DYNAMICS OF SOMATIC CELLLINEAGE COMPETITION IN CHIMERAS OF Hydractinia symbiolongicarpus (CNIDARIA: HYDROZOA)

Dinámica de competencia entre líneas celulares somáticas en quimeras de Hydractinia symbiolongicarpus (Cnidaria: Hydrozoa)

RYAN S. SCHWARZ, BSc¹; LUIS F. CADAVID, MD, Ph D^2*

¹Department of Biology, University of New Mexico, Albuquerque, NM 87131, USA.
²Departamento de Biología e Instituto de Genética, Universidad Nacional de Colombia, Bogotá, Colombia. *To whom correspondence should be addressed: Luis F. Cadavid, Instituto de Genética, Of. 213, Universidad Nacional de Colombia, Carrera 30 No. 4508, Bogotá DC, Colombia, Tel.: (571) 316 50 00 Ext. 11615. Fax: (571) 316 55 26. Email: lfcadavidg@unal.edu.co

Presentado el 30 de mayo de 2007, aceptado el 12 de agosto de 2007, correcciones 22 de agosto de 2007.

ABSTRACT

Sessile colonial invertebrates often fuse with conspecifics to form chimeras. Chimerism represents an unequivocal instance of withinindividual selection where genetically different celllineages compete for representation in the somatic and gametic pools. We analyzed temporal and spatial variations in somatic celllineage composition of laboratoryestablished chimeras of the hydroid Hydractinia symbiolongicarpus (Cnidaria: Hydrozoa). Using three clones with different allotypic specificities (i.e., two rejecting one another but fusing with a third one), we established two classes of twoway chimeras, a single threeway chimera class, and an incompatible interaction as control. Chimeras were sampled at five time intervals for a year. Celllineages in samples were identified by polyp fusibility assays against tester colonies of known fusibility. The cell lineages composing the chimeras showed a differential competitive ability, with one of them representing close to 80% by the end of the study. Rare celllineages stabilized at low frequencies but preserved their ability to gain somatic representation and to colonize distant parts of the chimera. This behavior characterizes cell parasites. As a consequence of the reproductive plasticity of most colonial invertebrates, celllineage variability may be transmitted to the offspring both sexually and asexually. Successful somatic competitors are expected to be preferentially transmitted asexually, whereas cell parasites would be preferentially transmitted sexually.

Key words: Allorecognition, Hydractinia, unitsofselection, celllineage competition, chimera

RESUMEN

Los invertebrados coloniales y sésiles con frecuencia se fusionan con conespecíficos para formar quimeras. Estas quimeras son un ejemplo de selección natural actuando al interior del individuo en donde células genéticamente distintas compiten por acceso tanto a la línea somática como a la germinal. En este estudio se analizaron las variaciones temporal y espacial de linajes celulares somáticos en quimeras establecidas en el laboratorio del hidroide colonial Hydractinia symbiolongicarpus (Cnidaria: Hydrozoa). Usando tres clones con distintas especificidades alotípicas (dos de ellas se rechazaban pero ambas se fusionaban a una tercera), se establecieron dos clases de biquimeras, una triquimera y una interacción incompatible como control. Muestras de tejido de quimeras se obtuvieron en cinco intervalos de tiempo durante 50 semanas. La identidad celular de cada muestra se determinó por ensayos de fusibilidad de pólipos con colonias estándar de fusibilidad conocida. Los distintos linajes celulares de cada quimera mostraron una habilidad competitiva diferencial, con una de ellas representando cerca del 80% de las quimeras hacia el final del estudio. Las líneas celulares con menor representación se estabilizaron a bajas frecuencias pero mantuvieron la capacidad de aumentar en frecuencia y de colonizar partes distantes en la quimera. Este comportamiento caracteriza los parásitos celulares. Como consecuencia de la plasticidad reproductiva de la mayoría de invertebrados coloniales, la variabilidad de los linajes celulares puede ser trasmitida a la descendencia tanto sexualmente como asexualmente. Linajes celulares somáticos con alta capacidad competitiva serían heredados asexualmente, mientras que los linajes celulares parásitos se transmitirían preferencialmente por reproducción sexual.

Palabras clave: alorreconocimiento, Hydractinia, unidades de selección, competición, quimeras

INTRODUCTION

Sessile colonial invertebrates frequently encounter members of their own species as they grow on hard surfaces, with the encounter typically culminating in either fusion of kin or rejection of unrelated colonies (Grosberg, 1988). These allorecognition responses have been unambiguously described in sponges, cnidarians, bryozoans, and ascidians (reviewed in Grosberg, 1988) and have excited interest in various disciplines. Specifically, these phenomena have attracted the attention of evolutionary theorists as a paradigm of conflicts between unitsofselection (Buss and Shenk, 1990), have represented a challenge to conventional notions on the origin and maintenance of genetic variability in natural populations (Grosberg et al., 1996; Grosberg and Hart, 2000), and have led to debates amongst comparative immunologists as to whether such phenomena represent the ancestral state of key elements of the vertebrate immune system (Burnet, 1971; Scofield et al., 1982; Buss and Green, 1985).

