Akt1 deletion prevents lung tumorigenesis by mutant K-ras.

K-ras mutations are associated with smoking-induced lung cancer and poor clinical outcomes. In mice, K-ras mutations are sufficient to induce lung tumors, which require phosphoinoside-3-kinase (PI3K) and further downstream, mammalian target of rapamycin (mTOR) activation. However, the roles of individual Akt isoforms that link PI3K and mTOR are unknown. Here, we show that deletion of Akt1 but not Akt2 or Akt3 prevents lung tumorigenesis in a tobacco carcinogen-induced model and a genetic model. Akt1 deletion prevented tumor initiation as well as tumor progression, coincident with decreased Akt signaling in tumor tissues. In contrast, deletion of Akt3 increased tumor multiplicity in the carcinogen model and increased tumor size in the genetic model. Fibroblasts lacking Akt1 are resistant to transformation by mutant K-ras and stimulation by epidermal growth factor. Human lung cancer cells with mutant K-ras and diminished Akt1 levels fail to grow in vivo. These data suggest that Akt1 is the primary Akt isoform activated by mutant K-ras in lung tumors, and that Akt3 may oppose Akt1 in lung tumorigenesis and lung tumor progression. Given that Akt inhibitors in clinical development as cancer therapeutics are not isoform selective, these studies support specific targeting of Akt1 to mitigate the effects of mutant K-ras in lung cancer.

Inherited polymorphisms in the RNA-mediated interference machinery affect microRNA expression and lung cancer survival.

BACKGROUND: MicroRNAs (miRs) have an important role in lung carcinogenesis and progression. Single-nucleotide polymorphisms (SNPs) in genes involved in miR biogenesis may affect miR expression in lung tissue and be associated with lung carcinogenesis and progression. METHODS: we analysed 12 SNPs in POLR2A, RNASEN and DICER1 genes in 1984 cases and 2073 controls from the Environment And Genetics in Lung cancer Etiology (EAGLE) study. We investigated miR expression profiles in 165 lung adenocarcinoma (AD) and 125 squamous cell carcinoma tissue samples from the same population. We used logistic and Cox regression models to examine the association of individual genotypes and haplotypes with lung cancer risk and with lung cancer-specific survival, respectively. SNPs-miR expression associations in cases were assessed using two-sample t-tests and global permutation tests. RESULTS: a haplotype in RNASEN (Drosha) was significantly associated with shorter lung cancer survival (hazard ratio=1.86, 95% CI=1.19-2.92, P=0.007). In AD cases, a SNP within the same haplotype was associated with reduced RNASEN mRNA expression (P=0.013) and with miR expression changes (global P=0.007) of miRs known to be associated with cancer (e.g., let-7 family, miR-21, miR-25, miR-126 and miR15a). CONCLUSION: inherited variation in the miR-processing machinery can affect miR expression levels and lung cancer-specific survival.

Genetic changes in contralateral bronchioloalveolar carcinomas of the lung.

OBJECTIVE: The pattern of metastases and recurrence of bronchioloalveolar carcinoma (BAC) differs from adenocarcinoma of the lung, occurring more frequently within the lung without extrapulmonary involvement. Analyses of genetic differences of contralateral BACs may help to explain these clinical differences. METHODS: We compared paired tumors from 5 patients with contralateral metachronous BACs for loss of heterozygosity (LOH) on 6 chromosomal arms (2q, 3p, 5q, 9p, 13q and 17p) and mutational analysis of p53 and K-ras. RESULTS: Two patients, patients 1 and 2, had discordant patterns of LOH on 2 and 3 of the chromosome arms, respectively. In addition, patient 2 had a detectable K-ras mutation in his initial tumor but not in his second. These results suggest that in patients 1 and 2, the contralateral tumors were clonally unrelated. Patient 3 had no mutations in the K-ras or p53 gene and no LOH on any of the 5 informative chromosome arms. Patient 4 had LOH of 9p and mutated K-ras in both the first and the second tumor, with a mutation in the p53 gene in the first but not in the second tumor. Patient 5 had LOH of 17p and the same p53 mutations in both the first and the second tumor, with a mutation of K-ras in the first tumor but not in the second. CONCLUSIONS: The preponderance of evidence suggests that in patients 3, 4 and 5, the paired tumors were clonally related. The different patterns of LOH and mutations in clinically similar contralateral metachronous BACs provide evidence of genetic heterogeneity in the tumors of this patient group.

