1999 sanders - art no sex please we are fungi

7/30/2019 1999 Sanders - ART No Sex Please We Are Fungi

1/31999 Macmillan Magazines Ltd

According to evolutionary theory, theadvantage of sex is that recombina-tion shuffles together new combina-

tions of genes, thereby producing geneticvariation and allowing deleterious muta-tions to be purged13. But some extremelysuccessful organisms are both asexual andancient4: the very existence of such scan-dalous asexuals5 flies in the face of theory.

Among them are counted the arbuscularmycorrhizal fungi, the Glomales, studies ofthe genome structure and organization ofwhich now reveal some remarkablephenomena. The papers concerned appearin Fungal Genetics and Biology6 and Gene7.Most notably, they identify a striking degreeof divergence in the ribosomal sequences ofindividuals among the Glomales.

As the boundaries continue to blurbetween p53 and its relatives, compelling dif-ferences remain. For example, despite theirsimilarities, the three genes seem to have verydifferent functions. Whereas deletion of

p6315,16 andp73 (F. McKeon, personal com-munication) has dramatic developmentalconsequences in mice,p53 null mice developnormally1 (with some interesting excep-

tions). But p53 null mice are highly prone totumours. That p53 is unique in serving as atumour suppressor is supported by the factthat, in human cancers, loss or mutation of

p73orp63seems to be infrequent2. Addition-ally, only p53 is susceptible to inactivation bythe SV40 T antigen, the adenovirus E1B 55Kprotein and the human papillomavirus E6protein. Finally, although Mdm2 binds top53 and p73 (Fig. 1), only p53 is degraded as aresult of this interaction2.

Why might p53 alone be a tumour sup-pressor? There are several possible answers,all supported by published reports2. First, thetissue distribution of p63 and p73 is more

restricted than that of p53. Second, thedownstream targets of p53 and its relativesmay differ, and perhaps p53 is more effectivein inducing key apoptosis-related targetsunder some circumstances. Third, there maybe a broader range of upstream regulatorsthat can signal to p53. And fourth, interac-tions between some forms of p53 and its rela-tives have been documented. For instance,tumour-derived p53 mutants can represstransactivation and apoptosis induced byp73, and amino-terminally-truncated formsof p63 can repress wild-type p53 (ref. 2). So,cross-talk between p53 and its relatives may

contribute to their different roles in tumori-genesis.It has been suggested that p53 is the most

recently evolved member of this family. Wemight speculate that, with the developmentof more complex multicellular organisms, aneed arose for a more versatile stress-response factor one that can respond toand transmit a more diverse set of signals to amore complex set of targets. Whatever theexplanation for the differences among p53family members, there is much work to bedone to find out how these genes regulateimportant processes in cells, and why onlyp53 suppresses tumour formation.

Eileen White is at the Howard Hughes MedicalInstitute, Center for Advanced Biotechnology and

Medicine, Department of Molecular Biology and

Biochemistry, Cancer Institute of New Jersey, and

Rutgers University, 679 Hoes Lane, Piscataway, New

Jersey 08854, USA.

e-mail: [email protected]

Carol Prives is in the Department of Biological

Sciences, Columbia University, Sherman Fairchild

Center for the Life Sciences, 816 Fairchild Building,

New York, New York 10027, USA.


1. Ko, L. J. & Prives, C. Genes Dev. 10, 10541072 (1996).

2. Kaelin, W. G. Jr Oncogene(in the press).

3. Gong, J.et al. Nature399, 806809 (1999).

4. Agami, R., Blandino, G., Oren, M. & Shaul, Y.Nature399,

809813 (1999).

5. Yuan, Z.-M. et al. Nature399, 814817 (1999).

6. Van Etten, R. A. Trends Cell Biol. 9, 179186 (1999).

7. Kharbanda, S. et al. Nature376, 785788 (1995).

8. Rotman, G. & Shiloh, Y. Hum. Mol. Genet. 7, 15551563

(1998).

9. Baskaran, R. et al. Nature387, 516519 (1997).

10.Shafman, T. et al. Nature387, 520523 (1997).

11.Yuan, Z.-M. et al. Nature382, 272274 (1996).

