Blood-based genomics of triple-negative breast cancer progression in patients treated with neoadjuvant chemotherapy.

BACKGROUND: As neoadjuvant chemotherapy (NAC) is increasingly used in triple-negative breast cancer (TNBC), we investigated the value of circulating tumor DNA (ctDNA) for patient monitoring prior, during, and after NAC, and circulating tumor cells (CTCs) for disease characterization at clinical progression. MATERIALS AND METHODS: Forty-two TNBC patients undergoing NAC were prospectively enrolled. Primary tumor mutations identified by targeted-gene sequencing were validated and tracked in 168 plasma samples longitudinally collected at multiple time-points by droplet digital polymerase chain reaction. At progression, plasma DNA underwent direct targeted-gene assay, and CTCs were collected and analyzed for copy number alterations (CNAs) by low-pass whole genome sequencing. RESULTS: ctDNA detection after NAC was associated with increased risk of relapse, with 2-year event-free survival estimates being 44.4% [95% confidence interval (CI) 21.4%-92.3%] versus 77.4% (95% CI 57.8%-100%). ctDNA prognostic value remained worthy even after adjusting for age, residual disease, systemic inflammatory indices, and Ki-67 [hazard ratio (HR) 1.91; 95% CI 0.51-7.08]. During follow-up, ctDNA was undetectable in non-recurrent cases with the unique exception of one showing a temporary peak over eight samples. Conversely, ctDNA was detected in 8/11 recurrent cases, and predated the clinical diagnosis up to 13 months. Notably, recurrent cases without ctDNA developed locoregional, contralateral, and bone-only disease. At clinical progression, CTCs presented chromosome 10 and 21q CNAs whose network analysis showed connected modules including HER/PI3K/Ras/JAK signaling and immune response. CONCLUSION: ctDNA is not only associated with but is also predictive of prognosis in TNBC patients receiving NAC, and represents an exploitable tool, either alone or with CTCs, for personalized TNBC management.

Analysis of gene expression identifies PLAB as a mediator of the apoptotic activity of fenretinide in human ovarian cancer cells.

Fenretinide (4-HPR) is a synthetic retinoid with antitumor activity, which induces apoptosis in cancer cell lines of different histotypes. To identify genes contributing to its apoptotic activity in ovarian cancer cells, we monitored, by cDNA arrays, gene expression changes after 4-HPR exposure in A2780, a human ovarian carcinoma cell line sensitive to the retinoid. Among the differentially expressed transcripts, PLAcental Bone morphogenetic protein (PLAB), a proapoptotic gene, was the most highly induced. In a panel of ovarian carcinoma cell lines with different 4-HPR sensitivities, PLAB upregulation was associated with cellular response to 4-HPR, its overexpression increased basal apoptosis and its silencing by small interfering RNA decreased the ability of 4-HPR to induce apoptosis. PLAB induction by 4-HPR was p53- and EGR-1 independent and was regulated, at least in part, by increased stability of PLAB mRNA. PLAB up-modulation by 4-HPR also occurred in vivo: in ascitic cells collected from patients with ovarian cancer before and after 4-HPR treatment, PLAB was upmodulated in 2/4 patients. Our results in certain ovarian cancer cell lines indicate a role for PLAB as a mediator of 4-HPR-induced apoptosis. The correlation of increased PLAB in vivo with antitumor activity remains to be established.

Involvement of c-Fos in fenretinide-induced apoptosis in human ovarian carcinoma cells.

Fenretinide (HPR), a synthetic retinoid that exhibits lower toxicity than other retinoids, has shown preventive and therapeutic activity against ovarian tumors. Although the growth inhibitory effects of HPR have been ascribed to its ability to induce apoptosis, little is known about the molecular mechanisms involved. Since the proto-oncogene c-Fos has been implicated in apoptosis induction, we analyzed its role in mediating HPR response in a human ovarian carcinoma cell line (A2780) sensitive to HPR apoptotic effect. In these cells, HPR treatment caused induction of c-Fos expression, whereas such an effect was not observed in cells made resistant to HPR-induced apoptosis (A2780/HPR). Moreover, in a panel of other human ovarian carcinoma cell lines, c-Fos inducibility and HPR sensitivity were closely associated. Ceramide, which is involved in HPR-induced apoptosis, was also involved in c-Fos induction because its upregulation by HPR was reduced by fumonisin B(1), a ceramide synthase inhibitor. The causal relationship between c-Fos induction and apoptosis was established by the finding of an increased apoptotic rate in cells overexpressing c-Fos. Similarly to that observed for c-Fos expression, HPR treatment increased c-Jun expression in HPR-sensitive but not in HPR-resistant cells, suggesting the involvement of the transcription factor activating protein 1 (AP-1) in HPR-induced apoptosis. In gene reporter experiments, HPR stimulated AP-1 transcriptional activity and potentiated the AP-1 activity induced by 12-tetradecanoylphorbol 13-acetate. Furthermore, inhibition of AP-1 DNA binding, by transfecting A2780 cells with a dominant-negative Fos gene, caused decreased sensitivity to HPR apoptotic effects. Overall, the results indicate that c-Fos plays a role in mediating HPR-induced growth inhibition and apoptosis in ovarian cancer cells and suggest that c-Fos regulates these processes as a member of the AP-1 transcription factor.