Attenuation of pulmonary neuroendocrine differentiation in mice lacking Clara cell secretory protein.

During development and injury, pulmonary neuroendocrine (NE) cells may transiently express Clara cell 10 kD protein (CC10), a major product of the nonciliated progenitor cells for normal and neoplastic airway epithelia suggesting a close relationship between the cells. To assess the role of CC10 during NE differentiation, we studied CC10-deficient mouse lungs by immunohistochemistry and digital imaging. The knockout model revealed a lack of the disrupted gene product in the lung. Because NE cells, which occur as solitary cells or in neuroepithelial bodies (NEBS), comprise less than 1% of airway epithelia, we counted foci positive for each of the three NE markers--synaptophysin, calcitonin gene-related peptide (CGRP), and protein gene product (PGP) 9.5--and developed a method to analyze numerous airways in serial sections. Digitized images of slides were segmented with Photoshop imaging software. The length of airway epithelium and total section areas were then measured using MetaMorph image analysis software. A comparable range of NE foci was observed regardless of CC10 expression patterns with all three markers, suggesting that CC10 is not critical for NE ontogenesis. However, discrimination according to size revealed that wild-type lungs harbored 30% to 40% greater synaptophysin- and CGRP-containing NEBs relative to CC10 deficient lungs. We posit that an attenuation of pulmonary NE differentiation afflicts the CC10-deficient state. Our imaging application greatly facilitates the acquisition and analysis of complex structures such as the lung and promises to be a widely applicable technique for assessments of tissue sections.

Constitutive achaete-scute homologue-1 promotes airway dysplasia and lung neuroendocrine tumors in transgenic mice.

The transcription factor achaete-scute homologue-1 (ASH1) is essential for neural differentiation during fetal development and is a cardinal feature of neuroendocrine (NE) tumors such as small cell lung cancer. To explore the potential of ASH1 to promote NE differentiation and tumorigenesis in the lung, we constitutively expressed the factor in nonendocrine airway epithelial cells using transgenic mice. Progressive airway hyperplasia and metaplasia developed beginning at 3 weeks of life. ASH1 potently enhanced the tumorigenic effect of SV40 large T antigen in airway epithelium. These doubly transgenic animals developed massive NE lung tumors, implying that ASH1 may cooperate with defects in p53, pRb, or related pathways in promoting NE lung carcinogenesis.

A clinicopathologic study of 100 cases of pulmonary sclerosing hemangioma with immunohistochemical studies: TTF-1 is expressed in both round and surface cells, suggesting an origin from primitive respiratory epithelium.

Pulmonary sclerosing hemangioma (SH) is a lung neoplasm of uncertain histogenesis that is composed of two major cell types: surface and round cells. The authors studied 100 cases of pulmonary SH that presented as a peripheral (95%), solitary (96%) mass of less than 3 cm in diameter (74%) in asymptomatic patients who were mostly women (83%) with a mean age of 46.2 years. Immunohistochemistry of multiple epithelial, mesothelial, pneumocyte, neuroendocrine, and mesenchymal markers was performed on 47 cases to investigate the histogenesis of this neoplasm. Both surface and round cells stained with epithelial membrane antigen (EMA) and thyroid transcription factor-1 (TTF-1) in more than 90% of cases; however, the round cells were uniformly negative for pancytokeratin and positive for cytokeratin-7 and CAM 5.2 in only 31% and 17% of cases, respectively. Surfactant proteins A and B as well as Clara cell antigen were positive in varying numbers of surface cells but they were negative in the round cells. Neuroendocrine cells either as isolated scattered cells or as a tumorlet within the center of SH were detected (chromogranin, Leu-7, synaptophysin positive) in three cases. The expression of TTF-1 in the absence of surfactant proteins A and B and Clara cell antigens in the round cells of SH suggests that they are derived from primitive respiratory epithelium. The alveolar pneumocytes and neuroendocrine cells may either represent phenotypic differentiation of a primitive respiratory epithelial component or they may correspond to non-neoplastic entrapped or hyperplastic elements. The concomitant positivity of both cell types in SH for TTF-1 and EMA, and the negativity of round cells for pancytokeratin and neuroendocrine markers, provide useful clues not only for histogenesis but also for the diagnosis of this lung neoplasm.

Morphometric analysis of CC10-hASH1 transgenic mouse lung: a model for bronchiolization of alveoli and neuroendocrine carcinoma.

