Serum antibodies to the HPV16 proteome as biomarkers for head and neck cancer.

BACKGROUND: Human papillomavirus (HPV) type 16 is associated with oropharyngeal carcinomas (OPC). Antibodies (Abs) to HPV16 E6 and E7 oncoproteins have been detected in patient sera; however, Abs to other early HPV-derived proteins have not been well explored. METHODS: Antibodies to the HPV16 proteome were quantified using a novel multiplexed bead assay, using C-terminal GST-fusion proteins captured onto Luminex beads. Sera were obtained from untreated patients with OPC (N=40), partners of patients with HPV16+ OPC (N=11), and healthy controls (N=50). RESULTS: Oropharyngeal carcinomas patients with known virus-like capsid particle+ Abs had elevated serum Abs to HPV16 E1, E2, E4, E6, and E7, and L1 antibody levels, but not E5. The ratios of specific median fluorescence intensity to p21-GST compared with controls were E1: 50.7 vs 2.1; E4: 14.6 vs 1.3; E6: 11.3 vs 2.4; E7: 43.1 vs 2.6; and L1: 10.3 vs 2.6 (each P≤0.01). In a validation cohort, HPV16 E1, E2, and E7 antibody levels were significantly elevated compared with healthy control samples (P≤0.02) and partners of OPC patients (P≤0.01). CONCLUSION: Patients with HPV16+ OPC have detectable Abs to E1, E2, and E7 proteins, which are potential biomarkers for HPV-associated OPC.

Screens using RNAi and cDNA expression as surrogates for genetics in mammalian tissue culture cells.

We have developed methods for the automation of transfection-grade DNA preparation, high-throughput retroviral preparation, and highly parallel phenotypic screens to establish approaches that will allow investigators to examine in an unbiased manner the roles of proteins in mammalian cells. These methods have been used to raise or lower the levels of individual kinases in individual micro-well cultures either by cDNA or short hairpin RNA expression and will allow investigators to treat mammalian cells in culture in manners that are analogous to genetic screens in yeast. Our proof-of-principle experiments have been performed in human cells using repositories that represent over 75% of the protein, nucleotide, carbohydrate, lipid, and amino acid kinases in the human genome. These initial experiments have demonstrated the feasibility of two general types of screens. We have performed phenotypic screens to identify proteins with specific roles in a chosen function and genetic interaction screens to establish epistatic relations between different proteins. The results suggest that any phenotype that can be scored by a robust assay in tissue culture is amenable to these types of screens and that interactions between mammalian proteins can be established. These results point to the near-term goal of establishing comprehensive, unbiased screens that will allow queries on the roles of all human proteins.

FLEXGene repository: from sequenced genomes to gene repositories for high-throughput functional biology and proteomics.

The vast amount of information generated by the human genome sequencing project and related projects has given rise to a new paradigm in experimental biology. This new paradigm invokes the experimentation and data analysis at genome-wide scales, as well as the generation of new technologies and resources that take full advantage of the available sequence information. The Institute of Proteomics at Harvard Medical School is building a comprehensive, characterized, arrayed and flexible gene repository that will allow full exploitation of the genomic information by enabling functional genomics as well as protein expression, purification and analysis at genome wide scale. The FLEXGene repository (Full Length EXpression-ready) will contain clones representing the complete set of open reading frames (ORFs) of different organisms including H. sapiens and several pathogens and model organisms. The clones are constructed using recombination-based cloning technology so that hundreds or thousands of coding regions can be transferred into any expression vector in a parallel and timely mode, allowing the broadest variety of experiments to be carried out.

Regulated ectopic expression of cyclin D1 induces transcriptional activation of the cdk inhibitor p21 gene without altering cell cycle progression.

