Deciphering the swordtail’stale: a molecular andevolutionary questAdam S. Wilkins

SummaryThe power of sexual selection to influence the evolutionof morphological traits was first proposed more than130 years ago by Darwin. Though long a controversialidea, it has been documented in recent decades for a hostof animal species. Yet few of the established sexuallyselected features have been explored at the level of theirgenetic or molecular foundations. In a recent report,Zauner et al.(1) describe some of the molecular featuresassociated with one of the best characterized of sexuallyselected traits, the male-specific tail ‘‘sword’’ seen incertain species of the fish genus Xiphophorus. Zauneret al. find that the msxC gene, a gene previouslyimplicated in fin development from work in zebrafish, isdramatically and specifically upregulated in the deve-lopment of the ventral caudal fin rays, which give riseto the sword, in males. The results provide the firstmolecular insight into the development of this sexuallyselected trait while prompting new questions about thestructure of the entire genetic network that underliesthis trait. To fully understand the molecular-genetic andevolutionary history of this network, however, it will beessential to determine whether sword-development is abasal or derived trait in Xiphophorus. BioEssays26:116–119, 2004. � 2004 Wiley Periodicals, Inc.

Introduction

The concept of sexual selection, the idea that morphological

traits can evolve in response to mating preferences and/or

competition formates,was first put forward byCharlesDarwin,

in his book The Descent of Man and Selection in Relation to

Sex.(2) The hypothesis was intended to explain those striking

sexually dimorphic features in animals that cannot be readily

understood within the general adaptationist framework of the

theory of natural selection. The archetypal example of sexual

selection is the male peacock’s tail which confers no general

adaptive advantage, indeed a probable disadvantage for

survival, but which has utility in the male peacock’s wooing of

the female. In principle, all pronounced secondary sexually

dimorphic traits might be the products of sexual selection, and

serve either to enhance chances of mate selection or to

function in intra-sex competition for mates.

Although the idea of sexual selection, in particular Darwin’s

belief that female choice of mates on the basis of the

‘‘aesthetic’’ appeal of certain male traits can influence the

evolution of the latter, was treated for many decades with

skepticism, sometimes bordering on derision, the general

phenomenon of sexual selection and Darwin’s hypothesis

about the importance of female preference are now well

accepted.(3,4) Yet, the study of sexually selected traits has

remained, for the most part, at the purely morphological and

behavioral levels. In contrast to many features that are shared

equally between both sexes, such as external sensory organs,

appendages, various internal organs, segmentation patterns,

and bristle arrays, the analysis of sexually selected traits in

terms of their genetic or molecular foundations is still in its

comparative infancy. One of the few such studies of this kind

explored the basis of the control of sexually dimorphic pigment

patterns in an invertebrate group, the Drosophilids.(5)

In a recent report, Zauner et al.(1) have now extended the

study of sexual selection at the molecular level to the ver-

tebrates, specifically the fish genus Xiphophorus. The trait

studied is a classic one in the field of sexual selection—it was

noted by Darwin—and consists of the extension of the ventral

caudal fin-rays into a decorative ‘‘sword’’-like structure in the

male tail. Specifically, Zauner et al. analyzed sword formation

in juvenile males and in tail-fin regeneration with respect

to expression of a gene, msxC, that had previously been

implicated in fin-ray development in the zebrafish. They find

that msxC expression is elevated specifically in the ventral

caudal fin rays that give rise to the sword, both in normal

development and in fin regeneration.

This work will be briefly reviewed here and discussed with

respect to some of the wider genetic and evolutionary

questions and issues that it touches upon.

A short history of the Xiphophorus ‘‘sword’’

The Xiphophorus sword is a long extension of the ventral fin

rays that develops during sexual maturation of males, at

approximately 4–6 months. It is exclusively a male trait in

normal development, never spontaneously appearing in

116 BioEssays 26.2 BioEssays 26:116–119, � 2004 Wiley Periodicals, Inc.

BioEssays Editorial Office, 10/11 Tredgold Lane, Napier Street,

Cambridge CB1 1HN, UK. E-mail: awilkins@bioessays.demon.co.uk

DOI 10.1002/bies.10414

Published online in Wiley InterScience (www.interscience.wiley.com).

