“…Following that, we used the sequence d(G 6 ) to link S1 and CS1, S2 and CS2, which formed the sequences d(CGACGAG 6 AGCAGC) (SG1) and d(GCTGCTG 6 TCGTCG) (CSG1), respectively; and used d(T 2 G 6 T 2 ) to link S1 and S2, CS1 and CS2 which formed the sequences d(CGACGATTG 6 TTAGCAGC) (SG2) and d(GCTGCTTTG 6 TTTCGTCG) (CSG2), respectively (table 1). Thus, the strands SG1, SG2, CSG1 and CSG2 containing consecutive six Gs in the centre were supposed to form stable parallel tetramolecular G-quadruplex core in KOAc buffer solution (structures A–C in figure 2; electronic supplementary material, figure S1 a – f ) [26]. Specifically, SG2 could form a G-quadruplex dimer (A in figure 2) connected by II-DNA structure of 5′-(CGA) n sequences at pH 4.5, and a mixture of SG2 and CSG2 (at a molar ration of 1 : 1) could generate a G-quadruplex tetramer in acidic solution (D in figure 2).…”