Constitutive expression of human achaete-scute homolog-1 (hASH1) alone or in combination with Simian virus 40 (SV40) large T antigen (TAg) under the Clara cell 10-kDa secretory protein (CC10) promoter results in bronchiolization of alveoli and enhanced tumorigenesis, respectively. A novel morphometric system composed of series of fixed distance concentric rings originating at the bronchioloalveolar junction was used to determine spatial growth patterns. hASH1 mice exhibited progressive airway epithelial hyperplasia near this junction, and minimal changes further away in the alveolar compartment. TAg animals shared this increase, but exhibited variable distance-dependent growth. By 2 months TAg/hASH1 animals showed highly increased growth at all points measured. Remarkably, TAg/hASH1 animals expressed both CC10 and extensive neuroendocrine differentiation in airways and tumors. The results suggest that these transgenic mice provide a useful model with many similarities to human lung carcinogenesis, which originates in airway epithelium, and often reveals neuroendocrine differentiation.

Transforming growth factor-beta 1 and its receptors in human lung cancer and mouse lung carcinogenesis.

The transforming growth factor-betas (TGF-beta s) are multifunctional proteins that inhibit the proliferation of many epithelial cells through a set of cell protein receptors that includes the TGF-beta type I (RI) and type II (RII) receptors. Loss of growth inhibition by TGF-beta is thought to contribute to the development of many types of tumors. In the present study, we have examined expression of the proteins and mRNAs for TGF-beta 1, TGF-beta RI, and TGF-beta RII in normal human lung, well-characterized non-small cell lung cancer (NSCLC) cell lines, and primary NSCLC specimens. Immunohistochemical staining for TGF-beta 1, TGF-beta RI, and TGF-beta RII using specific antibodies in normal human lung showed expression of the 3 proteins in the epithelium of bronchi and bronchioles as well as in alveoli. Differential expression of TGF-beta RI and TGF-beta RII proteins was detected in 5 NSCLC cell lines using Western blot analysis, with reduced levels in 3 cell lines. A panel of 45 formalin-fixed and paraffin-embedded NSCLC specimens showed positive immunostaining for TGF-beta 1, TGF-beta RI, and TGF-beta RII, with reduced TGF-beta RII in poorly differentiated adenocarcinomas and squamous cell carcinomas and some moderately differentiated adenocarcinomas. In situ hybridization studies conducted with specific riboprobes for TGF-beta 1, TGF-beta RI, and TGF-beta RII showed corresponding localization of expression of the mRNAs in the specimens that showed positive immunostaining for the proteins. To investigate the roles of TGF-beta 1, TGF-beta RI, and TGF-beta RII in chemically induced mouse lung tumorigenesis, we examined the expression of their proteins and mRNAs in 2 mouse model systems. Whereas expression of the proteins and mRNAs for TGF-beta 1 and TGF-beta RI was comparable in lung adenomas and bronchioles of A/J mice treated with benzo(alpha)pyrene, decreased immunostaining and hybridization for TGF-beta RII protein and mRNA was detected in 50% of lung adenomas in these mice. Interestingly, expression of TGF-beta 1 and the TGF-beta receptor proteins was similar to that of bronchioles in C57B1/6 mice and their littermates heterozygous for deletion of the TGF-beta 1 gene treated with diethylnitrosamine. These data show that reduced levels of expression of TGF-beta RII occur in some, but not all, human and mouse lung tumors. This suggests that different mechanisms of action, some of which may involve the TGF-beta signaling pathway, may contribute to the progression of lung tumorigenesis.

The role of CC10 in pulmonary carcinogenesis: from a marker to tumor suppression.

CC10 is infrequently expressed in human non-small cell lung cancers (NSCLCs), despite being abundantly produced by progenitor cells for normal and neoplastic epithelium. Many abnormalities in the surrounding lung associated with field carcinogenesis, which reflect prolonged exposure to such carcinogens as tobacco smoke, also revealed altered expression of CC10. Exposure of hamsters and mice to the tobacco-specific carcinogen NNK led to reduced CC10 expression, which was partially reversible. Overexpression of CC10 in immortalized bronchial epithelial cells delayed the induction of anchorage-independent growth in response to NNK. The data suggest that downregulation of CC10 contributes to carcinogenesis because CC10 antagonizes the neoplastic phenotype.

