Heparin, a glycosaminoglycan, is a stable source of carbon that supports the growth of microorganisms in the human intestine. It is also a commonly used anticoagulant drug in clinical practice, with significant therapeutic effects. Low molecular weight heparin (LMWH) is a highly active low molecular weight fragment obtained via enzymatic reaction or the chemical degradation of heparin. LMWH has been applied globally in the prevention and treatment of venous thromboembolism in thrombosis patients. Simultaneously, as a potential prebiotic, because of its low molecular weight, LMWH can be well degraded by the gut microbiota to maintain intestinal balance. Enzymatic heparin degradation has recently emerged as a viable disposal method for LMWH preparation; however, only very few benchmark enzymes have been thoroughly described and subjected to protein engineering to improve their properties over the past few years. The commercialization of enzymes will require the development of robustly engineered enzymes that meet the demands of industrial processes. Herein, we report a rational protein engineering strategy that includes molecular dynamic simulations of flexible amino acid mutations and disulfide bond screening. Several Bacteroides thetaiotaomicron heparanase Ⅰ (Bt-HepI) mutants were obtained and screened for high thermal stability. We obtained the Bt-HepID204C/K208C/H189W/Q198R variant, which features a stabilized protein surface structure, with a 1.3-fold increase in catalytic constant/michaelis-menten constant (kcat/Km), a 2.44-fold increase in thermal stability at 50 ℃, and a 1.8-fold decrease in the average molecular weight of LMWH produced at 40 ℃ compared with that seen with Bt-HepIWT. Our study establishes a strategy to engineer thermostable HepI to underpin its industrial applications.