CsCl Density Gradient Centrifugation Research Articles

Only trace amounts of fatty acids are found in pure mucus glycoproteins

The presence of noncovalently associated lipids and covalently bound fatty acids was investigated in preparations of mucus glycoproteins obtained by using density-gradient centrifugation in CsCl/guanidinium chloride. No phospholipids, glycolipids, cholesterol, or triglycerides could be detected. However, small amounts of extractable fatty acids were consistently found, the sum of which ranged from 0.3 to 0.9 μg/mg of glycoprotein. The amount of fatty acid released after subsequent treatment with KOH ranged from 0 to 27 ng/mg of glycoprotein. We conclude that density-gradient centrifugation in CsCl/ guanidinium chloride is very efficient in removing noncovalently associated lipids from mucus glycoproteins and that covalently bound fatty acids are probably not present in the macromolecules.

Isolation and characterization of chondroitin 6-sulfate proteoglycans present in the extracellular matrix of rabbit bone marrow

The authors' previous studies showed that the glycosaminoglycans present in rabbit bone marrow were composed of chondroitin 6-sulfate (79%) and hyaluronic acid (16%). Immunohistochemically the chondroitin 6-sulfate was demonstrated to be in bone marrow matrix constructing hematopoietic microenvironment. In this study the authors isolated and characterized these glycosaminoglycans in their macromolecular form (i.e., proteoglycans). Bone marrow of 3-month-old rabbits was defatted with organic solvents containing proteinase inhibitors at -20 degrees C, and proteoglycans were extracted from the defatted tissue with 4 mol/L guanidine HCl containing the proteinase inhibitors. After extensive dialysis of the extract against 7 mol/L urea, more than 90% of hexuronate-containing materials was recovered in the urea-soluble fraction. The proteoglycans were purified from the urea-soluble fraction by diethyl aminoethyl (DEAE)-Sephacel chromatography, CsCl density-gradient centrifugation, and Bio-Gel A-5m gel filtration, then were rechromatographed on DEAE-Sephacel and on Bio-Gel A-5m. The proteoglycans were separated into three molecular species with different mol wts that were assessed to be 46,000, 16,000, and 8,300 by sedimentation-equilibrium centrifugation. Amino acid analyses of these proteoglycans revealed that serine and glycine accounted for approximately 60% of the total amino acids common to the three proteoglycans. The glycosaminoglycan side chains of these proteoglycans were converted stoichiometrically into unsaturated 6-sulfated disaccharide by digestion with chondroitinase AC-II, indicating that they were fully sulfated chondroitin 6-sulfate. Their apparent mol wts were estimated by gel filtration on Bio-Gel A-0.5m to be 10,900, 14,400, and 7,700. Computation of these results, taken together with their biochemical composition, revealed that the largest proteoglycan, PG-II, consisted of four chains of the chondroitin 6-sulfate and the core peptide with mol wt of approximately 4,000. The smaller two proteoglycan subpopulations, PG-I and PG-III, consisted of a single glycosaminoglycan chain linked to a small peptide with mol wts of approximately 500 and 1,000, respectively.

Isolation and Characterization of Chondroitin 6-Sulfate Proteoglycans Present in the Extracellular Matrix of Rabbit Bone Marrow

The authors' previous studies showed that the glycosaminoglycans present in rabbit bone marrow were composed of chondroitin 6-sulfate (79%) and hyaluronic acid (16%). Immunohistochemically the chondroitin 6-sulfate was demonstrated to be in bone marrow matrix constructing hematopoietic microenvironment. In this study the authors isolated and characterized these glycosaminoglycans in their macromolecular form (i.e., proteoglycans). Bone marrow of 3-month-old rabbits was defatted with organic solvents containing proteinase inhibitors at -20 degrees C, and proteoglycans were extracted from the defatted tissue with 4 mol/L guanidine HCl containing the proteinase inhibitors. After extensive dialysis of the extract against 7 mol/L urea, more than 90% of hexuronate-containing materials was recovered in the urea-soluble fraction. The proteoglycans were purified from the urea-soluble fraction by diethyl aminoethyl (DEAE)-Sephacel chromatography, CsCl density-gradient centrifugation, and Bio-Gel A-5m gel filtration, then were rechromatographed on DEAE-Sephacel and on Bio-Gel A-5m. The proteoglycans were separated into three molecular species with different mol wts that were assessed to be 46,000, 16,000, and 8,300 by sedimentation-equilibrium centrifugation. Amino acid analyses of these proteoglycans revealed that serine and glycine accounted for approximately 60% of the total amino acids common to the three proteoglycans. The glycosaminoglycan side chains of these proteoglycans were converted stoichiometrically into unsaturated 6-sulfated disaccharide by digestion with chondroitinase AC-II, indicating that they were fully sulfated chondroitin 6-sulfate. Their apparent mol wts were estimated by gel filtration on Bio-Gel A-0.5m to be 10,900, 14,400, and 7,700. Computation of these results, taken together with their biochemical composition, revealed that the largest proteoglycan, PG-II, consisted of four chains of the chondroitin 6-sulfate and the core peptide with mol wt of approximately 4,000. The smaller two proteoglycan subpopulations, PG-I and PG-III, consisted of a single glycosaminoglycan chain linked to a small peptide with mol wts of approximately 500 and 1,000, respectively.