Clinical features of patients with stage IIIB and IV bronchioloalveolar carcinoma of the lung.

BACKGROUND: The incidence of bronchioloalveolar carcinoma of the lung (BAC), a pathologically distinct type of nonsmall cell lung carcinoma (NSCLC), appears to be rising. In this study, the authors compared data on the clinical presentation and clinical courses of patients with Stage IIIB and IV BAC with data on other types of NSCLC. METHODS: The authors collected clinical, radiographic, and pathology information about 28 patients with Stage IIIB and IV BAC and 124 patients with other histologic types of NSCLC. RESULTS: Twelve of 28 BAC patients (43%) were women, compared with 40 of 124 control patients (32%). Nine (32%) of the patients with BAC had never smoked cigarettes, versus 20 controls (16%) (P = 0.02). Eighteen patients (64%) with BAC had bilateral multilobar or multicentric pulmonary involvement, compared with 13 controls (15%) (P < 0.001). Patients with advanced stage (IIIB and IV) BAC had a median survival of 15 months from the time of diagnosis; for patients with other types of Stage IIIB and IV NSCLC, had a median survival of 10 months (P = 0.01). CONCLUSIONS: Patients with BAC of the lung have clinical, radiographic, and pathologic characteristics that distinguish them from patients with other types of NSCLC. A greater proportion of women and nonsmokers present with BAC than with other types of NSCLC. Patients with advanced stage BAC are more likely to have bilateral diffuse pulmonary involvement, are less likely to develop brain metastases, and have longer survival than patients with other types of Stage IIIB and IV NSCLC. Further research is warranted to define etiology, molecular abnormalities, and more effective therapeutic interventions.

MUC1 is a novel marker for the type II pneumocyte lineage during lung carcinogenesis.

Abnormalities in mucin-type glycoprotein expression have been documented in a variety of cancers, identifying these molecules as targets for immunologically based therapies and prognostic/diagnostic assays. We examined the expression of the membrane-bound MUC1 mucin in normal, histologically atypical, and neoplastic lung to determine its potential contribution to lung carcinogenesis. In vivo, intense MUC1 immunoreactivity was present in normal type II pneumocytes as well as in a range of atypical lesions derived from type II cells and >60% of primary and metastatic non-small cell lung cancers. Expression was not associated with altered survival, although it was highly correlated with the adenocarcinoma histology. A carcinogenesis model using 4-(N-methyl-N-nitrosamino)-1-(3-pyridyl)-1-butanone-exposed hamsters revealed that MUC1 mRNA increased prior to the histological appearance of tumors. In vitro studies using MUC1 expressing non-small cell lung cancer cell lines revealed that differentiation away from a type II cell lineage was associated with dramatic down-regulation of MUC1. We propose that MUC1 is a powerful new marker for the type II pneumocyte cell lineage that allows us to follow the type II pneumocyte lineage during the process of lung carcinogenesis.

Overexpression of CC10 modifies neoplastic potential in lung cancer cells.

CC10 is infrequently expressed in non-small cell lung cancer cell lines, despite being abundantly produced by progenitor cells for normal and neoplastic airway epithelium. We overexpressed CC10 cDNA in the non-small cell lung cancer cell line A549 to determine its effect on the neoplastic phenotype. A549 cells transfected with CC10 demonstrated a marked reduction in invasiveness that was paralleled by diminished 92-kDa and absent 72-kDa metalloproteinase activity by zymography. Western analysis revealed the near absence of the corresponding matrix metalloproteinases (MMPs) MMP-2 and MMP-9 in the CC10-transfected cell lines, but not in the vector-transfected cell lines. The CC10-transfected cell lines also demonstrated decreased adhesiveness to fibronectin compared with the controls. CC10 expression was associated with decreased anchorage-independent growth but not with decreased anchorage-dependent growth. These data suggest that loss of CC10 may contribute to carcinogenesis, because CC10 antagonizes the neoplastic phenotype.

Specific primer design and exonuclease III treatment for the reduction of nonspecific staining in direct in situ PCR.