What the papers say

females. A typical male swordtail is shown in Fig. 1, along with

a swordless female. Not all sword-producing species, how-

ever, produce the impressive sword shown in the figure; the

males of some species produce relatively fore-shortened

swords.(6–8) Furthermore, not all Xiphophorus species pro-

duce swords; it is a trait found in some sub-clades and not in

others.(7,8) The sub-genera that do not produce swords are

designated as ‘‘platyfish’’ and sword production is evidently

a trait that has either been gained or lost more than once

within this genus (see below). Intriguingly, however, sword

production can be artificially induced both in females and in

some species that normally lack them, or in juvenile males

prematurely, by treatment with exogenous testosterone.(6)

Evidently, much of the genetic machinery for producing the

sword is present more widely within the genus than its normal

phenotypic manifestation would suggest.

Although the evolutionary history of the sword within

Xiphophorus is thus still somewhat murky, the evidence

that it is involved in sexual selection is strong. In particular,

female Xiphophorus have been shown to prefer males with

larger swords.(9) In this respect, swordtails are reminiscent of

peacocks; females of both groups seem to prefer exagge-

rated male traits, a possible instance of ‘‘runaway sexual

selection’’.(3) Furthermore, female platyfish that normally

never see swords in conspecific mates, are more inclined to

mate with sword-bearing males, when a sword has been

artificially attached.(10)

The evidence for a role of msxC in the

development of the male ‘‘sword’’

Since sword formation, or its absence, is a species char-

acteristic, the traitmust have a genetic basis. Several decades

ago, hybridization experiments between swordtails and platys

indicated that the trait was polygenic, involving several

unlinked genes.(6) The precise identity of the actual ‘‘sword

genes’’, however, remained unknown. For their molecular

analysis, Zauner et al. built on some previous observations on

fin regeneration in zebrafish. In that work, Akimenko et al.(11)

had reported that four genes of the msx (muscle segment

homeobox) subfamily of homeobox genes, zebrafish msxA,

msxB, msxC, andmsxD—a gene family previously implicated

in various epithelial-mesenchymal developmental processes

in vertebrates—are strongly upregulated during caudal fin

regeneration. All fourmsx genes show characteristic temporal

patterns and positional and tissue locations in the blastema of

the regenerating tail fin.

Zauner et al., therefore, decided to investigate msx

gene expression in relationship to sword development, in

both normally maturing Xiphophorus males of sword-bearing

species and in caudal fin regeneration, following amputa-

tion of the distal part of the fin including the sword, in those

species. They used RT-PCR to isolate msx homologs and

identified two, one an apparent msxC ortholog, the other

showing sequence relatedness to both Xenopus msx1 and a

fifth fish msx gene, msxE (which had been identified in both

zebrafish and puffer fish). The latter gene they designated

msxE/1.

The expression of msxC gives a clear and unambigous

correlation with sword development. Although it is expressed

at low but significant levels in the distal tips of all developing

caudal fin rays in normal development, it is expressed at much

higher levels specifically in the distal tips of the ventral fin rays

that will give rise to the sword. Furthermore, testosterone

treatment of juvenile males, a procedure that can prematurely

induce sword development, induced a small caudal protrusion

at 10 days which was accompanied by strong msxC expres-

sion in the developing sword rays specifically. In females,

in contrast, there was no elevation ofmsxC expression in any

of the developing rays during caudal fin regeneration. Similar

results were found in two sword-bearingXiphophorus species,

from different clades, X. helleri and X. montezumae although

the precise kinetics of sword development and of msxC

expression differed slightly between them. In both species,

two domains of expression were seen in regenerating fins:

Figure 1. Amale, sword-bearing Xiphophorus and a female.

(Photograph courtesy of Dr Manfred K. Meyer).

What the papers say

BioEssays 26.2 117

the distal, undifferentiated cells of the blastemata, and in a

more proximal mesenchymal domain that may correspond to

cells destined to become scleroblasts. The latter expression

domain during fin regeneration is not seen in zebrafish,(11) thus

marking one difference in fin development between the two

fish groups.

