NEWS & VIEWS
NATURE|Vol 466|26 August 2010
condensation minimizes drag and potential breakage while the chromosomes are pushed and pulled during segregation to the opposite sides of the dividing cell. Coschi et al.3 demonstrate in mice that Rb facilitates chromoGiovanni Bosco some condensation, independently of its role in DNA replication. When combined with an The retinoblastoma protein is essential for accurate DNA replication, and additional mutation in another tumour-supits loss is commonly associated with cancer. It emerges that this protein pressor protein, mitotic defects caused by the also regulates another stage of the cell cycle. loss of Rb resulted in highly aggressive cancers. The involvement of Rb in chromosome conJust as a team of explorers needs a detailed map cycle (Fig. 1). There have been hints2 that Rb densation therefore seems to be crucial for its to guide them on their journey, a cell uses infor- also plays a direct part in segregating chromo- tumour-suppressive activity. mation held within its DNA, and packaged into somes during mitosis, although it remained Manning and colleagues4 report that Rb is chromosomes, for guidance throughout its life. unclear whether mitotic defects associated also required for the functioning of specialized It is therefore crucial that genetic information with loss of this protein were an indirect con- chromosomal regions called centromeres. is faithfully maintained for the lifetime of the sequence of errors committed during the pre- Chromosomes attach to the mitotic spindle cell and of its progeny. The tumour-suppressor ceding S phase. Could it be that Rb mediates through centromeres, which move along the protein retinoblastoma is known for its role in both accurate and timely DNA replication and spindle to the opposite ends of the dividing cell. regulating DNA replication and so maintaining proper distribution of chromosomes in mito- Rb mediates the proper condensation and origenome stability1,2. Three papers3–5 published sis? The answer is yes, according to the three entation of centromeres on duplicated chromoin Genes & Development now also implicate new studies3–5. somes and thus their attachment to the spindle. this protein in mediating proper distribution of For chromosomes to physically move Strikingly, Manning et al. find that, when the Rb genetic information between the two daughter through the viscous environment of the cell, function is compromised, the levels of chromocells during mitotic cell division. they must condense into relatively small and some mis-segregation rise to those measured in Once a cell divides, its genetic information stiff shapes. Just like folding a map into a com- tumour cells. instructs it what shape to assume, what func- pact form allows its easy storage and protects The stability of chromosomes must also be tions to perform, how quickly to proliferate it against damage during travel, presumably maintained during mitosis, when chromoagain or whether to stop dividing altosomes can break. Van Harn et al.5 gether. Indeed, some of the genes that provide evidence that mouse fibroregulate the cell-division cycle serve as blast cells lacking functional Rb and brakes to halt the cell’s journey, allowits related proteins p107 and p130 are Cell ing it to differentiate and settle into more prone to chromosome breakage. division its final tissue destination. Incorrect On cell division, broken chromosomes copying of the genetic instructions are haphazardly distributed between DNA or damage to the DNA harbouring the two daughter cells, which can lead synthesis these instructions could lead the cell to cells that are missing genetic inforG1 to the wrong destination or, worse, mation on some chromosomes or to an endless journey — a crisis that that carry duplicate copies of some warrants cell death. chromosome regions (Fig. 1). Tumour-suppressor proteins stop Apart from implicating Rb in the M phase cell division under such circumcontrol of mitotic chromosomes, the G2 stances. For instance, the retinothree studies have another theme in S phase blastoma (Rb) protein — the loss of common: loss of Rb function does not which can cause human eye tumours, doom cells to death. Consequently, as well as most other types of cancer1 — when the function of this protein is can stop cell proliferation by limiting