Although direct in situ PCR is more rapid than indirect method for the in situ identification of low copy number genes, several reports indicate serious nonspecific signals with this method. Some procedures have been reported in an effort to eliminate the nonspecific signals, but the results have not been satisfactory. Exonuclease III can progressively digest blunt or recessed 3' termini of double-stranded DNA whereas DNA with 3' overhanging end is resistant to digestion. DNA fragments amplified by PCR with primers incorporating the recognition site for Sph I generate 3' four bases extensions at both end after digestion. These fragments are expected to be resistant to exonuclease III digestion. It is also expected that nonspecifically incorporated digoxigenin would be released by treatment with exonuclease III thereby reducing background. We succeeded in digesting selectively with the samples after standard PCR by Sph I and exonuclease III treatment. However, we failed to eliminate the nonspecific signals of direct in situ PCR. Southern blotting revealed that the amount of nonspecific incorporation was so huge that exonuclease III was unable to release all of nonspecifically incorporated digoxigenin and maintain specific incorporated digoxigenin simultaneously. Direct in situ PCR is a sensitive method. However, its specificity has a significant problem.

K-ras point mutation occurs in the early stage of carcinogenesis in lung cancer.

In order to determine the topographical distribution of the K-ras codon 12 mutations in carcinoma and preneoplastic lesions of the lung, selective ultraviolet radiation fractionation, as well as microdissection followed by polymerase chain reaction-restriction fragment length polymorphism (PCR-RELP), was performed. Fourteen of 61 samples amplified (23.0%) had a mutation in the K-ras codon 12. Of 41 adenocarcinoma, 12 samples (29.3%) had a mutation, whereas none of the squamous cell carcinomas had a mutation. One of six large-cell carcinomas, one of three carcinoid tumours and none of three other carcinomas had a mutation. Direct sequencing revealed that K-ras codon 12 of six samples were TGT (Cys), five samples were GTT (Val), two samples were GCT (Ala) and one sample was TTT (Phe). A total of 113 lesions of 13 cases covered by dot were amplified after UV radiation. All of 74 carcinoma lesions had the mutation, and intratumour heterogeneity was not observed. Of 39 non-malignant lesions, one type II cell hyperplasia had the mutation, which suggests that the K-ras mutation occurs in the early stage of carcinogenesis. The lack of intratumour heterogeneity supports the hypothesis.

Bronchiolization of the alveoli in lung cancer: pathology, patterns of differentiation and oncogene expression.

We examined the incidence and association of bronchiolization of the alveoli with non-small cell lung cancer in lung resection specimens from 2 patient groups: those with non-small cell lung cancer and those diagnosed with a variety of non-neoplastic lung conditions. We observed marked variation in bronchiolization of the alveoli morphology ranging from normal to severely atypical and developed a classification scheme based on growth pattern, cell number and cytologic criteria. Patterns of differentiation, proliferation and growth factor receptor and oncogene expression were studied using immuno-histochemical and in situ hybridization techniques. While low-grade (0-I) bronchiolization of the alveoli lesions demonstrated markers similar to normal bronchiolar epithelium, a significant decrease in the Clara cell 10 kDa protein and tubulin and an increase in surfactant protein-A expression were observed in high-grade (II-III) lesions. Focal p53 expression was detected in 2 high-grade lesions, while c-myc mRNA and cJun protein were observed in all grades. No correlation was observed between bronchiolization of the alveoli incidence and histologic tumor type. A comparison of marker expression in lesions and tumors from the same case revealed a negative correlation between cytokeratin-14 and c-erbB-2 immuno-reactivity. Only one bronchialization of the alveoli lesion was found in the non-neoplastic patient group. We conclude that up to 12% of non-small cell lung cancer resection specimens contain bronchiolization of the alveoli lesions which exhibit altered morphology and patterns of differentiation.

Antitumor activity of a monoclonal antibody directed against gastrin-releasing peptide in patients with small cell lung cancer.

BACKGROUND: Small cell lung cancer (SCLC) cells express and secrete gastrin-releasing peptide (GRP) which binds to receptors and stimulates growth of these cells. A murine monoclonal antibody, 2A11, which binds GRP with high affinity, decreased growth of SCLC cells in vitro and in athymic nude mice. A phase 1 trial and pharmacokinetic modeling in patients with lung cancer has defined the phase 2 dose of 2A11 but the antitumor activity in patients is unknown. METHODS: Thirteen patients with previously treated SCLC received 2A11 at 250 mg/m2 over 1 h three times per week for 4 weeks. Serum GRP, urine GRP, serum levels of 2A11, and human antimouse antibodies (HAMA) were determined. RESULTS: One of 12 (8%; 95% confidence interval, 0 to 38%) evaluable patients had complete resolution of radiographically detectable tumor lasting 4 months. Four patients (33%) had stable disease. No toxic reactions were observed. The pretreatment serum GRP level of the responding patient was 3.1 fmol/mL and the median of nine nonresponding patients was 7.3 fmol/mL (range, <1.0 to 29.0). The mean trough serum 2A11 level was 49+/-18 microg/mL in the responding patient and 32 to 487 mg/mL (median, 117) in 10 nonresponding patients. HAMA did not increase during 2A11 administration in any patient. CONCLUSIONS: Interruption of the GRP autocrine growth factor loop with 2A11 results in clinical antitumor activity in a minority of patients with previously treated SCLC. Further evaluation of the antitumor effects of 2A11 is warranted to define characteristics associated with response to 2A11.

