DNA chips: the future of biomarkers.

DNA chips are small, solid supports such as microscope slides onto which thousands of cDNAs or oligonucleotides are arrayed, representing known genes or simply EST clones, or covering the entire sequence of a gene with all its possible mutations. Fluorescently labeled DNA or RNA extracted from tissues is hybridized to the array. Laser scanning of the chip permits quantitative evaluation of each individual complementary sequence present in the sample. DNA chip technology is currently being proposed for qualitative and quantitative applications, firstly for the detection of point mutations, small deletions and insertions in genes involved in human diseases or affected during cancer progression; secondly, to determine on a genome-wide basis the pattern of gene expression in tumors, as well as in a number of experimental situations. The extraordinary power of DNA chips will have a strong impact on medicine in the near future, both in the molecular characterization of tumors and genetic diseases and in drug discovery and evaluation. Quantitative applications will soon spread through all fields of biology.

Anti-p53 antibodies in patients with Barrett's esophagus or esophageal carcinoma can predate cancer diagnosis.

BACKGROUND & AIMS: We previously discovered anti-p53 antibodies predating a cancer diagnosis in subjects at increased risk for liver, lung, breast, and prostate cancer. Recently, we reported a significant correlation (P < 0.017) between p53 antibodies and p53 mutations in patients with late-stage esophageal carcinoma. Because others have reported p53 mutations and overexpression of p53 protein in Barrett's esophagus, we studied p53 antibodies in plasma of 88 serially endoscoped patients: 36 with Barrett's metaplasia, 23 with esophageal squamous cell carcinoma, 10 with esophageal adenocarcinoma, and 19 with esophagitis or normal esophagus. METHODS: We used enzyme immunoassay, immunoblotting, and immunoprecipitation assays for p53 antibodies; polymerase chain reaction, denaturant gradient gel electrophoresis, and sequencing for p53 mutations; and immunohistochemistry for p53 protein. RESULTS: p53 antibodies were detected in 4 patients with Barrett's esophagus, including 1 with dysplasia that later progressed to adenocarcinoma, and in 10 cancer patients (P = 0.002) (8 squamous and 2 adenocarcinoma), 2 of whom (1 squamous, 1 adenocarcinoma) had antibodies before cancer was diagnosed. Other patient groups were too small for informative statistical analysis. Six antibody-positive cancer patients had p53 mutations, whereas 2 patients with cancer and 1 with Barrett's esophagus with antibodies had p53 protein overexpressed in esophageal tissues. CONCLUSIONS: Patients with Barrett's esophagus and esophageal cancer can develop p53 antibodies that may predate the clinical diagnosis of malignancy.

Characterization of ten novel and 13 recurring BRCA1 and BRCA2 germline mutations in Italian breast and/or ovarian carcinoma patients. Mutations in brief no. 178. Online.

Germline mutations in the BRCA1 and BRCA2 genes are associated with approximately 80% of families with a high incidence of breast and/or ovarian cancers (OMIM database reference numbers: 113705, 600185). Furthermore, constitutional mutations in the these genes have been reported in women with early-onset breast carcinoma and without family history of cancer. We analyzed by protein truncation test (PTT) and single strand conformation polymorphism (SSCP) followed by sequence analysis, BRCA1 exons 11 and 20 and BRCA2 exons 10 and 11 in 142 Italian cancer patients. These included six male breast cancer cases, 61 women with breast carcinoma diagnosed before 36 years old and selected independently of family history of breast cancer and 75 familial breast and/or ovarian cancer patients. In a previous report, we described 11 different BRCA1 mutations in a subset of 70 cases. Here, we report the characterization of 23 additional mutations, 14 in BRCA1 and 9 in BRCA2, subsequently identified. Ten mutations were not previously described, while the other 13 were recurrent. Of the 61 women with early-onset breast cancer, 11 carried a germline mutation in BRCA1 (18.0%) and four in BRCA2 (6.6%). These frequencies indicate that BRCA1/BRCA2 genetic tests should be advised to women with breast cancer diagnosed at early age, independently of family history of cancer.

