widely different epitopes could be recognized at a similar level by a given antibody;
(c) because antibodies were raised against protein mixtures, some of which were
immunogenic, some of which not, the final signal may have been contorted and
biased. In spite of these limitations, the immunological pattern obtained suggested,
on one hand, that the secretory repertoire used by different phylogenetically distant
molluscs could exhibit some similarities; On the other hand, some secretory
repertoires appeared to be extremely different from group to group, the best
illustration being the clear gap between pteriomorphids and heterodonts bivalves.
13.4.3 Molecular Data on the Shell of P. nobilis
Although the serological approach gave indications on the overall homologies of
matrices in comparison to that of P. nobilis, sequence information on the macromolecular constituents of the prisms and nacre matrices were still badly missing.
This gap in the knowledge started to be filled by finding tricks to accurately purify
proteins directly from shell extracts, and, as an alternative approach, by using
molecular biology techniques to fish out randomly transcripts encoding shell
proteins. The first approach gave us two proteins, caspartin and calprismin, while
the second one, mucoperlin (Marin et al. 2003b; Marin and Luquet 2005). The first
two proteins belong to the group of highly acidic shell proteins (Fig. 13.5), while
the third one, which is moderately acidic, appears to belong to a completely
different protein family (Fig. 13.6). In addition, new proteins of the shell of
P. nobilis are now being discovered in our lab by one of us (P. Narayanappa).
13.4.3.1 Molecular Data on the Prisms
The way two prism matrix proteins were isolated relies on the following observation: first, most of the shell matrices proteins have a low content of amino acids
that contain aromatic groups (tyrosine, tryptophane, phenylalanine). This implies
that they cannot be easily detected with the classical spectrophotometric methods
at 280 nm. Second, in our numerous attempts to visualize proteins in a monodimensional gel, we often obtained sharper pictures on Western blots when the blot
Fig. 13.4 (continued) of P. nobilis. (a) western blot of the different shell matrices of P. nobilis
(prisms, lanes 1 and 2; nacre, lanes 3 and 4) with this antibody. (b) Cross-reactivities measured
with ELISA. Although the highest cross-reactivities are recorded with nacro-prismatic
pteriomorphid bivalves, the obtained signal cannot be considered to be “taxonomic”: some
distantly related genera (Haliotis, Littorina) cross-react strongly. The signal is neither
“microstructural”: some genera with shell microstructures that are different from the nacroprismatic microstructure of P. nobilis (Glycymeris, Pecten) also give high cross-reactivities.
Except Ruditapes, none of the veneroid bivalves cross-react. Similarly, none of the tested extragroups (cnidarian, brachiopod) give cross-reactivity. Min mineralogy, A aragonite, C calcite
◂
372
F. Marin et al.
(c) because antibodies were raised against protein mixtures, some of which were
immunogenic, some of which not, the final signal may have been contorted and
biased. In spite of these limitations, the immunological pattern obtained suggested,
on one hand, that the secretory repertoire used by different phylogenetically distant
molluscs could exhibit some similarities; On the other hand, some secretory
repertoires appeared to be extremely different from group to group, the best
illustration being the clear gap between pteriomorphids and heterodonts bivalves.
13.4.3 Molecular Data on the Shell of P. nobilis
Although the serological approach gave indications on the overall homologies of
matrices in comparison to that of P. nobilis, sequence information on the macromolecular constituents of the prisms and nacre matrices were still badly missing.
This gap in the knowledge started to be filled by finding tricks to accurately purify
proteins directly from shell extracts, and, as an alternative approach, by using
molecular biology techniques to fish out randomly transcripts encoding shell
proteins. The first approach gave us two proteins, caspartin and calprismin, while
the second one, mucoperlin (Marin et al. 2003b; Marin and Luquet 2005). The first
two proteins belong to the group of highly acidic shell proteins (Fig. 13.5), while
the third one, which is moderately acidic, appears to belong to a completely
different protein family (Fig. 13.6). In addition, new proteins of the shell of
P. nobilis are now being discovered in our lab by one of us (P. Narayanappa).
13.4.3.1 Molecular Data on the Prisms
The way two prism matrix proteins were isolated relies on the following observation: first, most of the shell matrices proteins have a low content of amino acids
that contain aromatic groups (tyrosine, tryptophane, phenylalanine). This implies
that they cannot be easily detected with the classical spectrophotometric methods
at 280 nm. Second, in our numerous attempts to visualize proteins in a monodimensional gel, we often obtained sharper pictures on Western blots when the blot
Fig. 13.4 (continued) of P. nobilis. (a) western blot of the different shell matrices of P. nobilis
(prisms, lanes 1 and 2; nacre, lanes 3 and 4) with this antibody. (b) Cross-reactivities measured
with ELISA. Although the highest cross-reactivities are recorded with nacro-prismatic
pteriomorphid bivalves, the obtained signal cannot be considered to be “taxonomic”: some
distantly related genera (Haliotis, Littorina) cross-react strongly. The signal is neither
“microstructural”: some genera with shell microstructures that are different from the nacroprismatic microstructure of P. nobilis (Glycymeris, Pecten) also give high cross-reactivities.
Except Ruditapes, none of the veneroid bivalves cross-react. Similarly, none of the tested extragroups (cnidarian, brachiopod) give cross-reactivity. Min mineralogy, A aragonite, C calcite
◂
372
F. Marin et al.