compromised, low-level defects in M phase the expression of genes that promote any of these mitotic processes can Rb loss DNA replication during the S phase ensue, setting the stage for inheritof the cell cycle1,2 (Fig. 1). It thus puts ance of altered genetic instructions a strong brake on cell division that at each cycle of cell division, and is removed only when the cell has incrementally increasing the likelilittle chance to pass on erroneous or hood of generating a renegade cell incomplete genetic instructions. with uncontrollable proliferative Just as essential as faithful DNA potential. Chromosome Chromosome Chromosome replication is the accurate distribuHow exactly does Rb facilitate gain loss breakage, mutations and deletions tion of the replicated chromosomes chromosome condensation, ensure between the two daughter cells during centromere function and protect cell division. (By analogy, if members chromosomes from breakage? Of of the explorers’ team decide to go Figure 1 | Retinoblastoma and the cell cycle. Dividing cells undergo particular interest is the molecular their separate ways, accurate copies of cycles of division, in which the S phase (when DNA replicates) is machinery that works on chromothe map would be of little use if they separated from the M phase (when mitotic cell division occurs) by somes to condense them. What precise two gap phases, G1 and G2. The role of the retinoblastoma (Rb) were not equally distributed among protein in regulating the transition from the G1 to the S phase is well role does condensation have on centhe explorers.) tromere function? Although general established. Three studies3–5 now implicate this protein in mediating Chromosomes are equally segre- chromosome dynamics during the M phase. Outcomes of Rb chromosome condensation is abnorgated into the daughter cells during deficiency during mitosis include chromosome gain, chromosome mal in Rb-deficient cells, chromomitosis, or the M phase of the cell loss and chromosome breakage. somes nevertheless achieve sufficient CELL CYCLE
Retinoblastoma, a trip organizer
© 2010 Macmillan Publishers Limited. All rights reserved
1051
NEWS & VIEWS compaction and progress through many cell divisions. To treat cancer, might it be possible to target the chromosome-condensation machinery to completely stall mitosis and so stop tumour-cell division? The new findings3–5 give us reason to explore this possibility. ■ Giovanni Bosco is in the Department of Molecular and Cell Biology, University of Arizona,
NATURE|Vol 466|26 August 2010
between the native proteins, which suggests that the interaction does not normally occur and may require an initiating stress, including any that may lead to an age-dependent accumulation of either protein. Does ataxin-2 also modulate TDP-43 toxicity in humans, and does it have a role in neurodegeneration? Apparently, on both counts it does. Ataxin-2 contains a polyQ tract, which can be of variable length — most commonly 22 or 23 residues. An expansion of this tract to 34 or more repeats causes, in most instances, spinocerebellar ataxia type 2 (SCA2)2–4 and, in some instances, parkinsonism. Elden et al.1 demonstrate that the length of the ataxin-2 polyQ tract in humans influences TDP-43 toxicity in motor neurons, thereby also contributing to ALS (Fig. 1). The authors report that intermediate-length polyQ-tract expansions are a risk factor for ALS. Expansions ranging from 27 to 33 repeats occurred in 4.7% of around 900 patients with ALS, but in only 1.4% of healthy individuals. Association of ataxin-2 intermediate-length polyQ expansions with ALS was also accompanied by a significantly earlier age of disease onset. A common feature of SCA2 and ALS is the presence of cytoplasmic aggregates of TDP-43 and ataxin-2 in the affected cells. Indeed, Elden et al.1 find TDP-43 cytoplasmic inclusions in the cerebellum and brainstem of patients with SCA2, adding this disorder to the list of conditions involving abnormalities in TDP-43. The
Tucson, Arizona 85721-0106, USA. e-mail:
[email protected] Weinberg, R. A. Cell 81, 323–330 (1995). Sage, J. & Straight, A. F. Genes Dev. 24, 1329–1333 (2010). Coschi, C. H. et al. Genes Dev. 24, 1351–1363 (2010). Manning, A. L., Longworth, M. S. & Dyson, N. J. Genes Dev. 24, 1364–1376 (2010). 5. van Harn, T. et al. Genes Dev. 24, 1377–1388 (2010). 1. 2. 3. 4.