In a different pattern to msxC, expression of the msxE/1

gene showed detectable but lower levels of expression in the

future sword rays than in the middle or dorsal rays of re-

generating caudal fins. Furthermore, in data reported but not

shown by Zauner et al.,(1) msxE expression was not detected

at all in regenerating caudal fins of the zebrafish, a second

apparent difference between these two relatively distantly

related actinopterygians. The precise significance of this

observation, however, is difficult to evaluate.

Some developmental genetic

and evolutionary questions

These findings provide an important entry point into under-

standing the developmental basis of a sexually selected

morphological trait in a vertebrate species, probably constitut-

ing the first report of this kind. Yet, clearly, this is still only a

beginning. The results serve to highlight how much is still

unknown.

The first obvious question, of course, concerns the devel-

opmental role of the msx genes in fin development generally

and in sword fin ray development specifically. As homeobox

genes, msx genes are transcription factor genes and the

question transforms, in part, to one about the target genes and

processes that they govern. Given their predominant distal

expression in fin rays that have exaggerated distal growth,

one might surmise that they function as regulators of growth

outwards along the proximodistal axis. Intriguingly, the msx

genes are fairly closely related in sequence to another homeo-

box gene subfamily, the Distalless (Dlx) genes,(12) many of

whose members are involved in proximodistal patterning.

Expression data implicate FGFs and signalling through FGF

receptors as part of this process.(13,14)

Yet, such connection can only be part of the regulatory

story. Both themsx genes andFGF signallingmust be part of a

larger regulatory structure, a genetic network of some kind.

Since exposure to testosterone can trigger sword develop-

ment in females,(6) it must be that much of the network exists

in both sexes but is only activated by the testosterone signal

in male animals, which induces male secondary sexual

characteristics. Such considerations are connected to ques-

tions about the evolutionary origins of the network itself and

how its structure has been altered in different Xiphophorus

sub-clades. A fundamental question in this regard is whether

the sword has been gained in some lineages, relative to an

ancestor that lacked this appendage, or whether it was lost in

several lineages, in independent events, after divergence from

an ancestral stock that possessed swords.

Perhaps surprisingly, given the extent of phylogenetic

analysis given to this problem over the years, there is still no

firm consensus. The sister group of Xiphophorus, Priapella,

is swordless and the traditional interpretation is that sword

development in Xiphophorus was not ancestral but was

independently acquired several times within the group.(7) In

contrast to this view, Meyer et al.,(15) using molecular phylo-

genetic analysis in a report published in 1994, concluded

that sword-formation was an ancestral, autapomorphic trait of

Xiphophorus. In this interpretation, sword development has

been lost independently in several lineages within the genus.

Subsequent detailed analysis of both morphological and

molecular material has not, however, succeeded in resolving

the issue.(16) Nor do a priori arguments really help since either

pattern can be rationalized in selective terms. Multiple

instances of sword acquisition, for instance, would testify to

the value of this trait in sexual selection, as sword-bearing

males compete more favorably for females than swordless

ones. In contrast, multiple losses of the sword in evolution of

the genus could be explained in terms of a natural selective

‘‘cost’’ to sword possession.

To the extent that the different phylogenies are compatible

with either multiple gains or losses, however, both possibilities

would be instances of parallel evolution, the phenomenon in

which the ‘‘same’’ evolutionary change occurs independently

in several related lineages. As noted by Simpson(17) and by

Meyer,(18) parallel evolution almost certainly implies the

existence of unexpressed ‘‘genetic potential’’—genetic net-

works, in today’s terminology—that can be activated in related

lineages in a relatively simple fashion.

In this particular case, the apparent multiplicity of occur-

rences of the transition between sword development and

swordlessness within the genus implies that only a small

number of genetic changes are involved. In principle, multiple

loss—if sword-development is a basal trait(8,15)—could involve

mutations in several points of the network.On theother hand, if

the basal state was swordlessness, the ‘‘invention’’ of swords

independently in several lineages would, almost certainly,

involve comparatively minor changes, in the form of new links,

to a network that already existed and which was active in

another role. Indeed, Zauner et al. present some evidence

for the existence of a related, but slightly different, network in

the development of a second male morphological feature,

the gonopodium, a modified anal fin. They show that msxC

expression also takes place during (and is presumably

required for) the formation of the gonopodium. Indeed, from

the comparative probable respective evolutionary ages of

gonopodia and swords, the development of the sword may

have involved recruitment of some of the genetic machinery

of the former when the latter first evolved.(1) Ultimately, the

elucidation and comparisons of the various networks involving

msxC expression, both within fish and in other vertebrates,

will provide clues to the probable evolutionary relationships of

What the papers say

118 BioEssays 26.2

these networks and the relationships between these networks

and the specific phenotypic traits that they underlie.

