PDZ domain protein GIPC interacts with the cytoplasmic tail of melanosomal membrane protein gp75 (tyrosinase-related protein-1).

Tyrosinase and tyrosinase-related proteins (TRPs) are a family of melanosomal membrane proteins involved in mammalian pigmentation. Whereas the melanogenic functions of TRPs are localized in their amino-terminal domains that reside within the lumen of melanosomes, the sorting and targeting of these proteins to melanosomes is mediated by signals in their cytoplasmic domains. To identify proteins that interact with the cytoplasmic tail of gp75 (TRP-1), the most abundant melanosomal membrane protein, we performed yeast two-hybrid screening of a melanocyte cDNA library. Here, we show that the cytoplasmic domain of gp75 interacts with a PDZ domain-containing protein. The gp75-interacting protein is identical to GIPC, an RGS (regulator of G protein signaling)/GAIP-interacting protein, and to SEMCAP-1, a transmembrane semaphorin-binding protein. Carboxyl-terminal amino acid residues, Ser-Val-Val, of gp75 are necessary and sufficient for interaction of gp75 with the single PDZ domain in GIPC. Although endogenous and transfected GIPCs bind efficiently to transiently expressed gp75, only a small amount of GIPC is found associated with gp75 at steady state. Using a strategy to selectively synchronize the biosynthesis of endogenous gp75, we demonstrate that only newly synthesized gp75 associates with GIPC, primarily in the juxtanuclear Golgi region. Our data suggest that GIPC/SEMCAP-1 plays a role in biosynthetic sorting of proteins, specifically gp75, to melanosomes.     

Cryo-EM structure of a bacteriophage T4 gp24 bypass mutant: the evolution of pentameric vertex proteins in icosahedral viruses

Many large viral capsids require special pentameric proteins at their fivefold vertices. Nevertheless, deletion of the special vertex protein gene product 24 (gp24) in bacteriophage T4 can be compensated by mutations in the homologous major capsid protein gp23. The structure of such a mutant virus, determined by cryo-electron microscopy to 26 angstroms, shows that the gp24 pentamers are replaced by mutant major capsid protein (gp23) pentamers at the vertices, thus re-creating a viral capsid prior to the evolution of specialized major capsid proteins and vertex proteins. The mutant gp23* pentamer is structurally similar to the wild-type gp24* pentamer but the insertion domain is slightly more distant from the gp23* pentamer center. There are additional SOC molecules around the gp23* pentamers in the mutant virus that were not present around the gp24* pentamers in the wild-type virus.     

Simultaneous Purification of Murine Mammary Tumor Virus Structural Proteins: Analysis of Antigenic Reactivities of Native gp34 by Radioimmunocompetition Assays

All the structural proteins (gp47, gp34, p27, p23, p16, and p12) of the murine mammary tumor virus (MuMTV) were simultaneously purified utilizing alkylagarose chromatography as the initial fractionation step. Least-hydrophobic MuMTV polypeptides (p23, p16) and the slightly hydrophobic p27 were separated from moderately hydrophobic proteins gp47 and p12 by passage through octylimino (C(8))-agarose; the gp47 and p12 could be removed from the matrix by elution with ethylene glycol, whereas the most hydrophobic MuMTV protein, gp34, was eluted using nonionic detergent together with ethylene glycol. Subsequent purification steps involved ion-exchange or gel filtration chromatography. The resulting protein preparations appeared near-homogeneous on analysis by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. Recoveries of MuMTV proteins, based on their approximate individual contribution to total virus protein, ranged from about 20% for gp47 to greater than 100% for the minor structural component p23, the major phosphoprotein of MuMTV. Antiserum against purified C3H MuMTV gp34, together with purified, radioiodinated gp34, was used to develop a radioimmunoassay which showed that from 13 to 14% of total MuMTV protein by weight is gp34. Using this assay system, the group-specific antigenic reactivity of gp34 was also demonstrated. When solubilized preparations of C3H, RIII, and GR MuMTV's were used as competing antigens in gp34 radioimmunoassays with anti-C3H MuMTV serum, both group- and type-specific differences in antigenic reactivity were found.     