Allorecognition responses in the colonial hydroid of the genus Hydractinia (Cnidaria: Hydrozoa) are better understood than in perhaps any other colonial invertebrate (e.g.,Teissier, 1929; Schijfsma, 1939; Crowell, 1950; Hauenschild, 1954; Hauenschild, 1956; Müller, 1964; Müller, 1967; Toth, 1967; Ivker, 1972; Gallien and Govaere, 1974; Buss et al., 1984; Buss et al., 1985; Buss, 1987; Müller et al., 1987; Grosberg, 1988; Lange et al., 1989; Buss and Grosberg, 1990; Shenk and Buss, 1991; Lange et al., 1992; Grosberg et al., 1996; Mokady and Buss, 1996; Hart and Grosberg, 1999). Hydractinia is a colonial athecate hydroid found in nearshore oceanic waters growing as a surface incrustation on gastropod shells inhabited by pagurid hermit crabs (Buss and Yund, 1989). Colonies are diploblastic and composed of three morphological modules: polyps, stolons and the stolonal mat. Polyps are feeding structures and gamete carriers, and are embedded in the stolonal mat, a twodimensional basal plate that consists of two ectodermal layers sandwiching a network of endodermal canals. These endodermal canals provide vascular continuity between the polyps' gastric cavities and may extend beyond the stolonal mat, in which case they are called stolons. Colonies release their gametes to the ocean where fertilization occurs. Fertilized eggs develop into crawling planula larvae which settle on hermit craboccupied shells, and subsequently metamorphose into primary polyps. As stolons extend, bifurcate, and anastomose, new polyps bud from the stolons yielding thus a mature colony (Ballard, 1942; Berking, 1991).

Contacts between Hydractinia colonies results in one of three outcomes: (i) Fusion: After contact, compatible colonies dissolve their periderm coat and, within an hour, they adhere to one another by their epithelial cells. Two to four hours postcontact, colonies establish a common gastrovascular system forming a permanent chimera. (ii) Rejection: Upon contact, incompatible colonies fail to adhere, and within the first 12 hours, interacting tissues begin to swell due to massive migration of "stinging cells" or nematocytes, phylumdefining cells containing specialized organelles called nematocysts. Nematocysts discharge a harpoonlike thread which delivers toxins causing extensive tissue destruction in the opponent (Müller, 1964; Buss et al., 1984). Subsequent rejection takes two forms depending upon the colony morphology (Buss and Grosberg, 1990). Encounters between incompatible stoloniferous colonies (i.e., those with predominance of free stolons over stolonal mat) result in aggressive rejections and are characterized by the induction of a specialized organ of defense, the hyperplastic stolon (Ivker, 1972). In these reactions, differentiation and recruitment of nematocytes continue until one colony has eliminated the other (Buss et al., 1984; Lange et al., 1989). Confrontations between stolonless colonies (i.e., those with predominance of stolonal mat over free stolons), produce passive rejections. These responses are characterized by the secretion of a fibrous matrix by both colonies, accompanied by cessation of growth along the contact margin (Buss and Grosberg, 1990). (iii) Transitory fusion: In this reaction, colonies initially fuse only to separate days or weeks later (Hauenschild, 1954; Shenk and Buss, 1991; Grosberg et al., 1996; Gild et al., 2003; Cadavid et al., 2004). When initial fusion is established through the stolons, the reaction is characterized by initial occlusion of vascular spaces, followed by local necrosis and separation of interacting stolons (Shenk and Buss, 1991). When initial fusion is established through the stolonal mats, a necrotic band appears at the point where colonies initially contacted. This band subsequently spreads to form a line spanning the original contact zone. The emergence of the necrotic line is accompanied by occlusion of the once fused endodermal canals. Within days or weeks after the first appearance of the necrotic line, colonies separate from one another. From this point on, the response is indistinguishable from a passive rejection, except at the growing edges of the contact zone which, upon contact, display the same time course and phenomenology described above (Cadavid et al., 2004). Transitory fusion may in fact represent a composite of phenotypes varying in time course, progression and developmental regulation. In H. symbiolongicarpus, these allorecognition responses are mediated by a dosedependent interaction of two linked allorecognition loci, with fusion occurring when colonies share at least one allele at each locus, rejection resulting when they share no alleles at either locus, and transitory fusion occurring when colonies share a single allele at only one of the two loci (Cadavid, 2004; Cadavid et al., 2004).

