RESEARCH

HIGHLIGHTS
HIGHLIGHT ADVISORS SEAN B. CARROLL UNIVERSITY OF WISCONSIN, USA NANCY J. COX UNIVERSITY OF CHICAGO, USA RALPH J. GREENSPAN THE NEUROSCIENCES INSTITUTE, CALIFORNIA, USA YOSHIHIDE HAYASHIZAKI RIKEN GENOMIC SCIENCES CENTER, JAPAN PETER KOOPMAN UNIVERSITY OF QUEENSLAND, AUSTRALIA LEONID KRUGLYAK FRED HUTCHINSON CANCER RESEARCH CENTER, USA STANLEY MALOY SAN DIEGO STATE UNIVERSITY, USA BARBARA MEYER UNIVERSITY OF CALIFORNIA, BERKELEY, USA JOHN QUAKENBUSH THE INSTITUTE FOR GENOMIC RESEARCH, USA NADIA ROSENTHAL EMBL MONTEROTONDO, ITALY NORIYUKI SATOH KYOTO UNIVERSITY, JAPAN MARC VIDAL DANA-FARBER CANCER INSTITUTE, BOSTON, USA VIRGINIA WALBOT STANFORD UNIVERSITY, USA DETLEF WEIGEL MAX PLANCK INSTITUTE FOR DEVELOPMENTAL BIOLOGY, GERMANY PHIL ZAMORE UNIVERSITY OF MASSACHUSETTS, USA LEONARD I. ZON CHILDRENS HOSPITAL, BOSTON, USA HUMAN GENOMICS

Enjoy the complete coverage

2001 was a landmark year in genomics in February, the draft sequence of the human genome was published. Although the word draft was dropped by many, the International Human Genome Sequencing Consortium (IHGSC) did not forget that there was work left to do. In a follow-up paper in Nature, they now report the results of the effort to finish what they had left undone in 2001. The 2001 draft sequence omitted 10% of the euchromatin and was interrupted by ~150,000 gaps; moreover, many local segments remained unordered and the corresponding clones were not oriented. The finishing process involved producing finished maps continuous and accurate paths of overlapping large-insert clones spanning the euchromatic region of each chromosome arm and finished clones continuous and accurate nucleotide sequence across each large-insert clone. The finished product (known as Build 35) now contains 2.85 billion nucleotides interrupted by only 341 gaps. The euchromatic human genome is now ~99% complete, with an error rate of only 1 in 105 bases. Among the many interesting facts about the genome that the report considers are the gaps that remain in the current sequence. Thirty three of them fall within heterochromatin, which was not targeted by the human genome project owing to its repetitive nature. A further 35 gaps are at the boundaries between euchromatin and heterochromatin, but the remaining 273 gaps lie within euchromatin and are mostly associated with segmental duplications. The IHGSC intend to close all of these euchromatic gaps, but rather than persevere with the existing strategies, which have so far failed in these cases, they propose to develop focused strategies, such as using DNA from a single haploid source and new kinds of large insert libraries. As the quality of the sequence went up, the estimate of the total number of protein coding genes went down. The 2001 prediction was in the region of 30,000; now, the IHGSC is talking about 20,00025,000; the figures are derived from a combination of high quality sequence and gene annotation, which has now been manually augmented, taking into consideration gene structure, ESTs and transcript evidence. As the members of the IHGSC point out, this number refers to protein-coding genes only and point out that almost nothing is known about the non-coding transcripts, despite the recent flood of reports about them. So here we are, the euchromatic sequence of the human genome has essentially been completed. But what of its heterochromatic portion? Will that be sequenced? The IHGSC says that to overcome the cloning and sequencing problems, new technologies would have to be developed. But will knowing the sequence of heterochromatic regions actually be that helpful in understanding this portion of the genome?
Magdalena Skipper References and links
ORIGINAL RESEARCH PAPER International Human Genome Sequencing Consortium. Finishing the euchromatic sequence of the human genome. Nature 21 October 2004 (doi: 10.1038/nature03001) FURTHER READING Eichler, E. E. et al. An assessment of the sequence gaps: unfinished business in a finished human genome. Nature Rev. Genet. 5, 345354 (2004)