12.Banin, S. et al. Science281, 16741677 (1998).

13.Canman, C. E. et al. Science281, 16771679 (1998).

14.Khanna, K. K. et al. Nature Genet. 20, 398400 (1998).

15.Mills, A. A. et al. Nature398, 708713 (1999).

16.Yang, A. et al. Nature398, 714718 (1999).

news and views

NATURE | VOL 399 | 24 JUNE 1999 | www.nature.com 737

Evolutionary genetics

No sex please, were fungiIan R. Sanders

Nuclear suspensions of singlespores diluted down to one

nucleus per tube foranalysis with PCR

PCR performed on single nuclei confirm that individual sporescontain a population of genetically divergent nuclei

Entrophospora

Acaulospora

Gigaspora

Scutellospora

Scutellospora castaneaGlomus

Endogone pisiformis(non-mycorrhizal)

x ya

b

Figure 1 Genetics of asexual mycorrhizal fungi of the order Glomales.a, Hijri et al.6 demonstrate that

different spores ofScutellospora castanea(xandy) contain different sequences of ribosomal DNA,

but that each spore does not have the same complement of sequences. Further, they show that rDNA

differs between nuclei from the same fungal individual, which may be because lack of recombination

has allowed the multiple copies of rDNA to diverge. b, Hosnyet al.7 find that different sequences of

the 18S gene from an isolate ofS. castaneagroup into two different Glomales genera, Scutellospora

and Glomus. The Glomales consists of three different families: Glomaceae (red), Gigasporaceae (light

blue) and Acaulosporaceae (green), and the first two are thought to have diverged 353367 million

years ago15. The discovery of rDNA sequences in S. castaneathat match genera in two different

families suggests that genetic diversity in these fungi may have been increased by acquisition of DNA

from their ancestors.



These fungi form symbioses mycor-rhizas with plant roots which help plantsto acquire mineral nutrients from the soiland also determine plant biodiversity andecosystem function8. They are truly ancient,having remained largely morphologicallyunchanged since plants first colonized theland around 400 million years ago9. Nosexual stage in the Glomales life cycle has

been observed.Polymorphism in the ribosomal DNAencoding the internal transcribed spacer(ITS), and the 5.8S and 18S subunits, occursinside individual spores of fungi of the genusGlomus10,11. This is unusual because it is wide-ly thought that sequences of multiple copiesof ribosomal DNA are kept the same by aprocess known as concerted evolution12, inwhich the joint evolution of two or morerelated genes occurs as if they constitute a sin-gle locus. It is on the assumption of concertedevolution that the ribosomal sequences are sowidely used for taxonomy and phylogenet-ics13. The existence of different sequences of

ribosomal genes in Glomus implies thatrecombination does not occur, because themechanisms that keep the sequences ofrDNA copies the same operate most fre-quently during recombination events.

Both of the new studies6,7 looked at

another fungus within the Glomales, Scutel-lospora castanea, and worked with exactly thesame isolate14 which was maintained in cul-ture in the roots of plants. Using the poly-merase chain reaction (PCR), six differentITS sequences and 13 of the 18S gene werepicked out, respectively, from genomic DNAand from clones in an S. castaneagenomicDNA library. This does not mean that varia-

tion is limited to six different sequences ofeach. Hijri et al.6 confirmed that several dif-ferent ITS sequences (including those of the5.8S gene) occur in a single spore, but thateach spore of this isolate does not have thesame complement of the different sequences(Fig. 1a).

The Glomales are coenocytes that is,many nuclei are enclosed within one cell wall and it could be that different rDNAsequences occur on different nuclei in whichthe rDNA sequences have diverged due to alack of recombination. To test this possibility,Hijri et al.6 diluted nuclear suspensions fromsingle spores to an expected one-nucleus-

per-sample and performed PCR with specificprimers that amplify the different rDNAsequences. Their results indicate that, indeed,different rDNA sequences are carried ondifferent nuclei (Fig. 1a); so it seems that anindividual of these fungi is, in genetic terms,

