The use of Chaetopterus [Fig. 1] as a system for study in developmental biology apparently began with E.B. Wilson (Wilson, 1882), who briefly described the early development ofChaetopterus along with that of several other annelids. Experimental analysis of the development of this organism began with Jacques Loeb in a paper in which he incorrectly described "differentiation without cleavage" resulting from K+-activation as true parthenogenesis. Other early studies were performed by such legendary developmental and cellular biologists as F.R. Lillie, T.H. Morgan, A. Tyler, E.B. Harvey, J. Brachet, N.J. Berrill, and J. Pasteels. More recently, significant use of this species has been made by S. Inoué, W.R. Jeffery, J. Henry and M. Martindale, among others, as well as the present authors. A comprehensive compilation of Chaetopterus literature, including that in the Literature Cited section of this paper, is provided in Appendix 1.
Experimental studies of development in Chaetopterus have emphasized four areas: the regulation of the cell cycle (germinal vesicle breakdown, GVBD), mechanisms of fertilization and egg activation, the effects of egg organization on development, and regeneration of the adult worm.
The primary advantages of Chaetopterus for the study of GVBD and the cell cycle are that a large number of oocytes (>106 cells or 2 ml) can be obtained from a female at one time, and that all can be induced to undergo GVBD synchronously in response either to their natural trigger (an unknown trace component in seawater) or to certain cellular agonists/antagonists of known biological activity. A further advantage is that the oocytes then arrest at metaphase I of meiosis until they are fertilized or artificially activated. In other words, the cells can be induced to undergo the G2/M phase transition at will without continuing to cycle. Furthermore, these oocytes can be easily labeled with isotopic markers. The availability of large numbers of synchronized, easily labeled eggs is also an important consideration in studies of fertilization and egg organization.
The unique advantages of Chaetopterus in studies of fertilization are that the fertilizing sperm interact with morphologically definable structures on the egg surface (Anderson and Eckberg, 1983) and that the egg is at least as metabolically active before fertilization as it is after. In fact, the unfertilized egg uses much more O2 than does the fertilized (Whitaker, 1933a,b) or artificially activated (Brachet, 1938) egg. Despite this, the fertilized or artificially activated egg undergoes waves of calcium release (Eckberg and Miller, 1995) demonstrating that the processes that initiate development in these organisms are probably similar to those in deuterostomes. The fact that the metabolic response of Chaetopterus eggs to fertilization differs so from that of sea urchins despite similar initial signals should make them the object of more detailed study.
The unique feature that makes Chaetopterus of particular interest for studies of egg and embryo organization in development is the ability of the artificially activated or fertilized egg to undergo differentiation without cleavage (DWC) (Lillie, 1902) [Fig. 2]. This phenomenon can be used to ascertain the extent to which segregation of ooplasmic determinants is essential for normal development, because eggs undergoing DWC do not segregate ooplasmic substances with cell membranes.
Chaetopterus eggs also undergo a series of reorganizations during meiosis and mitosis and produce a small polar lobe [Fig. 3].
Probably the most tantalizing questions facing students of regeneration are the following: (1) After injury, how does an animal "know" what is missing? (2) How does the remaining part reinitiate developmental processes? (3) How is the development of the blastema controlled so that the missing portion, and only that portion, is replaced? These questions are especially amenable to study in Chaetopterus, in which one segment is capable of regenerating an entire worm, while the adjacent segment is not.
Chaetopterus, like many polychaetes, is capable of extensive regeneration [Fig. 4]. Amputations made at almost any level of the body lead to the replacement of missing parts. Many other polychaetes, including capitellids and immature nereids, can regenerate posterior segments (Clark and Bonney, 1960; Golding, 1967; Hill et al, 1982; Hill et al, 1988). Chaetopterus, however, in addition to undergoing posterior regeneration from almost any level of the body, can undergo extensive anterior regeneration. This regenerative capacity is particularly striking because of the worm's high degree of specialization for a tubiculous life. The unique segmental specialization of Chaetopterus allows each segment to be recognized as it regenerates and makes the regenerative process particularly spectacular.