804

| NOVEMBER 2004 | VOLUME 5

2004 Nature Publishing Group

www.nature.com/reviews/genetics

RESEARCH HIGHLIGHTS

IN THE NEWS
A barcode for life? According to two recent studies, a DNA barcode could revolutionize taxonomy, potentially saving hours of peering down microscopes or poring over lists of morphological features to identify species. The barcode in question is a 648-bp stretch of the mitochondrial gene cytochrome c oxidase-I. As mitochondrial genes mutate at a high rate, enough changes should have taken place in this gene to provide a unique sequence for each species, allowing taxonomists to quickly and accurately identify specimens. A group led by Paul Hebert at the University of Guelph tested the technique in a study of 260 bird species. The barcoding approach proved an accurate way of distinguishing between species, and even identified four potential new species that might have been missed previously. [Birds are] big, theyre coloured differently, and they sing different songs yet even in that easy to identify group, there are hidden species, commented Hebert (CBC News Online). In a second study, the same technique revealed that the skipper butterfly, Astrapes fulgerator, is actually made up of at least ten species that look similar as adults, but have different characteristics as caterpillars. Taxonomist Felix Sperling, who wasnt involved in the study, is enthusiastic, describing this work as an excellent demonstration of the power of DNA barcoding to make sense of a confusing welter of ecological and color pattern variation (The Scientist Online). But the method is less popular among some taxonomists, and even those who are in favour are far from suggesting that barcoding is the solution to all taxonomic problems. Theres strong debate about whether one size fits all, stresses ecologist Craig Moritz, We have to be a little bit cynical about where it works and where it doesnt (news@nature.com).

EVO-DEVO

Breaking up the family

A new study that reveals Hox genes scattered throughout the genome of the tunicate Oikopleura dioica has put to rest the perceived wisdom that the genes in this family need to be clustered. Hox genes are crucial for specifying differential development along the anteroposterior axis in bilaterally symmetrical animals. These genes have always been found in a genomic cluster, usually ordered in a way that corresponds to the sequential expression of the genes along the anteroposterior axis during development. Seo and colleagues studied the Hox genes of O. dioica, a representative of a chordate lineage that diverged at an early stage from the lineage that gave rise to vertebrates. We already knew that the Hox genes of tunicates might be slightly unusual after the sequencing of Ciona intestinalis revealed a Hox cluster with a number of strange features. However, even considering these previous findings, Seo and coworkers results were unexpected, to say the least. The authors identified, cloned and phylogenetically classified full-length cDNAs for all O. dioicas nine Hox genes. Using in situ hybridization, they studied the expression patterns of these genes during early development. As for other chordates, expression patterns varied among tissues distributed along the anteroposterior axis, with a subset of the Hox gene complement being expressed in each tissue; but in this case, the Hox gene expression domains were mostly non-overlapping. However, the order of expression of these genes along the axis was generally correlated with the position in the Hox gene cluster of the paralogous genes in other chordates. So far, so good: the differences that the authors identified between the Hox gene complement and expression patterns in O. dioica compared with other chordates, although interesting and notable, were not particularly unusual considering the overall variation between bilateral lineages. The unexpected result came when the authors took the next obvious step and looked at the genomic organization of the O. dioica Hox genes. Seo and colleagues screened an O. dioica genomic BAC library with their nine Hox cDNA probes, no doubt expecting to be able to quickly locate the clone that contained the standard Hox cluster. Instead, they found nine separate BAC clones that contained individual Hox genes. Follow-up sequencing of the clones confirmed that none of the initial set of nine genes was located anywhere near the others, and indeed many unrelated genes surround each of these at the high density that is expected in this compact genome. The disintegration of the Hox cluster in this tunicate might be most surprising to vertebrate developmental biologists, as the integrity of this cluster has been clearly demonstrated to be essential for temporal coordination of Hox gene expression in the mouse. By contrast, partially fragmented Hox clusters have already been found in the fly, worm and C. intestinalis. The authors raise the intriguing possibility that in the tunicates C. intestinalis and O. dioica, it is the transition to determinative development, since divergence from their common ancestor with vertebrates, that has allowed the Hox gene family to be split apart: a hypothesis that is also consistent with worm data. In determinative development, the destiny of most cell lineages is engaged during the first division of the egg, so the authors postulate that the usual function of Hox genes after axis formation might have become superfluous, and with it the genomic clustering of these genes. Regardless of the underlying cause, these striking new results deal the most important blow so far to the pervasive and persistent idea that the Hox cluster is ultimately required to build a complex animal.
Nick Campbell, Nature Publishing Group References and links
ORIGINAL RESEARCH PAPER Seo, H.-C. et al. Hox cluster disintegration with persistent anteroposterior order of expression in Oikopleura dioica. Nature 431, 6771 (2004) WEB SITE

