Uncoupling of sequential heteromorphic developmental programs.

The regulatory basis for differences among species in the developmental rate at which successive life stages are reached ("heterochrony") is a subject of much controversy among vertebrate and invertebrate developmental biologists. The heterochrony in development of different insect species is characterized in part by the intercalcation between the embryo and adult of a (varied) number of heteromorphic larval instars. These heteromorphic larval instars exhibit changes of body form from one larval instar to the next, prior to the final metamorphic molt to the pupal form. The intractability of larval heteromorphosis to experimental dissection is due in part to the lack of suitable experimental probes that can test the nature of the coupling of each heterochronically expressed instar-specific program. The epistatic basis of expression of heteromorphic developmental programs was assessed by two-dimensional electrophoretic analysis of hemolymph proteins during the normal and experimentally manipulated feeding stages of the 3rd, 4th, and 5th (final) instar larvae, and during the prepupal stage, of Trichoplusia ni. Of the several hundred protein spots tracked, some were identified that were uniquely detected during a single stage, while others were observed during combinations of certain stages. The nature of coupling of sequential heterochronic expression of these proteins during successive instars or stages was tested by use of a parasite (Chelonus sp.) that injects regulatory material into the host embryo that later causes the subsequent precocious expression of the final instar larval program. Following the expression of a normal 3rd instar pattern, such larvae were observed to omit expression of the 4th instar program, including omission of the proteins heteromorphically specific to that instar, and instead then express an essentially normal final instar pattern. Thus, normal expression of the final instar feeding stage pattern was not invariantly coupled to prior expression of the penultimate instar-specific proteins or pattern. Also, expression of the full program of the final instar feeding stage was epistatic to teh penultimate instar program, i.e., the protein pattern unique to the penultimate larval instar was not co-expressed with the precociously expressed final instar pattern. Larvae developmentally redirected in this manner failed to fully express the final instar prepupal stage pattern of protein expression, due at least in part to failed expression of prepupal ecdysteroids, but this was shown not to arise from omission of any of the first 4 larval instars per se. The nature of the redirections in host development caused by this parasite finally provides means of probing the coupling of successive expression on heteromorphic programs during larval development.

Preultimate 4th/5th instar Trichoplusia ni naturally-injected with venom/calyx fluid from Chelonus curvimaculatus precociously metamorphose, rather than obey the metamorphic size threshold that would normally compel molting to a 5th/6th instar.

In normally regulated larval metamorphosis of Trichoplusia ni, a 4th, 5th or other numbered instar is a 'preultimate' instar, and will normally continue larval molting, if the larva has not yet surpassed the critical (minimal) size threshold corresponding to attainment of the 'ultimate' (metamorphic) instar. Natural injection of T. ni embryos with venom/calyx fluid of female Chelonus sp. near curvimaculatus caused 'penultimate' 4th or 5th instar larvae that would normally molt at least once more, to a 5th/6th instar, to instead precociously metamorphose without another larval molt. These effects were observed in naturally-injected insects that never contained either a parasite larva, a viable parasite embryo, or a parasite egg. These data demonstrate that this effect of venom/calyx fluid of this wasp to induce precocious metamorphosis, at an instar earlier than would otherwise have typically occurred under normal growth conditions, does not require the presence of a parasite larva. Other data did indicate the parasite larva contributes an additional effect that causes a 4th instar host (that from its size would normally require not just one, but at least two more larval molts to reach the metamorphic instar) to not grow to the size metamorphic threshold, but to instead, precociously metamorphose at an even smaller size than occurs with the venom/calyx fluid alone. Additionally, arylphorin was precociously highly expressed in parasitized hosts in a manner independent of a decline in the host JH titer. Therefore, the main target of the venom/calyx fluid activity to induce precocious metamorphosis appears to be an event upstream of the decline in JH production by the corpora allata. Pseudoparasitized hosts become developmentally arrested as precocious prepupae and express a 2.7kb polydnavirus transcript. The larger (but still subthreshold size) larvae showed less suppressed prepupal ecdysteroid titers, less developmental suppression, and a much weaker expression of that transcript. A general model for mechanisms of action of chelonine venom/calyx fluid, and larvae, to cause precocious host metamorphosis and suppressed prepupal development is presented that is based on the current 'size threshold' model of normal lepidopteran development, rather than the older, displaced 'instar count' model. By basing the model for chelonine regulation of host development on the current 'size threshold' model for normal development, the proposed model for chelonine action both accounts for observations reported on various species of that subfamily and makes useful, testable predictions.

Toxins produced by arthropod parasites: salivary gland proteins of human body lice and venom proteins of chelonine wasps.

A review is presented of our ongoing research projects on the protein components of the saliva of human body lice and of the non-paralyzing venom of wasps in the subfamily Cheloninae. Sodium dodecyl sulfate-polyacryamide gel electrophoretic analysis of lice salivary gland proteins showed a predominance of high and intermediate mol. wt proteins. Immunoblotting with a low titer polyclonal antiserum to lice salivary proteins indicated that some, but not all, of the predominant high mol. wt salivary gland proteins are injected into the host during feeding. The venom of a Chelonus sp. wasp contains a chitinase, and a 33,000 mol. wt protein with a primary structure composed mostly of a series of 12 tandem repeats of a 14-residue sequence. The N-terminus of this protein and its homologs in a related species of Ascogaster share a conserved adjacent pair of acidic residues. Epitope mapping/immunoprecipitation experiments now in progress will provide information on which linear motifs are on the surface of the protein, and will thereby provide information on the tertiary structure of the protein.