Yet, leaving aside this longer-term and broader perspec-

tive, the work of Zauner et al. is, in itself, an important advance

in the study of sexual selection. It provides an important insight

into a classic instance of the phenomenon, the sword of the

Xiphophorus swordtail fishes. More generally, it illustrates

the potential to extend molecular-genetic analysis of devel-

opment to vertebrate sexually selected traits. Darwin would

surely have been pleased.

References1. Zauner H, Begemann G, Mari-Beffa M, Meyer A. Differential regulation

of msx genes in the development of the gonopodium, an intromittant

organ, and of the ‘‘sword’’, a sexually selected trait of swordtail fishes

(Xiphophorus). Evol Dev 2003;5:466–477.

2. Darwin C. 1871. The Descent of Man and Selection in Relation to Sex.

London: John Murray and Sons.

3. Andersson M. Sexual Selection. 1994. Princeton: Princeton University

Press.

4. Zuk M. Sexual selection. In: Pagel M, editor. Encyclopaedia of Evolution.

Oxford: Oxford University Press. Vol. II, pp. 1047–1051.

5. Kopp A, Duncan I, Carroll SB. Genetic control and evolution of sexually

dimorphic characters in Drosophila. Nature 2000;408:553–558.

6. Zander CD, Dzwillo M. Untersuchungen zur Entwicklung und Vererbung

des Caudalfortsatzes der Xiphophorus-Arten (Pisces). Zeitschr.

Wissensch. Zool 1969;178:275–315.

7. Rauchenberger M, Kallman KD, Morizot DC. Monophyly and geography

of the Rio Panuco Basic swordtails (genus Xiphophorus) with descrip-

tions of four new species. Am Mus Novit 1990;2975:1–41.

8. Meyer A. The evolution of sexually selected traits in male swordtail fishes

(Xiphophorus: Poeciliidae). Heredity 1997;79:329–337.

9. Basola AL. Female preference for male sword length in the green

swordtail, Xiphophorus helleri (Pisces: Poeciliidae). Anim Behav 1990a;

40:332–338.

10. Basola AL. Female preference predates the evolution of the sword in

swordtail fish. Science 1990b;250:808–810.

11. Akimenko MA, Johnson SL, Westerfield M, Ekker M. Differential

regulation of four msx homeobox genes during fin development and

regeneration in zebrafish. Development 1995;121:347–357.

12. Bendall AJ, Abate-Shen C. Roles for Msx and Dlx homeoproteins in

vertebrate development. Gene 2000;247:17–31.

13. Poss KD, Shen J, Nechiporuk J, McMahon G, Thisse B, Thisse C,

Keating MT. Roles for FGF signalling during zebrafish fin regeneration.

Dev Biol 2000;222:347–358.

14. Wang Y, Sassoon D. Ectoderm-mesenchyme and mesenchyme-

mesenchyme interactions regulate Msx-1 expression and cellular

differentiation in the murine limb bud. Dev Biol 1995;168:374–382.

15. Meyer A, Morrissey JM, Schartl M. Recurrent origin of a sexually selected

trait in Xiphophorus fishes inferred from a molecular phylogeny. Nature

1994;368:539–542.

16. Marcus JM, McCune A. Ontogeny and phylogeny in the northern

swordtail clade of Xiphophorus. System Biol 1999;48:491–522.

17. Simpson GG. Fossils and the History of Life. 1983. San Francisco: WH

Freeman.

18. Meyer A. 1999. Homology and homoplasy: the retention of genetic

programmes. In: Bock GR and Cardew G, editors. Homology. pp. 141–157.

Novartis Foundation Symposium 222. John Wiley & Sons: Chichester.

What the papers say

BioEssays 26.2 119