Daniel Chourrouts laboratory: http://www.sars.no

Louisa Flintoft

NATURE REVIEWS | GENETICS 2004 Nature Publishing Group

VOLUME 5 | NOVEMBER 2004 | 8 0 5

RESEARCH HIGHLIGHTS

IN THE NEWS
Dung DNA set to foil ivory poachers Elephant dung could be more valuable than ivory to the elephants at least. A test that compares DNA from illegal ivory with maps of genetic variation based on dung samples might hold the key to tracking down poachers. The ban on ivory trading has driven poachers into forests, where their activities are more difficult to detect. Tom Milliken of TRAFFIC, an organization that monitors trade in ivory, thinks the dung-based maps could help to pinpoint poaching hotspots: the largest uncertainty in our chain is where it is coming from, and this method will help with that (news@nature.com). The group that constructed the map, from the University of Washington in Seattle, took skin and dung samples from 16 African countries. the most important breakthrough is the ability to get it (DNA) from feces because we can sample many countries very quickly now, said Samuel Wasser, who led the study (Reuters.com). Because forest elephants live in isolated communities, genetic variation is sufficient to distinguish between animals from different areas. We have incredible precision at telling one forest location from another, Wasser explained (newscientist.com). But there is room for improvement, as the map isnt as good at distinguishing man-made boundaries: Right now, its probably not precise enough because it might not tell us if a consignment comes from one side of a national border or another, commented Julian Blanc of the World Conservation Union (The Guardian, UK, 28 September, 2004). However, increased accuracy should make the map a valuable tool in the future. But this requires more dung, and Wasser has a plea for those who patrol the areas affected: please, just ask them to pick up the poop, (news@nature.com).

AGEI NG

Yeast: a death foretold?

A new study adds fuel to the controversy over whether yeast cells undergo programmed cell death by suggesting that it occurs for the good of the species. Death spares no one, not even yeast cells. But the how and why of death in Saccharomyces cerevisiae is dividing the community. In a recent paper, Paola Fabrizio and colleagues propose that the death of older yeast cells occurs to allow younger cells to thrive in essence, therefore, they suggest that apoptosis occurs in yeast as an adaptive process that benefits other members of the group. This conclusion is controversial for two reasons first, because it presupposes that death is actually programmed in yeast and that it occurs by apoptosis, and second, because it invokes the maligned theory of group selection, which runs counter to the well-established idea that selection occurs at the level of the individual. Fabrizio and colleagues first addressed how cell death occurs. The ageing yeast cells they studied showed many features of mammalian apoptosis, including chromatin condensation and acidification of the cytosol. Also consistent with programmed ageing in yeast was the cellular pathway used to bring about death: the authors found that ageing and dying yeast cells downregulate an inhibitor of superoxide (SOD2, a superoxide dismutase) and become sensitive to superoxide, both of which mediate cell death in higher eukaryotes. Indeed, mutant yeast colonies in which SOD genes are upregulated live longer. To see whether this programmed death had an adaptive purpose, the authors assayed the ability of various

