Isotope-edited ESEEM: A new method for probing copper binding sites in neurodegenerative proteins

ESEEM–A unique probe of IDR-Cu²⁺ interactions

The interactions between intrinsically disordered domains in neurodegenerative proteins are often stabilized by the inclusion of physiologic metal ions such as copper or zinc. Characterizing the metal ion coordination environment in such cases is critical for assessing the stability and organization of these relevant protein–protein interactions but is challenging given the lack of regular molecular order of the relevant protein domains. We recently developed a new pulsed electron paramagnetic resonance approach that takes advantage of electron spin echo envelope modulation (ESEEM) with selective ¹⁵N isotope labeling. Histidine side chains with ¹⁴N or ¹⁵N coordinated to Cu²⁺ give distinct, fully resolvable signals in both ESEEM and its two-dimensional companion method HYSCORE. The approach was applied to the cellular prion protein, PrP^C, to investigate how its otherwise neurotoxic, disordered N-terminal domain is regulated through a Cu²⁺ linkage to its globular C-terminal domain. Sortase-mediated ligation created an expressed murine prion protein with segmental ¹⁵N-labeling exclusive to the N-terminal domain. Isotope-edited ESEEM identifies the specific His residues from both domains that participate in this essential regulatory process. Extending this approach to hetero-protein complexes focused on how Cu²⁺ may facilitate the interaction between PrP^C and the Aβ peptide, a dominant component of senile plaques in Alzheimer's disease. Combining ¹⁵N-labeled PrP^C with unlabeled Aβ shows that both proteins simultaneously coordinate Cu²⁺ with Aβ contributing a single His side chain. Collectively, this review highlights isotope-edited ESEEM as an effective method for achieving residue-specific insights into metal coordination within structured and disordered protein complexes.

Prevalent neurodegenerative diseases arise from the accumulation and subsequent plaque formation of specific peptides or proteins. For example, the early stages of Alzheimer’s disease (AD) are marked by the accumulation of the 40 to 43 residue amyloid β (Aβ) peptide resulting in senile plaques, followed by intracellular tau inclusions leading to neurofibrillary tangles and neuronal death (1, 2, 3, 4). A consequence of AD and other neurodegenerative diseases is the redistribution of physiologic metal ions in the brain (5, 6). Early studies found that AD leads to a tissue region-dependent reduction of CNS copper but an increase in iron and zinc (7, 8). However, animal studies (5, 9) and postmortem human tissue analyses (6, 10, 11) find that copper accumulates in the Aβ-rich senile plaques. Similarly, in prion diseases, which are caused by the misfolding and aggregation of a normal cellular prion protein (PrP^C) into a harmful, infectious form called PrP-scrapie (PrP^Sc), elevated levels of copper and zinc are observed at the PrP^Sc aggregation sites (12). Several outstanding reviews provide detailed discussions of the potential role for copper in neurodegenerative diseases (6, 13, 14, 15, 16).

Both Aβ (17, 18, 19, 20) and PrP^C (21, 22, 23, 24, 25, 26, 27, 28) are high-affinity copper binding proteins, as is α-synuclein (29, 30, 31), the intracellular protein that causes Parkinson’s disease. These proteins interact preferentially with Cu²⁺, the dominant extracellular copper species (32). However, characterization of the specific Cu²⁺ binding modes is complicated by the fact that each protein’s putative binding sites are in flexible, intrinsically disordered regions (IDRs). It’s not surprising, therefore, that characterization of the copper binding affinity, stoichiometry and coordination mode(s) for Aβ and PrP^C have led to disparate results. For example, K_d measurements of the Aβ-Cu²⁺ complex range from 10 pM to 100 nM, a range covering four orders of magnitude (16, 33). Despite these uncertainties, if copper, specifically Cu²⁺, plays a role in neurodegeneration, for example by predisposing monomeric peptides to form plaques or by conferring deleterious redox activity within these plaques, then characterizing its affinity and coordination modes is essential.