Allorecognition responses in colonial invertebrates mediate competitive interactions at two different levels of biological organization: the colony and the celllineage (Buss, 1990; Buss and Shenk, 1990). Rejecting colonies compete for habitable space by interfering with their competitor's growth, where the outcome of such competition is a function of colony size and morphology (Buss and Grosberg, 1990). Fusing colonies, on the other hand, form discrete physiological entities composed by genetically different celllineages which compete for representation in the chimera's somatic and germ cell lines (Stoner et al., 1999). Chimerism represents an example of withinindividual selection and potentially plays a prominent role in the adaptive evolution of colonial invertebrates (Buss, 1987). Natural selection acts on differences in cell growth and survival between cell lineages within the chimera. Yet, for celllineage selection to have an evolutionary significance, this variation ought to be inherited to the offspring. In contrast to solitary organisms for which evolutionary important variation is restricted to the germ line, colonial invertebrates may transmit both somatic and germ line variation to the progeny. These organisms often lack discrete germ lines but possess multipotent stem cells that are able to differentiate into somatic and germ cells at any time during ontogeny. Thus, variation in fitnessrelated traits between celllineages composing a chimera can be sexually inherited after their eventual differentiation into gametes, or asexually transmitted through fragmentation, fission or budding. Chimerism, therefore, is an important mechanism to introduce heritable celllineage variation in colonial invertebrates. The evolutionary significance of celllineage selection due to somatic fusion largely depends on both the frequency of chimerism in natural populations and the competitive behavior of celllineages. While a number of studies have shown that chimerism is of relatively common occurrence among colonial invertebrates (Stoner and Weissman, 1996; Hart and Grosberg, 1999; Stoner et al., 1999; Sommerfeldt et al., 2003), our understanding of the nature and dynamics of celllineage competition and selection in chimeras of colonial invertebrates is still fragmentary. In this study, we analyzed differences in fitnessrelated traits, such as replication and survival rates, between somatic celllineages in laboratoryestablished chimeras of H. symbiolongicarpus. Using unique tissue compatibility specificities as markers for individual celllineages, we measured the celllineage composition of different classes of chimeras and determined how such composition varies in time and space.

MATERIALS AND METHODS

ANIMALS AND CHIMERA GENERATION

Three Hydractinia symbiolongicarpus clones were used in this study (8338, 41172, and 43163). They were derived from the inbreeding program designed to genetically characterize the Hydractinia allorecognition complex (Mokady and Buss, 1996; Cadavid et al., 2004). Colony 8338 was the product of eight generations of brothersister matings from two wildtype colonies. Clone 41172 resulted from seven generations of inbreeding from a wildtype animal and a sixgeneration clone from the 8338's inbred line. Clone 43163 was derived from three generations of backcrossing from the 8338's line into the 41172's line. These clones differed in their allorecognition specificities such that 8338 and 41172 reject one another and both fuse to 43163. Animals were grown on microscope glass slides and maintained in 35 liter aquaria in recirculating artificial seawater (Reef Crystals) at 16 ± 2 °C with three changes per week of onequarter of the water volume. Colonies were fed three times a week to repletion with 34 day old nauplii of Artemia salina. Colonies 8338 and 41172 also served as testers to identify individual celllineages in the chimeras by fusibility assays (see below). Two classes of twoway chimeras, 8338/43163 and 41172/43163, and one threeway chimera, 8338/43163/41172, were established. Twoway chimeras were generated by placing a fivepolyp fragment of stolonal mat from each partner at 0.5 cm from one another on a glass slide, and held in position with a thread. Threeway chimeras were established in a linear disposition with colony 43163 placed in the center between 8338 and 41172 colonies. An incompatible interaction (8338/41172) was also established as control. Explants attached to the glass slide within 3648 hours and the thread was then removed. Clones grew into contact within 36 days, and were subsequently maintained as described above.

