The kinase PINK1 and the E3 ubiquitin ligase Parkin are mutated in some forms of Parkinson's disease and it has been of interest to understand how these genes function to prevent neurodegeneration. One hypothesis is that these proteins control the targeted degradation of dysfunctional (i.e. likely ROS-overproducing) mitochondria via mitophagy. Now Richard Youle's lab at the NIH has performed a really beautiful and well-controlled set of genome-wide, high-content RNAi screens to find new modulators of Parkin mitochondrial translocation, including TOMM7, HSPA1L and BAG4. A good illustration of the power of high-content screening.
Fifteen years ago Lee Hartwell, Stephan Friend and colleagues published a classic paper suggesting that synthetic lethal approaches in yeast and other model organisms could by applied profitably to the identification of genotype-selective anticancer agents. This concept has been immensely influential in the field and upon my own work at the University of Toronto and at Columbia. But...has the potential of this approach been realized as far as developing successful new therapies for treating cancer? We get an update on progress made so far (spoiler: not much) and directions for future work in Science.
There is a lot of interest in how neurons die during various neurodegenerative processes. Is it always via apoptosis? This paper from the Dawson lab implicates parthanatos (PARP-dependent non-apoptotic death) in the degenerative death of dopaminergic neurons in a mouse model of Parkinson's disease. This has important therapeutic implications as it suggests that we should consider drugging PARP to prevent Parkinson's in some cases.
One of the more interesting findings in the ROS field in recent years has been that certain stem cell populations must maintain low levels of ROS to retain stem-ness; high ROS can deplete stem cell pools by favoring terminal differentiation (and death?). This recent paper described an interesting mechanism involving an interaction between the thioredoxin interacting protein (TXNIP) and p53 in the regulation of blood stem cell number.
A few months ago Bert Vogelstein published an excellent analysis of the accumulated cancer genome sequencing data suggesting that we had identified basically all the cancer driver mutations (~140) that are likely to be found in most tumors. In essence, we have saturated the screen. However, now the Elledge lab suggests an alternative perspective, that a many more genes are involved in driving tumor formation but each in a graded manner. In either case, it seems clear that the age of new oncogene identification is drawing to a close. Next task: figuring out what many of these genes do and how they can be targetted to induce tumor-specific cell death.
Why do cancer cells actually die in response to chemotherapy? Cancer cells are generally said to be resistant to apoptosis (e.g. due to p53 mutation, BCL-2 overexpression, etc) so you'd think chemo would never work. But it does work very well in some cases but not others. Why? Lately the Letai lab has been looking at this question using BH3 profiling and made a number of exciting and high profile observations summarized by Kris Sarosiek here. Also, in eLife, Jim Wells' lab at UCSF takes an interesting 'omics approach to understanding how proteasomal inhibition leads to the death of myeloma cancer cells.
Michael White's lab at UT Southwestern performed a large siRNA screen looking for genes that were preferentially required for cell viability in a lung tumor cell line model. The results are interesting. Few of the essential genes identified in one cell line were also essential across a panel of other lung tumor cell lines. Further work revealed specific addictions to FLIP (an undruggable target, currently) in the small subset of NLRP3-mutated tumor cell lines, and addiction to lysosomal function in the small subset of KRAS + LBK1 mutant tumor cells. It is becoming apparent that finding useful synthetic lethal interactions is very difficult and will require the search space to be broadened significantly, especially with respect to the mutations considered actionable.
Jason Moffat's lab in Toronto has published a synthetic lethal genetic interaction map for human cancer cells using panels of isogenic human HCT-116 cancer cell lines and a pooled shRNA library targeting most genes in the genome. They report a number of novel gene-specific interactions for PTEN -/-, KRAS G12D and other genotypes and do some very nice genetic and biochemical follow-up studies. Interestingly, the apoptosis inhibitor XIAP emerges as a common synthetic partner in several mutant backgrounds, although this was not a focus of the paper. Many years ago in the Boone Lab I shared a bench with Franco Vizeacoumar, the first author on this report. Nice to see what he has been doing!
A nice review by of the various ways that apoptotic cell death pathways are inactivated in cancer cells (genetic, epigenetic, post-translational, etc), thereby contributing to tumor formation in vivo. Does the inactivation of non-apoptotic cell death pathways also contribute to tumor formation or viability in some circumstances?
BCL-2 family proteins (Bcl-2, Bcl-xL, Mcl-1, etc) are typically thought to prevent cell death by inhibiting the pro-apoptotic action of other BH3 family proteins. This paper (as well as previous work from the Hardwick lab) demonstrates how an N-terminal fragment of Bcl-xL, presumably generated by caspase cleavage, can actually act as a lethal stimulus to kill neurons following ischemia. The complex anti- and pro-death activities of Bcl-2 family members is fascinating.
There was a lot of excitement surrounding the development of small molecule PARP inhibitors as synthetic lethal agents targeting BRCA-mutant cancers, then disappointment about the results of a phase III trial that failed to show any benefit and now the finding that the first generation clinical PARP inhibitor Iniparib maybe wasn't that good at inhibiting PARP after all. The linked article provides a terrific overview of where improvements in the process of drug development need to be made, starting with better preclinical research.
Like iron, copper is an important transition element required for normal cell function and thought to contribute to cell death via Fenton chemistry when in excess. Two recent papers demonstrate how Cu is required normally for MAPK signaling via interesting mechanisms - one mechanism involves the modulation of physiological ROS levels (via Cu-dependent SOD1 activity) and the other involves the direct binding of MEK to two Cu atoms which enhances MEK interaction with ERK.
Lots of excitement about the role of PKM2 expression in cancer since the 2008 paper from the Cantley lab. Now seems to be much more complex in vivo and may not be such a hot target for cancer therapy. Will this dampen the general enthusiasm for cancer therapies targeting metabolism?
This is a very interesting perspective on the role of caspases in cell death. Long thought to be responsible for the execution of apoptosis, it is increasingly clear that, at least in mammals, caspase activation is not required for the cell to die. The authors propose instead that caspase-mediated cleavage dictates whether or not a given stimulus triggers non-inflammatory apoptosis versus pro-inflammatory necrosis; that is, caspases are activated to dismantle what would otherwise represent pro-inflammatory proteins.