actually a population of discrete nuclei.Perhaps more astounding, however, is

just how divergent the ribosomal genesequences are. Hosny et al.7 carried out aphylogenetic analysis of 18S sequences fromS. castaneaand from species of each of theother Glomales genera. Most sequencesgrouped in the genus Scutellospora, but twoothers from the same S. castaneaisolate were

so divergent that they clustered in the genusGlomus(Fig. 1b).A contaminant in the genomic library,

perhaps? It seems not, because the Glomus-like rDNA sequences were successfullyamplified again from single spores of thisisolate ofS. castanea. According to previousphylogenetic analyses of the Glomales, thefamily known as Glomaceae, which containsthe genus Glomus, diverged from the othermycorrhizal fungi that were later to form theAcaulosporaceae and the Gigasporaceae(containing the genus Scutellospora),between 353 and 367 million years ago15.

How then can we explain the observa-

tions reported in these two studies? Theresults support the idea that genetic drift the random change of allelic frequencies in apopulation has occurred in the absence ofrecombination, leading to genetically diver-gent nuclei. But at the same time it seems

news and views

738 NATURE | VOL 399 | 24 JUNE 1999 | www.nature.com

Almost exactly 85 years ago, Sir ErnestShackleton set out to cross the Antarctic

continent. After his ship was frozen intothe ice pack, the expedition waited and

worked through the long polar nights,then the long polar days, beforeShackleton managed to get them all

rescued, a story that continues to generatepopular books1. In honour of his exploits,Shackleton had a lunar crater (just to the

right of centre in this image of the lunarsouth pole) named after him. That crater

is now getting attention as a part ofanother story that involves ice, long daysand nights and perhaps, ultimately,

human exploration.Lunar scientists have long known that

because the Moons axis of rotation is

almost perpendicular to its path around

the Sun, the polar regions could have highpoints that are permanently sunlit, and

crater floors that are permanently inshadow. This false-colour image (inside the

yellow line, and overlaid on a radar imageof the region) shows the percentage of thelunar day for which a given location is

illuminated (white, orange and red receivethe most illumination, the clear regionsreceive none), based on a newly reported

analysis of images taken in 1994 by theClementine spacecraft2. A separate radar

study3 has confirmed that many regions,

including the bottom of Shackleton Crater,

are likely to be permanently shadowed.Conversely, some regions are sunlit most ofthe time and would be prime locations forlunar bases, or at least for solar arrays to

support a polar base. For example, thewhite region at the left (poleward) edge ofShackleton Crater is sunlit more than 80%

of the time, and there is a ridge 10 km awaythat is sunlit 90% of the time that the firstspot is not.

A polar base would be more attractive ifthe permanently shadowed crater floors

have managed to cold-trap water ice, which

could be mined for life support or fuel.Data from the neutron spectrometer

aboard the Lunar Prospector spacecraft4

apparently showed abundant hydrogen at

the poles, possibly, though not definitely,in the form of water ice; similarly, aClementine radar experiment may5 or may

not6 have provided evidence for ice.At the end of July, with its funding

expired and its fuel nearly gone, Lunar

Prospector will be crashed into a craternear the pole (the larger crater above andto the right of Shackleton) and Earth-

bound telescopes will search for evidenceof water released by the impact7. The

chances of the experiment detecting waterare no better than 10%, but this is just thenext step in a search for the conditions that

might make exploration of the lunar poles

possible. Timothy D. SwindleTimothy D. Swindle is in the Lunar and Planetary

Laboratory, University of Arizona, Tucson,

Arizona 85721-0092, USA.


1.Alexander, C. The Endurance: Shackletons Legendary Antarctic

Expedition(Knopf, New York, 1998).

2.Bussey, D. B. J., Spudis, P. D. & Robinson, M. Geophys. Res.

Lett. 26, 11871190 (1999).

3.Margot, J. L.et al. Science284, 16581660 (1999).

4.Feldman, W. C. et al. Science281,14961500 (1998).

5.Nozette, S.et al. Science274, 14951498 (1996).

6.Simpson, R. A. & Tyler, G. L. J. Geophys. Res. 104, 38453862

(1999).

7.Goldstein, D. B. et al. Geophys. Res. Lett. 26, 16521656 (1999).