Our interests focus on PrP^C, the protein responsible for mad cow disease, chronic wasting disease, which continues to spread among cervids in the Rocky Mountain regions, and the human diseases kuru and Creutzfeldt-Jakob disease (34). The ∼210-residue protein avidly takes up Cu²⁺, an interaction important for regulating inherent PrP^C toxicity (35, 36, 37, 38, 39). PrP^C also plays a central role in AD by functioning as a receptor for both monomeric and oligomeric Aβ (40), the latter of which is designated Aβ_o (41, 42). Monomeric Aβ binding to PrP^C results in the translocation of Aβ through endocytosis, possibly leading to its intracellular accumulation (40). Aβ_o binding to PrP^C is implicated in transmembrane signaling ultimately driving tau phosphorylation and intraneuronal aggregation (43). Both processes are active in the early stages of AD (44). Moreover, PrP^C is found enriched in amyloid plaques of AD patients (45) and promotes the formation of Aβ aggregates (46).

This review is aimed at new electron paramagnetic resonance (EPR) methods for revealing how Cu²⁺ regulates PrP^C toxicity and, in addition, facilitates the protein’s interaction with Aβ. We will briefly discuss current models of PrP^C- and Aβ-Cu²⁺ binding derived from EPR and complementary methods, such as affinity studies, optical spectroscopy and X-ray absorption. Next, taking advantage of protein expression with uniform ¹⁵N labeling, along with segmental labeling using sortase-mediated ligation (47), we describe how the distinct pulsed EPR signals of ¹⁴N and ¹⁵N resolved through electron spin echo envelope modulation (ESEEM) (48) are used to unambiguously assign site-specific histidine residues in the Cu²⁺ coordination environment. We then examine how PrP^C interacts with Aβ through simultaneous interactions at a copper center. This review will demonstrate how advanced EPR techniques, isotope-edited ESEEM, resolves longstanding ambiguities in copper coordination.

Residues implicated in Cu²⁺ binding to the prion protein and the Alzheimer’s Aβ peptide

Sequences for PrP^C and Aβ are shown in Figure 1. PrP^C possesses two independent domains: an IDR from residue 23 to 125 and a predominantly helical globular domain from 126 to 230 (mouse sequence after removal of the N-terminal and C-terminal signal peptides) (37, 49). Original EPR experiments identified multiple copper binding modes, depending on the ratio of Cu²⁺ to protein (23). At a low Cu²⁺:PrP^C ratio, Cu²⁺ is coordinated primarily by His imidazoles in the octapeptide repeat (OR) domain, a segment characterized by four His-containing repeats of the eight-residue PHGGXWGQ sequence (where X in repeats 2 and 3 represents either Gly or Ser depending on species). With additional equivalents, Cu²⁺ is also taken up by the non-OR His residues at 95 and 110 (50). Interestingly, recent work with transgenic mice finds that mutation of His95 to Tyr results in spontaneous prion disease, spongiform degeneration, and protease-resistant PrP accumulation consistent with prion disease (51). At high Cu²⁺:PrP^C ratios and pH > 6.5, the OR domain saturates with each repeat taking up a single Cu²⁺ equivalent, coordinated by His and backbone Gly amide nitrogens (23, 25, 26, 27). Finally, there are His residues at positions 139 and 176 in the globular domain, also implicated in copper binding by NMR experiments (39).

Figure 1.

Sequences of the cellular prion protein, PrP^C (top), and the Alzheimer’s Aβ peptide (bottom). PrP^C possesses two domains, an IDR from residue 23 to 125 (sequence to residue 115 shown) and a predominantly helical globular domain from 126 to 230. The locations of relevant histidines in both polypeptides, implicated as possible Cu²⁺ binding sites, are indicated by red dots. The PrP^C globular domain has two His residues proposed to serve as Cu²⁺ mediated anchor points bringing the two domains together, regulating the protein’s toxicity. PrP^C, cellular isoform of prion protein residues 23 to 230; Aβ, amyloid-β peptide; IDR, intrinsically disordered region.

