Memory retention through metamorphosis

Verworn cited Kowalevsky's work on insects to show that leucocytes take part in metamorphosis. In the fly larva, organs such as the creeping muscles become superfluous and are broken down, and Verworn held that the chemotaxis of leucocytes plays an important role in the development of many animals.¹ Loeb asked what coordinates metamorphosis. In Amblystoma larvae with the spinal cord cut at any level, the gills at head and tail still changed together. He concluded that the central nervous system is not the regulator and that the blood or substances in it must be.² Child noted that where metamorphosis builds many new parts, larval stages may show a high degree of mosaic behavior.³

Blackiston, Silva Casey and Weiss (2008) began from the idea that, if synaptic connections persist through metamorphosis, carryover of larval memory might depend on when the larval experience occurred. In Drosophila, mushroom body neurons pruned before pupation form early in larval development, and those that persist form later. Later experience might therefore be kept and early experience lost.⁴ In M. sexta, adults from larvae trained at fifth instar recalled their larval experience. Adults from larvae trained at third instar did not.⁵ The authors noted that M. sexta passes through five instars before pupation and Drosophila through three, so third instar training probably comes before the neurogenesis of the α′/β′ lobe.⁵

The moth interpretation leans on fly anatomy. Blackiston and colleagues wrote that Drosophila is the only holometabolous insect in which individual mushroom body neurons have been tracked through metamorphosis. The γ lobe forms in the embryo, the α′/β′ lobe just before pupation, and the α/β lobe from the onset of pupation.⁵ During pupation, γ lobe neurons are pruned to the main process before adult-specific projections form. The α′/β′ neurons keep intact projections throughout.⁵ The authors called their moth findings consistent with this picture if moth mushroom body development is analogous to the fly's. The match is an inference by analogy.⁵

Blackiston and colleagues (2010) dedicated an automated training device paper to Inna M. Sheiman and colleagues for pioneering work on memory and its dynamics during metamorphosis.⁶ Sengupta, Stemmler and Friston (2013) suggested that free-energy optimisation might explain synaptic pruning in the brain during neurodevelopment.⁷ Kirchhoff and colleagues (2018) described the pupa, in which most of the caterpillar's body breaks down to a mass of undifferentiated cells, as an example of a malleable boundary.⁸ Pio-Lopez and colleagues (2023) observed that studies of caterpillars usually treat the separate phases of life and seldom study the transitional states.⁹