NEURODEGENERATION
An expansion in ALS genetics Clotilde Lagier-Tourenne and Don W. Cleveland Aggregates and mutations of the proteins ataxin-2 and TDP-43 have been implicated in distinct neurodegenerative disorders. An interplay between these proteins is now reported for amyotrophic lateral sclerosis. Amyotrophic lateral sclerosis (ALS) is a fatal, adult-onset motor-neuron disease. Although the cause of this neuromuscular disorder is not well understood, genetic factors have been implicated in roughly 10% of cases. Elden et al.1 present evidence on page 1069 of this issue that short expansions of glutamine (Q) amino-acid residues — a polyglutamine, or polyQ tract — in the ataxin-2 protein are associated with increased risk of ALS. This unexpected finding comes 15 years after the discovery 2–4 that long polyQ expansions in ataxin-2 cause spinocerebellar ataxia type 2, a neurodegenerative disorder involving abnormalities of gait. The neurotoxic effects of ataxin-2 seem to be RNA dependent and involve another protein, TDP-43. That TDP-43, an RNA/DNA-binding protein, is central to the development of neurodegeneration is well documented. Normally found in the nucleus, TDP-43 mislocalizes to the cytoplasm of neurons and glial cells to form protein aggregates in most cases of sporadic ALS5,6, and such mislocalization has been found in a growing number of other neurological disorders, for example frontotemporal lobar dementia (FTLD), Alzheimer’s disease and parkinsonism (reviewed in ref. 7). In addition, TDP-43 mutations have been reported in cases of sporadic and familial ALS8–10 and in patients with FTLD. How TDP-43 forms aggregates and how this then leads to neuronal death have been largely unknown, although for a few patients mutation in another gene (such as that encoding granulin in FTLD) is one component leading to TDP-43 aggregates. Elden et al.1 report that the cytoplasmic protein ataxin-2 may also enhance TDP-43dependent toxicity to drive the death of motor neurons in ALS. The initial evidence for the role of ataxin-2 as a modulator of TDP-43 damage came from a genetic screen in yeast. The authors then powerfully extended this lesson in yeast to the fruitfly, establishing that TDP-43 toxicity is respectively enhanced or 1052
alleviated by increased or reduced levels of the fly ataxin-2. Elden and colleagues further show that, when wild-type and mutant ataxin-2 or TDP-43 are expressed in yeast and in human cells in culture, these proteins interact physically. Intriguingly, treatment with RNase enzymes, which degrade RNA, or expression of a TDP-43 mutant that cannot bind to RNA abolishes this interaction, indicating that the ataxin-2– TDP-43 association is RNA dependent. Notably, the authors could not detect interaction
Cytoplasmic aggregates
Ataxin-2 polyQ length
a
≥34 Q PolyQ aggregates
Long polyQ expansion
TDP-43 aggregates
Predominant site of aggregation Cerebellum and brainstem Basal ganglia
Disorder
SCA2 Parkinsonism
b 27–33 Q Ataxin-2 accumulation
Intermediate-length polyQ
c 22–23 Q Normal-length polyQ
Spinal cord
Increased risk of ALS
TDP-43 aggregates Factors leading to TDP-43 mislocalization
+
Ataxin-2 accumulation TDP-43 aggregates
Spinal cord Frontotemporal cortex
ALS FTLD
Figure 1 | PolyQ-tract size matters. a, Previous studies have shown that long polyQ-tract expansion (≥34 glutamines (Q)) in ataxin-2 leads to neurodegeneration, which — depending on the brain site predominantly affected — causes spinocerebellar ataxia type 2 (SCA2) or parkinsonism. b, Elden et al.1 show that polyQ-tract expansions of intermediate length (27–33 Q) in this protein drive TDP-43 aggregation in motor neurons in the spinal cord, increasing the risk of ALS. c, Abnormal accumulation of ataxin-2 containing the normal number (22–23 Q) of glutamines, presumably mediated by other factors, has been observed in sporadic ALS and frontotemporal lobar dementia (FTLD). In at least three cases (SCA2, ALS and FTLD), TDP-43 aggregation in the affected neurons accompanies ataxin-2 accumulation. The thickness of the arrows on the right reflects the apparent frequency with which ataxin-2 accumulation is observed in associated disorders. © 2010 Macmillan Publishers Limited. All rights reserved
