Mutations in ribosomal proteins: Apoptosis, cell competition, and cancer

Abstract

Mutations affecting multiple ribosomal proteins are implicated in cancer. Using genetic mosaics in the fruit fly Drosophila, we describe 3 apoptotic mechanisms that affect Rp/Rp homozygous mutant cells, Rp/+ heterozygous cells, or Rp/+ heterozygous cells in competition with nearby wild type cells, and discuss how apoptosis might be related to cancer predisposition.

Ribosomes are essential for growth, yet paradoxically ribosomal protein mutations are associated with cancer. Individuals who inherit mutations in ribosomal protein genes typically develop Diamond-Blackfan anemia (DBA), which, among other features, is associated with a 5.4-fold enhanced cancer risk across all tissues.¹ Mutations can also arise somatically, resulting in mosaicism. Somatic deletions encompassing ribosomal protein S14 (RpS14) cause 5q syndrome, a myelodysplastic syndrome with predisposition to acute myeloid leukemia²; sporadic mutations in RpL5 and RpL10 are thought to contribute to T-cell acute lymphoblastic leukemia;³ and mutations in RpS20 are associated with colon cancer.⁴ Although the reason why ribosomal protein mutations are oncogenic remains uncertain, ribosomal protein mutations are associated with apoptosis, and apoptosis can contribute to cancer. Using the fruit fly Drosophila melanogaster to study ribosomal protein mutant cells in vivo, we recently described 3 apoptotic mechanisms that affect ribosomal protein mutant cells.⁵

The experimental paradigm (Fig. 1) made use of mitotic recombination in heterozygous animals (Rp/+). It has long been known that heterozygotes for mutations at any of the 79 unlinked ribosomal protein gene loci survive, but most exhibit a delay in development and maturation of up to ∼30%. These mutations were known historically as ‘Minutes’, not because they are small (they are not) but because they have small adult chaetae, presumably because of a limitation in the rate of sclerotin synthesis. Mitotic recombination events, achieved through recombinase-stimulated exchange of chromosome arms, simultaneously generate wild type (+/+) and homozygous mutant (Rp/Rp) sister cells. The survival and apoptosis of these various genotypes was studied in the developing wing primordium, a rapidly proliferating tissue in the fly.⁵

Figure 1.

Apoptosis in genetic mosaics containing wild type, Rp/+, and Rp/Rp mutant cells. Cartoon showing cells dying in a mosaic tissue containing 3 cellular genotypes that arise as the recombinant sisters of +/+ cells: heterozygotes for a ribosomal protein mutation (Rp/+, magenta); wild type cells descended from Rp/+ by mitotic recombination (+/+, blue); homozygotes for ribosomal protein mutation (Rp/Rp, black). Distinct genetic requirements for apoptosis are indicated. See text for details.

The baseline level of apoptosis within wild type wing cells was very low, as expected from the virtual absence of cell death during wild type wing development. The death that did occur seemed to depend on p53. In Drosophila, as in mammals, the p53 pathway triggers apoptosis in response to DNA damage, suggesting that spontaneous DNA damage might occur in a very few wild type cells.

Perhaps surprisingly, Rp/Rp cells were found to die by apoptosis. When the pan-caspase inhibitor p35 protein derived from baculovirus was expressed transgenically, small clones of Rp/Rp cells that had survived and divided a few times were evident. The Rp/Rp cells activated Dronc, an important initiator caspase, which in turn cleaved the executioner procaspase Drice in Rp/Rp cells. We do not know whether p35 can protect Rp/Rp cells indefinitely, but their prolonged survival indicates that apoptosis is the primary mechanism that normally removes them. These observations do not imply that protein synthesis is dispensable if cell death is prevented. Recombinant Rp/Rp cells contain ribosomes inherited from the parent Rp/+ cell, they are just unable to replenish them. As numbers become depleted, apoptosis must be triggered before the cells become unable to grow at all. Which aspect of Rp/Rp cells triggers apoptosis remains to be determined, but it is independent of p53 and dependent on Dronc.

Rp/+ cells also undergo apoptosis, but sporadically. It is possible that occasional Rp/+ cells drop below the same survival threshold that is implicated in Rp/Rp cells because the genetic requirements are the same: apoptosis of Rp/+ cells is p53-independent, but depends on the initiator caspase Dronc and is prevented by baculovirus p35.

A second mechanism also affects Rp/+ cells, but only those that are in close contact with +/+ cells in the mosaic tissues (Fig. 1). This is ‘cell competition’, which has been defined as the conditional loss of a cell genotype only in the presence of cells of a different genotype.⁶ Cell competition is an interesting phenomenon with potential implications for tissue growth and surveillance in Drosophila that is only recently being described in mammals.⁷ Competitive cell death is also p53-independent, but unlike noncompetitive apoptosis only low levels of Dronc activity are involved and competitive apoptosis can be initiated almost equally as well by another initiator caspase, Dream/Strica.⁵ The distinct genetic requirements for competitive apoptosis might reflect its dependence on a close proximity to wild type cells.

Dronc, the most important initiator caspase in flies, resembles mammalian caspase-9, and its incorporation enables the apoptosome to cleave the procaspase Drice.⁸ Because cytochrome c does not activate the Drosophila apoptosome, inhibitor of apoptosis (IAP) proteins are critical regulators, and activation relies on proapoptotic proteins including Reaper, Grim, and Hid to antagonize DIAP1.⁸ Reaper is required for competitive apoptosis of Rp/+ cells, but since it is transcribed regardless of proximity to wild type cells it may not be the specific trigger for cell competition.^9,5 Whether and how Dream/Strica interacts with the apoptosome is not yet known.

There has been a recent spate of genes described to have roles in cell competition but so far none of these explain the basis of the interaction between the cell genotypes or have known connections to Dream/Strica.⁶ One hypothesis is that cells compete for Dpp, a growth factor of the transforming growth factor-β (TGF-β) family. An exciting recent finding is that cell competition requires components of the innate immune system.⁹ Since Toll-like receptors can be activated by peptides activated by extracellular proteolysis, interactions between ligands, proteases, and inhibitors from neighboring cells might occur although direct evidence for this is lacking at present.

These studies do not explain how ribosomal protein mutations lead to cancer but they suggest additional hypotheses. It is already thought that patients with DBA experience chronic, p53-dependent apoptosis in the bone marrow. This may create selection for p53 mutations, or increase the proliferation of remaining cells to such an extent that spontaneous mutations accumulate.^1,10 If the novel pathway that triggers apoptosis in Rp/Rp cells is conserved, this may create another parallel stress. K. Golic, M. Brodsky and their colleagues have suggested that cell competition may have evolved to remove aneuploid cells, since this would often alter ribosomal protein gene dose.⁵ Because such a mechanism would be ineffective in tissues that were already Rp/+, this hypothetical aneuploidy surveillance would be lost. It is apparent, however, that much more remains to be discovered regarding the consequences of ribosomal mutations in vivo.

Funding

Supported by grants from the NIH (GM061230 and GM104213) and by an unrestricted grant from Research to Prevent Blindness to the Department of Ophthalmology and Visual Sciences.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.