566 | AUGUST 2001 | VOLUME 2 www.nature.com/reviews/genetics

H I G H L I G H T S

Facing up to antisense

G E N E E X P R E S S I O N

Gene expression is primarilyregulated at the levels oftranscription, RNA splicing ortranslation. But evidence isaccumulating for another way toregulate gene expression through the interaction ofendogenous antisense RNA withthe corresponding sensetranscript. In a recent issue ofPNAS, Blin-Wakkach et al.present evidence that Msx1 akey transcription factor involvedin craniofacial skeleton formation is regulated by an endogenousantisense transcript, and that thismode of regulation is conservedbetween rodents and humans.

In tissues such as cartilage andmuscle, the downregulation ofMsx1 expression duringdevelopment is associated with atransition from an undifferentiated,proliferative state to a terminallydifferentiated one. When Blin-Wakkach and colleaguesgenerated Msx1-deficient mice bymeans of a lacZ insertion, theynoticed that lacZ was notexpressed in mice that wereheterozygous for the insertion,even though the transcript wasdetectable by PCR. To testwhether the lack of lacZexpression was due to regulationby an antisense RNA, the authorsidentified, by Northern blotanalysis, an endogenoustranscript the sequence of whichwas complementary to part of theMsx1 transcript and had noobvious coding potential. Theantisense transcript was ~2.2 kband overlapped with at least oneMsx1 intron–exon boundary. Theauthors also mapped thetranscription initiation site of theantisense transcript andidentified a 66-bp putativepromoter region that is conservedin five mammalian species.

The antisense RNA was muchmore abundant than the senseRNA and its transcriptioncoincided with terminaldifferentiation in the craniofacialregion. Simultaneousdisappearance of the Msx1protein indicates that Msx1expression might be regulated bythe ratio of its sense and antisensetranscripts.

Interaction between the senseand antisense transcripts of Msx1seems to be important in thetransition from proliferation toterminal differentiation duringcraniofacial development.Although the exact mechanism isunclear, the authors propose thatsense–antisense interactionsprevent the correct splicing of theMsx1 mRNA. More generally, thisstudy highlights an importantadditional level of generegulation.

Magdalena Skipper

References and linksORIGINAL RESEARCH PAPER Blin-Wakkach, C.et al. Endogenous Msx1 antisense transcript: invivo and in vitro evidences, structure, andpotential involvement in skeleton developmentin mammals. Proc. Natl Acad. Sci. USA 98,7336–7341 (2001)FURTHER READING Kumar, M. & Carmichael, G.Antisense RNA: function and fate of duplex RNAin cells of higher eukaryotes. Microbiol. Mol.Biol. Rev. 62, 1415–1434 (1998)

IN BRIEF

Evolutionary EST analysis identifies rapidly evolvingmale reproductive proteins in Drosophila.Swanson, W. J. et al. Proc. Natl Acad. Sci. USA 98, 7375–7379 (2001)

Reproductive proteins are likely candidates for causing the rapid divergence between closely related species,as their modification favours adaptive divergence. To test this hypothesis, Swanson et al. compared the sequences of 285 ESTs (expressed sequence tags) isolated from the male accessory glands — which secrete seminal-fluid proteins — of Drosophila simulans and D. melanogaster.11% of the 176 genes identified showed evidence ofpositive selection, and the 57 newly identified accessory gland genes should become candidate genes for furtherpositive selection studies.

Mitotic recombination effects homozygosity for NF1germline mutations in neurofibromas.Serra, E. et al. Nature Genet. 28, 294–296 (2001)

Neurofibromatosis type 1 (NF1) is an autosomal-dominantdisorder caused by mutations in the NF1 tumour suppressorgene. Of the many cell types seen in NF1 nerve-sheath tumours,only Schwann cells are homozygous for NF1 mutations.Through molecular and in situ hybridization of culturedSchwann cells, this paper shows that loss of heterozygosity forNF1 occurs by mitotic recombination. As there is inter-individual variability in mitotic recombination rates, genes thatcontrol this phenomenon might act as modifier genes forsusceptibility to NF1 and other cancers.

Charcot–Marie–Tooth disease type 2A caused bymutation in a microtubule motor K1F1Bβ.Zhao, C. et al. Cell 105, 587–597 (2001).

CMT2A is a dominant subtype of Charcot–Marie–Toothdisease (CMT) — the most commonly inherited humanperipheral neuropathy — that maps to 1p35–36. When Zhao et al. knocked out the mouse β isoform of K1f1B —which encodes a motor protein belonging to the kinesinsuperfamily — null mutants died at birth with multipleneurological abnormalities, and heterozygotes developedprogressive muscle weakness and defective synaptic vesicletransport. This CMT-like phenotype and the chromosome-1map location of K1f1B lead the authors to analyse the gene in CMT2A patients, in whom they found loss-of-functionmutations in the motor domain of K1F1B. As axons lackprotein-synthesis machinery, their survival depends on thetransport of essential proteins. Haploinsufficiency for K1f1B might therefore reduce levels of synaptic vesicletransport of essential factors and receptors, leading to theimpairment of nerve endings and to the decreased survival of peripheral neurons.

H U M A N G E N E T I C S

C A N C E R G E N E T I C S

E V O L U T I O N

© 2001 Macmillan Magazines Ltd