TWEAK induces angiogenesis and proliferation of endothelial cells.

TWEAK is a recently described member of the Tumor Necrosis Factor (TNF) ligand family whose transcripts are present in a wide variety of human tissues (Chicheportiche, Y., Bourdon, P. R., Xu, H., Hsu Y. M., Scott, H., Hession, C., Garcia, I., and Browning, J. L. (1997) J. Biol. Chem. 272, 32401-32410). TWEAK is a weak inducer of apoptosis in transformed cells when administered with interferon-gamma or cycloheximide (Chicheportiche, Y., Bourdon, P. R., Xu, H., Hsu Y. M., Scott, H., Hession, C., Garcia, I., and Browning, J. L. (1997) J. Biol. Chem. 272, 32401-32410; Masters, S. A., Sheridan, J. P., Pitti, R. M., Brush, A. G., and Ashkenazi, A. (1998) Curr. Biol. 8, 525-528) and also promotes IL-8 secretion in cultured cells. We report here that picomolar concentrations of recombinant soluble TWEAK induce proliferation in a variety of normal human endothelial cells and in aortic smooth muscle cells and reduce culture requirements for serum and growth factors. Blocking antibodies to Vascular Endothelial Growth Factor (VEGF) do not significantly inhibit TWEAK-induced proliferation, indicating that TWEAK does not function indirectly through up-regulation of VEGF. Pellets containing TWEAK induce a strong angiogenic response when implanted in rat corneas, suggesting a role for TWEAK in vasculature formation in vivo.

Serologic and phylogenetic characterization of HIV-1 subtypes in Uganda.

OBJECTIVES: To determine the HIV genetic subtypes present in HIV-1-infected asymptomatic blood donors in Uganda and to evaluate serologic detection of infection by commercial immunoassays; to evaluate samples for HIV-1 group O infections. METHODS: Sixty-four HIV-seropositive plasma samples were collected from the Nakasero Blood Bank, Kampala, Uganda. The plasma were evaluated using commercial HIV enzyme immunoassays (EIA) and a research immunoblot. HIV-1 group M and O infections were identified on the basis of discordant seroreactivity in EIA and reactivity to group M and O antigens on the immunoblot. Regions of gag p24 and env gp41 were amplified using reverse transcriptase polymerase chain reaction, and genetic subtypes were determined by phylogenetic analysis. RESULTS: Serologic testing confirmed that 63 out of 64 plasma units were positive for HIV-1 group M infection and showed no evidence of HIV-1 group O infections. Genetic subtyping determined that 25 samples were subtype A, three subtype C, 22 subtype D, and nine were heterogeneous for subtypes A and D. CONCLUSIONS: Despite the sequence variation observed in Uganda, commercial EIA based on HIV-1 subtype B proteins detected all the infections. In contrast, a peptide-based assay failed to detect three infections by subtype D viruses. This emphasizes the negative impact of HIV genetic variation on assays that rely on peptides to detect HIV infections. The number of infections with heterogeneous subtype (due to mixed infections or recombinant viruses) is high and reflects the growing complexity of the HIV epidemic in endemic regions where multiple subtypes are present in the population.

PIP: Extensive sequence heterogeneity between HIV-1 isolates has led to the classification of HIV-1 into group M (major) subtypes A-J, and group O (outlier). Some isolates have also been found to be the result of recombination between different group M subtypes. Findings are reported from a study conducted to determine the various HIV genetic subtypes in HIV-1-infected asymptomatic blood donors in Uganda and to evaluate the serologic detection of infection by commercial immunoassays. 64 HIV-seropositive plasma samples were collected from the Nakasero Blood Bank in Kampala and evaluated using commercial HIV enzyme immunoassays (EIA) and a research immunoblot. 63 of 64 plasma units were positive for HIV-1 group M infection and showed no evidence of group O infections. According to phylogenetic analysis, 25 samples were subtype A, 3 subtype C, 22 subtype D, and 9 heterogenous for subtypes A and D. Despite the sequence variation observed in this study population, commercial EIA based upon HIV-1 subtype B proteins detected all of the infections. A peptide-based assay failed to detect 3 infections by subtype D viruses.

Envelope sequence variability and serologic characterization of HIV type 1 group O isolates from equatorial guinea.

