(a) Superposition of Cγ1 and Cμ1 domain structures. The different amino-acid sequence and conformation of the μ-chain constant domain (coloured yellow) compared to correspondent region in the γ-chain results in the loss of hydrogen bonds and polar contacts with subset no. 4 VH CDR3, thus hampering the formation of the homotypic interactions. The positively charged amino acid Lys214H, involved in hydrogen bonding interactions with the BcR combining site, is not conserved in either IgE or IgM CH1 domains. Moreover, the conformation of the –PNG– loop in Cμ1 prevents the formation of a hydrogen bond to the main chain nitrogen of Tyr110H, and of van der Waals contacts between Tyr109H and the domain surface. (b) Amino-acid sequence alignment for IgG, IgE and IgM constant domains. The residues that contribute to the conformational epitope recognized by subset no. 4 CLL BcR IGs are highlighted in yellow, and in red when endowed with different chemical properties. (c) Substitution of the wild-type Cγ1/Cκ domains with the mouse IgG1 leads to loss of autonomous signalling to the subset no. 4 BcRs derived from the CLL240 (CLL240-mIgG1) and CLL183 (CLL183-mIgG1) cases. The mouse Cγ1 mutant lacks the side chain of Lys214H substituted with a Pro, and is thus unable to establish hydrogen bonds with the VH CDR3 residues Tyr110H and Tyr111H (see Fig. 1). Both mutated receptors signal on crosslinking with an anti-light-chain antibody. Positive and negative controls are a CLL-derived BcR that signals autonomously (UM-CLL2) and an unrelated BcR (53), respectively. Expression was similar for all receptors, as judged from fluorescence-activated cell sorting analysis.