Predictive Modeling of Riboflavin Production in Lactiplantibacillus plantarum MTCC 25432 Using Fuzzy Inference System

Area of Science:

• Microbiology
• Biotechnology
• Metabolic Engineering

Background:

• Riboflavin (Vitamin B2) is a vital nutrient and a microbial metabolite produced by lactic acid bacteria (LAB).
• Optimizing riboflavin production in LAB is crucial for biotechnological applications and nutritional fortification.
• Selected *Lactiplantibacillus plantarum* strains were investigated for their potential in overproducing riboflavin.

Purpose of the Study:

• To optimize riboflavin overproduction in *Lactiplantibacillus plantarum* using a combination of One Factor At A Time (OFAT) and Fuzzy Inference System (FIS).
• To validate the optimized conditions through fermentative production in a semi-defined medium.
• To identify key media components influencing riboflavin yield.

Main Methods:

• Screening of three *Lactiplantibacillus* strains to select the highest riboflavin producer.
• Application of OFAT coupled with FIS to optimize media components (casamino acid, GTP, sodium acetate, glycine).
• Fermentative production and validation of optimized conditions in a semi-defined medium.

Main Results:

• *Lactiplantibacillus plantarum* MTCC 25432 was identified as the highest riboflavin producer, yielding 346 µg/L.
• The OFAT-FIS optimization predicted and achieved a 24% increase in riboflavin production, reaching 429 µg/L.
• The FIS model accurately estimated riboflavin production within an acceptable limit of 3.4%.

Conclusions:

• An FIS-based predictive model is effective for optimizing riboflavin production in *Lactiplantibacillus plantarum*.
• Optimized levels of casamino acid, GTP, and sodium acetate significantly enhance riboflavin yield.
• The findings provide a basis for redesigning media matrices for improved microbial production of riboflavin.