Ecological Tools Developed for Nicotiana attenuata

Biotic and abiotic stresses influence plant performances under natural conditions, conditions which are difficult to simulate when plants are grown in the glasshouse. Our research program examines the influence of altering the expression of key genes on plant performance in natural and semi-natural habitats (see field site at Brigham Young University's Lytle Ranch Preserve), under simulated natural conditions (tent) as well as under standardized glasshouse-conditions and even the Hansson Department’s  wind tunnels of the Schneider house of our institute (Department of Evolutionary Neuroethology) . For most of the transformants, the plants lack visible phenotypes when grown in the glasshouse in single pots, but show clear phenotypes when they are planted back into their natural habitats in Utah, where we can use the behavior of all the different organisms that live from their interactions with this host plant and are therefore better at phenotyping N. attenuata plants than humans are, and thereby uncover the function of the genes that we study.

Semi natural
Semi natural
Glasshouse conditions
Glasshouse conditions
Semi natural
Glasshouse conditions

Our facilities are state of the art and offer the maximum bandwidth for conducting “deep phenotyping” of transformed plants. By conducting these experiments under natural, semi natural and glasshouse conditions, we can cover the entire spectrum of environmental control, from highly controlled conditions which are ideal for the quantification of subtle changes in growth to environments that contain the full panoply of environmental challenges (abiotic and biotic) as well as all of the complexities of seasonal changes that occur in the rough and tumble of nature (Utah planting procedure Video). Our glasshouse was designed by Baldwin to provide ideal conditions for the study of plant-herbivore interactions, which due to the particular whole-plant nature of their responses to insect attack, is designed to produce highly uniform growth conditions so as to minimize variance amongst replicate plants. Specifically, high pressure lamps affixed to light movers supplement the natural light to ensure equal light conditions at each position across all glasshouse tables, which in turn are watered by an ebb-and-flood automatic watering system, and a heating/cooling system which ensures that all plants are not exposed to any thigmomorphogenic stimulations from wind.


Plant performance in the native habitat - herbivory

Plants in the field may be attacked by a large number of different herbivores. It is essential to determine which herbivore species are responsible for the observed damage. By training students to use standardized protocols for herbivore screens, we are able to correlate herbivore abundance as well as herbivore-specific damage estimates with gene expression in paired wildtype and gene-silenced plants  (link to Molecular platform) to understand the genes which allow this native plant to cope with its enemies, competitors, and manage its mutualists. We developed an easy-to-use database for this purpose, which continues to evolve with the two decades of fieldwork with this plant. Each season we perform natural history observations, which lead to the discovery of new species which interact with N. attenuata. This large and continually evolving array of possible interactions with N. attenuata is one of the most important ways in which we conduct “deep phenotyping” of this plant.

First row from left to right: tobacco hornworm (Manduca sexta), beet armyworm (Spodoptera exigua), tobacco budworm (Heliothis virescence), tomato hornworm (Manduca quinquemaculata); second row: suckfly (Tupiocoris notatus), potatoe leafhopper (Empoasca fabae),tobacco flea beetle(Epitrix hirtipennis), snowy tree cricket (Oecanthus fultoni); third row: tobacco stalk borer (Trichobaris mucorea), negro bug (Corimelaena extensa), tobacco aphid (Myzus nicotianae), spotted cucumber beetle (Diabrotica undecimpunctata)

Plant-microbial interactions in the field

Nicotiana attenuata plants occasionally showed sudden wilt symptoms in the field for which pathogenic fungi belonging to Alternaria and Fusarium genera found to be responsible as they were found with high abundance in the roots of diseased plants (reference to Stefan Schuck’s paper). We used a library containing hundreds of bacterial strains isolated from field grown N. attenuata plants (reference Long’s paper) for re-inoculation experiments in the glasshouse to study plant growth promoting effects but also the efficiency of bacteria to be used as biocontrols. The re-inoculation with a core-set of five bacterial strains showed protective effects against a wilting disease occurring on the field plot. Currently we are working with Bacillus megaterium (B55), Bacillus mojavensis (k1), Arthrobacter nitroguajacolicus (E46), Pseudomonas azotoformans (A70), and Pseudomonas frederiksbergensis (A176) (reference Rakesh et al in preparation). To explore the influence of N. attenuata´s microbiome under natural conditions we use transgenic plants expressing small cationic antimicrobial peptides (AMPs) (Arne’s paper in preparation). We measures plant growth, plant fitness and analyze root-associated bacterial communities using high throughput pyrosequencing. Furthermore we investigate mycorrhizal symbiotic interactions using lines lacking the calcium/calmodulin-dependent protein kinase (CCaMK), an important element in the arbuscular mycorrhiza fungi sym-signaling pathway (reference Groten et al).

