Source: PURDUE UNIVERSITY submitted to
A MODEL PLANT-SOIL SYSTEM FOR ASSESSING GENETIC BIOCONTAINMENT AND ITS USE IN EVALUATING THE EVASION RISK OF A `KILL SWITCH` CIRCUIT
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
ACTIVE
Funding Source
OTHER GRANTS
Reporting Frequency
Annual
Accession No.
1030924
Grant No.
2023-33522-40410
Cumulative Award Amt.
$640,897.00
Proposal No.
2023-02446
Multistate No.
(N/A)
Project Start Date
Jul 1, 2023
Project End Date
Jun 30, 2027
Grant Year
2023
Program Code
[HX]- Biotechnology Risk Assessment
Project Director
Wilhelm, R. C.
Recipient Organization
PURDUE UNIVERSITY
(N/A)
WEST LAFAYETTE,IN 47907
Performing Department
(N/A)
Non Technical Summary
We propose new methods to evaluate genetic containment strategies for engineered microorganisms (GEMs) in plants and soils, fulfilling the program goal to evaluate management, monitoring, and mitigation methodologies. Genetic containment of cell viability ('kill switch' circuits) offer safeguards against the uncontrolled proliferation of GEMs outside their intended use. The risk of evasion depends on multiple interacting environmental and ecological factors that influence kill switch efficacy and the fitness of survivor populations. Therefore, an evaluation of evasion risk requires experimentation and modeling in naturalistic settings that capture these dynamics. We propose to develop a model system to evaluate the efficacy of kill switches using a model legume and bacterial symbiont, where infection and nodule formation is used to assess risk. Our first objective is to establish the molecular methods needed to quantify kill switch efficacy and the prevalence of escape mutants in soil and plant samples. Subsequently, we will use the model to study the factors impacting evasion risk in different plant and soil compartments (objective 2), and the relative importance of environmental stress, resources, and microbial competition on evasion risk (objective 3). Our aim is to link evasion risk to factors relevant to management practices, such as nutrient management and soil type, to inform the use of kill switches in agroecosystems. Overall, our project will provide researchers and bioengineers with a model for risk assessment to validate new genetic containment strategies and advance our understanding of the environmental and ecological factors impacting containment.
Animal Health Component
50%
Research Effort Categories
Basic
40%
Applied
50%
Developmental
10%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
1025299106030%
1025299104035%
1025299107035%
Knowledge Area
102 - Soil, Plant, Water, Nutrient Relationships;

Subject Of Investigation
5299 - Agricultural supplies, general/other;

Field Of Science
1040 - Molecular biology; 1060 - Biology (whole systems); 1070 - Ecology;
Goals / Objectives
Genetic controls on the viability of genetically engineered microbes (GEM) offer safeguards against uncontrolled proliferation and represent an essential layer of safety as part of a containment strategy. However, the efficacy of genetic controls, specifically the efficacy of kill switch circuits, has yet been demonstrated in naturalistic settings, where environmental and ecological factors can cause GEMs to evade controls. As a result, we lack sufficient knowledge of the nature of these risks stemming from the factors impacting kill switch efficacy and the risk of uncontrolled proliferation in scenarios where kill switches fail. Our project will advance understanding of the risks posed by the variable efficacy of genetic containment in plant-soil contexts as it relates to the evasion and proliferation of GEMs. We will develop a model plant-soil system designed to evaluate the efficacy of various genetic containment strategies and use this system to test the factors in soil and plant compartments that impact the efficacy of kill switches deployed in an endophytic bacterium. We will develop tools to identify and model the risk factors associated with kill switch evasion, fulfilling the program needs to evaluate management, monitoring, and mitigation methodologies for the containment of GEMs in agroecosystems.The proposed research aims to develop and validate a model Phaseolus - Rhizobium system for assessing the evasion risk genetic containment strategies (Obj. 1). We will then evaluate the factors influencing the evasion risk of a toxin-antitoxin kill switch in plant and soil compartments (Obj. 2), and the relative importance of environmental and ecological factors (Obj. 3). Our overarching objectives are to better understand the environmental and ecological parameters influencing the probability of evasion. For this project, we will consider the uncontrolled proliferation of GEM in a successive generation of potted P. vulgaris following non-permissive conditions (i.e., containment) as the risk outcome.Objective 1. To establish standardized procedures and quantitative methods for assessing evasion risk in a Phaseolus-Rhizobium model system.Objective 2: To compare the efficacy of programmed death and the frequency and type of escape mutants in plant (nodule, root, leaf, and seed) and soil compartments (rhizosphere, soil leachate, and bulk soil).Objective 3: To assess the influence of environmental (resources and stress) and ecological drivers (competition) on evasion risk.
