Welcome to our new website

Robert Josephson

Robert Josephson writes about Carl Pantin's 1935 ground breaking publications on sea anemone neurophysiology. Pdf files of Pantin's papers can be accessed as supplemental data at jeb.biologists.org

In the early 1930s Carl Pantin spent several months at the Stazione Zoologica in Naples studying neuromuscular transmission in crustacea. These investigations reached a convenient stopping point several weeks before Pantin was to return to England, and he cast about for something else to do in the meantime. The previous occupant of the laboratory in which he was working had left some sea anemones in an aquarium, and Pantin chose to examine the neural control of muscular contraction in these (Pantin, 1968). The results of this fortuitous change in direction was a series of papers on the behavioral machinery of coelenterates that changed, in a major way, views on the neural control of behavior in coelenterates and on the early evolution of nervous systems in general. The first and arguably the most important of these papers is the principal subject of this column (Pantin, 1935a).

Coelenterates (Phyla Cnidaria and Ctenophora) are the most simply constructed animals with nervous systems, and they offer the most simply organized nervous systems in the animal kingdom. Both cnidarians and ctenphores are basically composed of two epithelial layers, an outer ectoderm and an inner endoderm, separated by a largely acellular, gelatinous mesoglea. Histological studies in the 19th and early 20th centuries established that there are nerve cells in coelenterates, and that these form diffuse, two-dimensional nerve nets that lie in the basal regions of the epithelial layers. In the sea anemone polyps studied by Pantin all the nerve cells occur as parts of diffuse nerve nets. In the late 19th century, G. J. Romanes, Darwin's protégé, had examined conduction in jellyfish as part of his exploration of the evolution of mental processes in animals. Through clever cutting experiments Romanes demonstrated that contractile waves are conducted diffusely across the subumbrella epithelium of the jellyfish Aurelia and will spread between any two blocks of subumbrellar tissue so long as these are joined by a bridge of intact tissue larger than a millimetre or so in width (Romanes, 1885). The diffuse conduction demonstrated physiologically by Romanes was consistent with the diffuse distribution of the nerve cells found in histological studies. G. H. Parker, another giant in early neurophysiology, used the cut and stimulate approach of Romanes and found that conduction in the column of anemones is also diffuse (Parker, 1919). Parker stimulated coelenterate tissue with mechanical prodding, light and chemicals. He noted, as had others before him, that contractile responses in coelenterates were sometimes local and often increased in amplitude with increasing strength of stimulation. To account for local, graded responses, Parker proposed that conduction in the cells of the coelenterate nerve net was graded, and not all-or-nothing as in axons of higher animals. Thus when Pantin began his studies of anemones there was a question as to whether conduction in coelenterate neurons was graded or all-or-nothing, and it was not clear how to account for local, graded behavioral responses thought to be mediated through a widely distributed, diffusely conducting nerve net.

Pantin brought to his studies of anemones an appreciation for the importance of facilitation in neuronal functioning obtained from his work on neuromuscular transmission in crustacea. He also brought to this project electrical stimulators, which, though quite primitive by later standards, were adequate to produce single shocks and trains of shocks of controlled intensity and frequency for activating anemone conducting systems.

The work described in the first anemone paper from Pantin's time in Naples demonstrated that conduction in the column is all-or-nothing. An effective stimulus initiated a wave of activity that reached all parts of the column. The contractile responses evoked by appropriate column stimulation were symmetrical contractions of columnar muscles and of the sphincter at the top of the column. The column conducting system had a clear threshold, and increasing the intensity of stimuli above the threshold caused no change in evoked responses. A conducted wave in the column was followed by a refractory period lasting a few tens of milliseconds. The units of the conducting system, presumably the neurons of the nerve net, appeared to have the same basic physiological properties as neurons elsewhere in the animal kingdom.

The column conducting system acted as a single unit. Conduction in the oral disk of the anemone was rather different. A single stimulus to the disk initiated a wave of activity that spread tangentially over only a small part of the disk. A second stimulus given shortly after the first led to greater tangential spread, but many sequential stimuli were required to activate the entire disk. Pantin attributed this progressive spread of activity to `interneural facilitation'. The wave of activity initiated by a stimulus was thought to have died out at synapses between neurons of the net that were initially non-transmissive, and to have left these synapses temporarily tranmissive to subsequent impulses arriving at them. Interneural facilitation results in responses whose spread increases with increasing number of imposed stimuli or with increasing number of impulses generated by sensory structures during natural behaviour. Further, because muscles near the stimulus origin receive more impulses than do those further away, the contractile responses are greatest at the stimulus site and decline with distance from this site, thus accounting for local responses whose spread and intensity are graded with stimulus number or, with natural stimuli, with stimulus strength.

Although the wave of activity initiated by a single shock to the column apparently reaches all the muscles of the column, it generally results in no overt muscle response. Two or more stimuli at an appropriate frequency are required to initiate muscle contraction, presumably because of requirements for neuromuscular facilitation at the junctions between the nerve net and the muscles. The second paper in the anemone series considered the varying requirements for facilitation among the different muscles of the column (Pantin, 1935b). Very low frequency of stimulation, one shock each 10 s or so, activated just the circular muscles. Increasing the stimulus frequency brought in, progressively, the parietal muscles, the mesenterics and finally the powerful sphincter muscle. Thus different frequencies of activation caused qualitatively different behavioral responses.

Following the publication of Pantin's research on anemones in 1935 (Pantin, 1935a,b,c) one could have entertained, at least for a while, the idea that the behavioral machinery of anemones was largely understood. It appeared that a reasonably complete model of anemone behavior could be constructed with a few, simple components: a through-conducting nerve net, a locally conducting nerve net with interneural facilitation, and a group of muscles with differing requirements for neuromuscular facilitation. But the inadequacies of this simple view soon became apparent. Unstimulated anemones are not inert. Batham and Pantin, in a series of papers published in 1950, used slow mechanical recording and time-lapse cinematography to view the behavior of unstimulated anemones. They found that even under constant conditions animals periodically expand, contract, sway, and even move about by gliding on the pedal disk (reviewed in Pantin, 1952). These behaviors occur on a time scale so slow that that are not easily perceived by direct observation. How these slow behaviors are coordinated is still quite unknown. More recently it has been shown that there is not just one but rather several conducting systems in the column of anemones; a rapidly conducting system, probably the column nerve net, and at least two slow systems (see, for example, McFarlane, 1969, 1975). The slow systems may reflect activity in a subset of the neurons of the general nerve net or they may represent conduction in the epithelial cells themselves. Impulses in the slow systems appear to modulate such activities as oral disk expansion and retraction, and foot detachment. The goal of totally explaining the behavior of anemones seems to be receding as more is becoming known about the behavior of these interesting creatures. But it should be emphasized that the newer findings on slow and often spontaneous behaviors, and the presence of multiple conducting systems in anemone tissues, are additions to and not changes in the basic plan for the behavioral machinery of anemones given to us by Pantin in his classic papers published nearly 70 years ago.