Chaetopterus offers several features that make it particularly well suited to studies of regeneration. (1) Its segments are morphologically distinct [Fig. 1]. (2) It regenerates both anterior and posterior segments. (3) The capacity to regenerate anterior segments is restricted to the anterior of the worm, while the capacity for posterior regeneration is more widespread. This allows the investigator to evoke posterior regeneration in the presence of either existing cephalic structures, regenerating cephalic structures, or the complete absence of cephalic structures by altering the level of amputation. (4) The thirteenth segment, which can regenerate an entire worm, is extremely long. This allows operations to be performed on a single isolated segment, thus eliminating the influence of adjacent segments [Fig. 4].
Fertilized eggs undergo typical spiral cleavage. The first five cleavages are synchronous; thereafter, cleavage becomes highly asynchronous. While the cell lineage has not been followed as completely in this organism as in some other spiralians (Mead, 1897; Henry 1986), there is no reason to expect that it differs significantly in Chaetopterus.
The first cleavage is unequal due in part to the presence of a small polar lobe [Fig. 3] and more importantly to asymmetric placement of the metaphase spindle. In some embryos, polar lobes do not form, but the cleavage is still unequal and subsequent development is normal. Polar lobes in spirally cleaving embryos typically have substantial morphogenetic significance as shown by the fact that removal of the polar lobe results in severely deficient embryos, although this is less true for Chaetopterus than for most other lobe-bearing spiralians that have been studied (Henry, 1986, 1989; Henry and Martindale, 1987).
Adult Chaetopterus pergamentaceus live in U-shaped tubes buried in the sand [Fig. 5; Appendix 2]. The leathery tube gives the animals one of their common names, "the parchment worm." Their other common name, "the innkeeper worm," derives from the fact that most Chaetopterus share their tubes with commensal crabs [one of which is shown in Fig. 7]. When animals are first brought into the lab, we leave them in their tubes. Sex can be determined by either cutting off one end of the tube and squeezing the worm to the cut end of the tube until the gonadal (posterior) parapodia [Fig. 1] are visible or, alternatively, by cutting a slit in the middle of the tube and viewing the parapodia through the slit. Gonadal parapodia from fertile males will be a uniform milky-to-yellowish white; those from fertile females will contain yellow ovaries that appear coiled in the living animal. In either case, the worms are not injured and will reseal the cut tube. If observation of the specimens is desired, they may be removed from their tubes and placed in curved glass or Tygon® tubing. They will eventually secrete opaque tube material that will make them invisible, but they can be observed for weeks before that happens.
To prevent spontaneous spawning, we keep the worms at 15-18°C. This seems to have no effect on their fertility. If running seawater is available, the animals need not be fed as long as there is sufficient water flow for them to obtain an adequate supply of plankton from the water. If they are kept in closed aquaria, they can be fed commercial filter-feeder diets (available from marine aquarium stores). If they are not fed, they begin to resorb their gametes within 2-3 weeks.
We have also kept fertile adults for at least 2 weeks, singly or in groups, in finger bowls of running seawater after they are removed from their tubes. This is convenient and especially valuable for specimen conservation if the studies under way do not require a large number of eggs. Because the sperm from one parapodium are more than sufficient to fertilize even a very large number of eggs, this is especially convenient for obtaining sperm over the course of several experiments. If the sexes are not kept segregated, the females should be briefly rinsed with distilled (or tap) water to kill any adhering sperm before eggs are obtained.
Information on sources of fertile Chaetopterus is provided in Appendix 3.
Chaetopterus eggs suitable for use in studies involving fertilization or development may be obtained essentially as described (Costello and Henley, 1971) . The methods below work on both species, but the adult C. variopedatus from the Pacific coast are much smaller than the Atlantic C. pergamentaceus and yield only ~10% as many eggs per adult female. Sperm and egg structure and morphology are similar in the two species, though the egg sizes differ slightly. We have never been able to force the sperm of either species to fertilize the eggs of the other [Appendix 2].