TISSUE SAMPLING AND POLYP FUSIBILITY ASSAYS

Cell lineage composition in each chimera was determined by polyp fusibility assays on defined sampling regions and at various time intervals, as follows. Microscope glass slides serving as substrata for chimera growth were divided into fourteen regions of 1 x 1.25 cm each, distributed in two rows and seven columns (Fig 1A). From each sampling region containing tissue, two to four polyps were surgically removed with a scalpel and used for polyp fusibility assays (Lange et al., 1992) against tester clones 8338 or 41172. Briefly, an excised polyp from a sampling region was held with its cut aboral end in contact with an excised polyp from either 8338 or 41172 tester clones. Interacting polyps were maintained in position by passing a human hair through their gastric cavities. Compatible polyps developed continuous endodermal and ectodermal cell layers forming a common gastric cavity within 1224 hours whereas incompatible polyps remained separated (Fig 1B). Accordingly with the different allorecognition specificities, if a polyp from the 41172/43163 chimera failed to fuse with the 8338 tester it was considered to be derived from the 41172 celllineage. If, however, the assay resulted in fusion, the polyp was considered to be derived from the 43163 celllineage. Likewise, if a polyp sample from the 8338/43163 chimera failed to fuse with the 41172 tester, it was assigned to the 8338 celllineage, but if the interaction resulted in fusion, the polyp was considered to be derived from the 43163 celllineage. In threeway chimeras four to six polyps were excised from each sampling region and divided into two groups of equal size. One was assayed against the 41172 tester and other against 8338 tester. Assignment of polyps to a given celllineage in the threeway chimera as described may underestimate the actual contribution of 8338 and 41172 cell lineages. That is, if a polyp from the threeway chimera fails to fuse with the 41172 tester, it is considered to be derived from the 8338 celllineage. But if the interaction results in fusion, it could be derived from either 41172 or 43163 celllineages. We, therefore, registered the minimum number of samples derived form 8338 and 41172 cell lineages in the threeway chimera. For the incompatible control (8338/41172), two to four polyps from each sampling region were tested with polyps from the 41172 tester. Assays resulting in fusion indicated that the polyp was derived from cell lineage 41172, whereas failure to fuse indicated that it was derived from cell lineage 8338. As sampling destroys tissue and might bias cell lineage composition in time, temporal changes in cell lineage composition was evaluated on chimera replicas. Five replicas of each chimera and incompatible control were established. Each replica was sampled only once at either 2, 8, 37.5, 44.5, or 50 weeks postcontact. No data was collected at week 50 postcontact for the threeway chimera. To evaluate reproducibility, three experimental trials of identical design were performed simultaneously. Only one trial was performed for the incompatible interaction (8338/41172).

STATISTICAL ANALYSIS

As repeated sampling destroys tissue and can bias cell lineage composition and distribution in time, replicas of each chimera class were established to be sampled only once at either 2, 8, 37.5, 44.5, or 50 weeks postcontact. Two or three replicas were established for each sampling time. The General Lineal Model on an analysis of variance (ANOVA) design was used to test the null hypothesis that the mean number of polyps contributed by each cell lineage in a chimera was the same at the different time intervals. Multiple comparisons of means was performed with the TukeyKramer method to obtain confidence intervals and P values for all pairwise differences between level means. Statistical analysis was performed with the Minitab package.

RESULTS

Table 1 shows the mean number of polyps derived from the different cell lineages in each chimera at the different time intervals. The 8338/43163 chimera grew linearly in time and showed a significant difference in the mean number of polyps across time intervals (F=18.89, DF=4, P=0.0). Examining pairwise comparisons of mean number of polyps at different weeks postcontact revealed that the overall difference was explained by the difference in mean number of polyps between 2 and 44.5 weeks (T=6.1, P=0.0), 2 and 50 weeks (T=7.4, P=0.0), 8 and 44.5 weeks (T=4.6, P=0.0), and 8 and 50 weeks (T=5.7, P=0.0). In this chimera, the mean number of polyps contributed by cell lineage 43163 was significantly grater than that contributed by cell lineage 8338 across time intervals (F=105.1, DF=1, P=0.0). While there was no significant difference between the mean number of polyps contributed by 8338 and 43163 cell lineages at 2, 8, and 37.5 weeks postcontact, the number of polyps from 43163 was significantly higher than those of 8338 at 44.5 weeks (T=8.9, P=0.0) and 50 weeks (T=9.7, P=0.0) postcontact. At this two time intervals, cell lineage 43163 represented the 93% and 87%, respectively, of the chimera. The 41172/43163 chimera also showed a linear growth with significant differences in mean number of polyps across time (F=23.8, DF=4, P=0.0). In pairwise comparisons of mean number of polyps for different time intervals, all differences were significant at the 5% level, except those for 2 and 8 weeks, 8 and 37.5 weeks, 37.5 and 50 weeks and 44.5 and 50 weeks post contact. The 41172/43163 chimera also was primarily composed by cell lineage 43163 (F=48.0, DF=1, P=0.0). The mean number of polyps derived from cell lineage 43163 was significantly higher than that derived from cell lineage 41172 at 44.5 (T=9.1, P=0.0) and 50 (T=5.0, P=0.001) weeks post contact. Cell lineage 43163 represented the 88% and 71% of the chimera at 44.5 and 50 weeks post contact, respectively. Thus, in both twoway chimeras, cell lineage 43163 grew at a higher rate than its partner to become the predominant composite of the chimera. Cell lineages 8338 and 41172 remained at low frequency and were relatively stable through time. Additionally, they preserved the ability to increase in frequency after a critical low at 44.5 weeks postcontact. Figure 2A compares the fraction of 8338 and 41172 cell lineages in their respective chimeras.

