The lateral line system at metamorphosis in Xenopus laevis (Daudin)

Just how sense organs of developing animals are adapted to the concurrent changes of behaviour is unknown. The Amphibia are especially suitable for studies in this field because, wherever metamorphosis occurs, drastic changes in anatomy and mode of life take place in a short time. In Xenopus laevis the larval lateral line system develops smoothly into that of the adult. This paper describes some major changes in this system during development and tentatively correlates them with changes in behaviour.

The whole function of the lateral line systems of fishes and Amphibia is subject to debate but some features are known. For example, animals possessing lateral line organs can accurately locate objects in their environment (Dijkgraaf, 1963). Characteristics of the system which help to make this function possible are: (1) organs are spread in rows over large parts of the surface of the animals; (2) individual organs are maximally sensitive to water currents in one plane only (Flock, 1965); and (3) different organs are orientated so that their planes of maximal sensitivity are in different directions.

During the metamorphosis of X. laevis the supra-orbital and post-orbital lines migrate to a new position around the orbit of the eye (Paterson, 1939).

This movement involves changes in the orientation and location of organs so that the plane in which they show maximal sensitivity to vibration changes at metamorphosis. Therefore, the system as a whole must compensate for the new distribution of individual organs. The present account shows that this conclusion applies generally to the whole lateral line system.

A group of lateral line organs of adult X. laevis is called a plaque (Murray, 1955) within which the organs are separated from each other by raised tactile organs (Calabresi, 1924). The latter, by virtue of their position, direct water currents on to the lateral line organs (Gömer, 1961) and may have a protective function. Rows of plaques make up each lateral line and a common nerve trunk supplies each row with branches to each plaque.

Each organ contains ciliated receptor cells (neuromasts) of two classes. These are sensitive to water displacements in opposite directions as shown by electrophysiological distinction between units (Gömer, 1963). The maximum sensitivity of the plaque is approximately at right angles to its long axis although there is some variation in the orientation of organs in one plaque (Gömer, 1961, 1963). Each neuromast cell has a single kinocilium projecting from its apical surface. The kinocilium possesses the standard ‘9 + 2’ arrangement of ciliary filaments. In addition, neuromast cells have a group of stereocilia to one side of the kinocilium. These stereocilia are about two-thirds of the diameter of kinocilia and have no ‘9 + 2’ pattern of filaments. They do contain fine fibrils but there is no obvious order in their arrangement. The directional sensitivity of organs is correlated with an anatomical orientation of stereocilia to the single kinocilium of each neuromast cell; half of the cells have the kinocilium on one side of the stereocilia and half have the kinocilium on the opposite side (Gömer, 1963). In adult X. laevis the nerve running from each plaque to the main nerve trunk always contains (a) two large myelinated nerve fibres (8–10 μin diameter) ; (b) a bundle of non-myelinated fibres, and (c) often but not always, one or more small myelinated fibres (0·5–1·0 μin diameter). The two large fibres are sensory (Gömer, 1961) and synaptic-type contact between the unmyelinated ends of sensory fibres and receptor cells has been shown in the lateral line system of Lota vulgaris (Flock, 1965) although the exact pattern of innervation is unknown. The small myelinated fibres are efferent and inhibitory (Russell, 1968) and inhibitory fibres have been shown to make synaptic-type contact in L. vulgaris (Flock, 1965). The function and destination of the non-myelinated fibres is unknown.

The position of the lateral line rows in larval X. laevis has been described most accurately by Nieuwkoop & Faber (1967), and in the adult by Escher (1925), Horst (1934), Paterson (1939) and Murray (1955). The anterior organs are supplied by the anterior lateralis nerve and the posterior ones by the posterior lateralis nerve. Peripherally the anterior lateralis nerve supplies three lateralis rows, the supra-orbital, the infra-orbital and the hyomandibular, each of which has further subdivisions. The posterior lateralis nerve divides peripherally to supply five rows, the occipital, the aortic (previously undescribed), the lower lateral, the middle lateral and the upper lateral lines. The lower lateral line has several subdivisions. In addition to the rows there are anterior and posterior auditory groups. All these parts of the system have been examined during metamorphosis.