The 40 to 43 residue Aβ peptide (Fig. 1) is a cleavage product resulting from β- and γ-secretases that act to release the Amyloid Precursor Protein (APP) ectodomain from the extracellular membrane surface (1). Whether Aβ itself possesses physiological function remains an area of investigation. However, there is broad consensus that Aβ takes up Cu²⁺ with high affinity (16, 17, 20, 52). The interaction has been carefully investigated by a range of methods including electrochemical titrations (53), UV-Vis Absorption spectroscopy (54), NMR, continuous wave EPR (17), pulsed EPR (55), X-ray absorption and extended X-ray absorption fine structure spectroscopy (19, 56) and density functional theory calculations (57). The emerging model proposes an equatorial coordination environment composed of two His imidazoles, the N-terminal amine and an oxygen, which could come from the Asp1 side chain or backbone carbonyl (58). Common to both PrP^C and Aβ, of course, is the critical role of histidine, with its unique capacity for participating in the formation of high affinity Cu²⁺ binding sites (59, 60).

ESEEM–A unique probe of IDR-Cu²⁺ interactions

ESEEM, first demonstrated by Mims et al. in their development of pulsed EPR (61), continues to find new applications in both electronic materials and biological research. If there are weak nuclear couplings to a paramagnetic center, sweeping out pulse delays in a two or three pulse spin echo sequence will generate time-dependent oscillations of the echo intensity related to the coupled nuclear frequencies (48, 62, 63). As discovered by Peisach and Mims, Cu²⁺-imidazole samples probed at 9 to 10 GHz give pronounced ESEEM modulations with frequencies labeled ν₋, ν₀ and ν₊, approximately matching the ¹⁴N (I = 1) quadrupolar transitions (64, 65, 66). Moreover, the ¹⁴N coupling is from the remote, non-coordinated, imidazole nitrogen, which is unique to histidine residues. Consequently, the method is incredibly valuable for identifying His coordination to copper centers, as are typically found in neurodegenerative proteins. ¹⁵N (I = ½) His also produces pronounced ESEEM signals but with transitions distinct from those of ¹⁴N His (65). Specifically, the ¹⁵N nucleus exists in one of two spin states that, when mixed with ¹⁵N hyperfine coupling, gives rise to two observable transitions labeled ν_α and ν_β. Energy level diagrams and simulated spectra for copper coordinated to ¹⁴N or ¹⁵N His, with typical coupling parameters, are shown in Figure 2. The detailed ESEEM theories for these cases are well described elsewhere (48, 63).

Figure 2.

Simulated 3-pulse ESEEM and HYSCORE spectra demonstrating the distinct signals from ¹⁴N and ¹⁵N His coordinated to Cu²⁺.A, the electron spin energy level diagram for ¹⁴N His, indicating the low frequency quadrupole transitions ν₀, ν₋ and ν₊, along with a higher frequency “double quantum” transition, ν_dq, at approximately 4.0 MHz. B, Cu²⁺ ESEEM spectra with one, two, and four coupled ¹⁴N His showing diagnostic increases of the 4.0 MHz signal. (Note: all spectra were simulated using the EasySpin package (75), version 5.1). C, the ¹⁵N energy level diagram shows observed transitions from coupling to a single ¹⁵N His, ν_α and ν_β, and transitions from two ¹⁵N His, revealing an additional signal at 2ν_α. D, ESEEM spectra showing how the 2ν_α arises only with more than one coupled His residue. E, the two-dimensional HYSCORE Cu²⁺ spectrum resulting from four coordinated ¹⁵N-His. Cross peaks arise at multiples of ν_α. ESEEM, electron spin echo envelope modulation; HYSCORE, hyperfine sublevel correlation.

Beyond the assignment of His imidazoles coordinated to a copper center, ESEEM also yields spectroscopic features related to the number of coupled nitrogen nuclei. Importantly, the nature of these signals is distinct for each of the two nitrogen isotopes. For ¹⁴N, multiple His coordination results in a dramatic increase in the intensity of so-called “double quantum” signal at approximately 4.0 MHz (Fig. 2) (23, 64). By contrast, multiple ¹⁵N His gives rise to an additional signal at 2ν_α, arising from the transition between the |½, ½> and the |-½, -½> states, as recently characterized by Smart et al. (67, 68). The two-dimensional version of ESEEM, referred to as hyperfine sublevel correlation spectroscopy (HYSCORE), further resolves the 2ν_α signal as a separate peak (Fig. 2) (67). Moreover, additional ¹⁵N couplings are evident as further multiples of the fundamental ν_α frequency. Given the potential value of the 2ν_α ESEEM peak, Smart et al. evaluated the timing in three-pulse experiment to simultaneously optimize this signal but without loss of the fundamental ¹⁴N and ¹⁵N transitions (67).