Shape and DNA packaging activity of bacteriophage SPP1 procapsid: protein components and interactions during assembly

The procapsid of the Bacillus subtilis bacteriophage SPP1 is formed by the major capsid protein gp13, the scaffolding protein gp11, the portal protein gp6, and the accessory protein gp7. The protein stoichiometry suggests a T=7 symmetry for the SPP1 procapsid. Overexpression of SPP1 procapsid proteins in Escherichia coli leads to formation of biologically active procapsids, procapsid-like, and aberrant structures. Co-production of gp11, gp13 and gp6 is essential for assembly of procapsids competent for DNA packaging in vitro. Presence of gp7 in the procapsid increases the yield of viable phages assembled during the reaction in vitro five- to tenfold. Formation of closed procapsid-like structures requires uniquely the presence of the major head protein and the scaffolding protein. The two proteins interact only when co-produced but not when mixed in vitro after separate synthesis. Gp11 controls the polymerization of gp13 into normal (T=7) and small sized (T=4?) procapsids. Predominant formation of T=7 procapsids requires presence of the portal protein. This implies that the portal protein has to be integrated at an initial stage of the capsid assembly process. Its presence, however, does not have a detectable effect on the rate of procapsid assembly during SPP1 infection. A stable interaction between gp6 and the two major procapsid proteins was only detected when the three proteins are co-produced. Efficient incorporation of a single portal protein in the procapsid appears to require a structural context created by gp11 and gp13 early during assembly, rather than strong interactions with any of those proteins. Gp7, which binds directly to gp6 both in vivo and in vitro, is not necessary for incorporation of the portal protein in the procapsid structure.     

Terminating a macromolecular helix. Structural model for the minor proteins of bacteriophage M13.

Analysis of the results of X-ray diffraction, electron microscopy and s sequence studies of filamentous bacteriophage M13 are used to construct structural models for the minor proteins gp7 and gp9 at the end of the virus assembled first, and a portion of gp6 at the end of the virus that binds host. Comparison of the sequence of the major coat protein, gp8, with those of gp7, gp9 and gp6 indicates that significant portions of these three proteins have sequences similar to that of gp8. Assuming that sequence similarity is indicative of structural similarity, gp7, gp9 and portions of gp6 are modeled based on what is known about the structure of gp8. These molecular models are analyzed to predict the packing of the minor proteins with the terminal gp8 proteins (the last gp8 proteins at either end of the helix). This analysis indicates that the gp8 proteins integrated into the virus first may have a structure distinct from those in the body of the virus particle. The gp8 proteins at the end assembled last appear to have a conformation very similar to that of the integral coat proteins. These models place specific constraints on models for the process of viral assembly.     

The N-terminal 31 amino acids of human immunodeficiency virus type 1 envelope protein gp120 contain a potential gp41 contact site.

We have compared the expression of full-length gp160 envelope protein from human immunodeficiency virus type 1 with that of a deletion mutant lacking the N-terminal 31 amino acids of the mature protein (gp160 delta 32). The gp160 and gp160 delta 32 proteins are processed to yield gp41 and gp120 or gp120 delta 32, respectively. In contrast to full-length gp120, gp120 delta 32 failed to associate with gp41 at the cell surface, despite conformational integrity as judged by soluble CD4 binding. Thus, the N-terminal 31 amino acids of gp120, which contain hyperconserved sequences, are likely involved in forming a contact site for gp41.     

Viral protein synthesis in mouse hepatitis virus strain A59-infected cells: effect of tunicamycin.

We identified eight protein species in virions of mouse hepatitis virus strain A59. Based on their sizes, prosthetic groups, and locations in virions, these proteins were designated gp180/E2, gp90/E2, pp54/N, gp26.5/E1, gp25.5/E1, p24/E1, p22/X, and p14.5/Y. The positions of the last two proteins in virions are not known. Host protein synthesis in Sac(-) cells infected with mouse hepatitis virus strain A59 was inhibited, and the following novel proteins appeared: gp150, gp90, p54, gp26.5, gp25.5, p24, p22, and p14.5. Except for gp150, these polypeptides all co-electrophoresed with mouse hepatitis virus strain A59 structural proteins. In addition, all of these proteins could be immunoprecipitated with a convalescent mouse serum or a rabbit antiserum raised against purified disrupted virus. After a 15-min pulse of infected cells with radioactive amino acids at 7h postinfection, gp90 was not detected, whereas gp26.5 and gp25.5 were only labeled to a small extent. During a subsequent chase period gp150 was processed to gp90, whereas the radioactivity in gp26.5 and gp25.5 increased concomitantly with a reduction of label in p24. Tunicamycin, an antibiotic which inhibits the synthesis of glycopeptides bearing N glycosidically linked oligosaccharides, prevented the appearance of gp150 in mouse hepatitis virus strain A59-infected cells. Instead, a 110,000-dalton protein accumulated. In contrast, the syntheses of the smaller viral glycoproteins gp26.5 and gp25.5 were resistant to this drug, indicating that these glycosylations were of the O glycosidical type. Although the production of infectious virus in tunicamycin-treated cells was inhibited by more than 99%, release of noninfectious viral particles continued. An analysis of these particles revealed that they lacked the peplomeric glycoproteins gp90/E2 and gp180/E2. Obviously, although the surface projections were not essential for budding of virus particles from the cells, they were required for infectivity.     