Lunar exploration

Polar endeavours



unlikely that the nuclei have also harbouredrDNA sequences that have then remainedconserved for 400 million years. Probably themost likely way to explain their occurrence ishyphal anastomosis (cross-connection), andthe exchange of nuclei. Clearly, more infor-mation is now needed on DNA polymor-phisms within and among spores of Gloma-les species from regions of DNA other than

the multi-copy ribosomal genes. One possi-bility is that infrequent recombinationoccurs among nuclei. Using such sequencesas markers in population-genetic modelswould enable a test of whether nuclei from asingle spore form a recombinant population.

Meantime we are left with the distinctpossibility that individuals in the Glomalescontain a population of highly divergentnuclei that are subject to the accumulation ofmutations and additional genetic materialfrom distant sources. How do these fungicope with the accumulation of deleteriousmutations? How can they perform and regu-late their cellular, developmental and meta-

bolic processes? Phylogeneticists are facedwith equally fundamental questions. Isconcerted evolution in fact in operation?Is it valid to apply phylogenetic techniquesto these organisms at all?

The evolution of genomes in the absenceof recombination is not yet understood. Butperhaps the way that these asexual fungi have

existed for 400 million years is in some wayconnected to the maintenance of high geneticdiversity in their populations and their abilityto have exchanged genetic material with theirancestors. What is clear is that evolutionarybiologists and fungal geneticists are facedwith a happily disconcerting puzzle.Ian R. Sanders is at the Botanisches Institut,

Universitt Basel, Hebelstrasse 1, 4056 Basel,

Switzerland.email: [email protected]

1. Stearns, S. C. (ed.) The Evolution of Sex and its Consequences

(Birkhuser, Basel, 1987).

2. Fain, H. D. Evol. Theory11, 1529 (1995).

3. Lynch, M., Butcher, R. B. D. & Gabriel, W.J. Hered. 84, 339344

(1993).

4. Judson, O. P. & Normark, B. B. Trends Ecol. Evol. 11, 4146

(1996).

5. Maynard-Smith, J. Nature324, 300301 (1986).

6. Hijri, M., Hosny, M., van Tuinen, D. & Dulieu, H. Fungal Genet.

Biol. 26, 141151 (1999).

7. Hosny, M., Hijri, M., Passerieux, E. & Dulieu, H. Gene226,

6171 (1999).

8. van der Heijden, M. G. A. et al. Nature396, 6972 (1998).

9. Remy, W., Taylor, T. N., Hass, H. & Herp, H. Proc. Natl Acad.

Sci. USA91, 1184111843 (1995).

10.Sanders, I. R., Alt, M., Groppe, K., Boller, T. & Wiemken, A.

New Phytol. 130, 419427 (1995).11.Lloyd-McGilp, S. A. et al. New Phytol. 133, 103111 (1996).

12.Hoelzel, A. R. & Dover, G. A. Molecular Genetic Ecology

(IRL, Oxford, 1991).

13.Winker, S. & Woese, C. R. System. Appl. Microbiol. 14, 305310

(1991).

14.Banque Europenne des Glomales (BEG)

http://wwwbio.ukc.ac.uk/beg/

15.Simon, L., Bousquet, J., Lvesque, R. C. & Lalonde, M. Nature

363, 6769 (1993).

news and views

NATURE | VOL 399 | 24 JUNE 1999 | www.nature.com 739

Alzheimers disease

Pinning down phosphorylated tauMichel Goedert

Adefining characteristic of severalneurodegenerative diseases, includ-ing Alzheimers disease and Picks

disease, is the formation of filamentousdeposits of a microtubule-associated proteincalled tau in an abnormally hyperphospho-rylated form. The discovery of mutations inthe taugene in a condition known as familialfrontotemporal dementia and parkinsonismlinked to chromosome 17 has renewedinterest in the mechanisms by which dys-function of tau causes neurodegeneration.