Large-scale isolation of covalently closed circular DNA using gel filtration chromatography

The isolation of covalently closed circular (ccc) DNA free of contamination by RNA and other forms of DNA is fundamental to molecular biology. A variety of methods have been explored but CsCl density-gradient centrifugation remains the method most widely used for preparative scale resolution. The process is expensive, time-consuming, requires the use of large amounts of the carcinogen ethidium bromide, and is subject to considerable variation in yield and purity. To avoid these problems, we have devised a procedure for the preparation of cell lysates which results in consistently good yields of biologically active ccc DNA minimally contaminated with chromosomal DNA fragments and RNA. Lysates are deproteinized, precipitated with CaCl 2 to remove rRNA, concentrated by ethanol precipitation, and applied to a Sephacryl S-1000 column which resolves chromosomal fragments, open circular plasmid DNA, and residual RNA from the ccc DNA. We have found that substituting the gel filtration column for CsCl density-gradient centrifugation results in substantially better purification as well as reducing processing time, cost, and degree of difficulty. The time required from harvest of cells to final recovery of DNA is about 16 h. We have used the method to isolate plasmids from 4.4 to 12 kb and, with slight modifications, recombinant M 13 replicative form DNAs.

Selectively Enhanced Procollagen Gene Expression in Sclerosing (Morphea-Like) Basal Cell Carcinoma as Reflected by Elevated Proα1(I) and Proα1(III) Procollagen Messenger RNA Steady-State Levels

Sclerosing or morphea-like variant of basal cell carcinoma (BCC) is characterized by an extensive connective tissue stroma, and histopathology has suggested that the extracellular matrix is largely composed of collagen. In addition, fibronectin deposition has been proposed to modulate tumor growth in BCC. In this study, we examined the expression of genes coding for type I, III, and IV procollagens, as well as for fibronectin, in tissue from 10 patients with sclerosing BCC. For comparison, tissues from 5 patients with nodular BCC and 4 controls were examined. Total RNA was isolated by CsCl density gradient centrifugation, and messenger RNA (mRNA) steady-state levels were determined by slot-blot hybridizations with human sequence specific complementary DNAs (cDNAs). The abundance of type I procollagen mRNA in sclerosing BCC tissue was increased to 233.6 +/- 36.7% of the controls (mean +/- SEM). The corresponding value for type III procollagen mRNA in sclerosing BCC was 281.8 +/- 54.8% of the controls. Consequently, the steady-state ratio of type I/III procollagen mRNAs in sclerosing BCCs (5.0 +/- 1.2; mean +/- SD) was within the control range. Thus, there is a coordinate increase in type I and type III procollagen mRNA levels in sclerosing BCC. In contrast, the values for type I and type III procollagen mRNAs in nodular BCC were not different from the controls. In addition, type IV procollagen and fibronectin mRNA levels were not different from the controls either in sclerosing or nodular BCCs, attesting to the selectivity of the increase in type I and III procollagen mRNA levels in sclerosing BCC. These observations may relate to the excessive deposition of the extracellular matrix stroma surrounding the tumor cells in sclerosing BCC.