R NA SI LE NCI NG

Simple, but effective

An important question about silencing by small RNAs concerns why some target genes are regulated at the level of transcription, whereas others are regulated post-transcriptionally. Phillip Zamore and colleagues suggest that the simple system that operates in the yeast Schizosaccharomyces pombe could help to find the answer. Argonaute (Ago) proteins are crucial for the execution of gene silencing by small RNAs. Many species express multiple Ago proteins, and some of these induce silencing at the transcriptional level (transcriptional gene silencing (TGS)), whereas others specifically mediate posttranscriptional gene silencing (PTGS). Yeasts are poor relations in this respect; for example, S. pombe encodes a single Ago protein, Ago1, which functions in TGS. However, despite lacking other members of the Ago family, previous studies have hinted that PTGS also occurs in this species. To confirm this, the authors replaced the S. pombe adh1 gene with an adh1:GFP fusion. They then used a GFP hairpin RNA to induce silencing of the transgene. An intron was included in the hairpin construct, a strategy that had been shown to increase the

Louisa Flintoft

efficiency of PTGS in other studies. Expression of the hairpin decreased GFP fluorescence, indicating successful silencing of the fusion gene. To test whether this was mediated by TGS or PTGS, the authors measured both the total levels of adh1:GFP mRNA and the levels of its transcription. Overall, adh1:GFP mRNA levels decreased in cells expressing the hairpin, but the rate of transcription remained the same, indicating that silencing was post-transcriptional. As Ago1 is the only S. pombe Ago protein, is it responsible for PTGS as well as TGS? Silencing of adh1:GFP was abolished in Ago mutants, confirming that this is the case. Interestingly, the authors went on to show that proteins that interact with Ago1 to mediate TGS are not required for PTGS,

806

| NOVEMBER 2004 | VOLUME 5

2004 Nature Publishing Group

www.nature.com/reviews/genetics

RESEARCH HIGHLIGHTS

yeast strains of varying longevity to repopulate a yeast colony after a substantial proportion of the colony had died. This phenomenon, called adaptive regrowth, occurs normally in wild type strains however, the longerlived strains, which overexpress SOD enzymes, could not repopulate the colony in the long term, and eventually died out. So, although longevity might confer an immediate advantage, it is detrimental to the species as a whole. Short-lived strains that lack the superoxide inhibitors were even better than the wild type at adaptive regrowth; the authors show that the ability of superoxide to promote adaptive regrowth depends on its ability to release nutrients from dying cells into the growth media, which in turn would allow younger cells to thrive and reproduce. In addition, because superoxide induces DNA

mutations, its presence favours the selection and growth of mutants that are better adapted to the environment. The results were not peculiar to laboratory strains, as the same phenomena occurred in three strains newly collected from the wild. Computational simulations tell a similar story that a population that undergoes premature death and has a high mutation frequency is more likely to adapt to a changing environment. So in yeast, at least, apoptosis is an altruistic act, as not dying damages the chances of survival of the whole group. If the theory stands up to scrutiny then what consequences does it have for humans? Should we thwart any attempt to extend our lives for the sake of our species? Whatever the eventual answer, this is a debate that isnt being laid to rest.
Tanita Casci References and links
ORIGINAL RESEARCH PAPER Fabrizio, P. et al.

IN BRIEF
P L A N T D E V E LO P M E N T

The PLETHORA genes mediate patterning of the Arabidopsis root stem cell niche.
Aida, M. et al. Cell 119, 109120 (2004)

In Arabidopsis thaliana, root stem-cells are maintained by a small set of organizing cells, known as the quiescent centre (QC), the location of which depends on auxin accumulation. By using a promoter-trap screen, the authors identified two putative transcription factors, PLETHORA 1 (PLT1) and PLT2, which are required for QC specification and for maintaining root stem-cells during embryonic pattern formation; in addition, evidence indicates that their expression in the QC responds to auxin.
D E V E LO P M E N TA L B I O LO G Y

Foxa2 is required for transition to air breathing at birth.

