupled with aminopeptidase M (EC 3.4.11.2). (2) A detergent form of the pig kidney enzyme has been purified by immuno-adsorbent chromatography and its molecular properties compared with other forms of the enzyme from rabbit kidney and pig intestine. (3) Rat kidney microvilli contain two endopeptidases of about equal activity when assayed with [125I]iodo-insulin B chain as substrate. One is similar to the rabbit and pig endopeptidases in being sensitive to inhibition by phosphoamidon. The other is insensitive to the inhibitor, though susceptible to chelating agents. The two enzymes are resolvable and have been partially characterized. (4) Endopeptidases of the phosphoramidon-sensitive type are present in various tissues in addition to the principal locations in brush borders of kidney and intestine.     

Immunochemical analysis of membrane vesicles from Escherichia coli.

Membrane vesicles isolated from Escherichia coli ML 308--225 have been analyzed by crossed immunoelectrophoresis, and immunoprecipitates corresponding to the following cellular components have been identified: ATPase (EC 3.6.1,3), two or three NADH dehydrogenases (EC 1.6.99.3), D-lactate dehydrogenase (EC 1.1.1.27), glutamate dehydrogenase (EC 1.4.1.4), dihydro-orotate dehydrogenase (EC 1.3.3.1), 6-phosphogluconate dehydrogenase (EC 1.1.1.43), polynucleotide phosphorylase (EC 2.3.7.8), beta-galactosidase (EC 3.2.1.23), lipopolysaccharide, and Braun's lipoprotein. The cellular origin of many of the vesicle immunogens is determined, and Braun's lipoprotein is used as a marker to quantitate the extent of outer membrane contamination (less than 3%). Membrane antigens are also characterized with regard to their amphiphilic or hydrophilic properties by charge-shift crossed immunoelectrophoresis. Furthermore, the following immunogens cross-react with components in membrane vesicles prepared from Salmonella typhimurium: one of the three NADH dehydrogenases, ATPase, polynucleotide phosphorylase, 6-phosphogluconate dehydrogenase, Braun's lipoprotein, and three unidentified antigens. In the accompanying paper [Owen, P., & Kaback, H. R. (1979) Biochemistry 18 (following paper in this issue)] quantitative immunoadsorption is utilized to establish the topology of the vesicles with respect to the distribution of antigens on the inner and outer faces of the membrane.