Cyclin D1 plays a key regulatory role during the G1 phase of the cell cycle and its gene is amplified and overexpressed in many cancers. To address the relationship between cyclin D1 and other cell cycle regulatory proteins, we established human glioma and rodent fibroblast cell lines in which cyclin D1 expression could be regulated ectopically with tetracycline. In both of these cell lines, we found that ectopic expression of cyclin D1 in asynchronously growing cells was accompanied by increased levels of the p53 tumor suppressor protein and the cyclin/cdk inhibitor p21. Despite the induction of these cell cycle inhibitory proteins, cyclin D1-associated cdk kinase remained activated and the cells grew essentially like that of the parent cells. Although growth parameters were unchanged in these cells, morphological changes were clearly identifiable and anchorage independent growth was observed in NIH3T3 cells. In a first step toward elaborating the mechanism for cyclin D1-mediated induction of p21 gene expression we show that co-expression of E2F-1 and DP-1 can specifically transactivate the p21 promoter. In support of these findings and a direct effect of E2F on induction of p21 gene expression a putative E2F binding site was identified within the p21 promoter. In summary, our results demonstrate that ectopic expression of cyclin D1 can induce gene expression of the cdk inhibitor p21 through an E2F mechanism the consequences of which are not to growth arrest cells but possibly to stabilize cyclin D1/cdk function.

New functional activities for the p21 family of CDK inhibitors.

The association of cdk4 with D-type cyclins to form functional kinase complexes is comparatively inefficient. This has led to the suggestion that assembly might be a regulated step. In this report we demonstrate that the CDK inhibitors p21(CIP), p27(KIP), and p57(KIP2) all promote the association of cdk4 with the D-type cyclins. This effect is specific and does not occur with other cdk inhibitors or cdk-binding proteins. Both in vivo and in vitro, the abundance of assembled cdk4/cyclin D complex increases directly with increasing inhibitor levels. The promotion of assembly is not attributable to a simple cell cycle block and requires the function of both the cdk and cyclin-binding domains. Kinetic studies demonstrate that p21 and p27 lead to a 35- and 80-fold increase in K(a), respectively, mostly because of a decrease in K(off). At low concentrations, p21 promotes the assembly of active kinase complexes, whereas at higher concentrations, it inhibits activity. Moreover, immunodepletion experiments demonstrate that most of the active cdk4-associated kinase activity also associates with p21. To confirm these results in a natural setting, we examine the assembly of endogenous complexes in mammary epithelial cells after release from a G(0) arrest. In agreement with our other data, cyclin D1 and p21 bind concomitantly to cdk4 during the in vivo assembly of cdk4/cyclin D1 complexes. This complex assembly occurs in parallel to an increase in cyclin D1-associated kinase activity. Immunodepletion experiments demonstrate that most of the cellular cyclin D1-associated kinase activity is also p21 associated. Finally, we find that all three CIP/KIP inhibitors target cdk4 and cyclin D1 to the nucleus. We suggest that in addition to their roles as inhibitors, the p21 family of proteins, originally identified as inhibitors, may also have roles as adaptor proteins that assemble and program kinase complexes for specific functions.

Stimulation of the antigen receptor on WEHI-231 B lymphoma cells results in a voltage-independent increase in cytoplasmic calcium.

WEHI-231, a lymphoma-derived murine B cell line, responded to anti-IgM antibodies by increasing the concentration of free calcium in the cytoplasm from 140 nM to 590 nM within 15 sec. This is very similar to the response observed previously in normal B cells (Pozzan et al., 1982, J. Cell Biol. 94:335). Only antibodies specific for mIgM stimulated this response; control antibodies had no effect. In addition, anti-IgM did not stimulate a response by a mutant with a greatly decreased amount of membrane IgM. The relationship of this increase in cytoplasmic calcium to the plasma membrane potential was examined. Anti-IgM did not cause a rapid depolarization of the cells, suggesting that a voltage-dependent calcium channel was not responsible for the calcium increase. Furthermore, experimental depolarization of WEHI-231 cells did not cause a calcium influx, and the calcium increase caused by anti-IgM was not greatly affected by previous depolarization or by prevention of depolarization. These experiments argue strongly that the increase in cytoplasmic calcium was not mediated by a depolarization-activated calcium channel, such as the one found in cardiac muscle and in some neurons. Indeed, a significant portion of the initial increase in cytoplasmic calcium was due to the release of calcium from internal stores, suggesting the involvement of a soluble mediator. Examination of these internal storage sites in permeabilized cells revealed that inositol 1,4,5-trisphosphate could induce the release of calcium. These results are consistent with the hypothesis that the calcium increase in B cells stimulated by anti-IgM is caused by breakdown of phosphatidylinositol 4,5-bisphosphate, generating diacylglycerol and inositol trisphosphate, with the latter compound mediating calcium mobilization.