Wan, H. et al. Proc. Natl Acad. Sci. USA 101, 1444914454 (2004)

Superoxide is a mediator of an altruistic aging program in Saccharomyces cerevisiae. J. Cell Biol. 166, 10551067 (2004) WEB SITE Longos laboratory: http://www.usc.edu/ programs/pibbs/site/faculty/longo_v.htm

A fundamental adaptation faced by a newborn mammal is the ability to breathe in air through its lungs. Now, by knocking out gene function in the epithelial cells of the developing mouse lung, Wan and colleagues show that Foxa2, which encodes a forkhead transcription factor, is a master gene required for lung maturation at birth. This finding could inform treatments for premature babies and for individuals with lung disease or injury.
D E V E LO P M E N TA L B I O LO G Y

Hmx2 and Hmx3 homeobox genes direct development of the murine inner ear and hypothalamus and can be functionally replaced by Drosophila Hmx.
Wang, W. et al. Dev. Cell 7, 439453 (2004)

suggesting that Ago1 functions as part of distinct complexes to mediate the two types of silencing. The fact that both TGS and PTGS are mediated by Ago1 in S. pombe indicates that it is not simply the availability of specialized Ago proteins that determines which pathway is used to silence specific genes. Gene-specific characteristics are also likely to be important, such as the chromosomal context of the gene or its rate of transcription. The simple system provided by silencing in S. pombe should be a useful tool for dissecting these requirements.
Louisa Flintoft References and links
ORIGINAL RESEARCH PAPER Sigova, A.,

The authors show that the roles of mouse homeobox genes Hmx2 and Hmx3 in the development of the vestibular system are overlapping and distinct, but that their roles in the central nervous system (CNS) are interchangeable. Moreover, the single fly Hmx can rescue the CNS and inner-ear phenotype in double-knockout mice, despite differences in morphology. The authors propose that evolution of complex organs such as the vertebrate inner ear might involve cooption of primitive genetic programmes to new locations, not just from acquisition and modification of protein domains.
GENE EXPRESSION

Genome-wide mRNA surveillance is coupled to mRNA export.

Hieronymus, H. et al. Genes Dev., 1 November 2004 (doi:10.1101/gad.1241204)

Rhind, N. & Zamore, P. D. A single Argonaute protein mediates both transcriptional and post-transcriptional silencing in Schizosaccharomyces pombe. Genes Dev. 18, 23592367 (2004)
WEB SITE

http://www.umassmed.edu/bmp/faculty/ zamore.cfm?start=0&

The authors found evidence to suggest that there are links between DNA and RNA surveillance and mRNA export. A screen of annotated, non-essential Saccharomyces cerevisiae genes identified new factors required for mRNA export, including Rrp6, an mRNA surveillance factor, and Lrp1, a DNA-repair protein. The authors found that Lrp1 can mediate mRNA degradation and requires Rrp6 for nuclear localization to the genes that encode their target mRNAs.

NATURE REVIEWS | GENETICS 2004 Nature Publishing Group

VOLUME 5 | NOVEMBER 2004 | 8 0 7

RESEARCH HIGHLIGHTS
How does wild-type tsc1 control the rate of differentiation? Mutations that activate the InR pathway cause precocious differentiation in the eye, as with mutations in tsc1. The converse experiment, in which the InR and Tor pathways were inactivated, led to delays in neuronal differentiation. This work further supports the involvement of tsc1 in InR and Tor pathways and, importantly, implicates these pathways in the control of developmental timing. The cells that make up each eye unit in the fly are recruited to their fate by reiterative signalling through the Ras/MAPK pathway; however, lack of tsc1 does not seem to affect this signalling, indicating that tsc1 acts downstream of known components of this pathway or in parallel to them. Uncoupling the execution of cellfate decisions from the time at which the decisions are made might allow

