CitS is a member of 2-hydroxycarboxylate transporters and has been extensively studied for more than two decades. It imports two sodium ions and a citrate molecule from outside of the cell. CitS is known to exist as a dimer and each monomer is composed of the N- and C-terminal domains whose structures are related by two-fold pseudo-symmetry. CitS contains 11 transmembrane alpha helices and two re-entrant loops that are indispensable for substrate transportation. Similar with many transmembrane proteins, crystallographic study of CitS has been hampered by hydrophobic nature of the protein. In order to overcome this problem, I have developed a novel method that can improve crystallization property of CitS. Diabodies are bispecific antibody fragments that are composed of two Fv domains. Using previously reported structures of antibodies against CitS and human TLR3, I designed a diabody that can simultaneously interact with CitS and TLR3. The resulting diabody can make a stable complex with CitS and TLR3 that can be crystallized in a buffer containing PEG4000. In order to improve diffraction limit of the CitS-diabody-TLR3 crystals, diverse set of detergents have been tested. Among them, I found that Cymal 7 could substantially improve diffraction limit of the crystals. Careful optimization of dehydration condition of the crystals greatly improved diffraction limit of the crystal presumably by reducing solvent content of the crystals. To further improve x-ray diffraction, I have introduced a disulfide bridge between two heavy chains of the diabody. The disulfide bridge can have significant impact in the crystal packing because it can modulate the angle between the two heavy chains of the diabody. Several amino acid pairs that can form disulfide bridges were chosen using the known crystal structures of the antibodies and mutated to cysteines. Among them, the disulfide bridge made by mutations of T87C of anti-CitS and S88C of anti-hTLR3 produced the best crystals that can diffract x-ray beyond 2.3 $\AA$.