The threeway chimera 8338/43163/41172 was set with two rejecting clones bridged with a mutually fusible clone. Surprisingly, this chimera did not show signs of rejection at any time during the study and grew linearly across time intervals (F=9.43 DF=3 P=0.0). This overall difference in growing was explained by differences in the mean number of polyps between 2 and 8 weeks (T=3.1, P=0.02), 2 and 37.5 weeks (T=4.5, P=0), and 2 and 44.5 weeks (T=4.7, P=0) postcontact. A significant difference between the relative contributions of cell lineages to the chimera was also observed across time intervals (F=6.38, DF=3, P=0.002).

This overall difference was accounted only by the differences in number of polyps at 37.5 weeks post contact between cell lineages 8338 and [8338 OR 43163] (T=3.7, P=0.05), 8338 and [41172 OR 43163] (T=5.5, P=0.0004), and 41172 and [41172 OR 43163] (T=4.1, P=0.02). Thus, contrary to what was observed in the twoway chimeras, all celllineages contributed nearly equally to the threeway chimera. Figure 2B shows the fraction of positively identified 8338 and 41172 cell lineages that contributed to the threeway chimera. While 41172 showed higher representation during the last three time intervals than its 8338 counterpart, the difference was not significant. As it was the case for the twoway chimeras, no total replacement of celllineages was observed in the threeway chimera at the end of the study. The incompatible control 8338/4117 elicited a strong rejection response with clone 41172 displaying the most aggressive behavior. This clone developed a dense thread of hyperplastic stolons which progressed slowly to eliminate clone 8338. Tabla 1 shows the number of polyps derived from both competitors. At the onset, polyps from both colonies were detected at equal numbers, but as the rejection progressed, most of the polyps derived from clone 41172. After week 37.5 postcontact, clone 41172 completely eradicated clone 8338. Figure 2C depicts the fraction of each competitor detected at the different time intervals.

Spatial distribution of celllineages within chimeras was evaluated by counting the number of polyps derived from a given cell lineage for each of the 14 sampling regions and over all experimental trials. Figure 3 shows the location of each cell lineage at each time interval. In this figure, squares represent polyps derived from a given celllineage and areas of 4x3 squares correspond to sampling regions. Within a sampling region with more than one celllineage, distribution of polyps is arbitrary. The overrepresentation of celllineage 43163 in the twoway chimeras is evident from week 8 postcontact. Although cell lineages 8338 and 41172 were identified at low frequencies, they were highly mobile. Indeed, polyps derived from these two cell lineages were found throughout the chimeras. A similar situation was observed in the threeway chimeras, where polyps derived from 8338 and 41172 celllineages were randomly distributed in the clone. While in some cases polyps derived from incompatible celllineages occupied the same region, no signs of incompatibility were observed.

DISSCUSION

Most levels of biological organization function as units of natural selection in the evolutionary process, provided that they display heritable variation in fitness (Lewontin, 1970). The heritability nature of celllineage variation correlates with certain lifehistory traits and differ between solitary and colonial organisms (Orive, 2001). Most solitary organisms that reproduce gametically separate the germ and somatic lines early in ontogeny. Thus, in these organisms only variation introduced in the germ line, typically via mutation, is represented in the gametic pool. Heritability of celllineage variation in colonial organisms differs from that of solitary animals in at least three fundamental ways. First, celllineage variation in colonial organisms is introduced not only by mutation, but also by somatic fusion between conspecifics. In chimeras of colonial organisms, therefore, two or more cell lineages may compete for access to somatic and germ pools. Second, colonial organisms which reproduce gametically do not sequester the germ line. Instead, they possess multipotent stem cells which are competent to produce gametes and the various somatic cell types throughout ontogeny. As a consequence, celllineage variants arising either by mitotic mutation or somatic fusion might be represented in the gametic pool. Third, colonial invertebrates may reproduce asexually by fission, budding or fragmentation. Propagules produced by these means might differ to their parental clone in celllineage composition.