Copper regulation of prion protein neurotoxicity

Figure 3 highlights the two distinct domains of PrP^C, the predominantly helical C-terminal domain and the extended N-terminal IDR with an equivalent of Cu²⁺ coordinated by the OR His residues. Deletion of specific polypeptide segments between the OR and the C-terminal domain are highly toxic, leading to neonatal death in transgenic mice and spontaneous transmembrane currents in cultured cells and primary neurons (37, 69, 70). Consequently, it is now hypothesized that the toxicity of monomeric PrP^C arises from its N terminus, referred to as the “toxic effector” domain (37, 71). Similar N-terminally driven toxic responses may be generated by treatment of WT PrP^C with monoclonal antibodies that target specific epitopes on the C-terminal domain (71). The C-terminal domain, therefore, regulates the N-terminal toxic effector domain. Moreover, Cu²⁺ is required for this regulatory interaction since the addition of copper-pentaglycine to a toxic PrP^C deletion mutant abrogates toxic transmembrane ionic currents (37).

Figure 3.

Identification of His residues available to coordinate Cu²⁺. Previous work suggested that Cu²⁺ formed a coordination link bringing together the two PrP^C domains, shifting the equilibrium to the right, tempering the protein’s intrinsic neurotoxicity. However, as indicated by the inset, the specific His residues involved in this regulatory mechanism, as well as the number of His residues from the respective domains, remained unclear. PrP^C, cellular isoform of prion protein residues 23 to 230.

The PrP^C C-terminal regulatory domain possesses two His residues at positions 139 and 176. NMR experiments suggest that the Cu²⁺-occupied OR docks to a C-terminal surface encompassing these two His residues (39). Moreover, simultaneous deletion of both His residues leads to enhanced spontaneous currents in cultured neuroblastoma cells (39). These observations suggest that the regulatory interaction involves copper linking the two PrP^C domains through mutual His coordination, as shown schematically in Figure 3 (39, 72). However, the identity of the specific C-terminal His and the respective numbers of C-terminal versus OR His remained unknown.

By traditional EPR methods, His coordination to Cu²⁺ centers gives signals with little variation; there is no way to distinguish one ¹⁴N His from another. However, the unique ¹⁴N versus ¹⁵N His ESEEM spectra suggested that selective placement of isotopically distinct histidines in a segmentally labeled PrP^C construct would critically test for mutual copper coordination between the two domains. As demonstrated by Pavlovici et al., segmental labeling was achieved through sortase-mediated ligation (47), as shown in Figure 4 (68). The N-terminal domain through residue 117 was expressed with uniform ¹⁵N labeling and ligated to the unlabeled C-terminal domain. The resulting construct retained ¹⁴N at His positions 139 and 176.

Figure 4.

Strategy for segmental labeling of PrP^C and resulting pulsed EPR spectra.A, sortase-mediated ligation results in PrP^C with uniform ¹⁵N labeling of residues 23 to 117. B, ESEEM (top) and HYSCORE (bottom) show a superposition of ¹⁴N- and ¹⁵N-His Cu²⁺ spectra. The presence of the 2ν_α peak indicates multiple N-terminal His at the copper center (detailed analysis in ref (68)). C, AlphaFold 3 model showing how Cu²⁺ coordinates simultaneously to H176 of the globular domain and three histidines of the N-terminal OR domain. PrP^C, cellular isoform of prion protein residues 23 to 230; EPR, electron paramagnetic resonance; ESEEM, electron spin echo envelope modulation; HYSCORE, hyperfine sublevel correlation; OR, octapeptide repeat.

ESEEM and HYSCORE of segmentally isotopically labeled PrP^C with one equivalent of copper clearly show simultaneous signals of both ¹⁴N and ¹⁵N-His (68). In addition, both spectra exhibit 2ν_α signals indicating coordination from multiple OR His residues. By contrast, analysis of the ¹⁴N 4.0 MHz peak suggests coordination by a single C-terminal His residue. Follow up mutagenesis studies identified His 176 as a critical participant in copper coordination and, hence, regulation of the otherwise toxic N-terminal domain. AlphaFold 3 simulations with a single Cu²⁺ ion produced a structure consistent with our EPR experiments (Fig. 4) (68).

