l-Threonine　aldolase　from　Actinocorallia　herbida　(AhLTA)　is　an　ideal　catalyst　for　producing　l-threo-4-methylsulfonylphenylserine　[(2S,3R)-1　b],　a　key　chiral　precursor　for　florfenicol　and　thiamphenicol.　The　moderate　C-beta　stereoselectivity　is　the　main　obstacle　to　the　industrial　application　of　AhLTA.　To　address　this　issue,　a　combinatorial　active-site　saturation　test　(CAST)　together　with　sequence　conservatism　analysis　was　applied　to　engineer　the　AhLTA　toward　improved　C-beta　stereoselectivity.　The　optical　mutant　Y314R　could　asymmetrically　synthesize　l-threo-4-methylsulfonylphenylserine　with　81　%　diastereomeric　excess　(de),　which　is　23　%　higher　than　wild-type　AhLTA.　Molecular　dynamic　(MD)　simulations　revealed　that　the　mechanism　for　the　improvement　in　C-beta　stereoselectivity　of　Y314R　is　due　to　the　acylamino　group　of　residues　Arg314　controlling　the　orientation　of　substrate　4-methylsulfonyl　benzaldehyde　(1　a)　in　the　active　pocket　by　directed　interaction　with　the　methylsulfonyl　group;　this　leads　to　asymmetric　synthesis　of　l-threo-4-methylsulfonylphenylserine.　The　success　in　this　study　demonstrates　that　direct　control　of　substrates　in　an　active　pocket　is　an　attract　strategy　to　address　the　C-beta　stereoselectivity　problem　of　LTA　and　contribute　to　the　industrial　application　of　LTA.