Four sera from Equatorial Guinea (EG) suspected to contain antibody against HIV-1 group O-related viruses were identified on the basis of unusual and differential serologic reactivity in selected commercial assays and Western blot. Degenerate primers, designed from HIV-1 group O published sequences, were used to PCR amplify envelope (env) gene sequences from the suspect EG sera. A complete envelope gene sequence from each serum was determined from the overlapping env gene fragments. Analysis (PHYLIP package of programs) of Env amino acid sequences (translated from nucleotide sequences) indicated that the amino acid sequences obtained from EG sera clustered more closely with HIV Env sequences of group O compared to group M. The amino acid sequences at the octameric tip of the V3 loop were either RIGPLAWY (one isolate), RIGPMAWY (two isolates), or GLGPLAVY (one isolate). The V3 tip tetrameric sequence GPLA is represented only once in the 1995 HIV (Los Alamos) database, but was present in two of our group O-related EG samples. The gp41 immunodominant regions (IDR) protein sequences were identical for sequences from three of the sera, RLLALETLIQNQQLLNLWGCKGR(K)L(I)VCYTSVK(T)W, whereas sequence from the fourth serum contained three changes as noted in parentheses. IDR sequences derived from EG sera were unique compared to those reported for other HIV-1 group O isolate ANT70, VAU, or MVP5180. Antibody in each EG serum directed against the IDR could be detected using synthetic peptides comprising sequences from the ANT70 or MVP5180 IDRs, but were most reactive against the sequences derived from the samples themselves. Little or no serologic reactivity was detected when EG sera were reacted against peptides comprising the IDR of HIV-1 group M (subtype B consensus) or HIV-2 (consensus).

PIP: The genetic variation and epidemiology of HIV-1 group O isolates are of considerable importance to the design of HIV-1 diagnostic and screening assays, especially since current serologic and genetic methods to detect HIV-1 have been developed mainly on the basis of sequences from isolates belonging to HIV-1 group M. The HIV envelope protein, especially the gp41 immunodominant region, plays a major antigenic role in the detection of HIV infection and for discriminating HIV-1 from HIV-2 antibody. This paper reports upon genetic variation and the serologic characterization of env sequences from 4 people living in Equatorial Guinea (EG) who were infected with HIV-1 group O. Selected commercial assays and Western blot were first used to identify the sera, then degenerate primers, designed from HIV-1 group O published sequences, were used to PCR amplify envelope (env) gene sequences. A complete envelope gene sequence from each serum was determined from the overlapping env gene fragments. The env amino acid sequence analysis found the EG sera sequences to be clustered more closely with the HIV env sequences of group O rather than to group M. The amino acid sequences at the octameric tip of the V3 loop were either RIGPLAWY, RIGPMAWY, or GLGPLAVY. Although the V3 tip tetrameric sequence GPLA is represented only once in the 1995 HIV database, it was present in 2 of the group O-related EG samples. The gp41 immunodominant regions (IDR) protein sequences were identical for sequences from 3 of the sera. IDR sequences derived from the EG sera were unique compared to those reported for other HIV-1 group O isolates ANT70, VAU, or MVP5180. Other findings are discussed in detail.

Sequence of gp41env immunodominant region of HIV type 1 group O from west central Africa.

PIP: HIV-1 group O is endemic in the west central region of Africa, where the frequency of infection is estimated to be 3-10% of all HIV-1-infected individuals. However, international travel and immigration have led to group O cases being identified in France, Germany, Belgium, Spain, and the US. With the exception of an infected French woman, all reported group O-infected individuals originate from or have a connection to west central Africa. Since most immunoassay reagents are based upon HIV-1 group M, many HIV immunoassays have lower sensitivity for the detection of group O infections. Serum samples were collected from patients at hospitals, tuberculosis (TB) clinics, and STD clinics in endemic regions of Cameroon and Equatorial Guinea in a study of the sequence divergence with group O isolate infections. Screening of the 1086 samples using a range of research and commercial immunoassays found 255 to be HIV-1 seropositive. On the basis of differential reactivity in the various immunoassays, 8 individuals were identified as potentially being infected with group O virus, of which 4 were drawn from TB patients. 7 of the group-O samples were then subjected to polymerase chain reaction (PCR) amplification to verify group O infection. The gp41(env) immunodominant region was successfully amplified and sequenced from 4 of the 7 samples, 2 of which were from the TB patients; 4 of 1086 samples were definitely infected with HIV-1 group O.^ieng

Molecular analyses of HIV-1 group O and HIV-2 variants from Africa.

