Autophagy suppresses Ras-driven epithelial tumourigenesis by limiting the accumulation of reactive oxygen species.

This corrects the article DOI: 10.1038/onc.2017.175.

Autophagy suppresses Ras-driven epithelial tumourigenesis by limiting the accumulation of reactive oxygen species.

Activation of Ras signalling occurs in ~30% of human cancers; however, activated Ras alone is not sufficient for tumourigenesis. In a screen for tumour suppressors that cooperate with oncogenic Ras (RasV12) in Drosophila, we identified genes involved in the autophagy pathway. Bioinformatic analysis of human tumours revealed that several core autophagy genes, including GABARAP, correlate with oncogenic KRAS mutations and poor prognosis in human pancreatic cancer, supporting a potential tumour-suppressive effect of the pathway in Ras-driven human cancers. In Drosophila, we demonstrate that blocking autophagy at any step of the pathway enhances RasV12-driven epithelial tissue overgrowth via the accumulation of reactive oxygen species and activation of the Jun kinase stress response pathway. Blocking autophagy in RasV12 clones also results in non-cell-autonomous effects with autophagy, cell proliferation and caspase activation induced in adjacent wild-type cells. Our study has implications for understanding the interplay between perturbations in Ras signalling and autophagy in tumourigenesis, which might inform the development of novel therapeutics targeting Ras-driven cancers.

Control of tumourigenesis by the Scribble/Dlg/Lgl polarity module.

The neoplastic tumour suppressors, Scribble, Dlg and Lgl, originally discovered in the vinegar fly Drosophila melanogaster, are currently being actively studied for their potential role in mammalian tumourigenesis. In Drosophila, these tumour suppressors function in a common genetic pathway to regulate apicobasal cell polarity and also play important roles in the control of cell proliferation, survival, differentiation and in cell migration/invasion. The precise mechanism by which Scribble, Dlg and Lgl function is not clear; however, they have been implicated in the regulation of signalling pathways, vesicle trafficking and in the Myosin II-actin cytoskeleton. We review the evidence for the involvement of Scribble, Dlg, and Lgl in cancer, and how the various functions ascribed to these tumour suppressors in Drosophila and mammalian systems may impact on the process of tumourigenesis.

Loss of human Scribble cooperates with H-Ras to promote cell invasion through deregulation of MAPK signalling.

Activating mutations in genes of the Ras-mitogen-activated protein kinase (MAPK) pathway occur in approximately 30% of all human cancers; however, mutation of Ras alone is rarely sufficient to induce tumour development. Scribble is a polarity regulator recently isolated from a Drosophila screen for events that cooperate with Ras mutation to promote tumour progression and cell invasion. In mammals, Scribble regulates directed cell migration and wound healing in vivo; however, no role has been identified for mammalian Scribble in oncogenic transformation. Here we show that in human epithelial cells expressing oncogenic Ras or Raf, loss of Scribble promotes invasion of cells through extracellular matrix in an organotypic culture system. Further, we show that the mechanism by which this occurs is in the regulation of MAPK signalling by Scribble. The suppression of MAPK signalling is a highly conserved function of Scribble as it also prevents Raf-mediated defects in Drosophila wing development. Our data identify Scribble as an important mediator of MAPK signalling and provide a molecular basis for the observation that Scribble expression is decreased in many invasive human cancers.

A Drosophila G1-specific cyclin E homolog exhibits different modes of expression during embryogenesis.