Project Methods
Obj. 1 setup. We will engineer two genetic elements into the genome of Rhizobium etli CFN42T to create a model endophytic GEM ('ReGEM'): (A) an inducible ReGEM specific marker ('ReMark') to measure the abundance of living cells, and (B) a CcdB-CcdA toxin-antitoxin kill switch circuit triggered at low temperatures. The ccdB-ccdA circuit ('controlled cell death'; ccd) is among the most common TAS used in kill switch design.?The engineering of R. etli will be performed with a modular cloning library that uses Golden Gate and Gibson (3G) assembly methods. Using a library of genetic components, we will construct a transcriptional unit with inducible promoter sequences to drive the expression of ccdB and ccdA. The first step will be to build and optimize the cryodeath circuit using exogenous chemical inducers. We will arrange the CcdB toxin to be constitutive expressed ('always on'), while the CcdA antitoxin genetic components to be constitutively expressed in the presence of aTc using an independent aTc-sensitive promoter (PTet). The cryodeath circuit will cause cell mortality unless the induction of CcdA recovers R. etli growth. The ratio of the CcdB and CcdA components determines the efficiency of the on-off dynamics of the TAS kill switch. We will optimize the dynamics of our circuit by screening a pool of RBS sequences with varying expression strengths upstream of each TAS gene. We will characterize the on/off dynamics of the TAS kill switch by titrating inducer concentrations and measuring optical density 700 and fluorescent readout in a plate reader.Obj. 2 setup. We will perform a potted plant experiment with a small cohort of plants. In brief, ReGEM will be inoculated into the system via seed stock and via a soil amendment at 2 weeks to ensure the colonization of both plant and soil compartments. The soil will be constructed using agricultural soil and sterilized. After growing to V7, half of the cohort will be exposed to non-permissive, cooler temperatures for 7 days, triggering the containment of ReGEM populations. Both cohorts will be grown for an additional 7 days in permissive conditions, to allow survivor populations to establish. The pots will be subdivided into three soil compartments (soil leachate, bulk soil, and rhizosphere soil) and four plant compartments (nodule, root, leaf, and seed) and sampled as described in section A.3. The remaining plant material will be cut into 1" pieces and incorporated into the top 5" of soil, and a second generation of seeds, this time axenic (surface sterilized), will be planted in the soil. Nodule formation by survivor populations will be monitored in a non-destructive manner using an X-ray CT root scanner.Obj. 3 setup. The experimental design will be identical to Obj. 2, except for the use of smaller ¼ gallon pots (~1 L) due to the increased sample number, and a shorter growing period, since we do not anticipate harvesting seed. A cohort of 48 plants (2 stress levels x 2 resource levels x 2 sterile/non-sterile soil x 2 temperature conditions x 3 reps) will be grown and exposed to one of two sets of treatment conditions. If above-ground compartments are assessed, the treatment factors will be phosphoruslimitation and UV exposure. Half of the plants ('low stress') will be grown in a chamber where lights have been sheathed in UV filter sheet to reduce UV radiation by 99%, while the other half grown in normal exposure ('high stress'). Plants will be further divided into two phosphorus fertilizer conditions: P-limiting (4 ? mg L-1 K2HPO4) or replete concentrations (8 mg ? L-1), according to prior methods. If below-ground compartments are assessed, the treatment factors will be soil texture and soil organic matter content. We selected soil texture to serve as a master variable over water retention to induce mild water stress. Half of the cohort will be potted in artificially constructed loam ('low stress') and loamy