Chimerism is of common occurrence amongst colonial invertebrates. The frequency of chimeras in natural populations of the ascidian Botryllus schlosseri ranges from 68% (Karakashian and Milkman, 1967; BenShlomo et al., 2001), whereas in another ascidian species, Diplosoma listerianum, it ranges from 361% (Sommerfeldt et al., 2003). In natural populations of H. symbiolongicarpus, the observed frequency of chimerism is 67% (Hart and Grosberg, 1999). Celllineage competition is known to occur in chimeras of colonial ascidians. Using microsatellites to trace celllineages, Stoner et al. (1999) showed that celllineages from Botryllus schlosseri chimeras compete for access to the gonads and somatic tissues, and that the competitive ability of each lineage is hierarchical and heritable. In this study, we have documented variation in celllineage composition through time and space in laboratoryestablished chimeras of the hydroid H. symbiolongicarpus. This variation is largely explained by differences in cell growth rates between celllineages composing the chimeras. In all classes of chimeras, celllineages from the original partners were detected after nearly a year of establishment, indicating that they are stable through time and that their survival rates are equivalent. They, however, had a different replication rate which resulted in a differential competitive ability. In both twoway chimeras, celllineage 43163 was significantly overrepresented, constituting the majority of the colony early in the study. Thus, cell lineage 43163 had a higher replication rate than both 8338 and 41172. While twoway chimeras grew linearly, the number of polyps derived from 8338 and 41172 were relatively constant during the study period. A comparison between celllineages 8338 and 41172 showed that the latter displayed a higher rate of replication than the former, although no significant difference was detected. In the threeway chimera, differences in competitive abilities of cell lineages were not as evident as in twoway chimeras. As suggested by Stoner et al. (1999), competitive interactions between celllineages might be modified as a function of the number of competing genotypes. Constancy across replicates suggests that the competitive ability of cell lineages is genetically controlled, and therefore, heritable.

Celllineages composing a chimera might differ in their somatic versus gametic investments (Buss, 1982). Indeed, a celllineage may be disproportionately represented in the gametes at the expense of the chimera's soma. This "cell parasitism" has been observed in various colonial organisms (Buss, 1982; Stoner and Weissman, 1996; Rinkevich, 2002), and might provide an adaptive explanation for the origin and evolution of allorecognition systems in colonial invertebrates (Buss, 1982). As none of the chimeras used in this study reach sexual maturity, it was not possible to evaluate differential somatic versus gametic investment schemes between celllineages. Yet, there was indirect evidence that celllineages underrepresented in frequency (i.e., 8338 and 41172) behaved as cell parasites. On one hand, they were stable at low frequencies for nearly a year and thoroughly distributed in the chimera. On the other hand, during chimera regression when they were at the lowest frequency (week 44.5 postcontact), they displayed the ability to increase its representation. Thus, 8338 and 41172 cell

chimera (third panel) at the different time intervals. Chimeras are represented by a grid where each cell is a sampled polyp. Areas of 3 x 4 cells represent a sampling region in the chimera. Black cells indicate polyps derived from celllineage 41172, gray cells are polyps derived from 43163, and white cells are polyps derived from 8338 celllineages. In threeway chimeras, grey cells represent polyps derived from 41172 OR 43163 and 8338 OR 43163 (see Materials and Methods for an explanation). Numbers in the horizontal axis represent the five time intervals in weeks.

lineages showed a frequencydependent proliferation ability, which characterizes cell parasites (Buss, 1982). Despite the fact that rare celllineages might access the gametic pool in high proportions when they behave as parasites, asexual reproduction confers an advantage to the more frequent celllineage. Fragmentation processes by water movements is common in sessile colonial invertebrates (Fautin, 2002), and, in colonies of Hydractinia, have been shown to induce a rapid and abundant production of propagules (Bavestrello et al., 2000). Fragments of a chimera are likely to be composed by the most represented celllineage of the parental chimera.

Selection acting upon celllineages within individuals has long been recognized as a critical phenomenon in development and evolution. For example, celllineage selection is thought to act as a sieve eliminating deleterious and spreading beneficial mutations in populations (Otto and Hastings, 1998), and has been invoked as a key factor in the evolutionary transitions leading to multicellularity (Buss, 1987). Furthermore, it largely explains why mutation rates in humans have a paternalage effect and appear to be higher in male than female gametes (Extavour and García Bellindo, 2001; Goriely et al., 2003). In colonial invertebrates celllineage selection is likely to play an important role. First, the frequency of chimerism in natural populations of various colonial organisms is appreciable. Second, somatic celllineages composing a chimera display differential competitive abilities that results in a biased chimeric composition. More predominant celllineages have a better chance of being represented in progeny produced by asexual reproduction. If, on the other hand, rare celllineages behave as parasites, their representation might be higher when reproduction is sexual. Chimerism may confer competitive advantages to the colony by virtue of size increase. Mortality in sessile colonial invertebrates is sizedependent with smaller colonies displaying the highest mortality (Buss, 1990). Chimerism, as shown here, may provide an advantage to small colonies in early stages postfusion, when celllineages contribute equally to the colony. However, beyond a critical size, chimerism did not greatly contribute to increase colony size because competition resulted in a predominance of one celllineage over the other.

ACKNOWLEDGMENTS

We thank Mona Razaei and JohnPaul Larrieu for technical support, William Usaquén for statistical assistance, and Leo Buss for providing animals. This work was supported by University of New Mexico.