Genetic variation among HIV isolates creates challenges for their detection by serologic and genetic techniques. To characterize the sequence variation and its correlation to serologic diversity of HIV-1 Group O and HIV-2 isolates, samples were identified by differential reactivity in selected commercial and research assays. Analysis of sera from Equatorial Guinea (EG) led to identification of 4 HIV-1 Group O variants. Viral RNA, extracted from these samples was used to PCR amplify overlapping sequences of the entire envelope gene using multiple primer pairs. Sequence analysis indicated that the V3 loop nucleotide and protein sequences aligned more closely with HIVANT70 compared to other Group O sequences. The amino acid sequences at the octameric tip of the V3 loop were RIGPLAWY, RIGPMAWY, or GLGPLAVY. The tetrameric tip GPLA is represented only once in the published 1994 HIV database (Los Alamos) but was present in 2 of 4 of EG samples. The immuno-dominant region (IDR) sequences derived from EG sera were unique in that none of the sequences were completely homologous to other HIV-1 group O variants. Further, the HIV-1 group O sequence variation could be correlated with differential serologic reactivity using IDR peptides. Compared to HIV-1, the sequence information on HIV-2 isolates is relatively limited, though the HIV-2 isolates also show genetic variation similar to HIV-1. To further establish a correlation between the genetic diversity and serologic detection of HIV-2, plasma samples from Western Africa were evaluated. Eight samples were selected based on weak serologic reactivity to env proteins. PCR amplification and sequence analysis of the gag, env V3 loop, and env IDR regions indicated that the samples could be classified as subtypes A (4 samples), B (3 samples) and D (1 sample). Across the subtypes, there was conservation in the IDR region of the sequence WGCAFRQVCHT. This region is absolutely conserved among the majority of currently known HIV-2 and related SIV viruses (1994 HIV database). One subtype B sample had a unique sequence immediately adjacent to the IDR, however, this did not change the serologic detection using a HIV-2 IDR specific monoclonal antibody.

PCR amplification from paraffin-embedded tissues: recommendations on fixatives for long-term storage and prospective studies.

The development of polymerase chain reaction (PCR) DNA amplification methods has afforded molecular studies of fixed paraffin-embedded tissue samples and other archival material. Some fixation methods damage DNA, and thus deleteriously affect subsequent PCR analysis. This study addressed the effect of short- and long-term storage (2 hr to 30 days) in a variety of fixatives (10% buffered-neutral formalin [BNF], 95% ethanol, acetone, and OmniFix) before paraffin embedding. We tested the ability of prepared tissue sections to yield DNA amplification products ranging from 268 to 1327 bp. Results indicated that tissues fixed for 8 days in BNF were able to amplify 536-bp but very little 989-bp DNA fragments; after 30 days of BNF fixation only a 268-bp fragment was amplifiable. Samples fixed in OmniFix and acetone yielded products of 989 and 1327 bp, respectively, after 8 days of fixation; both fixatives yielded 989-bp amplification products after 30 days of fixation. Tissues fixed in 95% ethanol for up to 30 days efficiently produced DNA amplification fragments of up to 1327 bp in length. The results provide important information for prospective studies that involve PCR analysis from archival material. Furthermore, fixation and long-term storage in ethanol should prove particularly useful in remote areas where refrigeration or immediate sample-processing is unavailable.

Development and transformation of the ductal plate in the developing human liver.

The livers of 15 embryos and fetuses measuring from 0.5 cm to 21 cm in crown-rump length were examined. The liver of the smallest embryo showed a sheet of hepatocytes without any ductal plates or intrahepatic bile ducts. Transformation of the hepatocytes of a hilar ductal plate was first observed in a 1.8-cm embryo. A 10.0-cm fetus had ductal plates virtually throughout the liver. Focal transformation of the flattened cells of the ductal plates into tubules composed of columnar or cuboidal ductal cells was observed. Mature interlobular ducts were observed predominantly in the hilum whereas scattered primitive interlobular ducts were scattered throughout the parenchyma. The transformation of hepatocytes into interlobular bile ducts thus seemed to occur in two stages: in the first, the hepatocytes of the ductal plate became flattened and developed increased cytokeratin; in the second, these flattened cells became focally cuboidal or columnar, lost their carcinoembryonic antigen, became strongly positive for epithelial membrane antigen, and formed tubules and primitive interlobular bile ducts.

A comparison of six monoclonal antibodies for detection of cytokeratins in normal and neoplastic canine tissues.