We have isolated a Drosophila homolog of the human G1-specific cyclin E gene. Cyclin E proteins thus constitute an evolutionarily conserved subfamily of metazoan cyclins. The Drosophila cyclin E gene, DmcycE, encodes two proteins with a common C-terminal region and unique N-terminal regions. Unlike other Drosophila cyclins, DmcycE exhibits a dynamic pattern of expression during development. DmcycE is supplied maternally, but at the completion of the cleavage divisions and prior to mitosis 14, the maternal transcripts are rapidly degraded in all cells except the pole (germ) cells. Two modes of DmcycE expression are observed in the subsequent divisions. During cycles 14, 15 and 16 in non-neural cells, DmcycE mRNA levels show no cell-cycle-associated variation. DmcycE expression in these cells is therefore independent of the cell cycle phase. In contrast, expression in proliferating embryonic peripheral nervous system cells occurs during interphase as a brief pulse that initiates before and overlaps with S phase, demonstrating the presence of a G1 phase in these embryonic neural cell cycles. DmcycE appears not to be expressed in cells that undergo endoreplication cycles during polytenization. The structural homology to human cyclin E, the ability of DmcycE to rescue a G1 cyclin-deficient yeast strain, the presence of multiple PEST sequences characteristic of G1-specific cyclins and expression during G1 phase in proliferating peripheral nervous system cells all argue that Drosophila cyclin E is a G1 cyclin. Constitutive DmcycE expression in embryonic cycles lacking a G1 phase, in contrast to expression during the G1-S phase transition in cycles exhibiting a G1 phase, implicates DmcycE expression in the regulation of the G1 to S phase transition during Drosophila embryogenesis.

G1 control in yeast and animal cells.

In budding yeast, Saccharomyces cerevisiae, the cell cycle is controlled at the G1/S phase transition by regulating the activity of the CDC28 protein kinase. This is the budding yeast homologue of the cdc2 protein kinase associated in most organisms with control of mitosis. In budding yeast CDC28 controls both the G1/S phase transition and the G2/M phase transition by being differentially activated by two distinct classes of positive regulatory subunits known as G1 cyclins or CLNs and B-type cyclins or CLBs, respectively. To establish whether a similar dual role for Cdc2-related kinases exists in animal cells, we and others have sought human homologues of yeast G1 cyclins. Of several candidates, cyclin E is the most promising in that it accumulates prior to S phase and is associated with a pre-S phase protein kinase activity. The kinetics of accumulation of cyclin E-associated protein kinase activity is consistent with a role at the mammalian cell cycle restriction point.

A cyclin B homolog in S. cerevisiae: chronic activation of the Cdc28 protein kinase by cyclin prevents exit from mitosis.

A cyclin B homolog was identified in Saccharomyces cerevisiae using degenerate oligonucleotides and the polymerase chain reaction. The protein, designated Scb1, has a high degree of similarity with B-type cyclins from organisms ranging from fission yeast to human. Levels of SCB1 mRNA and protein were found to be periodic through the cell cycle, with maximum accumulation late, most likely in the G2 interval. Deletion of the gene was found not to be lethal, and subsequently other B-type cyclins have been found in yeast functionally redundant with Scb1. A mutant allele of SCB1 that removes an amino-terminal fragment of the encoded protein thought to be required for efficient degradation during mitosis confers a mitotic arrest phenotype. This arrest can be reversed by inactivation of the Cdc28 protein kinase, suggesting that cyclin-mediated arrest results from persistent protein kinase activation.

Human cDNAs encoding homologs of the small p34Cdc28/Cdc2-associated protein of Saccharomyces cerevisiae and Schizosaccharomyces pombe.

The Cks1 protein is a component of the Cdc28 protein kinase in the budding yeast Saccharomyces cerevisiae. This paper reports the cloning of two homologs of the S. cerevisiae CKS1 gene from human cells. These homologs, CKShs1 and CKShs2, both encode proteins of 79 amino acids that share considerable homology at the amino acid level with the products of CKS1 from S. cerevisiae and suc1+ from the fission yeast Schizosaccharomyces pombe. Both human homologs are capable of rescuing a null mutation of the S. cerevisiae CKS1 gene when expressed from the S. cerevisiae GAL1 promoter. S. pombe suc1+ expressed from the GAL1 promoter is also capable of rescuing a S. cerevisiae cks1 null mutation. Ckshs1 or Ckshs2 protein linked to Sepharose beads can bind the Cdc28/Cdc2 protein kinase from both S. cerevisiae and human cells. The CKShs1 and CKShs2 mRNAs are expressed in different patterns through the cell cycle in HeLa cells, which may reflect specialized roles for the encoded proteins.