Specific targeting of a DNA-binding protein to the SPP1 procapsid by interaction with the portal oligomer

The icosahedral procapsid of tailed bacteriophages is composed of a large number of identical subunits and of minor proteins found in a few copies. Proteins present in a very low copy number are targeted to the viral procapsid by an unknown mechanism. Bacteriophage SPP1 procapsids and mature virions contain two copies of gp7 on average. Gp7 forms stable complexes with the SPP1 portal protein gp6. Deletion of the gp6 carboxyl-terminus and the mutation Y467-->C localized in the same region prevent gp6-gp7 complex formation. Gp7 binds double-stranded and single-stranded DNA. Gp6 competes for this interaction, and purified gp6-gp7 complexes do not bind DNA. Procapsid structures assembled in the absence of gp6 or carrying the mutant gp6 Y467-->C lack gp7. The gp6-gp7 interaction thus targets gp7 to the procapsid where the portal protein is localized asymmetrically at a single vertex of the icosahedral structure. The interaction between the two proteins is disrupted during viral assembly. Proteins homologous to gp6 and gp7 are coded by contiguous genes in a variety of phage genomes from Gram-positive bacteria, suggesting that the gp6-gp7 complex is widespread in this group of phages. Transient association with the portal protein, an essential component of tailed bacteriophages and herpes viruses, provides a novel strategy to target minor proteins to the virion structure that might be operative in a large number of viruses.     

Leucine zipper domain of HIV-1 gp41 interacted specifically with alpha-catenin

Interactions between viral and cellular proteins could explain the molecular mechanisms behind the viral life cycle of HIV-1. The envelope protein gp41 of HIV-1 specifically interacted with alpha-catenin, not with beta-catenin. This interaction was shown by in vitro protein assay and in vivo transfected cell systems. Microinjection of the DNA expressing HIV-1 gp160 and alpha-catenin, into the HeLa cell, resulted in the colocalization of gp41 and alpha-catenin. Interestingly the noncleavable mutant of gp160 and alpha-catenin were found to be colocalized in the cell membrane. Mapping of the interaction sites between these two proteins revealed that the leucine zipper-like structure, located between the first and second alpha-helix domains from the carboxy terminus of HIV-1 gp41, interacted strongly with the carboxy terminus of alpha-catenin.     

An Epstein-Barr virus DNA fragment encodes messages for the two major envelope glycoproteins (gp350/300 and gp220/200).

The genes encoding the two major Epstein-Barr virus glycoproteins (gp350/300 and gp220/200) have been mapped to a 5-kilobase fragment of the viral genome (BamHI-L). This fragment encodes 3.4- and 2.8-kilobase RNAs which translate proteins of 135 and 100 kilodaltons, respectively. Both proteins react with antiserum specific for gp350/300 and gp220/200. The 135-kilodalton protein is identical in size to the nascent polypeptide precursor to gp350/300, and the 100-kilodalton protein is the expected size of the polypeptide precursor to gp220/200.     