sand ('high stress') and inoculated with the same field soil. Soils will be further constructed with either ~ 0.5% ('low') or ~ 2.5% ('high resource') soil organic matter using an amendment compost.Inoculum preparation. All inoculum used in plant-pot experiments will be prepared from cells grow in 100 mL of medium in 250 mL Erlenmeyer flasks, shaking (225 rpm) at room temperature. Cell densities will be standardized using optical density measurements (OD600). All inoculum will be prepared from fresh cultures growing at mid-log phase growth and adjusted to a consistent OD with buffer (0.85% NaCl), corresponding to ~5 × 107 colony forming units (CFU), as described previously for legume-symbioses experiments.Plant growing conditions. Plants will be grown in 0.5-gallon (~ 2 L) pots with constructed soil filled to ~1" below the lip. For Obj. 2, soil will be constructed from a 1:1:1 mix of farmland soil, sand and perlite (bulk density ~1.1g cm-3) and autoclave sterilized, as previously described [76]. Soil will be sourced from a soybean field in an on-going field trial at the Purdue Agronomy Farm. For Obj. 3, if above-ground factors are assessed, the same soil will be used as Obj. 2. Where necessary, seeds of P. vulgaris will be surface sterilized with 5 % bleach solution. Three seeds will be sown per pot but will be thinned to one healthy plant. Plants will be grown in a growth chamber using a consistent light intensity (260 μmol ? m-2 ? s-1), temperature (28 day /20 night °C), photoperiod (14 / 10 h), and a relative humidity of ~60%. Plants will be fertigated for the first month with filter-sterilized 300 ppm 5-10-10 to encourage nodule formation, then 10-30-20. Plants will be sampled at V7 stage when the first seed pods have reached maturity (~63 das). Experiments will be set up in a blocked design to account for spatial variation in growth chambers. Plant performance will be measured by regular imaging of above- and below-ground biomass.Soil and plant sampling. Soils and plant tissues will be destructively harvested. Each plant pot will be subdivided into three soil compartments: soil leachate, bulk soil, and rhizosphere soil and four plant compartments: nodule, root, leaf, and seed. Sample from each compartment will be divided in half: one part for molecular analysis and the other reserved for repeat molecular analysis, chemical / physical analyses, or enrichment culturing. The sample for molecular analyses will be immediately homogenized in a buffer solution containing IPTG and incubated to induce ReMark expression. To release endophyte cells, plant tissues will be gently disrupted using tissue homogenizer on low for 30 s. Following the incubation period with IPTG (determined in Obj. 1.2.A), samples will be flash frozen on dry ice and stored at -80 °C. Each plant pot will be sampled as follows: Plants will be gently removed from pots and shaken to remove non-adhering soil back into the pot. Plant roots will be placed in a sterile bag containing IPTG buffer and vortexed for 1 min to remove adhering 'rhizosphere' soil, which will be portioned. Plant roots will be transferred to a sterile bag containing 5% bleach solution and agitated for 10 min, removed and severed from plant. Nodules will be gently removed, and non-nodulated lateral roots will be portioned. The first set of seeds pods and leaves composited from the 1st, 3rd, and 5th leaf node will be sampled separately and portioned. Soil remaining in the pot will be mixed by hand for 1 min, then 200 g will be portioned. Remaining soil will be evenly wetted with 0.5 L of IPTG solution, and a representative sample of leachate will be collected from drainage holes and portioned. Prior to storage, rhizosphere wash and leachate will be pelleted by centrifugation and the supernatant will be discarded. Prior to homogenization, nodules will be counted and weighed. All samples will be lyophilized prior to downstream molecular analyses.

