Fragments of worms can be readily cultured in small aquaria through which seawater is gently flowing or that are equipped with bubbling stones. Instant Ocean®, previously "conditioned" by living organisms, works well if natural seawater is not available. It is important that large volumes of clean, well-aerated water be used; several pieces of worm will survive much better in a one- or two-gallon aquarium equipped with well-washed quartz sand and a bubbling stone than will isolated pieces in small finger bowls. We have routinely carried out experiments without the use of antibiotics; however, penicillin (100 U/ml) or sodium sulfadiazine (50 µg/ml) in combination with streptomycin sulfate (100 µg/ml) can be used to retard bacterial growth without adversely affecting regeneration. All water should be filtered and changed frequently, and strands or coatings of mucus removed, to prevent accumulations of protozoa that, at times, attack regenerating fragments.
To investigate the participation of tissues in anterior regeneration, the following excisions can be performed on the elongated 13th segment [Fig. 1]:
a) The anterior portion of the gut and overlying epidermis can be dissected off the muscle mass.
b) The anterior half of the ventral muscle including the ventral epidermis and the anterior portion of the nerve cord of segment 13 can be removed.
We have found that removal of either the gut or the ventral muscle mass leads to a sharp flexion of the body in the direction of the excised tissue. In fragments in which the dorsal gut is removed, the ventral muscle mass curves sharply upward so that the cut ends of the muscle and gut are in close apposition. The anterior blastema forms at the junction of the cut tissue. Because of the muscle flexion, regenerating heads continue to curl sharply caudally along the dorsal surface of the gut. In segments in which ventral muscle was removed, the gut curves ventrally so that again, the cut ends of the tissue are next to each other. Within a few days, muscle tissue gradually extends forward from the old muscle mass. The regeneration blastema forms at the junction of the severed gut and the muscle extensions. Simultaneous posterior regeneration occurs. Both operations initiate a similar response. In neither does the regenerate form at the cut end of one tissue; rather, the original organization of the worm is first reestablished and the regenerate forms in the region where tissues of all three germ layers are present. This experiment was performed on the small Pacific species in which wound closure and healing occur more readily than in the larger Atlantic species.
The role of the nervous system in regeneration remains poorly understood. While there is little doubt that the nervous system is crucial, its role in supporting tissue regeneration remains unclear. In some other polychaetes, ablation and implantation studies showed that the brain is essential for posterior regeneration; because implanted brains that support regeneration lacked neural connections, the influence of one or more neurosecretory hormones was strongly indicated (Golding, 1967; Clark and Bonney, 1969). This is easily tested in Chaetopterus, because anterior regeneration does not occur caudal to segment 14. When segments 14 and 15 are isolated in pairs, no anterior regeneration occurs; however, posterior regeneration occurs normally in the complete absence of a brain or regenerating brain (Hill, 1972).
Again working with segment 13, one can test whether ventral ganglia play a role in regeneration by obtaining fragments that contain different parts of the nerve cord. The ventral ganglia (i.e., the enlarged areas from which segmental connectives emerge) lie ventral to the ciliated cup. If segment 13 is isolated and cut just anterior to the ciliated cup, the posterior half segment is ganglion-free. Under these conditions, the posterior halves regenerate both anteriorly and posteriorly; the anterior halves produce only new anterior segments (Hill, 1972) (Fig. 4), indicating that the ventral ganglia are probably essential for posterior regeneration.
On the basis of these results and those from other laboratories, it appears that something produced by the ganglia, probably a neurosecretory hormone, is necessary for posterior regeneration; anterior regeneration is cued differently. We have been unable to remove the ventral ganglia from the posterior half of the segment and implant them into the anterior half. However, if the anterior part of segment 13, which is incapable of undergoing posterior regeneration in isolation, is left attached to the anterior 12 segments, posterior regeneration occurs normally (Hill, 1972). Consequently, it appears that something not present in the anterior part of segment 13 is provided by the attached anterior segments. (In Chaetopterus the nerve cell bodies are not restricted to the ganglia, but extend along the connectives.) This last experiment has been performed on both species.