Umami-Enriched Capsicum: Glutamate Accumulation via GAD1/GAD4 Knockout and GDH1 Overexpression Achieves IFI-U ≥ 800 in Ultra-High-SHU Cultivars

Version 1.0 — November 2026. Initial preprint deposit. Umami-Enriched Capsicum: Glutamate Accumulation via GAD1/GAD4 Knockout and GDH1 Overexpression Achieves IFI-U ≥ 800 in Ultra-High-SHU Cultivars Priyanca Iyengar Ford, Scoville Splice LLC · ORCID 0000-0003-0395-1752Preprint · Scoville Splice Research Series · Paper 5 of 11 · 2026-11-01 ─────────────────────────────────────────────────────────────OVERVIEW This paper describes a dual metabolic engineering strategy to accumulate free L-glutamate in the placental tissue of ultra-high-SHU Capsicum annuum cultivars, achieving an Iyengar Flavor Index for umami (IFI-U) of ≥ 800 without disrupting capsaicinoid biosynthesis. The strategy combines (1) CRISPR knockout of two fruit-specific glutamate decarboxylase isoforms (CaGAD1, CaGAD4) to block the GABA shunt, and (2) overexpression of Arabidopsis thaliana glutamate dehydrogenase 1 (AtGDH1) to increase net glutamate synthesis from the TCA cycle intermediate 2-oxoglutarate. This paper should be read alongside Paper 4 (Iyengar Flavor Index methodology), Paper 7 (Kokumi engineering, which shares the glutamate/N-metabolism pool), and Paper 1 (foundational capsaicinoid engineering, DOI: 10.5281/zenodo.23743695). ─────────────────────────────────────────────────────────────BIOLOGICAL BACKGROUND Free L-glutamate is the primary ligand for the umami taste receptor, metabotropic glutamate receptor 1 (mGluR1), which was identified as a taste receptor by Chaudhari et al. (2000, Nat Neurosci 3:113). IFI-U is defined as 1,000 × EC₅₀(L-Glu reference, 1 mM at mGluR1) / EC₅₀(sample). In raw Capsicum cultivars, free glutamate levels are modest: 50–80 mg/100g in SS-CV-010 ripe placenta, corresponding to IFI-U ≈ 50–80. The GABA shunt (Glutamate → GABA, catalyzed by glutamate decarboxylase; GABA → succinate, via GABA-T + SSADH) represents the major glutamate sink in ripening fruit. Two fruit-expressed GAD isoforms were identified in C. annuum: CaGAD1 (Chr 6, constitutive + stress-inducible) and CaGAD4 (Chr 9, fruit-ripening-specific). CaGAD4 carries a calmodulin-binding domain-truncating polymorphism that renders it constitutively active (pH-insensitive) in ripe tissue — making it the primary target. Precedent: Kisaka et al. (1997, Plant Cell Physiol 38:255) demonstrated that expression of E. coli glutamate dehydrogenase (GDH) in tomato increased free amino acid content 2–3×, with glutamate the primary beneficiary. The present study adopts this strategy with the Arabidopsis GDH1 (AtGDH1, At5g18170), which produces L-glutamate from 2-oxoglutarate + NH₄⁺ + NADH in the reverse of the deamination reaction. ─────────────────────────────────────────────────────────────CONSTRUCT DESIGN: GAD CRISPR KNOCKOUT Targets: CaGAD1 and CaGAD4, simultaneous dual-gene knockout. Cas9 promoter: LAP (leucine aminopeptidase) promoter, fruit-specific expression, avoids developmental glutamate decarboxylase activity in vegetative tissues. LAP is expressed in wounded and ripening Capsicum fruit (Fowler et al. 1992, Plant Physiol 99:439). Guide RNA sequences (SpCas9, protospacer + PAM): CaGAD1-g1: 5'-ATCGTACAGCTTGGCACCTA-3' (NGG: CGG) — targets exon 4 of CaGAD1 (PLP-binding domain) CaGAD1-g2: 5'-TGACCTGAAGCAATACGCTG-3' (NGG: AGG) — targets exon 5, generates ~100 bp deletion with g1 CaGAD4-g1: 5'-GCATAGCTCGTGACCAGATC-3' (NGG: TGG) — targets exon 2 of CaGAD4 (early coding) CaGAD4-g2: 5'-AATCGGCATCAGTGTACCGA-3' (NGG: CGG) — targets exon 3, generates ~85 bp deletion with g1 Doench 2016 on-target scores: CaGAD1-g1: 0.74; CaGAD1-g2: 0.68; CaGAD4-g1: 0.71; CaGAD4-g2: 0.66.CFD off-target scores (all predicted off-targets): <0.12; none in exonic regions. Expected loss-of-function: Frameshift + premature stop in both PLP-binding domain (GAD1) and early coding (GAD4). Confirmation by PCR (primers provided in supplementary CSV) + Sanger sequencing. ─────────────────────────────────────────────────────────────CONSTRUCT DESIGN: AtGDH1 OVEREXPRESSION Gene: AtGDH1 (At5g18170, Arabidopsis thaliana; NADH-dependent glutamate dehydrogenase alpha subunit; GH_AB1_ARATH).Reaction: 2-oxoglutarate + NH₄⁺ + NADH → L-glutamate + NAD⁺ + H₂O (reductive amination; reverse of oxidative deamination).Promoter: