Abstract

Ice nucleation proteins (INPs) form oligomeric structures by self-assembly and aggregation. We looked for the presence of potential aggregating sequences inside the INP from Pseudomonas syringae by a computational approach with the AGGRESCAN, FOMDAMYLOID and TANGO softwares. A total of 38 hot spots of aggregation were predicted in the INP sequence: 7 localized in the Nterminal domain, 2 in the C-terminal region, 28 in the highly repetitive central (HRC) region and 1 shared between the HRC and the Carboxyl-terminus regions of the protein. All the hot spots of aggregation identified in the HRC domain overlapped a 8-residue low fidelity repeat including a LIAGYrelated sequence. We confirmed the predictions by an experimental approach using synthetic peptides corresponding to different parts of the INP central sequence, absorbance spectroscopy and fluorescence spectroscopy in the presence of Congo red (CR) or Thioflavin T (ThT), respectively. Peptide 620-SFIIAGYG-627 predicted to aggregate by the three softwares induced an increase in fluorescence of ThT. Peptide 729-GFKSILTAGY-738 predicted to aggregate by AGGRESCAN and FOLDAMYLOID induced a shift in the maximum of absorbance of CR. Peptide 1124-SVLTAGA-1130 predicted to aggregate only by TANGO did not interfere with CR absorbance or ThT fluorescence. In conclusion, the use of three aggregation prediction algorithms and two biochemical assays showed that the hexapeptide repeated segment LIAGY, previously shown to form a hairpin loop may be involved in the aggregation of the P. syringae INP.

Keywords: Aggregation, ice nucleation, INP, Pseudomonas syringae.

Graphical Abstract