55K Protein Research Articles

Enteric adenovirus type 40: E1B transcription map and identification of novel E1A–E1B cotranscripts in lytically infected cells

Adenovirus 40 (Ad40) is defective for growth in tissue culture but is complemented when the Ad2/5 or Ad12 E1B 55K protein is supplied in trans. Ad40 E1B mRNA has not been detected in E1-transformed cells, or at early times in lytically infected cells. In cells constitutively expressing the E1B region of Ad2, Ad40 E1B mRNAs are detected at late times in infection, after the onset of DNA replication. We have determined the Ad40 E1B transcription map from RNA produced at late times in infected KB16 cells, using S1 nuclease, primer extension, PCR-cDNA analysis, and Northern blotting. E1B transcripts corresponding to Ad2 14 S, 22 S, and 9 S mRNAs were identified but no 13 S mRNA equivalent was detected, a pattern similar to that seen in the Adl2 transcription map. The coding potential for E1B 19K, 55K, and 15K proteins and for ppIX is retained in the Ad40 transcripts. In addition we find novel E1A–E1B cotranscript counterparts of the 14 S and 22 S mRNAs. These contain the first 40 codons of the E1A first exon linked to a site 4–5 nt downstream of the E1B cap site, retaining all the coding potential of the E1B mRNAs. No new open reading frames are created by the junction, and the E1A ORF terminates with one codon added after the junction. Each E1A–E1B cotranscript is present in abundance comparable to that of its authentic E1B counterpart. The E1A–E1B junction is unusual in that it does not conform to splice consensus sequences and thus may not be generated by a conventional splicing mechanism.

Dissection of functional domains in the adenovirus 2 early 1b 55k polypeptide by suppressor-linker insertional mutagenesis

To determine whether the viral replication functions of the adenovirus E1 B 55K protein play a role in its ability to transform cloned rat embryo fibroblast cells in culture, we constructed an extensive series of insertion mutations throughout the 55K gene. The mutations were recombined into infectious virus and characterized for their abilities to produce stable 55K protein in HeLa cells, replicate virus in HeLa cells, express late viral proteins efficiently, and transform CREF cells following infection. Mutant 55K transforming activity in primary baby rat kidney cells was also assayed following DNA transfection. The functions required for viral replication are encoded in several patches of the 55K linear sequence, while the CREF transforming functions are sensitive to all of the insertions. An insertion at amino acid 380 created a mutant virus which was reduced in transforming activity, but was not reduced for viral replication. Therefore, a function required for efficient transformation of CREF cells can be separated from functions required for late gene expression and viral replication. Transformation of BRK cells following DNA transfection was reduced by complete disruption of the 55K protein gene, but was not significantly affected by any of the insertions.

Domains required for in vitro association between the cellular p53 and the adenovirus 2 E1B 55k proteins

The 55K protein encoded by the adenovirus 2 E 1 B gene is required for complete cellular transformation and binds the cellular protein p53. Using an in vitro immunoprecipitation assay, we mapped the domains in both 55K and p53 required for the interaction of the two proteins. The domain in p53 mapped to the amino terminal 123 residues. There are several domains in the 495 residue 55K polypeptide which contribute to stable association with p53, with the most essential region mapping between residues 224 and 354. Mutations which prevented 55K-p53 binding were not more defective for transformation than other mutations which did not affect binding.

Nucleotide Sequence and Transcriptional Analysis of the HindIII P Region of a Temperature-sensitive Mutant of Autographa Californica Nuclear Polyhedrosis Virus

DNA sequence analysis of the HindIII P region of a temperature-sensitive mutant of Autographa californica nuclear polyhedrosis virus confirmed the specific amplification of 1.4 kb of viral DNA from this region of the genome. The sequenced region included an open reading frame, translated in a counterclockwise direction, which would potentially encode a 74K protein. The amplified DNA was contained within this open reading frame, resulting in in-frame amplifications of a domain within the protein. Transcription studies revealed the presence of a ladder of viral RNA species corresponding to a 2.5 kb transcript carrying tandem repeats of about 1.4 kb. This indicated that the duplicated DNA was transcribed in the same orientation as the p10 gene. We predict that transcripts synthesized from the opposite DNA strand also consist of a ladder of related mRNAs which would be translated to produce a family of p74 proteins with multiple internal domains.