First shorten the tube by careful snipping and determine the posterior end of the worm. Worms will back out of their tubes into a bowl of seawater if the head is gently prodded with a blunt probe. They are reluctant to advance anteriorly. From one to all of the gonadal parapodia may then be snipped off into a beaker or finger bowl [Fig. 6].
The simplest way to obtain eggs is to cut the tips off several gonadal parapodia [Fig. 7] and leave the worm in a dish of seawater for 10-15 min. The worm will release a few hundred to a few thousand eggs into the dish. These may be collected by a Pasteur pipet and washed with seawater prior to insemination. To increase yield, cut the tips off the ovarian parapodia [Fig. 7], and vigorously shake the oocytes out into seawater [Fig. 8, Fig. 9]. Such eggs must be filtered through cheesecloth [Fig. 10] to remove debris. These eggs will be suspended in a dense mucus that passes through the cheesecloth with the eggs, but which is removed by washing the eggs in seawater. Washing can be speeded up dramatically by centrifuging the eggs briefly at low speed (<2,000 rev/~500 x g<>). Care should be taken, as overcentrifuging the eggs will stratify their contents. Nearly all of the mucus is lost in the first wash, so subsequent washes can (and should) be done at unit gravity. This method can be expected to yield a homogeneous population of metaphase I-arrested cells approximately 105 µm in diameter [Fig. 11, Appendix 5]. If a significant fraction (5% or more) of the eggs is significantly smaller or their germinal vesicles do not break down within 30 min of contact with seawater, the egg batch should be discarded. Oocytes of this organism normally will NOT undergo germinal vesicle breakdown in artificial seawater (ASW) [Appendix 4] (Ikegami et al., 1976), so it is essential to have some natural seawater (NSW) available. Once the eggs reach metaphase I, however, fertilization and development will occur as well in ASW as in NSW. Eggs may then be vitally stained [Appendix 6] or microinjected [Appendix 7].
Sperm are obtained "dry" by cutting a single parapodium into a small (ca. 10 ml) beaker that is then filled with seawater [Fig. 12]. Contrary to some earlier workers, we have found that freshly diluted sperm of Chaetopterus are never motile, and that they become motile within 2-5 min after dilution. All sperm preparations should be observed by phase-contrast microscopy to ensure motility. A significant number of non-gametic cells will be observed in a typical sperm suspension. They do not appear to interfere with fertilization, but because they are much larger than the sperm, they can be removed by differential centrifugation if necessary or desired.
Eggs are shed as primary oocytes arrested at early prophase with nine pairs of small, partially condensed chromosomes in the large germinal vesicle (oocyte nucleus) [Fig. 13]. Upon contact with a trace component in NSW - but absent from ASW (Ikegami et al., 1976) - the chromosomes finish condensing, the germinal vesicle breaks down (GVBD), the meiotic spindle forms, attaches to the metaphase chromosomes and migrates to the animal pole [Fig. 11]. This entire process takes 12-15 min, with GVBD occurring about 8-9 min after the contact with seawater.
The process of GVBD can be studied if the oocytes are shed into calcium-free ASW (CaFASW) [Appendix 4]. When the oocytes are shed into ASW containing Ca++, a small but variable percentage of the oocytes will undergo GVBD; spawning them into CaFASW ensures that less than 5% will undergo GVBD spontaneously; any batch of oocytes spawned into CaFASW in which greater than 5% undergo GVBD should be considered abnormal and discarded. Other procedures involved in obtaining primary oocytes are as described above, except that all washes are carried out in CaFASW. This procedure should yield a homogeneous population of oocytes, >95% of which are arrested at the G2/M border.