A simple and efficient procedure for the isolation of high-quality phage λ DNA using a DEAE-cellulose column

A simple and rapid procedure for purifying large quantities of bacteriophage λ particles and DNA is described. The procedure involves DEAE-cellulose column chromatography of the phage particles and elution of the phage particles from the column with a low-ionic-strength buffer. The resulting phage were well separated from RNA, DNA, and proteins derived from Escherichia coli host cells. The λ DNA was prepared from the purified phage particles by the conventional method of phenol extraction and ethanol precipitation. This procedure did not use nucleases, proteases, detergents, or CsCl density gradient centrifugation. The λ DNA obtained by this method was equivalent in purity to the material prepared by CsCl density gradient centrifugation and amenable to restriction enzyme digestion, ligation, radiolabeling, and double-stranded DNA sequencing. A detailed protocol is described for obtaining 0.5 to 1.0 mg DNA from a 1-liter liquid lysate in less than 5 h. This procedure is simple, inexpensive, and timesaving, and is particularly suitable for large-scale isolation of λ DNA.

Structural polypeptides of feline parvovirus subspecies viruses.

Structural polypeptides of feline parvovirus (FPV) subspecies viruses, that is, feline panleukopenia virus (FPLV), mink enteritis virus (MEV) and canine parvovirus (CPV), were comparatively investigated. Viruses propagated in the Crandell feline kidney (CRFK) cell culture were purified by equilibrium density gradient centrifugation in CsCl and were examined by electron-microscopy and sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE). Virus infected CRFK cells were also analyzed by SDS-PAGE to determine virus-specific polypeptides. The purified virion of each subspecies contained three size classes of polypeptide which were almost identical for all three viruses: VP 1 (84, 000), VP 2 (64, 000) and VP 3 (63, 000). The characteristics of FPV polypeptides revealed in the present study were very similar to those of other members in genus Parvovirus and the results indicate that FPLV, MEV and CPV can not be distinguished by the SDS-PAGE analysis of purified virions.

Expression of the Human Serum Apolipoprotein AI and AII Genes inXenopus laevisOocytes. Lipid-Associated Secretion of Gene Products

The two major apolipoproteins of plasma high-density lipoproteins (HDL) are apolipoprotein AI (apo AI) and AII (apo AII). The apo AI and the correctly oriented apo CIII genes separated by 2.6 kb were obtained by fusion of two human lambda-genomic clones. The apo AII gene was isolated as a 3 kb clone. These apolipoprotein genes have been injected independently and together into Xenopus laevis oocytes and their expression studied. Both apolipoprotein genes were transcribed and translated into their preproforms and processed in Xenopus laevis oocytes to their proforms. They were secreted into the medium associated with newly synthesized phospholipids and neutral lipids as particles floating in the high-density lipoprotein range between 1.12 and 1.21 g/ml. Secreted apo AI is associated mainly with newly synthesized phosphatidylethanolamine and little triglyceride, apo AII with phosphatidylethanolamine, lysophosphatidylethanolamine and neutral lipids. Simultaneous injection of the apo AI and apo AII genes led to the secretion of both apoproteins which separated into two bands during CsCl-density gradient centrifugation. The heavier particles were associated with proapo AI and AII, phosphatidylethanolamine (greater than 90%) and traces of lysophosphatidylethanolamine as lipid components. Proapo AII was immunoprecipitated from the less dense fraction and found to be mainly associated with lysophosphatidylethanolamine. Radiolabelled newly synthesized apolipoproteins in secreted particles were characterized by immunoprecipitation after delipidation of the secreted lipoprotein particles. The oocyte-system proved very suitable for studies of the expression of serum apolipoprotein genes, the assembly of the apolipoproteins with specific lipids to lipoprotein particles and their secretion.

Isolation and characterization of a collagen fibril-associated dermatan sulphate proteoglycan from bovine lung

Dermatan sulphate proteoglycans have been extracted from bovine lung with 2.0 M CaCl 2 and isolated using CsCl density gradient centrifugation, DEAE ion-exchange chromatography, gel chromatography and preparative sodium dodecyl sulphate-polyacrylamide gel electrophoresis. Ultrastructurally these proteoglycans are specifically associated with collagen fibrils. Dermatan sulphate ( M r 15·10 3–35·10 3, with a strong prevalence for the higher M r) is linked via an O-glycosidic bond to a protein core, which is rich in Asx, Glx and Leu. Of the total uronic acid, 91% is iduronic acid. A part of the glucuronic acid residues is located near the protein core and a large cluster of disaccharides is devoid of glucuronic acid residues. An inhibition enzyme immunoassay has been developed to quantitate the proteoglycan. A model for the interaction between dermatan sulphate proteoglycans and collagen fibrils is proposed.

Characterization of Bdellovibrio bacteriovorus bacteriophage MAC-1.