Enteric adenovirus type 40: Expression of E1B mRNA and proteins in permissive and nonpermissive cells

The enteric adenovirus type 40 (strain Dugan) grows well in tissue culture only when the E1B 55K protein of Ad5 or Ad1 2 is supplied in trans, either constitutively expressed in an established cell line or by coinfection with an appropriate helper virus (V. Mautner, N. Mackay, and V. Steinthorsdottir, 1989, Virology 171, 619–622). The synthesis of Ad40 E1B mRNAs and proteins has been examined under permissive and nonpermissive conditions: At late times postinfection in permissive cells, E1B-specific mRNA species of 22 and 13–14 S are made, as well as 15 and 9 S messages for the late IVa2 and ppIX proteins. None of these are detected before the onset of DNA replication and none of them accumulate in the presence of a cytosine arabinoside block to DNA replication. The failure to detect cytoplasmic mRNAs at early times cannot be attributed to a failure of mRNA transport from the nucleus as there is no accumulation of nuclear E1 RNA. In nonpermissive Hela cells only traces of E1B- and ppIX-specific mRNAs are detectable, at very late times postinfection. Antibodies raised to synthetic oligopeptides corresponding to the N- and C-terminal domains of the putative E1B 19K and 55K proteins show a high titer against the cognate peptide by ELISA, but only the E1B 19K C-terminus-specific sera have detected a unique polypeptide in Ad40-infected cells, at late times postinfection. There is no shut-off of host protein synthesis in permissive cells, despite the expression of Ad2 55K protein.

Expression of plant virus genes in animal cells: high-level synthesis of cowpea mosaic virus B-RNA-encoded proteins with baculovirus expression vectors

The baculovirus expression system has been used to produce non-structural proteins encoded by bottom-component RNA (B-RNA) of cowpea mosaic virus (CPMV). For this, cDNAs containing the 60K, 87K, 110K and 170K protein coding sequences were each provided with an ATG start codon and the cDNA containing the 60K coding sequence with a TAA stop codon immediately downstream of the coding sequence. Recombinant baculoviruses were retrieved which harboured the modified B-cDNA sequences under the control of the polyhedrin promoter of Autographa californica nuclear polyhedrosis virus (AcNPV). Upon infection of Spodoptera frugiperda cells with these recombinant baculoviruses, proteins were produced which were indistinguishable from the viral proteins found in CPMV-infected plants as judged by their migration in polyacrylamide gels and their reactivity with CPMV-specific antisera. Specific processing of CPMV polyproteins in cells infected with the 110K- and 170K-encoding baculovirus recombinants proved that the CPMV-encoded 24K protease activity contained in these polyproteins is active in these cells. Approximately 10% of the 110K protein was processed into 87K and 24K proteins and the 170K protein almost completely into the 110K, 87K, 84K, 60K and 24K polypeptides. In S. frugiperda cells infected by recombinant AcNPVs harbouring the 87K or 110K coding sequences, the CPMV-specific proteins amounted to 10 to 20% of the total cellular protein content, whereas in cells infected by recombinants encoding the 60K and 170K polypeptides the amounts of CPMV-specific proteins synthesized were much lower. Northern blot analysis indicated that the low-level synthesis of the 60K and 170K polypeptides was not due to inferior transcription of the cloned genes but was probably the result of inefficient translation of the RNAs derived from these constructs. It is concluded that plant virus genes can be efficiently expressed in an animal cell expression system to yield proteins that are structurally and, in at least one case (24K protein), functionally identical to the authentic plant virus proteins.

Nucleotide sequence and characterization of the 39K gene region of Autographa californica nuclear polyhedrosis virus

The complete nucleotide sequence of the Pstl-K fragment of the Autographa californica nuclear polyhedrosis virus (AcMNPV) genome was determined. This region of the genome contains the delayed early 39K gene and V-ubi, a viral protein with homology to ubiquitin. In addition, P stl-K potentially encodes five other proteins. Hybrid-select translation mapped nine viral-specific proteins to this fragment. Three proteins were observed with both early and late RNA; two of them comigrated with proteins directed by transcripts made from the 39K open reading frame (ORF). One protein which was only seen with hybrid-selected late RNA comigrated with ubiquitin. A correction of previously published sequence data indicates that there is an additional AUG codon upstream of the 39K ORF. This AUG codon is located 8 nucleotides downstream of the early transcription initiation site. The methionine codon is followed by codons for three amino acids and a termination codon. The 39K ORF is located 102 nucleotides downstream of the minicistron. In vitro transcription-translation experiments confirmed that the downstream AUG serves as the initiation codon for the 39K ORF. Radioimmunoprecipitation experiments indicated that the Pstl-K-encoded 39K protein reacted with antiserum raised against a previously described nuclear matrix-associated 39K protein.

