CD70 (TNFSF7) is expressed at high prevalence in renal cell carcinomas and is rapidly internalised on antibody binding.

In order to identify potential markers of renal cancer, the plasma membrane protein content of renal cell carcinoma (RCC)-derived cell lines was annotated using a proteomics process. One unusual protein identified at high levels in A498 and 786-O cells was CD70 (TNFSF7), a type II transmembrane receptor normally expressed on a subset of B, T and NK cells, where it plays a costimulatory role in immune cell activation. Immunohistochemical analysis of CD70 expression in multiple carcinoma types demonstrated strong CD70 staining in RCC tissues. Metastatic tissues from eight of 11 patients with clear cell RCC were positive for CD70 expression. Immunocytochemical analysis demonstrated that binding of an anti-CD70 antibody to CD70 endogenously expressed on the surface of A498 and 786-O cell lines resulted in the rapid internalisation of the antibody-receptor complex. Coincubation of the internalising anti-CD70 antibody with a saporin-conjugated secondary antibody before addition to A498 cells resulted in 50% cell killing. These data indicate that CD70 represents a potential target antigen for toxin-conjugated therapeutic antibody treatment of RCC.

Proteomic analysis of the cell-surface membrane in chronic lymphocytic leukemia: identification of two novel proteins, BCNP1 and MIG2B.

B-cell-specific plasma-membrane proteins are potential targets for either small molecule or antibody-based therapies. We have sought to annotate proteins expressed at the cell surface membrane in patients with chronic lymphocytic leukemia (CLL) using plasma-membrane-based proteomic analysis to identify previously uncharacterized and potentially B-cell-specific proteins. Proteins from plasma-membrane fractions were separated on one-dimensional gels and trypsinized fractions subjected to high-throughput MALDI-TOF mass spectrometry. Using this method, many known B-cell surface antigens were detected, but also known proteins not previously described in this disease or in this cellular compartment, including cell surface receptors, membrane-associated enzymes and secreted proteins, and completely unknown proteins. To validate the method, we show that BLK, a B-cell-specific kinase, is located in the CLL-plasma-membrane fraction. We also describe two novel proteins (MIG2B and B-cell novel protein #1, BCNP1), which are expressed preferentially in B cells. MIG2B is in a highly conserved and defined gene family containing two plasma-membrane-binding ezrin/radixin/moesin domains and a pleckstrin homology domain; the Caenorhabditis elegans homolog (UNC-112) is a membrane-associated protein that colocalizes with integrin at cell-matrix adhesion complexes. BCNP1 is a completely unknown protein with three predicted transmembrane domains, with three alternatively spliced final exons. Proteomic analysis may thus define new potential therapeutic targets.

A human cDNA sequence with homology to non-mammalian lysophosphatidic acid acyltransferases.

A novel human homologue of Escherichia coli, yeast and plant 1-acylglycerol-3-phosphate acyltransferase has been isolated from U937 cell cDNA. Expression of the cloned sequence in 1-acylglycerol-3-phosphate acyltransferase-deficient E. coli resulted in increased incorporation of oleic acid into cellular phospholipids. Membranes made from COS7 cells transfected with the cDNA exhibited higher acyltransferase activity towards a range of donor fatty acyl-CoAs and lysophosphatidic acid. Northern-blot analysis of the cDNA sequence indicated high levels of expression in immune cells and epithelium. Rapid amplification of cDNA ends revealed differentially expressed splice variants, which suggests regulation of the enzyme by alternative splicing. This cDNA therefore represents the first described sequence of a mammalian gene homologous to non-mammalian lysophosphatidic acid acyltransferases.

Analysis of proviral integration in human mammary epithelial cell lines immortalized by retroviral infection with a temperature-sensitive SV40 T-antigen construct.

A panel of eight conditionally immortal lines derived by infection of human breast epithelial cells with an amphotropic retrovirus transducing a ts mutant of SV40 large T-antigen was analyzed with respect to individual retroviral integration patterns. Each line contained multiple integration sites which were clonal and stable over extended passage. Similar integration patterns were observed between individual lines arising separately from the same stock of pre-immortal cells, suggesting a common progenitor. Retroviral integration analysis of pre-immortal cells at different stages of pre-crisis growth showed changes indicative of a progressive transition from polyclonality to clonality as the cells approached crisis. Each of the immortal lines contained a sub-set of the integration sites of their pre-immortal progenitors, with individual combinations and copy numbers of sites. Since all the cell lines appeared to originate from single foci in separate flasks, it is likely that each set arose from a common clone of pre-immortal cells as the result of separate genetic events. There was no evidence from this analysis to suggest that specific integration sites played any part either in the selection of pre-crisis clones or in the subsequent establishment of immortal lines.

Hepatitis B virus-associated hepatocellular carcinoma in African patients.

We have examined several tumors from South African patients with hepatitis B virus (HBV)-associated hepatocellular carcinoma for the presence of integrated viral DNA. In contrast with our findings in patients from Taiwan, few copies of the viral genome were integrated in each tumor. Furthermore, Southern hybridization showed similarities in integration patterns between different tumors. We are presently constructing genomic libraries from selected single-copy tumors in order to make a more detailed analysis of HBV DNA integration at the molecular level.

Mapping of two novel transcripts of bovine papillomavirus type 4.

A cDNA library was synthesized using RNA from bovine papillomavirus type 4 (BPV-4)-induced papillomas. The major viral transcript was characterized by sequencing of its cDNA, primer extension mapping and S1 nuclease protection studies. The transcript initiates at multiple sites between nucleotide (nt) 777 and nt 902, contains a splice junction between nt 1016 and nt 3376 and terminates at nt 4034, using the early polyadenylation signal at nt 4009. Sequencing of viral cDNAs and mRNA revealed deviations from the reported genomic sequence between nt 3412 and nt 3460. These resulted in a temporary frameshift, abolished the translation termination codon of the E5a open reading frame (ORF), and introduced a new termination codon in the E4 ORF. The new uninterrupted E5 ORF was found to have greater homology to the E4 ORFs of other papillomaviruses than to their E5 ORFs. The E5 ORF of BPV-4 is joined to a five codon ORF in the leader exon of the major transcript, in the same way as the E4 ORF in the major transcripts of BPV-1 and human papillomavirus type 11. On this basis, it has been redesignated E4, and the viral genomic sequence revised. Two novel transcriptional promoters of BPV-4 were defined: a putative controller of the multiple major RNA start sites around nt 870 and a minor TATA box at nt 691. In addition, minor early region transcripts were mapped which initiate between nt 3071 and nt 3152. None of these transcripts utilizes the splice site at nt 3376. These messages may express E2-encoded functions.