Photoaffinity labeling of the gonadotropin receptor with native, asialo, and deglycosylated choriogonadotropin

Multiple regions within EBNA1 can link DNAs

Epstein-Barr virus nuclear antigen 1 (EBNA1) can bind specifically to two clusters of sites within the Epstein-Barr virus plasmid origin of DNA replication (oriP). EBNA1 activates DNA replication mediated by oriP and can also activate transcription and retain DNA in cells when bound site specifically. EBNA1 bound to oriP physically links the two clusters of EBNA1-binding sites, resulting in loop formation by the intervening DNA. To elucidate the contribution of DNA linking by EBNA1 to its biological activities, we identified regions within it that can independently link DNAs to which they are bound. An electrophoretic mobility shift assay was used to detect this activity. Proteins which link DNA aggregate that DNA into large lattices. Proteins which cannot link DNA but still bind to DNA retard the mobility of that DNA but do not cause it to form lattices. Amino-terminal truncations were used to map the amino-terminal limit of a minimal DNA-linking domain approximately to amino acid 372 of EBNA1. To map the carboxy-terminal limit of this minimal domain, fusion proteins containing the DNA-binding domain of GAL4 and fragments of EBNA1 were generated and studied. This approach identified the carboxy-terminal limit of this minimal domain to be approximately amino acid 391 and verified its amino-terminal limit. Internal deletions within a truncated EBNA1 derivative verified the importance of this region. Two additional fragments of EBNA1, each of which independently conferred DNA-linking activity on the domain of GAL4 which binds DNA, were identified within amino acids 54 to 89 and amino acids 331 to 361. Therefore, EBNA1 contains at least three regions that can act independently to link DNAs and that may act in concert within intact EBNA1.

EBNA1 can link the enhancer element to the initiator element of the Epstein-Barr virus plasmid origin of DNA replication

The plasmid origin of DNA replication of Epstein-Barr virus, oriP, is replicated once per cell division, employing cellular replication machinery and only one viral protein. To understand how replication from this origin is initiated and regulated, we purified this viral protein, EBNA1. EBNA1 was expressed in CV-1p cells by using an infectious simian virus 40 vector containing the EBNA1 gene. It was purified in two chromatographic steps to apparent homogeneity. The purified protein is capable of supporting transcription of the luciferase gene from a reporter plasmid carrying the FR enhancer element to which EBNA1 binds. EBNA1 does not have oriP-dependent ATPase activity, indicating that it does not carry out an energy-dependent step in the initiation of DNA replication. However, EBNA1 does mediate an association between the two elements of oriP. We measured this association by binding one of the elements, the enhancer element, to a solid matrix and measuring retention by this element of the other one, the initiator element, in the presence of EBNA1. This retention is specific for DNA fragments containing EBNA1-binding sites. EBNA1 thus can link the two elements of the origin, providing a locally high concentration of EBNA1 at the site of initiation of DNA replication. We propose that this association is important either (i) to affect DNA structure to allow a cellular helicase to initiate DNA strand separation or (ii) to bind replication proteins to bring them to the origin of replication.

Transformation by the oncogenic latent membrane protein correlates with its rapid turnover, membrane localization, and cytoskeletal association

The latent membrane protein (LMP) of Epstein-Barr virus (EBV) has a short half-life (V. R. Baichwal and B. Sugden, J. Virol, 61:866-875, 1987; K.P. Mann and D. Thorley-Lawson, J. Virol, 61:2100-2108, 1987), is localized in patches in the membrane (D. Liebowitz, D. Wang, and E, Kieff, J. Virol, 58:233-237, 1986), and associates with the cytoskeleton in EBV-immortalized B lymphocytes (D. Liebowitz, R. Kopan, E. Fuchs, J. Sample, and E. Kieff, Mol. Cell. Biol. 7:2299-2308, 1987; K. P. Mann and D. Thorley-Lawson, J. Virol. 61:2100-2108, 1987). Deletion mutants of LMP that are either positive or negative in the induction both of anchorage-independent growth of BALB/c 3T3 cells (V. R. Baichwal and B. Sugden, Oncogene 4:67-74, 1989) and of cytotoxicity in a variety of cells (W. Hammerschmidt, B. Sugden, and V. R. Baichwal, J. Virol. 63:2469-2475, 1989) have been studied to identify the biochemical properties of this protein that correlate with its effects on cell growth. Mutant LMP proteins that are metabolically stable, do not associate with the cytoskeleton, and exhibit a diffuse plasma membrane localization also do not induce anchorage-independent growth in rodent cells or cytotoxicity in B lymphoblastoid cells. In contrast, a mutant of LMP that is functionally identical to the wild-type protein has a half-life, membrane localization, and cytoskeletal association similar or identical to those of LMP. These results are consistent with the hypothesis that LMP's rapid turnover, association with the cytoskeleton, and patching in the membrane are required for it to affect cell growth.