Progress 07/01/23 to 06/30/24

Outputs
Target Audience:We are targeting several audiences for sharing our materials and methods for the plant-model system (ex. bacterial strains) and the knowledge we gain into the escape risk of genetic containment strategies in natural settings. Our target audiences include: 1. Biotechnology companies: Our project will serve biotech researchers in agriculture by providing a framework for the evaluation of genetic containment strategies. Our genetically tractable and traceable Rhizobium species and plant-soil system will provide the means to evaluate and compare evasion risk among containment strategies. These tools will improve confidence in the deployment of future biotechnologies in plants and soil systems. PD Wilhelm is a member of the Coordinating Committee of the Phytobiomes Alliance and has been familiarizing industrial partners with the aims of our project, including Syngenta, PivotBio, Earth Microbial, and Ginkgo Bioworks. We have established a relationship with researchers at Ginkgo Bioworks to explore the use of our model system, as well as share resources and expertise. 2. Agronomists and Growers: Biotechnologies are low cost and scalable for agricultural applications. Currently in the USA and globally, there is a surge of interest in their efficacy for biocontrol, biofertilization etc. We will engage with growers and agronomist to learn about their needs and to share what we learn about developments in biocontainment strategies. In the USA, we will reach these audiences at the national level at the annual Tri-Societies meeting (Agronomy, Crop, and Soil Science Societies of America) and via PD Wilhelm's annual extension outreach events in Indiana, including the Indiana Certified Crop Advisors Conference, and the Indiana Soil Health and Sustainability Course. We also plan to engage audiences outside of the USA, particularly in Latin America, where beans are a staple crop. The PhD student leading the work, Túlio Machado, is a Brazilian national who has developed collaborations with researchers in Mexico (Dr. David Romero), Brazil (Dr. Mariangela Hungria da Cunha), and Spain (Dr. Beatriz Jorrin Rubio). 3. Classroom and laboratory instruction: The expertise we develop in our project will contribute to foundational growth in synthetic biology at Purdue. Co-PD Green is the supervisor of Purdue's International Genetically Engineered Machine (iGEM) team and both incoming MS students will be closely involved in mentoring the team. PD Wilhelm and Co-PD Green are board members of the Purdue Applied Microbiome Sciences and are steering the group to develop graduate training workshops in synthetic biology. While this ambition will not have an immediate impact beyond undergraduate and graduate students on our campus, the goal is to build expertise that serves early career scientists in the Midwest. This ambition is manifested in the growth of the Purdue Microbiome Symposium into the Midwest Microbiome Symposium in the past 3 years, which has been led by PD Wilhelm and Co-PD Green who sit on the organizing committee. Changes/Problems:We have not yet encountered any problems that have altered our plans. We have changed our initial timeline by approximately 6 - 8 months due to delays in our recruitment of a MS student to undertake R. etli engineering in Objective 1. Students with training in synthetic biology are in high demand. We sought to entice talented students with entrance scholarships and conducted an extensive search for suitably trained students. From these efforts, we managed to recruit two MS students (Grace Cook and Luis González) to work on Objective 1 and were successful in obtaining a one-year fellowship for Luis González. Both incoming MS students have extensive backgrounds in bioengineering. Túlio Machado, the PhD responsible for Objectives 2 and 3, was also awarded a three-year fellowship. With awards and budgeted funds, we have the capacity to support both MS for two years and our PhD candidate for a potential 5 years. What opportunities for training and professional development has the project provided?Incoming MS student, Grace Cook, was admitted to the prestigious Cold Spring Harbor Laboratory Synthetic Biology 2-week summer course (Summer 2024). Funds from this grant are supporting this training opportunity. How have the results been disseminated to communities of interest?The current PhD student has given seminars on two topics: Machado, Túlio. The use of PNA clamps for surveys of plant bacterial diversity. Agronomy Graduate Student Organization Seminar Series. Purdue University. April 29th, 2024. Machado, Túlio. Kill Switches for GEM in Agriculture. Purdue Applied Microbiome Sciences. Purdue University. December 1st, 2024. What do you plan to do during the next reporting period to accomplish the goals?In Year 2, our team will make significant advances in the engineering of the cold temperature-sensitive Rhizobium strain. MS students Grace Cook and Luis Gonzalez will be trained to engineer Rhizobium. MS student, Grace Cook, will be responsible for engineering and testing the ccdB-ccdA kill switch circuit in Rhizobium (currently in E. coli) and for developing a Cas9-mediated kill switch circuit. MS student, Luis Gonzalez, will be responsible for tuning a library of temperature sensitive promoters in Rhizobium and using them to regulate the kill switches. PhD student, Tulio Machado, will finalize the methodological setup for version 1 of the model plant-soil system. He will complete the development and testing of qPCR primers targeting our isolate. He will have performed in soil and in plant competition assays with our GEM and close relatives of Rhizobium.