The production of human parvovirus capsid proteins in Escherichia coli and their potential as diagnostic antigens

We have expressed a number of polypeptides derived from the capsid proteins of the human parvovirus B19 in Escherichia coli. These include native VP1 (84K) and VP2 (58K) proteins and also fusions to beta-galactosidase containing differing amounts of the amino terminus of the VP1/2 polypeptide. Although each of these was expressed at high levels and the majority were produced as full-length proteins, only one was soluble. This soluble polypeptide, p132, is a beta-galactosidase fusion protein that includes 145 amino acids from B19 which are entirely derived from the region unique to VP1. Despite containing such a small portion of VP1, which itself constitutes only 4% of total capsid protein, p132 reacted with all our known anti-B19 IgM-positive human serum samples. We conclude that this region contains epitopes which must be prominently exposed on the intact virus. We have demonstrated the use of this recombinant antigen in a simple diagnostic assay for B19-specific antibodies which can be used for initial screening of human serum samples. In a survey of 103 serum specimens, our ELISA positively identified all samples (19/19) which were positive by IgM antibody capture radioimmunoassay. The recombinant p132 antigen is efficiently produced and readily purified from E. coli, and its use as a diagnostic antigen should increase the availability of routine clinical testing for human parvovirus infection.

Characterization of a potato leafroll luteovirus subgenomic RNA: differential expression by internal translation initiation and UAG suppression

Northern blot analysis of Solanum tuberosum infected with potato leafroll luteovirus revealed the 6 kb genomic RNA and a major 2.3 kb subgenomic RNA. The 5' end of the subgenomic RNA was located at nucleotide 3653 in an intergenic region located at the centre of the viral genome upstream of three open reading frames (ORFs). Transient expression in tobacco and potato protoplasts of the beta-glucuronidase reporter gene fused to various putative regulatory sequences present in the subgenomic RNA was used to study their influence on expression levels. We observed a suppression of the amber stop codon separating the coat protein (CP) gene from a downstream ORF (56K protein), to a level of 0.9% to 1.3%. Translation initiation at the AUG of an ORF (17K protein) which is nested within the CP gene, exceeds translation of the CP gene itself by a factor of 7.

Translation of cucumber necrosis virus RNA in vitro

The in vitro translation products directed by cucumber necrosis virus (CNV) RNA were analysed in both rabbit reticulocyte lysate and wheatgerm extract cell-free translation systems. In rabbit reticulocyte lysates, one major protein of approximate Mr 34.6K was produced. In wheatgerm extracts, four proteins of approximate Mr values 41.6K, 34.6K, 24K and 20K were produced. The genomic locations of the CNV in vitro translation products were determined using several experimental approaches including, first, hybrid-arrested translation using negative-sense RNA corresponding to selected regions of the CNV genome, second, in vitro translation of synthetic positive-sense CNV transcripts and third, in vitro translation of CNV virion RNA fractionated according to size. Together these experiments demonstrated that the protein of Mr 34.6K is derived from the 5'-proximal coding region, the 41.6K protein is derived from an internal coding region, and that at least one but probably both the 24K and 20K proteins are derived from the 3'-terminal coding region. In addition, immunoprecipitation of in vitro translation products using anti-CNV polyclonal serum demonstrated that the 41.6K protein is the coat protein. The templates for the expression of CNV cistrons were investigated by in vitro translation of sucrose gradient-fractionated CNV virion RNA as well as in vitro translation of positive-sense synthetic transcripts.

Differential expression of heat shock proteins in Drosophila embryonic cells following metal ion exposure

Drosophila embryonic cells were exposed to a number of metal ions that have been previously reported to act as teratogens in mammalian systems, including some known to induce heat shock (stress) proteins in a variety of model systems. This study examined the effects of these ions both on differentiation of muscles and neurons and on the induction of heat shock proteins. Metals such as arsenate, cadmium, and mercury all inhibited neuron and/or muscle differentiation in Drosophila embryonic cultures, while they also induced the entire set of heat shock proteins. Two metal ions, nickel and zinc, were shown to induce only the 22-and 23-K proteins, a pattern similar to that seen in “classical” teratogens reported previously. None of the metals tested induced only the 26-and 27-K proteins. These results suggest that there exist different regulatory mechanisms responsible for the heat shock response.