Binding-induced stabilization and assembly of the phage P22 tail accessory factor gp4

To infect and replicate, bacteriophage P22 injects its 43 kbp genome across the cell wall of Salmonella enterica serovar Typhimurium. The attachment of phage P22 to the host cell as well as the injection of the viral DNA into the host is mediated by the virion's tail complex. This 2.8 MDa molecular machine is formed by five proteins, which include the portal protein gp1, the adhesion tailspike protein gp9, and three tail accessory factors: gp4, gp10, gp26. We have isolated the tail accessory factor gp4 and characterized its structure and binding interactions with portal protein. Interestingly, gp4 exists in solution as a monomer, which displays an exceedingly low structural stability (Tm 34 degrees C). Unfolded gp4 is prone to aggregation within a narrow range of temperatures both in vitro and in Salmonella extracts. In the virion the thermal unfolding of gp4 is prevented by the interaction with the dodecameric portal protein, which stabilizes the structure of gp4 and suppresses unfolded gp4 from irreversibly aggregating in the Salmonella milieu. The structural stabilization of gp4 is accompanied by the concomitant oligomerization of the protein to form a ring of 12 subunits bound to the lower end of the portal ring. The interaction of gp4 with portal protein is complex and likely involves the distinct binding of two non-equivalent sets of six gp4 proteins. Binding of the first set of six gp4 equivalents to dodecameric portal protein yields a gp(1)12:gp(4)6 assembly intermediate, which is stably populated at 30 degrees C and can be resolved by native gel electrophoresis. The final product of the assembly reaction is a bi-dodecameric gp(1)12:gp(4)12 complex, which appears hollow by electron microscopy, suggesting that gp4 does not physically plug the DNA entry/exit channel, but acts as a structural adaptor for the other tail accessory factors: gp10 and gp26.     

Models for the binary complex of bacteriophage T4 gp59 helicase loading protein: gp32 single-stranded DNA-BINDING protein and ternary complex with pseudo-Y junction DNA

Bacteriophage T4 gp59 helicase assembly protein (gp59) is required for loading of gp41 replicative helicase onto DNA protected by gp32 single-stranded DNA-binding protein. The gp59 protein recognizes branched DNA structures found at replication and recombination sites. Binding of gp32 protein (full-length and deletion constructs) to gp59 protein measured by isothermal titration calorimetry demonstrates that the gp32 protein C-terminal A-domain is essential for protein-protein interaction in the absence of DNA. Sedimentation velocity experiments with gp59 protein and gp32ΔB protein (an N-terminal B-domain deletion) show that these proteins are monomers but form a 1:1 complex with a dissociation constant comparable with that determined by isothermal titration calorimetry. Small angle x-ray scattering (SAXS) studies indicate that the gp59 protein is a prolate monomer, consistent with the crystal structure and hydrodynamic properties determined from sedimentation velocity experiments. SAXS experiments also demonstrate that gp32ΔB protein is a prolate monomer with an elongated A-domain protruding from the core. Fitting structures of gp59 protein and the gp32 core into the SAXS-derived molecular envelope supports a model for the gp59 protein-gp32ΔB protein complex. Our earlier work demonstrated that gp59 protein attracts full-length gp32 protein to pseudo-Y junctions. A model of the gp59 protein-DNA complex, modified to accommodate new SAXS data for the binary complex together with mutational analysis of gp59 protein, is presented in the accompanying article (Dolezal, D., Jones, C. E., Lai, X., Brister, J. R., Mueser, T. C., Nossal, N. G., and Hinton, D. M. (2012) J. Biol. Chem. 287, 18596-18607).     

Effect of single-stranded DNA-binding proteins on the helicase and primase activities of the bacteriophage T7 gene 4 protein

Gene 4 protein (gp4) of bacteriophage T7 provides two essential functions at the T7 replication fork, primase and helicase activities. Previous studies have shown that the single-stranded DNA-binding protein of T7, encoded by gene 2.5, interacts with gp4 and modulates its multiple functions. To further characterize the interactions between gp4 and gene 2.5 protein (gp2.5), we have examined the effect of wild-type and altered gene 2.5 proteins as well as Escherichia coli single-stranded DNA-binding (SSB) protein on the ability of gp4 to synthesize primers, hydrolyze dTTP, and unwind duplex DNA. Wild-type gp2.5 and E. coli SSB protein stimulate primer synthesis and DNA-unwinding activities of gp4 at low concentrations but do not significantly affect single-stranded DNA-dependent hydrolysis of dTTP. Neither protein inhibits the binding of gp4 to single-stranded DNA. The variant gene 2.5 proteins, gp2.5-F232L and gp2.5-Delta26C, inhibit primase, dTTPase, and helicase activities proportional to their increased affinities for DNA. Interestingly, wild-type gp2.5 stimulates the unwinding activity of gp4 except at very high concentrations, whereas E. coli SSB protein is highly inhibitory at relative low concentrations.     