Copper anchors Aβ to PrP^C through mutual copper coordination

Accumulation of intraneuronal Aβ (iAβ) is emerging as a distinct contributor to AD perhaps paralleling the more conventional neurodegenerative mechanism of extracellular Aβ aggregation and plaque formation (73, 74). iAβ may associate at organelles such as the mitochondria, ultimately compromising cellular function. Experiments with cultured cells demonstrates that PrP^C transports fluorescently-labeled Aβ across the plasma membrane through endocytosis (40). Given that both PrP^C and Aβ exhibit high affinity copper binding, it is reasonable to explore whether the prion protein’s Cu²⁺ center provides an anchor point for binding Aβ.

Isotope-edited ESEEM was applied to uniformly ¹⁵N-labeled PrP^C in combination with unlabeled Aβ (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30), a nonaggregating segment of full-length Aβ (67). Each protein yields a distinct ESEEM spectrum, as shown previously in reference (67). Aβ alone with one equivalent of copper gives a pronounced 4.0 MHz signal indicating multiple His coordination, consistent with participation of the peptide’s multiple His residues. The ESEEM spectrum of ¹⁵N-PrP^C exhibits the 2ν_α signal indicating multiple His coordination; ESEEM of the ternary mixture of 1:1:1 PrP^C, Aβ and Cu²⁺ shows the simultaneous signals of both isotopes (Fig. 5). These features are also present in the companion HYSCORE spectrum (Fig. 5). However, the ¹⁴N 4.0 MHz ESEEM signal in the ternary mixture is greatly reduced relative to that from Aβ alone. Along with persistence of the 2ν_α transition, these findings point to a coordination environment composed of multiple His from PrP^C but only a single His from Aβ, as shown schematically in Figure 5 (67). Reduction of the copper concentration to below one equivalent in the otherwise 1:1 PrP^C:Aβ mixture results in ESEEM spectra equivalent to that observed in Figure 5. This suggests that the ternary complex is thermodynamically stable, thus providing compelling evidence that the coordination model in Figure 5 represents a plausible complex for PrP^C-mediated Aβ transport.

Figure 5.

Isotope-edited ESEEM and HYSCORE of the PrP^C-Cu²⁺-Aβ complex. PrP^C was uniformly ¹⁵N-labeled, while Aβ is unlabeled. A, ESEEM and HYSCORE of a 1:1:1 mixture reveals the ¹⁵N 2ν_α peak, consistent with multiple prion protein His residues, while the 4.0 MHz peak from ¹⁴N is clearly present but weak relative to the Cu²⁺-Aβ complex alone. A thorough analysis of these spectra (67) suggests the model in (B) where Cu²⁺ anchors the two proteins together, with Aβ (red) contributing a single His to the copper center. ESEEM, electron spin echo envelope modulation; HYSCORE, hyperfine sublevel correlation; PrP^C, cellular isoform of prion protein residues 23 to 230; Aβ, amyloid-β peptide.

These experiments provide a molecular perspective, and testable hypotheses, for how the prion protein interacts with Aβ, as necessary for PrP^C-mediated Aβ endocytosis. Aβ transport by PrP^C endocytosis might be part of normal cellular physiology in which the peptide is ultimately degraded through the endo-lysosome pathway. However, it is also plausible that high levels of extracellular Aβ could overwhelm this degradation pathway, ultimately leading to the accumulation of iAβ. Elucidation of the relevant Aβ pathways, along with further clarification of the PrP^C:Aβ:Cu²⁺ complex could provide a new level of understanding and identification of therapeutic targets for treating AD.

Conclusions

ESEEM and HYSCORE are ideal for assessing structural details in IDR-Cu²⁺ complexes, relevant to neurodegenerative proteins. With isotope-edited pulsed EPR, we have expanded the capabilities of these robust methods providing new insights into both the regulation of PrP^C toxicity and PrP^C-mediated Aβ transport. It is our hope that the methods reviewed here contribute to the development of new concepts and strategies for treating the neurodegenerative maladies that plague our aging population.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Reviewed by members of the JBC Editorial Board. Edited by Joan B. Broderick