Circulating anti-p53 antibodies in esophageal cancer patients are found predominantly in individuals with p53 core domain mutations in their tumors.

Serum antibodies reacting with the tumor suppressor protein p53 have been detected previously in cancer patients with a variety of neoplasms. Two initial (although insufficient) prerequisites for a B-cell response to occur have been proposed: p53 protein accumulation in the tumor or a mutant p53 gene, or both. We have examined 65 esophageal cancer cases (42 from Guangzhou and Shenyang, People's Republic of China, and 23 from Paris, France) to obtain a prevalence estimate of anti-p53 antibodies for this type of cancer and to define the relationship of p53 tumor status to B-cell immune response. Sera were analyzed in a triplicate assay (enzyme-linked immunoassay, immunoprecipitation, and immunoblot) for anti-p53 antibodies. Tumor DNA was screened for mutations in exons 5-8, and tumor tissue was examined by immunohistochemistry for abnormal p53 protein accumulation. p53 mutations were found in 36 (58%) of 62 cases analyzed. Sixteen patients (25%) had circulating antibodies to the tumor suppressor protein. All but two (88%) of the tumors from seropositive cases had a mutation in the DNA binding region of the p53 gene, and with one exception, these tumors also showed nuclear accumulation of the p53 protein. In contrast, tumor mutations were found in just 22 (46%) of the 48 individuals in whom we did not detect anti-p53 antibodies. Among the 22 seronegative cases for which we found no tumor mutations, 11 revealed p53 protein accumulation by immunohistochemical analysis. Thus, circulating anti-p53 antibodies may be present in one-fourth of esophageal cancer patients, most of whom also would be expected to have a p53 gene mutation in their tumors. Patients without such mutations appear considerably less likely to mount a B-cell response to the p53 tumor suppressor protein than those that do (P < 0.01).

Anti-p53 antibodies in sera from patients with chronic obstructive pulmonary disease can predate a diagnosis of cancer.

Serum anti-p53 antibodies (p53-Abs) may be surrogate markers for both p53 alterations and preclinical cancer. Ancillary to a prospective trial to abate progressive development of clinical stages of chronic obstructive pulmonary disease, we conducted a retrospective, nested case-control study. Twenty-three cases were diagnosed with cancer during the trial. Enzyme immunoassay, immunoblotting, and immunoprecipitation were used to detect p53-Abs in serum, immunohistochemistry (IHC) to detect p53 accumulation, and single-strand conformation polymorphism and DNA sequencing to detect p53 mutations in tumor samples. p53-Abs were detected by three types of assays in five (23%) of the cancer patients, 80% of whom had detectable p53-Abs before diagnosis: 2 lung cancers (7 and 6 months before), 1 prostate cancer (11 months), and 1 breast cancer (5 months). Four Ab-positive patients had IHC-positive tumors. Two of 4 Ab-positive patients and 2 of 14 Ab-negative had p53 missense mutations or base pair deletion and IHC-positive tumors. The 44 noncancer COPD controls, matched with the cancer cases for age, gender, and smoking habits, were negative for p53-Abs. These results indicate that p53-Abs may facilitate the early diagnosis of cancer in a subset of smokers with chronic obstructive pulmonary disease who are at an increased cancer risk.

Screening for mutations in exon 11 of the BRCA1 gene in 70 Italian breast and ovarian cancer patients by protein truncation test.

The most common mutations in the familial breast and ovarian cancer susceptibility gene BRCA1 are frameshift and nonsense mutations, which lead to the synthesis of truncated proteins. On this ground, we have analysed BRCA1 exon 11, which includes about 61% of coding region, in germline DNA from 70 Italian breast and/or ovarian cancer patients, using the protein truncation test (PTT). BRCA1 mutations were identified in nine of 29 (approximately 31%) patients with a family history of cancer and in three of 41 (approximately 7%) women with early-onset breast carcinomas, and were subsequently characterized by sequence analysis. In addition, BRCA1 mutations were also detected in six affected relatives of two positive index cases. The observed frequencies of mutations were not significantly different from those expected on the basis of the phenotypic characteristics of patients and their families, indicating that PTT is a rapid and sensitive method that can be used for a first BRCA1 mutational screening. The histological findings in BRCA1 mutated cases showed that eight of nine (approximately 89%) breast carcinomas were of grade III and nine of 9 (100%) ovarian carcinomas were of the endometrioid type (eight of grade III and one of grade II). This suggests that specific histological characteristics may represent additional criteria for selection of cases eligible to BRCA1 mutational analysis.

p53 mutations in lung cancer following radiation therapy for Hodgkin's disease.

