in situ localization), the aim of which is to provide a comparison between matrices.
Thus, although the method does not provide any structural information on the
studied shell matrices, it gives indirectly an informative picture on the degree of
structural similarity between macromolecules.
In the present case, polyclonal antibodies were produced in rabbits against nonfractionated acetic acid-soluble matrix macromolecules extracted from prisms and
from nacre of Pinna nobilis. These antibodies were tested against prisms and nacre
matrices (Fig. 13.4a for the western blot results obtained with the antibodies raised
against prisms matrix). They cross-reacted in a more or less symmetrical way,
i.e., antibodies elicited against the nacre matrix recognized, in a lesser extent,
epitopes in the prisms matrix, and antibodies elicited against the prisms matrix
crossed-reacted with nacre matrix. This suggested that both matrices shared several
epitopes. This idea was – and still is – very appealing: it may indicate that shell
microstructures as different as prisms and nacre can be produced from different
secretory protein repertoires, which contain however some identical epitopes (short
domains). It is interesting to notice that before our study, by using a different
approach – a combination of ion-exchange and HPLC chromatography on the
soluble nacre and prisms matrices of the mussel Mytilus californianus – Weiner
(1983) observed that some of the numerous peaks were shared by the two matrices.
He consequently came to the conclusion of a partial similarity between the two
matrices.
As the developed anti-prism/nacre matrices antibodies exhibited good titers
and gave satisfactory results, we tested them by ELISA technique against a broad
range of shell matrices extracted from numerous molluscs, including bivalves,
gastropods, and cephalopods. In addition, the antibodies were tested against skeletal matrices of extra-group samples, brachiopods, and corals (Marin 1992; Marin
et al. 1999, 2007a). The results were surprising (Fig. 13.4b). First, none of the two
antibodies gave immunological patterns, which could be simply correlated to
mollusc phylogeny. Consequently, these antibodies could not be used for serotaxonomy purposes. To give an example, both antibodies gave strong crossreactivities with the shell matrices of the two gastropods Haliotis and Littorina,
while they did not cross-react with the pteriomorphid bivalve Arca (Fig. 13.4). One
of the two antibodies (anti-nacre) also reacted strongly with the shell matrix of the
cephalopod Nautilus. Second, the obtained signals were not strictly correlated to
shell microstructures, neither to the mineralogy (calcite vs. aragonite). For the antiprisms antiserum, we observed that it strongly cross-reacted with the matrices of
nacro-prismatic pteriomorphid bivalves (Mytilus, Pteria, Pteria, Isognomon,
Brachidontes), with those of foliated pteriomorphids (Pecten, Crassostrea) but
also with the fully aragonitic nacro-prismatic paleoheterodont bivalves (Unio,
Anodonta, Neotrigonia). There were no cross-reactivities with the matrices of
“crossed-lamellar” aragonitic veneroids (Venus, Mercenaria, Tridacna, Ensis),
but curiously, cross-reactivities with aragonitic crossed-lamellar myoids (Mya,
Pholas). This pattern was not expressed with the second antibody (anti-nacre). Of
course, we were fully aware that the serological approach has some drawbacks: (a)
it did not give any indication on the chemical nature of the target antigens; (b)
370
F. Marin et al.
Thus, although the method does not provide any structural information on the
studied shell matrices, it gives indirectly an informative picture on the degree of
structural similarity between macromolecules.
In the present case, polyclonal antibodies were produced in rabbits against nonfractionated acetic acid-soluble matrix macromolecules extracted from prisms and
from nacre of Pinna nobilis. These antibodies were tested against prisms and nacre
matrices (Fig. 13.4a for the western blot results obtained with the antibodies raised
against prisms matrix). They cross-reacted in a more or less symmetrical way,
i.e., antibodies elicited against the nacre matrix recognized, in a lesser extent,
epitopes in the prisms matrix, and antibodies elicited against the prisms matrix
crossed-reacted with nacre matrix. This suggested that both matrices shared several
epitopes. This idea was – and still is – very appealing: it may indicate that shell
microstructures as different as prisms and nacre can be produced from different
secretory protein repertoires, which contain however some identical epitopes (short
domains). It is interesting to notice that before our study, by using a different
approach – a combination of ion-exchange and HPLC chromatography on the
soluble nacre and prisms matrices of the mussel Mytilus californianus – Weiner
(1983) observed that some of the numerous peaks were shared by the two matrices.
He consequently came to the conclusion of a partial similarity between the two
matrices.
As the developed anti-prism/nacre matrices antibodies exhibited good titers
and gave satisfactory results, we tested them by ELISA technique against a broad
range of shell matrices extracted from numerous molluscs, including bivalves,
gastropods, and cephalopods. In addition, the antibodies were tested against skeletal matrices of extra-group samples, brachiopods, and corals (Marin 1992; Marin
et al. 1999, 2007a). The results were surprising (Fig. 13.4b). First, none of the two
antibodies gave immunological patterns, which could be simply correlated to
mollusc phylogeny. Consequently, these antibodies could not be used for serotaxonomy purposes. To give an example, both antibodies gave strong crossreactivities with the shell matrices of the two gastropods Haliotis and Littorina,
while they did not cross-react with the pteriomorphid bivalve Arca (Fig. 13.4). One
of the two antibodies (anti-nacre) also reacted strongly with the shell matrix of the
cephalopod Nautilus. Second, the obtained signals were not strictly correlated to
shell microstructures, neither to the mineralogy (calcite vs. aragonite). For the antiprisms antiserum, we observed that it strongly cross-reacted with the matrices of
nacro-prismatic pteriomorphid bivalves (Mytilus, Pteria, Pteria, Isognomon,
Brachidontes), with those of foliated pteriomorphids (Pecten, Crassostrea) but
also with the fully aragonitic nacro-prismatic paleoheterodont bivalves (Unio,
Anodonta, Neotrigonia). There were no cross-reactivities with the matrices of
“crossed-lamellar” aragonitic veneroids (Venus, Mercenaria, Tridacna, Ensis),
but curiously, cross-reactivities with aragonitic crossed-lamellar myoids (Mya,
Pholas). This pattern was not expressed with the second antibody (anti-nacre). Of
course, we were fully aware that the serological approach has some drawbacks: (a)
it did not give any indication on the chemical nature of the target antigens; (b)
370
F. Marin et al.