Plant growth measurements in the field

We have standardized growth measures of plants (above and below ground) in the field and use procedures to alter light conditions, such as UVB (UV-B opaque films) or far-red (LED clusters) radiation, designed for both above and below ground illumination. Furthermore we are able to mimic different competitive situations by using different planting designs, in which resources, such as water and nutrients are delivered into the competition arenas in spatially discrete areas. In addition to above-ground phenotyping, we also examine below-ground growth, as well as defense responses. We use different methods to investigate root structure, as well as the responses of the plant’s native bacterial and fungal communities. Even below-ground we are capable of using far red treatments to evaluate signal transduction between roots and shoot. We developed a micrografting method which allows us to graft shoots and roots of different transformed plants together during the seedling stage. Recently we micrografted wildtype N. attenuata shoots to the roots of transgenic plants impaired in jasmonic acid signaling and evaluated the regulation of herbivore-induced shoot metabolites in the roots (Fragoso et al NEW PHYTOLOGIST 2014). Another valuable tool which we have in hand are dexamethasone (DEX) inducible lines (Molecular platform) to perform real-time genetic manipulation (silencing or overexpression of certain genes) in the field. This technique allows the expression of genes to be specifically altered in certain tissues at particular times, which is particularly advantageous when studying for example developmental changes in gene expression or tissue-specific direct/ indirect defenses within a single plant (Schäfer et al TPJ 2013).

UV-B opaque boxes
Far red treatment
Circle planting design

Volatile organic compound profiling in the field

Volatile organic compounds emitted from leaves or flowers can be collected dynamically on charcoal traps with a pump or statically, with PDMS tubing, which allows for well replicated kinetic and spatially explicit sampling. These traps are analyzed by thermal desorbtion GC-q-MS (link to Analytical platform). Real time measurements of GLVs and floral volatiles under field conditions are also routinely performed with a portable Z-nose GC detector.

Charcoal flower trapping
Dynamic charcoal trap design
Silicone tubing setup

Predation assay

Caterpillar attacked plants emit volatiles which attract predacious big-eyed-bugs. To quantify how predation rates are influenced by volatile emissions, we glue Manduca sexta eggs on leaves and quantify their rate of predation by the native populations of Geocoris spp. (Fig. 3a; Kessler & Baldwin SCIENCE 2001; Allmann & Baldwin SCIENCE 2010). In addition, caterpillars of M. sexta fed different N. attenuata transformants that influence the body and frass odor, or alkaloid content of the larvae are used in feeding assays with diverse predators, such as ants or lizards (Fig.3b). Furthermore scent addition experiments are used in conjunction with choice assays to unravel the diversity of tritrophic interactions mediated by volatiles, which plants as well as herbivores use to manipulate the outcome of their interactions. For example metabolites of glandular trichomes, the first meal of caterpillars, were found to impart a distinct volatile profile to the body and frass of larvae which attracted predacious ants (Pogonomyrmex rugosus; Weinhold & Baldwin PNAS 2011) and possible also lizards (Stork , Plant Signaling and Behavior 2012).

From left to right: 1. Egg predation assay. M. sexta eggs are glued on the underside of a leaf to quantify the attractive function of volatiles induced by herbivore damage on predacious Geocoris big eyed bugs. 2. Lizard predation assay: M. sexta caterpillars fed on N. attenuata lines altered in genes changing the detoxification of alkaloids or glycosides in the midgut of the caterpillar (plant mediated RNAi) are used in choice assays with lizards (Aspidoscelis tigris) 3. Caterpillar perfuming assay. Reconstituted blends of the M. sexta caterpillar body are used to measure ant attraction.

Plant mediated RNAi in the field

We use this revolutionary reverse genetics approach to silence genes in native insects by feeding insects N. attenuata plants transformed to express large quantities of double stranded fragments of the insect genes of interest, which when ingested, silence the cognate gene in free ranging insects feeding on these plants (Kumar et al PLOS ONE 2014). This procedure was used in the field (Kumar et al PNAS 2014) to uncover the role of a midgut expressed P450 gene in the repurposing of ingested nicotine by the caterpillar into a defensive halitosis that protects larvae from attack by predaceous spiders. We have also recently used the procedure in both field and glasshouse experiments to silence a M. sexta midgut expressed β-glucosidase 1 which plays a central role in diterpene glycoside detoxification. The malonylated forms of lyciumoside IV, Nicotianoside I and II from N. attenuata leaves is demalonylated by the alkalinity of larval oral secretion to form lyciumoside IV. Further it is deglycosylated by β-glucosidase 1 which is expressed in the midgut of M. sexta caterpillars. When this β-glucosidase 1 was silenced by PMRi, larvae were impaired in lyciumoside IV deglycosylation and lyciumoside IV was accumulated in their body. Such larvae were unpalatable to the native predatory spiders, suggesting that defensive co-option of lyciumoside IV could be ecologically advantageous to M.sexta larvae.