The bacteriophage MAC-1, which specifically infects Bdellovibrio bacteriovorus, was plaque purified and raised to high titre. The phage was purified by NaCl/polyethylene glycol precipitation, followed by two cycles of isopycnic density gradient centrifugation in CsCl. The purified phage exhibited a density of 1.363 g cm-3 and a sedimentation coefficient of 94S. Nucleic acid isolated from purified phage was resistant to hydrolysis under alkaline conditions and to digestion with RNAase, but it was hydrolysed by DNAase, providing evidence that the phage genome is made up of DNA. The lack of hyperchromic effect upon denaturation, hydrolysis of phage DNA by S1 nuclease, characteristic fluorescent staining with acridine orange, and resistance to digestion with a variety of restriction endonucleases are consistent with the DNA being single-stranded. A buoyant density of 1.722 g cm-3 and a sedimentation coefficient of 17.9S were obtained for the phage DNA. The molecular mass of phage DNA was determined as 1.58 MDa by agarose gel electrophoresis with single-stranded DNA as standards. Electron microscopy of the DNA showed that the genome is circular in nature. In addition, using Southern blots, the two replicative forms, RF1 (supercoiled) and RF2 (circular) have been identified and isolated from infected cell extracts.

Proteoglycans from human umbilical vein endothelial cells.

Human umbilical vein endothelial cells were incubated with [35S]sulphate and investigated for their proteoglycan production. By gel chromatography, ion-exchange chromatography and CsCl density-gradient centrifugation we obtained preparative amounts of the endothelial proteoheparan sulphate HSI and of proteochondroitin sulphate from the conditioned medium of mass-cultured human umbilical vein endothelial cells. Approximately 90% of the 35S-labeled material in the endothelial cell conditioned medium was proteochondroitin sulphate. This molecule, with a molecular mass of 180-200 kDa, contains four side-chains of 35-40 kDa and a core protein of 35-40 kDa. Two proteoheparan sulphate forms (HSI and HSII) from the conditioned medium were distinguished by molecular mass and transport kinetics from the cell layer to the medium in pulse-chase experiments. One major form (HSI), with an approximate molecular mass of 160-200 kDa a core protein of 55-60 kDa and three to four polysaccharide side-chains of 35 kDa each, was found enriched in the cellular membrane pellet. Another proteoheparan sulphate (HSII), with polysaccharide moieties of 20 kDa, is enriched in the subendothelial matrix (substratum).

Physical state of the latent herpes simplex virus genome in a mouse model system: Evidence suggesting an episomal state

Herpes simplex virus (HSV) can establish latent infections in tissues of the nervous system. We have examined the HSV-1 genome in both acutely and latently infected mice by CsCI buoyant density gradient centrifugation. Viral sequences, in gradient fractions, were detected by a spot blot technique using nick-translated HSV-1 cloned DNA as probe, and mouse chromosomal DNA was located by measuring optical density at 260 nm. Most HSV-1-specific DNA from both acutely and latently infected mouse brains was found to band at the buoyant density of virion DNA. However, some HSV-1-specific hybridization banded at the density of the mouse chromosomal DNA. Further rounds of CsCl density gradient centrifugation of this chromosomal DNA from acutely infected brains released most of the HSV-1-specific hybridizing material, suggesting that the association was due to trapping. In the case of the recycled chromosomal DNA from a latent infection, all the HSV-1-specific hybridization remained chromosomal DNA associated. However, the amount of hybridization was not significantly greater in quantity than the cross-hybridization between HSV-1 and chromosomal DNA from uninfected mice. AW-Ramos, an EBV-transformed cell line containing one integrated copy of the viral genome, was centrifuged under similar conditions and showed little if any shearing; the EBV DNA banded at the density of the host cell chromosomal DNA. We conclude that the majority of the latent HSV-1 DNA exists in an extrachromosomal state in the mouse model.

Formation of reactive metabolites from benzene.

Rat liver mitoplasts were incubated first with [3H]dGTP, to form DNA labeled in G, and then with [14C]benzene. The DNA was isolated and upon isopycnic density gradient centrifugation in CsCl yielded a single fraction of DNA labeled with both [3H] and [14C]. These data are consistent with the covalent binding of one or more metabolites of benzene to DNA. The DNA was enzymatically hydrolyzed to deoxynucleosides and chromatographed to reveal at least seven deoxyguanosine adducts. Further studies with labeled deoxyadenine revealed one adduct on deoxyadenine. [3H]Deoxyguanosine was reacted with [14C]hydroquinone or benzoquinone. The product was characterized using uv, fluorescence, mass and NMR spectroscopy. A proposed structure is described.