Differential interactions of human choriogonadotropin and its antagonistic aglycosylated analog with their receptor

Human choriogonadotropin (hCG) is a heterodimeric hormone consisting of an alpha subunit and a beta subunit. hCG and aglycosylated hCG (aghCG) have similar receptor binding affinities but differ in their ability to activate hormone-responsive adenylate cyclase. aghCG is an effective antagonist. The mechanisms of this antagonism and interactions of antagonistic aghCG with the receptor are not understood. To address this critical question, we have examined the interaction of this hormone analog with the receptor. The hormone receptor on porcine granulosa cells is a glycoprotein of 86 kDa and thas three domains of 24 kDa, 28 kDa, and 34 kDa, which are disulfide-linked. They undergo proteolysis, particularly when bound to the hormones, to produce three polypeptide components. These three receptor components can readily be identified through the use of affinity labeling with the hormones. Affinity labeling with an amino-specific homobifunctional reagent and subsequent cleavage indicate that hCG is cross-linked directly to the 24-kDa receptor component. In contrast, aghCG is cross-linked directly to the 34-kDa component. The peptide map of the cross-linked aghCG-34-kDa receptor component produced by papain treatment is different from the peptide map of the cross-linked complex of hCG-24-kDa component. This difference in receptor binding may be a factor determining the success or failure of signal transduction from the receptor to the effector system, guanine nucleotide-binding regulatory protein, and adenylate cyclase.

The average number of molecules of Epstein-Barr nuclear antigen 1 per cell does not correlate with the average number of Epstein-Barr virus (EBV) DNA molecules per cell among different clones of EBV-immortalized cells

Epstein-Barr nuclear antigen 1 (EBNA-1) is the only viral protein required to support latent replication of Epstein-Barr virus (EBV). To assess the likelihood that EBNA-1 regulates the amount of EBV DNA in a cell, we measured the average numbers of EBNA-1 molecules and EBV DNA molecules per cell in different clones of cells. The amount of EBNA-1 protein present in recently established lymphoblastoid cell lines was measured with affinity-purified anti-EBNA-1 antibodies, and viral DNA was measured by nucleic acid hybridization. The average levels of EBNA-1 protein varied little between these cell lines, whereas the average amount of viral DNA present varied substantially; consequently, these numbers were not correlated. There is no apparent relationship between amounts of EBNA-1 and viral DNA.

Evidence for monomeric and oligomeric hormone-binding domains in affinity-purified gonadotropin receptor from rat ovary

Rat ovarian lutropin/choriogonadotropin receptor was purified from a Triton X-100-solubilized membrane preparation by affinity chromatography with Affi-Gel 10 coupled to purified human choriogonadotropin. The affinity-purified receptor preparations contained a single class of high-affinity binding sites for 125I-labeled human choriogonadotropin, with an equilibrium dissociation constant (Kd) of 2.5 x 10(-9) M, which is comparable to the Kd values for membrane-bound and solubilized receptors. The purified receptor appeared as two dominant bands with molecular weights of 135,000 and 92,000 after sodium dodecyl sulfate/polyacrylamide gel electrophoresis (SDS/PAGE) under nonreducing conditions. These two bands were also detected in subsequent direct ligand blotting analysis when the purified receptor was electrophoretically transferred to a nitrocellulose membrane after SDS/PAGE under nonreducing conditions. When the individual affinity-purified receptor bands were electroeluted from the gel and analyzed again by SDS/PAGE under nonreducing conditions, both the Mr 92,000 and the 135,000 proteins retained their original molecular form even when 8 M urea was included in the gel. However, when the electrophoretically purified Mr 92,000 and 135,000 bands were subjected to SDS/PAGE under reducing conditions, the Mr 135,000 species was almost completely converted to a Mr 92,000 band, but the Mr 92,000 species did not undergo any alteration in molecular weight. The results suggest that the lutropin/choriogonadotropin receptor from rat ovary exists in two molecular forms, and the higher molecular weight form appears to be composed of disulfide-linked Mr 92,000 subunit, which comprises the hormone-binding domain.

Alterations in antigenic structure of gonadotropins following deglycosylation

Radioimmunological techniques were utilized to probe possible changes in conformation of gonadotropins (human chorionic gonadotropin--hCG; and ovine luteinizing hormone--oLH) following chemical deglycosylation (DG-hCG and DG-LH). All antisera produced in rabbits, rats or mice contained antibodies that were specific to the deglycosylated hormones with the native hormones showing weak and non-parallel cross-reaction (less than 5%), but with rabbit antibodies to native hormones the deglycosylated hormones were fully reactive. Using hCG, asialo-hCG (A-hCG) and DG-hCG, we have shown that removal of sugars internal to sialic acid is required to produce these specific antibodies. These are in complete agreement with the observations that extensive deglycosylation of these hormones is necessary to induce changes in biological activity at the cellular level. Based on these data, we suggest that chemical deglycosylation results in changes in antigenic structure of these hormones by generation of new determinants or exposure of previously buried sites and these changes are of no consequence to receptor recognition.