Developing a gene-edited crop variety from initial target identification to commercial market entry is a complex, multi-year process spanning disciplines from molecular biology and computational science to agronomy, regulatory affairs, and commercial development. The path is not linear — it involves iterative cycles of hypothesis, experiment, and refinement — and it requires decisions at each stage that affect not just scientific outcomes but commercial timelines, regulatory risk, and ultimately the probability that a product reaches the farmers who need it.
This article describes the stages of ClimateCrop's crop development pipeline in detail, explaining the rationale behind our approach at each phase and identifying where the critical decision points lie. Understanding this pipeline is important for evaluating the maturity of any gene editing program — the scientific literature is full of impressive early-stage results that never progressed to commercial deployment because the development infrastructure to carry them forward was not in place.
Stage 1: Target Identification and Prioritization
The pipeline begins with identifying the specific genomic loci where edits are most likely to produce the desired agronomic phenotype. This is not a trivial step. Crops have tens of thousands of genes, and the literature describing their functions ranges from precisely characterized to entirely speculative. A poorly selected target can waste 12 to 24 months of editing and greenhouse work before the phenotypic evaluation reveals no agronomic benefit.
ClimateCrop's target identification process integrates three primary inputs. First, genome-wide association studies (GWAS) applied to large germplasm panels phenotyped under stress conditions identify genomic regions statistically associated with tolerance traits. Second, transcriptomics data — RNA sequencing comparing stressed and unstressed plant tissues — identifies genes that are differentially regulated under the stress conditions of interest, providing candidate lists of functionally responsive loci within GWAS-associated regions. Third, ortholog analysis maps known functional genes from model plants (primarily Arabidopsis) and well-studied crops (rice) onto the target crop genome, providing mechanistic hypotheses for candidates lacking direct functional validation in the crop species.
Candidates emerging from this integrated analysis are scored on a prioritization matrix that weights: strength of GWAS association, consistency of expression response across multiple experiments, quality of mechanistic evidence from model systems, editability (estimated difficulty of achieving the desired edit with current tools), and regulatory risk profile (whether natural variants of the desired modification exist in the species' gene pool). Typically, three to six high-confidence targets are selected per crop-trait combination to enter the editing phase simultaneously, because the probability that any single target produces the expected phenotype is less than certain even with strong prior evidence.
Stage 2: Guide RNA Design and Editing Construct Development
For each selected target, guide RNA (gRNA) sequences are designed to direct Cas9 or base editor complexes to the precise genomic location where the edit is required. Guide RNA design must satisfy multiple criteria simultaneously: high on-target activity at the intended editing site, minimal predicted off-target editing at other genomic loci, compatibility with the editing modality being used (standard nuclease-based editing versus base editing versus prime editing), and avoidance of single-nucleotide polymorphisms in the target sequence that could reduce gRNA binding efficiency across different genetic backgrounds of the crop species.
Off-target prediction uses computational tools (Cas-OFFinder, CRISPOR, and related software) to identify sites in the genome with sequence similarity to the target site, weighted by mismatches in the guide-target duplex and the PAM sequence context. Predicted off-target sites in protein-coding sequences or known regulatory elements receive additional scrutiny. Where off-target risk is elevated, alternative gRNA sequences targeting adjacent positions at the same functional locus are evaluated. The goal is to select gRNA sequences that provide high editing efficiency at the target with the fewest unintended editing events in the rest of the genome.
The editing construct — the complete package of gRNA and Cas9 or base editor protein sequences, promoters, and delivery elements — is assembled and validated in a model system (typically Nicotiana benthamiana or rice protoplasts) before moving to the target crop species. This protoplast validation step confirms that the construct is expressed, that the gRNA directs editing to the predicted location, and that the intended molecular outcome — the specific nucleotide change or indel — is produced at acceptable efficiency.
Stage 3: Crop Transformation and Event Generation
Delivering the editing construct into crop cells and regenerating fertile plants from successfully edited cells is the most technically demanding step of the pipeline — and historically the most variable in throughput and cost across different crop species. Wheat and sorghum are notoriously difficult to transform; soybean and canola are substantially more tractable. ClimateCrop's transformation protocols are optimized per species, reflecting years of iterative refinement to current standards.
For wheat, we use Agrobacterium-mediated immature embryo transformation with morphogenic regulator co-delivery (Wuschel/Baby Boom) to increase callus quality and regeneration frequency, as described by transformation optimization research published by multiple groups in 2023. For soybean, we use both Agrobacterium cotyledonary node transformation and hairy root transformation for rapid early-stage validation. For maize, we use biolistic delivery into embryogenic suspension cultures, a protocol that has been well-established in the industry for over a decade.
Transformation campaigns generate several hundred primary transgenic or edited plants (T0 events) per target per crop species. Each event is a genetically unique individual carrying edits at a potentially different combination of loci and with potentially different tissue culture-induced background variation. Events must be screened comprehensively to identify those carrying the intended on-target edit in the desired configuration — biallelic, homozygous, or hemizygous depending on the specific target and editing strategy.