High risks of lung cancer occur after successful treatment of Hodgkin's disease. In addition to tobacco smoking, other risk factors include radiotherapy, chemotherapy, and immunosuppression, although the relative contributions of each are unknown. We conducted p53 mutational spectrum analysis in second lung cancers after radiation therapy for Hodgkin's disease in the Netherlands and in Ontario, Canada. Lung cancer tissues from 11 patients were analyzed by p53 immunohistochemistry and DNA sequence analysis. All were male cigarette smokers, all received radiation therapy, and six also received chemotherapy. The lung cancers occurred 9.8 years (mean) after treatment. Radiation doses to lung lobes that developed the tumors averaged 5.7 Gy (range, 3.7-11.7 Gy). Sequence analysis showed four missense and two silent p53 point mutations in five patients. There were four G:C-->A:T transitions; three of four mutated deoxyguanines occurred on the coding strand, and one was a CpG site. There were two transversions: one G:C-->C:G and one A:T-->C:G. Despite moderate or heavy smoking histories in all patients, the mutational spectrum appears to differ from usual smoking-related lung cancers in which G:C-->T:A transversions predominate. The absence of G:C-->T:A mutations and the prominence of G:C-->A:T transitions, which are characteristic of radiation and oxidative damage, suggest that radiotherapy might have caused some of the p53 mutations. These data illustrate the potential of mutation analysis to determine causes of human cancer. If confirmed in a larger series, these results imply that some radiation-induced cancers can be distinguished from those caused by other factors.

p53 mutations in hepatocellular carcinoma related to oral contraceptive use.

Oral contraceptives (OCs) are implicated in the development of hepatocellular carcinoma (HCC). Mitogenic stimulation may be the primary mechanism of tumorigenesis, but other factors may also contribute. Mutational spectrum analysis can provide insights into pathogenesis, therefore we analyzed the p53 tumor suppressor gene in 10 HCCs from women with a history of OC use. All were non-Asians whose average OC use was 6.7 years (range 2 months-13 years) and whose mean age at HCC diagnosis was 48.8 years (range 21-67 years). Each tumor was analyzed by immunohistochemistry, DNA sequencing and allelic deletion analysis. Three tumors were positive by p53 immunohistochemistry; allelic deletion analysis identified loss of heterozygosity in one of four informative cases. Two p53 point mutations were found in one tumor containing moderately and well-differentiated components; this patient was negative for all serological markers of hepatitis B and C infections. Both components showed p53 protein accumulation and a GTTval-->GCTala mutation at codon 274. In addition, a silent mutation (ACCthr-->ACTthr) at codon 140 of the p53 gene was detected in the moderately differentiated component of the tumor. These preliminary data indicate that p53 mutations are uncommon in OC-related HCCs. One of the two detected mutations was a G:C-->A:T transition at a non-CpG site, which is characteristic of DNA damage by free radicals. These data support a model whereby estrogens contribute to HCC development primarily through mitogen stimulation and secondarily by mutagenesis via hydroxyl radicals produced during estrogen metabolism. Confirmational analysis of a larger series is warranted.

Lung carcinoma and malignant mesothelioma in patients exposed to Thorotrast: incidence, histology and p53 status.