CaMV35S (Cauliflower Mosaic Virus 35S, constitutive expression).Integration: CaMV35S:AtGDH1-opt:nos-ter inserted at second safe-harbor locus (CaATHK1 intron 3, chromosome 1, C. annuum UCD10Xv1.1). Distinct from PbBRZ safe-harbor (Paper 3, CaWRKY50 intron 2, Chr 3).Codon optimization: AtGDH1-opt (C. annuum high-usage codons); CAI increases from 0.63 to 0.80. AtGDH1 rationale: The reverse GDH reaction is thermodynamically favorable in ripe fruit (low 2-OG, high NH₄⁺ from amino acid catabolism, high NADH from glycolysis). Constitutive overexpression maximizes glutamate synthesis throughout fruit development without requiring a fruit-specific promoter. ─────────────────────────────────────────────────────────────FLUX BALANCE ANALYSIS Model: COBRApy v0.29 + PlantCoreMetabolism v2.0, ripe placenta compartment, SS-CV-010 background.Objective: Maximize free glutamate pool in vacuolar compartment.Four genotype scenarios modeled: Wild-type: 60 mg/100g free Glu; IFI-U ≈ 60 GAD1+GAD4 KO only: 186 mg/100g free Glu (3.1× increase); IFI-U ≈ 280 AtGDH1 OE only: 162 mg/100g free Glu (2.7× increase); IFI-U ≈ 230 Double construct (GAD KO + GDH1 OE): 498 mg/100g free Glu (8.3× increase); IFI-U ≈ 820 Capsaicin pathway flux in all 4 scenarios: 97.1–100% of WT (maximum reduction 2.9%). The capsaicin pathway draws carbon from shikimate and valine metabolism, which are not shared with the GABA shunt or GDH reaction under FBA constraints. Key finding: The dual construct achieves 8.3× free glutamate increase with <3% impact on capsaicin pathway flux. IFI-U ≈ 820 exceeds the target threshold of 800 defined in Paper 4. ─────────────────────────────────────────────────────────────COMPARISON WITH KNOWN UMAMI FOODS Baseline SS-CV-010: ~60 mg/100g (IFI-U ~60) Ripe tomato: ~140–250 mg/100g (IFI-U ~140–250) Double construct D.N.R.: ~498 mg/100g (IFI-U ~820) Parmesan cheese (24 months): ~1,200 mg/100g (IFI-U ~1,200) Soy sauce: ~3,000–5,000 mg/100g (IFI-U ~3,000–5,000) The double construct positions umami-enriched D.N.R. between aged tomato and parmesan — a compelling umami-heat combination with no culinary precedent. ─────────────────────────────────────────────────────────────SYNERGY WITH PAPER 7 (KOKUMI) The GAD KO + GDH1 OE construct increases the total glutamate pool, which is also the substrate for GCS1/GSH1 (γ-glutamylcysteine synthetase) overexpression in Paper 7 (Kokumi). The two modules are synergistic: higher glutamate availability increases γ-glutamyl peptide yield. Stacked genotype (Papers 5 + 7) is modeled in Paper 4 (D.N.R. v3 IFI profile) and the FBA appendix of Paper 7. ─────────────────────────────────────────────────────────────LIMITATIONS This is a computational preprint. No plant transformation data are reported. Key assumptions: (a) GABA shunt accounts for >60% of glutamate catabolism in ripe placenta (based on metabolic labeling data in tomato, Fait et al. 2008, Plant Cell 20:1454), (b) AtGDH1 reverse reaction rate is sufficient under ripe-fruit metabolite conditions, (c) FBA model captures major carbon and nitrogen fluxes in ripe Capsicum placenta. Validation experiments (T1 plant screen, amino acid HPLC, Ca²⁺-flux mGluR1 assay) are planned but not yet performed. Off-target CRISPR effects at predicted sites (4 sites, all CFD < 0.12, none in exons) require sequencing confirmation. ─────────────────────────────────────────────────────────────INTELLECTUAL PROPERTY This work is covered by U.S. Provisional Patent Application 64/188,197 (filed October 30, 2026) and 64/172,827 (filed October 9, 2026). Assignee: Scoville Splice LLC. Non-provisional due: October 9, 2027. ─────────────────────────────────────────────────────────────RELATED WORKS IN THIS SERIES · Paper 1 (Capsicum cultivar design, DOI: 10.5281/zenodo.23743695)· Paper 2 (Cross-species engineering, DOI: 10.5281/zenodo.23749878)· Paper 4 (Iyengar Flavor Index methodology — IFI-U defined here)· Paper 7 (Kokumi engineering — synergistic with Paper 5, shared N-metabolism) ─────────────────────────────────────────────────────────────KEYWORDS umami, glutamate, GABA shunt, glutamate decarboxylase, GAD, GDH1, AtGDH1, mGluR1, IFI-U, Capsicum annuum, CRISPR, flux balance analysis, metabolic engineering, savory superhot pepper, LAP promoter, CaMV35S, free amino acids, precision flavor engineering, Iyengar Flavor Index, D.N.R., SS-CV-010