When most of the sperm are motile (2-5 min after dilution), they are used to inseminate eggs. The morphological, molecular and physiological aspects of fertilization have been reviewed (Eckberg and Anderson, 1996). Because the eggs are resistant to polyspermy (Eckberg and Anderson, 1985) , the actual amount of sperm added to the eggs [Appendix 8] is no more crucial than that for the more familiar sea urchin. Chaetopterus have both rapid and slow blocks to polyspermy. The rapid block is probably electrical in nature (Jaffe, 1983; Eckberg and Anderson, 1985) . The slow block is somewhat unique in that it results from the retraction of egg microvilli from the vitelline envelope in the absence of a cortical reaction comparable to that in many other organisms (Anderson and Eckberg, 1983) . A similar mechanism also acts in another annelid, Neanthes (Sato and Osanai, 1986) .
Excess sperm are washed away by allowing the eggs to settle at unit gravity and decanting and replacing the seawater a couple of times. Zygotes can then be cultured. Because there is no cortical reaction or vitelline envelope (VE) elevation comparable to that of the sea urchin, it is not possible to tell immediately if eggs have been fertilized. However, 15 min after fertilization, the first polar bodies form, and concomitantly the VE begins to ruffle due to the formation and retraction of very long microvilli from the zygote surface. A thin perivitelline space forms when the microvilli retract, leaving a "fertilization envelope"; however, because exposure of the eggs to slightly hypertonic seawater can cause the same effect, we recommend that polar body formation be used as the criterion for fertilization.
The lack of a cortical reaction comparable to that of sea urchins, etc., may frighten investigators interested in examining the earliest phases of development. This should not be disturbing, however, as polar body formation occurs within 15 min, and a sample of a culture can be kept to ensure full activation. The cortical reaction in a sea urchin egg correlates in time with its initiation of development because both result from the same cell-cell interaction and probably from some of the same intracellular signals. However, it has no obligatory role in development and is, thus, not the best possible criterion for egg activation. The lack of such a response in Chaetopterus can then be seen as an advantage.
Although the vitelline envelope contains receptors for sperm (Anderson and Eckberg, 1983), it has no obligatory role in fertilization (Eckberg and Anderson, 1985; Polk et al., 1987) . Oocytes from which the VE has been removed [Appendix 9] can be fertilized as efficiently as intact eggs, but become polyspermic (Polk, et al., 1987) .
The metabolic changes that occur in these eggs upon fertilization are ripe for study. The term "egg activation" has a slightly different meaning in Chaetopterus than it does in sea urchins. In direct opposition to the metabolic responses of sea urchin eggs, when Chaetopterus eggs are fertilized, their oxygen consumption decreases (Whitaker, 1933; Brachet, 1938) as do their protein synthetic and some metabolite transport activities (unpub. data), indicating that metabolic activity is maximal prior to fertilization and may actually be depressed somewhat after fertilization. Clearly, studies to compare metabolism before and after fertilization should be fruitful.
Fertilized eggs can be cultured at room temperature and will develop normally either in monolayer (Petri or Syracuse dish) or suspension culture (spinner flask or screw-cap centrifuge tube on a rocker). They can even be cultured for at least an hour at a concentration of at least 10% (v/v, eggs/seawater) [Appendix 5] without deleterious effects on development. This reduces the amount of isotopes needed for labeling studies. Development is very rapid. Swimming blastulae are produced in 4-5 h and fully differentiated trochophore larvae are produced in less than 24 h at 22°C. Embryos do not appear to "hatch," but cilia grow out through the vitelline envelope. The trochophore is slightly atypical, lacking a prototrochal ciliary band.
For long-term culture or studies in which the embryos are to be labeled with isotopes, embryos should be cultured in ASW containing 100 U/ml penicillin and 100 µg/ml streptomycin.