Sex determination in loggerhead turtles: differential expression of two hnRNP proteins

Sex determination in the loggerhead turtle, Caretta caretta, is controlled by incubation temperature during a critical period of embryogenesis. As heat-shock gene expression is temperature-dependent and has been shown to be associated with early developmental regulation in several organisms, we studied the constitutive expression of hsp70 and hsp90 in embryonic brain and urinogenital tissues to see if these proteins are differentially expressed during the sex-determining period in embryos incubated at male- (26 degrees C) and female- (32 degrees C) determining temperatures. The level of expression of hsp70 and hsp90, as determined from monoclonal antibody staining, is similar in both sexes during the sex-determining period. However, AC88, a monoclonal antibody that identifies hsp90 in several systems, recognised two additional protein bands (Mr 42 and 46 x 10(3)), which are differentially expressed in the urinogenital tissue of developing male and female embryos during the sex-determining period. While the 42K and 46K proteins appear in the urinogenital tissue of developing female (32 degrees C) embryos until stage 25, they are not expressed in the male (26 degrees C) urinogenital system after stage 24. Subsequent experiments have identified both turtle proteins as heterogeneous nuclear ribonucleoprotein particles (hnRNPs). As several hnRNP proteins have specific RNA-binding sites and are involved in mRNA processing reactions, the 46K protein may mediate post-transcriptional control of specific RNA transcripts required for sexual differentiation in C. caretta.

Anti-Platelet Action of GU-7, A 3-Arylcoumarin Derivative, Purified from Glycyrrhizae Radix

We have purified GU-7, a 3-arylcoumarin derivative, from glycyrrhizae radix, which is a crude drug of kampo herbal medicines. This was identified to be a new chemical compound and was found to have an anti-platelet action. GU-7 inhibited platelet aggregation, phosphorylation of 40K and 20K dalton proteins, inositol 1,4,5-trisphosphate production, intraplatelet calcium increase and phosphodiesterase activity in vitro. The data indicate that GU-7 inhibits platelet aggregation by increasing intraplatelet cAMP concentration. We have already reported the existence of aldose reductase inhibitors in some kampo medicines. In addition to the aldose reductase inhibiting action, anti-platelet action may also explain the mechanism by which kampo medicines are effective to diabetic neuropathy.

Separate regions of an adenovirus E1B protein critical for different biological functions

The E1B region of Ad12 encodes two major proteins, the 482R (55K) and 163R (19K). In this report we showed that the E1B 482R is multifunctional, in that its structure may somehow contribute to its own stability, in viral DNA and virus replication, in transformation of primary cells, and in tumorigenicity. Deletion of the first 24 amino acids and of as residues 114–155 (d142) results in an instability of the 55K protein. The N-terminal 24 as residues (pml852) or amino acids residues 80–96 (d117) are not required for viral DNA or virus replication, whereas amino acid residues between 114 and 155 (d142) are absolutely necessary for viral DNA synthesis. Deletion of amino acid residues 1–24, 80–96, and 114–155 (d142) greatly reduces the transforming ability of both virus and plasmids containing any one of these deletions. One of the critical regions for tumorigenicity resides within amino acid residues 80–96, since cells transformed by this plasmid are nontumorigenic. On the other hand the region bounded by amino acid residues 114–155 (d142) is not required for tumorigenicity in immunocompetent animals.

Characterization and Immunological Properties of Influenza A Virus Nucleoprotein (NP): Cell-associated NP Isolated from Infected Cells or Viral NP Expressed by Vaccinia Recombinant Virus do not Confer Protection