Detection of K-ras point mutation by in situ PCR in cell suspensions: comparison of the indirect and direct methods.

In situ PCR is a new technique for the localization of low copy number sequences. We report here a method for the in situ visualization of a point mutation in K-ras codon 12 by indirect in situ PCR. Twenty-five primers were examined to select mutant-specific primers. Harvested cell lines were fixed and suspended in PCR mixture. Forty cycles of PCR in cell suspension was performed in a thermal cycler using a hot start method. Cells were cytocentrifuged onto slides, and post-fixation was performed. The specimens on the slides were then hybridized with a digoxigenin-labeled probe, followed by color reaction. Both Calu-1 (mutated: TGT) and NCI-H460 (wild type: GGT) cells had strong hybridization signals in the nuclei with general primers. But with mutant-specific primers, only Calu-1 cells had hybridization signals. No signal was observed without primers or Taq DNA polymerase. Southern blotting of the same preparation confirmed desired amplification. We also applied direct in situ PCR, but this method failed to detect the point mutation. We conclude that our indirect in situ PCR method shows the feasibility of in situ identification of single cells carrying point mutations.

Analysis of Ki-ras, p53, and MDM2 genes in uterine leiomyomas and leiomyosarcomas.

Uterine sarcomas are unusual neoplasms of the female genital tract whose molecular etiology is largely unknown. We examined 20 leiomyomas as well as 23 uterine leiomyosarcomas for the presence of mutations in the Ki-ras and p53 genes, and overexpression of the MDM2 gene. Codons 12, 13, and 61 from the Ki-ras gene were characterized for the presence of mutations by restriction enzyme polymorphisms using mismatched primers and nucleic acid sequencing as appropriate. Activated Ki-ras genes were identified in 3/20 leiomyomas and 0/23 leiomyosarcomas. The p53 gene was analyzed by SSCP, nucleic acid sequencing, and immunohistochemical staining. None of 20 leiomyomas and 6/23 leiomyosarcomas exhibited p53 abnormalities. The SSCP/sequencing results did not consistently correlate with the IHC staining. MDM2 overexpression occurred in 0/20 leiomyomas and 3/23 sarcomas. Clinical correlation suggested that tumors with p53 mutations have a higher histologic grade or stage at presentation. We conclude that leiomyomas and leiomyosarcomas have different patterns of molecular alterations and are separate biologic entities. In addition, p53 and MDM2 overexpression may play a role in the development of a subset of leiomyosarcomas.

An achaete-scute homologue essential for neuroendocrine differentiation in the lung.

In Drosophila and in vertebrates, the achaete-scute family of basic helix-loop-helix transcription factors plays a critical developmental role in neuronal commitment and differentiation. Relatively little is known, however, about the transcriptional control of neural features in cells outside a neuronal context. A minority of normal bronchial epithelial cells and many lung cancers, especially small-cell lung cancer, exhibit a neuroendocrine phenotype that may reflect a common precursor cell population. We show here that human achaete-scute homologue-1 (hASH1) is selectively expressed in normal fetal pulmonary neuroendocrine cells, as well as in the diverse range of lung cancers with neuroendocrine features. Strikingly, newborn mice bearing a disruption of the ASH1 gene have no detectable pulmonary neuroendocrine cells. Depletion of this transcription factor from lung cancer cells by antisense oligonucleotides results in a significant decrease in the expression of neuroendrocrine markers. Thus, a homologue of Drosophila neural fate determination genes seems to be necessary for progression of lung epithelial cells through a neuroendocrine differentiation pathway that is a feature of small-cell lung cancer, the most lethal form of human lung cancer.