Decrease in drug accumulation and in tumour aggressiveness marker expression in a fenretinide-induced resistant ovarian tumour cell line.

We investigated whether the efficacy of fenretinide (HPR) against ovarian tumours may be limited by induction of resistance. The human ovarian carcinoma cell line A2780, which is sensitive to a pharmacologically achievable HPR concentration (IC(50)= 1 microM), became 10-fold more resistant after exposure to increasing HPR concentrations. The cells (A2780/HPR) did not show cross-resistance to the synthetic retinoid 6-[3-adamantyl-4-hydroxyphenyl]-2-naphthalene carboxylic acid (CD437) and were not sensitive, similarly to the parent line, to all-trans-retinoic acid, 13-cis-retinoic acid or N-(4-methoxyphenyl)retinamide. A2780/HPR cells showed, compared to parental cells, a 3-fold reduction in colony-forming ability in agar. The development of HPR resistance was associated with a marked increase in retinoic acid receptor beta (RARbeta) mRNA and protein levels, which decreased, together with drug resistance, after drug removal. The expression of cell surface molecules associated with tumour progression including HER-2, laminin receptor and beta1 integrin was markedly reduced. The increase in the levels of reactive oxygen species is not involved in HPR-resistance because it was similar in parental and resistant cells. Conversely differences in pharmacokinetics may account for resistance because, in A2780/HPR cells, intracellular peak drug levels were 2 times lower than in A2780 cells and an as yet unidentified polar metabolite was present. These data suggest that acquired resistance to HPR is associated with changes in marker expression, suggestive of a more differentiated status and may be explained, at least in part, by reduced drug accumulation and increased metabolism.

Role of retinoic acid receptor overexpression in sensitivity to fenretinide and tumorigenicity of human ovarian carcinoma cells.

The role of retinoic acid receptor (RAR) expression in sensitivity to N-(4-hydroxyphenyl)retinamide (4HPR or fenretinide) as well as on the tumorigenicity of human ovarian carcinoma cells was examined. Two human ovarian cancer cell lines, A2780 and IGROV-1, with a 10-fold difference in sensitivity to 4HPR were chosen to study RAR involvement in the response to 4HPR. To determine which RAR was effective, RARalpha, beta and gamma were individually overexpressed in A2780 cells, which are the most sensitive to 4HPR. Sensitivity to 4HPR was increased in RARbeta-overexpressing clones, whereas it was slightly decreased in RARalpha transfectants (which had diminished RARbeta expression) and was unchanged in clones transfected with RARgamma. IGROV-1 cells, which are RARbeta negative, were transfected with RARbeta. Surprisingly, none of the obtained IGROV-1 RARbeta transfectants expressed RARbeta protein, in spite of RARbeta mRNA transcription. All clones were similar to the parental IGROV-1 cells in their sensitivity to 4HPR. Treatment with a pharmacologically achievable concentration of 4HPR (1 microM) led to a rapid 2-fold increase in RARbeta mRNA levels in A2780 cells, but it did not induce RARbeta expression in IGROV-1 cells. Analysis of the tumorigenicity of A2780-transfected clones revealed that overexpression of RARalpha was associated with a significant reduction in tumor takes (50% and 67%, respectively, vs. 96% for the parent line) and with a reduced growth rate. Oncogenicity was clearly decreased in only 1 of the 2 RARbeta-overexpressing clones (33% takes) and was unchanged in the 2 clones with increased RARgamma expression. Our results demonstrate that basal expression and 4HPR inducibility of RARbeta play a role in mediating 4HPR response in ovarian cancer cells. The findings of reduced oncogenicity of clones overexpressing RARalpha and of one clone overexpressing RARbeta indicate that RARalpha and RARbeta might have a tumor-suppressive effect in ovarian tumors.

Expression of the epidermal growth factor receptor gene in human intestinal metaplasia: a preliminary report.

BACKGROUND: The role of growth factors/receptors in the etiopathology and/or development of gastric cancer has recently come under scrutiny, since overexpression or amplification of the EGF system has been found in many intestinal type gastric cancers and related to a more aggressive behavior. Since these gastric carcinomas appear to develop from intestinal metaplasia, a study was planned to investigate whether overexpression of the EGF-receptor gene also occurred in intestinal metaplastic mucosa. METHODS: Patients underwent upper GL endoscopy. Gastric biopsies for routine histology, Helicobacter pylori detection, quantification of intestinal metaplasia and EGF-R expression analysis were performed. A 30mer EGF-R specific oligonucleotide was end-labeled and used to probe a dot blot filter containing the RNA from the bioptic samples. RESULTS: Though all the gastric samples transcribed the EGF-R gene to a detectable level, overexpression of the EGF-R gene was found in the metaplastic mucosa in a minority of patients. CONCLUSIONS: These preliminary findings suggest that overexpression of the EGF-R gene is infrequent in the metaplastic gastric mucosa.