A nucleoprotein (NP) preparation purified from the chorioallantoic membrane of chicken eggs infected with fowl plague virus (A/FPV/Rostock/34, H7N1) yielded, in addition to the commonly known 56K protein, a 42K component that could not be detected in virus particles. After testing with a series of NP-specific monoclonal antibodies it was found that some reacted with both proteins and others were bound only by the 56K protein. Among both types of NP-specific monoclonal antibodies only a limited number were bound to infected murine cells. Some antibodies bound to cells infected with a given subtype failed to react with the surface of cells infected with a different subtype. Binding was demonstrated by cellular ELISA, radioimmunoassay and immunofluorescence. The results indicate that only restricted antigenic domains of the native NP and perhaps NP fragments are exposed at the surface of infected murine cells. Additionally, the purified NP preparation was used to immunize mice in order to determine the protective capacity of cell-associated NP. In parallel, and as a relevant control, mice were immunized with a vaccinia virus recombinant containing the gene for NP prior to challenge with infectious virus. High levels of monospecific antibodies and a cytotoxic T cell activity was found in mice immunized with purified NP or infected with the vaccinia recombinant after secondary restimulation in vitro. After treatment with specific antibodies the cytotoxic cells were shown to be classical CD8+ cytotoxic T lymphocytes. Despite the elicitation of a humoral and a cellular immune response by the forms of NP employed mice were not protected from influenza virus infection.

Identification and characterization of a new replication region in the Neisseria gonorrhoeae beta-lactamase plasmid pFA3

The 7.1-kilobase-pair (kbp) plasmid pFA3 specifies TEM beta-lactamase production in Neisseria gonorrhoeae. We studied the minimal region required for replication of this plasmid in Escherichia coli by constructing a set of nested deletions of the 3.4-kbp PstI-HindIII fragment. The smallest fragment capable of maintenance in E. coli when ligated to a streptomycin-spectinomycin resistance cassette was 2.0 kbp in size and was different from another autonomously replicating fragment of pFA3 reported by K. H. Yeung and J. Dillon (Plasmid 20:232-240, 1988). The fragment contained single BamHI and XbaI sites and specified a 39-K protein. Fragments subcloned from the minimal region or constructed by deletion from the 3' or 5' ends were not capable of autonomous replication. Mutants constructed by end filling and religating DNA cleaved at the BamHI or XbaI sites were not capable of autonomous replication and no longer produced the 39K protein. These results suggest that replication is dependent on the 39K protein. DNA sequence analysis of the region showed an A-T-rich region followed by four 22-bp direct repeats followed by an open reading frame encoding a 39K basic protein.

Linkage and cytogenetic maps of genes controlling endosperm storage proteins and isozymes in rye (Secale cereale L.)

An F1 plant fromSecale cereale ssp.ancestrale xtelocentric substitution lines3R of the cultivated rye "Petkus spring" was used as female in a cross with the inbred line Riodeva (I28), which has the standard chromosome arrangement. Single plants from this backcross progeny were analyzed for chromosome constitution, storage protein, and isozymic patterns. The seed protein loci were identified asSec-1a andSec-1b loci controlling 40-Kγ-secalins andω-secalins, respectively. These loci are located on the short arm of chromosome1R. TheSec-3 locus controlling high-molecular-weight secalins is located on the long arm of chromosome1R. A further seed protein locus,Pr-3 (55-K protein), was located on the short arm of chromosome1R. A linkage was found between the6Pgd-2 isozyme locus controlling 6-phosphogluconate dehydrogenase isozymes located on the long arm of chromosome1R and the four seed protein loci. The results favor the gene order:6Pgd-2 ...Sec-3 ... [centromere] ...Pr-3 ...Sec-1b ...Sec-1a. Other linkages detected werePer-3a andPer-3b (0.33±0.33 cM),Est-8 andEst-12 (0.33±0.33 cM), andGot-3 and centromere (20.57±2.42 cM). The proxidase (Per), glutamate oxaloacetate transaminase (Got), and esterase (Est) loci were located on chromosome arms2RS,3RL, and6RL, respectively. The distances and the maps obtained are compared with data available in the literature.