D E V E LO P M E N TA L B I O LO G Y

Keeping an eye on the tempo

Developmental signalling pathways defy simple job descriptions; partly because nature has effectively recycled them in many capacities, but largely because we are ignorant of just how varied their jobs are. This is illustrated in a recent paper by Joseph Bateman and Helen McNeill. By analysing mutations in the fly eye, they show that the insulin receptor (InR) and Tor pathways which have a well-conserved function in controlling cell and organ size have an unexpected role in determining when cells differentiate. This conclusion was derived from work on a gene, tsc1 (tuberous sclerosis complex 1), which the authors recovered in a genetic screen for fly mutants with defective planar polarity in the eye. As expected from the mutant phenotype of the vertebrate homologue of tsc1, which encodes a tumour suppressor, homozygous mutant photoreceptor cells are larger than normal, but otherwise differentiate normally. What was peculiar, however, was that several types of mutant photoreceptor cells which are studied precisely because their differentiation is so stereotypical differentiated more precociously than their genetically wild-type neighbours. This speededup development did not lead to an abnormal eye unit or to ectopic cell fates, but simply to an acceleration of the normal differentiation program.

S Y S T E M S B I O LO G Y

Rewiring the network

Although systems biology helps to make sense of the complex interactions between genes, proteins and other biologically relevant molecules, most studies have provided only snapshots of how these networks operate in specific conditions. A recent paper in Nature describes the first genome-scale study of how biological networks are rewired according to the needs of the cell, revealing some important insights into network dynamics. Luscombe and colleagues first defined a network of known interactions between 142 transcription factors and 3,420 genes in the yeast Saccharomyces cerevisieae. They then incorporated data from previous studies that had examined patterns of gene expression during the cell cycle, sporulation, diauxic shift (the switch from anaerobic to aerobic respiration), DNA damage and the stress response, and used an algorithm to identify sub-networks of interactions that are active in these different conditions. To characterize these sub-networks, the authors devised a statistical method SANDY (statistical analysis of network dynamics) for the analysis of interactions both within and between conditions. Overall comparisons grouped the five sub-networks into two categories. The exogenous diauxic shift, DNA-damage and stress-response subnetworks were characterized by the regulation of several genes by each transcription factor, and by shorter pathways leading from transcription factors to their target genes. This fits in with their ability to produce large and rapid responses to changes in the environment. By contrast, the endogenous cell-cycle and sporulation subnetworks allow more precise, multi-stage control, with longer pathways to activation and more inter-regulation between transcription factors. Although these results might not seem surprising, given the biological functions of the sub-networks, more unexpected patterns emerged when other characteristics were investigated. Static gene-regulatory networks are characterized by the existence of hubs individual transcription factors that regulate a disproportionately large number of genes. The importance of these hubs suggests that they are likely to function across a range of conditions, as they regulate key pathways, and this is supported by theoretical simulations. However, the

authors found the reverse to be true: most hubs (78%) were important in only a single set of conditions and were therefore dubbed transient hubs. Another surprising result was seen when the interactions made by those hubs that do function across several conditions, known as permanent hubs, were examined. Rather than using a similar set of interactions in each condition, these hubs redefined their interactions just as frequently as transient hubs further evidence that networks are more dynamic than was previously thought. As Luscombe and colleagues point out, their study was limited to results that were available from previous experiments, although the robustness of the features they describe in response to random noise suggests that similar patterns are likely to emerge from direct studies of S. cereviseae network dynamics. The increasing availability of genome-wide data on regulatory interactions in cell types should allow future studies to determine whether these features apply on a wider scale.
Louisa Flintoft References and links
ORIGINAL RESEARCH PAPER Luscombe, N. M. et al. Genomic analysis of regulatory network dynamics reveals large topological changes. Nature 431, 308312 (2004) FURTHER READING Barabsi, A. L. & Oltvai, Z. N. Network biology: understanding the cells functional organization. Nature Rev. Genet. 5, 101113 (2004) WEB SITE

Web supplement to original research paper: http://SANDY.TopNet.GersteinLab.org

808

| NOVEMBER 2004 | VOLUME 5

2004 Nature Publishing Group

www.nature.com/reviews/genetics

better control over the execution of the developmental program. For example, in the case of tsc1 the temporal control is linked to nutrient conditions through its connection with the insulin pathway when nutrients are scarce the organism could then coordinate a slow down in its development, in line with its reduced growth. The authors also showed that the temporal control function of InR/Tor pathways holds true for neuronal cell types outside the fly eye, but just how broadly it applies in flies and beyond, and precisely how the control is effected, is not yet known.
Tanita Casci References and links
ORIGINAL RESEARCH PAPER Bateman, J. M.