Functional characterization of a human DNase-like protein encoded by a gene positioned in Xq28.

Xib, a gene recently reported to reside on the q28 region of the human X chromosome [Pergolizzi et al. (1996) Gene 168, 267-270], contains an open reading frame homologous to those of the DNase I family enzymes. The full open reading frame of this gene has been fused to the E. coli gene of the maltose binding protein and expressed in bacteria as a chimeric protein. The partially purified chimeric protein is enzymatically active. It introduces single and double stranded breaks into supercoiled DNA, at 30 degrees C in the absence of divalent cations and at a pH optimum of 5.2. To our knowledge this enzyme represents the first cloned human endonuclease with characteristics similar to those of acidic DNase II.

Identification of a U7snRNA homologue mapping to the human Xq27.1 region, between the DXS1232 and DXS119 loci.

To contribute to the identification and analysis of novel genes, we undertook the study of a cosmid clone in the Xq27 region of human DNA. The cloned fragment was previously observed to have a high number of evolutionarily conserved sequences. In this genomic stretch of DNA we have identified sequence homologous to the U7 RNA gene including its potential regulatory elements. This paper describes the genomic organisation of this gene and its mapping to the Xq27.1 genomic sub-interval between the DXS1232 and DXS119 loci.

Sequence and gene content in 35 kb genomic clone mapping in the human Xq27.1 region.

This paper presents detailed analysis of the entire sequence of a cosmid clone, 26H7, containing 35 kb of human DNA. This cosmid resides on the q27.1 region of the human X chromosome between, DXS1232 and DXS119 loci. Novel potential small exons were detected for which conventional gene identification strategies (Northern blot analysis and extensive cDNA library screening) proved to be inefficient. Of the standard repetitive elements we found: 8 Alu's making up 6.2% of the sequence; 10 MIR segments (4.1%); 5 LINE1 elements (4.8%), 3 MIR2 (1.0%); 2 MLT (2.9%), and 1 MSTA (0.7%) representing about 20% of the total sequence. The overall GC content was rather low, only 42% and no CpG island was detected using rare restriction enzymes. However, a CpG-rich region was identified. Computer aided analysis of the sequence inferred the presence of three possible genes: one of them was found to be homologous to the U7 RNA family elements; a second is reported in this paper, however at the moment no significant homology has been found in the data bank. The third predicted gene has not as yet been found to be detectable by RT-PCR. We also report in this paper the identification of X-chromosome specific repeated sequences.

Isolation of a pancreas-specific gene located on human chromosome 14q31: expression analysis in human pancreatic ductal carcinomas.

We have isolated a novel cDNA (SEL1L) that shows sequence similarities to SEL-1, a gene identified as an extragenic suppressor of the lin-12 hypomorphic mutant from Caenorhabditis elegans (7, 8). SEL1L exhibits a tissue-specific pattern of expression: a single poly(A)+ RNA species of 7.5 kb is abundantly expressed only in the pancreas of healthy individuals, whereas low to undetectable levels are observed in other adult and in some fetal tissues. Somatic hybrid panel and fluorescence in situ hybridization positioned this gene in the q31 band of human chromosome 14. The tissue-specific expression of this gene induced us to study its role in human pancreatic carcinomas. Our analysis revealed that 17% of adenocarcinomas of the pancreas did not express SEL1L to a detectable level; however, no gross genomic alterations were apparent in the few hundred kilobases of the relevant region.

Cloning of a gene encoding a DNase I-like endonuclease in the human Xq28 region.

To contribute to the isolation of genes within the q-24-qter region of the human X chromosome,we screened three cDNA libraries (human fetal brain, liver and skeletal muscle) with a cosmid clone containing a CpG island previously mapped in the q28 region. A full-length 2.1-kb cDNA clone was isolated (XIB); DNA databank searches revealed identity with an EST fragment (XAP-1), residing between the RCP/GCP and G6PD loci. The XIB coding region (909 bp) showed 44% amino acid (aa) identity to pig DNase I. Several conserved residues have been observed between these two genes including aa in the active site. XIB expressed a single transcript in adult heart and skeletal muscle, whereas, in some fetal tissues, two different-sized transcripts were seen. Zoo blot analysis showed a remarkable cross-species conservation. Expression and sequence of this novel gene are reported.

Improved procedure for the preparation of DNA restriction fragments suitable for sequencing.

A rapid and integrated procedure was developed for the preparation of small DNA restriction fragments (< or = 1000 bp) starting from a large cosmid (35,000 bp) containing exogenous DNA. The process is based on restriction enzymatic digestion followed by HPLC separation and fractions collection. All DNA fragments are separated in a single run, detected "on-line" by UV absorption, and straightforward collected with very high recovery. Small fragments can be directly subjected to the sequence procedure, whereas those larger than 1000 bp are redigested with a second enzyme, the fractionated subfragments are separated, ligated to plasmid vector, and sequenced. A human genomic cosmid of 35,000 bp (26H7) has been chosen as a model.