On page 784 of this issue, Lu et al .1

describe an intriguing interaction between

phosphorylated tau and a prolyl isomerase,Pin1. Prolyl isomerases enhance the rate ofcisto transisomerization of the peptide bondon the amino-terminal side of proline. Pin1is an essential nuclear protein belonging tothe parvulin family of prolyl isomerases2.This group is distinct from two other prolylisomerase families, the cyclophilins and theFK506-binding proteins, which are targets ofthe immunosuppressive drugs cyclosporineand FK506, respectively. Pin1 consists of acarboxy-terminal catalytic domain, as wellas an amino-terminal proteinproteininteraction region called a WW domain

that specifically recognizes phosphorylatedserine or threonine residues preceding aproline residue (the S/TP motif)2,3.

In its short history, Pin1 has generatedmuch interest because it regulates entry andprogression through mitosis. It does this byinteracting with a large set of mitosis-specificphosphoproteins, most of which can bedetected by a monoclonal antibody calledMPM2 (ref. 3). Lu et al .1 started off armedwith the knowledge that MPM2 also recog-nizes hyperphosphorylated tau in the brainsof people with Alzheimers disease 4,5. More-over, during mitosis, tau is phosphorylated

at a number of the S/TP sites that are hyper-phosphorylated in Alzheimers6,7. Thisprompted the authors to examine whetherphosphorylated tau interacts with Pin1. Andthey found that tau (phosphorylated eitherby a mitotic cell extract or by Cdc2 kinase)does, indeed, bind to the WW domain ofPin1.

The longest isoform of tau in the humanbrain has 17 S/TP sites. Of these, Lu and col-leagues found that only one phosphory-lated threonine 231 (T231) was requiredfor the interaction with Pin1. This residue islocated upstream of the microtubule-

binding repeats in a proline-rich region thatis required for full activity of tau. TheT231 residue is hyperphosphorylated inAlzheimers disease and is also phosphory-lated, to a certain extent, in the normalbrain8,9. This residue can also be phosphory-lated by glycogen synthase kinase-3, butonly after phosphorylation of serine 235 bycyclin-dependent kinase-5 or mitogen-acti-

vated protein kinase

10,11

.Lu and colleagues went on to gather moreevidence for the interaction between tau andPin1. First, using a pull-down assay, theyshowed that Pin1 binds to hyperphosphory-lated tau from the brains of people withAlzheimers disease, but not to tau from age-matched healthy brains. Tau from normalbrain is notorious for the speed with which itis dephosphorylated after death12, so it maybe premature to conclude that Pin1 interactsonly with hyperphosphorylated (pathologi-cal) tau.

Second, by immunoblotting, the authorsdetected endogenous Pin1 in the paired heli-

cal filaments (PHFs) from diseased brains.(The PHFs, which are composed of hyper-phosphorylated tau, make up the pathologi-cal neurofibrillary tangles of Alzheimersdisease.) Third, using immunohistochem-istry, Lu etal. found that recombinant Pin1binds to pathological tau. Finally, theauthors looked at localization of Pin1. Incontrol brains they observed nuclear stain-ing for endogenous Pin1, consistent with itsknown localization in non-neuronal cells.But in the brains of people with Alzheimersdisease, Pin1 staining was also associatedwith pathological tau in neuronal cells

(although it is not clear what percentage ofthe tau-positive structures was alsoimmunoreactive for Pin1).

The tau protein in PHFs from the brainsof patients with Alzheimers disease is phos-phorylated at more than 20 residues, many(but not all) of which are S/TP sites13. Inhealthy brains, tau is heterogeneously phos-phorylated at between eight and ten of theseresidues9. Because of its abnormal hyper-phosphorylation, tau from PHFs cannotbind to microtubules or promote micro-tubule assembly. Hyperphosphorylation oftau is believed to be an early event that pre-cedes assembly into PHFs. Yet there is no

experimental evidence linking hyperphos-phorylation of tau to PHF assembly14 syn-thetic, paired helical-like filaments can, infact, be produced in a phosphorylation-independent way.

To examine the functional effects of theinteraction between Pin1 (Fig. 1, overleaf)and tau phosphorylated at T231, Lu et al .used recombinant tau phosphorylated byCdc2 kinase, with or without added Pin1. Asexpected, phosphorylated tau could neitherbind microtubules nor promote micro-tubule assembly properly. But in the pres-ence of Pin1, biological activity of the phos-

1999 sanders - art no sex please we are fungi

Documents