GENE EXPRESSION

The true purpose

Mendell and co-workers have uncovered the physiological function of the nonsense mediated decay (NMD) pathway in higher eukaryotes it is a crucial mechanism for post-transcriptional regulation, which is interlinked with essential homeostatic mechanisms. Whose line is it anyway a successful comedycum-game show has entertained audiences on both sides of the Atlantic. The contestants actors, actresses and other celebrities are asked to perform a series of tasks. In one, they are presented with an object and asked to indicate through acting as many uses for it as possible. The intended use of the object is not always clear, but in the game, this is beside the point. The real purpose or function of biological phenomena can be frustratingly elusive; this is often because the experimental conditions that are used are artificial. But unlike in the example above, uncovering the natural function is essential in biology. As its name suggests, NMD is a mechanism that removes mRNAs that carry nonsense mutations. But as Dietz and colleagues point out, this role alone could not account for the evolutionary conservation of the pathway it must, therefore, have another function. To uncover it, the authors knocked down the pathway in HeLa cells and, using microarray analysis, they compared transcription profiles of these cellswith those in which the pathway was intact. The results revealed that almost 5% of genes were upregulated the transcripts of these genes are normally eliminated by the NMD pathway. Among them are transcripts that harbour upstream open reading frames that lie in 5 UTRs, transcripts in which nonsense codons or frameshift mutations have been introduced by alternative splicing, those that contain introns in their 3 UTRs and transcripts that are derived from ancient transposons and endogenous retroviruses. A common feature of most of these transcripts is the presence of a spliced intron located at least 50 nucleotides downstream of the termination codon a feature that is sufficient to activate the NMD response. The authors noted that many of the NMDpathway substrates are involved in amino-acid metabolism and the cellular response to aminoacid starvation. This observation revealed an interesting homeostatic feedback mechanism. Amino-acid starvation inhibits translation, so as the authors say: Since NMD requires ongoing translation, it is likely that regulation of these transcripts by nonsense surveillance couples their expression level to translational efficiency. Thus, under conditions of amino-acid

& McNeill, H. Temporal control of differentiation by the Insulin receptor/Tor pathway in Drosophila. Cell 119, 8796 (2004) WEB SITE Helen McNeills laboratory: http://science.cancerresearchuk.org/research/loc/ london/lifch/mcneillh/

starvation, inhibition of translation and NMD would increase expression of transcripts that promote restoration of amino-acid homeostasis. This mechanism of preserving amino-acid homeostasis is evolutionarily conserved as revealed by the authors analysis of previously published data on NMD-regulated gene expression in yeast. The work of Mendell et al. has put the role of the NMD pathway in an interesting perspective. Its predominant physiological function seems to involve the regulation of many transcripts, whereas its role in human disease caused by nonsense mutation, although medically important, seems evolutionarily insignificant.
Magdalena Skipper References and links
ORIGINAL RESEARCH PAPER Mendell, J. T. et al. Nonsense surveillance regulates expression of diverse classes of mammalian transcripts and mutes genomic noise. Nature Genet. 36, 10731078 (2004) FURTHER READING He, F. et al. Genome-wide analysis of mRNAs regulated by the nonsense-mediated and 5 to 3 mRNA decay pathways in yeast. Mol. Cell 12, 14391452 (2003)