Nicotiana Velutina Mosaic Virus: Evidence for a Bipartite Genome Comprising 3 kb and 8 kb RNAs

DNA complementary to Nicotiana velutina mosaic virus (NVMV) RNA was cloned and five segments larger than 0.9 kb were used in Northern blot hybridization analysis to identify two virus-specific RNAs, approximately 8 kb (RNA 1) and 3 kb (RNA 2) in size. The clones selected as probes did not hybridize with RNA from various tobamoviruses, or from beet necrotic yellow vein (BNYVV) and peanut clump furoviruses. In an attempt to determine the taxonomic position of the virus, about 75% of the NVMV RNA 2 was sequenced and four open reading frames (ORFs) were identified. ORFs 1, 2 and 3 encode proteins of Mr 20K, 39K and 13K, whereas ORF 4 was incomplete. ORFs 2, 3 and 4 overlapped in an arrangement closely resembling the triple gene block identified in BNYVV RNA 2, barley stripe mosaic virus (BSMV) RNA 2, potato virus X and potato virus M RNA. The presumed coat protein gene of NVMV RNA 2 (ORF 1) is situated to the 5' side of the triple gene block as for BNYVV and BSMV RNA 2. Amino acid homologies were detected among the 13K and 14K proteins of NVMV RNA 2, BNYVV RNA 2 and BSMV RNA 2. Significant homology was also detected between the 39K protein of NVMV RNA 2 and the 42K protein of BNYVV RNA 2, with a motif specific for ATP- and GTP-binding (NTP-binding motif), and a conserved viral DNA polymerase domain. The presence of a triple gene block in NVMV RNA 2 indicates that NVMV has affinities with members of the hordei-, furo-, potex- and carlavirus groups but not with the tobamovirus group. The divided RNA genome of NVMV, and the sizes of the two RNAs suggest that NVMV is most closely allied to the furoviruses, but the unique nature of its different biological properties and lack of any serological relationships with furoviruses lead us to conclude that NVMV has no clear relatedness to any taxonomic group of plant viruses.

Cloning of the genes for AF/R1 pili from rabbit enteroadherent Escherichia coli RDEC-1 and DNA sequence of the major structural subunit

AF/R1 pili on the surface of Escherichia coli RDEC-1 promote attachment of the bacteria to rabbit intestinal brush borders. In order to characterize AF/R1 pili and manipulate their expression, we cloned the genes necessary for AF/R1 expression; determined the size of proteins produced in minicells; located the gene encoding the major structural subunit, named AfrA; and determined the DNA sequence of afrA as well as the sequence of 700 additional nucleotides upstream of afrA. Two contiguous EcoRI fragments spanning 7.9 kilobases were cloned from the 86-megadalton plasmid of RDEC-1 into vector pUC19 to make plasmid pW1. Bacteria carrying pW1 produced AF/R1 pili that were recognized by AF/R1-specific antiserum and promoted adherence of bacteria to brush borders prepared from rabbit intestine. Proteins with a molecular weight of 17,000 (17K proteins), which was the size of AfrA, as well as 15K, 15.5K, 26K, 28K, and 80K proteins were detected in minicells carrying pW1. The gene afrA was located by using an oligonucleotide probe, and its DNA sequence was determined. The DNA sequence of 700 additional nucleotides upstream was determined because this sequence may be important in the regulation of AF/R1 expression.

Structural and chemical characterization of the S layer of a Pseudomonas-like bacterium

Sections and freeze-fractured preparations showed an S layer on the surface of Pseudomonas-like strain EU2. Polyacrylamide gel electrophoresis of cell envelopes extracted with 1% sodium dodecyl sulfate (SDS) at room temperature showed three proteins (45K, 55K, and 110K). The 55K protein was identified as the S-layer protein. Incubation in 1.5 M guanidine hydrochloride removed the S layer from cell envelopes and dissociated the structure into subunits. The soluble 55K protein reassembled into planar sheets upon removal of the guanidine hydrochloride by dialysis. Electron microscopy and image processing indicated that these sheets had p4 symmetry in projection with a lattice constant of 13.2 +/- 0.1 nm (corresponding to 9.3 nm between adjacent fourfold axes). In some instances these reassemblies appeared to form small three-dimensional crystals which gave particularly clear views of the structure in projection because of the superimposition of information from a number of layers. A model is proposed with molecules having rounded lobes connected by a narrower linker region and joining at the lobes to form the fourfold axes of the array. The pattern superficially resembles those of other bacterial S layers, such as those of Aeromonas salmonicida, Aeromonas hydrophila, and Azotobacter vinelandii. Extraction of cell envelopes with 1% SDS at 50 degrees C released the 110K protein from the envelopes and removed an amorphous backing layer from the S layer. The 45K protein displayed heat-modifiable migration in SDS-polyacrylamide gel electrophoresis and was insoluble in SDS at 50 degrees C or in high concentrations of guanidine hydrochloride, suggesting that it was associated with the peptidoglycan.