In a previous registry-based survey of 999 patients injected with alpha-emitting 232ThO2 (Thorotrast), we identified elevated risks for lung carcinoma and malignant mesothelioma. Since injected Thorotrast is retained lifelong mostly in liver, spleen and lymph nodes, the mesothelial surfaces of these organs are constantly irradiated. Thorotrast-administered patients also perpetually exhale 220Rn, a 232Th-daughter. Study of Thorotrast-exposed patients may, therefore, provide data with regard to carcinogenicity of radon exposure, a current public health concern, as well as the pathogenesis of malignant mesothelioma. The incidence and histologic types of lung carcinoma and malignant mesothelioma within the cohort were examined by review of available histopathologic material and medical records. Further, mutations of the p53 gene were analyzed whenever possible as it has previously been suggested that radon-associated lung carcinomas exhibit specific mutational patterns. The cumulative risk for lung carcinoma reached 11.0% based on 20 confirmed cases. Nine were small cell lung cancer (SCLC), whereas the expected frequency was 18%. The risk for malignant mesothelioma reached 2.5% based on 7 cases. The actuarial risk of malignant mesothelioma for patients given more than 20 ml Thorotrast was 7.8% compared to 1.4% for patients administered smaller amounts. Seven lung carcinomas and 5 malignant mesotheliomas were analyzed for p53 mutations; only 1 (in a lung adenocarcinoma) was detected. A possible association between Thorotrast and SCLC is suggested. In addition, a possible dose-response gradient exists for Thorotrast and malignant mesothelioma.

Gender comparisons in human lung cancer: analysis of p53 mutations, anti-p53 serum antibodies and C-erbB-2 expression.

Little is known about the molecular mechanisms of lung carcinogenesis in women. We initiated an investigation of the role of gender in pulmonary carcinogenesis by analysis of p53 mutations, immunohistochemistry, serum antibodies and c-erbB-2 expression in a series of 63 male and 44 female lung cancer patients whose tumors were resected at the Mayo Clinic between 1991 and 1992. There were 102 smokers and 5 never smoked. Adenocarcinoma was the more frequent histological type in women (62%) than in men (41%). Sequence analysis of exons 5-8 in 42 females and 49 males identified 44 p53 mutations in 42 tumors (46%). Base substitution mutations showed a preponderance of G:C-->T:A transversions, which were more frequent in women than men (40 versus 25%) and in individuals exposed to asbestos. c-erbB-2 immunohistochemical staining was identified more frequently in females (nine cases) than males (two cases). Marked immunohistochemical staining for p53 positively correlated with the presence of missense mutations in exons 5-8 (81%, P < 0.001). Seven missense mutations (four in exon 5, two in exon 6, one in exon 8) were identified in five of nine patients who had serum antibodies recognizing p53; tumors from these patients were also strongly positive for p53 by immunohistochemistry. These and other results indicate gender differences in the genetic and biochemical alterations in lung cancer and generate hypothesis regarding gender differences in lung cancer susceptibility.

p53 is not mutated in hepatocellular carcinomas from Alaska Natives.

Hepatocellular carcinoma is common among Alaska Natives. The known risk factor in this population is hepatitis B viral infection; fungal toxins, including aflatoxin B1, have not been detected in foodstuffs. In this series of 14 patients (including 4 siblings and 2 second cousins), 3 patients were less than 12 years old at diagnosis of hepatocellular carcinoma, 8 patients were 13-24 years old, and 3 patients were more than 60 years old. Since p53 mutations occur in 29% of hepatocellular carcinomas worldwide, we tested the tumors for p53 mutations and serum samples for anti-p53 antibodies. Serum samples from these 14 patients did not contain detectable levels of anti-p53 antibodies. Loss of heterozygosity within the p53 locus was not detected in any of 9 informative cases. Immunohistochemical analysis for p53 protein accumulation was negative in all of 11 tumors. DNA sequence analysis of 12 tumor samples showed no evidence of p53 mutation in the highly conserved regions included in exons 5-8. These data, combined with one case from a previous report, indicate a mutation frequency of 0 of 13, which differs significantly from the worldwide frequency of 29% (chi 2 3.9; P = 0.048). These results indicate that liver carcinogenesis among Alaska Natives occurs independently of a traditional p53 pathway. The familial clustering and early onset in this population strongly suggest an inherited genetic predisposition to develop liver cancer. Germline mutations in a tumor suppressor or a cancer susceptibility gene are likely. Future studies of these samples should include investigations of candidate suppressor or susceptibility genes which map to chromosomal regions commonly deleted in liver cancers.