The changes below were made to the article on the date noted. These changes have been incorporated in the article and the details are documented here.

8 October 2002:

In the caption to Table 1, the last sentence was corrected from:

The composition of the diet is given in g/100 g body weight, as reported in Reeves [31].

to:

The composition of the diet is given in g/100 g, as reported in Reeves [31].

8 January 2003:

In the Methods section under the heading "DTT-induced aggregation of insulin", the last sentence was corrected from:

An unrelated protein, BSA, did not slow down the aggregation of the insulin B-chain indicating the specificity of α-crystallin in suppressing insulin aggregation (Figure 6A).

to:

An unrelated protein, lysozyme, did not slow down the aggregation of the insulin B-chain indicating the specificity of α-crystallin in suppressing insulin aggregation (Figure 6A).

In the Results section under the heading "Heat-induced aggregation of βL- and γ-crystallin", the last sentence was corrected from:

BSA was used as a positive control and it did not minimize heat-induced aggregation of βL-crystallin (Figure 6B).

to:

Lysozyme was used as a positive control and it did not minimize heat-induced aggregation of βL-crystallin (Figure 6B).

In Figure 6, the last sentence in the description of panel A was corrected from:

BSA (0.5 mg/ml, trace 6) did not prevent insulin aggregation.

to:

Lysozyme (0.5 mg/ml, trace 6) did not prevent insulin aggregation.

In Figure 6, the last sentence in the description of panel B was corrected from:

BSA (0.5 mg/ml, trace 6) did not suppress βL-crystallin aggregation.

to:

Lysozyme (0.5 mg/ml, trace 6) did not suppress βL-crystallin aggregation.