Intracellular sorting and targeting of melanosomal membrane proteins: identification of signals for sorting of the human brown locus protein, gp75

The structural and functional integrity of cytoplasmic organelles is maintained by intracellular mechanisms that sort and target newly synthesized proteins to their appropriate cellular locations. In melanocytic cells, melanin pigment is synthesized in specialized organelles, melanosomes. A family of melanocyte-specific proteins, known as tyrosinase-related proteins that regulate melanin pigment synthesis, is localized to the melanosomal membrane. The human brown locus protein, tyrosinase-related protein-1 or gp75, is the most abundant glycoprotein in melanocytic cells, and is a prototype for melanosomal membrane proteins. To investigate the signals that allow intracellular retention and sorting of glycoprotein (gp)75, we constructed protein chimeras containing the amino-terminal extracellular domain of the T lymphocyte surface protein CD8, and transmembrane and cytoplasmic domains of gp75. In fibroblast transfectants, chimeric CD8 molecules containing the 36-amino acid cytoplasmic domain of gp75 were retained in cytoplasmic organelles. Signals in the gp75 cytoplasmic tail alone, were sufficient for intracellular retention and targeting of the chimeric proteins to the endosomal/lysosomal compartment. Analysis of subcellular localization of carboxy-terminal deletion mutants of gp75 and the CD8/gp75 chimeras showed that deletion of up amino acids from the gp75 carboxyl terminus did not affect intracellular retention and sorting, whereas both gp75 and CD8/gp75 mutants lacking the carboxyl-terminal 27 amino acids were transported to the cell surface. This region contains the amino acid sequence, asn-gln-pro-leu-leu-thr, and this hexapeptide is conserved among other melanosomal proteins. Further evidence showed that this hexapeptide sequence is necessary for intracellular sorting of gp75 in melanocytic cells, and suggested that a signal for sorting melanosomal proteins along the endosomal/lysosomal pathway lies within this sequence. These data provide evidence for common signals for intracellular sorting of melanosomal and lysosomal proteins, and support the notion that lysosomes and melanosomes share a common endosomal pathway of biogenesis.     

A cell surface protein that binds avian hepatitis B virus particles.

We have identified a 180-kDa cellular glycoprotein (gp180) that binds with high affinity to duck hepatitis B virus (DHBV) particles. The protein was detected by coprecipitating labeled duck hepatocyte proteins with virions or recombinant DHBV envelope proteins, using nonneutralizing monoclonal antibodies to the virion envelope. Binding of gp180 requires only the pre-S region of the viral large envelope protein, since recombinant fusion proteins bearing only this region efficiently coprecipitate gp180. The DHBV-gp180 interaction is blocked by two independent neutralizing monoclonal antibodies. The protein is found on both internal and surface membranes of the cell, and the species distribution of gp180 binding activity mirrors the known host range of DHBV infection. Functional gp180 is expressed in a wide variety of tissues in susceptible ducks.     

Crystallization of a trimeric human T cell leukemia virus type 1 gp21 ectodomain fragment as a chimera with maltose-binding protein.

We present a novel protein crystallization strategy, applied to the crystallization of human T cell leukemia virus type 1 (HTLV-1) transmembrane protein gp21 lacking the fusion peptide and the transmembrane domain, as a chimera with the Escherichia coli maltose binding protein (MBP). Crystals could not be obtained with a MBP/gp21 fusion protein in which fusion partners were separated by a flexible linker, but were obtained after connecting the MBP C-terminal alpha-helix to the predicted N-terminal alpha-helical sequence of gp21 via three alanine residues. The gp21 sequences conferred a trimeric structure to the soluble fusion proteins as assessed by sedimentation equilibrium and X-ray diffraction, consistent with the trimeric structures of other retroviral transmembrane proteins. The envelope protein precursor, gp62, is likewise trimeric when expressed in mammalian cells. Our results suggest that MBP may have a general application for the crystallization of proteins containing N-terminal alpha-helical sequences.     

Simian retrovirus-D serotype 1 (SRV-1) envelope glycoproteins gp70 and gp20: expression in yeast cells and identification of specific antibodies in sera from monkeys that recovered from SRV-1 infection.