Molecular Characterization of Events
Molecular screening of T0 events uses a combination of PCR-based assays, Sanger sequencing of the target amplicon, and — for events advancing to the T1 generation — amplicon deep sequencing to characterize the full spectrum of editing outcomes at the target site. Events with the intended edit and no detectable large structural rearrangements at the target locus advance to T1 seed production.
For events where the editing construct was introduced via Agrobacterium or biolistics and may be present as a stably integrated transgene, off-target integration screening is required. We use digital droplet PCR to confirm transgene copy number and whole-genome sequencing of lead events to confirm clean on-target editing without large insertions or deletions at predicted off-target sites. Events where the Cas9 construct is present as a stably integrated transgene are progressed through backcross generations to segregate away the construct, retaining only the intended genomic edit.
Stage 4: Greenhouse Phenotyping and Initial Characterization
The first functional test of each edited event is greenhouse phenotyping under controlled stress conditions. Plants are grown in climate-controlled growth chambers capable of imposing precise drought or heat stress regimens at defined crop growth stages. Stressed and unstressed control treatments are applied to edited events, unedited control plants of the same genetic background, and commercial check varieties to provide reference performance standards.
Phenotypic measurements at this stage focus on stress response physiology: relative water content under drought, chlorophyll fluorescence (Fv/Fm) as an indicator of photosystem function under heat stress, leaf senescence scores, stomatal conductance, and agronomic traits measured at harvest (seed number, seed weight, biomass). Events that show physiologically consistent stress tolerance improvements in two or more independent greenhouse experiments advance to field evaluation. Events where stress tolerance is not confirmed at the greenhouse stage are terminated unless there are compelling reasons to believe the greenhouse protocol failed to capture the relevant stress.
Stage 5: Multi-Environment Field Trials
Field evaluation is the longest and most resource-intensive stage of the pipeline. Events that clear greenhouse phenotyping enter preliminary field trials at two or three locations in the first field season, evaluated against local commercial checks and unedited controls under managed drought stress (withheld irrigation) and natural stress conditions. Data from the first field season is used to make a stage-gate advancement decision: events showing consistent yield advantage under stress at multiple locations advance to an expanded multi-environment trial network the following season.
Expanded trials span the full seven-location global network described in our field trials overview. Trials at this stage include managed stress, natural stress, and well-watered treatments to characterize both stress tolerance and yield potential under adequate water. Data from two or more field seasons across multiple environments is analyzed using mixed model frameworks that partition genetic effects from environmental and G×E effects, producing robust estimates of the trait advantage attributable to the edit and its stability across environments.
Events that demonstrate statistically significant, commercially relevant, and environmentally stable yield advantages in multi-environment trials are advanced to the regulatory characterization phase. Events where performance is marginal, environment-specific, or accompanied by agronomic deficiencies that reduce practical value are terminated at this stage. The typical rate of advancement from field entry to regulatory qualification is approximately 15 to 25 percent of events, reflecting the combination of scientific attrition and the high performance bar set by the commercial check varieties against which edited events are compared.
Stage 6: Regulatory Clearance and Commercial Deployment
For programs targeting US markets under the USDA SECURE rule, regulatory activities begin in parallel with late-stage field evaluation. The voluntary self-determination process with USDA-APHIS is initiated with a complete package of molecular characterization data, target gene identity and function, and documentation of the natural variation basis for the edit. FDA voluntary consultation for food and feed safety is initiated following completion of molecular characterization, typically in the second or third field season.
Backcrossing programs run in parallel with field trials to introgress the editing trait from the initial transformation event background into multiple elite locally adapted variety backgrounds. No single variety serves all production environments, and commercial deployment requires that the trait be available in germplasm that is regionally adapted and preferred by farmers in each target market. Backcross programs typically require three to five generations, which with double-haploid and speed-breeding protocols can be completed in 18 to 30 months.
Commercial deployment involves licensing the trait to seed companies or national variety registration programs, depending on the market and crop. ClimateCrop's commercial model is trait licensing — we develop and characterize the edited genetic material, manage regulatory clearance, and license the trait for incorporation into commercial seed products — rather than direct variety sales. This approach allows broad deployment across the diversity of locally adapted germplasm needed to address the full geographic scope of climate adaptation challenges, without requiring ClimateCrop to build the distribution infrastructure of a commercial seed company.
The Development Timeline in Practice
From target identification to commercial variety availability, the pipeline typically spans five to eight years for a new trait in an established crop species with functional transformation protocols. The longest components are multi-environment field trials (requiring multiple growing seasons) and regulatory clearance (12 to 36 months depending on jurisdiction and trait complexity). Speed-breeding protocols and doubled haploid systems have compressed the backcrossing component substantially. The molecular characterization and greenhouse phenotyping components now run faster than they did a decade ago due to advances in sequencing costs and high-throughput phenotyping capacity.
ClimateCrop's ClimateWheat drought tolerance program entered the pipeline in 2021 with initial target identification, completed greenhouse phenotyping in 2022, entered field trials in 2023, and is currently in the expanded multi-environment trial phase with regulatory activities underway. This timeline places commercial variety availability in the 2027 to 2028 window — well within the urgency defined by climate projections for yield losses in major wheat production regions over the same period. The pipeline is designed not just to generate scientifically credible results but to deliver those results as commercial products in time to be relevant to the farmers and food systems that need them.