On finishing military service in 1955, Gaze was appointed to a lectureship in physiology at the University of Edinburgh. He became Reader in Physiology in 1966. In 1970 Sir Peter Medawar invited Gaze to become director of a new division in developmental biology at The National Institute for Medical Research in London. Gaze accepted the invitation. He returned to Edinburgh in 1984 and led a Medical Research Council unit on neural development and regeneration. In 1992 he retired, continuing to live in Edinburgh.
Gaze began his post-doctoral research by repeating the experiments of Roger Sperry.[4] The experiments involved cutting the optic nerve in frogs, rotating the eye, and allowing the nerve to grow back. Vision was restored. Sperry showed that the frogs behaved as though they saw the world upside down through the rotated eye. Whereas Sperry had had to assume the anatomy from the frogs’ behaviour Gaze used electrophysiological recording to directly examine the precise anatomy – the exact pattern of functional connexions between the retina (eye) and the tectum (the visual part of the frog brain with which the retinal fibres connect).[5] This was the first time that electrophysiological techniques had been used to examine nerve regeneration. These initial results showed that Sperry had been correct in his anatomical assumptions.
On the basis of his experiments, Sperry proposed the chemoaffinity theory (what he then called the theory of neuronal specificity).[6][7] Sperry proposed that every retinal fibre has a unique chemical label and every part of the tectum has a unique label. According to Sperry’s theory there is a rigid rule (the mapping rule) that determines the connexions made: each retinal fibre label can connect with one, and only one, tectal label. A central feature of the theory is that there is no interaction between retinal fibres themselves in forming the pattern of connexions.
In a number of experiments Gaze and his colleagues showed that Sperry’s mapping rule could not account for all the results.[1] The first such experiment was published in 1963.[8] György Székely had developed a method of creating ‘compound’ eyes at embryological stages in frogs. Such eyes were made from putting together two half-eyes from two different eyes. For example a compound eye might be the result of fusing two front (nasal) half-eyes. Gaze and his colleagues showed that the retinal fibres from each half eye covered the whole of the tectum (rather than being limited to the half tectum that normally received fibres from the nasal half eye). Gaze and his colleagues went on to explore the pattern of regenerated connexions between retina and tectum after removal of part of the retina, or part of the tectum, or part of both retina and tectum. In all cases, in time, the whole of whatever remained of the retina connected in order with the whole of whatever remained of the tectum.[9][10]
Gaze went on to explore the growth of retina and tectum during normal development. He showed that the retina grows concentrically; the tectum grows linearly.[11][12] During this period of growth the retinal fibres form functional connexions with the tectum and the frog can see. Gaze showed that the connexions between the eye and the brain must be forming and breaking, and then new connexions made, in a continual process. In other words, during development the retinal fibres are constantly ‘sliding’, making and breaking functional connexions with the tectum.[13] In 1972 ‘this idea was quite revolutionary’.[14] Despite this ‘plasticity’ the projection from eye to brain was always ordered: adjacent parts of the retina connected with adjacent parts of the tectum. The ‘mapping rule’ appeared to be less rigid than that envisaged by Sperry: not cell-to-cell but system-to-system. To describe this process Gaze used the term ‘systems matching’.[15]
The visual fields through the two eyes in frogs overlap to some extent. That is, frogs have binocular vision. The retinal fibres from the left eye connect with cells in the right tectum (and vice versa). The nerve impulse is then passed on to further cells in the tectum. Some of these further cells connect with specific points in the other tectum. Thus there are parts of the tectum that receive a nerve impulse directly from the opposite (contralateral) eye, and an impulse indirectly from the other (ipsilateral) eye. In the normal animal both these impulses are stimulated by an object in the same part of the visual field. Gaze and his colleagues observed that in animals with one compound eye the ipsilateral projection through the normal eye was frequently abnormal.[16] Gaze’s student, and long-term collaborator, Mike Keating, suggested a possible theory to account for these findings.[17] He proposed that the connexions from one tectum to the other were not the result of processes such as chemoaffinity or systems matching. Instead the connexions are formed by linking those nerves with similar spatiotemporal excitation patterns: what he called the functional hypothesis. Gaze and colleagues tested this hypothesis in a variety of situations and found that it accounted for all the results.[18] Although Hebb had proposed that functional interaction is crucial in learning,[19] and Hubel and Wiesel had shown that it plays a role in the preservation of binocular vision in cats,[20] it had not previously been realised that functional interaction is important in the formation of precise nerve connexions. The degree of binocular overlap of the visual fields in frogs changes as the frog develops and the eyes move further apart. This results in the connexions between the tecta, that are determined by the functional hypothesis, changing. These connexions, like those of the connexions between the eye and its contralateral tectum, are ‘sliding’ during normal development.[18]
