24 Feb 2012

How do glutamine-rich proteins cause neurodegenerative disease? A common theory holds that the excess glutamines make proteins glom together into toxic aggregates that trigger cellular dysfunction and death. But could polyQ proteins activate an ancient physiological death pathway that goes awry, killing cells inappropriately? So suggests a paper in today’s Science. Shai Shaham and colleagues at Rockefeller University in New York identified a polyglutamine-repeat protein that induces programmed, non-apoptotic death in worms. Certain morphological features of the dying linker cell resemble neuronal pathology in mammalian polyQ disease models, raising the possibility that unraveling the process of linker cell death could provide insight into glutamine repeat disorders in people.

Previously, Shaham and colleagues discovered that Caenorhabditis elegans roundworms have a specialized linker cell that dies in predictable fashion during male gonadal development. These cells bow out independently of caspases and other known cell death genes (Abraham et al., 2007). Electron microscopy of dying linker cells revealed ultrastructural pathology—for example, indented nuclear envelope, non-condensed chromatin, swollen organelles—also seen in neurons that die during normal vertebrate development, suggesting that the pathways driving linker cell killing may operate in mammals.

First author Elyse Blum and colleagues used interfering RNAs to hunt for worm genes that promote linker cell demise. Screening roughly 18,000 clones covering 89 percent of the C. elegans genome, Blum and colleagues turned up LIN-29 (a gene previously identified as critical for linker cell death) and five additional hits, several affecting other activities (e.g., linker cell migration) in addition to death. The present study focused on an RNAi clone that exclusively interrupted linker cell death.

The interfering RNA inactivated the PQN-41 gene, which is expressed in many cell types during development, but turns on in linker cells only when the cell starts dying (see image below). PQN-41 encodes a 427-amino acid polyglutamine protein (PQN-41C) that exists in at least two other isoforms (PQN-41A and PQN-41B) with additional upstream sequences. Of the three variants, only PQN-41C formed cytoplasmic aggregates and promoted linker cell death when transfected into wild-type worms or mutants with a PQN-41 deletion. The A and B isoforms did not aggregate and, curiously enough, enhanced linker cell survival, suggesting their N-termini may override the death-promoting activity of the glutamine-rich domain shared by all three PQN-41 variants, the authors wrote.

Death by PolyQ? The death of linker cells during male C. elegans gonadal development is controlled in part by the polyglutamine-repeat protein PQN-41. A phagocyte (green) is gobbling the dying linker cell (red), with time progressing top to bottom. Image credit: Mary Abraham and Shai Shaham, Rockefeller University, New York

PQN-41 does not appear essential for linker cell death. When the gene was inactivated in the RNAi screen, 80 percent of animals still lost their linker cells. The partial defect indicates that “PQN-41 is clearly not the only gene that participates in the death process. There must be additional genes that contribute to the killing,” Shaham said. From their RNAi screen, he and colleagues identified that stress-activated protein kinase (SEK-1) contributes to linker cell death—but determined that it acts upstream of PQN-41. LIN-29 triggers linker cell death independently of PQN-41.

The study is “interesting and somewhat unexpected. It implicates a glutamine-rich protein in an evolved cell death pathway,” said Chris Link of the University of Colorado, Boulder, whose editorial on the new findings appears in the same issue of Science. The findings “raise the possibility that some polyQ diseases may result from dysregulation of a natural death pathway, and not just because of toxic, aggregating proteins,” Link told Alzforum.

The notion finds support from other research showing that mammalian cells die by non-apoptotic means (see, e.g., ARF related news story; Degterev et al., 2005), and that blocking apoptosis in neurodegenerative disease models does not stop disease. Furthermore, one feature of linker cell death—nuclear envelope indentation—also shows up prominently in several polyQ expansion diseases, the authors found.

Whether mammals sport non-apoptotic death pathways regulated by polyQ proteins may be challenging to sort out. No known mammalian homologs of PQN-41 exist, as is often the case for proteins with repeated motifs, since their key feature is structure rather than sequence, Link noted. Huntingtin, which causes Huntington’s disease when it undergoes polyglutamine expansion, bears no resemblance to PQN-41. If, for example, another glutamine-rich gene came up as a hit in a genomewide association study for HD, “that would pique your interest,” Link said.

In the meantime, Shaham’s team has used bioinformatics programs to find mammalian proteins that structurally resemble PQN-41. Their search turned up two candidates, MED12 and p400. Both have glutamine-rich C-termini and roles in tumor formation—a process that relies on inactivation of cell death pathways. Shaham and colleagues are also trying to learn more about the mechanisms of linker cell death by studying other hits from their RNAi screen and figuring out which pathways they regulate, and if they work with PQN-41 to promote linker cell death.—Esther Landhuis