Internship: Computational chemistry and modern machine learning techniques for the selection of short amphiphilic peptide sequences and the conception of new theranostic nanovectors: from structure to self-assembling properties

The development of new alternative therapeutic strategies for the treatment of cancer while limiting undesirable side effects of conventional chemotherapy is essential. Nanomedicine is an effective approach to meet this requirement, as the encapsulation of antitumor molecules in biocompatible nanovectors should make it possible to reduce their toxicity while improving their circulation and targeting in vivo1.

In this context, the SEISAD team has expertise in the design of new nanovectors based on short peptide self-assembly in monodisperse and size-controlled cavitary supramolecular nanostructures. Indeed, such supramolecular architectures offer the possibility of encapsulating small molecules for better preservation of their activity but also better control of their release in response to different stimuli (pH, thermal, photonic, enzymatic …)2 They also feature surface chemical reactivity via amino- and carboxyl groups, enabling chemical grafting of targeting ligands and imaging probes (optical, MRI, PET), in addition to their intrinsic biocompatibility and biodegradability.

So far, we have studied a peptide sequence with an amphiphilic character including an RGD motif for active targeting of αvβ3 integrin receptors overexpressed in cancer cells, and able to self-assemble into pH-responsive micelle-like nanostructures3. A continuous-flow solid-phase peptide synthesis (SPPS) was developed and optimized in the laboratory through an iterative process based on CE-ESI-MS analysis of the peptide product. The exact sequence was then functionalized with a gadolinium complex and the nanovector successfully formulated based on the co-assembly of the peptide and its conjugate in different proportions. Nonetheless, spontaneously organized nano-rods could only be characterized (DLS, zetametry, TDA, TEM and CE-UV) in strongly basic media (pH < 11), without confirmation of the stoechiometry of assemblies4.

The combination of traditional computational chemistry and modern machine learning techniques has already demonstrated its potential to predict the properties/reactivity of short synthetic peptides and their derivatives5 but also of peptide self-assemblies6. The scope of this internship will be to benefit from the first set of experimental data acquired in the SEISAD team together with already published data on self-assembling short amphiphilic peptides to build high-performance machine learning models, enriched with knowledge from the field of theoretical and computational chemistry, to predict self-assembly7. The aim will be to rapidly identify one or more sequences likely to lead to the right nano-vector characteristics: cavitary nano-architectures thermodynamically stable in the pH 6-7.4 range. Complementary, molecular dynamics will be performed to assess (i) the thermodynamic stability of self-assembled structures resulting from selected sequences in physiological environments and over time, and (ii) the impact of grafting imaging probes on self-assembly properties of these peptides. Eventually, the self-assembly will be evaluated experimentally in the lab.

The intern will have the excellent opportunity to integrate the Synthesis, Electrochemistry, Imaging and Analytical Systems for Diagnosis (SEISAD) team recognized for its expertise in the design and development of innovative tools for the early detection of pathological signals using chemical and analytical methodologies and work in collaboration with the Chemical Theory and Modelling (CTM) team recognized for its expertise in the development of new approaches for in-silico modeling of the reactivity and properties of molecules and extended systems in complex environments.

Keywords: peptide sequence selection; self-assemblying peptidenanostructured biomaterials; Computational chemistry; Machine learning.

Qualifications:

Proactivity, interest in computational design, ability to work in multidisciplinary research and good interpersonal skills are essential.
Compulsory: Full English language proficiency.
The position is available from February 2025 for 5 months.
Salary: About 600 euros per month according to French regulation (to be updated in January 2025).

Application:

Send a CV including at least two references and a cover letter to fanny.dorlye@chimieparistech.psl.eu and davide.avagliano@chimieparistech.psl.eu