Developmental stages of laboratory bred animals were classified according to the system of Mieuwkoop & Faber (1967). They were anaesthetized in a 1:10000 solution of MS 222 (Sandoz) in distilled water and then pithed. All material for examination was fixed at 0 °C in 1% osmium tetroxide buffered to pH 7·4 with veronal acetate and made up to a concentration of 300 m-osmoles with sucrose (after Palade, 1952). Fixative was pipetted under and on top of the skin. Material being prepared for whole mounts was allowed to darken noticeably before being removed and dehydrated in acetone. Skins were laid out on microscope slides and covered with small pieces of glass to keep them flat during dehydration. They were finally mounted in Araldite under cover slips on microscope slides. The osmophilic properties of the organs and nerves made them stand out against the paler background. Whole mounts of skin were photographed and the distribution of organs and their rows was transferred to diagrams of the complete system. Five adult animals and five stage 55 tadpoles were examined and combined data concerning each lateral line row were tabulated. Material for electron microscopy was removed after 10 min in situ fixation and fixed for a further 20 min in covered watch-glasses. After dehydration in acetone, material was embedded in Araldite. Blocks of tissue were cut on an L.K.B. ultramicrotome and sections were stained with lead citrate and uranyl acetate for examination with an A.E.L EM 6B electron microscope. For light microscopy 1 μ sections were stained for 1 min with toluidine blue. For conventional electrophysiological recording tadpoles were held down and bathed with Ringer solution (see Russell, 1968) and nerve trunks were picked up on silver wire hook electrodes.

The distribution of lateral line rows in larva and adult is shown in Figs. 1 and 2 respectively. The names of the subdivisions of the rows are listed in Table 1 together with data regarding the position of the rows, the number of plaques found in each row and the angles at which plaques are orientated with respect to an antero-posterior median axis.

The number of plaques in many of the rows is approximately the same in larva and adult, in others there is a drastic reduction in numbers. The anterior auditory groups, the aortic row and the tail portions of the middle lateral and upper lateral lines are lost completely. Numbers of plaques in the hyomandibular, anterior lower lateral and caudal rows are severely reduced (see Table 1). In certain rows there appears to be an increase in plaque number at metamorphosis, this apparent increase is probably due to confusion in knowing where one row begins and another ends and does not represent formation of new organs (there is, for instance, a large reduction in organs of the anterior lower lateral line and an increase in the posterior lower lateral).

Lines which are dorsal, lateral or ventral in the larva often maintain these positions in the adult; however, the infra-orbital, tentacular, aortic, anterior lower lateral, posterior lower lateral, caudal and posterior auditory rows show distinct changes of position at metamorphosis (see Table 1).

The orientation of the long axis of plaques with respect to the animal’s median antero-posterior axis was measured in a number of larvae and adults. For each row the range of angles to the reference axis was ascertained and the extremes noted. While plaques in certain lines approximately maintain their larval orientation, many do not. These include the post-orbital, hyomandibular, occipital, anterior lower lateral, median ventral and caudal lines (see Table 1).

In the anterior lower lateral and caudal lines there are changes in position, number and orientation of organs. In the hyomandibular and posterior lower lateral lines there are changes in two of these parameters. In all lines there is some change in at least one of the three features examined.

While it was found impossible to record nervous activity from single plaques, recordings from whole nerve trunks were made using silver wire hook electrodes. All recording was done from the middle lateral nerve trunk as a relatively long length of nerve was available. Spontaneous afferent nervous activity was recorded from proximally cut larval nerves as early as stage 54, at least 20 days before the functional loss of the tail. The activity showed a clear increase when a vibrating rod was introduced into the water (Fig. 3).

Sections of the sensory hairs of larval organs at stage 54 reveal a clearly orientated arrangement of groups of stereo- and kinocilia similar to that found in the adult. There are two populations of sense cells identified by the relative position of the kinocilium to the stereocilia in each group. Some cells have their kinocilium on one side of the stereocilia, the others have their kinocilium on the opposite side (Fig. 4). Ciliary groups are arranged so that the plane of maximum sensitivity would be approximately at right angles to the plaque’s long axis.