Twenty normal canine tissue specimens, both fetal and adult; 19 epithelial neoplasms; and 18 nonepithelial neoplasms were examined using 6 commercially available monoclonal antibodies differing in their recognition of various molecular weight cytokeratins in human tissues. Fresh tissue samples were fixed in 100% ethanol and paraffin embedded prior to sectioning. The intermediate filament proteins were identified by an avidin-biotin-immunoperoxidase method. Primary antisera used included AE1/AE3, CAM-5.2, 35BH11, 34BE12, PKK1, MAK-6 cytokeratins, and vimentin. Monoclonal antibodies detected cytokeratins in a wide variety of canine epithelial tissues and neoplasms. Normal mesenchymal tissues and neoplasms, and stromal elements of epithelial tissues, showed no reactivity with anti-cytokeratins, but reacted positively with vimentin. Although PKK1, CAM-5.2, and MAK-6 were the most consistently reactive anti-cytokeratins, the full panel of monoclonals was required to detect cytokeratins in all of the epithelia evaluated.

Immunohistochemical evaluation of intermediate filament expression in canine and feline neoplasms.

Specimens of neoplastic tissues from 19 dogs and 4 cats were examined immunohistochemically for intermediate filament expression, using commercially available antibodies. Staining was observed in a wide range of tumor tissues and in normal internal controls by use of antibodies to vimentin, desmin, glial fibrillary acidic protein, and low and high molecular weight cytokeratins. Intermediate filament expression was found to be consistent with light and/or electron microscopic findings, and hence believed to be an accurate indicator of tumor histogenesis in cats and dogs. Three fixatives were evaluated for their relative abilities to preserve antigenicity. Absolute alcohol was superior to B5 fixative and both were superior to formalin. Some tissues that clearly displayed intermediate filament antigens with alcohol and B5 fixative failed to stain when fixed in formalin.

Characterization of two monoclonal antibodies in an immunohistochemical study of keratin 8 and 18 expression.

Two monoclonal antibodies, UCD/AB 6.11 and UCD/PR 10.11, were evaluated for their patterns of immunohistochemical reactivity in a survey of paraffin-embedded human tissues by the avidin-biotin-immunoperoxidase technique. By two-dimensional immunoblotting, UCD/AB 6.11 reacts with keratin number 18 and UCD/PR 10.11 identifies both keratin numbers 8 and 18, the major keratins found in human simple epithelial cells. Both antibodies have excellent signal-to-noise ratios, and specifically react with simple epithelia, transitional epithelia, mesothelia and tumors derived from such tissues, making them superior immunological reagents. They do not react with neural, muscle, hematopoietic, connective or most epidermal tissues. However, within a given tissue or cell type, differences in the distributions of the antigens recognized by these two antibodies can be observed, raising the possibility of differential expression, modification, or masking of the keratin epitopes revealed by UCD/AB 6.11 and UCD/PR 10.11.

Detection of acquired provirus sequences in mammary tumors from low-expressor, low-risk mice.

Murine mammary tumor virus (MuMTV) provirus sequences in the DNA from early-occurring (average age 10 mo) and late occurring (age greater than 20 mo) tumors in BALB/cfC3H mice were analyzed by Eco RI restriction endonuclease mapping procedures. All early tumors were MuMTV antigen-positive mammary adenocarcinomas that contained the 0.92- and 4.0-kilo base (kb) exogenous C3H MuMTV-specific Pst I restriction endonuclease fragments. All but 1 of the late mammary adenocarcinomas had MuMTV antigens detected by peroxidase antiperoxidase staining, and all contained the 0.92- and 4.0-kb exogenous virus Pst I fragments. Three late nonmammary tumors lacked both MuMTV antigens and acquired provirus sequences. Greater numbers of MuMTV sequences were detected in both early and late-arising mammary tumors by Eco RI restriction endonuclease mapping than were detected in tissues from uninfected BALB/c mice. However, neither the number nor the location of MuMTV proviruses correlated with tumor latent period.

Development and characterization of mouse hyperplastic mammary outgrowth lines from BALB/cfC3H hyperplastic alveolar nodules.

