**2. The N-Term: Function and Structural Determinants**

This naturally unfolded domain contains the major part of the so-called transmembrane domain (termed TM1, comprising roughly residues 112–135) and the preceding "stop transfer effector" (STE, a hydrophilic region containing roughly residues 104–111) [39,40] (Figure 1B). STE and TM1 act in concert to control the co-translational translocation at the endoplasmic reticulum (ER) during the biosynthesis of the protein [41,42].

N-term\_HuPrPC functions as a broad-spectrum molecular sensor [43]. Along with the highly homologous protein from mouse (N-term\_MoPrPC, 93% sequence identity), it interacts with copper ions (see below) and sulphated glycosaminoglycans [44]. In addition, N-term\_MoPrP<sup>C</sup> interacts with vitronectin [45], the stress-inducible protein 1 (STI1) [46], amyloid-β (Aβ) multimers [47–49], lipoprotein receptor-related protein 1 (LRP1) [50], and the neural cell adhesion molecule (NCAM) [51].

Because experimental structural information on the full-length N-term\_HuPrPC is currently lacking, one has to resort to biocomputing-based predictions. Recently, some of us have used a combination of bioinformatics along with replica-exchange-based Monte Carlo simulation at room temperature, based on a simplified force field, to predict the conformational ensemble on the full-length N-term\_MoPrPC [31,52].

This is expected to be quite similar to that from *Homo sapiens*, given the extremely high sequence identity (93%) with N-term\_HuPrPC [31,52]. Monte Carlo simulations suggest that the N-term\_MoPrPC consists of several regions characterized by different secondary structure elements, consistently with biophysical data [53–57]. Specifically, it contains 19 ± 8% α-helix, 8 ± 5% β-sheet, 7 ± 3% β-bridge, 27 ± 5% β-turn, 12 ± 4% bend, 4 ± 3% 310-helix, and 1 ± 1% π-helix. The secondary structure elements are distributed among the N-term in a highly heterogeneous manner (Figure 2A): residues 23-30 are mainly coil/β-turn/bend; residues 31-50 are mainly β-turn/coil/bend/β-bridge; and residues 59-90 form four sequential octarepeat (OR) peptides, with sequence PHGGGWGQ, and are mainly β-turn/coil/bend/β-sheet conformations. In particular, the loop/β-turn conformations in the OR region resemble (backbone RMSD < 2.5 Å) those identified by NMR [57]; residues 89-98 are mainly coil/β-turn/bend/β-sheet; residues 99-117 feature the highest content α-helix of N-term\_MoPrP<sup>C</sup> regions; and residues 118-125 display a comparable percentage of α-helix and β-turn. Residues 105-125, the "amyloidogenic region", feature transient helical structures (the last eight residues also have a comparable content of beta turn). This is consistent with circular dichroism (CD), nuclear magnetic resonance (NMR), and Fourier transform infrared (FTIR) studies on HuPrP<sup>C</sup> fragments [54–56] (Figure 2). The same simulation procedure can be carried out for the known disease-linked mutations (Figure 2).

**Figure 2.** Selected conformations of (**A**) WT N-term\_MoPrPC and (**B**) one PM (N-term\_MoPrPC\_Q52P) emerging from molecular simulation [31,52]. These contain transient α-helix (in violet), β-sheet (yellow), β-bridge (orange), β-turn (cyan), 310-helix (blue), and p-helix (red) elements. (**C**) Superimposition of our conformational ensemble (orange) with available fragments of N-term deposited structures. Readapted from [31,52].

While many PMs in the GD are known to modify significantly the folded structure and to increase its flexibility [58–61], our Monte Carlo calculations suggest that those in the N-term do not impact significantly the global structural properties of the N-term. This finding is consistent with experimental findings showing that PMs in N-term\_HuPrPC do not affect the thermostability or misfolding kinetics of the protein [58,62–64]. On the contrary, our Monte Carlo simulations show that the PMs at the N-term modify local features at the binding sites for known cellular partners, as well as of interdomain interactions. This points to an interference of the PMs with the related physiological functions.

The major differences in the presence of the PMs were observed in the residues binding Cu2+ and sulphated GAG (i.e., the OR region and the H110 Cu2+-binding site mouse sequence, H111 in the human sequence). In addition, the PMs affect the SS and the flexibility and increase the hydrophobicity of STE/TM1. The latter contains the putative binding sites for in vivo binding partner proteins such as vitronectin [45] and STI1 [46]. This might affect the biological function of these interactions, which involves the signaling for axonal growth [45] and that for neuroprotection [46], respectively.

The PM Q52P in the OR region, interestingly, affects the flexibility of STE/TM1, while the other six PMs in STE/TM1 also alter the intra-molecular contacts in the OR region suggesting a role played by PMs in altering transient interdomain interactions between the OR region and STE/TM1. Recent studies suggest that N-Term and GD interactions might also serve to regulate the activity and/or toxicity of the PrP<sup>C</sup> N-term [65]. Unfortunately, in the reported Monte Carlo study [52], the GD was not taken into account.

The altered local features in STE/TM1 might also impact the interactions of the protein with trans-acting factors in the cytosol and in the ER membrane [66]. This result is consistent with the in vitro data that PMs P101L, P104L, and A116V increase the interactions between MoPrPC STE/TM1 and a membrane mimetic at pH 7 [67].