Genetic variation in native populations

We regularly search for different N. attenuata phenotypes in native populations focusing on morphological differences and also on differences in metabolism. For the latter, we use native herbivore species to identify natural variation in signaling or secondary metabolites, in essence using insects as "bloodhounds" to ferret out genetic variation hidden within native Nicotiana attenuata populations. As an example, see our recent PNAS publication in which we used native Empoasca leaf hoppers to identify jasmonate mutants in native populations (Kallenbach et al PNAS 2012). Other insects such as the Trichobaris weevils seem to prefer natural clock mutants.

Photosynthetic capacity

We developed methods to collect data on photosynthetic rates (IRGA), chlorophyll content (chlorophyll-meter), and the efficiency of Photosystem II (fluorescence-meter), under field conditions. Measuring photosynthetic activity (such as photosynthesis, stomatal conductance, and chlorophyll) especially in clock silenced lines is a commonly used tool. These measurements allow visualizing the ability of plants altered in their internal clock to respond for example to light cues in the middle of the night.

Photosynthetic rate measurements in a native N. attenuata population and on the plot

Nectar removal and pollination rates

A standardized protocol is available to measure nectar volume and nectar sugar concentration in flowers of N. attenuata. Nectar removal provides a measure of nectar quality or floral attraction. To quantify pollen transfer, we antherectomize flowers (removal of anthers) just before anthesis, and subsequently expose these flowers to either night or day-active pollinators. The resulting capsule set quantifies outcrossing rates, which can be further quantified by genotyping by microsatellite analysis (Molecular platform) to reveal the number of different paternal genotypes that have sired seeds. This approach allows us to evaluate the function of individual floral traits, such as floral scent or nectar, by directly measuring the consequences for plant fitness (Kessler et al SCIENCE 2008) of a lacking a particular trait. In addition we collect behavioral data on floral visitors which are not limited to mutualistic pollinators, but also include antagonists such as nectar robbing carpenter bees (Xylocopa spp.) or florivores, such as cucumber beetles (Diabrotica undecimpunctata), tree crickets (Oecanthus fultoni), or the tobacco budworm (Heliothis virescence), which frequently damage flowers. In short, by using plants altered in a specific floral trait in the field, we can investigate the whole community of floral visitors at the same time, measure fitness consequences, and thus draw conclusions about the function and evolution of particular floral traits.

Anther removal
Hummingbird pollination
Mesh cones to exclude day pollinators

Video observation

Different sophisticated camera systems allow for the observation of the herbivore community as well as floral visitors of different plants at the same time. Passive infrared sensors provide the opportunity to observe pollinator and herbivore interactions during twilight and night. We use high resolution cameras for close ups, as well as numerous webcams for quantifying visitations and behavior of floral visitors, such as pollinators or nectar robbers. A special developed “in-flower” video setup enables us to record a moth’s behavior not only on the flower, but also inside the flower which allows the movement of the proboscis inside the flower to be investigated. In addition, we use a Gigapan photo stitch systems to create high-resolution images to qualitatively and quantitatively characterize growth and the movement of different plant organs over time and observe herbivore movement over a range of spatial scales.

Field video observation
Video image at dusk
Video setup imaging in-flower proboscis behavior

Choice assays

Fitness data, as well as observational data are frequently supplemented with inferences drawn from simple choice tests. For pollinators, we use sets of artificial flowers filled with sugar solutions, which we spike with relevant concentrations of compounds found in nectar. We use standardized perfuming experiments in native populations to test the function of individual leaf volatiles, or use leaves or flowers of our transformed lines in choice assays with WT or EV plants with native florivores, herbivores or even pollinators.

Hummingbird assay
Ant nectar (robbing) assay
Cotton swap perfuming


Pollination and oviposition assays in the Isserstedt tent

We have the possibility to conduct pollination and oviposition experiments in a large tent (8x24m) in Isserstedt (picture one), a village in the proximity of Jena. Its mesh cover allows for continues air exchange. The tent enables hawkmoths to fly and behave almost as they would in their native habitats in Utah. For example, nectaring and oviposition behavior of adult Manduca moths is comparable between experiments conducted at the field station and ones conducted in the tent. The advantage of the tent is the ability of conducting pollination trials with single pollinator species, something which is challenging to do in the field.