Five hyperplastic outgrowth lines were developed by serial transplantation of hyperplastic alveolar nodules from BALB/cfC3H mice into BALB/c hosts. The lines were used to study the biology, morphology, and virology of premalignant tissue originating in mouse mammary tumor virus (MuMTV)-positive animals. The five lines differed with respect to tumor potential and growth characteristics, corroborating the previous evidence that MuMTV-positive hyperplastic alveolar nodules are biologically heterogeneous. Subgross and microscopic examination of outgrowths and tumors revealed that each line had unique morphological characteristics. The presence of atypical lobules within the hyperplastic outgrowth appeared to be correlated with tumor risk, and a morphological continuum of atypical lesions ending in overt cancer was suggested. Viral expression was detected by nucleic acid hybridization and immunoperoxidase staining for MuMTV structural antigens. While the MuMTV RNA in certain tissues appeared to vary qualitatively with tumor potential of the outgrowth line, no correlation between viral antigens detected by immunoperoxidase staining and tumor potential was observed.

Correlation between the detection of specific mouse mammary tumor proviral sequences and the presence of pulmonary metastases in mice bearing spontaneous mammary tumors.

Pulmonary metastases in C3H/He mice bearing spontaneous mammary tumors were detected and characterized by histological criteria and immunocytochemical staining for mouse mammary tumor virus antigens. The same lungs containing metastases were also positive when assayed for a specific subset of mouse mammary tumor virus proviral DNA sequences. These sequences, termed tumor-associated sequences, have previously been shown to be present in the DNA of spontaneous mammary tumors that arise before 1 yr of age in C3H/He mice but are absent in DNA's of apparently normal tissues of C3H/He mice. Reconstruction experiments demonstrated that the nucleic acid hybridization method will detect at least one mammary tumor cell/250 cells. While DNA from 13 lungs of apparently normal C3H/He mice did not contain sequences homologous to mouse mammary tumor virus tumor-associated-sequence RNA, DNA from lungs of 9 of 12 C3H/He mice bearing spontaneous mammary tumors did contain these sequences. Since the entire DNA content of the lung can be assayed as one sample, the hybridization method minimizes false negatives resulting from histological analysis of random biopsy sampling. The hybridization procedure described here thus represents a sensitive and quantitative element as an adjunct for the detection of micrometastatic lesions in mice bearing viral-mediated spontaneous mammary carcinomas.

Prevalence and distribution of murine mammary tumor virus antigen detectable by immunocytochemistry in spontaneous breast tumors of wild mice.

Mouse mammary tumor virus (MuMTV) antigens were detected by immunoperoxidase cytochemistry in spontaneous breast tumors of wild mice from two widely separated areas of southern California. Eleven of 25 (44%) tumors from Lake Casitas (LC) mice and 4 of 5 tumors from Bouquet Canyon mice were positive. Included among the tumors lacking detectable MuMTV antigen were well-differentiated type A and type B carcinomas as well as tumors with an atypical pattern. In the antigen-positive tumors the distribution of staining was patchy and extremely variable in extent (less than 1-70% stained cells). The intensity and extent of staining were generally greater in breast tumors from hybrids of LC wild mice and C5LBL/10Sn or AKR inbred mice. A good correlation was found in the same tumors between immunoperoxidase staining, detection of MuMTV gp52 antigen by radioimmunoassay, and detection of type B particles by electron microscopy. All of the breast tumors in LC mice were positive for type C virus particles.

Distribution of dengue-2 antigens by electron immunocytochemistry.

The distribution of dengue-2 antigens was studied in infected monkey kidney cells (LLC MK2) using an indirect, horseradish peroxidase-conjugated immunoglobulin technique. This procedure allowed both light and electron microscopic examination of serial-step sections of individual cells cut in a plane perpendicular to the monolayer. Both virion and nonvirion antigens were identified on the plasma membrane with this technique. A diffuse cytoplasmic reaction product was also present. The intensity and distribution of the cytoplasmic reaction product was related to disruption of the plasma membrane.

Structure of the mouse mammary tumor virus: characterization of bald particles.

The polypeptide, antigenic, and morphological structure of the mouse mammary tumor virus was studied following protease digestion of intact virions. Intact, untreated virions (rho = 1.17 g/ml) had characteristic envelope spikes, five major polypeptides, and were precipitated by antisera against gp52. Two of the major polypeptides, with molecular weights of 52,000 (gp52) and 36,000 (gp36), had carbohydrate moieties. Protease treatment resulted in spikeless, "bald" particles (rho = 1.14 g/ml), which had altered surface antigenicity and which contained neither gp52 nor gp36. These data indicated that gp52 and gp36 were on the viral envelope. Bald particles retained a 28,000 dalton polypeptide (p28) which was proposed as the major internal polypeptide.