Fig 2: Proposed model of EBOV entry into cells expressing HER2 and TYRO3.HER2 interacts with all members of the TAM (TYRO3, AXL, and MERTK) receptor family to form heterodimers. In TAM receptor-mediated cellular entry, EBOV particles attach to the cell surface by binding to a TAM receptor (e.g., TYRO3) via the serum protein GAS6 (1). HER2 also interacts with the virus-bound TYRO3, forming a HER2/TYRO3/EBOV complex on the cell surface (2). The formation of this complex leads to HER2 phosphorylation (either autophosphorylation or phosphorylation by an as yet unknown cellular factor) and the phosphorylation of TYRO3 by HER2 (3), leading to downstream activation of PI3K/AKT signaling pathway (HER2-mediated AKT1 activation) (4) and subsequent induction of macropinocytosis (5). After internalization of the HER2/TYRO3/EBOV complexes into early endosomes via macropinocytosis (6), the complexes are disassociated, and only EBOV particles are delivered to late the endosomes/lysosomes, where later events such as membrane fusion occur. HER2 and TYRO3 are most likely recycled back to the cell surface.

Fig 3: Behavioral improvements in Tyro3−/− mice. (A) Adult Tyro3−/− mice take less time to detect the submerged platform in a Morris water maze test (see Methods) than their WT counterparts beginning at 3 days of testing. (B) When returned to the training (‘conditioning’) context, Tyro3−/− mice exhibit enhanced memory recall, as monitored by elevated freezing behavior (left). No difference in freezing is detected when mice are placed in a novel context (right). n is total number of mice tested per genotype. P-values: * <0.05.

Fig 4: Immature ultrastructure of Tyro3−/− synapses. (A) Representative negatively stained transmission electron microscopic image of the stratum radiatum from the Tyro3−/− hippocampus at P60. Synaptic vesicle-filled pre-synaptic axonal boutons are blue. Dendritic spines with post-synaptic densities are pink. Some of these display a non-contiguous post-synaptic density, or ‘perforated synapse’ (ps; curved arrows). (B) 3D image of a perforated synapse reconstructed from serial EM sections, showing a discontinuous post-synaptic density. (C–E), Quantification of the incidence of perforated synapses (C), multiple synapse boutons (D, MSB), and conventional synapses (E) in WT versus Tyro3−/− stratum radiatum, from P60 EM images as in (A). The total number of fields quantified (n) from 3 separate mice of each genotype (6 sections per mouse) is indicated: for WT, 28 quantifications were performed from 18 sections, and for Tyro3−/−, 26 quantifications were performed from 18 sections. p-values: ** <0.01, **** <0.0001, Mann–Whitney test. Scale bar: 1 μm.

Fig 5: Negative and positive electrostatic potentials mapped onto the molecular surface of predicted ProS1 structure and Tyro3. (a) The predicted structure for this region of ProS1 is shown in a cartoon representation. The calcium ion is shown as a green sphere and the three N-glycosylation sites are highlighted. (b) Negative and positive electrostatic potentials mapped onto the molecular surface of the ProS1 model. Face 1 is visible in this orientation. The potentials are on a −5, +5 red-white-blue colour map in units of kJ/mol/e. A major electropositive region is observed in the LG1 domain of ProS1. (c) The other side of ProS1 is shown (Face 2). (d) ProS1 electrostatic potentials; Face 1. The negative regions are in red and the positive areas in blue. The potentials are on a −5, +5 red-white-blue colour map in units of kJ/mol/e. (e) Tyro3 electrostatic potentials. The orientation is similar to the one shown in A. A key electronegative surface is noticed in the Ig1 domain of Tyro3, next to the electropositive surface of ProS1. (f) View of the other face (Face 2). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)