Association of three chicken proteins with the 34 kD target of rous sarcoma virus tyrosine kinase

Transformation of chicken embryo fibroblasts by avian retroviruses induces the tyrosine phosphorylation of a 34–39 kD cellular protein (p34). In vitro, p34 isolated from intestine interacts with F-actin in a Ca 2+-dependent manner. We report here that, in the absence of Ca 2+ chelators, three proteins co-purified with p34 extracted from a cytosolic or membrane fraction of chicken embryo fibroblasts; these two fractions account respectively for 10–20% and 50% of the total cellular p34. Isolated from the cytosoluble fraction of fibroblasts by sucrose gradient centrifugation and hydrophobic chromatography, p34 and the other proteins behaved as a homogeneous species upon non-denaturing gel electrophoresis, gel filtration, and CsCl density gradient centrifugation, thus indicating a strong association. Moreover, an analysis by electron microscopy following uranyl acetate staining revealed particles with a raspberry-like shape. This association was always disrupted by the calcium-chelating agent, EGTA.

L-A double-stranded RNA viruslike particle replication cycle in Saccharomyces cerevisiae: particle maturation in vitro and effects of mak10 and pet18 mutations.

Previously, we found that log-phase cells of Saccharomyces cerevisiae contain a new type of viruslike particles containing only plus- strand L-A single-stranded RNA (ssRNA). These particles synthesize minus-strand RNA in an in vitro RNA polymerase reaction to produce L-A double-stranded RNA (dsRNA). The major class of particles contains L-A dsRNA and synthesizes plus-strand L-A ssRNA by a conservative mechanism. In this paper, we show that mutations in mak10 or the pet18 locus, which result in temperature-dependent replication of L-A dsRNA in vivo, also result in instability of the L-A dsRNA-containing (major class) viruslike particles in vitro. The L-A dsRNA (minus-strand)-synthesizing particles isolated by CsCl density gradient centrifugation synthesize plus-strand L-A ssRNA after completion of dsRNA (minus-strand) synthesis and have the same major coat protein as that of the major-class particles. Furthermore, the density of the dsRNA-synthesizing particles from wild-type cells shifts to that of the major-class dsRNA-containing particles as a result of the in vitro RNA polymerase reaction. Thus, L-A dsRNA-synthesizing particles undergo functional and structural maturation in vitro.

Sequence and transcription of tRNA Val gene from Xenopus laevis

A DNA fraction enriched in tRNA genes has been prepared by CsCl density gradient centrifugation of Xenopus laevis DNA in the presence of actinomycin D. This DNA fraction was cut with the restriction endonuclease EcoRI and the fragments 800–900 base pairs in size were cloned into the plasmid pBR325. Recombinant DNAs were screened by hybridization to labeled tRNA and for the ability to support transcription in vitro. The entire sequence of one fragment was determined by sequencing the ends of an overlapping set of deletion fragments. A sequence homologous to tRNA Val from mammalian sources was found in this fragment and it was shown that this sequence corresponds to the region of the fragment that is transcribed. The cloned fragment was also transcribed in vivo after injection into X. laevis oocytes. The RNA that was synthesized in the oocytes was digested with ribonuclease T1 and the oligonucleotides were separated to produce a two-dimensional fingerprint. The results of the analysis of the oligonucleotides are consistent with the sequence determined for the tRNA Val gene. The X. laevis genome has 200–250 copies of the 892 base pair EcoRI fragment and additional copies of a 4100 base pair EcoRI fragment that each contain a tRNA Val gene. Digestion of X. laevis DNA with several other restriction endonucleases reveals that the cloned fragment that contains the tRNA Val gene is part of a longer sequence element that is tandemly repeated in the genome.

L-A double-stranded RNA viruslike particle replication cycle in Saccharomyces cerevisiae: particle maturation in vitro and effects of mak10 and pet18 mutations.