Impacts
What was accomplished under these goals? We have accomplished several tasks in each of our Objective areas: Objective 1. To establish standardized procedures and quantitative methods for assessing evasion risk in a Phaseolus-Rhizobium model system. We identified and finalized several design elements of our model system to ensure it can be readily adopted and widely used for testing and risk assessment. To this end, we performed two major surveys to select the optimal cultivars of Phaseolus (for growth in a containment facility) and strains of Rhizobium (for colonization performance and ease of transformation) as described in Activity 2. We are currently running our first potted plant experiment testing the colonization of P. vulgaris Benton by our three candidate Rhizobium: the two engineered strains (R. etli CN42 and R. leguminosarum 3841) and one native strain R. tropici, which is known to be among most effective nodulators of P. vulgaris. The experiment will test whether there are differences in the characteristics of colonization by our different Rhizobium strains, and to generate samples for testing our quantification methods. We have made progress on two quantification methods for calculating the survivor ratio (cells that survive under non-permissive conditions): (i) qPCR targeting our engineering bacteria ('reMarker') and (ii) the use of fluorescent protein markers (GFP) and flow cytometry. We have designed and are currently testing qPCR primer sets for the family Rhizobiaceae (targeting the recA gene) and targeting the GFP gene (a marker of our engineered strains). We have begun designing our unique mRNA construct ('reMarker') to optimize detection of surviving cells by strengthening expression and the RNA sequence that has the greatest stability (i.e. the lowest minimum-free-energy change). Objective 2: To compare the efficacy of programmed death and the frequency and type of escape mutants in plant (nodule, root, leaf, and seed) and soil compartments (rhizosphere, soil leachate, and bulk soil). In Year 1, our main goal for Objective 2 was to engineer inducible ccdB expression. To that end, we accomplished two tasks: (1) to establish protocols for engineering kill switches in the E. coli Marionette strain system and (2) to acquire and develop the genetic components and workflow to transform R. etli strains. Task 1: The Marionette strain contains a 12.6 kb insert that expresses 12 small-molecule inducible transcription factors, or "sensors," and has been integrated into the glvC locus of three E. coli strains (MG1655, DH10B, and BL21). This system is valuable for optimizing gene expression dynamics in synthetic bacteria. We modified the plasmid pSC101 by introducing the ccdB gene, which produces a toxin that inhibits DNA gyrase activity that causes cell death. Specifically, we cloned the ccdB gene under the control of the pLac promoter into the MG1655 strain of the Marionette using single-day construction of 3G assembly to create a modular plasmid for heat shock transformation. We confirmed the efficacy of the kill switch by performing a growth inhibition assay using our new E. coli marionette chassis with no IPTG or 1mM of IPTG. We measured and plotted the RFP fluorescence normalized by OD 600 over a 24-hour incubation period. The IPTG-inducible ccdB circuit demonstrated the ability to constrain the bacterial culture population to a low density when introduced at the start and at a subsequent stage of growth. Gyrase inhibitors inhibit cell growth, resulting in residual cell remnants in the solution. In Year 2, we will be optimizing the cell death components in E. coli, including lysis and Cas-based kill switches. Task 2: In Year 1, We have obtained genetically competent Rhizobium strains, which will be transformed with our kill switch circuit. We have acquired Rhizobium etli CFN42 with integrase that catalyzes the direct recombination into the chromosome, creating a stable construct. This will serve our aim of integrating our kill switch circuit into the host genome, to improve the stability of the chassis being used in our model plant-soil system. To hedge our bets, we also have a strain of R. leguminosarum 3841 harboring a mini-Tn7-Gm cassette with GFP integrated in the chromosome with gentamycin as an antibiotic marker. Over the next three months (July through September 2024), the Green Lab will develop a library of genetic components for engineering responsive kill switches in R. etli. We will identify and order (via IDT gBlocks) sequences for inducible quorum sensing receptors and protomers, a library of ribosome binding sequences, and toxin output genes (e.g., optimizing ccdB or expressing yacG in Rhizobium, viral lytic broad-range Rhizobium genes). The engineering of our chassis should be robust for our model system to be useful to our intended users. In Year 2, we will also optimize the protocol for transforming Rhizobium strains to enhance the uptake of synthetic DNA using electrocompetent, chemically competent, and a novel protocol the Green Lab is establishing for DNA nanopore-competent transformation. We aim to demonstrate robust and efficient methods for transferring synthetic DNA plasmids across R. etli membranes for engineering novel circuits in the GEM chassis. Objective 3: To assess the influence of environmental (resources and stress) and ecological drivers (competition) on evasion risk. This objective will be focused on in Year 3, but we have begun preparation for the experiments. We demonstrated that the PNA clamp can be used to determine diversity and composition of the endophyte and rhizosphere microbiome of P. vulgaris, which is necessary to study the ecological drivers of evasion risk. We have also started a collection of Rhizobium and closely related members of Rhizobiaceae for use in controlled experiments to test the influence of competition on the survival of our target GEMs in non-permissive conditions.

Publications