REFERENCES

BALLARD WW. The Mechanisms of Synchronous Spawning in Hydractinia and Perannia. Biol Bull. 1942;82:329-339.         [ Links ]

BAVESTRELLO G, PUCE S, CERRANO C, CASTELLANO L, ARILLO A. Water Movement Activating Fragmentation: A New Dispersal Strategy for hydractiniid hydroids. J Mar Biol Ass UK. 2000;80(2):361-362.         [ Links ]

BENSHLOMO R, DOUEK J and RINKEVICH B. Heterozygote deficiency and chimerism in remote populations of a colonial ascidian from New Zeland. Mar Ecol Prog Series. 2001;209:109-117.         [ Links ]

BERKING S. Control of metamorphosis and pattern formation in Hydractinia (Hydrozoa, Cnidaria). BioEssays. 1991;13:323-329.         [ Links ]

BURNET M. "Selfrecognition" in colonial marine forms and flowering plants in relation to the evolution of immunity. Nature. 1971;232:230-235.         [ Links ]

BUSS LW. Somatic cell parasitism and the evolution of somatic tissue compatibility. Proc Natl Acad Sci U S A. 1982;79:5337-5341.         [ Links ]

BUSS LW. The Evolution of Individuality. Princeton, Princeton University Press; 1987.         [ Links ]

BUSS LW. Competition within and between encrusting invertebrates. Trends Ecol Evol. 1990;5:352-356.         [ Links ]

BUSS LW, GREEN DR. Histocompatibility in vertebrates: the relict hypothesis. Dev Comp Immunol. 1985;9:191-201.         [ Links ]

BUSS LW, GROSBERG RK. Morphogenetic basis for phenotypic differences in hydroid competitive behaviour. Nature. 1990;343:63-66.         [ Links ]

BUSS LW, MCFADDEN CS, KEENE DR. Biology of hydractiniid hydroids. 2. Histocompatibility effector system/competitive mechanism mediated by nematocyst discharge. Biol. Bull. 1984;167:139-158.         [ Links ]

BUSS LW, MOORE JL, GREEN DR. Autoreactivity and selftolerance in an invertebrate. Nature. 1985;313:400-402.         [ Links ]

BUSS LW, SHENK MA. Hydroid allorecognition regulates competition at both the level of the colony and at the level of the cell lineage. Defense Molecules. J. J. Marchalonis and C. Reinisch. New York, Alan R. Liss: 1990;85-105.         [ Links ]

BUSS LW, YUND PO. A sibling group of Hydractinia in the northeastern United Stated. J Mar Biol Ass UK. 1989;69:875-895.         [ Links ]

CADAVID LF. Selfdiscrimination in colonial invertebrates: Genetic control of allorecognition in the hydroid Hydractinia. Dev Comp Immunol. 2004;28:871-879.         [ Links ]

CADAVID LF, POWELL AE, NICOTRA ML, MORENO M, BUSS LW. An invertebrate histocompatibility complex. Genetics. 2004;167:357-365.         [ Links ]

CROWELL S. Individual Specificity in the Fusion of Hydroid Stolons and the Relationship Between Stolonic Growth and Colony Growth. Anat Rec. 1950;108(560-561).         [ Links ]

EXTAVOUR C, GARCÍABELLINDO A. Germ cell selection in genetic mosaics in Drosophila melanogaster. Proc Natl Acad Sci USA. 2001;98:11341-11346.         [ Links ]

FAUTIN DG. Reproduction of cnidaria. Can J Zool. 2002;80:1735-1754.         [ Links ]

GALLIEN L, GOVAERE MC. Incompatibilite entre cultures intergeneriques d'explants chez les Hydraires Hydractinia echinata Fleming et Podocoryne carnea Sars. C R Acad Sc Paris. 1974;278:107-110.         [ Links ]

GILD S, FRANK U, MOKADY O. Allogeneic interactions in Hydractinia: is the transitory chimera beneficial? Int J Dev Biol. 2003;47(3):433-437.         [ Links ]

GORIELY A, MCVEAN GAT, ROJMYR M, INGEMARSSON B, WILKIE AOM. Evidence for selective adventage of pathogenic FGFR2 mutations in the male germ line. Science. 2003;301:643-646.         [ Links ]

GROSBERG RK. The evolution of allorecognition specificity in clonal invertebrates. Q Rev Biol. 1988;63:377-412.         [ Links ]

GROSBERG RK, HART MW.Mate selection and the evolution of highly polymorphic self/nonself recognition genes. Science. 2000;289:2111-2114.         [ Links ]

GROSBERG RK, LEVITAN DR, CAMERON BB. Evolutionary Genetics of the Allorecognition in the Colonial Hydroid Hydractinia symbiolongicarpus. Evolution. 1996;50:2221-2240.         [ Links ]