Chaetopterus oocytes that are activated parthenogenetically do not cleave, but they do undergo substantial reorganization, produce cilia and develop a gross morphology deceptively like that of a normal trochophore [Fig. 2]. This process is called differentiation without cleavage (DWC) and consists of several reasonably discrete phases: pseudocleavage (ameboid contractions that occur when controls cleave); unicellular gastrulation (redistribution of the granular and agranular endoplasms that results in a reorganization equivalent to that which segregates the ectoderm from the endoderm); ciliation (the differentiation of cilia) and swimming.
Differentiation without cleavage is traditionally initiated by adding excess KCl to seawater (Loeb, 1901; Lillie, 1902) . Exposure of the oocytes to excess K+ results in nearly 100% of the oocytes undergoing DWC. Details of this method and an alternative to elicit DWC from fertilized eggs are given in Appendix 10.
Although the egg is opaque and densely pigmented, all conventional light microscopic techniques can be applied, including phase- and differential interference-contrast, epifluorescence and polarization. However, because of the size and opacity of the egg, a significant fraction of any fluorescent signal from cytosolic molecules in parts of the egg that are distant from the objective lens will be quenched (Eckberg and Miller, 1995) . The presence of refringent granules in the egg makes visualization of spindles by polarization optics slightly more complex than in some other cells. However, this can be compensated by slightly flattening the eggs with a coverslip (Eckberg and Palazzo, 1992) or by centrifuging them into fragments (Lutz et al., 1988) [Appendix 11]. The eggs can also be microinjected [Appendix 7], impaled with electrodes (Hagiwara and Miyasaki, 1977; Hagiwara and Jaffe, 1979; Jaffe, 1983; Eckberg et al., 1993), and vital stained [Appendix 6], and pieces of the eggs can be removed microsurgically (Henry, 1986) [Appendix 12].
Because neither impaling these eggs with electrodes nor microinjecting them with markers for subsequent observation prevented their being fertilized, similar manipulations should be effective procedures for examining other phenomena.
General procedures for fixing specimens for microscopic examination are given in Appendix 13.
The role of the egg cytoskeleton in development has been reviewed (Eckberg and Anderson, 1995); it is responsible for localizing mRNA molecules in the egg (Jeffery and Wilson, 1983; Jeffery, 1985; Swalla et al., 1985; Jeffery et al., 1986) and for reorganizing the egg during differentiation without cleavage (Brachet and Denis-Donini, 1977; Brachet and Donini-Denis, 1978; Eckberg, 1981a; Eckberg and Kang, 1981) . Procedures have been developed for the isolation of the cytoskeleton and its analysis [Appendix 14].
Because large populations of oocytes arrested at metaphase can be obtained, these oocytes have also been used frequently for the isolation of the meiotic spindle [Appendix 15] and its analysis (Inoué et al., 1974; Salmon, 1975; Goode and Sarma, 1982, 1986; Lutz et al., 1988).
A regenerative response can be elicited readily by isolating somite (segment) combinations and maintaining them in filtered, gently flowing or well-aerated seawater. Worms should be removed gently from their tubes [Appendix 16]. We routinely perform operations without the use of anesthesia, but if required, motion may be slowed down by dropwise addition of seawater saturated with 1,1,1-trichloro-2-methyl-2-propanol hydrate (chloretone) or by chilling.
Segments can be isolated readily and without damage by snipping intersegmentally with sharp scissors. Segment 13 is very long [Fig. 1] and, if isolated, can regenerate an entire worm. This provides investigators with the distinct advantage of having all of the regenerative capacities contained in one segment instead of repetitive units and with a single segment that is large enough to allow complex manipulations [Appendix 16] [Fig. 4].
Wound-healing in Chaetopterus has been briefly described (Faulkner, 1932; Berrill, 1928, 1952), but modern methods such as electron microscopy and autoradiography have not been used. At the National Vibrating Probe Facility located at the Marine Biological Laboratory, Woods Hole, Massachusetts, we have measured electrical currents associated with wound closure in sabellid fanworms (Hill et al, 1993, 1994). Attempts to extend this work to Chaetopterus have been unsuccessful to date, because of the activity level of the worms.