NATURE REVIEWS | GENETICS 2004 Nature Publishing Group

VOLUME 5 | NOVEMBER 2004 | 8 0 9

RESEARCH HIGHLIGHTS

IN BRIEF
EVOLUT ION ARY G E N ETICS

C O N S E R V AT I O N G E N E T I C S

Genes feel the heat

A recent study suggests that climate change affects wildlife at the genetic level, with implications for protecting biodiversity in the face of changing weather patterns. Hadly and colleagues studied historical changes in genetic variation in northern pocket gophers and montane voles. The two species thrive in wet climates, so population numbers of both are expected to decrease during periods of dry weather. But there are also differences between the two species: whereas pocket gophers are home-loving animals and tend not to venture far from the sub-populations they live in, montane voles are more adventurous, with more migration between groups. What impact might these ecological differences have on genetic variation in these species during periods of climate change? Over the past 2,500 years, two periods of climate change occurred the Medieval Warm Period (1,150650 years ago) and the Little Ice Age (65050 years ago). To investigate the effects of these periods on gopher and vole genetic diversity, the authors took advantage of the abundance of fossils for the two species that have been found in Yellowstone National Park, in the United States. They used mitochondrial DNA from these fossils to estimate effective population sizes and levels of genetic variation at different times over the past 2,500 years. In keeping with their preference for damper climes, both species underwent population decreases during the Medieval Warm Period. However, this similarity didnt hold up for genetic variation: whereas the gophers showed reduced genetic diversity during this period, the reverse was true for the voles. This fits in well with the ecological strategies of the two species. The high level of migration between vole sub-populations is expected to contribute to genetic diversity in this species, and a recent study has shown that migration increases at times of low population density. By contrast, as gophers live in closed populations, this means of maintaining variation would not have applied to them. Another explanation is that genetic variation increased in the voles due to stronger selective pressure, a possibility that needs further investigation. Whatever the case, this study indicates that climate change affects genetic variation, with varying effects on different species. As levels of genetic variation can contribute to the likelihood of extinction, further studies could provide important pointers as to which species are likely to be hit hardest by climate change.
Louisa Flintoft References and links
ORIGINAL RESEARCH PAPER Hadly, E. A. et al. Genetic response to climatic change: insights from ancient DNA and phylochronology. PloS Biol. 2, e290 (2004) WEB SITE Hadlys laboratory: http://www.stanford.edu/group/hadlylab/index.html

Birth and adaptive evolution of a hominoid gene that supports high neurotransmitter flux.
Burki, F. & Kaessmann, H. Nature Genet. 36, 10611063 (2004)

There are two glutamate dehydrogenase genes in humans: GLUD1 a housekeeping gene and GLUD2, which is expressed specifically in neural tissue and testis. The authors show that GLUD2 originated by retrotransposition from GLUD1 in our ancestor, 23 Mya. The changes that give GLUD2 its tissue specific properties are a result of positive selection, following a duplication event. GLUD2 probably contributed to enhanced brain function in humans and apes, and has also been implicated in late memory formation.
P O P U L AT I O N G E N E T I C S

Global patterns of human mitochondrial DNA and Y-chromosome structure are not influenced by higher migration rates of females versus males.
Wilder, J. A. et al. Nature Genet. 36, 11221125 (2004)

Genetic evidence supports demic diffusion of Han culture.

Wen, B. et al. Nature 431, 302305 (2004)

These studies analyse the effects of historical migration on human population structure. Wen et al. examined patterns of Y-chromosome and mitochondrial DNA (mtDNA) variation to study the spread of the Han Chinese culture. Their results indicate that this spread followed migration of the Han people, rather than diffusion of the culture through social interchange without genetic mixing, and show that males had a greater role than females in this expansion. Wilder et al. tested the theory that population structures have been more strongly influenced by female migration than that of males due to patrilocality, which occurs when females move to the locality of their spouses following marriage. Analysis of genetic variation between ten populations from different global regions showed that this is not the case, at least at the continental and global level, as similar levels of variation for Y-chromosomes and mtDNA indicate roughly equal contributions of male and female migration.
EVOLUT ION ARY G E N ETICS

Regulatory evolution across the protein interaction network.

Lemos, B. et al. Nature Genet. 36, 10591060 (2004)

These authors showed that, for a specific gene, the number of interactions that its protein product participates in is negatively correlated with the level of variation in gene expression, both within and between species. Furthermore, for pairs of interacting genes, levels of variation in gene expression were more similar than for randomly assigned pairs. These results indicate that proteinprotein interactions might have an important role in constraining evolutionary changes in gene regulation.

810

| NOVEMBER 2004 | VOLUME 5

2004 Nature Publishing Group

www.nature.com/reviews/genetics