HART MW, GROSBERG RK. Kin interactions in a colonial hydrozoan (Hydractinia symbiolongicarpus): Population structure on a mobile landscape. Evolution. 1999;53:793-805.         [ Links ]

HAUENSCHILD CV. Genetische und entwicklungphysiologische Untersuchungen über Intersexuälitat und Gewebeverträlichkeit bei Hydractinia echinata Flem. Wilhem Roux' Archiv. 1954;147:1-41.         [ Links ]

HAUENSCHILD CV. Uber die vererbung einer geweberverträglichkeits eigenschaft bei dem hydroidpolypen Hydractini echinata. Z. Naturforsch. 1956;11b:132-138.         [ Links ]

IVKER FB. A hierarchy of histoincompatibility in Hydractinia echinata. Biol Bull. 1972;143:162-174.         [ Links ]

KARAKASHIAN S, MILKMAN R. Colony fusion compatibility types in Botryllus schlosseri. Biol Bull. 1967;133:473-490.         [ Links ]

LANGE R, PLICKERT G, MÜLLER WA. Histocompatibility in a low invertebrate, Hydractinia echinata: Analysis of the mechanism of rejection. J Exp Zool. 1989;249:284-292.         [ Links ]

LANGE RG, DICK MH, MÜLLER WA. Specificity and early ontogeny of historecognition in the hydroid Hydractinia. J Exp Zool. 1992;262:307-316.         [ Links ]

LEWONTIN RC. The units of selection. Annu Rev Ecol Syst. 1970;1:1-18.         [ Links ]

MOKADY O, BUSS LW. Transmission genetics of allorecognition in Hydractinia symbiolongicarpus (Cnidaria: Hydrozoa). Genetics. 1996;143:823-827.         [ Links ]

MÜLLER W. Experimentelle Untersuchungen über Stockentwicklung, Polypendifferenzierung und Sexualchimären bei Hydractinia echinata. Wilhem Roux' Arch EntwicklMech Org. 1964;155:182-268.         [ Links ]

MÜLLER W. Differenzierungspotenzen und Geschlechtstabilitat der IZellen von Hydractinia echinata. Wilhem Roux' Arch EntwicklMech Org. 1967;159:412-432.         [ Links ]

MÜLLER WA, HAUCH A , PLICKERT G. Morphogenetic Factors in Hydroids: I. Stolon Tip Activation and Inhibition. J Exp Zool. 1987;243:111-124.         [ Links ]

ORIVE ME. Somatic Mutations in Organisms with Complex Life Histories. Theor Popul Biol. 2001;59:235-249.         [ Links ]

OTTO SP, HASTINGS IM. Mutation and Selection within the Individual. Genetica. 1998;102/103:507-524.         [ Links ]

RINKEVICH B. Germ Cell Parasitism as an Ecological and Evolutionary Puzzle: Hitchiking with Positively Selected Genotypes. Oikos. 2002;96:25-30.         [ Links ]

SCHIJFSMA K. Preliminary Notes on the Early Stages in the Growth of Colonies of Hydractinia echinata (Flem.). Arch Neerland Zool. 1939;4:93-102.         [ Links ]

SCOFIELD VL, SCHLUMPBERGER JM, WEST LA, WEISSMAN IL. Protochordate Allorecognition is Controlled by a MHClike Gene System. Nature. 1982;295:499-502.         [ Links ]

SHENK MA, BUSS LW. Ontogenetic Changes in Fusibility in the Colonial Hydroid Hydractinia symbiolongicarpus. J Exp Zool. 1991;257:80-86.         [ Links ]

SOMMERFELDT AD, BISHOP JDD, WOOD CA. Chimerism Following Fusion in a Clonal Ascidian (Urochordata). Biol J Linn Soc. 2003;79:183-192.         [ Links ]

STONER DS, RINKEVICH B, WEISSMAN I. Heritable Germ and Somatic Cell Lineage Competitions in Chimeric Colonial Protochordates. Proc Natl Acad Sci USA. 1999;96:9149-9153.         [ Links ]

STONER DS, WEISSMAN IL. Somatic and Germ Cell Parasitism in a Colonial Ascidian: Possible Role for a Highly Polymorphic Allorecognition System. Proc Natl Acad Sci USA. 1996;93:15254-15259.         [ Links ]

TEISSIER G. L'Origine multiple de certaines colonies d'Hydractinia echinata (Flem.) et ses consequences possibles. Bul Soc Zool Fr. 1929;54:645647. TOTH SE. Tissue Compatibility in Regenerating Explant from the Colonial Marine Hydroid Hydractinia echinata. J Cell Physiol. 1967;69:125-131.         [ Links ]