Nature © Macmillan Publishers Ltd 1997

correspondence

NATURE | VOL 389 | 9 OCTOBER 1997 537

Sir — The News item on NASA’s Bionprogramme of flying restrained and heavilyinstrumented rhesus macaques into spaceportrayed it in a light that the programmedoes not deserve (Nature 387, 4; 1997). Infact, Bion’s design is such that meaningfulresults about “how musculoskeletal andregulatory mechanisms respond to andrecover from space flight” (Bion 11/12Integrated Science Proposal and DisciplineProposals, 1996) are virtually precluded.

Both restraint and microgravity result inosteopenia and disuse atrophy of muscles,which makes it impossible to determine thespecific impact of microgravity (or anyother cosmic factor) on bones and musclesin the restrained animals sent into space.

The ‘regulatory physiology’ protocolcontains no hypotheses, which, inconjunction with the interference fromrestraint and psychological stress, thesample size of two, and dramatic individualdifferences in response (as documented byprevious missions), reduces Bion to‘parametric tinkering’, generating a spiral of

inconclusive animal studies that are onlyvaguely, if at all, applicable to free-moving,relatively comfortable astronauts.

To be most useful for future missions,experimental data of this nature should beobtained from humans. The head-down tiltbed rest model simulates the physiologicaleffects of microgravity and accurate non-invasive techniques, and osteoblast culturesare now available to study the cellular andbiochemical mechanisms of osteopenia. Allthe musculoskeletal tests and eight out ofnine ‘regulatory physiology’ tests couldhave been done in consenting humans.Eight measurements taken from astronautsare far more valuable than nine from strait-jacketed monkeys in pain and distress.

We believe that sending animals intospace is one of the most ineffective ways ofstudying the effects of microgravity onastronauts, and that subjecting consciousmammals to extreme trauma in the name oftechnological progress is unethical.

Bion involved some of the most invasiveinstrumentation ever imposed on a

conscious primate. In addition to extensivewiring of the body and 10 incisions(including a perforation of abdominalmuscles), each of the two flight and 10control macaques had 11 holes (eight forEvarts crown, three for epidural cannulae)drilled in the skull roof and two depressionsand canals in the zygomatic bones for anelectro-oculogram. As well as the pain andterror of being strait-jacketed in thesecircumstances, the flight animals wereexposed to motion sickness and shock fromthe dramatic concomitance of launch andlanding. After returning to Earth, onemacaque choked to death under anaesthesia.David O. Wiebers Andrzej ElzanowskiPaul W. GikasJennifer LeaningRoger D. White Scientific Advisory Council to The Humane Society of the United States,2100 L Street NW,Washington, DC 20037, USA e-mail: elzanow@ix.netcom.com

Monkey business in space

Bet you didn’t know thatSir — Contrary to Richard Wassersug’sdeclaration in News and Views (Nature 388,827–828; 1997), the penis of the nine-banded armadillo is not the first biologicalstructure known to have an investment oforthogonal fibres instead of the helicalarrangement classical in a hydrostat.

The sac of dura mater containing thehuman cerebrospinal fluid and spinal cordhas an orthogonal array of elastin fibres thatwas demonstrated by electron microscopyeight years ago (B. R. Fink & S. Walker,Anesth. Analg. 69, 768–772; 1989).

In the spinal dura the orthogonalpattern apparently helps to accommodatethe increase in longitudinal tensionoccasioned by flexion of the vertebralcolumn. The orthogonal investment of thepenis, however, is composed of collagenfibres which, as Wassersug pointed out,provide the penis with maximum flexuralstiffness when erect.Raymond Fink Department of Anesthesiology, University of Washington, Seattle, Washington 98195, USA

Definition of a geneSir — The term ‘gene’ is widely used bybiologists and is also in the publicvocabulary, yet it has no precise universallyaccepted molecular definition.

Some authors consider that a geneincludes the regulatory sequences requiredfor its expression1. But genes do not have tobe expressed to be present. The somaticcells of a multicellular organism all have thesame genes, but particular cell types expressonly some of them.

Inclusion of regulatory sequencesexpands the term ‘gene’ from a specificationof ‘what it is’ to indicate also ‘how it is used’.It may be useful to separate these twoconcepts. It is thought that majorevolutionary changes in the morphology oforganisms have resulted from alteredpatterns of synthesis of the same geneproducts2. It seems reasonable to considerthat the genes themselves have not changed,but rather the ways in which they have beenused.

Inclusion of regulatory sequencesintroduces considerable complexity to theterm ‘gene’. There are many different typesof regulatory elements, and they generallyoperate in complex combinations. Some ofthem, such as enhancers, are nonspecificand influence any compatible promoterwithin their range.

Others, such as the promoters ofpolycistronic mRNAs, control the synthesisof several gene products. Furthermore,there is already an acceptable term(‘operon’) for a unit of gene expression(page 339 of ref. 3).

Should introns be considered as parts ofgenes? When first discovered, genes withintrons were described as ‘interrupted’. This

implies that a gene consists only of itsexons; otherwise it would be continuous.An ‘exons only’ definition highlights theevolutionary theory that new genes can becreated by shuffling existing exons into newcombinations. There is already anacceptable term (‘transcription unit’) forthe entire region, including introns,between a promoter and a transcriptionalterminator(page 366 of ref. 4). In cases ofalternative splicing of primary transcripts,it has been suggested that, unless theprotein products are very different, onegene can encode a series of protein isoforms(page 457 of ref. 4).

In my opinion, the single best moleculardefinition of the term ‘gene’ is thefollowing: it is the nucleotide sequence thatstores the information which specifies theorder of the monomers in a final functionalpolypeptide or RNA molecule, or set ofclosely related isoforms. This definition issimple and concise. Geneticists can readilyfind other names for more complexgenetic entities such as ‘operons’ and‘transcription units’.Christopher D. EppCollege of Sciences,Massey University,Palmerston North,New Zealand

1. Lodish, H. et al. J. Mol. Cell Biol., 3rd edn, 308–310 (Scientific

American Books, 1995).

2. Pennisi, E. & Roush, W. Science 277, 34–37 (1997).

3. Lewin, B. Genes VI (Oxford Univ. Press,1997).

4. Alberts, B. et al. Molecular Biology of the Cell, 3rd edn (Garland

Publishing, 1994).