The origin of the cells that give rise to the new outgrowths has not been investigated recently. Faulkner (1932) attributed the formation of the posterior blastema to the accumulation of neoblasts, a population of cells believed to retain the embryonic characteristics of totipotency and cell division. This is not the case in sabellid fanworms (Hill, 1970), so this problem should be investigated in Chaetopterus.
Regeneration of anterior somites occurs readily back to the level of somite 13. Berrill (1928) reported anterior regeneration from segment 14 (one case); we have found that anterior regeneration from the 14th segment occurs only rarely and is usually incomplete. Replacement of anterior segments does not occur behind segment 14.
Chaetopterus is one of the polychaetes that seems to possess the ability to "count." Each somite has its own identity [Fig. 1] and is replaced directly during regeneration. For example, if the body is severed between somites 12 and 13, the anterior part of the body will regenerate new segments 13, 14, 15, etc., posteriorly, while the posterior part regenerates new segments 12 through 1 in the anterior direction.
Segmentation of anterior and posterior regenerates occurs in fascinatingly different ways in Chaetopterus. During anterior regeneration, the blastema elongates into a small, delicate tube-like structure. Numerous somites appear simultaneously by segmentation of the blastema. Rudiments of tentacles and of the long aliform notopodia appear. Thus all of the missing segments are replaced by direct outgrowth and differentiation of the blastema. This method of regeneration (epimorphosis) is quite different from the regenerative response of many other polychaetes that regenerate anteriorly (Gross and Huxley, 1935; Berrill and Mees, 1936).
Posterior regeneration is like that in other polychaetes in that a pygidium forms from the blastema, and a growth zone is established immediately anterior to it (Faulkner, 1932). New segments are added anterior to the pygidium, and these differentiate to form whatever has been lost. Because mature animals seem to continue to add genital segments slowly, the process of posterior regeneration finally becomes indistinguishable from that of normal growth.
WRE is currently supported by grants from the National Institutes of Health and the Council for Tobacco Research, USA, Inc..
NOTE: See Appendix 1 for an extensive compilation of Chaetopterus literature (both cited and non-cited).
Appendix 1: Chaetopterus Literature
Appendix 2: Habits, collection, and adult anatomy
Appendix 3: American species
Appendix 4: Seawater formulae
Appendix 5: Determining egg concentration
Appendix 6: Vital staining
Appendix 7: Microinjection
Appendix 8: Determining sperm concentration and sperm:egg ratio
Appendix 9: Vitelline envelope removal
Appendix 10: Differentiation without cleavage
Appendix 11: Production of transparent mini-cells lacking yolk
Appendix 12: Microsurgery on embryos
Appendix 13: Fixation procedures
Appendix 14: Cytoskeleton isolation
Appendix 15: Isolation of meiotic spindles
Appendix 16: Regeneration
Fig. 1: Adult Chaetopterus pergamentaceus.
Fig. 2: Differentiation without cleavage; micrographs of semi-thin sections.
Fig. 3: Differential interference contrast micrographs of movements associated with polar body formation, cytoplasmic localization and cleavage.
Fig. 4: Amputation and regeneration of segment 13 in Chaetopterus.
Fig. 5: Tube of an adult worm as it looks when removed from the sand.
Fig. 6: Removal of gonadal parapodia from a female Chaetopterus.
Fig. 7: The ends of the gonadal parapodia are cut off to release the eggs.
Fig. 8: After their tips have been cut off, gonadal segments are shaken in a beaker of seawater using forceps to release the eggs.
Fig. 9: An egg suspension, obtained by shaking the cut gonadal segments.
Fig. 10: The egg suspension is poured through a double layer of cheesecloth.
Fig. 11: A phase-contrast micrograph of an unfertilized egg after germinal vesicle breakdown and spindle formation.
Fig. 12: Collection of sperm.
Fig. 13: Fluorescence micrograph of a Hoechst-stained primary oocyte prior to germinal vesicle breakdown.