Figure 3

Analysis of compensatory mutations in the signal peptide of murine GDF9. A. Alignment of consensus and murine (Mm) signal peptide sequences of GDF9. The consensus sequence is the same as presented in Figure 2. The divergences between the two sequences are identified by numbers that are subsequently used in the analysis. B. Signal peptide probability of the different sequences, as predicted by Phobius. Top panel: predicted signal peptide probability for the consensus and murine sequences. Middle panel: drastic diminution of the predicted signal peptide probability when introducing 3 of the 6 murine divergent sites in the consensus sequence. Bottom panel: restoration of the signal peptide probability when introducing the I21S (compensatory) mutation. C. Diagram depicting all possible combinations of the 6 divergent sites (and their effects when introduced in the consensus sequence). The 6 divergent sites are represented by their numbers, as shown in A, at the vertices of the hexagon. The lines link the mutations together and each black dot at the intersections represents a specific combination. The combination at the center of the hexagon contains all the mutations (i.e. the present state of the sequence in the relevant species). The labels are color-coded according to the signal peptide probabilities predicted by Phobius (as % with respect to the score of the consensus with no mutations). For the murine sequence, the predicted impact varies from slighlty 'hypermorphic' (up to 113% of the score of the consensus) to very deleterious (down to 21% of the score of the consensus).