The gp70 and transmembrane gp20 envelope proteins of simian retrovirus-D serotype 1 (SRV-1) were expressed in Saccharomyces cerevisiae as fusion proteins with human superoxide dismutase (SOD). Expression of the SOD-gp70 and SOD-gp20 sequences yielded fusion proteins of 52 and 29 kilodaltons, respectively. The yeast-expressed SRV-1 envelope proteins were used in an enzyme-linked immunosorbent assay (ELISA) for the detection of antibodies in the sera of rhesus macaques that recovered from SRV-1. Sera from 47 of 49 such monkeys tested positive for antibodies to the SOD-gp70 fusion protein, while 45 of 49 reacted positively to SOD-gp20. None of 26 SRV-1-nonexposed monkeys tested positive in either ELISA. Monkeys immunized with the recombinant SRV-1 gp20 and gp70 proteins made good ELISA and Western blot (immunoblot) antibodies to whole SRV-1. This antibody was not neutralizing in vitro, however.     

AAA ATPase p97/valosin-containing protein interacts with gp78, a ubiquitin ligase for endoplasmic reticulum-associated degradation

Endoplasmic reticulum-associated degradation (ERAD) is a protein quality control mechanism that eliminates unwanted proteins from the endoplasmic reticulum (ER) through a ubiquitin-dependent proteasomal degradation pathway. gp78 is a previously described ER membrane-anchored ubiquitin ligase (E3) involved in ubiquitination of ER proteins. AAA ATPase (ATPase associated with various cellular activities) p97/valosin-containing protein (VCP) subsequently dislodges the ubiquitinated proteins from the ER and chaperones them to the cytosol, where they undergo proteasomal degradation. We now report that gp78 physically interacts with p97/VCP and enhances p97/VCP-polyubiquitin association. The enhanced association correlates with decreases in ER stress-induced accumulation of polyubiquitinated proteins. This effect is abolished when the p97/VCP-interacting domain of gp78 is removed. Further, using ERAD substrate CD3delta, gp78 consistently enhances p97/VCP-CD3delta binding and facilitates CD3delta degradation. Moreover, inhibition of endogenous gp78 expression by RNA interference markedly increases the levels of total polyubiquitinated proteins, including CD3delta, and abrogates VCP-CD3delta interactions. The gp78 mutant with deletion of its p97/VCP-interacting domain fails to increase CD3delta degradation and leads to accumulation of polyubiquitinated CD3delta, suggesting a failure in delivering ubiquitinated CD3delta for degradation. These data suggest that gp78-p97/VCP interaction may represent one way of coupling ubiquitination with retrotranslocation and degradation of ERAD substrates.     

E proteins of bacteriophage P22. I. Identification and ejection from wild-type and defective particles.

Of the nine proteins found in the virion of phage P22, four are ejected into the cell after adsorption. The four ejected proteins, termed E proteins, are gp16, gp20, gp26, and gp7. This was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis of radioactively labeled phage that had been adsorbed to cells and then eluted off the surface with distilled water. Phage particles that lack gp7 (7- particles) or gp20 (20- particles) successfully eject all their E proteins. The 16- particles do not eject gp7. Analysis of phage ghosts showed that they lack gp16, gp20, and gp7, but they have gp26 in close to normal quantities. Our results suggest roles for gp16 and gp26 in DNA and E protein ejection. All four E proteins are possible candidates for roles in helping the phage DNA cross the plasma membrane.     

Bacteriophage P22 tail accessory factor GP26 is a long triple-stranded coiled-coil

P22 is a well characterized tailed bacteriophage that infects Salmonella enterica serovar Typhimurium. It is characterized by a "short" tail, which is formed by five proteins: the dodecameric portal protein (gp1), three tail accessory factors (gp4, gp10, gp26), and six trimeric copies of the tail-spike protein (gp9). We have isolated the gene encoding tail accessory factor gp26, which is responsible for stabilization of viral DNA within the mature phage, and using a variety of biochemical and biophysical techniques we show that gp26 is very likely a triple stranded coiled-coil protein. Electron microscopic examination of purified gp26 indicates that the protein adopts a rod-like structure approximately 210 angstroms in length. This trimeric rod displays an exceedingly high intrinsic thermostability (T(m) approximately 85 degrees C), which suggests a potentially important structural role within the phage tail apparatus. We propose that gp26 forms the thin needle-like fiber emanating from the base of the P22 neck that has been observed by electron microscopy of negatively stained P22 virions. By analogy with viral trimeric coiled-coil class I membrane fusion proteins, gp26 may represent the membrane-penetrating device used by the phage to pierce the host outer membrane.