At metamorphosis the skin thickens, the distance between the basement membrane and the surface increasing from 30μ in the larva to 100 μ in the adult. The lateral line organs in the larva are about 50 μ deep and in consequence their outer surface protrudes above the level of the surrounding skin (Figs. 5, 7). The adult organs increase in size to 80 μ deep but sink down into the skin. At the same time tactile organs develop between the organs in each plaque (Figs. 6, 7), clearly separating neighbouring organs. The result is that the larval sensory hairs are situated well above the skin surface while the adult ones are partly below it. These changes in morphology of the organ plaque may have functional significance.

Examination of the nerve bundles innervating single organ plaques in the larva consistently reveals two morphologically distinct groups of fibre; two myelinated fibres approximately 2–4μ in diameter and a number of unmyelinated fibres up to 0·5μ in diameter (Fig. 8 A). At stage 60 another class of smaller myelinated fibres is present occasionally. At later stages the additional class of small myelinated fibre is found more frequently, and at stage 66 (adult) is nearly always present (see Table 2, Fig. 8B). This small myelinated fibre is known to be of efferent function in the adult, the pair of larger myelinated fibres is of known afferent function.

The larval lateral line system is clearly functional at least by stage 54. Myelination of the nerve pairs running to plaques is complete by stage 52 and afferent activity has been recorded from tadpole nerve trunks as early as stage 54. The orientation of cilia at the apex of the sensory cells suggests that organs have the same type of directional sensitivity in the larva as has been shown for the adult (Gömer, 1963). The plane of maximum sensitivity may be less clearly defined, however, because larval ciliary groups do not lie in channels but are raised above the skin surface. An orientated arrangement of cilia has been noted in Rana pipiens tadpoles (Jande, 1966), an animal which loses its lateral line system at metamorphosis. It would seem, therefore, that the orientated cilia in larvae are of functional significance to the tadpole rather than just a stage in the development of the adult.

The redistribution of organ plaques and their changes of orientation must result in changed planes of maximum sensitivity at metamorphosis. If the tadpole lateral line system is used as in the adult to locate objects in the environment, the peripheral modifications necessitate changes in the central nervous system. It is possible that the larval system is not used as a ‘distant touch’ receptor at all and is merely a generalized vibration receptor used in conjunction with the Mauthner cells in rapid escape movements. This seems unlikely when an obviously orientated arrangement of hair cells occurs in both R. pipiens and Xenopus tadpoles.

The small myelinated inhibitory class of fibres may be present in larvae in a non-myelinated form or it may grow out as a completely new element of the nervous system at metamorphosis. Its maturation corresponds to a stage in development when there is a distinct change in locomotory behaviour. The tadpole swims in a fish-like manner and the adult moves using a characteristic hind-leg kick. It has already been noted that the level of efferent activity sharply increases when adults attempt to move (Gömer, 1967). It may be that the adult animal develops the myelinated inhibitory system to ‘switch off’ peripherally the lateral line organs during swimming movements. This peripheral control by a fast conducting myelinated system may not be necessary in Xenopus tadpoles where locomotion depends upon a sinusoidal tail movement.

Des changements prononcés surviennent au cours de la métamorphose de Xenopus, dans l’anatomie du système de la ligne latérale. La répartition des canaux diffère chez la larve et chez l’adulte, en outre l’orientation et le nombre des organes sont modifiés à la métamorphose.

Des plaques larvaires sont fonctionnelles, ainsi que l’a montré l’enregistrement de leur activité nerveuse.

Deux espèces cellulaires, avec cils polarisés, sont présentes chez le têtard bien avant que l’orientation des plaques individuelles d’organes se soit modifiée à la métamorphose.

La topographie de la surface cutanée autour des plaques individuelles change à la métamorphose. Cette modification peut réduire la sensibilité directionnelle des organes.

On trouve des axones inhibiteurs myélinisés dans le nerf latéral seulement lorsque le têtard devient mature. Ce changement a lieu au moment où le mode de locomotion adulte se développe.

This work was supported by a scholarship from the Medical Research Council. Thanks are due to Dr David Sandeman for help during the work and for his valuable suggestions.