Molluscan Shell Matrix Characterization by Preparative SDS-PAGE

The glycoproteinaceous constituents of molluscan shell matrices usually resist chromatographical fractionation. We describe a protocol that overcomes this difficulty and permits collection of a large amount of shell proteins for further in vitro characterization. After dissolution of the mineral phase, the glycoproteins are fractionated “blind” on a preparative electrophoresis. They are subsequently detected with a polyclonal antibody raised against the whole matrix.


INTRODUCTION
The secretion of molluscan shell is finely regulated by a complex array of glycoproteins, polysaccharides, and chitin, which self-assemble during the calcification process and stay entrapped within the shell [1]. These shell constituents exert a strict control on the nucleation and growth of calcium carbonate crystals. By terminating crystallization, they also determine the final shape of the crystals. Because of these multiple roles, they can be used in different ways, i.e., as constituents of biomimetic organomineral materials with superior mechanical properties [2,3] and as natural biodegradable antiscalant additives [4]. At last, they offer promising possibilities in tissue engineering and bone reconstruction [5,6].
However, due to their polydispersity, their polyanionic properties, and their glycosylation, these proteins usually resist classical chromatographical fractionation, stain very poorly on polyacrylamide gels, and absorb very little at 280 nm [7]. A consequence of these technical obstacles is that only few shell proteins are known at present [8], and their exact function during calcification remains elusive.
A way to improve the detection of shell matrix components consists of raising polyclonal antibodies against the nonfractionated matrix, and in using these antibodies on western-blots. Whatever the epitopes (proteinaceous or saccharidic) are, we observe by experience that this often results in a better discrimination of discrete bands than any gel staining. Using this property, we have set up a two-step procedure for obtaining large amount of shell proteins: a blind fractionation of molluscan shell matrices on a preparative denaturing gel electrophoresis, and the detection of the eluted proteins by dot-blot. This approach permits us to acquire structural information on the isolated proteins and to determine their putative functions in biomineralization.

Shell Matrix Extraction and Polyclonal Antibody Production
The extraction is performed at 4˚C.

2.
Suspend the powder (5 to 100 g) in milli-Q water, in a beaker.

3.
Add progressively cold acetic acid (5 to 20% vol/vol), until pH 4. The decalcification is controlled by a titrimeter. The decalcification is over when the pH does not vary (overnight decalcification).

5.
Filter the supernatant on a 0.45-µm filter, and discard the pellet.

6.
Reduce the volume of the solution by ultrafiltration (Amicon stirred cell, cutoff 10 kDa) to 10 to 30 ml. 7.
Dialyze the solution against milli-Q water (several water changes). 8.
Freeze dry the solution. Expect 0.02 to 0.3% of the initial weight of the powder. 9.
Prepare a polyclonal antibody, by injecting an emulsion containing the antigen and the Freunds adjuvant, in a rabbit (see Note 1).

1.
A standard immunization procedure is performed with injections at 0, 14, 28, and 56 days, and bleedings at 0 (preimmune), 38, 66, and 80 days. The respective titers of the collected antisera are determined by ELISA, and their specificity is verified on westernblots. Because the antigen (the shell matrix) is a mixture of glycoproteins and polysaccharides, the resulting antibodies may be raised against proteinaceous and saccharidic epitopes. This does not have any influence on the subsequent preparative fractionation. A way to determine the nature of the epitope consists in degrading the matrix with proteinase K (proteinaceous epitopes) or with Na-periodate (saccharidic epitopes), and to measure, by ELISA or dot-blot, the loss of reactivity of the treated matrix.

2.
A 10 to 12% acrylamide gel is prepared according to a standard procedure [10]. During polymerization, the separation gel is cooled down by circulating water in the cooling core. Polymerization is performed overnight. The volume used for the stacking gel should be at least twice that of the sample. The elution buffer (Tris/glycine) has the same molarity as the running buffer, but does not contain SDS.

3.
Because the shell matrix is a mixture of components, which exhibit very different immunogenicities, the antiserum used does not permit us to quantify specific bands on Western-blots nor on dot-blots. Quantification may be performed from the freeze-dried fractions, by weighing the lyophilisates or by redissolving them and performing a micro-BCA.  The arrows indicate the two main fractions, which are also visualized in Fig. 1. Some minor and less immunogenic fractions have also been collected. The fact that these fractions are visualized on the final western-blot and not on the initial blot of the whole matrix (lane WB, Fig. 1, step 2) is explained by a "concentration" effect. On the final blot, the fractions are much more concentrated, and are subsequently detected by the antimatrix serum, in spite of their low immunogenicity