An essential G1 function for cyclin-like proteins in yeast.

Cyclins were discovered in marine invertebrates based on their dramatic cell cycle periodicity. Recently, the products of three genes associated with cell cycle progression in S. cerevisiae were found to share limited homology with cyclins. Mutational elimination of the CLN1, CLN2, and DAF1/WHI1 products leads to cell cycle arrest independent of cell type, while expression of any one of the genes allows cell proliferation. Using strains where CLN1 was expressed conditionally, the essential function of Cln proteins was found to be limited to the G1 phase. Furthermore, the ability of the Cln proteins to carry out this function was found to decay rapidly upon cessation of Cln biosynthesis. The data are consistent with the hypothesis that Cln proteins activate the Cdc28 protein kinase, shown to be essential for the G1 to S phase transition in S. cerevisiae. Because of the apparent functional redundancy of these genes, DAF1/WHI1 has been renamed CLN3.

A family of cyclin homologs that control the G1 phase in yeast.

Two Saccharomyces cerevisiae genes were isolated based upon their dosage-dependent rescue of a temperature-sensitive mutation of the gene CDC28, which encodes a protein kinase involved in control of cell division. CLN1 and CLN2 encode closely related proteins that also share homology with cyclins. Cyclins, characterized by a dramatic periodicity of abundance through the cell cycle, are thought to be involved in mitotic induction in animal cells. A dominant mutation in the CLN2 gene, CLN2-1, advances the G1- to S-phase transition in cycling cells and impairs the ability of cells to arrest in G1 phase in response to external signals, suggesting that the encoded protein is involved in G1 control of the cell cycle in Saccharomyces.

Analysis of the Cdc28 protein kinase complex by dosage suppression.

In the interest of identifying components of the Cdc28 protein kinase complex, dosage suppression analysis was performed on temperature-sensitive and dominant negative CDC28 mutations. Dosage suppression is based on a rationale in which elevated expression of wild-type genes can rescue mutations in a target gene as a result of interaction between the respective encoded proteins. Three sequences capable of rescuing a temperature sensitive cdc28 mutation were isolated from a library of wild-type genomic DNA segments in the high copy vector YEp13. Two of these, named CLN1 and CLN2 were found to encode closely related proteins with homology to cyclins. The third, CKS1, encodes an 18K (K = 10(3) Mr) protein that has been shown to be a component of the Cdc28 protein kinase complex and is a homolog of the suc1+ product of fission yeast. A number of dosage suppressors of the CDC28-dn1 dominant negative mutation have been isolated. The one analyzed to date encodes a truncated subunit of the mitochondrial enzyme succinyl-CoA synthetase. The basis for suppression in this case remains to be elucidated.

Dominant negative protein kinase mutations that confer a G1 arrest phenotype.

The CDC28 gene of Saccharomyces cerevisiae encodes a protein kinase that is required for passage through the G1 phase of the cell cycle. We have used an inducible promoter fused to the CDC28 coding sequence to isolate conditionally dominant mutant alleles of CDC28. Overexpression of these dominant alleles causes arrest in the G1 phase of the cell cycle but permits the distinctive asymmetric growth that is characteristic of recessive temperature-sensitive cdc28 mutants. The dominant alleles encode products with no detectable protein kinase activity, and their phenotypic effects can be suppressed by simultaneous overproduction of the wild-type protein. DNA sequence analysis showed that the mutant site in at least one of the dominant alleles is in a residue that is highly conserved among protein kinases. These properties are best understood if the dominant mutation results in the catalytic inactivation of the protein kinase but still allows the binding of another component needed for CDC28 function. By this model, high levels of the mutant protein arrest cell division by denying the wild-type protein access to this other component. Suppressors that may encode this other component have been isolated on high-copy-number plasmids.