Galling insects are plant parasites. The gall is an organ formed by the plant under the control of the galling organism, which may include viruses, fungi, bacteria, nematodes, and mites. Galling is ancient. At least ten insect lineages have independently evolved the ability to manipulate their host’s development to provide shelter and nutritionally enhanced food. Unlike microbial galls, insect galls are often as highly organized and complex as normal plant organs and are not simply neoplasms. And unlike some microbial gallers (e.g., Agrobacterium), insects do not genetically transform plant cells. Removing the insect halts gall growth and development. Insects direct gall development chemically, but how this is done and what the chemical cues are remain mysteries.
Agrobacterium tumefaciens gall.
Gall elicitation by insects is highly species-specific and includes the only known examples of gene-for-gene interactions between insect and plant. Gall morphology, tissue, and host plant species are determined primarily by the insect’s genotype and can be used to identify a galling insect to species.
As Darwin pointed out, the development and design of insect galls are convergent on plant reproductive organs. Some galls can be described as superficially ‘petaloid’, ‘pistillate’, ‘carpeloid’, and many are strikingly fruit-like; taxonomists have even mistaken galls for fruits. They are usually strong sinks like developing fruits. They have meristems from which growth and development proceed, they have highly organized and specialized cell and tissue types, including storage, transfer, and vascular tissues, and they may have biochemistries specific to fruits or seeds.
Both galls and flower carpels/gynoecia are structures that envelope, protect, and provide specialized food for an ‘alien’ organism, embryo or insect. Typically, nutritive cells rich in carbohydrates and proteins are consumed by the insect, just as endosperm provides nutrition to embryos. The nutritive tissue accumulates protein and carbohydrate, is often polyploid, and is supplied by transfer cells. At least one gall’s nutritive layer includes proteins normally found only in the seed. Growth and regeneration of the nutritive layer depend on chemical cues from the insect, much as embryos manage surrounding tissues hormonally. At least one galler maintains normal endosperm development in the absence of the plant embryo. Defensive chemistry is sequestered away from the galling insect as it is from embryos. Rolled leaf edges are considered ancestral elements in the evolution of some gall lineages and in the evolution of the carpel. Gall traits and development suggest that galling insects appropriate and elicit expression of organ developmental programs ectopically.
We hypothesize that galls form on vegetative organs (leaves) by appropriation and ectopic expression of flower/fruit development genes.
Testing this hypothesis
If gall development and flower or fruit development have common elements, we should see this in shared expression of genes as they originate and grow. We are using RNAseq to profile gene expression in developing galls produced by phylloxera (Daktulosphaira vitifoliae) on grape leaves and comparing those profiles with those of developmentally-matched leaves, and with those of developing flowers. We expect to see greater or exclusive expression of genes involved in reproduction in the galls compared with leaves. Reproductive genes should include those involved in the transition from vegetative to reproductive development, origination and identity of floral organs, and development of the gynoecium (from which fruit would normally elaborate).
Leaf gall on the wild grape, Vitis riparia, induced by the Phylloxera aphid, Daktulosphaira vitifoliae.
Finding reproductive genes ectopically expressed in galls on leaves comprises an association between galling and flowering, but does not comprise proof that galling depends on ectopic expression of those genes. To determine whether galling requires expression of reproductive genes, we are collaborating with Dennis Gray at the University of Florida (LINK) who is developing transgenic grapevines in which expression of key genes is modified (up or down). Dennis will also produce reporter constructs to elucidate the micro-genetic anatomy of gall development. Establishing infestations on vines is easy, so we expect to see galling fail on plants with modified expression of key reproductive genes.
Cross-section of a leaf gall on the wild grape, Vitis riparia, induced by the Phylloxera aphid, Daktulosphaira vitifoliae (stage 4).
What we’ve learned so far
In preliminary experiments, Dr. Sarah Melissa Witiak (LINK) used rtqPCR to focus on selected flowering genes, such as the Vitis vinifera orthologs of the Arabidopsis LEAFY (LFY), AGAMOUS (AG), SEEDSTICK (STK), SHATTERPROOF (SHP), and others. She found LFY, AG and SHP elevated in galls compared with leaves. Besides being limited to a few targeted genes, this approach suffered from difficulty is sampling gall tissue without surrounding leaf tissue, which dilutes and alters the apparent expression pattern of presumptive gall cells.
Leaf gall on the wild grape, Vitis riparia, induced by the Phylloxera aphid, Daktulosphaira vitifoliae at different developmental stages.
What lies ahead
Current work employs RNAseq to develop a picture of transcription during gall development. We expect to find expression of genes involved in reproduction statistically over-represented in this data set. The number and identity of these genes should tell us to what extent a gall resembles flowers or fruits compared with leaves, and may provide clues to the signals involved in eliciting galls.
Once we have finished interpreting transcriptome results in the context of reproductive development, we will return to the data to see what gene expression patterns can tell us about chemical signals – hormones – that may be involved. The literature is replete with demonstration that levels of auxin, cytokinins, gibberellins, abscisic acid, and jasmonate in galls differ from those in ungalled tissues. We plan chemical analyses of hormone concentrations to link with gene expression evidence. These patterns may tell us something about the insect’s elicitors. For example, elevated expression of hormone-responsive genes at the same time as depressed expression of biosynthesis genes may suggest that the given hormone is being supplied by the insect. We plan to sample insect heads and/or saliva as well. Eventually we will attempt to manipulate activity of key hormones pharmacologically during insect bioassays.
We are passing sequences of key reproductive genes to the Gray lab for production of transgenics. When those are established, we will be doing bioassays with the insects. Plants with modified expression of LFY have been created and should be ready for bioassay trials sometime in 2015.
People involved in this research topic
Chris Pires, University of Missouri
Patrick Edger, University of Missouri
Michele Warmund, University of Missouri
Chung-Ho Lin, University of Missouri
Dennis Gray, University of Florida
Trudi Grant, University of Florida