Previously, we found that log-phase cells of Saccharomyces cerevisiae contain a new type of viruslike particles containing only plus- strand L-A single-stranded RNA (ssRNA). These particles synthesize minus-strand RNA in an in vitro RNA polymerase reaction to produce L-A double-stranded RNA (dsRNA). The major class of particles contains L-A dsRNA and synthesizes plus-strand L-A ssRNA by a conservative mechanism. In this paper, we show that mutations in mak10 or the pet18 locus, which result in temperature-dependent replication of L-A dsRNA in vivo, also result in instability of the L-A dsRNA-containing (major class) viruslike particles in vitro. The L-A dsRNA (minus-strand)-synthesizing particles isolated by CsCl density gradient centrifugation synthesize plus-strand L-A ssRNA after completion of dsRNA (minus-strand) synthesis and have the same major coat protein as that of the major-class particles. Furthermore, the density of the dsRNA-synthesizing particles from wild-type cells shifts to that of the major-class dsRNA-containing particles as a result of the in vitro RNA polymerase reaction. Thus, L-A dsRNA-synthesizing particles undergo functional and structural maturation in vitro.

Sulfated proteoglycans produced by rat dorsal root ganglion cells.

Primary organotypic cultures of embryonic rat dorsal root ganglia, which contain sensory neurons, Schwann cells, and fibroblasts, produce an extensive extracellular matrix. These cultures actively incorporated 35SO4 into glycosaminoglycans, of which 30% were heparan sulfate, 65% chondroitin sulfate, and 5% dermatan sulfate. Sulfate-labeled proteoglycans made by these cells were extracted with 4 M guanidine-hydrochloride and purified by CsCl density gradient centrifugation, gel filtration chromatography, and DEAE-cellulose chromatography. Eight individual species were resolved, of which five were subjected to further analysis. One low-density (less than 1.35 g/ml) proteoglycan, with a Kav on Sepharose Cl-4B = 0.34 contained heparan sulfate chains of Mr = 20,000. The other four proteoglycans, with Kav on Sepharose Cl-4B of 0.04, 0.35 (two), and 0.44, contained predominantly chondroitin sulfate chains with Mr ranging from 30,000 to 40,000. To determine possible functions of these proteoglycans, cultures were grown in medium containing 1 mM 4-methyl-umbelliferyl-beta-D-xyloside. The drug inhibited 35SO4 proteoglycan accumulation in the cell layer by approximately 90%. Drug-treated cultures exhibited several developmental abnormalities, including decreased migration of fibroblast-like cells, abnormal morphology of these cells, and decreased myelination of axons by Schwann cells.

Gene distribution and nucleotide sequence organization in the mouse genome.

Mouse DNA was fractionated by preparative centrifugation in density gradients of Cs2SO4 containing 3,6-bis(acetatomercurimethyl)dioxane (BAMD). The effects of temperature, BAMD/nucleotide molar ratio and solvent on the fractionation were explored. The fractions so obtained were investigated by analytical centrifugation in CsCl density gradient and by hybridization with a number of gene probes. These approaches led to the definition of satisfactory conditions for the rapid fractionation of mouse DNA; to the localization of a number of genes in mouse DNA fractions; and to a better understanding of the mosaic organization of the mouse genome and, more specifically, to a better estimate of both the intermolecular and intramolecular compositional heterogeneity of mouse DNA in the (75-150) X 10(3)-base size range.

Analysis of Escherichia coli ribosomal proteins by an improved two dimensional gel electrophoresis. II. Characterization of four new proteins.

Four new proteins, A, B, C, and D, found in Escherichia coli ribosomes by an improved two dimensional gel electrophoresis were characterized by oxidation, reduction, and carboxymethylation of cysteine residues, and CsCl fractionation. The cysteine contents of proteins A, B, C, and D were determined to be 1 +/- 0, 3 +/- 1, 5 +/- 1, and 0 +/- 0 by carboxymethylation with iodoacetic acid. The components of protein complexes, which formed numerously under non-reducing conditions, were analyzed. Including protein A, B, and C, every ribosomal protein (r-protein) having cysteine residue(s) except unconfirmed S1 was proved to form such complexes with various combinations. The cysteine residue in protein A, in particular, was highly reactive to make intermolecular S-S bridges so that spot A almost disappeared on the second dimension gel under the non-reducing conditions. Proteins B and C shifted their spots by reduction towards upper left side as do all known r-proteins having plural cysteine residues except S1. This suggests that proteins B and C change their conformation by intramolecular S-S bridges. The CsCl density gradient centrifugation of high salt washed 70S ribosomes showed that protein A belonged to the insoluble split proteins, proteins B and C to the core particles, and protein D and a small population of B to the soluble split proteins. The electrophoretic behaviors, CsCl fractionation and stoichiometry of the four new proteins suggested strongly that they were intrinsic ribosomal constituents different from known ribosomal proteins or factors.