PhD defense of Solmaz HOSSEIN KHANI

Congratulations to Solmaz, who brilliantly defended her thesis as part of a project conducted at the FARE lab with support from the Grand Est region and the FillingGaps project of the PEPR B-BEST program, under the supervision of Yassin Refahi and Gabriel Paës.

The jury memberes were:

Mr Grégoire MALANDAIN, examiner, Director of Research at INRIA, President of the jury
Ms Estelle BONNIN, rapporteur, Research Engineer at INRAE
Mr Philippe ANDREY, rapporteur, Director of Research at INRAE
Mr Paul DURU, examiner, Professor at INP Toulouse

Abstract:

Plant cell wall is a renewable source of biopolymers and an alternative for fossil-based resources.
However, its resistance to enzymatic deconstruction, known as recalcitrance, needs to be overcome
to achieve a cost-effective transformation into bioproducts. This requires a comprehensive
understanding of enzymatic deconstruction which remains under-explored at microscale. In this
PhD, a robust 4D (space + time) computational pipeline has been developed which comprehensively
addresses the challenges to quantify cell wall deconstruction at microscale from time-lapse 3D
images of wood samples during enzymatic hydrolysis. The pipeline combines a divide-and-conquer
strategy with temporal propagation of spatial information to track extensively deconstructed cell
walls and provides dynamics of cell wall autofluorescence intensity and morphological parameters.
In parallel, the sugar conversion yields were also measured. The analysis revealed a strong negative
correlation between dynamics of autofluorescence intensity and sugar conversions, which
demonstrated that cell wall autofluorescence can serve as a predictor of hydrolysis yield. In parallel,
strong correlations between morphological parameter dynamics and sugar conversion yields were
observed. Building further on these results, a hierarchical family of dynamic models is developed
that integrates polymer-level substrate description, enzyme adsorption, and product inhibition into a
mathematical framework to quantitatively study enzymatic hydrolysis. Overall, in this PhD an
original combination of time-lapse imaging, 4D image processing and computational modeling has
been used to better understand enzymatic deconstruction.