Mechanism of interaction of subgroup B2 adenovirus to the desmosomal cadherin DSG2 and applications – Ad-Cadh

Protein expression and purification
Biochemistry
Interaction measurements (SPR, BLI)
Biophysics
SEC-MALLS
Analytical ultracentrifugation AUC
X-ray crystallography
Cryo-elecron-microscopy
In-vivo validation

A major result at this stage of our ANR is the discovery of a non-classical mechanism of interaction of adenoviruses of subgroup B (i.e. Ad3) with its receptor desmogléine 2 (DSG2). Indeed, the trimeric 'fibre' protein of other adenoviruses has always been reported to interact with three receptor molecules (i.e. Ad12/CAR(3X) or Ad11/CD46(3X)). However, the fibre of Ad3 interacts surprisingly with only one or to a lesser extent two DSG2 molecules (Ad3/DSG2 (1X) or Ad3/DSG2(2X)).
The complex thus formed is very small (96kDa) but has nevertheless been resolved by cryo-EM using the ESRF Krios microscope equipped with a 'phase-plate'. This is one of the smallest non-symmetric complexes resolved by this technique to date.
This technological feat combined with an unexpected result on viruses used as oncolytic vectors has been published in a high-level journal (Nature Communications 2019) and has been the subject of numerous national and European press releases. (CNRS-INSB; Défis du CEA (Adenovirus le 3 en 1 contre les tumeurs); PSB Highlights (Adenovirus binding to its receptor visualized by Cryo-EM); Instruct (Optimised cancer therapy, one step closer thanks to Cryo-EM); ESRF...).

Completion of tasks 8, 9 and 10 (creation of vectors, modification of vector tropism, study of biodistribution).

1. Stermann et al., Nature Communications 2019.
doi.org/10.1038/s41467-019-09220-y

2. Stermann et al., Acta Cryst 2019.
doi.org/10.1107/S2053230X19015784

3. Hograindleur et al., Viruses. 2020.
doi.org/10.3390/v12101075

Background:
Adenoviruses (Ads) are a family of non-enveloped linear double stranded-DNA viruses infecting a wide range of hosts, including humans. In humans, the 60 serotypes are divided in seven sub- groups (A to G). The first identified adenovirus receptor, called CAR was identified in 1997 and was shown to interact with an elongated protein located at the twelve capsid vertices: the fibre. This receptor is recognized by all sub-groups except subgroup B adenoviruses (B-Ad) which interact with either CD46 (B1-Ads) or desmoglein 2: DSG2 (B2-Ads) which we identified. Until recently, most of Ad vectors used in preclinical and clinical studies were derived from serotype 5 (subgroup C). However, there is a renewed interest in other serotypes, including those belonging to subgroup B because of their different tropism and their ability to escape from serotype 5 pre-existing immunity. However, a major challenge in the use of B-Ads as vectors is the relatively limited understanding of their biology, especially as compared to C-Ad5. In collaboration with Prof. Andre Lieber (Seattle, USA), our team has previously reported that the tropism of B2-Ad (Ad3, Ad7, Ad11 and Ad14) is mainly dictated by their interaction with DSG-2, a desmosomal component.

Relevance:
Biophysically characterizing and solving the high-resolution structure of a new virus/receptor interaction is often challenging, but the results will provide new understanding on how this virus is able to specifically target and interact with host cells. The fundamental data that we will generate on the B2-Ad/DSG2 interaction will not only provide a foundation for understanding virus-receptor interactions, but also facilitate the design of the next generation of cancer–targeted oncolytic adenoviruses. In addition, Pr Lieber team has shown in a collaborative work with our team that B2-Ad can act as a junction opener and thus as adjuvant for chemotherapeutics and monoclonal antibodies (MAbs) already approved in human use. The resolution of the structure of the B2-Ad/DSG2 interaction will enable a better understanding of the mode of action of junction opening and enable the design of improved B2-Ad adjuvants for therapy. Altogether, this integrative project will bridge the gap between fundamental research in virology and applied research in vectorology.

Methodology:
Our research plan will be divided in three mains steps:
1/ Expression and purification of different DSG2 ectodomain constructs by two expression means (mammalian and prokaryotic) and their screening for B2-Ad fibre binding by surface plasmon resonance (SPR) in order to determine the minimal DSG2 domain required for Ad binding (DSG2min)
2/ Co-crystallisation and cryo-EM imaging of the B2-Ad fiber/DSG2min complex to identify at the atomic level which residues of both partners are critical in the interaction
3/ Design of mutations in the Ad-fibre abrogating or increasing DSG2 binding; in vitro characterisation of adenoviral vectors displaying these mutations and their impact in the in vivo bio-distribution in transgenic mice expressing human DSG2 (hDSG2).