M. sexta oviposition
M. sexta pollination
H. lineata pollination


Glasshouse feeding assays

To evaluate the ecological significance of N. attenuata genotypes, standardized feeding assays are conducted. In addition to caterpillars from Manduca sexta, Spodoptera littoralis, Spodoptera exigua, or Heliothis virescens moths, we maintain colonies of myrids (Tupiocoris notatus), aphids (Myzus nicotianae), leaf hoppers (Epitrix fabae), flea beetles (Epitrix hirtipennis), stalk boring weevils (Trichobaris mucorea), and seed feeding negro bugs (Corimelena extensa) all of which can be used as indicator for food quality of a plant. We conduct growth measurements (caterpillars), reproduction assays (aphids), and choice tests with all of these insects. Additionally Waldbauer nutritional assays are used to obtain detailed data on food assimilation (Rayapuram and Baldwin Plant, Cell, Environment 2006; SOM) processing and growth as well as quantitative measures of secondary metabolite flux through the insects. Furthermore, we have a experimental mesocosm arena situated in the glasshouse in Isserstedt that bridges the gap between field and glasshouse studies, allowing experiments to be conducted on controlled communities in a semi-natural environment. The mesocosm comprises 12 cylinders filled with soil or other substrate and independently watered by an automated system; the aboveground area is surrounded by a fine steel mesh which permits light penetration and gas exchange, and free movement of insects within, but not outside of the mesocosm. Each cylinder can accommodate populations of up to 8 Nicotiana attenuata plants. During experiments, the surface between cylinders is also covered with soil or substrate to provide an appropriate surface for the movement of insects between populations. The mesocosm projects runs in close collaboration with the German Centre for Integrative Biodiversity Research (IDIV).

Mesocosm conceptual overview
Mesocosm population

Glasshouse predation assays

We maintain colonies of big eyed bugs (Geocoris pallens), as well as predatory wolf spiders (Camptocosa parallela), which allows us to perform predation assays even under glasshouse conditions. We use predation assays similar to those used in the field with Manduca eggs, and extend these by smaller experimental arenas in which we use paired plants or y-tube designs to investigate the attractiveness of certain volatiles for Geocoris. Wolf spiders in contrast are used to test “caterpillar quality”. These spiders are sensitive to caterpillars containing toxic compounds (e.g. nicotine or diterpene glycosides). Caterpillar quality is altered by using different larval diets, such as transformed N. attenuata plants or artificial diets supplemented with particular secondary metabolites, or by changing a caterpillar’s ability to metabolize certain compounds by using plant mediated RNAi (Kumar et al PNAS 2014).

From left to right: 1. Spider predation assay. Wolfspiders are allowed to choose between two M. sexta caterpillars feed on different diets. 2. Plant competition assay. Different insect species are allowed to choose between two plant phenotypes which are located in a mesh box. 3. Nine different insect colonies are available for the use in feeding or choice assays, as well as predation assays (see herbivores of N. attenuata on top of page)

Glasshouse growth analysis

In addition to the traditional measurements of rosette diameter, stalk length, or flower number, we use photographic plant growth analysis which allows for high throughput measurements, which can provide additional detailed data on the dynamics of plant growth. For this, we collaborate with the Leibniz-Institut für Pflanzengenetik und Kulturpflanzenforschung (IPK) in Gatersleben to perform high throughput plant phenotyping with an indoor Lemnatec system. In addition to the aboveground plant parts, we can measure root growth continuously by using a sophisticated scanner system.

From left to right: 1. Video monitoring of single flowers to investigate differences in flower opening and movement in clock altered lines. 2. Description of floral phenotypes: Flower opening kinetic for night opening flowers (upper row) and morning opening flowers (lower row, Kessler et al. 2010)

Wind tunnel

We have the possibility of phenotyping plants in the wind tunnel of the Department of Evolutionary Neuroethology. This advanced technique allows us to conduct choice assays with moths such as M. sexta or H. lineata. We are able to measure the moth's behavioral response to attractive traits between two plants, two odor sources (either living plants or artificial scent compounds) or even perfuming one plant with the scent of another, with a sophisticated air pump and channeling system. A permanent airflow within the tunnel guarantees a permanent air exchange and a directed movement of the odor.